\input texinfo @c -*-texinfo-*-
@c %**start of header
@setfilename scm.info
@settitle scm
@include version.txi
@setchapternewpage on
@c Choices for setchapternewpage are {on,off,odd}.
@paragraphindent 0
@defcodeindex ft
@syncodeindex ft tp
@c %**end of header
@copying
@noindent
This manual is for SCM (version @value{SCMVERSION}, @value{SCMDATE}),
an implementation of the algorithmic language Scheme.
@noindent
Copyright @copyright{} 1990-2007 Free Software Foundation, Inc.
@quotation
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.2 or
any later version published by the Free Software Foundation; with no
Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A
copy of the license is included in the section entitled ``GNU Free
Documentation License.''
@end quotation
@end copying
@dircategory The Algorithmic Language Scheme
@direntry
* SCM: (scm). A Scheme interpreter.
@end direntry
@iftex
@finalout
@c DL: lose the egregious vertical whitespace, esp. around examples
@c but paras in @defun-like things don't have parindent
@parskip 4pt plus 1pt
@end iftex
@titlepage
@title SCM
@subtitle Scheme Implementation
@subtitle Version @value{SCMVERSION}
@author Aubrey Jaffer
@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage
@contents
@ifnottex
@node Top, Overview, (dir), (dir)
@top SCM
@insertcopying
@menu
* Overview::
* Installing SCM:: How to
* Operational Features::
* The Language:: Reference.
* Packages:: Optional Capabilities.
* The Implementation:: How it works.
* Index::
@end menu
@end ifnottex
@node Overview, Installing SCM, Top, Top
@chapter Overview
@noindent
SCM is a portable Scheme implementation written in C. SCM provides a
machine independent platform for [JACAL], a symbolic algebra system.
SCM supports and requires the SLIB Scheme library. SCM, SLIB, and
JACAL are GNU projects.
@iftex
@noindent
The most recent information about SCM can be found on SCM's @dfn{WWW}
home page:
@ifset html
@end ifset
@center @url{http://swiss.csail.mit.edu/~jaffer/SCM}
@ifset html
@end ifset
@end iftex
@menu
* SCM Features::
* SCM Authors::
* Copying::
* Bibliography::
@end menu
@node SCM Features, SCM Authors, Overview, Overview
@section Features
@itemize @bullet
@item
Conforms to Revised^5 Report on the Algorithmic Language Scheme [R5RS]
and the [IEEE] P1178 specification.
@item
Support for [SICP], [R2RS], [R3RS], and [R5RS] scheme code.
@item
Runs under Amiga, Atari-ST, MacOS, MS-DOS, OS/2, NOS/VE, Unicos, VMS,
Unix and similar systems. Supports ASCII and EBCDIC character sets.
@item
Is fully documented in @TeX{}info form, allowing documentation to be
generated in info, @TeX{}, html, nroff, and troff formats.
@item
Supports inexact real and complex numbers, 30 bit immediate integers and
large precision integers.
@item
Many Common Lisp functions: @code{logand}, @code{logor}, @code{logxor},
@code{lognot}, @code{ash}, @code{logcount}, @code{integer-length},
@code{bit-extract}, @code{defmacro}, @code{macroexpand},
@code{macroexpand1}, @code{gentemp}, @code{defvar}, @code{force-output},
@code{software-type}, @code{get-decoded-time},
@code{get-internal-run-time}, @code{get-internal-real-time},
@code{delete-file}, @code{rename-file}, @code{copy-tree}, @code{acons},
and @code{eval}.
@item
@code{Char-code-limit}, @code{most-positive-fixnum},
@code{most-negative-fixnum}, @code{and internal-time-units-per-second}
constants. @code{slib:features} and @code{*load-pathname*} variables.
@item
Arrays and bit-vectors. String ports and software emulation ports.
I/O extensions providing ANSI C and POSIX.1 facilities.
@item
Interfaces to standard libraries including REGEX string regular
expression matching and the CURSES screen management package.
@item
Available add-on packages including an interactive debugger, database,
X-window graphics, BGI graphics, Motif, and Open-Windows packages.
@item
The Hobbit compiler and dynamic linking of compiled modules.
@item
User definable responses to interrupts and errors,
Process-syncronization primitives. Setable levels of monitoring and
timing information printed interactively (the @code{verbose} function).
@code{Restart}, @code{quit}, and @code{exec}.
@end itemize
@node SCM Authors, Copying, SCM Features, Overview
@section Authors
@table @b
@item Aubrey Jaffer (agj @@ alum.mit.edu)
Most of SCM.
@item Radey Shouman
Arrays, @code{gsubr}s, compiled closures, records, Ecache, syntax-rules
macros, and @dfn{safeport}s.
@item Jerry D. Hedden
Real and Complex functions. Fast mixed type arithmetics.
@item Hugh Secker-Walker
Syntax checking and memoization of special forms by evaluator. Storage
allocation strategy and parameters.
@item George Carrette
@dfn{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
(@pxref{Garbage Collection}).
@end table
@noindent
There are many other contributors to SCM. They are acknowledged in the
file @file{ChangeLog}, a log of changes that have been made to scm.
@node Copying, Bibliography, SCM Authors, Overview
@section Copyright
@noindent
Authors have assigned their SCM copyrights to:
@sp 1
@center Free Software Foundation, Inc.
@center 59 Temple Place, Suite 330, Boston, MA 02111, USA
@menu
* The SCM License::
* SIOD copyright::
* GNU Free Documentation License:: Copying this Manual
@end menu
@node The SCM License, SIOD copyright, Copying, Copying
@subsection The SCM License
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Lesser General Public License as
published by the Free Software Foundation, either version 3 of the
License, 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this program. If not, see
@url{http://www.gnu.org/licenses/}.
@node SIOD copyright, GNU Free Documentation License, The SCM License, Copying
@subsection SIOD copyright
@sp 1
@center COPYRIGHT @copyright{} 1989 BY
@center PARADIGM ASSOCIATES INCORPORATED, CAMBRIDGE, MASSACHUSETTS.
@center ALL RIGHTS RESERVED
@noindent
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.
@noindent
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.
@noindent
gjc@@paradigm.com
@flushright
Phone: 617-492-6079
@end flushright
@flushleft
Paradigm Associates Inc
29 Putnam Ave, Suite 6
Cambridge, MA 02138
@end flushleft
@node GNU Free Documentation License, , SIOD copyright, Copying
@subsection GNU Free Documentation License
@include fdl.texi
@node Bibliography, , Copying, Overview
@section Bibliography
@table @asis
@item [IEEE]
@cindex IEEE
@cite{IEEE Standard 1178-1990. IEEE Standard for the Scheme
Programming Language.} IEEE, New York, 1991.
@item [R4RS]
@cindex R4RS
William Clinger and Jonathan Rees, Editors.
@ifset html
@end ifset
Revised(4) Report on the Algorithmic Language Scheme.
@ifset html
@end ifset
@cite{ACM Lisp Pointers} Volume IV, Number 3 (July-September 1991),
pp. 1-55.
@ifinfo
@ref{Top, , , r4rs, Revised(4) Report on the Algorithmic Language Scheme}.
@end ifinfo
@item [R5RS]
@cindex R5RS
Richard Kelsey and William Clinger and Jonathan (Rees, editors)
@ifset html
@end ifset
Revised(5) Report on the Algorithmic Language Scheme.
@ifset html
@end ifset
@cite{Higher-Order and Symbolic Computation} Volume 11, Number 1 (1998),
pp. 7-105, and
@cite{ACM SIGPLAN Notices} 33(9), September 1998.
@ifinfo
@ref{Top, , , r5rs, Revised(5) Report on the Algorithmic Language Scheme}.
@end ifinfo
@item [Exrename]
@cindex Exrename
William Clinger
@ifset html
@end ifset
Hygienic Macros Through Explicit Renaming
@ifset html
@end ifset
@cite{Lisp Pointers} Volume IV, Number 4 (December 1991),
pp 17-23.
@item [SICP]
@cindex SICP
Harold Abelson and Gerald Jay Sussman with Julie Sussman.
@cite{Structure and Interpretation of Computer Programs.}
MIT Press, Cambridge, 1985.
@item [Simply]
@cindex Simply
Brian Harvey and Matthew Wright.
@ifset html
@end ifset
@cite{Simply Scheme: Introducing Computer Science}
@ifset html
@end ifset
MIT Press, 1994 ISBN 0-262-08226-8
@item [SchemePrimer]
@cindex SchemePrimer
犬飼大(Dai Inukai)
@ifset html
@end ifset
@cite{入門Scheme}
@ifset html
@end ifset
1999年12月初版 ISBN4-87966-954-7
@c @item [GUILE]
@c @cindex GUILE
@c Free Software Foundation
@c @ifset html
@c
@c @end ifset
@c Guile: Project GNU's extension language
@c @ifset html
@c
@c @end ifset
@item [SLIB]
@cindex SLIB
Todd R. Eigenschink, Dave Love, and Aubrey Jaffer.
@ifset html
@end ifset
SLIB, The Portable Scheme Library.
@ifset html
@end ifset
Version 2c8, June 2000.
@ifinfo
@ref{Top, , , slib, SLIB}.
@end ifinfo
@item [JACAL]
@cindex JACAL
Aubrey Jaffer.
@ifset html
@end ifset
JACAL Symbolic Mathematics System.
@ifset html
@end ifset
Version 1b0, Sep 1999.
@ifinfo
@ref{Top, , , jacal, JACAL}.
@end ifinfo
@end table
@table @file
@item scm.texi
@itemx scm.info
Documentation of @code{scm} extensions (beyond Scheme standards).
Documentation on the internal representation and how to extend or
include @code{scm} in other programs.
@item Xlibscm.texi
@itemx Xlibscm.info
Documentation of the Xlib - SCM Language X Interface.
@end table
@node Installing SCM, Operational Features, Overview, Top
@chapter 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::
@end menu
@node Making SCM, SLIB, Installing SCM, Installing SCM
@section Making SCM
The SCM distribution has @dfn{Makefile} which contains rules for
making @dfn{scmlit}, a ``bare-bones'' version of SCM sufficient for
running @file{build}. @file{build} is used to compile (or create
scripts to compile) full featured versions (@pxref{Building SCM}).
Makefiles are not portable to the majority of platforms. If
@file{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:
@itemize @bullet
@item
Use the @uref{http://swiss.csail.mit.edu/~jaffer/buildscm.html, build}
web page to create custom batch scripts for compiling SCM.
@item
Use SCM on a different platform to run @file{build} to create a script
to build SCM;
@item
Use another implementation of Scheme to run @file{build} to create a
script to build SCM;
@item
Create your own script or @file{Makefile}.
@end itemize
@node SLIB, Building SCM, Making SCM, Installing SCM
@section SLIB
@noindent
[SLIB] is a portable Scheme library meant to provide compatibility and
utility functions for all standard Scheme implementations. Although
SLIB is not @emph{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:
@ifclear html
@itemize @bullet
@item
swiss.csail.mit.edu:/pub/scm/slib-3b1.tar.gz
@item
ftp.gnu.org:/pub/gnu/jacal/slib-3b1.tar.gz
@end itemize
@end ifclear
@ifset html
@itemize @bullet
@item
http://swiss.csail.mit.edu/ftpdir/scm/slib-3b1.zip
@item
ftp.gnu.org:/pub/gnu/jacal/slib-3b1.tar.gz
@end itemize
@end ifset
@noindent
Unpack SLIB (@samp{tar xzf slib-3b1.tar.gz} or @samp{unzip -ao
slib-3b1.zip}) in an appropriate directory for your system; both
@code{tar} and @code{unzip} will create the directory @file{slib}.
@noindent
Then create a file @file{require.scm} in the SCM
@dfn{implementation-vicinity} (this is the same directory as where the
file @file{Init@value{SCMVERSION}.scm} is installed).
@file{require.scm} should have the contents:
@example
(define (library-vicinity) "/usr/local/lib/slib/")
@end example
@noindent
where the pathname string @file{/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 (@pxref{I/O-Extensions, chmod}).
The way to specify a relative pathname is to append it to the
implementation-vicinity, which is absolute:
@example
(define library-vicinity
(let ((lv (string-append (implementation-vicinity) "../slib/")))
(lambda () lv)))
@end example
@noindent
Alternatively, you can set the (shell) environment variable
@code{SCHEME_LIBRARY_PATH} to the pathname of the SLIB directory
(@pxref{SCM Variables, SCHEME_LIBRARY_PATH, Environment Variables}). If
set, the environment variable overrides @file{require.scm}. Again,
absolute pathnames are recommended.
@node Building SCM, Installing Dynamic Linking, SLIB, Installing SCM
@section Building SCM
@cindex build
@cindex build.scm
The file @dfn{build} loads the file @dfn{build.scm}, which constructs a
relational database of how to compile and link SCM executables.
@file{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::
@end menu
@node Invoking Build, Build Options, Building SCM, Building SCM
@subsection Invoking Build
@noindent
The @emph{all} method will also work for MS-DOS and unix. Use
the @emph{all} method if you encounter problems with @file{build}.
@table @asis
@item MS-DOS
From the SCM source directory, type @samp{build} followed by up to 9
command line arguments.
@item unix
From the SCM source directory, type @samp{./build} followed by command
line arguments.
@item @emph{all}
From the SCM source directory, start @samp{scm} or @samp{scmlit} and
type @code{(load "build")}. Alternatively, start @samp{scm} or
@samp{scmlit} with the command line argument @samp{-ilbuild}.
@end table
@noindent
Invoking build without the @samp{-F} option will build or create a shell
script with the @code{arrays}, @code{inexact}, and @code{bignums}
options as defaults.
@example
bash$ ./build
@print{}
#! /bin/sh
# unix (linux) script created by SLIB/batch
# ================ Write file with C defines
rm -f scmflags.h
echo '#define IMPLINIT "Init@value{SCMVERSION}.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
@end example
@noindent
To cross compile for another platform, invoke build with the @samp{-p}
or @samp{--platform=} option. This will create a script for the
platform named in the @samp{-p} or @samp{--platform=} option.
@example
bash$ ./build -o scmlit -p darwin -F lit
@print{}
#! /bin/sh
# unix (darwin) script created by SLIB/batch
# ================ Write file with C defines
rm -f scmflags.h
echo '#define IMPLINIT "Init@value{SCMVERSION}.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
@end example
@node Build Options, Compiling and Linking Custom Files, Invoking Build, Building SCM
@subsection Build Options
@noindent
The options to @dfn{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.
@deffn {Build Option} -p @var{platform-name}
@deffnx {Build Option} ---platform=@var{platform-name}
specifies that the compilation should be for a computer/operating-system
combination called @var{platform-name}. @emph{Note@:} The case of
@var{platform-name} is distinguised. The current @var{platform-name}s
are all lower-case.
The platforms defined by table @dfn{platform} in @file{build.scm} are:
@end deffn
@example
@include platform.txi
@end example
@deffn {Build Option} -f @var{pathname}
specifies that the build options contained in @var{pathname} be
spliced into the argument list at this point. The use of option files
can separate functional features from platform-specific ones.
The @file{Makefile} calls out builds with the options in @samp{.opt}
files:
@table @file
@item dlls.opt
Options for Makefile targets mydlls, myturtle, and x.so.
@item gdb.opt
Options for udgdbscm and gdbscm.
@item libscm.opt
Options for libscm.a.
@item pg.opt
Options for pgscm, which instruments C functions.
@item udscm4.opt
Options for targets udscm4 and dscm4 (scm).
@item udscm5.opt
Options for targets udscm5 and dscm5 (scm).
@end table
The Makefile creates options files it depends on only if they do not
already exist.
@end deffn
@deffn {Build Option} -o @var{filename}
@deffnx {Build Option} ---outname=@var{filename}
specifies that the compilation should produce an executable or object
name of @var{filename}. The default is @samp{scm}. Executable suffixes
will be added if neccessary, e.g. @samp{scm} @result{} @samp{scm.exe}.
@end deffn
@deffn {Build Option} -l @var{libname} @dots{}
@deffnx {Build Option} ---libraries=@var{libname}
specifies that the @var{libname} should be linked with the executable
produced. If compile flags or include directories (@samp{-I}) are
needed, they are automatically supplied for compilations. The @samp{c}
library is always included. SCM @dfn{features} specify any libraries
they need; so you shouldn't need this option often.
@end deffn
@deffn {Build Option} -D @var{definition} @dots{}
@deffnx {Build Option} ---defines=@var{definition}
specifies that the @var{definition} should be made in any C source
compilations. If compile flags or include directories (@samp{-I}) are
needed, they are automatically supplied for compilations. SCM
@dfn{features} specify any flags they need; so you shouldn't need this
option often.
@end deffn
@deffn {Build Option} ---compiler-options=@var{flag}
specifies that that @var{flag} will be put on compiler command-lines.
@end deffn
@deffn {Build Option} ---linker-options=@var{flag}
specifies that that @var{flag} will be put on linker command-lines.
@end deffn
@deffn {Build Option} -s @var{pathname}
@deffnx {Build Option} ---scheme-initial=@var{pathname}
specifies that @var{pathname} should be the default location of the SCM
initialization file @file{Init@value{SCMVERSION}.scm}. SCM tries
several likely locations before resorting to @var{pathname}
(@pxref{File-System Habitat}). If not specified, the current directory
(where build is building) is used.
@end deffn
@deffn {Build Option} -c @var{pathname} @dots{}
@deffnx {Build Option} ---c-source-files=@var{pathname}
specifies that the C source files @var{pathname} @dots{} are to be
compiled.
@end deffn
@deffn {Build Option} -j @var{pathname} @dots{}
@deffnx {Build Option} ---object-files=@var{pathname}
specifies that the object files @var{pathname} @dots{} are to be linked.
@end deffn
@deffn {Build Option} -i @var{call} @dots{}
@deffnx {Build Option} ---initialization=@var{call}
specifies that the C functions @var{call} @dots{} are to be
invoked during initialization.
@end deffn
@deffn {Build Option} -t @var{build-what}
@deffnx {Build Option} ---type=@var{build-what}
specifies in general terms what sort of thing to build. The choices
are:
@table @samp
@item exe
executable program.
@item lib
library module.
@item dlls
archived dynamically linked library object files.
@item dll
dynamically linked library object file.
@end table
The default is to build an executable.
@end deffn
@deffn {Build Option} -h @var{batch-syntax}
@deffnx {Build Option} --batch-dialect=@var{batch-syntax}
specifies how to build. The default is to create a batch file for the
host system. The SLIB file @file{batch.scm} knows how to create batch
files for:
@itemize @bullet
@item
unix
@item
dos
@item
vms
@item
amigaos (was amigados)
@item
system
This option executes the compilation and linking commands through the
use of the @code{system} procedure.
@item
*unknown*
This option outputs Scheme code.
@end itemize
@end deffn
@deffn {Build Option} -w @var{batch-filename}
@deffnx {Build Option} --script-name=@var{batch-filename}
specifies where to write the build script. The default is to display it
on @code{(current-output-port)}.
@end deffn
@deffn {Build Option} -F @var{feature} @dots{}
@deffnx {Build Option} ---features=@var{feature}
specifies to build the given features into the executable. The defined
features are:
@table @dfn
@c @itemx none
@c @cindex none
@c Lightweight -- no features
@include features.txi
@end table
@end deffn
@node Compiling and Linking Custom Files, , Build Options, Building SCM
@subsection Compiling and Linking Custom Files
@noindent
A correspondent asks:
@quotation
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?
@end quotation
@noindent
(@pxref{Changing Scm} has instructions describing the C code format).
@cindex foo.c
@cindex Extending Scm
Suppose a C file @dfn{foo.c} has functions you wish to add to SCM. To
compile and link your file at compile time, use the @samp{-c} and
@samp{-i} options to build:
@example
bash$ ./build -c foo.c -i init_foo
@print{}
#! /bin/sh
rm -f scmflags.h
echo '#define IMPLINIT "/home/jaffer/scm/Init@value{SCMVERSION}.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
@end example
@noindent
To make a dynamically loadable object file use the @code{-t dll} option:
@example
bash$ ./build -t dll -c foo.c
@print{}
#! /bin/sh
rm -f scmflags.h
echo '#define IMPLINIT "/home/jaffer/scm/Init@value{SCMVERSION}.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
@end example
@noindent
Once @file{foo.c} compiles correctly (and your SCM build supports
dynamic-loading), you can load the compiled file with the Scheme command
@code{(load "./foo.so")}. See @ref{Configure Module Catalog} for how to
add a compiled dll file to SLIB's catalog.
@node Installing Dynamic Linking, Configure Module Catalog, Building SCM, Installing SCM
@section Installing Dynamic Linking
@noindent
Dynamic linking has not been ported to all platforms. Operating systems
in the BSD family (a.out binary format) can usually be ported to
@dfn{DLD}. The @dfn{dl} library (@code{#define SUN_DL} for SCM) was a
proposed POSIX standard and may be available on other machines with
@dfn{COFF} binary format. For notes about porting to MS-Windows and
finishing the port to VMS @ref{VMS Dynamic Linking}.
@noindent
@dfn{DLD} is a library package of C functions that performs
@dfn{dynamic link editing} on GNU/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:
@ifclear html
@itemize @bullet
@item
ftp.gnu.org:pub/gnu/dld-3.3.tar.gz
@end itemize
@end ifclear
@ifset html
ftp.gnu.org:pub/gnu/dld-3.3.tar.gz
@end ifset
@noindent
These notes about using libdl on SunOS are from @file{gcc.info}:
@quotation
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 @file{libdl.a} with some
versions of SunOS (mainly 4.1). This results in undefined symbols when
linking static binaries (that is, if you use @samp{-static}). If you
see undefined symbols @samp{_dlclose}, @samp{_dlsym} or @samp{_dlopen}
when linking, compile and link against the file
@file{mit/util/misc/dlsym.c} from the MIT version of X windows.
@end quotation
@node Configure Module Catalog, Saving Images, Installing Dynamic Linking, Installing SCM
@section Configure Module Catalog
@noindent
The SLIB module @dfn{catalog} can be extended to define other
@code{require}-able packages by adding calls to the Scheme source file
@file{mkimpcat.scm}. Within @file{mkimpcat.scm}, the following
procedures are defined.
@defun add-link feature object-file lib1 @dots{}
@var{feature} should be a symbol. @var{object-file} should be a string
naming a file containing compiled @dfn{object-code}. Each @var{lib}n
argument should be either a string naming a library file or @code{#f}.
If @var{object-file} exists, the @code{add-link} procedure registers
symbol @var{feature} so that the first time @code{require} is called
with the symbol @var{feature} as its argument, @var{object-file} and the
@var{lib1} @dots{} are dynamically linked into the executing SCM
session.
If @var{object-file} exists, @code{add-link} returns @code{#t},
otherwise it returns @code{#f}.
For example, to install a compiled dll @file{foo}, add these lines to
@file{mkimpcat.scm}:
@example
(add-link 'foo
(in-vicinity (implementation-vicinity) "foo"
link:able-suffix))
@end example
@end defun
@defun add-alias alias feature
@var{alias} and @var{feature} are symbols. The procedure
@code{add-alias} registers @var{alias} as an alias for @var{feature}.
An unspecified value is returned.
@code{add-alias} causes @code{(require '@var{alias})} to behave like
@code{(require '@var{feature})}.
@end defun
@defun add-source feature filename
@var{feature} is a symbol. @var{filename} is a string naming a file
containing Scheme source code. The procedure @code{add-source}
registers @var{feature} so that the first time @code{require} is called
with the symbol @var{feature} as its argument, the file @var{filename}
will be @code{load}ed. An unspecified value is returned.
@end defun
@noindent
Remember to delete the file @file{slibcat} after modifying the file
@file{mkimpcat.scm} in order to force SLIB to rebuild its cache.
@node Saving Images, Automatic C Preprocessor Definitions, Configure Module Catalog, Installing SCM
@section Saving Images
In SCM, the ability to save running program images is called @dfn{dump}
(@pxref{Dump}). In order to make @code{dump} available to SCM, build
with feature @samp{dump}. @code{dump}ed executables are compatible with
dynamic linking.
Most of the code for @dfn{dump} is taken from
@file{emacs-19.34/src/unex*.c}. No modifications to the emacs source
code were required to use @file{unexelf.c}. Dump has not been ported to
all platforms. If @file{unexec.c} or @file{unexelf.c} don't work for
you, try using the appropriate @file{unex*.c} file from emacs.
The @samp{dscm4} and @samp{dscm5} targets in the SCM @file{Makefile}
save images from @file{udscm4} and @file{udscm5} executables
respectively.
Recent GNU/Linux innovations interfere with @code{dump}. For:
@table @asis
@item Fedora-Core-1
Remove the @samp{#} from the line @samp{#SETARCH = setarch i386} in
the @file{Makefile}.
@item Fedora-Core-3
@url{http://jamesthornton.com/writing/emacs-compile.html} writes:
[For FC3] combreloc has become the default for recent GNU ld, which
breaks the unexec/undump on all versions of both Emacs and
XEmacs...
Override by adding the following to @file{udscm5.opt}:
@samp{--linker-options="-z nocombreloc"}
@item Kernels later than 2.6.11
@url{http://www.opensubscriber.com/message/emacs-devel@@gnu.org/1007118.html}
mentions the @dfn{exec-shield} feature. Kernels later than 2.6.11
must do (as root):
@example
echo 0 > /proc/sys/kernel/randomize_va_space
@end example
before dumping. @file{Makefile} has this @file{randomize_va_space}
stuffing scripted for targets @samp{dscm4} and @samp{dscm5}. You must
either set @file{randomize_va_space} to 0 or run as root to dump.
@end table
@node Automatic C Preprocessor Definitions, Problems Compiling, Saving Images, Installing SCM
@section Automatic C Preprocessor Definitions
These @samp{#defines} are automatically provided by preprocessors of
various C compilers. SCM uses the presence or absence of these
definitions to configure @dfn{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 @code{-D@var{flag}} to
the compilation command lines or add a @code{#define @var{flag}} line to
@file{scmfig.h} or the beginning of @file{scmfig.h}.
@example
#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
__DragonflyBSD__ DragonflyBSD
__FreeBSD__ FreeBSD
GNUDOS DJGPP (obsolete in version 1.08)
__GO32__ DJGPP (future?)
hpux HP-UX
linux GNU/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
__OpenBSD__ OpenBSD
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
@end example
@node Problems Compiling, Problems Linking, Automatic C Preprocessor Definitions, Installing SCM
@section Problems Compiling
@multitable @columnfractions .10 .45 .45
@item FILE
@tab PROBLEM / MESSAGE
@tab HOW TO FIX
@item *.c
@tab include file not found.
@tab Correct the status of @t{STDC_HEADERS} in scmfig.h.
@item
@tab
@tab fix @t{#include} statement or add @t{#define} for system type to scmfig.h.
@item *.c
@tab Function should return a value.
@tab Ignore.
@item
@tab Parameter is never used.
@tab
@item
@tab Condition is always false.
@tab
@item
@tab Unreachable code in function.
@tab
@item scm.c
@tab assignment between incompatible types.
@tab Change @t{SIGRETTYPE} in scm.c.
@item time.c
@tab CLK_TCK redefined.
@tab incompatablility between and .
@item
@tab
@tab Remove @t{STDC_HEADERS} in scmfig.h.
@item
@tab
@tab Edit to remove incompatability.
@item subr.c
@tab Possibly incorrect assignment in function lgcd.
@tab Ignore.
@item sys.c
@tab statement not reached.
@tab Ignore.
@item
@tab constant in conditional expression.
@tab
@item sys.c
@tab undeclared, outside of functions.
@tab @t{#undef STDC_HEADERS} in scmfig.h.
@item scl.c
@tab syntax error.
@tab @t{#define SYSTNAME} to your system type in scl.c (softtype).
@end multitable
@node Problems Linking, Problems Running, Problems Compiling, Installing SCM
@section Problems Linking
@multitable @columnfractions .5 .5
@item PROBLEM
@tab HOW TO FIX
@item _sin etc. missing.
@tab Uncomment @t{LIBS} in makefile.
@end multitable
@node Problems Running, Testing, Problems Linking, Installing SCM
@section Problems Running
@multitable @columnfractions .5 .5
@item PROBLEM
@tab HOW TO FIX
@item Opening message and then machine crashes.
@tab Change memory model option to C compiler (or makefile).
@item
@tab Make sure @t{sizet} definition is correct in scmfig.h.
@item
@tab Reduce the size of @t{HEAP_SEG_SIZE} in setjump.h.
@item Input hangs.
@tab @t{#define NOSETBUF}
@item ERROR: heap: need larger initial.
@tab Increase initial heap allocation using -a or @t{INIT_HEAP_SIZE}.
@item ERROR: Could not allocate.
@tab Check @t{sizet} definition.
@item
@tab Use 32 bit compiler mode.
@item
@tab Don't try to run as subproccess.
@item remove in scmfig.h and recompile scm.
@tab Do so and recompile files.
@item add in scmfig.h and recompile scm.
@tab
@item ERROR: Init@value{SCMVERSION}.scm not found.
@tab Assign correct @t{IMPLINIT} in makefile or scmfig.h.
@item
@tab Define environment variable @t{SCM_INIT_PATH} to be the full pathname of Init@value{SCMVERSION}.scm.
@item WARNING: require.scm not found.
@tab Define environment variable @t{SCHEME_LIBRARY_PATH} to be the full pathname of the scheme library [SLIB].
@item
@tab Change @t{library-vicinity} in Init@value{SCMVERSION}.scm to point to library or remove.
@item
@tab Make sure the value of @t{(library-vicinity)} has a trailing file separator (like @t{/} or @t{\}).
@end multitable
@node Testing, Reporting Problems, Problems Running, Installing SCM
@section Testing
@noindent
Loading @file{r4rstest.scm} in the distribution will run an [R4RS]
conformance test on @code{scm}.
@example
> (load "r4rstest.scm")
@print{}
;loading "r4rstest.scm"
SECTION(2 1)
SECTION(3 4)
#
#
#
#
@dots{}
@end example
@noindent
Loading @file{pi.scm} in the distribution will enable you to compute
digits of pi.
@example
> (load "pi")
;loading "pi"
;done loading "pi.scm"
;Evaluation took 20 ms (0 in gc) 767 cells work, 233.B other
#
> (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
#
@end example
@noindent
Loading @file{bench.scm} will compute and display performance statistics
of SCM running @file{pi.scm}. @samp{make bench} or @samp{make benchlit}
appends the performance report to the file @file{BenchLog}, facilitating
tracking effects of changes to SCM on performance.
@multitable @columnfractions .5 .5
@item PROBLEM
@tab HOW TO FIX
@item Runs some and then machine crashes.
@tab See above under machine crashes.
@item Runs some and then ERROR: @dots{} (after a GC has happened).
@tab Remove optimization option to C compiler and recompile.
@item
@tab @t{#define SHORT_ALIGN} in @file{scmfig.h}.
@item Some symbol names print incorrectly.
@tab Change memory model option to C compiler (or makefile).
@item
@tab Check that @t{HEAP_SEG_SIZE} fits within @t{sizet}.
@item
@tab Increase size of @t{HEAP_SEG_SIZE} (or @t{INIT_HEAP_SIZE} if it is smaller than @t{HEAP_SEG_SIZE}).
@item ERROR: Rogue pointer in Heap.
@tab See above under machine crashes.
@item Newlines don't appear correctly in output files.
@tab Check file mode (define OPEN_@dots{} in @file{Init@value{SCMVERSION}.scm}).
@item Spaces or control characters appear in symbol names.
@tab Check character defines in @file{scmfig.h}.
@item Negative numbers turn positive.
@tab Check SRS in @file{scmfig.h}.
@item ;ERROR: bignum: numerical overflow
@tab Increase NUMDIGS_MAX in @file{scmfig.h} and recompile.
@item VMS: Couldn't unwind stack.
@tab @t{#define CHEAP_CONTINUATIONS} in @file{scmfig.h}.
@item VAX: botched longjmp.
@end multitable
@table @asis
@item 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 @emph{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 @file{continue.c}.
@end table
@node Reporting Problems, , Testing, Installing SCM
@section Reporting Problems
@noindent
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 @code{agj @@ alum.mit.edu}. The bug report
should include:
@enumerate
@item
The version of SCM (printed when SCM is invoked with no arguments).
@item
The type of computer you are using.
@item
The name and version of your computer's operating system.
@item
The values of the environment variables @code{SCM_INIT_PATH} and
@code{SCHEME_LIBRARY_PATH}.
@item
The name and version of your C compiler.
@item
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.
@end enumerate
@node Operational Features, The Language, Installing SCM, Top
@chapter 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::
@end menu
@node Invoking SCM, SCM Options, Operational Features, Operational Features
@section Invoking SCM
@example
@exdent @b{ scm } [-a @i{kbytes}] [-muvbiq] @w{[--version]} @w{[--help]}
@w{[[-]-no-init-file]} @w{[--no-symbol-case-fold]}
@w{[-p @i{int}]} @w{[-r @i{feature}]} @w{[-h @i{feature}]}
@w{[-d @i{filename}]} @w{[-f @i{filename}]} @w{[-l @i{filename}]}
@w{[-c @i{expression}]} @w{[-e @i{expression}]} @w{[-o @i{dumpname}]}
@w{[-- | - | -s]} @w{[@i{filename}]} @w{[@i{arguments} @dots{}]}
@end example
@noindent
Upon startup @code{scm} loads the file specified by by the environment
variable @var{SCM_INIT_PATH}.
@noindent
If @var{SCM_INIT_PATH} is not defined or if the file it names is not
present, @code{scm} tries to find the directory containing the
executable file. If it is able to locate the executable, @code{scm}
looks for the initialization file (usually
@file{Init@value{SCMVERSION}.scm}) in platform-dependent directories
relative to this directory. See @ref{File-System Habitat} for a
blow-by-blow description.
@noindent
As a last resort (if initialization file cannot be located), the C
compile parameter @var{IMPLINIT} (defined in the makefile or
@file{scmfig.h}) is tried.
@noindent
Unless the option @code{-no-init-file} or @code{--no-init-file} occurs
in the command line, or if @code{scm} is being invoked as a script,
@file{Init@value{SCMVERSION}.scm} checks to see if there is file
@file{ScmInit.scm} in the path specified by the environment variable
@var{HOME} (or in the current directory if @var{HOME} is undefined).
If it finds such a file, then it is loaded.
@noindent
@file{Init@value{SCMVERSION}.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.
@noindent
This explanation applies to SCMLIT or other builds of SCM.
@noindent
Scheme-code files can also invoke SCM and its variants.
@xref{Lexical Conventions, #!}.
@node SCM Options, Invocation Examples, Invoking SCM, Operational Features
@section Options
@noindent
The options are processed in the order specified on the command line.
@deffn {Command Option} -a k
specifies that @code{scm} should allocate an initial heapsize of @var{k}
kilobytes. This option, if present, must be the first on the command
line. If not specified, the default is @code{INIT_HEAP_SIZE} in source
file @file{setjump.h} which the distribution sets at
@code{25000*sizeof(cell)}.
@end deffn
@deffn {Command Option} -no-init-file
@deffnx {Command Option} ---no-init-file
Inhibits the loading of @file{ScmInit.scm} as described above.
@end deffn
@deffn {Command Option} --no-symbol-case-fold
Symbol (and identifier) names will be case sensitive.
@end deffn
@deffn {Command Option} ---help
prints usage information and URI; then exit.
@end deffn
@deffn {Command Option} ---version
prints version information and exit.
@end deffn
@deffn {Command Option} -r feature
requires @var{feature}. This will load a file from [SLIB] if that
@var{feature} is not already provided. If @var{feature} is 2, 2rs, or
r2rs; 3, 3rs, or r3rs; 4, 4rs, or r4rs; 5, 5rs, or r5rs; @code{scm}
will require the features neccessary to support [R2RS]; [R3RS];
[R4RS]; or [R5RS], respectively.
@end deffn
@deffn {Command Option} -h feature
provides @var{feature}.
@end deffn
@deffn {Command Option} -l filename
@deffnx {Command Option} -f filename
loads @var{filename}. @code{Scm} will load the first (unoptioned)
file named on the command line if no @code{-c}, @code{-e}, @code{-f},
@code{-l}, or @code{-s} option preceeds it.
@end deffn
@deffn {Command Option} -d filename
Loads SLIB @code{databases} feature and opens @var{filename} as a
database.
@end deffn
@deffn {Command Option} -e expression
@deffnx {Command Option} -c expression
specifies that the scheme expression @var{expression} is to be
evaluated. These options are inspired by @code{perl} and @code{sh}
respectively. On Amiga systems the entire option and argument need to be
enclosed in quotes. For instance @samp{"-e(newline)"}.
@end deffn
@deffn {Command Option} -o dumpname
saves the current SCM session as the executable program @file{dumpname}.
This option works only in SCM builds supporting @code{dump}
(@pxref{Dump}).
If options appear on the command line after @samp{-o @var{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.
@end deffn
@deffn {Command Option} -p level
sets the prolixity (verboseness) to @var{level}. This is the same as
the @code{scm} command (verobse @var{level}).
@end deffn
@deffn {Command Option} -v
(verbose mode) specifies that @code{scm} will print prompts, evaluation
times, notice of loading files, and garbage collection statistics. This
is the same as @code{-p3}.
@end deffn
@deffn {Command Option} -q
(quiet mode) specifies that @code{scm} will print no extra
information. This is the same as @code{-p0}.
@end deffn
@deffn {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
@code{-r} @var{macropackage} before @code{-m} on the command line.
@end deffn
@deffn {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 @code{-m} on the command line or from
Scheme code.
@end deffn
@deffn {Command Option} -i
specifies that @code{scm} should run interactively. That means that
@code{scm} will not terminate until the @code{(quit)} or @code{(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 @code{scm} is started from a tty, it will assume
that it should be interactive unless given a subsequent @code{-b}
option.
@end deffn
@deffn {Command Option} -b
specifies that @code{scm} should run non-interactively. That means that
@code{scm} will terminate after processing the command line or if there
are errors.
@end deffn
@deffn {Command Option} -s
specifies, by analogy with @code{sh}, that @code{scm} should run
interactively and that further options are to be treated as program
aguments.
@end deffn
@deffn {Command Option} -
@deffnx {Command Option} ---
specifies that further options are to be treated as program aguments.
@end deffn
@node Invocation Examples, SCM Variables, SCM Options, Operational Features
@section Invocation Examples
@table @code
@item % scm foo.scm
Loads and executes the contents of @file{foo.scm} and then enters
interactive session.
@item % scm -f foo.scm arg1 arg2 arg3
Parameters @code{arg1}, @code{arg2}, and @code{arg3} are stored in the
global list @code{*argv*}; Loads and executes the contents of
@file{foo.scm} and exits.
@item % scm -s foo.scm arg1 arg2
Sets *argv* to @code{("foo.scm" "arg1" "arg2")} and enters interactive
session.
@item % scm -e `(write (list-ref *argv* *optind*))' bar
Prints @samp{"bar"}.
@item % scm -rpretty-print -r format -i
Loads @code{pretty-print} and @code{format} and enters interactive
session.
@item % scm -r5
Loads @code{dynamic-wind}, @code{values}, and syntax-rules macros and
enters interactive (with macros) session.
@item % scm -r5 -r4
Like above but @code{rev4-optional-procedures} are also loaded.
@end table
@node SCM Variables, SCM Session, Invocation Examples, Operational Features
@section Environment Variables
@defvr {Environment Variable} SCM_INIT_PATH
is the pathname where @code{scm} will look for its initialization
code. The default is the file @file{Init@value{SCMVERSION}.scm} in the
source directory.
@end defvr
@defvr {Environment Variable} SCHEME_LIBRARY_PATH
is the [SLIB] Scheme library directory.
@end defvr
@defvr {Environment Variable} HOME
is the directory where @file{Init@value{SCMVERSION}.scm} will look for
the user initialization file @file{ScmInit.scm}.
@end defvr
@defvr {Environment Variable} EDITOR
is the name of the program which @code{ed} will call. If @var{EDITOR}
is not defined, the default is @samp{ed}.
@end defvr
@section Scheme Variables
@defvar *argv*
contains the list of arguments to the program. @code{*argv*} can change
during argument processing. This list is suitable for use as an argument
to [SLIB] @code{getopt}.
@end defvar
@defvar *syntax-rules*
controls whether loading and interaction support syntax-rules
macros. Define this in @file{ScmInit.scm} or files specified on the
command line. This can be overridden by subsequent @code{-m} and
@code{-u} options.
@end defvar
@defvar *interactive*
controls interactivity as explained for the @code{-i} and @code{-b}
options. Define this in @file{ScmInit.scm} or files specified on the
command line. This can be overridden by subsequent @code{-i} and
@code{-b} options.
@end defvar
@node SCM Session, Editing Scheme Code, SCM Variables, Operational Features
@section SCM Session
@itemize @bullet
@item
Options, file loading and features can be specified from the command
line.
@xref{System interface, , , scm, SCM}. @xref{Require, , , slib, SLIB}.
@item
Typing the end-of-file character at the top level session (while SCM is
not waiting for parenthesis closure) causes SCM to exit.
@item
Typing the interrupt character aborts evaluation of the current form
and resumes the top level read-eval-print loop.
@end itemize
@defun quit
@defunx quit n
@defunx exit
@defunx exit n
Aliases for @code{exit} (@pxref{System, exit, , slib, SLIB}). On many
systems, SCM can also tail-call another program.
@xref{I/O-Extensions, execp}.
@end defun
@deffn {Callback procedure} boot-tail dumped?
@code{boot-tail} is called by @code{scm_top_level} just before entering
interactive top-level. If @code{boot-tail} calls @code{quit}, then
interactive top-level is not entered.
@end deffn
@defun program-arguments
Returns a list of strings of the arguments scm was called with.
@end defun
@defun 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.
@end defun
@noindent
For documentation of the procedures @code{getenv} and @code{system}
@xref{System Interface, , , slib, SLIB}.
@defun vms-debug
If SCM is compiled under VMS this @code{vms-debug} will invoke the VMS
debugger.
@end defun
@node Editing Scheme Code, Debugging Scheme Code, SCM Session, Operational Features
@section Editing Scheme Code
@defun ed arg1 @dots{}
The value of the environment variable @code{EDITOR} (or just @code{ed}
if it isn't defined) is invoked as a command with arguments @var{arg1}
@dots{}.
@defunx ed filename
If SCM is compiled under VMS @code{ed} will invoke the editor with a
single the single argument @var{filename}.
@end defun
@table @asis
@item 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 @samp{M-x run-scheme} with SCM. Otherwise, use the emacs
command @samp{M-x suspend-emacs}; or see ``other systems'' below.
@item Epsilon (MS-DOS):
There is lisp (and scheme) mode available by use of the package
@samp{LISP.E}. It offers several different indentation formats. With
this package, buffers holding files ending in @samp{.L}, @samp{.LSP},
@samp{.S}, and @samp{.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 @samp{-e512 -m0} are neccessary to manage
RAM properly. It has been reported that when compiling SCM with Turbo
C, you need to @samp{#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.
@item other systems:
Define the environment variable @samp{EDITOR} to be the name of the
editing program you use. The SCM procedure @code{(ed arg1 @dots{})}
will invoke your editor and return to SCM when you exit the editor. The
following definition is convenient:
@example
(define (e) (ed "work.scm") (load "work.scm"))
@end example
Typing @samp{(e)} will invoke the editor with the file of interest.
After editing, the modified file will be loaded.
@end table
@node Debugging Scheme Code, Debugging Continuations, Editing Scheme Code, Operational Features
@section Debugging Scheme Code
@noindent
The @code{cautious} option of @code{build}
(@pxref{Build Options}) supports debugging in Scheme.
@table @dfn
@item CAUTIOUS
If SCM is built with the @samp{CAUTIOUS} flag, then when an error
occurs, a @dfn{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 @ref{Memoized Expressions} for how to read memoized
expressions.
Also as the result of the @samp{CAUTIOUS} flag, both @code{error} and
@code{user-interrupt} (invoked by @key{C-c}) to print stack traces and
conclude by calling @code{breakpoint}
(@pxref{Breakpoints, , , slib, SLIB}) instead of aborting to top
level. Under either condition, program execution can be resumed by
@code{(continue)}.
In this configuration one can interrupt a running Scheme program with
@key{C-c}, inspect or modify top-level values, trace or untrace
procedures, and continue execution with @code{(continue)}.
@end table
If @code{verbose} (@pxref{Internal State, verbose}) is called with an
argument greater than 2, then the interpreter will check stack size
periodically. If the size of stack in use exceeds the C #define
@code{STACK_LIMIT} (default is @code{HEAP_SEG_SIZE}), SCM generates a
@samp{stack} @code{segment violation}.
@noindent
There are several SLIB macros which so useful that SCM automatically
loads the appropriate module from SLIB if they are invoked.
@defmac trace proc1 @dots{}
Traces the top-level named procedures given as arguments.
@defmacx 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.
@end defmac
@defmac untrace proc1 @dots{}
Turns tracing off for its arguments.
@defmacx untrace
With no arguments, untraces all currently traced identifiers and returns
a list of these formerly traced identifiers.
@end defmac
The routines I use most frequently for debugging are:
@defun print arg1 @dots{}
@code{Print} writes all its arguments, separated by spaces.
@code{Print} outputs a @code{newline} at the end and returns the value
of the last argument.
One can just insert @samp{(print '} and @samp{)} around an
expression in order to see its values as a program operates.
@end defun
@defun pprint arg1 @dots{}
@code{Pprint} pretty-prints (@pxref{Pretty-Print, , , slib, SLIB}) all
its arguments, separated by newlines. @code{Pprint} returns the value
of the last argument.
One can just insert @samp{(pprint '} and @samp{)} around an
expression in order to see its values as a program operates.
@emph{Note@:} @code{pretty-print} does @emph{not} format procedures.
@end defun
@noindent
When typing at top level, @code{pprint} is not a good way to see
nested structure because it will return the last object
pretty-printed, which could be large. @code{pp} is a better choice.
@deffn Procedure pp arg1 @dots{}
@code{Pprint} pretty-prints (@pxref{Pretty-Print, , , slib, SLIB}) all
its arguments, separated by newlines. @code{pp} returns
@code{#}.
@end deffn
@deffn Syntax print-args name
@deffnx Syntax print-args
Writes @var{name} if supplied; then writes the names and values of the
closest lexical bindings enclosing the call to @code{Print-args}.
@example
(define (foo a b) (print-args foo) (+ a b))
(foo 3 6)
@print{} In foo: a = 3; b = 6;
@result{} 9
@end example
@end deffn
@noindent
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, @ref{Quick Print, , , slib, SLIB}, can be
used.
When @code{trace} is not sufficient to find program flow problems,
@ifset html
@end ifset
SLIB-PSD, the Portable Scheme Debugger
@ifset html
@end ifset
offers source code debugging from
GNU Emacs. PSD runs slowly, so start by instrumenting only a few
functions at a time.
@lisp
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
@end lisp
@node Debugging Continuations, Errors, Debugging Scheme Code, Operational Features
@section Debugging Continuations
These functions are defined in @file{debug.c}, all operate on captured
continuations:
@deffn Procedure frame-trace cont n
Prints information about the code being executed and the environment
scopes active for continuation frame @var{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.
@end deffn
@deffn Procedure frame->environment cont n
Prints the environment for continuation frame @var{n} of continuation
@var{cont}. This contains just the names, not the values, of the
environment.
@end deffn
@deffn Procedure scope-trace env
will print information about active lexical scopes for environment
@var{env}.
@end deffn
@deffn Procedure frame-eval cont n expr
Evaluates @var{expr} in the environment defined by continuation frame
@var{n} of continuation CONT and returns the result. Values in the
environment may be returned or SET!.
@end deffn
@ref{Errors, stack-trace} also now accepts an optional continuation
argument. @code{stack-trace} differs from @code{frame-trace} in that
it truncates long output using safeports and prints code from all
available frames.
@example
(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)
@print{}
;STACK TRACE
1; ((#@@set! #@@k (#@@call-with-current-continuation #@@identity)) #f ...
2; (#@@let ((a 3) (b 4)) (#@@foo #@@a #@@b) #f)
@dots{}
#t
@end example
@example
(frame-trace k 0)
@print{}
(#@@call-with-current-continuation #@@identity)
; in scope:
; (x y) procedure foo#
@end example
@example
(frame-trace k 1)
@print{}
((#@@set! #@@k (#@@call-with-current-continuation #@@identity)) #f)
; in scope:
; (x y) procedure foo#
@end example
@example
(frame-trace k 2)
@print{}
(#@@let ((a 3) (b 4)) (#@@foo #@@a #@@b) #f)
; in scope:
; (a b . #@@let)#
@end example
@example
(frame-trace k 3)
@print{}
(#@@let ((a 3) (b 4)) (#@@foo #@@a #@@b) #f)
; in top level environment.
@end example
@example
(frame->environment k 0)
@print{}
((x y) 2 foo)
@end example
@example
(scope-trace (frame->environment k 0))
@print{}
; in scope:
; (x y) procedure foo#
@end example
@example
(frame-eval k 0 'x) @result{} 3
(frame-eval k 0 '(set! x 8))
(frame-eval k 0 'x) @result{} 8
@end example
@node Errors, Memoized Expressions, Debugging Continuations, Operational Features
@section Errors
@noindent
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.
@noindent
The following common error and conditions are handled by C code. Those
with callback names after them can also be handled by Scheme code
(@pxref{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.
@table @dfn
@item ARGn
Wrong type in argument
@item ARG1
Wrong type in argument 1
@item ARG2
Wrong type in argument 2
@item ARG3
Wrong type in argument 3
@item ARG4
Wrong type in argument 4
@item ARG5
Wrong type in argument 5
@item WNA
Wrong number of args
@item OVFLOW
numerical overflow
@item OUTOFRANGE
Argument out of range
@item NALLOC
@code{(out-of-storage)}
@item THRASH
GC is @code{(thrashing)}
@item EXIT
@code{(end-of-program)}
@item HUP_SIGNAL
@code{(hang-up)}
@item INT_SIGNAL
@code{(user-interrupt)}
@item FPE_SIGNAL
@code{(arithmetic-error)}
@item BUS_SIGNAL
bus error
@item SEGV_SIGNAL
segment violation
@item ALRM_SIGNAL
@code{(alarm-interrupt)}
@item VTALRM_SIGNAL
@code{(virtual-alarm-interrupt)}
@item PROF_SIGNAL
@code{(profile-alarm-interrupt)}
@end table
@defvar 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 @var{errobj} is the offending
object if appropriate. The builtin procedure @code{error} does
@emph{not} set @var{errobj}.
@end defvar
@noindent
@code{errno} and @code{perror} report ANSI C errors encountered during a
call to a system or library function.
@defun errno
@defunx errno n
With no argument returns the current value of the system variable
@code{errno}. When given an argument, @code{errno} sets the system
variable @code{errno} to @var{n} and returns the previous value of
@code{errno}. @code{(errno 0)} will clear outstanding errors. This is
recommended after @code{try-load} returns @code{#f} since this occurs
when the file could not be opened.
@end defun
@defun perror string
Prints on standard error output the argument @var{string}, a colon,
followed by a space, the error message corresponding to the current
value of @code{errno} and a newline. The value returned is unspecified.
@end defun
@noindent
@code{warn} and @code{error} provide a uniform way for Scheme code to
signal warnings and errors.
@defun warn arg1 arg2 arg3 @dots{}
Alias for @ref{System, slib:warn, , slib, SLIB}. Outputs an error
message containing the arguments. @code{warn} is defined in
@file{Init@value{SCMVERSION}.scm}.
@end defun
@defun error arg1 arg2 arg3 @dots{}
Alias for @ref{System, slib:error, , slib, SLIB}. Outputs an error
message containing the arguments, aborts evaluation of the current form
and resumes the top level read-eval-print loop. @code{Error} is defined
in @file{Init@value{SCMVERSION}.scm}.
@end defun
@noindent
If SCM is built with the @samp{CAUTIOUS} flag, then when an error
occurs, a @dfn{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 @ref{Memoized Expressions} for how to read memoized
expressions.
@noindent
Also as the result of the @samp{CAUTIOUS} flag, both @code{error} and
@code{user-interrupt} (invoked by @key{C-c}) are defined to print stack
traces and conclude by calling @code{breakpoint}
(@pxref{Breakpoints, , , slib, SLIB}). This allows the user to
interract with SCM as with Lisp systems.
@defun stack-trace
Prints information describing the stack of partially evaluated
expressions. @code{stack-trace} returns @code{#t} if any lines were
printed and @code{#f} otherwise. See @file{Init@value{SCMVERSION}.scm}
for an example of the use of @code{stack-trace}.
@end defun
@node Memoized Expressions, Internal State, Errors, Operational Features
@section Memoized Expressions
@noindent
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.
@noindent
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.
@itemize @bullet
@item
The names of memoized lexically bound identifiers are replaced with
@r{#@@}@i{}@r{-}@i{}, where @i{} is the number of binding
contours back and @i{} is the index of the value in that
binding countour.
@item
The names of identifiers which are not lexiallly bound but defined at
top-level have @r{#@@} prepended.
@end itemize
@noindent
For instance, @code{open-input-file} is defined as follows in
@file{Init@value{SCMVERSION}.scm}:
@example
(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)))
@end example
@noindent
If @code{open-input-file} has not yet been used, the displayed procedure
is similar to the original definition (lines wrapped for readability):
@example
open-input-file @result{}
#
@end example
@noindent
If we open a file using @code{open-input-file}, the sections of code
used become memoized:
@example
(open-input-file "r4rstest.scm") @result{} #
open-input-file @result{}
#
@end example
@noindent
If we cause @code{open-input-file} to execute other sections of code,
they too become memoized:
@example
(open-input-file "foo.scm") @result{}
ERROR: No such file or directory
ERROR: OPEN-INPUT-FILE couldn't open file "foo.scm"
open-input-file @result{}
#
@end example
@node Internal State, Scripting, Memoized Expressions, Operational Features
@section Internal State
@defvar *interactive*
The variable @var{*interactive*} determines whether the SCM session is
interactive, or should quit after the command line is processed.
@var{*interactive*} is controlled directly by the command-line options
@samp{-b}, @samp{-i}, and @samp{-s} (@pxref{Invoking SCM}). If none of
these options are specified, the rules to determine interactivity are
more complicated; see @file{Init@value{SCMVERSION}.scm} for details.
@end defvar
@defun abort
Resumes the top level Read-Eval-Print loop.
@end defun
@defun restart
Restarts the SCM program with the same arguments as it was originally
invoked. All @samp{-l} loaded files are loaded again; If those files
have changed, those changes will be reflected in the new session.
@emph{Note@:} When running a saved executable (@pxref{Dump}),
@code{restart} is redefined to be @code{exec-self}.
@end defun
@defun 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 @emph{new} executable will be
invoked. This differentiates @code{exec-self} from @code{restart}.
@end defun
@defun verbose n
Controls how much monitoring information is printed.
If @var{n} is:
@table @asis
@item 0
no prompt or information is printed.
@item >= 1
a prompt is printed.
@item >= 2
messages bracketing file loading are printed.
@item >= 3
the CPU time is printed after each top level form evaluated;
notifications of heap growth printed; the interpreter checks stack
depth periodically.
@item >= 4
a garbage collection summary is printed after each top level form
evaluated;
@item >= 5
a message for each GC (@pxref{Garbage Collection}) is printed;
warnings issued for top-level symbols redefined.
@end table
@end defun
@defun gc
Scans all of SCM objects and reclaims for further use those that are
no longer accessible.
@end defun
@defun room
@defunx room #t
Prints out statistics about SCM's current use of storage. @code{(room #t)}
also gives the hexadecimal heap segment and stack bounds.
@end defun
@defvr Constant *scm-version*
Contains the version string (e.g. @file{@value{SCMVERSION}}) of SCM.
@end defvr
@subsection Executable path
@noindent
In order to dump a saved executable or to dynamically-link using DLD,
SCM must know where its executable file is. Sometimes SCM
(@pxref{Executable Pathname}) guesses incorrectly the location of the
currently running executable. In that case, the correct path can be set
by calling @code{execpath} with the pathname.
@defun 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.
@defunx execpath #f
@defunx execpath newpath
Sets the path to @code{#f} or @var{newpath}, respectively. The old path
is returned.
@end defun
@noindent
For other configuration constants and procedures
@xref{Configuration, , , slib, SLIB}.
@node Scripting, , Internal State, Operational Features
@section 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
@end menu
@node Unix Scheme Scripts, MS-DOS Compatible Scripts, Scripting, Scripting
@subsection Unix Scheme Scripts
@noindent
In reading this section, keep in mind that the first line of a script
file has (different) meanings to SCM and the operating system
(@code{execve}).
@deftp file #! interpreter \ @dots{}
@tindex Scheme Script
@tindex Scheme-Script
@tindex meta-argument
On unix systems, a @dfn{Shell-Script} is a file (with execute
permissions) whose first two characters are @samp{#!}. The
@var{interpreter} argument must be the pathname of the program to
process the rest of the file. The directories named by environment
variable @code{PATH} are @emph{not} searched to find @var{interpreter}.
When executing a shell-script, the operating system invokes
@var{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 @samp{#!} and the first character of
@var{interpreter} (@samp{/}). The @var{interpreter} name is followed by
@samp{ \}; SCM substitutes the second line of @var{file} for @samp{\}
(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 @var{*script*} will be
set to the script pathname. The last argument before @samp{!#} on the
second line should be @samp{-}; SCM will load the script file, preserve
the unprocessed arguments, and set @var{*argv*} to a list of the script
pathname and the unprocessed arguments.
Note that the interpreter, not the operating system, provides the
@samp{\} substitution; this will only take place if @var{interpreter} is
a SCM or SCSH interpreter.
@end deftp
@c @deffn {Read syntax} #! ignored
@c When the first two characters of the file being loaded are @code{#!},
@c the first line of that file will be ignored.
@deffn {Read syntax} #! ignored !#
When the first two characters of the file being loaded are @code{#!} and
a @samp{\} is present before a newline in the file, all characters up
to @samp{!#} will be ignored by SCM @code{read}.
@end deffn
@noindent
This combination of interpretatons allows SCM source files to be used as
POSIX shell-scripts if the first line is:
@example
#! /usr/local/bin/scm \
@end example
@noindent
The following Scheme-Script prints factorial of its argument:
@example
#! /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*))))
@end example
@example
./fact 32
@result{}
263130836933693530167218012160000000
@end example
@noindent
If the wrong number of arguments is given, @code{fact} prints its
@var{argv} with usage information.
@example
./fact 3 2
@print{}
("./fact" "3" "2")
Usage: fact N
Returns the factorial of N.
@end example
@node MS-DOS Compatible Scripts, Unix Shell Scripts, Unix Scheme Scripts, Scripting
@subsection MS-DOS Compatible Scripts
@noindent
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:
@code{#!.bat} and @code{!#.exe},
@cindex !#
@cindex !#.exe
@cindex #!
@cindex #!.bat
which are available from:
@url{http://swiss.csail.mit.edu/ftpdir/scm/sharpbang.zip}
@noindent
With these two programs installed in a @code{PATH} directory, we have
the following syntax for @var{.BAT} files.
@deftp file #! interpreter \ %0 %*
@tindex Scheme Script
@tindex Scheme-Script
The first two characters of the Scheme-Script are @samp{#!}. The
@var{interpreter} can be either a unix style program path (using
@samp{/} between filename components) or a DOS program name or path.
The rest of the first line of the Scheme-Script should be literally
@w{@samp{\ %0 %*}}, as shown.
If @var{interpreter} has @samp{/} in it, @var{interpreter} is converted
to a DOS style filename (@samp{/} @result{} @samp{\}).
In looking for an executable named @var{interpreter}, @code{#!} first
checks this (converted) filename; if @var{interpreter} doesn't exist, it
then tries to find a program named like the string starting after the
last @samp{\} (or @samp{/}) in @var{interpreter}. When searching for
executables, @code{#!} tries all directories named by environment
variable @code{PATH}.
Once the @var{interpreter} executable path is found, arguments are
processed in the manner of scheme-shell, with all the text after the
@samp{\} taken as part of the meta-argument. More precisely, @code{#!}
calls @var{interpreter} with any options on the second line of the
Scheme-Script up to @samp{!#}, the name of the Scheme-Script file, and
then any of at most 8 arguments given on the command line invoking this
Scheme-Script.
@end deftp
@noindent
The previous example Scheme-Script works in both MS-DOS and unix
systems.
@node Unix Shell Scripts, , MS-DOS Compatible Scripts, Scripting
@subsection Unix Shell Scripts
@noindent
Scheme-scripts suffer from two drawbacks:
@itemize @bullet
@item
Some Unixes limit the length of the @samp{#!} interpreter line to the
size of an object file header, which can be as small as 32 bytes.
@item
A full, explicit pathname must be specified, perhaps requiring more than
32 bytes and making scripts vulnerable to breakage when programs are
moved.
@end itemize
@noindent
The following approach solves these problems at the expense of slower
startup. Make @samp{#! /bin/sh} the first line and prepend every
subsequent line to be executed by the shell with @code{:;}. The last
line to be executed by the shell should contain an @dfn{exec} command;
@code{exec} tail-calls its argument.
@noindent
@code{/bin/sh} is thus invoked with the name of the script file, which
it executes as a *sh script. Usually the second line starts
@samp{:;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.
@noindent
The second line of the script file does not have the length restriction
mentioned above. Also, @code{/bin/sh} searches the directories listed
in the `PATH' environment variable for @samp{scm}, eliminating the need
to use absolute locations in order to invoke a program.
@noindent
The following example additionally sets @var{*script*} to the script
argument, making it compatible with the scheme code of the previous
example.
@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*))))
@end example
@example
./fact 6
@result{} 720
@end example
@node The Language, Packages, Operational Features, Top
@chapter 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
@end menu
@node Standards Compliance, Storage, The Language, The Language
@section Standards Compliance
@noindent
Scm conforms to the
@ifset html
[IEEE],
@end ifset
@cite{IEEE Standard 1178-1990. IEEE Standard for the Scheme Programming
Language.}
@ifclear html
(@pxref{Bibliography}),
@end ifclear
and
@ifset html
[R5RS],
@end ifset
@cite{Revised(5) Report on the Algorithmic Language Scheme}.
@ifset html
@end ifset
@ifinfo
@ref{Top, , , r5rs, Revised(5) Report on the Algorithmic Language Scheme}.
@end ifinfo
All the required features of these specifications are supported.
Many of the optional features are supported as well.
@subheading Optionals of [R5RS] Supported by SCM
@table @asis
@item @code{-} and @code{/} of more than 2 arguments
@itemx @code{exp}
@itemx @code{log}
@itemx @code{sin}
@itemx @code{cos}
@itemx @code{tan}
@itemx @code{asin}
@itemx @code{acos}
@itemx @code{atan}
@itemx @code{sqrt}
@itemx @code{expt}
@itemx @code{make-rectangular}
@itemx @code{make-polar}
@itemx @code{real-part}
@itemx @code{imag-part}
@itemx @code{magnitude}
@itemx @code{angle}
@itemx @code{exact->inexact}
@itemx @code{inexact->exact}
@xref{Numerical operations, , , r5rs, Revised(5) Scheme}.
@itemx @code{with-input-from-file}
@itemx @code{with-output-to-file}
@xref{Ports, , , r5rs, Revised(5) Scheme}.
@itemx @code{load}
@itemx @code{transcript-on}
@itemx @code{transcript-off}
@xref{System interface, , , r5rs, Revised(5) Scheme}.
@end table
@subheading Optionals of [R5RS] not Supported by SCM
@table @asis
@item @code{numerator}
@itemx @code{denominator}
@itemx @code{rationalize}
@xref{Numerical operations, , , r5rs, Revised(5) Scheme}.
@end table
@subheading [SLIB] Features of SCM and SCMLIT
@table @code
@item delay
@itemx full-continuation
@itemx ieee-p1178
@itemx object-hash
@itemx rev4-report
@itemx source
See SLIB file @file{Template.scm}.
@item current-time
@xref{Time and Date, , , slib, SLIB}.
@item defmacro
@xref{Defmacro, , , slib, SLIB}.
@item getenv
@itemx system
@xref{System Interface, , , slib, SLIB}.
@item hash
@xref{Hashing, , , slib, SLIB}.
@item logical
@xref{Bit-Twiddling, , , slib, SLIB}.
@item multiarg-apply
@xref{Multi-argument Apply, , , slib, SLIB}.
@item multiarg/and-
@xref{Multi-argument / and -, , , slib, SLIB}.
@item rev4-optional-procedures
@xref{Rev4 Optional Procedures, , , slib, SLIB}.
@item string-port
@xref{String Ports, , , slib, SLIB}.
@item tmpnam
@xref{Input/Output, , , slib, SLIB}.
@item transcript
@xref{Transcripts, , , slib, SLIB}.
@item vicinity
@xref{Vicinity, , , slib, SLIB}.
@item with-file
@xref{With-File, , , slib, SLIB}.
@end table
@subheading [SLIB] Features of SCM
@table @code
@item array
@xref{Arrays, , , slib, SLIB}.
@item array-for-each
@xref{Array Mapping, , , slib, SLIB}.
@item bignum
@itemx complex
@itemx inexact
@itemx rational
@itemx real
@xref{Require, , , slib, SLIB}.
@end table
@node Storage, Time, Standards Compliance, The Language
@section Storage
@defun vector-set-length! object length
Change the length of string, vector, bit-vector, or uniform-array
@var{object} to @var{length}. If this shortens @var{object} then the
remaining contents are lost. If it enlarges @var{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.
@end defun
@defun copy-tree obj
@defunx @@copy-tree obj
@xref{Tree Operations, copy-tree, , slib, SLIB}. This extends the SLIB
version by also copying vectors. Use @code{@@copy-tree} if you
depend on this feature; @code{copy-tree} could get redefined.
@end defun
@defun acons obj1 obj2 obj3
Returns (cons (cons obj1 obj2) obj3).
@lisp
(set! a-list (acons key datum a-list))
@end lisp
Adds a new association to a-list.
@end defun
@deffn {Callback procedure} gc-hook @dots{}
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, @code{(set! gc-hook #f)}.
@end deffn
@defun add-finalizer object finalizer
@var{object} may be any garbage collected object, that is, any object
other than an immediate integer, character, or special token such
as @code{#f} or @code{#t}, @xref{Immediates}. @var{finalizer} is
a thunk, or procedure taking no arguments.
@var{finalizer} will be invoked asynchronously exactly once some time
after @var{object} becomes eligible for garbage collection. A reference
to @var{object} in the environment of @var{finalizer} will not prevent
finalization, but will delay the reclamation of @var{object} at least
until the next garbage collection. A reference to @var{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.
@end defun
@node Time, Interrupts, Storage, The Language
@section Time
@defvr Constant internal-time-units-per-second
Is the integer number of internal time units in a second.
@end defvr
@defun get-internal-run-time
Returns the integer run time in internal time units from an unspecified
starting time. The difference of two calls to
@code{get-internal-run-time} divided by
@code{internal-time-units-per-second} will give elapsed run time in
seconds.
@end defun
@defun get-internal-real-time
Returns the integer time in internal time units from an unspecified
starting time. The difference of two calls to
@code{get-internal-real-time} divided by
@code{interal-time-units-per-second} will give elapsed real time in
seconds.
@end defun
@defun current-time
Returns the time since 00:00:00 GMT, January 1, 1970, measured in
seconds. @xref{Time and Date, current-time, , slib, SLIB}. @code{current-time} is
used in @ref{Time and Date, , , slib, SLIB}.
@end defun
@node Interrupts, Process Synchronization, Time, The Language
@section Interrupts
@defun 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 @var{n} is 0, any ticks request is canceled. Otherwise a
@code{ticks-interrupt} will be signaled @var{n} from the current time.
@code{ticks} is supported if SCM is compiled with the @code{ticks} flag
defined.
@end defun
@deffn {Callback procedure} ticks-interrupt @dots{}
Establishes a response for tick interrupts. Another tick interrupt will
not occur unless @code{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.
@end deffn
@defun alarm secs
Returns the number of seconds remaining till the next alarm interrupt.
If @var{secs} is 0, any alarm request is canceled. Otherwise an
@code{alarm-interrupt} will be signaled @var{secs} from the current
time. ALARM is not supported on all systems.
@end defun
@defun milli-alarm millisecs interval
@defunx virtual-alarm millisecs interval
@defunx profile-alarm millisecs interval
@code{milli-alarm} is similar to @code{alarm}, except that the first
argument @var{millisecs}, and the return value are measured in
milliseconds rather than seconds. If the optional argument
@var{interval} is supplied then alarm interrupts will be scheduled every
@var{interval} milliseconds until turned off by a call to
@code{milli-alarm} or @code{alarm}.
@code{virtual-alarm} and @code{profile-alarm} are similar.
@code{virtual-alarm} decrements process execution time rather than real
time, and causes @code{SIGVTALRM} to be signaled.
@code{profile-alarm} decrements both process execution time and
system execution time on behalf of the process, and causes
@code{SIGPROF} to be signaled.
@code{milli-alarm}, @code{virtual-alarm}, and @code{profile-alarm} are
supported only on systems providing the @code{setitimer} system call.
@end defun
@deffn {Callback procedure} user-interrupt @dots{}
@deffnx {Callback procedure} alarm-interrupt @dots{}
@deffnx {Callback procedure} virtual-alarm-interrupt @dots{}
@deffnx {Callback procedure} profile-alarm-interrupt @dots{}
Establishes a response for @code{SIGINT} (control-C interrupt) and
@code{SIGALRM}, @code{SIGVTALRM}, and @code{SIGPROF} interrupts.
Program execution will resume if the handler returns. This procedure
should @code{(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 @code{system} and
@code{ed} procedures.
To unestablish a response for an interrupt set the handler symbol to
@code{#f}. For instance, @code{(set! user-interrupt #f)}.
@end deffn
@deffn {Callback procedure} out-of-storage @dots{}
@deffnx {Callback procedure} could-not-open @dots{}
@deffnx {Callback procedure} end-of-program @dots{}
@deffnx {Callback procedure} hang-up @dots{}
@deffnx {Callback procedure} arithmetic-error @dots{}
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 @var{hang-up}
then @var{end-of-program} (if defined) will be called.
To unestablish a response for an error set the handler symbol to
@code{#f}. For instance, @code{(set! could-not-open #f)}.
@end deffn
@node Process Synchronization, Files and Ports, Interrupts, The Language
@section Process Synchronization
@noindent
An @dfn{exchanger} is a procedure of one argument regulating mutually
@cindex exchanger
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.
@defun make-exchanger obj
Returns a new exchanger with the argument @var{obj} as its initial
content.
@example
(define queue (make-exchanger (list a)))
@end example
A queue implemented as an exchanger holding a list can be protected from
reentrant execution thus:
@example
(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) @result{} a
(pop queue) @result{} #f
@end example
@end defun
@defun make-arbiter name
Returns an object of type arbiter and name @var{name}. Its state is
initially unlocked.
@end defun
@defun try-arbiter arbiter
Returns @code{#t} and locks @var{arbiter} if @var{arbiter} was unlocked.
Otherwise, returns @code{#f}.
@end defun
@defun release-arbiter arbiter
Returns @code{#t} and unlocks @var{arbiter} if @var{arbiter} was locked.
Otherwise, returns @code{#f}.
@end defun
@node Files and Ports, Eval and Load, Process Synchronization, The Language
@section Files and Ports
@noindent
These procedures generalize and extend the standard capabilities in
@ref{Ports, , ,r5rs, Revised(5) Scheme}.
@menu
* Opening and Closing::
* Port Properties::
* Port Redirection::
* Soft Ports::
@end menu
@node Opening and Closing, Port Properties, Files and Ports, Files and Ports
@subsection Opening and Closing
@defun open-file string modes
@defunx try-open-file string modes
Returns a port capable of receiving or delivering characters as
specified by the @var{modes} string. If a file cannot be opened
@code{#f} is returned.
Internal functions opening files @dfn{callback} to the SCM function
@code{open-file}. You can extend @code{open-file} by redefining it.
@code{try-open-file} is the primitive procedure; Do not redefine
@code{try-open-file}!
@end defun
@defvr Constant open_read
@defvrx Constant open_write
@defvrx 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 two-argument file-position done on the port
between a read operation and a write operation or vice-versa.
@end defvr
@defun _ionbf modestr
Returns a version of @var{modestr} which when @code{open-file} is called
with it as the second argument will return an unbuffered port. An
input-port must be unbuffered in order for @code{char-ready?} and
@code{wait-for-input} to work correctly on it. The initial value of
@code{(current-input-port)} is unbuffered if the platform supports it.
@end defun
@defun _tracked modestr
Returns a version of @var{modestr} which when @code{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 @code{port-line} and @code{port-column}.
@end defun
@defun _exclusive modestr
Returns a version of @var{modestr} which when @code{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
@code{try-create-file} @xref{I/O-Extensions}, which is not available
for all platforms.
@end defun
@defun open-ports
Returns a list of all currently open ports, excluding string ports,
see @xref{String Ports, , , slib, SLIB}. This may be useful after
a fork @xref{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.
@end defun
@defun close-port port
Closes @var{port}. The same as close-input-port and close-output-port.
@end defun
@node Port Properties, Port Redirection, Opening and Closing, Files and Ports
@subsection Port Properties
@defun port-closed? port
Returns #t if @var{port} is closed.
@end defun
@defun port-type obj
If @var{obj} is not a port returns false, otherwise returns
a symbol describing the port type, for example string or pipe.
@end defun
@defun port-filename port
Returns the filename @var{port} was opened with. If @var{port} is
not open to a file the result is unspecified.
@end defun
@defun file-position port
@defunx file-position port #f
Returns the current position of the character in @var{port} which will
next be read or written. If @var{port} is open to a non-file then
@code{#f} is returned.
@defunx file-position port k
Sets the current position in @var{port} which will next be read or
written. If successful, @code{#f} is returned. If @var{port} is open
to a non-file, then @code{file-position} returns @code{#f}.
@end defun
@defun port-line port
@defunx port-column port
If @var{port} is a tracked port, return the current line (column) number,
otherwise return @code{#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.
@end defun
@defun freshline port
Outputs a newline to optional argument @var{port} unless the current
output column number of @var{port} is known to be zero, ie output will
start at the beginning of a new line. @var{port} defaults to
@code{current-output-port}. If @var{port} is not a tracked port
@code{freshline} is equivalent to @code{newline}.
@end defun
@defun isatty? port
Returns @code{#t} if @var{port} is input or output to a serial non-file
device.
@end defun
@deffn {procedure} char-ready?
@deffnx {procedure} char-ready? port
Returns @code{#t} if a character is ready on the input @var{port} and
returns @code{#f} otherwise. If @code{char-ready?} returns @code{#t}
then
@findex char-ready
the next @code{read-char} operation on the given @var{port} is
guaranteed
@findex read-char
not to hang. If the @var{port} is at end of file then
@code{char-ready?} returns @code{#t}.
@findex char-ready?
@var{Port} may be omitted, in which case it defaults to
the value returned by @code{current-input-port}.
@findex current-input-port
@emph{Rationale@:} @code{Char-ready?} exists to make it possible for a
program to
@findex char-ready?
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 @code{char-ready?}
@findex char-ready?
cannot be rubbed out. If @code{char-ready?} were to return @code{#f} at
end of file, a port at end of file would be indistinguishable from an
interactive port that has no ready characters.
@c end rationale
@end deffn
@deffn {procedure} wait-for-input x
@deffnx {procedure} wait-for-input x port1 @dots{}
Returns a list those ports @var{port1} @dots{} which are @code{char-ready?}.
@findex char-ready?
If none of @var{port1} @dots{} become @code{char-ready?} within the time
interval of @var{x} seconds, then #f is returned. The
@var{port1} @dots{} arguments may be omitted, in which case they default
to the list of the value returned by @code{current-input-port}.
@findex current-input-port
@end deffn
@node Port Redirection, Soft Ports, Port Properties, Files and Ports
@subsection Port Redirection
@defun current-error-port
Returns the current port to which diagnostic output is directed.
@end defun
@defun with-error-to-file string thunk
@var{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 @var{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 @var{thunk}.
@end defun
@defun with-input-from-port port thunk
@defunx with-output-to-port port thunk
@defunx 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.
@end defun
@defun call-with-outputs thunk proc
Calls the @var{thunk} procedure while the current-output-port and
current-error-port are directed to string-ports. If @var{thunk}
returns, the @var{proc} procedure is called with the output-string, the
error-string, and the value returned by @var{thunk}. If @var{thunk}
does not return a value (perhaps because of error), @var{proc} is called
with just the output-string and the error-string as arguments.
@end defun
@node Soft Ports, , Port Redirection, Files and Ports
@subsection Soft Ports
@noindent
A @dfn{soft-port} is a port based on a vector of procedures capable of
accepting or delivering characters. It allows emulation of I/O ports.
@defun make-soft-port vector modes
Returns a port capable of receiving or delivering characters as
specified by the @var{modes} string (@pxref{Files and Ports, open-file}).
@var{vector} must be a vector of length 5. Its components are as
follows:
@enumerate 0
@item
procedure accepting one character for output
@item
procedure accepting a string for output
@item
thunk for flushing output
@item
thunk for getting one character
@item
thunk for closing port (not by garbage collection)
@end enumerate
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 @code{#f} if there is no useful
operation for them to perform.
If thunk 3 returns @code{#f} or an @code{eof-object}
(@pxref{Input, eof-object?, ,r5rs, Revised(5) Scheme}) 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, (@pxref{Interrupts, add-finalizer}).
@example
(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) @result{} #
@end example
@end defun
@node Eval and Load, Lexical Conventions, Files and Ports, The Language
@section Eval and Load
@defun try-load filename
If the string @var{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 @code{#t}. If not,
try-load returns @code{#f}. The try-load procedure does not affect the
values returned by @code{current-input-port} and
@code{current-output-port}.
@end defun
@defvar *load-pathname*
Is set to the pathname given as argument to @code{load},
@code{try-load}, and @code{dyn:link}
(@pxref{Compiling And Linking, , , hobbit, Hobbit}).
@code{*load-pathname*} is used to compute the value of
@ref{Vicinity, program-vicinity, , slib, SLIB}.
@end defvar
@defun eval obj
Alias for @ref{System, eval, , slib, SLIB}.
@end defun
@defun eval-string str
Returns the result of reading an expression from @var{str} and
evaluating it. @code{eval-string} does not change
@code{*load-pathname*} or @code{line-number}.
@end defun
@defun load-string str
Reads and evaluates all the expressions from @var{str}. As with
@code{load}, the value returned is unspecified. @code{load-string} does
not change @code{*load-pathname*} or @code{line-number}.
@end defun
@defun line-number
Returns the current line number of the file currently being loaded.
@end defun
@menu
* Line Numbers::
@end menu
@node Line Numbers, , Eval and Load, Eval and Load
@subsection 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.
@defun try-load pathname reader
This is the primitive for loading, @var{pathname} is the name of
a file containing Scheme code, and optional argument @var{reader} is
a function of one argument, a port. @var{reader} should read and
return Scheme code as list structure. The default value is @code{read},
which is used if @var{reader} is not supplied or is false.
@end defun
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.
@defun read-numbered port
Behaves like @code{read}, except that
@itemize bullet
@item
Load (read) sytnaxes are enabled.
@item
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 @var{port} is a tracked port
See @xref{Files and Ports}.
@end itemize
@end defun
@defun integer->line-number int
Returns a line-number object with value @var{int}. @var{int} should
be an exact non-negative integer.
@end defun
@defun line-number->integer linum
Returns the value of line-number object @var{linum} as an integer.
@end defun
@defun line-number? obj
Returns true if and only if @var{obj} is a line-number object.
@end defun
@defun read-for-load port
Behaves like @code{read}, except that load syntaxes are enabled.
@end defun
@defvar *load-reader*
@defvarx *slib-load-reader*
The value of @code{*load-reader*} should be a value acceptable as
the second argument to @code{try-load} (note that #f is acceptable).
This value will be used to read code during calls to @code{scm:load}.
The value of @code{*slib-load-reader*} will similarly be used during
calls to @code{slib:load} and @code{require}.
In order to disable all line-numbering, it is sufficient to set!
@code{*load-reader*} and @code{*slib-load-reader*} to #f.
@end defvar
@node Lexical Conventions, Syntax, Eval and Load, The Language
@section Lexical Conventions
@menu
* Common-Lisp Read Syntax::
* Load Syntax::
* Documentation and Comments::
* Modifying Read Syntax::
@end menu
@node Common-Lisp Read Syntax, Load Syntax, Lexical Conventions, Lexical Conventions
@subsection Common-Lisp Read Syntax
@deffn {Read syntax} #\token
If @var{token} is a sequence of two or more digits, then this syntax is
equivalent to @code{#.(integer->char (string->number token 8))}.
If @var{token} is @code{C-}, @code{c-}, or @code{^} followed by a
character, then this syntax is read as a control character. If
@var{token} is @code{M-} or @code{m-} followed by a character, then a
meta character is read. @code{c-} and @code{m-} prefixes may be
combined.
@end deffn
@deffn {Read syntax} #+ feature form
If feature is @code{provided?} then @var{form} is read as a scheme
expression. If not, then @var{form} is treated as whitespace.
Feature is a boolean expression composed of symbols and @code{and},
@code{or}, and @code{not} of boolean expressions.
For more information on @code{provided?},
@xref{Require, , , slib, SLIB}.
@end deffn
@deffn {Read syntax} #- feature form
is equivalent to @code{#+(not feature) expression}.
@end deffn
@deffn {Read syntax} #| any thing |#
Is a balanced comment. Everything up to the matching @code{|#} is
ignored by the @code{read}. Nested @code{#|@dots{}|#} can occur inside
@var{any thing}.
@end deffn
@noindent
@dfn{Load sytax} is Read syntax enabled for @code{read} only when that
@code{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 @samp{#.}.
@deffn {Load syntax} #. expression
Is read as the object resulting from the evaluation of @var{expression}.
This substitution occurs even inside quoted structure.
In order to allow compiled code to work with @code{#.} it is good
practice to define those symbols used inside of @var{expression} with
@code{#.(define @dots{})}. For example:
@example
#.(define foo 9) @result{} #
'(#.foo #.(+ foo foo)) @result{} (9 18)
@end example
@end deffn
@deffn {Load syntax} #' form
is equivalent to @var{form} (for compatibility with common-lisp).
@end deffn
@node Load Syntax, Documentation and Comments, Common-Lisp Read Syntax, Lexical Conventions
@subsection Load Syntax
@noindent
@dfn{#!} is the unix mechanism for executing scripts.
See @ref{Unix Scheme Scripts} for the full description of how this
comment supports scripting.
@deffn {Load syntax} #?line
@deffnx {Load syntax} #?column
Return integers for the current line and column being read during a
load.
@end deffn
@deffn {Load syntax} #?file
Returns the string naming the file currently being loaded. This path
is the string passed to @code{load}, possibly with @samp{.scm}
appended.
@end deffn
@node Documentation and Comments, Modifying Read Syntax, Load Syntax, Lexical Conventions
@subsection Documentation and Comments
@deffn {procedure} procedure-documentation proc
Returns the documentation string of @var{proc} if it exists, or
@code{#f} if not.
If the body of a @code{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
@dfn{documentation string} of that procedure.
@cindex documentation string
@example
(procedure-documentation (lambda (x) "Identity" x)) @result{} "Identity"
(define (square x)
"Return the square of X."
(* x x))
@result{} #
(procedure-documentation square) @result{} "Return the square of X."
@end example
@end deffn
@defun comment string1 @dots{}
Appends @var{string1} @dots{} to the strings given as arguments to
previous calls @code{comment}.
@defunx comment
Returns the (appended) strings given as arguments to previous calls
@code{comment} and empties the current string collection.
@end defun
@deffn {Load syntax} #;text-till-end-of-line
Behaves as @code{(comment "@var{text-till-end-of-line}")}.
@end deffn
@node Modifying Read Syntax, , Documentation and Comments, Lexical Conventions
@subsection Modifying Read Syntax
@deffn {Callback procedure} read:sharp c port
If a @key{#} followed by a character (for a non-standard syntax) is
encountered by @code{read}, @code{read} will call the value of the
symbol @code{read:sharp} with arguments the character and the port being
read from. The value returned by this function will be the value of
@code{read} for this expression unless the function returns
@code{#} in which case the expression will be treated as
whitespace. @code{#} is the value returned by the
expression @code{(if #f #f)}.
@deffnx {Callback procedure} load:sharp c port
Dispatches like @code{read:sharp}, but only during @code{load}s. The
read-syntaxes handled by @code{load:sharp} are a superset of those
handled by @code{read:sharp}. @code{load:sharp} calls
@code{read:sharp} if none of its syntaxes match @var{c}.
@end deffn
@deffn {Callback procedure} char:sharp token
If the sequence @key{#\} followed by a non-standard character name is
encountered by @code{read}, @code{read} will call the value of the
symbol @code{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 @code{read} for this expression, otherwise an error
will be signaled.
@end deffn
@emph{Note@:} When adding new @key{#} syntaxes, have your code save the
previous value of @code{load:sharp}, @code{read:sharp}, or
@code{char:sharp} when defining it. Call this saved value if an
invocation's syntax is not recognized. This will allow @code{#+},
@code{#-}, and @ref{Uniform Array}s to still be supported (as they
dispatch from @code{read:sharp}).
@node Syntax, , Lexical Conventions, The Language
@section Syntax
SCM provides a native implementation of @dfn{defmacro}.
@xref{Defmacro, , , slib, SLIB}.
When built with @samp{-F macro} build option (@pxref{Build Options}) and
@samp{*syntax-rules*} is non-false, SCM also supports [R5RS]
@code{syntax-rules} macros. @xref{Macros, , ,r5rs, Revised(5) Scheme}.
Other Scheme Syntax Extension Packages from SLIB can be employed through
the use of @samp{macro:eval} and @samp{macro:load}; Or by using the SLIB
read-eval-print-loop:
@example
(require 'repl)
(repl:top-level macro:eval)
@end example
With the appropriate catalog entries
(@pxref{Library Catalogs, , , slib, SLIB}), files using macro
packages will automatically use the correct macro loader when
@samp{require}d.
@menu
* Define and Set::
* Defmacro::
* Syntax-Rules::
* Macro Primitives::
* Environment Frames::
* Syntactic Hooks for Hygienic Macros::
@end menu
@node Define and Set, Defmacro, Syntax, Syntax
@subsection Define and Set
@defspec defined? symbol
Equivalent to @code{#t} if @var{symbol} is a syntactic keyword (such as
@code{if}) or a symbol with a value in the top level environment
(@pxref{Variables and regions, , ,r5rs, Revised(5) Scheme}). Otherwise
equivalent to @code{#f}.
@end defspec
@defspec defvar identifier initial-value
If @var{identifier} is unbound in the top level environment, then
@var{identifier} is @code{define}d to the result of evaluating the form
@var{initial-value} as if the @code{defvar} form were instead the form
@code{(define identifier initial-value)} . If @var{identifier} already
has a value, then @var{initial-value} is @emph{not} evaluated and
@var{identifier}'s value is not changed. @code{defvar} is valid only
when used at top-level.
@end defspec
@defspec defconst identifier value
If @var{identifier} is unbound in the top level environment, then
@var{identifier} is @code{define}d to the result of evaluating the form
@var{value} as if the @code{defconst} form were instead the form
@code{(define identifier value)} . If @var{identifier} already has a
value, then @var{value} is @emph{not} evaluated, @var{identifier}'s
value is not changed, and an error is signaled. @code{defconst} is
valid only when used at top-level.
@end defspec
@defspec set! (variable1 variable2 @dots{}) @r{}
The identifiers @var{variable1}, @var{variable2}, @dots{} must be bound
either in some region enclosing the @samp{set!} expression or at top
level.
@r{} is evaluated, and the elements of the resulting list
are stored in the locations to which each corresponding @var{variable}
is bound. The result of the @samp{set!} expression is unspecified.
@example
(define x 2)
(define y 3)
(+ x y) @result{} 5
(set! (x y) (list 4 5)) @result{} @emph{unspecified}
(+ x y) @result{} 9
@end example
@end defspec
@defspec qase key clause1 clause2 @dots{}
@code{qase} is an extension of standard Scheme @code{case}: Each
@var{clause} of a @code{qase} statement must have as first element a
list containing elements which are:
@itemize @bullet
@item
literal datums, or
@item
a comma followed by the name of a symbolic constant, or
@item
a comma followed by an at-sign (@@) followed by the name of a symbolic
constant whose value is a list.
@end itemize
A @code{qase} statement is equivalent to a @code{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, @code{unquote}) is
similar to that of @code{quasiquote} except that the unquoted
expressions must be @dfn{symbolic constants}.
Symbolic constants are defined using @code{defconst}, their values are
substituted in the head of each @code{qase} clause during macro
expansion. @code{defconst} constants should be defined before use.
@code{qase} can be substituted for any correct use of @code{case}.
@format
@t{(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)) ==> @emph{unspecified}
(qase (car '(c d))
((a e i o u) 'vowel)
((,@@semivowels) 'semivowel)
(else 'consonant)) ==> consonant
}
@end format
@end defspec
@node Defmacro, Syntax-Rules, Define and Set, Syntax
@subsection Defmacro
@noindent
@findex defmacro
@findex macroexpand
@findex macroexpand-1
@findex gentemp
SCM supports the following constructs from Common Lisp:
@code{defmacro}, @code{macroexpand}, @code{macroexpand-1}, and
@code{gentemp}. @xref{Defmacro, , , slib, SLIB}.
SCM @code{defmacro} is extended over that described for SLIB:
@lisp
(defmacro (macro-name . arguments) body)
@end lisp
is equivalent to
@lisp
(defmacro macro-name arguments body)
@end lisp
As in Common Lisp, an element of the formal argument list for
@code{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:
@lisp
(defmacro (let1 ((name value)) . body)
`((lambda (,name) ,@@body) ,value))
(let1 ((x (foo))) (print x) x) @equiv{} ((lambda (x) (print x) x) (foo))
(let1 not legal syntax) @error{} not "does not match" ((name value))
@end lisp
@node Syntax-Rules, Macro Primitives, Defmacro, Syntax
@subsection Syntax-Rules
@findex syntax-rules
SCM supports [R5RS] @code{syntax-rules} macros
@xref{Macros, , ,r5rs, Revised(5) Scheme}.
The pattern language is extended by the syntax @code{(... )}, which
is identical to @code{} except that ellipses in @code{} are
treated as ordinary identifiers in a template, or as literals in a
pattern. In particular, @code{(... ...)} quotes the ellipsis token
@code{...} in a pattern or template.
For example:
@lisp
(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))) @result{} #t
(check-tree ((a b) ...) '((1 2) (3 4) not-a-2list) @result{} #f
@end lisp
Note that although the ellipsis is matched as a literal token in the
defined macro it is not included in the literals list for
@code{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:
@lisp
(define-syntax eight
(syntax-rules ()
(_ 8)))
(+ 3 eight) @result{} 11
(eight) @result{} ERROR
(set! eight 9) @result{} ERROR
@end lisp
@node Macro Primitives, Environment Frames, Syntax-Rules, Syntax
@subsection Macro Primitives
@defun procedure->syntax proc
Returns a @dfn{macro} which, when a symbol defined to this value appears
as the first symbol in an expression, returns the result of applying
@var{proc} to the expression and the environment.
@end defun
@defun procedure->macro proc
@defunx procedure->memoizing-macro proc
@defunx procedure->identifier-macro
Returns a @dfn{macro} which, when a symbol defined to this value appears
as the first symbol in an expression, evaluates the result of applying
@var{proc} to the expression and the environment. The value returned
from @var{proc} which has been passed to
@code{PROCEDURE->MEMOIZING-MACRO} replaces the form passed to
@var{proc}. For example:
@example
(defsyntax trace
(procedure->macro
(lambda (x env) `(set! ,(cadr x) (tracef ,(cadr x) ',(cadr x))))))
(trace @i{foo}) @equiv{} (set! @i{foo} (tracef @i{foo} '@i{foo})).
@end example
@code{PROCEDURE->IDENTIFIER-MACRO} is similar to
@code{PROCEDURE->MEMOIZING-MACRO} except that @var{proc} is also
called in case the symbol bound to the macro appears in an expression
but @emph{not} as the first symbol, that is, when it looks like a
variable reference. In that case, the form passed to @var{proc} is
a single identifier.
@end defun
@defspec defsyntax name expr
Defines @var{name} as a macro keyword bound to the result of evaluating
@var{expr}, which should be a macro. Using @code{define} for this
purpose may not result in @var{name} being interpreted as a macro
keyword.
@end defspec
@node Environment Frames, Syntactic Hooks for Hygienic Macros, Macro Primitives, Syntax
@subsection Environment Frames
An @dfn{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:
@table @code
@item ((lambda (variable1 @dots{}) @dots{}) value1 @dots{})
@itemx (let ((variable1 value1) (variable2 value2) @dots{}) @dots{})
@itemx (letrec ((variable1 value1) @dots{}) @dots{})
result in a single enviroment frame:
@example
(variable1 variable2 @dots{})
@end example
@item (let ((variable1 value1)) @dots{})
@itemx (let* ((variable1 value1) @dots{}) @dots{})
result in an environment frame for each variable:
@example
variable1 variable2 @dots{}
@end example
@item (let-syntax ((key1 macro1) (key2 macro2)) @dots{})
@itemx (letrec-syntax ((key1 value1) (key2 value2)) @dots{})
Lexically bound macros result in environment frames consisting of
a marker and an alist of keywords and macro objects:
@example
( (key1 . value1) (key2 . value2))
@end example
Currently is the integer 6.
@item line numbers
Line numbers (@pxref{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.
@example
#
@end example
@item 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.
@example
"foo.scm" foo @dots{}
@end example
Currently is the integer 1 and
the integer 2.
@end table
@defspec @@apply procedure argument-list
Returns the result of applying @var{procedure} to @var{argument-list}.
@code{@@apply} differs from @code{apply} when the identifiers bound by
the closure being applied are @code{set!}; setting affects
@var{argument-list}.
@example
(define lst (list 'a 'b 'c))
(@@apply (lambda (v1 v2 v3) (set! v1 (cons v2 v3))) lst)
lst @result{} ((b . c) b c)
@end example
Thus a mutable environment can be treated as both a list and local
bindings.
@end defspec
@node Syntactic Hooks for Hygienic Macros, , Environment Frames, Syntax
@subsection Syntactic Hooks for Hygienic Macros
SCM provides a synthetic identifier type for efficient implementation of
hygienic macros (for example, @code{syntax-rules}
@pxref{Macros, , , r5rs, Revised(5) Scheme}) 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 @emph{identifiers}.
@defun identifier? obj
Returns @code{#t} if @var{obj} is a symbol or a synthetic
identifier, and @code{#f} otherwise.
@end defun
If it is necessary to distinguish between symbols and synthetic identifiers,
use the predicate @code{symbol?}.
A synthetic identifier includes two data: a parent, which is an
identifier, and an environment, which is either @code{#f} or a lexical
environment which has been passed to a @dfn{macro expander}
(a procedure passed as an argument to @code{procedure->macro},
@code{procedure->memoizing-macro}, or @code{procedure->syntax}).
@defun renamed-identifier parent env
Returns a synthetic identifier. @var{parent} must be an identifier, and
@var{env} must either be @code{#f} or a lexical environment passed to a
macro expander. @code{renamed-identifier} returns a distinct object for
each call, even if passed identical arguments.
@end defun
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.
@defun identifier->symbol id
Returns the symbol obtained by recursively extracting the parent of
@var{id}, which must be an identifier.
@end defun
@subsection Use of Synthetic Identifiers
@code{renamed-identifier} may be used as a replacement for @code{gentemp}:
@lisp
(define gentemp
(let ((name (string->symbol "An unlikely variable")))
(lambda ()
(renamed-identifier name #f))))
@end lisp
If an identifier returned by this version of @code{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
@code{renamed-identifier}) may denote that variable. If an identifier
returned by @code{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:
@lisp
(define top-level-foo
(procedure->memoizing-macro
(lambda (exp env)
(renamed-identifier 'foo #f))))
@end lisp
Defines a macro which may always be used to refer to the top-level binding
of @code{foo}.
@lisp
(define foo 'top-level)
(let ((foo 'local))
(top-level-foo)) @result{} top-level
@end lisp
In other words, we can avoid capturing @code{foo}.
If a lexical environment is passed as the second argument to
@code{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 @emph{must}
be restricted to the lexical scope of its environment.
There is another restriction imposed for implementation convenience:
Macros passing their lexical environments to @code{renamed-identifier}
may be lexically bound only by the special forms @code{let-syntax} or
@code{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
@code{eq?} in order to denote the same binding.
@defun identifier-equal? id1 id2 env
Returns @code{#t} if identifiers @var{id1} and @var{id2} denote the same
binding in lexical environment @var{env}, and @code{#f} otherwise.
@var{env} must either be a lexical environment passed to a macro transformer
during macro expansion or the empty list.
For example,
@lisp
(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) @result{} #t
(let ((foo 'local))
(top-level-foo? foo)) @result{} #f
@end lisp
@end defun
@defun @@macroexpand1 expr env
If the @code{car} of @var{expr} denotes a macro in @var{env}, then
if that macro is a primitive, @var{expr} will be returned, if the
macro was defined in Scheme, then a macro expansion will be returned.
If the @code{car} of @var{expr} does not denote a macro, the @code{#f}
is returned.
@end defun
@defun extended-environment names values env
Returns a new environment object, equivalent to @var{env}, which must
either be an environment object or null, extended by one frame.
@var{names} must be an identifier, or an improper list of identifiers,
usable as a formals list in a @code{lambda} expression. @var{values}
must be a list of objects long enough to provide a binding for each of
the identifiers in @var{names}. If @var{names} is an identifier or an
improper list then @var{vals} may be, respectively, any object or an
improper list of objects.
@end defun
@defspec syntax-quote obj
Synthetic identifiers are converted to their parent symbols by @code{quote}
and @code{quasiquote} so that literal data in macro definitions will be
properly transcribed. @code{syntax-quote} behaves like @code{quote}, but
preserves synthetic identifier intact.
@end defspec
@defspec the-macro mac
@code{the-macro} is the simplest of all possible macro transformers:
@var{mac} may be a syntactic keyword (macro name) or an expression
evaluating to a macro, otherwise an error is signaled. @var{mac} is
evaluated and returned once only, after which the same memoizied value is
returned.
@code{the-macro} may be used to protect local copies of macros against
redefinition, for example:
@lisp
(@@let-syntax ((let (the-macro let)))
;; code that will continue to work even if LET is redefined.
@dots{})
@end lisp
@end defspec
@defspec 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 @code{define-syntax}, @code{let-syntax}, and @code{letrec-syntax}
using @code{renaming-transformer} instead of @code{syntax-rules}.
@var{proc} should evaluate to a procedure accepting three arguments:
@var{expr}, @var{rename}, and @var{compare}. @var{expr} is a
representation of Scheme code to be expanded, as list structure.
@var{rename} is a procedure accepting an identifier and returning an
identifier renamed in the definition environment of the new syntax.
@var{compare} accepts two identifiers and returns true if and only if
both denote the same binding in the usage environment of the new syntax.
@end defspec
@node Packages, The Implementation, The Language, Top
@chapter 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
@end menu
@cindex Xlib
@cindex Xlibscm
@cindex xlib
@cindex xlibscm
@cindex x
@cindex X
@cindex graphics
@cindex hobbit
@menu
* Xlib: (Xlibscm). X Window Graphics.
* Hobbit: (hobbit). Scheme-to-C Compiler
@end menu
@node Dynamic Linking, Dump, Packages, Packages
@section Dynamic Linking
@noindent
If SCM has been compiled with @file{dynl.c} then the additional
properties of load and ([SLIB]) require specified here are supported.
The @code{require} form is preferred.
@defun require feature
If the symbol @var{feature} has not already been given as an argument to
@code{require}, then the object and library files associated with
@var{feature} will be dynamically-linked, and an unspecified value
returned. If @var{feature} is not found in @code{*catalog*}, then an
error is signaled.
@end defun
@defun usr:lib lib
Returns the pathname of the C library named @var{lib}. For example:
@code{(usr:lib "m")} returns @code{"/usr/lib/libm.a"}, the path of the C
math library.
@end defun
@defun x:lib lib
Returns the pathname of the X library named @var{lib}. For example:
@code{(x:lib "X11")} returns @code{"/usr/X11/lib/libX11.sa"}, the path
of the X11 library.
@end defun
@defun load filename lib1 @dots{}
In addition to the [R5RS] requirement of loading Scheme expressions if
@var{filename} is a Scheme source file, @code{load} will also
dynamically load/link object files (produced by @code{compile-file}, for
instance). The object-suffix need not be given to load. For example,
@example
(load (in-vicinity (implementation-vicinity) "sc2"))
or (load (in-vicinity (implementation-vicinity) "sc2.o"))
or (require 'rev2-procedures)
@ftindex rev2-procedures
or (require 'rev3-procedures)
@ftindex rev3-procedures
@end example
will load/link @file{sc2.o} if it exists.
The @var{lib1} @dots{} pathnames specify additional libraries which may
be needed for object files not produced by the Hobbit compiler. For
instance, crs is linked on GNU/Linux by
@example
(load (in-vicinity (implementation-vicinity) "crs.o")
(usr:lib "ncurses") (usr:lib "c"))
or (require 'curses)
@ftindex curses
@end example
Turtlegr graphics library is linked by:
@example
(load (in-vicinity (implementation-vicinity) "turtlegr")
(usr:lib "X11") (usr:lib "c") (usr:lib "m"))
or (require 'turtle-graphics)
@ftindex turtle-graphics
@end example
And the string regular expression
(@pxref{Regular Expression Pattern Matching}) package is linked by:
@example
(load (in-vicinity (implementation-vicinity) "rgx") (usr:lib "c"))
@end example
or
@example
(require 'regex)
@ftindex regex
@end example
@end defun
@noindent
The following functions comprise the low-level Scheme interface to
dynamic linking. See the file @file{Link.scm} in the SCM distribution
for an example of their use.
@defun dyn:link filename
@var{filename} should be a string naming an @dfn{object} or
@dfn{archive} file, the result of C-compiling. The @code{dyn:link}
procedure links and loads @var{filename} into the current SCM session.
If successfull, @code{dyn:link} returns a @dfn{link-token} suitable for
passing as the second argument to @code{dyn:call}. If not successful,
@code{#f} is returned.
@end defun
@defun dyn:call name link-token
@var{link-token} should be the value returned by a call to
@code{dyn:link}. @var{name} should be the name of C function of no
arguments defined in the file named @var{filename} which was succesfully
@code{dyn:link}ed in the current SCM session. The @code{dyn:call}
procedure calls the C function corresponding to @var{name}. If
successful, @code{dyn:call} returns @code{#t}; If not successful,
@code{#f} is returned.
@code{dyn:call} is used to call the @dfn{init_@dots{}} function after
loading SCM object files. The init_@dots{} function then makes the
identifiers defined in the file accessible as Scheme procedures.
@end defun
@defun dyn:main-call name link-token arg1 @dots{}
@var{link-token} should be the value returned by a call to
@code{dyn:link}. @var{name} should be the name of C function of 2
arguments, @code{(int argc, const char **argv)}, defined in the file named
@var{filename} which was succesfully @code{dyn:link}ed in the current
SCM session. The @code{dyn:main-call} procedure calls the C function
corresponding to @var{name} with @code{argv} style arguments, such as
are given to C @code{main} functions. If successful,
@code{dyn:main-call} returns the integer returned from the call to
@var{name}.
@code{dyn:main-call} can be used to call a @code{main} procedure from
SCM. For example, I link in and @code{dyn:main-call} a large C program,
the low level routines of which callback (@pxref{Callbacks}) into SCM
(which emulates PCI hardware).
@end defun
@defun dyn:unlink link-token
@var{link-token} should be the value returned by a call to
@code{dyn:link}. The @code{dyn:unlink} procedure removes the previously
loaded file from the current SCM session. If successful,
@code{dyn:unlink} returns @code{#t}; If not successful, @code{#f} is
returned.
@end defun
@node Dump, Numeric, Dynamic Linking, Packages
@section Dump
@ftindex dump
@ftindex unexec
@dfn{Dump}, (also known as @dfn{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.
@noindent
There are constraints on which sessions are savable using @code{dump}
@itemize @bullet
@item
Saved continuations are invalid in subsequent invocations; they cause
segmentation faults and other unpleasant side effects.
@item
Although DLD (@pxref{Dynamic Linking}) can be used to load compiled
modules both before and after dumping, @samp{SUN_DL} ELF systems can
load compiled modules only after dumping. This can be worked around by
compiling in those features you wish to @code{dump}.
@item
Ports (other than @code{current-input-port}, @code{current-output-port},
@code{current-error-port}), X windows, etc. are invalid in subsequent
invocations.
This restriction could be removed; @xref{Improvements To Make}.
@item
@code{Dump} should only be called from a loading file when the call to
dump is the last expression in that file.
@item
@code{Dump} can be called from the command line.
@end itemize
@defun dump newpath
@defunx dump newpath #f
@defunx dump newpath #t
@defunx dump newpath thunk
@itemize @bullet
@item
Calls @code{gc}.
@item
@findex boot-tail
Creates an executable program named @var{newpath} which continues the
state of the current SCM session when invoked. The optional argument
@var{thunk}, if provided, should be a procedure of no arguments;
@var{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 @code{dump} is @code{#t}, argument processing
will continue from the command line passed to the dumping session. If
the second argument is missing or @code{#f} then the command line
arguments of the restoring invocation will be processed.
@item
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.
@end itemize
@code{dump} may set the values of @code{boot-tail}, @code{*argv*},
@code{restart}, and @var{*interactive*}. @code{dump} returns an
unspecified value.
@end defun
When a dumped executable is invoked, the variable @var{*interactive*}
(@pxref{Internal State}) has the value it possessed when @code{dump}
created it. Calling @code{dump} with a single argument sets
@var{*interactive*} to @code{#f}, which is the state it has at the
beginning of command line processing.
The procedure @code{program-arguments} returns the command line
arguments for the curent invocation. More specifically,
@code{program-arguments} for the restored session are @emph{not} saved
from the dumping session. Command line processing is done on
the value of the identifier @code{*argv*}.
The following example shows how to create @samp{rscm}, which is like
regular scm, but which loads faster and has the @samp{random} package
alreadly provided.
@example
bash$ scm -rrandom
> (dump "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$
@end example
This task can also be accomplished using the @samp{-o} command line
option (@pxref{SCM Options}).
@example
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$
@end example
@node Numeric, Arrays, Dump, Packages
@section Numeric
@defvr Constant most-positive-fixnum
The immediate integer closest to positive infinity.
@xref{Configuration, , , slib, SLIB}.
@end defvr
@defvr Constant most-negative-fixnum
The immediate integer closest to negative infinity.
@end defvr
@defvr Constant $pi
@defvrx Constant pi
The ratio of the circumference to the diameter of a circle.
@end defvr
@noindent
These procedures augment the standard capabilities in
@ref{Numerical operations, , ,r5rs, Revised(5) Scheme}.
@defun pi* z
@code{(* pi @var{z})}
@end defun
@defun pi/ z
@code{(/ pi @var{z})}
@end defun
@defun sinh z
@defunx cosh z
@defunx tanh z
Return the hyperbolic sine, cosine, and tangent of @var{z}
@end defun
@defun asinh z
@defunx acosh z
@defunx atanh z
Return the inverse hyperbolic sine, cosine, and tangent of @var{z}
@end defun
@defun real-sqrt x
@defunx real-exp x
@defunx real-ln x
@defunx real-sin x
@defunx real-cos x
@defunx real-tan x
@defunx real-asin x
@defunx real-acos x
@defunx real-atan x
@defunx real-sinh x
@defunx real-cosh x
@defunx real-tanh x
@defunx real-asinh x
@defunx real-acosh x
@defunx real-atanh x
Real-only versions of these popular functions. The argument @var{x}
must be a real number. It is an error if the value which should be
returned by a call to these procedures is @emph{not} real.
@end defun
@defun real-log10 x
Real-only base 10 logarithm.
@end defun
@defun $atan2 y x
Computes @code{(angle (make-rectangular x y))} for real numbers @var{y}
and @var{x}.
@end defun
@defun real-expt x1 x2
Returns real number @var{x1} raised to the real power @var{x2}. It is
an error if the value which should be returned by a call to @code{real-expt}
is not real.
@end defun
@node Arrays, Records, Numeric, Packages
@section Arrays
@menu
* Conventional Arrays::
* Uniform Array::
* Bit Vectors::
* Array Mapping:: array-for-each
@end menu
@node Conventional Arrays, Uniform Array, Arrays, Arrays
@subsection Conventional Arrays
The following syntax and procedures are SCM extensions to feature
@code{array} in @ref{Arrays, , , slib, SLIB}.
@dfn{Arrays} read and write as a @code{#} followed by the @dfn{rank}
@cindex array
(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.
@example
(make-array '#(ho) 4 3) @result{}
#2A((ho ho ho) (ho ho ho) (ho ho ho) (ho ho ho))
@end example
Unshared, conventional (not uniform) 0-based arrays of rank 1 are
equivalent to (and can't be distinguished from) scheme vectors.
@example
(make-array '#(ho) 3) @result{} #(ho ho ho)
@end example
@defun transpose-array array dim0 dim1 @dots{}
Returns an array sharing contents with @var{array}, but with dimensions
arranged in a different order. There must be one @var{dim} argument for
each dimension of @var{array}. @var{dim0}, @var{dim1}, @dots{} 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 @var{dim0}, @var{dim1}, @dots{} correspond to dimensions
in the array to be returned, their positions in the argument list to
dimensions of @var{array}. Several @var{dim}s may have the same value,
in which case the returned array will have smaller rank than
@var{array}.
examples:
@example
(transpose-array '#2A((a b) (c d)) 1 0) @result{} #2A((a c) (b d))
(transpose-array '#2A((a b) (c d)) 0 0) @result{} #1A(a d)
(transpose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1 1 0) @result{}
#2A((a 4) (b 5) (c 6))
@end example
@end defun
@defun enclose-array array dim0 dim1 @dots{}
@var{dim0}, @var{dim1} @dots{} should be nonnegative integers less than
the rank of @var{array}. @var{enclose-array} returns an array
resembling an array of shared arrays. The dimensions of each shared
array are the same as the @var{dim}th 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 @var{dim}.
An enclosed array is not a general Scheme array. Its elements may not
be set using @code{array-set!}. Two references to the same element of
an enclosed array will be @code{equal?} but will not in general be
@code{eq?}. The value returned by @var{array-prototype} when given an
enclosed array is unspecified.
examples:
@example
(enclose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1) @result{}
#
(enclose-array '#3A(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1 0) @result{}
#
@end example
@end defun
@defun array->list array
Returns a list consisting of all the elements, in order, of @var{array}.
In the case of a rank-0 array, returns the single element.
@end defun
@defun array-contents array
@defunx array-contents array strict
If @var{array} may be @dfn{unrolled} into a one dimensional shared
array without changing their order (last subscript changing fastest),
then @code{array-contents} returns that shared array, otherwise it
returns @code{#f}. All arrays made by @var{make-array} may be
unrolled, some arrays made by @var{make-shared-array} may not be.
If the optional argument @var{strict} is provided, a shared array will
be returned only if its elements are stored internally contiguous in
memory.
@end defun
@node Uniform Array, Bit Vectors, Conventional Arrays, Arrays
@subsection Uniform Array
@noindent
@dfn{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 @code{make-array}, detailed in @xref{Arrays, , , slib, SLIB}.
@noindent
Unshared uniform character 0-based arrays of rank 1 (dimension)
are equivalent to (and can't be distinguished from) strings.
@example
(make-array "" 3) @result{} "$q2"
@end example
@noindent
Unshared uniform boolean 0-based arrays of rank 1 (dimension) are
equivalent to (and can't be distinguished from)
@ref{Bit Vectors, bit-vectors}.
@example
(make-array '#1at() 3) @result{} #*000
@equiv{}
#1At(#f #f #f) @result{} #*000
@end example
@noindent
@var{prototype} arguments in the following procedures are interpreted
according to the table:
@example
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
@end example
@noindent
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, @code{'#1A:fixZ32b(3 5 9)}
returns a uniform vector of signed integers.
@defun array? obj prototype
Returns @code{#t} if the @var{obj} is an array of type corresponding to
@var{prototype}, and @code{#f} if not.
@end defun
@defun array-prototype array
Returns an object that would produce an array of the same type as
@var{array}, if used as the @var{prototype} for
@code{list->uniform-array}.
@end defun
@defun list->uniform-array rank prot lst
Returns a uniform array of the type indicated by prototype @var{prot}
with elements the same as those of @var{lst}. Elements must be of the
appropriate type, no coercions are done.
In, for example, the case of a rank-2 array, @var{lst} must be a list of
lists, all of the same length. The length of @var{lst} will be the
first dimension of the result array, and the length of each element the
second dimension.
If @var{rank} is zero, @var{lst}, which need not be a list, is the
single element of the returned array.
@end defun
@defun uniform-array-read! ura
@defunx uniform-array-read! ura port
Attempts to read all elements of @var{ura}, in lexicographic order, as
binary objects from @var{port}. If an end of file is encountered during
uniform-array-read! the objects up to that point only are put into @var{ura}
(starting at the beginning) and the remainder of the array is
unchanged.
@code{uniform-array-read!} returns the number of objects read.
@var{port} may be omitted, in which case it defaults to the value
returned by @code{(current-input-port)}.
@end defun
@defun uniform-array-write ura
@defunx uniform-array-write ura port
Writes all elements of @var{ura} as binary objects to @var{port}. The
number of of objects actually written is returned. @var{port} may be
omitted, in which case it defaults to the value returned by
@code{(current-output-port)}.
@end defun
@defun logaref array index1 index2 @dots{}
If an @var{index} is provided for each dimension of @var{array} returns
the @var{index1}, @var{index2}, @dots{}'th element of @var{array}. If
one more @var{index} is provided, then the last index specifies bit
position of the twos-complement representation of the array element
indexed by the other @var{index}s returning @code{#t} if the bit is 1,
and @code{#f} if 0. It is an error if this element is not an exact
integer.
@example
(logaref '#(#b1101 #b0010) 0) @result{} #b1101
(logaref '#(#b1101 #b0010) 0 1) @result{} #f
(logaref '#2((#b1101 #b0010)) 0 0) @result{} #b1101
@end example
@end defun
@defun logaset! array val index1 index2 @dots{}
If an @var{index} is provided for each dimension of @var{array} sets the
@var{index1}, @var{index2}, @dots{}'th element of @var{array} to
@var{val}. If one more @var{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 @var{val} is @code{#t}
and to 0 if @var{val} is @code{#f}. In this case it is an error
if the array element is not an exact integer or if @var{val} is not
boolean.
@end defun
@node Bit Vectors, Array Mapping, Uniform Array, Arrays
@subsection Bit Vectors
@noindent
Bit vectors can be written and read as a sequence of @code{0}s and
@code{1}s prefixed by @code{#*}.
@example
#1At(#f #f #f #t #f #t #f) @result{} #*0001010
@end example
@noindent
Some of these operations will eventually be generalized to other
uniform-arrays.
@defun bit-count bool bv
Returns the number of occurrences of @var{bool} in @var{bv}.
@end defun
@defun bit-position bool bv k
Returns the minimum index of an occurrence of @var{bool} in @var{bv}
which is at least @var{k}. If no @var{bool} occurs within the specified
range @code{#f} is returned.
@end defun
@defun bit-invert! bv
Modifies @var{bv} by replacing each element with its negation.
@end defun
@defun bit-set*! bv uve bool
If uve is a bit-vector @var{bv} and uve must be of the same length. If
@var{bool} is @code{#t}, uve is OR'ed into @var{bv}; If @var{bool} is @code{#f}, the
inversion of uve is AND'ed into @var{bv}.
If uve is a unsigned integer vector all the elements of uve must be
between 0 and the @code{LENGTH} of @var{bv}. The bits of @var{bv}
corresponding to the indexes in uve are set to @var{bool}.
The return value is unspecified.
@end defun
@defun bit-count* bv uve bool
Returns
@example
(bit-count (bit-set*! (if bool bv (bit-invert! bv)) uve #t) #t).
@end example
@var{bv} is not modified.
@end defun
@node Array Mapping, , Bit Vectors, Arrays
@subsection Array Mapping
@code{(require 'array-for-each)}
@ftindex array-for-each
SCM has some extra functions in feature @code{array-for-each}:
@defun array-fill! array fill
Stores @var{fill} in every element of @var{array}. The value returned
is unspecified.
@end defun
@defun serial-array:copy! destination source
Same as @code{array:copy!} but guaranteed to copy in row-major order.
@end defun
@defun array-equal? array0 array1 @dots{}
Returns @code{#t} iff all arguments are arrays with the same shape,
the same type, and have corresponding elements which are either
@code{equal?} or @code{array-equal?}. This function differs from
@code{equal?} in that a one dimensional shared array may be
@var{array-equal?} but not @var{equal?} to a vector or uniform vector.
@end defun
@defun array-map! array0 proc array1 @dots{}
If @var{array1}, @dots{} are arrays, they must have the same number of
dimensions as @var{array0} and have a range for each index which
includes the range for the corresponding index in @var{array0}. If
they are scalars, that is, not arrays, vectors, or strings, then they
will be converted internally to arrays of the appropriate shape.
@var{proc} is applied to each tuple of elements of @var{array1}
@dots{} and the result is stored as the corresponding element in
@var{array0}. The value returned is unspecified. The order of
application is unspecified.
Handling non-array arguments is a SCM extension of
@ref{Array Mapping, array-map!, , slib, SLIB}
@end defun
@defun serial-array-map! array0 proc array1 @dots{}
Same as @var{array-map!}, but guaranteed to apply @var{proc} in
row-major order.
@end defun
@defun array-map prototype proc array1 array2 @dots{}
@var{array2}, @dots{} must have the same number of dimensions as
@var{array1} and have a range for each index which includes the
range for the corresponding index in @var{array1}. @var{proc} is
applied to each tuple of elements of @var{array1}, @var{array2},
@dots{} and the result is stored as the corresponding element in a
new array of type @var{prototype}. The new array is returned. The
order of application is unspecified.
@end defun
@defun scalar->array scalar array prototype
@defunx scalar->array scalar array
Returns a uniform array of the same shape as @var{array}, having only
one shared element, which is @code{eqv?} to @var{scalar}.
If the optional argument @var{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 @code{array} if that is possible, and
a conventional array if it is not. This function is used internally
by @code{array-map!} and friends to handle scalar arguments.
@end defun
@node Records, I/O-Extensions, Arrays, Packages
@section Records
SCM provides user-definable datatypes with the same interface as SLIB,
see @xref{Records, , , slib, SLIB}, with the following extension.
@defun record-printer-set! rtd printer
Causes records of type @var{rtd} to be printed in a user-specified format.
@var{rtd} must be a record type descriptor returned by @code{make-record-type},
@var{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 @code{write} and false if for @code{display}.
If @var{printer} returns #f, the default record printer will be called.
A @var{printer} value of #f means use the default printer.
Only the default printer will be used when printing error messages.
@end defun
@node I/O-Extensions, Posix Extensions, Records, Packages
@section I/O-Extensions
@noindent
If @code{'i/o-extensions} is provided (by linking in @file{ioext.o}),
@ref{Line I/O, , , slib, SLIB}, and the following functions are defined:
@defun stat
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 @code{#f} is returned.
The elements of the returned vector are as follows:
@table @r
@item 0 st_dev
ID of device containing a directory entry for this file
@item 1 st_ino
Inode number
@item 2 st_mode
File type, attributes, and access control summary
@item 3 st_nlink
Number of links
@item 4 st_uid
User ID of file owner
@item 5 st_gid
Group ID of file group
@item 6 st_rdev
Device ID; this entry defined only for char or blk spec files
@item 7 st_size
File size (bytes)
@item 8 st_atime
Time of last access
@item 9 st_mtime
Last modification time
@item 10 st_ctime
Last file status change time
@end table
@end defun
@defun getpid
Returns the process ID of the current process.
@end defun
@defun try-create-file name modes perms
If the file with name @var{name} already exists, return @code{#f},
otherwise try to create and open the file like @code{try-open-file},
@xref{Files and Ports}. If the optional integer argument @var{perms} is
provided, it is used as the permissions of the new file (modified by
the current umask).
@end defun
@defun reopen-file filename modes port
Closes port @var{port} and reopens it with @var{filename} and
@var{modes}. @code{reopen-file} returns @code{#t} if successful,
@code{#f} if not.
@end defun
@defun duplicate-port port modes
Creates and returns a @dfn{duplicate} port from @var{port}. Duplicate
@emph{unbuffered} ports share one file position. @var{modes} are as for
@ref{Files and Ports, open-file}.
@end defun
@defun redirect-port! from-port to-port
Closes @var{to-port} and makes @var{to-port} be a duplicate of
@var{from-port}. @code{redirect-port!} returns @var{to-port} if
successful, @code{#f} if not. If unsuccessful, @var{to-port} is not
closed.
@end defun
@defun opendir dirname
Returns a @dfn{directory} object corresponding to the file system
directory named @var{dirname}. If unsuccessful, returns @code{#f}.
@end defun
@defun readdir dir
Returns the string name of the next entry from the directory @var{dir}.
If there are no more entries in the directory, @code{readdir} returns a
@code{#f}.
@end defun
@defun rewinddir dir
Reinitializes @var{dir} so that the next call to @code{readdir} with
@var{dir} will return the first entry in the directory again.
@end defun
@defun closedir dir
Closes @var{dir} and returns @code{#t}. If @var{dir} is already
closed,, @code{closedir} returns a @code{#f}.
@end defun
@defun directory-for-each proc directory
@var{proc} must be a procedure taking one argument.
@samp{Directory-For-Each} applies @var{proc} to the (string) name of
each file in @var{directory}. The dynamic order in which @var{proc} is
applied to the filenames is unspecified. The value returned by
@samp{directory-for-each} is unspecified.
@defunx directory-for-each proc directory pred
Applies @var{proc} only to those filenames for which the procedure
@var{pred} returns a non-false value.
@defunx directory-for-each proc directory match
Applies @var{proc} only to those filenames for which
@code{(filename:match?? @var{match})} would return a non-false value
(@pxref{Filenames, , , slib, SLIB}).
@example
(require 'directory)
(directory-for-each print "." "[A-Z]*.scm")
@print{}
"Init.scm"
"Iedline.scm"
"Link.scm"
"Macro.scm"
"Transcen.scm"
"Init@value{SCMVERSION}.scm"
@end example
@end defun
@defun mkdir path mode
The @code{mkdir} function creates a new, empty directory whose name is
@var{path}. The integer argument @var{mode} specifies the file
permissions for the new directory.
@xref{The Mode Bits for Access Permission, , , libc, Gnu C Library},
for more information about this.
@code{mkdir} returns if successful, @code{#f} if not.
@end defun
@defun rmdir path
The @code{rmdir} function deletes the directory @var{path}. The
directory must be empty before it can be removed. @code{rmdir} returns
if successful, @code{#f} if not.
@end defun
@defun chdir filename
Changes the current directory to @var{filename}. If @var{filename} does not
exist or is not a directory, @code{#f} is returned. Otherwise, @code{#t} is
returned.
@end defun
@defun getcwd
The function @code{getcwd} returns a string containing the absolute file
name representing the current working directory. If this string cannot
be obtained, @code{#f} is returned.
@end defun
@defun rename-file oldfilename newfilename
Renames the file specified by @var{oldfilename} to @var{newfilename}.
If the renaming is successful, @code{#t} is returned. Otherwise,
@code{#f} is returned.
@end defun
@defun chmod file mode
The function @code{chmod} sets the access permission bits for the file
named by @var{file} to @var{mode}. The @var{file} argument may be a
string containing the filename or a port open to the file.
@code{chmod} returns if successful, @code{#f} if not.
@end defun
@defun utime pathname acctime modtime
Sets the file times associated with the file named @var{pathname} to
have access time @var{acctime} and modification time @var{modtime}.
@code{utime} returns if successful, @code{#f} if not.
@end defun
@defun umask mode
The function @code{umask} sets the file creation mask of the current
process to @var{mask}, and returns the previous value of the file
creation mask.
@end defun
@defun fileno port
Returns the integer file descriptor associated with the port @var{port}.
If an error is detected, @code{#f} is returned.
@end defun
@defun access pathname how
Returns @code{#t} if the file named by @var{pathname} can be accessed in
the way specified by the @var{how} argument. The @var{how} argument can
be the @code{logior} of the flags:
@enumerate 0
@item
File-exists?
@item
File-is-executable?
@item
File-is-writable?
@end enumerate
@enumerate 4
@item
File-is-readable?
@end enumerate
Or the @var{how} argument can be a string of 0 to 3 of the following
characters in any order. The test performed is the @code{and} of the
associated tests and @code{file-exists?}.
@table @key
@item x
File-is-executable?
@item w
File-is-writable?
@item r
File-is-readable?
@end table
@end defun
@defun execl command arg0 @dots{}
@defunx execlp command arg0 @dots{}
Transfers control to program @var{command} called with arguments
@var{arg0} @dots{}. For @code{execl}, @var{command} must be an exact
pathname of an executable file. @code{execlp} searches for
@var{command} in the list of directories specified by the environment
variable @var{PATH}. The convention is that @var{arg0} is the same name
as @var{command}.
If successful, this procedure does not return. Otherwise an error
message is printed and the integer @code{errno} is returned.
@defunx execv command arglist
@defunx execvp command arglist
Like @code{execl} and @code{execlp} except that the set of arguments to
@var{command} is @var{arglist}.
@end defun
@defun putenv string
adds or removes definitions from the @dfn{environment}. If the
@var{string} is of the form @samp{NAME=VALUE}, the definition is added
to the environment. Otherwise, the @var{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 @code{=}. System-defined environment variables are
invariably uppercase.
@code{Putenv} is used to set up the environment before calls to
@code{execl}, @code{execlp}, @code{execv}, @code{execvp}, @code{system},
or @code{open-pipe} (@pxref{Posix Extensions, open-pipe}).
To access environment variables, use @code{getenv}
(@pxref{System Interface, getenv, , slib, SLIB}).
@end defun
@node Posix Extensions, Unix Extensions, I/O-Extensions, Packages
@section Posix Extensions
@cindex Posix
@cindex posix
@noindent
If @code{'posix} is provided (by linking in @file{posix.o}), the
following functions are defined:
@defun open-pipe string modes
If the string @var{modes} contains an @key{r}, returns an input port
capable of delivering characters from the standard output of the system
command @var{string}. Otherwise, returns an output port capable of
receiving characters which become the standard input of the system
command @var{string}. If a pipe cannot be created @code{#f} is
returned.
@end defun
@defun open-input-pipe string
Returns an input port capable of delivering characters from the
standard output of the system command @var{string}. If a pipe cannot be
created @code{#f} is returned.
@end defun
@defun open-output-pipe string
Returns an output port capable of receiving characters which become
the standard input of the system command @var{string}. If a pipe cannot
be created @code{#f} is returned.
@end defun
@defun 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 @var{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.
@end defun
@defun close-port pipe
Closes the @var{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.
@end defun
@defun pipe
Returns @code{(cons @var{rd} @var{wd})} where @var{rd} and @var{wd} are
the read and write (port) ends of a @dfn{pipe} respectively.
@end defun
@defun fork
Creates a copy of the process calling @code{fork}. Both processes
return from @code{fork}, but the calling (@dfn{parent}) process's
@code{fork} returns the @dfn{child} process's ID whereas the child
process's @code{fork} returns 0.
@end defun
@noindent
For a discussion of @dfn{ID}s
@xref{Process Persona, , , GNU C Library, libc}.
@defun getppid
Returns the process ID of the parent of the current process.
For a process's own ID @xref{I/O-Extensions, getpid}.
@end defun
@defun getuid
Returns the real user ID of this process.
@end defun
@defun getgid
Returns the real group ID of this process.
@end defun
@defun getegid
Returns the effective group ID of this process.
@end defun
@defun geteuid
Returns the effective user ID of this process.
@end defun
@defun setuid id
Sets the real user ID of this process to @var{id}.
Returns @code{#t} if successful, @code{#f} if not.
@end defun
@defun setgid id
Sets the real group ID of this process to @var{id}.
Returns @code{#t} if successful, @code{#f} if not.
@end defun
@defun setegid id
Sets the effective group ID of this process to @var{id}.
Returns @code{#t} if successful, @code{#f} if not.
@end defun
@defun seteuid id
Sets the effective user ID of this process to @var{id}.
Returns @code{#t} if successful, @code{#f} if not.
@end defun
@defun kill pid sig
The @code{kill} function sends the signal @var{signum} to the process or
process group specified by @var{pid}. Besides the signals listed in
@ref{Standard Signals, , ,libc , GNU C Library}, @var{signum} can also
have a value of zero to check the validity of the @var{pid}.
The @var{pid} specifies the process or process group to receive the
signal:
@table @asis
@item > 0
The process whose identifier is @var{pid}.
@item 0
All processes in the same process group as the sender. The
sender itself does not receive the signal.
@item -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.
@item < -1
The process group whose identifier is @code{(abs @var{pid})}.
@end table
A process can send a signal to itself with @code{(kill (getpid)
@var{signum})}. If @code{kill} is used by a process to send a signal to
itself, and the signal is not blocked, then @code{kill} delivers at
least one signal (which might be some other pending unblocked signal
instead of the signal @var{signum}) to that process before it returns.
The return value from @code{kill} is zero if the signal can be sent
successfully. Otherwise, no signal is sent, and a value of @code{-1} is
returned. If @var{pid} specifies sending a signal to several processes,
@code{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.
@end defun
@defun waitpid pid options
The @code{waitpid} function suspends execution of the current process
until a child as specified by the @var{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 @var{pid} has already exited by the time of the call (a so-called
@dfn{zombie} process), the function returns immediately. Any system
resources used by the child are freed.
The value of @var{pid} can be:
@table @asis
@item < -1
which means to wait for any child process whose process group ID is
equal to the absolute value of @var{pid}.
@item -1
which means to wait for any child process; this is the same behaviour
which wait exhibits.
@item 0
which means to wait for any child process whose process group ID is
equal to that of the calling process.
@item > 0
which means to wait for the child whose process ID is equal to the value
of @var{pid}.
@end table
The value of @var{options} is one of the following:
@enumerate 0
@item
Nothing special.
@item
(@code{WNOHANG}) which means to return immediately if no child is there
to be waited for.
@item
(@code{WUNTRACED}) which means to also return for children which are
stopped, and whose status has not been reported.
@item
Which means both of the above.
@end enumerate
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
@code{WNOHANG} option was specified and no child process is waiting to
be noticed, the value is zero. A value of @code{#f} is returned in case
of error and @code{errno} is set. For information about the
@code{errno} codes @xref{Process Completion, , , GNU C Library, libc}.
@end defun
@defun uname
You can use the @code{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:
@enumerate 0
@item
The name of the operating system in use.
@item
The network name of this particular computer.
@item
The current release level of the operating system implementation.
@item
The current version level within the release of the operating system.
@item
Description of the type of hardware that is in use.
Some examples are @samp{"i386-ANYTHING"}, @samp{"m68k-hp"},
@samp{"sparc-sun"}, @samp{"m68k-sun"}, @samp{"m68k-sony"} and @samp{"mips-dec"}.
@end enumerate
@end defun
@defun getpw name
@defunx getpw uid
@defunx getpw
Returns a vector of information for the entry for @code{NAME},
@code{UID}, or the next entry if no argument is given. The
information is:
@enumerate 0
@item
The user's login name.
@item
The encrypted password string.
@item
The user ID number.
@item
The user's default group ID number.
@item
A string typically containing the user's real name, and
possibly other information such as a phone number.
@item
The user's home directory, initial working directory, or @code{#f}, in
which case the interpretation is system-dependent.
@item
The user's default shell, the initial program run when the user logs in,
or @code{#f}, indicating that the system default should be used.
@end enumerate
@end defun
@defun setpwent #t
Rewinds the pw entry table back to the begining.
@defunx setpwent #f
@defunx setpwent
Closes the pw table.
@end defun
@defun getgr name
@defunx getgr uid
@defunx getgr
Returns a vector of information for the entry for @code{NAME},
@code{UID}, or the next entry if no argument is given. The
information is:
@enumerate 0
@item
The name of the group.
@item
The encrypted password string.
@item
The group ID number.
@item
A list of (string) names of users in the group.
@end enumerate
@end defun
@defun setgrent #t
Rewinds the group entry table back to the begining.
@defunx setgrent #f
@defunx setgrent
Closes the group table.
@end defun
@defun getgroups
Returns a vector of all the supplementary group IDs of the process.
@end defun
@defun link oldname newname
The @code{link} function makes a new link to the existing file named by
@var{oldname}, under the new name @var{newname}.
@code{link} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun chown filename owner group
The @code{chown} function changes the owner of the file @var{filename}
to @var{owner}, and its group owner to @var{group}.
@code{chown} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun ttyname port
If port @var{port} is associated with a terminal device, returns a
string containing the file name of termainal device; otherwise
@code{#f}.
@end defun
@node Unix Extensions, Sequence Comparison, Posix Extensions, Packages
@section Unix Extensions
@cindex Unix
@cindex unix
@noindent
If @code{'unix} is provided (by linking in @file{unix.o}), the following
functions are defined:
@noindent
These @dfn{privileged} and symbolic link functions are not in Posix:
@defun symlink oldname newname
The @code{symlink} function makes a symbolic link to @var{oldname} named
@var{newname}.
@code{symlink} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun readlink filename
Returns the value of the symbolic link @var{filename} or @code{#f} for
failure.
@end defun
@defun lstat filename
The @code{lstat} function is like @code{stat}, except that it does not
follow symbolic links. If @var{filename} is the name of a symbolic
link, @code{lstat} returns information about the link itself; otherwise,
@code{lstat} works like @code{stat}. @xref{I/O-Extensions}.
@end defun
@defun nice increment
Increment the priority of the current process by @var{increment}.
@code{chown} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun 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
@var{filename} as it terminates. An argument of @code{#f} causes
accounting to be turned off.
@code{acct} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun mknod filename mode dev
The @code{mknod} function makes a special file with name @var{filename}
and modes @var{mode} for device number @var{dev}.
@code{mknod} returns a value of @code{#t} if it is successful and
@code{#f} on failure.
@end defun
@defun sync
@code{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.
@end defun
@node Sequence Comparison, Regular Expression Pattern Matching, Unix Extensions, Packages
@section Sequence Comparison
@code{(require 'diff)}
@ftindex regex
A blazing fast implementation of the sequence-comparison module in
SLIB, see @xref{Sequence Comparison, , , slib, SLIB}.
@node Regular Expression Pattern Matching, Line Editing, Sequence Comparison, Packages
@section Regular Expression Pattern Matching
These functions are defined in @file{rgx.c} using a POSIX or GNU
@dfn{regex} library. If your computer does not support regex, a package
is available via ftp from
@file{ftp.gnu.org:/pub/gnu/regex-0.12.tar.gz}. For a description of
regular expressions,
@xref{syntax, , , regex, "regex" regular expression matching library}.
@defun regcomp @var{pattern} [@var{flags}]
Compile a @dfn{regular expression}. Return a compiled regular
expression, or an integer error code suitable as an argument to
@code{regerror}.
@var{flags} in @code{regcomp} is a string of option letters used to
control the compilation of the regular expression. The letters may
consist of:
@table @samp
@item n
newlines won't be matched by @code{.} or hat lists; ( @code{[^...]} )
@item i
ignore case.
@exdent only when compiled with @var{_GNU_SOURCE}:
@item 0
allows dot to match a null character.
@item f
enable GNU fastmaps.
@end table
@end defun
@defun regerror @var{errno}
Returns a string describing the integer @var{errno} returned when
@code{regcomp} fails.
@end defun
@defun regexec @var{re} @var{string}
Returns @code{#f} or a vector of integers. These integers are in
doublets. The first of each doublet is the index of @var{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
@var{string} of the end of that expression. @code{#f} is returned if
the string does not match.
@end defun
@defun regmatch? @var{re} @var{string}
Returns @code{#t} if the @var{pattern} such that @var{regexp} = (regcomp
@var{pattern}) matches @var{string} as a POSIX extended regular
expressions. Returns @code{#f} otherwise.
@end defun
@defun regsearch @var{re} @var{string} [@var{start} [@var{len}]]
@defunx regsearchv @var{re} @var{string} [@var{start} [@var{len}]]
@defunx regmatch @var{re} @var{string} [@var{start} [@var{len}]]
@defunx regmatchv @var{re} @var{string} [@var{start} [@var{len}]]
@code{Regsearch} searches for the pattern within the string.
@code{Regmatch} anchors the pattern and begins matching it against
string.
@code{Regsearch} returns the character position where @var{re} starts,
or @code{#f} if not found.
@code{Regmatch} returns the number of characters matched, @code{#f} if
not matched.
@code{Regsearchv} and @code{regmatchv} return the match vector is
returned if @var{re} is found, @code{#f} otherwise.
@table @var
@item re
may be either:
@enumerate
@item
a compiled regular expression returned by @code{regcomp};
@item
a string representing a regular expression;
@item
a list of a string and a set of option letters.
@end enumerate
@item string
The string to be operated upon.
@item start
The character position at which to begin the search or match. If absent,
the default is zero.
@exdent @emph{Compiled _GNU_SOURCE and using GNU libregex only@:}
When searching, if @var{start} is negative, the absolute value of
@var{start} will be used as the start location and reverse searching
will be performed.
@item len
The search is allowed to examine only the first @var{len} characters of
@var{string}. If absent, the entire string may be examined.
@end table
@end defun
@defun string-split @var{re} @var{string}
@defunx string-splitv @var{re} @var{string}
@code{String-split} splits a string into substrings that are separated
by @var{re}, returning a vector of substrings.
@code{String-splitv} returns a vector of string positions that indicate
where the substrings are located.
@end defun
@defun string-edit @var{re} @var{edit-spec} @var{string} [@var{count}]
Returns the edited string.
@table @var
@item edit-spec
Is a string used to replace occurances of @var{re}. Backquoted integers
in the range of 1-9 may be used to insert subexpressions in @var{re}, as
in @code{sed}.
@item count
The number of substitutions for @code{string-edit} to perform. If
@code{#t}, all occurances of @var{re} will be replaced. The default is
to perform one substitution.
@end table
@end defun
@node Line Editing, Curses, Regular Expression Pattern Matching, Packages
@section Line Editing
@noindent
These procedures provide input line editing and recall.
@noindent
These functions are defined in @file{edline.c} and @file{Iedline.scm}
using the @dfn{editline} or GNU @dfn{readline}
(@pxref{Top, , Overview ,readline ,GNU Readline Library}) libraries
available from:
@itemize @bullet
@item
@ifset html
@end ifset
@code{ftp.sys.toronto.edu:/pub/rc/editline.shar}
@ifset html
@end ifset
@item
@ifset html
@end ifset
@code{ftp.gnu.org:/pub/gnu/readline-2.0.tar.gz}
@ifset html
@end ifset
@end itemize
@noindent
When @file{Iedline.scm} is loaded, if the current input port is the
default input port and the environment variable @var{EMACS} is not
defined, line-editing mode will be entered.
@defun default-input-port
Returns the initial @code{current-input-port} SCM was invoked with
(stdin).
@end defun
@defun default-output-port
Returns the initial @code{current-output-port} SCM was invoked with
(stdout).
@end defun
@defun make-edited-line-port
Returns an input/output port that allows command line editing and
retrieval of history.
@end defun
@defun line-editing
Returns the current edited line port or @code{#f}.
@defunx line-editing bool
If @var{bool} is false, exits line-editing mode and returns the previous
value of @code{(line-editing)}. If @var{bool} is true, sets the current
input and output ports to an edited line port and returns the previous
value of @code{(line-editing)}.
@end defun
@node Curses, Sockets, Line Editing, Packages
@section Curses
@noindent
These functions are defined in @file{crs.c} using the @dfn{curses}
library. Unless otherwise noted these routines return @code{#t} for
successful completion and @code{#f} for failure.
@defun initscr
Returns a port for a full screen window. This routine must be called to
initialize curses.
@end defun
@defun endwin
A program should call @code{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 @ref{Window Manipulation, refresh}.
@end defun
@menu
* Output Options Setting::
* Terminal Mode Setting::
* Window Manipulation::
* Output::
* Input::
* Curses Miscellany::
@end menu
@node Output Options Setting, Terminal Mode Setting, Curses, Curses
@subsection Output Options Setting
@noindent
These routines set options within curses that deal with output. All
options are initially @code{#f}, unless otherwise stated. It is not
necessary to turn these options off before calling @code{endwin}.
@defun clearok win bf
If enabled (@var{bf} is @code{#t}), the next call to @code{force-output}
or @code{refresh} with @var{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.
@end defun
@defun idlok win bf
If enabled (@var{bf} is @code{#t}), curses will consider using the
hardware ``insert/delete-line'' feature of terminals so equipped. If
disabled (@var{bf} is @code{#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.
@end defun
@defun 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.
@end defun
@defun scrollok win bf
This option controls what happens when the cursor of window @var{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 (@var{bf} is @code{#f}), the cursor is left on the
bottom line at the location where the offending character was entered.
If enabled (@var{bf} is @code{#t}), @code{force-output} is called on the
window @var{win}, and then the physical terminal and window @var{win}
are scrolled up one line.
@emph{Note@:} in order to get the physical scrolling effect on the
terminal, it is also necessary to call @code{idlok}.
@end defun
@defun 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.
@end defun
@node Terminal Mode Setting, Window Manipulation, Output Options Setting, Curses
@subsection Terminal Mode Setting
@noindent
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 @code{endwin}. The routines in this section all return an
unspecified value.
@defun cbreak
@defunx nocbreak
These two routines put the terminal into and out of @code{CBREAK} mode,
respectively. In @code{CBREAK} mode, characters typed by the user are
immediately available to the program and erase/kill character
processing is not performed. When in @code{NOCBREAK} mode, the tty driver
will buffer characters typed until a @key{LFD} or @key{RET} is typed.
Interrupt and flowcontrol characters are unaffected by this mode.
Initially the terminal may or may not be in @code{CBREAK} mode, as it is
inherited, therefore, a program should call @code{cbreak} or @code{nocbreak}
explicitly. Most interactive programs using curses will set @code{CBREAK}
mode.
@emph{Note@:} @code{cbreak} overrides @code{raw}. For a discussion of
how these routines interact with @code{echo} and @code{noecho}
@xref{Input, read-char}.
@end defun
@defun raw
@defunx noraw
The terminal is placed into or out of @code{RAW} mode. @code{RAW} mode
is similar to @code{CBREAK} mode, in that characters typed are
immediately passed through to the user program. The differences are
that in @code{RAW} mode, the interrupt, quit, suspend, and flow control
characters are passed through uninterpreted, instead of generating a
signal. @code{RAW} mode also causes 8-bit input and output. The
behavior of the @code{BREAK} key depends on other bits in the terminal
driver that are not set by curses.
@end defun
@defun echo
@defunx noecho
These routines control whether characters typed by the user are echoed
by @code{read-char} as they are typed. Echoing by the tty driver is
always disabled, but initially @code{read-char} is in @code{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 @code{noecho}.
For a discussion of how these routines interact with @code{echo} and
@code{noecho} @xref{Input, read-char}.
@end defun
@defun nl
@defunx nonl
These routines control whether @key{LFD} is translated into @key{RET}
and @code{LFD} on output, and whether @key{RET} is translated into
@key{LFD} on input. Initially, the translations do occur. By disabling
these translations using @code{nonl}, curses is able to make better use
of the linefeed capability, resulting in faster cursor motion.
@end defun
@defun resetty
@defunx savetty
These routines save and restore the state of the terminal modes.
@code{savetty} saves the current state of the terminal in a buffer and
@code{resetty} restores the state to what it was at the last call to
@code{savetty}.
@end defun
@node Window Manipulation, Output, Terminal Mode Setting, Curses
@subsection Window Manipulation
@defun newwin nlines ncols begy begx
Create and return a new window with the given number of lines (or rows),
@var{nlines}, and columns, @var{ncols}. The upper left corner of the
window is at line @var{begy}, column @var{begx}. If either @var{nlines}
or @var{ncols} is 0, they will be set to the value of
@code{LINES}-@var{begy} and @code{COLS}-@var{begx}. A new full-screen
window is created by calling @code{newwin(0,0,0,0)}.
@end defun
@defun subwin orig nlines ncols begy begx
Create and return a pointer to a new window with the given number of
lines (or rows), @var{nlines}, and columns, @var{ncols}. The window is
at position (@var{begy}, @var{begx}) on the screen. This position is
relative to the screen, and not to the window @var{orig}. The window is
made in the middle of the window @var{orig}, so that changes made to one
window will affect both windows. When using this routine, often it will
be necessary to call @code{touchwin} or @code{touchline} on @var{orig}
before calling @code{force-output}.
@end defun
@defun close-port win
Deletes the window @var{win}, freeing up all memory associated with it.
In the case of sub-windows, they should be deleted before the main
window @var{win}.
@end defun
@defun refresh
@defunx force-output win
These routines are called to write output to the terminal, as most other
routines merely manipulate data structures. @code{force-output} copies
the window @var{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
@code{leaveok} has been enabled, the physical cursor of the terminal is
left at the location of window @var{win}'s cursor. With @code{refresh},
the number of characters output to the terminal is returned.
@end defun
@defun mvwin win y x
Move the window @var{win} so that the upper left corner will be at position
(@var{y}, @var{x}). If the move would cause the window @var{win} to be off the
screen, it is an error and the window @var{win} is not moved.
@end defun
@defun overlay srcwin dstwin
@defunx overwrite srcwin dstwin
These routines overlay @var{srcwin} on top of @var{dstwin}; that is, all
text in @var{srcwin} is copied into @var{dstwin}. @var{srcwin} and
@var{dstwin} need not be the same size; only text where the two windows
overlap is copied. The difference is that @code{overlay} is
non-destructive (blanks are not copied), while @code{overwrite} is
destructive.
@end defun
@defun touchwin win
@defunx touchline win start count
Throw away all optimization information about which parts of the window
@var{win} have been touched, by pretending that the entire window
@var{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. @code{touchline} only pretends that
@var{count} lines have been changed, beginning with line @var{start}.
@end defun
@defun wmove win y x
The cursor associated with the window @var{win} is moved to line (row) @var{y},
column @var{x}. This does not move the physical cursor of the terminal
until @code{refresh} (or @code{force-output}) is called. The position
specified is relative to the upper left corner of the window @var{win},
which is (0, 0).
@end defun
@node Output, Input, Window Manipulation, Curses
@subsection Output
@noindent
These routines are used to @dfn{draw} text on windows
@defun display ch win
@defunx display str win
@defunx wadd win ch
@defunx wadd win str
The character @var{ch} or characters in @var{str} are put into the
window @var{win} at the current cursor position of the window and the
position of @var{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 @var{ch} is a @key{TAB}, @key{LFD}, or backspace, the cursor will be
moved appropriately within the window @var{win}. A @key{LFD} also does a
@code{wclrtoeol} before moving. @key{TAB} characters are considered to
be at every eighth column. If @var{ch} is another control character, it
will be drawn in the @kbd{C-x} notation. (Calling @code{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 @code{standout},
below.
@emph{Note@:} For @code{wadd} @var{ch} can be an integer and will insert
the character of the corresponding value.
@end defun
@defun werase win
This routine copies blanks to every position in the window @var{win}.
@end defun
@defun wclear win
This routine is like @code{werase}, but it also calls
@ref{Output Options Setting, clearok}, arranging that the screen will
be cleared completely on the next call to @code{refresh} or
@code{force-output} for window @var{win}, and repainted from scratch.
@end defun
@defun wclrtobot win
All lines below the cursor in window @var{win} are erased. Also, the
current line to the right of the cursor, inclusive, is erased.
@end defun
@defun wclrtoeol win
The current line to the right of the cursor, inclusive, is erased.
@end defun
@defun wdelch win
The character under the cursor in the window @var{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.
@end defun
@defun wdeleteln win
The line under the cursor in the window @var{win} is deleted. All lines
below the current line are moved up one line. The bottom line @var{win}
is cleared. The cursor position does not change. This does not imply
use of the hardware ``deleteline'' feature.
@end defun
@defun winsch win ch
The character @var{ch} is inserted before the character under the
cursor. All characters to the right are moved one @key{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.
@end defun
@defun 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.
@end defun
@defun scroll win
The window @var{win} is scrolled up one line. This involves moving the
lines in @var{win}'s data structure. As an optimization, if @var{win}
is stdscr and the scrolling region is the entire window, the physical
screen will be scrolled at the same time.
@end defun
@node Input, Curses Miscellany, Output, Curses
@subsection Input
@defun read-char win
A character is read from the terminal associated with the window
@var{win}. Depending on the setting of @code{cbreak}, this will be
after one character (@code{CBREAK} mode), or after the first newline
(@code{NOCBREAK} mode). Unless @code{noecho} has been set, the
character will also be echoed into @var{win}.
When using @code{read-char}, do not set both @code{NOCBREAK} mode
(@code{nocbreak}) and @code{ECHO} mode (@code{echo}) at the same time.
Depending on the state of the terminal driver when each character is
typed, the program may produce undesirable results.
@end defun
@defun winch win
The character, of type chtype, at the current position in window
@var{win} is returned. If any attributes are set for that position,
their values will be OR'ed into the value returned.
@end defun
@defun getyx win
A list of the y and x coordinates of the cursor position of the window
@var{win} is returned
@end defun
@node Curses Miscellany, , Input, Curses
@subsection Curses Miscellany
@defun wstandout win
@defunx wstandend win
These functions set the current attributes of the window @var{win}. The
current attributes of @var{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.
@code{wstandout} sets the current attributes of the window @var{win} to
be visibly different from other text. @code{wstandend} turns off the
attributes.
@end defun
@defun box win vertch horch
A box is drawn around the edge of the window @var{win}. @var{vertch}
and @var{horch} are the characters the box is to be drawn with. If
@var{vertch} and @var{horch} are 0, then appropriate default characters,
@code{ACS_VLINE} and @code{ACS_HLINE}, will be used.
@emph{Note@:} @var{vertch} and @var{horch} can be an integers and will
insert the character (with attributes) of the corresponding values.
@end defun
@defun unctrl c
This macro expands to a character string which is a printable
representation of the character @var{c}. Control characters are
displayed in the @kbd{C-x} notation. Printing characters are displayed
as is.
@end defun
@node Sockets, SCMDB, Curses, Packages
@section Sockets
@noindent
These procedures (defined in @file{socket.c}) provide a Scheme interface
to most of the C @dfn{socket} library. For more information on sockets,
@xref{Sockets, , , libc, The GNU C Library Reference Manual}.
@menu
* Host and Other Inquiries::
* Internet Addresses and Socket Names::
* Socket::
@end menu
@node Host and Other Inquiries, Internet Addresses and Socket Names, Sockets, Sockets
@subsection Host and Other Inquiries
@defvr Constant af_inet
@defvrx Constant af_unix
Integer family codes for Internet and Unix sockets, respectively.
@end defvr
@defun gethost host-spec
@defunx gethost
Returns a vector of information for the entry for @code{HOST-SPEC} or the
next entry if @code{HOST-SPEC} isn't given. The information is:
@enumerate 0
@item
host name string
@item
list of host aliases strings
@item
integer address type (@code{AF_INET})
@item
integer size of address entries (in bytes)
@item
list of integer addresses
@end enumerate
@end defun
@defun sethostent stay-open
@defunx sethostent
Rewinds the host entry table back to the begining if given an argument.
If the argument @var{stay-open} is @code{#f} queries will be be done
using @code{UDP} datagrams. Otherwise, a connected @code{TCP} socket
will be used. When called without an argument, the host table is
closed.
@end defun
@defun getnet name-or-number
@defunx getnet
Returns a vector of information for the entry for @var{name-or-number} or
the next entry if an argument isn't given. The information is:
@enumerate 0
@item
official network name string
@item
list of network aliases strings
@item
integer network address type (@code{AF_INET})
@item
integer network number
@end enumerate
@end defun
@defun setnetent stay-open
@defunx setnetent
Rewinds the network entry table back to the begining if given an
argument. If the argument @var{stay-open} is @code{#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.
@end defun
@defun getproto name-or-number
@defunx getproto
Returns a vector of information for the entry for @var{name-or-number} or
the next entry if an argument isn't given. The information is:
@enumerate
@item
official protocol name string
@item
list of protocol aliases strings
@item
integer protocol number
@end enumerate
@end defun
@defun setprotoent stay-open
@defunx setprotoent
Rewinds the protocol entry table back to the begining if given an
argument. If the argument @var{stay-open} is @code{#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.
@end defun
@defun getserv name-or-port-number protocol
@defunx getserv
Returns a vector of information for the entry for @var{name-or-port-number}
and @var{protocol} or the next entry if arguments aren't given. The
information is:
@enumerate 0
@item
official service name string
@item
list of service aliases strings
@item
integer port number
@item
protocol
@end enumerate
@end defun
@defun setservent stay-open
@defunx setservent
Rewinds the service entry table back to the begining if given an
argument. If the argument @var{stay-open} is @code{#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.
@end defun
@node Internet Addresses and Socket Names, Socket, Host and Other Inquiries, Sockets
@subsection Internet Addresses and Socket Names
@defun inet:string->address string
Returns the host address number (integer) for host @var{string} or
@code{#f} if not found.
@end defun
@defun inet:address->string address
Converts an internet (integer) address to a string in numbers and dots
notation.
@end defun
@defun inet:network address
Returns the network number (integer) specified from @var{address} or
@code{#f} if not found.
@end defun
@defun inet:local-network-address address
Returns the integer for the address of @var{address} within its local
network or @code{#f} if not found.
@end defun
@defun inet:make-address network local-address
Returns the Internet address of @var{local-address} in @var{network}.
@end defun
@noindent
The type @dfn{socket-name} is used for inquiries about open sockets in
the following procedures:
@defun getsockname socket
Returns the socket-name of @var{socket}. Returns @code{#f} if
unsuccessful or @var{socket} is closed.
@end defun
@defun getpeername socket
Returns the socket-name of the socket connected to @var{socket}.
Returns @code{#f} if unsuccessful or @var{socket} is closed.
@end defun
@defun socket-name:family socket-name
Returns the integer code for the family of @var{socket-name}.
@end defun
@defun socket-name:port-number socket-name
Returns the integer port number of @var{socket-name}.
@end defun
@defun socket-name:address socket-name
Returns the integer Internet address for @var{socket-name}.
@end defun
@node Socket, , Internet Addresses and Socket Names, Sockets
@subsection Socket
@noindent
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):
@example
(require 'i/o-extensions)
@ftindex i/o-extensions
(define i-port (duplicate-port sock-port "r"))
(define o-port (duplicate-port sock-port "w"))
@end example
@defun make-stream-socket family
@defunx make-stream-socket family protocol
Returns a @code{SOCK_STREAM} socket of type @var{family} using
@var{protocol}. If @var{family} has the value @code{AF_INET},
@code{SO_REUSEADDR} will be set. The integer argument @var{protocol}
corresponds to the integer protocol numbers returned (as vector
elements) from @code{(getproto)}. If the @var{protocol} argument is not
supplied, the default (0) for the specified @var{family} is used. SCM
sockets look like ports opened for neither reading nor writing.
@end defun
@defun make-stream-socketpair family
@defunx make-stream-socketpair family protocol
Returns a pair (cons) of connected @code{SOCK_STREAM} (socket) ports of
type @var{family} using @var{protocol}. Many systems support only
socketpairs of the @code{af-unix} @var{family}. The integer argument
@var{protocol} corresponds to the integer protocol numbers returned (as
vector elements) from (getproto). If the @var{protocol} argument is
not supplied, the default (0) for the specified @var{family} is used.
@end defun
@defun socket:shutdown socket how
Makes @var{socket} no longer respond to some or all operations depending on
the integer argument @var{how}:
@enumerate 0
@item
Further input is disallowed.
@item
Further output is disallowed.
@item
Further input or output is disallowed.
@end enumerate
@code{Socket:shutdown} returns @var{socket} if successful, @code{#f} if
not.
@end defun
@defun socket:connect inet-socket host-number port-number
@defunx socket:connect unix-socket pathname
Returns @var{socket} (changed to a read/write port) connected to the
Internet socket on host @var{host-number}, port @var{port-number} or
the Unix socket specified by @var{pathname}. Returns @code{#f} if not
successful.
@end defun
@defun socket:bind inet-socket port-number
@defunx socket:bind unix-socket pathname
Returns @var{inet-socket} bound to the integer @var{port-number} or the
@var{unix-socket} bound to new socket in the file system at location
@var{pathname}. Returns @code{#f} if not successful. Binding a
@var{unix-socket} creates a socket in the file system that must be
deleted by the caller when it is no longer needed (using
@code{delete-file}).
@end defun
@defun socket:listen socket backlog
The bound (@pxref{Socket, bind}) @var{socket} is readied to
accept connections. The positive integer @var{backlog} specifies how
many pending connections will be allowed before further connection
requests are refused. Returns @var{socket} (changed to a read-only
port) if successful, @code{#f} if not.
@end defun
@defun char-ready? listen-socket
The input port returned by a successful call to @code{socket:listen} can
be polled for connections by @code{char-ready?}
(@pxref{Files and Ports, char-ready?}). This avoids blocking on
connections by @code{socket:accept}.
@end defun
@defun socket:accept socket
Accepts a connection on a bound, listening @var{socket}. Returns an
input/output port for the connection.
@end defun
@noindent
The following example is not too complicated, yet shows the use of
sockets for multiple connections without input blocking.
@example
;;;; 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)
@ftindex socket
(require 'i/o-extensions)
@ftindex 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-position con 0)
(write-char c con)
(file-position con 0))
connections)
(cons con (next (cdr con-list)))))))
(else (cons con (next (cdr con-list)))))))))))))
@end example
@noindent
You can use @samp{telnet localhost 8001} to connect to the chat server,
or you can use a client written in scheme:
@example
;;;; 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)
@ftindex socket
(require 'i/o-extensions)
@ftindex 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-position con 0)
(display ct con)
(file-position con 0)))
(go)))))
(cond (con (display "Connecting to ")
(display (getpeername con))
(newline)
(go))
(else (display "Server not listening on port 8001")
(newline)))
@end example
@node SCMDB, , Sockets, Packages
@section SCMDB
@code{(require 'mysql)}
@ifset html
@end ifset
@dfn{SCMDB}
@ifset html
@end ifset
is an add-on for SCM that ports the MySQL C-library to SCM.
@ifclear html
It is available from:
@url{http://www.dedecker.net/jessie/scmdb/}
@end ifclear
@iftex
@section Xlibscm
@code{(require 'Xlib)}
@ifset html
@dfn{Xlibscm}
is a SCM interface to the
X Window System.
@end ifset
@ifclear html
@xref{Top, ,SCM Language X Interface , Xlibscm, Xlibscm} for the SCM
interface to the @dfn{X Window System}.
@end ifclear
@section Hobbit
@code{(require 'hobbit)}
@code{(require 'compile)}
@ifset html
@dfn{Hobbit}
is a small optimizing scheme-to-C compiler
generating C modules for use with the SCM interpreter.
@end ifset
@ifclear html
@xref{Top, , SCM Compiler, hobbit, hobbit} for a small optimizing
scheme-to-C compiler for use with the SCM interpreter.
@end ifclear
@end iftex
@node The Implementation, Index, Packages, Top
@chapter The Implementation
@menu
* Data Types::
* Operations::
* Program Self-Knowledge:: What SCM needs to know about itself.
* Improvements To Make::
@end menu
@node Data Types, Operations, The Implementation, The Implementation
@section Data Types
@noindent
In the descriptions below it is assumed that @code{long int}s are 32
bits in length. Acutally, SCM is written to work with any @code{long
int} size larger than 31 bits. With some modification, SCM could work
with word sizes as small as 24 bits.
@noindent
All SCM objects are represented by type @dfn{SCM}. Type @code{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
@end menu
@node Immediates, Cells, Data Types, Data Types
@subsection Immediates
@noindent
An @dfn{immediate} is a data type contained in type @code{SCM}
(@code{long int}). The type codes distinguishing immediate types from
each other vary in length, but reside in the low order bits.
@defmac IMP x
@defmacx NIMP x
Return non-zero if the @code{SCM} object @var{x} is an immediate or
non-immediate type, respectively.
@end defmac
@deftp Immediate inum
immediate 30 bit signed integer. An INUM is flagged by a @code{1} in
the second to low order bit position. The high order 30 bits are used
for the integer's value.
@defmac INUMP x
@defmacx NINUMP x
Return non-zero if the @code{SCM} @var{x} is an immediate integer or not
an immediate integer, respectively.
@end defmac
@defmac INUM x
Returns the C @code{long integer} corresponding to @code{SCM} @var{x}.
@end defmac
@defmac MAKINUM x
Returns the @code{SCM} inum corresponding to C @code{long integer} x.
@end defmac
@defvr {Immediate Constant} INUM0
is equivalent to @code{MAKINUM(0)}.
@end defvr
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 @file{scmfig.h} and a signed right shift,
@code{SRS}, is constructed in terms of unsigned right shift.
@end deftp
@deftp Immediate ichr
characters.
@defmac ICHRP x
Return non-zero if the @code{SCM} object @var{x} is a character.
@end defmac
@defmac ICHR x
Returns corresponding @code{unsigned char}.
@end defmac
@defmac MAKICHR x
Given @code{char} @var{x}, returns @code{SCM} character.
@end defmac
@end deftp
@deftp Immediate iflags
These are frequently used immediate constants.
@deftypevr {Immediate Constant} SCM BOOL_T
@code{#t}
@end deftypevr
@deftypevr {Immediate Constant} SCM BOOL_F
@code{#f}
@end deftypevr
@deftypevr {Immediate Constant} SCM EOL
@code{()}. If @code{SICP} is @code{#define}d, @code{EOL} is
@code{#define}d to be identical with @code{BOOL_F}. In this case, both
print as @code{#f}.
@end deftypevr
@deftypevr {Immediate Constant} SCM EOF_VAL
end of file token, @code{#}.
@end deftypevr
@deftypevr {Immediate Constant} SCM UNDEFINED
@code{#} used for variables which have not been defined and
absent optional arguments.
@end deftypevr
@deftypevr {Immediate Constant} SCM UNSPECIFIED
@code{#} is returned for those procedures whose return
values are not specified.
@end deftypevr
@end deftp
@defmac IFLAGP n
Returns non-zero if @var{n} is an ispcsym, isym or iflag.
@end defmac
@defmac ISYMP n
Returns non-zero if @var{n} is an ispcsym or isym.
@end defmac
@defmac ISYMNUM n
Given ispcsym, isym, or iflag @var{n}, returns its index in the C array
@code{isymnames[]}.
@end defmac
@defmac ISYMCHARS n
Given ispcsym, isym, or iflag @var{n}, returns its @code{char *}
representation (from @code{isymnames[]}).
@end defmac
@defmac MAKSPCSYM n
Returns @code{SCM} ispcsym @var{n}.
@end defmac
@defmac MAKISYM n
Returns @code{SCM} iisym @var{n}.
@end defmac
@defmac MAKIFLAG n
Returns @code{SCM} iflag @var{n}.
@end defmac
@defvar isymnames
An array of strings containing the external representations of all the
ispcsym, isym, and iflag immediates. Defined in @file{repl.c}.
@end defvar
@defvr Constant NUM_ISPCSYM
@defvrx Constant NUM_ISYMS
The number of ispcsyms and ispcsyms+isyms, respectively. Defined in
@file{scm.h}.
@end defvr
@deftp Immediate isym
@code{and}, @code{begin}, @code{case}, @code{cond}, @code{define},
@code{do}, @code{if}, @code{lambda}, @code{let}, @code{let*},
@code{letrec}, @code{or}, @code{quote}, @code{set!}, @code{#f},
@code{#t}, @code{#}, @code{#}, @code{()}, and
@code{#}.
@deftpx {CAR Immediate} ispcsym
special symbols: syntax-checked versions of first 14 isyms
@end deftp
@deftp {CAR Immediate} iloc
indexes to a variable's location in environment
@end deftp
@deftp {CAR Immediate} gloc
pointer to a symbol's value cell
@end deftp
@deftp 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 @code{0}.
There is one exception to this rule, @emph{CAR Immediate}s, described
next.
@end deftp
@noindent
A @dfn{CAR Immediate} is an Immediate point which can only occur in the
@code{CAR}s of evaluated code (as a result of @code{ceval}'s memoization
process).
@node Cells, Header Cells, Immediates, Data Types
@subsection Cells
@noindent
@dfn{Cell}s represent all SCM objects other than immediates. A cell has
a @code{CAR} and a @code{CDR}. Low-order bits in @code{CAR} identify
the type of object. The rest of @code{CAR} and @code{CDR} hold object
data. The number after @code{tc} specifies how many bits are in the
type code. For instance, @code{tc7} indicates that the type code is 7
bits.
@defmac NEWCELL x
Allocates a new cell and stores a pointer to it in @code{SCM} local
variable @var{x}.
Care needs to be taken that stores into the new cell pointed to by
@var{x} do not create an inconsistent object. @xref{Signals}.
@end defmac
@noindent
All of the C macros decribed in this section assume that their argument
is of type @code{SCM} and points to a cell (@code{CELLPTR}).
@defmac CAR x
@defmacx CDR x
Returns the @code{car} and @code{cdr} of cell @var{x}, respectively.
@end defmac
@defmac TYP3 x
@defmacx TYP7 x
@defmacx TYP16 x
Returns the 3, 7, and 16 bit type code of a cell.
@end defmac
@deftp Cell tc3_cons
scheme cons-cell returned by (cons arg1 arg2).
@defmac CONSP x
@defmacx NCONSP x
Returns non-zero if @var{x} is a @code{tc3_cons} or isn't, respectively.
@end defmac
@end deftp
@deftp Cell tc3_closure
applicable object returned by (lambda (args) @dots{}).
@code{tc3_closure}s have a pointer to the body of the procedure in the
@code{CAR} and a pointer to the environment in the @code{CDR}. Bits 1
and 2 (zero-based) in the @code{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 @code{CDR}: In the case of
an empty environment all bits above 2 in the @code{CDR} are zero.
@defmac CLOSUREP x
Returns non-zero if @var{x} is a @code{tc3_closure}.
@end defmac
@defmac CODE x
@defmacx ENV x
Returns the code body or environment of closure @var{x}, respectively.
@end defmac
@defmac ARGC x
Returns the a lower bound on the number of required arguments to closure
@var{x}, it cannot exceed 3.
@end defmac
@end deftp
@node Header Cells, Subr Cells, Cells, Data Types
@subsection Header Cells
@noindent
@dfn{Header}s are Cells whose @code{CDR}s point elsewhere in memory,
such as to memory allocated by @code{malloc}.
@deftp Header spare
spare @code{tc7} type code
@end deftp
@deftp Header tc7_vector
scheme vector.
@defmac VECTORP x
@defmacx NVECTORP x
Returns non-zero if @var{x} is a @code{tc7_vector} or if not, respectively.
@end defmac
@defmac VELTS x
@defmacx LENGTH x
Returns the C array of @code{SCM}s holding the elements of vector
@var{x} or its length, respectively.
@end defmac
@end deftp
@deftp Header tc7_ssymbol
static scheme symbol (part of initial system)
@deftpx Header tc7_msymbol
@code{malloc}ed scheme symbol (can be GCed)
@defmac SYMBOLP x
Returns non-zero if @var{x} is a @code{tc7_ssymbol} or
@code{tc7_msymbol}.
@end defmac
@defmac CHARS x
@defmacx UCHARS x
@defmacx LENGTH x
Returns the C array of @code{char}s or as @code{unsigned char}s holding
the elements of symbol @var{x} or its length, respectively.
@end defmac
@end deftp
@deftp Header tc7_string
scheme string
@defmac STRINGP x
@defmacx NSTRINGP x
Returns non-zero if @var{x} is a @code{tc7_string} or isn't,
respectively.
@end defmac
@defmac CHARS x
@defmacx UCHARS x
@defmacx LENGTH x
Returns the C array of @code{char}s or as @code{unsigned char}s holding
the elements of string @var{x} or its length, respectively.
@end defmac
@end deftp
@deftp Header tc7_Vbool
uniform vector of booleans (bit-vector)
@end deftp
@deftp Header tc7_VfixZ32
uniform vector of integers
@end deftp
@deftp Header tc7_VfixN32
uniform vector of non-negative integers
@end deftp
@deftp Header tc7_VfixN16
uniform vector of non-negative short integers
@end deftp
@deftp Header tc7_VfixZ16
uniform vector of short integers
@end deftp
@deftp Header tc7_VfixN8
uniform vector of non-negative bytes
@end deftp
@deftp Header tc7_VfixZ8
uniform vector of signed bytes
@end deftp
@deftp Header tc7_VfloR32
uniform vector of short inexact real numbers
@end deftp
@deftp Header tc7_VfloR64
uniform vector of double precision inexact real numbers
@end deftp
@deftp Header tc7_VfloC64
uniform vector of double precision inexact complex numbers
@end deftp
@deftp Header tc7_contin
applicable object produced by call-with-current-continuation
@end deftp
@deftp Header tc7_specfun
subr that is treated specially within the evaluator
@code{apply} and @code{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:
@enumerate
@item
the cclo itself is consed onto the head of the argument list
@item
the first element of the cclo is applied to that list. Cclo invocation
is currently not tail recursive when given 2 or more arguments.
@end enumerate
@defun makcclo proc len
makes a closure from the @emph{subr} @var{proc} with @var{len}-1 extra
locations for @code{SCM} data. Elements of a @var{cclo} are referenced
using @code{VELTS(cclo)[n]} just as for vectors.
@end defun
@defmac CCLO_LENGTH cclo
Expands to the length of @var{cclo}.
@end defmac
@end deftp
@node Subr Cells, Ptob Cells, Header Cells, Data Types
@subsection Subr Cells
@noindent
A @dfn{Subr} is a header whose @code{CDR} points to a C code procedure.
Scheme primitive procedures are subrs. Except for the arithmetic
@code{tc7_cxr}s, the C code procedures will be passed arguments (and
return results) of type @code{SCM}.
@deftp Subr tc7_asubr
associative C function of 2 arguments. Examples are @code{+}, @code{-},
@code{*}, @code{/}, @code{max}, and @code{min}.
@end deftp
@deftp Subr tc7_subr_0
C function of no arguments.
@end deftp
@deftp Subr tc7_subr_1
C function of one argument.
@end deftp
@deftp Subr tc7_cxr
These subrs are handled specially. If inexact numbers are enabled, the
@code{CDR} should be a function which takes and returns type
@code{double}. Conversions are handled in the interpreter.
@code{floor}, @code{ceiling}, @code{truncate}, @code{round},
@code{real-sqrt}, @code{real-exp}, @code{real-ln}, @code{real-sin},
@code{real-cos}, @code{real-tan}, @code{real-asin}, @code{real-acos}, @code{real-atan},
@code{real-sinh}, @code{real-cosh}, @code{real-tanh}, @code{real-asinh}, @code{real-acosh},
@code{real-atanh}, and @code{exact->inexact} are defined this way.
If the @code{CDR} is @code{0} (@code{NULL}), the name string of the
procedure is used to control traversal of its list structure argument.
@code{car}, @code{cdr}, @code{caar}, @code{cadr}, @code{cdar},
@code{cddr}, @code{caaar}, @code{caadr}, @code{cadar}, @code{caddr},
@code{cdaar}, @code{cdadr}, @code{cddar}, @code{cdddr}, @code{caaaar},
@code{caaadr}, @code{caadar}, @code{caaddr}, @code{cadaar},
@code{cadadr}, @code{caddar}, @code{cadddr}, @code{cdaaar},
@code{cdaadr}, @code{cdadar}, @code{cdaddr}, @code{cddaar},
@code{cddadr}, @code{cdddar}, and @code{cddddr} are defined this way.
@end deftp
@deftp Subr tc7_subr_3
C function of 3 arguments.
@end deftp
@deftp Subr tc7_subr_2
C function of 2 arguments.
@end deftp
@deftp Subr tc7_rpsubr
transitive relational predicate C function of 2 arguments. The C
function should return either @code{BOOL_T} or @code{BOOL_F}.
@end deftp
@deftp Subr tc7_subr_1o
C function of one optional argument. If the optional argument is not
present, @code{UNDEFINED} is passed in its place.
@end deftp
@deftp Subr tc7_subr_2o
C function of 1 required and 1 optional argument. If the optional
argument is not present, @code{UNDEFINED} is passed in its place.
@end deftp
@deftp Subr tc7_lsubr_2
C function of 2 arguments and a list of (rest of) @code{SCM} arguments.
@end deftp
@deftp Subr tc7_lsubr
C function of list of @code{SCM} arguments.
@end deftp
@node Ptob Cells, Smob Cells, Subr Cells, Data Types
@subsection Ptob Cells
@noindent
A @dfn{ptob} is a port object, capable of delivering or accepting
@tindex ptob
characters.
@xref{Ports, , , r5rs, Revised(5) Report on the Algorithmic Language Scheme}.
Unlike the types described so far, new varieties of ptobs can be
defined dynamically (@pxref{Defining Ptobs}). These are the initial
ptobs:
@deftp ptob tc16_inport
input port.
@end deftp
@deftp ptob tc16_outport
output port.
@end deftp
@deftp ptob tc16_ioport
input-output port.
@end deftp
@deftp ptob tc16_inpipe
input pipe created by @code{popen()}.
@end deftp
@deftp ptob tc16_outpipe
output pipe created by @code{popen()}.
@end deftp
@deftp ptob tc16_strport
String port created by @code{cwos()} or @code{cwis()}.
@end deftp
@deftp ptob tc16_sfport
Software (virtual) port created by @code{mksfpt()} (@pxref{Soft Ports}).
@end deftp
@defmac PORTP x
@defmacx OPPORTP x
@defmacx OPINPORTP x
@defmacx OPOUTPORTP x
@defmacx INPORTP x
@defmacx OUTPORTP x
Returns non-zero if @var{x} is a port, open port, open input-port, open
output-port, input-port, or output-port, respectively.
@end defmac
@defmac OPENP x
@defmacx CLOSEDP x
Returns non-zero if port @var{x} is open or closed, respectively.
@end defmac
@defmac STREAM x
Returns the @code{FILE *} stream for port @var{x}.
@end defmac
@noindent
Ports which are particularly well behaved are called @dfn{fport}s.
Advanced operations like @code{file-position} and @code{reopen-file}
only work for fports.
@defmac FPORTP x
@defmacx OPFPORTP x
@defmacx OPINFPORTP x
@defmacx OPOUTFPORTP x
Returns non-zero if @var{x} is a port, open port, open input-port, or
open output-port, respectively.
@end defmac
@node Smob Cells, Data Type Representations, Ptob Cells, Data Types
@subsection Smob Cells
@noindent
A @dfn{smob} is a miscellaneous datatype. The type code and GCMARK bit
@tindex smob
occupy the lower order 16 bits of the @code{CAR} half of the cell. The
rest of the @code{CAR} can be used for sub-type or other information.
The @code{CDR} contains data of size long and is often a pointer to
allocated memory.
@noindent
Like ptobs, new varieties of smobs can be defined dynamically
(@pxref{Defining Smobs}). These are the initial smobs:
@deftp smob tc_free_cell
unused cell on the freelist.
@end deftp
@deftp smob tc16_flo
single-precision float.
Inexact number data types are subtypes of type @code{tc16_flo}. If the
sub-type is:
@enumerate 0
@item
a single precision float is contained in the @code{CDR}.
@item
@code{CDR} is a pointer to a @code{malloc}ed double.
@end enumerate
@enumerate 3
@item
@code{CDR} is a pointer to a @code{malloc}ed pair of doubles.
@end enumerate
@deftp smob tc_dblr
double-precision float.
@end deftp
@deftp smob tc_dblc
double-precision complex.
@end deftp
@end deftp
@deftp smob tc16_bigpos
@deftpx 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 @code{malloc}ed array
of type @code{BIGDIG} which must be an unsigned integral type with size
smaller than @code{long}. @code{BIGRAD} is the radix associated with
@code{BIGDIG}.
@code{NUMDIGS_MAX} (defined in @file{scmfig.h}) limits the number of
digits of a bignum to 1000. These digits are base @code{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 @emph{some multiple} of the product of the
number of digits of the two multiplicands.
@example
digits * digits ==> operations
5 x
50 100 * x
500 10000 * x
5000 1000000 * x
@end example
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:
@center @url{ftp://megrez.math.u-bordeaux.fr/pub/pari/}
@end deftp
@deftp smob tc16_promise
made by DELAY. @xref{Control features, , , r5rs, Revised(5) Scheme}.
@end deftp
@deftp smob tc16_arbiter
synchronization object. @xref{Process Synchronization}.
@end deftp
@deftp smob tc16_macro
macro expanding function. @xref{Macro Primitives}.
@end deftp
@deftp smob tc16_array
multi-dimensional array. @xref{Arrays}.
This type implements both conventional arrays (those with arbitrary data
as elements @pxref{Conventional Arrays}) and uniform arrays (those with
elements of a uniform type @pxref{Uniform Array}).
Conventional Arrays have a pointer to a vector for their @code{CDR}.
Uniform Arrays have a pointer to a Uniform Vector type (string, Vbool,
VfixZ32, VfixN32, VfloR32, VfloR64, or VfloC64) in their @code{CDR}.
@end deftp
@node Data Type Representations, , Smob Cells, Data Types
@subsection Data Type Representations
@format
@r{IMMEDIATE: B,D,E,F=data bit, C=flag code, P=pointer address bit}
@t{ ................................
inum BBBBBBBBBBBBBBBBBBBBBBBBBBBBBB10
ichr BBBBBBBBBBBBBBBBBBBBBBBB11110100
iflag CCCCCCC101110100
isym CCCCCCC001110100}
@r{ IMCAR: only in car of evaluated code, cdr has cell's GC bit}
@t{ispcsym 000CCCC00CCCC100
iloc 0DDDDDDDDDDDEFFFFFFFFFFF11111100
pointer PPPPPPPPPPPPPPPPPPPPPPPPPPPPP000
gloc PPPPPPPPPPPPPPPPPPPPPPPPPPPPP001}
@r{ 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@:}
@t{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...........
Vbool .........long length....G0010101 ..........long *words...........
spare 00010111
VfixN8 .........long length....G0011101 ......unsigned char *words......
VfixZ8 .........long length....G0011111 ..........char *words...........
VfixN16 .........long length....G0100101 ......unsigned short *words.....
VfixZ16 .........long length....G0100111 ........ short *words...........
VfixN32 .........long length....G0101101 ......unsigned long *words......
VfixZ32 .........long length....G0101111 ..........long *words...........
VfloR32 .........long length....G0110101 .........float *words...........
VfloC32 .........long length....G0110111 .........float *words...........
VfloR64 .........long length....G0111101 ........double *words...........
VfloC64 .........long length....G0111111 ........double *words...........
spare 01000101
contin .........long length....G1001101 .............*regs..............
specfun ................xxxxxxxxG1001111 ...........SCM name.............
cclo ..short length..xxxxxx10G1001111 ...........SCM **elts...........}
@r{ PTOBs@:}
@t{ port int portnum.CwroxxxxxxxxG1000111 ..........FILE *stream..........
socket int portnum.C001xxxxxxxxG1000111 ..........FILE *stream..........
inport int portnum.C011xxxxxxxxG1000111 ..........FILE *stream..........
outport int portnum.0101xxxxxxxxG1000111 ..........FILE *stream..........
ioport int portnum.C111xxxxxxxxG1000111 ..........FILE *stream..........
fport int portnum.C 00000000G1000111 ..........FILE *stream..........
pipe int portnum.C 00000001G1000111 ..........FILE *stream..........
strport 00000000000.0 00000010G1000111 ..........FILE *stream..........
sfport int portnum.C 00000011G1000111 ..........FILE *stream..........}
@r{ SUBRs@:}
@t{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)()...........}
@r{ SMOBs@:}
@t{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..............}
@end format
@node Operations, Program Self-Knowledge, Data Types, The Implementation
@section 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
@end menu
@node Garbage Collection, Memory Management for Environments, Operations, Operations
@subsection Garbage Collection
The garbage collector is in the latter half of @file{sys.c}. The
primary goal of @dfn{garbage collection} (or @dfn{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 @dfn{heap} (composed of @dfn{heap segments}).
Note that this is different from what Computer Science usually defines
as a heap.
@menu
* Marking Cells::
* Sweeping the Heap::
@end menu
@node Marking Cells, Sweeping the Heap, Garbage Collection, Garbage Collection
@subsubsection Marking Cells
The first step in garbage collection is to @dfn{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 @file{scm.h}
to allow easy manipulation when GC bits are possibly set. @code{CAR},
@code{TYP3}, and @code{TYP7} can be used on GC marked cells as they are.
@defmac GCCDR x
Returns the CDR of a cons cell, even if that cell has been GC marked.
@end defmac
@defmac GCTYP16 x
Returns the 16 bit type code of a cell.
@end defmac
We need to (recursively) mark only a few objects in order to assure that
all accessible objects are marked. Those objects are
@code{sys_protects[]} (for example, @code{dynwinds}), the current
C-stack and the hash table for symbols, @dfn{symhash}.
@deftypefun void gc_mark (SCM @var{obj})
The function @code{gc_mark()} is used for marking SCM cells. If
@var{obj} is marked, @code{gc_mark()} returns. If @var{obj} is
unmarked, gc_mark sets the mark bit in @var{obj}, then calls
@code{gc_mark()} on any SCM components of @var{obj}. The last call to
@code{gc_mark()} is tail-called (looped).
@end deftypefun
@deftypefun void mark_locations (STACKITEM @var{x}[], sizet @var{len})
The function @code{mark_locations} is used for marking segments of
C-stack or saved segments of C-stack (marked continuations). The
argument @var{len} is the size of the stack in units of size
@code{(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 @code{gc_mark}. If the stack is word rather than
longword aligned @code{(#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.
@end deftypefun
@node Sweeping the Heap, , Marking Cells, Garbage Collection
@subsubsection 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.
@deftypefun static void gc_sweep ()
The function @code{gc_sweep} scans through all heap segments. The mark
bit is cleared from marked cells. Unmarked cells are spliced into
@var{freelist}, where they can again be returned by invocations of
@code{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
@code{free} procedure is called to free its storage.
@end deftypefun
@node Memory Management for Environments, Signals, Garbage Collection, Operations
@subsection Memory Management for Environments
@ifset html
}} routine at the end of the file
which calls @code{init_iprocs} with the correct type for each of the
@code{iproc}s created in step 5.
@example
void init_@i{foo}()
@{
init_iprocs(subr1s, tc7_subr_1);
init_iprocs(subr3s, tc7_subr_3);
@}
@end example
If your package needs to have a @dfn{finalization} routine called to
free up storage, close files, etc, then also have a line in
@code{init_@i{foo}} like:
@example
add_final(final_@i{foo});
@end example
@code{final_@i{foo}} should be a (void) procedure of no arguments. The
finals will be called in opposite order from their definition.
The line:
@example
add_feature("@i{foo}");
@end example
will append a symbol @code{'@i{foo}} to the (list) value of
@code{slib:features}.
@item
put any scheme code which needs to be run as part of your package into
@file{I@i{foo}.scm}.
@item
put an @code{if} into @file{Init@value{SCMVERSION}.scm} which loads
@file{I@i{foo}.scm} if your package is included:
@example
(if (defined? twiddle-bits!)
(load (in-vicinity (implementation-vicinity)
"I@i{foo}"
(scheme-file-suffix))))
@end example
or use @code{(provided? '@i{foo})} instead of @code{(defined?
twiddle-bits!)} if you have added the feature.
@item
put documentation of the new procedures into @file{@i{foo}.doc}
@item
add lines to your @file{Makefile} to compile and link SCM with your
object file. Add a @code{init_@i{foo}\(\)\;} to the @code{INITS=@dots{}}
line at the beginning of the makefile.
@end enumerate
@noindent
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.
@enumerate
@item
define a new @code{MAKISYM} in @file{scm.h} and increment
@code{NUM_ISYMS}.
@item
add a string with the new name in the corresponding place in
@code{isymnames} in @file{repl.c}.
@item
add @code{case@:} clause to @code{ceval()} near @code{i_quasiquote} (in
@file{eval.c}).
@end enumerate
@noindent
New syntax can now be added without recompiling SCM by the use of the
@code{procedure->syntax}, @code{procedure->macro},
@code{procedure->memoizing-macro}, and @code{defmacro}. For details,
@xref{Syntax}.
@node Defining Subrs, Defining Smobs, Changing Scm, Operations
@subsection Defining Subrs
@noindent
If @dfn{CCLO} is @code{#define}d when compiling, the compiled closure
feature will be enabled. It is automatically enabled if dynamic linking
is enabled.
@noindent
The SCM interpreter directly recognizes subrs taking small numbers of
arguments. In order to create subrs taking larger numbers of arguments
use:
@defun make_gsubr name req opt rest fcn
returns a cclo (compiled closure) object of name @code{char *}
@var{name} which takes @code{int} @var{req} required arguments,
@code{int} @var{opt} optional arguments, and a list of rest arguments if
@code{int} @var{rest} is 1 (0 for not).
@code{SCM (*fcn)()} is a pointer to a C function to do the work.
The C function will always be called with @var{req} + @var{opt} +
@var{rest} arguments, optional arguments not supplied will be passed
@code{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.
@example
/* A silly example, taking 2 required args,
1 optional, and a list of rest args */
#include
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);
@}
@end example
@end defun
@node Defining Smobs, Defining Ptobs, Defining Subrs, Operations
@subsection Defining Smobs
@noindent
Here is an example of how to add a new type named @code{@i{foo}} to SCM.
The following lines need to be added to your code:
@table @code
@item long tc16_@i{foo};
The type code which will be used to identify the new type.
@item static smobfuns @i{foo}smob = @{mark@i{foo},free@i{foo},print@i{foo},equalp@i{foo}@};
smobfuns is a structure composed of 4 functions:
@example
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;
@end example
@table @code
@item smob.mark
is a function of one argument of type @code{SCM} (the cell to mark) and
returns type @code{SCM} which will then be marked. If no further
objects need to be marked then return an immediate object such as
@code{BOOL_F}. The smob cell itself will already have been marked.
@emph{Note@:} This is different from SCM versions prior to 5c5. Only
additional data specific to a smob type need be marked by @code{smob.mark}.
2 functions are provided:
@table @code
@item markcdr(ptr)
returns @code{CDR(ptr)}.
@item mark0(ptr)
is a no-op used for smobs containing no additional @code{SCM} data. 0
may also be used in this case.
@end table
@item smob.free
is a function of one argument of type @code{CELLPTR} (the cell to
collected) and returns type @code{sizet} which is the number of
@code{malloc}ed bytes which were freed. @code{Smob.free} should free
any @code{malloc}ed storage associated with this object. The function
free0(ptr) is provided which does not free any storage and returns 0.
@item smob.print
is 0 or a function of 3 arguments. The first, of type @code{SCM}, is
the smob object. The second, of type @code{SCM}, is the stream on which
to write the result. The third, of type int, is 1 if the object should
be @code{write}n, 0 if it should be @code{display}ed, and 2 if it should
be @code{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).
@item smob.equalp
is 0 or a function of 2 @code{SCM} arguments. Both of these arguments
will be of type @code{tc16@i{foo}}. This function should return
@code{BOOL_T} if the smobs are equal, @code{BOOL_F} if they are not. If
@code{smob.equalp} is 0, @code{equal?} will return @code{BOOL_F} if they
are not @code{eq?}.
@end table
@item tc16_@i{foo} = newsmob(&@i{foo}smob);
Allocates the new type with the functions from @code{@i{foo}smob}. This
line goes in an @code{init_} routine.
@end table
@noindent
Promises and macros in @file{eval.c} and arbiters in @file{repl.c}
provide examples of SMOBs. There are a maximum of 256 SMOBs.
Smobs that must allocate blocks of memory should use, for example,
@code{must_malloc} rather than @code{malloc} @xref{Allocating memory}.
@node Defining Ptobs, Allocating memory, Defining Smobs, Operations
@subsection Defining Ptobs
@noindent
@dfn{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 @code{ptobfuns}:
@example
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;
@end example
@noindent
The @code{.free} component to the structure takes a @code{FILE *} or
other C construct as its argument, unlike @code{.free} in a smob, which
takes the whole smob cell. Often, @code{.free} and @code{.fclose} can be
the same function. See @code{fptob} and @code{pipob} in @file{sys.c}
for examples of how to define ptobs.
Ptobs that must allocate blocks of memory should use, for example,
@code{must_malloc} rather than @code{malloc} @xref{Allocating memory}.
@node Allocating memory, Embedding SCM, Defining Ptobs, Operations
@subsection 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, @code{malloc} and @code{free} should
not be called directly to manage memory freeable by garbage collection.
The following functions are provided for that purpose:
@deftypefun SCM must_malloc_cell (long @var{len}, SCM @var{c}, char *@var{what})
@deftypefunx {char *} must_malloc (long @var{len}, char *@var{what})
@var{len} is the number of bytes that should be allocated, @var{what} is
a string to be used in error or gc messages. @code{must_malloc} returns
a pointer to newly allocated memory. @code{must_malloc_cell} returns a
newly allocated cell whose @code{car} is @var{c} and whose @code{cdr} is
a pointer to newly allocated memory.
@end deftypefun
@deftypefun void must_realloc_cell (SCM @var{z}, long @var{olen}, long @var{len}, char *@var{what})
@deftypefunx {char *} must_realloc (char *@var{where}, long @var{olen}, long @var{len}, char *@var{what})
@code{must_realloc_cell} takes as argument @var{z} a cell whose
@code{cdr} should be a pointer to a block of memory of length @var{olen}
allocated with @code{must_malloc_cell} and modifies the @code{cdr} to point
to a block of memory of length @var{len}. @code{must_realloc} takes as
argument @var{where} the address of a block of memory of length @var{olen}
allocated by @code{must_malloc} and returns the address of a block of
length @var{len}.
The contents of the reallocated block will be unchanged up to the
minimum of the old and new sizes.
@var{what} is a pointer to a string used for error and gc messages.
@end deftypefun
@code{must_malloc}, @code{must_malloc_cell}, @code{must_realloc}, and
@code{must_realloc_cell} must be called with interrupts deferred
@xref{Signals}. @code{must_realloc} and @code{must_realloc_cell} must
not be called during initialization (non-zero errjmp_bad) -- the initial
allocations must be large enough.
@deftypefun void must_free (char *@var{ptr}, sizet @var{len})
@code{must_free} is used to free a block of memory allocated by the
above functions and pointed to by @var{ptr}. @var{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.
@end deftypefun
@node Embedding SCM, Callbacks, Allocating memory, Operations
@subsection Embedding SCM
@cindex Embedding SCM
@noindent
The file @file{scmmain.c} contains the definition of main().
When SCM is compiled as a library @file{scmmain.c} is not included in
the library; a copy of @file{scmmain.c} can be modified to use SCM as an
embedded library module.
@deftypefun int main (int @var{argc}, char **@var{argv})
This is the top level C routine. The value of the @var{argc} argument
is the number of command line arguments. The @var{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:
@code{@var{argv}[@var{argc}]} is this null pointer.
@end deftypefun
@deftypevar char *execpath
This string is the pathname of the executable file being run. This
variable can be examined and set from Scheme (@pxref{Internal State}).
@var{execpath} must be set to executable's path in order to use DUMP
(@pxref{Dump}) or DLD.
@end deftypevar
@noindent
Rename main() and arrange your code to call it with an @var{argv} which
sets up SCM as you want it.
@noindent
If you need more control than is possible through @var{argv}, here are
descriptions of the functions which main() calls.
@deftypefun void init_sbrk (void)
Call this before SCM calls malloc(). Value returned from sbrk() is used
to gauge how much storage SCM uses.
@end deftypefun
@deftypefun {char *} scm_find_execpath (int @var{argc}, char **@var{argv}, char *@var{script_arg})
@var{argc} and @var{argv} are as described in main(). @var{script_arg}
is the pathname of the SCSH-style script (@pxref{Scripting}) being
invoked; 0 otherwise. @code{scm_find_execpath} returns the pathname of
the executable being run; if @code{scm_find_execpath} cannot determine
the pathname, then it returns 0.
@end deftypefun
@noindent
@code{scm_find_implpath} is defined in @file{scmmain.c}. Preceeding
this are definitions of@var{GENERIC_NAME} and @var{INIT_GETENV}. These,
along with @var{IMPLINIT} and @var{dirsep} control scm_find_implpath()'s
operation.
@noindent
If your application has an easier way to locate initialization code for
SCM, then you can replace @code{scm_find_implpath}.
@deftypefun {char *} scm_find_implpath (char *@var{execpath})
Returns the full pathname of the Scheme initialization file or 0 if it
cannot find it.
The string value of the preprocessor variable @var{INIT_GETENV} names an
environment variable (default @samp{"SCM_INIT_PATH"}). If this
environment variable is defined, its value will be returned from
@code{scm_find_implpath}. Otherwise find_impl_file() is called with the
arguments @var{execpath}, @var{GENERIC_NAME} (default "scm"),
@var{INIT_FILE_NAME} (default "Init@value{SCMVERSION}_scm"), and the
directory separator string @var{dirsep}. If find_impl_file() returns 0
and @var{IMPLINIT} is defined, then a copy of the string @var{IMPLINIT}
is returned.
@end deftypefun
@deftypefun int init_buf0 (FILE *@var{inport})
Tries to determine whether @var{inport} (usually stdin) is an
interactive input port which should be used in an unbuffered mode. If
so, @var{inport} is set to unbuffered and non-zero is returned.
Otherwise, 0 is returned.
@code{init_buf0} should be called before any input is read from
@var{inport}. Its value can be used as the last argument to
scm_init_from_argv().
@end deftypefun
@deftypefun void scm_init_from_argv (int @var{argc}, char **@var{argv}, char *@var{script_arg}, int @var{iverbose}, int @var{buf0stdin})
Initializes SCM storage and creates a list of the argument strings
@var{program-arguments} from @var{argv}. @var{argc} and @var{argv} must
already be processed to accomodate Scheme Scripts (if desired). The
scheme variable @var{*script*} is set to the string @var{script_arg}, or
#f if @var{script_arg} is 0.
@var{iverbose} is the initial prolixity level. If @var{buf0stdin} is
non-zero, stdin is treated as an unbuffered port.
@end deftypefun
@noindent
Call @code{init_signals} and @code{restore_signals} only if you want SCM
to handle interrupts and signals.
@deftypefun void init_signals (void)
Initializes handlers for @code{SIGINT} and @code{SIGALRM} if they are
supported by the C implementation. All of the signal handlers
immediately reestablish themselves by a call to @code{signal()}.
@end deftypefun
@deftypefun void restore_signals (void)
Restores the handlers in effect when @code{init_signals} was called.
@end deftypefun
@deftypefun SCM scm_top_level (char *@var{initpath}, SCM (*toplvl_fun)())
This is SCM's top-level. Errors longjmp here. @var{toplvl_fun} is a
callback function of zero arguments that is called by
@code{scm_top_level} to do useful work -- if zero, then @code{repl},
which implements a read-eval-print loop, is called.
If @var{toplvl_fun} returns, then @code{scm_top_level} will return as
well. If the return value of @var{toplvl_fun} is an immediate integer
then it will be used as the return value of @code{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 @var{initpath} is @samp{;}, @samp{(} or
whitespace, then scm_ldstr() is called with @var{initpath} to initialize
SCM; otherwise @var{initpath} names a file of Scheme code to be loaded
to initialize SCM.
When a Scheme error is signaled; control will pass into
@code{scm_top_level} by @code{longjmp}, error messages will be printed
to @code{current-error-port}, and then @var{toplvl_fun} will be called
again. @var{toplvl_fun} must maintain enough state to prevent errors
from being resignalled. If @code{toplvl_fun} can not recover from an
error situation it may simply return.
@end deftypefun
@deftypefun void final_scm (int @var{freeall})
Calls all finalization routines registered with add_final(). If
@var{freeall} is non-zero, then all memory which SCM allocated with
malloc() will be freed.
@end deftypefun
@noindent
You can call indivdual Scheme procedures from C code in the
@var{toplvl_fun} argument passed to scm_top_level(), or from module
subrs (registered by an @code{init_} function, @pxref{Changing Scm}).
@noindent
Use @code{apply} to call Scheme procedures from your C code. For
example:
@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);
@end example
Functions for loading Scheme files and evaluating Scheme code given
as C strings are described in the next section, (@pxref{Callbacks}).
Here is a minimal embedding program @file{libtest.c}:
@example
/* 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);
@}
@print{}
dld_find_executable(./libtest): /home/jaffer/scm/libtest
implpath: /home/jaffer/scm/Init@value{SCMVERSION}.scm
This is libtest_init_user_scm
hello world
@end example
@node Callbacks, Type Conversions, Embedding SCM, Operations
@subsection Callbacks
@cindex callbacks
@cindex rope
@noindent
SCM now has routines to make calling back to Scheme procedures easier.
The source code for these routines are found in @file{rope.c}.
@deftypefun int scm_ldfile (char *@var{file})
Loads the Scheme source file @var{file}. Returns 0 if successful, non-0
if not. This function is used to load SCM's initialization file
@file{Init@value{SCMVERSION}.scm}.
@end deftypefun
@deftypefun int scm_ldprog (char *@var{file})
Loads the Scheme source file @code{(in-vicinity (program-vicinity)
@var{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. @code{program-vicinity} is the
directory from which the calling code was loaded
(@pxref{Vicinity, , , slib, SLIB}).
@end deftypefun
@deftypefun SCM scm_evstr (char *@var{str})
Returns the result of reading an expression from @var{str} and
evaluating it.
@end deftypefun
@deftypefun void scm_ldstr (char *@var{str})
Reads and evaluates all the expressions from @var{str}.
@end deftypefun
@noindent
If you wish to catch errors during execution of Scheme code, then you
can use a wrapper like this for your Scheme procedures:
@example
(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))
@end example
@noindent
Calls to procedures so wrapped will return even if an error occurs.
@node Type Conversions, Continuations, Callbacks, Operations
@subsection Type Conversions
@cindex rope
These type conversion functions are very useful for connecting SCM and C
code. Most are defined in @file{rope.c}.
@deftypefun SCM long2num (long @var{n})
@deftypefunx SCM ulong2num (unsigned long @var{n})
Return an object of type @code{SCM} corresponding to the @code{long} or
@code{unsigned long} argument @var{n}. If @var{n} cannot be converted,
@code{BOOL_F} is returned. Which numbers can be converted depends on
whether SCM was compiled with the @code{BIGDIG} or @code{FLOATS} flags.
To convert integer numbers of smaller types (@code{short} or
@code{char}), use the macro @code{MAKINUM(n)}.
@end deftypefun
@deftypefun long num2long (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
@deftypefunx unsigned long num2ulong (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
@deftypefunx short num2short (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
@deftypefunx unsigned short num2ushort (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
@deftypefunx unsigned char num2uchar (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
@deftypefunx double num2dbl (SCM @var{num}, char *@var{pos}, char *@var{s_caller})
These functions are used to check and convert @code{SCM} arguments to
the named C type. The first argument @var{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 @code{ASRTER} macro calls @code{wta}
with @var{num} and strings @var{pos} and @var{s_caller}. For a listing
of useful predefined @var{pos} macros, @xref{C Macros}.
@emph{Note@:} Inexact numbers are accepted only by @code{num2dbl},
@code{num2long}, and @code{num2ulong} (for when @code{SCM} is compiled
without bignums). To convert inexact numbers to exact numbers,
@xref{Numerical operations, inexact->exact, , r5rs, Revised(5) Scheme}.
@end deftypefun
@deftypefun unsigned long scm_addr (SCM @var{args}, char *@var{s_name})
Returns a pointer (cast to an @code{unsigned long}) to the storage
corresponding to the location accessed by
@code{aref(CAR(args),CDR(args))}. The string @var{s_name} is used in
any messages from error calls by @code{scm_addr}.
@code{scm_addr} is useful for performing C operations on strings or
other uniform arrays (@pxref{Uniform Array}).
@deftypefunx unsigned long scm_base_addr(SCM @var{ra}, char *@var{s_name})
Returns a pointer (cast to an @code{unsigned long}) to the beginning
of storage of array @var{ra}. Note that if @var{ra} is a
shared-array, the strorage accessed this way may be much larger than
@var{ra}.
@emph{Note@:} While you use a pointer returned from @code{scm_addr} or
@code{scm_base_addr} you must keep a pointer to the associated
@code{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. @xref{Changing Scm, scm_gc_protect}.
@end deftypefun
@deftypefun SCM makfrom0str (char *@var{src})
@deftypefunx SCM makfromstr (char *@var{src}, sizet @var{len})
Return a newly allocated string @code{SCM} object copy of the
null-terminated string @var{src} or the string @var{src} of length
@var{len}, respectively.
@end deftypefun
@deftypefun SCM makfromstrs (int @var{argc}, char **@var{argv})
Returns a newly allocated @code{SCM} list of strings corresponding to
the @var{argc} length array of null-terminated strings @var{argv}. If
@var{argv} is less than @code{0}, @var{argv} is assumed to be
@code{NULL} terminated. @code{makfromstrs} is used by
@code{scm_init_from_argv} to convert the arguments SCM was called with
to a @code{SCM} list which is the value of SCM procedure calls to
@code{program-arguments} (@pxref{SCM Session, program-arguments}).
@end deftypefun
@deftypefun {char **} makargvfrmstrs (SCM @var{args}, char *@var{s_name})
Returns a @code{NULL} terminated list of null-terminated strings copied
from the @code{SCM} list of strings @var{args}. The string @var{s_name}
is used in messages from error calls by @code{makargvfrmstrs}.
@code{makargvfrmstrs} is useful for constructing argument lists suitable
for passing to @code{main} functions.
@end deftypefun
@deftypefun void must_free_argv (char **@var{argv})
Frees the storage allocated to create @var{argv} by a call to
@code{makargvfrmstrs}.
@end deftypefun
@node Continuations, Evaluation, Type Conversions, Operations
@subsection Continuations
@cindex continuations
@noindent
The source files @file{continue.h} and @file{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
@ref{Control features, call-with-current-continuation, , r5rs, Revised(5) Scheme}.
@noindent
The C constructs @code{jmp_buf}, @code{setjmp}, and @code{longjmp}
implement escape continuations. On VAX and Cray platforms, the setjmp
provided does not save all the registers. The source files
@file{setjump.mar}, @file{setjump.s}, and @file{ugsetjump.s} provide
implementations which do meet this criteria.
@noindent
SCM uses the names @code{jump_buf}, @code{setjump}, and @code{longjump}
in lieu of @code{jmp_buf}, @code{setjmp}, and @code{longjmp} to prevent
name and declaration conflicts.
@deftp {Data type} CONTINUATION jmpbuf length stkbse other parent
is a @code{typedef}ed structure holding all the information needed to
represent a continuation. The @var{other} slot can be used to hold any
data the user wishes to put there by defining the macro
@code{CONTINUATION_OTHER}.
@end deftp
@defmac SHORT_ALIGN
If @code{SHORT_ALIGN} is @code{#define}d (in @file{scmfig.h}), then the
it is assumed that pointers in the stack can be aligned on @code{short
int} boundaries.
@end defmac
@deftp {Data type} STACKITEM
is a pointer to objects of the size specified by @code{SHORT_ALIGN}
being @code{#define}d or not.
@end deftp
@defmac CHEAP_CONTINUATIONS
If @code{CHEAP_CONTINUATIONS} is @code{#define}d (in @file{scmfig.h})
each @code{CONTINUATION} has size @code{sizeof CONTINUATION}.
Otherwise, all but @dfn{root} @code{CONTINUATION}s have additional
storage (immediately following) to contain a copy of part of the stack.
@emph{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 @code{CHEAP_CONTINUATIONS} in
@file{scmfig.h}.
@end defmac
@defmac STACK_GROWS_UP
Expresses which way the stack grows by its being @code{#define}d or not.
@end defmac
@deftypevar long thrown_value
Gets set to the @var{value} passed to @code{throw_to_continuation}.
@end deftypevar
@deftypefun long stack_size (STACKITEM *@var{start})
Returns the number of units of size @code{STACKITEM} which fit between
@var{start} and the current top of stack. No check is done in this
routine to ensure that @var{start} is actually in the current stack
segment.
@end deftypefun
@deftypefun {CONTINUATION *} make_root_continuation (STACKITEM *@var{stack_base})
Allocates (@code{malloc}) storage for a @code{CONTINUATION} of the
current extent of stack. This newly allocated @code{CONTINUATION} is
returned if successful, @code{0} if not. After
@code{make_root_continuation} returns, the calling routine still needs
to @code{setjump(@var{new_continuation}->jmpbuf)} in order to complete
the capture of this continuation.
@end deftypefun
@deftypefun {CONTINUATION *} make_continuation (CONTINUATION *@var{parent_cont})
Allocates storage for the current @code{CONTINUATION}, copying (or
encapsulating) the stack state from @code{@var{parent_cont}->stkbse} to
the current top of stack. The newly allocated @code{CONTINUATION} is
returned if successful, @code{0}q if not. After
@code{make_continuation} returns, the calling routine still needs to
@code{setjump(@var{new_continuation}->jmpbuf)} in order to complete the
capture of this continuation.
@end deftypefun
@deftypefun void free_continuation (CONTINUATION *@var{cont})
Frees the storage pointed to by @var{cont}. Remember to free storage
pointed to by @code{@var{cont}->other}.
@end deftypefun
@deftypefun void throw_to_continuation (CONTINUATION *@var{cont}, long @var{value}, CONTINUATION *@var{root_cont})
Sets @code{thrown_value} to @var{value} and returns from the
continuation @var{cont}.
If @code{CHEAP_CONTINUATIONS} is @code{#define}d, then
@code{throw_to_continuation} does @code{longjump(@var{cont}->jmpbuf, val)}.
If @code{CHEAP_CONTINUATIONS} is not @code{#define}d, the CONTINUATION
@var{cont} contains a copy of a portion of the C stack (whose bound must
be @code{CONT(@var{root_cont})->stkbse}). Then:
@itemize @bullet
@item
the stack is grown larger than the saved stack, if neccessary.
@item
the saved stack is copied back into it's original position.
@item
@code{longjump(@var{cont}->jmpbuf, val)};
@end itemize
@end deftypefun
@node Evaluation, , Continuations, Operations
@subsection Evaluation
SCM uses its type representations to speed evaluation. All of the
@code{subr} types (@pxref{Subr Cells}) are @code{tc7} types. Since the
@code{tc7} field is in the low order bit position of the @code{CAR} it
can be retrieved and dispatched on quickly by dereferencing the SCM
pointer pointing to it and masking the result.
All the SCM @dfn{Special Forms} get translated to immediate symbols
(@code{isym}) the first time they are encountered by the interpreter
(@code{ceval}). The representation of these immediate symbols is
engineered to occupy the same bits as @code{tc7}. All the @code{isym}s
occur only in the @code{CAR} of lists.
If the @code{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 @code{ILOC} or @code{GLOC}
(@pxref{Immediates}). The codes for @code{ILOC} and @code{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,
@code{ceval} need only retrieve and dispatch on the low order 7 bits of
the @code{CAR} of that cell, regardless of whether that cell is a
closure, header, or subr, or a cons containing @code{ILOC} or
@code{GLOC}.
In order to be able to convert a SCM symbol pointer to an immediate @code{ILOC}
or @code{GLOC}, the evaluator must be holding the pointer to the list in which
that symbol pointer occurs. Turning this requirement to an advantage,
@code{ceval} does not recursively call itself to evaluate symbols in
lists; It instead calls the macro @dfn{EVALCAR}. @code{EVALCAR} does
symbol lookup and memoization for symbols, retrieval of values for @code{ILOC}s
and @code{GLOC}s, returns other immediates, and otherwise recursively calls
itself with the @code{CAR} of the list.
@code{ceval} inlines evaluation (using @code{EVALCAR}) of almost all
procedure call arguments. When @code{ceval} needs to evaluate a list of
more than length 3, the procedure @code{eval_args} is called. So
@code{ceval} can be said to have one level lookahead. The avoidance of
recursive invocations of @code{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 @code{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 @code{evalcar}.
@defvar symhash
Top level symbol values are stored in the @code{symhash} table.
@code{symhash} is an array of lists of @code{ISYM}s and pairs of symbols
and values.
@end defvar
@deftp 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
(@code{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.
@end deftp
@code{ILOC}s work up to a maximum depth of 4096 frames or 4096
identifiers in a frame. Radey Shouman added @dfn{FARLOC}
@tindex FARLOC
to handle cases exceeding these limits. A @code{FARLOC} consists of a
pair whose CAR is the immediate type @code{IM_FARLOC_CAR} or
@code{IM_FARLOC_CDR}, and whose CDR is a pair of INUMs specifying the
frame and distance with a larger range than @code{ILOC}s span.
Adding @code{#define TEST_FARLOC} to @file{eval.c} causes @code{FARLOC}s
to be generated for all local identifiers; this is useful only for
testing memoization.
@deftp 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 @code{GLOC}. The low order bit is normally reserved for
GCmark; But, since references to variables in the code always occur in
the @code{CAR} position and the GCmark is in the @code{CDR}, there is no
conflict.
@end deftp
If the compile FLAG @code{CAUTIOUS} is #defined then the number of
arguments is always checked for application of closures. If the compile
FLAG @code{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
@code{ILOC} or @code{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 @code{ILOC} or @code{GLOC}.
@defmac EVAL expression env
@defmacx SIDEVAL expression env
@code{EVAL} Returns the result of evaluating @var{expression} in
@var{env}. @code{SIDEVAL} evaluates @var{expression} in @var{env} when
the value of the expression is not used.
Both of these macros alter the list structure of @var{expression} as it
is memoized and hence should be used only when it is known that
@var{expression} will not be referenced again. The C function
@code{eval} is safe from this problem.
@end defmac
@deftypefun SCM eval (SCM @var{expression})
Returns the result of evaluating @var{expression} in the top-level
environment. @code{eval} copies @code{expression} so that memoization
does not modify @code{expression}.
@end deftypefun
@node Program Self-Knowledge, Improvements To Make, Operations, The Implementation
@section Program Self-Knowledge
@menu
* File-System Habitat::
* Executable Pathname::
* Script Support::
@end menu
@node File-System Habitat, Executable Pathname, Program Self-Knowledge, Program Self-Knowledge
@subsection File-System Habitat
@noindent
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.
@noindent
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.
@noindent
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.
@quotation
Programs of the world unite! You have nothing to lose but loss itself.
@end quotation
@noindent
The function @code{find_impl_file} in @file{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.
@deftypefun {char *} find_impl_file (char *@var{exec_path}, char *@var{generic_name}, char *@var{initname}, char *@var{sep})
Given the pathname of this executable (@var{exec_path}), test for the
existence of @var{initname} in the implementation-vicinity of this
program. Return a newly allocated string of the path if successful, 0
if not. The @var{sep} argument is a @emph{null-terminated string} of
the character used to separate directory components.
@end deftypefun
@itemize @bullet
@item
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.
@item
Another convention is to install the executables in a directory named
@file{bin}, @file{BIN}, @file{exe}, or @file{EXE} and support files in a
directroy named @file{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
@samp{/usr/local/bin/} and initialization file in
@samp{/usr/local/lib/}.
If the executable directory name matches, the peer directroy @file{lib}
is tested for @var{initname}.
@item
Sometimes @file{lib} directories become too crowded. So we look in any
subdirectories of @file{lib} or @file{src} having the name (sans type
suffix such as @samp{.EXE}) of the program we are running. For example,
the executable might be @samp{/usr/local/bin/foo} and initialization
file in @samp{/usr/local/lib/foo/}.
@item
But the executable name may not be the usual program name; So also look
in any @var{generic_name} subdirectories of @file{lib} or @file{src}
peers.
@item
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 @var{generic_name}
subdirectory. For example, the executable might be
@samp{C:\foo\bar.exe} and the initialization file in @samp{C:\foo\bar\}.
@end itemize
@node Executable Pathname, Script Support, File-System Habitat, Program Self-Knowledge
@subsection Executable Pathname
@noindent
For purposes of finding @file{Init@value{SCMVERSION}.scm}, dumping an
executable, and dynamic linking, a SCM session needs the pathname of its
executable image.
@noindent
When a program is executed by MS-DOS, the full pathname of that
executable is available in @code{argv[0]}. This value can be passed
directly to @code{find_impl_file} (@pxref{File-System Habitat}).
@noindent
In order to find the habitat for a unix program, we first need to know
the full pathname for the associated executable file.
@deftypefun {char *} dld_find_executable (const char *@var{command})
@code{dld_find_executable} returns the absolute path name of the file
that would be executed if @var{command} were given as a command. It
looks up the environment variable @var{PATH}, searches in each of the
directory listed for @var{command}, and returns the absolute path name
for the first occurrence. Thus, it is advisable to invoke
@code{dld_init} as:
@example
main (int argc, const char **argv)
@{
@dots{}
if (dld_init (dld_find_executable (argv[0]))) @{
@dots{}
@}
@dots{}
@}
@end example
@quotation
@strong{Note@@:} If the current process is executed using the
@code{execve} call without passing the correct path name as argument 0,
@code{dld_find_executable (argv[0]) } will also fail to locate the
executable file.
@end quotation
@code{dld_find_executable} returns zero if @code{command} is not found
in any of the directories listed in @code{PATH}.
@end deftypefun
@node Script Support, , Executable Pathname, Program Self-Knowledge
@subsection Script Support
@noindent
Source code for these C functions is in the file @file{script.c}.
@ref{Scripting} for a description of script argument processing.
@noindent
@code{script_find_executable} is only defined on unix systems.
@deftypefun {char *} script_find_executable (const char *@var{name})
@code{script_find_executable} returns the path name of the
executable which is invoked by the script file @var{name};
@var{name} if it is a binary executable (not a script); or 0 if
@var{name} does not exist or is not executable.
@end deftypefun
@deftypefun {char **} script_process_argv (int @var{argc}; char **@var{argv})
Given an @dfn{main} style argument vector @var{argv} and the number of
arguments, @var{argc}, @code{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.
@tindex 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.
@code{script_process_argv} correctly processes argument vectors of
nested script invocations.
@end deftypefun
@deftypefun int script_count_argv (char **@var{argv})
Returns the number of argument strings in @var{argv}.
@end deftypefun
@node Improvements To Make, , Program Self-Knowledge, The Implementation
@section Improvements To Make
@itemize @bullet
@item
Allow users to set limits for @code{malloc()} storage.
@item
Prefix and make more uniform all C function, variable, and constant
names. Provide a file full of #define's to provide backward
compatability.
@item
@code{lgcd()} @emph{needs} to generate at most one bignum, but currently
generates more.
@item
@code{divide()} could use shifts instead of multiply and divide when
scaling.
@item
Currently, @code{dump}ing an executable does not preserve ports. When
loading a @code{dump}ed executable, disk files could be reopened to the
same file and position as they had when the executable was dumped.
@item
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.
@item
In the quest for speed, there has been some discussion about a "Forth"
style Scheme interpreter.
@quotation
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 @code{GLOC} or @code{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 @code{CAR},
@code{CDR}, etc. Since the actual operation code is localized in the
interpreter, it is much easier than uncompilation and then recompilation
to handle @code{(trace car)}; For instance a switch gets set which tells
the interpreter to instead always look up the values of the associated
symbols.
@end quotation
@item
Scott Schwartz suggests: One way to
tidy up the dynamic loading stuff would be to grab the code from perl5.
@end itemize
@menu
* VMS Dynamic Linking:: Finishing the job.
@end menu
@node VMS Dynamic Linking, , Improvements To Make, Improvements To Make
@subsection VMS Dynamic Linking
@noindent
George Carrette (gjc@@mitech.com) outlines how to dynamically link on
VMS. There is already some code in @file{dynl.c} to do this, but
someone with a VMS system needs to finish and debug it.
@enumerate
@item
Say you have this @file{main.c} program:
@format
@t{main()
@{init_lisp();
lisp_repl();@}}
@end format
@item
and you have your lisp in files @file{repl.c}, @file{gc.c},
@code{eval.c} and there are some toplevel non-static variables in use
called @code{the_heap}, @code{the_environment}, and some read-only
toplevel structures, such as @code{the_subr_table}.
@format
@t{$ LINK/SHARE=LISPRTL.EXE/DEBUG REPL.OBJ,GC.OBJ,EVAL.OBJ,LISPRTL.OPT/OPT}
@end format
@item
where @file{LISPRTL.OPT} must contain at least this:
@format
@t{SYS$LIBRARY:VAXCRTL/SHARE
UNIVERSAL=init_lisp
UNIVERSAL=lisp_repl
PSECT_ATTR=the_subr_table,SHR,NOWRT,LCL
PSECT_ATTR=the_heap,NOSHR,LCL
PSECT_ATTR=the_environment,NOSHR,LCL}
@end format
@emph{Notice@:} The @dfn{psect} (Program Section) attributes.
@table @code
@item LCL
means to keep the name local to the shared library. You almost always
want to do that for a good clean library.
@item SHR,NOWRT
means shared-read-only. Which is the default for code, and is also good
for efficiency of some data structures.
@item NOSHR,LCL
is what you want for everything else.
@end table
Note: If you do not have a handy list of all these toplevel variables,
do not dispair. Just do your link with the /MAP=LISPRTL.MAP/FULL
and then search the map file,
@format
@t{$SEARCH/OUT=LISPRTL.LOSERS LISPRTL.MAP ", SHR,NOEXE, RD, WRT"}
@end format
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.
@item
@noindent
Now MAIN.EXE would be linked thusly:
@format
@t{$ DEFINE LISPRTL USER$DISK:[JAFFER]LISPRTL.EXE
$LINK MAIN.OBJ,SYS$INPUT:/OPT
SYS$LIBRARY:VAXCRTL/SHARE
LISPRTL/SHARE}
@end format
Note the definition of the @code{LISPRTL} logical name. Without such a
definition you will need to copy @file{LISPRTL.EXE} over to
@file{SYS$SHARE@:} (aka @file{SYS$LIBRARY@:}) in order to invoke the main
program once it is linked.
@item
Now say you have a file of optional subrs, @file{MYSUBRS.C}. And there
is a routine @code{INIT_MYSUBRS} that must be called before using it.
@format
@t{$ CC MYSUBRS.C
$ LINK/SHARE=MYSUBRS.EXE MYSUBRS.OBJ,SYS$INPUT:/OPT
SYS$LIBRARY:VAXCRTL/SHARE
LISPRTL/SHARE
UNIVERSAL=INIT_MYSUBRS}
@end format
Ok. Another hint is that you can avoid having to add the @code{PSECT}
declaration of @code{NOSHR,LCL} by declaring variables @code{status} in
the C language source. That works great for most things.
@item
Then the dynamic loader would have to do this:
@format
@t{@{void (*init_fcn)();
long retval;
retval = lib$find_image_symbol("MYSUBRS","INIT_MYSUBRS",&init_fcn,
"SYS$DISK:[].EXE");
if (retval != SS$_NORMAL) error(@dots{});
(*init_fcn)();@}}
@end format
But of course all string arguments must be @code{(struct dsc$descriptor
*)} and the last argument is optional if @code{MYSUBRS} is defined as a
logical name or if @file{MYSUBRS.EXE} has been copied over to
@file{SYS$SHARE}. The other consideration is that you will want to turn
off @key{C-c} or other interrupt handling while you are inside most
@code{lib$} calls.
As far as the generation of all the @code{UNIVERSAL=@dots{}}
declarations. Well, you could do well to have that automatically
generated from the public @file{LISPRTL.H} file, of course.
VMS has a good manual called the @cite{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 @code{LISPRTL} you probably don't need that hair.
One fancier option that is useful under VMS for @file{LISPLIB.EXE} is to
define all your exported procedures through an @dfn{call vector} instead
of having them just be pointers into random places in the image, which
is what you get by using @code{UNIVERSAL}.
If you set up the call vector thing correctly it will allow you to
modify and relink @file{LISPLIB.EXE} without having to relink programs
that have been linked against it.
@end enumerate
@ifinfo
@node Index, , The Implementation, Top
@unnumbered Index
@end ifinfo
@include indexes.texi
@bye