START-INFO-DIR-ENTRY
* Gambit: (gambit).		A portable implementation of Scheme.
* gsi: (gambit) interpreter.	Gambit interpreter.
* gsc: (gambit) compiler.	Gambit compiler.
END-INFO-DIR-ENTRY

   This file documents Gambit, a portable implementation of Scheme.

   Copyright (C) 1994-2016 Marc Feeley.

   Permission is granted to make and distribute verbatim copies of this
manual provided the copyright notice and this permission notice are
preserved on all copies.

   Permission is granted to copy and distribute modified versions of
this manual under the conditions for verbatim copying, provided that
the entire resulting derived work is distributed under the terms of a
permission notice identical to this one.

   Permission is granted to copy and distribute translations of this
manual into another language, under the above conditions for modified
versions, except that this permission notice may be stated in a
translation approved by the copyright holder.

Gambit
******

This manual documents Gambit.  It covers release v4.8.5.

1 The Gambit system
*******************

The Gambit programming system is a full implementation of the Scheme
language which conforms to the R4RS, R5RS and IEEE Scheme standards.  It
consists of two main programs: `gsi', the Gambit Scheme interpreter,
and `gsc', the Gambit Scheme compiler.

   The Gambit compiler generates portable C code, making the whole
Gambit system and the programs compiled with it easily portable to many
computer architectures for which a C compiler is available.  With
appropriate declarations in the source code the executable programs
generated by the compiler run roughly as fast as equivalent C programs.

   For the most up to date information on Gambit and add-on packages
please check the Gambit web page at `http://gambitscheme.org'.  The web
page has links to the Gambit mailing list, the bug reporting system,
and the source code repository.

1.1 Accessing the system files
==============================

Files related to Gambit, such as executables, libraries and header
files, are stored in multiple "Gambit installation directories".
Gambit may be installed on a system according to two different
installation models.

   In the first model there is a single directory where all the Gambit
installation directories are stored.  This "central installation
directory" is typically `/usr/local/Gambit' under UNIX,
`/Library/Gambit' under Mac OS X and `C:/Program Files/Gambit' under
Microsoft Windows.  This may have been overridden when the system was
built with the command `configure --prefix=/my/Gambit'.  If the system
was built with the command `configure --enable-multiple-versions' then
the central installation directory is `prefix/version', where `version'
is the system version string (e.g. `v4.8.5' for Gambit v4.8.5).
Moreover, `prefix/current' will be a symbolic link which points to the
central installation directory.  In this model, the Gambit installation
directory named X is simply the subdirectory X of the central
installation directory.

   In the second model some or all of the Gambit installation
directories are stored in installation specific directories.  The
location of these directories is assigned when the system is built
using the command `configure --bindir=/my/bin --includedir=/my/include
--libdir=/my/lib'.

   The advantage of the first model is that it is easy to have multiple
versions of Gambit coexist and to remove all the files of a given
version.  However, the second model may be necessary to conform to the
package installation conventions of some operating systems.

   Executable programs such as the interpreter `gsi' and compiler `gsc'
can be found in the `bin' installation directory.  Adding this
directory to the `PATH' environment variable allows these programs to
be started by simply entering their name.  This is done automatically
by the Mac OS X and Microsoft Windows installers.

   The runtime library is located in the `lib' installation directory.
When the system's runtime library is built as a shared-library (with
the command `configure --enable-shared') all programs built with
Gambit, including the interpreter and compiler, need to find this
library when they are executed and consequently this directory must be
in the path searched by the system for shared-libraries.  This path is
normally specified through an environment variable which is
`LD_LIBRARY_PATH' on most versions of UNIX, `LIBPATH' on AIX,
`SHLIB_PATH' on HPUX, `DYLD_LIBRARY_PATH' on Mac OS X, and `PATH' on
Microsoft Windows.  If the shell is `sh', the setting of the path can be
made for a single execution by prefixing the program name with the
environment variable assignment, as in:

     $ LD_LIBRARY_PATH=/usr/local/Gambit/lib gsi

   A similar problem exists with the Gambit header file `gambit.h',
located in the `include' installation directory.  This header file is
needed for compiling Scheme programs with the Gambit compiler.  When
the C compiler is being called explicitly it may be necessary to use a
`-I<DIR>' command line option to indicate where to find header files
and a `-L<DIR>' command line option to indicate where to find libraries.

   Access to both of these files can be simplified by creating a link to
them in the appropriate system directories (special privileges may
however be required):

     $ ln -s /usr/local/Gambit/lib/libgambit.a /usr/lib # name may vary
     $ ln -s /usr/local/Gambit/include/gambit.h /usr/include

   Alternatively these files can be copied or linked in the directory
where the C compiler is invoked (this requires no special privileges).

   Another approach is to set some environment variables which are used
to tell the C compiler where to find header files and libraries.  For
example, the following settings can be used for the `gcc' C compiler:

     $ export LIBRARY_PATH=/usr/local/Gambit/lib
     $ export CPATH=/usr/local/Gambit/include

   Note that this may have been done by the installation process.  In
particular, the Mac OS X and Microsoft Windows prebuilt installers set
up the environment so that the `gcc' compiler finds these files
automatically.

2 The Gambit Scheme interpreter
*******************************

Synopsis:

     gsi [-:RUNTIMEOPTION,...] [-i] [-f] [-v] [[-] [-e EXPRESSIONS] [FILE]]...

   The interpreter is executed in "interactive mode" when no file or
`-' or `-e' option is given on the command line.  Otherwise the
interpreter is executed in "batch mode".  The `-i' option is ignored by
the interpreter.  The initialization file will be examined unless the
`-f' option is present (*note GSI customization::).  The `-v' option
prints the system version string, system time stamp, operating system
type, and configure script options on standard output and exits.
Runtime options are explained in *Note Runtime options::.

2.1 Interactive mode
====================

In interactive mode a read-eval-print loop (REPL) is started for the
user to interact with the interpreter.  At each iteration of this loop
the interpreter displays a prompt, reads a command and executes it.
The commands can be expressions to evaluate (the typical case) or
special commands related to debugging, for example `,q' to terminate
the process (for a complete list of commands see *Note Debugging::).
Most commands produce some output, such as the value or error message
resulting from an evaluation.

   The input and output of the interaction is done on the "interaction
channel".  The interaction channel can be specified through the runtime
options but if none is specified the system uses a reasonable default
that depends on the system's configuration.  When the system's runtime
library was built with support for GUIDE, the Gambit Universal IDE
(with the command `configure --enable-guide') the interaction channel
corresponds to the "console window" of the primordial thread (for
details see *Note GUIDE::), otherwise the interaction channel is the
user's "console", also known as the "controlling terminal" in the UNIX
world.  When the REPL starts, the ports associated with
`(current-input-port)', `(current-output-port)' and
`(current-error-port)' all refer to the interaction channel.

   Expressions are evaluated in the global "interaction environment".
The interpreter adds to this environment any definition entered using
the `define' and `define-macro' special forms.  Once the evaluation of
an expression is completed, the value or values resulting from the
evaluation are output to the interaction channel by the pretty printer.
The special "void" object is not output.  This object is returned by
most procedures and special forms which the Scheme standard defines as
returning an unspecified value (e.g. `write', `set!', `define').

   Here is a sample interaction with `gsi':

     $ gsi
     Gambit v4.8.5

     > (define (fact n) (if (< n 2) 1 (* n (fact (- n 1)))))
     > (map fact '(1 2 3 4 5 6))
     (1 2 6 24 120 720)
     > (values (fact 10) (fact 40))
     3628800
     815915283247897734345611269596115894272000000000
     > ,q

   What happens when errors occur is explained in *Note Debugging::.

2.2 Batch mode
==============

In batch mode the command line arguments denote files to be loaded,
REPL interactions to start (`-' option), and expressions to be
evaluated (`-e' option).  Note that the `-' and `-e' options can be
interspersed with the files on the command line and can occur multiple
times.  The interpreter processes the command line arguments from left
to right, loading files with the `load' procedure and evaluating
expressions with the `eval' procedure in the global interaction
environment.  After this processing the interpreter exits.

   When the file name has no extension the `load' procedure first
attempts to load the file with no extension as a Scheme source file.
If that file doesn't exist it will search for both a source file and an
object file.  The object file's name is obtained by adding to the file
name a `.oN' extension with the highest consecutive version number
starting with 1.  The source file's name is obtained by adding to the
file name the file extensions `.scm' and `.six' (the first found is the
source file).  If both a source file and an object file exist, then the
one with the latest modification time is loaded.  Otherwise the file
that is found is loaded.  When the file name has an extension, the
`load' procedure will only attempt to load the file with that specific
name.

   When the extension of the file loaded is `.scm' the content of the
file will be parsed using the normal Scheme prefix syntax.  When the
extension of the file loaded is `.six' the content of the file will be
parsed using the Scheme infix syntax extension (see *Note Scheme infix
syntax extension::).  Otherwise, `gsi' will parse the file using the
normal Scheme prefix syntax.

   The ports associated with `(current-input-port)',
`(current-output-port)' and `(current-error-port)' initially refer
respectively to the standard input (`stdin'), standard output
(`stdout') and the standard error (`stderr') of the interpreter.  This
is true even in REPLs started with the `-' option.  The usual
interaction channel (console or IDE's console window) is still used to
read expressions and commands and to display results.  This makes it
possible to use REPLs to debug programs which read the standard input
and write to the standard output, even when these have been redirected.

   Here is a sample use of the interpreter in batch mode, under UNIX:

     $ cat h.scm
     (display "hello") (newline)
     $ cat w.six
     display("world"); newline();
     $ gsi h.scm - w.six -e "(pretty-print 1)(pretty-print 2)"
     hello
     > (define (display x) (write (reverse (string->list x))))
     > ,(c 0)
     (#\d #\l #\r #\o #\w)
     1
     2

2.3 Customization
=================

There are two ways to customize the interpreter.  When the interpreter
starts off it tries to execute a `(load "~~lib/gambext")' (for an
explanation of how file names are interpreted see *Note Host
environment::).  An error is not signaled when the file does not exist.
Interpreter extensions and patches that are meant to apply to all
users and all modes should go in that file.

   Extensions which are meant to apply to a single user or to a specific
working directory are best placed in the "initialization file", which
is a file containing Scheme code.  In all modes, the interpreter first
tries to locate the initialization file by searching the following
locations: `.gambini' and `~/.gambini' (with no extension, a `.scm'
extension, and a `.six' extension in that order).  The first file that
is found is examined as though the expression `(include
INITIALIZATION-FILE)' had been entered at the read-eval-print loop
where INITIALIZATION-FILE is the file that was found.  Note that by
using an `include' the macros defined in the initialization file will
be visible from the read-eval-print loop (this would not have been the
case if `load' had been used).  The initialization file is not searched
for or examined when the `-f' option is specified.

2.4 Process exit status
=======================

The status is zero when the interpreter exits normally and is nonzero
when the interpreter exits due to an error.  Here is the meaning of the
exit statuses:

`0'
     The execution of the primordial thread (i.e. the main thread) did
     not encounter any error.  It is however possible that other threads
     terminated abnormally (by default threads other than the primordial
     thread terminate silently when they raise an exception that is not
     handled).

`64'
     The runtime options or the environment variable `GAMBOPT'
     contained a syntax error or were invalid.

`70'
     This normally indicates that an exception was raised in the
     primordial thread and the exception was not handled.

`71'
     There was a problem initializing the runtime system, for example
     insufficient memory to allocate critical tables.


   For example, if the shell is `sh':

     $ gsi -:d0 -e "(pretty-print (expt 2 100))"
     1267650600228229401496703205376
     $ echo $?
     0
     $ gsi -:d0,unknown # try to use an unknown runtime option
     $ echo $?
     64
     $ gsi -:d0 nonexistent.scm # try to load a file that does not exist
     $ echo $?
     70
     $ gsi nonexistent.scm
     *** ERROR IN ##main -- No such file or directory
     (load "nonexistent.scm")
     $ echo $?
     70

     $ gsi -:m4000000 # ask for a 4 gigabyte heap
     *** malloc: vm_allocate(size=528384) failed (error code=3)
     *** malloc[15068]: error: Can't allocate region
     $ echo $?
     71

   Note the use of the runtime option `-:d0' that prevents error
messages from being output, and the runtime option `-:m4000000' which
sets the minimum heap size to 4 gigabytes.

2.5 Scheme scripts
==================

The `load' procedure treats specially files that begin with the two
characters `#!' and `@;'.  Such files are called "script files" and the
first line is called the "script line".  In addition to indicating that
the file is a script, the script line provides information about the
source code language to be used by the `load' procedure.  After the two
characters `#!' and `@;' the system will search for the first substring
matching one of the following language specifying tokens:

`scheme-r4rs'
     R4RS language with prefix syntax, case-insensitivity, keyword
     syntax not supported

`scheme-r5rs'
     R5RS language with prefix syntax, case-insensitivity, keyword
     syntax not supported

`scheme-ieee-1178-1990'
     IEEE 1178-1990 language with prefix syntax, case-insensitivity,
     keyword syntax not supported

`scheme-srfi-0'
     R5RS language with prefix syntax and SRFI 0 support (i.e.
     `cond-expand' special form), case-insensitivity, keyword syntax
     not supported

`gsi-script'
     Full Gambit Scheme language with prefix syntax, case-sensitivity,
     keyword syntax supported

`gsc-script'
     Full Gambit Scheme language with prefix syntax, case-sensitivity,
     keyword syntax supported

`six-script'
     Full Gambit Scheme language with infix syntax, case-sensitivity,
     keyword syntax supported


   If a language specifying token is not found, `load' will use the
same language as a nonscript file (i.e. it uses the file extension and
runtime system options to determine the language).

   After processing the script line, `load' will parse the rest of the
file (using the syntax of the language indicated) and then execute it.
When the file is being loaded because it is an argument on the
interpreter's command line, the interpreter will:

   * Setup the `command-line' procedure so that it returns a list
     containing the expanded file name of the script file and the
     arguments following the script file on the command line.  This is
     done before the script is executed.  The expanded file name of the
     script file can be used to determine the directory that contains
     the script (i.e. `(path-directory (car (command-line)))').

   * After the script is loaded the procedure `main' is called with the
     command-line arguments.  The way this is done depends on the
     language specifying token.  For `scheme-r4rs', `scheme-r5rs',
     `scheme-ieee-1178-1990', and `scheme-srfi-0', the `main' procedure
     is called with the equivalent of `(main (cdr (command-line)))' and
     `main' is expected to return a process exit status code in the
     range 0 to 255.  This conforms to the "Running Scheme Scripts on
     Unix SRFI" (SRFI 22).  For `gsi-script' and `six-script' the `main'
     procedure is called with the equivalent of `(apply main (cdr
     (command-line)))' and the process exit status code is 0 (`main''s
     result is ignored).  The Gambit system has a predefined `main'
     procedure which accepts any number of arguments and returns 0, so
     it is perfectly valid for a script to not define `main' and to do
     all its processing with top-level expressions (examples are given
     in the next section).

   * When `main' returns, the interpreter exits.  The command-line
     arguments after a script file are consequently not processed
     (however they do appear in the list returned by the `command-line'
     procedure, after the script file's expanded file name, so it is up
     to the script to process them).


2.5.1 Scripts under UNIX and Mac OS X
-------------------------------------

Under UNIX and Mac OS X, the Gambit installation process creates the
executable `gsi' and also the executables `six', `gsi-script',
`six-script', `scheme-r5rs', `scheme-srfi-0', etc as links to `gsi'.  A
Scheme script need only start with the name of the desired Scheme
language variant prefixed with `#!' and the directory where the Gambit
executables are stored.  This script should be made executable by
setting the execute permission bits (with a `chmod +x SCRIPT').  Here
is an example of a script which lists on standard output the files in
the current directory:

     #!/usr/local/Gambit/bin/gsi-script
     (for-each pretty-print (directory-files))

   Here is another UNIX script, using the Scheme infix syntax extension,
which takes a single integer argument and prints on standard output the
numbers from 1 to that integer:

     #!/usr/local/Gambit/bin/six-script

     void main (obj n_str)
     {
       int n = \string->number(n_str);
       for (int i=1; i<=n; i++)
         \pretty-print(i);
     }

   For maximal portability it is a good idea to start scripts indirectly
through the `/usr/bin/env' program, so that the executable of the
interpreter will be searched in the user's `PATH'.  This is what SRFI
22 recommends.  For example here is a script that mimics the UNIX `cat'
utility for text files:

     #!/usr/bin/env gsi-script

     (define (display-file filename)
       (display (call-with-input-file filename
                  (lambda (port)
                    (read-line port #f)))))

     (for-each display-file (cdr (command-line)))

2.5.2 Scripts under Microsoft Windows
-------------------------------------

Under Microsoft Windows, the Gambit installation process creates the
executable `gsi.exe' and `six.exe' and also the batch files
`gsi-script.bat', `six-script.bat', `scheme-r5rs.bat',
`scheme-srfi-0.bat', etc which simply invoke `gsi.exe' with the same
command line arguments.  A Scheme script need only start with the name
of the desired Scheme language variant prefixed with `@;'.  A UNIX
script can be converted to a Microsoft Windows script simply by
changing the script line and storing the script in a file whose name
has a `.bat' or `.cmd' extension:

     @;gsi-script %~f0 %*
     (display "files:\n")
     (pretty-print (directory-files))

   Note that Microsoft Windows always searches executables in the user's
`PATH', so there is no need for an indirection such as the UNIX
`/usr/bin/env'.  However the script line must end with `%~f0 %*' to
pass the expanded filename of the script and command line arguments to
the interpreter.

2.5.3 Compiling scripts
-----------------------

A script file can be compiled using the Gambit Scheme compiler (*note
GSC::) into a standalone executable.  The script line will provide
information to the compiler on which language to use.  The script line
also provides information on which runtime options to use when
executing the compiled script.  This is useful to set the default
runtime options of an executable program.

   The compiled script will be executed similarly to an interpreted
script (i.e. the list of command line arguments returned by the
`command-line' procedure and the invocation of the `main' procedure).

   For example:

     $ cat square.scm
     #!/usr/local/Gambit/bin/gsi-script -:d0
     (define (main arg)
       (pretty-print (expt (string->number arg) 2)))
     $ gsi square 30        # gsi will load square.scm
     900
     $ gsc -exe square      # compile the script to a standalone program
     $ ./square 30
     900
     $ ./square 1 2 3       # too many arguments to main
     $ echo $?
     70
     $ ./square -:d1 1 2 3  # ask for error message
     *** ERROR -- Wrong number of arguments passed to procedure
     (main "1" "2" "3")

3 The Gambit Scheme compiler
****************************

Synopsis:

     gsc [-:RUNTIMEOPTION,...] [-i] [-f] [-v]
         [-prelude EXPRESSIONS] [-postlude EXPRESSIONS]
         [-dynamic] [-exe] [-obj] [-cc-options OPTIONS]
         [-ld-options-prelude OPTIONS] [-ld-options OPTIONS]
         [-warnings] [-verbose] [-report] [-expansion] [-gvm]
         [-debug] [-debug-location] [-debug-source]
         [-debug-environments] [-track-scheme]
         [-o OUTPUT] [-c] [-keep-c] [-link] [-flat] [-l BASE]
         [[-] [-e EXPRESSIONS] [FILE]]...

3.1 Interactive mode
====================

When no command line argument is present other than options the
compiler behaves like the interpreter in interactive mode.  The only
difference with the interpreter is that the compilation related
procedures listed in this chapter are also available (i.e.
`compile-file', `compile-file-to-target', etc).

3.2 Customization
=================

Like the interpreter, the compiler will examine the initialization file
unless the `-f' option is specified.

3.3 Batch mode
==============

In batch mode `gsc' takes a set of file names (with either no
extension, or a C file extension, or some other extension) on the
command line and compiles each Scheme file into a C file.  The
recognized C file extensions are `.c', `.C', `.cc', `.cp', `.cpp',
`.CPP', `.cxx', `.c++', `.m', `.M', and `.mm'.  The extension can be
omitted from FILE when the Scheme file has a `.scm' or `.six'
extension.  When the extension of the Scheme file is `.six' the content
of the file will be parsed using the Scheme infix syntax extension (see
*Note Scheme infix syntax extension::). Otherwise, `gsc' will parse the
Scheme file using the normal Scheme prefix syntax.  Files with a C file
extension must have been previously produced by `gsc', with the `-c'
option, and are used by Gambit's linker.

   For each Scheme file a C file `FILE.c' will be produced.  The C
file's name is the same as the Scheme file, but the extension is
changed to `.c'.  By default the C file is created in the same
directory as the Scheme file.  This default can be overridden with the
compiler's `-o' option.

   The C files produced by the compiler serve two purposes.  They will
be processed by a C compiler to generate object files, and they also
contain information to be read by Gambit's linker to generate a "link
file".  The link file is a C file that collects various linking
information for a group of modules, such as the set of all symbols and
global variables used by the modules.  The linker is only invoked when
the `-link' or `-exe' options appear on the command line.

   Compiler options must be specified before the first file name and
after the `-:' runtime option (*note Runtime options::).  If present,
the `-i', `-f', and `-v' compiler options must come first.  The
available options are:

`-i'
     Force interpreter mode.

`-f'
     Do not examine the initialization file.

`-v'
     Print the system version string, system time stamp, operating
     system type, and configure script options on standard output and
     exit.

`-prelude EXPRESSIONS'
     Add expressions to the top of the source code being compiled.

`-postlude EXPRESSIONS'
     Add expressions to the bottom of the source code being compiled.

`-cc-options OPTIONS'
     Add options to the command that invokes the C compiler.

`-ld-options-prelude OPTIONS'
     Add options to the command that invokes the C linker.

`-ld-options OPTIONS'
     Add options to the command that invokes the C linker.

`-warnings'
     Display warnings.

`-verbose'
     Display a trace of the compiler's activity.

`-report'
     Display a global variable usage report.

`-expansion'
     Display the source code after expansion.

`-gvm'
     Generate a listing of the GVM code.

`-debug'
     Include all debugging information in the code generated.

`-debug-location'
     Include source code location debugging information in the code
     generated.

`-debug-source'
     Include the source code debugging information in the code
     generated.

`-debug-environments'
     Include environment debugging information in the code generated.

`-track-scheme'
     Generate `#line' directives referring back to the Scheme code.

`-o OUTPUT'
     Set name of output file or directory where output file(s) are
     written.

`-dynamic'
     Compile Scheme source files to dynamically loadable object files
     (this is the default).

`-exe'
     Compile Scheme source files into an executable program.

`-obj'
     Compile Scheme source files to object files.

`-keep-c'
     Keep any intermediate `.c' files that are generated.

`-c'
     Compile Scheme source files to C without generating link file.

`-link'
     Compile Scheme source files to C and generate a link file.

`-flat'
     Generate a flat link file instead of the default incremental link
     file.

`-l BASE'
     Specify the link file of the base library to use for the link.

`-'
     Start REPL interaction.

`-e EXPRESSIONS'
     Evaluate expressions in the interaction environment.

   The `-i' option forces the compiler to process the remaining command
line arguments like the interpreter.

   The `-prelude' option adds the specified expressions to the top of
the source code being compiled.  The main use of this option is to
supply declarations on the command line.  For example the following
invocation of the compiler will compile the file `bench.scm' in unsafe
mode:

     $ gsc -prelude "(declare (not safe))" bench.scm

   The `-postlude' option adds the specified expressions to the bottom
of the source code being compiled.  The main use of this option is to
supply the expression that will start the execution of the program.  For
example:

     $ gsc -postlude "(start-bench)" bench.scm

   The `-cc-options' option is only meaningful when a dynamically
loadable object file is being generated (neither the `-c' or `-link'
options are used).  The `-cc-options' option adds the specified options
to the command that invokes the C compiler.  The main use of this
option is to specify the include path, some symbols to define or
undefine, the optimization level, or any C compiler option that is
different from the default.  For example:

     $ gsc -cc-options "-U___SINGLE_HOST -O2 -I../include" bench.scm

   The `-ld-options-prelude' and `-ld-options' options are only
meaningful when a dynamically loadable object file is being generated
(neither the `-c' or `-link' options are used).  The
`-ld-options-prelude' and `-ld-options' options add the specified
options to the command that invokes the C linker (the options in
LD-OPTIONS-PRELUDE are passed to the C linker before the input file and
the options in LD-OPTIONS are passed after).  The main use of this
option is to specify additional object files or libraries that need to
be linked, or any C linker option that is different from the default
(such as the library search path and flags to select between static and
dynamic linking).  For example:

     $ gsc -ld-options "-L/usr/X11R6/lib -lX11 -dynamic" bench.scm

   The `-warnings' option displays on standard output all warnings that
the compiler may have.

   The `-verbose' option displays on standard output a trace of the
compiler's activity.

   The `-report' option displays on standard output a global variable
usage report.  Each global variable used in the program is listed with
4 flags that indicate whether the global variable is defined,
referenced, mutated and called.

   The `-expansion' option displays on standard output the source code
after expansion and inlining by the front end.

   The `-gvm' option generates a listing of the intermediate code for
the "Gambit Virtual Machine" (GVM) of each Scheme file on `FILE.gvm'.

   The `-debug' option causes all kinds of debugging information to be
saved in the code generated.  See the documentation of the `debug'
declaration for details.

   The `-debug-location' option causes source code location debugging
information to be saved in the code generated.  See the documentation
of the `debug-location' declaration for details.

   The `-debug-source' option causes source code debugging information
to be saved in the code generated.  See the documentation of the
`debug-source' declaration for details.

   The `-debug-environments' option causes environment debugging
information to be saved in the code generated.  See the documentation
of the `debug-environments' declaration for details.

   The `-track-scheme' options causes the generation of `#line'
directives that refer back to the Scheme source code.  This allows the
use of a C debugger or profiler to debug Scheme code.

   The `-o' option sets the filename of the output file, or the
directory in which the output file(s) generated by the compiler are
written.

   If the `-link' or `-exe' options appear on the command line, the
Gambit linker is invoked to generate the link file from the set of C
files specified on the command line or produced by the Gambit compiler.
By default the link file is `LAST_.c', where `LAST.c' is the last file
in the set of C files.  When the `-c' option is specified, the Scheme
source files are compiled to C files.  When the `-exe' option is
specified, the generated C files and link file are compiled and linked
using the C compiler to produce an executable program whose name
defaults to `LAST.exe'.  When the `-obj' option is specified, the
generated C files are compiled using the C compiler to produce object
files (`.o' or `.obj' extensions).  If neither the `-link', `-c',
`-exe', `-obj' options appear on the command line, the Scheme source
files are compiled to dynamically loadable object files (`.oN'
extension).  The `-keep-c' option will prevent the deletion of any
intermediate `.c' file that is generated.  Note that in this case the
intermediate `.c' file will be generated in the same directory as the
Scheme source file even if the `-o' option is used.

   The `-flat' option is only meaningful when a link file is being
generated (i.e. the `-link' or `-exe' options also appear on the
command line).  The `-flat' option directs the Gambit linker to
generate a flat link file.  By default, the linker generates an
incremental link file (see the next section for a description of the
two types of link files).

   The `-l' option is only meaningful when an incremental link file is
being generated (i.e. the `-link' or `-exe' options appear on the
command line and the `-flat' option is absent).  The `-l' option
specifies the link file (without the `.c' extension) of the base
library to use for the incremental link.  By default the link file of
the Gambit runtime library is used (i.e. `~~lib/_gambit.c').

   The `-' option starts a REPL interaction.

   The `-e' option evaluates the specified expressions in the
interaction environment.

3.4 Link files
==============

Gambit can be used to create programs and libraries of Scheme modules.
This section explains the steps required to do so and the role played
by the link files.

   In general, a program is composed of a set of Scheme modules and C
modules.  Some of the modules are part of the Gambit runtime library and
the other modules are supplied by the user.  When the program is
started it must setup various global tables (including the symbol table
and the global variable table) and then sequentially execute the Scheme
modules (more or less as though they were being loaded one after
another).  The information required for this is contained in one or
more "link files" generated by the Gambit linker from the C files
produced by the Gambit compiler.

   The order of execution of the Scheme modules corresponds to the
order of the modules on the command line which produced the link file.
The order is usually important because most modules define variables and
procedures which are used by other modules (for this reason the
program's main computation is normally started by the last module).

   When a single link file is used to contain the linking information of
all the Scheme modules it is called a "flat link file".  Thus a program
built with a flat link file contains in its link file both information
on the user modules and on the runtime library.  This is fine if the
program is to be statically linked but is wasteful in a shared-library
context because the linking information of the runtime library can't be
shared and will be duplicated in all programs (this linking information
typically takes hundreds of kilobytes).

   Flat link files are mainly useful to bundle multiple Scheme modules
to make a runtime library (such as the Gambit runtime library) or to
make a single file that can be loaded with the `load' procedure.

   An "incremental link file" contains only the linking information
that is not already contained in a second link file (the "base" link
file).  Assuming that a flat link file was produced when the runtime
library was linked, a program can be built by linking the user modules
with the runtime library's link file, producing an incremental link
file.  This allows the creation of a shared-library which contains the
modules of the runtime library and its flat link file.  The program is
dynamically linked with this shared-library and only contains the user
modules and the incremental link file.  For small programs this
approach greatly reduces the size of the program because the
incremental link file is small.  A "hello world" program built this way
can be as small as 5 Kbytes.  Note that it is perfectly fine to use an
incremental link file for statically linked programs (there is very
little loss compared to a single flat link file).

   Incremental link files may be built from other incremental link
files.  This allows the creation of shared-libraries which extend the
functionality of the Gambit runtime library.

3.4.1 Building an executable program
------------------------------------

The simplest way to create an executable program is to invoke `gsc'
with the `-exe' option.  The compiler will transparently perform all
the steps necessary, including compiling Scheme source files to C
files, generating the link file, compiling the C files generated to
object files, and creating the final executable file using the C
linker.  The following example shows how to build the executable
program `hello.exe' which contains the two Scheme modules `h.scm' and
`w.six'.

     $ cat h.scm
     (display "hello") (newline)
     $ cat w.six
     display("world"); newline();
     $ gsc -o hello.exe -exe h.scm w.six
     h.scm:
     /Users/feeley/gambit/doc/h.c:
     w.six:
     /Users/feeley/gambit/doc/w.c:
     /Users/feeley/gambit/doc/w_.c:
     $ ./hello.exe
     hello
     world

   The detailed steps which are performed can be viewed by setting the
`GAMBCOMP_VERBOSE' environment variable to a nonnull value.  For
example:

     $ export GAMBCOMP_VERBOSE=yes
     $ gsc -o hello.exe -exe h.scm w.six
     h.scm:
     /Users/feeley/gambit/doc/h.c:
     gcc -no-cpp-precomp -Wno-unused -O1 -fno-math-errno -fschedule-insns2
      -fno-trapping-math -fno-strict-aliasing -fwrapv -fomit-frame-pointer
      -fPIC -fno-common -mieee-fp -I"/usr/local/Gambit/include" -c -o "h.o" h.c
     w.six:
     /Users/feeley/gambit/doc/w.c:
     gcc -no-cpp-precomp -Wno-unused -O1 -fno-math-errno -fschedule-insns2
      -fno-trapping-math -fno-strict-aliasing -fwrapv -fomit-frame-pointer
      -fPIC -fno-common -mieee-fp -I"/usr/local/Gambit/include" -c -o "w.o" w.c
     /Users/feeley/gambit/doc/w_.c:
     gcc -no-cpp-precomp -Wno-unused -O1 -fno-math-errno -fschedule-insns2
      -fno-trapping-math -fno-strict-aliasing -fwrapv -fomit-frame-pointer
      -fPIC -fno-common -mieee-fp -I"/usr/local/Gambit/include" -c -o "w_.o" w_.c
     gcc  -no-cpp-precomp -Wno-unused -O1 -fno-math-errno -fschedule-insns2
      -fno-trapping-math -fno-strict-aliasing -fwrapv -fomit-frame-pointer
      -fPIC -fno-common -mieee-fp -I"/usr/local/Gambit/include"
      -o "hello.exe" h.o w.o w_.o "/usr/local/Gambit/lib/libgambit.a"

   Using a single invocation of `gsc' with the `-exe' option is
sometimes inappropriate when the build process is more complex, for
example when the program is composed of several seperately compiled
modules.  In such a case it is useful to decompose the build process
into smaller compilation steps.  The `hello.exe' executable program
could have been built by seperating the generation of C files from the
C compilation and linking:

     $ gsc -c h.scm
     $ gsc -c w.six
     $ gsc -o hello.exe -exe h.c w.c

   When even finer control is desired the build process can be
decomposed into smaller steps that invoke the C compiler and linker
explicitly.  This is described in the rest of this section.

   The `gsc' compiler can be invoked to compile each Scheme module into
a C file and to create an incremental link file.  The C files and the
link file must then be compiled with a C compiler and linked (at the
object file level) with the Gambit runtime library and possibly other
libraries (such as the math library and the dynamic loading library).

   Here is for example how a program with three modules (one in C and
two in Scheme) can be built.  The content of the three source files
(`m1.c', `m2.scm' and `m3.scm') is:

     /* File: "m1.c" */
     int power_of_2 (int x) { return 1<<x; }

     ; File: "m2.scm"
     (c-declare "extern int power_of_2 ();")
     (define pow2 (c-lambda (int) int "power_of_2"))
     (define (twice x) (cons x x))

     ; File: "m3.scm"
     (write (map twice (map pow2 '(1 2 3 4)))) (newline)

   The compilation of the two Scheme source files can be done with
three invocations of `gsc':

     $ gsc -c m2.scm        # create m2.c (note: .scm is optional)
     $ gsc -c m3.scm        # create m3.c (note: .scm is optional)
     $ gsc -link m2.c m3.c  # create the incremental link file m3_.c

   Alternatively, the three invocations of `gsc' can be replaced by a
single invocation:

     $ gsc -link m2 m3
     m2:
     m3:

   At this point there will be 4 C files: `m1.c', `m2.c', `m3.c', and
`m3_.c'.  To produce an executable program these files must be compiled
with a C compiler and linked with the Gambit runtime library.  The C
compiler options needed will depend on the C compiler and the operating
system (in particular it may be necessary to add the options
`-I/usr/local/Gambit/include -L/usr/local/Gambit/lib' to access the
`gambit.h' header file and the Gambit runtime library).

   Here is an example under Mac OS X:

     $ uname -srmp
     Darwin 8.1.0 Power Macintosh powerpc
     $ gsc -obj m1.c m2.c m3.c m3_.c
     m1.c:
     m2.c:
     m3.c:
     m3_.c:
     $ gcc m1.o m2.o m3.o m3_.o -lgambit
     $ ./a.out
     ((2 . 2) (4 . 4) (8 . 8) (16 . 16))

   Here is an example under Linux:

     $ uname -srmp
     Linux 2.6.8-1.521 i686 athlon
     $ gsc -obj m1.c m2.c m3.c m3_.c
     m1.c:
     m2.c:
     m3.c:
     m3_.c:
     $ gcc m1.o m2.o m3.o m3_.o -lgambit -lm -ldl -lutil
     $ ./a.out
     ((2 . 2) (4 . 4) (8 . 8) (16 . 16))

3.4.2 Building a loadable library
---------------------------------

To bundle multiple modules into a single object file that can be
dynamically loaded with the `load' procedure, a flat link file is
needed.  The compiler's `-o' option must be used to name the C file
generated as follows.  If the dynamically loadable object file is to be
named `MYFILE.oN' then the `-o' option must set the name of the link
file generated to `MYFILE.oN.c' (note that the `.c' extension could
also be `.cc', `.cpp' or whatever extension is appropriate for C/C++
source files).  The three modules of the previous example can be
bundled by generating a link file in this way:

     $ gsc -link -flat -o foo.o1.c m2 m3
     m2:
     m3:
     *** WARNING -- "cons" is not defined,
     ***            referenced in: ("m2.c")
     *** WARNING -- "map" is not defined,
     ***            referenced in: ("m3.c")
     *** WARNING -- "newline" is not defined,
     ***            referenced in: ("m3.c")
     *** WARNING -- "write" is not defined,
     ***            referenced in: ("m3.c")

   The warnings indicate that there are no definitions (`define's or
`set!'s) of the variables `cons', `map', `newline' and `write' in the
set of modules being linked.  Before `foo.o1' is loaded, these
variables will have to be bound; either implicitly (by the runtime
library) or explicitly.

   When compiling the C files and link file generated, the flag
`-D___DYNAMIC' must be passed to the C compiler and the C compiler and
linker must be told to generate a dynamically loadable shared library.

   Here is an example under Mac OS X:

     $ uname -srmp
     Darwin 10.5.0 i386 i386
     $ gsc -link -flat -o foo.o1.c m2 m3 > /dev/null
     m2:
     m3:
     $ gsc -cc-options "-D___DYNAMIC" -obj m1.c m2.c m3.c foo.o1.c
     m1.c:
     m2.c:
     m3.c:
     foo.o1.c:
     $ gcc -bundle m1.o m2.o m3.o foo.o1.o -o foo.o1
     $ gsi foo.o1
     ((2 . 2) (4 . 4) (8 . 8) (16 . 16))

   Here is an example under Linux:

     $ uname -srmp
     Linux 2.6.8-1.521 i686 athlon
     $ gsc -link -flat -o foo.o1.c m2 m3 > /dev/null
     m2:
     m3:
     $ gsc -cc-options "-D___DYNAMIC" -obj m1.c m2.c m3.c foo.o1.c
     m1.c:
     m2.c:
     m3.c:
     foo.o1.c:
     $ gcc -shared m1.o m2.o m3.o foo.o1.o -o foo.o1
     $ gsi foo.o1
     ((2 . 2) (4 . 4) (8 . 8) (16 . 16))

   Here is a more complex example, under Solaris, which shows how to
build a loadable library `mymod.o1' composed of the files `m4.scm',
`m5.scm' and `x.c' that links to system shared libraries (for
X-windows):

     $ uname -srmp
     SunOS ungava 5.6 Generic_105181-05 sun4m sparc SUNW,SPARCstation-20
     $ gsc -link -flat -o mymod.o1.c m4 m5
     m4:
     m5:
     *** WARNING -- "*" is not defined,
     ***            referenced in: ("m4.c")
     *** WARNING -- "+" is not defined,
     ***            referenced in: ("m5.c")
     *** WARNING -- "display" is not defined,
     ***            referenced in: ("m5.c" "m4.c")
     *** WARNING -- "newline" is not defined,
     ***            referenced in: ("m5.c" "m4.c")
     *** WARNING -- "write" is not defined,
     ***            referenced in: ("m5.c")
     $ gsc -cc-options "-D___DYNAMIC" -obj m4.c m5.c x.c mymod.o1.c
     m4.c:
     m5.c:
     x.c:
     mymod.o1.c:
     $ /usr/ccs/bin/ld -G -o mymod.o1 mymod.o1.o m4.o m5.o x.o -lX11 -lsocket
     $ gsi mymod.o1
     hello from m4
     hello from m5
     (f1 10) = 22
     $ cat m4.scm
     (define (f1 x) (* 2 (f2 x)))
     (display "hello from m4")
     (newline)

     (c-declare #<<c-declare-end
     #include "x.h"
     c-declare-end
     )
     (define x-initialize (c-lambda (char-string) bool "x_initialize"))
     (define x-display-name (c-lambda () char-string "x_display_name"))
     (define x-bell (c-lambda (int) void "x_bell"))
     $ cat m5.scm
     (define (f2 x) (+ x 1))
     (display "hello from m5")
     (newline)

     (display "(f1 10) = ")
     (write (f1 10))
     (newline)

     (x-initialize (x-display-name))
     (x-bell 50) ; sound the bell at 50%
     $ cat x.c
     #include <X11/Xlib.h>

     static Display *display;

     int x_initialize (char *display_name)
     {
       display = XOpenDisplay (display_name);
       return display != NULL;
     }

     char *x_display_name (void)
     {
       return XDisplayName (NULL);
     }

     void x_bell (int volume)
     {
       XBell (display, volume);
       XFlush (display);
     }
     $ cat x.h
     int x_initialize (char *display_name);
     char *x_display_name (void);
     void x_bell (int);

3.4.3 Building a shared-library
-------------------------------

A shared-library can be built using an incremental link file or a flat
link file.  An incremental link file is normally used when the Gambit
runtime library (or some other library) is to be extended with new
procedures.  A flat link file is mainly useful when building a "primal"
runtime library, which is a library (such as the Gambit runtime
library) that does not extend another library.  When compiling the C
files and link file generated, the flags `-D___LIBRARY' and
`-D___SHARED' must be passed to the C compiler.  The flag `-D___PRIMAL'
must also be passed to the C compiler when a primal library is being
built.

   A shared-library `mylib.so' containing the two first modules of the
previous example can be built this way:

     $ uname -srmp
     Linux bailey 1.2.13 #2 Wed Aug 28 16:29:41 GMT 1996 i586
     $ gsc -link -o mylib.c m2
     $ gsc -obj -cc-options "-D___SHARED" m1.c m2.c mylib.c
     m1.c:
     m2.c:
     mylib.c:
     $ gcc -shared  m1.o m2.o mylib.o -o mylib.so

   Note that this shared-library is built using an incremental link file
(it extends the Gambit runtime library with the procedures `pow2' and
`twice').  This shared-library can in turn be used to build an
executable program from the third module of the previous example:

     $ gsc -link -l mylib m3
     $ gsc -obj m3.c m3_.c
     m3.c:
     m3_.c:
     $ gcc m3.o m3_.o mylib.so -lgambit
     $ LD_LIBRARY_PATH=.:/usr/local/lib ./a.out
     ((2 . 2) (4 . 4) (8 . 8) (16 . 16))

3.4.4 Other compilation options
-------------------------------

The performance of the code can be increased by passing the
`-D___SINGLE_HOST' flag to the C compiler.  This will merge all the
procedures of a module into a single C procedure, which reduces the
cost of intra-module procedure calls.  In addition the `-O' option can
be passed to the C compiler.  For large modules, it will not be
practical to specify both `-O' and `-D___SINGLE_HOST' for typical C
compilers because the compile time will be high and the C compiler
might even fail to compile the program for lack of memory.  It has been
observed that lower levels of optimization (e.g. `-O1') often give
faster compilation and also generate faster code.  It is a good idea to
experiment.

   Normally C compilers will not automatically search
`/usr/local/Gambit/include' for header files so the flag
`-I/usr/local/Gambit/include' should be passed to the C compiler.
Similarly, C compilers/linkers will not automatically search
`/usr/local/Gambit/lib' for libraries so the flag
`-L/usr/local/Gambit/lib' should be passed to the C compiler/linker.
Alternatives are given in *Note Accessing the system files::.

   A variety of flags are needed by some C compilers when compiling a
shared-library or a dynamically loadable library.  Some of these flags
are: `-shared', `-call_shared', `-rdynamic', `-fpic', `-fPIC', `-Kpic',
`-KPIC', `-pic', `+z', `-G'.  Check your compiler's documentation to see
which flag you need.

3.5 Procedures specific to compiler
===================================

The Gambit Scheme compiler features the following procedures that are
not available in the Gambit Scheme interpreter.

 -- procedure: compile-file-to-target FILE [`options:' OPTIONS]
          [`output:' OUTPUT] [`module-name:' MODULE-NAME]
     The FILE parameter must be a string naming an existing file
     containing Scheme source code.  The extension can be omitted from
     FILE when the Scheme file has a `.scm' or `.six' extension.  This
     procedure compiles the source file into a file containing C code.
     By default, this file is named after FILE with the extension
     replaced with `.c'.  The name of the generated file can be
     specified with the OUTPUT parameter.  If OUTPUT is a string naming
     a directory then the C file is created in that directory.
     Otherwise the name of the C file is OUTPUT.  The name of the
     generated module can be specified with the MODULE-NAME parameter.
     If MODULE-NAME is `#f' or is not specified, then the name of the
     module is derived from the name of the C file generated, without
     the extension.

     Compilation options are specified through the OPTIONS parameter
     which must be a list of symbols.  Any combination of the following
     options can be used: `verbose', `report', `expansion', `gvm', and
     `debug'.

     When the compilation is successful, `compile-file-to-target'
     returns the name of the C file generated.  When there is a
     compilation error, `#f' is returned.

          $ cat h.scm
          (display "hello") (newline)
          $ gsc
          Gambit v4.8.5

          > (compile-file-to-target "h")
          "/Users/feeley/gambit/doc/h.c"

 -- procedure: compile-file FILE [`options:' OPTIONS] [`output:'
          OUTPUT] [`cc-options:' CC-OPTIONS] [`ld-options-prelude:'
          LD-OPTIONS-PRELUDE] [`ld-options:' LD-OPTIONS]
     The FILE, OPTIONS, and OUTPUT parameters have the same meaning as
     for the `compile-file-to-target' procedure, except that FILE may
     be a Scheme source file or a C file possibly generated by the
     Gambit Scheme compiler (for example with the
     `compile-file-to-target' procedure).  The CC-OPTIONS parameter is
     a string containing the options to pass to the C compiler and the
     LD-OPTIONS-PRELUDE and LD-OPTIONS parameters are strings
     containing the options to pass to the C linker (the options in
     LD-OPTIONS-PRELUDE are passed to the C linker before the input
     file and the options in LD-OPTIONS are passed after).

     The `compile-file' procedure compiles the source file FILE into an
     object file, which is either a file dynamically loadable using the
     `load' procedure, or a C linkable object file destined to be
     linked with the C linker (for example to create a standalone
     executable program).  The presence of the `obj' option in OPTIONS
     will cause the creation of a C linkable object file and therefore
     the options LD-OPTIONS-PRELUDE and LD-OPTIONS are ignored,
     otherwise a dynamically loadable file is created.  In both cases,
     if FILE is a Scheme source file, the compiler first compiles FILE
     to a C file which is created in the same directory as FILE
     regardless of the OUTPUT parameter.  Then the C file is compiled
     with the C compiler.

     When the compilation is successful, `compile-file' returns the
     name of the object file generated.  When there is a compilation
     error, `#f' is returned.

     The name of the object file can be specified with the OUTPUT
     parameter.  If OUTPUT is a string naming a directory then the
     object file is created in that directory.  Otherwise the name of
     the object file is OUTPUT.

     In the case of a dynamically loadable object file, by default the
     object file is named after FILE with the extension replaced with
     `.oN', where N is a positive integer that acts as a version
     number.  The next available version number is generated
     automatically by `compile-file'.

     When dynamically loaded object files are loaded using the `load'
     procedure, the `.oN' extension can be specified (to select a
     particular version) or omitted (to load the file with a `.oN'
     extension with the highest N consecutively from 1).  When the
     `.oN' extension is not specified and older versions are no longer
     needed, all versions must be deleted and the compilation must be
     repeated (this is necessary because the file name, including the
     extension, is used to name some of the exported symbols of the
     object file).

     Note that dynamically loadable object files can only be generated
     on host operating systems that support dynamic loading.

          $ cat h.scm
          (display "hello") (newline)
          $ gsc
          Gambit v4.8.5

          > (compile-file "h")
          "/Users/feeley/gambit/doc/h.o1"
          > (load "h")
          hello
          "/Users/feeley/gambit/doc/h.o1"
          > (compile-file-to-target "h" output: "h.o99.c")
          "/Users/feeley/gambit/doc/h.o99.c"
          > (compile-file "h.o99.c")
          "/Users/feeley/gambit/doc/h.o99"
          > (load "h.o99")
          hello
          "/Users/feeley/gambit/doc/h.o99"
          > (compile-file-to-target "h")
          "/Users/feeley/gambit/doc/h.c"
          > (compile-file "h.c" options: '(obj))
          "/Users/feeley/gambit/doc/h.o"


 -- procedure: link-incremental MODULE-LIST [`output:' OUTPUT] [`base:'
          BASE] [`warnings?:' WARNINGS?]
     The first parameter must be a non empty list of strings naming
     Scheme modules to link (the file extension may be omitted).  An
     incremental link file is generated for the modules specified in
     MODULE-LIST.  By default the link file generated is named
     `LAST_.c', where LAST is the name of the last module, without the
     file extension.  The name of the generated link file can be
     specified with the OUTPUT parameter.  If OUTPUT is a string naming
     a directory then the link file is created in that directory.
     Otherwise the name of the link file is OUTPUT.

     The base link file is specified by the BASE parameter, which must
     be a string.  By default the base link file is the Gambit runtime
     library link file `~~lib/_gambit.c'.  However, when BASE is
     supplied it is the name of the base link file (the file extension
     may be omitted).

     The WARNINGS? parameter controls whether warnings are generated
     for undefined references.

     The following example shows how to build the executable program
     `hello' which contains the two Scheme modules `h.scm' and `w.six'.

          $ uname -srmp
          Darwin 8.1.0 Power Macintosh powerpc
          $ cat h.scm
          (display "hello") (newline)
          $ cat w.six
          display("world"); newline();
          $ gsc
          Gambit v4.8.5

          > (compile-file-to-target "h")
          "/Users/feeley/gambit/doc/h.c"
          > (compile-file-to-target "w")
          "/Users/feeley/gambit/doc/w.c"
          > (link-incremental '("h" "w") output: "hello.c")
          "/Users/feeley/gambit/doc/hello_.c"
          > ,q
          $ gsc -obj h.c w.c hello.c
          h.c:
          w.c:
          hello.c:
          $ gcc h.o w.o hello.o -lgambit -o hello
          $ ./hello
          hello
          world


 -- procedure: link-flat MODULE-LIST [`output:' OUTPUT] [`warnings?:'
          WARNINGS?]
     The first parameter must be a non empty list of strings naming
     Scheme modules to link (the file extension may be omitted).  The
     first string must be the name of a Scheme module or the name of a
     link file and the remaining strings must name Scheme modules.  A
     flat link file is generated for the modules specified in
     MODULE-LIST.  By default the link file generated is named
     `LAST_.c', where LAST is the name of the last module.  The name of
     the generated link file can be specified with the OUTPUT
     parameter.  If OUTPUT is a string naming a directory then the link
     file is created in that directory.  Otherwise the name of the link
     file is OUTPUT.  If a dynamically loadable object file is produced
     from the link file `OUTPUT', then the name of the dynamically
     loadable object file must be `OUTPUT' stripped of its file
     extension.

     The WARNINGS? parameter controls whether warnings are generated
     for undefined references.

     The following example shows how to build the dynamically loadable
     object file `lib.o1' which contains the two Scheme modules
     `m6.scm' and `m7.scm'.

          $ uname -srmp
          Darwin 8.1.0 Power Macintosh powerpc
          $ cat m6.scm
          (define (f x) (g (* x x)))
          $ cat m7.scm
          (define (g y) (+ n y))
          $ gsc
          Gambit v4.8.5

          > (compile-file-to-target "m6")
          "/Users/feeley/gambit/doc/m6.c"
          > (compile-file-to-target "m7")
          "/Users/feeley/gambit/doc/m7.c"
          > (link-flat '("m6" "m7") output: "lib.o1.c")
          *** WARNING -- "*" is not defined,
          ***            referenced in: ("m6.c")
          *** WARNING -- "+" is not defined,
          ***            referenced in: ("m7.c")
          *** WARNING -- "n" is not defined,
          ***            referenced in: ("m7.c")
          "/Users/feeley/gambit/doc/lib.o1.c"
          > ,q
          $ gcc -bundle -D___DYNAMIC m6.c m7.c lib.o1.c -o lib.o1
          $ gsc
          Gambit v4.8.5

          > (load "lib")
          *** WARNING -- Variable "n" used in module "m7" is undefined
          "/Users/feeley/gambit/doc/lib.o1"
          > (define n 10)
          > (f 5)
          35
          > ,q

     The warnings indicate that there are no definitions (`define's or
     `set!'s) of the variables `*', `+' and `n' in the modules
     contained in the library.  Before the library is used, these
     variables will have to be bound; either implicitly (by the runtime
     library) or explicitly.


4 Runtime options
*****************

Both `gsi' and `gsc' as well as executable programs compiled and linked
using `gsc' take a `-:' option which supplies parameters to the runtime
system.  This option must appear first on the command line.  The colon
is followed by a comma separated list of options with no intervening
spaces.  The available options are:

`mHEAPSIZE'
     Set minimum heap size in kilobytes.

`hHEAPSIZE'
     Set maximum heap size in kilobytes.

`lLIVEPERCENT'
     Set heap occupation after garbage collection.

`s'
     Select standard Scheme mode.

`S'
     Select Gambit Scheme mode.

`d[OPT...]'
     Set debugging options.

`@[INTF][:PORT]'
     Override the configuration of the main RPC server.

`=DIRECTORY'
     Override the central installation directory.

`~~DIR=DIRECTORY'
     Override the DIR installation directory.

`+ARGUMENT'
     Add ARGUMENT to the command line before other arguments.

`f[OPT...]'
     Set file options.

`t[OPT...]'
     Set terminal options.

`-[OPT...]'
     Set standard input and output options.

   The `m' option specifies the minimum size of the heap.  The `m' is
immediately followed by an integer indicating the number of kilobytes
of memory.  The heap will not shrink lower than this size.  By default,
the minimum size is 0.

   The `h' option specifies the maximum size of the heap.  The `h' is
immediately followed by an integer indicating the number of kilobytes
of memory.  The heap will not grow larger than this size.  By default,
there is no limit (i.e.  the heap will grow until the virtual memory is
exhausted).

   The `l' option specifies the percentage of the heap that will be
occupied with live objects after the heap is resized at the end of a
garbage collection.  The `l' is immediately followed by an integer
between 1 and 100 inclusively indicating the desired percentage.  The
garbage collector resizes the heap to reach this percentage occupation.
By default, the percentage is 50.

   The `s' option selects standard Scheme mode.  In this mode the
reader is case-insensitive and keywords are not recognized.  The `S'
option selects Gambit Scheme mode (the reader is case-sensitive and
recognizes keywords which end with a colon).  By default Gambit Scheme
mode is used.

   The `d' option sets various debugging options.  The letter `d' is
followed by a sequence of letters indicating suboptions.

`p'
     Uncaught exceptions will be treated as "errors" in the primordial
     thread only.

`a'
     Uncaught exceptions will be treated as "errors" in all threads.

`r'
     When an "error" occurs a new REPL will be started.

`s'
     When an "error" occurs a new REPL will be started.  Moreover the
     program starts in single-stepping mode.

`q'
     When an "error" occurs the program will terminate with a nonzero
     exit status.

`R'
     When a user interrupt occurs a new REPL will be started.  User
     interrupts are typically obtained by typing <^C>.  Note that with
     some system configurations <^C> abruptly terminates the process.
     For example, under Microsoft Windows, <^C> works fine with the
     standard console but with the MSYS terminal window it terminates
     the process.

`D'
     When a user interrupt occurs it will be deferred until the
     parameter `current-user-interrupt-handler' is bound.

`Q'
     When a user interrupt occurs the program will terminate with a
     nonzero exit status.

`LEVEL'
     The verbosity level is set to LEVEL (a digit from 0 to 9).  At
     level 0 the runtime system will not display error messages and
     warnings.

`i'
     The REPL interaction channel will be the IDE REPL window (if the
     IDE is available).

`c'
     The REPL interaction channel will be the console.

`-'
     The REPL interaction channel will be standard input and standard
     output.

`@[HOST][:PORT]'
     The REPL interaction channel will be connected to the remote
     debugger at address HOST:PORT (if there is a remote debugger at
     that address).  The default HOST is 127.0.0.1 and the default PORT
     is 44555.  THIS OPTION IS NOT YET IMPLEMENTED!


   The default debugging options are equivalent to `-:dpqQ1i' (i.e. an
uncaught exception in the primordial thread terminates the program
after displaying an error message).  When the letter `d' is not
followed by suboptions, it is equivalent to `-:dprR1i' (i.e. a new REPL
is started only when an uncaught exception occurs in the primordial
thread).  When `gsi' and `gsc' are running the main REPL, the debugging
options are changed to cause errors in the primordial thread and user
interrupts to start a nested REPL.

   The `@[INTF][:PORT]' option overrides the configuration of the main
RPC server.  The default INTF is 127.0.0.1 and the default PORT is
44556.  THIS OPTION IS NOT YET IMPLEMENTED!

   The `=DIRECTORY' option overrides the setting of the central
installation directory.

   The `~~DIR=DIRECTORY' option overrides the setting of the DIR
installation directory.

   The `+' option adds the text that follows to the command line before
other arguments.

   The `f', `t' and `-' options specify the default settings of the
ports created for files, terminals and standard input and output
respectively.  The default character encoding, end-of-line encoding and
buffering can be set.  Moreover, for terminals the line-editing feature
can be enabled or disabled.  The `f', `t' and `-' must be followed by a
sequence of these options:

`A'
     ASCII character encoding.

`1'
     ISO-8859-1 character encoding.

`2'
     UCS-2 character encoding.

`4'
     UCS-4 character encoding.

`6'
     UTF-16 character encoding.

`8'
     UTF-8 character encoding.

`U'
     UTF character encoding with fallback to UTF-8 on input if no BOM
     is present.

`UA'
     UTF character encoding with fallback to ASCII on input if no BOM
     is present.

`U1'
     UTF character encoding with fallback to ISO-8859-1 on input if no
     BOM is present.

`U6'
     UTF character encoding with fallback to UTF-16 on input if no BOM
     is present.

`U8'
     UTF character encoding with fallback to UTF-8 on input if no BOM
     is present.

`c'
     End-of-line is encoded as CR (carriage-return).

`l'
     End-of-line is encoded as LF (linefeed)

`cl'
     End-of-line is encoded as CR-LF.

`u'
     Unbuffered I/O.

`n'
     Line buffered I/O (`n' for "at newline").

`f'
     Fully buffered I/O.

`r'
     Illegal character encoding is treated as an error (exception
     raised).

`R'
     Silently replace illegal character encodings with Unicode
     character #xfffd (replacement character).

`e'
     Enable line-editing (applies to terminals only).

`E'
     Disable line-editing (applies to terminals only).

   When a program's execution starts, the runtime system obtains the
runtime options by processing in turn various sources of runtime
options: the defaults, the environment variable `GAMBOPT', the script
line of the source code, and, unless the program is an interpreted
script, the first command line argument of the program.  Any runtime
option can be overriden by a subsequent source of runtime options.  It
is sometimes useful to prevent overriding the runtime options of the
script line.  This can be achieved by starting the script line runtime
options with `-::'.  In this case the environment variable `GAMBOPT' is
ignored, and the first command line argument of the program is not used
for runtime options (it is treated like a normal command line argument
even if it starts with `-:').

   For example:

     $ GAMBOPT=d0,=~/my-gambit2
     $ export GAMBOPT
     $ gsi -e '(pretty-print (path-expand "~~")) (/ 1 0)'
     "/Users/feeley/my-gambit2/"
     $ echo $?
     70
     $ gsi -:d1 -e '(pretty-print (path-expand "~~")) (/ 1 0)'
     "/Users/feeley/my-gambit2/"
     *** ERROR IN (string)@1.3 -- Divide by zero
     (/ 1 0)

5 Debugging
***********

5.1 Debugging model
===================

The evaluation of an expression may stop before it is completed for the
following reasons:

  a. An evaluation error has occured, such as attempting to divide by
     zero.

  b. The user has interrupted the evaluation (usually by typing <^C>).

  c. A breakpoint has been reached or `(step)' was evaluated.

  d. Single-stepping mode is enabled.


   When an evaluation stops, a message is displayed indicating the
reason and location where the evaluation was stopped.  The location
information includes, if known, the name of the procedure where the
evaluation was stopped and the source code location in the format
`STREAM@LINE.COLUMN', where STREAM is either a string naming a file or
a symbol within parentheses, such as `(console)'.

   A "nested REPL" is then initiated in the context of the point of
execution where the evaluation was stopped.  The nested REPL's
continuation and evaluation environment are the same as the point where
the evaluation was stopped.  For example when evaluating the expression
`(let ((y (- 1 1))) (* (/ x y) 2))', a "divide by zero" error is
reported and the nested REPL's continuation is the one that takes the
result and multiplies it by two.  The REPL's lexical environment
includes the lexical variable `y'.  This allows the inspection of the
evaluation context (i.e. the lexical and dynamic environments and
continuation), which is particularly useful to determine the exact
location and cause of an error.

   The prompt of nested REPLs includes the nesting level; `1>' is the
prompt at the first nesting level, `2>' at the second nesting level,
and so on.  An end of file (usually <^D>) will cause the current REPL
to be terminated and the enclosing REPL (one nesting level less) to be
resumed.

   At any time the user can examine the frames in the REPL's
continuation, which is useful to determine which chain of procedure
calls lead to an error.  A backtrace that lists the chain of active
continuation frames in the REPL's continuation can be obtained with the
`,b' command.  The frames are numbered from 0, that is frame 0 is the
most recent frame of the continuation where execution stopped, frame 1
is the parent frame of frame 0, and so on.  It is also possible to move
the REPL to a specific parent continuation (i.e. a specific frame of
the continuation where execution stopped) with the `,N', `,N+', `,N-',
`,+', `,-', `,++', and `,--' commands.  When the frame number of the
frame being examined is not zero, it is shown in the prompt after the
nesting level, for example `1\5>' is the prompt when the REPL nesting
level is 1 and the frame number is 5.

   Expressions entered at a nested REPL are evaluated in the environment
(both lexical and dynamic) of the continuation frame currently being
examined if that frame was created by interpreted Scheme code.  If the
frame was created by compiled Scheme code then expressions get evaluated
in the global interaction environment.  This feature may be used in
interpreted code to fetch the value of a variable in the current frame
or to change its value with `set!'.  Note that some special forms
(`define' in particular) can only be evaluated in the global
interaction environment.

5.2 Debugging commands
======================

In addition to expressions, the REPL accepts the following special
"comma" commands:

`,?'
     Give a summary of the REPL commands.

`,(h SUBJECT)'
     This command will show the section of the Gambit manual with the
     definition of the procedure or special form SUBJECT, which must be
     a symbol.  For example `,(h time)' will show the section
     documenting the `time' special form.  Please see the `help'
     procedure for additional information.

`,h'
     This command will show the section of the Gambit manual with the
     definition of the procedure which raised the exception for which
     this REPL was started.

`,q'
     Terminate the process with exit status 0.  This is equivalent to
     calling `(exit 0)'.

`,qt'
     Terminate the current thread (note that terminating the primordial
     thread terminates the process).

`,t'
     Return to the outermost REPL, also known as the "top-level REPL".

`,d'
     Leave the current REPL and resume the enclosing REPL.  This
     command does nothing in the top-level REPL.

`,(c EXPR)'
     Leave the current REPL and continue the computation that initiated
     the REPL with a specific value.  This command can only be used to
     continue a computation that signaled an error.  The expression
     EXPR is evaluated in the current context and the resulting value
     is returned as the value of the expression which signaled the
     error.  For example, if the evaluation of the expression `(* (/ x
     y) 2)' signaled an error because `y' is zero, then in the nested
     REPL a `,(c (+ 4 y))' will resume the computation of `(* (/ x y)
     2)' as though the value of `(/ x y)' was 4.  This command must be
     used carefully because the context where the error occured may
     rely on the result being of a particular type.  For instance a
     `,(c #f)' in the previous example will cause `*' to signal a type
     error (this problem is the most troublesome when debugging Scheme
     code that was compiled with type checking turned off so be
     careful).

`,c'
     Leave the current REPL and continue the computation that initiated
     the REPL.  This command can only be used to continue a computation
     that was stopped due to a user interrupt, breakpoint or a
     single-step.

`,s'
     Leave the current REPL and continue the computation that initiated
     the REPL in single-stepping mode.  The computation will perform an
     evaluation step (as defined by `step-level-set!') and then stop,
     causing a nested REPL to be entered.  Just before the evaluation
     step is performed, a line is displayed (in the same format as
     `trace') which indicates the expression that is being evaluated.
     If the evaluation step produces a result, the result is also
     displayed on another line.  A nested REPL is then entered after
     displaying a message which describes the next step of the
     computation.  This command can only be used to continue a
     computation that was stopped due to a user interrupt, breakpoint
     or a single-step.

`,l'
     This command is similar to `,s' except that it "leaps" over
     procedure calls, that is procedure calls are treated like a single
     step.  Single-stepping mode will resume when the procedure call
     returns, or if and when the execution of the called procedure
     encounters a breakpoint.

`,N'
     Move to frame number N of the continuation.  After changing the
     current frame, a one-line summary of the frame is displayed as if
     the `,y' command was entered.

`,N+'
     Move forward by N frames in the chain of continuation frames (i.e.
     towards older continuation frames).  After changing the current
     frame, a one-line summary of the frame is displayed as if the `,y'
     command was entered.

`,N-'
     Move backward by N frames in the chain of continuation frames
     (i.e.  towards more recent continuation frames).  After changing
     the current frame, a one-line summary of the frame is displayed as
     if the `,y' command was entered.

`,+'
     Equivalent to `,1+'.

`,-'
     Equivalent to `,1-'.

`,++'
     Equivalent to `,N+' where N is the number of continuation frames
     displayed at the head of a backtrace.

`,--'
     Equivalent to `,N-' where N is the number of continuation frames
     displayed at the head of a backtrace.

`,y'
     Display a one-line summary of the current frame.  The information
     is displayed in four fields.  The first field is the frame number.
     The second field is the procedure that created the frame or
     `(interaction)' if the frame was created by an expression entered
     at the REPL.  The remaining fields describe the subproblem
     associated with the frame, that is the expression whose value is
     being computed.  The third field is the location of the
     subproblem's source code and the fourth field is a reproduction of
     the source code, possibly truncated to fit on the line.  The last
     two fields may be missing if that information is not available.
     In particular, the third field is missing when the frame was
     created by a user call to the `eval' procedure or by a compiled
     procedure not compiled with the declaration `debug-location', and
     the last field is missing when the frame was created by a compiled
     procedure not compiled with the declaration `debug-source'.

`,b'
     Display a backtrace summarizing each frame in the chain of
     continuation frames starting with the current frame.  For each
     frame, the same information as for the `,y' command is displayed
     (except that location information is displayed in the format
     `STREAM@LINE:COLUMN').  If there are more than 15 frames in the
     chain of continuation frames, some of the middle frames will be
     omitted.

`,be'
     Like the `,b' command but also display the environment.

`,bed'
     Like the `,be' command but also display the dynamic environment.

`,(b EXPR)'
     Display the backtrace of EXPR's value, X, which is obtained by
     evaluating EXPR in the current frame.  X must be a continuation or
     a thread.  When X is a continuation, the frames in that
     continuation are displayed.  When X is a thread, the backtrace of
     the current continuation of that thread is displayed.

`,(be EXPR)'
     Like the `,(b EXPR)' command but also display the environment.

`,(bed EXPR)'
     Like the `,(be EXPR)' command but also display the dynamic
     environment.

`,i'
     Pretty print the procedure that created the current frame or
     `(interaction)' if the frame was created by an expression entered
     at the REPL.  Compiled procedures will only be pretty printed when
     they are compiled with the declaration `debug-source'.

`,e'
     Display the environment which is accessible from the current frame.
     The lexical environment is displayed, followed by the dynamic
     environment if the parameter object
     `repl-display-dynamic-environment?' is not false.  Global lexical
     variables are not displayed.  Moreover the frame must have been
     created by interpreted code or code compiled with the declaration
     `debug-environments'.  Due to space safety considerations and
     compiler optimizations, some of the lexical variable bindings may
     be missing.  Lexical variable bindings are displayed using the
     format `VARIABLE = EXPRESSION' (when VARIABLE is mutable) or
     `VARIABLE == EXPRESSION' (when VARIABLE is immutable, which may
     happen in compiled code due to compiler optimization) and
     dynamically-bound parameter bindings are displayed using the
     format `(PARAMETER) = EXPRESSION'.  Note that EXPRESSION can be a
     self-evaluating expression (number, string, boolean, character,
     ...), a quoted expression, a lambda expression or a global
     variable (the last two cases, which are only used when the value
     of the variable or parameter is a procedure, simplifies the
     debugging of higher-order procedures).  A PARAMETER can be a
     quoted expression or a global variable.  Lexical bindings are
     displayed in inverse binding order (most deeply nested first) and
     shadowed variables are included in the list.

`,ed'
     Like the `,e' command but the dynamic environment is always
     displayed.

`,(e EXPR)'
     Display the environment of EXPR's value, X, which is obtained by
     evaluating EXPR in the current frame.  X must be a continuation, a
     thread, a procedure, or a nonnegative integer.  When X is a
     continuation, the environment at that point in the code is
     displayed.  When X is a thread, the environment of the current
     continuation of that thread is displayed. When X is a procedure,
     the lexical environment where X was created is combined with the
     current continuation and this combined environment is displayed.
     When X is an integer, the environment at frame number X of the
     continuation is displayed.

`,(ed EXPR)'
     Like the `,(e EXPR)' command but the dynamic environment is always
     displayed.

`,st'
     Display the state of the threads in the current thread's thread
     group.  A thread can be: uninitialized, initialized, active, and
     terminated (normally or abnormally).  Active threads can be
     running, sleeping and waiting on a synchronization object (mutex,
     condition variable or port) possibly with a timeout.

`,(st EXPR)'
     Display the state of a specific thread or thread group.  The value
     of EXPR must be a thread or thread group.

`,(v EXPR)'
     Start a new REPL visiting EXPR's value, X, which is obtained by
     evaluating EXPR in the current frame.  X must be a continuation, a
     thread, a procedure, or a nonnegative integer.  When X is a
     continuation, the new REPL's continuation is X and evaluations are
     done in the environment at that point in the code.  When X is a
     thread, the thread is interrupted and the new REPL's continuation
     is the point where the thread was interrupted.  When X is a
     procedure, the lexical environment where X was created is combined
     with the current continuation and evaluations are done in this
     combined environment.  When X is an integer, the REPL is started
     in frame number X of the continuation.


5.3 Debugging example
=====================

Here is a sample interaction with `gsi':

     $ gsi
     Gambit v4.8.5

     > (define (invsqr x) (/ 1 (expt x 2)))
     > (define (mymap fn lst)
         (define (mm in)
           (if (null? in)
               '()
               (cons (fn (car in)) (mm (cdr in)))))
         (mm lst))
     > (mymap invsqr '(5 2 hello 9 1))
     *** ERROR IN invsqr, (console)@1.25 -- (Argument 1) NUMBER expected
     (expt 'hello 2)
     1> ,i
     #<procedure #2 invsqr> =
     (lambda (x) (/ 1 (expt x 2)))
     1> ,e
     x = 'hello
     1> ,b
     0  invsqr                    (console)@1:25          (expt x 2)
     1  #<procedure #4>           (console)@6:17          (fn (car in))
     2  #<procedure #4>           (console)@6:31          (mm (cdr in))
     3  #<procedure #4>           (console)@6:31          (mm (cdr in))
     4  (interaction)             (console)@8:1           (mymap invsqr '(5 2 hel...
     1> ,+
     1  #<procedure #4>           (console)@6.17          (fn (car in))
     1\1> (pp #4)
     (lambda (in) (if (null? in) '() (cons (fn (car in)) (mm (cdr in)))))
     1\1> ,e
     in = '(hello 9 1)
     mm = (lambda (in) (if (null? in) '() (cons (fn (car in)) (mm (cdr in)))))
     fn = invsqr
     lst = '(5 2 hello 9 1)
     1\1> ,(e mm)
     mm = (lambda (in) (if (null? in) '() (cons (fn (car in)) (mm (cdr in)))))
     fn = invsqr
     lst = '(5 2 hello 9 1)
     1\1> fn
     #<procedure #2 invsqr>
     1\1> (pp fn)
     (lambda (x) (/ 1 (expt x 2)))
     1\1> ,+
     2  #<procedure #4>           (console)@6.31          (mm (cdr in))
     1\2> ,e
     in = '(2 hello 9 1)
     mm = (lambda (in) (if (null? in) '() (cons (fn (car in)) (mm (cdr in)))))
     fn = invsqr
     lst = '(5 2 hello 9 1)
     1\2> ,(c (list 3 4 5))
     (1/25 1/4 3 4 5)
     > ,q

5.4 Procedures related to debugging
===================================

 -- procedure: help SUBJECT
 -- procedure: help-browser [NEW-VALUE]
     The `help' procedure displays the section of the Gambit manual
     with the definition of the procedure or special form SUBJECT,
     which must be a procedure or symbol.  For example the call `(help
     gensym)' will show the section documenting the `gensym' procedure
     and the call `(help 'time)' will show the section documenting the
     `time' special form.  The `help' procedure returns the void object.

     The parameter object `help-browser' is bound to a string naming
     the external program that is used by the `help' procedure to view
     the documentation.  Initially it is bound to the empty string.  In
     normal circumstances when `help-browser' is bound to an empty
     string the `help' procedure runs the script `~~bin/gambdoc.bat'
     which searches for a suitable web browser to open the
     documentation in HTML format.  Unless the system was built with
     the command `configure --enable-help-browser=...', the text-only
     browser `lynx' (see `http://lynx.isc.org/') will be used by
     default if it is available.  We highly recommend that you install
     this browser if you are interested in viewing the documentation
     within the console in which the REPL is running.  You can exit
     `lynx' conveniently by typing an end of file (usually <^D>).

     For example:

          > (help-browser "firefox") ; use firefox instead of lynx
          > (help 'gensym)
          > (help gensym) ; OK because gensym is a procedure
          > (help 'time)
          > (help time) ; not OK because time is a special form
          *** ERROR IN (console)@5.7 -- Macro name can't be used as a variable: time
          >


 -- procedure: repl-result-history-ref I
 -- procedure: repl-result-history-max-length-set! N
     The REPL keeps a history of the last few results printed by the
     REPL. The call `(repl-result-history-ref I)' returns the Ith
     previous result (the last for I=0, the next to last for I=1, etc).
     By default the REPL result history remembers up to 3 results.
     The maximal length of the history can be set to N between 0 and 10
     by a call to `(repl-result-history-max-length-set! N)'.

     For convenience the reader defines an abbreviation for calling
     `repl-result-history-ref'.  Tokens formed by a sequence of one or
     more hash signs, such as ``#'', ``##'', etc, are expanded by the
     reader into the list `(repl-result-history-ref I)', where I is the
     number of hash signs minus 1.  In other words, ``#'' will return
     the last result printed by the REPL, ``##'' will return the next
     to last, etc.

     For example:

          > (map (lambda (x) (* x x)) '(1 2 3))
          (1 4 9)
          > (reverse #)
          (9 4 1)
          > (append # ##)
          (9 4 1 1 4 9)
          > 1
          1
          > 1
          1
          > (+ # ##)
          2
          > (+ # ##)
          3
          > (+ # ##)
          5
          > ####
          *** ERROR IN (console)@9.1 -- (Argument 1) Out of range
          (repl-result-history-ref 3)
          1>


 -- procedure: trace PROC...
 -- procedure: untrace PROC...
     The `trace' procedure starts tracing calls to the specified
     procedures.  When a traced procedure is called, a line containing
     the procedure and its arguments is displayed (using the procedure
     call expression syntax).  The line is indented with a sequence of
     vertical bars which indicate the nesting depth of the procedure's
     continuation.  After the vertical bars is a greater-than sign
     which indicates that the evaluation of the call is starting.

     When a traced procedure returns a result, it is displayed with the
     same indentation as the call but without the greater-than sign.
     This makes it easy to match calls and results (the result of a
     given call is the value at the same indentation as the
     greater-than sign).  If a traced procedure P1 performs a tail call
     to a traced procedure P2, then P2 will use the same indentation as
     P1.  This makes it easy to spot tail calls.  The special handling
     for tail calls is needed to preserve the space complexity of the
     program (i.e. tail calls are implemented as required by Scheme
     even when they involve traced procedures).

     The `untrace' procedure stops tracing calls to the specified
     procedures.  When no argument is passed to the `trace' procedure,
     the list of procedures currently being traced is returned.  The
     void object is returned by the `trace' procedure when it is passed
     one or more arguments.  When no argument is passed to the
     `untrace' procedure stops all tracing and returns the void object.
     A compiled procedure may be traced but only if it is bound to a
     global variable.

     For example:

          > (define (fact n) (if (< n 2) 1 (* n (fact (- n 1)))))
          > (trace fact)
          > (fact 5)
          | > (fact 5)
          | | > (fact 4)
          | | | > (fact 3)
          | | | | > (fact 2)
          | | | | | > (fact 1)
          | | | | | 1
          | | | | 2
          | | | 6
          | | 24
          | 120
          120
          > (trace -)
          *** WARNING -- Rebinding global variable "-" to an interpreted procedure
          > (define (fact-iter n r) (if (< n 2) r (fact-iter (- n 1) (* n r))))
          > (trace fact-iter)
          > (fact-iter 5 1)
          | > (fact-iter 5 1)
          | | > (- 5 1)
          | | 4
          | > (fact-iter 4 5)
          | | > (- 4 1)
          | | 3
          | > (fact-iter 3 20)
          | | > (- 3 1)
          | | 2
          | > (fact-iter 2 60)
          | | > (- 2 1)
          | | 1
          | > (fact-iter 1 120)
          | 120
          120
          > (trace)
          (#<procedure #2 fact-iter> #<procedure #3 -> #<procedure #4 fact>)
          > (untrace)
          > (fact 5)
          120


 -- procedure: step
 -- procedure: step-level-set! LEVEL
     The `step' procedure enables single-stepping mode.  After the call
     to `step' the computation will stop just before the interpreter
     executes the next evaluation step (as defined by
     `step-level-set!').  A nested REPL is then started.  Note that
     because single-stepping is stopped by the REPL whenever the prompt
     is displayed it is pointless to enter `(step)' by itself.  On the
     other hand entering `(begin (step) EXPR)' will evaluate EXPR in
     single-stepping mode.

     The procedure `step-level-set!' sets the stepping level which
     determines the granularity of the evaluation steps when
     single-stepping is enabled.  The stepping level LEVEL must be an
     exact integer in the range 0 to 7.  At a level of 0, the
     interpreter ignores single-stepping mode.  At higher levels the
     interpreter stops the computation just before it performs the
     following operations, depending on the stepping level:

       1. procedure call

       2. `delay' special form and operations at lower levels

       3. `lambda' special form and operations at lower levels

       4. `define' special form and operations at lower levels

       5. `set!' special form and operations at lower levels

       6. variable reference and operations at lower levels

       7. constant reference and operations at lower levels


     The default stepping level is 7.

     For example:

          > (define (fact n) (if (< n 2) 1 (* n (fact (- n 1)))))
          > (step-level-set! 1)
          > (begin (step) (fact 5))
          *** STOPPED IN (console)@3.15
          1> ,s
          | > (fact 5)
          *** STOPPED IN fact, (console)@1.22
          1> ,s
          | | > (< n 2)
          | | #f
          *** STOPPED IN fact, (console)@1.43
          1> ,s
          | | > (- n 1)
          | | 4
          *** STOPPED IN fact, (console)@1.37
          1> ,s
          | | > (fact (- n 1))
          *** STOPPED IN fact, (console)@1.22
          1> ,s
          | | | > (< n 2)
          | | | #f
          *** STOPPED IN fact, (console)@1.43
          1> ,s
          | | | > (- n 1)
          | | | 3
          *** STOPPED IN fact, (console)@1.37
          1> ,l
          | | | > (fact (- n 1))
          *** STOPPED IN fact, (console)@1.22
          1> ,l
          | | > (* n (fact (- n 1)))
          | | 24
          *** STOPPED IN fact, (console)@1.32
          1> ,l
          | > (* n (fact (- n 1)))
          | 120
          120


 -- procedure: break PROC...
 -- procedure: unbreak PROC...
     The `break' procedure places a breakpoint on each of the specified
     procedures.  When a procedure is called that has a breakpoint, the
     interpreter will enable single-stepping mode (as if `step' had
     been called).  This typically causes the computation to stop soon
     inside the procedure if the stepping level is high enough.

     The `unbreak' procedure removes the breakpoints on the specified
     procedures.  With no argument, `break' returns the list of
     procedures currently containing breakpoints.  The void object is
     returned by `break' if it is passed one or more arguments.  With
     no argument `unbreak' removes all the breakpoints and returns the
     void object.  A breakpoint can be placed on a compiled procedure
     but only if it is bound to a global variable.

     For example:

          > (define (double x) (+ x x))
          > (define (triple y) (- (double (double y)) y))
          > (define (f z) (* (triple z) 10))
          > (break double)
          > (break -)
          *** WARNING -- Rebinding global variable "-" to an interpreted procedure
          > (f 5)
          *** STOPPED IN double, (console)@1.21
          1> ,b
          0  double                    (console)@1:21          +
          1  triple                    (console)@2:31          (double y)
          2  f                         (console)@3:18          (triple z)
          3  (interaction)             (console)@6:1           (f 5)
          1> ,e
          x = 5
          1> ,c
          *** STOPPED IN double, (console)@1.21
          1> ,c
          *** STOPPED IN f, (console)@3.29
          1> ,c
          150
          > (break)
          (#<procedure #3 -> #<procedure #4 double>)
          > (unbreak)
          > (f 5)
          150


 -- procedure: generate-proper-tail-calls [NEW-VALUE]
     [Note: this procedure is DEPRECATED and will be removed in a
     future version of Gambit.  Use the `proper-tail-calls' declaration
     instead.]

     The parameter object `generate-proper-tail-calls' is bound to a
     boolean value controlling how the interpreter handles tail calls.
     When it is bound to `#f' the interpreter will treat tail calls
     like nontail calls, that is a new continuation will be created for
     the call.  This setting is useful for debugging, because when a
     primitive signals an error the location information will point to
     the call site of the primitive even if this primitive was called
     with a tail call.  The initial value of this parameter object is
     `#t', which means that a tail call will reuse the continuation of
     the calling function.

     This parameter object only affects code that is subsequently
     processed by `load' or `eval', or entered at the REPL.

     For example:

          > (generate-proper-tail-calls)
          #t
          > (let loop ((i 1)) (if (< i 10) (loop (* i 2)) oops))
          *** ERROR IN #<procedure #2>, (console)@2.47 -- Unbound variable: oops
          1> ,b
          0  #<procedure #2>           (console)@2:47          oops
          1  (interaction)             (console)@2:1           ((letrec ((loop (lambda...
          1> ,t
          > (generate-proper-tail-calls #f)
          > (let loop ((i 1)) (if (< i 10) (loop (* i 2)) oops))
          *** ERROR IN #<procedure #3>, (console)@6.47 -- Unbound variable: oops
          1> ,b
          0  #<procedure #3>           (console)@6:47          oops
          1  #<procedure #3>           (console)@6:32          (loop (* i 2))
          2  #<procedure #3>           (console)@6:32          (loop (* i 2))
          3  #<procedure #3>           (console)@6:32          (loop (* i 2))
          4  #<procedure #3>           (console)@6:32          (loop (* i 2))
          5  (interaction)             (console)@6:1           ((letrec ((loop (lambda...


 -- procedure: display-environment-set! DISPLAY?
     [Note: this procedure is DEPRECATED and will be removed in a
     future version of Gambit.  Use the parameter object
     `repl-display-environment?' instead.]

     This procedure sets a flag that controls the automatic display of
     the environment by the REPL.  If DISPLAY? is true, the environment
     is displayed by the REPL before the prompt.  The default setting is
     not to display the environment.


 -- procedure: repl-display-environment? DISPLAY?
     The parameter object `repl-display-environment?' is bound to a
     boolean value that controls the automatic display of the
     environment by the REPL.  If DISPLAY? is true, the environment is
     displayed by the REPL before the prompt.  This is particularly
     useful in single-stepping mode.  The default setting is not to
     display the environment.


 -- procedure: display-dynamic-environment? DISPLAY?
     The parameter object `display-dynamic-environment?' is bound to a
     boolean value that controls wether the dynamic environment is
     displayed when the environment is displayed.  The default setting
     is not to display the dynamic environment.


 -- procedure: pretty-print OBJ [PORT]
     This procedure pretty-prints OBJ on the port PORT.  If it is not
     specified, PORT defaults to the current output-port.

     For example:

          > (pretty-print
              (let* ((x '(1 2 3 4)) (y (list x x x))) (list y y y)))
          (((1 2 3 4) (1 2 3 4) (1 2 3 4))
           ((1 2 3 4) (1 2 3 4) (1 2 3 4))
           ((1 2 3 4) (1 2 3 4) (1 2 3 4)))


 -- procedure: pp OBJ [PORT]
     This procedure pretty-prints OBJ on the port PORT.  When OBJ is a
     procedure created by the interpreter or a procedure created by
     code compiled with the declaration `debug-source', the procedure's
     source code is displayed.  If it is not specified, PORT defaults
     to the interaction channel (i.e. the output will appear at the
     REPL).

     For example:

          > (define (f g) (+ (time (g 100)) (time (g 1000))))
          > (pp f)
          (lambda (g)
            (+ (##time (lambda () (g 100)) '(g 100))
               (##time (lambda () (g 1000)) '(g 1000))))


 -- procedure: gc-report-set! REPORT?
     This procedure controls the generation of reports during garbage
     collections.  If the argument is true, a brief report of memory
     usage is generated after every garbage collection.  It contains:
     the time taken for this garbage collection, the amount of memory
     allocated in megabytes since the program was started, the size of
     the heap in megabytes, the heap memory in megabytes occupied by
     live data, the proportion of the heap occupied by live data, and
     the number of bytes occupied by movable and nonmovable objects.


5.5 Console line-editing
========================

The console implements a simple Scheme-friendly line-editing
user-interface that is enabled by default.  It offers parentheses
balancing, a history of previous commands, symbol completion, and
several emacs-compatible keyboard commands.  The user's input is
displayed in a bold font and the output produced by the system is in a
plain font.  The history of previous commands is saved in the file
`~/.gambit_history'.  It is restored when a REPL is started.

   Symbol completion is triggered with the tab key.  When the cursor is
after a sequence of characters that can form a symbol, typing the tab
key will search the symbol table for the first symbol (in alphabetical
order) that begins with that sequence and insert that symbol.  Typing
the tab key in succession will cycle through all symbols with that
prefix.  When all possible symbols have been shown or there are no
possible completions, the text reverts to the uncompleted symbol and
the bell is rung.

   Here are the keyboard commands available (where the ``M-'' prefix
means the escape key is typed and the ``C-'' prefix means the control
key is pressed):

`C-d'
     Generate an end-of-file when the line is empty, otherwise delete
     character at cursor.

`delete or backspace'
     Delete character before cursor.

`M-C-d'
     Delete word forward and keep a copy of this text on the clipboard.

`M-delete'
     Delete word backward and keep a copy of this text on the clipboard.

`M-backspace'
     Delete S-expression backward and keep a copy of this text on the
     clipboard.

`C-a'
     Move cursor to beginning of line.

`C-e'
     Move cursor to end of line.

`C-b or left-arrow'
     Move cursor left one character.

`M-b'
     Move cursor left one word.

`M-C-b or `M-'left-arrow'
     Move cursor left one S-expression.

`C-f or right-arrow'
     Move cursor right one character.

`M-f'
     Move cursor right one word.

`M-C-f or `M-'right-arrow'
     Move cursor right one S-expression.

`C-p or `M-p' or up-arrow'
     Move to previous line in history.

`C-n or `M-n' or down-arrow'
     Move to next line in history.

`C-t'
     Transpose character at cursor with previous character.

`M-t'
     Transpose word after cursor with previous word.

`M-C-t'
     Transpose S-expression after cursor with previous S-expression.

`C-l'
     Clear console and redraw line being edited.

`C-nul'
     Set the mark to the cursor.

`C-w'
     Delete the text between the cursor and the mark and keep a copy of
     this text on the clipboard.

`C-k'
     Delete the text from the cursor to the end of the line and keep a
     copy of this text on the clipboard.

`C-y'
     Paste the text that is on the clipboard.

`F8'
     Same as typing `#||#,c;' (REPL command to continue the
     computation).

`F9'
     Same as typing `#||#,-;' (REPL command to move to newer frame).

`F10'
     Same as typing `#||#,+;' (REPL command to move to older frame).

`F11'
     Same as typing `#||#,s;' (REPL command to step the computation).

`F12'
     Same as typing `#||#,l;' (REPL command to leap the computation).


   On Mac OS X, depending on your configuration, you may have to press
the `fn' key to access the function key `F12' and the `option' key to
access the other function keys.

   On Microsoft Windows the clipboard is the system clipboard.  This
allows text to be copied and pasted between the program and other
applications.  On other operating systems the clipboard is internal to
the program (it is not integrated with the operating system).

5.6 Emacs interface
===================

Gambit comes with the Emacs package `gambit.el' which provides a nice
environment for running Gambit from within the Emacs editor.  This
package filters the standard output of the Gambit process and when it
intercepts a location information (in the format `STREAM@LINE.COLUMN'
where STREAM is either `(stdin)' when the expression was obtained from
standard input, `(console)' when the expression was obtained from the
console, or a string naming a file) it opens a window to highlight the
corresponding expression.

   To use this package, make sure the file `gambit.el' is accessible
from your load-path and that the following lines are in your `.emacs'
file:

     (autoload 'gambit-inferior-mode "gambit" "Hook Gambit mode into cmuscheme.")
     (autoload 'gambit-mode "gambit" "Hook Gambit mode into scheme.")
     (add-hook 'inferior-scheme-mode-hook (function gambit-inferior-mode))
     (add-hook 'scheme-mode-hook (function gambit-mode))
     (setq scheme-program-name "gsi -:d-")

   Alternatively, if you don't mind always loading this package, you
can simply add this line to your `.emacs' file:

     (require 'gambit)

   You can then start an inferior Gambit process by typing `M-x
run-scheme'.  The commands provided in `cmuscheme' mode will be
available in the Gambit interaction buffer (i.e. `*scheme*') and in
buffers attached to Scheme source files.  Here is a list of the most
useful commands (for a complete list type `C-h m' in the Gambit
interaction buffer):
`C-x C-e'
     Evaluate the expression which is before the cursor (the expression
     will be copied to the Gambit interaction buffer).

`C-c C-z'
     Switch to Gambit interaction buffer.

`C-c C-l'
     Load a file (file attached to current buffer is default) using
     `(load FILE)'.

`C-c C-k'
     Compile a file (file attached to current buffer is default) using
     `(compile-file FILE)'.

   The file `gambit.el' provides these additional commands:

`F8 or C-c c'
     Continue the computation (same as typing `#||#,c;' to the REPL).

`F9 or C-c ]'
     Move to newer frame (same as typing `#||#,-;' to the REPL).

`F10 or C-c ['
     Move to older frame (same as typing `#||#,+;' to the REPL).

`F11 or C-c s'
     Step the computation (same as typing `#||#,s;' to the REPL).

`F12 or C-c l'
     Leap the computation (same as typing `#||#,l;' to the REPL).

`C-c _'
     Removes the last window that was opened to highlight an expression.

   The two keystroke version of these commands can be shortened to
`M-c', `M-[', `M-]', `M-s', `M-l', and `M-_' respectively by adding
this line to your `.emacs' file:

     (setq gambit-repl-command-prefix "\e")

   This is more convenient to type than the two keystroke `C-c' based
sequences but the purist may not like this because it does not follow
normal Emacs conventions.

   Here is what a typical `.emacs' file will look like:

     (setq load-path ; add directory containing gambit.el
       (cons "/usr/local/Gambit/share/emacs/site-lisp"
             load-path))
     (setq scheme-program-name "/tmp/gsi -:d-") ; if gsi not in executable path
     (setq gambit-highlight-color "gray") ; if you don't like the default
     (setq gambit-repl-command-prefix "\e") ; if you want M-c, M-s, etc
     (require 'gambit)

5.7 GUIDE
=========

The implementation and documentation for GUIDE, the Gambit Universal
IDE, are not yet complete.

6 Scheme extensions
*******************

6.1 Extensions to standard procedures
=====================================

 -- procedure: transcript-on FILE
 -- procedure: transcript-off
     These procedures do nothing.


 -- procedure: call-with-current-continuation PROC
 -- procedure: call/cc PROC
     The procedure `call-with-current-continuation' is bound to the
     global variables `call-with-current-continuation' and `call/cc'.


6.2 Extensions to standard special forms
========================================

 -- special form: lambda lambda-formals body
 -- special form: define (variable define-formals) body
          lambda-formals = `(' formal-argument-list `)' |
          r4rs-lambda-formals

          define-formals = formal-argument-list | r4rs-define-formals

          formal-argument-list = dsssl-formal-argument-list |
          rest-at-end-formal-argument-list

          dsssl-formal-argument-list = reqs opts rest keys

          rest-at-end-formal-argument-list = reqs opts keys rest | reqs
          opts keys `.' rest-formal-argument

          reqs = required-formal-argument*

          required-formal-argument = variable

          opts = `#!optional' optional-formal-argument* | empty

          optional-formal-argument = variable | `(' variable
          initializer `)'

          rest = `#!rest' rest-formal-argument | empty

          rest-formal-argument = variable

          keys = `#!key' keyword-formal-argument* | empty

          keyword-formal-argument = variable | `(' variable initializer
          `)'

          initializer = expression

          r4rs-lambda-formals = `(' variable* `)' | `(' variable+ `.'
          variable `)' | variable

          r4rs-define-formals = variable* | variable* `.' variable

     These forms are extended versions of the `lambda' and `define'
     special forms of standard Scheme.  They allow the use of optional
     formal arguments, either positional or named, and support the
     syntax and semantics of the DSSSL standard.

     When the procedure introduced by a `lambda' (or `define') is
     applied to a list of actual arguments, the formal and actual
     arguments are processed as specified in the R4RS if the
     lambda-formals (or define-formals) is a r4rs-lambda-formals (or
     r4rs-define-formals).

     If the formal-argument-list matches dsssl-formal-argument-list or
     extended-formal-argument-list they are processed as follows:

       a. Variables in required-formal-arguments are bound to
          successive actual arguments starting with the first actual
          argument.  It shall be an error if there are fewer actual
          arguments than required-formal-arguments.

       b. Next variables in optional-formal-arguments are bound to
          remaining actual arguments.  If there are fewer remaining
          actual arguments than optional-formal-arguments, then the
          variables are bound to the result of evaluating initializer,
          if one was specified, and otherwise to `#f'.  The initializer
          is evaluated in an environment in which all previous formal
          arguments have been bound.

       c. If `#!key' does not appear in the formal-argument-list and
          there is no rest-formal-argument then it shall be an error if
          there are any remaining actual arguments.

       d. If `#!key' does not appear in the formal-argument-list and
          there is a rest-formal-argument then the rest-formal-argument
          is bound to a list of all remaining actual arguments.

       e. If `#!key' appears in the formal-argument-list and there is
          no rest-formal-argument then there shall be an even number of
          remaining actual arguments.  These are interpreted as a
          series of pairs, where the first member of each pair is a
          keyword specifying the argument name, and the second is the
          corresponding value.  It shall be an error if the first
          member of a pair is not a keyword.  It shall be an error if
          the argument name is not the same as a variable in a
          keyword-formal-argument.  If the same argument name occurs
          more than once in the list of actual arguments, then the
          first value is used.  If there is no actual argument for a
          particular keyword-formal-argument, then the variable is
          bound to the result of evaluating initializer if one was
          specified, and otherwise to `#f'.  The initializer is
          evaluated in an environment in which all previous formal
          arguments have been bound.

       f. If `#!key' appears in the formal-argument-list and there is a
          rest-formal-argument before the `#!key' then there may be an
          even or odd number of remaining actual arguments and the
          rest-formal-argument is bound to a list of all remaining
          actual arguments.  Then, these remaining actual arguments are
          scanned from left to right in pairs, stopping at the first
          pair whose first element is not a keyword.  Each pair whose
          first element is a keyword matching the name of a
          keyword-formal-argument gives the value (i.e. the second
          element of the pair) of the corresponding formal argument.  If
          the same argument name occurs more than once in the list of
          actual arguments, then the first value is used.  If there is
          no actual argument for a particular keyword-formal-argument,
          then the variable is bound to the result of evaluating
          initializer if one was specified, and otherwise to `#f'.  The
          initializer is evaluated in an environment in which all
          previous formal arguments have been bound.

       g. If `#!key' appears in the formal-argument-list and there is a
          rest-formal-argument after the `#!key' then there may be an
          even or odd number of remaining actual arguments.  The
          remaining actual arguments are scanned from left to right in
          pairs, stopping at the first pair whose first element is not
          a keyword.  Each pair shall have as its first element a
          keyword matching the name of a keyword-formal-argument; the
          second element gives the value of the corresponding formal
          argument.  If the same argument name occurs more than once in
          the list of actual arguments, then the first value is used.
          If there is no actual argument for a particular
          keyword-formal-argument, then the variable is bound to the
          result of evaluating initializer if one was specified, and
          otherwise to `#f'.  The initializer is evaluated in an
          environment in which all previous formal arguments have been
          bound.  Finally, the rest-formal-argument is bound to the
          list of the actual arguments that were not scanned (i.e.
          after the last keyword/value pair).

     In all cases it is an error for a variable to appear more than
     once in a formal-argument-list.

     Note that this specification is compatible with the DSSSL language
     standard (i.e. a correct DSSSL program will have the same semantics
     when run with Gambit).

     It is unspecified whether variables receive their value by binding
     or by assignment.  Currently the compiler and interpreter use
     different methods, which can lead to different semantics if
     `call-with-current-continuation' is used in an initializer.  Note
     that this is irrelevant for DSSSL programs because
     `call-with-current-continuation' does not exist in DSSSL.

     For example:

          > ((lambda (#!rest x) x) 1 2 3)
          (1 2 3)
          > (define (f a #!optional b) (list a b))
          > (define (g a #!optional (b a) #!key (k (* a b))) (list a b k))
          > (define (h1 a #!rest r #!key k) (list a k r))
          > (define (h2 a #!key k #!rest r) (list a k r))
          > (f 1)
          (1 #f)
          > (f 1 2)
          (1 2)
          > (g 3)
          (3 3 9)
          > (g 3 4)
          (3 4 12)
          > (g 3 4 k: 5)
          (3 4 5)
          > (g 3 4 k: 5 k: 6)
          (3 4 5)
          > (h1 7)
          (7 #f ())
          > (h1 7 k: 8 9)
          (7 8 (k: 8 9))
          > (h1 7 k: 8 z: 9)
          (7 8 (k: 8 z: 9))
          > (h2 7)
          (7 #f ())
          > (h2 7 k: 8 9)
          (7 8 (9))
          > (h2 7 k: 8 z: 9)
          *** ERROR IN (console)@17.1 -- Unknown keyword argument passed to procedure
          (h2 7 k: 8 z: 9)


6.3 Miscellaneous extensions
============================

 -- procedure: vector-copy VECTOR
     This procedure returns a newly allocated vector with the same
     content as the vector VECTOR.  Note that the elements are not
     recursively copied.

     For example:

          > (define v1 '#(1 2 3))
          > (define v2 (vector-copy v1))
          > v2
          #(1 2 3)
          > (eq? v1 v2)
          #f


 -- procedure: subvector VECTOR START END
     This procedure is the vector analog of the `substring' procedure.
     It returns a newly allocated vector formed from the elements of
     the vector VECTOR beginning with index START (inclusive) and
     ending with index END (exclusive).

     For example:

          > (subvector '#(a b c d e f) 3 5)
          #(d e)


 -- procedure: vector-append VECTOR...
     This procedure is the vector analog of the `string-append'
     procedure.  It returns a newly allocated vector whose elements
     form the concatenation of the given vectors.

     For example:

          > (define v '#(1 2 3))
          > (vector-append v v v)
          #(1 2 3 1 2 3 1 2 3)


 -- procedure: append-vectors LST
     This procedure returns a newly allocated vector whose elements form
     the concatenation of all the vectors in the list LST. It is
     equivalent to `(apply vector-append LST)'.

     For example:

          > (define v '#(1 2 3))
          > (append-vectors (list v v v))
          #(1 2 3 1 2 3 1 2 3)


 -- procedure: subvector-fill! VECTOR START END FILL
     This procedure is like `vector-fill!', but fills a selected part
     of the given vector. It sets the elements of the vector VECTOR,
     beginning with index START (inclusive) and ending with index END
     (exclusive) to FILL.  The value returned is unspecified.

     For example:

          > (define v (vector 'a 'b 'c 'd 'e 'f))
          > (subvector-fill! v 3 5 'x)
          > v
          #(a b c x x f)


 -- procedure: subvector-move! SRC-VECTOR SRC-START SRC-END DST-VECTOR
          DST-START
     This procedure replaces part of the contents of vector DST-VECTOR
     with part of the contents of vector SRC-VECTOR. It copies elements
     from SRC-VECTOR, beginning with index SRC-START (inclusive) and
     ending with index SRC-END (exclusive) to DST-VECTOR beginning with
     index DST-START (inclusive).  The value returned is unspecified.

     For example:

          > (define v1 '#(1 2 3 4 5 6))
          > (define v2 (vector 'a 'b 'c 'd 'e 'f))
          > (subvector-move! v1 3 5 v2 1)
          > v2
          #(a 4 5 d e f)


 -- procedure: vector-shrink! VECTOR K
     This procedure shortens the vector VECTOR so that its new size is
     K.  The value returned is unspecified.

     For example:

          > (define v (vector 'a 'b 'c 'd 'e 'f))
          > v
          #(a b c d e f)
          > (vector-shrink! v 3)
          > v
          #(a b c)


 -- procedure: append-strings LST
     This procedure returns a newly allocated string whose elements form
     the concatenation of all the strings in the list LST. It is
     equivalent to `(apply string-append LST)'.

     For example:

          > (define s "abc")
          > (append-strings (list s s s))
          "abcabcabc"


 -- procedure: substring-fill! STRING START END FILL
     This procedure is like `string-fill!', but fills a selected part
     of the given string. It sets the elements of the string STRING,
     beginning with index START (inclusive) and ending with index END
     (exclusive) to FILL.  The value returned is unspecified.

     For example:

          > (define s (string #\a #\b #\c #\d #\e #\f))
          > (substring-fill! s 3 5 #\x)
          > s
          "abcxxf"


 -- procedure: substring-move! SRC-STRING SRC-START SRC-END DST-STRING
          DST-START
     This procedure replaces part of the contents of string DST-STRING
     with part of the contents of string SRC-STRING. It copies elements
     from SRC-STRING, beginning with index SRC-START (inclusive) and
     ending with index SRC-END (exclusive) to DST-STRING beginning with
     index DST-START (inclusive).  The value returned is unspecified.

     For example:

          > (define s1 "123456")
          > (define s2 (string #\a #\b #\c #\d #\e #\f))
          > (substring-move! s1 3 5 s2 1)
          > s2
          "a45def"


 -- procedure: string-shrink! STRING K
     This procedure shortens the string STRING so that its new size is
     K.  The value returned is unspecified.

     For example:

          > (define s (string #\a #\b #\c #\d #\e #\f))
          > s
          "abcdef"
          > (string-shrink! s 3)
          > s
          "abc"


 -- procedure: box OBJ
 -- procedure: box? OBJ
 -- procedure: unbox BOX
 -- procedure: set-box! BOX OBJ
     These procedures implement the "box" data type.  A box is a cell
     containing a single mutable field.  The lexical syntax of a box
     containing the object OBJ is `#&OBJ' (*note Box syntax::).

     The procedure `box' returns a new box object whose content is
     initialized to OBJ.  The procedure `box?' returns `#t' if OBJ is a
     box, and otherwise returns `#f'.  The procedure `unbox' returns
     the content of the box BOX.  The procedure `set-box!' changes the
     content of the box BOX to OBJ.  The procedure `set-box!' returns
     an unspecified value.

     For example:

          > (define b (box 0))
          > b
          #&0
          > (define (inc!) (set-box! b (+ (unbox b) 1)))
          > (inc!)
          > b
          #&1
          > (unbox b)
          1


 -- procedure: keyword? OBJ
 -- procedure: keyword->string KEYWORD
 -- procedure: string->keyword STRING
     These procedures implement the "keyword" data type.  Keywords are
     similar to symbols but are self evaluating and distinct from the
     symbol data type.  The lexical syntax of keywords is specified in
     *Note Keyword syntax::.

     The procedure `keyword?' returns `#t' if OBJ is a keyword, and
     otherwise returns `#f'.  The procedure `keyword->string' returns
     the name of KEYWORD as a string.  The procedure `string->keyword'
     returns the keyword whose name is STRING.

     For example:

          > (keyword? 'color)
          #f
          > (keyword? color:)
          #t
          > (keyword->string color:)
          "color"
          > (string->keyword "color")
          color:


 -- procedure: gensym [PREFIX]
     This procedure returns a new "uninterned symbol".  Uninterned
     symbols are guaranteed to be distinct from the symbols generated
     by the procedures `read' and `string->symbol'.  The symbol PREFIX
     is the prefix used to generate the new symbol's name.  If it is
     not specified, the prefix defaults to `g'.

     For example:

          > (gensym)
          #:g0
          > (gensym)
          #:g1
          > (gensym 'star-trek-)
          #:star-trek-2


 -- procedure: string->uninterned-symbol NAME [HASH]
 -- procedure: uninterned-symbol? OBJ
     The procedure `string->uninterned-symbol' returns a new uninterned
     symbol whose name is NAME and hash is HASH.  The name must be a
     string and the hash must be a nonnegative fixnum.

     The procedure `uninterned-symbol?' returns `#t' when OBJ is a
     symbol that is uninterned and `#f' otherwise.

     For example:

          > (uninterned-symbol? (gensym))
          #t
          > (string->uninterned-symbol "foo")
          #:foo:
          > (uninterned-symbol? (string->uninterned-symbol "foo"))
          #t
          > (uninterned-symbol? 'hello)
          #f
          > (uninterned-symbol? 123)
          #f


 -- procedure: string->uninterned-keyword NAME [HASH]
 -- procedure: uninterned-keyword? OBJ
     The procedure `string->uninterned-keyword' returns a new uninterned
     keyword whose name is NAME and hash is HASH.  The name must be a
     string and the hash must be a nonnegative fixnum.

     The procedure `uninterned-keyword?' returns `#t' when OBJ is a
     keyword that is uninterned and `#f' otherwise.

     For example:

          > (string->uninterned-keyword "foo")
          #:foo:
          > (uninterned-keyword? (string->uninterned-keyword "foo"))
          #t
          > (uninterned-keyword? hello:)
          #f
          > (uninterned-keyword? 123)
          #f


 -- procedure: void
     This procedure returns the void object.  The read-eval-print loop
     prints nothing when the result is the void object.


 -- procedure: eval EXPR [ENV]
     The first parameter is a datum representing an expression.  The
     `eval' procedure evaluates this expression in the global
     interaction environment and returns the result.  If present, the
     second parameter is ignored (it is provided for compatibility with
     R5RS).

     For example:

          > (eval '(+ 1 2))
          3
          > ((eval 'car) '(1 2))
          1
          > (eval '(define x 5))
          > x
          5


 -- special form: include file
     The file parameter must be a string naming an existing file
     containing Scheme source code.  The `include' special form splices
     the content of the specified source file.  This form can only
     appear where a `define' form is acceptable.

     For example:

          (include "macros.scm")

          (define (f lst)
            (include "sort.scm")
            (map sqrt (sort lst)))


 -- special form: define-macro (name define-formals) body
     Define name as a macro special form which expands into body.  This
     form can only appear where a `define' form is acceptable.  Macros
     are lexically scoped.  The scope of a local macro definition
     extends from the definition to the end of the body of the
     surrounding binding construct.  Macros defined at the top level of
     a Scheme module are only visible in that module.  To have access
     to the macro definitions contained in a file, that file must be
     included using the `include' special form.  Macros which are
     visible from the REPL are also visible during the compilation of
     Scheme source files.

     For example:

          (define-macro (unless test . body)
            `(if ,test #f (begin ,@body)))

          (define-macro (push var #!optional val)
            `(set! ,var (cons ,val ,var)))

     To examine the code into which a macro expands you can use the
     compiler's `-expansion' option or the `pp' procedure.  For example:

          > (define-macro (push var #!optional val)
              `(set! ,var (cons ,val ,var)))
          > (pp (lambda () (push stack 1) (push stack) (push stack 3)))
          (lambda ()
            (set! stack (cons 1 stack))
            (set! stack (cons #f stack))
            (set! stack (cons 3 stack)))


 -- special form: define-syntax name expander
     Define name as a macro special form whose expansion is specified
     by expander.  This form is available only when the runtime option
     `-:s' is used.  This option causes the loading of the
     `~~lib/syntax-case' support library, which is the Hieb and Dybvig
     portable `syntax-case' implementation which has been ported to the
     Gambit interpreter and compiler.  Note that this implementation of
     `syntax-case' does not support special forms that are specific to
     Gambit.

     For example:

          $ gsi -:s
          Gambit v4.8.5

          > (define-syntax unless
              (syntax-rules ()
                ((unless test body ...)
                 (if test #f (begin body ...)))))
          > (let ((test 111)) (unless (= 1 2) (list test test)))
          (111 111)
          > (pp (lambda () (let ((test 111)) (unless (= 1 2) (list test test)))))
          (lambda () ((lambda (%%test14) (if (= 1 2) #f (list %%test14 %%test14))) 111))
          > (unless #f (pp xxx))
          *** ERROR IN (console)@7.16 -- Unbound variable: xxx


 -- special form: declare declaration...
     This form introduces declarations to be used by the compiler
     (currently the interpreter ignores the declarations).  This form
     can only appear where a `define' form is acceptable.  Declarations
     are lexically scoped in the same way as macros.  The following
     declarations are accepted by the compiler:

    `(DIALECT)'
          Use the given dialect's semantics.  DIALECT can be:
          `ieee-scheme', `r4rs-scheme', `r5rs-scheme' or
          `gambit-scheme'.

    `(STRATEGY)'
          Select block compilation or separate compilation.  In block
          compilation, the compiler assumes that global variables
          defined in the current file that are not mutated in the file
          will never be mutated.  STRATEGY can be: `block' or
          `separate'.

    `([not] inline)'
          Allow (or disallow) inlining of user procedures.

    `([not] inline-primitives PRIMITIVE...)'
          The given primitives should (or should not) be inlined if
          possible (all primitives if none specified).

    `(inlining-limit N)'
          Select the degree to which the compiler inlines user
          procedures.  N is the upper-bound, in percent, on code
          expansion that will result from inlining.  Thus, a value of
          300 indicates that the size of the program will not grow by
          more than 300 percent (i.e. it will be at most 4 times the
          size of the original).  A value of 0 disables inlining.  The
          size of a program is the total number of subexpressions it
          contains (i.e. the size of an expression is one plus the size
          of its immediate subexpressions).  The following conditions
          must hold for a procedure to be inlined: inlining the
          procedure must not cause the size of the call site to grow
          more than specified by the inlining limit, the site of
          definition (the `define' or `lambda') and the call site must
          be declared as `(inline)', and the compiler must be able to
          find the definition of the procedure referred to at the call
          site (if the procedure is bound to a global variable, the
          definition site must have a `(block)' declaration).  Note
          that inlining usually causes much less code expansion than
          specified by the inlining limit (an expansion around 10% is
          common for N=350).

    `([not] lambda-lift)'
          Lambda-lift (or don't lambda-lift) locally defined procedures.

    `([not] constant-fold)'
          Allow (or disallow) constant-folding of primitive procedures.

    `([not] standard-bindings VAR...)'
          The given global variables are known (or not known) to be
          equal to the value defined for them in the dialect (all
          variables defined in the standard if none specified).

    `([not] extended-bindings VAR...)'
          The given global variables are known (or not known) to be
          equal to the value defined for them in the runtime system
          (all variables defined in the runtime if none specified).

    `([not] run-time-bindings VAR...)'
          The given global variables will be tested at run time to see
          if they are equal to the value defined for them in the
          runtime system (all variables defined in the runtime if none
          specified).

    `([not] safe)'
          Generate (or don't generate) code that will prevent fatal
          errors at run time.  Note that in `safe' mode certain
          semantic errors will not be checked as long as they can't
          crash the system.  For example the primitive `char=?' may
          disregard the type of its arguments in `safe' as well as `not
          safe' mode.

    `([not] interrupts-enabled)'
          Generate (or don't generate) interrupt checks.  Interrupt
          checks are used to detect user interrupts and also to check
          for stack overflows.  Interrupt checking should not be turned
          off casually.

    `([not] debug)'
          Enable (or disable) the generation of debugging information.
          The kind of debugging information that is generated depends
          on the declarations `debug-location', `debug-source', and
          `debug-environments'.  If any of the command line options
          `-debug', `-debug-location', `-debug-source' and
          `-debug-environments' are present, the `debug' declaration is
          initially enabled, otherwise it is initially disabled.  When
          all kinds of debugging information are generated there is a
          substantial increase in the C compilation time and the size
          of the generated code.  When compiling a 3000 line Scheme
          file it was observed that the total compilation time is 500%
          longer and the executable code is 150% bigger.

    `([not] debug-location)'
          Select (or deselect) source code location debugging
          information.  When this declaration and the `debug'
          declaration are in effect, run time error messages indicate
          the location of the error in the source code file.  If any of
          the command line options `-debug-source' and
          `-debug-environments' are present and `-debug-location' is
          absent, the `debug-location' declaration is initially
          disabled, otherwise it is initially enabled.  When compiling
          a 3000 line Scheme file it was observed that the total
          compilation time is 200% longer and the executable code is
          60% bigger.

    `([not] debug-source)'
          Select (or deselect) source code debugging information.  When
          this declaration and the `debug' declaration are in effect,
          run time error messages indicate the source code, the
          backtraces are more precise, and the `pp' procedure will
          display the source code of compiled procedures.  If any of
          the command line options `-debug-location' and
          `-debug-environments' are present and `-debug-source' is
          absent, the `debug-source' declaration is initially disabled,
          otherwise it is initially enabled.  When compiling a 3000
          line Scheme file it was observed that the total compilation
          time is 90% longer and the executable code is 90% bigger.

    `([not] debug-environments)'
          Select (or deselect) environment debugging information.  When
          this declaration and the `debug' declaration are in effect,
          the debugger will have access to the environments of the
          continuations.  In other words the local variables defined in
          compiled procedures (and not optimized away by the compiler)
          will be shown by the `,e' REPL command.  If any of the
          command line options `-debug-location' and `-debug-source'
          are present and `-debug-environments' is absent, the
          `debug-environments' declaration is initially disabled,
          otherwise it is initially enabled.  When compiling a 3000
          line Scheme file it was observed that the total compilation
          time is 70% longer and the executable code is 40% bigger.

    `([not] proper-tail-calls)'
          Generate (or don't generate) proper tail calls.  When proper
          tail calls are turned off, tail calls are handled like
          non-tail calls, that is a continuation frame will be created
          for all calls regardless of their kind.  This is useful for
          debugging because the caller of a procedure will be visible
          in the backtrace produced by the REPL's `,b' command even
          when the call is a tail call.  Be advised that this does
          cause stack space to be consumed for tail calls which may
          cause the stack to overflow when performing long iterations
          with tail calls (whether they are expressed with a `letrec',
          named `let', `do', or other form).

    `([not] generative-lambda)'
          Force (or don't force) the creation of fresh closures when
          evaluating lambda-expressions.  A fresh closure is always
          created when a lambda-expression has at least one free
          variable (that has not been eliminated by dead-code
          elimination or other compiler optimization) or when the
          generative-lambda declaration is turned on.  When a
          lambda-expression has no free variables and the
          generative-lambda declaration is turned off, the value of the
          lambda-expression may be the same procedure (in the sense of
          `eq?').

    `([not] optimize-dead-local-variables)'
          Remove (or preserve) the dead local variables in the
          environment.  Preserving the dead local variables is useful
          for debugging because continuations will contain the dead
          variables.  Thus, if the code is also compiled with the
          declaration `debug-environments' the `,e', `,ed', `,be', and
          `,bed' REPL commands will display the dead variables.  On the
          other hand, preserving the dead local variables may change
          the space complexity of the program (i.e. some of the data
          that would normally be reclaimed by the garbage collector
          will not be).  Note that due to other compiler optimizations
          some dead local variables may be removed regardless of this
          declaration.

    `([not] optimize-dead-definitions VAR...)'
          Remove (or preserve) the dead toplevel definitions of the
          given global variables (all global variables if none
          specified).  A toplevel definition is dead if the evaluation
          of its expression can't possibly cause a side-effect and it
          is not referenced by toplevel expressions of the programs or
          toplevel definitions that aren't dead.  When a module is
          separately compiled and some of its definitions are only used
          by other modules, this declaration must be used with care to
          keep definitions that are used by other modules, for example
          if `foo' is referenced in another module the following
          declaration should be used: `(declare (not
          optimize-dead-definitions foo))'.

    `(NUMBER-TYPE PRIMITIVE...)'
          Numeric arguments and result of the specified primitives are
          known to be of the given type (all primitives if none
          specified).  NUMBER-TYPE can be: `generic', `fixnum', or
          `flonum'.

    `(MOSTLY-NUMBER-TYPE PRIMITIVE...)'
          Numeric arguments and result of the specified primitives are
          expected to be most often of the given type (all primitives
          if none specified).  MOSTLY-NUMBER-TYPE can be:
          `mostly-generic', `mostly-fixnum', `mostly-fixnum-flonum',
          `mostly-flonum', or `mostly-flonum-fixnum'.


     The default declarations used by the compiler are equivalent to:

          (declare
            (gambit-scheme)
            (separate)
            (inline)
            (inline-primitives)
            (inlining-limit 350)
            (constant-fold)
            (lambda-lift)
            (not standard-bindings)
            (not extended-bindings)
            (run-time-bindings)
            (safe)
            (interrupts-enabled)
            (not debug)           ;; depends on debugging command line options
            (debug-location)      ;; depends on debugging command line options
            (debug-source)        ;; depends on debugging command line options
            (debug-environments)  ;; depends on debugging command line options
            (proper-tail-calls)
            (not generative-lambda)
            (optimize-dead-local-variables)
            (not optimize-dead-definitions)
            (generic)
            (mostly-fixnum-flonum)
          )

     These declarations are compatible with the semantics of R5RS Scheme
     and includes a few procedures from R6RS (mainly fixnum specific and
     flonum specific procedures).  Typically used declarations that
     enhance performance, at the cost of violating the R5RS Scheme
     semantics, are: `(standard-bindings)', `(block)', `(not safe)' and
     `(fixnum)'.


6.4 Undocumented extensions
===========================

The procedures in this section are not yet documented.

 -- procedure: continuation? OBJ
 -- procedure: continuation-capture PROC
 -- procedure: continuation-graft CONT PROC OBJ...
 -- procedure: continuation-return CONT OBJ...
     These procedures provide access to internal first-class
     continuations which are represented using continuation objects
     distinct from procedures.

     The procedure `continuation?' returns `#t' when OBJ is a
     continuation object and `#f' otherwise.

     The procedure `continuation-capture' is similar to the `call/cc'
     procedure but it represents the continuation with a continuation
     object.  The PROC parameter must be a procedure accepting a single
     argument.  The procedure `continuation-capture' reifies its
     continuation and calls PROC with the corresponding continuation
     object as its sole argument.  Like for `call/cc', the implicit
     continuation of the call to PROC is the implicit continuation of
     the call to `continuation-capture'.

     The procedure `continuation-graft' performs a procedure call to
     the procedure PROC with arguments OBJ... and the implicit
     continuation corresponding to the continuation object CONT.  The
     current continuation of the call to procedure `continuation-graft'
     is ignored.

     The procedure `continuation-return' invokes the implicit
     continuation corresponding to the continuation object CONT with
     the result(s) OBJ....  This procedure can be easily defined in
     terms of `continuation-graft':

          (define (continuation-return cont . objs)
            (continuation-graft (lambda () (apply values objs))))

     For example:

          > (define x #f)
          > (define p (make-parameter 11))
          > (pp (parameterize ((p 22))
                  (cons 33 (continuation-capture
                            (lambda (c) (set! x c) 44)))))
          (33 . 44)
          > x
          #<continuation #2>
          > (continuation-return x 55)
          (33 . 55)
          > (continuation-graft x (lambda () (expt 2 10)))
          (33 . 1024)
          > (continuation-graft x expt 2 10)
          (33 . 1024)
          > (continuation-graft x (lambda () (p)))
          (33 . 22)
          > (define (map-sqrt1 lst)
              (call/cc
               (lambda (k)
                 (map (lambda (x)
                        (if (< x 0)
                            (k 'error)
                            (sqrt x)))
                      lst))))
          > (map-sqrt1 '(1 4 9))
          (1 2 3)
          > (map-sqrt1 '(1 -1 9))
          error
          > (define (map-sqrt2 lst)
              (continuation-capture
               (lambda (c)
                 (map (lambda (x)
                        (if (< x 0)
                            (continuation-return c 'error)
                            (sqrt x)))
                      lst))))
          > (map-sqrt2 '(1 4 9))
          (1 2 3)
          > (map-sqrt2 '(1 -1 9))
          error


 -- procedure: display-exception EXC [PORT]
 -- procedure: display-exception-in-context EXC CONT [PORT]
 -- procedure: display-procedure-environment PROC [PORT]
 -- procedure: display-continuation-environment CONT [PORT]
 -- procedure: display-continuation-dynamic-environment CONT [PORT]

 -- procedure: display-continuation-backtrace CONT [PORT [ALL-FRAMES?
          [DISPLAY-ENV? [MAX-HEAD [MAX-TAIL [DEPTH]]]]]]
     The procedure `display-continuation-backtrace' displays the frames
     of the continuation corresponding to the continuation object CONT
     on the port PORT.  If it is not specified, PORT defaults to the
     current output-port.  The frames are displayed in the same format
     as the REPL's `,b' command.

     The parameter ALL-FRAMES?, which defaults to `#f', controls which
     frames are displayed.  Some frames of ancillary importance, such
     as internal frames created by the interpreter, are not displayed
     when ALL-FRAMES? is `#f'.  Otherwise all frames are displayed.

     The parameter DISPLAY-ENV?, which defaults to `#f', controls if
     the frames are displayed with its environment (the variables
     accessible and their bindings).

     The parameters MAX-HEAD and MAX-TAIL, which default to 10 and 4
     respectively, control how many frames are displayed at the head
     and tail of the continuation.

     The parameter DEPTH, which defaults to 0, causes the frame numbers
     to be offset by that value.

     For example:

          > (define x #f)
          > (define (fib n)
              (if (< n 2)
                  (continuation-capture
                   (lambda (c) (set! x c) 1))
                  (+ (fib (- n 1))
                     (fib (- n 2)))))
          > (fib 10)
          89
          > (display-continuation-backtrace x)
          0  fib             (console)@7:12     (fib (- n 2))
          1  fib             (console)@7:12     (fib (- n 2))
          2  fib             (console)@7:12     (fib (- n 2))
          3  fib             (console)@7:12     (fib (- n 2))
          4  fib             (console)@7:12     (fib (- n 2))
          5  (interaction)   (console)@8:1      (fib 10)
          #f
          > (display-continuation-backtrace x (current-output-port) #t)
          0  fib             (console)@7:12     (fib (- n 2))
          1  fib             (console)@6:9      (+ (fib (- n 1)) (fib (- ...
          2  fib             (console)@7:12     (fib (- n 2))
          3  fib             (console)@6:9      (+ (fib (- n 1)) (fib (- ...
          4  fib             (console)@7:12     (fib (- n 2))
          5  fib             (console)@6:9      (+ (fib (- n 1)) (fib (- ...
          6  fib             (console)@7:12     (fib (- n 2))
          7  fib             (console)@6:9      (+ (fib (- n 1)) (fib (- ...
          8  fib             (console)@7:12     (fib (- n 2))
          9  fib             (console)@6:9      (+ (fib (- n 1)) (fib (- ...
          ...
          13 ##with-no-result-expected-toplevel
          14 ##repl-debug
          15 ##repl-debug-main
          16 ##kernel-handlers
          #f
          > (display-continuation-backtrace x (current-output-port) #f #t)
          0  fib             (console)@7:12     (fib (- n 2))
                  n = 2
          1  fib             (console)@7:12     (fib (- n 2))
                  n = 4
          2  fib             (console)@7:12     (fib (- n 2))
                  n = 6
          3  fib             (console)@7:12     (fib (- n 2))
                  n = 8
          4  fib             (console)@7:12     (fib (- n 2))
                  n = 10
          5  (interaction)   (console)@8:1      (fib 10)
          #f
          > (display-continuation-backtrace x (current-output-port) #f #f 2 1 100)
          100 fib            (console)@7:12     (fib (- n 2))
          101 fib            (console)@7:12     (fib (- n 2))
          ...
          105 (interaction)  (console)@8:1      (fib 10)
          #f


 -- procedure: make-thread-group [NAME [THREAD-GROUP]]
 -- procedure: thread-group? OBJ
 -- procedure: thread-group-name THREAD-GROUP
 -- procedure: thread-group-parent THREAD-GROUP
 -- procedure: thread-group-resume! THREAD-GROUP
 -- procedure: thread-group-suspend! THREAD-GROUP
 -- procedure: thread-group-terminate! THREAD-GROUP
 -- procedure: thread-group->thread-group-list THREAD-GROUP
 -- procedure: thread-group->thread-group-vector THREAD-GROUP
 -- procedure: thread-group->thread-list THREAD-GROUP
 -- procedure: thread-group->thread-vector THREAD-GROUP

 -- procedure: thread-state THREAD
 -- procedure: thread-state-uninitialized? THREAD-STATE
 -- procedure: thread-state-initialized? THREAD-STATE
 -- procedure: thread-state-active? THREAD-STATE
 -- procedure: thread-state-active-waiting-for THREAD-STATE
 -- procedure: thread-state-active-timeout THREAD-STATE
 -- procedure: thread-state-normally-terminated? THREAD-STATE
 -- procedure: thread-state-normally-terminated-result THREAD-STATE
 -- procedure: thread-state-abnormally-terminated? THREAD-STATE
 -- procedure: thread-state-abnormally-terminated-reason THREAD-STATE
 -- procedure: top [TIMEOUT [THREAD-GROUP [PORT]]]

 -- procedure: thread-interrupt! THREAD [THUNK]

 -- procedure: thread-suspend! THREAD
 -- procedure: thread-resume! THREAD

 -- procedure: thread-thread-group THREAD

 -- special form: define-type-of-thread name field...

 -- procedure: thread-init! THREAD THUNK [NAME [THREAD-GROUP]]

 -- procedure: initialized-thread-exception? OBJ
 -- procedure: initialized-thread-exception-procedure EXC
 -- procedure: initialized-thread-exception-arguments EXC

 -- procedure: uninitialized-thread-exception? OBJ
 -- procedure: uninitialized-thread-exception-procedure EXC
 -- procedure: uninitialized-thread-exception-arguments EXC

 -- procedure: inactive-thread-exception? OBJ
 -- procedure: inactive-thread-exception-procedure EXC
 -- procedure: inactive-thread-exception-arguments EXC

 -- procedure: rpc-remote-error-exception? OBJ
 -- procedure: rpc-remote-error-exception-procedure EXC
 -- procedure: rpc-remote-error-exception-arguments EXC
 -- procedure: rpc-remote-error-exception-message EXC

 -- procedure: timeout->time TIMEOUT

 -- procedure: open-dummy

 -- procedure: port-settings-set! PORT SETTINGS

 -- procedure: port-io-exception-handler-set! PORT HANDLER

 -- procedure: input-port-bytes-buffered PORT

 -- procedure: input-port-characters-buffered PORT

 -- procedure: nonempty-input-port-character-buffer-exception? OBJ
 -- procedure: nonempty-input-port-character-buffer-exception-arguments
          EXC
 -- procedure: nonempty-input-port-character-buffer-exception-procedure
          EXC

 -- procedure: repl-input-port
 -- procedure: repl-output-port
 -- procedure: console-port

 -- procedure: current-user-interrupt-handler [HANDLER]
 -- procedure: defer-user-interrupts

 -- procedure: primordial-exception-handler EXC

 -- procedure: err-code->string CODE

 -- procedure: foreign? OBJ
 -- procedure: foreign-tags FOREIGN
 -- procedure: foreign-address FOREIGN
 -- procedure: foreign-release! FOREIGN
 -- procedure: foreign-released? FOREIGN

 -- procedure: invalid-hash-number-exception? OBJ
 -- procedure: invalid-hash-number-exception-procedure EXC
 -- procedure: invalid-hash-number-exception-arguments EXC

 -- procedure: tcp-client-peer-socket-info TCP-CLIENT-PORT
 -- procedure: tcp-client-self-socket-info TCP-CLIENT-PORT

 -- procedure: tcp-server-socket-info TCP-SERVER-PORT

 -- procedure: socket-info? OBJ
 -- procedure: socket-info-address SOCKET-INFO
 -- procedure: socket-info-family SOCKET-INFO
 -- procedure: socket-info-port-number SOCKET-INFO

 -- procedure: system-version
 -- procedure: system-version-string

 -- procedure: system-type
 -- procedure: system-type-string
 -- procedure: configure-command-string

 -- procedure: system-stamp

 -- special form: future EXPR
 -- procedure: touch OBJ

 -- procedure: tty? OBJ
 -- procedure: tty-history TTY
 -- procedure: tty-history-set! TTY HISTORY
 -- procedure: tty-history-max-length-set! TTY N
 -- procedure: tty-paren-balance-duration-set! TTY DURATION
 -- procedure: tty-text-attributes-set! TTY ATTRIBUTES
 -- procedure: tty-mode-set! TTY MODE
 -- procedure: tty-type-set! TTY TYPE

 -- procedure: with-input-from-port PORT THUNK
 -- procedure: with-output-to-port PORT THUNK

 -- procedure: input-port-char-position PORT
 -- procedure: output-port-char-position PORT

 -- procedure: open-event-queue N

 -- procedure: main ...

 -- special form: define-record-type ...
 -- special form: define-type ...

 -- special form: namespace ...

 -- special form: this-source-file

 -- special form: receive ...

 -- special form: cond-expand ...

 -- special form: define-cond-expand-feature IDENT

 -- procedure: finite? X
 -- procedure: infinite? X
 -- procedure: nan? X

 -- undefined: six.!
 -- special form: six.!x X
 -- special form: six.&x X
 -- special form: six.*x X
 -- special form: six.++x X
 -- special form: six.+x X
 -- special form: six.-x X
 -- special form: six.-x X
 -- special form: six.arrow EXPR IDENT
 -- undefined: six.break
 -- special form: six.call FUNC ARG...
 -- undefined: six.case
 -- undefined: six.clause
 -- special form: six.compound STATEMENT...
 -- special form: six.cons X Y
 -- undefined: six.continue
 -- special form: six.define-procedure IDENT PROC
 -- special form: six.define-variable IDENT TYPE DIMS INIT
 -- special form: six.do-while STAT EXPR
 -- special form: six.dot EXPR IDENT
 -- special form: six.for STAT1 EXPR2 EXPR3 STAT2
 -- undefined: six.goto
 -- special form: six.identifier IDENT
 -- special form: six.if EXPR STAT1 [STAT2]
 -- special form: six.index EXPR1 EXPR2
 -- undefined: six.label
 -- special form: six.list X Y
 -- special form: six.literal VALUE
 -- procedure: six.make-array INIT DIM...
 -- special form: six.new IDENT ARG...
 -- special form: six.null
 -- special form: six.prefix DATUM
 -- special form: six.procedure TYPE PARAMS STAT
 -- special form: six.procedure-body STAT...
 -- undefined: six.return
 -- undefined: six.switch
 -- special form: six.while EXPR STAT...
 -- special form: six.x!=y X Y
 -- special form: six.x%=y X Y
 -- special form: six.x%y X Y
 -- special form: six.x&&y X Y
 -- special form: six.x&=y X Y
 -- special form: six.x&y X Y
 -- special form: six.x*=y X Y
 -- special form: six.x*y X Y
 -- special form: six.x++ X
 -- special form: six.x+=y X Y
 -- special form: six.x+y X Y
 -- special form: |six.x,y| X Y
 -- special form: six.x- X
 -- special form: six.x-=y X Y
 -- special form: six.x-y X Y
 -- special form: six.x/=y X Y
 -- special form: six.x/y X Y
 -- undefined: six.x:-y X Y
 -- special form: six.x:=y X Y
 -- special form: six.x:y X Y
 -- special form: six.x<<=y X Y
 -- special form: six.x<<y X Y
 -- special form: six.x<=y X Y
 -- special form: six.x<y X Y
 -- special form: six.x==y X Y
 -- special form: six.x=y X Y
 -- special form: six.x>=y X Y
 -- special form: six.x>>=y X Y
 -- special form: six.x>>y X Y
 -- special form: six.x>y X Y
 -- special form: six.x?y:z X Y Z
 -- special form: six.x^=y X Y
 -- special form: six.x^y X Y
 -- special form: |six.x\|=y| X Y
 -- special form: |six.x\|y| X Y
 -- special form: |six.x\|\|y| X Y
 -- special form: six.~x X

7 Namespaces
************

TO DO!

8 Characters and strings
************************

Gambit supports the Unicode character encoding standard.  Scheme
characters can be any of the characters whose Unicode encoding is in
the range 0 to #x10ffff (inclusive) but not in the range #xd800 to
#xdfff.  Source code can also contain any Unicode character, however to
read such source code properly `gsi' and `gsc' must be told which
character encoding to use for reading the source code (i.e. ASCII,
ISO-8859-1, UTF-8, etc).  This can be done by specifying the runtime
option `-:f' when `gsi' and `gsc' are started.

8.1 Extensions to character procedures
======================================

 -- procedure: char->integer CHAR
 -- procedure: integer->char N
     The procedure `char->integer' returns the Unicode encoding of the
     character CHAR.

     The procedure `integer->char' returns the character whose Unicode
     encoding is the exact integer N.

     For example:

          > (char->integer #\!)
          33
          > (integer->char 65)
          #\A
          > (integer->char (char->integer #\u1234))
          #\u1234
          > (integer->char #xd800)
          *** ERROR IN (console)@4.1 -- (Argument 1) Out of range
          (integer->char 55296)


 -- procedure: char=? CHAR1...
 -- procedure: char<? CHAR1...
 -- procedure: char>? CHAR1...
 -- procedure: char<=? CHAR1...
 -- procedure: char>=? CHAR1...
 -- procedure: char-ci=? CHAR1...
 -- procedure: char-ci<? CHAR1...
 -- procedure: char-ci>? CHAR1...
 -- procedure: char-ci<=? CHAR1...
 -- procedure: char-ci>=? CHAR1...
     These procedures take any number of arguments including no
     argument.  This is useful to test if the elements of a list are
     sorted in a particular order.  For example, testing that the list
     of characters `lst' is sorted in nondecreasing order can be done
     with the call `(apply char<? lst)'.


8.2 Extensions to string procedures
===================================

 -- procedure: string=? STRING1...
 -- procedure: string<? STRING1...
 -- procedure: string>? STRING1...
 -- procedure: string<=? STRING1...
 -- procedure: string>=? STRING1...
 -- procedure: string-ci=? STRING1...
 -- procedure: string-ci<? STRING1...
 -- procedure: string-ci>? STRING1...
 -- procedure: string-ci<=? STRING1...
 -- procedure: string-ci>=? STRING1...
     These procedures take any number of arguments including no
     argument.  This is useful to test if the elements of a list are
     sorted in a particular order.  For example, testing that the list
     of strings `lst' is sorted in nondecreasing order can be done with
     the call `(apply string<? lst)'.


9 Numbers
*********

9.1 Extensions to numeric procedures
====================================

 -- procedure: = Z1...
 -- procedure: < X1...
 -- procedure: > X1...
 -- procedure: <= X1...
 -- procedure: >= X1...
     These procedures take any number of arguments including no
     argument.  This is useful to test if the elements of a list are
     sorted in a particular order.  For example, testing that the list
     of numbers `lst' is sorted in nondecreasing order can be done with
     the call `(apply < lst)'.


9.2 IEEE floating point arithmetic
==================================

To better conform to IEEE floating point arithmetic the standard
numeric tower is extended with these special inexact reals:

`+inf.0'
     positive infinity

`-inf.0'
     negative infinity

`+nan.0'
     "not a number"

`-0.'
     negative zero (`0.' is the positive zero)

   The infinities and "not a number" are reals (i.e. `(real?  +inf.0)'
is `#t') but are not rational (i.e. `(rational?  +inf.0)' is `#f').

   Both zeros are numerically equal (i.e. `(= -0. 0.)' is `#t') but are
not equivalent (i.e. `(eqv? -0. 0.)' and `(equal?  -0. 0.)' are `#f').
All numerical comparisons with "not a number", including `(= +nan.0
+nan.0)', are `#f'.

9.3 Integer square root and nth root
====================================

 -- procedure: integer-sqrt N
     This procedure returns the integer part of the square root of the
     nonnegative exact integer N.

     For example:

          > (integer-sqrt 123)
          11


 -- procedure: integer-nth-root N1 N2
     This procedure returns the integer part of N1 raised to the power
     1/N2, where N1 is a nonnegative exact integer and N2 is a positive
     exact integer.

     For example:

          > (integer-nth-root 100 3)
          4


9.4 Bitwise-operations on exact integers
========================================

The procedures defined in this section are compatible with the
withdrawn "Integer Bitwise-operation Library SRFI" (SRFI 33).  Note
that some of the procedures specified in SRFI 33 are not provided.

   Most procedures in this section are specified in terms of the binary
representation of exact integers.  The two's complement representation
is assumed where an integer is composed of an infinite number of bits.
The upper section of an integer (the most significant bits) are either
an infinite sequence of ones when the integer is negative, or they are
an infinite sequence of zeros when the integer is nonnegative.

 -- procedure: arithmetic-shift N1 N2
     This procedure returns N1 shifted to the left by N2 bits, that is
     `(floor (* N1 (expt 2 N2)))'.  Both N1 and N2 must be exact
     integers.

     For example:

          > (arithmetic-shift 1000 7)  ; n1=...0000001111101000
          128000
          > (arithmetic-shift 1000 -6) ; n1=...0000001111101000
          15
          > (arithmetic-shift -23 -3)  ; n1=...1111111111101001
          -3


 -- procedure: bitwise-merge N1 N2 N3
     This procedure returns an exact integer whose bits combine the bits
     from N2 and N3 depending on N1.  The bit at index I of the result
     depends only on the bits at index I in N1, N2 and N3: it is equal
     to the bit in N2 when the bit in N1 is 0 and it is equal to the
     bit in N3 when the bit in N1 is 1.  All arguments must be exact
     integers.

     For example:

          > (bitwise-merge -4 -11 10) ; ...11111100 ...11110101 ...00001010
          9
          > (bitwise-merge 12 -11 10) ; ...00001100 ...11110101 ...00001010
          -7


 -- procedure: bitwise-and N...
     This procedure returns the bitwise "and" of the exact integers
     N....  The value -1 is returned when there are no arguments.

     For example:

          > (bitwise-and 6 12)  ; ...00000110 ...00001100
          4
          > (bitwise-and 6 -4)  ; ...00000110 ...11111100
          4
          > (bitwise-and -6 -4) ; ...11111010 ...11111100
          -8
          > (bitwise-and)
          -1


 -- procedure: bitwise-ior N...
     This procedure returns the bitwise "inclusive-or" of the exact
     integers N....  The value 0 is returned when there are no
     arguments.

     For example:

          > (bitwise-ior 6 12)  ; ...00000110 ...00001100
          14
          > (bitwise-ior 6 -4)  ; ...00000110 ...11111100
          -2
          > (bitwise-ior -6 -4) ; ...11111010 ...11111100
          -2
          > (bitwise-ior)
          0


 -- procedure: bitwise-xor N...
     This procedure returns the bitwise "exclusive-or" of the exact
     integers N....  The value 0 is returned when there are no
     arguments.

     For example:

          > (bitwise-xor 6 12)  ; ...00000110 ...00001100
          10
          > (bitwise-xor 6 -4)  ; ...00000110 ...11111100
          -6
          > (bitwise-xor -6 -4) ; ...11111010 ...11111100
          6
          > (bitwise-xor)
          0


 -- procedure: bitwise-not N
     This procedure returns the bitwise complement of the exact integer
     N.

     For example:

          > (bitwise-not 3)  ; ...00000011
          -4
          > (bitwise-not -1) ; ...11111111
          0


 -- procedure: bit-count N
     This procedure returns the bit count of the exact integer N.  If N
     is nonnegative, the bit count is the number of 1 bits in the two's
     complement representation of N.  If N is negative, the bit count
     is the number of 0 bits in the two's complement representation of
     N.

     For example:

          > (bit-count 0)   ; ...00000000
          0
          > (bit-count 1)   ; ...00000001
          1
          > (bit-count 2)   ; ...00000010
          1
          > (bit-count 3)   ; ...00000011
          2
          > (bit-count 4)   ; ...00000100
          1
          > (bit-count -23) ; ...11101001
          3


 -- procedure: integer-length N
     This procedure returns the bit length of the exact integer N.  If
     N is a positive integer the bit length is one more than the index
     of the highest 1 bit (the least significant bit is at index 0).
     If N is a negative integer the bit length is one more than the
     index of the highest 0 bit.  If N is zero, the bit length is 0.

     For example:

          > (integer-length 0)   ; ...00000000
          0
          > (integer-length 1)   ; ...00000001
          1
          > (integer-length 2)   ; ...00000010
          2
          > (integer-length 3)   ; ...00000011
          2
          > (integer-length 4)   ; ...00000100
          3
          > (integer-length -23) ; ...11101001
          5


 -- procedure: bit-set? N1 N2
     This procedure returns a boolean indicating if the bit at index N1
     of N2 is set (i.e. equal to 1) or not.  Both N1 and N2 must be
     exact integers, and N1 must be nonnegative.

     For example:

          > (map (lambda (i) (bit-set? i -23)) ; ...11101001
                 '(7 6 5 4 3 2 1 0))
          (#t #t #t #f #t #f #f #t)


 -- procedure: any-bits-set? N1 N2
     This procedure returns a boolean indicating if the bitwise and of
     N1 and N2 is different from zero or not.  This procedure is
     implemented more efficiently than the naive definition:

          (define (any-bits-set? n1 n2) (not (zero? (bitwise-and n1 n2))))

     For example:

          > (any-bits-set? 5 10)   ; ...00000101 ...00001010
          #f
          > (any-bits-set? -23 32) ; ...11101001 ...00100000
          #t


 -- procedure: all-bits-set? N1 N2
     This procedure returns a boolean indicating if the bitwise and of
     N1 and N2 is equal to N1 or not.  This procedure is implemented
     more efficiently than the naive definition:

          (define (all-bits-set? n1 n2) (= n1 (bitwise-and n1 n2)))

     For example:

          > (all-bits-set? 1 3) ; ...00000001 ...00000011
          #t
          > (all-bits-set? 7 3) ; ...00000111 ...00000011
          #f


 -- procedure: first-bit-set N
     This procedure returns the bit index of the least significant bit
     of N equal to 1 (which is also the number of 0 bits that are below
     the least significant 1 bit).  This procedure returns `-1' when N
     is zero.

     For example:

          > (first-bit-set 24) ; ...00011000
          3
          > (first-bit-set 0)  ; ...00000000
          -1


 -- procedure: extract-bit-field N1 N2 N3
 -- procedure: test-bit-field? N1 N2 N3
 -- procedure: clear-bit-field N1 N2 N3
 -- procedure: replace-bit-field N1 N2 N3 N4
 -- procedure: copy-bit-field N1 N2 N3 N4
     These procedures operate on a bit-field which is N1 bits wide
     starting at bit index N2.  All arguments must be exact integers
     and N1 and N2 must be nonnegative.

     The procedure `extract-bit-field' returns the bit-field of N3
     shifted to the right so that the least significant bit of the
     bit-field is the least significant bit of the result.

     The procedure `test-bit-field?' returns `#t' if any bit in the
     bit-field of N3 is equal to 1, otherwise `#f' is returned.

     The procedure `clear-bit-field' returns N3 with all bits in the
     bit-field replaced with 0.

     The procedure `replace-bit-field' returns N4 with the bit-field
     replaced with the least-significant N1 bits of N3.

     The procedure `copy-bit-field' returns N4 with the bit-field
     replaced with the (same index and size) bit-field in N3.

     For example:

          > (extract-bit-field 5 2 -37)    ; ...11011011
          22
          > (test-bit-field? 5 2 -37)      ; ...11011011
          #t
          > (test-bit-field? 1 2 -37)      ; ...11011011
          #f
          > (clear-bit-field 5 2 -37)      ; ...11011011
          -125
          > (replace-bit-field 5 2 -6 -37) ; ...11111010 ...11011011
          -21
          > (copy-bit-field 5 2 -6 -37)    ; ...11111010 ...11011011
          -5


9.5 Fixnum specific operations
==============================

 -- procedure: fixnum? OBJ

 -- procedure: fx* N1...

 -- procedure: fx+ N1...

 -- procedure: fx- N1 N2...

 -- procedure: fx< N1...

 -- procedure: fx<= N1...

 -- procedure: fx= N1...

 -- procedure: fx> N1...

 -- procedure: fx>= N1...

 -- procedure: fxabs N

 -- procedure: fxand N1...

 -- procedure: fxarithmetic-shift N1 N2

 -- procedure: fxarithmetic-shift-left N1 N2

 -- procedure: fxarithmetic-shift-right N1 N2

 -- procedure: fxbit-count N

 -- procedure: fxbit-set? N1 N2

 -- procedure: fxeven? N

 -- procedure: fxfirst-bit-set N

 -- procedure: fxif N1 N2 N3

 -- procedure: fxior N1...

 -- procedure: fxlength N

 -- procedure: fxmax N1 N2...

 -- procedure: fxmin N1 N2...

 -- procedure: fxmodulo N1 N2

 -- procedure: fxnegative? N

 -- procedure: fxnot N

 -- procedure: fxodd? N

 -- procedure: fxpositive? N

 -- procedure: fxquotient N1 N2

 -- procedure: fxremainder N1 N2

 -- procedure: fxwrap* N1...

 -- procedure: fxwrap+ N1...

 -- procedure: fxwrap- N1 N2...

 -- procedure: fxwrapabs N

 -- procedure: fxwraparithmetic-shift N1 N2

 -- procedure: fxwraparithmetic-shift-left N1 N2

 -- procedure: fxwraplogical-shift-right N1 N2

 -- procedure: fxwrapquotient N1 N2

 -- procedure: fxxor N1...

 -- procedure: fxzero? N

 -- procedure: fixnum-overflow-exception? OBJ
 -- procedure: fixnum-overflow-exception-procedure EXC
 -- procedure: fixnum-overflow-exception-arguments EXC
     Fixnum-overflow-exception objects are raised by some of the fixnum
     specific procedures when the result is larger than can fit in a
     fixnum.  The parameter EXC must be a fixnum-overflow-exception
     object.

     The procedure `fixnum-overflow-exception?' returns `#t' when OBJ
     is a fixnum-overflow-exception object and `#f' otherwise.

     The procedure `fixnum-overflow-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `fixnum-overflow-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (fixnum-overflow-exception? exc)
                  (list (fixnum-overflow-exception-procedure exc)
                        (fixnum-overflow-exception-arguments exc))
                  'not-fixnum-overflow-exception))
          > (with-exception-catcher
              handler
              (lambda () (fx* 100000 100000)))
          (#<procedure #2 fx*> (100000 100000))


9.6 Flonum specific operations
==============================

 -- procedure: flonum? OBJ

 -- procedure: fixnum->flonum N

 -- procedure: fl* X1...

 -- procedure: fl+ X1...

 -- procedure: fl- X1 X2...

 -- procedure: fl/ X1 X2

 -- procedure: fl< X1...

 -- procedure: fl<= X1...

 -- procedure: fl= X1...

 -- procedure: fl> X1...

 -- procedure: fl>= X1...

 -- procedure: flabs X

 -- procedure: flacos X

 -- procedure: flasin X

 -- procedure: flatan X
 -- procedure: flatan Y X

 -- procedure: flceiling X

 -- procedure: flcos X

 -- procedure: fldenominator X

 -- procedure: fleven? X

 -- procedure: flexp X

 -- procedure: flexpt X Y

 -- procedure: flfinite? X

 -- procedure: flfloor X

 -- procedure: flinfinite? X

 -- procedure: flinteger? X

 -- procedure: fllog X

 -- procedure: flmax X1 X2...

 -- procedure: flmin X1 X2...

 -- procedure: flnan? X

 -- procedure: flnegative? X

 -- procedure: flnumerator X

 -- procedure: flodd? X

 -- procedure: flpositive? X

 -- procedure: flround X

 -- procedure: flsin X

 -- procedure: flsqrt X

 -- procedure: fltan X

 -- procedure: fltruncate X

 -- procedure: flzero? X

9.7 Pseudo random numbers
=========================

The procedures and variables defined in this section are compatible
with the "Sources of Random Bits SRFI" (SRFI 27).  The implementation
is based on Pierre L'Ecuyer's MRG32k3a pseudo random number generator.
At the heart of SRFI 27's interface is the random source type which
encapsulates the state of a pseudo random number generator.  The state
of a random source object changes every time a pseudo random number is
generated from this random source object.

 -- variable: default-random-source
     The global variable `default-random-source' is bound to the random
     source object which is used by the `random-integer',
     `random-real', `random-u8vector' and `random-f64vector' procedures.


 -- procedure: random-integer N
     This procedure returns a pseudo random exact integer in the range
     0 to N-1.  The random source object in the global variable
     `default-random-source' is used to generate this number.  The
     parameter N must be a positive exact integer.

     For example:

          > (random-integer 100)
          24
          > (random-integer 100)
          2
          > (random-integer 10000000000000000000000000000000000000000)
          6143360270902284438072426748425263488507


 -- procedure: random-real
     This procedure returns a pseudo random inexact real between, but
     not including, 0 and 1.  The random source object in the global
     variable `default-random-source' is used to generate this number.

     For example:

          > (random-real)
          .24230672079133753
          > (random-real)
          .02317001922506932


 -- procedure: random-u8vector N
     This procedure returns a u8vector of length N containing pseudo
     random exact integers in the range 0 to 255.  The random source
     object in the global variable `default-random-source' is used to
     generate these numbers.  The parameter N must be a nonnegative
     exact integer.

     For example:

          > (random-u8vector 10)
          #u8(200 53 29 202 3 85 208 187 73 219)


 -- procedure: random-f64vector N
     This procedure returns a f64vector of length N containing pseudo
     random inexact reals between, but not including, 0 and 1.  The
     random source object in the global variable
     `default-random-source' is used to generate these numbers.  The
     parameter N must be a nonnegative exact integer.

     For example:

          > (random-f64vector 3)
          #f64(.7145854494613069 .47089632669147946 .5400124875182746)


 -- procedure: make-random-source
     This procedure returns a new random source object initialized to a
     predetermined state (to initialize to a pseudo random state the
     procedure `random-source-randomize!' should be called).

     For example:

          > (define rs (make-random-source))
          > ((random-source-make-integers rs) 10000000)
          8583952


 -- procedure: random-source? OBJ
     This procedure returns `#t' when OBJ is a random source object and
     `#f' otherwise.

     For example:

          > (random-source? default-random-source)
          #t
          > (random-source? 123)
          #f


 -- procedure: random-source-state-ref RANDOM-SOURCE
 -- procedure: random-source-state-set! RANDOM-SOURCE STATE
     The procedure `random-source-state-ref' extracts the state of the
     random source object RANDOM-SOURCE and returns a vector containing
     the state.

     The procedure `random-source-state-set!' restores the state of the
     random source object RANDOM-SOURCE to STATE which must be a vector
     returned from a call to the procedure `random-source-state-ref'.

     For example:

          > (define s (random-source-state-ref default-random-source))
          > (random-integer 10000000000000000000000000000000000000000)
          7583880188903074396261960585615270693321
          > (random-source-state-set! default-random-source s)
          > (random-integer 10000000000000000000000000000000000000000)
          7583880188903074396261960585615270693321


 -- procedure: random-source-randomize! RANDOM-SOURCE
 -- procedure: random-source-pseudo-randomize! RANDOM-SOURCE I J
     These procedures change the state of the random source object
     RANDOM-SOURCE.  The procedure `random-source-randomize!' sets the
     random source object to a state that depends on the current time
     (which for typical uses can be considered to randomly initialize
     the state).  The procedure `random-source-pseudo-randomize!' sets
     the random source object to a state that is determined only by the
     current state and the nonnegative exact integers I and J.  For
     both procedures the value returned is unspecified.

     For example:

          > (define s (random-source-state-ref default-random-source))
          > (random-source-pseudo-randomize! default-random-source 5 99)
          > (random-integer 10000000000000000000000000000000000000000)
          9816755163910623041601722050112674079767
          > (random-source-state-set! default-random-source s)
          > (random-source-pseudo-randomize! default-random-source 5 99)
          > (random-integer 10000000000000000000000000000000000000000)
          9816755163910623041601722050112674079767
          > (random-source-pseudo-randomize! default-random-source 5 99)
          > (random-integer 10000000000000000000000000000000000000000)
          9816755163910623041601722050112674079767
          > (random-source-state-set! default-random-source s)
          > (random-source-randomize! default-random-source)
          > (random-integer 10000000000000000000000000000000000000000)
          2271441220851914333384493143687768110622
          > (random-source-state-set! default-random-source s)
          > (random-source-randomize! default-random-source)
          > (random-integer 10000000000000000000000000000000000000000)
          6247966138948323029033944059178072366895


 -- procedure: random-source-make-integers RANDOM-SOURCE
     This procedure returns a procedure for generating pseudo random
     exact integers using the random source object RANDOM-SOURCE.  The
     returned procedure accepts a single parameter N, a positive exact
     integer, and returns a pseudo random exact integer in the range 0
     to N-1.

     For example:

          > (define rs (make-random-source))
          > (define ri (random-source-make-integers rs))
          > (ri 10000000)
          8583952
          > (ri 10000000)
          2879793


 -- procedure: random-source-make-reals RANDOM-SOURCE [PRECISION]
     This procedure returns a procedure for generating pseudo random
     inexact reals using the random source object RANDOM-SOURCE.  The
     returned procedure accepts no parameters and returns a pseudo
     random inexact real between, but not including, 0 and 1.  The
     optional parameter PRECISION specifies an upper bound on the
     minimum amount by which two generated pseudo-random numbers can be
     separated.

     For example:

          > (define rs (make-random-source))
          > (define rr (random-source-make-reals rs))
          > (rr)
          .857402537562821
          > (rr)
          .2876463473845367


 -- procedure: random-source-make-u8vectors RANDOM-SOURCE
     This procedure returns a procedure for generating pseudo random
     u8vectors using the random source object RANDOM-SOURCE.  The
     returned procedure accepts a single parameter N, a nonnegative
     exact integer, and returns a u8vector of length N containing
     pseudo random exact integers in the range 0 to 255.

     For example:

          > (define rs (make-random-source))
          > (define rv (random-source-make-u8vectors rs))
          > (rv 10)
          #u8(200 53 29 202 3 85 208 187 73 219)
          > (rv 10)
          #u8(113 8 182 120 138 103 53 192 40 176)


 -- procedure: random-source-make-f64vectors RANDOM-SOURCE [PRECISION]
     This procedure returns a procedure for generating pseudo random
     f64vectors using the random source object RANDOM-SOURCE.  The
     returned procedure accepts a single parameter N, a nonnegative
     exact integer, and returns an f64vector of length N containing
     pseudo random inexact reals between, but not including, 0 and 1.
     The optional parameter PRECISION specifies an upper bound on the
     minimum amount by which two generated pseudo-random numbers can be
     separated.

     For example:

          > (define rs (make-random-source))
          > (define rv (random-source-make-f64vectors rs))
          > (rv 3)
          #f64(.7342236104231586 .2876463473845367 .8574025375628211)
          > (rv 3)
          #f64(.013863292728449427 .33449296573515447 .8162050798467028)


10 Homogeneous vectors
**********************

Homogeneous vectors are vectors containing raw numbers of the same type
(signed or unsigned exact integers or inexact reals).  There are 10
types of homogeneous vectors: `s8vector' (vector of exact integers in
the range -2^7 to 2^7-1), `u8vector' (vector of exact integers in the
range 0 to 2^8-1), `s16vector' (vector of exact integers in the range
-2^15 to 2^15-1), `u16vector' (vector of exact integers in the range 0
to 2^16-1), `s32vector' (vector of exact integers in the range -2^31 to
2^31-1), `u32vector' (vector of exact integers in the range 0 to
2^32-1), `s64vector' (vector of exact integers in the range -2^63 to
2^63-1), `u64vector' (vector of exact integers in the range 0 to
2^64-1), `f32vector' (vector of 32 bit floating point numbers), and
`f64vector' (vector of 64 bit floating point numbers).

   The lexical syntax of homogeneous vectors is specified in *Note
Homogeneous vector syntax::.

   The procedures available for homogeneous vectors, listed below, are
the analog of the normal vector/string procedures for each of the
homogeneous vector types.

 -- procedure: s8vector? OBJ
 -- procedure: make-s8vector K [FILL]
 -- procedure: s8vector EXACT-INT8...
 -- procedure: s8vector-length S8VECTOR
 -- procedure: s8vector-ref S8VECTOR K
 -- procedure: s8vector-set! S8VECTOR K EXACT-INT8
 -- procedure: s8vector->list S8VECTOR
 -- procedure: list->s8vector LIST-OF-EXACT-INT8
 -- procedure: s8vector-fill! S8VECTOR FILL
 -- procedure: subs8vector-fill! VECTOR START END FILL
 -- procedure: append-s8vectors LST
 -- procedure: s8vector-copy S8VECTOR
 -- procedure: s8vector-append S8VECTOR...
 -- procedure: subs8vector S8VECTOR START END
 -- procedure: subs8vector-move! SRC-S8VECTOR SRC-START SRC-END
          DST-S8VECTOR DST-START
 -- procedure: s8vector-shrink! S8VECTOR K

 -- procedure: u8vector? OBJ
 -- procedure: make-u8vector K [FILL]
 -- procedure: u8vector EXACT-INT8...
 -- procedure: u8vector-length U8VECTOR
 -- procedure: u8vector-ref U8VECTOR K
 -- procedure: u8vector-set! U8VECTOR K EXACT-INT8
 -- procedure: u8vector->list U8VECTOR
 -- procedure: list->u8vector LIST-OF-EXACT-INT8
 -- procedure: u8vector-fill! U8VECTOR FILL
 -- procedure: subu8vector-fill! VECTOR START END FILL
 -- procedure: append-u8vectors LST
 -- procedure: u8vector-copy U8VECTOR
 -- procedure: u8vector-append U8VECTOR...
 -- procedure: subu8vector U8VECTOR START END
 -- procedure: subu8vector-move! SRC-U8VECTOR SRC-START SRC-END
          DST-U8VECTOR DST-START
 -- procedure: u8vector-shrink! U8VECTOR K

 -- procedure: s16vector? OBJ
 -- procedure: make-s16vector K [FILL]
 -- procedure: s16vector EXACT-INT16...
 -- procedure: s16vector-length S16VECTOR
 -- procedure: s16vector-ref S16VECTOR K
 -- procedure: s16vector-set! S16VECTOR K EXACT-INT16
 -- procedure: s16vector->list S16VECTOR
 -- procedure: list->s16vector LIST-OF-EXACT-INT16
 -- procedure: s16vector-fill! S16VECTOR FILL
 -- procedure: subs16vector-fill! VECTOR START END FILL
 -- procedure: append-s16vectors LST
 -- procedure: s16vector-copy S16VECTOR
 -- procedure: s16vector-append S16VECTOR...
 -- procedure: subs16vector S16VECTOR START END
 -- procedure: subs16vector-move! SRC-S16VECTOR SRC-START SRC-END
          DST-S16VECTOR DST-START
 -- procedure: s16vector-shrink! S16VECTOR K

 -- procedure: u16vector? OBJ
 -- procedure: make-u16vector K [FILL]
 -- procedure: u16vector EXACT-INT16...
 -- procedure: u16vector-length U16VECTOR
 -- procedure: u16vector-ref U16VECTOR K
 -- procedure: u16vector-set! U16VECTOR K EXACT-INT16
 -- procedure: u16vector->list U16VECTOR
 -- procedure: list->u16vector LIST-OF-EXACT-INT16
 -- procedure: u16vector-fill! U16VECTOR FILL
 -- procedure: subu16vector-fill! VECTOR START END FILL
 -- procedure: append-u16vectors LST
 -- procedure: u16vector-copy U16VECTOR
 -- procedure: u16vector-append U16VECTOR...
 -- procedure: subu16vector U16VECTOR START END
 -- procedure: subu16vector-move! SRC-U16VECTOR SRC-START SRC-END
          DST-U16VECTOR DST-START
 -- procedure: u16vector-shrink! U16VECTOR K

 -- procedure: s32vector? OBJ
 -- procedure: make-s32vector K [FILL]
 -- procedure: s32vector EXACT-INT32...
 -- procedure: s32vector-length S32VECTOR
 -- procedure: s32vector-ref S32VECTOR K
 -- procedure: s32vector-set! S32VECTOR K EXACT-INT32
 -- procedure: s32vector->list S32VECTOR
 -- procedure: list->s32vector LIST-OF-EXACT-INT32
 -- procedure: s32vector-fill! S32VECTOR FILL
 -- procedure: subs32vector-fill! VECTOR START END FILL
 -- procedure: append-s32vectors LST
 -- procedure: s32vector-copy S32VECTOR
 -- procedure: s32vector-append S32VECTOR...
 -- procedure: subs32vector S32VECTOR START END
 -- procedure: subs32vector-move! SRC-S32VECTOR SRC-START SRC-END
          DST-S32VECTOR DST-START
 -- procedure: s32vector-shrink! S32VECTOR K

 -- procedure: u32vector? OBJ
 -- procedure: make-u32vector K [FILL]
 -- procedure: u32vector EXACT-INT32...
 -- procedure: u32vector-length U32VECTOR
 -- procedure: u32vector-ref U32VECTOR K
 -- procedure: u32vector-set! U32VECTOR K EXACT-INT32
 -- procedure: u32vector->list U32VECTOR
 -- procedure: list->u32vector LIST-OF-EXACT-INT32
 -- procedure: u32vector-fill! U32VECTOR FILL
 -- procedure: subu32vector-fill! VECTOR START END FILL
 -- procedure: append-u32vectors LST
 -- procedure: u32vector-copy U32VECTOR
 -- procedure: u32vector-append U32VECTOR...
 -- procedure: subu32vector U32VECTOR START END
 -- procedure: subu32vector-move! SRC-U32VECTOR SRC-START SRC-END
          DST-U32VECTOR DST-START
 -- procedure: u32vector-shrink! U32VECTOR K

 -- procedure: s64vector? OBJ
 -- procedure: make-s64vector K [FILL]
 -- procedure: s64vector EXACT-INT64...
 -- procedure: s64vector-length S64VECTOR
 -- procedure: s64vector-ref S64VECTOR K
 -- procedure: s64vector-set! S64VECTOR K EXACT-INT64
 -- procedure: s64vector->list S64VECTOR
 -- procedure: list->s64vector LIST-OF-EXACT-INT64
 -- procedure: s64vector-fill! S64VECTOR FILL
 -- procedure: subs64vector-fill! VECTOR START END FILL
 -- procedure: append-s64vectors LST
 -- procedure: s64vector-copy S64VECTOR
 -- procedure: s64vector-append S64VECTOR...
 -- procedure: subs64vector S64VECTOR START END
 -- procedure: subs64vector-move! SRC-S64VECTOR SRC-START SRC-END
          DST-S64VECTOR DST-START
 -- procedure: s64vector-shrink! S64VECTOR K

 -- procedure: u64vector? OBJ
 -- procedure: make-u64vector K [FILL]
 -- procedure: u64vector EXACT-INT64...
 -- procedure: u64vector-length U64VECTOR
 -- procedure: u64vector-ref U64VECTOR K
 -- procedure: u64vector-set! U64VECTOR K EXACT-INT64
 -- procedure: u64vector->list U64VECTOR
 -- procedure: list->u64vector LIST-OF-EXACT-INT64
 -- procedure: u64vector-fill! U64VECTOR FILL
 -- procedure: subu64vector-fill! VECTOR START END FILL
 -- procedure: append-u64vectors LST
 -- procedure: u64vector-copy U64VECTOR
 -- procedure: u64vector-append U64VECTOR...
 -- procedure: subu64vector U64VECTOR START END
 -- procedure: subu64vector-move! SRC-U64VECTOR SRC-START SRC-END
          DST-U64VECTOR DST-START
 -- procedure: u64vector-shrink! U64VECTOR K

 -- procedure: f32vector? OBJ
 -- procedure: make-f32vector K [FILL]
 -- procedure: f32vector INEXACT-REAL...
 -- procedure: f32vector-length F32VECTOR
 -- procedure: f32vector-ref F32VECTOR K
 -- procedure: f32vector-set! F32VECTOR K INEXACT-REAL
 -- procedure: f32vector->list F32VECTOR
 -- procedure: list->f32vector LIST-OF-INEXACT-REAL
 -- procedure: f32vector-fill! F32VECTOR FILL
 -- procedure: subf32vector-fill! VECTOR START END FILL
 -- procedure: append-f32vectors LST
 -- procedure: f32vector-copy F32VECTOR
 -- procedure: f32vector-append F32VECTOR...
 -- procedure: subf32vector F32VECTOR START END
 -- procedure: subf32vector-move! SRC-F32VECTOR SRC-START SRC-END
          DST-F32VECTOR DST-START
 -- procedure: f32vector-shrink! F32VECTOR K

 -- procedure: f64vector? OBJ
 -- procedure: make-f64vector K [FILL]
 -- procedure: f64vector INEXACT-REAL...
 -- procedure: f64vector-length F64VECTOR
 -- procedure: f64vector-ref F64VECTOR K
 -- procedure: f64vector-set! F64VECTOR K INEXACT-REAL
 -- procedure: f64vector->list F64VECTOR
 -- procedure: list->f64vector LIST-OF-INEXACT-REAL
 -- procedure: f64vector-fill! F64VECTOR FILL
 -- procedure: subf64vector-fill! VECTOR START END FILL
 -- procedure: append-f64vectors LST
 -- procedure: f64vector-copy F64VECTOR
 -- procedure: f64vector-append F64VECTOR...
 -- procedure: subf64vector F64VECTOR START END
 -- procedure: subf64vector-move! SRC-F64VECTOR SRC-START SRC-END
          DST-F64VECTOR DST-START
 -- procedure: f64vector-shrink! F64VECTOR K

   For example:

     > (define v (u8vector 10 255 13))
     > (u8vector-set! v 2 99)
     > v
     #u8(10 255 99)
     > (u8vector-ref v 1)
     255
     > (u8vector->list v)
     (10 255 99)
     > (u8vector-shrink! v 2)
     > (v)
     #u8(10 255)

 -- procedure: object->u8vector OBJ [ENCODER]
 -- procedure: u8vector->object U8VECTOR [DECODER]
     The procedure `object->u8vector' returns a u8vector that contains
     the sequence of bytes that encodes the object OBJ.  The procedure
     `u8vector->object' decodes the sequence of bytes contained in the
     u8vector U8VECTOR, which was produced by the procedure
     `object->u8vector', and reconstructs an object structurally equal
     to the original object.  In other words the procedures
     `object->u8vector' and `u8vector->object' respectively perform
     serialization and deserialization of Scheme objects.  Note that
     some objects are non-serializable (e.g. threads, wills, some types
     of ports, and any object containing a non-serializable object).

     The optional ENCODER and DECODER parameters are single parameter
     procedures which default to the identity function.  The ENCODER
     procedure is called during serialization.  As the serializer walks
     through OBJ, it calls the ENCODER procedure on each sub-object X
     that is encountered.  The ENCODER transforms the object X into an
     object Y that will be serialized instead of X.  Similarly the
     DECODER procedure is called during deserialization.  When an
     object Y is encountered, the DECODER procedure is called to
     transform it into the object X that is the result of
     deserialization.

     The ENCODER and DECODER procedures are useful to customize the
     serialized representation of objects.  In particular, it can be
     used to define the semantics of serializing objects, such as
     threads and ports, that would otherwise not be serializable.  The
     DECODER procedure is typically the inverse of the ENCODER
     procedure, i.e. `(DECODER (ENCODER X))' = `X'.

     For example:

          > (define (make-adder x) (lambda (y) (+ x y)))
          > (define f (make-adder 10))
          > (define a (object->u8vector f))
          > (define b (u8vector->object a))
          > (u8vector-length a)
          1639
          > (f 5)
          15
          > (b 5)
          15
          > (pp b)
          (lambda (y) (+ x y))


11 Hashing and weak references
******************************

11.1 Hashing
============

 -- procedure: object->serial-number OBJ
 -- procedure: serial-number->object N [DEFAULT]
     All Scheme objects are uniquely identified with a serial number
     which is a nonnegative exact integer.  The `object->serial-number'
     procedure returns the serial number of object OBJ.  This serial
     number is only allocated the first time the `object->serial-number'
     procedure is called on that object.  Objects which do not have an
     external textual representation that can be read by the `read'
     procedure, use an external textual representation that includes a
     serial number of the form `#N'.  Consequently, the procedures
     `write', `pretty-print', etc will call the `object->serial-number'
     procedure to get the serial number, and this may cause the serial
     number to be allocated.

     The `serial-number->object' procedure takes an exact integer
     parameter N and returns the object whose serial number is N.  If
     no object currently exists with that serial number, DEFAULT is
     returned if it is specified, otherwise an
     unbound-serial-number-exception object is raised.  The reader
     defines the following abbreviation for calling
     `serial-number->object': the syntax `#N', where N is a sequence of
     decimal digits and it is not followed by ``='' or ``#'', is
     equivalent to the list `(serial-number->object N)'.

     For example:

          > (define z (list (lambda (x) (* x x)) (lambda (y) (/ 1 y))))
          > z
          (#<procedure #2> #<procedure #3>)
          > (#3 10)
          1/10
          > '(#3 10)
          ((serial-number->object 3) 10)
          > car
          #<procedure #4 car>
          > (#4 z)
          #<procedure #2>


 -- procedure: unbound-serial-number-exception? OBJ
 -- procedure: unbound-serial-number-exception-procedure EXC
 -- procedure: unbound-serial-number-exception-arguments EXC
     Unbound-serial-number-exception objects are raised by the procedure
     `serial-number->object' when no object currently exists with that
     serial number.  The parameter EXC must be an
     unbound-serial-number-exception object.

     The procedure `unbound-serial-number-exception?' returns `#t' when
     OBJ is a unbound-serial-number-exception object and `#f' otherwise.

     The procedure `unbound-serial-number-exception-procedure' returns
     the procedure that raised EXC.

     The procedure `unbound-serial-number-exception-arguments' returns
     the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (unbound-serial-number-exception? exc)
                  (list (unbound-serial-number-exception-procedure exc)
                        (unbound-serial-number-exception-arguments exc))
                  'not-unbound-serial-number-exception))
          > (with-exception-catcher
              handler
              (lambda () (serial-number->object 1000)))
          (#<procedure #2 serial-number->object> (1000))


 -- procedure: symbol-hash SYMBOL
     The `symbol-hash' procedure returns the hash number of the symbol
     SYMBOL.  The hash number is a small exact integer (fixnum).  When
     SYMBOL is an interned symbol the value returned is the same as
     `(string=?-hash (symbol->string SYMBOL))'.

     For example:

          > (symbol-hash 'car)
          444471047


 -- procedure: keyword-hash KEYWORD
     The `keyword-hash' procedure returns the hash number of the
     keyword KEYWORD.  The hash number is a small exact integer
     (fixnum).  When KEYWORD is an interned keyword the value returned
     is the same as `(string=?-hash (keyword->string KEYWORD))'.

     For example:

          > (keyword-hash car:)
          444471047


 -- procedure: string=?-hash STRING
     The `string=?-hash' procedure returns the hash number of the
     string STRING.  The hash number is a small exact integer (fixnum).
     For any two strings S1 and S2, `(string=?  S1 S2)' implies `(=
     (string=?-hash S1) (string=?-hash S2))'.

     For example:

          > (string=?-hash "car")
          444471047


 -- procedure: string-ci=?-hash STRING
     The `string-ci=?-hash' procedure returns the hash number of the
     string STRING.  The hash number is a small exact integer (fixnum).
     For any two strings S1 and S2, `(string-ci=?  S1 S2)' implies `(=
     (string-ci=?-hash S1) (string-ci=?-hash S2))'.

     For example:

          > (string-ci=?-hash "CaR")
          444471047


 -- procedure: eq?-hash OBJ
     The `eq?-hash' procedure returns the hash number of the object
     OBJ.  The hash number is a small exact integer (fixnum).  For any
     two objects O1 and O2, `(eq? O1 O2)' implies `(= (eq?-hash O1)
     (eq?-hash O2))'.

     For example:

          > (eq?-hash #t)
          536870910


 -- procedure: eqv?-hash OBJ
     The `eqv?-hash' procedure returns the hash number of the object
     OBJ.  The hash number is a small exact integer (fixnum).  For any
     two objects O1 and O2, `(eqv? O1 O2)' implies `(= (eqv?-hash O1)
     (eqv?-hash O2))'.

     For example:

          > (eqv?-hash 1.5)
          496387656


 -- procedure: equal?-hash OBJ
     The `equal?-hash' procedure returns the hash number of the object
     OBJ.  The hash number is a small exact integer (fixnum).  For any
     two objects O1 and O2, `(equal? O1 O2)' implies `(= (equal?-hash
     O1) (equal?-hash O2))'.

     For example:

          > (equal?-hash (list 1 2 3))
          442438567


11.2 Weak references
====================

The garbage collector is responsible for reclaiming objects that are no
longer needed by the program.  This is done by analyzing the
reachability graph of all objects from the roots (i.e. the global
variables, the runnable threads, permanently allocated objects such as
procedures defined in a compiled file, nonexecutable wills, etc).  If a
root or a reachable object X contains a reference to an object Y then Y
is reachable.  As a general rule, unreachable objects are reclaimed by
the garbage collector.

   There are two types of references: strong references and weak
references.  Most objects, including pairs, vectors, records and
closures, contain strong references.  An object X is "strongly
reachable" if there is a path from the roots to X that traverses only
strong references.  Weak references only occur in wills and tables.
There are two types of weak references: will-weak references and
table-weak references.  If all paths from the roots to an object Y
traverse at least one table-weak reference, then Y will be reclaimed by
the garbage collector.  The will-weak references are used for
finalization and are explained in the next section.

11.2.1 Wills
------------

The following procedures implement the "will" data type.  Will objects
provide support for finalization.  A will is an object that contains a
will-weak reference to a TESTATOR object (the object attached to the
will), and a strong reference to an ACTION procedure which is a one
parameter procedure which is called when the will is executed.

 -- procedure: make-will TESTATOR ACTION
 -- procedure: will? OBJ
 -- procedure: will-testator WILL
 -- procedure: will-execute! WILL
     The `make-will' procedure creates a will object with the given
     TESTATOR object and ACTION procedure.  The `will?' procedure tests
     if OBJ is a will object.  The `will-testator' procedure gets the
     testator object attached to the WILL.  The `will-execute!'
     procedure executes WILL.

     A will becomes "executable" when its TESTATOR object is not
     strongly reachable (i.e. the TESTATOR object is either unreachable
     or only reachable using paths from the roots that traverse at
     least one weak reference).  Some objects, including symbols, small
     exact integers (fixnums), booleans and characters, are considered
     to be always strongly reachable.

     When the runtime system detects that a will has become executable
     the current computation is interrupted, the will's testator is set
     to `#f' and the will's action procedure is called with the will's
     testator as the sole argument.  Currently only the garbage
     collector detects when wills become executable but this may change
     in future versions of Gambit (for example the compiler could
     perform an analysis to infer will executability at compile time).
     The garbage collector builds a list of all executable wills.
     Shortly after a garbage collection, the action procedures of these
     wills will be called.  The link from the will to the action
     procedure is severed when the action procedure is called.

     Note that the testator object will not be reclaimed during the
     garbage collection that determined executability of the will.  It
     is only when an object is not reachable from the roots that it is
     reclaimed by the garbage collector.

     A remarkable feature of wills is that an action procedure can
     "resurrect" an object.  An action procedure could for example
     assign the testator object to a global variable or create a new
     will with the same testator object.

     For example:

          > (define a (list 123))
          > (set-cdr! a a) ; create a circular list
          > (define b (vector a))
          > (define c #f)
          > (define w
              (let ((obj a))
                (make-will obj
                           (lambda (x) ; x will be eq? to obj
                             (display "executing action procedure")
                             (newline)
                             (set! c x)))))
          > (will? w)
          #t
          > (car (will-testator w))
          123
          > (##gc)
          > (set! a #f)
          > (##gc)
          > (set! b #f)
          > (##gc)
          executing action procedure
          > (will-testator w)
          #f
          > (car c)
          123


11.2.2 Tables
-------------

The following procedures implement the "table" data type.  Tables are
heterogenous structures whose elements are indexed by keys which are
arbitrary objects.  Tables are similar to association lists but are
abstract and the access time for large tables is typically smaller.
Each key contained in the table is bound to a value.  The length of the
table is the number of key/value bindings it contains.  New key/value
bindings can be added to a table, the value bound to a key can be
changed, and existing key/value bindings can be removed.

   The references to the keys can either be all strong or all table-weak
and the references to the values can either be all strong or all
table-weak.  The garbage collector removes key/value bindings from a
table when 1) the key is a table-weak reference and the key is
unreachable or only reachable using paths from the roots that traverse
at least one table-weak reference, or 2) the value is a table-weak
reference and the value is unreachable or only reachable using paths
from the roots that traverse at least one table-weak reference.
Key/value bindings that are removed by the garbage collector are
reclaimed immediately.

   Although there are several possible ways of implementing tables, the
current implementation uses hashing with open-addressing.  This is
space efficient and provides constant-time access.  Hash tables are
automatically resized to maintain the load within specified bounds.
The load is the number of active entries (the length of the table)
divided by the total number of entries in the hash table.

   Tables are parameterized with a key comparison procedure.  By default
the `equal?' procedure is used, but `eq?', `eqv?', `string=?',
`string-ci=?', or a user defined procedure can also be used.  To
support arbitrary key comparison procedures, tables are also
parameterized with a hashing procedure accepting a key as its single
parameter and returning a fixnum result.  The hashing procedure HASH
must be consistent with the key comparison procedure TEST, that is, for
any two keys K1 and K2 in the table, `(TEST K1 K2)' implies `(= (HASH
K1) (HASH K2))'.  A default hashing procedure consistent with the key
comparison procedure is provided by the system.  The default hashing
procedure generally gives good performance when the key comparison
procedure is `eq?', `eqv?', `equal?', `string=?', and `string-ci=?'.
However, for user defined key comparison procedures, the default
hashing procedure always returns 0.  This degrades the performance of
the table to a linear search.

   Tables can be compared for equality using the `equal?' procedure.
Two tables `X' and `Y' are considered equal by `equal?' when they have
the same weakness attributes, the same key comparison procedure, the
same hashing procedure, the same length, and for all the keys `K' in
`X', `(equal?  (table-ref X K) (table-ref Y K))'.

 -- procedure: make-table [`size:' SIZE] [`init:' INIT] [`weak-keys:'
          WEAK-KEYS] [`weak-values:' WEAK-VALUES] [`test:' TEST]
          [`hash:' HASH] [`min-load:' MIN-LOAD] [`max-load:' MAX-LOAD]
     The procedure `make-table' returns a new table.  The optional
     keyword parameters specify various parameters of the table.

     The SIZE parameter is a nonnegative exact integer indicating the
     expected length of the table.  The system uses SIZE to choose an
     appropriate initial size of the hash table so that it does not
     need to be resized too often.

     The INIT parameter indicates a value that is associated to keys
     that are not in the table.  When INIT is not specified, no value
     is associated to keys that are not in the table.

     The WEAK-KEYS and WEAK-VALUES parameters are extended booleans
     indicating respectively whether the keys and values are table-weak
     references (true) or strong references (false).  By default the
     keys and values are strong references.

     The TEST parameter indicates the key comparison procedure.  The
     default key comparison procedure is `equal?'.  The key comparison
     procedures `eq?', `eqv?', `equal?', `string=?', and `string-ci=?'
     are special because the system will use a reasonably good hash
     procedure when none is specified.

     The HASH parameter indicates the hash procedure.  This procedure
     must accept a single key parameter, return a fixnum, and be
     consistent with the key comparison procedure.  When HASH is not
     specified, a default hash procedure is used.  The default hash
     procedure is reasonably good when the key comparison procedure is
     `eq?', `eqv?', `equal?', `string=?', or `string-ci=?'.

     The MIN-LOAD and MAX-LOAD parameters are real numbers that
     indicate the minimum and maximum load of the table respectively.
     The table is resized when adding or deleting a key/value binding
     would bring the table's load outside of this range.  The MIN-LOAD
     parameter must be no less than 0.05 and the MAX-LOAD parameter
     must be no greater than 0.95.  Moreover the difference between
     MIN-LOAD and MAX-LOAD must be at least 0.20.  When MIN-LOAD is not
     specified, the value 0.45 is used.  When MAX-LOAD is not
     specified, the value 0.90 is used.

     For example:

          > (define t (make-table))
          > (table? t)
          #t
          > (table-length t)
          0
          > (table-set! t (list 1 2) 3)
          > (table-set! t (list 4 5) 6)
          > (table-ref t (list 1 2))
          3
          > (table-length t)
          2


 -- procedure: table? OBJ
     The procedure `table?' returns `#t' when OBJ is a table and `#f'
     otherwise.

     For example:

          > (table? (make-table))
          #t
          > (table? 123)
          #f


 -- procedure: table-length TABLE
     The procedure `table-length' returns the number of key/value
     bindings contained in the table TABLE.

     For example:

          > (define t (make-table weak-keys: #t))
          > (define x (list 1 2))
          > (define y (list 3 4))
          > (table-set! t x 111)
          > (table-set! t y 222)
          > (table-length t)
          2
          > (table-set! t x)
          > (table-length t)
          1
          > (##gc)
          > (table-length t)
          1
          > (set! y #f)
          > (##gc)
          > (table-length t)
          0


 -- procedure: table-ref TABLE KEY [DEFAULT]
     The procedure `table-ref' returns the value bound to the object
     KEY in the table TABLE.  When KEY is not bound and DEFAULT is
     specified, DEFAULT is returned.  When DEFAULT is not specified but
     an INIT parameter was specified when TABLE was created, INIT is
     returned.  Otherwise an unbound-table-key-exception object is
     raised.

     For example:

          > (define t1 (make-table init: 999))
          > (table-set! t1 (list 1 2) 3)
          > (table-ref t1 (list 1 2))
          3
          > (table-ref t1 (list 4 5))
          999
          > (table-ref t1 (list 4 5) #f)
          #f
          > (define t2 (make-table))
          > (table-ref t2 (list 4 5))
          *** ERROR IN (console)@7.1 -- Unbound table key
          (table-ref '#<table #2> '(4 5))


 -- procedure: table-set! TABLE KEY [VALUE]
     The procedure `table-set!' binds the object KEY to VALUE in the
     table TABLE.  When VALUE is not specified, if TABLE contains a
     binding for KEY then the binding is removed from TABLE.  The
     procedure `table-set!' returns an unspecified value.

     For example:

          > (define t (make-table))
          > (table-set! t (list 1 2) 3)
          > (table-set! t (list 4 5) 6)
          > (table-set! t (list 4 5))
          > (table-set! t (list 7 8))
          > (table-ref t (list 1 2))
          3
          > (table-ref t (list 4 5))
          *** ERROR IN (console)@7.1 -- Unbound table key
          (table-ref '#<table #2> '(4 5))


 -- procedure: table-search PROC TABLE
     The procedure `table-search' searches the table TABLE for a
     key/value binding for which the two parameter procedure PROC
     returns a non false result.  For each key/value binding visited by
     `table-search' the procedure PROC is called with the key as the
     first parameter and the value as the second parameter.  The
     procedure `table-search' returns the first non false value
     returned by PROC, or `#f' if PROC returned `#f' for all key/value
     bindings in TABLE.

     The order in which the key/value bindings are visited is
     unspecified and may vary from one call of `table-search' to the
     next.  While a call to `table-search' is being performed on TABLE,
     it is an error to call any of the following procedures on TABLE:
     `table-ref', `table-set!', `table-search', `table-for-each',
     `table-copy', `table-merge', `table-merge!', and `table->list'.
     It is also an error to compare with `equal?' (directly or
     indirectly with `member', `assoc', `table-ref', etc.) an object
     that contains TABLE.  All these procedures may cause TABLE to be
     reordered and resized.  This restriction allows a more efficient
     iteration over the key/value bindings.

     For example:

          > (define square (make-table))
          > (table-set! square 2 4)
          > (table-set! square 3 9)
          > (table-search (lambda (k v) (and (odd? k) v)) square)
          9


 -- procedure: table-for-each PROC TABLE
     The procedure `table-for-each' calls the two parameter procedure
     PROC for each key/value binding in the table TABLE.  The procedure
     PROC is called with the key as the first parameter and the value
     as the second parameter.  The procedure `table-for-each' returns
     an unspecified value.

     The order in which the key/value bindings are visited is
     unspecified and may vary from one call of `table-for-each' to the
     next.  While a call to `table-for-each' is being performed on
     TABLE, it is an error to call any of the following procedures on
     TABLE: `table-ref', `table-set!', `table-search',
     `table-for-each', and `table->list'.  It is also an error to
     compare with `equal?' (directly or indirectly with `member',
     `assoc', `table-ref', etc.) an object that contains TABLE.  All
     these procedures may cause TABLE to be reordered and resized.
     This restriction allows a more efficient iteration over the
     key/value bindings.

     For example:

          > (define square (make-table))
          > (table-set! square 2 4)
          > (table-set! square 3 9)
          > (table-for-each (lambda (k v) (write (list k v)) (newline)) square)
          (2 4)
          (3 9)


 -- procedure: table->list TABLE
     The procedure `table->list' returns an association list containing
     the key/value bindings in the table TABLE.  Each key/value binding
     yields a pair whose car field is the key and whose cdr field is
     the value bound to that key.  The order of the bindings in the
     list is unspecified.

     For example:

          > (define square (make-table))
          > (table-set! square 2 4)
          > (table-set! square 3 9)
          > (table->list square)
          ((3 . 9) (2 . 4))


 -- procedure: list->table LIST [`size:' SIZE] [`init:' INIT]
          [`weak-keys:' WEAK-KEYS] [`weak-values:' WEAK-VALUES]
          [`test:' TEST] [`hash:' HASH] [`min-load:' MIN-LOAD]
          [`max-load:' MAX-LOAD]
     The procedure `list->table' returns a new table containing the
     key/value bindings in the association list LIST.  The optional
     keyword parameters specify various parameters of the table and have
     the same meaning as for the `make-table' procedure.

     Each element of LIST is a pair whose car field is a key and whose
     cdr field is the value bound to that key.  If a key appears more
     than once in LIST (tested using the table's key comparison
     procedure) it is the first key/value binding in LIST that has
     precedence.

     For example:

          > (define t (list->table '((b . 2) (a . 1) (c . 3) (a . 4))))
          > (table->list t)
          ((a . 1) (b . 2) (c . 3))


 -- procedure: unbound-table-key-exception? OBJ
 -- procedure: unbound-table-key-exception-procedure EXC
 -- procedure: unbound-table-key-exception-arguments EXC
     Unbound-table-key-exception objects are raised by the procedure
     `table-ref' when the key does not have a binding in the table.
     The parameter EXC must be an unbound-table-key-exception object.

     The procedure `unbound-table-key-exception?' returns `#t' when OBJ
     is a unbound-table-key-exception object and `#f' otherwise.

     The procedure `unbound-table-key-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `unbound-table-key-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define t (make-table))
          > (define (handler exc)
              (if (unbound-table-key-exception? exc)
                  (list (unbound-table-key-exception-procedure exc)
                        (unbound-table-key-exception-arguments exc))
                  'not-unbound-table-key-exception))
          > (with-exception-catcher
              handler
              (lambda () (table-ref t '(1 2))))
          (#<procedure #2 table-ref> (#<table #3> (1 2)))


 -- procedure: table-copy TABLE
     The procedure `table-copy' returns a new table containing the same
     key/value bindings as TABLE and the same table parameters (i.e.
     hash procedure, key comparison procedure, key and value weakness,
     etc).

     For example:

          > (define t (list->table '((b . 2) (a . 1) (c . 3))))
          > (define x (table-copy t))
          > (table-set! t 'b 99)
          > (table->list t)
          ((a . 1) (b . 99) (c . 3))
          > (table->list x)
          ((a . 1) (b . 2) (c . 3))


 -- procedure: table-merge! TABLE1 TABLE2 [TABLE2-TAKES-PRECEDENCE?]
     The procedure `table-merge!' returns TABLE1 after the key/value
     bindings contained in TABLE2 have been added to it.  When a key
     exists both in TABLE1 and TABLE2, then the parameter
     TABLE2-TAKES-PRECEDENCE? indicates which binding will be kept (the
     one in TABLE1 if TABLE2-TAKES-PRECEDENCE?  is false, and the one
     in TABLE2 otherwise).  If TABLE2-TAKES-PRECEDENCE? is not
     specified the binding in TABLE1 is kept.

     For example:

          > (define t1 (list->table '((a . 1) (b . 2) (c . 3))))
          > (define t2 (list->table '((a . 4) (b . 5) (z . 6))))
          > (table->list (table-merge! t1 t2))
          ((a . 1) (b . 2) (c . 3) (z . 6))
          > (define t1 (list->table '((a . 1) (b . 2) (c . 3))))
          > (define t2 (list->table '((a . 4) (b . 5) (z . 6))))
          > (table->list (table-merge! t1 t2 #t))
          ((a . 4) (b . 5) (c . 3) (z . 6))


 -- procedure: table-merge TABLE1 TABLE2 [TABLE2-TAKES-PRECEDENCE?]
     The procedure `table-merge' returns a copy of TABLE1 (created with
     `table-copy') to which the key/value bindings contained in TABLE2
     have been added using `table-merge!'.  When a key exists both in
     TABLE1 and TABLE2, then the parameter TABLE2-TAKES-PRECEDENCE?
     indicates which binding will be kept (the one in TABLE1 if
     TABLE2-TAKES-PRECEDENCE?  is false, and the one in TABLE2
     otherwise).  If TABLE2-TAKES-PRECEDENCE? is not specified the
     binding in TABLE1 is kept.

     For example:

          > (define t1 (list->table '((a . 1) (b . 2) (c . 3))))
          > (define t2 (list->table '((a . 4) (b . 5) (z . 6))))
          > (table->list (table-merge t1 t2))
          ((a . 1) (b . 2) (c . 3) (z . 6))
          > (table->list (table-merge t1 t2 #t))
          ((a . 4) (b . 5) (c . 3) (z . 6))


12 Records
**********

 -- special form: define-structure name field...
     Record data types similar to Pascal records and C `struct' types
     can be defined using the `define-structure' special form.  The
     identifier name specifies the name of the new data type.  The
     structure name is followed by K identifiers naming each field of
     the record.  The `define-structure' expands into a set of
     definitions of the following procedures:

        * `make-name' - A K argument procedure which constructs a new
          record from the value of its K fields.

        * `name?' - A procedure which tests if its single argument is
          of the given record type.

        * `name-field' - For each field, a procedure taking as its
          single argument a value of the given record type and returning
          the content of the corresponding field of the record.

        * `name-field-set!' - For each field, a two argument procedure
          taking as its first argument a value of the given record
          type.  The second argument gets assigned to the corresponding
          field of the record and the void object is returned.


     Record data types have a printed representation that includes the
     name of the type and the name and value of each field.  Record
     data types can not be read by the `read' procedure.

     For example:

          > (define-structure point x y color)
          > (define p (make-point 3 5 'red))
          > p
          #<point #2 x: 3 y: 5 color: red>
          > (point-x p)
          3
          > (point-color p)
          red
          > (point-color-set! p 'black)
          > p
          #<point #2 x: 3 y: 5 color: black>


13 Threads
**********

Gambit supports the execution of multiple Scheme threads.  These
threads are managed entirely by Gambit's runtime and are not related to
the host operating system's threads.  Gambit's runtime does not
currently take advantage of multiprocessors (i.e. at most one thread is
running).

13.1 Introduction
=================

Multithreading is a paradigm that is well suited for building complex
systems such as: servers, GUIs, and high-level operating systems.
Gambit's thread system offers mechanisms for creating threads of
execution and for synchronizing them.  The thread system also supports
features which are useful in a real-time context, such as priorities,
priority inheritance and timeouts.

   The thread system provides the following data types:

   * Thread (a virtual processor which shares object space with all
     other threads)

   * Mutex (a mutual exclusion device, also known as a lock and binary
     semaphore)

   * Condition variable (a set of blocked threads)


13.2 Thread objects
===================

A "running thread" is a thread that is currently executing.  A
"runnable thread" is a thread that is ready to execute or running.  A
thread is "blocked" if it is waiting for a mutex to become unlocked, an
I/O operation to become possible, the end of a "sleep" period, etc.  A
"new thread" is a thread that has been allocated but has not yet been
initialized.  An "initialized thread" is a thread that can be made
runnable.  A new thread becomes runnable when it is started by calling
`thread-start!'.  A "terminated thread" is a thread that can no longer
become runnable (but "deadlocked threads" are not considered
terminated).  The only valid transitions between the thread states are
from new to initialized, from initialized to runnable, between runnable
and blocked, and from any state except new to terminated as indicated in
the following diagram:

                                                 unblock
                               start            <-------
     NEW -------> INITIALIZED -------> RUNNABLE -------> BLOCKED
                            \             |      block  /
                             \            v            /
                              +-----> TERMINATED <----+

   Each thread has a "base priority", which is a real number (where a
higher numerical value means a higher priority), a "priority boost",
which is a nonnegative real number representing the priority increase
applied to a thread when it blocks, and a "quantum", which is a
nonnegative real number representing a duration in seconds.

   Each thread has a "specific field" which can be used in an
application specific way to associate data with the thread (some thread
systems call this "thread local storage").

   Each thread has a "mailbox" which is used for inter-thread
communication.

13.3 Mutex objects
==================

A mutex can be in one of four states: "locked" (either "owned" or "not
owned") and "unlocked" (either "abandoned" or "not abandoned").

   An attempt to lock a mutex only succeeds if the mutex is in an
unlocked state, otherwise the current thread will wait.  A mutex in the
locked/owned state has an associated "owner thread", which by
convention is the thread that is responsible for unlocking the mutex
(this case is typical of critical sections implemented as "lock mutex,
perform operation, unlock mutex").  A mutex in the locked/not-owned
state is not linked to a particular thread.

   A mutex becomes locked when a thread locks it using the
`mutex-lock!' primitive.  A mutex becomes unlocked/abandoned when the
owner of a locked/owned mutex terminates.  A mutex becomes
unlocked/not-abandoned when a thread unlocks it using the
`mutex-unlock!' primitive.

   The mutex primitives do not implement "recursive mutex semantics".
An attempt to lock a mutex that is locked implies that the current
thread waits even if the mutex is owned by the current thread (this can
lead to a deadlock if no other thread unlocks the mutex).

   Each mutex has a "specific field" which can be used in an
application specific way to associate data with the mutex.

13.4 Condition variable objects
===============================

A condition variable represents a set of blocked threads.  These blocked
threads are waiting for a certain condition to become true.  When a
thread modifies some program state that might make the condition true,
the thread unblocks some number of threads (one or all depending on the
primitive used) so they can check if the condition is now true.  This
allows complex forms of interthread synchronization to be expressed more
conveniently than with mutexes alone.

   Each condition variable has a "specific field" which can be used in
an application specific way to associate data with the condition
variable.

13.5 Fairness
=============

In various situations the scheduler must select one thread from a set of
threads (e.g. which thread to run when a running thread blocks or
expires its quantum, which thread to unblock when a mutex becomes
unlocked or a condition variable is signaled).  The constraints on the
selection process determine the scheduler's "fairness".  The selection
depends on the order in which threads become runnable or blocked and on
the "priority" attached to the threads.

   The definition of fairness requires the notion of time ordering,
i.e. "event A occured before event B".  For the purpose of establishing
time ordering, the scheduler uses a clock with a discrete, usually
variable, resolution (a "tick").  Events occuring in a given tick can
be considered to be simultaneous (i.e. if event A occured before event
B in real time, then the scheduler will claim that event A occured
before event B unless both events fall within the same tick, in which
case the scheduler arbitrarily chooses a time ordering).

   Each thread T has three priorities which affect fairness; the "base
priority", the "boosted priority", and the "effective priority".

   * The "base priority" is the value contained in T's "base priority"
     field (which is set with the `thread-base-priority-set!'
     primitive).

   * T's "boosted flag" field contains a boolean that affects T's
     "boosted priority".  When the boosted flag field is false, the
     boosted priority is equal to the base priority, otherwise the
     boosted priority is equal to the base priority plus the value
     contained in T's "priority boost" field (which is set with the
     `thread-priority-boost-set!' primitive).  The boosted flag field is
     set to false when a thread is created, when its quantum expires,
     and when "thread-yield!" is called.  The boosted flag field is set
     to true when a thread blocks.  By carefully choosing the base
     priority and priority boost, relatively to the other threads, it
     is possible to set up an interactive thread so that it has good
     I/O response time without being a CPU hog when it performs long
     computations.

   * The "effective priority" is equal to the maximum of T's boosted
     priority and the effective priority of all the threads that are
     blocked on a mutex owned by T.  This "priority inheritance" avoids
     priority inversion problems that would prevent a high priority
     thread blocked at the entry of a critical section to progress
     because a low priority thread inside the critical section is
     preempted for an arbitrary long time by a medium priority thread.


   Let P(T) be the effective priority of thread T and let R(T) be the
most recent time when one of the following events occurred for thread
T, thus making it runnable: T was started by calling `thread-start!', T
called `thread-yield!', T expired its quantum, or T became unblocked.
Let the relation NL(T1,T2), "T1 no later than T2", be true if
P(T1)<P(T2) or P(T1)=P(T2) and R(T1)>R(T2), and false otherwise.  The
scheduler will schedule the execution of threads in such a way that
whenever there is at least one runnable thread, 1) within a finite time
at least one thread will be running, and 2) there is never a pair of
runnable threads T1 and T2 for which NL(T1,T2) is true and T1 is not
running and T2 is running.

   A thread T expires its quantum when an amount of time equal to T's
quantum has elapsed since T entered the running state and T did not
block, terminate or call `thread-yield!'.  At that point T exits the
running state to allow other threads to run.  A thread's quantum is
thus an indication of the rate of progress of the thread relative to
the other threads of the same priority.  Moreover, the resolution of
the timer measuring the running time may cause a certain deviation from
the quantum, so a thread's quantum should only be viewed as an
approximation of the time it can run before yielding to another thread.

   Threads blocked on a given mutex or condition variable will unblock
in an order which is consistent with decreasing priority and increasing
blocking time (i.e. the highest priority thread unblocks first, and
among equal priority threads the one that blocked first unblocks first).

13.6 Memory coherency
=====================

Read and write operations on the store (such as reading and writing a
variable, an element of a vector or a string) are not atomic.  It is an
error for a thread to write a location in the store while some other
thread reads or writes that same location.  It is the responsibility of
the application to avoid write/read and write/write races through
appropriate uses of the synchronization primitives.

   Concurrent reads and writes to ports are allowed.  It is the
responsibility of the implementation to serialize accesses to a given
port using the appropriate synchronization primitives.

13.7 Timeouts
=============

All synchronization primitives which take a timeout parameter accept
three types of values as a timeout, with the following meaning:

   * a time object represents an absolute point in time

   * an exact or inexact real number represents a relative time in
     seconds from the moment the primitive was called

   * `#f' means that there is no timeout


   When a timeout denotes the current time or a time in the past, the
synchronization primitive claims that the timeout has been reached only
after the other synchronization conditions have been checked.  Moreover
the thread remains running (it does not enter the blocked state).  For
example, `(mutex-lock! m 0)' will lock mutex `m' and return `#t' if `m'
is currently unlocked, otherwise `#f' is returned because the timeout
is reached.

13.8 Primordial thread
======================

The execution of a program is initially under the control of a single
thread known as the "primordial thread".  The primordial thread has an
unspecified base priority, priority boost, boosted flag, quantum, name,
specific field, dynamic environment, `dynamic-wind' stack, and
exception-handler.  All threads are terminated when the primordial
thread terminates (normally or not).

13.9 Procedures
===============

 -- procedure: current-thread
     This procedure returns the current thread.  For example:

          > (current-thread)
          #<thread #1 primordial>
          > (eq? (current-thread) (current-thread))
          #t


 -- procedure: thread? OBJ
     This procedure returns `#t' when OBJ is a thread object and `#f'
     otherwise.

     For example:

          > (thread? (current-thread))
          #t
          > (thread? 'foo)
          #f


 -- procedure: make-thread THUNK [NAME [THREAD-GROUP]]
 -- procedure: make-root-thread THUNK [NAME [THREAD-GROUP [INPUT-PORT
          [OUTPUT-PORT]]]]
     The `make-thread' procedure creates and returns an initialized
     thread.  This thread is not automatically made runnable (the
     procedure `thread-start!' must be used for this).  A thread has the
     following fields: base priority, priority boost, boosted flag,
     quantum, name, specific, end-result, end-exception, and a list of
     locked/owned mutexes it owns.  The thread's execution consists of a
     call to THUNK with the "initial continuation".  This continuation
     causes the (then) current thread to store the result in its
     end-result field, abandon all mutexes it owns, and finally
     terminate.  The `dynamic-wind' stack of the initial continuation
     is empty.  The optional NAME is an arbitrary Scheme object which
     identifies the thread (useful for debugging); it defaults to an
     unspecified value.  The specific field is set to an unspecified
     value.  The optional THREAD-GROUP indicates which thread group this
     thread belongs to; it defaults to the thread group of the current
     thread.  The base priority, priority boost, and quantum of the
     thread are set to the same value as the current thread and the
     boosted flag is set to false.  The thread's mailbox is initially
     empty.  The thread inherits the dynamic environment from the
     current thread. Moreover, in this dynamic environment the
     exception-handler is bound to the "initial exception-handler"
     which is a unary procedure which causes the (then) current thread
     to store in its end-exception field an uncaught-exception object
     whose "reason" is the argument of the handler, abandon all mutexes
     it owns, and finally terminate.

     The `make-root-thread' procedure behaves like the `make-thread'
     procedure except the created thread does not inherit the dynamic
     environment from the current thread and the base priority is set
     to 0, the priority boost is set to 1.0e-6, and the quantum is set
     to 0.02.  The dynamic environment of the thread has the global
     bindings of the parameter objects, except `current-input-port'
     which is bound to INPUT-PORT, `current-output-port' which is bound
     to OUTPUT-PORT, and `current-directory' which is bound to the
     initial current working directory of the current process.  If
     INPUT-PORT is not specified it defaults to a port corresponding to
     the standard input (`stdin').  If OUTPUT-PORT is not specified it
     defaults to a port corresponding to the standard output (`stdout').

     For example:

          > (make-thread (lambda () (write 'hello)))
          #<thread #2>
          > (make-root-thread (lambda () (write 'world)) 'a-name)
          #<thread #3 a-name>


 -- procedure: thread-name THREAD
     This procedure returns the name of the THREAD.  For example:

          > (thread-name (make-thread (lambda () #f) 'foo))
          foo


 -- procedure: thread-specific THREAD
 -- procedure: thread-specific-set! THREAD OBJ
     The `thread-specific' procedure returns the content of the
     THREAD's specific field.

     The `thread-specific-set!' procedure stores OBJ into the THREAD's
     specific field and returns an unspecified value.

     For example:

          > (thread-specific-set! (current-thread) "hello")
          > (thread-specific (current-thread))
          "hello"


 -- procedure: thread-base-priority THREAD
 -- procedure: thread-base-priority-set! THREAD PRIORITY
     The procedure `thread-base-priority' returns a real number which
     corresponds to the base priority of the THREAD.

     The procedure `thread-base-priority-set!' changes the base
     priority of the THREAD to PRIORITY and returns an unspecified
     value.  The PRIORITY must be a real number.

     For example:

          > (thread-base-priority-set! (current-thread) 12.3)
          > (thread-base-priority (current-thread))
          12.3


 -- procedure: thread-priority-boost THREAD
 -- procedure: thread-priority-boost-set! THREAD PRIORITY-BOOST
     The procedure `thread-priority-boost' returns a real number which
     corresponds to the priority boost of the THREAD.

     The procedure `thread-priority-boost-set!' changes the priority
     boost of the THREAD to PRIORITY-BOOST and returns an unspecified
     value.  The PRIORITY-BOOST must be a nonnegative real.

     For example:

          > (thread-priority-boost-set! (current-thread) 2.5)
          > (thread-priority-boost (current-thread))
          2.5


 -- procedure: thread-quantum THREAD
 -- procedure: thread-quantum-set! THREAD QUANTUM
     The procedure `thread-quantum' returns a real number which
     corresponds to the quantum of the THREAD.

     The procedure `thread-quantum-set!' changes the quantum of the
     THREAD to QUANTUM and returns an unspecified value.  The QUANTUM
     must be a nonnegative real.  A value of zero selects the smallest
     quantum supported by the implementation.

     For example:

          > (thread-quantum-set! (current-thread) 1.5)
          > (thread-quantum (current-thread))
          1.5
          > (thread-quantum-set! (current-thread) 0)
          > (thread-quantum (current-thread))
          0.


 -- procedure: thread-start! THREAD
     This procedure makes THREAD runnable and returns the THREAD.  The
     THREAD must be an initialized thread.

     For example:

          > (let ((t (thread-start! (make-thread (lambda () (write 'a))))))
              (write 'b)
              (thread-join! t))
          ab> or ba>

     NOTE: It is useful to separate thread creation and thread
     activation to avoid the race condition that would occur if the
     created thread tries to examine a table in which the current
     thread stores the created thread.  See the last example of the
     `thread-terminate!' procedure which contains mutually recursive
     threads.


 -- procedure: thread-yield!
     This procedure causes the current thread to exit the running state
     as if its quantum had expired and returns an unspecified value.

     For example:

          ; a busy loop that avoids being too wasteful of the CPU

          (let loop ()
            (if (mutex-lock! m 0) ; try to lock m but don't block
                (begin
                  (display "locked mutex m")
                  (mutex-unlock! m))
                (begin
                  (do-something-else)
                  (thread-yield!) ; relinquish rest of quantum
                  (loop))))


 -- procedure: thread-sleep! TIMEOUT
     This procedure causes the current thread to wait until the timeout
     is reached and returns an unspecified value.  This blocks the
     thread only if TIMEOUT represents a point in the future.  It is an
     error for TIMEOUT to be `#f'.

     For example:

          ; a clock with a gradual drift:

          (let loop ((x 1))
            (thread-sleep! 1)
            (write x)
            (loop (+ x 1)))

          ; a clock with no drift:

          (let ((start (time->seconds (current-time)))
            (let loop ((x 1))
              (thread-sleep! (seconds->time (+ x start)))
              (write x)
              (loop (+ x 1))))


 -- procedure: thread-terminate! THREAD
     This procedure causes an abnormal termination of the THREAD.  If
     the THREAD is not already terminated, all mutexes owned by the
     THREAD become unlocked/abandoned and a terminated-thread-exception
     object is stored in the THREAD's end-exception field.  If THREAD
     is the current thread, `thread-terminate!' does not return.
     Otherwise `thread-terminate!' returns an unspecified value; the
     termination of the THREAD will occur at some point between the
     calling of `thread-terminate!'  and a finite time in the future
     (an explicit thread synchronization is needed to detect
     termination, see `thread-join!').

     For example:

          (define (amb thunk1 thunk2)
            (let ((result #f)
                  (result-mutex (make-mutex))
                  (done-mutex (make-mutex)))
              (letrec ((child1
                        (make-thread
                          (lambda ()
                            (let ((x (thunk1)))
                              (mutex-lock! result-mutex #f #f)
                              (set! result x)
                              (thread-terminate! child2)
                              (mutex-unlock! done-mutex)))))
                       (child2
                        (make-thread
                          (lambda ()
                            (let ((x (thunk2)))
                              (mutex-lock! result-mutex #f #f)
                              (set! result x)
                              (thread-terminate! child1)
                              (mutex-unlock! done-mutex))))))
                (mutex-lock! done-mutex #f #f)
                (thread-start! child1)
                (thread-start! child2)
                (mutex-lock! done-mutex #f #f)
                result)))

     NOTE: This operation must be used carefully because it terminates a
     thread abruptly and it is impossible for that thread to perform any
     kind of cleanup.  This may be a problem if the thread is in the
     middle of a critical section where some structure has been put in
     an inconsistent state.  However, another thread attempting to
     enter this critical section will raise an
     abandoned-mutex-exception object because the mutex is
     unlocked/abandoned.  This helps avoid observing an inconsistent
     state.  Clean termination can be obtained by polling, as shown in
     the example below.

     For example:

          (define (spawn thunk)
            (let ((t (make-thread thunk)))
              (thread-specific-set! t #t)
              (thread-start! t)
              t))

          (define (stop! thread)
            (thread-specific-set! thread #f)
            (thread-join! thread))

          (define (keep-going?)
            (thread-specific (current-thread)))

          (define count!
            (let ((m (make-mutex))
                  (i 0))
              (lambda ()
                (mutex-lock! m)
                (let ((x (+ i 1)))
                  (set! i x)
                  (mutex-unlock! m)
                  x))))

          (define (increment-forever!)
            (let loop () (count!) (if (keep-going?) (loop))))

          (let ((t1 (spawn increment-forever!))
                (t2 (spawn increment-forever!)))
            (thread-sleep! 1)
            (stop! t1)
            (stop! t2)
            (count!))  ==>  377290


 -- procedure: thread-join! thread [TIMEOUT [TIMEOUT-VAL]]
     This procedure causes the current thread to wait until the THREAD
     terminates (normally or not) or until the timeout is reached if
     TIMEOUT is supplied.  If the timeout is reached, THREAD-JOIN!
     returns TIMEOUT-VAL if it is supplied, otherwise a
     join-timeout-exception object is raised.  If the THREAD terminated
     normally, the content of the end-result field is returned,
     otherwise the content of the end-exception field is raised.

     For example:

          (let ((t (thread-start! (make-thread (lambda () (expt 2 100))))))
            (do-something-else)
            (thread-join! t))  ==>  1267650600228229401496703205376

          (let ((t (thread-start! (make-thread (lambda () (raise 123))))))
            (do-something-else)
            (with-exception-handler
              (lambda (exc)
                (if (uncaught-exception? exc)
                    (* 10 (uncaught-exception-reason exc))
                    99999))
              (lambda ()
                (+ 1 (thread-join! t)))))  ==>  1231

          (define thread-alive?
            (let ((unique (list 'unique)))
              (lambda (thread)
                ; Note: this procedure raises an exception if
                ; the thread terminated abnormally.
                (eq? (thread-join! thread 0 unique) unique))))

          (define (wait-for-termination! thread)
            (let ((eh (current-exception-handler)))
              (with-exception-handler
                (lambda (exc)
                  (if (not (or (terminated-thread-exception? exc)
                               (uncaught-exception? exc)))
                      (eh exc))) ; unexpected exceptions are handled by eh
                (lambda ()
                  ; The following call to thread-join! will wait until the
                  ; thread terminates.  If the thread terminated normally
                  ; thread-join! will return normally.  If the thread
                  ; terminated abnormally then one of these two exception
                  ; objects is raised by thread-join!:
                  ;   - terminated-thread-exception object
                  ;   - uncaught-exception object
                  (thread-join! thread)
                  #f)))) ; ignore result of thread-join!


 -- procedure: thread-send THREAD MSG
     Each thread has a mailbox which stores messages delivered to the
     thread in the order delivered.

     The procedure `thread-send' adds the message MSG at the end of the
     mailbox of thread THREAD and returns an unspecified value.

     For example:

          > (thread-send (current-thread) 111)
          > (thread-send (current-thread) 222)
          > (thread-receive)
          111
          > (thread-receive)
          222


 -- procedure: thread-receive [TIMEOUT [DEFAULT]]
 -- procedure: thread-mailbox-next [TIMEOUT [DEFAULT]]
 -- procedure: thread-mailbox-rewind
 -- procedure: thread-mailbox-extract-and-rewind
     To allow a thread to examine the messages in its mailbox without
     removing them from the mailbox, each thread has a "mailbox cursor"
     which normally points to the last message accessed in the mailbox.
     When a mailbox cursor is rewound using the procedure
     `thread-mailbox-rewind' or `thread-mailbox-extract-and-rewind' or
     `thread-receive', the cursor does not point to a message, but the
     next call to `thread-receive' and `thread-mailbox-next' will set
     the cursor to the oldest message in the mailbox.

     The procedure `thread-receive' advances the mailbox cursor of the
     current thread to the next message, removes that message from the
     mailbox, rewinds the mailbox cursor, and returns the message.  When
     TIMEOUT is not specified, the current thread will wait until a
     message is available in the mailbox.  When TIMEOUT is specified
     and DEFAULT is not specified, a mailbox-receive-timeout-exception
     object is raised if the timeout is reached before a message is
     available.  When TIMEOUT is specified and DEFAULT is specified,
     DEFAULT is returned if the timeout is reached before a message is
     available.

     The procedure `thread-mailbox-next' behaves like `thread-receive'
     except that the message remains in the mailbox and the mailbox
     cursor is not rewound.

     The procedures `thread-mailbox-rewind' or
     `thread-mailbox-extract-and-rewind' rewind the mailbox cursor of
     the current thread so that the next call to `thread-mailbox-next'
     and `thread-receive' will access the oldest message in the
     mailbox.  Additionally the procedure
     `thread-mailbox-extract-and-rewind' will remove from the mailbox
     the message most recently accessed by a call to
     `thread-mailbox-next'.  When `thread-mailbox-next' has not been
     called since the last call to `thread-receive' or
     `thread-mailbox-rewind' or `thread-mailbox-extract-and-rewind', a
     call to `thread-mailbox-extract-and-rewind' only resets the mailbox
     cursor (no message is removed).

     For example:

          > (thread-send (current-thread) 111)
          > (thread-receive 1 999)
          111
          > (thread-send (current-thread) 222)
          > (thread-send (current-thread) 333)
          > (thread-mailbox-next 1 999)
          222
          > (thread-mailbox-next 1 999)
          333
          > (thread-mailbox-next 1 999)
          999
          > (thread-mailbox-extract-and-rewind)
          > (thread-receive 1 999)
          222
          > (thread-receive 1 999)
          999


 -- procedure: mailbox-receive-timeout-exception? OBJ
 -- procedure: mailbox-receive-timeout-exception-procedure EXC
 -- procedure: mailbox-receive-timeout-exception-arguments EXC
     Mailbox-receive-timeout-exception objects are raised by the
     procedures `thread-receive' and `thread-mailbox-next' when a
     timeout expires before a message is available and no default value
     is specified.  The parameter EXC must be a
     mailbox-receive-timeout-exception object.

     The procedure `mailbox-receive-timeout-exception?' returns `#t'
     when OBJ is a mailbox-receive-timeout-exception object and `#f'
     otherwise.

     The procedure `mailbox-receive-timeout-exception-procedure'
     returns the procedure that raised EXC.

     The procedure `mailbox-receive-timeout-exception-arguments'
     returns the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (mailbox-receive-timeout-exception? exc)
                  (list (mailbox-receive-timeout-exception-procedure exc)
                        (mailbox-receive-timeout-exception-arguments exc))
                  'not-mailbox-receive-timeout-exception))
          > (with-exception-catcher
              handler
              (lambda () (thread-receive 1)))
          (#<procedure #2 thread-receive> (1))


 -- procedure: mutex? OBJ
     This procedure returns `#t' when OBJ is a mutex object and `#f'
     otherwise.

     For example:

          > (mutex? (make-mutex))
          #t
          > (mutex? 'foo)
          #f


 -- procedure: make-mutex [NAME]
     This procedure returns a new mutex in the unlocked/not-abandoned
     state.  The optional NAME is an arbitrary Scheme object which
     identifies the mutex (useful for debugging); it defaults to an
     unspecified value.  The mutex's specific field is set to an
     unspecified value.

     For example:

          > (make-mutex)
          #<mutex #2>
          > (make-mutex 'foo)
          #<mutex #3 foo>


 -- procedure: mutex-name MUTEX
     Returns the name of the MUTEX.  For example:

          > (mutex-name (make-mutex 'foo))
          foo


 -- procedure: mutex-specific MUTEX
 -- procedure: mutex-specific-set! MUTEX OBJ
     The `mutex-specific' procedure returns the content of the MUTEX's
     specific field.

     The `mutex-specific-set!' procedure stores OBJ into the MUTEX's
     specific field and returns an unspecified value.

     For example:

          > (define m (make-mutex))
          > (mutex-specific-set! m "hello")
          > (mutex-specific m)
          "hello"
          > (define (mutex-lock-recursively! mutex)
              (if (eq? (mutex-state mutex) (current-thread))
                  (let ((n (mutex-specific mutex)))
                    (mutex-specific-set! mutex (+ n 1)))
                  (begin
                    (mutex-lock! mutex)
                    (mutex-specific-set! mutex 0))))
          > (define (mutex-unlock-recursively! mutex)
              (let ((n (mutex-specific mutex)))
                (if (= n 0)
                    (mutex-unlock! mutex)
                    (mutex-specific-set! mutex (- n 1)))))
          > (mutex-lock-recursively! m)
          > (mutex-lock-recursively! m)
          > (mutex-lock-recursively! m)
          > (mutex-specific m)
          2


 -- procedure: mutex-state MUTEX
     Thos procedure returns information about the state of the MUTEX.
     The possible results are:

        * thread T: the MUTEX is in the locked/owned state and thread T
          is the owner of the MUTEX

        * symbol `not-owned': the MUTEX is in the locked/not-owned state

        * symbol `abandoned': the MUTEX is in the unlocked/abandoned
          state

        * symbol `not-abandoned': the MUTEX is in the
          unlocked/not-abandoned state


     For example:

          (mutex-state (make-mutex))  ==>  not-abandoned

          (define (thread-alive? thread)
            (let ((mutex (make-mutex)))
              (mutex-lock! mutex #f thread)
              (let ((state (mutex-state mutex)))
                (mutex-unlock! mutex) ; avoid space leak
                (eq? state thread))))


 -- procedure: mutex-lock! MUTEX [TIMEOUT [THREAD]]
     This procedure locks MUTEX.  If the MUTEX is currently locked, the
     current thread waits until the MUTEX is unlocked, or until the
     timeout is reached if TIMEOUT is supplied.  If the timeout is
     reached, `mutex-lock!' returns `#f'.  Otherwise, the state of the
     MUTEX is changed as follows:

        * if THREAD is `#f' the MUTEX becomes locked/not-owned,

        * otherwise, let T be THREAD (or the current thread if THREAD
          is not supplied),

             * if T is terminated the MUTEX becomes unlocked/abandoned,

             * otherwise MUTEX becomes locked/owned with T as the owner.



     After changing the state of the MUTEX, an
     abandoned-mutex-exception object is raised if the MUTEX was
     unlocked/abandoned before the state change, otherwise
     `mutex-lock!' returns `#t'.  It is not an error if the MUTEX is
     owned by the current thread (but the current thread will have to
     wait).

     For example:

          ; an implementation of a mailbox object of depth one; this
          ; implementation does not behave well in the presence of forced
          ; thread terminations using thread-terminate! (deadlock can occur
          ; if a thread is terminated in the middle of a put! or get! operation)

          (define (make-empty-mailbox)
            (let ((put-mutex (make-mutex)) ; allow put! operation
                  (get-mutex (make-mutex))
                  (cell #f))

              (define (put! obj)
                (mutex-lock! put-mutex #f #f) ; prevent put! operation
                (set! cell obj)
                (mutex-unlock! get-mutex)) ; allow get! operation

              (define (get!)
                (mutex-lock! get-mutex #f #f) ; wait until object in mailbox
                (let ((result cell))
                  (set! cell #f) ; prevent space leaks
                  (mutex-unlock! put-mutex) ; allow put! operation
                  result))

              (mutex-lock! get-mutex #f #f) ; prevent get! operation

              (lambda (msg)
                (case msg
                  ((put!) put!)
                  ((get!) get!)
                  (else (error "unknown message"))))))

          (define (mailbox-put! m obj) ((m 'put!) obj))
          (define (mailbox-get! m) ((m 'get!)))

          ; an alternate implementation of thread-sleep!

          (define (sleep! timeout)
            (let ((m (make-mutex)))
              (mutex-lock! m #f #f)
              (mutex-lock! m timeout #f)))

          ; a procedure that waits for one of two mutexes to unlock

          (define (lock-one-of! mutex1 mutex2)
            ; this procedure assumes that neither mutex1 or mutex2
            ; are owned by the current thread
            (let ((ct (current-thread))
                  (done-mutex (make-mutex)))
              (mutex-lock! done-mutex #f #f)
              (let ((t1 (thread-start!
                         (make-thread
                          (lambda ()
                            (mutex-lock! mutex1 #f ct)
                            (mutex-unlock! done-mutex)))))
                    (t2 (thread-start!
                         (make-thread
                          (lambda ()
                            (mutex-lock! mutex2 #f ct)
                            (mutex-unlock! done-mutex))))))
                (mutex-lock! done-mutex #f #f)
                (thread-terminate! t1)
                (thread-terminate! t2)
                (if (eq? (mutex-state mutex1) ct)
                    (begin
                      (if (eq? (mutex-state mutex2) ct)
                          (mutex-unlock! mutex2)) ; don't lock both
                      mutex1)
                    mutex2))))


 -- procedure: mutex-unlock! MUTEX [CONDITION-VARIABLE [TIMEOUT]]
     This procedure unlocks the MUTEX by making it
     unlocked/not-abandoned.  It is not an error to unlock an unlocked
     mutex and a mutex that is owned by any thread.  If
     CONDITION-VARIABLE is supplied, the current thread is blocked and
     added to the CONDITION-VARIABLE before unlocking MUTEX; the thread
     can unblock at any time but no later than when an appropriate call
     to `condition-variable-signal!' or `condition-variable-broadcast!'
     is performed (see below), and no later than the timeout (if
     TIMEOUT is supplied).  If there are threads waiting to lock this
     MUTEX, the scheduler selects a thread, the mutex becomes
     locked/owned or locked/not-owned, and the thread is unblocked.
     `mutex-unlock!' returns `#f' when the timeout is reached,
     otherwise it returns `#t'.

     NOTE: The reason the thread can unblock at any time (when
     CONDITION-VARIABLE is supplied) is that the scheduler, when it
     detects a serious problem such as a deadlock, must interrupt one of
     the blocked threads (such as the primordial thread) so that it can
     perform some appropriate action.  After a thread blocked on a
     condition-variable has handled such an interrupt it would be wrong
     for the scheduler to return the thread to the blocked state,
     because any calls to `condition-variable-broadcast!' during the
     interrupt will have gone unnoticed.  It is necessary for the
     thread to remain runnable and return from the call to
     `mutex-unlock!' with a result of `#t'.

     NOTE: `mutex-unlock!' is related to the "wait" operation on
     condition variables available in other thread systems.  The main
     difference is that "wait" automatically locks MUTEX just after the
     thread is unblocked.  This operation is not performed by
     `mutex-unlock!' and so must be done by an explicit call to
     `mutex-lock!'.  This has the advantages that a different timeout
     and exception-handler can be specified on the `mutex-lock!' and
     `mutex-unlock!' and the location of all the mutex operations is
     clearly apparent.

     For example:

          (let loop ()
            (mutex-lock! m)
            (if (condition-is-true?)
                (begin
                  (do-something-when-condition-is-true)
                  (mutex-unlock! m))
                (begin
                  (mutex-unlock! m cv)
                  (loop))))


 -- procedure: condition-variable? OBJ
     This procedure returns `#t' when OBJ is a condition-variable
     object and `#f' otherwise.

     For example:

          > (condition-variable? (make-condition-variable))
          #t
          > (condition-variable? 'foo)
          #f


 -- procedure: make-condition-variable [NAME]
     This procedure returns a new empty condition variable.  The
     optional NAME is an arbitrary Scheme object which identifies the
     condition variable (useful for debugging); it defaults to an
     unspecified value.  The condition variable's specific field is set
     to an unspecified value.

     For example:

          > (make-condition-variable)
          #<condition-variable #2>


 -- procedure: condition-variable-name CONDITION-VARIABLE
     This procedure returns the name of the CONDITION-VARIABLE.  For
     example:

          > (condition-variable-name (make-condition-variable 'foo))
          foo


 -- procedure: condition-variable-specific CONDITION-VARIABLE
 -- procedure: condition-variable-specific-set! CONDITION-VARIABLE OBJ
     The `condition-variable-specific' procedure returns the content of
     the CONDITION-VARIABLE's specific field.

     The `condition-variable-specific-set!' procedure stores OBJ into
     the CONDITION-VARIABLE's specific field and returns an unspecified
     value.

     For example:

          > (define cv (make-condition-variable))
          > (condition-variable-specific-set! cv "hello")
          > (condition-variable-specific cv)
          "hello"


 -- procedure: condition-variable-signal! CONDITION-VARIABLE
     This procedure unblocks a thread blocked on the CONDITION-VARIABLE
     (if there is at least one) and returns an unspecified value.

     For example:

          ; an implementation of a mailbox object of depth one; this
          ; implementation behaves gracefully when threads are forcibly
          ; terminated using thread-terminate! (an abandoned-mutex-exception
          ; object will be raised when a put! or get! operation is attempted
          ; after a thread is terminated in the middle of a put! or get!
          ; operation)

          (define (make-empty-mailbox)
            (let ((mutex (make-mutex))
                  (put-condvar (make-condition-variable))
                  (get-condvar (make-condition-variable))
                  (full? #f)
                  (cell #f))

              (define (put! obj)
                (mutex-lock! mutex)
                (if full?
                    (begin
                      (mutex-unlock! mutex put-condvar)
                      (put! obj))
                    (begin
                      (set! cell obj)
                      (set! full? #t)
                      (condition-variable-signal! get-condvar)
                      (mutex-unlock! mutex))))

              (define (get!)
                (mutex-lock! mutex)
                (if (not full?)
                    (begin
                      (mutex-unlock! mutex get-condvar)
                      (get!))
                    (let ((result cell))
                      (set! cell #f) ; avoid space leaks
                      (set! full? #f)
                      (condition-variable-signal! put-condvar)
                      (mutex-unlock! mutex)
                      result)))

              (lambda (msg)
                (case msg
                  ((put!) put!)
                  ((get!) get!)
                  (else (error "unknown message"))))))

          (define (mailbox-put! m obj) ((m 'put!) obj))
          (define (mailbox-get! m) ((m 'get!)))


 -- procedure: condition-variable-broadcast! CONDITION-VARIABLE
     This procedure unblocks all the thread blocked on the
     CONDITION-VARIABLE and returns an unspecified value.

     For example:

          (define (make-semaphore n)
            (vector n (make-mutex) (make-condition-variable)))

          (define (semaphore-wait! sema)
            (mutex-lock! (vector-ref sema 1))
            (let ((n (vector-ref sema 0)))
              (if (> n 0)
                  (begin
                    (vector-set! sema 0 (- n 1))
                    (mutex-unlock! (vector-ref sema 1)))
                  (begin
                    (mutex-unlock! (vector-ref sema 1) (vector-ref sema 2))
                    (semaphore-wait! sema))))

          (define (semaphore-signal-by! sema increment)
            (mutex-lock! (vector-ref sema 1))
            (let ((n (+ (vector-ref sema 0) increment)))
              (vector-set! sema 0 n)
              (if (> n 0)
                  (condition-variable-broadcast! (vector-ref sema 2)))
              (mutex-unlock! (vector-ref sema 1))))


14 Dynamic environment
**********************

The "dynamic environment" is the structure which allows the system to
find the value returned by the standard procedures `current-input-port'
and `current-output-port'.  The standard procedures
`with-input-from-file' and `with-output-to-file' extend the dynamic
environment to produce a new dynamic environment which is in effect for
the dynamic extent of the call to the thunk passed as their last
argument.  These procedures are essentially special purpose dynamic
binding operations on hidden dynamic variables (one for
`current-input-port' and one for `current-output-port').  Gambit
generalizes this dynamic binding mechanism to allow the user to
introduce new dynamic variables, called "parameter objects", and
dynamically bind them.  The parameter objects implemented by Gambit are
compatible with the specification of the "Parameter objects SRFI" (SRFI
39).

   One important issue is the relationship between the dynamic
environments of the parent and child threads when a thread is created.
Each thread has its own dynamic environment that is accessed when
looking up the value bound to a parameter object by that thread.  When
a thread's dynamic environment is extended it does not affect the
dynamic environment of other threads.  When a thread is created it is
given a dynamic environment whose bindings are inherited from the
parent thread.  In this inherited dynamic environment the parameter
objects are bound to the same cells as the parent's dynamic environment
(in other words an assignment of a new value to a parameter object is
visible in the other thread).

   Another important issue is the interaction between the
`dynamic-wind' procedure and dynamic environments.  When a thread
creates a continuation, the thread's dynamic environment and the
`dynamic-wind' stack are saved within the continuation (an alternate
but equivalent point of view is that the `dynamic-wind' stack is part
of the dynamic environment).  When this continuation is invoked the
required `dynamic-wind' before and after thunks are called and the
saved dynamic environment is reinstated as the dynamic environment of
the current thread.  During the call to each required `dynamic-wind'
before and after thunk, the dynamic environment and the `dynamic-wind'
stack in effect when the corresponding `dynamic-wind' was executed are
reinstated.  Note that this specification precisely defines the
semantics of calling `call-with-current-continuation' or invoking a
continuation within a before or after thunk.  The semantics are well
defined even when a continuation created by another thread is invoked.
Below is an example exercising the subtleties of this semantics.

     (with-output-to-file
      "foo"
      (lambda ()
        (let ((k (call-with-current-continuation
                  (lambda (exit)
                    (with-output-to-file
                     "bar"
                     (lambda ()
                       (dynamic-wind
                        (lambda ()
                          (write '(b1))
                          (force-output))
                        (lambda ()
                          (let ((x (call-with-current-continuation
                                    (lambda (cont) (exit cont)))))
                            (write '(t1))
                            (force-output)
                            x))
                        (lambda ()
                          (write '(a1))
                          (force-output)))))))))
          (if k
              (dynamic-wind
               (lambda ()
                 (write '(b2))
                 (force-output))
               (lambda ()
                 (with-output-to-file
                  "baz"
                  (lambda ()
                    (write '(t2))
                    (force-output)
                    ; go back inside (with-output-to-file "bar" ...)
                    (k #f))))
               (lambda ()
                 (write '(a2))
                 (force-output)))))))

   The following actions will occur when this code is executed:
`(b1)(a1)' is written to "bar", `(b2)' is then written to "foo", `(t2)'
is then written to "baz", `(a2)' is then written to "foo", and finally
`(b1)(t1)(a1)' is written to "bar".

 -- procedure: make-parameter OBJ [FILTER]
     The dynamic environment is composed of two parts: the "local
     dynamic environment" and the "global dynamic environment".  There
     is a single global dynamic environment, and it is used to lookup
     parameter objects that can't be found in the local dynamic
     environment.

     The `make-parameter' procedure returns a new "parameter object".
     The FILTER argument is a one argument conversion procedure.  If it
     is not specified, FILTER defaults to the identity function.

     The global dynamic environment is updated to associate the
     parameter object to a new cell.  The initial content of the cell
     is the result of applying the conversion procedure to OBJ.

     A parameter object is a procedure which accepts zero or one
     argument.  The cell bound to a particular parameter object in the
     dynamic environment is accessed by calling the parameter object.
     When no argument is passed, the content of the cell is returned.
     When one argument is passed the content of the cell is updated
     with the result of applying the parameter object's conversion
     procedure to the argument.  Note that the conversion procedure can
     be used for guaranteeing the type of the parameter object's
     binding and/or to perform some conversion of the value.

     For example:

          > (define radix (make-parameter 10))
          > (radix)
          10
          > (radix 2)
          > (radix)
          2
          > (define prompt
              (make-parameter
                123
                (lambda (x)
                  (if (string? x)
                      x
                      (object->string x)))))
          > (prompt)
          "123"
          > (prompt "$")
          > (prompt)
          "$"
          > (define write-shared
              (make-parameter
                #f
                (lambda (x)
                  (if (boolean? x)
                      x
                      (error "only booleans are accepted by write-shared")))))
          > (write-shared 123)
          *** ERROR IN ##make-parameter -- only booleans are accepted by write-shared


 -- special form: parameterize ((procedure value)...) body
     The `parameterize' form, evaluates all procedure and value
     expressions in an unspecified order.  All the procedure
     expressions must evaluate to procedures, either parameter objects
     or procedures accepting zero and one argument.  Then, for each
     procedure p and in an unspecified order:

        * If p is a parameter object a new cell is created,
          initialized, and bound to the parameter object in the local
          dynamic environment.  The value contained in the cell is the
          result of applying the parameter object's conversion
          procedure to value.  The resulting dynamic environment is
          then used for processing the remaining bindings (or the
          evaluation of body if there are no other bindings).

        * Otherwise p will be used according to the following protocol:
          we say that the call `(p)' "gets p's value" and that the call
          `(p x)' "sets p's value to x".  First, the `parameterize'
          form gets p's value and saves it in a local variable.  It
          then sets p's value to value.  It then processes the
          remaining bindings (or evaluates body if there are no other
          bindings).  Then it sets p's value to the saved value.  These
          steps are performed in a `dynamic-wind' so that it is
          possible to use continuations to jump into and out of the body
          (i.e. the `dynamic-wind''s before thunk sets p's value to
          value and the after thunk sets p's value to the saved value).


     The result(s) of the `parameterize' form are the result(s) of the
     body.

     Note that using procedures instead of parameter objects may lead to
     unexpected results in multithreaded programs because the before and
     after thunks of the `dynamic-wind' are not called when control
     switches between threads.

     For example:

          > (define radix (make-parameter 2))
          > (define prompt
              (make-parameter
                123
                (lambda (x)
                  (if (string? x)
                      x
                      (object->string x)))))
          > (radix)
          2
          > (parameterize ((radix 16)) (radix))
          16
          > (radix)
          2
          > (define (f n) (number->string n (radix)))
          > (f 10)
          "1010"
          > (parameterize ((radix 8)) (f 10))
          "12"
          > (parameterize ((radix 8) (prompt (f 10))) (prompt))
          "1010"
          > (define p
              (let ((x 1))
                (lambda args
                  (if (null? args) x (set! x (car args))))))
          > (let* ((a (p))
                   (b (parameterize ((p 2)) (list (p))))
                   (c (p)))
              (list a b c))
          (1 (2) 1)


15 Exceptions
*************

15.1 Exception-handling
=======================

Gambit's exception-handling model is inspired from the withdrawn
"Exception Handling SRFI" (SRFI 12), the "Multithreading support SRFI"
(SRFI 18), and the "Exception Handling for Programs SRFI" (SRFI 34).
The two fundamental operations are the dynamic binding of an exception
handler (i.e. the procedure `with-exception-handler') and the
invocation of the exception handler (i.e. the procedure `raise').

   All predefined procedures which check for errors (including type
errors, memory allocation errors, host operating-system errors, etc)
report these errors using the exception-handling system (i.e. they
"raise" an exception that can be handled in a user-defined exception
handler).  When an exception is raised and the exception is not handled
by a user-defined exception handler, the predefined exception handler
will display an error message (if the primordial thread raised the
exception) or the thread will silently terminate with no error message
(if it is not the primordial thread that raised the exception).  This
default behavior can be changed through the `-:d' runtime option (*note
Runtime options::).

   Predefined procedures normally raise exceptions by performing a
tail-call to the exception handler (the exceptions are "complex"
procedures such as `eval', `compile-file', `read', `write', etc).  This
means that the continuation of the exception handler and of the REPL
that may be started due to this is normally the continuation of the
predefined procedure that raised the exception.  By exiting the REPL
with the `,(c EXPRESSION)' command it is thus possible to resume the
program as though the call to the predefined procedure returned the
value of EXPRESSION.  For example:

     > (define (f x) (+ (car x) 1))
     > (f 2) ; typo... we meant to say (f '(2))
     *** ERROR IN f, (console)@1.18 -- (Argument 1) PAIR expected
     (car 2)
     1> ,(c 2)
     3

 -- procedure: current-exception-handler [NEW-EXCEPTION-HANDLER]
     The parameter object `current-exception-handler' is bound to the
     current exception-handler.  Calling this procedure with no argument
     returns the current exception-handler and calling this procedure
     with one argument sets the current exception-handler to
     NEW-EXCEPTION-HANDLER.

     For example:

          > (current-exception-handler)
          #<procedure #2 primordial-exception-handler>
          > (current-exception-handler (lambda (exc) (pp exc) 999))
          > (/ 1 0)
          #<divide-by-zero-exception #3>
          999


 -- procedure: with-exception-handler HANDLER THUNK
     Returns the result(s) of calling THUNK with no arguments.  The
     HANDLER, which must be a procedure, is installed as the current
     exception-handler in the dynamic environment in effect during the
     call to THUNK.  Note that the dynamic environment in effect during
     the call to HANDLER has HANDLER as the exception-handler.
     Consequently, an exception raised during the call to HANDLER may
     lead to an infinite loop.

     For example:

          > (with-exception-handler
              (lambda (e) (write e) 5)
              (lambda () (+ 1 (* 2 3) 4)))
          11
          > (with-exception-handler
              (lambda (e) (write e) 5)
              (lambda () (+ 1 (* 'foo 3) 4)))
          #<type-exception #2>10
          > (with-exception-handler
              (lambda (e) (write e 9))
              (lambda () (+ 1 (* 'foo 3) 4)))
          INFINITE LOOP


 -- procedure: with-exception-catcher HANDLER THUNK
     Returns the result(s) of calling THUNK with no arguments.  A new
     exception-handler is installed as the current exception-handler in
     the dynamic environment in effect during the call to THUNK.  This
     new exception-handler will call the HANDLER, which must be a
     procedure, with the exception object as an argument and with the
     same continuation as the call to `with-exception-catcher'.  This
     implies that the dynamic environment in effect during the call to
     HANDLER is the same as the one in effect at the call to
     `with-exception-catcher'.  Consequently, an exception raised
     during the call to HANDLER will not lead to an infinite loop.

     For example:

          > (with-exception-catcher
              (lambda (e) (write e) 5)
              (lambda () (+ 1 (* 2 3) 4)))
          11
          > (with-exception-catcher
              (lambda (e) (write e) 5)
              (lambda () (+ 1 (* 'foo 3) 4)))
          #<type-exception #2>5
          > (with-exception-catcher
              (lambda (e) (write e 9))
              (lambda () (+ 1 (* 'foo 3) 4)))
          *** ERROR IN (console)@7.1 -- (Argument 2) OUTPUT PORT expected
          (write '#<type-exception #3> 9)


 -- procedure: raise OBJ
     This procedure tail-calls the current exception-handler with OBJ
     as the sole argument.  If the exception-handler returns, the
     continuation of the call to `raise' is invoked.

     For example:

          > (with-exception-handler
              (lambda (exc)
                (pp exc)
                100)
              (lambda ()
                (+ 1 (raise "hello"))))
          "hello"
          101


 -- procedure: abort OBJ
 -- procedure: noncontinuable-exception? OBJ
 -- procedure: noncontinuable-exception-reason EXC
     The procedure `abort' calls the current exception-handler with OBJ
     as the sole argument.  If the exception-handler returns, the
     procedure `abort' will be tail-called with a
     noncontinuable-exception object, whose reason field is OBJ, as
     sole argument.

     Noncontinuable-exception objects are raised by the `abort'
     procedure when the exception-handler returns.  The parameter EXC
     must be a noncontinuable-exception object.

     The procedure `noncontinuable-exception?' returns `#t' when OBJ is
     a noncontinuable-exception object and `#f' otherwise.

     The procedure `noncontinuable-exception-reason' returns the
     argument of the call to `abort' that raised EXC.

     For example:

          > (call-with-current-continuation
              (lambda (k)
                (with-exception-handler
                  (lambda (exc)
                    (pp exc)
                    (if (noncontinuable-exception? exc)
                        (k (list (noncontinuable-exception-reason exc)))
                        100))
                  (lambda ()
                    (+ 1 (abort "hello"))))))
          "hello"
          #<noncontinuable-exception #2>
          ("hello")


15.2 Exception objects related to memory management
===================================================

 -- procedure: heap-overflow-exception? OBJ
     Heap-overflow-exception objects are raised when the allocation of
     an object would cause the heap to use more memory space than is
     available.

     The procedure `heap-overflow-exception?' returns `#t' when OBJ is
     a heap-overflow-exception object and `#f' otherwise.

     For example:

          > (define (handler exc)
              (if (heap-overflow-exception? exc)
                  exc
                  'not-heap-overflow-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (define (f x) (f (cons 1 x)))
                (f '())))
          #<heap-overflow-exception #2>


 -- procedure: stack-overflow-exception? OBJ
     Stack-overflow-exception objects are raised when the allocation of
     a continuation frame would cause the heap to use more memory space
     than is available.

     The procedure `stack-overflow-exception?' returns `#t' when OBJ is
     a stack-overflow-exception object and `#f' otherwise.

     For example:

          > (define (handler exc)
              (if (stack-overflow-exception? exc)
                  exc
                  'not-stack-overflow-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (define (f) (+ 1 (f)))
                (f)))
          #<stack-overflow-exception #2>


15.3 Exception objects related to the host environment
======================================================

 -- procedure: os-exception? OBJ
 -- procedure: os-exception-procedure EXC
 -- procedure: os-exception-arguments EXC
 -- procedure: os-exception-code EXC
 -- procedure: os-exception-message EXC
     Os-exception objects are raised by procedures which access the host
     operating-system's services when the requested operation fails.
     The parameter EXC must be a os-exception object.

     The procedure `os-exception?' returns `#t' when OBJ is a
     os-exception object and `#f' otherwise.

     The procedure `os-exception-procedure' returns the procedure that
     raised EXC.

     The procedure `os-exception-arguments' returns the list of
     arguments of the procedure that raised EXC.

     The procedure `os-exception-code' returns an exact integer error
     code that can be converted to a string by the `err-code->string'
     procedure.  Note that the error code is operating-system dependent.

     The procedure `os-exception-message' returns `#f' or a string
     giving details of the exception in a human-readable form.

     For example:

          > (define (handler exc)
              (if (os-exception? exc)
                  (list (os-exception-procedure exc)
                        (os-exception-arguments exc)
                        (err-code->string (os-exception-code exc))
                        (os-exception-message exc))
                  'not-os-exception))
          > (with-exception-catcher
              handler
              (lambda () (host-info "x.y.z")))
          (#<procedure #2 host-info> ("x.y.z") "Unknown host" #f)


 -- procedure: no-such-file-or-directory-exception? OBJ
 -- procedure: no-such-file-or-directory-exception-procedure EXC
 -- procedure: no-such-file-or-directory-exception-arguments EXC
     No-such-file-or-directory-exception objects are raised by
     procedures which access the filesystem (such as `open-input-file'
     and `directory-files') when the path specified can't be found on
     the filesystem.  The parameter EXC must be a
     no-such-file-or-directory-exception object.

     The procedure `no-such-file-or-directory-exception?' returns `#t'
     when OBJ is a no-such-file-or-directory-exception object and `#f'
     otherwise.

     The procedure `no-such-file-or-directory-exception-procedure'
     returns the procedure that raised EXC.

     The procedure `no-such-file-or-directory-exception-arguments'
     returns the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (no-such-file-or-directory-exception? exc)
                  (list (no-such-file-or-directory-exception-procedure exc)
                        (no-such-file-or-directory-exception-arguments exc))
                  'not-no-such-file-or-directory-exception))
          > (with-exception-catcher
              handler
              (lambda () (with-input-from-file "nofile" read)))
          (#<procedure #2 with-input-from-file> ("nofile" #<procedure #3 read>))


 -- procedure: unbound-os-environment-variable-exception? OBJ
 -- procedure: unbound-os-environment-variable-exception-procedure EXC
 -- procedure: unbound-os-environment-variable-exception-arguments EXC
     Unbound-os-environment-variable-exception objects are raised when
     an unbound operating-system environment variable is accessed by the
     procedures `getenv' and `setenv'.  The parameter EXC must be an
     unbound-os-environment-variable-exception object.

     The procedure `unbound-os-environment-variable-exception?' returns
     `#t' when OBJ is an unbound-os-environment-variable-exception
     object and `#f' otherwise.

     The procedure `unbound-os-environment-variable-exception-procedure'
     returns the procedure that raised EXC.

     The procedure `unbound-os-environment-variable-exception-arguments'
     returns the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (unbound-os-environment-variable-exception? exc)
                  (list (unbound-os-environment-variable-exception-procedure exc)
                        (unbound-os-environment-variable-exception-arguments exc))
                  'not-unbound-os-environment-variable-exception))
          > (with-exception-catcher
              handler
              (lambda () (getenv "DOES_NOT_EXIST")))
          (#<procedure #2 getenv> ("DOES_NOT_EXIST"))


15.4 Exception objects related to threads
=========================================

 -- procedure: scheduler-exception? OBJ
 -- procedure: scheduler-exception-reason EXC
     Scheduler-exception objects are raised by the scheduler when some
     operation requested from the host operating system failed (e.g.
     checking the status of the devices in order to wake up threads
     waiting to perform I/O on these devices).  The parameter EXC must
     be a scheduler-exception object.

     The procedure `scheduler-exception?' returns `#t' when OBJ is a
     scheduler-exception object and `#f' otherwise.

     The procedure `scheduler-exception-reason' returns the
     os-exception object that describes the failure detected by the
     scheduler.


 -- procedure: deadlock-exception? OBJ
     Deadlock-exception objects are raised when the scheduler discovers
     that all threads are blocked and can make no further progress.  In
     that case the scheduler unblocks the primordial-thread and forces
     it to raise a deadlock-exception object.

     The procedure `deadlock-exception?' returns `#t' when OBJ is a
     deadlock-exception object and `#f' otherwise.

     For example:

          > (define (handler exc)
              (if (deadlock-exception? exc)
                  exc
                  'not-deadlock-exception))
          > (with-exception-catcher
              handler
              (lambda () (read (open-vector))))
          #<deadlock-exception #2>


 -- procedure: abandoned-mutex-exception? OBJ
     Abandoned-mutex-exception objects are raised when the current
     thread locks a mutex that was owned by a thread which terminated
     (see `mutex-lock!').

     The procedure `abandoned-mutex-exception?' returns `#t' when OBJ
     is a abandoned-mutex-exception object and `#f' otherwise.

     For example:

          > (define (handler exc)
              (if (abandoned-mutex-exception? exc)
                  exc
                  'not-abandoned-mutex-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (let ((m (make-mutex)))
                  (thread-join!
                    (thread-start!
                      (make-thread
                        (lambda () (mutex-lock! m)))))
                  (mutex-lock! m))))
          #<abandoned-mutex-exception #2>


 -- procedure: join-timeout-exception? OBJ
 -- procedure: join-timeout-exception-procedure EXC
 -- procedure: join-timeout-exception-arguments EXC
     Join-timeout-exception objects are raised when a call to the
     `thread-join!' procedure reaches its timeout before the target
     thread terminates and a timeout-value parameter is not specified.
     The parameter EXC must be a join-timeout-exception object.

     The procedure `join-timeout-exception?' returns `#t' when OBJ is a
     join-timeout-exception object and `#f' otherwise.

     The procedure `join-timeout-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `join-timeout-exception-arguments' returns the list
     of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (join-timeout-exception? exc)
                  (list (join-timeout-exception-procedure exc)
                        (join-timeout-exception-arguments exc))
                  'not-join-timeout-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (thread-join!
                  (thread-start!
                    (make-thread
                      (lambda () (thread-sleep! 10))))
                  5)))
          (#<procedure #2 thread-join!> (#<thread #3> 5))


 -- procedure: started-thread-exception? OBJ
 -- procedure: started-thread-exception-procedure EXC
 -- procedure: started-thread-exception-arguments EXC
     Started-thread-exception objects are raised when the target thread
     of a call to the procedure `thread-start!' is already started.  The
     parameter EXC must be a started-thread-exception object.

     The procedure `started-thread-exception?' returns `#t' when OBJ is
     a started-thread-exception object and `#f' otherwise.

     The procedure `started-thread-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `started-thread-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (started-thread-exception? exc)
                  (list (started-thread-exception-procedure exc)
                        (started-thread-exception-arguments exc))
                  'not-started-thread-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (let ((t (make-thread (lambda () (expt 2 1000)))))
                  (thread-start! t)
                  (thread-start! t))))
          (#<procedure #2 thread-start!> (#<thread #3>))


 -- procedure: terminated-thread-exception? OBJ
 -- procedure: terminated-thread-exception-procedure EXC
 -- procedure: terminated-thread-exception-arguments EXC
     Terminated-thread-exception objects are raised when the
     `thread-join!' procedure is called and the target thread has
     terminated as a result of a call to the `thread-terminate!'
     procedure.  The parameter EXC must be a
     terminated-thread-exception object.

     The procedure `terminated-thread-exception?' returns `#t' when OBJ
     is a terminated-thread-exception object and `#f' otherwise.

     The procedure `terminated-thread-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `terminated-thread-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (terminated-thread-exception? exc)
                  (list (terminated-thread-exception-procedure exc)
                        (terminated-thread-exception-arguments exc))
                  'not-terminated-thread-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (thread-join!
                  (thread-start!
                    (make-thread
                      (lambda () (thread-terminate! (current-thread))))))))
          (#<procedure #2 thread-join!> (#<thread #3>))


 -- procedure: uncaught-exception? OBJ
 -- procedure: uncaught-exception-procedure EXC
 -- procedure: uncaught-exception-arguments EXC
 -- procedure: uncaught-exception-reason EXC
     Uncaught-exception objects are raised when an object is raised in a
     thread and that thread does not handle it (i.e. the thread
     terminated because it did not catch an exception it raised).  The
     parameter EXC must be an uncaught-exception object.

     The procedure `uncaught-exception?' returns `#t' when OBJ is an
     uncaught-exception object and `#f' otherwise.

     The procedure `uncaught-exception-procedure' returns the procedure
     that raised EXC.

     The procedure `uncaught-exception-arguments' returns the list of
     arguments of the procedure that raised EXC.

     The procedure `uncaught-exception-reason' returns the object that
     was raised by the thread and not handled by that thread.

     For example:

          > (define (handler exc)
              (if (uncaught-exception? exc)
                  (list (uncaught-exception-procedure exc)
                        (uncaught-exception-arguments exc)
                        (uncaught-exception-reason exc))
                  'not-uncaught-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (thread-join!
                  (thread-start!
                    (make-thread
                      (lambda () (open-input-file "data" 99)))))))
          (#<procedure #2 thread-join!>
           (#<thread #3>)
           #<wrong-number-of-arguments-exception #4>)


15.5 Exception objects related to C-interface
=============================================

 -- procedure: cfun-conversion-exception? OBJ
 -- procedure: cfun-conversion-exception-procedure EXC
 -- procedure: cfun-conversion-exception-arguments EXC
 -- procedure: cfun-conversion-exception-code EXC
 -- procedure: cfun-conversion-exception-message EXC
     Cfun-conversion-exception objects are raised by the C-interface
     when converting between the Scheme representation and the C
     representation of a value during a call from Scheme to C.  The
     parameter EXC must be a cfun-conversion-exception object.

     The procedure `cfun-conversion-exception?' returns `#t' when OBJ
     is a cfun-conversion-exception object and `#f' otherwise.

     The procedure `cfun-conversion-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `cfun-conversion-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     The procedure `cfun-conversion-exception-code' returns an exact
     integer error code that can be converted to a string by the
     `err-code->string' procedure.

     The procedure `cfun-conversion-exception-message' returns `#f' or
     a string giving details of the exception in a human-readable form.

     For example:

          $ cat test1.scm
          (define weird
            (c-lambda (char-string) nonnull-char-string
              "___return(___arg1);"))
          $ gsc test1.scm
          $ gsi
          Gambit v4.8.5

          > (load "test1")
          "/Users/feeley/gambit/doc/test1.o1"
          > (weird "hello")
          "hello"
          > (define (handler exc)
              (if (cfun-conversion-exception? exc)
                  (list (cfun-conversion-exception-procedure exc)
                        (cfun-conversion-exception-arguments exc)
                        (err-code->string (cfun-conversion-exception-code exc))
                        (cfun-conversion-exception-message exc))
                  'not-cfun-conversion-exception))
          > (with-exception-catcher
              handler
              (lambda () (weird 'not-a-string)))
          (#<procedure #2 weird>
           (not-a-string)
           "(Argument 1) Can't convert to C char-string"
           #f)
          > (with-exception-catcher
              handler
              (lambda () (weird #f)))
          (#<procedure #2 weird>
           (#f)
           "Can't convert result from C nonnull-char-string"
           #f)


 -- procedure: sfun-conversion-exception? OBJ
 -- procedure: sfun-conversion-exception-procedure EXC
 -- procedure: sfun-conversion-exception-arguments EXC
 -- procedure: sfun-conversion-exception-code EXC
 -- procedure: sfun-conversion-exception-message EXC
     Sfun-conversion-exception objects are raised by the C-interface
     when converting between the Scheme representation and the C
     representation of a value during a call from C to Scheme.  The
     parameter EXC must be a sfun-conversion-exception object.

     The procedure `sfun-conversion-exception?' returns `#t' when OBJ
     is a sfun-conversion-exception object and `#f' otherwise.

     The procedure `sfun-conversion-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `sfun-conversion-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     The procedure `sfun-conversion-exception-code' returns an exact
     integer error code that can be converted to a string by the
     `err-code->string' procedure.

     The procedure `sfun-conversion-exception-message' returns `#f' or
     a string giving details of the exception in a human-readable form.

     For example:

          $ cat test2.scm
          (c-define (f str) (nonnull-char-string) int "f" ""
            (string->number str))
          (define t1 (c-lambda () int "___return(f (\"123\"));"))
          (define t2 (c-lambda () int "___return(f (0));"))
          (define t3 (c-lambda () int "___return(f (\"1.5\"));"))
          $ gsc test2.scm
          $ gsi
          Gambit v4.8.5

          > (load "test2")
          "/u/feeley/test2.o1"
          > (t1)
          123
          > (define (handler exc)
              (if (sfun-conversion-exception? exc)
                  (list (sfun-conversion-exception-procedure exc)
                        (sfun-conversion-exception-arguments exc)
                        (err-code->string (sfun-conversion-exception-code exc))
                        (sfun-conversion-exception-message exc))
                  'not-sfun-conversion-exception))
          > (with-exception-catcher handler t2)
          (#<procedure #2 f>
           ()
           "(Argument 1) Can't convert from C nonnull-char-string"
           #f)
          > (with-exception-catcher handler t3)
          (#<procedure #2 f> () "Can't convert result to C int" #f)


 -- procedure: multiple-c-return-exception? OBJ
     Multiple-c-return-exception objects are raised by the C-interface
     when a C to Scheme procedure call returns and that call's stack
     frame is no longer on the C stack because the call has already
     returned, or has been removed from the C stack by a `longjump'.

     The procedure `multiple-c-return-exception?' returns `#t' when OBJ
     is a multiple-c-return-exception object and `#f' otherwise.

     For example:

          $ cat test3.scm
          (c-define (f str) (char-string) scheme-object "f" ""
            (pp (list 'entry 'str= str))
            (let ((k (call-with-current-continuation (lambda (k) k))))
              (pp (list 'exit 'k= k))
              k))
          (define scheme-to-c-to-scheme-and-back
            (c-lambda (char-string) scheme-object
              "___return(f (___arg1));"))
          $ gsc test3.scm
          $ gsi
          Gambit v4.8.5

          > (load "test3")
          "/Users/feeley/gambit/doc/test3.o1"
          > (define (handler exc)
              (if (multiple-c-return-exception? exc)
                  exc
                  'not-multiple-c-return-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (let ((c (scheme-to-c-to-scheme-and-back "hello")))
                  (pp c)
                  (c 999))))
          (entry str= "hello")
          (exit k= #<procedure #2>)
          #<procedure #2>
          (exit k= 999)
          #<multiple-c-return-exception #3>


15.6 Exception objects related to the reader
============================================

 -- procedure: datum-parsing-exception? OBJ
 -- procedure: datum-parsing-exception-kind EXC
 -- procedure: datum-parsing-exception-parameters EXC
 -- procedure: datum-parsing-exception-readenv EXC
     Datum-parsing-exception objects are raised by the reader (i.e. the
     `read' procedure) when the input does not conform to the grammar
     for datum.  The parameter EXC must be a datum-parsing-exception
     object.

     The procedure `datum-parsing-exception?' returns `#t' when OBJ is
     a datum-parsing-exception object and `#f' otherwise.

     The procedure `datum-parsing-exception-kind' returns a symbol
     denoting the kind of parsing error that was encountered by the
     reader when it raised EXC.  Here is a table of the possible return
     values:

     `datum-or-eof-expected'            Datum or EOF expected
     `datum-expected'                   Datum expected
     `improperly-placed-dot'            Improperly placed dot
     `incomplete-form-eof-reached'      Incomplete form, EOF reached
     `incomplete-form'                  Incomplete form
     `character-out-of-range'           Character out of range
     `invalid-character-name'           Invalid '#\' name
     `illegal-character'                Illegal character
     `s8-expected'                      Signed 8 bit exact integer
                                        expected
     `u8-expected'                      Unsigned 8 bit exact integer
                                        expected
     `s16-expected'                     Signed 16 bit exact integer
                                        expected
     `u16-expected'                     Unsigned 16 bit exact integer
                                        expected
     `s32-expected'                     Signed 32 bit exact integer
                                        expected
     `u32-expected'                     Unsigned 32 bit exact integer
                                        expected
     `s64-expected'                     Signed 64 bit exact integer
                                        expected
     `u64-expected'                     Unsigned 64 bit exact integer
                                        expected
     `inexact-real-expected'            Inexact real expected
     `invalid-hex-escape'               Invalid hexadecimal escape
     `invalid-escaped-character'        Invalid escaped character
     `open-paren-expected'              '(' expected
     `invalid-token'                    Invalid token
     `invalid-sharp-bang-name'          Invalid '#!' name
     `duplicate-label-definition'       Duplicate definition for label
     `missing-label-definition'         Missing definition for label
     `illegal-label-definition'         Illegal definition of label
     `invalid-infix-syntax-character'   Invalid infix syntax character
     `invalid-infix-syntax-number'      Invalid infix syntax number
     `invalid-infix-syntax'             Invalid infix syntax

     The procedure `datum-parsing-exception-parameters' returns a list
     of the parameters associated with the parsing error that was
     encountered by the reader when it raised EXC.

     For example:

          > (define (handler exc)
              (if (datum-parsing-exception? exc)
                  (list (datum-parsing-exception-kind exc)
                        (datum-parsing-exception-parameters exc))
                  'not-datum-parsing-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (with-input-from-string "(s #\\pace)" read)))
          (invalid-character-name ("pace"))


15.7 Exception objects related to evaluation and compilation
============================================================

 -- procedure: expression-parsing-exception? OBJ
 -- procedure: expression-parsing-exception-kind EXC
 -- procedure: expression-parsing-exception-parameters EXC
 -- procedure: expression-parsing-exception-source EXC
     Expression-parsing-exception objects are raised by the evaluator
     and compiler (i.e. the procedures `eval', `compile-file', etc)
     when the input does not conform to the grammar for expression.  The
     parameter EXC must be a expression-parsing-exception object.

     The procedure `expression-parsing-exception?' returns `#t' when
     OBJ is a expression-parsing-exception object and `#f' otherwise.

     The procedure `expression-parsing-exception-kind' returns a symbol
     denoting the kind of parsing error that was encountered by the
     evaluator or compiler when it raised EXC.  Here is a table of the
     possible return values:

     `id-expected'                      Identifier expected
     `ill-formed-namespace'             Ill-formed namespace
     `ill-formed-namespace-prefix'      Ill-formed namespace prefix
     `namespace-prefix-must-be-string'  Namespace prefix must be a string
     `macro-used-as-variable'           Macro name can't be used as a
                                        variable
     `variable-is-immutable'            Variable is immutable
     `ill-formed-macro-transformer'     Macro transformer must be a
                                        lambda expression
     `reserved-used-as-variable'        Reserved identifier can't be used
                                        as a variable
     `ill-formed-special-form'          Ill-formed special form
     `cannot-open-file'                 Can't open file
     `filename-expected'                Filename expected
     `ill-placed-define'                Ill-placed 'define'
     `ill-placed-**include'             Ill-placed '##include'
     `ill-placed-**define-macro'        Ill-placed '##define-macro'
     `ill-placed-**declare'             Ill-placed '##declare'
     `ill-placed-**namespace'           Ill-placed '##namespace'
     `ill-formed-expression'            Ill-formed expression
     `unsupported-special-form'         Interpreter does not support
     `ill-placed-unquote'               Ill-placed 'unquote'
     `ill-placed-unquote-splicing'      Ill-placed 'unquote-splicing'
     `parameter-must-be-id'             Parameter must be an identifier
     `parameter-must-be-id-or-default'  Parameter must be an identifier
                                        or default binding
     `duplicate-parameter'              Duplicate parameter in parameter
                                        list
     `ill-placed-dotted-rest-parameter' Ill-placed dotted rest parameter
     `parameter-expected-after-rest'    #!rest must be followed by a
                                        parameter
     `ill-formed-default'               Ill-formed default binding
     `ill-placed-optional'              Ill-placed #!optional
     `ill-placed-rest'                  Ill-placed #!rest
     `ill-placed-key'                   Ill-placed #!key
     `key-expected-after-rest'          #!key expected after rest
                                        parameter
     `ill-placed-default'               Ill-placed default binding
     `duplicate-variable-definition'    Duplicate definition of a variable
     `empty-body'                       Body must contain at least one
                                        expression
     `variable-must-be-id'              Defined variable must be an
                                        identifier
     `else-clause-not-last'             Else clause must be last
     `ill-formed-selector-list'         Ill-formed selector list
     `duplicate-variable-binding'       Duplicate variable in bindings
     `ill-formed-binding-list'          Ill-formed binding list
     `ill-formed-call'                  Ill-formed procedure call
     `ill-formed-cond-expand'           Ill-formed 'cond-expand'
     `unfulfilled-cond-expand'          Unfulfilled 'cond-expand'

     The procedure `expression-parsing-exception-parameters' returns a
     list of the parameters associated with the parsing error that was
     encountered by the evaluator or compiler when it raised EXC.

     For example:

          > (define (handler exc)
              (if (expression-parsing-exception? exc)
                  (list (expression-parsing-exception-kind exc)
                        (expression-parsing-exception-parameters exc))
                  'not-expression-parsing-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (eval '(+ do 1))))
          (reserved-used-as-variable (do))


 -- procedure: unbound-global-exception? OBJ
 -- procedure: unbound-global-exception-variable EXC
 -- procedure: unbound-global-exception-code EXC
 -- procedure: unbound-global-exception-rte EXC
     Unbound-global-exception objects are raised when an unbound global
     variable is accessed.  The parameter EXC must be an
     unbound-global-exception object.

     The procedure `unbound-global-exception?' returns `#t' when OBJ is
     an unbound-global-exception object and `#f' otherwise.

     The procedure `unbound-global-exception-variable' returns a symbol
     identifying the unbound global variable.

     For example:

          > (define (handler exc)
              (if (unbound-global-exception? exc)
                  (list 'variable= (unbound-global-exception-variable exc))
                  'not-unbound-global-exception))
          > (with-exception-catcher
              handler
              (lambda () foo))
          (variable= foo)


15.8 Exception objects related to type checking
===============================================

 -- procedure: type-exception? OBJ
 -- procedure: type-exception-procedure EXC
 -- procedure: type-exception-arguments EXC
 -- procedure: type-exception-arg-num EXC
 -- procedure: type-exception-type-id EXC
     Type-exception objects are raised when a primitive procedure is
     called with an argument of incorrect type (i.e. when a run time
     type-check fails).  The parameter EXC must be a type-exception
     object.

     The procedure `type-exception?' returns `#t' when OBJ is a
     type-exception object and `#f' otherwise.

     The procedure `type-exception-procedure' returns the procedure
     that raised EXC.

     The procedure `type-exception-arguments' returns the list of
     arguments of the procedure that raised EXC.

     The procedure `type-exception-arg-num' returns the position of the
     argument whose type is incorrect.  Position 1 is the first
     argument.

     The procedure `type-exception-type-id' returns an identifier of
     the type expected.  The type-id can be a symbol, such as `number'
     and `string-or-nonnegative-fixnum', or a record type descriptor.

     For example:

          > (define (handler exc)
              (if (type-exception? exc)
                  (list (type-exception-procedure exc)
                        (type-exception-arguments exc)
                        (type-exception-arg-num exc)
                        (type-exception-type-id exc))
                  'not-type-exception))
          > (with-exception-catcher
              handler
              (lambda () (vector-ref '#(a b c) 'foo)))
          (#<procedure #2 vector-ref> (#(a b c) foo) 2 exact-integer)
          > (with-exception-catcher
              handler
              (lambda () (time->seconds 'foo)))
          (#<procedure #3 time->seconds> (foo) 1 #<type #4 time>)


 -- procedure: range-exception? OBJ
 -- procedure: range-exception-procedure EXC
 -- procedure: range-exception-arguments EXC
 -- procedure: range-exception-arg-num EXC
     Range-exception objects are raised when a numeric parameter is not
     in the allowed range.  The parameter EXC must be a range-exception
     object.

     The procedure `range-exception?' returns `#t' when OBJ is a
     range-exception object and `#f' otherwise.

     The procedure `range-exception-procedure' returns the procedure
     that raised EXC.

     The procedure `range-exception-arguments' returns the list of
     arguments of the procedure that raised EXC.

     The procedure `range-exception-arg-num' returns the position of
     the argument which is not in the allowed range.  Position 1 is the
     first argument.

     For example:

          > (define (handler exc)
              (if (range-exception? exc)
                  (list (range-exception-procedure exc)
                        (range-exception-arguments exc)
                        (range-exception-arg-num exc))
                  'not-range-exception))
          > (with-exception-catcher
              handler
              (lambda () (string-ref "abcde" 10)))
          (#<procedure #2 string-ref> ("abcde" 10) 2)


 -- procedure: divide-by-zero-exception? OBJ
 -- procedure: divide-by-zero-exception-procedure EXC
 -- procedure: divide-by-zero-exception-arguments EXC
     Divide-by-zero-exception objects are raised when a division by
     zero is attempted.  The parameter EXC must be a
     divide-by-zero-exception object.

     The procedure `divide-by-zero-exception?' returns `#t' when OBJ is
     a divide-by-zero-exception object and `#f' otherwise.

     The procedure `divide-by-zero-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `divide-by-zero-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (divide-by-zero-exception? exc)
                  (list (divide-by-zero-exception-procedure exc)
                        (divide-by-zero-exception-arguments exc))
                  'not-divide-by-zero-exception))
          > (with-exception-catcher
              handler
              (lambda () (/ 5 0 7)))
          (#<procedure #2 /> (5 0 7))


 -- procedure: improper-length-list-exception? OBJ
 -- procedure: improper-length-list-exception-procedure EXC
 -- procedure: improper-length-list-exception-arguments EXC
 -- procedure: improper-length-list-exception-arg-num EXC
     Improper-length-list-exception objects are raised by the `map' and
     `for-each' procedures when they are called with two or more list
     arguments and the lists are not of the same length.  The parameter
     EXC must be an improper-length-list-exception object.

     The procedure `improper-length-list-exception?' returns `#t' when
     OBJ is an improper-length-list-exception object and `#f' otherwise.

     The procedure `improper-length-list-exception-procedure' returns
     the procedure that raised EXC.

     The procedure `improper-length-list-exception-arguments' returns
     the list of arguments of the procedure that raised EXC.

     The procedure `improper-length-list-exception-arg-num' returns the
     position of the argument whose length is the shortest.  Position 1
     is the first argument.

     For example:

          > (define (handler exc)
              (if (improper-length-list-exception? exc)
                  (list (improper-length-list-exception-procedure exc)
                        (improper-length-list-exception-arguments exc)
                        (improper-length-list-exception-arg-num exc))
                  'not-improper-length-list-exception))
          > (with-exception-catcher
              handler
              (lambda () (map + '(1 2) '(3) '(4 5))))
          (#<procedure #2 map> (#<procedure #3 +> (1 2) (3) (4 5)) 3)


15.9 Exception objects related to procedure call
================================================

 -- procedure: wrong-number-of-arguments-exception? OBJ
 -- procedure: wrong-number-of-arguments-exception-procedure EXC
 -- procedure: wrong-number-of-arguments-exception-arguments EXC
     Wrong-number-of-arguments-exception objects are raised when a
     procedure is called with the wrong number of arguments.  The
     parameter EXC must be a wrong-number-of-arguments-exception object.

     The procedure `wrong-number-of-arguments-exception?' returns `#t'
     when OBJ is a wrong-number-of-arguments-exception object and `#f'
     otherwise.

     The procedure `wrong-number-of-arguments-exception-procedure'
     returns the procedure that raised EXC.

     The procedure `wrong-number-of-arguments-exception-arguments'
     returns the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (wrong-number-of-arguments-exception? exc)
                  (list (wrong-number-of-arguments-exception-procedure exc)
                        (wrong-number-of-arguments-exception-arguments exc))
                  'not-wrong-number-of-arguments-exception))
          > (with-exception-catcher
              handler
              (lambda () (open-input-file "data" 99)))
          (#<procedure #2 open-input-file> ("data" 99))


 -- procedure: number-of-arguments-limit-exception? OBJ
 -- procedure: number-of-arguments-limit-exception-procedure EXC
 -- procedure: number-of-arguments-limit-exception-arguments EXC
     Number-of-arguments-limit-exception objects are raised by the
     `apply' procedure when the procedure being called is passed more
     than 8192 arguments.  The parameter EXC must be a
     number-of-arguments-limit-exception object.

     The procedure `number-of-arguments-limit-exception?' returns `#t'
     when OBJ is a number-of-arguments-limit-exception object and `#f'
     otherwise.

     The procedure `number-of-arguments-limit-exception-procedure'
     returns the target procedure of the call to `apply' that raised
     EXC.

     The procedure `number-of-arguments-limit-exception-arguments'
     returns the list of arguments of the target procedure of the call
     to `apply' that raised EXC.

     For example:

          > (define (iota n) (if (= n 0) '() (cons n (iota (- n 1)))))
          > (define (handler exc)
              (if (number-of-arguments-limit-exception? exc)
                  (list (number-of-arguments-limit-exception-procedure exc)
                        (length (number-of-arguments-limit-exception-arguments exc)))
                  'not-number-of-arguments-limit-exception))
          > (with-exception-catcher
              handler
              (lambda () (apply + 1 2 3 (iota 8190))))
          (#<procedure #2 +> 8193)


 -- procedure: nonprocedure-operator-exception? OBJ
 -- procedure: nonprocedure-operator-exception-operator EXC
 -- procedure: nonprocedure-operator-exception-arguments EXC
 -- procedure: nonprocedure-operator-exception-code EXC
 -- procedure: nonprocedure-operator-exception-rte EXC
     Nonprocedure-operator-exception objects are raised when a procedure
     call is executed and the operator position is not a procedure.  The
     parameter EXC must be an nonprocedure-operator-exception object.

     The procedure `nonprocedure-operator-exception?' returns `#t' when
     OBJ is an nonprocedure-operator-exception object and `#f'
     otherwise.

     The procedure `nonprocedure-operator-exception-operator' returns
     the value in operator position of the procedure call that raised
     EXC.

     The procedure `nonprocedure-operator-exception-arguments' returns
     the list of arguments of the procedure call that raised EXC.

     For example:

          > (define (handler exc)
              (if (nonprocedure-operator-exception? exc)
                  (list (nonprocedure-operator-exception-operator exc)
                        (nonprocedure-operator-exception-arguments exc))
                  'not-nonprocedure-operator-exception))
          > (with-exception-catcher
              handler
              (lambda () (11 22 33)))
          (11 (22 33))


 -- procedure: unknown-keyword-argument-exception? OBJ
 -- procedure: unknown-keyword-argument-exception-procedure EXC
 -- procedure: unknown-keyword-argument-exception-arguments EXC
     Unknown-keyword-argument-exception objects are raised when a
     procedure accepting keyword arguments is called and one of the
     keywords supplied is not among those that are expected.  The
     parameter EXC must be an unknown-keyword-argument-exception object.

     The procedure `unknown-keyword-argument-exception?' returns `#t'
     when OBJ is an unknown-keyword-argument-exception object and `#f'
     otherwise.

     The procedure `unknown-keyword-argument-exception-procedure'
     returns the procedure that raised EXC.

     The procedure `unknown-keyword-argument-exception-arguments'
     returns the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (unknown-keyword-argument-exception? exc)
                  (list (unknown-keyword-argument-exception-procedure exc)
                        (unknown-keyword-argument-exception-arguments exc))
                  'not-unknown-keyword-argument-exception))
          > (with-exception-catcher
              handler
              (lambda () ((lambda (#!key (foo 5)) foo) bar: 11)))
          (#<procedure #2> (bar: 11))


 -- procedure: keyword-expected-exception? OBJ
 -- procedure: keyword-expected-exception-procedure EXC
 -- procedure: keyword-expected-exception-arguments EXC
     Keyword-expected-exception objects are raised when a procedure
     accepting keyword arguments is called and a nonkeyword was supplied
     where a keyword was expected.  The parameter EXC must be an
     keyword-expected-exception object.

     The procedure `keyword-expected-exception?' returns `#t' when OBJ
     is an keyword-expected-exception object and `#f' otherwise.

     The procedure `keyword-expected-exception-procedure' returns the
     procedure that raised EXC.

     The procedure `keyword-expected-exception-arguments' returns the
     list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (keyword-expected-exception? exc)
                  (list (keyword-expected-exception-procedure exc)
                        (keyword-expected-exception-arguments exc))
                  'not-keyword-expected-exception))
          > (with-exception-catcher
              handler
              (lambda () ((lambda (#!key (foo 5)) foo) 11 22)))
          (#<procedure #2> (11 22))


15.10 Other exception objects
=============================

 -- procedure: error-exception? OBJ
 -- procedure: error-exception-message EXC
 -- procedure: error-exception-parameters EXC
 -- procedure: error MESSAGE OBJ...
     Error-exception objects are raised when the `error' procedure is
     called.  The parameter EXC must be an error-exception object.

     The procedure `error-exception?' returns `#t' when OBJ is an
     error-exception object and `#f' otherwise.

     The procedure `error-exception-message' returns the first argument
     of the call to `error' that raised EXC.

     The procedure `error-exception-parameters' returns the list of
     arguments, starting with the second argument, of the call to
     `error' that raised EXC.

     The `error' procedure raises an error-exception object whose
     message field is MESSAGE and parameters field is the list of
     values OBJ....

     For example:

          > (define (handler exc)
              (if (error-exception? exc)
                  (list (error-exception-message exc)
                        (error-exception-parameters exc))
                  'not-error-exception))
          > (with-exception-catcher
              handler
              (lambda () (error "unexpected object:" 123)))
          ("unexpected object:" (123))


16 Host environment
*******************

The host environment is the set of resources, such as the filesystem,
network and processes, that are managed by the operating system within
which the Scheme program is executing.  This chapter specifies how the
host environment can be accessed from within the Scheme program.

   In this chapter we say that the Scheme program being executed is a
process, even though the concept of process does not exist in some
operating systems supported by Gambit (e.g. MSDOS).

16.1 Handling of file names
===========================

Gambit uses a naming convention for files that is compatible with the
one used by the host environment but extended to allow referring to the
"home directory" of the current user or some specific user and the
"installation directories".

   A "path" is a string that denotes a file, for example
`"src/readme.txt"'.  Each component of a path is separated by a `/'
under UNIX and Mac OS X and by a `/' or `\' under MSDOS and Microsoft
Windows.  A leading separator indicates an absolute path under UNIX,
Mac OS X, MSDOS and Microsoft Windows.  A path which does not contain a
path separator is relative to the "current working directory" on all
operating systems.  A volume specifier such as `C:' may prefix a file
name under MSDOS and Microsoft Windows.

   A path which starts with the characters `~~' denotes a file in an
installation directory.  If nothing follows the `~~' then the directory
denoted is the central installation directory.  Otherwise what follows
the `~~' is the name of the installation directory, for example `~~lib'
denotes the `lib' installation directory.  Note that the location of
the installation directories may be overridden by using the
`-:=DIRECTORY' and `-:~~DIR=DIRECTORY' runtime options or by defining
the `GAMBOPT' environment variable.

   A path which starts with the character `~' not followed by `~'
denotes a file in the user's home directory.  The user's home directory
is contained in the `HOME' environment variable under UNIX, Mac OS X,
MSDOS and Microsoft Windows.  Under MSDOS and Microsoft Windows, if the
`HOME' environment variable is not defined, the environment variables
`HOMEDRIVE' and `HOMEPATH' are concatenated if they are defined.  If
this fails to yield a home directory, the central installation
directory is used instead.

   A path which starts with the characters `~USERNAME' denotes a file
in the home directory of the given user.  Under UNIX and Mac OS X this
is found using the password file.  There is no equivalent under MSDOS
and Microsoft Windows.

 -- procedure: current-directory [NEW-CURRENT-DIRECTORY]
     The parameter object `current-directory' is bound to the current
     working directory.  Calling this procedure with no argument returns
     the absolute "normalized path" of the directory and calling this
     procedure with one argument sets the directory to
     NEW-CURRENT-DIRECTORY.  The initial binding of this parameter
     object is the current working directory of the current process.
     The path returned by `current-directory' always contains a trailing
     directory separator.  Modifications of the parameter object do not
     change the current working directory of the current process (i.e.
     that is accessible with the UNIX `getcwd()' function and the
     Microsoft Windows `GetCurrentDirectory' function).  It is an error
     to mutate the string returned by `current-directory'.

     For example under UNIX:

          > (current-directory)
          "/Users/feeley/gambit/doc/"
          > (current-directory "..")
          > (current-directory)
          "/Users/feeley/gambit/"
          > (path-expand "foo" "~~")
          "/usr/local/Gambit/foo"
          > (parameterize ((current-directory "~~")) (path-expand "foo"))
          "/usr/local/Gambit/foo"


 -- procedure: path-expand PATH [ORIGIN-DIRECTORY]
     The procedure `path-expand' takes the path of a file or directory
     and returns an expanded path, which is an absolute path when PATH
     or ORIGIN-DIRECTORY are absolute paths.  The optional
     ORIGIN-DIRECTORY parameter, which defaults to the current working
     directory, is the directory used to resolve relative paths.
     Components of the paths PATH and ORIGIN-DIRECTORY need not exist.

     For example under UNIX:

          > (path-expand "foo")
          "/Users/feeley/gambit/doc/foo"
          > (path-expand "~/foo")
          "/Users/feeley/foo"
          > (path-expand "~~lib/foo")
          "/usr/local/Gambit/lib/foo"
          > (path-expand "../foo")
          "/Users/feeley/gambit/doc/../foo"
          > (path-expand "foo" "")
          "foo"
          > (path-expand "foo" "/tmp")
          "/tmp/foo"
          > (path-expand "this/file/does/not/exist")
          "/Users/feeley/gambit/doc/this/file/does/not/exist"
          > (path-expand "")
          "/Users/feeley/gambit/doc/"


 -- procedure: path-normalize PATH [ALLOW-RELATIVE? [ORIGIN-DIRECTORY]]
     The procedure `path-normalize' takes a path of a file or directory
     and returns its normalized path.  The optional ORIGIN-DIRECTORY
     parameter, which defaults to the current working directory, is the
     directory used to resolve relative paths.  All components of the
     paths PATH and ORIGIN-DIRECTORY must exist, except possibly the
     last component of PATH.  A normalized path is a path containing no
     redundant parts and which is consistent with the current structure
     of the filesystem.  A normalized path of a directory will always
     end with a path separator (i.e. `/', `\', or `:' depending on the
     operating system).  The optional ALLOW-RELATIVE? parameter, which
     defaults to `#f', indicates if the path returned can be expressed
     relatively to ORIGIN-DIRECTORY: a `#f' requests an absolute path,
     the symbol `shortest' requests the shortest of the absolute and
     relative paths, and any other value requests the relative path.
     The shortest path is useful for interaction with the user because
     short relative paths are typically easier to read than long
     absolute paths.

     For example under UNIX:

          > (path-expand "../foo")
          "/Users/feeley/gambit/doc/../foo"
          > (path-normalize "../foo")
          "/Users/feeley/gambit/foo"
          > (path-normalize "this/file/does/not/exist")
          *** ERROR IN (console)@3.1 -- No such file or directory
          (path-normalize "this/file/does/not/exist")


 -- procedure: path-extension PATH
 -- procedure: path-strip-extension PATH
 -- procedure: path-directory PATH
 -- procedure: path-strip-directory PATH
 -- procedure: path-strip-trailing-directory-separator PATH
 -- procedure: path-volume PATH
 -- procedure: path-strip-volume PATH
     These procedures extract various parts of a path, which need not
     be a normalized path.  The procedure `path-extension' returns the
     file extension (including the period) or the empty string if there
     is no extension.  The procedure `path-strip-extension' returns the
     path with the extension stripped off.  The procedure
     `path-directory' returns the file's directory (including the last
     path separator) or the empty string if no directory is specified
     in the path.  The procedure `path-strip-directory' returns the
     path with the directory stripped off.  The procedure
     `path-strip-trailing-directory-separator' returns the path with
     the directory separator stripped off if one is at the end of the
     path.  The procedure `path-volume' returns the file's volume
     (including the last path separator) or the empty string if no
     volume is specified in the path.  The procedure
     `path-strip-volume' returns the path with the volume stripped off.

     For example under UNIX:

          > (path-extension "/tmp/foo")
          ""
          > (path-extension "/tmp/foo.txt")
          ".txt"
          > (path-strip-extension "/tmp/foo.txt")
          "/tmp/foo"
          > (path-directory "/tmp/foo.txt")
          "/tmp/"
          > (path-strip-directory "/tmp/foo.txt")
          "foo.txt"
          > (path-strip-trailing-directory-separator "/usr/local/bin/")
          "/usr/local/bin"
          > (path-strip-trailing-directory-separator "/usr/local/bin")
          "/usr/local/bin"
          > (path-volume "/tmp/foo.txt")
          ""
          > (path-volume "C:/tmp/foo.txt")
          "" ; result is "C:" under Microsoft Windows
          > (path-strip-volume "C:/tmp/foo.txt")
          "C:/tmp/foo.txt" ; result is "/tmp/foo.txt" under Microsoft Windows


16.2 Filesystem operations
==========================

 -- procedure: create-directory PATH-OR-SETTINGS
     This procedure creates a directory.  The argument PATH-OR-SETTINGS
     is either a string denoting a filesystem path or a list of port
     settings which must contain a `path:' setting.  Here are the
     settings allowed:

        * `path:' STRING

          This setting indicates the location of the directory to
          create in the filesystem.  There is no default value for this
          setting.

        * `permissions:' 12-BIT-EXACT-INTEGER

          This setting controls the UNIX permissions that will be
          attached to the file if it is created.  The default value of
          this setting is `#o777'.


     For example:

          > (create-directory "newdir")
          > (create-directory "newdir")
          *** ERROR IN (console)@2.1 -- File exists
          (create-directory "newdir")


 -- procedure: create-fifo PATH-OR-SETTINGS
     This procedure creates a FIFO.  The argument PATH-OR-SETTINGS is
     either a string denoting a filesystem path or a list of port
     settings which must contain a `path:' setting.  Here are the
     settings allowed:

        * `path:' STRING

          This setting indicates the location of the FIFO to create in
          the filesystem.  There is no default value for this setting.

        * `permissions:' 12-BIT-EXACT-INTEGER

          This setting controls the UNIX permissions that will be
          attached to the file if it is created.  The default value of
          this setting is `#o666'.


     For example:

          > (create-fifo "fifo")
          > (define a (open-input-file "fifo"))
          > (define b (open-output-file "fifo"))
          > (display "1 22 333" b)
          > (force-output b)
          > (read a)
          1
          > (read a)
          22


 -- procedure: create-link SOURCE-PATH DESTINATION-PATH
     This procedure creates a hard link between SOURCE-PATH and
     DESTINATION-PATH.  The argument SOURCE-PATH must be a string
     denoting the path of an existing file.  The argument
     DESTINATION-PATH must be a string denoting the path of the link to
     create.


 -- procedure: create-symbolic-link SOURCE-PATH DESTINATION-PATH
     This procedure creates a symbolic link between SOURCE-PATH and
     DESTINATION-PATH.  The argument SOURCE-PATH must be a string
     denoting the path of an existing file.  The argument
     DESTINATION-PATH must be a string denoting the path of the
     symbolic link to create.


 -- procedure: rename-file SOURCE-PATH DESTINATION-PATH
     This procedure renames the file SOURCE-PATH to DESTINATION-PATH.
     The argument SOURCE-PATH must be a string denoting the path of an
     existing file.  The argument DESTINATION-PATH must be a string
     denoting the new path of the file.


 -- procedure: copy-file SOURCE-PATH DESTINATION-PATH
     This procedure copies the file SOURCE-PATH to DESTINATION-PATH.
     The argument SOURCE-PATH must be a string denoting the path of an
     existing file.  The argument DESTINATION-PATH must be a string
     denoting the path of the file to create.


 -- procedure: delete-file PATH
     This procedure deletes the file PATH.  The argument PATH must be a
     string denoting the path of an existing file.


 -- procedure: delete-directory PATH
     This procedure deletes the directory PATH.  The argument PATH must
     be a string denoting the path of an existing directory.


 -- procedure: directory-files [PATH-OR-SETTINGS]
     This procedure returns the list of the files in a directory.  The
     argument PATH-OR-SETTINGS is either a string denoting a filesystem
     path to a directory or a list of settings which must contain a
     `path:' setting.  If it is not specified, PATH-OR-SETTINGS
     defaults to the current directory (the value bound to the
     `current-directory' parameter object).  Here are the settings
     allowed:

        * `path:' STRING

          This setting indicates the location of the directory in the
          filesystem.  There is no default value for this setting.

        * `ignore-hidden:' ( `#f' | `#t' | `dot-and-dot-dot' )

          This setting controls whether hidden-files will be returned.
          Under UNIX and Mac OS X hidden-files are those that start
          with a period (such as `.', `..', and `.profile').  Under
          Microsoft Windows hidden files are the `.' and `..' entries
          and the files whose "hidden file" attribute is set.  A
          setting of `#f' will enumerate all the files.  A setting of
          `#t' will only enumerate the files that are not hidden.  A
          setting of `dot-and-dot-dot' will enumerate all the files
          except for the `.' and `..' hidden files.  The default value
          of this setting is `#t'.


     For example:

          > (directory-files)
          ("complex" "README" "simple")
          > (directory-files "../include")
          ("config.h" "config.h.in" "gambit.h" "makefile" "makefile.in")
          > (directory-files (list path: "../include" ignore-hidden: #f))
          ("." ".." "config.h" "config.h.in" "gambit.h" "makefile" "makefile.in")


16.3 Shell command execution
============================

 -- procedure: shell-command COMMAND [CAPTURE?]
     The procedure `shell-command' calls up the shell to execute
     COMMAND which must be a string.  The argument CAPTURE?, which
     defaults to `#f', indicates if the output of the command is
     captured as a string.  If CAPTURE? is `#f', this procedure returns
     the exit status of the shell in the form that the C library's
     `system' routine returns.  If CAPTURE? is not `#f', this procedure
     returns a pair consisting of the exit status of the shell in the
     `car' field, and the captured output in the `cdr' field.  Be
     advised that the shell that is used, and consequently the syntax
     of COMMAND, depends on the operating system.  On Unix, the shell
     `/bin/sh' is usually invoked.  On Windows, the shell `cmd.exe' is
     usually invoked.

     For example under UNIX:

          > (shell-command "ls -sk f*.scm")
          4 fact.scm   4 fib.scm
          0
          > (shell-command "ls -sk f*.scm" #t)
          (0 . "4 fact.scm   4 fib.scm\n")
          > (shell-command "echo x\\\\\\\\y $HOME" #t)
          (0 . "x\\y /Users/feeley\n")

     For example under Windows:

          > (shell-command "echo x\\\\\\\\y %HOME%" #t)
          (0 . "x\\\\\\\\y C:\\Users\\feeley\r\n")


16.4 Process termination
========================

 -- procedure: exit [STATUS]
     The procedure `exit' causes the process to terminate with the
     status STATUS which must be an exact integer in the range 0 to
     255.  If it is not specified, STATUS defaults to 0.

     For example under UNIX:

          $ gsi
          Gambit v4.8.5

          > (exit 42)
          $ echo $?
          42


16.5 Command line arguments
===========================

 -- procedure: command-line
     This procedure returns a list of strings corresponding to the
     command line arguments, including the program file name as the
     first element of the list.  When the interpreter executes a Scheme
     script, the list returned by `command-line' contains the script's
     absolute path followed by the remaining command line arguments.

     For example under UNIX:

          $ gsi -:d -e "(pretty-print (command-line))"
          ("gsi" "-e" "(pretty-print (command-line))")
          $ cat foo
          #!/usr/local/Gambit/bin/gsi-script
          (pretty-print (command-line))
          $ ./foo 1 2 "3 4"
          ("/u/feeley/./foo" "1" "2" "3 4")


16.6 Environment variables
==========================

 -- procedure: getenv NAME [DEFAULT]
 -- procedure: setenv NAME [NEW-VALUE]
     The procedure `getenv' returns the value of the environment
     variable NAME of the current process.  Variable names are denoted
     with strings.  A string is returned if the environment variable is
     bound, otherwise DEFAULT is returned if it is specified, otherwise
     an exception is raised.

     The procedure `setenv' changes the binding of the environment
     variable NAME to NEW-VALUE which must be a string.  If NEW-VALUE
     is not specified the binding is removed.

     For example under UNIX:

          > (getenv "HOME")
          "/Users/feeley"
          > (getenv "DOES_NOT_EXIST" #f)
          #f
          > (setenv "DOES_NOT_EXIST" "it does now")
          > (getenv "DOES_NOT_EXIST" #f)
          "it does now"
          > (setenv "DOES_NOT_EXIST")
          > (getenv "DOES_NOT_EXIST" #f)
          #f
          > (getenv "DOES_NOT_EXIST")
          *** ERROR IN (console)@7.1 -- Unbound OS environment variable
          (getenv "DOES_NOT_EXIST")


16.7 Measuring time
===================

Procedures are available for measuring real time (aka "wall" time) and
cpu time (the amount of time the cpu has been executing the process).
The resolution of the real time and cpu time clock is operating system
dependent.  Typically the resolution of the cpu time clock is rather
coarse (measured in "ticks" of 1/60th or 1/100th of a second).  Real
time is internally computed relative to some arbitrary point in time
using floating point numbers, which means that there is a gradual loss
of resolution as time elapses.  Moreover, some operating systems report
time in number of ticks using a 32 bit integer so the value returned by
the time related procedures may wraparound much before any significant
loss of resolution occurs (for example 2.7 years if ticks are 1/50th of
a second).

 -- procedure: current-time
 -- procedure: time? OBJ
 -- procedure: time->seconds TIME
 -- procedure: seconds->time X
     The procedure `current-time' returns a "time object" representing
     the current point in real time.

     The procedure `time?' returns `#t' when OBJ is a time object and
     `#f' otherwise.

     The procedure `time->seconds' converts the time object TIME into
     an inexact real number representing the number of seconds elapsed
     since the "epoch" (which is 00:00:00 Coordinated Universal Time
     01-01-1970).

     The procedure `seconds->time' converts the real number X
     representing the number of seconds elapsed since the "epoch" into a
     time object.

     For example:

          > (current-time)
          #<time #2>
          > (time? (current-time))
          #t
          > (time? 123)
          #f
          > (time->seconds (current-time))
          1083118758.63973
          > (time->seconds (current-time))
          1083118759.909163
          > (seconds->time (+ 10 (time->seconds (current-time))
          #<time #3>  ; a time object representing 10 seconds in the future


 -- procedure: process-times
 -- procedure: cpu-time
 -- procedure: real-time
     The procedure `process-times' returns a three element f64vector
     containing the cpu time that has been used by the program and the
     real time that has elapsed since it was started.  The first element
     corresponds to "user" time in seconds, the second element
     corresponds to "system" time in seconds and the third element is
     the elapsed real time in seconds.  On operating systems that can't
     differentiate user and system time, the system time is zero.  On
     operating systems that can't measure cpu time, the user time is
     equal to the elapsed real time and the system time is zero.

     The procedure `cpu-time' returns the cpu time in seconds that has
     been used by the program (user time plus system time).

     The procedure `real-time' returns the real time that has elapsed
     since the program was started.

     For example:

          > (process-times)
          #f64(.02794 .021754 .159926176071167)
          > (cpu-time)
          .051223
          > (real-time)
          .40660619735717773


 -- special form: time expr [PORT]
     The `time' special form evaluates expr and returns the result.  As
     a side effect it displays a message on the port PORT which
     indicates various statistics about the evaluation of expr
     including how long the evaluation took (in real time and cpu time),
     how much time was spent in the garbage collector, how much memory
     was allocated during the evaluation and how many minor and major
     page faults occured (0 is reported if not running under UNIX).  If
     it is not specified, PORT defaults to the interaction channel
     (i.e. the output will appear at the REPL).

     For example:

          > (define (f x)
              (let loop ((x x) (lst '()))
                (if (= x 0)
                    lst
                    (loop (- x 1) (cons x lst)))))
          > (length (time (f 100000)))
          (time (f 100000))
              683 ms real time
              558 ms cpu time (535 user, 23 system)
              8 collections accounting for 102 ms real time (70 user, 5 system)
              6400160 bytes allocated
              no minor faults
              no major faults
          100000


16.8 File information
=====================

 -- procedure: file-exists? PATH [CHASE?]
     The PATH argument must be a string.  This procedure returns `#t'
     when a file by that name exists, and returns `#f' otherwise.

     When CHASE? is present and `#f', symbolic links will not be
     chased, in other words if PATH refers to a symbolic link,
     `file-exists?' will return `#t' whether or not it points to an
     existing file.

     For example:

          > (file-exists? "nofile")
          #f


 -- procedure: file-info PATH [CHASE?]
     This procedure accesses the filesystem to get information about the
     file whose location is given by the string PATH.  A
     file-information record is returned that contains the file's type,
     the device number, the inode number, the mode (permission bits),
     the number of links, the file's user id, the file's group id, the
     file's size in bytes, the times of last-access, last-modification
     and last-change, the attributes, and the creation time.

     When CHASE? is present and `#f', symbolic links will not be
     chased, in other words if PATH refers to a symbolic link the
     `file-info' procedure will return information about the link
     rather than the file it links to.

     For example:

          > (file-info "/dev/tty")
          #<file-info #2
             type: character-special
             device: 19513156
             inode: 20728196
             mode: 438
             number-of-links: 1
             owner: 0
             group: 0
             size: 0
             last-access-time: #<time #3>
             last-modification-time: #<time #4>
             last-change-time: #<time #5>
             attributes: 128
             creation-time: #<time #6>>


 -- procedure: file-info? OBJ
     This procedure returns `#t' when OBJ is a file-information record
     and `#f' otherwise.

     For example:

          > (file-info? (file-info "/dev/tty"))
          #t
          > (file-info? 123)
          #f


 -- procedure: file-info-type FILE-INFO
     Returns the type field of the file-information record FILE-INFO.
     The type is denoted by a symbol.  The following types are possible:

    `regular'
          Regular file

    `directory'
          Directory

    `character-special'
          Character special device

    `block-special'
          Block special device

    `fifo'
          FIFO

    `symbolic-link'
          Symbolic link

    `socket'
          Socket

    `unknown'
          File is of an unknown type

     For example:

          > (file-info-type (file-info "/dev/tty"))
          character-special
          > (file-info-type (file-info "/dev"))
          directory


 -- procedure: file-info-device FILE-INFO
     Returns the device field of the file-information record FILE-INFO.

     For example:

          > (file-info-device (file-info "/dev/tty"))
          19513156


 -- procedure: file-info-inode FILE-INFO
     Returns the inode field of the file-information record FILE-INFO.

     For example:

          > (file-info-inode (file-info "/dev/tty"))
          20728196


 -- procedure: file-info-mode FILE-INFO
     Returns the mode field of the file-information record FILE-INFO.

     For example:

          > (file-info-mode (file-info "/dev/tty"))
          438


 -- procedure: file-info-number-of-links FILE-INFO
     Returns the number-of-links field of the file-information record
     FILE-INFO.

     For example:

          > (file-info-number-of-links (file-info "/dev/tty"))
          1


 -- procedure: file-info-owner FILE-INFO
     Returns the owner field of the file-information record FILE-INFO.

     For example:

          > (file-info-owner (file-info "/dev/tty"))
          0


 -- procedure: file-info-group FILE-INFO
     Returns the group field of the file-information record FILE-INFO.

     For example:

          > (file-info-group (file-info "/dev/tty"))
          0


 -- procedure: file-info-size FILE-INFO
     Returns the size field of the file-information record FILE-INFO.

     For example:

          > (file-info-size (file-info "/dev/tty"))
          0


 -- procedure: file-info-last-access-time FILE-INFO
     Returns the last-access-time field of the file-information record
     FILE-INFO.

     For example:

          > (file-info-last-access-time (file-info "/dev/tty"))
          #<time #2>


 -- procedure: file-info-last-modification-time FILE-INFO
     Returns the last-modification-time field of the file-information
     record FILE-INFO.

     For example:

          > (file-info-last-modification-time (file-info "/dev/tty"))
          #<time #2>


 -- procedure: file-info-last-change-time FILE-INFO
     Returns the last-change-time field of the file-information record
     FILE-INFO.

     For example:

          > (file-info-last-change-time (file-info "/dev/tty"))
          #<time #2>


 -- procedure: file-info-attributes FILE-INFO
     Returns the attributes field of the file-information record
     FILE-INFO.

     For example:

          > (file-info-attributes (file-info "/dev/tty"))
          128


 -- procedure: file-info-creation-time FILE-INFO
     Returns the creation-time field of the file-information record
     FILE-INFO.

     For example:

          > (file-info-creation-time (file-info "/dev/tty"))
          #<time #2>


 -- procedure: file-type PATH
 -- procedure: file-device PATH
 -- procedure: file-inode PATH
 -- procedure: file-mode PATH
 -- procedure: file-number-of-links PATH
 -- procedure: file-owner PATH
 -- procedure: file-group PATH
 -- procedure: file-size PATH
 -- procedure: file-last-access-time PATH
 -- procedure: file-last-modification-time PATH
 -- procedure: file-last-change-time PATH
 -- procedure: file-attributes PATH
 -- procedure: file-creation-time PATH
     These procedures combine a call to the `file-info' procedure and a
     call to a file-information record field accessor.  For instance
     `(file-type PATH)' is equivalent to `(file-info-type (file-info
     PATH))'.


 -- procedure: file-last-access-and-modification-times-set! PATH [ATIME
          [MTIME]]
     This procedure changes the last-access and last-modification times
     of the file whose location is given by the string PATH.  Time is
     specified either with a time object indicating an absolute point in
     time or a real number indicating the number of seconds relative to
     the moment the procedure is called.  When ATIME and MTIME are not
     specified, the last-access and last-modification times are set to
     the current time.  When MTIME is not specified, the last-access
     and last-modification times are set to ATIME.  Otherwise the
     last-access time is set to ATIME and the last-modification time is
     set to MTIME.

     For example:

          > (define (t path)
              (list (time->seconds (file-last-access-time path))
                    (time->seconds (file-last-modification-time path))))
          > (with-output-to-file "nl.txt" newline)
          > (t "nl.txt")
          (1429547027. 1429547027.)
          > (t "nl.txt")
          (1429547027. 1429547027.)
          > (file-last-access-and-modification-times-set! "nl.txt")
          > (t "nl.txt")
          (1429547039. 1429547039.)
          > (file-last-access-and-modification-times-set! "nl.txt" -60)
          > (t "nl.txt")
          (1429547006. 1429547006.)
          > (file-last-access-and-modification-times-set! "nl.txt" -60 0)
          > (t "nl.txt")
          (1429547049. 1429547109.)


16.9 Group information
======================

 -- procedure: group-info GROUP-NAME-OR-ID
     This procedure accesses the group database to get information
     about the group identified by GROUP-NAME-OR-ID, which is the
     group's symbolic name (string) or the group's GID (exact integer).
     A group-information record is returned that contains the group's
     symbolic name, the group's id (GID), and the group's members (list
     of symbolic user names).

     For example:

          > (group-info "staff")
          #<group-info #2 name: "staff" gid: 20 members: ("root")>
          > (group-info 29)
          #<group-info #3
             name: "certusers"
             gid: 29
             members: ("root" "jabber" "postfix" "cyrusimap")>
          > (group-info 5000)
          *** ERROR IN (console)@3.1 -- Resource temporarily unavailable
          (group-info 5000)


 -- procedure: group-info? OBJ
     This procedure returns `#t' when OBJ is a group-information record
     and `#f' otherwise.

     For example:

          > (group-info? (group-info "daemon"))
          #t
          > (group-info? 123)
          #f


 -- procedure: group-info-name GROUP-INFO
     Returns the symbolic name field of the group-information record
     GROUP-INFO.

     For example:

          > (group-info-name (group-info 29))
          "certusers"


 -- procedure: group-info-gid GROUP-INFO
     Returns the group id field of the group-information record
     GROUP-INFO.

     For example:

          > (group-info-gid (group-info "staff"))
          20


 -- procedure: group-info-members GROUP-INFO
     Returns the members field of the group-information record
     GROUP-INFO.

     For example:

          > (group-info-members (group-info "staff"))
          ("root")


16.10 User information
======================

 -- procedure: user-name
     This procedure returns the user's name as a string.

     For example:

          > (user-name)
          "feeley"


 -- procedure: user-info USER-NAME-OR-ID
     This procedure accesses the user database to get information about
     the user identified by USER-NAME-OR-ID, which is the user's
     symbolic name (string) or the user's UID (exact integer).  A
     user-information record is returned that contains the user's
     symbolic name, the user's id (UID), the user's group id (GID), the
     path to the user's home directory, and the user's login shell.

     For example:

          > (user-info "feeley")
          #<user-info #2
             name: "feeley"
             uid: 506
             gid: 506
             home: "/Users/feeley"
             shell: "/bin/bash">
          > (user-info 0)
          #<user-info #3 name: "root" uid: 0 gid: 0 home: "/var/root" shell: "/bin/sh">
          > (user-info 5000)
          *** ERROR IN (console)@3.1 -- Resource temporarily unavailable
          (user-info 5000)


 -- procedure: user-info? OBJ
     This procedure returns `#t' when OBJ is a user-information record
     and `#f' otherwise.

     For example:

          > (user-info? (user-info "feeley"))
          #t
          > (user-info? 123)
          #f


 -- procedure: user-info-name USER-INFO
     Returns the symbolic name field of the user-information record
     USER-INFO.

     For example:

          > (user-info-name (user-info 0))
          "root"


 -- procedure: user-info-uid USER-INFO
     Returns the user id field of the user-information record USER-INFO.

     For example:

          > (user-info-uid (user-info "feeley"))
          506


 -- procedure: user-info-gid USER-INFO
     Returns the group id field of the user-information record
     USER-INFO.

     For example:

          > (user-info-gid (user-info "feeley"))
          506


 -- procedure: user-info-home USER-INFO
     Returns the home directory field of the user-information record
     USER-INFO.

     For example:

          > (user-info-home (user-info 0))
          "/var/root"


 -- procedure: user-info-shell USER-INFO
     Returns the shell field of the user-information record USER-INFO.

     For example:

          > (user-info-shell (user-info 0))
          "/bin/sh"


16.11 Host information
======================

 -- procedure: host-name
     This procedure returns the machine's host name as a string.

     For example:

          > (host-name)
          "mega.iro.umontreal.ca"


 -- procedure: host-info HOST-NAME
     This procedure accesses the internet host database to get
     information about the machine whose name is denoted by the string
     HOST-NAME.  A host-information record is returned that contains
     the official name of the machine, a list of aliases (alternative
     names), and a non-empty list of IP addresses for this machine.  An
     exception is raised when HOST-NAME does not appear in the database.

     For example:

          > (host-info "www.google.com")
          #<host-info #2
             name: "www.l.google.com"
             aliases: ("www.google.com")
             addresses: (#u8(66 249 85 99) #u8(66 249 85 104))>
          > (host-info "unknown.domain")
          *** ERROR IN (console)@2.1 -- Unknown host
          (host-info "unknown.domain")


 -- procedure: host-info? OBJ
     This procedure returns `#t' when OBJ is a host-information record
     and `#f' otherwise.

     For example:

          > (host-info? (host-info "www.google.com"))
          #t
          > (host-info? 123)
          #f


 -- procedure: host-info-name HOST-INFO
     Returns the official name field of the host-information record
     HOST-INFO.

     For example:

          > (host-info-name (host-info "www.google.com"))
          "www.l.google.com"


 -- procedure: host-info-aliases HOST-INFO
     Returns the aliases field of the host-information record
     HOST-INFO.  This field is a possibly empty list of strings.

     For example:

          > (host-info-aliases (host-info "www.google.com"))
          ("www.google.com")


 -- procedure: host-info-addresses HOST-INFO
     Returns the addresses field of the host-information record
     HOST-INFO.  This field is a non-empty list of u8vectors denoting
     IP addresses.

     For example:

          > (host-info-addresses (host-info "www.google.com"))
          (#u8(66 249 85 99) #u8(66 249 85 104))


 -- procedure: address-infos [`host:' HOST] [`service:' SERVICE]
          [`family:' FAMILY] [`socket-type:' SOCKET-TYPE] [`protocol:'
          PROTOCOL]
     This procedure is an interface to the `getaddrinfo' system call.
     It accesses the internet host database to get information about the
     machine whose name is denoted by the string HOST and service is
     denoted by the string SERVICE and network address family is FAMILY
     (`INET' or `INET6') and network socket-type is SOCKET-TYPE
     (`STREAM' or `DGRAM' or `RAW') and network protocol is SOCKET-TYPE
     (`TCP' or `UDP').  A list of address-information records is
     returned.

     For example:

          > (address-infos host: "ftp.at.debian.org")
          (#<address-info #2
              family: INET6
              socket-type: DGRAM
              protocol: UDP
              socket-info:
               #<socket-info #3
                  family: INET6
                  port-number: 0
                  address: #u16(8193 2136 2 1 0 0 0 16)>>
           #<address-info #4
              family: INET6
              socket-type: STREAM
              protocol: TCP
              socket-info:
               #<socket-info #5
                  family: INET6
                  port-number: 0
                  address: #u16(8193 2136 2 1 0 0 0 16)>>
           #<address-info #6
              family: INET
              socket-type: DGRAM
              protocol: UDP
              socket-info:
               #<socket-info #7
                  family: INET
                  port-number: 0
                  address: #u8(213 129 232 18)>>
           #<address-info #8
              family: INET
              socket-type: STREAM
              protocol: TCP
              socket-info:
               #<socket-info #9
                  family: INET
                  port-number: 0
                  address: #u8(213 129 232 18)>>)
          > (address-infos host: "ftp.at.debian.org"
                           family: 'INET
                           protocol: 'TCP)
          (#<address-info #10
              family: INET
              socket-type: STREAM
              protocol: TCP
              socket-info:
               #<socket-info #11
                  family: INET
                  port-number: 0
                  address: #u8(213 129 232 18)>>)
          > (address-infos host: "unknown.domain")
          *** ERROR IN (console)@5.1 -- nodename nor servname provided, or not known
          (address-infos host: "unknown.domain")


 -- procedure: address-info? OBJ
     This procedure returns `#t' when OBJ is an address-information
     record and `#f' otherwise.

     For example:

          > (map address-info?
                 (address-infos host: "ftp.at.debian.org"))
          (#t #t #t #t)
          > (address-info? 123)
          #f


 -- procedure: address-info-family ADDRESS-INFO
     Returns the family field of the address-information record
     ADDRESS-INFO.

     For example:

          > (map address-info-family
                 (address-infos host: "ftp.at.debian.org"))
          (INET6 INET6 INET INET)


 -- procedure: address-info-socket-type ADDRESS-INFO
     Returns the socket-type field of the address-information record
     ADDRESS-INFO.

     For example:

          > (map address-info-socket-type
                 (address-infos host: "ftp.at.debian.org"))
          (DGRAM STREAM DGRAM STREAM)


 -- procedure: address-info-protocol ADDRESS-INFO
     Returns the protocol field of the address-information record
     ADDRESS-INFO.

     For example:

          > (map address-info-protocol
                 (address-infos host: "ftp.at.debian.org"))
          (UDP TCP UDP TCP)


 -- procedure: address-info-socket-info ADDRESS-INFO
     Returns the socket-info field of the address-information record
     ADDRESS-INFO.

     For example:

          > (map address-info-socket-info
                 (address-infos host: "ftp.at.debian.org"))
          (#<socket-info #2
              family: INET6
              port-number: 0
              address: #u16(8193 2136 2 1 0 0 0 16)>
           #<socket-info #3
              family: INET6
              port-number: 0
              address: #u16(8193 2136 2 1 0 0 0 16)>
           #<socket-info #4
              family: INET
              port-number: 0
              address: #u8(213 129 232 18)>
           #<socket-info #5
              family: INET
              port-number: 0
              address: #u8(213 129 232 18)>)


16.12 Service information
=========================

 -- procedure: service-info SERVICE-NAME-OR-ID
     This procedure accesses the service database to get information
     about the service identified by SERVICE-NAME-OR-ID, which is the
     service's symbolic name (string) or the service's port number
     (exact integer).  A service-information record is returned that
     contains the service's symbolic name, a list of aliases
     (alternative names), the port number (exact integer), and the
     protocol name (string).  An exception is raised when
     SERVICE-NAME-OR-ID does not appear in the database.

     For example:

          > (service-info "http")
          #<service-info #2
             name: "http"
             aliases: ("www" "www-http")
             port-number: 80
             protocol: "udp">
          > (service-info 80)
          #<service-info #3
             name: "http"
             aliases: ("www" "www-http")
             port-number: 80
             protocol: "udp">


 -- procedure: service-info? OBJ
     This procedure returns `#t' when OBJ is a service-information
     record and `#f' otherwise.

     For example:

          > (service-info? (service-info "http"))
          #t
          > (service-info? 123)
          #f


 -- procedure: service-info-name SERVICE-INFO
     Returns the symbolic name field of the service-information record
     SERVICE-INFO.

     For example:

          > (service-info-name (service-info 80))
          "http"


 -- procedure: service-info-aliases SERVICE-INFO
     Returns the aliases field of the service-information record
     SERVICE-INFO.  This field is a possibly empty list of strings.

     For example:

          > (service-info-aliases (service-info "http"))
          ("www" "www-http")


 -- procedure: service-info-port-number SERVICE-INFO
     Returns the service port number field of the service-information
     record SERVICE-INFO.

     For example:

          > (service-info-port-number (service-info "http"))
          80


 -- procedure: service-info-protocol SERVICE-INFO
     Returns the service protocol name field of the service-information
     record SERVICE-INFO.

     For example:

          > (service-info-protocol (service-info "http"))
          "udp"


16.13 Protocol information
==========================

 -- procedure: protocol-info PROTOCOL-NAME-OR-ID
     This procedure accesses the protocol database to get information
     about the protocol identified by PROTOCOL-NAME-OR-ID, which is the
     protocol's symbolic name (string) or the protocol's number (exact
     integer).  A protocol-information record is returned that contains
     the protocol's symbolic name, a list of aliases (alternative
     names), and the protocol number (32 bit unsigned exact integer).
     An exception is raised when PROTOCOL-NAME-OR-ID does not appear in
     the database.

     For example:

          > (protocol-info "tcp")
          #<protocol-info #2 name: "tcp" aliases: ("TCP") number: 6>
          > (protocol-info 6)
          #<protocol-info #2 name: "tcp" aliases: ("TCP") number: 6>


 -- procedure: protocol-info? OBJ
     This procedure returns `#t' when OBJ is a protocol-information
     record and `#f' otherwise.

     For example:

          > (protocol-info? (protocol-info "tcp"))
          #t
          > (protocol-info? 123)
          #f


 -- procedure: protocol-info-name PROTOCOL-INFO
     Returns the symbolic name field of the protocol-information record
     PROTOCOL-INFO.

     For example:

          > (protocol-info-name (protocol-info 6))
          "tcp"


 -- procedure: protocol-info-aliases PROTOCOL-INFO
     Returns the aliases field of the protocol-information record
     PROTOCOL-INFO.  This field is a possibly empty list of strings.

     For example:

          > (protocol-info-aliases (protocol-info "tcp"))
          ("TCP")


 -- procedure: protocol-info-number PROTOCOL-INFO
     Returns the protocol number field of the protocol-information
     record PROTOCOL-INFO.

     For example:

          > (protocol-info-number (protocol-info "tcp"))
          6


16.14 Network information
=========================

 -- procedure: network-info NETWORK-NAME-OR-ID
     This procedure accesses the network database to get information
     about the network identified by NETWORK-NAME-OR-ID, which is the
     network's symbolic name (string) or the network's number (exact
     integer).  A network-information record is returned that contains
     the network's symbolic name, a list of aliases (alternative
     names), and the network number (32 bit unsigned exact integer).
     An exception is raised when NETWORK-NAME-OR-ID does not appear in
     the database.

     For example:

          > (network-info "loopback")
          #<network-info #2
             name: "loopback"
             aliases: ("loopback-net")
             number: 127>
          > (network-info 127)
          #<network-info #3
             name: "loopback"
             aliases: ("loopback-net")
             number: 127>


 -- procedure: network-info? OBJ
     This procedure returns `#t' when OBJ is a network-information
     record and `#f' otherwise.

     For example:

          > (network-info? (network-info "loopback"))
          #t
          > (network-info? 123)
          #f


 -- procedure: network-info-name NETWORK-INFO
     Returns the symbolic name field of the network-information record
     NETWORK-INFO.

     For example:

          > (network-info-name (network-info 127))
          "loopback"


 -- procedure: network-info-aliases NETWORK-INFO
     Returns the aliases field of the network-information record
     NETWORK-INFO.  This field is a possibly empty list of strings.

     For example:

          > (network-info-aliases (network-info "loopback"))
          ("loopback-net")


 -- procedure: network-info-number NETWORK-INFO
     Returns the network number field of the network-information record
     NETWORK-INFO.

     For example:

          > (network-info-number (network-info "loopback"))
          127


17 I/O and ports
****************

17.1 Unidirectional and bidirectional ports
===========================================

Unidirectional ports allow communication between a producer of
information and a consumer.  An input-port's producer is typically a
resource managed by the operating system (such as a file, a process or
a network connection) and the consumer is the Scheme program.  The
roles are reversed for an output-port.

   Associated with each port are settings that affect I/O operations on
that port (encoding of characters to bytes, end-of-line encoding, type
of buffering, etc).  Port settings are specified when the port is
created.  Some port settings can be changed after a port is created.

   Bidirectional ports, also called input-output-ports, allow
communication in both directions.  They are best viewed as an object
that groups two separate unidirectional ports (one in each direction).
Each direction has its own port settings and can be closed
independently from the other direction.

17.2 Port classes
=================

The four classes of ports listed below form an inheritance hierarchy.
Operations possible for a certain class of port are also possible for
the subclasses.  Only device-ports are connected to a device managed by
the operating system.  For instance it is possible to create ports that
behave as a FIFO where the Scheme program is both the producer and
consumer of information (possibly one thread is the producer and
another thread is the consumer).

  1. An "object-port" (or simply a port) provides operations to read
     and write Scheme data (i.e. any Scheme object) to/from the port.
     It also provides operations to force output to occur, to change
     the way threads block on the port, and to close the port.  Note
     that the class of objects for which write/read invariance is
     guaranteed depends on the particular class of port.

  2. A "character-port" provides all the operations of an object-port,
     and also operations to read and write individual characters to/from
     the port.  When a Scheme object is written to a character-port, it
     is converted into the sequence of characters that corresponds to
     its external-representation.  When reading a Scheme object, an
     inverse conversion occurs.  Note that some Scheme objects do not
     have an external textual representation that can be read back.

  3. A "byte-port" provides all the operations of a character-port, and
     also operations to read and write individual bytes to/from the
     port.  When a character is written to a byte-port, some encoding
     of that character into a sequence of bytes will occur (for example,
     `#\newline' will be encoded as the 2 bytes CR-LF when using
     ISO-8859-1 character encoding and `cr-lf' end-of-line encoding, and
     a non-ASCII character will generate more than 1 byte when using
     UTF-8 character encoding).  When reading a character, a similar
     decoding occurs.

  4. A "device-port" provides all the operations of a byte-port, and
     also operations to control the operating system managed device
     (file, network connection, terminal, etc) that is connected to the
     port.


17.3 Port settings
==================

Some port settings are only valid for specific port classes whereas
some others are valid for all ports.  Port settings are specified when
a port is created.  The settings that are not specified will default to
some reasonable values.  Keyword objects are used to name the settings
to be set.  As a simple example, a device-port connected to the file
`"foo"' can be created using the call

     (open-input-file "foo")

   This will use default settings for the character encoding, buffering,
etc.  If the UTF-8 character encoding is desired, then the port could
be opened using the call

     (open-input-file (list path: "foo" char-encoding: 'UTF-8))

   Here the argument of the procedure `open-input-file' has been
replaced by a "port settings list" which specifies the value of each
port setting that should not be set to the default value.  Note that
some port settings have no useful default and it is therefore required
to specify a value for them, such as the `path:' in the case of the
file opening procedures.  All port creation procedures (i.e. named
`open-...') take a single argument that can either be a port settings
list or a value of a type that depends on the kind of port being
created (a path string for files, an IP port number for socket servers,
etc).

17.4 Object-ports
=================

17.4.1 Object-port settings
---------------------------

The following is a list of port settings that are valid for all types
of ports.

   * `direction:' ( `input' | `output' | `input-output' )

     This setting controls the direction of the port.  The symbol
     `input' indicates a unidirectional input-port, the symbol `output'
     indicates a unidirectional output-port, and the symbol
     `input-output' indicates a bidirectional port.  The default value
     of this setting depends on the port creation procedure.

   * `buffering:' ( `#f' | `#t' | `line' )

     This setting controls the buffering of the port.  To set each
     direction separately the keywords `input-buffering:' and
     `output-buffering:' must be used instead of `buffering:'.  The
     value `#f' selects unbuffered I/O, the value `#t' selects fully
     buffered I/O, and the symbol `line' selects line buffered I/O (the
     output buffer is drained when a `#\newline' character is written).
     Line buffered I/O only applies to character-ports.  The default
     value of this setting is operating system dependent except
     consoles which are unbuffered.


17.4.2 Object-port operations
-----------------------------

 -- procedure: input-port? OBJ
 -- procedure: output-port? OBJ
 -- procedure: port? OBJ
     The procedure `input-port?' returns `#t' when OBJ is a
     unidirectional input-port or a bidirectional port and `#f'
     otherwise.

     The procedure `output-port?' returns `#t' when OBJ is a
     unidirectional output-port or a bidirectional port and `#f'
     otherwise.

     The procedure `port?' returns `#t' when OBJ is a port (either
     unidirectional or bidirectional) and `#f' otherwise.

     For example:

          > (input-port? (current-input-port))
          #t
          > (call-with-input-string "some text" output-port?)
          #f
          > (port? (current-output-port))
          #t


 -- procedure: read [PORT]
     This procedure reads and returns the next Scheme datum from the
     input-port PORT.  The end-of-file object is returned when the end
     of the stream is reached.  If it is not specified, PORT defaults
     to the current input-port.

     For example:

          > (call-with-input-string "some text" read)
          some
          > (call-with-input-string "" read)
          #!eof


 -- procedure: read-all [PORT [READER]]
     This procedure repeatedly calls the procedure READER with PORT as
     the sole argument and accumulates a list of each value returned up
     to the end-of-file object.  The procedure `read-all' returns the
     accumulated list without the end-of-file object.  If it is not
     specified, PORT defaults to the current input-port.  If it is not
     specified, READER defaults to the procedure `read'.

     For example:

          > (call-with-input-string "3,2,1\ngo!" read-all)
          (3 ,2 ,1 go!)
          > (call-with-input-string "3,2,1\ngo!"
                                    (lambda (p) (read-all p read-char)))
          (#\3 #\, #\2 #\, #\1 #\newline #\g #\o #\!)
          > (call-with-input-string "3,2,1\ngo!"
                                    (lambda (p) (read-all p read-line)))
          ("3,2,1" "go!")


 -- procedure: write OBJ [PORT]
     This procedure writes the Scheme datum OBJ to the output-port PORT
     and the value returned is unspecified.  If it is not specified,
     PORT defaults to the current output-port.

     For example:

          > (write (list 'compare (list 'quote '@x) 'and (list 'unquote '@x)))
          (compare '@x and , @x)>


 -- procedure: newline [PORT]
     This procedure writes an "object separator" to the output-port
     PORT and the value returned is unspecified.  The separator ensures
     that the next Scheme datum written with the `write' procedure will
     not be confused with the latest datum that was written.  On
     character-ports this is done by writing the character `#\newline'.
     On ports where successive objects are implicitly distinct (such
     as "vector ports") this procedure does nothing.

     Regardless of the class of a port P and assuming that the external
     textual representation of the object X is readable, the expression
     `(begin (write X P) (newline P))' will write to P a representation
     of X that can be read back with the procedure `read'.  If it is
     not specified, PORT defaults to the current output-port.

     For example:

          > (begin (write 123) (newline) (write 456) (newline))
          123
          456


 -- procedure: force-output [PORT [LEVEL]]
     The procedure `force-output' causes the data that was written to
     the output-port PORT to be moved closer to its destination
     according to LEVEL, an exact integer in the range 0 to 2.  If PORT
     is not specified, the current output-port is used.  If LEVEL is
     not specified, it defaults to 0.  Values of LEVEL above 0 are
     equivalent to LEVEL = 0 except for device ports as explained below.

     When LEVEL is 0, the output buffers of PORT which are managed in
     the Scheme process are drained (i.e.  the output operation that
     was delayed due to buffering is actually performed).  In the case
     of a device port the data is passed to the operating system and it
     becomes its responsibility to transmit the data to the device.  The
     operating system may implement its own buffering approach which
     delays the transmission of the data to the device.

     When LEVEL is 1, in addition to the operations for LEVEL = 0 and
     if the operating system supports the functionality, the operating
     system is requested to transmit the data to the device.  On UNIX
     this corresponds to a `fsync' system call.

     When LEVEL is 2, in addition to the operations for LEVEL = 1 and
     if the operating system supports the functionality, the operating
     system is requested to wait until the device reports that the data
     was saved by the device (e.g. actually written to disk in the case
     of a file).  This operation can take a long time on some operating
     systems.  On Mac OS X this corresponds to a `fcntl' system call
     with operation `F_FULLFSYNC'.

     For example:

          > (define p (open-tcp-client "www.iro.umontreal.ca:80"))
          > (display "GET /\n" p)
          > (force-output p)
          > (read-line p)
          "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\""


 -- procedure: close-input-port PORT
 -- procedure: close-output-port PORT
 -- procedure: close-port PORT
     The PORT argument of these procedures must be a unidirectional or
     a bidirectional port.  For all three procedures the value returned
     is unspecified.

     The procedure `close-input-port' closes the input-port side of
     PORT, which must not be a unidirectional output-port.

     The procedure `close-output-port' closes the output-port side of
     PORT, which must not be a unidirectional input-port.  The ouput
     buffers are drained before PORT is closed.

     The procedure `close-port' closes all sides of the PORT.  Unless
     PORT is a unidirectional input-port, the output buffers are
     drained before PORT is closed.

     For example:

          > (define p (open-tcp-client "www.iro.umontreal.ca:80"))
          > (display "GET /\n" p)
          > (close-output-port p)
          > (read-line p)
          "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\""


 -- procedure: input-port-timeout-set! PORT TIMEOUT [THUNK]
 -- procedure: output-port-timeout-set! PORT TIMEOUT [THUNK]
     When a thread tries to perform an I/O operation on a port, the
     requested operation may not be immediately possible and the thread
     must wait.  For example, the thread may be trying to read a line of
     text from the console and the user has not typed anything yet, or
     the thread may be trying to write to a network connection faster
     than the network can handle.  In such situations the thread
     normally blocks until the operation becomes possible.

     It is sometimes necessary to guarantee that the thread will not
     block too long.  For this purpose, to each input-port and
     output-port is attached a "timeout" and "timeout-thunk".  The
     timeout indicates the point in time beyond which the thread should
     stop waiting on an input and output operation respectively.  When
     the timeout is reached, the thread calls the port's timeout-thunk.
     If the timeout-thunk returns `#f' the thread abandons trying to
     perform the operation (in the case of an input operation an
     end-of-file is read and in the case of an output operation an
     exception is raised).  Otherwise, the thread will block again
     waiting for the operation to become possible (note that if the
     port's timeout has not changed the thread will immediately call
     the timeout-thunk again).

     The procedure `input-port-timeout-set!' sets the timeout of the
     input-port PORT to TIMEOUT and the timeout-thunk to THUNK.  The
     procedure `output-port-timeout-set!' sets the timeout of the
     output-port PORT to TIMEOUT and the timeout-thunk to THUNK.  If it
     is not specified, the THUNK defaults to a thunk that returns `#f'.
     The TIMEOUT is either a time object indicating an absolute point
     in time, or it is a real number which indicates the number of
     seconds relative to the moment the procedure is called.  For both
     procedures the value returned is unspecified.

     When a port is created the timeout is set to infinity (`+inf.0').
     This causes the thread to wait as long as needed for the operation
     to become possible.  Setting the timeout to a point in the past
     (`-inf.0') will cause the thread to attempt the I/O operation and
     never block (i.e. the timeout-thunk is called if the operation is
     not immediately possible).

     The following example shows how to cause the REPL to terminate
     when the user does not enter an expression within the next 60
     seconds.

          > (input-port-timeout-set! (repl-input-port) 60)
          >
          *** EOF again to exit


17.5 Character-ports
====================

17.5.1 Character-port settings
------------------------------

The following is a list of port settings that are valid for
character-ports.

   * `readtable:' READTABLE

     This setting determines the readtable attached to the
     character-port.  To set each direction separately the keywords
     `input-readtable:' and `output-readtable:' must be used instead of
     `readtable:'.  Readtables control the external textual
     representation of Scheme objects, that is the encoding of Scheme
     objects using characters.  The behavior of the `read' procedure
     depends on the port's input-readtable and the behavior of the
     procedures `write', `pretty-print', and related procedures is
     affected by the port's output-readtable.  The default value of this
     setting is the value bound to the parameter object
     `current-readtable'.

   * `output-width:' POSITIVE-INTEGER

     This setting indicates the width of the character output-port in
     number of characters.  This information is used by the
     pretty-printer.  The default value of this setting is 80.


17.5.2 Character-port operations
--------------------------------

 -- procedure: input-port-line PORT
 -- procedure: input-port-column PORT
 -- procedure: output-port-line PORT
 -- procedure: output-port-column PORT
     The current character location of a character input-port is the
     location of the next character to read.  The current character
     location of a character output-port is the location of the next
     character to write.  Location is denoted by a line number (the
     first line is line 1) and a column number, that is the location on
     the current line (the first column is column 1).  The procedures
     `input-port-line' and `input-port-column' return the line location
     and the column location respectively of the character input-port
     PORT.  The procedures `output-port-line' and `output-port-column'
     return the line location and the column location respectively of
     the character output-port PORT.

     For example:

          > (call-with-output-string
              '()
              (lambda (p)
                (display "abc\n123def" p)
                (write (list (output-port-line p) (output-port-column p))
                       p)))
          "abc\n123def(2 7)"


 -- procedure: output-port-width PORT
     This procedure returns the width, in characters, of the character
     output-port PORT.  The value returned is the port's output-width
     setting.

     For example:

          > (output-port-width (repl-output-port))
          80


 -- procedure: read-char [PORT]
     This procedure reads the character input-port PORT and returns the
     character at the current character location and advances the
     current character location to the next character, unless the PORT
     is already at end-of-file in which case `read-char' returns the
     end-of-file object.  If it is not specified, PORT defaults to the
     current input-port.

     For example:

          > (call-with-input-string
              "some text"
              (lambda (p)
                (let ((a (read-char p))) (list a (read-char p)))))
          (#\s #\o)
          > (call-with-input-string "" read-char)
          #!eof


 -- procedure: peek-char [PORT]
     This procedure returns the same result as `read-char' but it does
     not advance the current character location of the input-port PORT.
     If it is not specified, PORT defaults to the current input-port.

     For example:

          > (call-with-input-string
              "some text"
              (lambda (p)
                (let ((a (peek-char p))) (list a (read-char p)))))
          (#\s #\s)
          > (call-with-input-string "" peek-char)
          #!eof


 -- procedure: write-char CHAR [PORT]
     This procedure writes the character CHAR to the character
     output-port PORT and advances the current character location of
     that output-port.  The value returned is unspecified.  If it is not
     specified, PORT defaults to the current output-port.

     For example:

          > (write-char #\=)
          =>


 -- procedure: read-line [PORT [SEPARATOR [INCLUDE-SEPARATOR?
          [MAX-LENGTH]]]]
     This procedure reads characters from the character input-port PORT
     until a specific SEPARATOR or the end-of-file is encountered and
     returns a string containing the sequence of characters read.  If
     it is specified, MAX-LENGTH must be a nonnegative exact integer
     and it places an upper limit on the number of characters that are
     read.

     The SEPARATOR is included at the end of the string only if it was
     the last character read and INCLUDE-SEPARATOR? is not `#f'.  The
     SEPARATOR must be a character or `#f' (in which case all the
     characters until the end-of-file are read).  If it is not
     specified, PORT defaults to the current input-port.  If it is not
     specified, SEPARATOR defaults to `#\newline'.  If it is not
     specified, INCLUDE-SEPARATOR? defaults to `#f'.

     For example:

          > (define (split sep)
              (lambda (str)
                (call-with-input-string
                  str
                  (lambda (p)
                    (read-all p (lambda (p) (read-line p sep)))))))
          > ((split #\,) "a,b,c")
          ("a" "b" "c")
          > (map (split #\,)
                 (call-with-input-string "1,2,3\n4,5"
                                         (lambda (p) (read-all p read-line))))
          (("1" "2" "3") ("4" "5"))
          > (read-line (current-input-port) #\newline #f 2)1234
          "12"
          > 34


 -- procedure: read-substring STRING START END [PORT [NEED]]
 -- procedure: write-substring STRING START END [PORT]
     These procedures support bulk character I/O.  The part of the
     string STRING starting at index START and ending just before index
     END is used as a character buffer that will be the target of
     `read-substring' or the source of the `write-substring'.  The
     `read-substring' also accepts a NEED parameter which must be a
     nonnegative fixnum.  Up to END-START characters will be
     transferred.  The number of characters transferred, possibly zero,
     is returned by these procedures.  Fewer characters will be read by
     `read-substring' if an end-of-file is read, or a timeout occurs
     before all the requested characters are transferred and the
     timeout thunk returns `#f' (see the procedure
     `input-port-timeout-set!'), or NEED is specified and at least that
     many characters have been read (in other words the procedure does
     not block for more characters but may transfer more characters if
     they are immediately available).  Fewer characters will be written
     by `write-substring' if a timeout occurs before all the requested
     characters are transferred and the timeout thunk returns `#f' (see
     the procedure `output-port-timeout-set!').  If it is not
     specified, PORT defaults to the current input-port and current
     output-port respectively.

     For example:

          > (define s (make-string 10 #\x))
          > (read-substring s 2 5)123456789
          3
          > 456789
          > s
          "xx123xxxxx"
          > (read-substring s 2 10 (current-input-port) 3)abcd
          5
          > s
          "xxabcd\nxxx"


 -- procedure: input-port-readtable PORT
 -- procedure: output-port-readtable PORT
     These procedures return the readtable attached to the
     character-port PORT.  The PORT parameter of `input-port-readtable'
     must be an input-port.  The PORT parameter of
     `output-port-readtable' must be an output-port.


 -- procedure: input-port-readtable-set! PORT READTABLE
 -- procedure: output-port-readtable-set! PORT READTABLE
     These procedures change the readtable attached to the
     character-port PORT to the readtable READTABLE.  The PORT parameter
     of `input-port-readtable-set!'  must be an input-port.  The PORT
     parameter of `output-port-readtable-set!' must be an output-port.
     The value returned is unspecified.


17.6 Byte-ports
===============

17.6.1 Byte-port settings
-------------------------

The following is a list of port settings that are valid for byte-ports.

   * `char-encoding:' ENCODING

     This setting controls the character encoding of the byte-port.  For
     bidirectional byte-ports, the character encoding for input and
     output is set.  To set each direction separately the keywords
     `input-char-encoding:' and `output-char-encoding:' must be used
     instead of `char-encoding:'.  The default value of this setting is
     operating system dependent, but this can be overridden through the
     runtime options (*note Runtime options::).  The following
     encodings are supported:

    `ISO-8859-1'
          ISO-8859-1 character encoding.  Each character is encoded by
          a single byte.  Only Unicode characters with a code in the
          range 0 to 255 are allowed.

    `ASCII'
          ASCII character encoding.  Each character is encoded by a
          single byte.  In principle only Unicode characters with a
          code in the range 0 to 127 are allowed but most types of
          ports treat this exactly like `ISO-8859-1'.

    `UTF-8'
          UTF-8 character encoding.  Each character is encoded by a
          sequence of one to four bytes.  The minimum length UTF-8
          encoding is used.  If a BOM is needed at the beginning of the
          stream then it must be explicitly written.

    `UTF-16'
          UTF-16 character encoding.  Each character is encoded by one
          or two 16 bit integers (2 or 4 bytes).  The 16 bit integers
          may be encoded using little-endian encoding or big-endian
          encoding.  If the port is an input-port and the first two
          bytes read are a BOM ("Byte Order Mark" character with
          hexadecimal code FEFF) then the BOM will be discarded and the
          endianness will be set accordingly, otherwise the endianness
          depends on the operating system and how the Gambit runtime was
          compiled.  If the port is an output-port then a BOM will be
          output at the beginning of the stream and the endianness
          depends on the operating system and how the Gambit runtime
          was compiled.

    `UTF-16LE'
          UTF-16 character encoding with little-endian endianness.  It
          is like `UTF-16' except the endianness is set to
          little-endian and there is no BOM processing.  If a BOM is
          needed at the beginning of the stream then it must be
          explicitly written.

    `UTF-16BE'
          UTF-16 character encoding with big-endian endianness.  It is
          like `UTF-16LE' except the endianness is set to big-endian.

    `UTF / UTF-fallback-ASCII / UTF-fallback-ISO-8859-1 / UTF-fallback-UTF-16 / UTF-fallback-UTF-16LE / UTF-fallback-UTF-16BE'
          These encodings combine the UTF-8 and UTF-16 encodings.  When
          one of these character encodings is used for an output port,
          characters will be encoded using the UTF-8 encoding.  The
          first character, if there is one, is prefixed with a UTF-8
          BOM (the three byte sequence EF BB BF in hexadecimal).  When
          one of these character encodings is used for an input port,
          the character encoding depends on the first few bytes.  If
          the first bytes of the stream are a UTF-16LE BOM (FF FE in
          hexadecimal), or a UTF-16BE BOM (FE FF in hexadecimal), or a
          UTF-8 BOM (EF BB BF in hexadecimal), then the BOM is
          discarded and the remaining bytes of the stream are decoded
          using the corresponding character encoding.  If a BOM is not
          present, then the stream is decoded using the fallback
          encoding specified.  The encoding `UTF' is a synonym for
          `UTF-fallback-UTF-8'.  Note that the `UTF' character encoding
          for input will correctly handle streams produced using the
          encodings `UTF', `UTF-8', `UTF-16', `ASCII', and if an
          explicit BOM is output, the encodings `UTF-16LE', and
          `UTF-16BE'.

    `UCS-2'
          UCS-2 character encoding.  Each character is encoded by a 16
          bit integer (2 bytes).  The 16 bit integers may be encoded
          using little-endian encoding or big-endian encoding.  If the
          port is an input-port and the first two bytes read are a BOM
          ("Byte Order Mark" character with hexadecimal code FEFF) then
          the BOM will be discarded and the endianness will be set
          accordingly, otherwise the endianness depends on the
          operating system and how the Gambit runtime was compiled.  If
          the port is an output-port then a BOM will be output at the
          beginning of the stream and the endianness depends on the
          operating system and how the Gambit runtime was compiled.

    `UCS-2LE'
          UCS-2 character encoding with little-endian endianness.  It
          is like `UCS-2' except the endianness is set to little-endian
          and there is no BOM processing.  If a BOM is needed at the
          beginning of the stream then it must be explicitly written.

    `UCS-2BE'
          UCS-2 character encoding with big-endian endianness.  It is
          like `UCS-2LE' except the endianness is set to big-endian.

    `UCS-4'
          UCS-4 character encoding.  Each character is encoded by a 32
          integer (4 bytes).  The 32 bit integers may be encoded using
          little-endian encoding or big-endian encoding.  If the port
          is an input-port and the first four bytes read are a BOM
          ("Byte Order Mark" character with hexadecimal code 0000FEFF)
          then the BOM will be discarded and the endianness will be set
          accordingly, otherwise the endianness depends on the
          operating system and how the Gambit runtime was compiled.  If
          the port is an output-port then a BOM will be output at the
          beginning of the stream and the endianness depends on the
          operating system and how the Gambit runtime was compiled.

    `UCS-4LE'
          UCS-4 character encoding with little-endian endianness.  It
          is like `UCS-4' except the endianness is set to little-endian
          and there is no BOM processing.  If a BOM is needed at the
          beginning of the stream then it must be explicitly written.

    `UCS-4BE'
          UCS-4 character encoding with big-endian endianness.  It is
          like `UCS-4LE' except the endianness is set to big-endian.


   * `char-encoding-errors:' ( `#f' | `#t' )

     This setting controls whether illegal character encodings are
     silently replaced with the Unicode character #xfffd (replacement
     character) or raise an error.  To set each direction separately
     the keywords `input-char-encoding-errors:' and
     `output-char-encoding-errors:' must be used instead of
     `char-encoding-errors:'.  The default value of this setting is
     `#t'.

   * `eol-encoding:' ENCODING

     This setting controls the end-of-line encoding of the byte-port.
     To set each direction separately the keywords `input-eol-encoding:'
     and `output-eol-encoding:' must be used instead of
     `eol-encoding:'.  The default value of this setting is operating
     system dependent, but this can be overridden through the runtime
     options (*note Runtime options::).  Note that for output-ports the
     end-of-line encoding is applied before the character encoding, and
     for input-ports it is applied after.  The following encodings are
     supported:

    `lf'
          For an output-port, writing a `#\newline' character outputs a
          `#\linefeed' character to the stream (Unicode character code
          10).  For an input-port, a `#\newline' character is read when
          a `#\linefeed' character is encountered on the stream.  Note
          that `#\linefeed' and `#\newline' are two names for the same
          character, so this end-of-line encoding is actually the
          identity function.  Text files created by UNIX applications
          typically use this end-of-line encoding.

    `cr'
          For an output-port, writing a `#\newline' character outputs a
          `#\return' character to the stream (Unicode character code
          13).  For an input-port, a `#\newline' character is read when
          a `#\linefeed' character or a `#\return' character is
          encountered on the stream.  Text files created by Classic Mac
          OS applications typically use this end-of-line encoding.

    `cr-lf'
          For an output-port, writing a `#\newline' character outputs to
          the stream a `#\return' character followed by a `#\linefeed'
          character.  For an input-port, a `#\newline' character is read
          when a `#\linefeed' character or a `#\return' character is
          encountered on the stream.  Moreover, if this character is
          immediately followed by the opposite character (`#\linefeed'
          followed by `#\return' or `#\return' followed by
          `#\linefeed') then the second character is ignored.  In other
          words, all four possible end-of-line encodings are read as a
          single `#\newline' character.  Text files created by DOS and
          Microsoft Windows applications typically use this end-of-line
          encoding.



17.6.2 Byte-port operations
---------------------------

 -- procedure: read-u8 [PORT]
     This procedure reads the byte input-port PORT and returns the byte
     at the current byte location and advances the current byte
     location to the next byte, unless the PORT is already at
     end-of-file in which case `read-u8' returns the end-of-file
     object.  If it is not specified, PORT defaults to the current
     input-port.

     This procedure must be called when the port's input character
     buffer is empty otherwise the character-stream and byte-stream may
     be out of sync due to buffering.  The input character buffer is
     used for bulk decoding of encoded characters (i.e. to translate
     the byte-stream into a character-stream).  The input character
     buffer is initially empty.  It is only when characters are read
     that it is filled with characters obtained by decoding the
     byte-stream.

     One way to ensure that the port's input character buffer is empty
     is to call `read-u8' strictly before any use of the port in a
     character input operation (i.e. a call to the procedures `read',
     `read-char', `peek-char', etc).  Alternatively
     `input-port-characters-buffered' can be used to get the number of
     characters in the port's input character buffer, and to empty the
     buffer with calls to `read-char' or `read-substring'.

     For example:

          > (call-with-input-u8vector
              '#u8(11 22 33 44)
              (lambda (p)
                (let ((a (read-u8 p))) (list a (read-u8 p)))))
          (11 22)
          > (call-with-input-u8vector '#u8() read-u8)
          #!eof


 -- procedure: write-u8 N [PORT]
     This procedure writes the byte N to the byte output-port PORT and
     advances the current byte location of that output-port.  The value
     returned is unspecified.  If it is not specified, PORT defaults to
     the current output-port.

     For example:

          > (call-with-output-u8vector '() (lambda (p) (write-u8 33 p)))
          #u8(33)


 -- procedure: read-subu8vector U8VECTOR START END [PORT [NEED]]
 -- procedure: write-subu8vector U8VECTOR START END [PORT]
     These procedures support bulk byte I/O.  The part of the u8vector
     U8VECTOR starting at index START and ending just before index END
     is used as a byte buffer that will be the target of
     `read-subu8vector' or the source of the `write-subu8vector'.  The
     `read-subu8vector' also accepts a NEED parameter which must be a
     nonnegative fixnum.  Up to END-START bytes will be transferred.
     The number of bytes transferred, possibly zero, is returned by
     these procedures.  Fewer bytes will be read by `read-subu8vector'
     if an end-of-file is read, or a timeout occurs before all the
     requested bytes are transferred and the timeout thunk returns `#f'
     (see the procedure `input-port-timeout-set!'), or NEED is
     specified and at least that many bytes have been read (in other
     words the procedure does not block for more bytes but may transfer
     more bytes if they are immediately available).  Fewer bytes will
     be written by `write-subu8vector' if a timeout occurs before all
     the requested bytes are transferred and the timeout thunk returns
     `#f' (see the procedure `output-port-timeout-set!').  If it is not
     specified, PORT defaults to the current input-port and current
     output-port respectively.

     The procedure `read-subu8vector' must be called before any use of
     the port in a character input operation (i.e. a call to the
     procedures `read', `read-char', `peek-char', etc) because
     otherwise the character-stream and byte-stream may be out of sync
     due to the port buffering.

     For example:

          > (define v (make-u8vector 10))
          > (read-subu8vector v 2 5)123456789
          3
          > 456789
          > v
          #u8(0 0 49 50 51 0 0 0 0 0)
          > (read-subu8vector v 2 10 (current-input-port) 3)abcd
          5
          > v
          #u8(0 0 97 98 99 100 10 0 0 0)


17.7 Device-ports
=================

17.7.1 Filesystem devices
-------------------------

 -- procedure: open-file PATH-OR-SETTINGS
 -- procedure: open-input-file PATH-OR-SETTINGS
 -- procedure: open-output-file PATH-OR-SETTINGS
 -- procedure: call-with-input-file PATH-OR-SETTINGS PROC
 -- procedure: call-with-output-file PATH-OR-SETTINGS PROC
 -- procedure: with-input-from-file PATH-OR-SETTINGS THUNK
 -- procedure: with-output-to-file PATH-OR-SETTINGS THUNK
     All of these procedures create a port to interface to a byte-stream
     device (such as a file, console, serial port, named pipe, etc)
     whose name is given by a path of the filesystem.  The `direction:'
     setting will default to the value `input' for the procedures
     `open-input-file', `call-with-input-file' and
     `with-input-from-file', to the value `output' for the procedures
     `open-output-file', `call-with-output-file' and
     `with-output-to-file', and to the value `input-output' for the
     procedure `open-file'.

     The procedures `open-file', `open-input-file' and
     `open-output-file' return the port that is created.  The
     procedures `call-with-input-file' and `call-with-output-file' call
     the procedure PROC with the port as single argument, and then
     return the value(s) of this call after closing the port.  The
     procedures `with-input-from-file' and `with-output-to-file'
     dynamically bind the current input-port and current output-port
     respectively to the port created for the duration of a call to the
     procedure THUNK with no argument.  The value(s) of the call to
     THUNK are returned after closing the port.

     The first argument of these procedures is either a string denoting
     a filesystem path or a list of port settings which must contain a
     `path:' setting.  Here are the settings allowed in addition to the
     generic settings of byte-ports:

        * `path:' STRING

          This setting indicates the location of the file in the
          filesystem.  There is no default value for this setting.

        * `append:' ( `#f' | `#t' )

          This setting controls whether output will be added to the end
          of the file.  This is useful for writing to log files that
          might be open by more than one process.  The default value of
          this setting is `#f'.

        * `create:' ( `#f' | `#t' | `maybe' )

          This setting controls whether the file will be created when
          it is opened.  A setting of `#f' requires that the file exist
          (otherwise an exception is raised).  A setting of `#t'
          requires that the file does not exist (otherwise an exception
          is raised).  A setting of `maybe' will create the file if it
          does not exist.  The default value of this setting is `maybe'
          for output-ports and `#f' for input-ports and bidirectional
          ports.

        * `permissions:' 12-BIT-EXACT-INTEGER

          This setting controls the UNIX permissions that will be
          attached to the file if it is created.  The default value of
          this setting is `#o666'.

        * `truncate:' ( `#f' | `#t' )

          This setting controls whether the file will be truncated when
          it is opened.  For input-ports, the default value of this
          setting is `#f'.  For output-ports, the default value of this
          setting is `#t' when the `append:' setting is `#f', and `#f'
          otherwise.


     For example:

          > (with-output-to-file
              (list path: "nofile"
                    create: #f)
              (lambda ()
                (display "hello world!\n")))
          *** ERROR IN (console)@1.1 -- No such file or directory
          (with-output-to-file '(path: "nofile" create: #f) '#<procedure #2>)


 -- procedure: input-port-byte-position PORT [POSITION [WHENCE]]
 -- procedure: output-port-byte-position PORT [POSITION [WHENCE]]
     When called with a single argument these procedures return the byte
     position where the next I/O operation would take place in the file
     attached to the given PORT (relative to the beginning of the
     file).  When called with two or three arguments, the byte position
     for subsequent I/O operations on the given PORT is changed to
     POSITION, which must be an exact integer.  When WHENCE is omitted
     or is 0, the POSITION is relative to the beginning of the file.
     When WHENCE is 1, the POSITION is relative to the current byte
     position of the file.  When WHENCE is 2, the POSITION is relative
     to the end of the file.  The return value is the new byte
     position.  On most operating systems the byte position for reading
     and writing of a given bidirectional port are the same.

     When `input-port-byte-position' is called to change the byte
     position of an input-port, all input buffers will be flushed so
     that the next byte read will be the one at the given position.

     When `output-port-byte-position' is called to change the byte
     position of an output-port, there is an implicit call to
     `force-output' before the position is changed.

     For example:

          > (define p  ; p is an input-output-port
              (open-file '(path: "test" char-encoding: ISO-8859-1 create: maybe)))
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (0 0)
          > (display "abcdefghij\n" p)
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (0 0)
          > (force-output p)
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (11 11)
          > (input-port-byte-position p 2)
          2
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (2 2)
          > (peek-char p)
          #\c
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (11 11)
          > (output-port-byte-position p -7 2)
          4
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (4 4)
          > (write-char #\! p)
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (4 4)
          > (force-output p)
          > (list (input-port-byte-position p) (output-port-byte-position p))
          (5 5)
          > (input-port-byte-position p 1)
          1
          > (read p)
          bcd!fghij


17.7.2 Process devices
----------------------

 -- procedure: open-process PATH-OR-SETTINGS
 -- procedure: open-input-process PATH-OR-SETTINGS
 -- procedure: open-output-process PATH-OR-SETTINGS
 -- procedure: call-with-input-process PATH-OR-SETTINGS PROC
 -- procedure: call-with-output-process PATH-OR-SETTINGS PROC
 -- procedure: with-input-from-process PATH-OR-SETTINGS THUNK
 -- procedure: with-output-to-process PATH-OR-SETTINGS THUNK
     All of these procedures start a new operating system process and
     create a bidirectional port which allows communication with that
     process on its standard input and standard output.  The
     `direction:' setting will default to the value `input' for the
     procedures `open-input-process', `call-with-input-process' and
     `with-input-from-process', to the value `output' for the
     procedures `open-output-process', `call-with-output-process' and
     `with-output-to-process', and to the value `input-output' for the
     procedure `open-process'.  If the `direction:' setting is `input',
     the output-port side is closed.  If the `direction:' setting is
     `output', the input-port side is closed.

     The procedures `open-process', `open-input-process' and
     `open-output-process' return the port that is created.  The
     procedures `call-with-input-process' and
     `call-with-output-process' call the procedure PROC with the port
     as single argument, and then return the value(s) of this call
     after closing the port and waiting for the process to terminate.
     The procedures `with-input-from-process' and
     `with-output-to-process' dynamically bind the current input-port
     and current output-port respectively to the port created for the
     duration of a call to the procedure THUNK with no argument.  The
     value(s) of the call to THUNK are returned after closing the port
     and waiting for the process to terminate.

     The first argument of this procedure is either a string denoting a
     filesystem path of an executable program or a list of port settings
     which must contain a `path:' setting.  Here are the settings
     allowed in addition to the generic settings of byte-ports:

        * `path:' STRING

          This setting indicates the location of the executable program
          in the filesystem.  There is no default value for this
          setting.

        * `arguments:' LIST-OF-STRINGS

          This setting indicates the string arguments that are passed
          to the program.  The default value of this setting is the
          empty list (i.e. no arguments).

        * `environment:' LIST-OF-STRINGS

          This setting indicates the set of environment variable
          bindings that the process receives.  Each element of the list
          is a string of the form "`VAR=VALUE'", where `VAR' is the
          name of the variable and `VALUE' is its binding.  When
          LIST-OF-STRINGS is `#f', the process inherits the environment
          variable bindings of the Scheme program.  The default value
          of this setting is `#f'.

        * `directory:' DIR

          This setting indicates the current working directory of the
          process.  When DIR is `#f', the process uses the value of
          `(current-directory)'.  The default value of this setting is
          `#f'.

        * `stdin-redirection:' ( `#f' | `#t' )

          This setting indicates how the standard input of the process
          is redirected.  A setting of `#t' will redirect the standard
          input from the process-port (i.e. what is written to the
          process-port will be available on the standard input).  A
          setting of `#f' will leave the standard input as-is, which
          typically results in input coming from the console.  The
          default value of this setting is `#t'.

        * `stdout-redirection:' ( `#f' | `#t' )

          This setting indicates how the standard output of the process
          is redirected.  A setting of `#t' will redirect the standard
          output to the process-port (i.e. all output to standard
          output can be read from the process-port).  A setting of `#f'
          will leave the standard output as-is, which typically results
          in the output going to the console.  The default value of
          this setting is `#t'.

        * `stderr-redirection:' ( `#f' | `#t' )

          This setting indicates how the standard error of the process
          is redirected.  A setting of `#t' will redirect the standard
          error to the process-port (i.e. all output to standard error
          can be read from the process-port).  A setting of `#f' will
          leave the standard error as-is, which typically results in
          error messages being output to the console.  The default
          value of this setting is `#f'.

        * `pseudo-terminal:' ( `#f' | `#t' )

          This setting applies to UNIX.  It indicates what type of
          device will be bound to the process' standard input and
          standard output.  A setting of `#t' will use a
          pseudo-terminal device (this is a device that behaves like a
          tty device even though there is no real terminal or user
          directly involved).  A setting of `#f' will use a pair of
          pipes.  The difference is important for programs which behave
          differently when they are used interactively, for example
          shells.  The default value of this setting is `#f'.

        * `show-console:' ( `#f' | `#t' )

          This setting applies to Microsoft Windows.  It controls
          whether the process' console window will be hidden or
          visible.  The default value of this setting is `#t' (i.e.
          show the console window).


     For example:

          > (with-input-from-process "date" read-line)
          "Sun Jun 14 15:06:41 EDT 2009"
          > (define p (open-process (list path: "ls"
                                          arguments: '("../examples"))))
          > (read-line p)
          "README"
          > (read-line p)
          "Xlib-simple"
          > (close-port p)
          > (define p (open-process "/usr/bin/dc"))
          > (display "2 100 ^ p\n" p)
          > (force-output p)
          > (read-line p)
          "1267650600228229401496703205376"


 -- procedure: process-pid PROCESS-PORT
     This procedure returns the PID (Process Identifier) of the process
     of PROCESS-PORT.  The PID is a small exact integer.

     For example:

          > (let ((p (open-process "sort")))
              (process-pid p))
          318


 -- procedure: process-status PROCESS-PORT [TIMEOUT [TIMEOUT-VAL]]
     This procedure causes the current thread to wait until the process
     of PROCESS-PORT terminates (normally or not) or until the timeout
     is reached if TIMEOUT is supplied.  If the timeout is reached,
     PROCESS-STATUS returns TIMEOUT-VAL if it is supplied, otherwise an
     unterminated-process-exception object is raised.  The procedure
     returns the process exit status as encoded by the operating
     system.  Typically, if the process exited normally the return
     value is the process exit status multiplied by 256.

     For example:

          > (let ((p (open-process "sort")))
              (for-each (lambda (x) (pretty-print x p))
                        '(22 11 33))
              (close-output-port p)
              (let ((r (read-all p)))
                (close-input-port p)
                (list (process-status p) r)))
          (0 (11 22 33))


 -- procedure: unterminated-process-exception? OBJ
 -- procedure: unterminated-process-exception-procedure EXC
 -- procedure: unterminated-process-exception-arguments EXC
     Unterminated-process-exception objects are raised when a call to
     the `process-status' procedure reaches its timeout before the
     target process terminates and a timeout-value parameter is not
     specified.  The parameter EXC must be an
     unterminated-process-exception object.

     The procedure `unterminated-process-exception?' returns `#t' when
     OBJ is an unterminated-process-exception object and `#f' otherwise.

     The procedure `unterminated-process-exception-procedure' returns
     the procedure that raised EXC.

     The procedure `unterminated-process-exception-arguments' returns
     the list of arguments of the procedure that raised EXC.

     For example:

          > (define (handler exc)
              (if (unterminated-process-exception? exc)
                  (list (unterminated-process-exception-procedure exc)
                        (unterminated-process-exception-arguments exc))
                  'not-unterminated-process-exception))
          > (with-exception-catcher
              handler
              (lambda ()
                (let ((p (open-process "sort")))
                  (process-status p 1))))
          (#<procedure #2 process-status> (#<input-output-port #3 (process "sort")>))


17.7.3 Network devices
----------------------

 -- procedure: open-tcp-client PORT-NUMBER-OR-ADDRESS-OR-SETTINGS
     This procedure opens a network connection to a socket server and
     returns a tcp-client-port (a subtype of device-port) that
     represents this connection and allows communication with that
     server.  The default value of the `direction:' setting is
     `input-output', i.e. the Scheme program can send information to
     the server and receive information from the server.  The sending
     direction can be "shutdown" using the `close-output-port'
     procedure and the receiving direction can be "shutdown" using the
     `close-input-port' procedure.  The `close-port' procedure closes
     both directions of the connection.

     The parameter of this procedure is an IP port number (16-bit
     nonnegative exact integer), a string of the form `"HOST:PORT"' or
     a list of port settings.  When the parameter is the number PORT it
     is handled as if it was the setting `port-number:' PORT.  When the
     parameter is the string `"HOST:PORT"' it is handled as if it was
     the setting `server-address:' `"HOST:PORT"'.

     Here are the settings allowed in addition to the generic settings
     of byte-ports:

        * `server-address:' STRING-OR-IP-ADDRESS

          This setting indicates the internet address of the server, and
          possibly the IP port number.  When this parameter is not
          specified or is `""', the connection requests are sent to the
          loopback interface (with IP address 127.0.0.1).  The
          parameter can be a string denoting a host name, which will be
          translated to an IP address by the `host-info' procedure, or
          a 4 element u8vector which contains the 32-bit IPv4 address
          or an 8 element u16vector which contains the 128-bit IPv6
          address.  A string of the form `"HOST:PORT"' is handled as if
          it was the combination of settings `server-address:' `"HOST"'
          `port-number:' PORT.

        * `port-number:' 16-BIT-EXACT-INTEGER

          This setting indicates the IP port number of the server to
          connect to (e.g. 80 for the standard HTTP server, 23 for the
          standard telnet server).  There is no default value for this
          setting.

        * `keep-alive:' ( `#f' | `#t' )

          This setting controls the use of the "keep alive" option on
          the connection.  The "keep alive" option will periodically
          send control packets on otherwise idle network connections to
          ensure that the server host is active and reachable.  The
          default value of this setting is `#f'.

        * `coalesce:' ( `#f' | `#t' )

          This setting controls the use of TCP's "Nagle algorithm" which
          reduces the number of small packets by delaying their
          transmission and coalescing them into larger packets.  A
          setting of `#t' will coalesce small packets into larger ones.
          A setting of `#f' will transmit packets as soon as possible.
          The default value of this setting is `#t'.  Note that this
          setting does not affect the buffering of the port.

        * `tls-context:' ( `#f' | TLS-CONTEXT )

          This setting controls the use of TLS encryption.  If
          provided, the client will use this configuration for setting
          up a TCP connection with TLS encryption, otherwise it will
          use a plain TCP connection as usual.  Please note that Gambit
          must be compiled with TLS support for this option to be
          implemented.  See `make-tls-context' for futher information.
          The default value of this setting is `#f'.


     Below is an example of the client-side code that opens a
     connection to an HTTP server on port 8080 of the loopback
     interface (with IP address 127.0.0.1).  For the server-side code
     see the example for the procedure `open-tcp-server'.

          > (define p (open-tcp-client (list port-number: 8080
                                             eol-encoding: 'cr-lf)))
          > p
          #<input-output-port #2 (tcp-client #u8(127 0 0 1) 8080)>
          > (display "GET /\n" p)
          > (force-output p)
          > (read-line p)
          "<HTML>"


 -- procedure: open-tcp-server PORT-NUMBER-OR-ADDRESS-OR-SETTINGS
     This procedure sets up a socket to accept network connection
     requests from clients and returns a tcp-server-port from which
     network connections to clients are obtained.  Tcp-server-ports are
     a direct subtype of object-ports (i.e. they are not
     character-ports) and are input-ports.  Reading from a
     tcp-server-port with the `read' procedure will block until a
     network connection request is received from a client.  The `read'
     procedure will then return a tcp-client-port (a subtype of
     device-port) that represents this connection and allows
     communication with that client.  Closing a tcp-server-port with
     either the `close-input-port' or `close-port' procedures will
     cause the network subsystem to stop accepting connections on that
     socket.

     The parameter of this procedure is an IP port number (16-bit
     nonnegative exact integer), a string of the form `"INTF:PORT"' or
     a list of port settings which must contain a `port-number:'
     setting.  When the parameter is the number PORT it is handled as
     if it was the setting `port-number:' PORT.  When the parameter is
     the string `"INTF:PORT"' it is handled as if it was the setting
     `server-address:' `"INTF:PORT"'.

     Below is a list of the settings allowed in addition to the settings
     `keep-alive:' and `coalesce:' allowed by the `open-tcp-client'
     procedure and the generic settings of byte-ports.  The settings
     which are not listed below apply to the tcp-client-port that is
     returned by `read' when a connection is accepted and have the same
     meaning as if they were used in a call to the `open-tcp-client'
     procedure.

        * `server-address:' STRING-OR-IP-ADDRESS

          This setting indicates the internet address of the network
          interface on which connections requests are accepted, and
          possibly the IP port number.  When this parameter is not
          specified or is `""', the connection requests are accepted
          only on the loopback interface (with IP address 127.0.0.1).
          When this parameter is `"*"', the connection requests are
          accepted on all network interfaces (i.e. address INADDR_ANY).
          The parameter can be a string denoting a host name, which
          will be translated to an IP address by the `host-info'
          procedure, or a 4 element u8vector which contains the 32-bit
          IPv4 address or an 8 element u16vector which contains the
          128-bit IPv6 address.  A string of the form `"INTF:PORT"' is
          handled as if it was the combination of settings
          `server-address:' `"INTF"' `port-number:' PORT.

        * `port-number:' 16-BIT-EXACT-INTEGER

          This setting indicates the IP port number assigned to the
          socket which accepts connection requests from clients.  So
          called "well-known ports", which are reserved for standard
          services, have a port number below 1024 and can only be
          assigned to a socket by a process with superuser priviledges
          (e.g. 80 for the HTTP service, 23 for the telnet service).
          No special priviledges are needed to assign higher port
          numbers to a socket.  The special value 0 requests that a
          currently unused port number be assigned to the socket (the
          port number assigned can be retrieved using the procedure
          `tcp-server-socket-info').  There is no default value for this
          setting.

        * `backlog:' POSITIVE-EXACT-INTEGER

          This setting indicates the maximum number of connection
          requests that can be waiting to be accepted by a call to
          `read' (technically it is the value passed as the second
          argument of the UNIX `listen()' function).  The default value
          of this setting is 128.

        * `reuse-address:' ( `#f' | `#t' )

          This setting controls whether it is possible to assign a port
          number that is currently active.  Note that when a server
          process terminates, the socket it was using to accept
          connection requests does not become inactive immediately.
          Instead it remains active for a few minutes to ensure clean
          termination of the connections.  A setting of `#f' will cause
          an exception to be raised in that case.  A setting of `#t'
          will allow a port number to be used even if it is active.
          The default value of this setting is `#t'.

        * `tls-context:' ( `#f' | TLS-CONTEXT )

          This setting controls the use of TLS encryption.  If
          provided, the server will use this configuration for
          accepting TCP connections with TLS encryption, otherwise it
          will accept plain TCP connections as usual.  Please note that
          Gambit must be compiled with TLS support for this option to be
          implemented.  See `make-tls-context' for futher information.
          The default value of this setting is `#f'.


     Below is an example of the server-side code that accepts
     connections on port 8080 of any network interface.  For the
     client-side code see the example for the procedure
     `open-tcp-client'.

          > (define s (open-tcp-server (list server-address: "*"
                                             port-number: 8080
                                             eol-encoding: 'cr-lf)))
          > (define p (read s))  ; blocks until client connects
          > p
          #<input-output-port #2 (tcp-client 8080)>
          > (read-line p)
          "GET /"
          > (display "<HTML>\n" p)
          > (force-output p)


 -- procedure: tcp-service-register! PORT-NUMBER-OR-ADDRESS-OR-SETTINGS
          THUNK [THREAD-GROUP]
 -- procedure: tcp-service-unregister!
          PORT-NUMBER-OR-ADDRESS-OR-SETTINGS
     The procedure `tcp-service-register!' sets up a socket to accept
     network connection requests from clients and creates a "service"
     thread which processes the incoming connections.  The parameter
     PORT-NUMBER-OR-ADDRESS-OR-SETTINGS has the same meaning as for the
     procedure `open-tcp-server'.

     For each connection established the service thread creates a
     "handler" thread which executes a call to the procedure THUNK with
     no argument.  The handler thread's current input-port and current
     output-port are both set to the tcp-client-port created for the
     connection.  There is no need for the THUNK to close the
     tcp-client-port, as this is done by the handler thread when the
     THUNK returns normally.

     The procedure `tcp-service-unregister!' terminates the service
     thread which was registered by `tcp-service-register!' with the
     same network interface and port number (if a service thread is
     still registered).  The procedure `tcp-service-register!'
     implicitly calls `tcp-service-unregister!' before registering the
     new service thread.

          > (tcp-service-register!
             8000
             (lambda () (display "hello\n")))
          > (define p (open-tcp-client 8000))
          > (read-line p)
          "hello"
          > (tcp-service-unregister! 8000)


 -- procedure: make-tls-context [OPTIONS]
     This procedure requires Gambit to be compiled with TLS support,
     which is currently provided by OpenSSL. The `--enable-openssl'
     flag of the configure script will activate it, provided that you
     have the OpenSSL library and headers installed. It is strongly
     recommended that versions above 1.x are used.  On OSX, this means
     updating the OpenSSL bundled by default. This can be achieved
     using Homebrew, but manual installation or any other package
     manager will do.  Some notes on Windows with MinGW are also
     relevant here. Once you have a sane MinGW environment, remember to
     decompress the OpenSSL tarball with the tar utility, otherwise
     links to files won't work during the compilation process. The
     recommended build procedure for MinGW is as follows.

     Configure OpenSSL on MinGW 32 bits:

          perl Configure mingw no-asm --prefix=/usr/local --openssldir=/usr/local/openssl

     Configure OpenSSL on MinGW 64 bits:

          perl Configure mingw64 no-asm --prefix=/usr/local --openssldir=/usr/local/openssl

     Build and install with the following commands:

          make depend
          make
          make install

     A TLS context describes the options that will be used for setting
     up a TLS connection. If no TLS context is provided to
     `open-tcp-client' or `open-tcp-server', regular TCP connections
     without encryption will be used instead. The result of this
     procedure is a `SSL_CTX' pointer, which can be further manipulated
     with custom OpenSSL bindings. The configuration options are:

        * `min-version:' SYMBOL

          Establish a minimum TLS version for the connection. If the
          other peer doesn't support or agree with it, the connection
          will fail. Possible options (support depends on linked
          OpenSSL version): `ssl-v2', `ssl-v3', `tls-v1', `tls-v1.1',
          `tls-v1.2'.

        * `options:' LIST-OF-SYMBOLS

          A list of flags enabling/disabling TLS options. `server-mode'
          is required for using the TLS context with `open-tcp-server'.
          `use-diffie-hellman' enables the Diffie-Hellman key exchange.
          `use-elliptic-curves' enables Elliptic Curves. If no curve
          name is provided (with `elliptic-curve:'), `prime256v1' will
          be used. `request-client-authentication' is used by a server
          to enable request of authentication to clients.
          `insert-empty-fragments' enables a countermeasure against a
          SSL 3.0/TLS 1.0 protocol vulnerability affecting CBC ciphers.
          If used, the resulting connection may not be handled by some
          broken SSL implementations. This option has no effect for
          connections using other ciphers.

        * `certificate:' PATH

          Path to PEM Certificate file. This is a recommended option.
          If not provided OpenSSL will try to use anonymous cipher
          suites.

        * `private-key:' PATH

          Path to PEM Private Key file. If not provided, the
          Certificate path will be used instead.

        * `client-ca:' PATH

          Path to PEM file containing Certificate Authorities allowed
          for client authentication. Used only if
          `request-client-authentication' option is enabled.

        * `elliptic-curve:' STRING

          Name of the Elliptic Curve to use, according to RFC 4492.
          Used only if `use-elliptic-curves' option is enabled.


     TCP Client example with TLS encryption.

          (define ctx (make-tls-context))
          (define c (open-tcp-client (list server-address: "twitter.com"
                                           port-number: 443 tls-context: ctx)))
          (display "GET / HTTP/1.1\nHost: twitter.com\n\n" c)
          (force-output c)
          (read-line c)

     TCP Server example with several options. These are not mandatory,
     except for `server-mode'.

          (define ctx (make-tls-context options: '(server-mode
                                                   use-diffie-hellman
                                                   use-elliptic-curves)
                                        certificate: "server.pem"
                                        diffie-hellman-parameters: "dh_param_1024.pem"
                                        elliptic-curve: "prime256v1"))

          (define s (open-tcp-server (list server-address: "localhost"
                                           port-number: 1443 tls-context: ctx)))
          (define p (read s))
          (display "<HTML></HTML>\n" p)
          (force-output p)

     A practical way of testing TLS options are the `s_server' and
     `s_client' commands of the `openssl' tool.


17.8 Directory-ports
====================

 -- procedure: open-directory PATH-OR-SETTINGS
     This procedure opens a directory of the filesystem for reading its
     entries and returns a directory-port from which the entries can be
     enumerated.  Directory-ports are a direct subtype of object-ports
     (i.e. they are not character-ports) and are input-ports.  Reading
     from a directory-port with the `read' procedure returns the next
     file name in the directory as a string.  The end-of-file object is
     returned when all the file names have been enumerated.  Another
     way to get the list of all files in a directory is the
     `directory-files' procedure which returns a list of the files in
     the directory.  The advantage of using directory-ports is that it
     allows iterating over the files in a directory in constant space,
     which is interesting when the number of files in the directory is
     not known in advance and may be large.  Note that the order in
     which the names are returned is operating-system dependent.

     The parameter of this procedure is either a string denoting a
     filesystem path to a directory or a list of port settings which
     must contain a `path:' setting.  Here are the settings allowed in
     addition to the generic settings of object-ports:

        * `path:' STRING

          This setting indicates the location of the directory in the
          filesystem.  There is no default value for this setting.

        * `ignore-hidden:' ( `#f' | `#t' | `dot-and-dot-dot' )

          This setting controls whether hidden-files will be returned.
          Under UNIX and Mac OS X hidden-files are those that start
          with a period (such as `.', `..', and `.profile').  Under
          Microsoft Windows hidden files are the `.' and `..' entries
          and the files whose "hidden file" attribute is set.  A
          setting of `#f' will enumerate all the files.  A setting of
          `#t' will only enumerate the files that are not hidden.  A
          setting of `dot-and-dot-dot' will enumerate all the files
          except for the `.' and `..' hidden files.  The default value
          of this setting is `#t'.


     For example:

          > (let ((p (open-directory (list path: "../examples"
                                           ignore-hidden: #f))))
              (let loop ()
                (let ((fn (read p)))
                  (if (string? fn)
                      (begin
                        (pp (path-expand fn))
                        (loop)))))
              (close-input-port p))
          "/u/feeley/examples/."
          "/u/feeley/examples/.."
          "/u/feeley/examples/complex"
          "/u/feeley/examples/README"
          "/u/feeley/examples/simple"
          > (define x (open-directory "../examples"))
          > (read-all x)
          ("complex" "README" "simple")


17.9 Vector-ports
=================

 -- procedure: open-vector [VECTOR-OR-SETTINGS]
 -- procedure: open-input-vector [VECTOR-OR-SETTINGS]
 -- procedure: open-output-vector [VECTOR-OR-SETTINGS]
 -- procedure: call-with-input-vector VECTOR-OR-SETTINGS PROC
 -- procedure: call-with-output-vector VECTOR-OR-SETTINGS PROC
 -- procedure: with-input-from-vector VECTOR-OR-SETTINGS THUNK
 -- procedure: with-output-to-vector VECTOR-OR-SETTINGS THUNK
     Vector-ports represent streams of Scheme objects.  They are a
     direct subtype of object-ports (i.e. they are not
     character-ports).  All of these procedures create vector-ports
     that are either unidirectional or bidirectional.  The `direction:'
     setting will default to the value `input' for the procedures
     `open-input-vector', `call-with-input-vector' and
     `with-input-from-vector', to the value `output' for the procedures
     `open-output-vector', `call-with-output-vector' and
     `with-output-to-vector', and to the value `input-output' for the
     procedure `open-vector'.  Bidirectional vector-ports behave like
     FIFOs: data written to the port is added to the end of the stream
     that is read.  It is only when a bidirectional vector-port's
     output-side is closed with a call to the `close-output-port'
     procedure that the stream's end is known (when the stream's end is
     reached, reading the port returns the end-of-file object).

     The procedures `open-vector', `open-input-vector' and
     `open-output-vector' return the port that is created.  The
     procedures `call-with-input-vector' and `call-with-output-vector'
     create a vector port, call the procedure PROC with the port as
     single argument and then close the port.  The procedures
     `with-input-from-vector' and `with-output-to-vector' create a
     vector port, dynamically bind the current input-port and current
     output-port respectively to the port created for the duration of a
     call to the procedure THUNK with no argument, and then close the
     port.  The procedures `call-with-input-vector' and
     `with-input-from-vector' return the value returned by the
     procedures PROC and THUNK respectively.  The procedures
     `call-with-output-vector' and `with-output-to-vector' return the
     vector accumulated in the port (see `get-output-vector').

     The first parameter of these procedures is either a vector of the
     elements used to initialize the stream or a list of port settings.
     If it is not specified, the parameter of the `open-vector',
     `open-input-vector', and `open-output-vector' procedures defaults
     to an empty list of port settings.  Here are the settings allowed
     in addition to the generic settings of object-ports:

        * `init:' VECTOR

          This setting indicates the initial content of the stream.
          The default value of this setting is an empty vector.

        * `permanent-close:' ( `#f' | `#t' )

          This setting controls whether a call to the procedures
          `close-output-port' will close the output-side of a
          bidirectional vector-port permanently or not.  A permanently
          closed bidirectional vector-port whose end-of-file has been
          reached on the input-side will return the end-of-file object
          for all subsequent calls to the `read' procedure.  A
          non-permanently closed bidirectional vector-port will return
          to its opened state when its end-of-file is read.  The
          default value of this setting is `#t'.


     For example:

          > (define p (open-vector))
          > (write 1 p)
          > (write 2 p)
          > (write 3 p)
          > (force-output p)
          > (read p)
          1
          > (read p)
          2
          > (close-output-port p)
          > (read p)
          3
          > (read p)
          #!eof
          > (with-output-to-vector '() (lambda () (write 1) (write 2)))
          #(1 2)


 -- procedure: open-vector-pipe [VECTOR-OR-SETTINGS1
          [VECTOR-OR-SETTINGS2]]
     The procedure `open-vector-pipe' creates two vector-ports and
     returns these two ports.  The two ports are interrelated as
     follows: the first port's output-side is connected to the second
     port's input-side and the first port's input-side is connected to
     the second port's output-side.  The value VECTOR-OR-SETTINGS1 is
     used to setup the first vector-port and VECTOR-OR-SETTINGS2 is
     used to setup the second vector-port.  The same settings as for
     `open-vector' are allowed.  The default `direction:' setting is
     `input-output' (i.e. a bidirectional port is created).  If it is
     not specified VECTOR-OR-SETTINGS1 defaults to the empty list.  If
     it is not specified VECTOR-OR-SETTINGS2 defaults to
     VECTOR-OR-SETTINGS1 but with the `init:' setting set to the empty
     vector and with the input and output settings exchanged (e.g. if
     the first port is an input-port then the second port is an
     output-port, if the first port's input-side is non-buffered then
     the second port's output-side is non-buffered).

     For example:

          > (define (server op)
              (receive (c s) (open-vector-pipe)  ; client-side and server-side ports
                (thread-start!
                  (make-thread
                    (lambda ()
                      (let loop ()
                        (let ((request (read s)))
                          (if (not (eof-object? request))
                              (begin
                                (write (op request) s)
                                (newline s)
                                (force-output s)
                                (loop))))))))
                c))
          > (define a (server (lambda (x) (expt 2 x))))
          > (define b (server (lambda (x) (expt 10 x))))
          > (write 100 a)
          > (write 30 b)
          > (read a)
          1267650600228229401496703205376
          > (read b)
          1000000000000000000000000000000


 -- procedure: get-output-vector VECTOR-PORT
     The procedure `get-output-vector' takes an output vector-port or a
     bidirectional vector-port as parameter and removes all the objects
     currently on the output-side, returning them in a vector.  The port
     remains open and subsequent output to the port and calls to the
     procedure `get-output-vector' are possible.

     For example:

          > (define p (open-vector '#(1 2 3)))
          > (write 4 p)
          > (get-output-vector p)
          #(1 2 3 4)
          > (write 5 p)
          > (write 6 p)
          > (get-output-vector p)
          #(5 6)


17.10 String-ports
==================

 -- procedure: open-string [STRING-OR-SETTINGS]
 -- procedure: open-input-string [STRING-OR-SETTINGS]
 -- procedure: open-output-string [STRING-OR-SETTINGS]
 -- procedure: call-with-input-string STRING-OR-SETTINGS PROC
 -- procedure: call-with-output-string STRING-OR-SETTINGS PROC
 -- procedure: with-input-from-string STRING-OR-SETTINGS THUNK
 -- procedure: with-output-to-string STRING-OR-SETTINGS THUNK
 -- procedure: open-string-pipe [STRING-OR-SETTINGS1
          [STRING-OR-SETTINGS2]]
 -- procedure: get-output-string STRING-PORT
     String-ports represent streams of characters.  They are a direct
     subtype of character-ports.  These procedures are the string-port
     analog of the procedures specified in the vector-ports section.
     Note that these procedures are a superset of the procedures
     specified in the "Basic String Ports SRFI" (SRFI 6).

     For example:

          > (define p (open-string))
          > (write 1 p)
          > (write 2 p)
          > (write 3 p)
          > (force-output p)
          > (read-char p)
          #\1
          > (read-char p)
          #\2
          > (close-output-port p)
          > (read-char p)
          #\3
          > (read-char p)
          #!eof
          > (with-output-to-string '() (lambda () (write 1) (write 2)))
          "12"


 -- procedure: object->string OBJ [N]
     This procedure converts the object OBJ to its external
     representation and returns it in a string.  The parameter N
     specifies the maximal width of the resulting string.  If the
     external representation is wider than N, the resulting string will
     be truncated to N characters and the last 3 characters will be set
     to periods.  Note that the current readtable is used.

     For example:

          > (object->string (expt 2 100))
          "1267650600228229401496703205376"
          > (object->string (expt 2 100) 30)
          "126765060022822940149670320..."
          > (object->string (cons car cdr))
          "(#<procedure #2 car> . #<procedure #3 cdr>)"


17.11 U8vector-ports
====================

 -- procedure: open-u8vector [U8VECTOR-OR-SETTINGS]
 -- procedure: open-input-u8vector [U8VECTOR-OR-SETTINGS]
 -- procedure: open-output-u8vector [U8VECTOR-OR-SETTINGS]
 -- procedure: call-with-input-u8vector U8VECTOR-OR-SETTINGS PROC
 -- procedure: call-with-output-u8vector U8VECTOR-OR-SETTINGS PROC
 -- procedure: with-input-from-u8vector U8VECTOR-OR-SETTINGS THUNK
 -- procedure: with-output-to-u8vector U8VECTOR-OR-SETTINGS THUNK
 -- procedure: open-u8vector-pipe [U8VECTOR-OR-SETTINGS1
          [U8VECTOR-OR-SETTINGS2]]
 -- procedure: get-output-u8vector U8VECTOR-PORT
     U8vector-ports represent streams of bytes.  They are a direct
     subtype of byte-ports.  These procedures are the u8vector-port
     analog of the procedures specified in the vector-ports section.

     For example:

          > (define p (open-u8vector))
          > (write 1 p)
          > (write 2 p)
          > (write 3 p)
          > (force-output p)
          > (read-u8 p)
          49
          > (read-u8 p)
          50
          > (close-output-port p)
          > (read-u8 p)
          51
          > (read-u8 p)
          #!eof
          > (with-output-to-u8vector '() (lambda () (write 1) (write 2)))
          #u8(49 50)


17.12 Other procedures related to I/O
=====================================

 -- procedure: current-input-port [NEW-VALUE]
 -- procedure: current-output-port [NEW-VALUE]
 -- procedure: current-error-port [NEW-VALUE]
 -- procedure: current-readtable [NEW-VALUE]
     These procedures are parameter objects which represent
     respectively: the current input-port, the current output-port, the
     current error-port, and the current readtable.


 -- procedure: print [`port:' PORT] OBJ...
 -- procedure: println [`port:' PORT] OBJ...
     The `print' procedure writes a representation of each OBJ, from
     left to right, to PORT.  When a compound object is encountered
     (pair, list, vector, box) the elements of that object are
     recursively written without the surrounding tokens (parentheses,
     spaces, dots, etc).  Strings, symbols, keywords and characters are
     written like the `display' procedure.  If there are more than one
     OBJ, the first OBJ must not be a keyword object.  If it is not
     specified, PORT defaults to the current output-port.  The
     procedure `print' returns an unspecified value.

     The `println' procedure does the same thing as the `print'
     procedure and then writes an end of line to PORT.

     For example:

          > (println "2*2 is " (* 2 2) " and 2+2 is " (+ 2 2))
          2*2 is 4 and 2+2 is 4
          > (define x (list "<i>" (list "<tt>" 123 "</tt>") "</i>"))
          > (println x)
          <i><tt>123</tt></i>
          > (define p (open-output-string))
          > (print port: p 1 #\2 "345")
          > (get-output-string p)
          "12345"


18 Lexical syntax and readtables
********************************

18.1 Readtables
===============

Readtables control the external textual representation of Scheme
objects, that is the encoding of Scheme objects using characters.
Readtables affect the behavior of the reader (i.e. the `read' procedure
and the parser used by the `load' procedure and the interpreter and
compiler) and the printer (i.e. the procedures `write', `display',
`print', `println', `pretty-print', and `pp', and the procedure used by
the REPL to print results).  To preserve write/read invariance the
printer and reader must be using compatible readtables.  For example a
symbol which contains upper case letters will be printed with special
escapes if the readtable indicates that the reader is case-insensitive.

   Readtables are immutable records whose fields specify various textual
representation aspects.  There are accessor procedures to retrieve the
content of specific fields.  There are also functional update
procedures that create a copy of a readtable, with a specific field set
to a new value.

 -- procedure: readtable? OBJ
     This procedure returns `#t' when OBJ is a readtable and `#f'
     otherwise.

     For example:

          > (readtable? (current-readtable))
          #t
          > (readtable? 123)
          #f


 -- procedure: readtable-case-conversion? READTABLE
 -- procedure: readtable-case-conversion?-set READTABLE NEW-VALUE
     The procedure `readtable-case-conversion?' returns the content of
     the `case-conversion?' field of READTABLE.  When the content of
     this field is `#f', the reader preserves the case of symbols and
     keyword objects that are read (i.e. `Ice' and `ice' are distinct
     symbols).  When the content of this field is the symbol `upcase',
     the reader converts lowercase letters to uppercase when reading
     symbols and keywords (i.e. `Ice' is read as the symbol
     `(string->symbol "ICE")').  Otherwise the reader converts
     uppercase letters to lowercase when reading symbols and keywords
     (i.e. `Ice' is read as the symbol `(string->symbol "ice")').

     The procedure `readtable-case-conversion?-set' returns a copy of
     READTABLE where only the `case-conversion?' field has been changed
     to NEW-VALUE.

     For example:

          > (output-port-readtable-set!
              (repl-output-port)
              (readtable-case-conversion?-set
                (output-port-readtable (repl-output-port))
                #f))
          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-case-conversion?-set
                (input-port-readtable (repl-input-port))
                #f))
          > 'Ice
          Ice
          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-case-conversion?-set
                (input-port-readtable (repl-input-port))
                #t))
          > 'Ice
          ice
          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-case-conversion?-set
                (input-port-readtable (repl-input-port))
                'upcase))
          > 'Ice
          ICE


 -- procedure: readtable-keywords-allowed? READTABLE
 -- procedure: readtable-keywords-allowed?-set READTABLE NEW-VALUE
     The procedure `readtable-keywords-allowed?' returns the content of
     the `keywords-allowed?' field of READTABLE.  When the content of
     this field is `#f', the reader does not recognize keyword objects
     (i.e. `:foo' and `foo:' are read as the symbols `(string->symbol
     ":foo")' and `(string->symbol "foo:")' respectively).  When the
     content of this field is the symbol `prefix', the reader
     recognizes keyword objects that start with a colon, as in Common
     Lisp (i.e. `:foo' is read as the keyword `(string->keyword
     "foo")').  Otherwise the reader recognizes keyword objects that
     end with a colon, as in DSSSL (i.e. `foo:' is read as the symbol
     `(string->symbol "foo")').

     The procedure `readtable-keywords-allowed?-set' returns a copy of
     READTABLE where only the `keywords-allowed?' field has been
     changed to NEW-VALUE.

     For example:

          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-keywords-allowed?-set
                (input-port-readtable (repl-input-port))
                #f))
          > (map keyword? '(foo :foo foo:))
          (#f #f #f)
          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-keywords-allowed?-set
                (input-port-readtable (repl-input-port))
                #t))
          > (map keyword? '(foo :foo foo:))
          (#f #f #t)
          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-keywords-allowed?-set
                (input-port-readtable (repl-input-port))
                'prefix))
          > (map keyword? '(foo :foo foo:))
          (#f #t #f)


 -- procedure: readtable-sharing-allowed? READTABLE
 -- procedure: readtable-sharing-allowed?-set READTABLE NEW-VALUE
     The procedure `readtable-sharing-allowed?' returns the content of
     the `sharing-allowed?' field of READTABLE.  The reader recognizes
     the `#N#' and `#N=DATUM' notation for circular structures and the
     printer uses this notation if and only if the content of the
     `sharing-allowed?' field is not `#f'.  Moreover when the content
     of the `sharing-allowed?' field is the symbol `serialize', the
     printer uses a special external representation that the reader
     understands and that extends write/read invariance to the
     following types: records, procedures and continuations.  Note that
     an object can be serialized and deserialized if and only if all of
     its components are serializable.

     The procedure `readtable-sharing-allowed?-set' returns a copy of
     READTABLE where only the `sharing-allowed?' field has been changed
     to NEW-VALUE.

     Here is a simple example:

          > (define (wr obj allow?)
              (call-with-output-string
                '()
                (lambda (p)
                  (output-port-readtable-set!
                    p
                    (readtable-sharing-allowed?-set
                      (output-port-readtable p)
                      allow?))
                  (write obj p))))
          > (define (rd str allow?)
              (call-with-input-string
                str
                (lambda (p)
                  (input-port-readtable-set!
                    p
                    (readtable-sharing-allowed?-set
                      (input-port-readtable p)
                      allow?))
                  (read p))))
          > (define x (list 1 2 3))
          > (set-car! (cdr x) (cddr x))
          > (wr x #f)
          "(1 (3) 3)"
          > (wr x #t)
          "(1 #0=(3) . #0#)"
          > (define y (rd (wr x #t) #t))
          > y
          (1 (3) 3)
          > (eq? (cadr y) (cddr y))
          #t
          > (define f #f)
          > (let ((free (expt 2 10)))
            (set! f (lambda (x) (+ x free))))
          > (define s (wr f 'serialize))
          > (string-length s)
          4196
          > (define g (rd s 'serialize))
          > (eq? f g)
          #f
          > (g 4)
          1028

     Continuations are tricky to serialize because they contain a
     dynamic environment and this dynamic environment may contain
     non-serializable objects, in particular ports attached to
     operating-system streams such as files, the console or standard
     input/output.  Indeed, all dynamic environments contain a binding
     for the `current-input-port' and `current-output-port'.  Moreover,
     any thread that has started a REPL has a continuation which refers
     to the "repl-context" object in its dynamic environment.  A
     repl-context object contains the interaction channel, which is
     typically connected to a non-serializable port, such as the
     console.  Another problem is that the `parameterize' form saves
     the old binding of the parameter in the continuation, so it is not
     possible to eliminate the references to these ports in the
     continuation by using the `parameterize' form alone.

     Serialization of continuations can be achieved dependably by taking
     advantage of string-ports, which are serializable objects (unless
     there is a blocked thread), and the following features of threads:
     they inherit the dynamic environment of the parent thread and they
     start with an initial continuation that contains only serializable
     objects.  So a thread created in a dynamic environment where
     `current-input-port' and `current-output-port' are bound to a
     dummy string-port has a serializable continuation.

     Here is an example where continuations are serialized:

          > (define (wr obj)
              (call-with-output-string
               '()
               (lambda (p)
                 (output-port-readtable-set!
                  p
                  (readtable-sharing-allowed?-set
                   (output-port-readtable p)
                   'serialize))
                 (write obj p))))
          > (define (rd str)
              (call-with-input-string
               str
               (lambda (p)
                 (input-port-readtable-set!
                  p
                  (readtable-sharing-allowed?-set
                   (input-port-readtable p)
                   'serialize))
                 (read p))))
          > (define fifo (open-vector))
          > (define (suspend-and-die!)
              (call-with-current-continuation
               (lambda (k)
                 (write (wr k) fifo)
                 (newline fifo)
                 (force-output fifo)
                 (thread-terminate! (current-thread)))))
          > (let ((dummy-port (open-string)))
              (parameterize ((current-input-port dummy-port)
                             (current-output-port dummy-port))
                (thread-start!
                 (make-thread
                  (lambda ()
                    (* 100
                       (suspend-and-die!)))))))
          #<thread #2>
          > (define s (read fifo))
          > (thread-join!
              (thread-start!
                (make-thread
                  (lambda ()
                    ((rd s) 111)))))
          11100
          > (thread-join!
              (thread-start!
                (make-thread
                  (lambda ()
                    ((rd s) 222)))))
          22200
          > (string-length s)
          13114


 -- procedure: readtable-eval-allowed? READTABLE
 -- procedure: readtable-eval-allowed?-set READTABLE NEW-VALUE
     The procedure `readtable-eval-allowed?' returns the content of the
     `eval-allowed?' field of READTABLE.  The reader recognizes the
     `#.EXPRESSION' notation for read-time evaluation if and only if
     the content of the `eval-allowed?' field is not `#f'.

     The procedure `readtable-eval-allowed?-set' returns a copy of
     READTABLE where only the `eval-allowed?' field has been changed to
     NEW-VALUE.

     For example:

          > (input-port-readtable-set!
              (repl-input-port)
              (readtable-eval-allowed?-set
                (input-port-readtable (repl-input-port))
                #t))
          > '(5 plus 7 is #.(+ 5 7))
          (5 plus 7 is 12)
          > '(buf = #.(make-u8vector 25))
          (buf = #u8(0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0))


 -- procedure: readtable-write-cdr-read-macros? READTABLE
 -- procedure: readtable-write-cdr-read-macros?-set READTABLE NEW-VALUE
 -- procedure: readtable-write-extended-read-macros? READTABLE
 -- procedure: readtable-write-extended-read-macros?-set READTABLE
          NEW-VALUE
     The procedure `readtable-write-cdr-read-macros?' returns the
     content of the `write-cdr-read-macros?' field of READTABLE.  The
     procedure `readtable-write-extended-read-macros?' returns the
     content of the `write-extended-read-macros?' field of READTABLE.

     At all times the printer uses read-macros in its output for datums
     of the form `(quote DATUM)', `(quasiquote DATUM)', `(unquote
     DATUM)', and `(unquote-splicing DATUM)'.  That is the following
     read-macro notations will be used respectively: `'DATUM', ``DATUM',
     `,DATUM', and `,@DATUM'.  Moreover, normally the read-macros will
     not be used when the form appears in the cdr of a list, for
     example `(foo quote bar)', `(foo . (quote bar))' and `(foo .
     'bar)' will all be printed as `(foo quote bar)'.

     When the content of the `write-cdr-read-macros?' field is not
     `#f', the printer will use read-macros when the forms appear in
     the cdr of a list.  For example `(foo quote bar)' will be printed
     as `(foo . 'bar)'.  When the content of the
     `write-extended-read-macros?' field is not `#f', the printer will
     also use extended read-macros, for example `#'DATUM' in place of
     `(syntax DATUM)'.

     The procedure `readtable-write-cdr-read-macros?-set' returns a
     copy of READTABLE where only the `write-cdr-read-macros?' field
     has been changed to NEW-VALUE.  The procedure
     `readtable-write-extended-read-macros?-set' returns a copy of
     READTABLE where only the `write-extended-read-macros?' field has
     been changed to NEW-VALUE.

     For example:

          > (output-port-readtable-set!
              (repl-output-port)
              (readtable-write-extended-read-macros?-set
                (output-port-readtable (repl-output-port))
                #t))
          > '(foo (syntax bar))
          (foo #'bar)
          > '(foo syntax bar)
          (foo syntax bar)
          > (output-port-readtable-set!
              (repl-output-port)
              (readtable-write-cdr-read-macros?-set
                (output-port-readtable (repl-output-port))
                #t))
          > '(foo syntax bar)
          (foo . #'bar)


 -- procedure: readtable-max-write-level READTABLE
 -- procedure: readtable-max-write-level-set READTABLE NEW-VALUE
     The procedure `readtable-max-write-level' returns the content of
     the `max-write-level' field of READTABLE.  The printer will
     display an ellipsis for the elements of lists and vectors that are
     nested deeper than this level.

     The procedure `readtable-max-write-level-set' returns a copy of
     READTABLE where only the `max-write-level' field has been changed
     to NEW-VALUE, which must be an nonnegative fixnum.

     For example:

          > (define (wr obj n)
              (call-with-output-string
                '()
                (lambda (p)
                  (output-port-readtable-set!
                    p
                    (readtable-max-write-level-set
                      (output-port-readtable p)
                      n))
                  (write obj p))))
          > (wr '(a #(b (c c) #u8(9 9 9) b) a) 3)
          "(a #(b (c c) #u8(9 9 9) b) a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) a) 2)
          "(a #(b (...) #u8(...) b) a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) a) 1)
          "(a #(...) a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) a) 0)
          "(...)"
          > (wr 'hello 0)
          "hello"


 -- procedure: readtable-max-write-length READTABLE
 -- procedure: readtable-max-write-length-set READTABLE NEW-VALUE
     The procedure `readtable-max-write-length' returns the content of
     the `max-write-length' field of READTABLE.  The printer will
     display an ellipsis for the elements of lists and vectors that are
     at an index beyond that length.

     The procedure `readtable-max-write-length-set' returns a copy of
     READTABLE where only the `max-write-length' field has been changed
     to NEW-VALUE, which must be an nonnegative fixnum.

     For example:

          > (define (wr obj n)
              (call-with-output-string
                '()
                (lambda (p)
                  (output-port-readtable-set!
                    p
                    (readtable-max-write-length-set
                      (output-port-readtable p)
                      n))
                  (write obj p))))
          > (wr '(a #(b (c c) #u8(9 9 9) b) . a) 4)
          "(a #(b (c c) #u8(9 9 9) b) . a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) . a) 3)
          "(a #(b (c c) #u8(9 9 9) ...) . a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) . a) 2)
          "(a #(b (c c) ...) . a)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) . a) 1)
          "(a ...)"
          > (wr '(a #(b (c c) #u8(9 9 9) b) . a) 0)
          "(...)"


 -- procedure: readtable-max-unescaped-char READTABLE
 -- procedure: readtable-max-unescaped-char-set READTABLE NEW-VALUE
     The procedure `readtable-max-unescaped-char' returns the content
     of the `max-unescaped-char' field of READTABLE.  The printer will
     display using an escape sequence any character within symbols,
     strings and character objects greater than `max-unescaped-char'.
     When `max-unescaped-char' is `#f', the default value, the printer
     will take into account the output port and use an escape sequence
     for any character that isn't supported by the port's character
     encoding.

     The procedure `readtable-max-unescaped-char-set' returns a copy of
     READTABLE where only the `max-unescaped-char' field has been
     changed to NEW-VALUE, which must be a character or `#f'.

     For example:

          > (define rt (output-port-readtable (repl-output-port)))
          > (readtable-max-unescaped-char rt)
          #\delete
          > (string (integer->char 233))
          "\351"
          > (define (f c)
              (with-output-to-string
               (list readtable: (readtable-max-unescaped-char-set rt c))
               (lambda () (write (string (integer->char 233))))))
          > (f #\delete)
          "\"\\351\""
          > (string-length (f #\delete))
          6
          > (f #\U0010ffff)
          "\"\351\""
          > (string-length (f #\U0010ffff))
          3
          > (output-port-readtable-set!
             (repl-output-port)
             (readtable-max-unescaped-char-set rt #\U0010ffff))
          > (string (integer->char 233))
          "e'"


 -- procedure: readtable-start-syntax READTABLE
 -- procedure: readtable-start-syntax-set READTABLE NEW-VALUE
     The procedure `readtable-start-syntax' returns the content of the
     `start-syntax' field of READTABLE.  The reader uses this field to
     determine in which syntax to start parsing the input.  When the
     content of this field is the symbol `six', the reader starts in
     the infix syntax.  Otherwise the reader starts in the prefix
     syntax.

     The procedure `readtable-start-syntax-set' returns a copy of
     READTABLE where only the `start-syntax' field has been changed to
     NEW-VALUE.

     For example:

          > (+ 2 3)
          5
          > (input-port-readtable-set!
             (repl-input-port)
             (readtable-start-syntax-set
               (input-port-readtable (repl-input-port))
               'six))
          > 2+3;
          5
          > exit();


18.2 Boolean syntax
===================

Booleans are required to be followed by a delimiter (i.e. `#f64()' is
not the boolean `#f' followed by the number `64' and the empty list).

18.3 Character syntax
=====================

Characters are required to be followed by a delimiter (i.e.
`#\spaceballs' is not the character `#\space' followed by the symbol
`balls').  The lexical syntax of characters is extended to allow the
following:

`#\nul'
     Unicode character 0

`#\alarm'
     Unicode character 7

`#\backspace'
     Unicode character 8

`#\tab'
     Unicode character 9

`#\newline'
     Unicode character 10 (newline character)

`#\linefeed'
     Unicode character 10

`#\vtab'
     Unicode character 11

`#\page'
     Unicode character 12

`#\return'
     Unicode character 13

`#\esc'
     Unicode character 27

`#\space'
     Unicode character 32 (space character)

`#\delete'
     Unicode character 127

`#\xHH'
     character encoded in hexadecimal (>= 1 hexadecimal digit)

`#\uHHHH'
     character encoded in hexadecimal (exactly 4 hexadecimal digits)

`#\UHHHHHHHH'
     character encoded in hexadecimal (exactly 8 hexadecimal digits)

18.4 String syntax
==================

The lexical syntax of quoted strings is extended to allow the following
escape codes:

`\a'
     Unicode character 7

`\b'
     Unicode character 8

`\t'
     Unicode character 9

`\n'
     Unicode character 10 (newline character)

`\v'
     Unicode character 11

`\f'
     Unicode character 12

`\r'
     Unicode character 13

`\"'
     `"'

`\\'
     `\'

`\|'
     `|'

`\?'
     `?'

`\OOO'
     character encoded in octal (1 to 3 octal digits, first digit must
     be less than 4 when there are 3 octal digits)

`\xHH'
     character encoded in hexadecimal (>= 1 hexadecimal digit)

`\uHHHH'
     character encoded in hexadecimal (exactly 4 hexadecimal digits)

`\UHHHHHHHH'
     character encoded in hexadecimal (exactly 8 hexadecimal digits)

`\<space>'
     Unicode character 32 (space character)

`\<newline><whitespace-except-newline>*'
     This sequence expands to nothing (it is useful for splitting a long
     string literal on multiple lines while respecting proper
     indentation of the source code)

   Gambit also supports a "here string" syntax that is similar to shell
"here documents".  For example:

     > (pretty-print #<<THE-END
     hello
     world
     THE-END
     )
     "hello\nworld"

   The here string starts with the sequence `#<<'.  The part of the
line after the `#<<' up to and including the newline character is the
key. The first line afterward that matches the key marks the end of the
here string.  The string contains all the characters between the start
key and the end key, with the exception of the newline character before
the end key.

18.5 Symbol syntax
==================

The lexical syntax of symbols is extended to allow a leading and
trailing vertical bar (e.g. `|a\|b"c:|').  The symbol's name
corresponds verbatim to the characters between the vertical bars except
for escaped characters.  The same escape sequences as for strings are
permitted except that `"' does not need to be escaped and `|' needs to
be escaped.

   For example:

     > (symbol->string '|a\|b"c:|)
     "a|b\"c:"

18.6 Keyword syntax
===================

The lexical syntax of keywords is like symbols, but with a colon at the
end (note that this can be changed to a leading colon by setting the
`keywords-allowed?' field of the readtable to the symbol `prefix').  A
colon by itself is not a keyword, it is a symbol.  Vertical bars can be
used like symbols but the colon must be outside the vertical bars.
Note that the string returned by the `keyword->string' procedure does
not include the colon.

   For example:

     > (keyword->string 'foo:)
     "foo"
     > (map keyword? '(|ab()cd:| |ab()cd|: : ||:))
     (#f #t #f #t)

18.7 Box syntax
===============

The lexical syntax of boxes is `#&OBJ' where OBJ is the content of the
box.

   For example:

     > (list '#&"hello" '#&123)
     (#&"hello" #&123)
     > (box (box (+ 10 20)))
     #&#&30

18.8 Number syntax
==================

The lexical syntax of the special inexact real numbers is as follows:

`+inf.0'
     positive infinity

`-inf.0'
     negative infinity

`+nan.0'
     "not a number"

`-0.'
     negative zero (`0.' is the positive zero)

18.9 Homogeneous vector syntax
==============================

Homogeneous vectors are vectors containing raw numbers of the same type
(signed or unsigned exact integers or inexact reals).  There are 10
types of homogeneous vectors: `s8vector' (vector of 8 bit signed
integers), `u8vector' (vector of 8 bit unsigned integers), `s16vector'
(vector of 16 bit signed integers), `u16vector' (vector of 16 bit
unsigned integers), `s32vector' (vector of 32 bit signed integers),
`u32vector' (vector of 32 bit unsigned integers), `s64vector' (vector
of 64 bit signed integers), `u64vector' (vector of 64 bit unsigned
integers), `f32vector' (vector of 32 bit floating point numbers), and
`f64vector' (vector of 64 bit floating point numbers).

   The external representation of homogeneous vectors is similar to
normal vectors but with the `#(' prefix replaced respectively with
`#s8(', `#u8(', `#s16(', `#u16(', `#s32(', `#u32(', `#s64(', `#u64(',
`#f32(', and `#f64('.

   The elements of the integer homogeneous vectors must be exact
integers fitting in the given precision.  The elements of the floating
point homogeneous vectors must be inexact reals.

18.10 Special `#!' syntax
=========================

The lexical syntax of the special `#!' objects is as follows:

`#!eof'
     end-of-file object

`#!void'
     void object

`#!optional'
     optional object

`#!rest'
     rest object

`#!key'
     key object

18.11 Multiline comment syntax
==============================

Multiline comments are delimited by the tokens `#|' and `|#'.  These
comments can be nested.

18.12 Scheme infix syntax extension
===================================

The reader supports an infix syntax extension which is called SIX
(Scheme Infix eXtension).  This extension is both supported by the
`read' procedure and in program source code.

   The backslash character is a delimiter that marks the beginning of a
single datum expressed in the infix syntax (the details are given
below).  One way to think about it is that the backslash character
escapes the prefix syntax temporarily to use the infix syntax.  For
example a three element list could be written as `(X \Y Z)', the
elements X and Z are expressed using the normal prefix syntax and Y is
expressed using the infix syntax.

   When the reader encounters an infix datum, it constructs a syntax
tree for that particular datum.  Each node of this tree is represented
with a list whose first element is a symbol indicating the type of node.
For example, `(six.identifier abc)' is the representation of the infix
identifier `abc' and `(six.index (six.identifier abc) (six.identifier
i))' is the representation of the infix datum `abc[i];'.

18.12.1 SIX grammar
-------------------

The SIX grammar is given below.  On the left hand side are the
production rules.  On the right hand side is the datum that is
constructed by the reader.  The notation $I denotes the datum that is
constructed by the reader for the Ith part of the production rule.

<infix datum> ::=                           
<stat>                                      $1
<stat> ::=                                  
<if stat>                                   $1
| <for stat>                                $1
| <while stat>                              $1
| <do stat>                                 $1
| <switch stat>                             $1
| <case stat>                               $1
| <break stat>                              $1
| <continue stat>                           $1
| <label stat>                              $1
| <goto stat>                               $1
| <return stat>                             $1
| <expression stat>                         $1
| <procedure definition>                    $1
| <variable definition> `;'                 $1
| <clause stat>                             $1
| <compound stat>                           $1
| `;'                                       `(six.compound)'
<if stat> ::=                               
`if' `(' <pexpr> `)' <stat>                 `(six.if $3 $5)'
| `if' `(' <pexpr> `)' <stat> `else'        `(six.if $3 $5 $7)'
<stat>                                      
<for stat> ::=                              
`for' `(' <stat> `;' <oexpr> `;' <oexpr>    `(six.for $3 $5 $7 $9)'
`)' <stat>                                  
<while stat> ::=                            
`while' `(' <pexpr> `)' <stat>              `(six.while $3 $5)'
<do stat> ::=                               
`do' <stat> `while' `(' <pexpr> `)' `;'     `(six.do-while $2 $5)'
<switch stat> ::=                           
`switch' `(' <pexpr> `)' <stat>             `(six.switch $3 $5)'
<case stat> ::=                             
`case' <expr> `:' <stat>                    `(six.case $2 $4)'
<break stat> ::=                            
`break' `;'                                 `(six.break)'
<continue stat> ::=                         
`continue' `;'                              `(six.continue)'
<label stat> ::=                            
<identifier> `:' <stat>                     `(six.label $1 $3)'
<goto stat> ::=                             
`goto' <expr> `;'                           `(six.goto $2)'
<return stat> ::=                           
`return' `;'                                `(six.return)'
| `return' <expr> `;'                       `(six.return $2)'
<expression stat> ::=                       
<expr> `;'                                  $1
<clause stat> ::=                           
<expr> `.'                                  `(six.clause $1)'
<pexpr> ::=                                 
<procedure definition>                      $1
| <variable definition>                     $1
| <expr>                                    $1
<procedure definition> ::=                  
<type> <id-or-prefix> `(' <parameters> `)'  `(six.define-procedure $2
<body>                                      (six.procedure $1 $4 $6))'
<variable definition> ::=                   
<type> <id-or-prefix> <dimensions> <iexpr>  `(six.define-variable $2 $1
                                            $3 $4)'
<iexpr> ::=                                 
`=' <expr>                                  $2
|                                           `#f'
<dimensions> ::=                            
| `[' <expr> `]' <dimensions>               `($2 . $4)'
|                                           `()'
<oexpr> ::=                                 
<expr>                                      $1
|                                           `#f'
<expr> ::=                                  
<expr18>                                    $1
<expr18> ::=                                
<expr17> `:-' <expr18>                      `(six.x:-y $1 $3)'
| <expr17>                                  $1
<expr17> ::=                                
<expr17> `,' <expr16>                       `(|six.x,y| $1 $3)'
| <expr16>                                  $1
<expr16> ::=                                
<expr15> `:=' <expr16>                      `(six.x:=y $1 $3)'
| <expr15>                                  $1
<expr15> ::=                                
<expr14> `%=' <expr15>                      `(six.x%=y $1 $3)'
| <expr14> `&=' <expr15>                    `(six.x&=y $1 $3)'
| <expr14> `*=' <expr15>                    `(six.x*=y $1 $3)'
| <expr14> `+=' <expr15>                    `(six.x+=y $1 $3)'
| <expr14> `-=' <expr15>                    `(six.x-=y $1 $3)'
| <expr14> `/=' <expr15>                    `(six.x/=y $1 $3)'
| <expr14> `<<=' <expr15>                   `(six.x<<=y $1 $3)'
| <expr14> `=' <expr15>                     `(six.x=y $1 $3)'
| <expr14> `>>=' <expr15>                   `(six.x>>=y $1 $3)'
| <expr14> `^=' <expr15>                    `(six.x^=y $1 $3)'
| <expr14> `|=' <expr15>                    `(|six.x\|=y| $1 $3)'
| <expr14>                                  $1
<expr14> ::=                                
<expr13> `:' <expr14>                       `(six.x:y $1 $3)'
| <expr13>                                  $1
<expr13> ::=                                
<expr12> `?' <expr> `:' <expr13>            `(six.x?y:z $1 $3 $5)'
| <expr12>                                  $1
<expr12> ::=                                
<expr12> `||' <expr11>                      `(|six.x\|\|y| $1 $3)'
| <expr11>                                  $1
<expr11> ::=                                
<expr11> `&&' <expr10>                      `(six.x&&y $1 $3)'
| <expr10>                                  $1
<expr10> ::=                                
<expr10> `|' <expr9>                        `(|six.x\|y| $1 $3)'
| <expr9>                                   $1
<expr9> ::=                                 
<expr9> `^' <expr8>                         `(six.x^y $1 $3)'
| <expr8>                                   $1
<expr8> ::=                                 
<expr8> `&' <expr7>                         `(six.x&y $1 $3)'
| <expr7>                                   $1
<expr7> ::=                                 
<expr7> `!=' <expr6>                        `(six.x!=y $1 $3)'
| <expr7> `==' <expr6>                      `(six.x==y $1 $3)'
| <expr6>                                   $1
<expr6> ::=                                 
<expr6> `<' <expr5>                         `(six.x<y $1 $3)'
| <expr6> `<=' <expr5>                      `(six.x<=y $1 $3)'
| <expr6> `>' <expr5>                       `(six.x>y $1 $3)'
| <expr6> `>=' <expr5>                      `(six.x>=y $1 $3)'
| <expr5>                                   $1
<expr5> ::=                                 
<expr5> `<<' <expr4>                        `(six.x<<y $1 $3)'
| <expr5> `>>' <expr4>                      `(six.x>>y $1 $3)'
| <expr4>                                   $1
<expr4> ::=                                 
<expr4> `+' <expr3>                         `(six.x+y $1 $3)'
| <expr4> `-' <expr3>                       `(six.x-y $1 $3)'
| <expr3>                                   $1
<expr3> ::=                                 
<expr3> `%' <expr2>                         `(six.x%y $1 $3)'
| <expr3> `*' <expr2>                       `(six.x*y $1 $3)'
| <expr3> `/' <expr2>                       `(six.x/y $1 $3)'
| <expr2>                                   $1
<expr2> ::=                                 
`&' <expr2>                                 `(six.&x $2)'
| `+' <expr2>                               `(six.+x $2)'
| `-' <expr2>                               `(six.-x $2)'
| `*' <expr2>                               `(six.*x $2)'
| `!' <expr2>                               `(six.!x $2)'
| `!'                                       `(six.!)'
| `++' <expr2>                              `(six.++x $2)'
| `--' <expr2>                              `(six.--x $2)'
| `~' <expr2>                               `(six.~x $2)'
| `new' <id-or-prefix> `(' <arguments> `)'  `(six.new $2 . $4)'
| <expr1>                                   $1
<expr1> ::=                                 
<expr1> `++'                                `(six.x++ $1)'
| <expr1> `--'                              `(six.x-- $1)'
| <expr1> `(' <arguments> `)'               `(six.call $1 . $3)'
| <expr1> `[' <expr> `]'                    `(six.index $1 $3)'
| <expr1> `->' <id-or-prefix>               `(six.arrow $1 $3)'
| <expr1> `.' <id-or-prefix>                `(six.dot $1 $3)'
| <expr0>                                   $1
<expr0> ::=                                 
<id-or-prefix>                              $1
| <string>                                  `(six.literal $1)'
| <char>                                    `(six.literal $1)'
| <number>                                  `(six.literal $1)'
| `(' <expr> `)'                            $2
| `(' <block stat> `)'                      $2
| <datum-starting-with-#-or-backquote>      `(six.prefix $1)'
| `[' <elements> `]'                        $2
| <type> `(' <parameters> `)' <body>        `(six.procedure $1 $3 $5)'
<block stat> ::=                            
`{' <stat list> `}'                         `(six.compound . $2)'
<body> ::=                                  
`{' <stat list> `}'                         `(six.procedure-body . $2)'
<stat list> ::=                             
<stat> <stat list>                          `($1 . $2)'
|                                           `()'
<parameters> ::=                            
<nonempty parameters>                       $1
|                                           `()'
<nonempty parameters> ::=                   
<parameter> `,' <nonempty parameters>       `($1 . $3)'
| <parameter>                               `($1)'
<parameter> ::=                             
<type> <id-or-prefix>                       `($2 $1)'
<arguments> ::=                             
<nonempty arguments>                        $1
|                                           `()'
<nonempty arguments> ::=                    
<expr> `,' <nonempty arguments>             `($1 . $3)'
| <expr>                                    `($1)'
<elements> ::=                              
<nonempty elements>                         $1
|                                           `(six.null)'
<nonempty elements> ::=                     
<expr>                                      `(six.list $1 (six.null))'
| <expr> `,' <nonempty elements>            `(six.list $1 $3)'
| <expr> `|' <expr>                         `(six.cons $1 $3)'
<id-or-prefix> ::=                          
<identifier>                                `(six.identifier $1)'
| `\' <datum>                               `(six.prefix $2)'
<type> ::=                                  
`int'                                       `int'
| `char'                                    `char'
| `bool'                                    `bool'
| `void'                                    `void'
| `float'                                   `float'
| `double'                                  `double'
| `obj'                                     `obj'

18.12.2 SIX semantics
---------------------

The semantics of SIX depends on the definition of the `six.XXX'
identifiers (as functions and macros).  Many of these identifiers are
predefined macros which give SIX a semantics that is close to C's.  The
user may override these definitions to change the semantics either
globally or locally.  For example, `six.x^y' is a predefined macro that
expands `(six.x^y x y)' into `(bitwise-xor x y)'.  If the user prefers
the `^' operator to express exponentiation in a specific function, then
in that function `six.x^y' can be redefined as a macro that expands
`(six.x^y x y)' into `(expt x y)'.  Note that the associativity and
precedence of operators cannot be changed as that is a syntactic issue.

   Note that the following identifiers are not predefined, and
consequently they do not have a predefined semantics: `six.label',
`six.goto', `six.switch', `six.case', `six.break', `six.continue',
`six.return', `six.clause', `six.x:-y', and `six.!'.

   The following is an example showing some of the predefined semantics
of SIX:

     > (list (+ 1 2) \3+4; (+ 5 6))
     (3 7 11)
     > \[ 1+2, \(+ 3 4), 5+6 ];
     (3 7 11)
     > (map (lambda (x) \(x*x-1)/log(x+1);) '(1 2 3 4))
     (0 2.730717679880512 5.7707801635558535 9.320024018394177)
     > \obj n = expt(10,5);
     > n
     100000
     > \obj t[3][10] = 88;
     > \t[0][9] = t[2][1] = 11;
     11
     > t
     #(#(88 88 88 88 88 88 88 88 88 11)
       #(88 88 88 88 88 88 88 88 88 88)
       #(88 11 88 88 88 88 88 88 88 88))
     > \obj radix = new parameter (10);
     > \radix(2);
     > \radix();
     2
     > \for (int i=0; i<5; i++) pp(1<<i*8);
     1
     256
     65536
     16777216
     4294967296
     > \obj \make-adder (obj x) { obj (obj y) { x+y; }; }
     > \map (new adder (100), [1,2,3,4]);
     (101 102 103 104)
     > (map (make-adder 100) (list 1 2 3 4))
     (101 102 103 104)

19 C-interface
**************

The Gambit Scheme system offers a mechanism for interfacing Scheme code
and C code called the "C-interface".  A Scheme program indicates which
C functions it needs to have access to and which Scheme procedures can
be called from C, and the C interface automatically constructs the
corresponding Scheme procedures and C functions.  The conversions needed
to transform data from the Scheme representation to the C representation
(and back), are generated automatically in accordance with the argument
and result types of the C function or Scheme procedure.

   The C-interface places some restrictions on the types of data that
can be exchanged between C and Scheme.  The mapping of data types
between C and Scheme is discussed in the next section.  The remaining
sections of this chapter describe each special form of the C-interface.

19.1 The mapping of types between C and Scheme
==============================================

Scheme and C do not provide the same set of built-in data types so it is
important to understand which Scheme type is compatible with which C
type and how values get mapped from one environment to the other.  To
improve compatibility a new type is added to Scheme, the `foreign'
object type, and the following data types are added to C:

`scheme-object'
     denotes the universal type of Scheme objects (type `___SCMOBJ'
     defined in `gambit.h')

`bool'
     denotes the C++ `bool' type or the C `int' type (type `___BOOL'
     defined in `gambit.h')

`int8'
     8 bit signed integer (type `___S8' defined in `gambit.h')

`unsigned-int8'
     8 bit unsigned integer (type `___U8' defined in `gambit.h')

`int16'
     16 bit signed integer (type `___S16' defined in `gambit.h')

`unsigned-int16'
     16 bit unsigned integer (type `___U16' defined in `gambit.h')

`int32'
     32 bit signed integer (type `___S32' defined in `gambit.h')

`unsigned-int32'
     32 bit unsigned integer (type `___U32' defined in `gambit.h')

`int64'
     64 bit signed integer (type `___S64' defined in `gambit.h')

`unsigned-int64'
     64 bit unsigned integer (type `___U64' defined in `gambit.h')

`float32'
     32 bit floating point number (type `___F32' defined in `gambit.h')

`float64'
     64 bit floating point number (type `___F64' defined in `gambit.h')

`ISO-8859-1'
     denotes ISO-8859-1 encoded characters (8 bit unsigned integer,
     type `___ISO_8859_1' defined in `gambit.h')

`UCS-2'
     denotes UCS-2 encoded characters (16 bit unsigned integer, type
     `___UCS_2' defined in `gambit.h')

`UCS-4'
     denotes UCS-4 encoded characters (32 bit unsigned integer, type
     `___UCS_4' defined in `gambit.h')

`char-string'
     denotes the C `char*' type when used as a null terminated string

`nonnull-char-string'
     denotes the nonnull C `char*' type when used as a null terminated
     string

`nonnull-char-string-list'
     denotes an array of nonnull C `char*' terminated with a null
     pointer

`ISO-8859-1-string'
     denotes ISO-8859-1 encoded strings (null terminated string of 8
     bit unsigned integers, i.e. `___ISO_8859_1*')

`nonnull-ISO-8859-1-string'
     denotes nonnull ISO-8859-1 encoded strings (null terminated string
     of 8 bit unsigned integers, i.e. `___ISO_8859_1*')

`nonnull-ISO-8859-1-stringlist'
     denotes an array of nonnull ISO-8859-1 encoded strings terminated
     with a null pointer

`UTF-8-string'
     denotes UTF-8 encoded strings (null terminated string of `char',
     i.e. `char*')

`nonnull-UTF-8-string'
     denotes nonnull UTF-8 encoded strings (null terminated string of
     `char', i.e. `char*')

`nonnull-UTF-8-string-list'
     denotes an array of nonnull UTF-8 encoded strings terminated with
     a null pointer

`UTF-16-string'
     denotes UTF-16 encoded strings (null terminated string of `char',
     i.e. `char*')

`nonnull-UTF-16-string'
     denotes nonnull UTF-16 encoded strings (null terminated string of
     `char', i.e. `char*')

`nonnull-UTF-16-string-list'
     denotes an array of nonnull UTF-16 encoded strings terminated with
     a null pointer

`UCS-2-string'
     denotes UCS-2 encoded strings (null terminated string of 16 bit
     unsigned integers, i.e. `___UCS_2*')

`nonnull-UCS-2-string'
     denotes nonnull UCS-2 encoded strings (null terminated string of
     16 bit unsigned integers, i.e. `___UCS_2*')

`nonnull-UCS-2-string-list'
     denotes an array of nonnull UCS-2 encoded strings terminated with
     a null pointer

`UCS-4-string'
     denotes UCS-4 encoded strings (null terminated string of 32 bit
     unsigned integers, i.e. `___UCS_4*')

`nonnull-UCS-4-string'
     denotes nonnull UCS-4 encoded strings (null terminated string of
     32 bit unsigned integers, i.e. `___UCS_4*')

`nonnull-UCS-4-string-list'
     denotes an array of nonnull UCS-4 encoded strings terminated with
     a null pointer

`wchar_t-string'
     denotes `wchar_t' encoded strings (null terminated string of
     `wchar_t', i.e. `wchar_t*')

`nonnull-wchar_t-string'
     denotes nonnull `wchar_t' encoded strings (null terminated string
     of `wchar_t', i.e. `wchar_t*')

`nonnull-wchar_t-string-list'
     denotes an array of nonnull `wchar_t' encoded strings terminated
     with a null pointer

   To specify a particular C type inside the `c-lambda', `c-define' and
`c-define-type' forms, the following "Scheme notation" is used:

`Scheme notation'
     C type

`void'
     `void'

`bool'
     `bool'

`char'
     `char'  (may be signed or unsigned depending on the C compiler)

`signed-char'
     `signed char'

`unsigned-char'
     `unsigned char'

`ISO-8859-1'
     `ISO-8859-1'

`UCS-2'
     `UCS-2'

`UCS-4'
     `UCS-4'

`wchar_t'
     `wchar_t'

`size_t'
     `size_t' (type `___SIZE_T' defined in `gambit.h')

`ssize_t'
     `ssize_t' (type `___SSIZE_T' defined in `gambit.h')

`ptrdiff_t'
     `ptrdiff_t' (type `___PTRDIFF_T' defined in `gambit.h')

`short'
     `short'

`unsigned-short'
     `unsigned short'

`int'
     `int'

`unsigned-int'
     `unsigned int'

`long'
     `long'

`unsigned-long'
     `unsigned long'

`long-long'
     `long long'

`unsigned-long-long'
     `unsigned long long'

`float'
     `float'

`double'
     `double'

`int8'
     `int8'

`unsigned-int8'
     `unsigned-int8'

`int16'
     `int16'

`unsigned-int16'
     `unsigned-int16'

`int32'
     `int32'

`unsigned-int32'
     `unsigned-int32'

`int64'
     `int64'

`unsigned-int64'
     `unsigned-int64'

`float32'
     `float32'

`float64'
     `float64'

`(struct "C-STRUCT-ID" [TAGS [RELEASE-FUNCTION]])'
     `struct C-STRUCT-ID'  (where C-STRUCT-ID is the name of a C
     structure; see below for the meaning of TAGS and RELEASE-FUNCTION)

`(union "C-UNION-ID" [TAGS [RELEASE-FUNCTION]])'
     `union C-UNION-ID'  (where C-UNION-ID is the name of a C union;
     see below for the meaning of TAGS and RELEASE-FUNCTION)

`(type "C-TYPE-ID" [TAGS [RELEASE-FUNCTION]])'
     `C-TYPE-ID'  (where C-TYPE-ID is an identifier naming a C type;
     see below for the meaning of TAGS and RELEASE-FUNCTION)

`(pointer TYPE [TAGS [RELEASE-FUNCTION]])'
     `T*'  (where T is the C equivalent of TYPE which must be the
     Scheme notation of a C type; see below for the meaning of TAGS and
     RELEASE-FUNCTION)

`(nonnull-pointer TYPE [TAGS [RELEASE-FUNCTION]])'
     same as `(pointer TYPE [TAGS [RELEASE-FUNCTION]])' except the
     `NULL' pointer is not allowed

`(function (TYPE1...) RESULT-TYPE)'
     function with the given argument types and result type

`(nonnull-function (TYPE1...) RESULT-TYPE)'
     same as `(function (TYPE1...) RESULT-TYPE)' except the `NULL'
     pointer is not allowed

`char-string'
     `char-string'

`nonnull-char-string'
     `nonnull-char-string'

`nonnull-char-string-list'
     `nonnull-char-string-list'

`ISO-8859-1-string'
     `ISO-8859-1-string'

`nonnull-ISO-8859-1-string'
     `nonnull-ISO-8859-1-string'

`nonnull-ISO-8859-1-string-list'
     `nonnull-ISO-8859-1-string-list'

`UTF-8-string'
     `UTF-8-string'

`nonnull-UTF-8-string'
     `nonnull-UTF-8-string'

`nonnull-UTF-8-string-list'
     `nonnull-UTF-8-string-list'

`UTF-16-string'
     `UTF-16-string'

`nonnull-UTF-16-string'
     `nonnull-UTF-16-string'

`nonnull-UTF-16-string-list'
     `nonnull-UTF-16-string-list'

`UCS-2-string'
     `UCS-2-string'

`nonnull-UCS-2-string'
     `nonnull-UCS-2-string'

`nonnull-UCS-2-string-list'
     `nonnull-UCS-2-string-list'

`UCS-4-string'
     `UCS-4-string'

`nonnull-UCS-4-string'
     `nonnull-UCS-4-string'

`nonnull-UCS-4-string-list'
     `nonnull-UCS-4-string-list'

`wchar_t-string'
     `wchar_t-string'

`nonnull-wchar_t-string'
     `nonnull-wchar_t-string'

`nonnull-wchar_t-string-list'
     `nonnull-wchar_t-string-list'

`scheme-object'
     `scheme-object'

`NAME'
     appropriate translation of NAME (where NAME is a C type defined
     with `c-define-type')

`"C-TYPE-ID"'
     C-TYPE-ID (this form is equivalent to `(type "C-TYPE-ID")')

   The `struct', `union', `type', `pointer' and `nonnull-pointer' types
are "foreign types" and they are represented on the Scheme side as
"foreign objects".  A foreign object is internally represented as a
pointer.  This internal pointer is identical to the C pointer being
represented in the case of the `pointer' and `nonnull-pointer' types.

   In the case of the `struct', `union' and `type' types, the internal
pointer points to a copy of the C data type being represented.  When an
instance of one of these types is converted from C to Scheme, a block
of memory is allocated from the C heap and initialized with the
instance and then a foreign object is allocated from the Scheme heap
and initialized with the pointer to this copy.  This approach may
appear overly complex, but it allows the conversion of C++ classes that
do not have a zero parameter constructor or an assignment method (i.e.
when compiling with a C++ compiler an instance is copied using `new
TYPE (INSTANCE)', which calls the copy-constructor of TYPE if it is a
class; TYPE's assignment operator is never used).  Conversion from
Scheme to C simply dereferences the internal pointer (no allocation
from the C heap is performed).  Deallocation of the copy on the C heap
is under the control of the release function attached to the foreign
object (see below).

   The optional TAGS field of foreign type specifications is used for
type checking on the Scheme side.  The TAGS field must be `#f', a
symbol or a non-empty list of symbols.  When it is not specified the
TAGS field defaults to a symbol whose name, as returned by
`symbol->string', is the C type declaration for that type.  For example
the symbol `char**' is the default for the type `(pointer (pointer
char))'.  A TAGS field that is a single symbol is equivalent to a list
containing only that symbol.  The first symbol in the list of tags is
the primary tag.  For example the primary tag of the type `(pointer
char)' is `char*' and the primary tag of the type `(pointer char (foo
bar))' is `foo'.

   Type compatibility between two foreign types depends on their tags.
An instance of a foreign type T can be used where a foreign type E is
expected if and only if

   * T's TAGS field is `#f', or

   * E's TAGS field is `#f', or

   * T's primary tag is a member of E's tags.


   For the safest code a TAGS field of `#f' should be used sparingly,
as it completely bypasses type checking.  The external representation
of Scheme foreign objects (used by the `write' procedure) contains the
primary tag (if the TAGS field is not `#f'), and the hexadecimal
address denoted by the internal pointer, for example `#<char** #2
0x2AAC535C>'.  Note that the hexadecimal address is in C notation,
which can be easily transferred to a C debugger with a "cut-and-paste".

   A RELEASE-FUNCTION can also be specified within a foreign type
specification.  The RELEASE-FUNCTION must be `#f' or a string naming a
C function with a single parameter of type `void*' (in which the
internal pointer is passed) and with a result of type `___SCMOBJ' (for
returning an error code).  When the RELEASE-FUNCTION is not specified
or is `#f' a default function is constructed by the C-interface.  This
default function does nothing in the case of the `pointer' and
`nonnull-pointer' types (deallocation is not the responsibility of the
C-interface) and returns the fixnum `___FIX(___NO_ERR)' to indicate no
error.  In the case of the `struct', `union' and `type' types, the
default function reclaims the copy on the C heap referenced by the
internal pointer (when using a C++ compiler this is done using `delete
(TYPE*)INTERNAL-POINTER', which calls the destructor of TYPE if it is a
class) and returns `___FIX(___NO_ERR)'.  In many situations the default
RELEASE-FUNCTION will perform the appropriate cleanup for the foreign
type.  However, in certain cases special operations (such as
decrementing a reference count, removing the object from a table, etc)
must be performed.  For such cases a user supplied RELEASE-FUNCTION is
needed.

   The RELEASE-FUNCTION is invoked at most once for any foreign object.
After the RELEASE-FUNCTION is invoked, the foreign object is
considered "released" and can no longer be used in a foreign type
conversion.  When the garbage collector detects that a foreign object
is no longer reachable by the program, it will invoke the
RELEASE-FUNCTION if the foreign object is not yet released.  When there
is a need to release the foreign object promptly, the program can
explicitly call `(foreign-release! OBJ)' which invokes the
RELEASE-FUNCTION if the foreign object is not yet released, and does
nothing otherwise.  The call `(foreign-released? OBJ)' returns a
boolean indicating whether the foreign object OBJ has been released yet
or not.  The call `(foreign-address OBJ)' returns the address denoted
by the internal pointer of foreign object OBJ or 0 if it has been
released.  The call `(foreign? OBJ)' tests that OBJ is a foreign
object.  Finally the call `(foreign-tags OBJ)' returns the list of tags
of foreign object OBJ, or `#f'.

   The following table gives the C types to which each Scheme type can
be converted:

Scheme type
     Allowed target C types

boolean `#f'
     `scheme-object'; `bool'; `pointer'; `function'; `char-string';
     `ISO-8859-1-string'; `UTF-8-string'; `UTF-16-string';
     `UCS-2-string'; `UCS-4-string'; `wchar_t-string'

boolean `#t'
     `scheme-object'; `bool'

character
     `scheme-object'; `bool'; [`[un]signed'] `char'; `ISO-8859-1';
     `UCS-2'; `UCS-4'; `wchar_t'

exact integer
     `scheme-object'; `bool'; [`unsigned-']
     `int8'/`int16'/`int32'/`int64'; [`unsigned'] `short'/`int'/`long';
     `size_t'/`ssize_t'/`ptrdiff_t'

inexact real
     `scheme-object'; `bool'; `float'; `double'; `float32'; `float64'

string
     `scheme-object'; `bool'; `char-string'; `nonnull-char-string';
     `ISO-8859-1-string'; `nonnull-ISO-8859-1-string'; `UTF-8-string';
     `nonnull-UTF-8-string'; `UTF-16-string'; `nonnull-UTF-16-string';
     `UCS-2-string'; `nonnull-UCS-2-string'; `UCS-4-string';
     `nonnull-UCS-4-string'; `wchar_t-string'; `nonnull-wchar_t-string'

foreign object
     `scheme-object'; `bool';
     `struct'/`union'/`type'/`pointer'/`nonnull-pointer' with the
     appropriate tags

vector
     `scheme-object'; `bool'

symbol
     `scheme-object'; `bool'

procedure
     `scheme-object'; `bool'; `function'; `nonnull-function'

other objects
     `scheme-object'; `bool'

   The following table gives the Scheme types to which each C type will
be converted:

C type
     Resulting Scheme type

scheme-object
     the Scheme object encoded

bool
     boolean

[`[un]signed'] `char'; `ISO-8859-1'; `UCS-2'; `UCS-4'; `wchar_t'
     character

[`unsigned-'] `int8'/`int16'/`int32'/`int64'; [`unsigned'] `short'/`int'/`long'; `size_t'/`ssize_t'/`ptrdiff_t'
     exact integer

`float'; `double'; `float32'; `float64'
     inexact real

`char-string'; `ISO-8859-1-string'; `UTF-8-string'; `UTF-16-string'; `UCS-2-string'; `UCS-4-string'; `wchar_t-string'
     string or `#f' if it is equal to `NULL'

`nonnull-char-string'; `nonnull-ISO-8859-1-string'; `nonnull-UTF-8-string'; `nonnull-UTF-16-string'; `nonnull-UCS-2-string'; `nonnull-UCS-4-string'; `nonnull-wchar_t-string'
     string

`struct'/`union'/`type'/`pointer'/`nonnull-pointer'
     foreign object with the appropriate tags or `#f' in the case of a
     `pointer' equal to `NULL'

`function'
     procedure or `#f' if it is equal to `NULL'

`nonnull-function'
     procedure

`void'
     void object

   All Scheme types are compatible with the C types `scheme-object' and
`bool'.  Conversion to and from the C type `scheme-object' is the
identity function on the object encoding.  This provides a low-level
mechanism for accessing Scheme's object representation from C (with the
help of the macros in the `gambit.h' header file).  When a C `bool'
type is expected, an extended Scheme boolean can be passed (`#f' is
converted to 0 and all other values are converted to 1).

   The Scheme boolean `#f' can be passed to the C environment where a
`char-string', `ISO-8859-1-string', `UTF-8-string', `UTF-16-string',
`UCS-2-string', `UCS-4-string', `wchar_t-string', `pointer' or
`function' type is expected.  In this case, `#f' is converted to the
`NULL' pointer.  C `bool's are extended booleans so any value different
from 0 represents true.  Thus, a C `bool' passed to the Scheme
environment is mapped as follows: 0 to `#f' and all other values to
`#t'.

   A Scheme character passed to the C environment where any C character
type is expected is converted to the corresponding character in the C
environment.  An error is signaled if the Scheme character does not fit
in the C character.  Any C character type passed to Scheme is converted
to the corresponding Scheme character.  An error is signaled if the C
character does not fit in the Scheme character.

   A Scheme exact integer passed to the C environment where a C integer
type (other than `char') is expected is converted to the corresponding
integral value.  An error is signaled if the value falls outside of the
range representable by that integral type.  C integer values passed to
the Scheme environment are mapped to the same Scheme exact integer.  If
the value is outside the fixnum range, a bignum is created.

   A Scheme inexact real passed to the C environment is converted to the
corresponding `float', `double', `float32' or `float64' value.  C
`float', `double', `float32' and `float64' values passed to the Scheme
environment are mapped to the closest Scheme inexact real.

   Scheme's rational numbers and complex numbers are not compatible with
any C numeric type.

   A Scheme string passed to the C environment where any C string type
is expected is converted to a null terminated string using the
appropriate encoding.  The C string is a fresh copy of the Scheme
string.  If the C string was created for an argument of a `c-lambda',
the C string will be reclaimed when the `c-lambda' returns.  If the C
string was created for returning the result of a `c-define' to C, the
caller is responsible for reclaiming the C string with a call to the
`___release_string' function (see below for an example).  Any C string
type passed to the Scheme environment causes the creation of a fresh
Scheme string containing a copy of the C string (unless the C string is
equal to `NULL', in which case it is converted to `#f').

   A foreign type passed to the Scheme environment causes the creation
and initialization of a Scheme foreign object with the appropriate tags
(except for the case of a `pointer' equal to `NULL' which is converted
to `#f').  A Scheme foreign object can be passed where a foreign type
is expected, on the condition that the tags are compatible and the
Scheme foreign object is not yet released.  The value `#f' is also
acceptable for a `pointer' type, and is converted to `NULL'.

   Scheme procedures defined with the `c-define' special form can be
passed where the `function' and `nonnull-function' types are expected.
The value `#f' is also acceptable for a `function' type, and is
converted to `NULL'.  No other Scheme procedures are acceptable.
Conversion from the `function' and `nonnull-function' types to Scheme
procedures is not currently implemented.

19.2 The `c-declare' special form
=================================

 -- special form: c-declare c-declaration
     Initially, the C file produced by `gsc' contains only an
     `#include' of `gambit.h'.  This header file provides a number of
     macro and procedure declarations to access the Scheme object
     representation.  The special form `c-declare' adds c-declaration
     (which must be a string containing the C declarations) to the C
     file.  This string is copied to the C file on a new line so it can
     start with preprocessor directives.  All types of C declarations
     are allowed (including type declarations, variable declarations,
     function declarations, `#include' directives, `#define's, and so
     on).  These declarations are visible to subsequent `c-declare's,
     `c-initialize's, and `c-lambda's, and `c-define's in the same
     module.  The most common use of this special form is to declare
     the external functions that are referenced in `c-lambda' special
     forms.  Such functions must either be declared explicitly or by
     including a header file which contains the appropriate C
     declarations.

     The `c-declare' special form does not return a value.  It can only
     appear at top level.

     For example:

          (c-declare #<<c-declare-end

          #include <stdio.h>

          extern char *getlogin ();

          #ifdef sparc
          char *host = "sparc";
          #else
          char *host = "unknown";
          #endif

          FILE *tfile;

          c-declare-end
          )


19.3 The `c-initialize' special form
====================================

 -- special form: c-initialize c-code
     Just after the program is loaded and before control is passed to
     the Scheme code, each C file is initialized by calling its
     associated initialization function.  The body of this function is
     normally empty but it can be extended by using the `c-initialize'
     form.  Each occurence of the `c-initialize' form adds code to the
     body of the initialization function in the order of appearance in
     the source file.  c-code must be a string containing the C code to
     execute.  This string is copied to the C file on a new line so it
     can start with preprocessor directives.

     The `c-initialize' special form does not return a value.  It can
     only appear at top level.

     For example:

          (c-initialize "tfile = tmpfile ();")


19.4 The `c-lambda' special form
================================

 -- special form: c-lambda (type1...) result-type c-name-or-code
     The `c-lambda' special form makes it possible to create a Scheme
     procedure that will act as a representative of some C function or
     C code sequence.  The first subform is a list containing the type
     of each argument.  The type of the function's result is given
     next.  Finally, the last subform is a string that either contains
     the name of the C function to call or some sequence of C code to
     execute.  Variadic C functions are not supported.  The resulting
     Scheme procedure takes exactly the number of arguments specified
     and delivers them in the same order to the C function.  When the
     Scheme procedure is called, the arguments will be converted to
     their C representation and then the C function will be called.
     The result returned by the C function will be converted to its
     Scheme representation and this value will be returned from the
     Scheme procedure call.  An error will be signaled if some
     conversion is not possible.  The temporary memory allocated from
     the C heap for the conversion of the arguments and result will be
     reclaimed whether there is an error or not.

     When c-name-or-code is not a valid C identifier, it is treated as
     an arbitrary piece of C code.  Within the C code the variables
     `___arg1', `___arg2', etc. can be referenced to access the
     converted arguments.  Note that the C `return' statement can't be
     used to return from the procedure.  Instead, the `___return' macro
     must be used.  A procedure whose result-type is not `void' must
     pass the procedure's result as the single argument to the
     `___return' macro, for example `___return(123);' to return the
     value 123.  When result-type is `void', the `___return' macro must
     be called without a parameter list, for example `___return;'.

     The C code is copied to the C file on a new line so it can start
     with preprocessor directives.  Moreover the C code is always
     placed at the head of a compound statement whose lifetime encloses
     the C to Scheme conversion of the procedure's result.
     Consequently, temporary storage (strings in particular) declared
     at the head of the C code can be returned with the `___return'
     macro.

     In the c-name-or-code, the macro `___AT_END' may be defined as the
     piece of C code to execute before control is returned to Scheme
     but after the procedure's result is converted to its Scheme
     representation.  This is mainly useful to deallocate temporary
     storage contained in the result.

     When passed to the Scheme environment, the C `void' type is
     converted to the void object.

     For example:

          (define fopen
            (c-lambda (nonnull-char-string nonnull-char-string)
                      (pointer "FILE")
             "fopen"))

          (define fgetc
            (c-lambda ((pointer "FILE"))
                      int
             "fgetc"))

          (let ((f (fopen "datafile" "r")))
            (if f (write (fgetc f))))

          (define char-code
            (c-lambda (char) int "___return(___arg1);"))

          (define host
            ((c-lambda () nonnull-char-string "___return(host);")))

          (define stdin
            ((c-lambda () (pointer "FILE") "___return(stdin);")))

          ((c-lambda () void
          #<<c-lambda-end
            printf( "hello\n" );
            printf( "world\n" );
          c-lambda-end
          ))

          (define pack-1-char
            (c-lambda (char)
                      nonnull-char-string
          #<<c-lambda-end
             char *s = (char *)malloc (2);
             if (s != NULL) { s[0] = ___arg1; s[1] = 0; }
             #define ___AT_END if (s != NULL) free (s);
          c-lambda-end
          ))

          (define pack-2-chars
            (c-lambda (char char)
                      nonnull-char-string
          #<<c-lambda-end
             char s[3];
             s[0] = ___arg1;
             s[1] = ___arg2;
             s[2] = 0;
             ___return(s);
          c-lambda-end
          ))


19.5 The `c-define' special form
================================

 -- special form: c-define (variable define-formals) (type1...)
          result-type c-name scope body
     The `c-define' special form makes it possible to create a C
     function that will act as a representative of some Scheme
     procedure.  A C function named c-name as well as a Scheme
     procedure bound to the variable variable are defined.  The
     parameters of the Scheme procedure are define-formals and its body
     is at the end of the form.  The type of each argument of the C
     function, its result type and c-name (which must be a string) are
     specified after the parameter specification of the Scheme
     procedure.  When the C function c-name is called from C, its
     arguments are converted to their Scheme representation and passed
     to the Scheme procedure.  The result of the Scheme procedure is
     then converted to its C representation and the C function c-name
     returns it to its caller.

     The scope of the C function can be changed with the scope
     parameter, which must be a string.  This string is placed
     immediately before the declaration of the C function.  So if scope
     is the string `"static"', the scope of c-name is local to the
     module it is in, whereas if scope is the empty string, c-name is
     visible from other modules.

     The `c-define' special form does not return a value.  It can only
     appear at top level.

     For example:

          (c-define (proc x #!optional (y x) #!rest z) (int int char float) int "f" ""
            (write (cons x (cons y z)))
            (newline)
            (+ x y))

          (proc 1 2 #\x 1.5) => 3 and prints (1 2 #\x 1.5)
          (proc 1)           => 2 and prints (1 1)

          ; if f is called from C with the call  f (1, 2, 'x', 1.5)
          ; the value 3 is returned and (1 2 #\x 1.5) is printed.
          ; f has to be called with 4 arguments.

     The `c-define' special form is particularly useful when the
     driving part of an application is written in C and Scheme
     procedures are called directly from C.  The Scheme part of the
     application is in a sense a "server" that is providing services to
     the C part.  The Scheme procedures that are to be called from C
     need to be defined using the `c-define' special form.  Before it
     can be used, the Scheme part must be initialized with a call to
     the function `___setup'.  Before the program terminates, it must
     call the function `___cleanup' so that the Scheme part may do final
     cleanup.  A sample application is given in the file
     `tests/server.scm'.


19.6 The `c-define-type' special form
=====================================

 -- special form: c-define-type name type [c-to-scheme scheme-to-c
          [cleanup]]
     This form associates the type identifier name to the C type type.
     The name must not clash with predefined types (e.g. `char-string',
     `ISO-8859-1', etc.) or with types previously defined with
     `c-define-type' in the same file.  The `c-define-type' special
     form does not return a value.  It can only appear at top level.

     If only the two parameters name and type are supplied then after
     this definition, the use of name in a type specification is
     synonymous to type.

     For example:

          (c-define-type FILE "FILE")
          (c-define-type FILE* (pointer FILE))
          (c-define-type time-struct-ptr (pointer (struct "tms")))
          (define fopen (c-lambda (char-string char-string) FILE* "fopen"))
          (define fgetc (c-lambda (FILE*) int "fgetc"))

     Note that identifiers are not case-sensitive in standard Scheme
     but it is good programming practice to use a name with the same
     case as in C.

     If four or more parameters are supplied, then type must be a
     string naming the C type, c-to-scheme and scheme-to-c must be
     strings suffixing the C macros that convert data of that type
     between C and Scheme.  If cleanup is supplied it must be a boolean
     indicating whether it is necessary to perform a cleanup operation
     (such as freeing memory) when data of that type is converted from
     Scheme to C (it defaults to `#t').  The cleanup information is
     used when the C stack is unwound due to a continuation invocation
     (see *Note continuations::).  Although it is safe to always
     specify `#t', it is more efficient in time and space to specify
     `#f' because the unwinding mechanism can skip C-interface frames
     which only contain conversions of data types requiring no cleanup.
     Two pairs of C macros need to be defined for conversions
     performed by `c-lambda' forms and two pairs for conversions
     performed by `c-define' forms:

          ___BEGIN_CFUN_scheme-to-c(___SCMOBJ, type, int)
          ___END_CFUN_scheme-to-c(___SCMOBJ, type, int)

          ___BEGIN_CFUN_c-to-scheme(type, ___SCMOBJ)
          ___END_CFUN_c-to-scheme(type, ___SCMOBJ)

          ___BEGIN_SFUN_c-to-scheme(type, ___SCMOBJ, int)
          ___END_SFUN_c-to-scheme(type, ___SCMOBJ, int)

          ___BEGIN_SFUN_scheme-to-c(___SCMOBJ, type)
          ___END_SFUN_scheme-to-c(___SCMOBJ, type)

     The macros prefixed with `___BEGIN' perform the conversion and
     those prefixed with `___END' perform any cleanup necessary (such as
     freeing memory temporarily allocated for the conversion).  The
     macro `___END_CFUN_scheme-to-c' must free the result of the
     conversion if it is memory allocated, and
     `___END_SFUN_scheme-to-c' must not (i.e. it is the responsibility
     of the caller to free the result).

     The first parameter of these macros is the C variable that
     contains the value to be converted, and the second parameter is
     the C variable in which to store the converted value.  The third
     parameter, when present, is the index (starting at 1) of the
     parameter of the `c-lambda' or `c-define' form that is being
     converted (this is useful for reporting precise error information
     when a conversion is impossible).

     To allow for type checking, the first three `___BEGIN' macros must
     expand to an unterminated compound statement prefixed by an `if',
     conditional on the absence of type check error:

          if ((___err = CONVERSION_OPERATION) == ___FIX(___NO_ERR)) {

     The last `___BEGIN' macro must expand to an unterminated compound
     statement:

          { ___err = CONVERSION_OPERATION;

     If type check errors are impossible then a `___BEGIN' macro can
     simply expand to an unterminated compound statement performing the
     conversion:

          { CONVERSION_OPERATION;

     The `___END' macros must expand to a statement, or to nothing if no
     cleanup is required, followed by a closing brace (to terminate the
     compound statement started at the corresponding `___BEGIN' macro).

     The CONVERSION_OPERATION is typically a function call that returns
     an error code value of type `___SCMOBJ' (the error codes are
     defined in `gambit.h', and the error code `___FIX(___UNKNOWN_ERR)'
     is available for generic errors).  CONVERSION_OPERATION can also
     set the variable `___errmsg' of type `___SCMOBJ' to a specific
     Scheme string error message.

     Below is a simple example showing how to interface to an `EBCDIC'
     character type.  Memory allocation is not needed for conversion
     and type check errors are impossible when converting EBCDIC to
     Scheme characters, but they are possible when converting from
     Scheme characters to EBCDIC since Gambit supports Unicode
     characters.

          (c-declare #<<c-declare-end

          typedef char EBCDIC; /* EBCDIC encoded characters */

          void put_char (EBCDIC c) { ... } /* EBCDIC I/O functions */
          EBCDIC get_char (void) { ... }

          char EBCDIC_to_ISO_8859_1[256] = { ... }; /* conversion tables */
          char ISO_8859_1_to_EBCDIC[256] = { ... };

          ___SCMOBJ SCMOBJ_to_EBCDIC (___SCMOBJ src, EBCDIC *dst)
          {
            int x = ___INT(src); /* convert from Scheme character to int */
            if (x > 255) return ___FIX(___UNKNOWN_ERR);
            *dst = ISO_8859_1_to_EBCDIC[x];
            return ___FIX(___NO_ERR);
          }

          #define ___BEGIN_CFUN_SCMOBJ_to_EBCDIC(src,dst,i) \
          if ((___err = SCMOBJ_to_EBCDIC (src, &dst)) == ___FIX(___NO_ERR)) {
          #define ___END_CFUN_SCMOBJ_to_EBCDIC(src,dst,i) }

          #define ___BEGIN_CFUN_EBCDIC_to_SCMOBJ(src,dst) \
          { dst = ___CHR(EBCDIC_to_ISO_8859_1[src]);
          #define ___END_CFUN_EBCDIC_to_SCMOBJ(src,dst) }

          #define ___BEGIN_SFUN_EBCDIC_to_SCMOBJ(src,dst,i) \
          { dst = ___CHR(EBCDIC_to_ISO_8859_1[src]);
          #define ___END_SFUN_EBCDIC_to_SCMOBJ(src,dst,i) }

          #define ___BEGIN_SFUN_SCMOBJ_to_EBCDIC(src,dst) \
          { ___err = SCMOBJ_to_EBCDIC (src, &dst);
          #define ___END_SFUN_SCMOBJ_to_EBCDIC(src,dst) }

          c-declare-end
          )

          (c-define-type EBCDIC "EBCDIC" "EBCDIC_to_SCMOBJ" "SCMOBJ_to_EBCDIC" #f)

          (define put-char (c-lambda (EBCDIC) void "put_char"))
          (define get-char (c-lambda () EBCDIC "get_char"))

          (c-define (write-EBCDIC c) (EBCDIC) void "write_EBCDIC" ""
            (write-char c))

          (c-define (read-EBCDIC) () EBCDIC "read_EBCDIC" ""
            (read-char))

     Below is a more complex example that requires memory allocation
     when converting from C to Scheme.  It is an interface to a 2D
     `point' type which is represented in Scheme by a pair of integers.
     The conversion of the `x' and `y' components is done by calls to
     the conversion macros for the `int' type (defined in `gambit.h').
     Note that no cleanup is necessary when converting from Scheme to C
     (i.e. the last parameter of the `c-define-type' is `#f').

          (c-declare #<<c-declare-end

          typedef struct { int x, y; } point;

          void line_to (point p) { ... }
          point get_mouse (void) { ... }
          point add_points (point p1, point p2) { ... }

          ___SCMOBJ SCMOBJ_to_POINT (___PSD ___SCMOBJ src, point *dst, int arg_num)
          {
            ___SCMOBJ ___err = ___FIX(___NO_ERR);
            if (!___PAIRP(src))
              ___err = ___FIX(___UNKNOWN_ERR);
            else
              {
                ___SCMOBJ car = ___CAR(src);
                ___SCMOBJ cdr = ___CDR(src);
                ___BEGIN_CFUN_SCMOBJ_TO_INT(car,dst->x,arg_num)
                ___BEGIN_CFUN_SCMOBJ_TO_INT(cdr,dst->y,arg_num)
                ___END_CFUN_SCMOBJ_TO_INT(cdr,dst->y,arg_num)
                ___END_CFUN_SCMOBJ_TO_INT(car,dst->x,arg_num)
              }
            return ___err;
          }

          ___SCMOBJ POINT_to_SCMOBJ (___processor_state ___ps, point src, ___SCMOBJ *dst, int arg_num)
          {
            ___SCMOBJ ___err = ___FIX(___NO_ERR);
            ___SCMOBJ x_scmobj;
            ___SCMOBJ y_scmobj;
            ___BEGIN_SFUN_INT_TO_SCMOBJ(src.x,x_scmobj,arg_num)
            ___BEGIN_SFUN_INT_TO_SCMOBJ(src.y,y_scmobj,arg_num)
            *dst = ___EXT(___make_pair) (___ps, x_scmobj, y_scmobj);
            if (___FIXNUMP(*dst))
              ___err = *dst; /* return allocation error */
            ___END_SFUN_INT_TO_SCMOBJ(src.y,y_scmobj,arg_num)
            ___END_SFUN_INT_TO_SCMOBJ(src.x,x_scmobj,arg_num)
            return ___err;
          }

          #define ___BEGIN_CFUN_SCMOBJ_to_POINT(src,dst,i) \
          if ((___err = SCMOBJ_to_POINT (___PSP src, &dst, i)) == ___FIX(___NO_ERR)) {
          #define ___END_CFUN_SCMOBJ_to_POINT(src,dst,i) }

          #define ___BEGIN_CFUN_POINT_to_SCMOBJ(src,dst) \
          if ((___err = POINT_to_SCMOBJ (___ps, src, &dst, ___RETURN_POS)) == ___FIX(___NO_ERR)) {
          #define ___END_CFUN_POINT_to_SCMOBJ(src,dst) \
          ___EXT(___release_scmobj) (dst); }

          #define ___BEGIN_SFUN_POINT_to_SCMOBJ(src,dst,i) \
          if ((___err = POINT_to_SCMOBJ (___ps, src, &dst, i)) == ___FIX(___NO_ERR)) {
          #define ___END_SFUN_POINT_to_SCMOBJ(src,dst,i) \
          ___EXT(___release_scmobj) (dst); }

          #define ___BEGIN_SFUN_SCMOBJ_to_POINT(src,dst) \
          { ___err = SCMOBJ_to_POINT (___PSP src, &dst, ___RETURN_POS);
          #define ___END_SFUN_SCMOBJ_to_POINT(src,dst) }

          c-declare-end
          )

          (c-define-type point "point" "POINT_to_SCMOBJ" "SCMOBJ_to_POINT" #f)

          (define line-to (c-lambda (point) void "line_to"))
          (define get-mouse (c-lambda () point "get_mouse"))
          (define add-points (c-lambda (point point) point "add_points"))

          (c-define (write-point p) (point) void "write_point" ""
            (write p))

          (c-define (read-point) () point "read_point" ""
            (read))

     Note that the pair is allocated using the `___make_pair' runtime
     library function.  The prototype of this function is

          ___SCMOBJ ___make_pair(___processor_state ___ps, ___SCMOBJ car, ___SCMOBJ cdr);

     The fields of the pair are initialized to the `car' and `cdr'
     parameters.  The `___ps' parameter indicates how the pair is
     allocated.  A `NULL' `___ps' parameter will allocate the pair
     permanently (i.e. the pair will only be deallocated when
     `___cleanup' is called).  Otherwise a "still" object is allocated
     and the `___ps' parameter indicates the processor in whose heap
     the object is allocated (this is to support multithreaded
     execution).  Still objects are reference counted and initially
     have a reference count equal to 1.  The call to
     `___release_scmobj' in the macros `___END_CFUN_POINT_to_SCMOBJ' and
     `___END_SFUN_POINT_to_SCMOBJ' decrement this reference count.  A
     still object whose reference count is zero will be deallocated
     when a garbage collection is performed and there are no references
     to it from the Scheme world.  Note that the use of the `___PSD'
     macro in the parameter list of `SCMOBJ_to_POINT' and the `___PSP'
     macro in the calls of `SCMOBJ_to_POINT', are necessary to
     propagate the current processor state to that function.

     An example that requires memory allocation when converting from C
     to Scheme and Scheme to C is shown below.  It is an interface to a
     "null-terminated array of strings" type which is represented in
     Scheme by a list of strings.  Note that some cleanup is necessary
     when converting from Scheme to C.

          (c-declare #<<c-declare-end

          #include <stdlib.h>
          #include <unistd.h>

          extern char **environ;

          char **get_environ (void) { return environ; }

          void free_strings (char **strings)
          {
            char **ptr = strings;
            while (*ptr != NULL)
              {
                ___EXT(___release_string) (*ptr);
                ptr++;
              }
            free (strings);
          }

          ___SCMOBJ SCMOBJ_to_STRINGS (___PSD ___SCMOBJ src, char ***dst, int arg_num)
          {
            /*
             * Src is a list of Scheme strings.  Dst will be a null terminated
             * array of C strings.
             */

            int i;
            ___SCMOBJ lst = src;
            int len = 4; /* start with a small result array */
            char **result = (char**) malloc (len * sizeof (char*));

            if (result == NULL)
              return ___FIX(___HEAP_OVERFLOW_ERR);

            i = 0;
            result[i] = NULL; /* always keep array null terminated */

            while (___PAIRP(lst))
              {
                ___SCMOBJ scm_str = ___CAR(lst);
                char *c_str;
                ___SCMOBJ ___err;

                if (i >= len-1) /* need to grow the result array? */
                  {
                    char **new_result;
                    int j;

                    len = len * 3 / 2;
                    new_result = (char**) malloc (len * sizeof (char*));
                    if (new_result == NULL)
                      {
                        free_strings (result);
                        return ___FIX(___HEAP_OVERFLOW_ERR);
                      }
                    for (j=i; j>=0; j--)
                      new_result[j] = result[j];
                    free (result);
                    result = new_result;
                  }

                ___err = ___EXT(___SCMOBJ_to_CHARSTRING) (___PSP scm_str, &c_str, arg_num);

                if (___err != ___FIX(___NO_ERR))
                  {
                    free_strings (result);
                    return ___err;
                  }

                result[i++] = c_str;
                result[i] = NULL;
                lst = ___CDR(lst);
              }

            if (!___NULLP(lst))
              {
                free_strings (result);
                return ___FIX(___UNKNOWN_ERR);
              }

            /*
             * Note that the caller is responsible for calling free_strings
             * when it is done with the result.
             */

            *dst = result;
            return ___FIX(___NO_ERR);
          }

          ___SCMOBJ STRINGS_to_SCMOBJ (___processor_state ___ps, char **src, ___SCMOBJ *dst, int arg_num)
          {
            ___SCMOBJ ___err = ___FIX(___NO_ERR);
            ___SCMOBJ result = ___NUL; /* start with the empty list */
            int i = 0;

            while (src[i] != NULL)
              i++;

            /* build the list of strings starting at the tail */

            while (--i >= 0)
              {
                ___SCMOBJ scm_str;
                ___SCMOBJ new_result;

                /*
                 * Invariant: result is either the empty list or a ___STILL pair
                 * with reference count equal to 1.  This is important because
                 * it is possible that ___CHARSTRING_to_SCMOBJ and ___make_pair
                 * will invoke the garbage collector and we don't want the
                 * reference in result to become invalid (which would be the
                 * case if result was a ___MOVABLE pair or if it had a zero
                 * reference count).
                 */

                ___err = ___EXT(___CHARSTRING_to_SCMOBJ) (___ps, src[i], &scm_str, arg_num);

                if (___err != ___FIX(___NO_ERR))
                  {
                    ___EXT(___release_scmobj) (result); /* allow GC to reclaim result */
                    return ___FIX(___UNKNOWN_ERR);
                  }

                /*
                 * Note that scm_str will be a ___STILL object with reference
                 * count equal to 1, so there is no risk that it will be
                 * reclaimed or moved if ___make_pair invokes the garbage
                 * collector.
                 */

                new_result = ___EXT(___make_pair) (___ps, scm_str, result);

                /*
                 * We can zero the reference count of scm_str and result (if
                 * not the empty list) because the pair now references these
                 * objects and the pair is reachable (it can't be reclaimed
                 * or moved by the garbage collector).
                 */

                ___EXT(___release_scmobj) (scm_str);
                ___EXT(___release_scmobj) (result);

                result = new_result;

                if (___FIXNUMP(result))
                  return result; /* allocation failed */
              }

            /*
             * Note that result is either the empty list or a ___STILL pair
             * with a reference count equal to 1.  There will be a call to
             * ___release_scmobj later on (in ___END_CFUN_STRINGS_to_SCMOBJ
             * or ___END_SFUN_STRINGS_to_SCMOBJ) that will allow the garbage
             * collector to reclaim the whole list of strings when the Scheme
             * world no longer references it.
             */

            *dst = result;
            return ___FIX(___NO_ERR);
          }

          #define ___BEGIN_CFUN_SCMOBJ_to_STRINGS(src,dst,i) \
          if ((___err = SCMOBJ_to_STRINGS (___PSP src, &dst, i)) == ___FIX(___NO_ERR)) {
          #define ___END_CFUN_SCMOBJ_to_STRINGS(src,dst,i) \
          free_strings (dst); }

          #define ___BEGIN_CFUN_STRINGS_to_SCMOBJ(src,dst) \
          if ((___err = STRINGS_to_SCMOBJ (___ps, src, &dst, ___RETURN_POS)) == ___FIX(___NO_ERR)) {
          #define ___END_CFUN_STRINGS_to_SCMOBJ(src,dst) \
          ___EXT(___release_scmobj) (dst); }

          #define ___BEGIN_SFUN_STRINGS_to_SCMOBJ(src,dst,i) \
          if ((___err = STRINGS_to_SCMOBJ (___ps, src, &dst, i)) == ___FIX(___NO_ERR)) {
          #define ___END_SFUN_STRINGS_to_SCMOBJ(src,dst,i) \
          ___EXT(___release_scmobj) (dst); }

          #define ___BEGIN_SFUN_SCMOBJ_to_STRINGS(src,dst) \
          { ___err = SCMOBJ_to_STRINGS (___PSP src, &dst, ___RETURN_POS);
          #define ___END_SFUN_SCMOBJ_to_STRINGS(src,dst) }

          c-declare-end
          )

          (c-define-type char** "char**" "STRINGS_to_SCMOBJ" "SCMOBJ_to_STRINGS" #t)

          (define execv (c-lambda (char-string char**) int "execv"))
          (define get-environ (c-lambda () char** "get_environ"))

          (c-define (write-strings x) (char**) void "write_strings" ""
            (write x))

          (c-define (read-strings) () char** "read_strings" ""
            (read))


19.7 Continuations and the C-interface
======================================

The C-interface allows C to Scheme calls to be nested.  This means that
during a call from C to Scheme another call from C to Scheme can be
performed.  This case occurs in the following program:

     (c-declare #<<c-declare-end

     int p (char *); /* forward declarations */
     int q (void);

     int a (char *x) { return 2 * p (x+1); }
     int b (short y) { return y + q (); }

     c-declare-end
     )

     (define a (c-lambda (char-string) int "a"))
     (define b (c-lambda (short) int "b"))

     (c-define (p z) (char-string) int "p" ""
       (+ (b 10) (string-length z)))

     (c-define (q) () int "q" ""
       123)

     (write (a "hello"))

   In this example, the main Scheme program calls the C function `a'
which calls the Scheme procedure `p' which in turn calls the C function
`b' which finally calls the Scheme procedure `q'.

   Gambit maintains the Scheme continuation separately from the C stack,
thus allowing the Scheme continuation to be unwound independently from
the C stack.  The C stack frame created for the C function `f' is only
removed from the C stack when control returns from `f' or when control
returns to a C function "above" `f'.  Special care is required for
programs which escape to Scheme (using first-class continuations) from
a Scheme to C (to Scheme) call because the C stack frame will remain on
the stack.  The C stack may overflow if this happens in a loop with no
intervening return to a C function.  To avoid this problem make sure
the C stack gets cleaned up by executing a normal return from a Scheme
to C call.

20 System limitations
*********************

   * On some systems floating point overflows will cause the program to
     terminate with a floating point exception.

   * On some systems floating point operations involving `+nan.0'
     `+inf.0', `-inf.0', or `-0.' do not return the value required by
     the IEEE 754 floating point standard.

   * The compiler will not properly compile files with more than one
     definition (with `define') of the same procedure.  Replace all but
     the first `define' with assignments (`set!').

   * The maximum number of arguments that can be passed to a procedure
     by the `apply' procedure is 8192.


21 Copyright and license
************************

The Gambit system release v4.8.5 is Copyright (C) 1994-2016 by Marc
Feeley, all rights reserved.  The Gambit system release v4.8.5 is
licensed under two licenses: the Apache License, Version 2.0, and the
GNU LESSER GENERAL PUBLIC LICENSE, Version 2.1.  You have the option to
choose which of these two licenses to abide by.  The licenses are
copied below.


                                   Apache License
                             Version 2.0, January 2004
                          http://www.apache.org/licenses/

     TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION

     1. Definitions.

        "License" shall mean the terms and conditions for use, reproduction,
        and distribution as defined by Sections 1 through 9 of this document.

        "Licensor" shall mean the copyright owner or entity authorized by
        the copyright owner that is granting the License.

        "Legal Entity" shall mean the union of the acting entity and all
        other entities that control, are controlled by, or are under common
        control with that entity. For the purposes of this definition,
        "control" means (i) the power, direct or indirect, to cause the
        direction or management of such entity, whether by contract or
        otherwise, or (ii) ownership of fifty percent (50%) or more of the
        outstanding shares, or (iii) beneficial ownership of such entity.

        "You" (or "Your") shall mean an individual or Legal Entity
        exercising permissions granted by this License.

        "Source" form shall mean the preferred form for making modifications,
        including but not limited to software source code, documentation
        source, and configuration files.

        "Object" form shall mean any form resulting from mechanical
        transformation or translation of a Source form, including but
        not limited to compiled object code, generated documentation,
        and conversions to other media types.

        "Work" shall mean the work of authorship, whether in Source or
        Object form, made available under the License, as indicated by a
        copyright notice that is included in or attached to the work
        (an example is provided in the Appendix below).

        "Derivative Works" shall mean any work, whether in Source or Object
        form, that is based on (or derived from) the Work and for which the
        editorial revisions, annotations, elaborations, or other modifications
        represent, as a whole, an original work of authorship. For the purposes
        of this License, Derivative Works shall not include works that remain
        separable from, or merely link (or bind by name) to the interfaces of,
        the Work and Derivative Works thereof.

        "Contribution" shall mean any work of authorship, including
        the original version of the Work and any modifications or additions
        to that Work or Derivative Works thereof, that is intentionally
        submitted to Licensor for inclusion in the Work by the copyright owner
        or by an individual or Legal Entity authorized to submit on behalf of
        the copyright owner. For the purposes of this definition, "submitted"
        means any form of electronic, verbal, or written communication sent
        to the Licensor or its representatives, including but not limited to
        communication on electronic mailing lists, source code control systems,
        and issue tracking systems that are managed by, or on behalf of, the
        Licensor for the purpose of discussing and improving the Work, but
        excluding communication that is conspicuously marked or otherwise
        designated in writing by the copyright owner as "Not a Contribution."

        "Contributor" shall mean Licensor and any individual or Legal Entity
        on behalf of whom a Contribution has been received by Licensor and
        subsequently incorporated within the Work.

     2. Grant of Copyright License. Subject to the terms and conditions of
        this License, each Contributor hereby grants to You a perpetual,
        worldwide, non-exclusive, no-charge, royalty-free, irrevocable
        copyright license to reproduce, prepare Derivative Works of,
        publicly display, publicly perform, sublicense, and distribute the
        Work and such Derivative Works in Source or Object form.

     3. Grant of Patent License. Subject to the terms and conditions of
        this License, each Contributor hereby grants to You a perpetual,
        worldwide, non-exclusive, no-charge, royalty-free, irrevocable
        (except as stated in this section) patent license to make, have made,
        use, offer to sell, sell, import, and otherwise transfer the Work,
        where such license applies only to those patent claims licensable
        by such Contributor that are necessarily infringed by their
        Contribution(s) alone or by combination of their Contribution(s)
        with the Work to which such Contribution(s) was submitted. If You
        institute patent litigation against any entity (including a
        cross-claim or counterclaim in a lawsuit) alleging that the Work
        or a Contribution incorporated within the Work constitutes direct
        or contributory patent infringement, then any patent licenses
        granted to You under this License for that Work shall terminate
        as of the date such litigation is filed.

     4. Redistribution. You may reproduce and distribute copies of the
        Work or Derivative Works thereof in any medium, with or without
        modifications, and in Source or Object form, provided that You
        meet the following conditions:

        (a) You must give any other recipients of the Work or
            Derivative Works a copy of this License; and

        (b) You must cause any modified files to carry prominent notices
            stating that You changed the files; and

        (c) You must retain, in the Source form of any Derivative Works
            that You distribute, all copyright, patent, trademark, and
            attribution notices from the Source form of the Work,
            excluding those notices that do not pertain to any part of
            the Derivative Works; and

        (d) If the Work includes a "NOTICE" text file as part of its
            distribution, then any Derivative Works that You distribute must
            include a readable copy of the attribution notices contained
            within such NOTICE file, excluding those notices that do not
            pertain to any part of the Derivative Works, in at least one
            of the following places: within a NOTICE text file distributed
            as part of the Derivative Works; within the Source form or
            documentation, if provided along with the Derivative Works; or,
            within a display generated by the Derivative Works, if and
            wherever such third-party notices normally appear. The contents
            of the NOTICE file are for informational purposes only and
            do not modify the License. You may add Your own attribution
            notices within Derivative Works that You distribute, alongside
            or as an addendum to the NOTICE text from the Work, provided
            that such additional attribution notices cannot be construed
            as modifying the License.

        You may add Your own copyright statement to Your modifications and
        may provide additional or different license terms and conditions
        for use, reproduction, or distribution of Your modifications, or
        for any such Derivative Works as a whole, provided Your use,
        reproduction, and distribution of the Work otherwise complies with
        the conditions stated in this License.

     5. Submission of Contributions. Unless You explicitly state otherwise,
        any Contribution intentionally submitted for inclusion in the Work
        by You to the Licensor shall be under the terms and conditions of
        this License, without any additional terms or conditions.
        Notwithstanding the above, nothing herein shall supersede or modify
        the terms of any separate license agreement you may have executed
        with Licensor regarding such Contributions.

     6. Trademarks. This License does not grant permission to use the trade
        names, trademarks, service marks, or product names of the Licensor,
        except as required for reasonable and customary use in describing the
        origin of the Work and reproducing the content of the NOTICE file.

     7. Disclaimer of Warranty. Unless required by applicable law or
        agreed to in writing, Licensor provides the Work (and each
        Contributor provides its Contributions) on an "AS IS" BASIS,
        WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
        implied, including, without limitation, any warranties or conditions
        of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
        PARTICULAR PURPOSE. You are solely responsible for determining the
        appropriateness of using or redistributing the Work and assume any
        risks associated with Your exercise of permissions under this License.

     8. Limitation of Liability. In no event and under no legal theory,
        whether in tort (including negligence), contract, or otherwise,
        unless required by applicable law (such as deliberate and grossly
        negligent acts) or agreed to in writing, shall any Contributor be
        liable to You for damages, including any direct, indirect, special,
        incidental, or consequential damages of any character arising as a
        result of this License or out of the use or inability to use the
        Work (including but not limited to damages for loss of goodwill,
        work stoppage, computer failure or malfunction, or any and all
        other commercial damages or losses), even if such Contributor
        has been advised of the possibility of such damages.

     9. Accepting Warranty or Additional Liability. While redistributing
        the Work or Derivative Works thereof, You may choose to offer,
        and charge a fee for, acceptance of support, warranty, indemnity,
        or other liability obligations and/or rights consistent with this
        License. However, in accepting such obligations, You may act only
        on Your own behalf and on Your sole responsibility, not on behalf
        of any other Contributor, and only if You agree to indemnify,
        defend, and hold each Contributor harmless for any liability
        incurred by, or claims asserted against, such Contributor by reason
        of your accepting any such warranty or additional liability.

     END OF TERMS AND CONDITIONS

     APPENDIX: How to apply the Apache License to your work.

        To apply the Apache License to your work, attach the following
        boilerplate notice, with the fields enclosed by brackets "[]"
        replaced with your own identifying information. (Don't include
        the brackets!)  The text should be enclosed in the appropriate
        comment syntax for the file format. We also recommend that a
        file or class name and description of purpose be included on the
        same "printed page" as the copyright notice for easier
        identification within third-party archives.

     Copyright [yyyy] [name of copyright owner]

     Licensed under the Apache License, Version 2.0 (the "License");
     you may not use this file except in compliance with the License.
     You may obtain a copy of the License at

         http://www.apache.org/licenses/LICENSE-2.0

     Unless required by applicable law or agreed to in writing, software
     distributed under the License is distributed on an "AS IS" BASIS,
     WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     See the License for the specific language governing permissions and
     limitations under the License.

                       GNU LESSER GENERAL PUBLIC LICENSE
                            Version 2.1, February 1999

      Copyright (C) 1991, 1999 Free Software Foundation, Inc.
          51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
      Everyone is permitted to copy and distribute verbatim copies
      of this license document, but changing it is not allowed.

     [This is the first released version of the Lesser GPL.  It also counts
      as the successor of the GNU Library Public License, version 2, hence
      the version number 2.1.]

                                 Preamble

       The licenses for most software are designed to take away your
     freedom to share and change it.  By contrast, the GNU General Public
     Licenses are intended to guarantee your freedom to share and change
     free software--to make sure the software is free for all its users.

       This license, the Lesser General Public License, applies to some
     specially designated software packages--typically libraries--of the
     Free Software Foundation and other authors who decide to use it.  You
     can use it too, but we suggest you first think carefully about whether
     this license or the ordinary General Public License is the better
     strategy to use in any particular case, based on the explanations below.

       When we speak of free software, we are referring to freedom of use,
     not price.  Our General Public Licenses are designed to make sure that
     you have the freedom to distribute copies of free software (and charge
     for this service if you wish); that you receive source code or can get
     it if you want it; that you can change the software and use pieces of
     it in new free programs; and that you are informed that you can do
     these things.

       To protect your rights, we need to make restrictions that forbid
     distributors to deny you these rights or to ask you to surrender these
     rights.  These restrictions translate to certain responsibilities for
     you if you distribute copies of the library or if you modify it.

       For example, if you distribute copies of the library, whether gratis
     or for a fee, you must give the recipients all the rights that we gave
     you.  You must make sure that they, too, receive or can get the source
     code.  If you link other code with the library, you must provide
     complete object files to the recipients, so that they can relink them
     with the library after making changes to the library and recompiling
     it.  And you must show them these terms so they know their rights.

       We protect your rights with a two-step method: (1) we copyright the
     library, and (2) we offer you this license, which gives you legal
     permission to copy, distribute and/or modify the library.

       To protect each distributor, we want to make it very clear that
     there is no warranty for the free library.  Also, if the library is
     modified by someone else and passed on, the recipients should know
     that what they have is not the original version, so that the original
     author's reputation will not be affected by problems that might be
     introduced by others.

       Finally, software patents pose a constant threat to the existence of
     any free program.  We wish to make sure that a company cannot
     effectively restrict the users of a free program by obtaining a
     restrictive license from a patent holder.  Therefore, we insist that
     any patent license obtained for a version of the library must be
     consistent with the full freedom of use specified in this license.

       Most GNU software, including some libraries, is covered by the
     ordinary GNU General Public License.  This license, the GNU Lesser
     General Public License, applies to certain designated libraries, and
     is quite different from the ordinary General Public License.  We use
     this license for certain libraries in order to permit linking those
     libraries into non-free programs.

       When a program is linked with a library, whether statically or using
     a shared library, the combination of the two is legally speaking a
     combined work, a derivative of the original library.  The ordinary
     General Public License therefore permits such linking only if the
     entire combination fits its criteria of freedom.  The Lesser General
     Public License permits more lax criteria for linking other code with
     the library.

       We call this license the "Lesser" General Public License because it
     does Less to protect the user's freedom than the ordinary General
     Public License.  It also provides other free software developers Less
     of an advantage over competing non-free programs.  These disadvantages
     are the reason we use the ordinary General Public License for many
     libraries.  However, the Lesser license provides advantages in certain
     special circumstances.

       For example, on rare occasions, there may be a special need to
     encourage the widest possible use of a certain library, so that it becomes
     a de-facto standard.  To achieve this, non-free programs must be
     allowed to use the library.  A more frequent case is that a free
     library does the same job as widely used non-free libraries.  In this
     case, there is little to gain by limiting the free library to free
     software only, so we use the Lesser General Public License.

       In other cases, permission to use a particular library in non-free
     programs enables a greater number of people to use a large body of
     free software.  For example, permission to use the GNU C Library in
     non-free programs enables many more people to use the whole GNU
     operating system, as well as its variant, the GNU/Linux operating
     system.

       Although the Lesser General Public License is Less protective of the
     users' freedom, it does ensure that the user of a program that is
     linked with the Library has the freedom and the wherewithal to run
     that program using a modified version of the Library.

       The precise terms and conditions for copying, distribution and
     modification follow.  Pay close attention to the difference between a
     "work based on the library" and a "work that uses the library".  The
     former contains code derived from the library, whereas the latter must
     be combined with the library in order to run.

                       GNU LESSER GENERAL PUBLIC LICENSE
        TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION

       0. This License Agreement applies to any software library or other
     program which contains a notice placed by the copyright holder or
     other authorized party saying it may be distributed under the terms of
     this Lesser General Public License (also called "this License").
     Each licensee is addressed as "you".

       A "library" means a collection of software functions and/or data
     prepared so as to be conveniently linked with application programs
     (which use some of those functions and data) to form executables.

       The "Library", below, refers to any such software library or work
     which has been distributed under these terms.  A "work based on the
     Library" means either the Library or any derivative work under
     copyright law: that is to say, a work containing the Library or a
     portion of it, either verbatim or with modifications and/or translated
     straightforwardly into another language.  (Hereinafter, translation is
     included without limitation in the term "modification".)

       "Source code" for a work means the preferred form of the work for
     making modifications to it.  For a library, complete source code means
     all the source code for all modules it contains, plus any associated
     interface definition files, plus the scripts used to control compilation
     and installation of the library.

       Activities other than copying, distribution and modification are not
     covered by this License; they are outside its scope.  The act of
     running a program using the Library is not restricted, and output from
     such a program is covered only if its contents constitute a work based
     on the Library (independent of the use of the Library in a tool for
     writing it).  Whether that is true depends on what the Library does
     and what the program that uses the Library does.

       1. You may copy and distribute verbatim copies of the Library's
     complete source code as you receive it, in any medium, provided that
     you conspicuously and appropriately publish on each copy an
     appropriate copyright notice and disclaimer of warranty; keep intact
     all the notices that refer to this License and to the absence of any
     warranty; and distribute a copy of this License along with the
     Library.

       You may charge a fee for the physical act of transferring a copy,
     and you may at your option offer warranty protection in exchange for a
     fee.

       2. You may modify your copy or copies of the Library or any portion
     of it, thus forming a work based on the Library, and copy and
     distribute such modifications or work under the terms of Section 1
     above, provided that you also meet all of these conditions:

         a) The modified work must itself be a software library.

         b) You must cause the files modified to carry prominent notices
         stating that you changed the files and the date of any change.

         c) You must cause the whole of the work to be licensed at no
         charge to all third parties under the terms of this License.

         d) If a facility in the modified Library refers to a function or a
         table of data to be supplied by an application program that uses
         the facility, other than as an argument passed when the facility
         is invoked, then you must make a good faith effort to ensure that,
         in the event an application does not supply such function or
         table, the facility still operates, and performs whatever part of
         its purpose remains meaningful.

         (For example, a function in a library to compute square roots has
         a purpose that is entirely well-defined independent of the
         application.  Therefore, Subsection 2d requires that any
         application-supplied function or table used by this function must
         be optional: if the application does not supply it, the square
         root function must still compute square roots.)

     These requirements apply to the modified work as a whole.  If
     identifiable sections of that work are not derived from the Library,
     and can be reasonably considered independent and separate works in
     themselves, then this License, and its terms, do not apply to those
     sections when you distribute them as separate works.  But when you
     distribute the same sections as part of a whole which is a work based
     on the Library, the distribution of the whole must be on the terms of
     this License, whose permissions for other licensees extend to the
     entire whole, and thus to each and every part regardless of who wrote
     it.

     Thus, it is not the intent of this section to claim rights or contest
     your rights to work written entirely by you; rather, the intent is to
     exercise the right to control the distribution of derivative or
     collective works based on the Library.

     In addition, mere aggregation of another work not based on the Library
     with the Library (or with a work based on the Library) on a volume of
     a storage or distribution medium does not bring the other work under
     the scope of this License.

       3. You may opt to apply the terms of the ordinary GNU General Public
     License instead of this License to a given copy of the Library.  To do
     this, you must alter all the notices that refer to this License, so
     that they refer to the ordinary GNU General Public License, version 2,
     instead of to this License.  (If a newer version than version 2 of the
     ordinary GNU General Public License has appeared, then you can specify
     that version instead if you wish.)  Do not make any other change in
     these notices.

       Once this change is made in a given copy, it is irreversible for
     that copy, so the ordinary GNU General Public License applies to all
     subsequent copies and derivative works made from that copy.

       This option is useful when you wish to copy part of the code of
     the Library into a program that is not a library.

       4. You may copy and distribute the Library (or a portion or
     derivative of it, under Section 2) in object code or executable form
     under the terms of Sections 1 and 2 above provided that you accompany
     it with the complete corresponding machine-readable source code, which
     must be distributed under the terms of Sections 1 and 2 above on a
     medium customarily used for software interchange.

       If distribution of object code is made by offering access to copy
     from a designated place, then offering equivalent access to copy the
     source code from the same place satisfies the requirement to
     distribute the source code, even though third parties are not
     compelled to copy the source along with the object code.

       5. A program that contains no derivative of any portion of the
     Library, but is designed to work with the Library by being compiled or
     linked with it, is called a "work that uses the Library".  Such a
     work, in isolation, is not a derivative work of the Library, and
     therefore falls outside the scope of this License.

       However, linking a "work that uses the Library" with the Library
     creates an executable that is a derivative of the Library (because it
     contains portions of the Library), rather than a "work that uses the
     library".  The executable is therefore covered by this License.
     Section 6 states terms for distribution of such executables.

       When a "work that uses the Library" uses material from a header file
     that is part of the Library, the object code for the work may be a
     derivative work of the Library even though the source code is not.
     Whether this is true is especially significant if the work can be
     linked without the Library, or if the work is itself a library.  The
     threshold for this to be true is not precisely defined by law.

       If such an object file uses only numerical parameters, data
     structure layouts and accessors, and small macros and small inline
     functions (ten lines or less in length), then the use of the object
     file is unrestricted, regardless of whether it is legally a derivative
     work.  (Executables containing this object code plus portions of the
     Library will still fall under Section 6.)

       Otherwise, if the work is a derivative of the Library, you may
     distribute the object code for the work under the terms of Section 6.
     Any executables containing that work also fall under Section 6,
     whether or not they are linked directly with the Library itself.

       6. As an exception to the Sections above, you may also combine or
     link a "work that uses the Library" with the Library to produce a
     work containing portions of the Library, and distribute that work
     under terms of your choice, provided that the terms permit
     modification of the work for the customer's own use and reverse
     engineering for debugging such modifications.

       You must give prominent notice with each copy of the work that the
     Library is used in it and that the Library and its use are covered by
     this License.  You must supply a copy of this License.  If the work
     during execution displays copyright notices, you must include the
     copyright notice for the Library among them, as well as a reference
     directing the user to the copy of this License.  Also, you must do one
     of these things:

         a) Accompany the work with the complete corresponding
         machine-readable source code for the Library including whatever
         changes were used in the work (which must be distributed under
         Sections 1 and 2 above); and, if the work is an executable linked
         with the Library, with the complete machine-readable "work that
         uses the Library", as object code and/or source code, so that the
         user can modify the Library and then relink to produce a modified
         executable containing the modified Library.  (It is understood
         that the user who changes the contents of definitions files in the
         Library will not necessarily be able to recompile the application
         to use the modified definitions.)

         b) Use a suitable shared library mechanism for linking with the
         Library.  A suitable mechanism is one that (1) uses at run time a
         copy of the library already present on the user's computer system,
         rather than copying library functions into the executable, and (2)
         will operate properly with a modified version of the library, if
         the user installs one, as long as the modified version is
         interface-compatible with the version that the work was made with.

         c) Accompany the work with a written offer, valid for at
         least three years, to give the same user the materials
         specified in Subsection 6a, above, for a charge no more
         than the cost of performing this distribution.

         d) If distribution of the work is made by offering access to copy
         from a designated place, offer equivalent access to copy the above
         specified materials from the same place.

         e) Verify that the user has already received a copy of these
         materials or that you have already sent this user a copy.

       For an executable, the required form of the "work that uses the
     Library" must include any data and utility programs needed for
     reproducing the executable from it.  However, as a special exception,
     the materials to be distributed need not include anything that is
     normally distributed (in either source or binary form) with the major
     components (compiler, kernel, and so on) of the operating system on
     which the executable runs, unless that component itself accompanies
     the executable.

       It may happen that this requirement contradicts the license
     restrictions of other proprietary libraries that do not normally
     accompany the operating system.  Such a contradiction means you cannot
     use both them and the Library together in an executable that you
     distribute.

       7. You may place library facilities that are a work based on the
     Library side-by-side in a single library together with other library
     facilities not covered by this License, and distribute such a combined
     library, provided that the separate distribution of the work based on
     the Library and of the other library facilities is otherwise
     permitted, and provided that you do these two things:

         a) Accompany the combined library with a copy of the same work
         based on the Library, uncombined with any other library
         facilities.  This must be distributed under the terms of the
         Sections above.

         b) Give prominent notice with the combined library of the fact
         that part of it is a work based on the Library, and explaining
         where to find the accompanying uncombined form of the same work.

       8. You may not copy, modify, sublicense, link with, or distribute
     the Library except as expressly provided under this License.  Any
     attempt otherwise to copy, modify, sublicense, link with, or
     distribute the Library is void, and will automatically terminate your
     rights under this License.  However, parties who have received copies,
     or rights, from you under this License will not have their licenses
     terminated so long as such parties remain in full compliance.

       9. You are not required to accept this License, since you have not
     signed it.  However, nothing else grants you permission to modify or
     distribute the Library or its derivative works.  These actions are
     prohibited by law if you do not accept this License.  Therefore, by
     modifying or distributing the Library (or any work based on the
     Library), you indicate your acceptance of this License to do so, and
     all its terms and conditions for copying, distributing or modifying
     the Library or works based on it.

       10. Each time you redistribute the Library (or any work based on the
     Library), the recipient automatically receives a license from the
     original licensor to copy, distribute, link with or modify the Library
     subject to these terms and conditions.  You may not impose any further
     restrictions on the recipients' exercise of the rights granted herein.
     You are not responsible for enforcing compliance by third parties with
     this License.

       11. If, as a consequence of a court judgment or allegation of patent
     infringement or for any other reason (not limited to patent issues),
     conditions are imposed on you (whether by court order, agreement or
     otherwise) that contradict the conditions of this License, they do not
     excuse you from the conditions of this License.  If you cannot
     distribute so as to satisfy simultaneously your obligations under this
     License and any other pertinent obligations, then as a consequence you
     may not distribute the Library at all.  For example, if a patent
     license would not permit royalty-free redistribution of the Library by
     all those who receive copies directly or indirectly through you, then
     the only way you could satisfy both it and this License would be to
     refrain entirely from distribution of the Library.

     If any portion of this section is held invalid or unenforceable under any
     particular circumstance, the balance of the section is intended to apply,
     and the section as a whole is intended to apply in other circumstances.

     It is not the purpose of this section to induce you to infringe any
     patents or other property right claims or to contest validity of any
     such claims; this section has the sole purpose of protecting the
     integrity of the free software distribution system which is
     implemented by public license practices.  Many people have made
     generous contributions to the wide range of software distributed
     through that system in reliance on consistent application of that
     system; it is up to the author/donor to decide if he or she is willing
     to distribute software through any other system and a licensee cannot
     impose that choice.

     This section is intended to make thoroughly clear what is believed to
     be a consequence of the rest of this License.

       12. If the distribution and/or use of the Library is restricted in
     certain countries either by patents or by copyrighted interfaces, the
     original copyright holder who places the Library under this License may add
     an explicit geographical distribution limitation excluding those countries,
     so that distribution is permitted only in or among countries not thus
     excluded.  In such case, this License incorporates the limitation as if
     written in the body of this License.

       13. The Free Software Foundation may publish revised and/or new
     versions of the Lesser General Public License from time to time.
     Such new versions will be similar in spirit to the present version,
     but may differ in detail to address new problems or concerns.

     Each version is given a distinguishing version number.  If the Library
     specifies a version number of this License which applies to it and
     "any later version", you have the option of following the terms and
     conditions either of that version or of any later version published by
     the Free Software Foundation.  If the Library does not specify a
     license version number, you may choose any version ever published by
     the Free Software Foundation.

       14. If you wish to incorporate parts of the Library into other free
     programs whose distribution conditions are incompatible with these,
     write to the author to ask for permission.  For software which is
     copyrighted by the Free Software Foundation, write to the Free
     Software Foundation; we sometimes make exceptions for this.  Our
     decision will be guided by the two goals of preserving the free status
     of all derivatives of our free software and of promoting the sharing
     and reuse of software generally.

                                 NO WARRANTY

       15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
     WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
     EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
     OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
     KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
     IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
     LIBRARY IS WITH YOU.  SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
     THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.

       16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
     WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
     AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
     FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
     CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
     LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
     RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
     FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
     SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
     DAMAGES.

                          END OF TERMS AND CONDITIONS

                How to Apply These Terms to Your New Libraries

       If you develop a new library, and you want it to be of the greatest
     possible use to the public, we recommend making it free software that
     everyone can redistribute and change.  You can do so by permitting
     redistribution under these terms (or, alternatively, under the terms of the
     ordinary General Public License).

       To apply these terms, attach the following notices to the library.  It is
     safest to attach them to the start of each source file to most effectively
     convey the exclusion of warranty; and each file should have at least the
     "copyright" line and a pointer to where the full notice is found.

         <one line to give the library's name and a brief idea of what it does.>
         Copyright (C) <year>  <name of author>

         This library is free software; you can redistribute it and/or
         modify it under the terms of the GNU Lesser General Public
         License as published by the Free Software Foundation; either
         version 2.1 of the License, or (at your option) any later version.

         This library is distributed in the hope that it will be useful,
         but WITHOUT ANY WARRANTY; without even the implied warranty of
         MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
         Lesser General Public License for more details.

         You should have received a copy of the GNU Lesser General Public
         License along with this library; if not, write to the Free Software
         Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA

     Also add information on how to contact you by electronic and paper mail.

     You should also get your employer (if you work as a programmer) or your
     school, if any, to sign a "copyright disclaimer" for the library, if
     necessary.  Here is a sample; alter the names:

       Yoyodyne, Inc., hereby disclaims all copyright interest in the
       library `Frob' (a library for tweaking knobs) written by James Random Hacker.

       <signature of Ty Coon>, 1 April 1990
       Ty Coon, President of Vice

     That's all there is to it!

General index
*************

#:                                             See 5.4.     (line  2080)
##:                                            See 5.4.     (line  2080)
+z:                                            See 3.4.4.   (line  1189)
,(b EXPR):                                     See 5.2.     (line  1896)
,(be EXPR):                                    See 5.2.     (line  1903)
,(bed EXPR):                                   See 5.2.     (line  1906)
,(c EXPR):                                     See 5.2.     (line  1789)
,(e EXPR):                                     See 5.2.     (line  1944)
,(ed EXPR):                                    See 5.2.     (line  1956)
,(h SUBJECT):                                  See 5.2.     (line  1762)
,(st EXPR):                                    See 5.2.     (line  1967)
,(v EXPR):                                     See 5.2.     (line  1971)
,+:                                            See 5.2.     (line  1850)
,++:                                           See 5.2.     (line  1856)
,-:                                            See 5.2.     (line  1853)
,--:                                           See 5.2.     (line  1860)
,?:                                            See 5.2.     (line  1759)
,b:                                            See 5.2.     (line  1881)
,be:                                           See 5.2.     (line  1890)
,bed:                                          See 5.2.     (line  1893)
,c:                                            See 5.2.     (line  1806)
,d:                                            See 5.2.     (line  1785)
,e:                                            See 5.2.     (line  1916)
,ed:                                           See 5.2.     (line  1940)
,h:                                            See 5.2.     (line  1769)
,i:                                            See 5.2.     (line  1910)
,l:                                            See 5.2.     (line  1826)
,N:                                            See 5.2.     (line  1833)
,N+:                                           See 5.2.     (line  1838)
,N-:                                           See 5.2.     (line  1844)
,q:                                            See 5.2.     (line  1774)
,qt:                                           See 5.2.     (line  1778)
,s:                                            See 5.2.     (line  1812)
,st:                                           See 5.2.     (line  1960)
,t:                                            See 5.2.     (line  1782)
,y:                                            See 5.2.     (line  1864)
-:                                             See 3.3.     (line   785)
-:+:                                           See 4.       (line  1578)
-:-:                                           See 4.       (line  1581)
-:::                                           See 4.       (line  1658)
-:=:                                           See 4.       (line  1572)
-:d:                                           See 4.       (line  1499)
-:d-:                                          See 4.       (line  1549)
-:d@[HOST][:PORT]:                             See 4.       (line  1553)
-:da:                                          See 4.       (line  1507)
-:dc:                                          See 4.       (line  1546)
-:dD:                                          See 4.       (line  1529)
-:di:                                          See 4.       (line  1542)
-:dLEVEL:                                      See 4.       (line  1537)
-:dp:                                          See 4.       (line  1503)
-:dQ:                                          See 4.       (line  1533)
-:dq:                                          See 4.       (line  1517)
-:dr:                                          See 4.       (line  1510)
-:dR:                                          See 4.       (line  1521)
-:ds:                                          See 4.       (line  1513)
-:f:                                           See 4.       (line  1581)
-:h:                                           See 4.       (line  1480)
-:l:                                           See 4.       (line  1486)
-:m:                                           See 4.       (line  1475)
-:S:                                           See 4.       (line  1493)
-:s <1>:                                       See 4.       (line  1493)
-:s:                                           See 6.3.     (line  3180)
-:t:                                           See 4.       (line  1581)
-:}:                                           See 4.       (line  1568)
-:~~:                                          See 4.       (line  1575)
-c:                                            See 3.3.     (line   753)
-call_shared:                                  See 3.4.4.   (line  1189)
-cc-options:                                   See 3.3.     (line   688)
-D___DYNAMIC:                                  See 3.4.2.   (line   984)
-D___LIBRARY:                                  See 3.4.3.   (line  1130)
-D___PRIMAL:                                   See 3.4.3.   (line  1130)
-D___SHARED:                                   See 3.4.3.   (line  1130)
-D___SINGLE_HOST:                              See 3.4.4.   (line  1169)
-debug <1>:                                    See 3.3.     (line   729)
-debug:                                        See 6.3.     (line  3291)
-debug-environments <1>:                       See 6.3.     (line  3331)
-debug-environments:                           See 3.3.     (line   741)
-debug-location <1>:                           See 3.3.     (line   733)
-debug-location:                               See 6.3.     (line  3305)
-debug-source <1>:                             See 6.3.     (line  3318)
-debug-source:                                 See 3.3.     (line   737)
-dynamic:                                      See 3.3.     (line   753)
-e:                                            See 3.3.     (line   787)
-exe:                                          See 3.3.     (line   753)
-expansion:                                    See 3.3.     (line   723)
-flat:                                         See 3.3.     (line   771)
-fPIC:                                         See 3.4.4.   (line  1189)
-fpic:                                         See 3.4.4.   (line  1189)
-G:                                            See 3.4.4.   (line  1189)
-gvm:                                          See 3.3.     (line   726)
-i:                                            See 3.3.     (line   670)
-I/usr/local/Gambit/include:                   See 3.4.4.   (line  1181)
-keep-c:                                       See 3.3.     (line   753)
-KPIC:                                         See 3.4.4.   (line  1189)
-Kpic:                                         See 3.4.4.   (line  1189)
-l BASE:                                       See 3.3.     (line   778)
-L/usr/local/Gambit/lib:                       See 3.4.4.   (line  1181)
-ld-options:                                   See 3.3.     (line   698)
-ld-options-prelude:                           See 3.3.     (line   698)
-link:                                         See 3.3.     (line   753)
-O:                                            See 3.4.4.   (line  1169)
-o OUTPUT:                                     See 3.3.     (line   749)
-obj:                                          See 3.3.     (line   753)
-pic:                                          See 3.4.4.   (line  1189)
-postlude:                                     See 3.3.     (line   681)
-prelude:                                      See 3.3.     (line   673)
-rdynamic:                                     See 3.4.4.   (line  1189)
-report:                                       See 3.3.     (line   718)
-shared:                                       See 3.4.4.   (line  1189)
-track-scheme:                                 See 3.3.     (line   745)
-verbose:                                      See 3.3.     (line   715)
-warnings:                                     See 3.3.     (line   712)
.c:                                            See 3.3.     (line   544)
.scm:                                          See 3.3.     (line   544)
.six:                                          See 3.3.     (line   544)
<:                                             See 9.1.     (line  3926)
<=:                                            See 9.1.     (line  3928)
=:                                             See 9.1.     (line  3925)
>:                                             See 9.1.     (line  3927)
>=:                                            See 9.1.     (line  3929)
^C <1>:                                        See 5.1.     (line  1696)
^C:                                            See 4.       (line  1521)
^D:                                            See 5.1.     (line  1722)
___cleanup:                                    See 19.5.    (line 13213)
___setup:                                      See 19.5.    (line 13213)
abandoned-mutex-exception?:                    See 15.4.    (line  7242)
abort:                                         See 15.1.    (line  7013)
absolute path:                                 See 16.1.    (line  8129)
address-info-family:                           See 16.11.   (line  9236)
address-info-protocol:                         See 16.11.   (line  9258)
address-info-socket-info:                      See 16.11.   (line  9269)
address-info-socket-type:                      See 16.11.   (line  9247)
address-info?:                                 See 16.11.   (line  9223)
address-infos:                                 See 16.11.   (line  9157)
all-bits-set?:                                 See 9.4.     (line  4167)
any-bits-set?:                                 See 9.4.     (line  4152)
append-f32vectors:                             See 10.      (line  4832)
append-f64vectors:                             See 10.      (line  4850)
append-s16vectors:                             See 10.      (line  4724)
append-s32vectors:                             See 10.      (line  4760)
append-s64vectors:                             See 10.      (line  4796)
append-s8vectors:                              See 10.      (line  4688)
append-strings:                                See 6.3.     (line  2939)
append-u16vectors:                             See 10.      (line  4742)
append-u32vectors:                             See 10.      (line  4778)
append-u64vectors:                             See 10.      (line  4814)
append-u8vectors:                              See 10.      (line  4706)
append-vectors:                                See 6.3.     (line  2882)
arithmetic-shift:                              See 9.4.     (line  4001)
bit-count:                                     See 9.4.     (line  4094)
bit-set?:                                      See 9.4.     (line  4140)
bitwise-and:                                   See 9.4.     (line  4032)
bitwise-ior:                                   See 9.4.     (line  4048)
bitwise-merge:                                 See 9.4.     (line  4016)
bitwise-not:                                   See 9.4.     (line  4082)
bitwise-xor:                                   See 9.4.     (line  4065)
block:                                         See 6.3.     (line  3219)
box:                                           See 6.3.     (line  2996)
box?:                                          See 6.3.     (line  2997)
boxes:                                         See 6.3.     (line  2999)
break:                                         See 5.4.     (line  2270)
c-declare:                                     See 19.2.    (line 13004)
c-define:                                      See 19.5.    (line 13175)
c-define-type:                                 See 19.6.    (line 13231)
c-initialize:                                  See 19.3.    (line 13047)
c-lambda:                                      See 19.4.    (line 13070)
call-with-current-continuation:                See 6.1.     (line  2653)
call-with-input-file:                          See 17.7.1.  (line 10336)
call-with-input-process:                       See 17.7.2.  (line 10480)
call-with-input-string:                        See 17.10.   (line 11242)
call-with-input-u8vector:                      See 17.11.   (line 11299)
call-with-input-vector:                        See 17.9.    (line 11092)
call-with-output-file:                         See 17.7.1.  (line 10337)
call-with-output-process:                      See 17.7.2.  (line 10481)
call-with-output-string:                       See 17.10.   (line 11243)
call-with-output-u8vector:                     See 17.11.   (line 11300)
call-with-output-vector:                       See 17.9.    (line 11093)
call/cc:                                       See 6.1.     (line  2654)
central installation directory:                See 16.1.    (line  8129)
cfun-conversion-exception-arguments:           See 15.5.    (line  7416)
cfun-conversion-exception-code:                See 15.5.    (line  7417)
cfun-conversion-exception-message:             See 15.5.    (line  7418)
cfun-conversion-exception-procedure:           See 15.5.    (line  7415)
cfun-conversion-exception?:                    See 15.5.    (line  7414)
char->integer:                                 See 8.1.     (line  3861)
char-ci<=?:                                    See 8.1.     (line  3890)
char-ci<?:                                     See 8.1.     (line  3888)
char-ci=?:                                     See 8.1.     (line  3887)
char-ci>=?:                                    See 8.1.     (line  3891)
char-ci>?:                                     See 8.1.     (line  3889)
char<=?:                                       See 8.1.     (line  3885)
char<?:                                        See 8.1.     (line  3883)
char=?:                                        See 8.1.     (line  3882)
char>=?:                                       See 8.1.     (line  3886)
char>?:                                        See 8.1.     (line  3884)
clear-bit-field:                               See 9.4.     (line  4198)
close-input-port:                              See 17.4.2.  (line  9764)
close-output-port:                             See 17.4.2.  (line  9765)
close-port:                                    See 17.4.2.  (line  9766)
command-line <1>:                              See 2.5.     (line   382)
command-line:                                  See 16.5.    (line  8492)
compile-file:                                  See 3.5.     (line  1237)
compile-file-to-target:                        See 3.5.     (line  1203)
compiler:                                      See 3.       (line   514)
compiler options:                              See 3.3.     (line   576)
cond-expand:                                   See 6.4.     (line  3758)
condition-variable-broadcast!:                 See 13.9.    (line  6638)
condition-variable-name:                       See 13.9.    (line  6558)
condition-variable-signal!:                    See 13.9.    (line  6583)
condition-variable-specific:                   See 13.9.    (line  6566)
condition-variable-specific-set!:              See 13.9.    (line  6567)
condition-variable?:                           See 13.9.    (line  6533)
configure-command-string:                      See 6.4.     (line  3723)
console-port:                                  See 6.4.     (line  3689)
constant-fold:                                 See 6.3.     (line  3258)
continuation-capture:                          See 6.4.     (line  3453)
continuation-graft:                            See 6.4.     (line  3454)
continuation-return:                           See 6.4.     (line  3455)
continuation?:                                 See 6.4.     (line  3452)
continuations:                                 See 19.7.    (line 13689)
copy-bit-field:                                See 9.4.     (line  4200)
copy-file:                                     See 16.2.    (line  8383)
cpu-time:                                      See 16.7.    (line  8594)
create-directory:                              See 16.2.    (line  8302)
create-fifo:                                   See 16.2.    (line  8329)
create-link:                                   See 16.2.    (line  8360)
create-symbolic-link:                          See 16.2.    (line  8368)
current exception-handler:                     See 15.1.    (line  6925)
current working directory:                     See 16.1.    (line  8129)
current-directory:                             See 16.1.    (line  8167)
current-error-port:                            See 17.12.   (line 11335)
current-exception-handler:                     See 15.1.    (line  6925)
current-input-port:                            See 17.12.   (line 11333)
current-output-port:                           See 17.12.   (line 11334)
current-readtable:                             See 17.12.   (line 11336)
current-thread:                                See 13.9.    (line  5837)
current-time:                                  See 16.7.    (line  8558)
current-user-interrupt-handler:                See 6.4.     (line  3691)
datum-parsing-exception-kind:                  See 15.6.    (line  7582)
datum-parsing-exception-parameters:            See 15.6.    (line  7583)
datum-parsing-exception-readenv:               See 15.6.    (line  7584)
datum-parsing-exception?:                      See 15.6.    (line  7581)
deadlock-exception?:                           See 15.4.    (line  7221)
debug <1>:                                     See 6.3.     (line  3291)
debug:                                         See 3.3.     (line   729)
debug-environments <1>:                        See 3.3.     (line   741)
debug-environments:                            See 6.3.     (line  3331)
debug-location <1>:                            See 6.3.     (line  3305)
debug-location:                                See 3.3.     (line   733)
debug-source <1>:                              See 3.3.     (line   737)
debug-source:                                  See 6.3.     (line  3318)
declare:                                       See 6.3.     (line  3207)
default-random-source:                         See 9.7.     (line  4443)
defer-user-interrupts:                         See 6.4.     (line  3692)
define <1>:                                    See 6.2.     (line  2663)
define:                                        See 20.      (line 13741)
define-cond-expand-feature:                    See 6.4.     (line  3760)
define-macro:                                  See 6.3.     (line  3148)
define-record-type:                            See 6.4.     (line  3749)
define-structure:                              See 12.      (line  5558)
define-syntax:                                 See 6.3.     (line  3180)
define-type:                                   See 6.4.     (line  3750)
define-type-of-thread:                         See 6.4.     (line  3648)
delete-directory:                              See 16.2.    (line  8395)
delete-file:                                   See 16.2.    (line  8390)
deserialization <1>:                           See 18.1.    (line 11496)
deserialization:                               See 10.      (line  4873)
directory-files:                               See 16.2.    (line  8400)
display-continuation-backtrace:                See 6.4.     (line  3537)
display-continuation-dynamic-environment:      See 6.4.     (line  3534)
display-continuation-environment:              See 6.4.     (line  3533)
display-dynamic-environment?:                  See 5.4.     (line  2377)
display-environment-set!:                      See 5.4.     (line  2357)
display-exception:                             See 6.4.     (line  3530)
display-exception-in-context:                  See 6.4.     (line  3531)
display-procedure-environment:                 See 6.4.     (line  3532)
divide-by-zero-exception-arguments:            See 15.8.    (line  7854)
divide-by-zero-exception-procedure:            See 15.8.    (line  7853)
divide-by-zero-exception?:                     See 15.8.    (line  7852)
Emacs:                                         See 5.6.     (line  2550)
eq?-hash:                                      See 11.1.    (line  5042)
equal?-hash:                                   See 11.1.    (line  5066)
eqv?-hash:                                     See 11.1.    (line  5054)
err-code->string:                              See 6.4.     (line  3696)
error:                                         See 15.10.   (line  8084)
error-exception-message:                       See 15.10.   (line  8082)
error-exception-parameters:                    See 15.10.   (line  8083)
error-exception?:                              See 15.10.   (line  8081)
eval:                                          See 6.3.     (line  3115)
exit:                                          See 16.4.    (line  8474)
expression-parsing-exception-kind:             See 15.7.    (line  7657)
expression-parsing-exception-parameters:       See 15.7.    (line  7658)
expression-parsing-exception-source:           See 15.7.    (line  7659)
expression-parsing-exception?:                 See 15.7.    (line  7656)
extended-bindings:                             See 6.3.     (line  3266)
extract-bit-field:                             See 9.4.     (line  4196)
f32vector:                                     See 10.      (line  4824)
f32vector->list:                               See 10.      (line  4828)
f32vector-append:                              See 10.      (line  4834)
f32vector-copy:                                See 10.      (line  4833)
f32vector-fill!:                               See 10.      (line  4830)
f32vector-length:                              See 10.      (line  4825)
f32vector-ref:                                 See 10.      (line  4826)
f32vector-set!:                                See 10.      (line  4827)
f32vector-shrink!:                             See 10.      (line  4838)
f32vector?:                                    See 10.      (line  4822)
f64vector:                                     See 10.      (line  4842)
f64vector->list:                               See 10.      (line  4846)
f64vector-append:                              See 10.      (line  4852)
f64vector-copy:                                See 10.      (line  4851)
f64vector-fill!:                               See 10.      (line  4848)
f64vector-length:                              See 10.      (line  4843)
f64vector-ref:                                 See 10.      (line  4844)
f64vector-set!:                                See 10.      (line  4845)
f64vector-shrink!:                             See 10.      (line  4856)
f64vector?:                                    See 10.      (line  4840)
FFI:                                           See 19.      (line 12395)
file names:                                    See 16.1.    (line  8129)
file-attributes:                               See 16.8.    (line  8875)
file-creation-time:                            See 16.8.    (line  8876)
file-device:                                   See 16.8.    (line  8865)
file-exists?:                                  See 16.8.    (line  8654)
file-group:                                    See 16.8.    (line  8870)
file-info:                                     See 16.8.    (line  8669)
file-info-attributes:                          See 16.8.    (line  8844)
file-info-creation-time:                       See 16.8.    (line  8854)
file-info-device:                              See 16.8.    (line  8750)
file-info-group:                               See 16.8.    (line  8796)
file-info-inode:                               See 16.8.    (line  8759)
file-info-last-access-time:                    See 16.8.    (line  8814)
file-info-last-change-time:                    See 16.8.    (line  8834)
file-info-last-modification-time:              See 16.8.    (line  8824)
file-info-mode:                                See 16.8.    (line  8768)
file-info-number-of-links:                     See 16.8.    (line  8777)
file-info-owner:                               See 16.8.    (line  8787)
file-info-size:                                See 16.8.    (line  8805)
file-info-type:                                See 16.8.    (line  8714)
file-info?:                                    See 16.8.    (line  8702)
file-inode:                                    See 16.8.    (line  8866)
file-last-access-and-modification-times-set!:  See 16.8.    (line  8884)
file-last-access-time:                         See 16.8.    (line  8872)
file-last-change-time:                         See 16.8.    (line  8874)
file-last-modification-time:                   See 16.8.    (line  8873)
file-mode:                                     See 16.8.    (line  8867)
file-number-of-links:                          See 16.8.    (line  8868)
file-owner:                                    See 16.8.    (line  8869)
file-size:                                     See 16.8.    (line  8871)
file-type:                                     See 16.8.    (line  8864)
FILE.c:                                        See 3.3.     (line   544)
FILE.scm:                                      See 3.3.     (line   544)
FILE.six:                                      See 3.3.     (line   544)
finite?:                                       See 6.4.     (line  3762)
first-bit-set:                                 See 9.4.     (line  4182)
fixnum:                                        See 6.3.     (line  3398)
fixnum->flonum:                                See 9.6.     (line  4355)
fixnum-overflow-exception-arguments:           See 9.5.     (line  4322)
fixnum-overflow-exception-procedure:           See 9.5.     (line  4321)
fixnum-overflow-exception?:                    See 9.5.     (line  4320)
fixnum?:                                       See 9.5.     (line  4240)
fl*:                                           See 9.6.     (line  4357)
fl+:                                           See 9.6.     (line  4359)
fl-:                                           See 9.6.     (line  4361)
fl/:                                           See 9.6.     (line  4363)
fl<:                                           See 9.6.     (line  4365)
fl<=:                                          See 9.6.     (line  4367)
fl=:                                           See 9.6.     (line  4369)
fl>:                                           See 9.6.     (line  4371)
fl>=:                                          See 9.6.     (line  4373)
flabs:                                         See 9.6.     (line  4375)
flacos:                                        See 9.6.     (line  4377)
flasin:                                        See 9.6.     (line  4379)
flatan:                                        See 9.6.     (line  4381)
flceiling:                                     See 9.6.     (line  4384)
flcos:                                         See 9.6.     (line  4386)
fldenominator:                                 See 9.6.     (line  4388)
fleven?:                                       See 9.6.     (line  4390)
flexp:                                         See 9.6.     (line  4392)
flexpt:                                        See 9.6.     (line  4394)
flfinite?:                                     See 9.6.     (line  4396)
flfloor:                                       See 9.6.     (line  4398)
flinfinite?:                                   See 9.6.     (line  4400)
flinteger?:                                    See 9.6.     (line  4402)
fllog:                                         See 9.6.     (line  4404)
flmax:                                         See 9.6.     (line  4406)
flmin:                                         See 9.6.     (line  4408)
flnan?:                                        See 9.6.     (line  4410)
flnegative?:                                   See 9.6.     (line  4412)
flnumerator:                                   See 9.6.     (line  4414)
floating point overflow:                       See 20.      (line 13734)
flodd?:                                        See 9.6.     (line  4416)
flonum:                                        See 6.3.     (line  3398)
flonum?:                                       See 9.6.     (line  4353)
flpositive?:                                   See 9.6.     (line  4418)
flround:                                       See 9.6.     (line  4420)
flsin:                                         See 9.6.     (line  4422)
flsqrt:                                        See 9.6.     (line  4424)
fltan:                                         See 9.6.     (line  4426)
fltruncate:                                    See 9.6.     (line  4428)
flzero?:                                       See 9.6.     (line  4430)
force-output:                                  See 17.4.2.  (line  9726)
foreign function interface:                    See 19.      (line 12395)
foreign-address:                               See 6.4.     (line  3700)
foreign-release!:                              See 6.4.     (line  3701)
foreign-released?:                             See 6.4.     (line  3702)
foreign-tags:                                  See 6.4.     (line  3699)
foreign?:                                      See 6.4.     (line  3698)
future:                                        See 6.4.     (line  3727)
fx*:                                           See 9.5.     (line  4242)
fx+:                                           See 9.5.     (line  4244)
fx-:                                           See 9.5.     (line  4246)
fx<:                                           See 9.5.     (line  4248)
fx<=:                                          See 9.5.     (line  4250)
fx=:                                           See 9.5.     (line  4252)
fx>:                                           See 9.5.     (line  4254)
fx>=:                                          See 9.5.     (line  4256)
fxabs:                                         See 9.5.     (line  4258)
fxand:                                         See 9.5.     (line  4260)
fxarithmetic-shift:                            See 9.5.     (line  4262)
fxarithmetic-shift-left:                       See 9.5.     (line  4264)
fxarithmetic-shift-right:                      See 9.5.     (line  4266)
fxbit-count:                                   See 9.5.     (line  4268)
fxbit-set?:                                    See 9.5.     (line  4270)
fxeven?:                                       See 9.5.     (line  4272)
fxfirst-bit-set:                               See 9.5.     (line  4274)
fxif:                                          See 9.5.     (line  4276)
fxior:                                         See 9.5.     (line  4278)
fxlength:                                      See 9.5.     (line  4280)
fxmax:                                         See 9.5.     (line  4282)
fxmin:                                         See 9.5.     (line  4284)
fxmodulo:                                      See 9.5.     (line  4286)
fxnegative?:                                   See 9.5.     (line  4288)
fxnot:                                         See 9.5.     (line  4290)
fxodd?:                                        See 9.5.     (line  4292)
fxpositive?:                                   See 9.5.     (line  4294)
fxquotient:                                    See 9.5.     (line  4296)
fxremainder:                                   See 9.5.     (line  4298)
fxwrap*:                                       See 9.5.     (line  4300)
fxwrap+:                                       See 9.5.     (line  4302)
fxwrap-:                                       See 9.5.     (line  4304)
fxwrapabs:                                     See 9.5.     (line  4306)
fxwraparithmetic-shift:                        See 9.5.     (line  4308)
fxwraparithmetic-shift-left:                   See 9.5.     (line  4310)
fxwraplogical-shift-right:                     See 9.5.     (line  4312)
fxwrapquotient:                                See 9.5.     (line  4314)
fxxor:                                         See 9.5.     (line  4316)
fxzero?:                                       See 9.5.     (line  4318)
Gambit <1>:
          See ``Gambit''.                                   (line    28)
Gambit:                                        See 1.       (line    33)
gambit-scheme:                                 See 6.3.     (line  3214)
gambit.el:                                     See 5.6.     (line  2550)
GAMBOPT, environment variable:                 See 4.       (line  1658)
GC:                                            See 5.4.     (line  2414)
gc-report-set!:                                See 5.4.     (line  2414)
generate-proper-tail-calls:                    See 5.4.     (line  2316)
generative-lambda:                             See 6.3.     (line  3358)
generic:                                       See 6.3.     (line  3398)
gensym:                                        See 6.3.     (line  3049)
get-output-string:                             See 17.10.   (line 11248)
get-output-u8vector:                           See 17.11.   (line 11305)
get-output-vector:                             See 17.9.    (line 11217)
getenv:                                        See 16.6.    (line  8513)
group-info:                                    See 16.9.    (line  8920)
group-info-gid:                                See 16.9.    (line  8964)
group-info-members:                            See 16.9.    (line  8974)
group-info-name:                               See 16.9.    (line  8954)
group-info?:                                   See 16.9.    (line  8942)
gsc <1>:                                       See 1.       (line    33)
gsc <2>:                                       See 3.3.     (line   544)
gsc <3>:                                       See 3.5.     (line  1314)
gsc <4>:                                       See 4.       (line  1430)
gsc:                                           See 3.5.     (line  1364)
gsc-script:                                    See 2.5.     (line   365)
gsi <1>:                                       See 4.       (line  1430)
gsi <2>:                                       See 2.       (line   139)
gsi:                                           See 1.       (line    33)
gsi-script:                                    See 2.5.     (line   361)
hashing:                                       See 11.      (line  4921)
heap-overflow-exception?:                      See 15.2.    (line  7052)
help:                                          See 5.4.     (line  2045)
help-browser:                                  See 5.4.     (line  2046)
home directory:                                See 16.1.    (line  8129)
homogeneous vectors <1>:                       See 10.      (line  4657)
homogeneous vectors:                           See 18.9.    (line 12055)
host-info:                                     See 16.11.   (line  9092)
host-info-addresses:                           See 16.11.   (line  9144)
host-info-aliases:                             See 16.11.   (line  9134)
host-info-name:                                See 16.11.   (line  9124)
host-info?:                                    See 16.11.   (line  9112)
host-name:                                     See 16.11.   (line  9083)
ieee-scheme:                                   See 6.3.     (line  3214)
improper-length-list-exception-arg-num:        See 15.8.    (line  7884)
improper-length-list-exception-arguments:      See 15.8.    (line  7883)
improper-length-list-exception-procedure:      See 15.8.    (line  7882)
improper-length-list-exception?:               See 15.8.    (line  7881)
inactive-thread-exception-arguments:           See 6.4.     (line  3662)
inactive-thread-exception-procedure:           See 6.4.     (line  3661)
inactive-thread-exception?:                    See 6.4.     (line  3660)
include:                                       See 6.3.     (line  3133)
infinite?:                                     See 6.4.     (line  3763)
initialized-thread-exception-arguments:        See 6.4.     (line  3654)
initialized-thread-exception-procedure:        See 6.4.     (line  3653)
initialized-thread-exception?:                 See 6.4.     (line  3652)
inline:                                        See 6.3.     (line  3226)
inline-primitives:                             See 6.3.     (line  3229)
inlining-limit:                                See 6.3.     (line  3233)
input-port-byte-position:                      See 17.7.1.  (line 10415)
input-port-bytes-buffered:                     See 6.4.     (line  3677)
input-port-char-position:                      See 6.4.     (line  3742)
input-port-characters-buffered:                See 6.4.     (line  3679)
input-port-column:                             See 17.5.2.  (line  9875)
input-port-line:                               See 17.5.2.  (line  9874)
input-port-readtable:                          See 17.5.2.  (line 10032)
input-port-readtable-set!:                     See 17.5.2.  (line 10040)
input-port-timeout-set!:                       See 17.4.2.  (line  9791)
input-port?:                                   See 17.4.2.  (line  9635)
installation directories:                      See 16.1.    (line  8129)
integer->char:                                 See 8.1.     (line  3862)
integer-length:                                See 9.4.     (line  4117)
integer-nth-root:                              See 9.3.     (line  3976)
integer-sqrt:                                  See 9.3.     (line  3966)
interpreter <1>:                               See 2.       (line   135)
interpreter:                                   See 3.       (line   514)
interrupts-enabled:                            See 6.3.     (line  3285)
invalid-hash-number-exception-arguments:       See 6.4.     (line  3706)
invalid-hash-number-exception-procedure:       See 6.4.     (line  3705)
invalid-hash-number-exception?:                See 6.4.     (line  3704)
join-timeout-exception-arguments:              See 15.4.    (line  7270)
join-timeout-exception-procedure:              See 15.4.    (line  7269)
join-timeout-exception?:                       See 15.4.    (line  7268)
keyword->string:                               See 6.3.     (line  3025)
keyword-expected-exception-arguments:          See 15.9.    (line  8050)
keyword-expected-exception-procedure:          See 15.9.    (line  8049)
keyword-expected-exception?:                   See 15.9.    (line  8048)
keyword-hash:                                  See 11.1.    (line  5006)
keyword?:                                      See 6.3.     (line  3024)
keywords:                                      See 6.3.     (line  3026)
lambda:                                        See 6.2.     (line  2662)
lambda-lift:                                   See 6.3.     (line  3255)
LAST_.c:                                       See 3.3.     (line   753)
limitations:                                   See 20.      (line 13734)
link-flat:                                     See 3.5.     (line  1364)
link-incremental:                              See 3.5.     (line  1314)
list->f32vector:                               See 10.      (line  4829)
list->f64vector:                               See 10.      (line  4847)
list->s16vector:                               See 10.      (line  4721)
list->s32vector:                               See 10.      (line  4757)
list->s64vector:                               See 10.      (line  4793)
list->s8vector:                                See 10.      (line  4685)
list->table:                                   See 11.2.2.  (line  5448)
list->u16vector:                               See 10.      (line  4739)
list->u32vector:                               See 10.      (line  4775)
list->u64vector:                               See 10.      (line  4811)
list->u8vector:                                See 10.      (line  4703)
load:                                          See 3.5.     (line  1237)
mailbox-receive-timeout-exception-arguments:   See 13.9.    (line  6262)
mailbox-receive-timeout-exception-procedure:   See 13.9.    (line  6261)
mailbox-receive-timeout-exception?:            See 13.9.    (line  6260)
main:                                          See 6.4.     (line  3747)
make-condition-variable:                       See 13.9.    (line  6545)
make-f32vector:                                See 10.      (line  4823)
make-f64vector:                                See 10.      (line  4841)
make-mutex:                                    See 13.9.    (line  6304)
make-parameter:                                See 14.      (line  6757)
make-random-source:                            See 9.7.     (line  4504)
make-root-thread:                              See 13.9.    (line  5860)
make-s16vector:                                See 10.      (line  4715)
make-s32vector:                                See 10.      (line  4751)
make-s64vector:                                See 10.      (line  4787)
make-s8vector:                                 See 10.      (line  4679)
make-table:                                    See 11.2.2.  (line  5232)
make-thread:                                   See 13.9.    (line  5858)
make-thread-group:                             See 6.4.     (line  3617)
make-tls-context:                              See 17.7.3.  (line 10912)
make-u16vector:                                See 10.      (line  4733)
make-u32vector:                                See 10.      (line  4769)
make-u64vector:                                See 10.      (line  4805)
make-u8vector:                                 See 10.      (line  4697)
make-will:                                     See 11.2.1.  (line  5110)
mostly-fixnum:                                 See 6.3.     (line  3404)
mostly-fixnum-flonum:                          See 6.3.     (line  3404)
mostly-flonum:                                 See 6.3.     (line  3404)
mostly-flonum-fixnum:                          See 6.3.     (line  3404)
mostly-generic:                                See 6.3.     (line  3404)
multiple-c-return-exception?:                  See 15.5.    (line  7535)
mutex-lock!:                                   See 13.9.    (line  6387)
mutex-name:                                    See 13.9.    (line  6319)
mutex-specific:                                See 13.9.    (line  6326)
mutex-specific-set!:                           See 13.9.    (line  6327)
mutex-state:                                   See 13.9.    (line  6359)
mutex-unlock!:                                 See 13.9.    (line  6483)
mutex?:                                        See 13.9.    (line  6292)
namespace:                                     See 6.4.     (line  3752)
nan?:                                          See 6.4.     (line  3764)
network-info:                                  See 16.14.   (line  9442)
network-info-aliases:                          See 16.14.   (line  9488)
network-info-name:                             See 16.14.   (line  9478)
network-info-number:                           See 16.14.   (line  9498)
network-info?:                                 See 16.14.   (line  9466)
newline:                                       See 17.4.2.  (line  9704)
no-such-file-or-directory-exception-arguments: See 15.3.    (line  7141)
no-such-file-or-directory-exception-procedure: See 15.3.    (line  7140)
no-such-file-or-directory-exception?:          See 15.3.    (line  7139)
noncontinuable-exception-reason:               See 15.1.    (line  7015)
noncontinuable-exception?:                     See 15.1.    (line  7014)
nonempty-input-port-character-buffer-exception-arguments:See 6.4.
                                                            (line  3683)
nonempty-input-port-character-buffer-exception-procedure:See 6.4.
                                                            (line  3685)
nonempty-input-port-character-buffer-exception?:See 6.4.    (line  3681)
nonprocedure-operator-exception-arguments:     See 15.9.    (line  7986)
nonprocedure-operator-exception-code:          See 15.9.    (line  7987)
nonprocedure-operator-exception-operator:      See 15.9.    (line  7985)
nonprocedure-operator-exception-rte:           See 15.9.    (line  7988)
nonprocedure-operator-exception?:              See 15.9.    (line  7984)
normalized path:                               See 16.1.    (line  8222)
number-of-arguments-limit-exception-arguments: See 15.9.    (line  7952)
number-of-arguments-limit-exception-procedure: See 15.9.    (line  7951)
number-of-arguments-limit-exception?:          See 15.9.    (line  7950)
object file:                                   See 3.5.     (line  1237)
object->serial-number:                         See 11.1.    (line  4925)
object->string:                                See 17.10.   (line 11275)
object->u8vector:                              See 10.      (line  4872)
open-directory:                                See 17.8.    (line 11025)
open-dummy:                                    See 6.4.     (line  3671)
open-event-queue:                              See 6.4.     (line  3745)
open-file:                                     See 17.7.1.  (line 10333)
open-input-file:                               See 17.7.1.  (line 10334)
open-input-process:                            See 17.7.2.  (line 10478)
open-input-string:                             See 17.10.   (line 11240)
open-input-u8vector:                           See 17.11.   (line 11297)
open-input-vector:                             See 17.9.    (line 11090)
open-output-file:                              See 17.7.1.  (line 10335)
open-output-process:                           See 17.7.2.  (line 10479)
open-output-string:                            See 17.10.   (line 11241)
open-output-u8vector:                          See 17.11.   (line 11298)
open-output-vector:                            See 17.9.    (line 11091)
open-process:                                  See 17.7.2.  (line 10477)
open-string:                                   See 17.10.   (line 11239)
open-string-pipe:                              See 17.10.   (line 11247)
open-tcp-client:                               See 17.7.3.  (line 10680)
open-tcp-server:                               See 17.7.3.  (line 10767)
open-u8vector:                                 See 17.11.   (line 11296)
open-u8vector-pipe:                            See 17.11.   (line 11304)
open-vector:                                   See 17.9.    (line 11089)
open-vector-pipe:                              See 17.9.    (line 11173)
optimize-dead-definitions:                     See 6.3.     (line  3384)
optimize-dead-local-variables:                 See 6.3.     (line  3370)
options, compiler:                             See 3.3.     (line   576)
options, runtime:                              See 4.       (line  1430)
os-exception-arguments:                        See 15.3.    (line  7101)
os-exception-code:                             See 15.3.    (line  7102)
os-exception-message:                          See 15.3.    (line  7103)
os-exception-procedure:                        See 15.3.    (line  7100)
os-exception?:                                 See 15.3.    (line  7099)
output-port-byte-position:                     See 17.7.1.  (line 10416)
output-port-char-position:                     See 6.4.     (line  3743)
output-port-column:                            See 17.5.2.  (line  9877)
output-port-line:                              See 17.5.2.  (line  9876)
output-port-readtable:                         See 17.5.2.  (line 10033)
output-port-readtable-set!:                    See 17.5.2.  (line 10041)
output-port-timeout-set!:                      See 17.4.2.  (line  9792)
output-port-width:                             See 17.5.2.  (line  9901)
output-port?:                                  See 17.4.2.  (line  9636)
overflow, floating point:                      See 20.      (line 13734)
parameterize:                                  See 14.      (line  6813)
path-directory:                                See 16.1.    (line  8254)
path-expand:                                   See 16.1.    (line  8194)
path-extension:                                See 16.1.    (line  8252)
path-normalize:                                See 16.1.    (line  8222)
path-strip-directory:                          See 16.1.    (line  8255)
path-strip-extension:                          See 16.1.    (line  8253)
path-strip-trailing-directory-separator:       See 16.1.    (line  8256)
path-strip-volume:                             See 16.1.    (line  8258)
path-volume:                                   See 16.1.    (line  8257)
peek-char:                                     See 17.5.2.  (line  9931)
port-io-exception-handler-set!:                See 6.4.     (line  3675)
port-settings-set!:                            See 6.4.     (line  3673)
port?:                                         See 17.4.2.  (line  9637)
pp:                                            See 5.4.     (line  2397)
pretty-print:                                  See 5.4.     (line  2384)
primordial-exception-handler:                  See 6.4.     (line  3694)
print:                                         See 17.12.   (line 11342)
println:                                       See 17.12.   (line 11343)
process-pid:                                   See 17.7.2.  (line 10611)
process-status:                                See 17.7.2.  (line 10622)
process-times:                                 See 16.7.    (line  8593)
proper tail-calls <1>:                         See 5.4.     (line  2316)
proper tail-calls:                             See 6.3.     (line  3345)
proper-tail-calls:                             See 6.3.     (line  3345)
protocol-info:                                 See 16.13.   (line  9379)
protocol-info-aliases:                         See 16.13.   (line  9419)
protocol-info-name:                            See 16.13.   (line  9409)
protocol-info-number:                          See 16.13.   (line  9429)
protocol-info?:                                See 16.13.   (line  9397)
r4rs-scheme:                                   See 6.3.     (line  3214)
r5rs-scheme:                                   See 6.3.     (line  3214)
raise:                                         See 15.1.    (line  6996)
random-f64vector:                              See 9.7.     (line  4491)
random-integer:                                See 9.7.     (line  4449)
random-real:                                   See 9.7.     (line  4465)
random-source-make-f64vectors:                 See 9.7.     (line  4635)
random-source-make-integers:                   See 9.7.     (line  4582)
random-source-make-reals:                      See 9.7.     (line  4599)
random-source-make-u8vectors:                  See 9.7.     (line  4618)
random-source-pseudo-randomize!:               See 9.7.     (line  4549)
random-source-randomize!:                      See 9.7.     (line  4548)
random-source-state-ref:                       See 9.7.     (line  4528)
random-source-state-set!:                      See 9.7.     (line  4529)
random-source?:                                See 9.7.     (line  4516)
random-u8vector:                               See 9.7.     (line  4478)
range-exception-arg-num:                       See 15.8.    (line  7820)
range-exception-arguments:                     See 15.8.    (line  7819)
range-exception-procedure:                     See 15.8.    (line  7818)
range-exception?:                              See 15.8.    (line  7817)
read:                                          See 17.4.2.  (line  9659)
read-all:                                      See 17.4.2.  (line  9673)
read-char:                                     See 17.5.2.  (line  9912)
read-line:                                     See 17.5.2.  (line  9960)
read-substring:                                See 17.5.2.  (line  9995)
read-subu8vector:                              See 17.6.2.  (line 10285)
read-u8:                                       See 17.6.2.  (line 10237)
readtable-case-conversion?:                    See 18.1.    (line 11405)
readtable-case-conversion?-set:                See 18.1.    (line 11406)
readtable-eval-allowed?:                       See 18.1.    (line 11638)
readtable-eval-allowed?-set:                   See 18.1.    (line 11639)
readtable-keywords-allowed?:                   See 18.1.    (line 11452)
readtable-keywords-allowed?-set:               See 18.1.    (line 11453)
readtable-max-unescaped-char:                  See 18.1.    (line 11786)
readtable-max-unescaped-char-set:              See 18.1.    (line 11787)
readtable-max-write-length:                    See 18.1.    (line 11751)
readtable-max-write-length-set:                See 18.1.    (line 11752)
readtable-max-write-level:                     See 18.1.    (line 11716)
readtable-max-write-level-set:                 See 18.1.    (line 11717)
readtable-sharing-allowed?:                    See 18.1.    (line 11495)
readtable-sharing-allowed?-set:                See 18.1.    (line 11496)
readtable-start-syntax:                        See 18.1.    (line 11827)
readtable-start-syntax-set:                    See 18.1.    (line 11828)
readtable-write-cdr-read-macros?:              See 18.1.    (line 11662)
readtable-write-cdr-read-macros?-set:          See 18.1.    (line 11663)
readtable-write-extended-read-macros?:         See 18.1.    (line 11664)
readtable-write-extended-read-macros?-set:     See 18.1.    (line 11666)
readtable?:                                    See 18.1.    (line 11393)
real-time:                                     See 16.7.    (line  8595)
receive:                                       See 6.4.     (line  3756)
relative path:                                 See 16.1.    (line  8129)
rename-file:                                   See 16.2.    (line  8376)
repl-display-environment?:                     See 5.4.     (line  2368)
repl-input-port:                               See 6.4.     (line  3687)
repl-output-port:                              See 6.4.     (line  3688)
repl-result-history-max-length-set!:           See 5.4.     (line  2080)
repl-result-history-ref:                       See 5.4.     (line  2079)
replace-bit-field:                             See 9.4.     (line  4199)
rpc-remote-error-exception-arguments:          See 6.4.     (line  3666)
rpc-remote-error-exception-message:            See 6.4.     (line  3667)
rpc-remote-error-exception-procedure:          See 6.4.     (line  3665)
rpc-remote-error-exception?:                   See 6.4.     (line  3664)
run-time-bindings:                             See 6.3.     (line  3271)
runtime options:                               See 4.       (line  1430)
s16vector:                                     See 10.      (line  4716)
s16vector->list:                               See 10.      (line  4720)
s16vector-append:                              See 10.      (line  4726)
s16vector-copy:                                See 10.      (line  4725)
s16vector-fill!:                               See 10.      (line  4722)
s16vector-length:                              See 10.      (line  4717)
s16vector-ref:                                 See 10.      (line  4718)
s16vector-set!:                                See 10.      (line  4719)
s16vector-shrink!:                             See 10.      (line  4730)
s16vector?:                                    See 10.      (line  4714)
s32vector:                                     See 10.      (line  4752)
s32vector->list:                               See 10.      (line  4756)
s32vector-append:                              See 10.      (line  4762)
s32vector-copy:                                See 10.      (line  4761)
s32vector-fill!:                               See 10.      (line  4758)
s32vector-length:                              See 10.      (line  4753)
s32vector-ref:                                 See 10.      (line  4754)
s32vector-set!:                                See 10.      (line  4755)
s32vector-shrink!:                             See 10.      (line  4766)
s32vector?:                                    See 10.      (line  4750)
s64vector:                                     See 10.      (line  4788)
s64vector->list:                               See 10.      (line  4792)
s64vector-append:                              See 10.      (line  4798)
s64vector-copy:                                See 10.      (line  4797)
s64vector-fill!:                               See 10.      (line  4794)
s64vector-length:                              See 10.      (line  4789)
s64vector-ref:                                 See 10.      (line  4790)
s64vector-set!:                                See 10.      (line  4791)
s64vector-shrink!:                             See 10.      (line  4802)
s64vector?:                                    See 10.      (line  4786)
s8vector:                                      See 10.      (line  4680)
s8vector->list:                                See 10.      (line  4684)
s8vector-append:                               See 10.      (line  4690)
s8vector-copy:                                 See 10.      (line  4689)
s8vector-fill!:                                See 10.      (line  4686)
s8vector-length:                               See 10.      (line  4681)
s8vector-ref:                                  See 10.      (line  4682)
s8vector-set!:                                 See 10.      (line  4683)
s8vector-shrink!:                              See 10.      (line  4694)
s8vector?:                                     See 10.      (line  4678)
safe:                                          See 6.3.     (line  3277)
scheduler-exception-reason:                    See 15.4.    (line  7206)
scheduler-exception?:                          See 15.4.    (line  7205)
Scheme:                                        See 1.       (line    33)
Scheme, implementation of:
          See ``Gambit''.                                   (line    28)
scheme-ieee-1178-1990:                         See 2.5.     (line   352)
scheme-r4rs:                                   See 2.5.     (line   344)
scheme-r5rs:                                   See 2.5.     (line   348)
scheme-srfi-0:                                 See 2.5.     (line   356)
seconds->time:                                 See 16.7.    (line  8561)
separate:                                      See 6.3.     (line  3219)
serial-number->object:                         See 11.1.    (line  4926)
serialization <1>:                             See 10.      (line  4873)
serialization:                                 See 18.1.    (line 11496)
service-info:                                  See 16.12.   (line  9298)
service-info-aliases:                          See 16.12.   (line  9346)
service-info-name:                             See 16.12.   (line  9336)
service-info-port-number:                      See 16.12.   (line  9356)
service-info-protocol:                         See 16.12.   (line  9366)
service-info?:                                 See 16.12.   (line  9324)
set!:                                          See 20.      (line 13741)
set-box!:                                      See 6.3.     (line  2999)
setenv:                                        See 16.6.    (line  8514)
sfun-conversion-exception-arguments:           See 15.5.    (line  7479)
sfun-conversion-exception-code:                See 15.5.    (line  7480)
sfun-conversion-exception-message:             See 15.5.    (line  7481)
sfun-conversion-exception-procedure:           See 15.5.    (line  7478)
sfun-conversion-exception?:                    See 15.5.    (line  7477)
shell-command:                                 See 16.3.    (line  8441)
six-script:                                    See 2.5.     (line   369)
six.!:                                         See 6.4.     (line  3766)
six.!x:                                        See 6.4.     (line  3767)
six.&x:                                        See 6.4.     (line  3768)
six.*x:                                        See 6.4.     (line  3769)
six.++x:                                       See 6.4.     (line  3770)
six.+x:                                        See 6.4.     (line  3771)
six.--x:                                       See 6.4.     (line  3772)
six.-x:                                        See 6.4.     (line  3773)
six.arrow:                                     See 6.4.     (line  3774)
six.break:                                     See 6.4.     (line  3775)
six.call:                                      See 6.4.     (line  3776)
six.case:                                      See 6.4.     (line  3777)
six.clause:                                    See 6.4.     (line  3778)
six.compound:                                  See 6.4.     (line  3779)
six.cons:                                      See 6.4.     (line  3780)
six.continue:                                  See 6.4.     (line  3781)
six.define-procedure:                          See 6.4.     (line  3782)
six.define-variable:                           See 6.4.     (line  3783)
six.do-while:                                  See 6.4.     (line  3784)
six.dot:                                       See 6.4.     (line  3785)
six.for:                                       See 6.4.     (line  3786)
six.goto:                                      See 6.4.     (line  3787)
six.identifier:                                See 6.4.     (line  3788)
six.if:                                        See 6.4.     (line  3789)
six.index:                                     See 6.4.     (line  3790)
six.label:                                     See 6.4.     (line  3791)
six.list:                                      See 6.4.     (line  3792)
six.literal:                                   See 6.4.     (line  3793)
six.make-array:                                See 6.4.     (line  3794)
six.new:                                       See 6.4.     (line  3795)
six.null:                                      See 6.4.     (line  3796)
six.prefix:                                    See 6.4.     (line  3797)
six.procedure:                                 See 6.4.     (line  3798)
six.procedure-body:                            See 6.4.     (line  3799)
six.return:                                    See 6.4.     (line  3800)
six.switch:                                    See 6.4.     (line  3801)
six.while:                                     See 6.4.     (line  3802)
six.x!=y:                                      See 6.4.     (line  3803)
six.x%=y:                                      See 6.4.     (line  3804)
six.x%y:                                       See 6.4.     (line  3805)
six.x&&y:                                      See 6.4.     (line  3806)
six.x&=y:                                      See 6.4.     (line  3807)
six.x&y:                                       See 6.4.     (line  3808)
six.x*=y:                                      See 6.4.     (line  3809)
six.x*y:                                       See 6.4.     (line  3810)
six.x++:                                       See 6.4.     (line  3811)
six.x+=y:                                      See 6.4.     (line  3812)
six.x+y:                                       See 6.4.     (line  3813)
six.x--:                                       See 6.4.     (line  3815)
six.x-=y:                                      See 6.4.     (line  3816)
six.x-y:                                       See 6.4.     (line  3817)
six.x/=y:                                      See 6.4.     (line  3818)
six.x/y:                                       See 6.4.     (line  3819)
six.x:-y:                                      See 6.4.     (line  3820)
six.x:=y:                                      See 6.4.     (line  3821)
six.x:y:                                       See 6.4.     (line  3822)
six.x<<=y:                                     See 6.4.     (line  3823)
six.x<<y:                                      See 6.4.     (line  3824)
six.x<=y:                                      See 6.4.     (line  3825)
six.x<y:                                       See 6.4.     (line  3826)
six.x==y:                                      See 6.4.     (line  3827)
six.x=y:                                       See 6.4.     (line  3828)
six.x>=y:                                      See 6.4.     (line  3829)
six.x>>=y:                                     See 6.4.     (line  3830)
six.x>>y:                                      See 6.4.     (line  3831)
six.x>y:                                       See 6.4.     (line  3832)
six.x?y:z:                                     See 6.4.     (line  3833)
six.x^=y:                                      See 6.4.     (line  3834)
six.x^y:                                       See 6.4.     (line  3835)
six.~x:                                        See 6.4.     (line  3839)
socket-info-address:                           See 6.4.     (line  3714)
socket-info-family:                            See 6.4.     (line  3715)
socket-info-port-number:                       See 6.4.     (line  3716)
socket-info?:                                  See 6.4.     (line  3713)
stack-overflow-exception?:                     See 15.2.    (line  7074)
standard-bindings:                             See 6.3.     (line  3261)
started-thread-exception-arguments:            See 15.4.    (line  7305)
started-thread-exception-procedure:            See 15.4.    (line  7304)
started-thread-exception?:                     See 15.4.    (line  7303)
step:                                          See 5.4.     (line  2193)
step-level-set!:                               See 5.4.     (line  2194)
string->keyword:                               See 6.3.     (line  3026)
string->uninterned-keyword:                    See 6.3.     (line  3089)
string->uninterned-symbol:                     See 6.3.     (line  3066)
string-ci<=?:                                  See 8.2.     (line  3910)
string-ci<?:                                   See 8.2.     (line  3908)
string-ci=?:                                   See 8.2.     (line  3907)
string-ci=?-hash:                              See 11.1.    (line  5030)
string-ci>=?:                                  See 8.2.     (line  3911)
string-ci>?:                                   See 8.2.     (line  3909)
string-shrink!:                                See 6.3.     (line  2982)
string<=?:                                     See 8.2.     (line  3905)
string<?:                                      See 8.2.     (line  3903)
string=?:                                      See 8.2.     (line  3902)
string=?-hash:                                 See 11.1.    (line  5018)
string>=?:                                     See 8.2.     (line  3906)
string>?:                                      See 8.2.     (line  3904)
subf32vector:                                  See 10.      (line  4835)
subf32vector-fill!:                            See 10.      (line  4831)
subf32vector-move!:                            See 10.      (line  4837)
subf64vector:                                  See 10.      (line  4853)
subf64vector-fill!:                            See 10.      (line  4849)
subf64vector-move!:                            See 10.      (line  4855)
subs16vector:                                  See 10.      (line  4727)
subs16vector-fill!:                            See 10.      (line  4723)
subs16vector-move!:                            See 10.      (line  4729)
subs32vector:                                  See 10.      (line  4763)
subs32vector-fill!:                            See 10.      (line  4759)
subs32vector-move!:                            See 10.      (line  4765)
subs64vector:                                  See 10.      (line  4799)
subs64vector-fill!:                            See 10.      (line  4795)
subs64vector-move!:                            See 10.      (line  4801)
subs8vector:                                   See 10.      (line  4691)
subs8vector-fill!:                             See 10.      (line  4687)
subs8vector-move!:                             See 10.      (line  4693)
substring-fill!:                               See 6.3.     (line  2951)
substring-move!:                               See 6.3.     (line  2966)
subu16vector:                                  See 10.      (line  4745)
subu16vector-fill!:                            See 10.      (line  4741)
subu16vector-move!:                            See 10.      (line  4747)
subu32vector:                                  See 10.      (line  4781)
subu32vector-fill!:                            See 10.      (line  4777)
subu32vector-move!:                            See 10.      (line  4783)
subu64vector:                                  See 10.      (line  4817)
subu64vector-fill!:                            See 10.      (line  4813)
subu64vector-move!:                            See 10.      (line  4819)
subu8vector:                                   See 10.      (line  4709)
subu8vector-fill!:                             See 10.      (line  4705)
subu8vector-move!:                             See 10.      (line  4711)
subvector:                                     See 6.3.     (line  2858)
subvector-fill!:                               See 6.3.     (line  2894)
subvector-move!:                               See 6.3.     (line  2909)
symbol-hash:                                   See 11.1.    (line  4994)
syntax-case:                                   See 6.3.     (line  3180)
syntax-rules:                                  See 6.3.     (line  3180)
system-stamp:                                  See 6.4.     (line  3725)
system-type:                                   See 6.4.     (line  3721)
system-type-string:                            See 6.4.     (line  3722)
system-version:                                See 6.4.     (line  3718)
system-version-string:                         See 6.4.     (line  3719)
table->list:                                   See 11.2.2.  (line  5429)
table-copy:                                    See 11.2.2.  (line  5497)
table-for-each:                                See 11.2.2.  (line  5400)
table-length:                                  See 11.2.2.  (line  5300)
table-merge:                                   See 11.2.2.  (line  5535)
table-merge!:                                  See 11.2.2.  (line  5514)
table-ref:                                     See 11.2.2.  (line  5325)
table-search:                                  See 11.2.2.  (line  5369)
table-set!:                                    See 11.2.2.  (line  5349)
table?:                                        See 11.2.2.  (line  5288)
tables:                                        See 11.      (line  4921)
tail-calls <1>:                                See 5.4.     (line  2316)
tail-calls:                                    See 6.3.     (line  3345)
tcp-client-peer-socket-info:                   See 6.4.     (line  3708)
tcp-client-self-socket-info:                   See 6.4.     (line  3709)
tcp-server-socket-info:                        See 6.4.     (line  3711)
tcp-service-register!:                         See 17.7.3.  (line 10879)
tcp-service-unregister!:                       See 17.7.3.  (line 10881)
terminated-thread-exception-arguments:         See 15.4.    (line  7337)
terminated-thread-exception-procedure:         See 15.4.    (line  7336)
terminated-thread-exception?:                  See 15.4.    (line  7335)
test-bit-field?:                               See 9.4.     (line  4197)
this-source-file:                              See 6.4.     (line  3754)
thread-base-priority:                          See 13.9.    (line  5930)
thread-base-priority-set!:                     See 13.9.    (line  5931)
thread-group->thread-group-list:               See 6.4.     (line  3624)
thread-group->thread-group-vector:             See 6.4.     (line  3625)
thread-group->thread-list:                     See 6.4.     (line  3626)
thread-group->thread-vector:                   See 6.4.     (line  3627)
thread-group-name:                             See 6.4.     (line  3619)
thread-group-parent:                           See 6.4.     (line  3620)
thread-group-resume!:                          See 6.4.     (line  3621)
thread-group-suspend!:                         See 6.4.     (line  3622)
thread-group-terminate!:                       See 6.4.     (line  3623)
thread-group?:                                 See 6.4.     (line  3618)
thread-init!:                                  See 6.4.     (line  3650)
thread-interrupt!:                             See 6.4.     (line  3641)
thread-join!:                                  See 13.9.    (line  6131)
thread-mailbox-extract-and-rewind:             See 13.9.    (line  6202)
thread-mailbox-next:                           See 13.9.    (line  6200)
thread-mailbox-rewind:                         See 13.9.    (line  6201)
thread-name:                                   See 13.9.    (line  5908)
thread-priority-boost:                         See 13.9.    (line  5946)
thread-priority-boost-set!:                    See 13.9.    (line  5947)
thread-quantum:                                See 13.9.    (line  5962)
thread-quantum-set!:                           See 13.9.    (line  5963)
thread-receive:                                See 13.9.    (line  6199)
thread-resume!:                                See 6.4.     (line  3644)
thread-send:                                   See 13.9.    (line  6182)
thread-sleep!:                                 See 13.9.    (line  6020)
thread-specific:                               See 13.9.    (line  5915)
thread-specific-set!:                          See 13.9.    (line  5916)
thread-start!:                                 See 13.9.    (line  5982)
thread-state:                                  See 6.4.     (line  3629)
thread-state-abnormally-terminated-reason:     See 6.4.     (line  3638)
thread-state-abnormally-terminated?:           See 6.4.     (line  3637)
thread-state-active-timeout:                   See 6.4.     (line  3634)
thread-state-active-waiting-for:               See 6.4.     (line  3633)
thread-state-active?:                          See 6.4.     (line  3632)
thread-state-initialized?:                     See 6.4.     (line  3631)
thread-state-normally-terminated-result:       See 6.4.     (line  3636)
thread-state-normally-terminated?:             See 6.4.     (line  3635)
thread-state-uninitialized?:                   See 6.4.     (line  3630)
thread-suspend!:                               See 6.4.     (line  3643)
thread-terminate!:                             See 13.9.    (line  6044)
thread-thread-group:                           See 6.4.     (line  3646)
thread-yield!:                                 See 13.9.    (line  6001)
thread?:                                       See 13.9.    (line  5846)
threads:                                       See 13.      (line  5603)
time:                                          See 16.7.    (line  8622)
time->seconds:                                 See 16.7.    (line  8560)
time?:                                         See 16.7.    (line  8559)
timeout->time:                                 See 6.4.     (line  3669)
top:                                           See 6.4.     (line  3639)
touch:                                         See 6.4.     (line  3728)
trace:                                         See 5.4.     (line  2120)
transcript-off:                                See 6.1.     (line  2649)
transcript-on:                                 See 6.1.     (line  2648)
tty-history:                                   See 6.4.     (line  3731)
tty-history-max-length-set!:                   See 6.4.     (line  3733)
tty-history-set!:                              See 6.4.     (line  3732)
tty-mode-set!:                                 See 6.4.     (line  3736)
tty-paren-balance-duration-set!:               See 6.4.     (line  3734)
tty-text-attributes-set!:                      See 6.4.     (line  3735)
tty-type-set!:                                 See 6.4.     (line  3737)
tty?:                                          See 6.4.     (line  3730)
type-exception-arg-num:                        See 15.8.    (line  7774)
type-exception-arguments:                      See 15.8.    (line  7773)
type-exception-procedure:                      See 15.8.    (line  7772)
type-exception-type-id:                        See 15.8.    (line  7775)
type-exception?:                               See 15.8.    (line  7771)
u16vector:                                     See 10.      (line  4734)
u16vector->list:                               See 10.      (line  4738)
u16vector-append:                              See 10.      (line  4744)
u16vector-copy:                                See 10.      (line  4743)
u16vector-fill!:                               See 10.      (line  4740)
u16vector-length:                              See 10.      (line  4735)
u16vector-ref:                                 See 10.      (line  4736)
u16vector-set!:                                See 10.      (line  4737)
u16vector-shrink!:                             See 10.      (line  4748)
u16vector?:                                    See 10.      (line  4732)
u32vector:                                     See 10.      (line  4770)
u32vector->list:                               See 10.      (line  4774)
u32vector-append:                              See 10.      (line  4780)
u32vector-copy:                                See 10.      (line  4779)
u32vector-fill!:                               See 10.      (line  4776)
u32vector-length:                              See 10.      (line  4771)
u32vector-ref:                                 See 10.      (line  4772)
u32vector-set!:                                See 10.      (line  4773)
u32vector-shrink!:                             See 10.      (line  4784)
u32vector?:                                    See 10.      (line  4768)
u64vector:                                     See 10.      (line  4806)
u64vector->list:                               See 10.      (line  4810)
u64vector-append:                              See 10.      (line  4816)
u64vector-copy:                                See 10.      (line  4815)
u64vector-fill!:                               See 10.      (line  4812)
u64vector-length:                              See 10.      (line  4807)
u64vector-ref:                                 See 10.      (line  4808)
u64vector-set!:                                See 10.      (line  4809)
u64vector-shrink!:                             See 10.      (line  4820)
u64vector?:                                    See 10.      (line  4804)
u8vector:                                      See 10.      (line  4698)
u8vector->list:                                See 10.      (line  4702)
u8vector->object:                              See 10.      (line  4873)
u8vector-append:                               See 10.      (line  4708)
u8vector-copy:                                 See 10.      (line  4707)
u8vector-fill!:                                See 10.      (line  4704)
u8vector-length:                               See 10.      (line  4699)
u8vector-ref:                                  See 10.      (line  4700)
u8vector-set!:                                 See 10.      (line  4701)
u8vector-shrink!:                              See 10.      (line  4712)
u8vector?:                                     See 10.      (line  4696)
unbound-global-exception-code:                 See 15.7.    (line  7744)
unbound-global-exception-rte:                  See 15.7.    (line  7745)
unbound-global-exception-variable:             See 15.7.    (line  7743)
unbound-global-exception?:                     See 15.7.    (line  7742)
unbound-os-environment-variable-exception-arguments:See 15.3.
                                                            (line  7173)
unbound-os-environment-variable-exception-procedure:See 15.3.
                                                            (line  7172)
unbound-os-environment-variable-exception?:    See 15.3.    (line  7171)
unbound-serial-number-exception-arguments:     See 11.1.    (line  4966)
unbound-serial-number-exception-procedure:     See 11.1.    (line  4965)
unbound-serial-number-exception?:              See 11.1.    (line  4964)
unbound-table-key-exception-arguments:         See 11.2.2.  (line  5469)
unbound-table-key-exception-procedure:         See 11.2.2.  (line  5468)
unbound-table-key-exception?:                  See 11.2.2.  (line  5467)
unbox:                                         See 6.3.     (line  2998)
unbreak:                                       See 5.4.     (line  2271)
uncaught-exception-arguments:                  See 15.4.    (line  7372)
uncaught-exception-procedure:                  See 15.4.    (line  7371)
uncaught-exception-reason:                     See 15.4.    (line  7373)
uncaught-exception?:                           See 15.4.    (line  7370)
uninitialized-thread-exception-arguments:      See 6.4.     (line  3658)
uninitialized-thread-exception-procedure:      See 6.4.     (line  3657)
uninitialized-thread-exception?:               See 6.4.     (line  3656)
uninterned-keyword?:                           See 6.3.     (line  3090)
uninterned-symbol?:                            See 6.3.     (line  3067)
unknown-keyword-argument-exception-arguments:  See 15.9.    (line  8019)
unknown-keyword-argument-exception-procedure:  See 15.9.    (line  8018)
unknown-keyword-argument-exception?:           See 15.9.    (line  8017)
unterminated-process-exception-arguments:      See 17.7.2.  (line 10646)
unterminated-process-exception-procedure:      See 17.7.2.  (line 10645)
unterminated-process-exception?:               See 17.7.2.  (line 10644)
untrace:                                       See 5.4.     (line  2121)
user-info:                                     See 16.10.   (line  8996)
user-info-gid:                                 See 16.10.   (line  9051)
user-info-home:                                See 16.10.   (line  9061)
user-info-name:                                See 16.10.   (line  9032)
user-info-shell:                               See 16.10.   (line  9071)
user-info-uid:                                 See 16.10.   (line  9042)
user-info?:                                    See 16.10.   (line  9020)
user-name:                                     See 16.10.   (line  8987)
vector-append:                                 See 6.3.     (line  2870)
vector-copy:                                   See 6.3.     (line  2843)
vector-shrink!:                                See 6.3.     (line  2925)
void:                                          See 6.3.     (line  3110)
weak references:                               See 11.      (line  4921)
will-execute!:                                 See 11.2.1.  (line  5113)
will-testator:                                 See 11.2.1.  (line  5112)
will?:                                         See 11.2.1.  (line  5111)
with-exception-catcher:                        See 15.1.    (line  6967)
with-exception-handler:                        See 15.1.    (line  6942)
with-input-from-file:                          See 17.7.1.  (line 10338)
with-input-from-port:                          See 6.4.     (line  3739)
with-input-from-process:                       See 17.7.2.  (line 10482)
with-input-from-string:                        See 17.10.   (line 11244)
with-input-from-u8vector:                      See 17.11.   (line 11301)
with-input-from-vector:                        See 17.9.    (line 11094)
with-output-to-file:                           See 17.7.1.  (line 10339)
with-output-to-port:                           See 6.4.     (line  3740)
with-output-to-process:                        See 17.7.2.  (line 10483)
with-output-to-string:                         See 17.10.   (line 11245)
with-output-to-u8vector:                       See 17.11.   (line 11302)
with-output-to-vector:                         See 17.9.    (line 11095)
write:                                         See 17.4.2.  (line  9693)
write-char:                                    See 17.5.2.  (line  9947)
write-substring:                               See 17.5.2.  (line  9996)
write-subu8vector:                             See 17.6.2.  (line 10286)
write-u8:                                      See 17.6.2.  (line 10273)
wrong-number-of-arguments-exception-arguments: See 15.9.    (line  7922)
wrong-number-of-arguments-exception-procedure: See 15.9.    (line  7921)
wrong-number-of-arguments-exception?:          See 15.9.    (line  7920)
|six.x:                                        See 6.4.     (line  3814)
|six.x\|=y|:                                   See 6.4.     (line  3836)
|six.x\|\|y|:                                  See 6.4.     (line  3838)
|six.x\|y|:                                    See 6.4.     (line  3837)
~:                                             See 16.1.    (line  8152)
~username:                                     See 16.1.    (line  8161)
~~:                                            See 16.1.    (line  8143)
Table of Contents
*****************

Gambit
1 The Gambit system
  1.1 Accessing the system files
2 The Gambit Scheme interpreter
  2.1 Interactive mode
  2.2 Batch mode
  2.3 Customization
  2.4 Process exit status
  2.5 Scheme scripts
    2.5.1 Scripts under UNIX and Mac OS X
    2.5.2 Scripts under Microsoft Windows
    2.5.3 Compiling scripts
3 The Gambit Scheme compiler
  3.1 Interactive mode
  3.2 Customization
  3.3 Batch mode
  3.4 Link files
    3.4.1 Building an executable program
    3.4.2 Building a loadable library
    3.4.3 Building a shared-library
    3.4.4 Other compilation options
  3.5 Procedures specific to compiler
4 Runtime options
5 Debugging
  5.1 Debugging model
  5.2 Debugging commands
  5.3 Debugging example
  5.4 Procedures related to debugging
  5.5 Console line-editing
  5.6 Emacs interface
  5.7 GUIDE
6 Scheme extensions
  6.1 Extensions to standard procedures
  6.2 Extensions to standard special forms
  6.3 Miscellaneous extensions
  6.4 Undocumented extensions
7 Namespaces
8 Characters and strings
  8.1 Extensions to character procedures
  8.2 Extensions to string procedures
9 Numbers
  9.1 Extensions to numeric procedures
  9.2 IEEE floating point arithmetic
  9.3 Integer square root and nth root
  9.4 Bitwise-operations on exact integers
  9.5 Fixnum specific operations
  9.6 Flonum specific operations
  9.7 Pseudo random numbers
10 Homogeneous vectors
11 Hashing and weak references
  11.1 Hashing
  11.2 Weak references
    11.2.1 Wills
    11.2.2 Tables
12 Records
13 Threads
  13.1 Introduction
  13.2 Thread objects
  13.3 Mutex objects
  13.4 Condition variable objects
  13.5 Fairness
  13.6 Memory coherency
  13.7 Timeouts
  13.8 Primordial thread
  13.9 Procedures
14 Dynamic environment
15 Exceptions
  15.1 Exception-handling
  15.2 Exception objects related to memory management
  15.3 Exception objects related to the host environment
  15.4 Exception objects related to threads
  15.5 Exception objects related to C-interface
  15.6 Exception objects related to the reader
  15.7 Exception objects related to evaluation and compilation
  15.8 Exception objects related to type checking
  15.9 Exception objects related to procedure call
  15.10 Other exception objects
16 Host environment
  16.1 Handling of file names
  16.2 Filesystem operations
  16.3 Shell command execution
  16.4 Process termination
  16.5 Command line arguments
  16.6 Environment variables
  16.7 Measuring time
  16.8 File information
  16.9 Group information
  16.10 User information
  16.11 Host information
  16.12 Service information
  16.13 Protocol information
  16.14 Network information
17 I/O and ports
  17.1 Unidirectional and bidirectional ports
  17.2 Port classes
  17.3 Port settings
  17.4 Object-ports
    17.4.1 Object-port settings
    17.4.2 Object-port operations
  17.5 Character-ports
    17.5.1 Character-port settings
    17.5.2 Character-port operations
  17.6 Byte-ports
    17.6.1 Byte-port settings
    17.6.2 Byte-port operations
  17.7 Device-ports
    17.7.1 Filesystem devices
    17.7.2 Process devices
    17.7.3 Network devices
  17.8 Directory-ports
  17.9 Vector-ports
  17.10 String-ports
  17.11 U8vector-ports
  17.12 Other procedures related to I/O
18 Lexical syntax and readtables
  18.1 Readtables
  18.2 Boolean syntax
  18.3 Character syntax
  18.4 String syntax
  18.5 Symbol syntax
  18.6 Keyword syntax
  18.7 Box syntax
  18.8 Number syntax
  18.9 Homogeneous vector syntax
  18.10 Special `#!' syntax
  18.11 Multiline comment syntax
  18.12 Scheme infix syntax extension
    18.12.1 SIX grammar
    18.12.2 SIX semantics
19 C-interface
  19.1 The mapping of types between C and Scheme
  19.2 The `c-declare' special form
  19.3 The `c-initialize' special form
  19.4 The `c-lambda' special form
  19.5 The `c-define' special form
  19.6 The `c-define-type' special form
  19.7 Continuations and the C-interface
20 System limitations
21 Copyright and license
General index


