path: root/scm.info

This is scm.info, produced by makeinfo version 4.7 from scm.texi.

INFO-DIR-SECTION The Algorithmic Language Scheme
START-INFO-DIR-ENTRY
* SCM: (scm).           A Scheme interpreter.
END-INFO-DIR-ENTRY


File: scm.info,  Node: Top,  Next: Overview,  Prev: (dir),  Up: (dir)

This manual documents the SCM Scheme implementation.  SCM version
5e2 was released February 2006.  The most recent information about SCM        |
can be found on SCM's "WWW" home page:                                        |

               `http://swiss.csail.mit.edu/~jaffer/SCM'

Copyright (C) 1990-1999 Free Software Foundation

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 author.

* Menu:

* Overview::
* Installing SCM::              How to
* Operational Features::
* The Language::                Reference.
* Packages::                    Optional Capabilities.
* The Implementation::          How it works.
* Index::


File: scm.info,  Node: Overview,  Next: Installing SCM,  Prev: Top,  Up: Top

1 Overview
**********

Scm is a portable Scheme implementation written in C.  Scm provides a
machine independent platform for [JACAL], a symbolic algebra system.

* Menu:

* SCM Features::
* SCM Authors::
* Copying::
* Bibliography::


File: scm.info,  Node: SCM Features,  Next: SCM Authors,  Prev: Overview,  Up: Overview

1.1 Features
============

   * Conforms to Revised^5 Report on the Algorithmic Language Scheme
     [R5RS] and the [IEEE] P1178 specification.

   * Support for [SICP], [R2RS], [R3RS], and [R5RS] scheme code.

   * Runs under Amiga, Atari-ST, MacOS, MS-DOS, OS/2, NOS/VE, Unicos,
     VMS, Unix and similar systems.  Supports ASCII and EBCDIC
     character sets.

   * Is fully documented in TeXinfo form, allowing documentation to be
     generated in info, TeX, html, nroff, and troff formats.

   * Supports inexact real and complex numbers, 30 bit immediate
     integers and large precision integers.

   * Many Common Lisp functions: `logand', `logor', `logxor', `lognot',
     `ash', `logcount', `integer-length', `bit-extract', `defmacro',
     `macroexpand', `macroexpand1', `gentemp', `defvar', `force-output',
     `software-type', `get-decoded-time', `get-internal-run-time',
     `get-internal-real-time', `delete-file', `rename-file',
     `copy-tree', `acons', and `eval'.

   * `Char-code-limit', `most-positive-fixnum', `most-negative-fixnum',
     `and internal-time-units-per-second' constants.  `*Features*' and
     `*load-pathname*' variables.

   * Arrays and bit-vectors.  String ports and software emulation ports.
     I/O extensions providing ANSI C and POSIX.1 facilities.

   * Interfaces to standard libraries including REGEX string regular
     expression matching and the CURSES screen management package.

   * Available add-on packages including an interactive debugger,
     database, X-window graphics, BGI graphics, Motif, and Open-Windows
     packages.

   * A compiler (HOBBIT, available separately) and dynamic linking of
     compiled modules.

   * User definable responses to interrupts and errors,
     Process-syncronization primitives.  Setable levels of monitoring
     and timing information printed interactively (the `verbose'
     function).  `Restart', `quit', and `exec'.


File: scm.info,  Node: SCM Authors,  Next: Copying,  Prev: SCM Features,  Up: Overview

1.2 Authors
===========

Aubrey Jaffer (agj @ alum.mit.edu)
     Most of SCM.

Radey Shouman
     Arrays, `gsubr's, compiled closures, records, Ecache, syntax-rules
     macros, and "safeport"s.

Jerry D. Hedden
     Real and Complex functions.  Fast mixed type arithmetics.

Hugh Secker-Walker
     Syntax checking and memoization of special forms by evaluator.
     Storage allocation strategy and parameters.

George Carrette
     "Siod", written by George Carrette, was the starting point for SCM.
     The major innovations taken from Siod are the evaluator's use of
     the C-stack and being able to garbage collect off the C-stack
     (*note Garbage Collection::).

There are many other contributors to SCM.  They are acknowledged in the
file `ChangeLog', a log of changes that have been made to scm.


File: scm.info,  Node: Copying,  Next: Bibliography,  Prev: SCM Authors,  Up: Overview

1.3 Copyright
=============

Authors have assigned their SCM copyrights to:

                    Free Software Foundation, Inc.
           59 Temple Place, Suite 330, Boston, MA 02111, USA

* Menu:

* The SCM License::
* SIOD copyright::


File: scm.info,  Node: The SCM License,  Next: SIOD copyright,  Prev: Copying,  Up: Copying

1.3.1 The SCM License
---------------------

The license of SCM consists of the GNU GPL plus a special statement
giving blanket permission to link with non-free software.  This is the
license statement as found in any individual file that it applies to:

     This program is free software; you can redistribute it and/or
     modify it under the terms of the GNU General Public License as
     published by the Free Software Foundation; either version 2, or
     (at your option) any later version.

     This program 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
     General Public License for more details.

     You should have received a copy of the GNU General Public License
     along with this software; see the file COPYING.  If not, write to
     the Free Software Foundation, Inc., 59 Temple Place, Suite 330,
     Boston, MA 02111-1307 USA

     As a special exception, the Free Software Foundation gives
     permission for additional uses of the text contained in its
     release of SCM.

     The exception is that, if you link the SCM library with other
     files to produce an executable, this does not by itself cause the
     resulting executable to be covered by the GNU General Public
     License.  Your use of that executable is in no way restricted on
     account of linking the SCM library code into it.

     This exception does not however invalidate any other reasons why
     the executable file might be covered by the GNU General Public
     License.

     This exception applies only to the code released by the Free
     Software Foundation under the name SCM.  If you copy code from
     other Free Software Foundation releases into a copy of SCM, as the
     General Public License permits, the exception does not apply to
     the code that you add in this way.  To avoid misleading anyone as
     to the status of such modified files, you must delete this
     exception notice from them.

     If you write modifications of your own for SCM, it is your choice
     whether to permit this exception to apply to your modifications.
     If you do not wish that, delete this exception notice.

Permission to use, copy, modify, distribute, and sell this software and
its documentation for any purpose is hereby granted without fee,
provided that the above copyright notice appear in all copies and that
both that copyright notice and this permission notice appear in
supporting documentation.

                              NO WARRANTY

BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR
THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW.  EXCEPT WHEN
OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES
PROVIDE THE PROGRAM "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 PROGRAM IS WITH
YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL
NECESSARY SERVICING, REPAIR OR CORRECTION.

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 PROGRAM 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 PROGRAM
(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 PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR
OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.


File: scm.info,  Node: SIOD copyright,  Prev: The SCM License,  Up: Copying

1.3.2 SIOD copyright
--------------------


                         COPYRIGHT (c) 1989 BY
      PARADIGM ASSOCIATES INCORPORATED, CAMBRIDGE, MASSACHUSETTS.
                          ALL RIGHTS RESERVED

Permission to use, copy, modify, distribute and sell this software and
its documentation for any purpose and without fee is hereby granted,
provided that the above copyright notice appear in all copies and that
both that copyright notice and this permission notice appear in
supporting documentation, and that the name of Paradigm Associates Inc
not be used in advertising or publicity pertaining to distribution of
the software without specific, written prior permission.

PARADIGM DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
EVENT SHALL PARADIGM BE LIABLE FOR ANY SPECIAL, INDIRECT OR
CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.

gjc@paradigm.com
                                                    Phone: 617-492-6079

Paradigm Associates Inc
29 Putnam Ave, Suite 6
Cambridge, MA 02138


File: scm.info,  Node: Bibliography,  Prev: Copying,  Up: Overview

1.4 Bibliography
================

[IEEE]
     `IEEE Standard 1178-1990.  IEEE Standard for the Scheme
     Programming Language.'  IEEE, New York, 1991.

[R4RS]
     William Clinger and Jonathan Rees, Editors.  Revised(4) Report on
     the Algorithmic Language Scheme.  `ACM Lisp Pointers' Volume IV,
     Number 3 (July-September 1991), pp. 1-55.

     *Note Top: (r4rs)Top.

[R5RS]
     Richard Kelsey and William Clinger and Jonathan (Rees, editors)
     Revised(5) Report on the Algorithmic Language Scheme.
     `Higher-Order and Symbolic Computation' Volume 11, Number 1 (1998),
     pp. 7-105, and `ACM SIGPLAN Notices' 33(9), September 1998.

     *Note Top: (r5rs)Top.

[Exrename]
     William Clinger Hygienic Macros Through Explicit Renaming `Lisp
     Pointers' Volume IV, Number 4 (December 1991), pp 17-23.

[SICP]
     Harold Abelson and Gerald Jay Sussman with Julie Sussman.
     `Structure and Interpretation of Computer Programs.'  MIT Press,
     Cambridge, 1985.

[Simply]
     Brian Harvey and Matthew Wright.  `Simply Scheme: Introducing
     Computer Science' MIT Press, 1994 ISBN 0-262-08226-8

[SchemePrimer]
     $B8$;tBg(B(Dai Inukai) `$BF~Lg(BScheme'
     1999$BG/(B12$B7n=iHG(B ISBN4-87966-954-7

[SLIB]
     Todd R. Eigenschink, Dave Love, and Aubrey Jaffer.  SLIB, The
     Portable Scheme Library.  Version 2c8, June 2000.

     *Note Top: (slib)Top.

[JACAL]
     Aubrey Jaffer.  JACAL Symbolic Mathematics System.  Version 1b0,
     Sep 1999.

     *Note Top: (jacal)Top.

`scm.texi'
`scm.info'
     Documentation of `scm' extensions (beyond Scheme standards).
     Documentation on the internal representation and how to extend or
     include `scm' in other programs.

`Xlibscm.texi'
`Xlibscm.info'
     Documentation of the Xlib - SCM Language X Interface.


File: scm.info,  Node: Installing SCM,  Next: Operational Features,  Prev: Overview,  Up: Top

2 Installing SCM
****************

* Menu:

* Making SCM::                  Bootstrapping.
* SLIB::                        REQUIREd reading.
* Building SCM::
* Installing Dynamic Linking::
* Configure Module Catalog::
* Saving Images::               Make Fast-Booting Executables
* Automatic C Preprocessor Definitions::
* Problems Compiling::
* Problems Linking::
* Problems Running::
* Testing::
* Reporting Problems::


File: scm.info,  Node: Making SCM,  Next: SLIB,  Prev: Installing SCM,  Up: Installing SCM

2.1 Making SCM
==============

The SCM distribution has "Makefile" which contains rules for making
"scmlit", a "bare-bones" version of SCM sufficient for running `build'.
`build' is used to compile (or create scripts to compile) full
featured versions.

Makefiles are not portable to the majority of platforms.  If `Makefile'
works for you, good; If not, I don't want to hear about it.  If you
need to compile SCM without build, there are several ways to proceed:

   * Use the build (http://swiss.csail.mit.edu/~jaffer/buildscm.html)
     web page to create custom batch scripts for compiling SCM.

   * Use SCM on a different platform to run `build' to create a script
     to build SCM;

   * Use another implementation of Scheme to run `build' to create a
     script to build SCM;

   * Create your own script or `Makefile'.


File: scm.info,  Node: SLIB,  Next: Building SCM,  Prev: Making SCM,  Up: Installing SCM

2.2 SLIB
========

[SLIB] is a portable Scheme library meant to provide compatibility and
utility functions for all standard Scheme implementations.  Although
SLIB is not _neccessary_ to run SCM, I strongly suggest you obtain and
install it.  Bug reports about running SCM without SLIB have very low
priority.  SLIB is available from the same sites as SCM:

   * swiss.csail.mit.edu:/pub/scm/slib3a3.tar.gz                              |

   * ftp.gnu.org:/pub/gnu/jacal/slib3a3.tar.gz                                |

   * ftp.cs.indiana.edu:/pub/scheme-repository/imp/slib3a3.tar.gz             |

Unpack SLIB (`tar xzf slib3a3.tar.gz' or `unzip -ao slib3a3.zip') in an       |
appropriate directory for your system; both `tar' and `unzip' will
create the directory `slib'.

Then create a file `require.scm' in the SCM "implementation-vicinity"
(this is the same directory as where the file `Init5e2.scm' is                |
installed).  `require.scm' should have the contents:

     (define (library-vicinity) "/usr/local/lib/slib/")

where the pathname string `/usr/local/lib/slib/' is to be replaced by
the pathname into which you installed SLIB.  Absolute pathnames are
recommended here; if you use a relative pathname, SLIB can get confused
when the working directory is changed (*note chmod: I/O-Extensions.).
The way to specify a relative pathname is to append it to the
implementation-vicinity, which is absolute:

     (define library-vicinity
       (let ((lv (string-append (implementation-vicinity) "../slib/")))
         (lambda () lv)))

Alternatively, you can set the (shell) environment variable
`SCHEME_LIBRARY_PATH' to the pathname of the SLIB directory (*note
SCHEME_LIBRARY_PATH: SCM Variables.).  If set, the environment variable
overrides `require.scm'.  Again, absolute pathnames are recommended.


File: scm.info,  Node: Building SCM,  Next: Installing Dynamic Linking,  Prev: SLIB,  Up: Installing SCM

2.3 Building SCM
================

The file "build" loads the file "build.scm", which constructs a
relational database of how to compile and link SCM executables.
`build.scm' has information for the platforms which SCM has been ported
to (of which I have been notified).  Some of this information is old,
incorrect, or incomplete.  Send corrections and additions to jaffer @
ai.mit.edu.

* Menu:

* Invoking Build::
* Build Options::
* Compiling and Linking Custom Files::


File: scm.info,  Node: Invoking Build,  Next: Build Options,  Prev: Building SCM,  Up: Building SCM

2.3.1 Invoking Build
--------------------

The _all_ method will also work for MS-DOS and unix.  Use the _all_
method if you encounter problems with `build'.

MS-DOS
     From the SCM source directory, type `build' followed by up to 9
     command line arguments.

unix
     From the SCM source directory, type `./build' followed by command
     line arguments.

_all_
     From the SCM source directory, start `scm' or `scmlit' and type
     `(load "build")'.  Alternatively, start `scm' or `scmlit' with the
     command line argument `-ilbuild'.


Invoking build without the `-F' option will build or create a shell
script with the `arrays', `inexact', and `bignums' options as defaults.

     bash$ ./build
     -|
     #! /bin/sh
     # unix (linux) script created by SLIB/batch
     # ================ Write file with C defines
     rm -f scmflags.h
     echo '#define IMPLINIT "Init5e2.scm"'>>scmflags.h                        |
     echo '#define BIGNUMS'>>scmflags.h
     echo '#define FLOATS'>>scmflags.h
     echo '#define ARRAYS'>>scmflags.h
     # ================ Compile C source files
     gcc -O2 -c continue.c scm.c scmmain.c findexec.c script.c time.c repl.c scl.c eval.c sys.c subr.c debug.c unif.c rope.c
     # ================ Link C object files
     gcc -rdynamic -o scm continue.o scm.o scmmain.o findexec.o script.o time.o repl.o scl.o eval.o sys.o subr.o debug.o unif.o rope.o -lm -lc

To cross compile for another platform, invoke build with the `-p' or
`--platform=' option.  This will create a script for the platform named
in the `-p' or `--platform=' option.

     bash$ ./build -o scmlit -p darwin -F lit
     -|
     #! /bin/sh
     # unix (darwin) script created by SLIB/batch
     # ================ Write file with C defines
     rm -f scmflags.h
     echo '#define IMPLINIT "Init5e2.scm"'>>scmflags.h                        |
     # ================ Compile C source files
     cc -O3 -c continue.c scm.c scmmain.c findexec.c script.c time.c repl.c scl.c eval.c sys.c subr.c debug.c unif.c rope.c
     # ================ Link C object files
     mv -f scmlit scmlit~
     cc -o scmlit continue.o scm.o scmmain.o findexec.o script.o time.o repl.o scl.o eval.o sys.o subr.o debug.o unif.o rope.o


File: scm.info,  Node: Build Options,  Next: Compiling and Linking Custom Files,  Prev: Invoking Build,  Up: Building SCM

2.3.2 Build Options
-------------------

The options to "build" specify what, where, and how to build a SCM
program or dynamically linked module.  These options are unrelated to
the SCM command line options.

 -- Build Option: -p PLATFORM-NAME
 -- Build Option: --platform=PLATFORM-NAME
     specifies that the compilation should be for a
     computer/operating-system combination called PLATFORM-NAME.
     _Note_ The case of PLATFORM-NAME is distinguised.  The current
     PLATFORM-NAMEs are all lower-case.

     The platforms defined by table "platform" in `build.scm' are:

     Table: platform
     name              processor         operating-system  compiler
     #f                processor-family  operating-system  #f
     symbol            processor-family  operating-system  symbol
     symbol            symbol            symbol            symbol
     ================= ================= ================= =================
     *unknown*         *unknown*         unix              cc
     acorn-unixlib     acorn             *unknown*         cc
     aix               powerpc           aix               cc
     alpha-elf         alpha             unix              cc
     alpha-linux       alpha             linux             gcc
     amiga-aztec       m68000            amiga             cc
     amiga-dice-c      m68000            amiga             dcc
     amiga-gcc         m68000            amiga             gcc
     amiga-sas         m68000            amiga             lc
     atari-st-gcc      m68000            atari.st          gcc
     atari-st-turbo-c  m68000            atari.st          tcc
     borland-c         i8086             ms-dos            bcc
     darwin            powerpc           unix              cc
     djgpp             i386              ms-dos            gcc
     freebsd           i386              unix              cc
     gcc               *unknown*         unix              gcc
     gnu-win32         i386              unix              gcc
     highc             i386              ms-dos            hc386
     hp-ux             hp-risc           hp-ux             cc
     irix              mips              irix              gcc
     linux             i386              linux             gcc
     linux-aout        i386              linux             gcc
     microsoft-c       i8086             ms-dos            cl
     microsoft-c-nt    i386              ms-dos            cl
     microsoft-quick-c i8086             ms-dos            qcl
     ms-dos            i8086             ms-dos            cc
     netbsd            *unknown*         unix              gcc
     openbsd           *unknown*         unix              gcc
     os/2-cset         i386              os/2              icc
     os/2-emx          i386              os/2              gcc
     osf1              alpha             unix              cc
     plan9-8           i386              plan9             8c
     sunos             sparc             sunos             cc
     svr4              *unknown*         unix              cc
     svr4-gcc-sun-ld   sparc             sunos             gcc
     turbo-c           i8086             ms-dos            tcc
     unicos            cray              unicos            cc
     unix              *unknown*         unix              cc
     vms               vax               vms               cc
     vms-gcc           vax               vms               gcc
     watcom-9.0        i386              ms-dos            wcc386p

 -- Build Option: -f PATHNAME
     specifies that the build options contained in PATHNAME be spliced
     into the argument list at this point.  The use of option files can
     separate functional features from platform-specific ones.

     The `Makefile' calls out builds with the options in `.opt' files:

    `dlls.opt'
          Options for Makefile targets mydlls, myturtle, and x.so.

    `gdb.opt'
          Options for udgdbscm and gdbscm.

    `libscm.opt'
          Options for libscm.a.

    `pg.opt'
          Options for pgscm, which instruments C functions.

    `udscm4.opt'
          Options for targets udscm4 and dscm4 (scm).

    `udscm5.opt'
          Options for targets udscm5 and dscm5 (scm).

     The Makefile creates options files it depends on only if they do
     not already exist.

 -- Build Option: -o FILENAME
 -- Build Option: --outname=FILENAME
     specifies that the compilation should produce an executable or
     object name of FILENAME.  The default is `scm'.  Executable
     suffixes will be added if neccessary, e.g. `scm' => `scm.exe'.

 -- Build Option: -l LIBNAME ...
 -- Build Option: --libraries=LIBNAME
     specifies that the LIBNAME should be linked with the executable
     produced.  If compile flags or include directories (`-I') are
     needed, they are automatically supplied for compilations.  The `c'
     library is always included.  SCM "features" specify any libraries
     they need; so you shouldn't need this option often.

 -- Build Option: -D DEFINITION ...
 -- Build Option: --defines=DEFINITION
     specifies that the DEFINITION should be made in any C source
     compilations.  If compile flags or include directories (`-I') are
     needed, they are automatically supplied for compilations.  SCM
     "features" specify any flags they need; so you shouldn't need this
     option often.

 -- Build Option: --compiler-options=FLAG
     specifies that that FLAG will be put on compiler command-lines.

 -- Build Option: --linker-options=FLAG
     specifies that that FLAG will be put on linker command-lines.

 -- Build Option: -s PATHNAME
 -- Build Option: --scheme-initial=PATHNAME
     specifies that PATHNAME should be the default location of the SCM
     initialization file `Init5e2.scm'.  SCM tries several likely             |
     locations before resorting to PATHNAME (*note File-System
     Habitat::).  If not specified, the current directory (where build
     is building) is used.

 -- Build Option: -c PATHNAME ...
 -- Build Option: --c-source-files=PATHNAME
     specifies that the C source files PATHNAME ... are to be compiled.

 -- Build Option: -j PATHNAME ...
 -- Build Option: --object-files=PATHNAME
     specifies that the object files PATHNAME ... are to be linked.

 -- Build Option: -i CALL ...
 -- Build Option: --initialization=CALL
     specifies that the C functions CALL ... are to be invoked during
     initialization.

 -- Build Option: -t BUILD-WHAT
 -- Build Option: --type=BUILD-WHAT
     specifies in general terms what sort of thing to build.  The
     choices are:
    `exe'
          executable program.

    `lib'
          library module.

    `dlls'
          archived dynamically linked library object files.

    `dll'
          dynamically linked library object file.

     The default is to build an executable.

 -- Build Option: -h BATCH-SYNTAX
 -- Build Option: -batch-dialect=BATCH-SYNTAX
     specifies how to build.  The default is to create a batch file for
     the host system.  The SLIB file `batch.scm' knows how to create
     batch files for:
        * unix

        * dos

        * vms

        * amigaos (was amigados)

        * system

          This option executes the compilation and linking commands
          through the use of the `system' procedure.

        * *unknown*

          This option outputs Scheme code.

 -- Build Option: -w BATCH-FILENAME
 -- Build Option: -script-name=BATCH-FILENAME
     specifies where to write the build script.  The default is to
     display it on `(current-output-port)'.

 -- Build Option: -F FEATURE ...
 -- Build Option: --features=FEATURE
     specifies to build the given features into the executable.  The
     defined features are:

    "array"
          Alias for ARRAYS

    "array-for-each"
          array-map! and array-for-each (arrays must also be featured).

    "arrays"
          Use if you want arrays, uniform-arrays and uniform-vectors.

    "bignums"
          Large precision integers.

    "byte"
          Treating strings as byte-vectors.

    "careful-interrupt-masking"
          Define this for extra checking of interrupt masking and some
          simple checks for proper use of malloc and free.  This is for
          debugging C code in `sys.c', `eval.c', `repl.c' and makes the
          interpreter several times slower than usual.

    "cautious"
          Normally, the number of arguments arguments to interpreted
          closures (from LAMBDA) are checked if the function part of a
          form is not a symbol or only the first time the form is
          executed if the function part is a symbol.  defining
          `reckless' disables any checking.  If you want to have SCM
          always check the number of arguments to interpreted closures
          define feature `cautious'.

    "cheap-continuations"
          If you only need straight stack continuations, executables
          compile with this feature will run faster and use less
          storage than not having it.  Machines with unusual stacks
          _need_ this.  Also, if you incorporate new C code into scm
          which uses VMS system services or library routines (which
          need to unwind the stack in an ordrly manner) you may need to
          use this feature.

    "compiled-closure"
          Use if you want to use compiled closures.

    "curses"
          For the "curses" screen management package.

    "debug"
          Turns on the features `cautious',
          `careful-interrupt-masking', and `stack-limit'; uses `-g'
          flags for debugging SCM source code.

    "differ"
          Sequence comparison

    "dump"
          Convert a running scheme program into an executable file.

    "dynamic-linking"
          Be able to load compiled files while running.

    "edit-line"
          interface to the editline or GNU readline library.

    "engineering-notation"
          Use if you want floats to display in engineering notation
          (exponents always multiples of 3) instead of scientific
          notation.

    "generalized-c-arguments"
          `make_gsubr' for arbitrary (< 11) arguments to C functions.

    "i/o-extensions"
          Commonly available I/O extensions: "exec", line I/O, file
          positioning, file delete and rename, and directory functions.

    "inexact"
          Use if you want floating point numbers.

    "lit"
          Lightweight - no features

    "macro"
          C level support for hygienic and referentially transparent
          macros (syntax-rules macros).

    "mysql"
          Client connections to the mysql databases.

    "no-heap-shrink"
          Use if you want segments of unused heap to not be freed up
          after garbage collection.  This may increase time in GC for
          *very* large working sets.

    "none"
          No features

    "posix"
          Posix functions available on all "Unix-like" systems.  fork
          and process functions, user and group IDs, file permissions,
          and "link".

    "reckless"
          If your scheme code runs without any errors you can disable
          almost all error checking by compiling all files with
          `reckless'.

    "record"
          The Record package provides a facility for user to define
          their own record data types.  See SLIB for documentation.

    "regex"
          String regular expression matching.

    "rev2-procedures"
          These procedures were specified in the `Revised^2 Report on
          Scheme' but not in `R4RS'.

    "sicp"
          Use if you want to run code from:

          Harold Abelson and Gerald Jay Sussman with Julie Sussman.
          `Structure and Interpretation of Computer Programs.'  The MIT
          Press, Cambridge, Massachusetts, USA, 1985.

          Differences from R5RS are:
             * (eq? '() '#f)

             * (define a 25) returns the symbol a.

             * (set! a 36) returns 36.

    "single-precision-only"
          Use if you want all inexact real numbers to be single
          precision.  This only has an effect if SINGLES is also
          defined (which is the default).  This does not affect complex
          numbers.

    "socket"
          BSD "socket" interface.  Socket addr functions require
          inexacts or bignums for 32-bit precision.

    "stack-limit"
          Use to enable checking for stack overflow.  Define value of
          the C preprocessor variable STACK_LIMIT to be the size to
          which SCM should allow the stack to grow.  STACK_LIMIT should
          be less than the maximum size the hardware can support, as
          not every routine checks the stack.

    "tick-interrupts"
          Use if you want the ticks and ticks-interrupt functions.

    "turtlegr"
          "Turtle" graphics calls for both Borland-C and X11 from
          sjm@ee.tut.fi.

    "unix"
          Those unix features which have not made it into the Posix
          specs: nice, acct, lstat, readlink, symlink, mknod and sync.

    "wb"
          WB database with relational wrapper.

    "windows"
          Microsoft Windows executable.

    "x"
          Alias for Xlib feature.

    "xlib"
          Interface to Xlib graphics routines.



File: scm.info,  Node: Compiling and Linking Custom Files,  Prev: Build Options,  Up: Building SCM

2.3.3 Compiling and Linking Custom Files
----------------------------------------

A correspondent asks:

     How can we link in our own c files to the SCM interpreter so that
     we can add our own functionality?  (e.g. we have a bunch of tcp
     functions we want access to).  Would this involve changing
     build.scm or the Makefile or both?

(*note Changing Scm:: has instructions describing the C code format).  Suppose
a C file "foo.c" has functions you wish to add to SCM.  To compile and
link your file at compile time, use the `-c' and `-i' options to build:

     bash$ ./build -c foo.c -i init_foo
     -|
     #! /bin/sh
     rm -f scmflags.h
     echo '#define IMPLINIT "/home/jaffer/scm/Init5e2.scm"'>>scmflags.h       |
     echo '#define COMPILED_INITS init_foo();'>>scmflags.h
     echo '#define BIGNUMS'>>scmflags.h
     echo '#define FLOATS'>>scmflags.h
     echo '#define ARRAYS'>>scmflags.h
     gcc -O2 -c continue.c scm.c findexec.c script.c time.c repl.c scl.c \
             eval.c sys.c subr.c unif.c rope.c foo.c
     gcc -rdynamic -o scm continue.o scm.o findexec.o script.o time.o \
             repl.o scl.o eval.o sys.o subr.o unif.o rope.o foo.o -lm -lc

To make a dynamically loadable object file use the `-t dll' option:

     bash$ ./build -t dll -c foo.c
     -|
     #! /bin/sh
     rm -f scmflags.h
     echo '#define IMPLINIT "/home/jaffer/scm/Init5e2.scm"'>>scmflags.h       |
     echo '#define BIGNUMS'>>scmflags.h
     echo '#define FLOATS'>>scmflags.h
     echo '#define ARRAYS'>>scmflags.h
     echo '#define DLL'>>scmflags.h
     gcc -O2 -fpic -c foo.c
     gcc -shared -o foo.so foo.o -lm -lc

Once `foo.c' compiles correctly (and your SCM build supports
dynamic-loading), you can load the compiled file with the Scheme command
`(load "./foo.so")'.  See *Note Configure Module Catalog:: for how to
add a compiled dll file to SLIB's catalog.


File: scm.info,  Node: Installing Dynamic Linking,  Next: Configure Module Catalog,  Prev: Building SCM,  Up: Installing SCM

2.4 Installing Dynamic Linking
==============================

Dynamic linking has not been ported to all platforms.  Operating systems
in the BSD family (a.out binary format) can usually be ported to "DLD".
The "dl" library (`#define SUN_DL' for SCM) was a proposed POSIX
standard and may be available on other machines with "COFF" binary
format.  For notes about porting to MS-Windows and finishing the port
to VMS *Note VMS Dynamic Linking::.

"DLD" is a library package of C functions that performs "dynamic link
editing" on Linux, VAX (Ultrix), Sun 3 (SunOS 3.4 and 4.0),
SPARCstation (SunOS 4.0), Sequent Symmetry (Dynix), and Atari ST.  It is
available from:

   * ftp.gnu.org:pub/gnu/dld-3.3.tar.gz

These notes about using libdl on SunOS are from `gcc.info':

     On a Sun, linking using GNU CC fails to find a shared library and
     reports that the library doesn't exist at all.

     This happens if you are using the GNU linker, because it does only
     static linking and looks only for unshared libraries.  If you have
     a shared library with no unshared counterpart, the GNU linker
     won't find anything.

     We hope to make a linker which supports Sun shared libraries, but
     please don't ask when it will be finished-we don't know.

     Sun forgot to include a static version of `libdl.a' with some
     versions of SunOS (mainly 4.1).  This results in undefined symbols
     when linking static binaries (that is, if you use `-static').  If
     you see undefined symbols `_dlclose', `_dlsym' or `_dlopen' when
     linking, compile and link against the file `mit/util/misc/dlsym.c'
     from the MIT version of X windows.


File: scm.info,  Node: Configure Module Catalog,  Next: Saving Images,  Prev: Installing Dynamic Linking,  Up: Installing SCM

2.5 Configure Module Catalog
============================

The SLIB module "catalog" can be extended to define other
`require'-able packages by adding calls to the Scheme source file
`mkimpcat.scm'.  Within `mkimpcat.scm', the following procedures are
defined.

 -- Function: add-link feature object-file lib1 ...
     FEATURE should be a symbol.  OBJECT-FILE should be a string naming
     a file containing compiled "object-code".  Each LIBn argument
     should be either a string naming a library file or `#f'.

     If OBJECT-FILE exists, the `add-link' procedure registers symbol
     FEATURE so that the first time `require' is called with the symbol
     FEATURE as its argument, OBJECT-FILE and the LIB1 ... are
     dynamically linked into the executing SCM session.

     If OBJECT-FILE exists, `add-link' returns `#t', otherwise it
     returns `#f'.

     For example, to install a compiled dll `foo', add these lines to
     `mkimpcat.scm':

                  (add-link 'foo
                            (in-vicinity (implementation-vicinity) "foo"
                                         link:able-suffix))


 -- Function: add-alias alias feature
     ALIAS and FEATURE are symbols.  The procedure `add-alias'
     registers ALIAS as an alias for FEATURE.  An unspecified value is
     returned.

     `add-alias' causes `(require 'ALIAS)' to behave like `(require
     'FEATURE)'.

 -- Function: add-source feature filename
     FEATURE is a symbol.  FILENAME is a string naming a file
     containing Scheme source code.  The procedure `add-source'
     registers FEATURE so that the first time `require' is called with
     the symbol FEATURE as its argument, the file FILENAME will be
     `load'ed.  An unspecified value is returned.

Remember to delete the file `slibcat' after modifying the file
`mkimpcat.scm' in order to force SLIB to rebuild its cache.


File: scm.info,  Node: Saving Images,  Next: Automatic C Preprocessor Definitions,  Prev: Configure Module Catalog,  Up: Installing SCM

2.6 Saving Images
=================

In SCM, the ability to save running program images is called "dump"
(*note Dump::).  In order to make `dump' available to SCM, build with
feature `dump'.  `dump'ed executables are compatible with dynamic
linking.

Most of the code for "dump" is taken from `emacs-19.34/src/unex*.c'.
No modifications to the emacs source code were required to use
`unexelf.c'.  Dump has not been ported to all platforms.  If `unexec.c'
or `unexelf.c' don't work for you, try using the appropriate `unex*.c'
file from emacs.


File: scm.info,  Node: Automatic C Preprocessor Definitions,  Next: Problems Compiling,  Prev: Saving Images,  Up: Installing SCM

2.7 Automatic C Preprocessor Definitions
========================================

These `#defines' are automatically provided by preprocessors of various
C compilers.  SCM uses the presence or absence of these definitions to
configure "include file" locations and aliases for library functions.
If the definition(s) corresponding to your system type is missing as
your system is configured, add `-DFLAG' to the compilation command
lines or add a `#define FLAG' line to `scmfig.h' or the beginning of
`scmfig.h'.

     #define         Platforms:
     -------         ----------
     ARM_ULIB        Huw Rogers free unix library for acorn archimedes
     AZTEC_C         Aztec_C 5.2a
     __CYGWIN__      Cygwin
     __CYGWIN32__    Cygwin                                                   |
     _DCC            Dice C on AMIGA
     __GNUC__        Gnu CC (and DJGPP)
     __EMX__         Gnu C port (gcc/emx 0.8e) to OS/2 2.0
     __HIGHC__       MetaWare High C
     __IBMC__        C-Set++ on OS/2 2.1
     _MSC_VER        MS VisualC++ 4.2
     MWC             Mark Williams C on COHERENT
     __MWERKS__      Metrowerks Compiler; Macintosh and WIN32 (?)
     _POSIX_SOURCE   ??
     _QC             Microsoft QuickC
     __STDC__        ANSI C compliant
     __TURBOC__      Turbo C and Borland C
     __USE_POSIX     ??
     __WATCOMC__     Watcom C on MS-DOS
     __ZTC__         Zortech C

     _AIX            AIX operating system
     __APPLE__       Apple Darwin
     AMIGA           SAS/C 5.10 or Dice C on AMIGA
     __amigaos__     Gnu CC on AMIGA
     atarist         ATARI-ST under Gnu CC
     __FreeBSD__     FreeBSD
     GNUDOS          DJGPP (obsolete in version 1.08)
     __GO32__        DJGPP (future?)
     hpux            HP-UX
     linux           Linux
     macintosh       Macintosh (THINK_C and __MWERKS__ define)
     MCH_AMIGA       Aztec_c 5.2a on AMIGA
     __MACH__        Apple Darwin
     __MINGW32__     MinGW - Minimalist GNU for Windows                       |
     MSDOS           Microsoft C 5.10 and 6.00A
     _MSDOS          Microsoft CLARM and CLTHUMB compilers.
     __MSDOS__       Turbo C, Borland C, and DJGPP
     __NetBSD__      NetBSD
     nosve           Control Data NOS/VE
     SVR2            System V Revision 2.
     sun             SunOS
     __SVR4          SunOS
     THINK_C         developement environment for the Macintosh
     ultrix          VAX with ULTRIX operating system.
     unix            most Unix and similar systems and DJGPP (!?)
     __unix__        Gnu CC and DJGPP
     _UNICOS         Cray operating system
     vaxc            VAX C compiler
     VAXC            VAX C compiler
     vax11c          VAX C compiler
     VAX11           VAX C compiler
     _Windows        Borland C 3.1 compiling for Windows
     _WIN32          MS VisualC++ 4.2 and Cygwin (Win32 API)
     _WIN32_WCE      MS Windows CE
     vms             (and VMS) VAX-11 C under VMS.

     __alpha         DEC Alpha processor
     __alpha__       DEC Alpha processor
     hp9000s800      HP RISC processor
     __ia64          GCC on IA64
     __ia64__        GCC on IA64
     _LONGLONG       GCC on IA64
     __i386__        DJGPP
     i386            DJGPP
     _M_ARM          Microsoft CLARM compiler defines as 4 for ARM.
     _M_ARMT         Microsoft CLTHUMB compiler defines as 4 for Thumb.
     MULTIMAX        Encore computer
     ppc             PowerPC
     __ppc__         PowerPC
     pyr             Pyramid 9810 processor
     __sgi__         Silicon Graphics Inc.
     sparc           SPARC processor
     sequent         Sequent computer
     tahoe           CCI Tahoe processor
     vax             VAX processor
     __x86_64        AMD Opteron                                              |


File: scm.info,  Node: Problems Compiling,  Next: Problems Linking,  Prev: Automatic C Preprocessor Definitions,  Up: Installing SCM

2.8 Problems Compiling
======================

FILE    PROBLEM / MESSAGE                HOW TO FIX
*.c     include file not found.          Correct the status of
                                         STDC_HEADERS in scmfig.h.
                                         fix #include statement or add
                                         #define for system type to
                                         scmfig.h.
*.c     Function should return a value.  Ignore.
        Parameter is never used.         
        Condition is always false.       
        Unreachable code in function.    
scm.c   assignment between incompatible  Change SIGRETTYPE in scm.c.
        types.                           
time.c  CLK_TCK redefined.               incompatablility between
                                         <stdlib.h> and <sys/types.h>.
                                         Remove STDC_HEADERS in scmfig.h.
                                         Edit <sys/types.h> to remove
                                         incompatability.
subr.c  Possibly incorrect assignment    Ignore.
        in function lgcd.                
sys.c   statement not reached.           Ignore.
        constant in conditional          
        expression.                      
sys.c   undeclared, outside of           #undef STDC_HEADERS in scmfig.h.
        functions.                       
scl.c   syntax error.                    #define SYSTNAME to your system
                                         type in scl.c (softtype).


File: scm.info,  Node: Problems Linking,  Next: Problems Running,  Prev: Problems Compiling,  Up: Installing SCM

2.9 Problems Linking
====================

PROBLEM                              HOW TO FIX
_sin etc. missing.                   Uncomment LIBS in makefile.


File: scm.info,  Node: Problems Running,  Next: Testing,  Prev: Problems Linking,  Up: Installing SCM

2.10 Problems Running
=====================

PROBLEM                              HOW TO FIX
Opening message and then machine     Change memory model option to C
crashes.                             compiler (or makefile).
                                     Make sure sizet definition is
                                     correct in scmfig.h.
                                     Reduce the size of HEAP_SEG_SIZE in
                                     setjump.h.
Input hangs.                         #define NOSETBUF
ERROR: heap: need larger initial.    Increase initial heap allocation
                                     using -a<kb> or INIT_HEAP_SIZE.
ERROR: Could not allocate.           Check sizet definition.
                                     Use 32 bit compiler mode.
                                     Don't try to run as subproccess.
remove <FLAG> in scmfig.h and        Do so and recompile files.
recompile scm.                       
add <FLAG> in scmfig.h and           
recompile scm.                       
ERROR: Init5e2.scm not found.        Assign correct IMPLINIT in makefile      |
                                     or scmfig.h.
                                     Define environment variable
                                     SCM_INIT_PATH to be the full
                                     pathname of Init5e2.scm.                 |
WARNING: require.scm not found.      Define environment variable
                                     SCHEME_LIBRARY_PATH to be the full
                                     pathname of the scheme library
                                     [SLIB].
                                     Change library-vicinity in
                                     Init5e2.scm to point to library or       |
                                     remove.
                                     Make sure the value of
                                     (library-vicinity) has a trailing
                                     file separator (like / or \).


File: scm.info,  Node: Testing,  Next: Reporting Problems,  Prev: Problems Running,  Up: Installing SCM

2.11 Testing
============

Loading `r4rstest.scm' in the distribution will run an [R4RS]
conformance test on `scm'.

     > (load "r4rstest.scm")
     -|
     ;loading "r4rstest.scm"
     SECTION(2 1)
     SECTION(3 4)
      #<primitive-procedure boolean?>
         #<primitive-procedure char?>
            #<primitive-procedure null?>
               #<primitive-procedure number?>
     ...

Loading `pi.scm' in the distribution will enable you to compute digits
of pi.

     > (load "pi")
     ;loading "pi"
     ;done loading "pi.scm"
     ;Evaluation took 20 ms (0 in gc) 767 cells work, 233.B other
     #<unspecified>
     > (pi 100 5)
     00003 14159 26535 89793 23846 26433 83279 50288 41971 69399
     37510 58209 74944 59230 78164 06286 20899 86280 34825 34211
     70679
     ;Evaluation took 550 ms (60 in gc) 36976 cells work, 1548.B other
     #<unspecified>

Loading `bench.scm' will compute and display performance statistics of
SCM running `pi.scm'.  `make bench' or `make benchlit' appends the
performance report to the file `BenchLog', facilitating tracking
effects of changes to SCM on performance.

PROBLEM                              HOW TO FIX
Runs some and then machine crashes.  See above under machine crashes.
Runs some and then ERROR: ...        Remove optimization option to C
(after a GC has happened).           compiler and recompile.
                                     #define SHORT_ALIGN in `scmfig.h'.
Some symbol names print incorrectly. Change memory model option to C
                                     compiler (or makefile).
                                     Check that HEAP_SEG_SIZE fits
                                     within sizet.
                                     Increase size of HEAP_SEG_SIZE (or
                                     INIT_HEAP_SIZE if it is smaller
                                     than HEAP_SEG_SIZE).
ERROR: Rogue pointer in Heap.        See above under machine crashes.
Newlines don't appear correctly in   Check file mode (define OPEN_... in
output files.                        `Init5e2.scm').                          |
Spaces or control characters appear  Check character defines in
in symbol names.                     `scmfig.h'.
Negative numbers turn positive.      Check SRS in `scmfig.h'.
VMS: Couldn't unwind stack.          #define CHEAP_CONTIUATIONS in
                                     `scmfig.h'.
VAX: botched longjmp.                

Sparc(SUN-4) heap is growing out of control
     You are experiencing a GC problem peculiar to the Sparc.  The
     problem is that SCM doesn't know how to clear register windows.
     Every location which is not reused still gets marked at GC time.
     This causes lots of stuff which should be collected to not be.
     This will be a problem with any _conservative_ GC until we find
     what instruction will clear the register windows.  This problem is
     exacerbated by using lots of call-with-current-continuations.  A
     possible fix for dynthrow() is commented out in `continue.c'.


File: scm.info,  Node: Reporting Problems,  Prev: Testing,  Up: Installing SCM

2.12 Reporting Problems
=======================

Reported problems and solutions are grouped under Compiling, Linking,
Running, and Testing.  If you don't find your problem listed there, you
can send a bug report to `agj @ alum.mit.edu'.  The bug report should
include:

  1. The version of SCM (printed when SCM is invoked with no arguments).

  2. The type of computer you are using.

  3. The name and version of your computer's operating system.

  4. The values of the environment variables `SCM_INIT_PATH' and
     `SCHEME_LIBRARY_PATH'.

  5. The name and version of your C compiler.

  6. If you are using an executable from a distribution, the name,
     vendor, and date of that distribution.  In this case,
     corresponding with the vendor is recommended.


File: scm.info,  Node: Operational Features,  Next: The Language,  Prev: Installing SCM,  Up: Top

3 Operational Features
**********************

* Menu:

* Invoking SCM::
* SCM Options::
* Invocation Examples::
* SCM Variables::
* SCM Session::
* Editing Scheme Code::
* Debugging Scheme Code::
* Debugging Continuations::
* Errors::
* Memoized Expressions::
* Internal State::
* Scripting::


File: scm.info,  Node: Invoking SCM,  Next: SCM Options,  Prev: Operational Features,  Up: Operational Features

3.1 Invoking SCM
================

 scm  [-a kbytes] [-muvbiq] [-version] [-help]
     [[-]-no-init-file] [-p int] [-r feature] [-h feature]
     [-d filename] [-f filename] [-l filename]
     [-c expression] [-e expression] [-o dumpname]
     [-- | - | -s] [filename] [arguments ...]

Upon startup `scm' loads the file specified by by the environment
variable SCM_INIT_PATH.

If SCM_INIT_PATH is not defined or if the file it names is not present,
`scm' tries to find the directory containing the executable file.  If
it is able to locate the executable, `scm' looks for the initialization
file (usually `Init5e2.scm') in platform-dependent directories relative       |
to this directory.  See *Note File-System Habitat:: for a blow-by-blow
description.

As a last resort (if initialization file cannot be located), the C
compile parameter IMPLINIT (defined in the makefile or `scmfig.h') is
tried.

Unless the option `-no-init-file' or `--no-init-file' occurs in the
command line, `Init5e2.scm' checks to see if there is file                    |
`ScmInit.scm' in the path specified by the environment variable HOME
(or in the current directory if HOME is undefined).  If it finds such a
file it is loaded.

`Init5e2.scm' then looks for command input from one of three sources:         |
From an option on the command line, from a file named on the command
line, or from standard input.

This explanation applies to SCMLIT or other builds of SCM.

Scheme-code files can also invoke SCM and its variants.  *Note #!:
Lexical Conventions.


File: scm.info,  Node: SCM Options,  Next: Invocation Examples,  Prev: Invoking SCM,  Up: Operational Features

3.2 Options
===========

The options are processed in the order specified on the command line.

 -- Command Option: -a k
     specifies that `scm' should allocate an initial heapsize of K
     kilobytes.  This option, if present, must be the first on the
     command line.  If not specified, the default is `INIT_HEAP_SIZE'
     in source file `setjump.h' which the distribution sets at
     `25000*sizeof(cell)'.

 -- Command Option: -no-init-file
 -- Command Option: --no-init-file
     Inhibits the loading of `ScmInit.scm' as described above.

 -- Command Option: --help
     prints usage information and URI; then exit.

 -- Command Option: --version
     prints version information and exit.

 -- Command Option: -r feature
     requires FEATURE.  This will load a file from [SLIB] if that
     FEATURE is not already provided.  If FEATURE is 2, 2rs, or r2rs;
     3, 3rs, or r3rs; 4, 4rs, or r4rs; 5, 5rs, or r5rs; `scm' will
     require the features neccessary to support [R2RS]; [R3RS]; [R4RS];
     or [R5RS], respectively.

 -- Command Option: -h feature
     provides FEATURE.

 -- Command Option: -l filename
 -- Command Option: -f filename
     loads FILENAME.  `Scm' will load the first (unoptioned) file named
     on the command line if no `-c', `-e', `-f', `-l', or `-s' option
     preceeds it.

 -- Command Option: -d filename
     Loads SLIB `databases' feature and opens FILENAME as a database.

 -- Command Option: -e expression
 -- Command Option: -c expression
     specifies that the scheme expression EXPRESSION is to be
     evaluated.  These options are inspired by `perl' and `sh'
     respectively.  On Amiga systems the entire option and argument
     need to be enclosed in quotes.  For instance `"-e(newline)"'.

 -- Command Option: -o dumpname
     saves the current SCM session as the executable program `dumpname'.
     This option works only in SCM builds supporting `dump' (*note
     Dump::).

     If options appear on the command line after `-o DUMPNAME', then
     the saved session will continue with processing those options when
     it is invoked.  Otherwise the (new) command line is processed as
     usual when the saved image is invoked.

 -- Command Option: -p level
     sets the prolixity (verboseness) to LEVEL.  This is the same as
     the `scm' command (verobse LEVEL).

 -- Command Option: -v
     (verbose mode) specifies that `scm' will print prompts, evaluation
     times, notice of loading files, and garbage collection statistics.
     This is the same as `-p3'.

 -- Command Option: -q
     (quiet mode) specifies that `scm' will print no extra information.
     This is the same as `-p0'.

 -- Command Option: -m
     specifies that subsequent loads, evaluations, and user
     interactions will be with syntax-rules macro capability.  To use a
     specific syntax-rules macro implementation from [SLIB] (instead of
     [SLIB]'s default) put `-r' MACROPACKAGE before `-m' on the command
     line.

 -- Command Option: -u
     specifies that subsequent loads, evaluations, and user
     interactions will be without syntax-rules macro capability.
     Syntax-rules macro capability can be restored by a subsequent `-m'
     on the command line or from Scheme code.

 -- Command Option: -i
     specifies that `scm' should run interactively.  That means that
     `scm' will not terminate until the `(quit)' or `(exit)' command is
     given, even if there are errors.  It also sets the prolixity level
     to 2 if it is less than 2.  This will print prompts, evaluation
     times, and notice of loading files.  The prolixity level can be
     set by subsequent options.  If `scm' is started from a tty, it
     will assume that it should be interactive unless given a
     subsequent `-b' option.

 -- Command Option: -b
     specifies that `scm' should run non-interactively.  That means that
     `scm' will terminate after processing the command line or if there
     are errors.

 -- Command Option: -s
     specifies, by analogy with `sh', that `scm' should run
     interactively and that further options are to be treated as program
     aguments.

 -- Command Option: -
 -- Command Option: --
     specifies that further options are to be treated as program
     aguments.


File: scm.info,  Node: Invocation Examples,  Next: SCM Variables,  Prev: SCM Options,  Up: Operational Features

3.3 Invocation Examples
=======================

`% scm foo.scm'
     Loads and executes the contents of `foo.scm' and then enters
     interactive session.

`% scm -f foo.scm arg1 arg2 arg3'
     Parameters `arg1', `arg2', and `arg3' are stored in the global
     list `*argv*'; Loads and executes the contents of `foo.scm' and
     exits.

`% scm -s foo.scm arg1 arg2'
     Sets *argv* to `("foo.scm" "arg1" "arg2")' and enters interactive
     session.

`% scm -e `(write (list-ref *argv* *optind*))' bar'
     Prints `"bar"'.

`% scm -rpretty-print -r format -i'
     Loads `pretty-print' and `format' and enters interactive session.

`% scm -r5'
     Loads `dynamic-wind', `values', and syntax-rules macros and enters
     interactive (with macros) session.

`% scm -r5 -r4'
     Like above but `rev4-optional-procedures' are also loaded.


File: scm.info,  Node: SCM Variables,  Next: SCM Session,  Prev: Invocation Examples,  Up: Operational Features

3.4 Environment Variables
=========================

 -- Environment Variable: SCM_INIT_PATH
     is the pathname where `scm' will look for its initialization code.
     The default is the file `Init5e2.scm' in the source directory.           |

 -- Environment Variable: SCHEME_LIBRARY_PATH
     is the [SLIB] Scheme library directory.

 -- Environment Variable: HOME
     is the directory where `Init5e2.scm' will look for the user              |
     initialization file `ScmInit.scm'.

 -- Environment Variable: EDITOR
     is the name of the program which `ed' will call.  If EDITOR is not
     defined, the default is `ed'.

3.5 Scheme Variables
====================

 -- Variable: *argv*
     contains the list of arguments to the program.  `*argv*' can change
     during argument processing.  This list is suitable for use as an
     argument to [SLIB] `getopt'.

 -- Variable: *syntax-rules*
     controls whether loading and interaction support syntax-rules
     macros.  Define this in `ScmInit.scm' or files specified on the
     command line.  This can be overridden by subsequent `-m' and `-u'
     options.

 -- Variable: *interactive*
     controls interactivity as explained for the `-i' and `-b' options.
     Define this in `ScmInit.scm' or files specified on the command
     line.  This can be overridden by subsequent `-i' and `-b' options.


File: scm.info,  Node: SCM Session,  Next: Editing Scheme Code,  Prev: SCM Variables,  Up: Operational Features

3.6 SCM Session
===============

   * Options, file loading and features can be specified from the
     command line.  *Note System interface: (scm)System interface.
     *Note Require: (slib)Require.

   * Typing the end-of-file character at the top level session (while
     SCM is not waiting for parenthesis closure) causes SCM to exit.

   * Typing the interrupt character aborts evaluation of the current
     form and resumes the top level read-eval-print loop.

 -- Function: quit
 -- Function: quit n
 -- Function: exit
 -- Function: exit n
     Aliases for `exit' (*note exit: (slib)System.).  On many systems,
     SCM can also tail-call another program.  *Note execp:
     I/O-Extensions.

 -- Callback procedure: boot-tail dumped?
     `boot-tail' is called by `scm_top_level' just before entering
     interactive top-level.  If `boot-tail' calls `quit', then
     interactive top-level is not entered.

 -- Function: program-arguments
     Returns a list of strings of the arguments scm was called with.

 -- Function: getlogin
     Returns the (login) name of the user logged in on the controlling
     terminal of the process, or #f if this information cannot be
     determined.

For documentation of the procedures `getenv' and `system' *Note System
Interface: (slib)System Interface.

 -- Function: vms-debug
     If SCM is compiled under VMS this `vms-debug' will invoke the VMS
     debugger.


File: scm.info,  Node: Editing Scheme Code,  Next: Debugging Scheme Code,  Prev: SCM Session,  Up: Operational Features

3.7 Editing Scheme Code
=======================

 -- Function: ed arg1 ...
     The value of the environment variable `EDITOR' (or just `ed' if it
     isn't defined) is invoked as a command with arguments ARG1 ....

 -- Function: ed filename
     If SCM is compiled under VMS `ed' will invoke the editor with a
     single the single argument FILENAME.

Gnu Emacs:
     Editing of Scheme code is supported by emacs.  Buffers holding
     files ending in .scm are automatically put into scheme-mode.

     If your Emacs can run a process in a buffer you can use the Emacs
     command `M-x run-scheme' with SCM.  Otherwise, use the emacs
     command `M-x suspend-emacs'; or see "other systems" below.

Epsilon (MS-DOS):
     There is lisp (and scheme) mode available by use of the package
     `LISP.E'.  It offers several different indentation formats.  With
     this package, buffers holding files ending in `.L', `.LSP', `.S',
     and `.SCM' (my modification) are automatically put into lisp-mode.

     It is possible to run a process in a buffer under Epsilon.  With
     Epsilon 5.0 the command line options `-e512 -m0' are neccessary to
     manage RAM properly.  It has been reported that when compiling SCM
     with Turbo C, you need to `#define NOSETBUF' for proper operation
     in a process buffer with Epsilon 5.0.

     One can also call out to an editor from SCM if RAM is at a
     premium; See "under other systems" below.

other systems:
     Define the environment variable `EDITOR' to be the name of the
     editing program you use.  The SCM procedure `(ed arg1 ...)' will
     invoke your editor and return to SCM when you exit the editor.  The
     following definition is convenient:

          (define (e) (ed "work.scm") (load "work.scm"))

     Typing `(e)' will invoke the editor with the file of interest.
     After editing, the modified file will be loaded.


File: scm.info,  Node: Debugging Scheme Code,  Next: Debugging Continuations,  Prev: Editing Scheme Code,  Up: Operational Features

3.8 Debugging Scheme Code
=========================

The `cautious' and `stack-limit' options of `build' (*note Build
Options::) support debugging in Scheme.

"CAUTIOUS"
     If SCM is built with the `CAUTIOUS' flag, then when an error
     occurs, a "stack trace" of certain pending calls are printed as
     part of the default error response.  A (memoized) expression and
     newline are printed for each partially evaluated combination whose
     procedure is not builtin.  See *Note Memoized Expressions:: for
     how to read memoized expressions.

     Also as the result of the `CAUTIOUS' flag, both `error' and
     `user-interrupt' (invoked by <C-c>) to print stack traces and
     conclude by calling `breakpoint' (*note Breakpoints:
     (slib)Breakpoints.) instead of aborting to top level.  Under
     either condition, program execution can be resumed by `(continue)'.

     In this configuration one can interrupt a running Scheme program
     with <C-c>, inspect or modify top-level values, trace or untrace
     procedures, and continue execution with `(continue)'.

"STACK_LIMIT"
     If SCM is built with the `STACK_LIMIT' flag, the interpreter will
     check stack size periodically.  If the size of stack exceeds a
     certain amount (default is `HEAP_SEG_SIZE/2'), SCM generates a
     `segment violation' interrupt.

     The usefulness of `STACK_LIMIT' depends on the user.  I don't use
     it; but the user I added this feature for got primarily this type
     of error.

There are several SLIB macros which so useful that SCM automatically
loads the appropriate module from SLIB if they are invoked.

 -- Macro: trace proc1 ...
     Traces the top-level named procedures given as arguments.

 -- Macro: trace
     With no arguments, makes sure that all the currently traced
     identifiers are traced (even if those identifiers have been
     redefined) and returns a list of the traced identifiers.

 -- Macro: untrace proc1 ...
     Turns tracing off for its arguments.

 -- Macro: untrace
     With no arguments, untraces all currently traced identifiers and
     returns a list of these formerly traced identifiers.

The routines I use most frequently for debugging are:

 -- Function: print arg1 ...
     `Print' writes all its arguments, separated by spaces.  `Print'
     outputs a `newline' at the end and returns the value of the last
     argument.

     One can just insert `(print '<label>' and `)' around an expression
     in order to see its values as a program operates.

 -- Function: pprint arg1 ...
     `Pprint' pretty-prints (*note Pretty-Print: (slib)Pretty-Print.)
     all its arguments, separated by newlines.  `Pprint' returns the
     value of the last argument.

     One can just insert `(pprint '<label>' and `)' around an
     expression in order to see its values as a program operates.
     _Note_ `pretty-print' does _not_ format procedures.

When typing at top level, `pprint' is not a good way to see nested
structure because it will return the last object pretty-printed, which
could be large.  `pp' is a better choice.

 -- Procedure: pp arg1 ...
     `Pprint' pretty-prints (*note Pretty-Print: (slib)Pretty-Print.)
     all its arguments, separated by newlines.  `pp' returns
     `#<unspecified>'.

 -- Syntax: print-args name
 -- Syntax: print-args
     Writes NAME if supplied; then writes the names and values of the
     closest lexical bindings enclosing the call to `Print-args'.

          (define (foo a b) (print-args foo) (+ a b))
          (foo 3 6)
          -| In foo: a = 3; b = 6;
          => 9

Sometimes more elaborate measures are needed to print values in a useful
manner.  When the values to be printed may have very large (or infinite)
external representations, *Note Quick Print: (slib)Quick Print, can be
used.

When `trace' is not sufficient to find program flow problems, SLIB-PSD,
the Portable Scheme Debugger offers source code debugging from GNU
Emacs.  PSD runs slowly, so start by instrumenting only a few functions
at a time.
     http://swiss.csail.mit.edu/ftpdir/scm/slib-psd1-3.tar.gz
     swiss.csail.mit.edu:/pub/scm/slib-psd1-3.tar.gz
     ftp.maths.tcd.ie:pub/bosullvn/jacal/slib-psd1-3.tar.gz
     ftp.cs.indiana.edu:/pub/scheme-repository/utl/slib-psd1-3.tar.gz


File: scm.info,  Node: Debugging Continuations,  Next: Errors,  Prev: Debugging Scheme Code,  Up: Operational Features

3.9 Debugging Continuations
===========================

These functions are defined in `debug.c', all operate on captured
continuations:

 -- Procedure: frame-trace cont n
     Prints information about the code being executed and the
     environment scopes active for continuation frame N of continuation
     CONT.  A "continuation frame" is an entry in the environment
     stack; a new frame is pushed when the environment is replaced or
     extended in a non-tail call context.  Frame 0 is the top of the
     stack.

 -- Procedure: frame->environment cont n
     Prints the environment for continuation frame N of continuation
     CONT.  This contains just the names, not the values, of the
     environment.

 -- Procedure: scope-trace env
     will print information about active lexical scopes for environment
     ENV.

 -- Procedure: frame-eval cont n expr
     Evaluates EXPR in the environment defined by continuation frame N
     of continuation CONT and returns the result.  Values in the
     environment may be returned or SET!.

*Note stack-trace: Errors. also now accepts an optional continuation
argument.  `stack-trace' differs from `frame-trace' in that it
truncates long output using safeports and prints code from all
available frames.

     (define k #f)
     (define (foo x y)
        (set! k (call-with-current-continuation identity))
        #f)
     (let ((a 3) (b 4))
       (foo a b)
       #f)
     (stack-trace k)
     -|
     ;STACK TRACE
     1; ((#@set! #@k (#@call-with-current-continuation #@identity)) #f ...
     2; (#@let ((a 3) (b 4))  (#@foo #@a #@b) #f)
     ...
     #t

     (frame-trace k 0)
     -|
     (#@call-with-current-continuation #@identity)
     ; in scope:
     ;   (x y)  procedure foo#<unspecified>

     (frame-trace k 1)
     -|
     ((#@set! #@k (#@call-with-current-continuation #@identity)) #f)
     ; in scope:
     ;   (x y)  procedure foo#<unspecified>

     (frame-trace k 2)
     -|
     (#@let ((a 3) (b 4))  (#@foo #@a #@b) #f)
     ; in scope:
     ;   (a b . #@let)#<unspecified>

     (frame-trace k 3)
     -|
     (#@let ((a 3) (b 4))  (#@foo #@a #@b) #f)
     ; in top level environment.

     (frame->environment k 0)
     -|
     ((x y) 2 foo)

     (scope-trace (frame->environment k 0))
     -|
     ; in scope:
     ;   (x y)  procedure foo#<unspecified>

     (frame-eval k 0 'x) => 3

     (frame-eval k 0 '(set! x 8))
     (frame-eval k 0 'x) => 8


File: scm.info,  Node: Errors,  Next: Memoized Expressions,  Prev: Debugging Continuations,  Up: Operational Features

3.10 Errors
===========

A computer-language implementation designer faces choices of how
reflexive to make the implementation in handling exceptions and errors;
that is, how much of the error and exception routines should be written
in the language itself.  The design of a portable implementation is
further constrained by the need to have (almost) all errors print
meaningful messages, even when the implementation itself is not
functioning correctly.  Therefore, SCM implements much of its error
response code in C.

The following common error and conditions are handled by C code.  Those
with callback names after them can also be handled by Scheme code
(*note Interrupts::).  If the callback identifier is not defined at top
level, the default error handler (C code) is invoked.  There are many
other error messages which are not treated specially.

"ARGn"
     Wrong type in argument

"ARG1"
     Wrong type in argument 1

"ARG2"
     Wrong type in argument 2

"ARG3"
     Wrong type in argument 3

"ARG4"
     Wrong type in argument 4

"ARG5"
     Wrong type in argument 5

"WNA"
     Wrong number of args

"OVFLOW"
     numerical overflow

"OUTOFRANGE"
     Argument out of range

"NALLOC"
     `(out-of-storage)'

"THRASH"
     GC is `(thrashing)'

"EXIT"
     `(end-of-program)'

"HUP_SIGNAL"
     `(hang-up)'

"INT_SIGNAL"
     `(user-interrupt)'

"FPE_SIGNAL"
     `(arithmetic-error)'

"BUS_SIGNAL"
     bus error

"SEGV_SIGNAL"
     segment violation

"ALRM_SIGNAL"
     `(alarm-interrupt)'

"VTALRM_SIGNAL"
     `(virtual-alarm-interrupt)'

"PROF_SIGNAL"
     `(profile-alarm-interrupt)'

 -- Variable: errobj
     When SCM encounters a non-fatal error, it aborts evaluation of the
     current form, prints a message explaining the error, and resumes
     the top level read-eval-print loop.  The value of ERROBJ is the
     offending object if appropriate.  The builtin procedure `error'
     does _not_ set ERROBJ.

`errno' and `perror' report ANSI C errors encountered during a call to
a system or library function.

 -- Function: errno
 -- Function: errno n
     With no argument returns the current value of the system variable
     `errno'.  When given an argument, `errno' sets the system variable
     `errno' to N and returns the previous value of `errno'.  `(errno
     0)' will clear outstanding errors.  This is recommended after
     `try-load' returns `#f' since this occurs when the file could not
     be opened.

 -- Function: perror string
     Prints on standard error output the argument STRING, a colon,
     followed by a space, the error message corresponding to the current
     value of `errno' and a newline.  The value returned is unspecified.

`warn' and `error' provide a uniform way for Scheme code to signal
warnings and errors.

 -- Function: warn arg1 arg2 arg3 ...
     Alias for *Note slib:warn: (slib)System.  Outputs an error message
     containing the arguments.  `warn' is defined in `Init5e2.scm'.           |

 -- Function: error arg1 arg2 arg3 ...
     Alias for *Note slib:error: (slib)System.  Outputs an error
     message containing the arguments, aborts evaluation of the current
     form and resumes the top level read-eval-print loop.  `Error' is
     defined in `Init5e2.scm'.                                                |

If SCM is built with the `CAUTIOUS' flag, then when an error occurs, a
"stack trace" of certain pending calls are printed as part of the
default error response.  A (memoized) expression and newline are
printed for each partially evaluated combination whose procedure is not
builtin.  See *Note Memoized Expressions:: for how to read memoized
expressions.

Also as the result of the `CAUTIOUS' flag, both `error' and
`user-interrupt' (invoked by <C-c>) are defined to print stack traces
and conclude by calling `breakpoint' (*note Breakpoints:
(slib)Breakpoints.).  This allows the user to interract with SCM as
with Lisp systems.

 -- Function: stack-trace
     Prints information describing the stack of partially evaluated
     expressions.  `stack-trace' returns `#t' if any lines were printed
     and `#f' otherwise.  See `Init5e2.scm' for an example of the use         |
     of `stack-trace'.


File: scm.info,  Node: Memoized Expressions,  Next: Internal State,  Prev: Errors,  Up: Operational Features

3.11 Memoized Expressions
=========================

SCM memoizes the address of each occurence of an identifier's value when
first encountering it in a source expression.  Subsequent executions of
that memoized expression is faster because the memoized reference
encodes where in the top-level or local environment its value is.

When procedures are displayed, the memoized locations appear in a format
different from references which have not yet been executed.  I find this
a convenient aid to locating bugs and untested expressions.

   * The names of memoized lexically bound identifiers are replaced with
     #@<m>-<n>, where <m> is the number of binding contours back and
     <n> is the index of the value in that binding countour.

   * The names of identifiers which are not lexiallly bound but defined
     at top-level have #@ prepended.

For instance, `open-input-file' is defined as follows in `Init5e2.scm':       |

     (define (open-input-file str)
       (or (open-file str OPEN_READ)
           (and (procedure? could-not-open) (could-not-open) #f)
           (error "OPEN-INPUT-FILE couldn't open file " str)))

If `open-input-file' has not yet been used, the displayed procedure is
similar to the original definition (lines wrapped for readability):

     open-input-file =>
     #<CLOSURE (str) (or (open-file str open_read)
      (and (procedure? could-not-open) (could-not-open) #f)
      (error "OPEN-INPUT-FILE couldn't open file " str))>

If we open a file using `open-input-file', the sections of code used
become memoized:

     (open-input-file "r4rstest.scm") => #<input-port 3>
     open-input-file =>
     #<CLOSURE (str) (#@or (#@open-file #@0+0 #@open_read)
      (and (procedure? could-not-open) (could-not-open) #f)
      (error "OPEN-INPUT-FILE couldn't open file " str))>

If we cause `open-input-file' to execute other sections of code, they
too become memoized:

     (open-input-file "foo.scm") =>

     ERROR: No such file or directory
     ERROR: OPEN-INPUT-FILE couldn't open file  "foo.scm"

     open-input-file =>
     #<CLOSURE (str) (#@or (#@open-file #@0+0 #@open_read)
      (#@and (#@procedure? #@could-not-open) (could-not-open) #f)
      (#@error "OPEN-INPUT-FILE couldn't open file " #@0+0))>


File: scm.info,  Node: Internal State,  Next: Scripting,  Prev: Memoized Expressions,  Up: Operational Features

3.12 Internal State
===================

 -- Variable: *interactive*
     The variable *INTERACTIVE* determines whether the SCM session is
     interactive, or should quit after the command line is processed.
     *INTERACTIVE* is controlled directly by the command-line options
     `-b', `-i', and `-s' (*note Invoking SCM::).  If none of these
     options are specified, the rules to determine interactivity are
     more complicated; see `Init5e2.scm' for details.                         |

 -- Function: abort
     Resumes the top level Read-Eval-Print loop.

 -- Function: restart
     Restarts the SCM program with the same arguments as it was
     originally invoked.  All `-l' loaded files are loaded again; If
     those files have changed, those changes will be reflected in the
     new session.

     _Note_ When running a saved executable (*note Dump::), `restart'
     is redefined to be `exec-self'.

 -- Function: exec-self
     Exits and immediately re-invokes the same executable with the same
     arguments.  If the executable file has been changed or replaced
     since the beginning of the current session, the _new_ executable
     will be invoked.  This differentiates `exec-self' from `restart'.

 -- Function: verbose n
     Controls how much monitoring information is printed.  If N is:

    0
          no prompt or information is printed.

    >= 1
          a prompt is printed.

    >= 2
          messages bracketing file loading are printed.

    >= 3
          the CPU time is printed after each top level form evaluated;
          notifications of heap growth printed.

    >= 4
          a garbage collection summary is printed after each top level
          form evaluated;

    >= 5
          a message for each GC (*note Garbage Collection::) is printed;
          warnings issued for top-level symbols redefined.

 -- Function: gc
     Scans all of SCM objects and reclaims for further use those that
     are no longer accessible.

 -- Function: room
 -- Function: room #t
     Prints out statistics about SCM's current use of storage.  `(room
     #t)' also gives the hexadecimal heap segment and stack bounds.

 -- Constant: *scm-version*
     Contains the version string (e.g. `5e2') of SCM.                         |

3.12.1 Executable path
----------------------

In order to dump a saved executable or to dynamically-link using DLD,
SCM must know where its executable file is.  Sometimes SCM (*note
Executable Pathname::) guesses incorrectly the location of the
currently running executable.  In that case, the correct path can be set
by calling `execpath' with the pathname.

 -- Function: execpath
     Returns the path (string) which SCM uses to find the executable
     file whose invocation the currently running session is, or #f if
     the path is not set.

 -- Function: execpath #f
 -- Function: execpath newpath
     Sets the path to `#f' or NEWPATH, respectively.  The old path is
     returned.

For other configuration constants and procedures *Note Configuration:
(slib)Configuration.


File: scm.info,  Node: Scripting,  Prev: Internal State,  Up: Operational Features

3.13 Scripting
==============

* Menu:

* Unix Scheme Scripts::         From Olin Shivers' Scheme Shell
* MS-DOS Compatible Scripts::   Run in MS-DOS and Unix
* Unix Shell Scripts::          Use /bin/sh to run Scheme


File: scm.info,  Node: Unix Scheme Scripts,  Next: MS-DOS Compatible Scripts,  Prev: Scripting,  Up: Scripting

3.13.1 Unix Scheme Scripts
--------------------------

In reading this section, keep in mind that the first line of a script
file has (different) meanings to SCM and the operating system
(`execve').

 -- file: #! interpreter \ ...
     On unix systems, a "Shell-Script" is a file (with execute
     permissions) whose first two characters are `#!'.  The INTERPRETER
     argument must be the pathname of the program to process the rest
     of the file.  The directories named by environment variable `PATH'
     are _not_ searched to find INTERPRETER.

     When executing a shell-script, the operating system invokes
     INTERPRETER with a single argument encapsulating the rest of the
     first line's contents (if not just whitespace), the pathname of the
     Scheme Script file, and then any arguments which the shell-script
     was invoked with.

     Put one space character between `#!' and the first character of
     INTERPRETER (`/').  The INTERPRETER name is followed by ` \'; SCM
     substitutes the second line of FILE for `\' (and the rest of the
     line), then appends any arguments given on the command line
     invoking this Scheme-Script.

     When SCM executes the script, the Scheme variable *SCRIPT* will be
     set to the script pathname.  The last argument before `!#' on the
     second line should be `-'; SCM will load the script file, preserve
     the unprocessed arguments, and set *ARGV* to a list of the script
     pathname and the unprocessed arguments.

     Note that the interpreter, not the operating system, provides the
     `\' substitution; this will only take place if INTERPRETER is a
     SCM or SCSH interpreter.

 -- Read syntax: #! ignored !#
     When the first two characters of the file being loaded are `#!' and
     a `\' is present before a newline in the file, all characters up
     to `!#' will be ignored by SCM `read'.

This combination of interpretatons allows SCM source files to be used as
POSIX shell-scripts if the first line is:

     #! /usr/local/bin/scm \

The following Scheme-Script prints factorial of its argument:

     #! /usr/local/bin/scm \ %0 %*
     - !#

     (define (fact.script args)
       (cond ((and (= 1 (length args))
                   (string->number (car args)))
              => (lambda (n) (print (fact n)) #t))
             (else (fact.usage))))

     (define (fact.usage)
       (print *argv*)
       (display "\
     Usage: fact N
       Returns the factorial of N.
     "
                (current-error-port))
       #f)

     (define (fact n) (if (< n 2) 1 (* n (fact (+ -1 n)))))

     (if *script* (exit (fact.script (list-tail *argv* *optind*))))

     ./fact 32
     =>
     263130836933693530167218012160000000

If the wrong number of arguments is given, `fact' prints its ARGV with
usage information.

     ./fact 3 2
     -|
     ("./fact" "3" "2")
     Usage: fact N
       Returns the factorial of N.


File: scm.info,  Node: MS-DOS Compatible Scripts,  Next: Unix Shell Scripts,  Prev: Unix Scheme Scripts,  Up: Scripting

3.13.2 MS-DOS Compatible Scripts
--------------------------------

It turns out that we can create scheme-scripts which run both under unix
and MS-DOS.  To implement this, I have written the MS-DOS programs:
`#!.bat' and `!#.exe', which are available from:
`http://swiss.csail.mit.edu/ftpdir/scm/sharpbang.zip'

With these two programs installed in a `PATH' directory, we have the
following syntax for <PROGRAM>.BAT files.

 -- file: #! interpreter \ %0 %*
     The first two characters of the Scheme-Script are `#!'.  The
     INTERPRETER can be either a unix style program path (using `/'
     between filename components) or a DOS program name or path.  The
     rest of the first line of the Scheme-Script should be literally
     `\ %0 %*', as shown.

     If INTERPRETER has `/' in it, INTERPRETER is converted to a DOS
     style filename (`/' => `\').

     In looking for an executable named INTERPRETER, `#!' first checks
     this (converted) filename; if INTERPRETER doesn't exist, it then
     tries to find a program named like the string starting after the
     last `\' (or `/') in INTERPRETER.  When searching for executables,
     `#!' tries all directories named by environment variable `PATH'.

     Once the INTERPRETER executable path is found, arguments are
     processed in the manner of scheme-shell, with all the text after
     the `\' taken as part of the meta-argument.  More precisely, `#!'
     calls INTERPRETER with any options on the second line of the
     Scheme-Script up to `!#', the name of the Scheme-Script file, and
     then any of at most 8 arguments given on the command line invoking
     this Scheme-Script.

The previous example Scheme-Script works in both MS-DOS and unix
systems.


File: scm.info,  Node: Unix Shell Scripts,  Prev: MS-DOS Compatible Scripts,  Up: Scripting

3.13.3 Unix Shell Scripts
-------------------------

Scheme-scripts suffer from two drawbacks:
   * Some Unixes limit the length of the `#!' interpreter line to the
     size of an object file header, which can be as small as 32 bytes.

   * A full, explicit pathname must be specified, perhaps requiring
     more than 32 bytes and making scripts vulnerable to breakage when
     programs are moved.

The following approach solves these problems at the expense of slower
startup.  Make `#! /bin/sh' the first line and prepend every subsequent
line to be executed by the shell with `:;'.  The last line to be
executed by the shell should contain an "exec" command; `exec'
tail-calls its argument.

`/bin/sh' is thus invoked with the name of the script file, which it
executes as a *sh script.  Usually the second line starts `:;exec scm
-f$0', which executes scm, which in turn loads the script file.  When
SCM loads the script file, it ignores the first and second lines, and
evaluates the rest of the file as Scheme source code.

The second line of the script file does not have the length restriction
mentioned above.  Also, `/bin/sh' searches the directories listed in
the `PATH' environment variable for `scm', eliminating the need to use
absolute locations in order to invoke a program.

The following example additionally sets *SCRIPT* to the script
argument, making it compatible with the scheme code of the previous
example.

     #! /bin/sh
     :;exec scm -e"(set! *script* \"$0\")" -l$0 "$@"

     (define (fact.script args)
       (cond ((and (= 1 (length args))
                   (string->number (car args)))
              => (lambda (n) (print (fact n)) #t))
             (else (fact.usage))))

     (define (fact.usage)
       (print *argv*)
       (display "\
     Usage: fact N
       Returns the factorial of N.
     "
                (current-error-port))
       #f)

     (define (fact n) (if (< n 2) 1 (* n (fact (+ -1 n)))))

     (if *script* (exit (fact.script (list-tail *argv* *optind*))))

     ./fact 6
     => 720


File: scm.info,  Node: The Language,  Next: Packages,  Prev: Operational Features,  Up: Top

4 The Language
**************

* Menu:

* Standards Compliance::        Links to sections in [R5RS] and [SLIB]
* Storage::                     Finalizers, GC-hook, vector-set-length!
* Time::                        Both real time and processor time
* Interrupts::                  and exceptions
* Process Synchronization::     Because interrupts are preemptive
* Files and Ports::
* Eval and Load::               and line-numbers
* Lexical Conventions::         Also called read-syntax
* Syntax::                      Macros


File: scm.info,  Node: Standards Compliance,  Next: Storage,  Prev: The Language,  Up: The Language

4.1 Standards Compliance
========================

Scm conforms to the `IEEE Standard 1178-1990.  IEEE Standard for the
Scheme Programming Language.'  (*note Bibliography::), and `Revised(5)
Report on the Algorithmic Language Scheme'.  *Note Top: (r5rs)Top.  All
the required features of these specifications are supported.  Many of
the optional features are supported as well.

Optionals of [R5RS] Supported by SCM
------------------------------------

`-' and `/' of more than 2 arguments
`exp'
`log'
`sin'
`cos'
`tan'
`asin'
`acos'
`atan'
`sqrt'
`expt'
`make-rectangular'
`make-polar'
`real-part'
`imag-part'
`magnitude'
`angle'
`exact->inexact'
`inexact->exact'
     *Note Numerical operations: (r5rs)Numerical operations.

`with-input-from-file'
`with-output-to-file'
     *Note Ports: (r5rs)Ports.

`load'
`transcript-on'
`transcript-off'
     *Note System interface: (r5rs)System interface.

Optionals of [R5RS] not Supported by SCM
----------------------------------------

`numerator'
`denominator'
`rationalize'
     *Note Numerical operations: (r5rs)Numerical operations.

[SLIB] Features of SCM and SCMLIT
---------------------------------

`delay'
`full-continuation'
`ieee-p1178'
`object-hash'
`rev4-report'
`source'
     See SLIB file `Template.scm'.

`current-time'
     *Note Time and Date: (slib)Time and Date.

`defmacro'
     *Note Defmacro: (slib)Defmacro.

`getenv'
`system'
     *Note System Interface: (slib)System Interface.

`hash'
     *Note Hashing: (slib)Hashing.

`logical'
     *Note Bit-Twiddling: (slib)Bit-Twiddling.

`multiarg-apply'
     *Note Multi-argument Apply: (slib)Multi-argument Apply.

`multiarg/and-'
     *Note Multi-argument / and -: (slib)Multi-argument / and -.

`rev4-optional-procedures'
     *Note Rev4 Optional Procedures: (slib)Rev4 Optional Procedures.

`string-port'
     *Note String Ports: (slib)String Ports.

`tmpnam'
     *Note Input/Output: (slib)Input/Output.

`transcript'
     *Note Transcripts: (slib)Transcripts.

`vicinity'
     *Note Vicinity: (slib)Vicinity.

`with-file'
     *Note With-File: (slib)With-File.

[SLIB] Features of SCM
----------------------

`array'
     *Note Arrays: (slib)Arrays.

`array-for-each'
     *Note Array Mapping: (slib)Array Mapping.

`bignum'
`complex'
`inexact'
`rational'
`real'
     *Note Require: (slib)Require.


File: scm.info,  Node: Storage,  Next: Time,  Prev: Standards Compliance,  Up: The Language

4.2 Storage
===========

 -- Function: vector-set-length! object length
     Change the length of string, vector, bit-vector, or uniform-array
     OBJECT to LENGTH.  If this shortens OBJECT then the remaining
     contents are lost.  If it enlarges OBJECT then the contents of the
     extended part are undefined but the original part is unchanged.
     It is an error to change the length of literal datums.  The new
     object is returned.

 -- Function: copy-tree obj
 -- Function: @copy-tree obj
     *Note copy-tree: (slib)Tree Operations.  This extends the SLIB
     version by also copying vectors.  Use `@copy-tree' if you depend
     on this feature; `copy-tree' could get redefined.

 -- Function: acons obj1 obj2 obj3
     Returns (cons (cons obj1 obj2) obj3).

          (set! a-list (acons key datum a-list))

     Adds a new association to a-list.

 -- Callback procedure: gc-hook ...
     Allows a Scheme procedure to be run shortly after each garbage
     collection.  This procedure will not be run recursively.  If it
     runs long enough to cause a garbage collection before returning a
     warning will be printed.

     To remove the gc-hook, `(set! gc-hook #f)'.

 -- Function: add-finalizer object finalizer
     OBJECT may be any garbage collected object, that is, any object
     other than an immediate integer, character, or special token such
     as `#f' or `#t', *Note Immediates::.  FINALIZER is a thunk, or
     procedure taking no arguments.

     FINALIZER will be invoked asynchronously exactly once some time
     after OBJECT becomes eligible for garbage collection.  A reference
     to OBJECT in the environment of FINALIZER will not prevent
     finalization, but will delay the reclamation of OBJECT at least
     until the next garbage collection.  A reference to OBJECT in some
     other object's finalizer will necessarily prevent finalization
     until both objects are eligible for garbage collection.

     Finalizers are not run in any predictable order.  All finalizers
     will be run by the time the program ends.

     This facility was based on the paper by Simon Peyton Jones, et al,
     "Stretching the storage manager: weak pointers and stable names in
     Haskell", Proc. 11th International Workshop on the Implementation
     of Functional Languages, The Netherlands, September 7-10 1999,
     Springer-Verlag LNCS.


File: scm.info,  Node: Time,  Next: Interrupts,  Prev: Storage,  Up: The Language

4.3 Time
========

 -- Constant: internal-time-units-per-second
     Is the integer number of internal time units in a second.

 -- Function: get-internal-run-time
     Returns the integer run time in internal time units from an
     unspecified starting time.  The difference of two calls to
     `get-internal-run-time' divided by
     `internal-time-units-per-second' will give elapsed run time in
     seconds.

 -- Function: get-internal-real-time
     Returns the integer time in internal time units from an unspecified
     starting time.  The difference of two calls to
     `get-internal-real-time' divided by
     `interal-time-units-per-second' will give elapsed real time in
     seconds.

 -- Function: current-time
     Returns the time since 00:00:00 GMT, January 1, 1970, measured in
     seconds.  *Note current-time: (slib)Time and Date.  `current-time'
     is used in *Note Time and Date: (slib)Time and Date.


File: scm.info,  Node: Interrupts,  Next: Process Synchronization,  Prev: Time,  Up: The Language

4.4 Interrupts
==============

 -- Function: ticks n
     Returns the number of ticks remaining till the next tick interrupt.
     Ticks are an arbitrary unit of evaluation.  Ticks can vary greatly
     in the amount of time they represent.

     If N is 0, any ticks request is canceled.  Otherwise a
     `ticks-interrupt' will be signaled N from the current time.
     `ticks' is supported if SCM is compiled with the `ticks' flag
     defined.

 -- Callback procedure: ticks-interrupt ...
     Establishes a response for tick interrupts.  Another tick
     interrupt will not occur unless `ticks' is called again.  Program
     execution will resume if the handler returns.  This procedure
     should (abort) or some other action which does not return if it
     does not want processing to continue.

 -- Function: alarm secs
     Returns the number of seconds remaining till the next alarm
     interrupt.  If SECS is 0, any alarm request is canceled.
     Otherwise an `alarm-interrupt' will be signaled SECS from the
     current time.  ALARM is not supported on all systems.

 -- Function: milli-alarm millisecs interval
 -- Function: virtual-alarm millisecs interval
 -- Function: profile-alarm millisecs interval
     `milli-alarm' is similar to `alarm', except that the first
     argument MILLISECS, and the return value are measured in
     milliseconds rather than seconds.  If the optional argument
     INTERVAL is supplied then alarm interrupts will be scheduled every
     INTERVAL milliseconds until turned off by a call to `milli-alarm'
     or `alarm'.

     `virtual-alarm' and `profile-alarm' are similar.  `virtual-alarm'
     decrements process execution time rather than real time, and
     causes `SIGVTALRM' to be signaled.  `profile-alarm' decrements
     both process execution time  and system execution time on behalf
     of the process, and causes `SIGPROF' to be signaled.

     `milli-alarm', `virtual-alarm', and `profile-alarm' are supported
     only on systems providing the `setitimer' system call.

 -- Callback procedure: user-interrupt ...
 -- Callback procedure: alarm-interrupt ...
 -- Callback procedure: virtual-alarm-interrupt ...
 -- Callback procedure: profile-alarm-interrupt ...
     Establishes a response for `SIGINT' (control-C interrupt) and
     `SIGALRM', `SIGVTALRM', and `SIGPROF' interrupts.  Program
     execution will resume if the handler returns.  This procedure
     should `(abort)' or some other action which does not return if it
     does not want processing to continue after it returns.

     Interrupt handlers are disabled during execution `system' and `ed'
     procedures.

     To unestablish a response for an interrupt set the handler symbol
     to `#f'.  For instance, `(set! user-interrupt #f)'.

 -- Callback procedure: out-of-storage ...
 -- Callback procedure: could-not-open ...
 -- Callback procedure: end-of-program ...
 -- Callback procedure: hang-up ...
 -- Callback procedure: arithmetic-error ...
     Establishes a response for storage allocation error, file opening
     error, end of program, SIGHUP (hang up interrupt) and arithmetic
     errors respectively.  This procedure should (abort) or some other
     action which does not return if it does not want the default error
     message to also be displayed.  If no procedure is defined for
     HANG-UP then END-OF-PROGRAM (if defined) will be called.

     To unestablish a response for an error set the handler symbol to
     `#f'.  For instance, `(set! could-not-open #f)'.


File: scm.info,  Node: Process Synchronization,  Next: Files and Ports,  Prev: Interrupts,  Up: The Language

4.5 Process Synchronization
===========================

An "exchanger" is a procedure of one argument regulating mutually exclusive
access to a resource.  When a exchanger is called, its current content
is returned, while being replaced by its argument in an atomic
operation.

 -- Function: make-exchanger obj
     Returns a new exchanger with the argument OBJ as its initial
     content.

          (define queue (make-exchanger (list a)))

     A queue implemented as an exchanger holding a list can be
     protected from reentrant execution thus:

          (define (pop queue)
            (let ((lst #f))
              (dynamic-wind
                  (lambda () (set! lst (queue #f)))
                  (lambda () (and lst (not (null? lst))
                                  (let ((ret (car lst)))
                                    (set! lst (cdr lst))
                                    ret)))
                  (lambda () (and lst (queue lst))))))

          (pop queue)         => a

          (pop queue)         => #f

 -- Function: make-arbiter name
     Returns an object of type arbiter and name NAME.  Its state is
     initially unlocked.

 -- Function: try-arbiter arbiter
     Returns `#t' and locks ARBITER if ARBITER was unlocked.
     Otherwise, returns `#f'.

 -- Function: release-arbiter arbiter
     Returns `#t' and unlocks ARBITER if ARBITER was locked.
     Otherwise, returns `#f'.


File: scm.info,  Node: Files and Ports,  Next: Eval and Load,  Prev: Process Synchronization,  Up: The Language

4.6 Files and Ports
===================

These procedures generalize and extend the standard capabilities in
*Note Ports: (r5rs)Ports.

* Menu:

* Opening and Closing::
* Port Properties::
* Port Redirection::
* Soft Ports::


File: scm.info,  Node: Opening and Closing,  Next: Port Properties,  Prev: Files and Ports,  Up: Files and Ports

4.6.1 Opening and Closing
-------------------------

 -- Function: open-file string modes
 -- Function: try-open-file string modes
     Returns a port capable of receiving or delivering characters as
     specified by the MODES string.  If a file cannot be opened `#f' is
     returned.

     Internal functions opening files "callback" to the SCM function
     `open-file'.  You can extend `open-file' by redefining it.
     `try-open-file' is the primitive procedure; Do not redefine
     `try-open-file'!

 -- Constant: open_read
 -- Constant: open_write
 -- Constant: open_both
     Contain modes strings specifying that a file is to be opened for
     reading, writing, and both reading and writing respectively.

     Both input and output functions can be used with io-ports.  An end
     of file must be read or a file-set-position done on the port
     between a read operation and a write operation or vice-versa.

 -- Function: _ionbf modestr
     Returns a version of MODESTR which when `open-file' is called with
     it as the second argument will return an unbuffered port.  An
     input-port must be unbuffered in order for `char-ready?' and
     `wait-for-input' to work correctly on it.  The initial value of
     `(current-input-port)' is unbuffered if the platform supports it.

 -- Function: _tracked modestr
     Returns a version of MODESTR which when `open-file' is called with
     it as the second argument will return a tracked port.  A tracked
     port maintains current line and column numbers, which may be
     queried with `port-line' and `port-column'.

 -- Function: _exclusive modestr
     Returns a version of MODESTR which when `open-file' is called with
     it as the second argument will return a port only if the named file
     does not already exist.  This functionality is provided by calling
     `try-create-file' *Note I/O-Extensions::, which is not available
     for all platforms.

 -- Function: open-ports
     Returns a list of all currently open ports, excluding string ports,
     see *Note String Ports: (slib)String Ports.  This may be useful
     after a fork *Note Posix Extensions::, or for debugging.  Bear in
     mind that ports that would be closed by gc will be kept open by a
     reference to this list.

 -- Function: close-port port
     Closes PORT.  The same as close-input-port and close-output-port.


File: scm.info,  Node: Port Properties,  Next: Port Redirection,  Prev: Opening and Closing,  Up: Files and Ports

4.6.2 Port Properties
---------------------

 -- Function: port-closed? port
     Returns #t if PORT is closed.

 -- Function: port-type obj
     If OBJ is not a port returns false, otherwise returns a symbol
     describing the port type, for example string or pipe.

 -- Function: port-filename port
     Returns the filename PORT was opened with.  If PORT is not open to
     a file the result is unspecified.

 -- Function: port-line port
 -- Function: port-column port
     If PORT is a tracked port, return the current line (column) number,
     otherwise return `#f'.  Line and column numbers begin with 1.  The
     column number applies to the next character to be read; if that
     character is a newline, then the column number will be one more
     than the length of the line.

 -- Function: freshline port
     Outputs a newline to optional argument PORT unless the current
     output column number of PORT is known to be zero, ie output will
     start at the beginning of a new line.  PORT defaults to
     `current-output-port'.  If PORT is not a tracked port `freshline'
     is equivalent to `newline'.

 -- Function: isatty? port
     Returns `#t' if PORT is input or output to a serial non-file
     device.

 -- procedure: char-ready?
 -- procedure: char-ready? port
     Returns `#t' if a character is ready on the input PORT and returns
     `#f' otherwise.  If `char-ready?' returns `#t' then the next
     `read-char' operation on the given PORT is guaranteed not to hang.
     If the PORT is at end of file then `char-ready?' returns `#t'.  PORT
     may be omitted, in which case it defaults to the value returned by
     `current-input-port'.  

     _Rationale_ `Char-ready?' exists to make it possible for a program
     to accept characters from interactive ports without getting stuck
     waiting for input.  Any input editors associated with such ports
     must ensure that characters whose existence has been asserted by
     `char-ready?' cannot be rubbed out.  If `char-ready?' were to
     return `#f' at end of file, a port at end of file would be
     indistinguishable from an interactive port that has no ready
     characters.

 -- procedure: wait-for-input x
 -- procedure: wait-for-input x port1 ...
     Returns a list those ports PORT1 ... which are `char-ready?'.  If
     none of PORT1 ... become `char-ready?' within the time interval of
     X seconds, then #f is returned.  The PORT1 ... arguments may be
     omitted, in which case they default to the list of the value
     returned by `current-input-port'.  


File: scm.info,  Node: Port Redirection,  Next: Soft Ports,  Prev: Port Properties,  Up: Files and Ports

4.6.3 Port Redirection
----------------------

 -- Function: current-error-port
     Returns the current port to which diagnostic output is directed.

 -- Function: with-error-to-file string thunk
     THUNK must be a procedure of no arguments, and string must be a
     string naming a file.  The file is opened for output, an output
     port connected to it is made the default value returned by
     current-error-port, and the THUNK is called with no arguments.
     When the thunk returns, the port is closed and the previous
     default is restored.  With-error-to-file returns the value yielded
     by THUNK.

 -- Function: with-input-from-port port thunk
 -- Function: with-output-to-port port thunk
 -- Function: with-error-to-port port thunk
     These routines differ from with-input-from-file,
     with-output-to-file, and with-error-to-file in that the first
     argument is a port, rather than a string naming a file.

 -- Function: call-with-outputs thunk proc
     Calls the THUNK procedure while the current-output-port and
     current-error-port are directed to string-ports.  If THUNK
     returns, the PROC procedure is called with the output-string, the
     error-string, and the value returned by THUNK.  If THUNK does not
     return a value (perhaps because of error), PROC is called with
     just the output-string and the error-string as arguments.


File: scm.info,  Node: Soft Ports,  Prev: Port Redirection,  Up: Files and Ports

4.6.4 Soft Ports
----------------

A "soft-port" is a port based on a vector of procedures capable of
accepting or delivering characters.  It allows emulation of I/O ports.

 -- Function: make-soft-port vector modes
     Returns a port capable of receiving or delivering characters as
     specified by the MODES string (*note open-file: Files and Ports.).
     VECTOR must be a vector of length 5.  Its components are as
     follows:

       0. procedure accepting one character for output

       1. procedure accepting a string for output

       2. thunk for flushing output

       3. thunk for getting one character

       4. thunk for closing port (not by garbage collection)

     For an output-only port only elements 0, 1, 2, and 4 need be
     procedures.  For an input-only port only elements 3 and 4 need be
     procedures.  Thunks 2 and 4 can instead be `#f' if there is no
     useful operation for them to perform.

     If thunk 3 returns `#f' or an `eof-object' (*note eof-object?:
     (r5rs)Input.) it indicates that the port has reached end-of-file.
     For example:

     If it is necessary to explicitly close the port when it is garbage
     collected, (*note add-finalizer: Interrupts.).

          (define stdout (current-output-port))
          (define p (make-soft-port
                     (vector
                      (lambda (c) (write c stdout))
                      (lambda (s) (display s stdout))
                      (lambda () (display "." stdout))
                      (lambda () (char-upcase (read-char)))
                      (lambda () (display "@" stdout)))
                     "rw"))

          (write p p) => #<input-output-soft#\space45d10#\>


File: scm.info,  Node: Eval and Load,  Next: Lexical Conventions,  Prev: Files and Ports,  Up: The Language

4.7 Eval and Load
=================

 -- Function: try-load filename
     If the string FILENAME names an existing file, the try-load
     procedure reads Scheme source code expressions and definitions
     from the file and evaluates them sequentially and returns `#t'.
     If not, try-load returns `#f'.  The try-load procedure does not
     affect the values returned by `current-input-port' and
     `current-output-port'.

 -- Variable: *load-pathname*
     Is set to the pathname given as argument to `load', `try-load',
     and `dyn:link' (*note Compiling And Linking: (hobbit)Compiling And
     Linking.).  `*load-pathname*' is used to compute the value of
     *Note program-vicinity: (slib)Vicinity.

 -- Function: eval obj
     Alias for *Note eval: (slib)System.

 -- Function: eval-string str
     Returns the result of reading an expression from STR and
     evaluating it.  `eval-string' does not change `*load-pathname*' or
     `line-number'.

 -- Function: load-string str
     Reads and evaluates all the expressions from STR.  As with `load',
     the value returned is unspecified.  `load-string' does not change
     `*load-pathname*' or `line-number'.

 -- Function: line-number
     Returns the current line number of the file currently being loaded.

* Menu:

* Line Numbers::


File: scm.info,  Node: Line Numbers,  Prev: Eval and Load,  Up: Eval and Load

4.7.1 Line Numbers
------------------

Scheme code defined by load may optionally contain line number
information.  Currently this information is used only for reporting
expansion time errors, but in the future run-time error messages may
also include line number information.

 -- Function: try-load pathname reader
     This is the primitive for loading, PATHNAME is the name of a file
     containing Scheme code, and optional argument READER is a function
     of one argument, a port.  READER should read and return Scheme
     code as list structure.  The default value is `read', which is
     used if READER is not supplied or is false.

Line number objects are disjoint from integers or other Scheme types.
When evaluated or loaded as Scheme code, an s-expression containing a
line-number in the car is equivalent to the cdr of the s-expression.  A
pair consisting of a line-number in the car and a vector in the cdr is
equivalent to the vector.  The meaning of s-expressions with
line-numbers in other positions is undefined.

 -- Function: read-numbered port
     Behaves like `read', except that

        bullet Load (read) sytnaxes are enabled.

        bullet every s-expression read will be replaced with a cons of
          a line-number object and the sexp actually read.  This
          replacement is done only if PORT is a tracked port See *Note
          Files and Ports::.


 -- Function: integer->line-number int
     Returns a line-number object with value INT.  INT should be an
     exact non-negative integer.

 -- Function: line-number->integer linum
     Returns the value of line-number object LINUM as an integer.

 -- Function: line-number? obj
     Returns true if and only if OBJ is a line-number object.

 -- Function: read-for-load port
     Behaves like `read', except that load syntaxes are enabled.

 -- Variable: *load-reader*
 -- Variable: *slib-load-reader*
     The value of `*load-reader*' should be a value acceptable as the
     second argument to `try-load' (note that #f is acceptable).  This
     value will be used to read code during calls to `scm:load'.  The
     value of `*slib-load-reader*' will similarly be used during calls
     to `slib:load' and `require'.

     In order to disable all line-numbering, it is sufficient to set!
     `*load-reader*' and `*slib-load-reader*' to #f.


File: scm.info,  Node: Lexical Conventions,  Next: Syntax,  Prev: Eval and Load,  Up: The Language

4.8 Lexical Conventions
=======================

* Menu:

* Common-Lisp Read Syntax::
* Load Syntax::
* Documentation and Comments::
* Modifying Read Syntax::


File: scm.info,  Node: Common-Lisp Read Syntax,  Next: Load Syntax,  Prev: Lexical Conventions,  Up: Lexical Conventions

4.8.1 Common-Lisp Read Syntax
-----------------------------

 -- Read syntax: #\token
     If TOKEN is a sequence of two or more digits, then this syntax is
     equivalent to `#.(integer->char (string->number token 8))'.

     If TOKEN is `C-', `c-', or `^' followed by a character, then this
     syntax is read as a control character.  If TOKEN is `M-' or `m-'
     followed by a character, then a meta character is read.  `c-' and
     `m-' prefixes may be combined.

 -- Read syntax: #+ feature form
     If feature is `provided?' (by `*features*') then FORM is read as a
     scheme expression.  If not, then FORM is treated as whitespace.

     Feature is a boolean expression composed of symbols and `and',
     `or', and `not' of boolean expressions.

     For more information on `provided?' and `*features*', *Note
     Require: (slib)Require.

 -- Read syntax: #- feature form
     is equivalent to `#+(not feature) expression'.

 -- Read syntax: #| any thing |#
     Is a balanced comment.  Everything up to the matching `|#' is
     ignored by the `read'.  Nested `#|...|#' can occur inside ANY
     THING.

"Load sytax" is Read syntax enabled for `read' only when that `read' is
part of loading a file or string.  This distinction was made so that
reading from a datafile would not be able to corrupt a scheme program
using `#.'.

 -- Load syntax: #. expression
     Is read as the object resulting from the evaluation of EXPRESSION.
     This substitution occurs even inside quoted structure.

     In order to allow compiled code to work with `#.' it is good
     practice to define those symbols used inside of EXPRESSION with
     `#.(define ...)'.  For example:

          #.(define foo 9)                        => #<unspecified>
          '(#.foo #.(+ foo foo))                  => (9 18)

 -- Load syntax: #' form
     is equivalent to FORM (for compatibility with common-lisp).


File: scm.info,  Node: Load Syntax,  Next: Documentation and Comments,  Prev: Common-Lisp Read Syntax,  Up: Lexical Conventions

4.8.2 Load Syntax
-----------------

"#!" is the unix mechanism for executing scripts.  See *Note Unix
Scheme Scripts:: for the full description of how this comment supports
scripting.

 -- Load syntax: #?line
 -- Load syntax: #?column
     Return integers for the current line and column being read during a
     load.

 -- Load syntax: #?file
     Returns the string naming the file currently being loaded.  This
     path is the string passed to `load', possibly with `.scm' appended.


File: scm.info,  Node: Documentation and Comments,  Next: Modifying Read Syntax,  Prev: Load Syntax,  Up: Lexical Conventions

4.8.3 Documentation and Comments
--------------------------------

 -- procedure: procedure-documentation proc
     Returns the documentation string of PROC if it exists, or `#f' if
     not.

     If the body of a `lambda' (or the definition of a procedure) has
     more than one expression, and the first expression (preceeding any
     internal definitions) is a string, then that string is the
     "documentation string" of that procedure.  

          (procedure-documentation (lambda (x) "Identity" x)) => "Identity"
          (define (square x)
              "Return the square of X."
              (* x x))
          => #<unspecified>
          (procedure-documentation square) => "Return the square of X."

 -- Function: comment string1 ...
     Appends STRING1 ... to the strings given as arguments to previous
     calls `comment'.

 -- Function: comment
     Returns the (appended) strings given as arguments to previous calls
     `comment' and empties the current string collection.

 -- Load syntax: #;text-till-end-of-line
     Behaves as `(comment "TEXT-TILL-END-OF-LINE")'.


File: scm.info,  Node: Modifying Read Syntax,  Prev: Documentation and Comments,  Up: Lexical Conventions

4.8.4 Modifying Read Syntax
---------------------------

 -- Callback procedure: read:sharp c port
     If a <#> followed by a character (for a non-standard syntax) is
     encountered by `read', `read' will call the value of the symbol
     `read:sharp' with arguments the character and the port being read
     from.  The value returned by this function will be the value of
     `read' for this expression unless the function returns
     `#<unspecified>' in which case the expression will be treated as
     whitespace.  `#<unspecified>' is the value returned by the
     expression `(if #f #f)'.

 -- Callback procedure: load:sharp c port
     Dispatches like `read:sharp', but only during `load's.  The
     read-syntaxes handled by `load:sharp' are a superset of those
     handled by `read:sharp'.  `load:sharp' calls `read:sharp' if none
     of its syntaxes match C.

 -- Callback procedure: char:sharp token
     If the sequence <#\> followed by a non-standard character name is
     encountered by `read', `read' will call the value of the symbol
     `char:sharp' with the token (a string of length at least two) as
     argument.  If the value returned is a character, then that will be
     the value of `read' for this expression, otherwise an error will
     be signaled.

_Note_ When adding new <#> syntaxes, have your code save the previous
value of `load:sharp', `read:sharp', or `char:sharp' when defining it.
Call this saved value if an invocation's syntax is not recognized.
This will allow `#+', `#-', and *Note Uniform Array::s to still be
supported (as they dispatch from `read:sharp').


File: scm.info,  Node: Syntax,  Prev: Lexical Conventions,  Up: The Language

4.9 Syntax
==========

SCM provides a native implementation of "defmacro".  *Note Defmacro:
(slib)Defmacro.

When built with `-F macro' build option (*note Build Options::) and
`*syntax-rules*' is non-false, SCM also supports [R5RS] `syntax-rules'
macros.  *Note Macros: (r5rs)Macros.

Other Scheme Syntax Extension Packages from SLIB can be employed through
the use of `macro:eval' and `macro:load'; Or by using the SLIB
read-eval-print-loop:

     (require 'repl)
     (repl:top-level macro:eval)

With the appropriate catalog entries (*note Library Catalogs:
(slib)Library Catalogs.), files using macro packages will automatically
use the correct macro loader when `require'd.

* Menu:

* Define and Set::
* Defmacro::
* Syntax-Rules::
* Macro Primitives::
* Environment Frames::
* Syntactic Hooks for Hygienic Macros::


File: scm.info,  Node: Define and Set,  Next: Defmacro,  Prev: Syntax,  Up: Syntax

4.9.1 Define and Set
--------------------

 -- Special Form: defined? symbol
     Equivalent to `#t' if SYMBOL is a syntactic keyword (such as `if')
     or a symbol with a value in the top level environment (*note
     Variables and regions: (r5rs)Variables and regions.).  Otherwise
     equivalent to `#f'.

 -- Special Form: defvar identifier initial-value
     If IDENTIFIER is unbound in the top level environment, then
     IDENTIFIER is `define'd to the result of evaluating the form
     INITIAL-VALUE as if the `defvar' form were instead the form
     `(define identifier initial-value)' .  If IDENTIFIER already has a
     value, then INITIAL-VALUE is _not_ evaluated and IDENTIFIER's
     value is not changed.  `defvar' is valid only when used at
     top-level.

 -- Special Form: defconst identifier value
     If IDENTIFIER is unbound in the top level environment, then
     IDENTIFIER is `define'd to the result of evaluating the form VALUE
     as if the `defconst' form were instead the form `(define
     identifier value)' .  If IDENTIFIER already has a value, then
     VALUE is _not_ evaluated, IDENTIFIER's value is not changed, and
     an error is signaled.  `defconst' is valid only when used at
     top-level.

 -- Special Form: set! (variable1 variable2 ...) <expression>
     The identifiers VARIABLE1, VARIABLE2, ... must be bound either in
     some region enclosing the `set!' expression or at top level.

     <Expression> is evaluated, and the elements of the resulting list
     are stored in the locations to which each corresponding VARIABLE
     is bound.  The result of the `set!' expression is unspecified.

          (define x 2)
          (define y 3)
          (+ x y)                              => 5
          (set! (x y) (list 4 5))              => _unspecified_
          (+ x y)                              => 9

 -- Special Form: qase key clause1 clause2 ...
     `qase' is an extension of standard Scheme `case': Each CLAUSE of a
     `qase' statement must have as first element a list containing
     elements which are:

        * literal datums, or

        * a comma followed by the name of a symbolic constant, or

        * a comma followed by an at-sign (@) followed by the name of a
          symbolic constant whose value is a list.

     A `qase' statement is equivalent to a `case' statement in which
     these symbolic constants preceded by commas have been replaced by
     the values of the constants, and all symbolic constants preceded by
     comma-at-signs have been replaced by the elements of the list
     values of the constants.  This use of comma, (or, equivalently,
     `unquote') is similar to that of `quasiquote' except that the
     unquoted expressions must be "symbolic constants".

     Symbolic constants are defined using `defconst', their values are
     substituted in the head of each `qase' clause during macro
     expansion.  `defconst' constants should be defined before use.
     `qase' can be substituted for any correct use of `case'.

     (defconst unit '1)
     (defconst semivowels '(w y))
     (qase (* 2 3)
       ((2 3 5 7) 'prime)
       ((,unit 4 6 8 9) 'composite))        ==>  composite
     (qase (car '(c d))
       ((a) 'a)
       ((b) 'b))                            ==>  _unspecified_
     (qase (car '(c d))
       ((a e i o u) 'vowel)
       ((,@semivowels) 'semivowel)
       (else 'consonant))                   ==>  consonant


File: scm.info,  Node: Defmacro,  Next: Syntax-Rules,  Prev: Define and Set,  Up: Syntax

4.9.2 Defmacro
--------------

SCM supports the following constructs from Common Lisp: `defmacro',
`macroexpand', `macroexpand-1', and `gentemp'.  *Note Defmacro:
(slib)Defmacro.

SCM `defmacro' is extended over that described for SLIB:

     (defmacro (macro-name . arguments) body)

is equivalent to

     (defmacro macro-name arguments body)

As in Common Lisp, an element of the formal argument list for
`defmacro' may be a possibly nested list, in which case the
corresponding actual argument must be a list with as many members as the
formal argument.  Rest arguments are indicated by improper lists, as in
Scheme.  It is an error if the actual argument list does not have the
tree structure required by the formal argument list.

For example:

     (defmacro (let1 ((name value)) . body)
         `((lambda (,name) ,@body) ,value))

     (let1 ((x (foo))) (print x) x) == ((lambda (x) (print x) x) (foo))

     (let1 not legal syntax) error--> not "does not match" ((name value))


File: scm.info,  Node: Syntax-Rules,  Next: Macro Primitives,  Prev: Defmacro,  Up: Syntax

4.9.3 Syntax-Rules
------------------

SCM supports [R5RS] `syntax-rules' macros *Note Macros: (r5rs)Macros.

The pattern language is extended by the syntax `(... <obj>)', which is
identical to `<obj>' except that ellipses in `<obj>' are treated as
ordinary identifiers in a template, or as literals in a pattern.  In
particular, `(... ...)' quotes the ellipsis token `...' in a pattern or
template.

For example:
     (define-syntax check-tree
       (syntax-rules ()
         ((_ (?pattern (... ...)) ?obj)
          (let loop ((obj ?obj))
            (or (null? obj)
                (and (pair? obj)
                     (check-tree ?pattern (car obj))
                     (loop (cdr obj))))))
         ((_ (?first . ?rest) ?obj)
          (let ((obj ?obj))
            (and (pair? obj)
                 (check-tree ?first (car obj))
                 (check-tree ?rest (cdr obj)))))
         ((_ ?atom ?obj) #t)))

     (check-tree ((a b) ...) '((1 2) (3 4) (5 6))) => #t

     (check-tree ((a b) ...) '((1 2) (3 4) not-a-2list) => #f

Note that although the ellipsis is matched as a literal token in the
defined macro it is not included in the literals list for
`syntax-rules'.

The pattern language is also extended to support identifier macros.  A
reference to an identifier macro keyword that is not the first
identifier in a form may expand into Scheme code, rather than raising a
"keyword as variable" error.  The pattern for expansion of such a bare
macro keyword is a single identifier, as in other syntax rules the
identifier is ignored.

For example:
     (define-syntax eight
         (syntax-rules ()
           (_ 8)))

     (+ 3 eight) => 11
     (eight) => ERROR
     (set! eight 9) => ERROR


File: scm.info,  Node: Macro Primitives,  Next: Environment Frames,  Prev: Syntax-Rules,  Up: Syntax

4.9.4 Macro Primitives
----------------------

 -- Function: procedure->syntax proc
     Returns a "macro" which, when a symbol defined to this value
     appears as the first symbol in an expression, returns the result
     of applying PROC to the expression and the environment.

 -- Function: procedure->macro proc
 -- Function: procedure->memoizing-macro proc
 -- Function: procedure->identifier-macro
     Returns a "macro" which, when a symbol defined to this value
     appears as the first symbol in an expression, evaluates the result
     of applying PROC to the expression and the environment.  The value
     returned from PROC which has been passed to
     `PROCEDURE->MEMOIZING-MACRO' replaces the form passed to PROC.
     For example:

          (defsyntax trace
            (procedure->macro
             (lambda (x env) `(set! ,(cadr x) (tracef ,(cadr x) ',(cadr x))))))

          (trace foo) == (set! foo (tracef foo 'foo)).

     `PROCEDURE->IDENTIFIER-MACRO' is similar to
     `PROCEDURE->MEMOIZING-MACRO' except that PROC is also called in
     case the symbol bound to the macro appears in an expression but
     _not_ as the first symbol, that is, when it looks like a variable
     reference.  In that case, the form passed to PROC is a single
     identifier.


 -- Special Form: defsyntax name expr
     Defines NAME as a macro keyword bound to the result of evaluating
     EXPR, which should be a macro.  Using `define' for this purpose
     may not result in NAME being interpreted as a macro keyword.


File: scm.info,  Node: Environment Frames,  Next: Syntactic Hooks for Hygienic Macros,  Prev: Macro Primitives,  Up: Syntax

4.9.5 Environment Frames
------------------------

An "environment" is a list of frames representing lexical bindings.
Only the names and scope of the bindings are included in environments
passed to macro expanders - run-time values are not included.

There are several types of environment frames:

`((lambda (variable1 ...) ...) value1 ...)'
`(let ((variable1 value1) (variable2 value2) ...) ...)'
`(letrec ((variable1 value1)  ...) ...)'
     result in a single enviroment frame:

          (variable1 variable2 ...)

`(let ((variable1 value1)) ...)'
`(let* ((variable1 value1) ...) ...)'
     result in an environment frame for each variable:

          variable1 variable2 ...

`(let-syntax ((key1 macro1) (key2 macro2)) ...)'
`(letrec-syntax ((key1 value1) (key2 value2)) ...)'
     Lexically bound macros result in environment frames consisting of
     a marker and an alist of keywords and macro objects:

          (<env-syntax-marker> (key1 . value1) (key2 . value2))
     Currently <env-syntax-marker> is the integer 6.

`line numbers'
     Line numbers (*note Line Numbers::) may be included in the
     environment as frame entries to indicate the line number on which
     a function is defined.  They are ignored for variable lookup.

          #<line 8>

`miscellaneous'
     Debugging information is stored in environments in a plist format:
     Any exact integer stored as an environment frame may be followed
     by any value.  The two frame entries are ignored when doing
     variable lookup.  Load file names, procedure names, and closure
     documentation strings are stored in this format.

          <env-filename-marker> "foo.scm" <env-procedure-name-marker> foo ...

     Currently <env-filename-marker> is the integer 1 and
     <env-procedure-name-marker> the integer 2.


 -- Special Form: @apply procedure argument-list
     Returns the result of applying PROCEDURE to ARGUMENT-LIST.
     `@apply' differs from `apply' when the identifiers bound by the
     closure being applied are `set!'; setting affects ARGUMENT-LIST.

          (define lst (list 'a 'b 'c))
          (@apply (lambda (v1 v2 v3) (set! v1 (cons v2 v3))) lst)
          lst           => ((b . c) b c)

     Thus a mutable environment can be treated as both a list and local
     bindings.


File: scm.info,  Node: Syntactic Hooks for Hygienic Macros,  Prev: Environment Frames,  Up: Syntax

4.9.6 Syntactic Hooks for Hygienic Macros
-----------------------------------------

SCM provides a synthetic identifier type for efficient implementation of
hygienic macros (for example, `syntax-rules' *note Macros:
(r5rs)Macros.) A synthetic identifier may be inserted in Scheme code by
a macro expander in any context where a symbol would normally be used.
Collectively, symbols and synthetic identifiers are _identifiers_.

 -- Function: identifier? obj
     Returns `#t' if OBJ is a symbol or a synthetic identifier, and
     `#f' otherwise.

If it is necessary to distinguish between symbols and synthetic
identifiers, use the predicate `symbol?'.

A synthetic identifier includes two data: a parent, which is an
identifier, and an environment, which is either `#f' or a lexical
environment which has been passed to a "macro expander" (a procedure
passed as an argument to `procedure->macro',
`procedure->memoizing-macro', or `procedure->syntax').

 -- Function: renamed-identifier parent env
     Returns a synthetic identifier.  PARENT must be an identifier, and
     ENV must either be `#f' or a lexical environment passed to a macro
     expander.  `renamed-identifier' returns a distinct object for each
     call, even if passed identical arguments.

There is no direct way to access all of the data internal to a synthetic
identifier, those data are used during variable lookup.  If a synthetic
identifier is inserted as quoted data then during macro expansion it
will be repeatedly replaced by its parent, until a symbol is obtained.

 -- Function: identifier->symbol id
     Returns the symbol obtained by recursively extracting the parent of
     ID, which must be an identifier.

4.9.7 Use of Synthetic Identifiers
----------------------------------

`renamed-identifier' may be used as a replacement for `gentemp':
     (define gentemp
       (let ((name (string->symbol "An unlikely variable")))
         (lambda ()
           (renamed-identifier name #f))))

If an identifier returned by this version of `gentemp' is inserted in a
binding position as the name of a variable then it is guaranteed that
no other identifier (except one produced by passing the first to
`renamed-identifier') may denote that variable.  If an identifier
returned by `gentemp' is inserted free, then it will denote the
top-level value bound to its parent, the symbol named "An unlikely
variable".  This behavior, of course, is meant to be put to good use:

     (define top-level-foo
       (procedure->memoizing-macro
        (lambda (exp env)
          (renamed-identifier 'foo #f))))

Defines a macro which may always be used to refer to the top-level
binding of `foo'.

     (define foo 'top-level)
     (let ((foo 'local))
       (top-level-foo))  => top-level

In other words, we can avoid capturing `foo'.

If a lexical environment is passed as the second argument to
`renamed-identifier' then if the identifier is inserted free its parent
will be looked up in that environment, rather than in the top-level
environment.  The use of such an identifier _must_ be restricted to the
lexical scope of its environment.

There is another restriction imposed for implementation convenience:
Macros passing their lexical environments to `renamed-identifier' may
be lexically bound only by the special forms `let-syntax' or
`letrec-syntax'.  No error is signaled if this restriction is not met,
but synthetic identifier lookup will not work properly.

In order to maintain referential transparency it is necessary to
determine whether two identifiers have the same denotation.  With
synthetic identifiers it is not necessary that two identifiers be `eq?'
in order to denote the same binding.

 -- Function: identifier-equal? id1 id2 env
     Returns `#t' if identifiers ID1 and ID2 denote the same binding in
     lexical environment ENV, and `#f' otherwise.  ENV must either be a
     lexical environment passed to a macro transformer during macro
     expansion or the empty list.

     For example,
          (define top-level-foo?
            (procedure->memoizing-macro
             (let ((foo-name (renamed-identifier 'foo #f)))
               (lambda (exp env)
                 (identifier-equal? (cadr exp) foo-name env)))))

          (top-level-foo? foo)  => #t

          (let ((foo 'local))
            (top-level-foo? foo))  => #f

 -- Function: @macroexpand1 expr env
     If the `car' of EXPR denotes a macro in ENV, then if that macro is
     a primitive, EXPR will be returned, if the macro was defined in
     Scheme, then a macro expansion will be returned.  If the `car' of
     EXPR does not denote a macro, the `#f' is returned.

 -- Function: extended-environment names values env
     Returns a new environment object, equivalent to ENV, which must
     either be an environment object or null, extended by one frame.
     NAMES must be an identifier, or an improper list of identifiers,
     usable as a formals list in a `lambda' expression.  VALUES must be
     a list of objects long enough to provide a binding for each of the
     identifiers in NAMES.  If NAMES is an identifier or an improper
     list then VALS may be, respectively, any object or an improper
     list of objects.

 -- Special Form: syntax-quote obj
     Synthetic identifiers are converted to their parent symbols by
     `quote' and `quasiquote' so that literal data in macro definitions
     will be properly transcribed.  `syntax-quote' behaves like
     `quote', but preserves synthetic identifier intact.

 -- Special Form: the-macro mac
     `the-macro' is the simplest of all possible macro transformers:
     MAC may be a syntactic keyword (macro name) or an expression
     evaluating to a macro, otherwise an error is signaled.  MAC is
     evaluated and returned once only, after which the same memoizied
     value is returned.

     `the-macro' may be used to protect local copies of macros against
     redefinition, for example:
          (@let-syntax ((let (the-macro let)))
             ;; code that will continue to work even if LET is redefined.
                  ...)

 -- Special Form: renaming-transformer proc
     A low-level "explicit renaming" macro facility very similar to that
     proposed by W. Clinger [Exrename] is supported.  Syntax may be
     defined in `define-syntax', `let-syntax', and `letrec-syntax'
     using `renaming-transformer' instead of `syntax-rules'.  PROC
     should evaluate to a procedure accepting three arguments: EXPR,
     RENAME, and COMPARE.  EXPR is a representation of Scheme code to
     be expanded, as list structure.  RENAME is a procedure accepting
     an identifier and returning an identifier renamed in the
     definition environment of the new syntax.  COMPARE accepts two
     identifiers and returns true if and only if both denote the same
     binding in the usage environment of the new syntax.


File: scm.info,  Node: Packages,  Next: The Implementation,  Prev: The Language,  Up: Top

5 Packages
**********

* Menu:

* Dynamic Linking::
* Dump::                        Create Fast-Booting Executables
* Numeric::                     Numeric Language Extensions
* Arrays::                      As in APL
* Records::                     Define new aggregate data types
* I/O-Extensions::              i/o-extensions
* Posix Extensions::            posix
* Unix Extensions::             non-posix unix
* Sequence Comparison::
* Regular Expression Pattern Matching::  regex
* Line Editing::                edit-line
* Curses::                      Screen Control
* Sockets::                     Cruise the Net
* SCMDB::                       interface to MySQL

* Menu:

* Xlib: (Xlibscm).              X Window Graphics.
* Hobbit: (hobbit).             Scheme-to-C Compiler


File: scm.info,  Node: Dynamic Linking,  Next: Dump,  Prev: Packages,  Up: Packages

5.1 Dynamic Linking
===================

If SCM has been compiled with `dynl.c' then the additional properties
of load and ([SLIB]) require specified here are supported.  The
`require' form is preferred.

 -- Function: require feature
     If the symbol FEATURE has not already been given as an argument to
     `require', then the object and library files associated with
     FEATURE will be dynamically-linked, and an unspecified value
     returned.  If FEATURE is not found in `*catalog*', then an error
     is signaled.

 -- Function: usr:lib lib
     Returns the pathname of the C library named LIB.  For example:
     `(usr:lib "m")' returns `"/usr/lib/libm.a"', the path of the C
     math library.

 -- Function: x:lib lib
     Returns the pathname of the X library named LIB.  For example:
     `(x:lib "X11")' returns `"/usr/X11/lib/libX11.sa"', the path of
     the X11 library.

 -- Function: load filename lib1 ...
     In addition to the [R5RS] requirement of loading Scheme
     expressions if FILENAME is a Scheme source file, `load' will also
     dynamically load/link object files (produced by `compile-file', for
     instance).  The object-suffix need not be given to load.  For
     example,

          (load (in-vicinity (implementation-vicinity) "sc2"))
          or (load (in-vicinity (implementation-vicinity) "sc2.o"))
          or (require 'rev2-procedures)
          or (require 'rev3-procedures)
          

     will load/link `sc2.o' if it exists.

     The LIB1 ... pathnames specify additional libraries which may be
     needed for object files not produced by the Hobbit compiler.  For
     instance, crs is linked on Linux by

          (load (in-vicinity (implementation-vicinity) "crs.o")
                (usr:lib "ncurses") (usr:lib "c"))
          or (require 'curses)
          

     Turtlegr graphics library is linked by:

          (load (in-vicinity (implementation-vicinity) "turtlegr")
                (usr:lib "X11") (usr:lib "c") (usr:lib "m"))
          or (require 'turtle-graphics)
          

     And the string regular expression (*note Regular Expression
     Pattern Matching::) package is linked by:

          (load (in-vicinity (implementation-vicinity) "rgx") (usr:lib "c"))
     or
          (require 'regex)
          

The following functions comprise the low-level Scheme interface to
dynamic linking.  See the file `Link.scm' in the SCM distribution for
an example of their use.

 -- Function: dyn:link filename
     FILENAME should be a string naming an "object" or "archive" file,
     the result of C-compiling.  The `dyn:link' procedure links and
     loads FILENAME into the current SCM session.  If successfull,
     `dyn:link' returns a "link-token" suitable for passing as the
     second argument to `dyn:call'.  If not successful, `#f' is
     returned.

 -- Function: dyn:call name link-token
     LINK-TOKEN should be the value returned by a call to `dyn:link'.
     NAME should be the name of C function of no arguments defined in
     the file named FILENAME which was succesfully `dyn:link'ed in the
     current SCM session.  The `dyn:call' procedure calls the C
     function corresponding to NAME.  If successful, `dyn:call' returns
     `#t'; If not successful, `#f' is returned.

     `dyn:call' is used to call the "init_..." function after loading
     SCM object files.  The init_... function then makes the
     identifiers defined in the file accessible as Scheme procedures.

 -- Function: dyn:main-call name link-token arg1 ...
     LINK-TOKEN should be the value returned by a call to `dyn:link'.
     NAME should be the name of C function of 2 arguments, `(int argc,
     const char **argv)', defined in the file named FILENAME which was        |
     succesfully `dyn:link'ed in the current SCM session.  The
     `dyn:main-call' procedure calls the C function corresponding to
     NAME with `argv' style arguments, such as are given to C `main'
     functions.  If successful, `dyn:main-call' returns the integer
     returned from the call to NAME.

     `dyn:main-call' can be used to call a `main' procedure from SCM.
     For example, I link in and `dyn:main-call' a large C program, the
     low level routines of which callback (*note Callbacks::) into SCM
     (which emulates PCI hardware).

 -- Function: dyn:unlink link-token
     LINK-TOKEN should be the value returned by a call to `dyn:link'.
     The `dyn:unlink' procedure removes the previously loaded file from
     the current SCM session.  If successful, `dyn:unlink' returns
     `#t'; If not successful, `#f' is returned.


File: scm.info,  Node: Dump,  Next: Numeric,  Prev: Dynamic Linking,  Up: Packages

5.2 Dump
========

"Dump", (also known as "unexec"), saves the continuation of an entire
SCM session to an executable file, which can then be invoked as a
program.  Dumped executables start very quickly, since no Scheme code
has to be loaded.

There are constraints on which sessions are savable using `dump'

   * Saved continuations are invalid in subsequent invocations; they
     cause segmentation faults and other unpleasant side effects.

   * Although DLD (*note Dynamic Linking::) can be used to load compiled
     modules both before and after dumping, `SUN_DL' ELF systems can
     load compiled modules only after dumping.  This can be worked
     around by compiling in those features you wish to `dump'.

   * Ports (other than `current-input-port', `current-output-port',
     `current-error-port'), X windows, etc. are invalid in subsequent
     invocations.

     This restriction could be removed; *Note Improvements To Make::.

   * `Dump' should only be called from a loading file when the call to
     dump is the last expression in that file.

   * `Dump' can be called from the command line.

 -- Function: dump newpath
 -- Function: dump newpath #f
 -- Function: dump newpath #t
 -- Function: dump newpath thunk
        * Calls `gc'.

        * Creates an executable program named NEWPATH which continues
          the state of the current SCM session when invoked.  The
          optional argument THUNK, if provided, should be a procedure
          of no arguments; BOOT-TAIL will be set to this procedure,
          causing it to be called in the restored executable.

          If the optional argument is missing or a boolean, SCM's
          standard command line processing will be called in the
          restored executable.

          If the second argument to `dump' is `#t', argument processing
          will continue from the command line passed to the dumping
          session.  If the second argument is missing or `#f' then the
          command line arguments of the restoring invocation will be
          processed.

        * Resumes the top level Read-Eval-Print loop.  This is done
          instead of continuing normally to avoid creating a saved
          continuation in the dumped executable.

     `dump' may set the values of `boot-tail', `*argv*', `restart', and
     *INTERACTIVE*.  `dump' returns an unspecified value.

When a dumped executable is invoked, the variable *INTERACTIVE* (*note
Internal State::) has the value it possessed when `dump' created it.
Calling `dump' with a single argument sets *INTERACTIVE* to `#f', which
is the state it has at the beginning of command line processing.

The procedure `program-arguments' returns the command line arguments
for the curent invocation.  More specifically, `program-arguments' for
the restored session are _not_ saved from the dumping session.  Command
line processing is done on the value of the identifier `*argv*'.

The following example shows how to create `rscm', which is like regular
scm, but which loads faster and has the `random' package alreadly
provided.

     bash$ scm -rrandom
     > (dump "rscm")
     #<unspecified>
     > (quit)
     bash$ ./rscm -lpi.scm -e"(pi (random 200) 5)"
     00003 14159 26535 89793 23846 26433 83279 50288 41971 69399
     37510 58209 74944 59230 78164 06286 20899 86280 34825 34211
     70679 82148 08651 32823 06647 09384 46095 50582 23172 53594
     08128 48111 74502 84102 70193 85211 05559 64462 29489
     bash$

This task can also be accomplished using the `-o' command line option
(*note SCM Options::).

     bash$ scm -rrandom -o rscm
     > (quit)
     bash$ ./rscm -lpi.scm -e"(pi (random 200) 5)"
     00003 14159 26535 89793 23846 26433 83279 50288 41971 69399
     37510 58209 74944 59230 78164 06286 20899 86280 34825 34211
     70679 82148 08651 32823 06647 09384 46095 50582 23172 53594
     08128 48111 74502 84102 70193 85211 05559 64462 29489
     bash$


File: scm.info,  Node: Numeric,  Next: Arrays,  Prev: Dump,  Up: Packages

5.3 Numeric
===========

 -- Constant: most-positive-fixnum
     The immediate integer closest to positive infinity.  *Note
     Configuration: (slib)Configuration.

 -- Constant: most-negative-fixnum
     The immediate integer closest to negative infinity.

 -- Constant: $pi
 -- Constant: pi
     The ratio of the circumference to the diameter of a circle.

These procedures augment the standard capabilities in *Note Numerical
operations: (r5rs)Numerical operations.

 -- Function: pi* z
     `(* pi Z)'

 -- Function: pi/ z
     `(/ pi Z)'

 -- Function: sinh z
 -- Function: cosh z
 -- Function: tanh z
     Return the hyperbolic sine, cosine, and tangent of Z

 -- Function: asinh z
 -- Function: acosh z
 -- Function: atanh z
     Return the inverse hyperbolic sine, cosine, and tangent of Z

 -- Function: $sqrt x
 -- Function: $abs x
 -- Function: $exp x
 -- Function: $log x
 -- Function: $sin x
 -- Function: $cos x
 -- Function: $tan x
 -- Function: $asin x
 -- Function: $acos x
 -- Function: $atan x
 -- Function: $sinh x
 -- Function: $cosh x
 -- Function: $tanh x
 -- Function: $asinh x
 -- Function: $acosh x
 -- Function: $atanh x
     Real-only versions of these popular functions.  The argument X
     must be a real number.  It is an error if the value which should be
     returned by a call to these procedures is _not_ real.

 -- Function: $log10 x
     Real-only base 10 logarithm.

 -- Function: $atan2 y x
     Computes `(angle (make-rectangular x y))' for real numbers Y and X.

 -- Function: $expt x1 x2
     Returns real number X1 raised to the real power X2.  It is an
     error if the value which should be returned by a call to `$expt'
     is not real.


File: scm.info,  Node: Arrays,  Next: Records,  Prev: Numeric,  Up: Packages

5.4 Arrays
==========

* Menu:

* Conventional Arrays::
* Uniform Array::
* Bit Vectors::
* Array Mapping::               array-for-each


File: scm.info,  Node: Conventional Arrays,  Next: Uniform Array,  Prev: Arrays,  Up: Arrays

5.4.1 Conventional Arrays
-------------------------

The following syntax and procedures are SCM extensions to feature
`array' in *Note Arrays: (slib)Arrays.

"Arrays" read and write as a `#' followed by the "rank" (number of
dimensions) followed by the character #\a or #\A and what appear as
lists (of lists) of elements.  The lists must be nested to the depth of
the rank.  For each depth, all lists must be the same length.
     (make-array '#(ho) 4 3) =>
     #2A((ho ho ho) (ho ho ho) (ho ho ho) (ho ho ho))

Unshared, conventional (not uniform) 0-based arrays of rank 1 are
equivalent to (and can't be distinguished from) scheme vectors.
     (make-array '#(ho) 3) => #(ho ho ho)

 -- Function: transpose-array array dim0 dim1 ...
     Returns an array sharing contents with ARRAY, but with dimensions
     arranged in a different order.  There must be one DIM argument for
     each dimension of ARRAY.  DIM0, DIM1, ... should be integers
     between 0 and the rank of the array to be returned.  Each integer
     in that range must appear at least once in the argument list.

     The values of DIM0, DIM1, ... correspond to dimensions in the
     array to be returned, their positions in the argument list to
     dimensions of ARRAY.  Several DIMs may have the same value, in
     which case the returned array will have smaller rank than ARRAY.

     examples:
          (transpose-array '#2A((a b) (c d)) 1 0) => #2A((a c) (b d))
          (transpose-array '#2A((a b) (c d)) 0 0) => #1A(a d)
          (transpose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1 1 0) =>
                          #2A((a 4) (b 5) (c 6))

 -- Function: enclose-array array dim0 dim1 ...
     DIM0, DIM1 ... should be nonnegative integers less than the rank
     of ARRAY.  ENCLOSE-ARRAY returns an array resembling an array of
     shared arrays.  The dimensions of each shared array are the same
     as the DIMth dimensions of the original array, the dimensions of
     the outer array are the same as those of the original array that
     did not match a DIM.

     An enclosed array is not a general Scheme array.  Its elements may
     not be set using `array-set!'.  Two references to the same element
     of an enclosed array will be `equal?' but will not in general be
     `eq?'.  The value returned by ARRAY-PROTOTYPE when given an
     enclosed array is unspecified.

     examples:
          (enclose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1) =>
             #<enclosed-array (#1A(a d) #1A(b e) #1A(c f)) (#1A(1 4) #1A(2 5) #1A(3 6))>

          (enclose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1 0) =>
             #<enclosed-array #2A((a 1) (d 4)) #2A((b 2) (e 5)) #2A((c 3) (f 6))>

 -- Function: array->list array
     Returns a list consisting of all the elements, in order, of ARRAY.
     In the case of a rank-0 array, returns the single element.

 -- Function: array-contents array
 -- Function: array-contents array strict
     If ARRAY may be "unrolled" into a one dimensional shared array
     without changing their order (last subscript changing fastest),
     then `array-contents' returns that shared array, otherwise it
     returns `#f'.  All arrays made by MAKE-ARRAY may be unrolled, some
     arrays made by MAKE-SHARED-ARRAY may not be.

     If the optional argument STRICT is provided, a shared array will
     be returned only if its elements are stored internally contiguous
     in memory.


File: scm.info,  Node: Uniform Array,  Next: Bit Vectors,  Prev: Conventional Arrays,  Up: Arrays

5.4.2 Uniform Array
-------------------

"Uniform Arrays" and vectors are arrays whose elements are all of the
same type.  Uniform vectors occupy less storage than conventional
vectors.  Uniform Array procedures also work on vectors,
uniform-vectors, bit-vectors, and strings.

SLIB now supports uniform arrys.  The primary array creation procedure
is `make-array', detailed in *Note Arrays: (slib)Arrays.

Unshared uniform character 0-based arrays of rank 1 (dimension) are
equivalent to (and can't be distinguished from) strings.
     (make-array "" 3) => "$q2"

Unshared uniform boolean 0-based arrays of rank 1 (dimension) are
equivalent to (and can't be distinguished from) *Note bit-vectors: Bit
Vectors.
     (make-array '#1at() 3) => #*000
     ==
     #1At(#f #f #f) => #*000

PROTOTYPE arguments in the following procedures are interpreted
according to the table:

     prototype       type                              display prefix

     ()              conventional vector                     #A
     +64i            complex (double precision)              #A:floC64b
     64.0            double (double precision)               #A:floR64b
     32.0            float (single precision)                #A:floR32b
     32              unsigned integer (32-bit)               #A:fixN32b
     -32             signed integer (32-bit)                 #A:fixZ32b
     -16             signed integer (16-bit)                 #A:fixZ16b
     #\a             char (string)                           #A:char
     #t              boolean (bit-vector)                    #A:bool

Other uniform vectors are written in a form similar to that of general
arrays, except that one or more modifying characters are put between the
#\A character and the contents list.  For example, `'#1A:fixZ32b(3 5 9)'
returns a uniform vector of signed integers.

 -- Function: array? obj prototype
     Returns `#t' if the OBJ is an array of type corresponding to
     PROTOTYPE, and `#f' if not.

 -- Function: array-prototype array
     Returns an object that would produce an array of the same type as
     ARRAY, if used as the PROTOTYPE for `list->uniform-array'.

 -- Function: list->uniform-array rank prot lst
     Returns a uniform array of the type indicated by prototype PROT
     with elements the same as those of LST.  Elements must be of the
     appropriate type, no coercions are done.

     In, for example, the case of a rank-2 array, LST must be a list of
     lists, all of the same length.  The length of LST will be the
     first dimension of the result array, and the length of each
     element the second dimension.

     If RANK is zero, LST, which need not be a list, is the single
     element of the returned array.

 -- Function: uniform-array-read! ura
 -- Function: uniform-array-read! ura port
     Attempts to read all elements of URA, in lexicographic order, as
     binary objects from PORT.  If an end of file is encountered during
     uniform-array-read! the objects up to that point only are put into
     URA (starting at the beginning) and the remainder of the array is
     unchanged.

     `uniform-array-read!' returns the number of objects read.  PORT
     may be omitted, in which case it defaults to the value returned by
     `(current-input-port)'.

 -- Function: uniform-array-write ura
 -- Function: uniform-array-write ura port
     Writes all elements of URA as binary objects to PORT.  The number
     of of objects actually written is returned.  PORT may be omitted,
     in which case it defaults to the value returned by
     `(current-output-port)'.

 -- Function: logaref array index1 index2 ...
     If an INDEX is provided for each dimension of ARRAY returns the
     INDEX1, INDEX2, ...'th element of ARRAY.  If one more INDEX is
     provided, then the last index specifies bit position of the
     twos-complement representation of the array element indexed by the
     other INDEXs returning `#t' if the bit is 1, and `#f' if 0.  It is
     an error if this element is not an exact integer.

          (logaref '#(#b1101 #b0010) 0)       => #b1101
          (logaref '#(#b1101 #b0010) 0 1)     => #f
          (logaref '#2((#b1101 #b0010)) 0 0)  => #b1101

 -- Function: logaset! array val index1 index2 ...
     If an INDEX is provided for each dimension of ARRAY sets the
     INDEX1, INDEX2, ...'th element of ARRAY to VAL.  If one more INDEX
     is provided, then the last index specifies bit position of the
     twos-complement representation of an exact integer array element,
     setting the bit to 1 if VAL is `#t' and to 0 if VAL is `#f'.  In
     this case it is an error if the array element is not an exact
     integer or if VAL is not boolean.


File: scm.info,  Node: Bit Vectors,  Next: Array Mapping,  Prev: Uniform Array,  Up: Arrays

5.4.3 Bit Vectors
-----------------

Bit vectors can be written and read as a sequence of `0's and `1's
prefixed by `#*'.

     #1At(#f #f #f #t #f #t #f) => #*0001010

Some of these operations will eventually be generalized to other
uniform-arrays.

 -- Function: bit-count bool bv
     Returns the number of occurrences of BOOL in BV.

 -- Function: bit-position bool bv k
     Returns the minimum index of an occurrence of BOOL in BV which is
     at least K.  If no BOOL occurs within the specified range `#f' is
     returned.

 -- Function: bit-invert! bv
     Modifies BV by replacing each element with its negation.

 -- Function: bit-set*! bv uve bool
     If uve is a bit-vector BV and uve must be of the same length.  If
     BOOL is `#t', uve is OR'ed into BV; If BOOL is `#f', the inversion
     of uve is AND'ed into BV.

     If uve is a unsigned integer vector all the elements of uve must be
     between 0 and the `LENGTH' of BV.  The bits of BV corresponding to
     the indexes in uve are set to BOOL.

     The return value is unspecified.

 -- Function: bit-count* bv uve bool
     Returns
          (bit-count (bit-set*! (if bool bv (bit-invert! bv)) uve #t) #t).
     BV is not modified.


File: scm.info,  Node: Array Mapping,  Prev: Bit Vectors,  Up: Arrays

5.4.4 Array Mapping
-------------------

`(require 'array-for-each)' 

SCM has some extra functions in feature `array-for-each':

 -- Function: array-fill! array fill
     Stores FILL in every element of ARRAY.  The value returned is
     unspecified.

 -- Function: serial-array:copy! destination source
     Same as `array:copy!' but guaranteed to copy in row-major order.

 -- Function: array-equal? array0 array1 ...
     Returns `#t' iff all arguments are arrays with the same shape, the
     same type, and have corresponding elements which are either
     `equal?'  or `array-equal?'.  This function differs from `equal?'
     in that a one dimensional shared array may be ARRAY-EQUAL? but not
     EQUAL? to a vector or uniform vector.

 -- Function: array-map! array0 proc array1 ...
     If ARRAY1, ... are arrays, they must have the same number of
     dimensions as ARRAY0 and have a range for each index which
     includes the range for the corresponding index in ARRAY0.  If they
     are scalars, that is, not arrays, vectors, or strings, then they
     will be converted internally to arrays of the appropriate shape.
     PROC is applied to each tuple of elements of ARRAY1 ... and the
     result is stored as the corresponding element in ARRAY0.  The
     value returned is unspecified.  The order of application is
     unspecified.

     Handling non-array arguments is a SCM extension of *Note
     array-map!: (slib)Array Mapping.

 -- Function: serial-array-map! array0 proc array1 ...
     Same as ARRAY-MAP!, but guaranteed to apply PROC in row-major
     order.

 -- Function: array-map prototype proc array1 array2 ...
     ARRAY2, ... must have the same number of dimensions as ARRAY1 and
     have a range for each index which includes the range for the
     corresponding index in ARRAY1.  PROC is applied to each tuple of
     elements of ARRAY1, ARRAY2, ... and the result is stored as the
     corresponding element in a new array of type PROTOTYPE.  The new
     array is returned.  The order of application is unspecified.

 -- Function: scalar->array scalar array prototype
 -- Function: scalar->array scalar array
     Returns a uniform array of the same shape as ARRAY, having only
     one shared element, which is `eqv?' to SCALAR.  If the optional
     argument PROTOTYPE is supplied it will be used as the prototype
     for the returned array.  Otherwise the returned array will be of
     the same type as `array' if that is possible, and a conventional
     array if it is not.  This function is used internally by
     `array-map!' and friends to handle scalar arguments.


File: scm.info,  Node: Records,  Next: I/O-Extensions,  Prev: Arrays,  Up: Packages

5.5 Records
===========

SCM provides user-definable datatypes with the same interface as SLIB,
see *Note Records: (slib)Records, with the following extension.

 -- Function: record-printer-set! rtd printer
     Causes records of type RTD to be printed in a user-specified
     format.  RTD must be a record type descriptor returned by
     `make-record-type', PRINTER a procedure accepting three arguments:
     the record to be printed, the port to print to, and a boolean
     which is true if the record is being written on behalf of `write'
     and false if for `display'.  If PRINTER returns #f, the default
     record printer will be called.

     A PRINTER value of #f means use the default printer.

     Only the default printer will be used when printing error messages.


File: scm.info,  Node: I/O-Extensions,  Next: Posix Extensions,  Prev: Records,  Up: Packages

5.6 I/O-Extensions
==================

If `'i/o-extensions' is provided (by linking in `ioext.o'), *Note Line
I/O: (slib)Line I/O, and the following functions are defined:

 -- Function: stat <port-or-string>
     Returns a vector of integers describing the argument.  The argument
     can be either a string or an open input port.  If the argument is
     an open port then the returned vector describes the file to which
     the port is opened; If the argument is a string then the returned
     vector describes the file named by that string.  If there exists
     no file with the name string, or if the file cannot be accessed
     `#f' is returned.  The elements of the returned vector are as
     follows:

    0  st_dev
          ID of device containing a directory entry for this file

    1  st_ino
          Inode number

    2  st_mode
          File type, attributes, and access control summary

    3  st_nlink
          Number of links

    4  st_uid
          User ID of file owner

    5  st_gid
          Group ID of file group

    6  st_rdev
          Device ID; this entry defined only for char or blk spec files

    7  st_size
          File size (bytes)

    8  st_atime
          Time of last access

    9  st_mtime
          Last modification time

    10 st_ctime
          Last file status change time

 -- Function: getpid
     Returns the process ID of the current process.

 -- Function: file-position port
     Returns the current position of the character in PORT which will
     next be read or written.  If PORT is not open to a file the result
     is unspecified.

 -- Function: file-set-position port integer
     Sets the current position in PORT which will next be read or
     written.  If PORT is not open to a file the action of
     `file-set-position' is unspecified.  The result of
     `file-set-position' is unspecified.

 -- Function: try-create-file name modes perms
     If the file with name NAME already exists, return `#f', otherwise
     try to create and open the file like `try-open-file', *Note Files
     and Ports::.  If the optional integer argument PERMS is provided,
     it is used as the permissions of the new file (modified by the
     current umask).

 -- Function: reopen-file filename modes port
     Closes port PORT and reopens it with FILENAME and MODES.
     `reopen-file' returns `#t' if successful, `#f' if not.

 -- Function: duplicate-port port modes
     Creates and returns a "duplicate" port from PORT.  Duplicate
     _unbuffered_ ports share one file position.  MODES are as for
     *Note open-file: Files and Ports.

 -- Function: redirect-port! from-port to-port
     Closes TO-PORT and makes TO-PORT be a duplicate of FROM-PORT.
     `redirect-port!' returns TO-PORT if successful, `#f' if not.  If
     unsuccessful, TO-PORT is not closed.

 -- Function: opendir dirname
     Returns a "directory" object corresponding to the file system
     directory named DIRNAME.  If unsuccessful, returns `#f'.

 -- Function: readdir dir
     Returns the string name of the next entry from the directory DIR.
     If there are no more entries in the directory, `readdir' returns a
     `#f'.

 -- Function: rewinddir dir
     Reinitializes DIR so that the next call to `readdir' with DIR will
     return the first entry in the directory again.

 -- Function: closedir dir
     Closes DIR and returns `#t'.  If DIR is already closed,,
     `closedir' returns a `#f'.

 -- Function: directory-for-each proc directory
     PROC must be a procedure taking one argument.
     `Directory-For-Each' applies PROC to the (string) name of each
     file in DIRECTORY.  The dynamic order in which PROC is applied to
     the filenames is unspecified.  The value returned by
     `directory-for-each' is unspecified.

 -- Function: directory-for-each proc directory pred
     Applies PROC only to those filenames for which the procedure PRED
     returns a non-false value.

 -- Function: directory-for-each proc directory match
     Applies PROC only to those filenames for which `(filename:match??
     MATCH)' would return a non-false value (*note Filenames:
     (slib)Filenames.).

          (require 'directory)
          (directory-for-each print "." "[A-Z]*.scm")
          -|
          "Init.scm"
          "Iedline.scm"
          "Link.scm"
          "Macro.scm"
          "Transcen.scm"
          "Init5e2.scm"                                                       |

 -- Function: mkdir path mode
     The `mkdir' function creates a new, empty directory whose name is
     PATH.  The integer argument MODE specifies the file permissions
     for the new directory.  *Note The Mode Bits for Access Permission:
     (libc)The Mode Bits for Access Permission, for more information
     about this.

     `mkdir' returns if successful, `#f' if not.

 -- Function: rmdir path
     The `rmdir' function deletes the directory PATH.  The directory
     must be empty before it can be removed.  `rmdir' returns if
     successful, `#f' if not.

 -- Function: chdir filename
     Changes the current directory to FILENAME.  If FILENAME does not
     exist or is not a directory, `#f' is returned.  Otherwise, `#t' is
     returned.

 -- Function: getcwd
     The function `getcwd' returns a string containing the absolute file
     name representing the current working directory.  If this string
     cannot be obtained, `#f' is returned.

 -- Function: rename-file oldfilename newfilename
     Renames the file specified by OLDFILENAME to NEWFILENAME.  If the
     renaming is successful, `#t' is returned.  Otherwise, `#f' is
     returned.

 -- Function: chmod file mode
     The function `chmod' sets the access permission bits for the file
     named by FILE to MODE.  The FILE argument may be a string
     containing the filename or a port open to the file.

     `chmod' returns if successful, `#f' if not.

 -- Function: utime pathname acctime modtime
     Sets the file times associated with the file named PATHNAME to
     have access time ACCTIME and modification time MODTIME.  `utime'
     returns if successful, `#f' if not.

 -- Function: umask mode
     The function `umask' sets the file creation mask of the current
     process to MASK, and returns the previous value of the file
     creation mask.

 -- Function: fileno port
     Returns the integer file descriptor associated with the port PORT.
     If an error is detected, `#f' is returned.

 -- Function: access pathname how
     Returns `#t' if the file named by PATHNAME can be accessed in the
     way specified by the HOW argument.  The HOW argument can be the
     `logior' of the flags:

       0. File-exists?

       1. File-is-executable?

       2. File-is-writable?

       4. File-is-readable?

     Or the HOW argument can be a string of 0 to 3 of the following
     characters in any order.  The test performed is the `and' of the
     associated tests and `file-exists?'.

    <x>
          File-is-executable?

    <w>
          File-is-writable?

    <r>
          File-is-readable?

 -- Function: execl command arg0 ...
 -- Function: execlp command arg0 ...
     Transfers control to program COMMAND called with arguments ARG0
     ....  For `execl', COMMAND must be an exact pathname of an
     executable file.  `execlp' searches for COMMAND in the list of
     directories specified by the environment variable PATH.  The
     convention is that ARG0 is the same name as COMMAND.

     If successful, this procedure does not return.  Otherwise an error
     message is printed and the integer `errno' is returned.

 -- Function: execv command arglist
 -- Function: execvp command arglist
     Like `execl' and `execlp' except that the set of arguments to
     COMMAND is ARGLIST.

 -- Function: putenv string
     adds or removes definitions from the "environment".  If the STRING
     is of the form `NAME=VALUE', the definition is added to the
     environment.  Otherwise, the STRING is interpreted as the name of
     an environment variable, and any definition for this variable in
     the environment is removed.

     Names of environment variables are case-sensitive and must not
     contain the character `='.  System-defined environment variables
     are invariably uppercase.

     `Putenv' is used to set up the environment before calls to
     `execl', `execlp', `execv', `execvp', `system', or `open-pipe'
     (*note open-pipe: Posix Extensions.).

     To access environment variables, use `getenv' (*note getenv:
     (slib)System Interface.).


File: scm.info,  Node: Posix Extensions,  Next: Unix Extensions,  Prev: I/O-Extensions,  Up: Packages

5.7 Posix Extensions
====================

If `'posix' is provided (by linking in `posix.o'), the following
functions are defined:

 -- Function: open-pipe string modes
     If the string MODES contains an <r>, returns an input port capable
     of delivering characters from the standard output of the system
     command STRING.  Otherwise, returns an output port capable of
     receiving characters which become the standard input of the system
     command STRING.  If a pipe cannot be created `#f' is returned.

 -- Function: open-input-pipe string
     Returns an input port capable of delivering characters from the
     standard output of the system command STRING.  If a pipe cannot be
     created `#f' is returned.

 -- Function: open-output-pipe string
     Returns an output port capable of receiving characters which become
     the standard input of the system command STRING.  If a pipe cannot
     be created `#f' is returned.

 -- Function: broken-pipe port
     If this function is defined at top level, it will be called when an
     output pipe is closed from the other side (this is the condition
     under which a SIGPIPE is sent).  The already closed PORT will be
     passed so that any necessary cleanup may be done.  An error is not
     signaled when output to a pipe fails in this way, but any further
     output to the closed pipe will cause an error to be signaled.

 -- Function: close-port pipe
     Closes the PIPE, rendering it incapable of delivering or accepting
     characters.  This routine has no effect if the pipe has already
     been closed.  The value returned is unspecified.

 -- Function: pipe
     Returns `(cons RD WD)' where RD and WD are the read and write
     (port) ends of a "pipe" respectively.

 -- Function: fork
     Creates a copy of the process calling `fork'.  Both processes
     return from `fork', but the calling ("parent") process's `fork'
     returns the "child" process's ID whereas the child process's
     `fork' returns 0.

For a discussion of "ID"s *Note Process Persona: (GNU C Library)Process
Persona.

 -- Function: getppid
     Returns the process ID of the parent of the current process.  For
     a process's own ID *Note getpid: I/O-Extensions.

 -- Function: getuid
     Returns the real user ID of this process.

 -- Function: getgid
     Returns the real group ID of this process.

 -- Function: getegid
     Returns the effective group ID of this process.

 -- Function: geteuid
     Returns the effective user ID of this process.

 -- Function: setuid id
     Sets the real user ID of this process to ID.  Returns `#t' if
     successful, `#f' if not.

 -- Function: setgid id
     Sets the real group ID of this process to ID.  Returns `#t' if
     successful, `#f' if not.

 -- Function: setegid id
     Sets the effective group ID of this process to ID.  Returns `#t'
     if successful, `#f' if not.

 -- Function: seteuid id
     Sets the effective user ID of this process to ID.  Returns `#t' if
     successful, `#f' if not.

 -- Function: kill pid sig
     The `kill' function sends the signal SIGNUM to the process or
     process group specified by PID.  Besides the signals listed in
     *Note Standard Signals: (libc)Standard Signals, SIGNUM can also
     have a value of zero to check the validity of the PID.

     The PID specifies the process or process group to receive the
     signal:

    > 0
          The process whose identifier is PID.

    0
          All processes in the same process group as the sender.  The
          sender itself does not receive the signal.

    -1
          If the process is privileged, send the signal to all
          processes except for some special system processes.
          Otherwise, send the signal to all processes with the same
          effective user ID.

    < -1
          The process group whose identifier is `(abs PID)'.

     A process can send a signal to itself with `(kill (getpid)
     SIGNUM)'.  If `kill' is used by a process to send a signal to
     itself, and the signal is not blocked, then `kill' delivers at
     least one signal (which might be some other pending unblocked
     signal instead of the signal SIGNUM) to that process before it
     returns.

     The return value from `kill' is zero if the signal can be sent
     successfully.  Otherwise, no signal is sent, and a value of `-1' is
     returned.  If PID specifies sending a signal to several processes,
     `kill' succeeds if it can send the signal to at least one of them.
     There's no way you can tell which of the processes got the signal
     or whether all of them did.

 -- Function: waitpid pid options
     The `waitpid' function suspends execution of the current process
     until a child as specified by the PID argument has exited, or
     until a signal is delivered whose action is to terminate the
     current process or to call a signal handling function.  If a child
     as requested by PID has already exited by the time of the call (a
     so-called "zombie" process), the function returns immediately.
     Any system resources used by the child are freed.

     The value of PID can be:

    < -1
          which means to wait for any child process whose process group
          ID is equal to the absolute value of PID.

    -1
          which means to wait for any child process; this is the same
          behaviour which wait exhibits.

    0
          which means to wait for any child process whose process group
          ID is equal to that of the calling process.

    > 0
          which means to wait for the child whose process ID is equal
          to the value of PID.

     The value of OPTIONS is one of the following:

       0. Nothing special.

       1. (`WNOHANG') which means to return immediately if no child is
          there to be waited for.

       2. (`WUNTRACED') which means to also return for children which
          are stopped, and whose status has not been reported.

       3. Which means both of the above.

     The return value normally is the exit status of the child process,
     including the exit value along with flags indicating whether a
     coredump was generated or the child terminated as a result of a
     signal.  If the `WNOHANG' option was specified and no child
     process is waiting to be noticed, the value is zero.  A value of
     `#f' is returned in case of error and `errno' is set.  For
     information about the `errno' codes *Note Process Completion: (GNU
     C Library)Process Completion.

 -- Function: uname
     You can use the `uname' procedure to find out some information
     about the type of computer your program is running on.

     Returns a vector of strings.  These strings are:

       0. The name of the operating system in use.

       1. The network name of this particular computer.

       2. The current release level of the operating system
          implementation.

       3. The current version level within the release of the operating
          system.

       4. Description of the type of hardware that is in use.

          Some examples are `"i386-ANYTHING"', `"m68k-hp"',
          `"sparc-sun"', `"m68k-sun"', `"m68k-sony"' and `"mips-dec"'.

 -- Function: getpw name
 -- Function: getpw uid
 -- Function: getpw
     Returns a vector of information for the entry for `NAME', `UID',
     or the next entry if no argument is given.  The information is:

       0. The user's login name.

       1. The encrypted password string.

       2. The user ID number.

       3. The user's default group ID number.

       4. A string typically containing the user's real name, and
          possibly other information such as a phone number.

       5. The user's home directory, initial working directory, or
          `#f', in which case the interpretation is system-dependent.

       6. The user's default shell, the initial program run when the
          user logs in, or `#f', indicating that the system default
          should be used.

 -- Function: setpwent #t
     Rewinds the pw entry table back to the begining.

 -- Function: setpwent #f
 -- Function: setpwent
     Closes the pw table.

 -- Function: getgr name
 -- Function: getgr uid
 -- Function: getgr
     Returns a vector of information for the entry for `NAME', `UID',
     or the next entry if no argument is given.  The information is:

       0. The name of the group.

       1. The encrypted password string.

       2. The group ID number.

       3. A list of (string) names of users in the group.

 -- Function: setgrent #t
     Rewinds the group entry table back to the begining.

 -- Function: setgrent #f
 -- Function: setgrent
     Closes the group table.

 -- Function: getgroups
     Returns a vector of all the supplementary group IDs of the process.

 -- Function: link oldname newname
     The `link' function makes a new link to the existing file named by
     OLDNAME, under the new name NEWNAME.

     `link' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: chown filename owner group
     The `chown' function changes the owner of the file FILENAME to
     OWNER, and its group owner to GROUP.

     `chown' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: ttyname port
     If port PORT is associated with a terminal device, returns a
     string containing the file name of termainal device; otherwise
     `#f'.


File: scm.info,  Node: Unix Extensions,  Next: Sequence Comparison,  Prev: Posix Extensions,  Up: Packages

5.8 Unix Extensions
===================

If `'unix' is provided (by linking in `unix.o'), the following
functions are defined:

These "privileged" and symbolic link functions are not in Posix:

 -- Function: symlink oldname newname
     The `symlink' function makes a symbolic link to OLDNAME named
     NEWNAME.

     `symlink' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: readlink filename
     Returns the value of the symbolic link FILENAME or `#f' for
     failure.

 -- Function: lstat filename
     The `lstat' function is like `stat', except that it does not
     follow symbolic links.  If FILENAME is the name of a symbolic
     link, `lstat' returns information about the link itself; otherwise,
     `lstat' works like `stat'.  *Note I/O-Extensions::.

 -- Function: nice increment
     Increment the priority of the current process by INCREMENT.
     `chown' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: acct filename
     When called with the name of an exisitng file as argument,
     accounting is turned on, records for each terminating process are
     appended to FILENAME as it terminates.  An argument of `#f' causes
     accounting to be turned off.

     `acct' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: mknod filename mode dev
     The `mknod' function makes a special file with name FILENAME and
     modes MODE for device number DEV.

     `mknod' returns a value of `#t' if it is successful and `#f' on
     failure.

 -- Function: sync
     `sync' first commits inodes to buffers, and then buffers to disk.
     sync() only schedules the writes, so it may return before the
     actual writing is done.  The value returned is unspecified.


File: scm.info,  Node: Sequence Comparison,  Next: Regular Expression Pattern Matching,  Prev: Unix Extensions,  Up: Packages

5.9 Sequence Comparison
=======================

`(require 'diff)' 

A blazing fast implementation of the sequence-comparison module in
SLIB, see *Note Sequence Comparison: (slib)Sequence Comparison.


File: scm.info,  Node: Regular Expression Pattern Matching,  Next: Line Editing,  Prev: Sequence Comparison,  Up: Packages

5.10 Regular Expression Pattern Matching
========================================

These functions are defined in `rgx.c' using a POSIX or GNU "regex"
library.  If your computer does not support regex, a package is
available via ftp from `ftp.gnu.org:/pub/gnu/regex-0.12.tar.gz'.  For a
description of regular expressions, *Note syntax: (regex)syntax.

 -- Function: regcomp PATTERN [FLAGS]
     Compile a "regular expression".  Return a compiled regular
     expression, or an integer error code suitable as an argument to
     `regerror'.

     FLAGS in `regcomp' is a string of option letters used to control
     the compilation of the regular expression.  The letters may
     consist of:

    `n'
          newlines won't be matched by `.' or hat lists; ( `[^...]' )

    `i'
          ignore case.only when compiled with _GNU_SOURCE:


    `0'
          allows dot to match a null character.

    `f'
          enable GNU fastmaps.

 -- Function: regerror ERRNO
     Returns a string describing the integer ERRNO returned when
     `regcomp' fails.

 -- Function: regexec RE STRING
     Returns `#f' or a vector of integers.  These integers are in
     doublets.  The first of each doublet is the index of STRING of the
     start of the matching expression or sub-expression (delimited by
     parentheses in the pattern).  The last of each doublet is index of
     STRING of the end of that expression.  `#f' is returned if the
     string does not match.

 -- Function: regmatch? RE STRING
     Returns `#t' if the PATTERN such that REGEXP = (regcomp PATTERN)
     matches STRING as a POSIX extended regular expressions.  Returns
     `#f' otherwise.

 -- Function: regsearch RE STRING [START [LEN]]
 -- Function: regsearchv RE STRING [START [LEN]]
 -- Function: regmatch RE STRING [START [LEN]]
 -- Function: regmatchv RE STRING [START [LEN]]
     `Regsearch' searches for the pattern within the string.

     `Regmatch' anchors the pattern and begins matching it against
     string.

     `Regsearch' returns the character position where RE starts, or
     `#f' if not found.

     `Regmatch' returns the number of characters matched, `#f' if not
     matched.

     `Regsearchv' and `regmatchv' return the match vector is returned
     if RE is found, `#f' otherwise.

    RE
          may be either:
            1. a compiled regular expression returned by `regcomp';

            2. a string representing a regular expression;

            3. a list of a string and a set of option letters.

    STRING
          The string to be operated upon.

    START
          The character position at which to begin the search or match.
          If absent, the default is zero.

     _Compiled _GNU_SOURCE and using GNU libregex only_


          When searching, if START is negative, the absolute value of
          START will be used as the start location and reverse searching
          will be performed.

    LEN
          The search is allowed to examine only the first LEN
          characters of STRING.  If absent, the entire string may be
          examined.

 -- Function: string-split RE STRING
 -- Function: string-splitv RE STRING
     `String-split' splits a string into substrings that are separated
     by RE, returning a vector of substrings.

     `String-splitv' returns a vector of string positions that indicate
     where the substrings are located.

 -- Function: string-edit RE EDIT-SPEC STRING [COUNT]
     Returns the edited string.

    EDIT-SPEC
          Is a string used to replace occurances of RE.  Backquoted
          integers in the range of 1-9 may be used to insert
          subexpressions in RE, as in `sed'.

    COUNT
          The number of substitutions for `string-edit' to perform.  If
          `#t', all occurances of RE will be replaced.  The default is
          to perform one substitution.


File: scm.info,  Node: Line Editing,  Next: Curses,  Prev: Regular Expression Pattern Matching,  Up: Packages

5.11 Line Editing
=================

These procedures provide input line editing and recall.

These functions are defined in `edline.c' and `Iedline.scm' using the
"editline" or GNU "readline" (*note Overview: (readline)Top.) libraries
available from:

   * `ftp.sys.toronto.edu:/pub/rc/editline.shar'

   * `ftp.gnu.org:/pub/gnu/readline-2.0.tar.gz'

When `Iedline.scm' is loaded, if the current input port is the default
input port and the environment variable EMACS is not defined,
line-editing mode will be entered.

 -- Function: default-input-port
     Returns the initial `current-input-port' SCM was invoked with
     (stdin).

 -- Function: default-output-port
     Returns the initial `current-output-port' SCM was invoked with
     (stdout).

 -- Function: make-edited-line-port
     Returns an input/output port that allows command line editing and
     retrieval of history.

 -- Function: line-editing
     Returns the current edited line port or `#f'.

 -- Function: line-editing bool
     If BOOL is false, exits line-editing mode and returns the previous
     value of `(line-editing)'.  If BOOL is true, sets the current
     input and output ports to an edited line port and returns the
     previous value of `(line-editing)'.


File: scm.info,  Node: Curses,  Next: Sockets,  Prev: Line Editing,  Up: Packages

5.12 Curses
===========

These functions are defined in `crs.c' using the "curses" library.
Unless otherwise noted these routines return `#t' for successful
completion and `#f' for failure.

 -- Function: initscr
     Returns a port for a full screen window.  This routine must be
     called to initialize curses.

 -- Function: endwin
     A program should call `endwin' before exiting or escaping from
     curses mode temporarily, to do a system call, for example.  This
     routine will restore termio modes, move the cursor to the lower
     left corner of the screen and reset the terminal into the proper
     non-visual mode.  To resume after a temporary escape, call *Note
     refresh: Window Manipulation.

* Menu:

* Output Options Setting::
* Terminal Mode Setting::
* Window Manipulation::
* Output::
* Input::
* Curses Miscellany::


File: scm.info,  Node: Output Options Setting,  Next: Terminal Mode Setting,  Prev: Curses,  Up: Curses

5.12.1 Output Options Setting
-----------------------------

These routines set options within curses that deal with output.  All
options are initially `#f', unless otherwise stated.  It is not
necessary to turn these options off before calling `endwin'.

 -- Function: clearok win bf
     If enabled (BF is `#t'), the next call to `force-output' or
     `refresh' with WIN will clear the screen completely and redraw the
     entire screen from scratch.  This is useful when the contents of
     the screen are uncertain, or in some cases for a more pleasing
     visual effect.

 -- Function: idlok win bf
     If enabled (BF is `#t'), curses will consider using the hardware
     "insert/delete-line" feature of terminals so equipped.  If
     disabled (BF is `#f'), curses will very seldom use this feature.
     The "insert/delete-character" feature is always considered.  This
     option should be enabled only if your application needs
     "insert/delete-line", for example, for a screen editor.  It is
     disabled by default because

     "insert/delete-line" tends to be visually annoying when used in
     applications where it is not really needed.  If
     "insert/delete-line" cannot be used, curses will redraw the
     changed portions of all lines.

 -- Function: leaveok win bf
     Normally, the hardware cursor is left at the location of the window
     cursor being refreshed.  This option allows the cursor to be left
     wherever the update happens to leave it.  It is useful for
     applications where the cursor is not used, since it reduces the
     need for cursor motions.  If possible, the cursor is made
     invisible when this option is enabled.

 -- Function: scrollok win bf
     This option controls what happens when the cursor of window WIN is
     moved off the edge of the window or scrolling region, either from a
     newline on the bottom line, or typing the last character of the
     last line.  If disabled (BF is `#f'), the cursor is left on the
     bottom line at the location where the offending character was
     entered.  If enabled (BF is `#t'), `force-output' is called on the
     window WIN, and then the physical terminal and window WIN are
     scrolled up one line.

     _Note_ in order to get the physical scrolling effect on the
     terminal, it is also necessary to call `idlok'.

 -- Function: nodelay win bf
     This option causes wgetch to be a non-blocking call.  If no input
     is ready, wgetch will return an eof-object.  If disabled, wgetch
     will hang until a key is pressed.


File: scm.info,  Node: Terminal Mode Setting,  Next: Window Manipulation,  Prev: Output Options Setting,  Up: Curses

5.12.2 Terminal Mode Setting
----------------------------

These routines set options within curses that deal with input.  The
options involve using ioctl(2) and therefore interact with curses
routines.  It is not necessary to turn these options off before calling
`endwin'.  The routines in this section all return an unspecified value.

 -- Function: cbreak
 -- Function: nocbreak
     These two routines put the terminal into and out of `CBREAK' mode,
     respectively.  In `CBREAK' mode, characters typed by the user are
     immediately available to the program and erase/kill character
     processing is not performed.  When in `NOCBREAK' mode, the tty
     driver will buffer characters typed until a <LFD> or <RET> is
     typed.  Interrupt and flowcontrol characters are unaffected by
     this mode.  Initially the terminal may or may not be in `CBREAK'
     mode, as it is inherited, therefore, a program should call
     `cbreak' or `nocbreak' explicitly.  Most interactive programs
     using curses will set `CBREAK' mode.

     _Note_ `cbreak' overrides `raw'.  For a discussion of how these
     routines interact with `echo' and `noecho' *Note read-char: Input.

 -- Function: raw
 -- Function: noraw
     The terminal is placed into or out of `RAW' mode.  `RAW' mode is
     similar to `CBREAK' mode, in that characters typed are immediately
     passed through to the user program.  The differences are that in
     `RAW' mode, the interrupt, quit, suspend, and flow control
     characters are passed through uninterpreted, instead of generating
     a signal.  `RAW' mode also causes 8-bit input and output.  The
     behavior of the `BREAK' key depends on other bits in the terminal
     driver that are not set by curses.

 -- Function: echo
 -- Function: noecho
     These routines control whether characters typed by the user are
     echoed by `read-char' as they are typed.  Echoing by the tty
     driver is always disabled, but initially `read-char' is in `ECHO'
     mode, so characters typed are echoed.  Authors of most interactive
     programs prefer to do their own echoing in a controlled area of
     the screen, or not to echo at all, so they disable echoing by
     calling `noecho'.  For a discussion of how these routines interact
     with `echo' and `noecho' *Note read-char: Input.

 -- Function: nl
 -- Function: nonl
     These routines control whether <LFD> is translated into <RET> and
     `LFD' on output, and whether <RET> is translated into <LFD> on
     input.  Initially, the translations do occur.  By disabling these
     translations using `nonl', curses is able to make better use of
     the linefeed capability, resulting in faster cursor motion.

 -- Function: resetty
 -- Function: savetty
     These routines save and restore the state of the terminal modes.
     `savetty' saves the current state of the terminal in a buffer and
     `resetty' restores the state to what it was at the last call to
     `savetty'.


File: scm.info,  Node: Window Manipulation,  Next: Output,  Prev: Terminal Mode Setting,  Up: Curses

5.12.3 Window Manipulation
--------------------------

 -- Function: newwin nlines ncols begy begx
     Create and return a new window with the given number of lines (or
     rows), NLINES, and columns, NCOLS.  The upper left corner of the
     window is at line BEGY, column BEGX.  If either NLINES or NCOLS is
     0, they will be set to the value of `LINES'-BEGY and `COLS'-BEGX.
     A new full-screen window is created by calling `newwin(0,0,0,0)'.

 -- Function: subwin orig nlines ncols begy begx
     Create and return a pointer to a new window with the given number
     of lines (or rows), NLINES, and columns, NCOLS.  The window is at
     position (BEGY, BEGX) on the screen.  This position is relative to
     the screen, and not to the window ORIG.  The window is made in the
     middle of the window ORIG, so that changes made to one window will
     affect both windows.  When using this routine, often it will be
     necessary to call `touchwin' or `touchline' on ORIG before calling
     `force-output'.

 -- Function: close-port win
     Deletes the window WIN, freeing up all memory associated with it.
     In the case of sub-windows, they should be deleted before the main
     window WIN.

 -- Function: refresh
 -- Function: force-output win
     These routines are called to write output to the terminal, as most
     other routines merely manipulate data structures.  `force-output'
     copies the window WIN to the physical terminal screen, taking into
     account what is already there in order to minimize the amount of
     information that's sent to the terminal (called optimization).
     Unless `leaveok' has been enabled, the physical cursor of the
     terminal is left at the location of window WIN's cursor.  With
     `refresh', the number of characters output to the terminal is
     returned.

 -- Function: mvwin win y x
     Move the window WIN so that the upper left corner will be at
     position (Y, X).  If the move would cause the window WIN to be off
     the screen, it is an error and the window WIN is not moved.

 -- Function: overlay srcwin dstwin
 -- Function: overwrite srcwin dstwin
     These routines overlay SRCWIN on top of DSTWIN; that is, all text
     in SRCWIN is copied into DSTWIN.  SRCWIN and DSTWIN need not be
     the same size; only text where the two windows overlap is copied.
     The difference is that `overlay' is non-destructive (blanks are
     not copied), while `overwrite' is destructive.

 -- Function: touchwin win
 -- Function: touchline win start count
     Throw away all optimization information about which parts of the
     window WIN have been touched, by pretending that the entire window
     WIN has been drawn on.  This is sometimes necessary when using
     overlapping windows, since a change to one window will affect the
     other window, but the records of which lines have been changed in
     the other window will not reflect the change.  `touchline' only
     pretends that COUNT lines have been changed, beginning with line
     START.

 -- Function: wmove win y x
     The cursor associated with the window WIN is moved to line (row) Y,
     column X.  This does not move the physical cursor of the terminal
     until `refresh' (or `force-output') is called.  The position
     specified is relative to the upper left corner of the window WIN,
     which is (0, 0).


File: scm.info,  Node: Output,  Next: Input,  Prev: Window Manipulation,  Up: Curses

5.12.4 Output
-------------

These routines are used to "draw" text on windows

 -- Function: display ch win
 -- Function: display str win
 -- Function: wadd win ch
 -- Function: wadd win str
     The character CH or characters in STR are put into the window WIN
     at the current cursor position of the window and the position of
     WIN's cursor is advanced.  At the right margin, an automatic
     newline is performed.  At the bottom of the scrolling region, if
     scrollok is enabled, the scrolling region will be scrolled up one
     line.

     If CH is a <TAB>, <LFD>, or backspace, the cursor will be moved
     appropriately within the window WIN.  A <LFD> also does a
     `wclrtoeol' before moving.  <TAB> characters are considered to be
     at every eighth column.  If CH is another control character, it
     will be drawn in the `C-x' notation.  (Calling `winch' after
     adding a control character will not return the control character,
     but instead will return the representation of the control
     character.)

     Video attributes can be combined with a character by or-ing them
     into the parameter.  This will result in these attributes also
     being set.  The intent here is that text, including attributes,
     can be copied from one place to another using inch and display.
     See `standout', below.

     _Note_ For `wadd' CH can be an integer and will insert the
     character of the corresponding value.

 -- Function: werase win
     This routine copies blanks to every position in the window WIN.

 -- Function: wclear win
     This routine is like `werase', but it also calls *Note clearok:
     Output Options Setting, arranging that the screen will be cleared
     completely on the next call to `refresh' or `force-output' for
     window WIN, and repainted from scratch.

 -- Function: wclrtobot win
     All lines below the cursor in window WIN are erased.  Also, the
     current line to the right of the cursor, inclusive, is erased.

 -- Function: wclrtoeol win
     The current line to the right of the cursor, inclusive, is erased.

 -- Function: wdelch win
     The character under the cursor in the window WIN is deleted.  All
     characters to the right on the same line are moved to the left one
     position and the last character on the line is filled with a
     blank.  The cursor position does not change.  This does not imply
     use of the hardware "delete-character" feature.

 -- Function: wdeleteln win
     The line under the cursor in the window WIN is deleted.  All lines
     below the current line are moved up one line.  The bottom line WIN
     is cleared.  The cursor position does not change.  This does not
     imply use of the hardware "deleteline" feature.

 -- Function: winsch win ch
     The character CH is inserted before the character under the
     cursor.  All characters to the right are moved one <SPC> to the
     right, possibly losing the rightmost character of the line.  The
     cursor position does not change .  This does not imply use of the
     hardware "insertcharacter" feature.

 -- Function: winsertln win
     A blank line is inserted above the current line and the bottom
     line is lost.  This does not imply use of the hardware
     "insert-line" feature.

 -- Function: scroll win
     The window WIN is scrolled up one line.  This involves moving the
     lines in WIN's data structure.  As an optimization, if WIN is
     stdscr and the scrolling region is the entire window, the physical
     screen will be scrolled at the same time.


File: scm.info,  Node: Input,  Next: Curses Miscellany,  Prev: Output,  Up: Curses

5.12.5 Input
------------

 -- Function: read-char win
     A character is read from the terminal associated with the window
     WIN.  Depending on the setting of `cbreak', this will be after one
     character (`CBREAK' mode), or after the first newline (`NOCBREAK'
     mode).  Unless `noecho' has been set, the character will also be
     echoed into WIN.

     When using `read-char', do not set both `NOCBREAK' mode
     (`nocbreak') and `ECHO' mode (`echo') at the same time.  Depending
     on the state of the terminal driver when each character is typed,
     the program may produce undesirable results.

 -- Function: winch win
     The character, of type chtype, at the current position in window
     WIN is returned.  If any attributes are set for that position,
     their values will be OR'ed into the value returned.

 -- Function: getyx win
     A list of the y and x coordinates of the cursor position of the
     window WIN is returned


File: scm.info,  Node: Curses Miscellany,  Prev: Input,  Up: Curses

5.12.6 Curses Miscellany
------------------------

 -- Function: wstandout win
 -- Function: wstandend win
     These functions set the current attributes of the window WIN.  The
     current attributes of WIN are applied to all characters that are
     written into it.  Attributes are a property of the character, and
     move with the character through any scrolling and insert/delete
     line/character operations.  To the extent possible on the
     particular terminal, they will be displayed as the graphic
     rendition of characters put on the screen.

     `wstandout' sets the current attributes of the window WIN to be
     visibly different from other text.  `wstandend' turns off the
     attributes.

 -- Function: box win vertch horch
     A box is drawn around the edge of the window WIN.  VERTCH and
     HORCH are the characters the box is to be drawn with.  If VERTCH
     and HORCH are 0, then appropriate default characters, `ACS_VLINE'
     and `ACS_HLINE', will be used.

     _Note_ VERTCH and HORCH can be an integers and will insert the
     character (with attributes) of the corresponding values.

 -- Function: unctrl c
     This macro expands to a character string which is a printable
     representation of the character C.  Control characters are
     displayed in the `C-x' notation.  Printing characters are displayed
     as is.


File: scm.info,  Node: Sockets,  Next: SCMDB,  Prev: Curses,  Up: Packages

5.13 Sockets
============

These procedures (defined in `socket.c') provide a Scheme interface to
most of the C "socket" library.  For more information on sockets, *Note
Sockets: (libc)Sockets.

* Menu:

* Host and Other Inquiries::
* Internet Addresses and Socket Names::
* Socket::


File: scm.info,  Node: Host and Other Inquiries,  Next: Internet Addresses and Socket Names,  Prev: Sockets,  Up: Sockets

5.13.1 Host and Other Inquiries
-------------------------------

 -- Constant: af_inet
 -- Constant: af_unix
     Integer family codes for Internet and Unix sockets, respectively.

 -- Function: gethost host-spec
 -- Function: gethost
     Returns a vector of information for the entry for `HOST-SPEC' or
     the next entry if `HOST-SPEC' isn't given.  The information is:

       0. host name string

       1. list of host aliases strings

       2. integer address type (`AF_INET')

       3. integer size of address entries (in bytes)

       4. list of integer addresses

 -- Function: sethostent stay-open
 -- Function: sethostent
     Rewinds the host entry table back to the begining if given an
     argument.  If the argument STAY-OPEN is `#f' queries will be be
     done using `UDP' datagrams.  Otherwise, a connected `TCP' socket
     will be used.  When called without an argument, the host table is
     closed.

 -- Function: getnet name-or-number
 -- Function: getnet
     Returns a vector of information for the entry for NAME-OR-NUMBER or
     the next entry if an argument isn't given.  The information is:

       0. official network name string

       1. list of network aliases strings

       2. integer network address type (`AF_INET')

       3. integer network number

 -- Function: setnetent stay-open
 -- Function: setnetent
     Rewinds the network entry table back to the begining if given an
     argument.  If the argument STAY-OPEN is `#f' the table will be
     closed between calls to getnet.  Otherwise, the table stays open.
     When called without an argument, the network table is closed.

 -- Function: getproto name-or-number
 -- Function: getproto
     Returns a vector of information for the entry for NAME-OR-NUMBER or
     the next entry if an argument isn't given.  The information is:

       1. official protocol name string

       2. list of protocol aliases strings

       3. integer protocol number

 -- Function: setprotoent stay-open
 -- Function: setprotoent
     Rewinds the protocol entry table back to the begining if given an
     argument.  If the argument STAY-OPEN is `#f' the table will be
     closed between calls to getproto.  Otherwise, the table stays
     open.  When called without an argument, the protocol table is
     closed.

 -- Function: getserv name-or-port-number protocol
 -- Function: getserv
     Returns a vector of information for the entry for
     NAME-OR-PORT-NUMBER and PROTOCOL or the next entry if arguments
     aren't given.  The information is:

       0. official service name string

       1. list of service aliases strings

       2. integer port number

       3. protocol

 -- Function: setservent stay-open
 -- Function: setservent
     Rewinds the service entry table back to the begining if given an
     argument.  If the argument STAY-OPEN is `#f' the table will be
     closed between calls to getserv.  Otherwise, the table stays open.
     When called without an argument, the service table is closed.


File: scm.info,  Node: Internet Addresses and Socket Names,  Next: Socket,  Prev: Host and Other Inquiries,  Up: Sockets

5.13.2 Internet Addresses and Socket Names
------------------------------------------

 -- Function: inet:string->address string
     Returns the host address number (integer) for host STRING or `#f'
     if not found.

 -- Function: inet:address->string address
     Converts an internet (integer) address to a string in numbers and
     dots notation.

 -- Function: inet:network address
     Returns the network number (integer) specified from ADDRESS or
     `#f' if not found.

 -- Function: inet:local-network-address address
     Returns the integer for the address of ADDRESS within its local
     network or `#f' if not found.

 -- Function: inet:make-address network local-address
     Returns the Internet address of LOCAL-ADDRESS in NETWORK.

The type "socket-name" is used for inquiries about open sockets in the
following procedures:

 -- Function: getsockname socket
     Returns the socket-name of SOCKET.  Returns `#f' if unsuccessful
     or SOCKET is closed.

 -- Function: getpeername socket
     Returns the socket-name of the socket connected to SOCKET.
     Returns `#f' if unsuccessful or SOCKET is closed.

 -- Function: socket-name:family socket-name
     Returns the integer code for the family of SOCKET-NAME.

 -- Function: socket-name:port-number socket-name
     Returns the integer port number of SOCKET-NAME.

 -- Function: socket-name:address socket-name
     Returns the integer Internet address for SOCKET-NAME.


File: scm.info,  Node: Socket,  Prev: Internet Addresses and Socket Names,  Up: Sockets

5.13.3 Socket
-------------

When a port is returned from one of these calls it is unbuffered.  This
allows both reading and writing to the same port to work.  If you want
buffered ports you can (assuming sock-port is a socket i/o port):
     (require 'i/o-extensions)
     (define i-port (duplicate-port sock-port "r"))
     (define o-port (duplicate-port sock-port "w"))

 -- Function: make-stream-socket family
 -- Function: make-stream-socket family protocol
     Returns a `SOCK_STREAM' socket of type FAMILY using PROTOCOL.  If
     FAMILY has the value `AF_INET', `SO_REUSEADDR' will be set.  The
     integer argument PROTOCOL corresponds to the integer protocol
     numbers returned (as vector elements) from `(getproto)'.  If the
     PROTOCOL argument is not supplied, the default (0) for the
     specified FAMILY is used.  SCM sockets look like ports opened for
     neither reading nor writing.

 -- Function: make-stream-socketpair family
 -- Function: make-stream-socketpair family protocol
     Returns a pair (cons) of connected `SOCK_STREAM' (socket) ports of
     type FAMILY using PROTOCOL.  Many systems support only socketpairs
     of the `af-unix' FAMILY.  The integer argument PROTOCOL
     corresponds to the integer protocol numbers returned (as vector
     elements) from (getproto).  If the PROTOCOL argument is not
     supplied, the default (0) for the specified FAMILY is used.

 -- Function: socket:shutdown socket how
     Makes SOCKET no longer respond to some or all operations depending
     on the integer argument HOW:

       0. Further input is disallowed.

       1. Further output is disallowed.

       2. Further input or output is disallowed.

     `Socket:shutdown' returns SOCKET if successful, `#f' if not.

 -- Function: socket:connect inet-socket host-number port-number
 -- Function: socket:connect unix-socket pathname
     Returns SOCKET (changed to a read/write port) connected to the
     Internet socket on host HOST-NUMBER, port PORT-NUMBER or the Unix
     socket specified by PATHNAME.  Returns `#f' if not successful.

 -- Function: socket:bind inet-socket port-number
 -- Function: socket:bind unix-socket pathname
     Returns INET-SOCKET bound to the integer PORT-NUMBER or the
     UNIX-SOCKET bound to new socket in the file system at location
     PATHNAME.  Returns `#f' if not successful.  Binding a UNIX-SOCKET
     creates a socket in the file system that must be deleted by the
     caller when it is no longer needed (using `delete-file').

 -- Function: socket:listen socket backlog
     The bound (*note bind: Socket.) SOCKET is readied to accept
     connections.  The positive integer BACKLOG specifies how many
     pending connections will be allowed before further connection
     requests are refused.  Returns SOCKET (changed to a read-only
     port) if successful, `#f' if not.

 -- Function: char-ready? listen-socket
     The input port returned by a successful call to `socket:listen' can
     be polled for connections by `char-ready?' (*note char-ready?:
     Files and Ports.).  This avoids blocking on connections by
     `socket:accept'.

 -- Function: socket:accept socket
     Accepts a connection on a bound, listening SOCKET.  Returns an
     input/output port for the connection.

The following example is not too complicated, yet shows the use of
sockets for multiple connections without input blocking.

     ;;;; Scheme chat server

     ;;; This program implements a simple `chat' server which accepts
     ;;; connections from multiple clients, and sends to all clients any
     ;;; characters received from any client.

     ;;; To connect to chat `telnet localhost 8001'

     (require 'socket)
     (require 'i/o-extensions)
     
     (let ((listener-socket (socket:bind (make-stream-socket af_inet) 8001))
           (connections '()))
       (socket:listen listener-socket 5)
       (do () (#f)
         (let ((actives (or (apply wait-for-input 5 listener-socket connections)
                            '())))
           (cond ((null? actives))
                 ((memq listener-socket actives)
                  (set! actives (cdr (memq listener-socket actives)))
                  (let ((con (socket:accept listener-socket)))
                    (display "accepting connection from ")
                    (display (getpeername con))
                    (newline)
                    (set! connections (cons con connections))
                    (display "connected" con)
                    (newline con))))
           (set! connections
                 (let next ((con-list connections))
                   (cond ((null? con-list) '())
                         (else
                          (let ((con (car con-list)))
                            (cond ((memq con actives)
                                   (let ((c (read-char con)))
                                     (cond ((eof-object? c)
                                            (display "closing connection from ")
                                            (display (getpeername con))
                                            (newline)
                                            (close-port con)
                                            (next (cdr con-list)))
                                           (else
                                            (for-each (lambda (con)
                                                        (file-set-position con 0)
                                                        (write-char c con)
                                                        (file-set-position con 0))
                                                      connections)
                                            (cons con (next (cdr con-list)))))))
                                  (else (cons con (next (cdr con-list)))))))))))))

You can use `telnet localhost 8001' to connect to the chat server, or
you can use a client written in scheme:

     ;;;; Scheme chat client

     ;;; this program connects to socket 8001.  It then sends all
     ;;; characters from current-input-port to the socket and sends all
     ;;; characters from the socket to current-output-port.

     (require 'socket)
     (require 'i/o-extensions)
     
     (define con (make-stream-socket af_inet))
     (set! con (socket:connect con (inet:string->address "localhost") 8001))

     (define (go)
       (define actives (wait-for-input (* 30 60) con (current-input-port)))
       (let ((cs (and actives (memq con actives) (read-char con)))
             (ct (and actives (memq (current-input-port) actives) (read-char))))
         (cond ((or (eof-object? cs) (eof-object? ct)) (close-port con))
               (else (cond (cs (display cs)))
                     (cond (ct (file-set-position con 0)
                               (display ct con)
                               (file-set-position con 0)))
                     (go)))))
     (cond (con (display "Connecting to ")
                (display (getpeername con))
                (newline)
                (go))
           (else (display "Server not listening on port 8001")
                 (newline)))


File: scm.info,  Node: SCMDB,  Prev: Sockets,  Up: Packages

5.14 SCMDB
==========

`(require 'mysql)'

"SCMDB" is an add-on for SCM that ports the MySQL C-library to SCM.

It is available from: `http://www.dedecker.net/jessie/scmdb/'


File: scm.info,  Node: The Implementation,  Next: Index,  Prev: Packages,  Up: Top

6 The Implementation
********************

* Menu:

* Data Types::
* Operations::
* Program Self-Knowledge::      What SCM needs to know about itself.
* Improvements To Make::


File: scm.info,  Node: Data Types,  Next: Operations,  Prev: The Implementation,  Up: The Implementation

6.1 Data Types
==============

In the descriptions below it is assumed that `long int's are 32 bits in
length.  Acutally, SCM is written to work with any `long int' size
larger than 31 bits.  With some modification, SCM could work with word
sizes as small as 24 bits.

All SCM objects are represented by type "SCM".  Type `SCM' come in 2
basic flavors, Immediates and Cells:

* Menu:

* Immediates::
* Cells::                       Non-Immediate types
* Header Cells::                Malloc objects
* Subr Cells::                  Built-in and Compiled Procedures
* Ptob Cells::                  I/O ports
* Smob Cells::                  Miscellaneous datatypes
* Data Type Representations::   How they all fit together


File: scm.info,  Node: Immediates,  Next: Cells,  Prev: Data Types,  Up: Data Types

6.1.1 Immediates
----------------

An "immediate" is a data type contained in type `SCM' (`long int').
The type codes distinguishing immediate types from each other vary in
length, but reside in the low order bits.

 -- Macro: IMP x
 -- Macro: NIMP x
     Return non-zero if the `SCM' object X is an immediate or
     non-immediate type, respectively.

 -- Immediate: inum
     immediate 30 bit signed integer.  An INUM is flagged by a `1' in
     the second to low order bit position.  The high order 30 bits are
     used for the integer's value.

      -- Macro: INUMP x
      -- Macro: NINUMP x
          Return non-zero if the `SCM' X is an immediate integer or not
          an immediate integer, respectively.

      -- Macro: INUM x
          Returns the C `long integer' corresponding to `SCM' X.

      -- Macro: MAKINUM x
          Returns the `SCM' inum corresponding to C `long integer' x.

      -- Immediate Constant: INUM0
          is equivalent to `MAKINUM(0)'.

     Computations on INUMs are performed by converting the arguments to
     C integers (by a shift), operating on the integers, and converting
     the result to an inum.  The result is checked for overflow by
     converting back to integer and checking the reverse operation.

     The shifts used for conversion need to be signed shifts.  If the C
     implementation does not support signed right shift this fact is
     detected in a #if statement in `scmfig.h' and a signed right shift,
     `SRS', is constructed in terms of unsigned right shift.

 -- Immediate: ichr
     characters.

      -- Macro: ICHRP x
          Return non-zero if the `SCM' object X is a character.

      -- Macro: ICHR x
          Returns corresponding `unsigned char'.

      -- Macro: MAKICHR x
          Given `char' X, returns `SCM' character.


 -- Immediate: iflags
     These are frequently used immediate constants.

      -- Immediate Constant: SCM BOOL_T
          `#t'

      -- Immediate Constant: SCM BOOL_F
          `#f'

      -- Immediate Constant: SCM EOL
          `()'.  If `SICP' is `#define'd, `EOL' is `#define'd to be
          identical with `BOOL_F'.  In this case, both print as `#f'.

      -- Immediate Constant: SCM EOF_VAL
          end of file token, `#<eof>'.

      -- Immediate Constant: SCM UNDEFINED
          `#<undefined>' used for variables which have not been defined
          and absent optional arguments.

      -- Immediate Constant: SCM UNSPECIFIED
          `#<unspecified>' is returned for those procedures whose return
          values are not specified.


 -- Macro: IFLAGP n
     Returns non-zero if N is an ispcsym, isym or iflag.

 -- Macro: ISYMP n
     Returns non-zero if N is an ispcsym or isym.

 -- Macro: ISYMNUM n
     Given ispcsym, isym, or iflag N, returns its index in the C array
     `isymnames[]'.

 -- Macro: ISYMCHARS n
     Given ispcsym, isym, or iflag N, returns its `char *'
     representation (from `isymnames[]').

 -- Macro: MAKSPCSYM n
     Returns `SCM' ispcsym N.

 -- Macro: MAKISYM n
     Returns `SCM' iisym N.

 -- Macro: MAKIFLAG n
     Returns `SCM' iflag N.

 -- Variable: isymnames
     An array of strings containing the external representations of all
     the ispcsym, isym, and iflag immediates.  Defined in `repl.c'.

 -- Constant: NUM_ISPCSYM
 -- Constant: NUM_ISYMS
     The number of ispcsyms and ispcsyms+isyms, respectively.  Defined
     in `scm.h'.

 -- Immediate: isym
     `and', `begin', `case', `cond', `define', `do', `if', `lambda',
     `let', `let*', `letrec', `or', `quote', `set!', `#f', `#t',
     `#<undefined>', `#<eof>', `()', and `#<unspecified>'.

 -- CAR Immediate: ispcsym
     special symbols: syntax-checked versions of first 14 isyms

 -- CAR Immediate: iloc
     indexes to a variable's location in environment

 -- CAR Immediate: gloc
     pointer to a symbol's value cell

 -- Immediate: CELLPTR
     pointer to a cell (not really an immediate type, but here for
     completeness).  Since cells are always 8 byte aligned, a pointer
     to a cell has the low order 3 bits `0'.

     There is one exception to this rule, _CAR Immediate_s, described
     next.

A "CAR Immediate" is an Immediate point which can only occur in the
`CAR's of evaluated code (as a result of `ceval''s memoization process).


File: scm.info,  Node: Cells,  Next: Header Cells,  Prev: Immediates,  Up: Data Types

6.1.2 Cells
-----------

"Cell"s represent all SCM objects other than immediates.  A cell has a
`CAR' and a `CDR'.  Low-order bits in `CAR' identify the type of
object.  The rest of `CAR' and `CDR' hold object data.  The number
after `tc' specifies how many bits are in the type code.  For instance,
`tc7' indicates that the type code is 7 bits.

 -- Macro: NEWCELL x
     Allocates a new cell and stores a pointer to it in `SCM' local
     variable X.

     Care needs to be taken that stores into the new cell pointed to by
     X do not create an inconsistent object.  *Note Signals::.

All of the C macros decribed in this section assume that their argument
is of type `SCM' and points to a cell (`CELLPTR').

 -- Macro: CAR x
 -- Macro: CDR x
     Returns the `car' and `cdr' of cell X, respectively.

 -- Macro: TYP3 x
 -- Macro: TYP7 x
 -- Macro: TYP16 x
     Returns the 3, 7, and 16 bit type code of a cell.

 -- Cell: tc3_cons
     scheme cons-cell returned by (cons arg1 arg2).

      -- Macro: CONSP x
      -- Macro: NCONSP x
          Returns non-zero if X is a `tc3_cons' or isn't, respectively.

 -- Cell: tc3_closure
     applicable object returned by (lambda (args) ...).  `tc3_closure's
     have a pointer to the body of the procedure in the `CAR' and a
     pointer to the environment in the `CDR'.  Bits 1 and 2
     (zero-based) in the `CDR' indicate a lower bound on the number of
     required arguments to the closure, which is used to avoid
     allocating rest argument lists in the environment cache.  This
     encoding precludes an immediate value for the `CDR':  In the case
     of an empty environment all bits above 2 in the `CDR' are zero.

      -- Macro: CLOSUREP x
          Returns non-zero if X is a `tc3_closure'.

      -- Macro: CODE x
      -- Macro: ENV x
          Returns the code body or environment of closure X,
          respectively.

      -- Macro: ARGC x
          Returns the a lower bound on the number of required arguments
          to closure X, it cannot exceed 3.



File: scm.info,  Node: Header Cells,  Next: Subr Cells,  Prev: Cells,  Up: Data Types

6.1.3 Header Cells
------------------

"Header"s are Cells whose `CDR's point elsewhere in memory, such as to
memory allocated by `malloc'.

 -- Header: spare
     spare `tc7' type code

 -- Header: tc7_vector
     scheme vector.

      -- Macro: VECTORP x
      -- Macro: NVECTORP x
          Returns non-zero if X is a `tc7_vector' or if not,
          respectively.

      -- Macro: VELTS x
      -- Macro: LENGTH x
          Returns the C array of `SCM's holding the elements of vector
          X or its length, respectively.

 -- Header: tc7_ssymbol
     static scheme symbol (part of initial system)

 -- Header: tc7_msymbol
     `malloc'ed scheme symbol (can be GCed)

      -- Macro: SYMBOLP x
          Returns non-zero if X is a `tc7_ssymbol' or `tc7_msymbol'.

      -- Macro: CHARS x
      -- Macro: UCHARS x
      -- Macro: LENGTH x
          Returns the C array of `char's or as `unsigned char's holding
          the elements of symbol X or its length, respectively.

 -- Header: tc7_string
     scheme string

      -- Macro: STRINGP x
      -- Macro: NSTRINGP x
          Returns non-zero if X is a `tc7_string' or isn't,
          respectively.

      -- Macro: CHARS x
      -- Macro: UCHARS x
      -- Macro: LENGTH x
          Returns the C array of `char's or as `unsigned char's holding
          the elements of string X or its length, respectively.

 -- Header: tc7_bvect
     uniform vector of booleans (bit-vector)

 -- Header: tc7_ivect
     uniform vector of integers

 -- Header: tc7_uvect
     uniform vector of non-negative integers

 -- Header: tc7_svect
     uniform vector of short integers

 -- Header: tc7_fvect
     uniform vector of short inexact real numbers

 -- Header: tc7_dvect
     uniform vector of double precision inexact real numbers

 -- Header: tc7_cvect
     uniform vector of double precision inexact complex numbers

 -- Header: tc7_contin
     applicable object produced by call-with-current-continuation

 -- Header: tc7_specfun
     subr that is treated specially within the evaluator

     `apply' and `call-with-current-continuation' are denoted by these
     objects.  Their behavior as functions is built into the evaluator;
     they are not directly associated with C functions.  This is
     necessary in order to make them properly tail recursive.

     tc16_cclo is a subtype of tc7_specfun, a cclo is similar to a
     vector (and is GCed like one), but can be applied as a function:

       1. the cclo itself is consed onto the head of the argument list

       2. the first element of the cclo is applied to that list.  Cclo
          invocation is currently not tail recursive when given 2 or
          more arguments.

      -- Function: makcclo proc len
          makes a closure from the _subr_ PROC with LEN-1 extra
          locations for `SCM' data.  Elements of a CCLO are referenced
          using `VELTS(cclo)[n]' just as for vectors.

      -- Macro: CCLO_LENGTH cclo
          Expands to the length of CCLO.


File: scm.info,  Node: Subr Cells,  Next: Ptob Cells,  Prev: Header Cells,  Up: Data Types

6.1.4 Subr Cells
----------------

A "Subr" is a header whose `CDR' points to a C code procedure.  Scheme
primitive procedures are subrs.  Except for the arithmetic `tc7_cxr's,
the C code procedures will be passed arguments (and return results) of
type `SCM'.

 -- Subr: tc7_asubr
     associative C function of 2 arguments.  Examples are `+', `-',
     `*', `/', `max', and `min'.

 -- Subr: tc7_subr_0
     C function of no arguments.

 -- Subr: tc7_subr_1
     C function of one argument.

 -- Subr: tc7_cxr
     These subrs are handled specially.  If inexact numbers are
     enabled, the `CDR' should be a function which takes and returns
     type `double'.  Conversions are handled in the interpreter.

     `floor', `ceiling', `truncate', `round', `$sqrt', `$abs', `$exp',
     `$log', `$sin', `$cos', `$tan', `$asin', `$acos', `$atan',
     `$sinh', `$cosh', `$tanh', `$asinh', `$acosh', `$atanh', and
     `exact->inexact' are defined this way.

     If the `CDR' is `0' (`NULL'), the name string of the procedure is
     used to control traversal of its list structure argument.

     `car', `cdr', `caar', `cadr', `cdar', `cddr', `caaar', `caadr',
     `cadar', `caddr', `cdaar', `cdadr', `cddar', `cdddr', `caaaar',
     `caaadr', `caadar', `caaddr', `cadaar', `cadadr', `caddar',
     `cadddr', `cdaaar', `cdaadr', `cdadar', `cdaddr', `cddaar',
     `cddadr', `cdddar', and `cddddr' are defined this way.

 -- Subr: tc7_subr_3
     C function of 3 arguments.

 -- Subr: tc7_subr_2
     C function of 2 arguments.

 -- Subr: tc7_rpsubr
     transitive relational predicate C function of 2 arguments.  The C
     function should return either `BOOL_T' or `BOOL_F'.

 -- Subr: tc7_subr_1o
     C function of one optional argument.  If the optional argument is
     not present, `UNDEFINED' is passed in its place.

 -- Subr: tc7_subr_2o
     C function of 1 required and 1 optional argument.  If the optional
     argument is not present, `UNDEFINED' is passed in its place.

 -- Subr: tc7_lsubr_2
     C function of 2 arguments and a list of (rest of) `SCM' arguments.

 -- Subr: tc7_lsubr
     C function of list of `SCM' arguments.


File: scm.info,  Node: Ptob Cells,  Next: Smob Cells,  Prev: Subr Cells,  Up: Data Types

6.1.5 Ptob Cells
----------------

A "ptob" is a port object, capable of delivering or accepting characters.
*Note Ports: (r5rs)Ports.  Unlike the types described so far, new
varieties of ptobs can be defined dynamically (*note Defining Ptobs::).
These are the initial ptobs:

 -- ptob: tc16_inport
     input port.

 -- ptob: tc16_outport
     output port.

 -- ptob: tc16_ioport
     input-output port.

 -- ptob: tc16_inpipe
     input pipe created by `popen()'.

 -- ptob: tc16_outpipe
     output pipe created by `popen()'.

 -- ptob: tc16_strport
     String port created by `cwos()' or `cwis()'.

 -- ptob: tc16_sfport
     Software (virtual) port created by `mksfpt()' (*note Soft Ports::).

 -- Macro: PORTP x
 -- Macro: OPPORTP x
 -- Macro: OPINPORTP x
 -- Macro: OPOUTPORTP x
 -- Macro: INPORTP x
 -- Macro: OUTPORTP x
     Returns non-zero if X is a port, open port, open input-port, open
     output-port, input-port, or output-port, respectively.

 -- Macro: OPENP x
 -- Macro: CLOSEDP x
     Returns non-zero if port X is open or closed, respectively.

 -- Macro: STREAM x
     Returns the `FILE *' stream for port X.

Ports which are particularly well behaved are called "fport"s.
Advanced operations like `file-position' and `reopen-file' only work
for fports.

 -- Macro: FPORTP x
 -- Macro: OPFPORTP x
 -- Macro: OPINFPORTP x
 -- Macro: OPOUTFPORTP x
     Returns non-zero if X is a port, open port, open input-port, or
     open output-port, respectively.


File: scm.info,  Node: Smob Cells,  Next: Data Type Representations,  Prev: Ptob Cells,  Up: Data Types

6.1.6 Smob Cells
----------------

A "smob" is a miscellaneous datatype.  The type code and GCMARK bit occupy
the lower order 16 bits of the `CAR' half of the cell.  The rest of the
`CAR' can be used for sub-type or other information.  The `CDR'
contains data of size long and is often a pointer to allocated memory.

Like ptobs, new varieties of smobs can be defined dynamically (*note
Defining Smobs::).  These are the initial smobs:

 -- smob: tc_free_cell
     unused cell on the freelist.

 -- smob: tc16_flo
     single-precision float.

     Inexact number data types are subtypes of type `tc16_flo'.  If the
     sub-type is:

       0. a single precision float is contained in the `CDR'.

       1. `CDR' is a pointer to a `malloc'ed double.

       3. `CDR' is a pointer to a `malloc'ed pair of doubles.

      -- smob: tc_dblr
          double-precision float.

      -- smob: tc_dblc
          double-precision complex.

 -- smob: tc16_bigpos
 -- smob: tc16_bigneg
     positive and negative bignums, respectively.

     Scm has large precision integers called bignums.  They are stored
     in sign-magnitude form with the sign occuring in the type code of
     the SMOBs bigpos and bigneg.  The magnitude is stored as a
     `malloc'ed array of type `BIGDIG' which must be an unsigned
     integral type with size smaller than `long'.  `BIGRAD' is the
     radix associated with `BIGDIG'.

     `NUMDIGS_MAX' (defined in `scmfig.h') limits the number of digits
     of a bignum to 1000.  These digits are base `BIGRAD', which is
     typically 65536, giving 4816 decimal digits.

     Why only 4800 digits?  The simple multiplication algorithm SCM
     uses is O(n^2); this means the number of processor instructions
     required to perform a multiplication is _some multiple_ of the
     product of the number of digits of the two multiplicands.

          digits * digits  ==> operations
           5                    x
           50                   100 * x
           500                  10000 * x
           5000                 1000000 * x

     To calculate numbers larger than this, FFT multiplication
     [O(n*log(n))] and other specialized algorithms are required.  You
     should obtain a package which specializes in number-theoretical
     calculations:

                `ftp://megrez.math.u-bordeaux.fr/pub/pari/'

 -- smob: tc16_promise
     made by DELAY.  *Note Control features: (r5rs)Control features.

 -- smob: tc16_arbiter
     synchronization object.  *Note Process Synchronization::.

 -- smob: tc16_macro
     macro expanding function.  *Note Macro Primitives::.

 -- smob: tc16_array
     multi-dimensional array.  *Note Arrays::.

     This type implements both conventional arrays (those with
     arbitrary data as elements *note Conventional Arrays::) and
     uniform arrays (those with elements of a uniform type *note
     Uniform Array::).

     Conventional Arrays have a pointer to a vector for their `CDR'.
     Uniform Arrays have a pointer to a Uniform Vector type (string,
     bvect, ivect, uvect, fvect, dvect, or cvect) in their `CDR'.


File: scm.info,  Node: Data Type Representations,  Prev: Smob Cells,  Up: Data Types

6.1.7 Data Type Representations
-------------------------------

IMMEDIATE:      B,D,E,F=data bit, C=flag code, P=pointer address bit
        ................................
inum    BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB10
ichr    BBBBBBBBBBBBBBBBBBBBBBBB11110100
iflag                   CCCCCCC101110100
isym                    CCCCCCC001110100
        IMCAR:  only in car of evaluated code, cdr has cell's GC bit
ispcsym                 000CCCC00CCCC100
iloc    0DDDDDDDDDDDEFFFFFFFFFFF11111100
pointer PPPPPPPPPPPPPPPPPPPPPPPPPPPPP000
gloc    PPPPPPPPPPPPPPPPPPPPPPPPPPPPP001

   HEAP CELL:   G=gc_mark; 1 during mark, 0 other times.
        1s and 0s here indicate type.     G missing means sys (not GC'd)
        SIMPLE
cons    ..........SCM car..............0  ...........SCM cdr.............G
closure ..........SCM code...........011  ...........SCM env...........CCG
        HEADERs:
ssymbol .........long length....G0000101  ..........char *chars...........
msymbol .........long length....G0000111  ..........char *chars...........
string  .........long length....G0001101  ..........char *chars...........
vector  .........long length....G0001111  ...........SCM **elts...........
bvect   .........long length....G0010101  ..........long *words...........
 spare                          G0010111
ivect   .........long length....G0011101  ..........long *words...........
uvect   .........long length....G0011111  ......unsigned long *words......
 spare                          G0100101
svect   .........long length....G0100111  ........ short *words...........
fvect   .........long length....G0101101  .........float *words...........
dvect   .........long length....G0101111  ........double *words...........
cvect   .........long length....G0110101  ........double *words...........

contin  .........long length....G0111101  .............*regs..............
specfun ................xxxxxxxxG1111111  ...........SCM name.............
cclo    ..short length..xxxxxx10G1111111  ...........SCM **elts...........
                        PTOBs
   port int portnum.CwroxxxxxxxxG0110111  ..........FILE *stream..........
 socket int portnum.C001xxxxxxxxG0110111  ..........FILE *stream..........
 inport int portnum.C011xxxxxxxxG0110111  ..........FILE *stream..........
outport int portnum.0101xxxxxxxxG0110111  ..........FILE *stream..........
 ioport int portnum.C111xxxxxxxxG0110111  ..........FILE *stream..........
fport   int portnum.C   00000000G0110111  ..........FILE *stream..........
pipe    int portnum.C   00000001G0110111  ..........FILE *stream..........
strport 00000000000.0   00000010G0110111  ..........FILE *stream..........
sfport  int portnum.C   00000011G0110111  ..........FILE *stream..........
        SUBRs
 spare                          010001x1
 spare                          010011x1
subr_0  ..........int hpoff.....01010101  ...........SCM (*f)()...........
subr_1  ..........int hpoff.....01010111  ...........SCM (*f)()...........
cxr     ..........int hpoff.....01011101  .........double (*f)()..........
subr_3  ..........int hpoff.....01011111  ...........SCM (*f)()...........
subr_2  ..........int hpoff.....01100101  ...........SCM (*f)()...........
asubr   ..........int hpoff.....01100111  ...........SCM (*f)()...........
subr_1o ..........int hpoff.....01101101  ...........SCM (*f)()...........
subr_2o ..........int hpoff.....01101111  ...........SCM (*f)()...........
lsubr_2 ..........int hpoff.....01110101  ...........SCM (*f)()...........
lsubr   ..........int hpoff.....01110111  ...........SCM (*f)()...........
rpsubr  ..........int hpoff.....01111101  ...........SCM (*f)()...........
                        SMOBs
free_cell
        000000000000000000000000G1111111  ...........*free_cell........000
flo     000000000000000000000001G1111111  ...........float num............
dblr    000000000000000100000001G1111111  ..........double *real..........
dblc    000000000000001100000001G1111111  .........complex *cmpx..........
bignum  ...int length...0000001 G1111111  .........short *digits..........
bigpos  ...int length...00000010G1111111  .........short *digits..........
bigneg  ...int length...00000011G1111111  .........short *digits..........
                        xxxxxxxx = code assigned by newsmob();
promise 000000000000000fxxxxxxxxG1111111  ...........SCM val..............
arbiter 000000000000000lxxxxxxxxG1111111  ...........SCM name.............
macro   000000000000000mxxxxxxxxG1111111  ...........SCM name.............
array   ...short rank..cxxxxxxxxG1111111  ............*array..............


File: scm.info,  Node: Operations,  Next: Program Self-Knowledge,  Prev: Data Types,  Up: The Implementation

6.2 Operations
==============

* Menu:

* Garbage Collection::          Automatically reclaims unused storage
* Memory Management for Environments::
* Signals::
* C Macros::
* Changing Scm::
* Defining Subrs::
* Defining Smobs::
* Defining Ptobs::
* Allocating memory::
* Embedding SCM::               In other programs
* Callbacks::
* Type Conversions::            For use with C code.
* Continuations::               For C and SCM
* Evaluation::                  Why SCM is fast


File: scm.info,  Node: Garbage Collection,  Next: Memory Management for Environments,  Prev: Operations,  Up: Operations

6.2.1 Garbage Collection
------------------------

The garbage collector is in the latter half of `sys.c'.  The primary
goal of "garbage collection" (or "GC") is to recycle those cells no
longer in use.  Immediates always appear as parts of other objects, so
they are not subject to explicit garbage collection.

All cells reside in the "heap" (composed of "heap segments").  Note
that this is different from what Computer Science usually defines as a
heap.

* Menu:

* Marking Cells::
* Sweeping the Heap::


File: scm.info,  Node: Marking Cells,  Next: Sweeping the Heap,  Prev: Garbage Collection,  Up: Garbage Collection

6.2.1.1 Marking Cells
.....................

The first step in garbage collection is to "mark" all heap objects in
use.  Each heap cell has a bit reserved for this purpose.  For pairs
(cons cells) the lowest order bit (0) of the CDR is used.  For other
types, bit 8 of the CAR is used.  The GC bits are never set except
during garbage collection.  Special C macros are defined in `scm.h' to
allow easy manipulation when GC bits are possibly set.  `CAR', `TYP3',
and `TYP7' can be used on GC marked cells as they are.

 -- Macro: GCCDR x
     Returns the CDR of a cons cell, even if that cell has been GC
     marked.

 -- Macro: GCTYP16 x
     Returns the 16 bit type code of a cell.

We need to (recursively) mark only a few objects in order to assure that
all accessible objects are marked.  Those objects are `sys_protects[]'
(for example, `dynwinds'), the current C-stack and the hash table for
symbols, "symhash".

 -- Function: void gc_mark (SCM OBJ)
     The function `gc_mark()' is used for marking SCM cells.  If OBJ is
     marked, `gc_mark()' returns.  If OBJ is unmarked, gc_mark sets the
     mark bit in OBJ, then calls `gc_mark()' on any SCM components of
     OBJ.  The last call to `gc_mark()' is tail-called (looped).

 -- Function: void mark_locations (STACKITEM X[], sizet LEN))
     The function `mark_locations' is used for marking segments of
     C-stack or saved segments of C-stack (marked continuations).  The
     argument LEN is the size of the stack in units of size
     `(STACKITEM)'.

     Each longword in the stack is tried to see if it is a valid cell
     pointer into the heap.  If it is, the object itself and any
     objects it points to are marked using `gc_mark'.  If the stack is
     word rather than longword aligned `(#define WORD_ALIGN)', both
     alignments are tried.  This arrangement will occasionally mark an
     object which is no longer used.  This has not been a problem in
     practice and the advantage of using the c-stack far outweighs it.


File: scm.info,  Node: Sweeping the Heap,  Prev: Marking Cells,  Up: Garbage Collection

6.2.1.2 Sweeping the Heap
.........................

After all found objects have been marked, the heap is swept.

The storage for strings, vectors, continuations, doubles, complexes, and
bignums is managed by malloc.  There is only one pointer to each malloc
object from its type-header cell in the heap.  This allows malloc
objects to be freed when the associated heap object is garbage
collected.

 -- Function: static void gc_sweep ()
     The function `gc_sweep' scans through all heap segments.  The mark
     bit is cleared from marked cells.  Unmarked cells are spliced into
     FREELIST, where they can again be returned by invocations of
     `NEWCELL'.

     If a type-header cell pointing to malloc space is unmarked, the
     malloc object is freed.  If the type header of smob is collected,
     the smob's `free' procedure is called to free its storage.


File: scm.info,  Node: Memory Management for Environments,  Next: Signals,  Prev: Garbage Collection,  Up: Operations

6.2.2 Memory Management for Environments
----------------------------------------

   * "Ecache" was designed and implemented by Radey Shouman.

   * This documentation of ecache was written by Tom Lord.

The memory management component of SCM contains special features which
optimize the allocation and garbage collection of environments.

The optimizations are based on certain facts and assumptions:

The SCM evaluator creates many environments with short lifetimes and
these account of a _large portion_ of the total number of objects
allocated.

The general purpose allocator allocates objects from a freelist, and
collects using a mark/sweep algorithm.  Research into garbage
collection suggests that such an allocator is sub-optimal for object
populations containing a large portion of short-lived members and that
allocation strategies involving a copying collector are more
appropriate.

It is a property of SCM, reflected throughout the source code, that a
simple copying collector can not be used as the general purpose memory
manager: much code assumes that the run-time stack can be treated as a
garbage collection root set using "conservative garbage collection"
techniques, which are incompatible with objects that change location.

Nevertheless, it is possible to use a mostly-separate
copying-collector, just for environments.  Roughly speaking, cons pairs
making up environments are initially allocated from a small heap that
is collected by a precise copying collector.  These objects must be
handled specially for the collector to work.  The (presumably) small
number of these objects that survive one collection of the copying heap
are copied to the general purpose heap, where they will later be
collected by the mark/sweep collector.  The remaining pairs are more
rapidly collected than they would otherwise be and all of this
collection is accomplished without having to mark or sweep any other
segment of the heap.

Allocating cons pairs for environments from this special heap is a
heuristic that approximates the (unachievable) goal:

     allocate all short-lived objects from the copying-heap, at no
     extra cost in allocation time.

Implementation Details
......................

A separate heap (`ecache_v') is maintained for the copying collector.
Pairs are allocated from this heap in a stack-like fashion.  Objects in
this heap may be protected from garbage collection by:

  1. Pushing a reference to the object on a stack specially maintained
     for that purpose.  This stack (`scm_estk') is used in place of the
     C run-time stack by the SCM evaluator to hold local variables
     which refer to the copying heap.

  2. Saving a reference to every object in the mark/sweep heap which
     directly references the copying heap in a root set that is
     specially maintained for that purpose (`scm_egc_roots').  If no
     object in the mark/sweep heap directly references an object from
     the copying heap, that object can be preserved by storing a direct
     reference to it in the copying-collector root set.

  3. Keeping no other references to these objects, except references
     between the objects themselves, during copying collection.

When the copying heap or root-set becomes full, the copying collector is
invoked.  All protected objects are copied to the mark-sweep heap.  All
references to those objects are updated.  The copying collector root-set
and heap are emptied.

References to pairs allocated specificly for environments are
inaccessible to the Scheme procedures evaluated by SCM.  These pairs
are manipulated by only a small number of code fragments in the
interpreter.  To support copying collection, those code fragments
(mostly in `eval.c') have been modified to protect environments from
garbage collection using the three rules listed above.

During a mark-sweep collection, the copying collector heap is marked
and swept almost like any ordinary segment of the general purpose heap.
The only difference is that pairs from the copying heap that become
free during a sweep phase are not added to the freelist.

The environment cache is disabled by adding `#define NO_ENV_CACHE' to
`eval.c'; all environment cells are then allocated from the regular
heap.

Relation to Other Work
......................

This work seems to build upon a considerable amount of previous work
into garbage collection techniques about which a considerable amount of
literature is available.


File: scm.info,  Node: Signals,  Next: C Macros,  Prev: Memory Management for Environments,  Up: Operations

6.2.3 Signals
-------------

 -- Function: init_signals
     (in `scm.c') initializes handlers for `SIGINT' and `SIGALRM' if
     they are supported by the C implementation.  All of the signal
     handlers immediately reestablish themselves by a call to
     `signal()'.

 -- Function: int_signal sig
 -- Function: alrm_signal sig
     The low level handlers for `SIGINT' and `SIGALRM'.

If an interrupt handler is defined when the interrupt is received, the
code is interpreted.  If the code returns, execution resumes from where
the interrupt happened.  `Call-with-current-continuation' allows the
stack to be saved and restored.

SCM does not use any signal masking system calls.  These are not a
portable feature.  However, code can run uninterrupted by use of the C
macros `DEFER_INTS' and `ALLOW_INTS'.

 -- Macro: DEFER_INTS
     sets the global variable `ints_disabled' to 1.  If an interrupt
     occurs during a time when `ints_disabled' is 1, then
     `deferred_proc' is set to non-zero, one of the global variables
     `SIGINT_deferred' or `SIGALRM_deferred' is set to 1, and the
     handler returns.

 -- Macro: ALLOW_INTS
     Checks the deferred variables and if set the appropriate handler is
     called.

     Calls to `DEFER_INTS' can not be nested.  An `ALLOW_INTS' must
     happen before another `DEFER_INTS' can be done.  In order to check
     that this constraint is satisfied `#define CAREFUL_INTS' in
     `scmfig.h'.


File: scm.info,  Node: C Macros,  Next: Changing Scm,  Prev: Signals,  Up: Operations

6.2.4 C Macros
--------------

 -- Macro: ASRTER cond arg pos subr
     signals an error if the expression (COND) is 0.  ARG is the
     offending object, SUBR is the string naming the subr, and POS
     indicates the position or type of error.  POS can be one of

        * `ARGn' (> 5 or unknown ARG number)

        * `ARG1'

        * `ARG2'

        * `ARG3'

        * `ARG4'

        * `ARG5'

        * `WNA' (wrong number of args)

        * `OVFLOW'

        * `OUTOFRANGE'

        * `NALLOC'

        * `EXIT'

        * `HUP_SIGNAL'

        * `INT_SIGNAL'

        * `FPE_SIGNAL'

        * `BUS_SIGNAL'

        * `SEGV_SIGNAL'

        * `ALRM_SIGNAL'

        * a C string `(char *)'

     Error checking is not done by `ASRTER' if the flag `RECKLESS' is
     defined.  An error condition can still be signaled in this case
     with a call to `wta(arg, pos, subr)'.

 -- Macro: ASRTGO cond label
     `goto' LABEL if the expression (COND) is 0.  Like `ASRTER',
     `ASRTGO' does is not active if the flag `RECKLESS' is defined.


File: scm.info,  Node: Changing Scm,  Next: Defining Subrs,  Prev: C Macros,  Up: Operations

6.2.5 Changing Scm
------------------

When writing C-code for SCM, a precaution is recommended.  If your
routine allocates a non-cons cell which will _not_ be incorporated into
a `SCM' object which is returned, you need to make sure that a `SCM'
variable in your routine points to that cell as long as part of it
might be referenced by your code.

In order to make sure this `SCM' variable does not get optimized out
you can put this assignment after its last possible use:

     SCM_dummy1 = foo;

or put this assignment somewhere in your routine:

     SCM_dummy1 = (SCM) &foo;

`SCM_dummy' variables are not currently defined.  Passing the address
of the local `SCM' variable to _any_ procedure also protects it.  The
procedure `scm_protect_temp' is provided for this purpose.

 -- Function: void scm_protect_temp (SCM *PTR)
     Forces the SCM object PTR to be saved on the C-stack, where it
     will be traced for GC.

Also, if you maintain a static pointer to some (non-immediate) `SCM'
object, you must either make your pointer be the value cell of a symbol
(see `errobj' for an example) or (permanently) add your pointer to
`sys_protects' using:

 -- Function: SCM scm_gc_protect (SCM OBJ)
     Permanently adds OBJ to a table of objects protected from garbage
     collection.  `scm_gc_protect' returns OBJ.

To add a C routine to scm:

  1. choose the appropriate subr type from the type list.

  2. write the code and put into `scm.c'.

  3. add a `make_subr' or `make_gsubr' call to `init_scm'.  Or put an
     entry into the appropriate `iproc' structure.

To add a package of new procedures to scm (see `crs.c' for example):

  1. create a new C file (`foo.c').

  2. at the front of `foo.c' put declarations for strings for your
     procedure names.

          static char s_twiddle_bits[]="twiddle-bits!";
          static char s_bitsp[]="bits?";

  3. choose the appropriate subr types from the type list in `code.doc'.

  4. write the code for the procedures and put into `foo.c'

  5. create one `iproc' structure for each subr type used in `foo.c'

          static iproc subr3s[]= {
                  {s_twiddle-bits,twiddle-bits},
                  {s_bitsp,bitsp},
                  {0,0} };

  6. create an `init_<name of file>' routine at the end of the file
     which calls `init_iprocs' with the correct type for each of the
     `iproc's created in step 5.

          void init_foo()
          {
            init_iprocs(subr1s, tc7_subr_1);
            init_iprocs(subr3s, tc7_subr_3);
          }

     If your package needs to have a "finalization" routine called to
     free up storage, close files, etc, then also have a line in
     `init_foo' like:

          add_final(final_foo);

     `final_foo' should be a (void) procedure of no arguments.  The
     finals will be called in opposite order from their definition.

     The line:

          add_feature("foo");

     will append a symbol `'foo' to the (list) value of `*features*'.

  7. put any scheme code which needs to be run as part of your package
     into `Ifoo.scm'.

  8. put an `if' into `Init5e2.scm' which loads `Ifoo.scm' if your            |
     package is included:

          (if (defined? twiddle-bits!)
              (load (in-vicinity (implementation-vicinity)
                                 "Ifoo"
                                 (scheme-file-suffix))))

     or use `(provided? 'foo)' instead of `(defined?  twiddle-bits!)'
     if you have added the feature.

  9. put documentation of the new procedures into `foo.doc'

 10. add lines to your `Makefile' to compile and link SCM with your
     object file.  Add a `init_foo\(\)\;' to the `INITS=...' line at
     the beginning of the makefile.

These steps should allow your package to be linked into SCM with a
minimum of difficulty.  Your package should also work with dynamic
linking if your SCM has this capability.

Special forms (new syntax) can be added to scm.

  1. define a new `MAKISYM' in `scm.h' and increment `NUM_ISYMS'.

  2. add a string with the new name in the corresponding place in
     `isymnames' in `repl.c'.

  3. add `case' clause to `ceval()' near `i_quasiquote' (in `eval.c').

New syntax can now be added without recompiling SCM by the use of the
`procedure->syntax', `procedure->macro', `procedure->memoizing-macro',
and `defmacro'.  For details, *Note Syntax::.


File: scm.info,  Node: Defining Subrs,  Next: Defining Smobs,  Prev: Changing Scm,  Up: Operations

6.2.6 Defining Subrs
--------------------

If "CCLO" is `#define'd when compiling, the compiled closure feature
will be enabled.  It is automatically enabled if dynamic linking is
enabled.

The SCM interpreter directly recognizes subrs taking small numbers of
arguments.  In order to create subrs taking larger numbers of arguments
use:

 -- Function: make_gsubr name req opt rest fcn
     returns a cclo (compiled closure) object of name `char *' NAME
     which takes `int' REQ required arguments, `int' OPT optional
     arguments, and a list of rest arguments if `int' REST is 1 (0 for
     not).

     `SCM (*fcn)()' is a pointer to a C function to do the work.

     The C function will always be called with REQ + OPT + REST
     arguments, optional arguments not supplied will be passed
     `UNDEFINED'.  An error will be signaled if the subr is called with
     too many or too few arguments.  Currently a total of 10 arguments
     may be specified, but increasing this limit should not be
     difficult.

          /* A silly example, taking 2 required args,
             1 optional, and a list of rest args */

          #include <scm.h>

          SCM gsubr_21l(req1,req2,opt,rst)
               SCM req1,req2,opt,rst;
          {
            lputs("gsubr-2-1-l:\n req1: ", cur_outp);
            display(req1,cur_outp);
            lputs("\n req2: ", cur_outp);
            display(req2,cur_outp);
            lputs("\n opt: ", cur_outp);
            display(opt,cur_outp);
            lputs("\n rest: ", cur_outp);
            display(rst,cur_outp);
            newline(cur_outp);
            return UNSPECIFIED;
          }

          void init_gsubr211()
          {
            make_gsubr("gsubr-2-1-l", 2, 1, 1, gsubr_21l);
          }


File: scm.info,  Node: Defining Smobs,  Next: Defining Ptobs,  Prev: Defining Subrs,  Up: Operations

6.2.7 Defining Smobs
--------------------

Here is an example of how to add a new type named `foo' to SCM.  The
following lines need to be added to your code:

`long tc16_foo;'
     The type code which will be used to identify the new type.

`static smobfuns foosmob = {markfoo,freefoo,printfoo,equalpfoo};'
     smobfuns is a structure composed of 4 functions:

          typedef struct {
            SCM   (*mark)P((SCM));
            sizet (*free)P((CELLPTR));
            int   (*print)P((SCM exp, SCM port, int writing));
            SCM   (*equalp)P((SCM, SCM));
          } smobfuns;

    `smob.mark'
          is a function of one argument of type `SCM' (the cell to
          mark) and returns type `SCM' which will then be marked.  If
          no further objects need to be marked then return an immediate
          object such as `BOOL_F'.  The smob cell itself will already
          have been marked.  _Note_ This is different from SCM versions
          prior to 5c5.  Only additional data specific to a smob type
          need be marked by `smob.mark'.

          2 functions are provided:

         `markcdr(ptr)'
               returns `CDR(ptr)'.

         `mark0(ptr)'
               is a no-op used for smobs containing no additional `SCM'
               data.  0 may also be used in this case.

    `smob.free'
          is a function of one argument of type `CELLPTR' (the cell to
          collected) and returns type `sizet' which is the number of
          `malloc'ed bytes which were freed.  `Smob.free' should free
          any `malloc'ed storage associated with this object.  The
          function free0(ptr) is provided which does not free any
          storage and returns 0.

    `smob.print'
          is 0 or a function of 3 arguments.  The first, of type `SCM',
          is the smob object.  The second, of type `SCM', is the stream
          on which to write the result.  The third, of type int, is 1
          if the object should be `write'n, 0 if it should be
          `display'ed, and 2 if it should be `write'n for an error
          report.  This function should return non-zero if it printed,
          and zero otherwise (in which case a hexadecimal number will
          be printed).

    `smob.equalp'
          is 0 or a function of 2 `SCM' arguments.  Both of these
          arguments will be of type `tc16foo'.  This function should
          return `BOOL_T' if the smobs are equal, `BOOL_F' if they are
          not.  If `smob.equalp' is 0, `equal?' will return `BOOL_F' if
          they are not `eq?'.

`tc16_foo = newsmob(&foosmob);'
     Allocates the new type with the functions from `foosmob'.  This
     line goes in an `init_' routine.

Promises and macros in `eval.c' and arbiters in `repl.c' provide
examples of SMOBs.  There are a maximum of 256 SMOBs.  Smobs that must
allocate blocks of memory should use, for example, `must_malloc' rather
than `malloc' *Note Allocating memory::.


File: scm.info,  Node: Defining Ptobs,  Next: Allocating memory,  Prev: Defining Smobs,  Up: Operations

6.2.8 Defining Ptobs
--------------------

"ptob"s are similar to smobs but define new types of port to which SCM
procedures can read or write.  The following functions are defined in
the `ptobfuns':

     typedef struct {
       SCM   (*mark)P((SCM ptr));
       int   (*free)P((FILE *p));
       int   (*print)P((SCM exp, SCM port, int writing));
       SCM   (*equalp)P((SCM, SCM));
       int   (*fputc)P((int c, FILE *p));
       int   (*fputs)P((char *s, FILE *p));
       sizet (*fwrite)P((char *s, sizet siz, sizet num, FILE *p));
       int   (*fflush)P((FILE *stream));
       int   (*fgetc)P((FILE *p));
       int   (*fclose)P((FILE *p));
     } ptobfuns;

The `.free' component to the structure takes a `FILE *' or other C
construct as its argument, unlike `.free' in a smob, which takes the
whole smob cell.  Often, `.free' and `.fclose' can be the same
function.  See `fptob' and `pipob' in `sys.c' for examples of how to
define ptobs.  Ptobs that must allocate blocks of memory should use,
for example, `must_malloc' rather than `malloc' *Note Allocating
memory::.


File: scm.info,  Node: Allocating memory,  Next: Embedding SCM,  Prev: Defining Ptobs,  Up: Operations

6.2.9 Allocating memory
-----------------------

SCM maintains a count of bytes allocated using malloc, and calls the
garbage collector when that number exceeds a dynamically managed limit.
In order for this to work properly, `malloc' and `free' should not be
called directly to manage memory freeable by garbage collection.  The
following functions are provided for that purpose:

 -- Function: SCM must_malloc_cell (long LEN, SCM C, char *WHAT)
 -- Function: char * must_malloc (long LEN, char *WHAT)
     LEN is the number of bytes that should be allocated, WHAT is a
     string to be used in error or gc messages.  `must_malloc' returns
     a pointer to newly allocated memory.  `must_malloc_cell' returns a
     newly allocated cell whose `car' is C and whose `cdr' is a pointer
     to newly allocated memory.

 -- Function: void must_realloc_cell (SCM Z, long OLEN, long LEN, char
          *WHAT)
 -- Function: char * must_realloc (char *WHERE, long OLEN, long LEN,
          char *WHAT)
     `must_realloc_cell' takes as argument Z a cell whose `cdr' should
     be a pointer to a block of memory of length OLEN allocated with
     `must_malloc_cell' and modifies the `cdr' to point to a block of
     memory of length LEN.  `must_realloc' takes as argument WHERE the
     address of a block of memory of length OLEN allocated by
     `must_malloc' and returns the address of a block of length LEN.

     The contents of the reallocated block will be unchanged up to the
     minimum of the old and new sizes.

     WHAT is a pointer to a string used for error and gc messages.

`must_malloc', `must_malloc_cell', `must_realloc', and
`must_realloc_cell' must be called with interrupts deferred *Note
Signals::.  `must_realloc' and `must_realloc_cell' must not be called
during initialization (non-zero errjmp_bad) - the initial allocations
must be large enough.

 -- Function: void must_free (char *PTR, sizet LEN)
     `must_free' is used to free a block of memory allocated by the
     above functions and pointed to by PTR.  LEN is the length of the
     block in bytes, but this value is used only for debugging purposes.
     If it is difficult or expensive to calculate then zero may be used
     instead.


File: scm.info,  Node: Embedding SCM,  Next: Callbacks,  Prev: Allocating memory,  Up: Operations

6.2.10 Embedding SCM
--------------------

The file `scmmain.c' contains the definition of main().  When SCM is
compiled as a library `scmmain.c' is not included in the library; a
copy of `scmmain.c' can be modified to use SCM as an embedded library
module.

 -- Function: int main (int ARGC, char **ARGV)
     This is the top level C routine.  The value of the ARGC argument
     is the number of command line arguments.  The ARGV argument is a
     vector of C strings; its elements are the individual command line
     argument strings.  A null pointer always follows the last element:
     `ARGV[ARGC]' is this null pointer.

 -- Variable: char *execpath
     This string is the pathname of the executable file being run.  This
     variable can be examined and set from Scheme (*note Internal
     State::).  EXECPATH must be set to executable's path in order to
     use DUMP (*note Dump::) or DLD.

Rename main() and arrange your code to call it with an ARGV which sets
up SCM as you want it.

If you need more control than is possible through ARGV, here are
descriptions of the functions which main() calls.

 -- Function: void init_sbrk (void)
     Call this before SCM calls malloc().  Value returned from sbrk()
     is used to gauge how much storage SCM uses.

 -- Function: char * scm_find_execpath (int ARGC, char **ARGV, char
          *SCRIPT_ARG)
     ARGC and ARGV are as described in main().  SCRIPT_ARG is the
     pathname of the SCSH-style script (*note Scripting::) being
     invoked; 0 otherwise.  `scm_find_execpath' returns the pathname of
     the executable being run; if `scm_find_execpath' cannot determine
     the pathname, then it returns 0.

`scm_find_implpath' is defined in `scmmain.c'.  Preceeding this are
definitions ofGENERIC_NAME and INIT_GETENV.  These, along with IMPLINIT
and DIRSEP control scm_find_implpath()'s operation.

If your application has an easier way to locate initialization code for
SCM, then you can replace `scm_find_implpath'.

 -- Function: char * scm_find_implpath (char *EXECPATH)
     Returns the full pathname of the Scheme initialization file or 0
     if it cannot find it.

     The string value of the preprocessor variable INIT_GETENV names an
     environment variable (default `"SCM_INIT_PATH"').  If this
     environment variable is defined, its value will be returned from
     `scm_find_implpath'.  Otherwise find_impl_file() is called with the
     arguments EXECPATH, GENERIC_NAME (default "scm"), INIT_FILE_NAME
     (default "Init5e2_scm"), and the directory separator string              |
     DIRSEP.  If find_impl_file() returns 0 and IMPLINIT is defined,
     then a copy of the string IMPLINIT is returned.

 -- Function: int init_buf0 (FILE *INPORT)
     Tries to determine whether INPORT (usually stdin) is an
     interactive input port which should be used in an unbuffered mode.
     If so, INPORT is set to unbuffered and non-zero is returned.
     Otherwise, 0 is returned.

     `init_buf0' should be called before any input is read from INPORT.
     Its value can be used as the last argument to
     scm_init_from_argv().

 -- Function: void scm_init_from_argv (int ARGC, char **ARGV, char
          *SCRIPT_ARG, int IVERBOSE, int BUF0STDIN)
     Initializes SCM storage and creates a list of the argument strings
     PROGRAM-ARGUMENTS from ARGV.  ARGC and ARGV must already be
     processed to accomodate Scheme Scripts (if desired).  The scheme
     variable *SCRIPT* is set to the string SCRIPT_ARG, or #f if
     SCRIPT_ARG is 0.  IVERBOSE is the initial prolixity level.  If
     BUF0STDIN is non-zero, stdin is treated as an unbuffered port.

Call `init_signals' and `restore_signals' only if you want SCM to
handle interrupts and signals.

 -- Function: void init_signals (void)
     Initializes handlers for `SIGINT' and `SIGALRM' if they are
     supported by the C implementation.  All of the signal handlers
     immediately reestablish themselves by a call to `signal()'.

 -- Function: void restore_signals (void)
     Restores the handlers in effect when `init_signals' was called.

 -- Function: SCM scm_top_level (char *INITPATH, SCM (*toplvl_fun)())
     This is SCM's top-level.  Errors longjmp here.  TOPLVL_FUN is a
     callback function of zero arguments that is called by
     `scm_top_level' to do useful work - if zero, then `repl', which
     implements a read-eval-print loop, is called.

     If TOPLVL_FUN returns, then `scm_top_level' will return as well.
     If the return value of TOPLVL_FUN is an immediate integer then it
     will be used as the return value of `scm_top_level'.  In the main
     function supplied with SCM, this return value is the exit status
     of the process.

     If the first character of string INITPATH is `;', `(' or
     whitespace, then scm_ldstr() is called with INITPATH to initialize
     SCM; otherwise INITPATH names a file of Scheme code to be loaded
     to initialize SCM.

     When a Scheme error is signaled; control will pass into
     `scm_top_level' by `longjmp', error messages will be printed to
     `current-error-port', and then TOPLVL_FUN will be called again.
     TOPLVL_FUN must maintain enough state to prevent errors from being
     resignalled.  If `toplvl_fun' can not recover from an error
     situation it may simply return.

 -- Function: void final_scm (int FREEALL)
     Calls all finalization routines registered with add_final().  If
     FREEALL is non-zero, then all memory which SCM allocated with
     malloc() will be freed.

You can call indivdual Scheme procedures from C code in the TOPLVL_FUN
argument passed to scm_top_level(), or from module subrs (registered by
an `init_' function, *note Changing Scm::).

Use `apply' to call Scheme procedures from your C code.  For example:

     /* If this apply fails, SCM will catch the error */
     apply(CDR(intern("srv:startup",sizeof("srv:startup")-1)),
           mksproc(srvproc),
           listofnull);

     func = CDR(intern(rpcname,strlen(rpcname)));
     retval = apply(func, cons(mksproc(srvproc), args), EOL);

Functions for loading Scheme files and evaluating Scheme code given as
C strings are described in the next section, (*note Callbacks::).

Here is a minimal embedding program `libtest.c':

     /* gcc -o libtest libtest.c libscm.a -ldl -lm -lc */
     #include "scm.h"
     /* include patchlvl.h for SCM's INIT_FILE_NAME. */
     #include "patchlvl.h"

     void libtest_init_user_scm()
     {
       fputs("This is libtest_init_user_scm\n", stderr); fflush(stderr);
       sysintern("*the-string*", makfrom0str("hello world\n"));
     }

     SCM user_main()
     {
       static int done = 0;
       if (done++) return MAKINUM(EXIT_FAILURE);
       scm_ldstr("(display *the-string*)");
       return MAKINUM(EXIT_SUCCESS);
     }

     int main(argc, argv)
          int argc;
          const char **argv;                                                  |
     {
       SCM retval;
       char *implpath, *execpath;

       init_user_scm = libtest_init_user_scm;
       execpath = dld_find_executable(argv[0]);
       fprintf(stderr, "dld_find_executable(%s): %s\n", argv[0], execpath);
       implpath = find_impl_file(execpath, "scm", INIT_FILE_NAME, dirsep);
       fprintf(stderr, "implpath: %s\n", implpath);
       scm_init_from_argv(argc, argv, 0L, 0, 0);

       retval = scm_top_level(implpath, user_main);

       final_scm(!0);
       return (int)INUM(retval);
     }

     -|
     dld_find_executable(./libtest): /home/jaffer/scm/libtest
     implpath: /home/jaffer/scm/Init5e2.scm                                   |
     This is libtest_init_user_scm
     hello world


File: scm.info,  Node: Callbacks,  Next: Type Conversions,  Prev: Embedding SCM,  Up: Operations

6.2.11 Callbacks
----------------

SCM now has routines to make calling back to Scheme procedures easier.
The source code for these routines are found in `rope.c'.

 -- Function: int scm_ldfile (char *FILE)
     Loads the Scheme source file FILE.  Returns 0 if successful, non-0
     if not.  This function is used to load SCM's initialization file
     `Init5e2.scm'.                                                           |

 -- Function: int scm_ldprog (char *FILE)
     Loads the Scheme source file `(in-vicinity (program-vicinity)
     FILE)'.  Returns 0 if successful, non-0 if not.

     This function is useful for compiled code init_ functions to load
     non-compiled Scheme (source) files.  `program-vicinity' is the
     directory from which the calling code was loaded (*note Vicinity:
     (slib)Vicinity.).

 -- Function: SCM scm_evstr (char *STR)
     Returns the result of reading an expression from STR and
     evaluating it.

 -- Function: void scm_ldstr (char *STR)
     Reads and evaluates all the expressions from STR.

If you wish to catch errors during execution of Scheme code, then you
can use a wrapper like this for your Scheme procedures:

     (define (srv:protect proc)
       (lambda args
         (define result #f)                  ; put default value here
         (call-with-current-continuation
          (lambda (cont)
            (dynamic-wind (lambda () #t)
                          (lambda ()
                            (set! result (apply proc args))
                            (set! cont #f))
                          (lambda ()
                            (if cont (cont #f))))))
         result))

Calls to procedures so wrapped will return even if an error occurs.


File: scm.info,  Node: Type Conversions,  Next: Continuations,  Prev: Callbacks,  Up: Operations

6.2.12 Type Conversions
-----------------------

These type conversion functions are very useful for connecting SCM and C
code.  Most are defined in `rope.c'.

 -- Function: SCM long2num (long N)
 -- Function: SCM ulong2num (unsigned long N)
     Return an object of type `SCM' corresponding to the `long' or
     `unsigned long' argument N.  If N cannot be converted, `BOOL_F' is
     returned.  Which numbers can be converted depends on whether SCM
     was compiled with the `BIGDIG' or `FLOATS' flags.

     To convert integer numbers of smaller types (`short' or `char'),
     use the macro `MAKINUM(n)'.

 -- Function: long num2long (SCM NUM, char *POS, char *S_CALLER)
 -- Function: unsigned long num2ulong (SCM NUM, char *POS, char
          *S_CALLER)
 -- Function: short num2short (SCM NUM, char *POS, char *S_CALLER)
 -- Function: unsigned short num2ushort (SCM NUM, char *POS, char
          *S_CALLER)
 -- Function: unsigned char num2uchar (SCM NUM, char *POS, char
          *S_CALLER)
 -- Function: double num2dbl (SCM NUM, char *POS, char *S_CALLER)
     These functions are used to check and convert `SCM' arguments to
     the named C type.  The first argument NUM is checked to see it it
     is within the range of the destination type.  If so, the converted
     number is returned.  If not, the `ASRTER' macro calls `wta' with
     NUM and strings POS and S_CALLER.  For a listing of useful
     predefined POS macros, *Note C Macros::.

     _Note_ Inexact numbers are accepted only by `num2dbl', `num2long',
     and `num2ulong' (for when `SCM' is compiled without bignums).  To
     convert inexact numbers to exact numbers, *Note inexact->exact:
     (r5rs)Numerical operations.

 -- Function: unsigned long scm_addr (SCM ARGS, char *S_NAME)
     Returns a pointer (cast to an `unsigned long') to the storage
     corresponding to the location accessed by
     `aref(CAR(args),CDR(args))'.  The string S_NAME is used in any
     messages from error calls by `scm_addr'.

     `scm_addr' is useful for performing C operations on strings or
     other uniform arrays (*note Uniform Array::).

 -- Function: unsigned long scm_base_addr(SCM RA, char *S_NAME)
     Returns a pointer (cast to an `unsigned long') to the beginning of
     storage of array RA.  Note that if RA is a shared-array, the
     strorage accessed this way may be much larger than RA.

     _Note_ While you use a pointer returned from `scm_addr' or
     `scm_base_addr' you must keep a pointer to the associated `SCM'
     object in a stack allocated variable or GC-protected location in
     order to assure that SCM does not reuse that storage before you
     are done with it.  *Note scm_gc_protect: Changing Scm.

 -- Function: SCM makfrom0str (char *SRC)
 -- Function: SCM makfromstr (char *SRC, sizet LEN)
     Return a newly allocated string `SCM' object copy of the
     null-terminated string SRC or the string SRC of length LEN,
     respectively.

 -- Function: SCM makfromstrs (int ARGC, char **ARGV)
     Returns a newly allocated `SCM' list of strings corresponding to
     the ARGC length array of null-terminated strings ARGV.  If ARGV is
     less than `0', ARGV is assumed to be `NULL' terminated.
     `makfromstrs' is used by `scm_init_from_argv' to convert the
     arguments SCM was called with to a `SCM' list which is the value
     of SCM procedure calls to `program-arguments' (*note
     program-arguments: SCM Session.).

 -- Function: char ** makargvfrmstrs (SCM ARGS, char *S_NAME)
     Returns a `NULL' terminated list of null-terminated strings copied
     from the `SCM' list of strings ARGS.  The string S_NAME is used in
     messages from error calls by `makargvfrmstrs'.

     `makargvfrmstrs' is useful for constructing argument lists suitable
     for passing to `main' functions.

 -- Function: void must_free_argv (char **ARGV)
     Frees the storage allocated to create ARGV by a call to
     `makargvfrmstrs'.


File: scm.info,  Node: Continuations,  Next: Evaluation,  Prev: Type Conversions,  Up: Operations

6.2.13 Continuations
--------------------

The source files `continue.h' and `continue.c' are designed to function
as an independent resource for programs wishing to use continuations,
but without all the rest of the SCM machinery.  The concept of
continuations is explained in *Note call-with-current-continuation:
(r5rs)Control features.

The C constructs `jmp_buf', `setjmp', and `longjmp' implement escape
continuations.  On VAX and Cray platforms, the setjmp provided does not
save all the registers.  The source files `setjump.mar', `setjump.s',
and `ugsetjump.s' provide implementations which do meet this criteria.

SCM uses the names `jump_buf', `setjump', and `longjump' in lieu of
`jmp_buf', `setjmp', and `longjmp' to prevent name and declaration
conflicts.

 -- Data type: CONTINUATION jmpbuf length stkbse other parent
     is a `typedef'ed structure holding all the information needed to
     represent a continuation.  The OTHER slot can be used to hold any
     data the user wishes to put there by defining the macro
     `CONTINUATION_OTHER'.

 -- Macro: SHORT_ALIGN
     If `SHORT_ALIGN' is `#define'd (in `scmfig.h'), then the it is
     assumed that pointers in the stack can be aligned on `short int'
     boundaries.

 -- Data type: STACKITEM
     is a pointer to objects of the size specified by `SHORT_ALIGN'
     being `#define'd or not.

 -- Macro: CHEAP_CONTINUATIONS
     If `CHEAP_CONTINUATIONS' is `#define'd (in `scmfig.h') each
     `CONTINUATION' has size `sizeof CONTINUATION'.  Otherwise, all but
     "root" `CONTINUATION's have additional storage (immediately
     following) to contain a copy of part of the stack.

     _Note_ On systems with nonlinear stack disciplines (multiple
     stacks or non-contiguous stack frames) copying the stack will not
     work properly.  These systems need to #define
     `CHEAP_CONTINUATIONS' in `scmfig.h'.

 -- Macro: STACK_GROWS_UP
     Expresses which way the stack grows by its being `#define'd or not.

 -- Variable: long thrown_value
     Gets set to the VALUE passed to `throw_to_continuation'.

 -- Function: long stack_size (STACKITEM *START)
     Returns the number of units of size `STACKITEM' which fit between
     START and the current top of stack.  No check is done in this
     routine to ensure that START is actually in the current stack
     segment.

 -- Function: CONTINUATION * make_root_continuation (STACKITEM
          *STACK_BASE)
     Allocates (`malloc') storage for a `CONTINUATION' of the current
     extent of stack.  This newly allocated `CONTINUATION' is returned
     if successful, `0' if not.  After `make_root_continuation'
     returns, the calling routine still needs to
     `setjump(NEW_CONTINUATION->jmpbuf)' in order to complete the
     capture of this continuation.

 -- Function: CONTINUATION * make_continuation (CONTINUATION
          *PARENT_CONT)
     Allocates storage for the current `CONTINUATION', copying (or
     encapsulating) the stack state from `PARENT_CONT->stkbse' to the
     current top of stack.  The newly allocated `CONTINUATION' is
     returned if successful, `0'q if not.  After `make_continuation'
     returns, the calling routine still needs to
     `setjump(NEW_CONTINUATION->jmpbuf)' in order to complete the
     capture of this continuation.

 -- Function: void free_continuation (CONTINUATION *CONT)
     Frees the storage pointed to by CONT.  Remember to free storage
     pointed to by `CONT->other'.

 -- Function: void throw_to_continuation (CONTINUATION *CONT, long
          VALUE, CONTINUATION *ROOT_CONT)
     Sets `thrown_value' to VALUE and returns from the continuation
     CONT.

     If `CHEAP_CONTINUATIONS' is `#define'd, then
     `throw_to_continuation' does `longjump(CONT->jmpbuf, val)'.

     If `CHEAP_CONTINUATIONS' is not `#define'd, the CONTINUATION CONT
     contains a copy of a portion of the C stack (whose bound must be
     `CONT(ROOT_CONT)->stkbse').  Then:

        * the stack is grown larger than the saved stack, if neccessary.

        * the saved stack is copied back into it's original position.

        * `longjump(CONT->jmpbuf, val)';


File: scm.info,  Node: Evaluation,  Prev: Continuations,  Up: Operations

6.2.14 Evaluation
-----------------

SCM uses its type representations to speed evaluation.  All of the
`subr' types (*note Subr Cells::) are `tc7' types.  Since the `tc7'
field is in the low order bit position of the `CAR' it can be retrieved
and dispatched on quickly by dereferencing the SCM pointer pointing to
it and masking the result.

All the SCM "Special Forms" get translated to immediate symbols
(`isym') the first time they are encountered by the interpreter
(`ceval').  The representation of these immediate symbols is engineered
to occupy the same bits as `tc7'.  All the `isym's occur only in the
`CAR' of lists.

If the `CAR' of a expression to evaluate is not immediate, then it may
be a symbol.  If so, the first time it is encountered it will be
converted to an immediate type `ILOC' or `GLOC' (*note Immediates::).
The codes for `ILOC' and `GLOC' lower 7 bits distinguish them from all
the other types we have discussed.

Once it has determined that the expression to evaluate is not immediate,
`ceval' need only retrieve and dispatch on the low order 7 bits of the
`CAR' of that cell, regardless of whether that cell is a closure,
header, or subr, or a cons containing `ILOC' or `GLOC'.

In order to be able to convert a SCM symbol pointer to an immediate
`ILOC' or `GLOC', the evaluator must be holding the pointer to the list
in which that symbol pointer occurs.  Turning this requirement to an
advantage, `ceval' does not recursively call itself to evaluate symbols
in lists; It instead calls the macro "EVALCAR".  `EVALCAR' does symbol
lookup and memoization for symbols, retrieval of values for `ILOC's and
`GLOC's, returns other immediates, and otherwise recursively calls
itself with the `CAR' of the list.

`ceval' inlines evaluation (using `EVALCAR') of almost all procedure
call arguments.  When `ceval' needs to evaluate a list of more than
length 3, the procedure `eval_args' is called.  So `ceval' can be said
to have one level lookahead.  The avoidance of recursive invocations of
`ceval' for the most common cases (special forms and procedure calls)
results in faster execution.  The speed of the interpreter is currently
limited on most machines by interpreter size, probably having to do
with its cache footprint.  In order to keep the size down, certain
`EVALCAR' calls which don't need to be fast (because they rarely occur
or because they are part of expensive operations) are instead calls to
the C function `evalcar'.

 -- Variable: symhash
     Top level symbol values are stored in the `symhash' table.
     `symhash' is an array of lists of `ISYM's and pairs of symbols and
     values.

 -- Immediate: ILOC
     Whenever a symbol's value is found in the local environment the
     pointer to the symbol in the code is replaced with an immediate
     object (`ILOC') which specifies how many environment frames down
     and how far in to go for the value.  When this immediate object is
     subsequently encountered, the value can be retrieved quickly.

`ILOC's work up to a maximum depth of 4096 frames or 4096 identifiers
in a frame.  Radey Shouman added "FARLOC" to handle cases exceeding
these limits.  A `FARLOC' consists of a pair whose CAR is the immediate
type `IM_FARLOC_CAR' or `IM_FARLOC_CDR', and whose CDR is a pair of
INUMs specifying the frame and distance with a larger range than
`ILOC's span.

Adding `#define TEST_FARLOC' to `eval.c' causes `FARLOC's to be
generated for all local identifiers; this is useful only for testing
memoization.

 -- Immediate: GLOC
     Pointers to symbols not defined in local environments are changed
     to one plus the value cell address in symhash.  This incremented
     pointer is called a `GLOC'.  The low order bit is normally
     reserved for GCmark; But, since references to variables in the
     code always occur in the `CAR' position and the GCmark is in the
     `CDR', there is no conflict.

If the compile FLAG `CAUTIOUS' is #defined then the number of arguments
is always checked for application of closures.  If the compile FLAG
`RECKLESS' is #defined then they are not checked.  Otherwise, number of
argument checks for closures are made only when the function position
(whose value is the closure) of a combination is not an `ILOC' or
`GLOC'.  When the function position of a combination is a symbol it
will be checked only the first time it is evaluated because it will
then be replaced with an `ILOC' or `GLOC'.

 -- Macro: EVAL expression env
 -- Macro: SIDEVAL expression env
     `EVAL' Returns the result of evaluating EXPRESSION in ENV.
     `SIDEVAL' evaluates EXPRESSION in ENV when the value of the
     expression is not used.

     Both of these macros alter the list structure of EXPRESSION as it
     is memoized and hence should be used only when it is known that
     EXPRESSION will not be referenced again.  The C function `eval' is
     safe from this problem.

 -- Function: SCM eval (SCM EXPRESSION)
     Returns the result of evaluating EXPRESSION in the top-level
     environment.  `eval' copies `expression' so that memoization does
     not modify `expression'.


File: scm.info,  Node: Program Self-Knowledge,  Next: Improvements To Make,  Prev: Operations,  Up: The Implementation

6.3 Program Self-Knowledge
==========================

* Menu:

* File-System Habitat::
* Executable Pathname::
* Script Support::


File: scm.info,  Node: File-System Habitat,  Next: Executable Pathname,  Prev: Program Self-Knowledge,  Up: Program Self-Knowledge

6.3.1 File-System Habitat
-------------------------

Where should software reside?  Although individually a minor annoyance,
cumulatively this question represents many thousands of frustrated user
hours spent trying to find support files or guessing where packages need
to be installed.  Even simple programs require proper habitat; games
need to find their score files.

Aren't there standards for this?  Some Operating Systems have devised
regimes of software habitats - only to have them violated by large
software packages and imports from other OS varieties.

In some programs, the expected locations of support files are fixed at
time of compilation.  This means that the program may not run on
configurations unanticipated by the authors.  Compiling locations into a
program also can make it immovable - necessitating recompilation to
install it.

     Programs of the world unite!  You have nothing to lose but loss
     itself.

The function `find_impl_file' in `scm.c' is an attempt to create a
utility (for inclusion in programs) which will hide the details of
platform-dependent file habitat conventions.  It takes as input the
pathname of the executable file which is running.  If there are systems
for which this information is either not available or unrelated to the
locations of support files, then a higher level interface will be
needed.

 -- Function: char * find_impl_file (char *EXEC_PATH, char
          *GENERIC_NAME, char *INITNAME, char *SEP)
     Given the pathname of this executable (EXEC_PATH), test for the
     existence of INITNAME in the implementation-vicinity of this
     program.  Return a newly allocated string of the path if
     successful, 0 if not.  The SEP argument is a _null-terminated
     string_ of the character used to separate directory components.

   * One convention is to install the support files for an executable
     program in the same directory as the program.  This possibility is
     tried first, which satisfies not only programs using this
     convention, but also uninstalled builds when testing new releases,
     etc.

   * Another convention is to install the executables in a directory
     named `bin', `BIN', `exe', or `EXE' and support files in a
     directroy named `lib', which is a peer the executable directory.
     This arrangement allows multiple executables can be stored in a
     single directory.  For example, the executable might be in
     `/usr/local/bin/' and initialization file in `/usr/local/lib/'.

     If the executable directory name matches, the peer directroy `lib'
     is tested for INITNAME.

   * Sometimes `lib' directories become too crowded.  So we look in any
     subdirectories of `lib' or `src' having the name (sans type suffix
     such as `.EXE') of the program we are running.  For example, the
     executable might be `/usr/local/bin/foo' and initialization file
     in `/usr/local/lib/foo/'.

   * But the executable name may not be the usual program name; So also
     look in any GENERIC_NAME subdirectories of `lib' or `src' peers.

   * Finally, if the name of the executable file being run has a (system
     dependent) suffix which is not needed to invoke the program, then
     look in a subdirectory (of the one containing the executable file)
     named for the executable (without the suffix); And look in a
     GENERIC_NAME subdirectory.  For example, the executable might be
     `C:\foo\bar.exe' and the initialization file in `C:\foo\bar\'.


File: scm.info,  Node: Executable Pathname,  Next: Script Support,  Prev: File-System Habitat,  Up: Program Self-Knowledge

6.3.2 Executable Pathname
-------------------------

For purposes of finding `Init5e2.scm', dumping an executable, and             |
dynamic linking, a SCM session needs the pathname of its executable
image.

When a program is executed by MS-DOS, the full pathname of that
executable is available in `argv[0]'.  This value can be passed
directly to `find_impl_file' (*note File-System Habitat::).

In order to find the habitat for a unix program, we first need to know
the full pathname for the associated executable file.

 -- Function: char * dld_find_executable (const char *COMMAND)
     `dld_find_executable' returns the absolute path name of the file
     that would be executed if COMMAND were given as a command.  It
     looks up the environment variable PATH, searches in each of the
     directory listed for COMMAND, and returns the absolute path name
     for the first occurrence.  Thus, it is advisable to invoke
     `dld_init' as:

          main (int argc, const char **argv)                                  |
          {
              ...
              if (dld_init (dld_find_executable (argv[0]))) {
                  ...
              }
              ...
          }

          *Note@:* If the current process is executed using the
          `execve' call without passing the correct path name as
          argument 0, `dld_find_executable (argv[0]) ' will also fail
          to locate the executable file.

     `dld_find_executable' returns zero if `command' is not found in
     any of the directories listed in `PATH'.


File: scm.info,  Node: Script Support,  Prev: Executable Pathname,  Up: Program Self-Knowledge

6.3.3 Script Support
--------------------

Source code for these C functions is in the file `script.c'.  *Note
Scripting:: for a description of script argument processing.

`script_find_executable' is only defined on unix systems.

 -- Function: char * script_find_executable (const char *NAME)
     `script_find_executable' returns the path name of the executable
     which is invoked by the script file NAME; NAME if it is a binary
     executable (not a script); or 0 if NAME does not exist or is not
     executable.

 -- Function: char ** script_process_argv (int ARGC; char **ARGV)
     Given an "main" style argument vector ARGV and the number of
     arguments, ARGC, `script_process_argv' returns a newly allocated
     argument vector in which the second line of the script being
     invoked is substituted for the corresponding meta-argument.  

     If the script does not have a meta-argument, or if the file named
     by the argument following a meta-argument cannot be opened for
     reading, then 0 is returned.

     `script_process_argv' correctly processes argument vectors of
     nested script invocations.

 -- Function: int script_count_argv (char **ARGV)
     Returns the number of argument strings in ARGV.


File: scm.info,  Node: Improvements To Make,  Prev: Program Self-Knowledge,  Up: The Implementation

6.4 Improvements To Make
========================

   * Allow users to set limits for `malloc()' storage.

   * Prefix and make more uniform all C function, variable, and constant
     names.  Provide a file full of #define's to provide backward
     compatability.

   * `lgcd()' _needs_ to generate at most one bignum, but currently
     generates more.

   * `divide()' could use shifts instead of multiply and divide when
     scaling.

   * Currently, `dump'ing an executable does not preserve ports.  When
     loading a `dump'ed executable, disk files could be reopened to the
     same file and position as they had when the executable was dumped.

   * Copying all of the stack is wasteful of storage.  Any time a
     call-with-current-continuation is called the stack could be
     re-rooted with a frame which calls the contin just created.  This
     in combination with checking stack depth could also be used to
     allow stacks deeper than 64K on the IBM PC.

   * In the quest for speed, there has been some discussion about a
     "Forth" style Scheme interpreter.

          Provided there is still type code space available in SCM, if
          we devote some of the IMCAR codes to "inlined" operations, we
          should get a significant performance boost.  What is
          eliminated is the having to look up a `GLOC' or `ILOC' and
          then dispatch on the subr type.  The IMCAR operation would be
          dispatched to directly.  Another way to view this is that we
          make available special form versions of `CAR', `CDR', etc.
          Since the actual operation code is localized in the
          interpreter, it is much easier than uncompilation and then
          recompilation to handle `(trace car)'; For instance a switch
          gets set which tells the interpreter to instead always look
          up the values of the associated symbols.

   * Scott Schwartz <schwartz@galapagos.cse.psu.edu> suggests: One way
     to tidy up the dynamic loading stuff would be to grab the code
     from perl5.


* Menu:

* VMS Dynamic Linking::         Finishing the job.


File: scm.info,  Node: VMS Dynamic Linking,  Prev: Improvements To Make,  Up: Improvements To Make

6.4.1 VMS Dynamic Linking
-------------------------

George Carrette (gjc@mitech.com) outlines how to dynamically link on
VMS.  There is already some code in `dynl.c' to do this, but someone
with a VMS system needs to finish and debug it.

  1. Say you have this `main.c' program:

     main()
     {init_lisp();
      lisp_repl();}

  2. and you have your lisp in files `repl.c', `gc.c', `eval.c' and
     there are some toplevel non-static variables in use called
     `the_heap', `the_environment', and some read-only toplevel
     structures, such as `the_subr_table'.

     $ LINK/SHARE=LISPRTL.EXE/DEBUG REPL.OBJ,GC.OBJ,EVAL.OBJ,LISPRTL.OPT/OPT

  3. where `LISPRTL.OPT' must contain at least this:

     SYS$LIBRARY:VAXCRTL/SHARE
     UNIVERSAL=init_lisp
     UNIVERSAL=lisp_repl
     PSECT_ATTR=the_subr_table,SHR,NOWRT,LCL
     PSECT_ATTR=the_heap,NOSHR,LCL
     PSECT_ATTR=the_environment,NOSHR,LCL

     _Notice_ The "psect" (Program Section) attributes.
    `LCL'
          means to keep the name local to the shared library.  You
          almost always want to do that for a good clean library.

    `SHR,NOWRT'
          means shared-read-only.  Which is the default for code, and
          is also good for efficiency of some data structures.

    `NOSHR,LCL'
          is what you want for everything else.

     Note: If you do not have a handy list of all these toplevel
     variables, do not dispair.  Just do your link with the
     /MAP=LISPRTL.MAP/FULL and then search the map file,

     $SEARCH/OUT=LISPRTL.LOSERS LISPRTL.MAP  ",  SHR,NOEXE,  RD,  WRT"

     And use an emacs keyboard macro to muck the result into the proper
     form.  Of course only the programmer can tell if things can be
     made read-only.  I have a DCL command procedure to do this if you
     want it.

  4. Now MAIN.EXE would be linked thusly:

     $ DEFINE LISPRTL USER$DISK:[JAFFER]LISPRTL.EXE

     $LINK MAIN.OBJ,SYS$INPUT:/OPT
      SYS$LIBRARY:VAXCRTL/SHARE
      LISPRTL/SHARE

     Note the definition of the `LISPRTL' logical name.  Without such a
     definition you will need to copy `LISPRTL.EXE' over to `SYS$SHARE'
     (aka `SYS$LIBRARY') in order to invoke the main program once it is
     linked.

  5. Now say you have a file of optional subrs, `MYSUBRS.C'.  And there
     is a routine `INIT_MYSUBRS' that must be called before using it.

     $ CC MYSUBRS.C
     $ LINK/SHARE=MYSUBRS.EXE MYSUBRS.OBJ,SYS$INPUT:/OPT
       SYS$LIBRARY:VAXCRTL/SHARE
       LISPRTL/SHARE
       UNIVERSAL=INIT_MYSUBRS

     Ok.  Another hint is that you can avoid having to add the `PSECT'
     declaration of `NOSHR,LCL' by declaring variables `status' in the
     C language source.  That works great for most things.

  6. Then the dynamic loader would have to do this:

     {void (*init_fcn)();
      long retval;
      retval = lib$find_image_symbol("MYSUBRS","INIT_MYSUBRS",&init_fcn,
                                     "SYS$DISK:[].EXE");
      if (retval != SS$_NORMAL) error(...);
      (*init_fcn)();}

     But of course all string arguments must be `(struct dsc$descriptor
     *)' and the last argument is optional if `MYSUBRS' is defined as a
     logical name or if `MYSUBRS.EXE' has been copied over to
     `SYS$SHARE'.  The other consideration is that you will want to turn
     off <C-c> or other interrupt handling while you are inside most
     `lib$' calls.

     As far as the generation of all the `UNIVERSAL=...' declarations.
     Well, you could do well to have that automatically generated from
     the public `LISPRTL.H' file, of course.

     VMS has a good manual called the `Guide to Writing Modular
     Procedures' or something like that, which covers this whole area
     rather well, and also talks about advanced techniques, such as a
     way to declare a program section with a pointer to a procedure
     that will be automatically invoked whenever any shared image is
     dynamically activated.  Also, how to set up a handler for normal
     or abnormal program exit so that you can clean up side effects
     (such as opening a database).  But for use with `LISPRTL' you
     probably don't need that hair.

     One fancier option that is useful under VMS for `LISPLIB.EXE' is to
     define all your exported procedures through an "call vector"
     instead of having them just be pointers into random places in the
     image, which is what you get by using `UNIVERSAL'.

     If you set up the call vector thing correctly it will allow you to
     modify and relink `LISPLIB.EXE' without having to relink programs
     that have been linked against it.


File: scm.info,  Node: Index,  Prev: The Implementation,  Up: Top

Procedure and Macro Index
*************************

This is an alphabetical list of all the procedures and macros in SCM.