1 Guile NEWS --- history of user-visible changes.
2 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send Guile bug reports to bug-guile@gnu.org.
8 Changes in 1.9.0 (changes since the 1.8.x series):
10 * New modules (see the manual for details)
12 ** `(srfi srfi-18)', more sophisticated multithreading support
13 ** `(ice-9 i18n)', internationalization support
14 ** `(rnrs bytevector)', the R6RS bytevector API
15 ** `(rnrs io ports)', a subset of the R6RS I/O port API
16 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
18 * Changes to the stand-alone interpreter
20 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
22 Compiled code loads much faster than Scheme source code, and runs around
23 3 or 4 times as fast, generating much less garbage in the process.
25 ** The stack limit is now initialized from the environment.
27 If getrlimit(2) is available and a stack limit is set, Guile will set
28 its stack limit to 80% of the rlimit. Otherwise the limit is 160000
29 words, a four-fold increase from the earlier default limit.
31 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
32 GUILE_SYSTEM_LOAD_COMPILED_PATH
34 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
35 for source files. It is a different path, however, because compiled
36 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
39 ** New read-eval-print loop (REPL) implementation
41 Running Guile with no arguments drops the user into the new REPL. While
42 it is self-documenting to an extent, the new REPL has not yet been
43 documented in the manual. This will be fixed before 2.0.
45 ** New `guile-tools' commands: `compile', `disassemble'
47 Pass the `--help' command-line option to these commands for more
50 * Changes to Scheme functions and syntax
52 ** Procedure removed: `the-environment'
54 This procedure was part of the interpreter's execution model, and does
55 not apply to the compiler.
57 ** Files loaded with `primitive-load-path' will now be compiled
60 If a compiled .go file corresponding to a .scm file is not found or is
61 not fresh, the .scm file will be compiled on the fly, and the resulting
62 .go file stored away. An advisory note will be printed on the console.
64 Note that this mechanism depends on preservation of the .scm and .go
65 modification times; if the .scm or .go files are moved after
66 installation, care should be taken to preserve their original
69 Autocompiled files will be stored in the user's ~/.guile-ccache
70 directory, which will be created if needed. This is analogous to
71 ccache's behavior for C files.
73 To inhibit autocompilation, set the GUILE_AUTO_COMPILE environment
74 variable to 0, or pass --no-autocompile on the Guile command line.
76 Note that there is currently a bug here: automatic compilation will
77 sometimes be attempted when it shouldn't.
79 For example, the old (lang elisp) modules are meant to be interpreted,
80 not compiled. This bug will be fixed before 2.0. FIXME 2.0: Should say
81 something here about module-transformer called for compile.
83 ** New POSIX procedures: `getrlimit' and `setrlimit'
85 Note however that the interface of these functions is likely to change
86 in the next prerelease.
88 ** New procedure in `(oops goops)': `method-formals'
90 ** BUG: (procedure-property func 'arity) does not work on compiled
93 This will be fixed one way or another before 2.0.
95 ** New procedures in (ice-9 session): `add-value-help-handler!',
96 `remove-value-help-handler!', `add-name-help-handler!'
97 `remove-name-help-handler!', `procedure-arguments',
99 The value and name help handlers provide some minimal extensibility to
100 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
101 example, to make stexinfo help documentation available. See those
102 procedures' docstrings for more information.
104 `procedure-arguments' describes the arguments that a procedure can take,
105 combining arity and formals. For example:
107 (procedure-arguments resolve-interface)
108 => ((required . (name)) (rest . args))
110 Additionally, `module-commentary' is now publically exported from
113 ** Deprecated: `procedure->memoizing-macro', `procedure->syntax'
115 These procedures will not work with syncase expansion, and indeed are
116 not used in the normal course of Guile. They are still used by the old
117 Emacs Lisp support, however.
119 ** New language: ECMAScript
121 Guile now ships with one other high-level language supported,
122 ECMAScript. The goal is to support all of version 3.1 of the standard,
123 but not all of the libraries are there yet. This support is not yet
124 documented; ask on the mailing list if you are interested.
126 ** Defmacros may now have docstrings.
128 Indeed, any macro may have a docstring. `object-documentation' from
129 `(ice-9 documentation)' may be used to retrieve the docstring, once you
130 have a macro value -- but see the above note about first-class macros.
131 Docstrings are associated with the syntax transformer procedures.
133 ** The psyntax expander now knows how to interpret the @ and @@ special
136 ** The psyntax expander is now hygienic with respect to modules.
138 Free variables in a macro are scoped in the module that the macro was
139 defined in, not in the module the macro is used in. For example, code
142 (define-module (foo) #:export (bar))
143 (define (helper x) ...)
145 (syntax-rules () ((_ x) (helper x))))
147 (define-module (baz) #:use-module (foo))
150 It used to be you had to export `helper' from `(foo)' as well.
151 Thankfully, this has been fixed.
153 ** New function, `procedure-module'
155 While useful on its own, `procedure-module' is used by psyntax on syntax
156 transformers to determine the module in which to scope introduced
159 ** `eval-case' has been deprecated, and replaced by `eval-when'.
161 The semantics of `eval-when' are easier to understand. It is still
162 missing documentation, however.
164 ** Guile is now more strict about prohibiting definitions in expression
167 Although previous versions of Guile accepted it, the following
168 expression is not valid, in R5RS or R6RS:
170 (if test (define foo 'bar) (define foo 'baz))
172 In this specific case, it would be better to do:
174 (define foo (if test 'bar 'baz))
176 It is certainly possible to circumvent this resriction with e.g.
177 `(module-define! (current-module) 'foo 'baz)'. We would appreciate
178 feedback about this change (a consequence of using psyntax as the
179 default expander), and may choose to revisit this situation before 2.0
180 in response to user feedback.
182 ** Defmacros must now produce valid Scheme expressions.
184 It used to be that defmacros could unquote in Scheme values, as a way of
185 supporting partial evaluation, and avoiding some hygiene issues. For
188 (define (helper x) ...)
189 (define-macro (foo bar)
192 Assuming this macro is in the `(baz)' module, the direct translation of
195 (define (helper x) ...)
196 (define-macro (foo bar)
197 `((@@ (baz) helper) ,bar))
199 Of course, one could just use a hygienic macro instead:
203 ((_ bar) (helper bar))))
205 ** Guile's psyntax now supports docstrings and internal definitions.
207 The following Scheme is not strictly legal:
214 However its intent is fairly clear. Guile interprets "bar" to be the
215 docstring of `foo', and the definition of `baz' is still in definition
218 ** Macros need to be defined before their first use.
220 It used to be that with lazy memoization, this might work:
224 (define-macro (ref x) x)
227 But now, the body of `foo' is interpreted to mean a call to the toplevel
228 `ref' function, instead of a macro expansion. The solution is to define
229 macros before code that uses them.
231 ** Functions needed by macros at expand-time need to be present at
234 For example, this code will work at the REPL:
236 (define (double-helper x) (* x x))
237 (define-macro (double-literal x) (double-helper x))
238 (double-literal 2) => 4
240 But it will not work when a file is compiled, because the definition of
241 `double-helper' is not present at expand-time. The solution is to wrap
242 the definition of `double-helper' in `eval-when':
244 (eval-when (load compile eval)
245 (define (double-helper x) (* x x)))
246 (define-macro (double-literal x) (double-helper x))
247 (double-literal 2) => 4
249 See the (currently missing) documentation for eval-when for more
252 ** New variable, %pre-modules-transformer
254 Need to document this one some more.
256 ** Temporarily removed functions: `macroexpand', `macroexpand-1'
258 `macroexpand' will be added back before 2.0. It is unclear how to
259 implement `macroexpand-1' with syntax-case, though PLT Scheme does prove
262 ** New reader macros: #' #` #, #,@
264 These macros translate, respectively, to `syntax', `quasisyntax',
265 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
266 These reader macros may be overridden by `read-hash-extend'.
268 ** Incompatible change to #'
270 Guile did have a #' hash-extension, by default, which just returned the
271 subsequent datum: #'foo => foo. In the unlikely event that anyone
272 actually used this, this behavior may be reinstated via the
273 `read-hash-extend' mechanism.
275 ** Scheme expresssions may be commented out with #;
277 #; comments out an entire expression. See SRFI-62 or the R6RS for more
280 ** `make-stack' with a tail-called procedural narrowing argument no longer
281 works (with compiled procedures)
283 It used to be the case that a captured stack could be narrowed to select
284 calls only up to or from a certain procedure, even if that procedure
285 already tail-called another procedure. This was because the debug
286 information from the original procedure was kept on the stack.
288 Now with the new compiler, the stack only contains active frames from
289 the current continuation. A narrow to a procedure that is not in the
290 stack will result in an empty stack. To fix this, narrow to a procedure
291 that is active in the current continuation, or narrow to a specific
292 number of stack frames.
294 ** backtraces through compiled procedures only show procedures that are
295 active in the current continuation
297 Similarly to the previous issue, backtraces in compiled code may be
298 different from backtraces in interpreted code. There are no semantic
299 differences, however. Please mail bug-guile@gnu.org if you see any
300 deficiencies with Guile's backtraces.
302 ** syntax-rules and syntax-case macros now propagate source information
303 through to the expanded code
305 This should result in better backtraces.
307 ** The currying behavior of `define' has been removed.
309 Before, `(define ((f a) b) (* a b))' would translate to
311 (define f (lambda (a) (lambda (b) (* a b))))
313 Now a syntax error is signaled, as this syntax is not supported by
314 default. If there is sufficient demand, this syntax can be supported
317 ** All modules have names now
319 Before, you could have anonymous modules: modules without names. Now,
320 because of hygiene and macros, all modules have names. If a module was
321 created without a name, the first time `module-name' is called on it, a
322 fresh name will be lazily generated for it.
324 ** Many syntax errors have different texts now
326 Syntax errors still throw to the `syntax-error' key, but the arguments
327 are often different now. Perhaps in the future, Guile will switch to
328 using standard SRFI-35 conditions.
330 ** Returning multiple values to compiled code will silently truncate the
331 values to the expected number
333 For example, the interpreter would raise an error evaluating the form,
334 `(+ (values 1 2) (values 3 4))', because it would see the operands as
335 being two compound "values" objects, to which `+' does not apply.
337 The compiler, on the other hand, receives multiple values on the stack,
338 not as a compound object. Given that it must check the number of values
339 anyway, if too many values are provided for a continuation, it chooses
340 to truncate those values, effectively evaluating `(+ 1 3)' instead.
342 The idea is that the semantics that the compiler implements is more
343 intuitive, and the use of the interpreter will fade out with time.
344 This behavior is allowed both by the R5RS and the R6RS.
346 ** Multiple values in compiled code are not represented by compound
349 This change may manifest itself in the following situation:
351 (let ((val (foo))) (do-something) val)
353 In the interpreter, if `foo' returns multiple values, multiple values
354 are produced from the `let' expression. In the compiler, those values
355 are truncated to the first value, and that first value is returned. In
356 the compiler, if `foo' returns no values, an error will be raised, while
357 the interpreter would proceed.
359 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
360 behavior is more correct, however. If you wish to preserve a potentially
361 multiply-valued return, you will need to set up a multiple-value
362 continuation, using `call-with-values'.
364 ** Defmacros are now implemented in terms of syntax-case.
366 The practical ramification of this is that the `defmacro?' predicate has
367 been removed, along with `defmacro-transformer', `macro-table',
368 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
369 `defmacro:transformer'. This is because defmacros are simply macros. If
370 any of these procedures provided useful facilities to you, we encourage
371 you to contact the Guile developers.
373 ** psyntax is now the default expander
375 Scheme code is now expanded by default by the psyntax hygienic macro
376 expander. Expansion is performed completely before compilation or
379 Notably, syntax errors will be signalled before interpretation begins.
380 In the past, many syntax errors were only detected at runtime if the
381 code in question was memoized.
383 As part of its expansion, psyntax renames all lexically-bound
384 identifiers. Original identifier names are preserved and given to the
385 compiler, but the interpreter will see the renamed variables, e.g.,
386 `x432' instead of `x'.
388 Note that the psyntax that Guile uses is a fork, as Guile already had
389 modules before incompatible modules were added to psyntax -- about 10
390 years ago! Thus there are surely a number of bugs that have been fixed
391 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
393 ** syntax-rules and syntax-case are available by default.
395 There is no longer any need to import the `(ice-9 syncase)' module
396 (which is now deprecated). The expander may be invoked directly via
397 `sc-expand', though it is normally searched for via the current module
400 Also, the helper routines for syntax-case are available in the default
401 environment as well: `syntax->datum', `datum->syntax',
402 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
403 `identifier?', and `syntax-violation'. See the R6RS for documentation.
405 ** Lexical bindings introduced by hygienic macros may not be referenced
406 by nonhygienic macros.
408 If a lexical binding is introduced by a hygienic macro, it may not be
409 referenced by a nonhygienic macro. For example, this works:
412 (define-macro (bind-x val body)
413 `(let ((x ,val)) ,body))
414 (define-macro (ref x)
421 (define-syntax bind-x
423 ((_ val body) (let ((x val)) body))))
424 (define-macro (ref x)
428 It is not normal to run into this situation with existing code. However,
429 as code is ported over from defmacros to syntax-case, it is possible to
430 run into situations like this. In the future, Guile will probably port
431 its `while' macro to syntax-case, which makes this issue one to know
434 ** Macros may no longer be referenced as first-class values.
436 In the past, you could evaluate e.g. `if', and get its macro value. Now,
437 expanding this form raises a syntax error.
439 Macros still /exist/ as first-class values, but they must be
440 /referenced/ via the module system, e.g. `(module-ref (current-module)
443 This decision may be revisited before the 2.0 release. Feedback welcome
444 to guile-devel@gnu.org (subscription required) or bug-guile@gnu.org (no
445 subscription required).
447 ** New macro type: syncase-macro
449 XXX Need to decide whether to document this for 2.0, probably should:
450 make-syncase-macro, make-extended-syncase-macro, macro-type,
451 syncase-macro-type, syncase-macro-binding
453 ** A new `memoize-symbol' evaluator trap has been added.
455 This trap can be used for efficiently implementing a Scheme code
458 ** Duplicate bindings among used modules are resolved lazily.
460 This slightly improves program startup times.
462 ** New thread cancellation and thread cleanup API
464 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
466 ** Fix bad interaction between `false-if-exception' and stack-call.
468 Exceptions thrown by `false-if-exception' were erronously causing the
469 stack to be saved, causing later errors to show the incorrectly-saved
470 backtrace. This has been fixed.
472 ** New global variables: %load-compiled-path, %load-compiled-extensions
474 These are analogous to %load-path and %load-extensions.
476 ** New procedure, `make-promise'
478 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
480 ** New entry into %guile-build-info: `ccachedir'
482 ** Fix bug in `module-bound?'.
484 `module-bound?' was returning true if a module did have a local
485 variable, but one that was unbound, but another imported module bound
486 the variable. This was an error, and was fixed.
488 ** `(ice-9 syncase)' has been deprecated.
490 As syntax-case is available by default, importing `(ice-9 syncase)' has
491 no effect, and will trigger a deprecation warning.
493 * Changes to the C interface
495 ** The GH interface (deprecated in version 1.6, 2001) was removed.
497 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
499 This makes these internal functions technically not callable from
502 ** Functions for handling `scm_option' now no longer require an argument
503 indicating length of the `scm_t_option' array.
505 ** scm_primitive_load_path has additional argument, exception_on_error
507 ** New C function: scm_module_public_interface
509 This procedure corresponds to Scheme's `module-public-interface'.
511 ** `scm_stat' has an additional argument, `exception_on_error'
512 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
514 * Changes to the distribution
516 ** Guile's license is now LGPLv3+
518 In other words the GNU Lesser General Public License, version 3 or
519 later (at the discretion of each person that chooses to redistribute
522 ** `guile-config' will be deprecated in favor of `pkg-config'
524 `guile-config' has been rewritten to get its information from
525 `pkg-config', so this should be a transparent change. Note however that
526 guile.m4 has yet to be modified to call pkg-config instead of
529 ** New installation directory: $(pkglibdir)/1.9/ccache
531 If $(libdir) is /usr/lib, for example, Guile will install its .go files
532 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
534 ** New dependency: GNU libunistring.
536 See http://www.gnu.org/software/libunistring/. We hope to merge in
537 Unicode support in the next prerelease.
541 Changes in 1.8.7 (since 1.8.6)
543 * New modules (see the manual for details)
545 ** `(srfi srfi-98)', an interface to access environment variables
549 ** Fix compilation with `--disable-deprecated'
550 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
551 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
552 ** Fix build problem when scm_t_timespec is different from struct timespec
553 ** Fix build when compiled with -Wundef -Werror
554 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
555 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
556 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
557 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
559 ** Allow @ macro to work with (ice-9 syncase)
561 Previously, use of the @ macro in a module whose code is being
562 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
563 Now it works as you would expect (giving the value of the specified
566 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
569 Changes in 1.8.6 (since 1.8.5)
571 * New features (see the manual for details)
573 ** New convenience function `scm_c_symbol_length ()'
575 ** Single stepping through code from Emacs
577 When you use GDS to evaluate Scheme code from Emacs, you can now use
578 `C-u' to indicate that you want to single step through that code. See
579 `Evaluating Scheme Code' in the manual for more details.
581 ** New "guile(1)" man page!
583 * Changes to the distribution
585 ** Automake's `AM_MAINTAINER_MODE' is no longer used
587 Thus, the `--enable-maintainer-mode' configure option is no longer
588 available: Guile is now always configured in "maintainer mode".
590 ** `ChangeLog' files are no longer updated
592 Instead, changes are detailed in the version control system's logs. See
593 the top-level `ChangeLog' files for details.
598 ** `symbol->string' now returns a read-only string, as per R5RS
599 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
600 ** `guile-config link' now prints `-L$libdir' before `-lguile'
601 ** Fix memory corruption involving GOOPS' `class-redefinition'
602 ** Fix possible deadlock in `mutex-lock'
603 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
604 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
605 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
606 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
607 ** Fix misleading output from `(help rationalize)'
608 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
609 ** Fix `gcd' when called with a single, negative argument.
610 ** Fix `Stack overflow' errors seen when building on some platforms
611 ** Fix bug when `scm_with_guile ()' was called several times from the
613 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
614 dynamic environment of the call to `raise'
615 ** Fix potential deadlock in `make-struct'
616 ** Fix compilation problem with libltdl from Libtool 2.2.x
617 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
620 Changes in 1.8.5 (since 1.8.4)
622 * Infrastructure changes
624 ** Guile repository switched from CVS to Git
626 The new repository can be accessed using
627 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
628 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
630 ** Add support for `pkg-config'
632 See "Autoconf Support" in the manual for details.
634 * New modules (see the manual for details)
638 * New features (see the manual for details)
640 ** New `postfix' read option, for SRFI-88 keyword syntax
641 ** Some I/O primitives have been inlined, which improves I/O performance
642 ** New object-based traps infrastructure
644 This is a GOOPS-based infrastructure that builds on Guile's low-level
645 evaluator trap calls and facilitates the development of debugging
646 features like single-stepping, breakpoints, tracing and profiling.
647 See the `Traps' node of the manual for details.
649 ** New support for working on Guile code from within Emacs
651 Guile now incorporates the `GDS' library (previously distributed
652 separately) for working on Guile code from within Emacs. See the
653 `Using Guile In Emacs' node of the manual for details.
657 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
658 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
660 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
661 would trigger an unbound variable error for `match:andmap'.
663 ** `(oop goops describe)' now properly provides the `describe' feature
664 ** Fixed `args-fold' from `(srfi srfi-37)'
666 Previously, parsing short option names of argument-less options would
667 lead to a stack overflow.
669 ** `(srfi srfi-35)' is now visible through `cond-expand'
670 ** Fixed type-checking for the second argument of `eval'
671 ** Fixed type-checking for SRFI-1 `partition'
672 ** Fixed `struct-ref' and `struct-set!' on "light structs"
673 ** Honor struct field access rights in GOOPS
674 ** Changed the storage strategy of source properties, which fixes a deadlock
675 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
676 ** Fixed build issue for GNU/Linux on IA64
677 ** Fixed build issues on NetBSD 1.6
678 ** Fixed build issue on Solaris 2.10 x86_64
679 ** Fixed build issue with DEC/Compaq/HP's compiler
680 ** Fixed `scm_from_complex_double' build issue on FreeBSD
681 ** Fixed `alloca' build issue on FreeBSD 6
682 ** Removed use of non-portable makefile constructs
683 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
684 ** Make sure all tests honor `$TMPDIR'
686 * Changes to the distribution
690 We've started collecting Frequently Asked Questions (FAQ), and will
691 distribute these (with answers!) in future Guile releases.
694 Changes in 1.8.4 (since 1.8.3)
698 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
699 ** Fixed a segmentation fault which occurred when displaying the
700 backtrace of a stack with a promise object (made by `delay') in it.
701 ** Make `accept' leave guile mode while blocking
702 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
703 ** Fixed a build problem on AIX (use of func_data identifier)
704 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
705 called with an associator proc that returns neither a pair nor #f.
706 ** Secondary threads now always return a valid module for (current-module).
707 ** Avoid MacOS build problems caused by incorrect combination of "64"
708 system and library calls.
709 ** `guile-snarf' now honors `$TMPDIR'
710 ** `guile-config compile' now reports CPPFLAGS used at compile-time
711 ** Fixed build with Sun Studio (Solaris 9)
712 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
713 uniform vectors on AIX.
714 ** Fixed a deadlock that occurs upon GC with multiple threads.
715 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
716 ** Fixed autotool-derived build problems on AIX 6.1.
717 ** Fixed NetBSD/alpha support
718 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
720 * New modules (see the manual for details)
724 * Documentation fixes and improvements
726 ** Removed premature breakpoint documentation
728 The features described are not available in the series of 1.8.x
729 releases, so the documentation was misleading and has been removed.
731 ** More about Guile's default *random-state* variable
733 ** GOOPS: more about how to use `next-method'
735 * Changes to the distribution
737 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
739 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
740 General Public License, and the few incorrect files have now been
741 fixed to agree with the rest of the Guile distribution.
743 ** Removed unnecessary extra copies of COPYING*
745 The distribution now contains a single COPYING.LESSER at its top level.
748 Changes in 1.8.3 (since 1.8.2)
750 * New modules (see the manual for details)
757 ** The `(ice-9 slib)' module now works as expected
758 ** Expressions like "(set! 'x #t)" no longer yield a crash
759 ** Warnings about duplicate bindings now go to stderr
760 ** A memory leak in `make-socket-address' was fixed
761 ** Alignment issues (e.g., on SPARC) in network routines were fixed
762 ** A threading issue that showed up at least on NetBSD was fixed
763 ** Build problems on Solaris and IRIX fixed
765 * Implementation improvements
767 ** The reader is now faster, which reduces startup time
768 ** Procedures returned by `record-accessor' and `record-modifier' are faster
771 Changes in 1.8.2 (since 1.8.1):
773 * New procedures (see the manual for details)
775 ** set-program-arguments
778 * Incompatible changes
780 ** The body of a top-level `define' no longer sees the binding being created
782 In a top-level `define', the binding being created is no longer visible
783 from the `define' body. This breaks code like
784 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
785 unbound in the body. However, such code was not R5RS-compliant anyway,
790 ** Fractions were not `equal?' if stored in unreduced form.
791 (A subtle problem, since printing a value reduced it, making it work.)
792 ** srfi-60 `copy-bit' failed on 64-bit systems
793 ** "guile --use-srfi" option at the REPL can replace core functions
794 (Programs run with that option were ok, but in the interactive REPL
795 the core bindings got priority, preventing SRFI replacements or
797 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
798 ** `kill' on mingw throws an error for a PID other than oneself
799 ** Procedure names are attached to procedure-with-setters
800 ** Array read syntax works with negative lower bound
801 ** `array-in-bounds?' fix if an array has different lower bounds on each index
802 ** `*' returns exact 0 for "(* inexact 0)"
803 This follows what it always did for "(* 0 inexact)".
804 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
805 ** SRFI-19: `date->julian-day' did not account for timezone offset
806 ** `ttyname' no longer crashes when passed a non-tty argument
807 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
808 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
809 ** GOOPS: Fixed a bug in `method-more-specific?'
810 ** Build problems on Solaris fixed
811 ** Build problems on HP-UX IA64 fixed
812 ** Build problems on MinGW fixed
815 Changes in 1.8.1 (since 1.8.0):
817 * LFS functions are now used to access 64-bit files on 32-bit systems.
819 * New procedures (see the manual for details)
821 ** primitive-_exit - [Scheme] the-root-module
822 ** scm_primitive__exit - [C]
823 ** make-completion-function - [Scheme] (ice-9 readline)
824 ** scm_c_locale_stringn_to_number - [C]
825 ** scm_srfi1_append_reverse [C]
826 ** scm_srfi1_append_reverse_x [C]
834 ** Build problems have been fixed on MacOS, SunOS, and QNX.
836 ** `strftime' fix sign of %z timezone offset.
838 ** A one-dimensional array can now be 'equal?' to a vector.
840 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
842 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
844 ** `record-accessor' and `record-modifier' now have strict type checks.
846 Record accessor and modifier procedures now throw an error if the
847 record type of the record they're given is not the type expected.
848 (Previously accessors returned #f and modifiers silently did nothing).
850 ** It is now OK to use both autoload and use-modules on a given module.
852 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
854 Previously there was no checking on primatives like make-vector that
855 accept "one or two" arguments. Now there is.
857 ** The srfi-1 assoc function now calls its equality predicate properly.
859 Previously srfi-1 assoc would call the equality predicate with the key
860 last. According to the SRFI, the key should be first.
862 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
864 ** The array-set! procedure no longer segfaults when given a bit vector.
866 ** Bugs in make-shared-array have been fixed.
868 ** string<? and friends now follow char<? etc order on 8-bit chars.
870 ** The format procedure now handles inf and nan values for ~f correctly.
872 ** exact->inexact should no longer overflow when given certain large fractions.
874 ** srfi-9 accessor and modifier procedures now have strict record type checks.
876 This matches the srfi-9 specification.
878 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
880 Previously the (ice-9 ftw) procedures would ignore any file that had
881 the same inode number as a file they had already seen, even if that
882 file was on a different device.
885 Changes in 1.8.0 (changes since the 1.6.x series):
887 * Changes to the distribution
889 ** Guile is now licensed with the GNU Lesser General Public License.
891 ** The manual is now licensed with the GNU Free Documentation License.
893 ** Guile now requires GNU MP (http://swox.com/gmp).
895 Guile now uses the GNU MP library for arbitrary precision arithmetic.
897 ** Guile now has separate private and public configuration headers.
899 That is, things like HAVE_STRING_H no longer leak from Guile's
902 ** Guile now provides and uses an "effective" version number.
904 Guile now provides scm_effective_version and effective-version
905 functions which return the "effective" version number. This is just
906 the normal full version string without the final micro-version number,
907 so the current effective-version is "1.8". The effective version
908 should remain unchanged during a stable series, and should be used for
909 items like the versioned share directory name
910 i.e. /usr/share/guile/1.8.
912 Providing an unchanging version number during a stable release for
913 things like the versioned share directory can be particularly
914 important for Guile "add-on" packages, since it provides a directory
915 that they can install to that won't be changed out from under them
916 with each micro release during a stable series.
918 ** Thread implementation has changed.
920 When you configure "--with-threads=null", you will get the usual
921 threading API (call-with-new-thread, make-mutex, etc), but you can't
922 actually create new threads. Also, "--with-threads=no" is now
923 equivalent to "--with-threads=null". This means that the thread API
924 is always present, although you might not be able to create new
927 When you configure "--with-threads=pthreads" or "--with-threads=yes",
928 you will get threads that are implemented with the portable POSIX
929 threads. These threads can run concurrently (unlike the previous
930 "coop" thread implementation), but need to cooperate for things like
933 The default is "pthreads", unless your platform doesn't have pthreads,
934 in which case "null" threads are used.
936 See the manual for details, nodes "Initialization", "Multi-Threading",
937 "Blocking", and others.
939 ** There is the new notion of 'discouraged' features.
941 This is a milder form of deprecation.
943 Things that are discouraged should not be used in new code, but it is
944 OK to leave them in old code for now. When a discouraged feature is
945 used, no warning message is printed like there is for 'deprecated'
946 features. Also, things that are merely discouraged are nevertheless
947 implemented efficiently, while deprecated features can be very slow.
949 You can omit discouraged features from libguile by configuring it with
950 the '--disable-discouraged' option.
952 ** Deprecation warnings can be controlled at run-time.
954 (debug-enable 'warn-deprecated) switches them on and (debug-disable
955 'warn-deprecated) switches them off.
957 ** Support for SRFI 61, extended cond syntax for multiple values has
960 This SRFI is always available.
962 ** Support for require-extension, SRFI-55, has been added.
964 The SRFI-55 special form `require-extension' has been added. It is
965 available at startup, and provides a portable way to load Scheme
966 extensions. SRFI-55 only requires support for one type of extension,
967 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
970 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
972 The (srfi srfi-26) module is an implementation of SRFI-26 which
973 provides the `cut' and `cute' syntax. These may be used to specialize
974 parameters without currying.
976 ** New module (srfi srfi-31)
978 This is an implementation of SRFI-31 which provides a special form
979 `rec' for recursive evaluation.
981 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
982 been merged with the core, making their functionality always
985 The modules are still available, tho, and you could use them together
986 with a renaming import, for example.
988 ** Guile no longer includes its own version of libltdl.
990 The official version is good enough now.
992 ** The --enable-htmldoc option has been removed from 'configure'.
994 Support for translating the documentation into HTML is now always
995 provided. Use 'make html'.
997 ** New module (ice-9 serialize):
999 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
1000 don't trust the thread safety of most of your program, but where you
1001 have some section(s) of code which you consider can run in parallel to
1002 other sections. See ice-9/serialize.scm for more information.
1004 ** The configure option '--disable-arrays' has been removed.
1006 Support for arrays and uniform numeric arrays is now always included
1009 * Changes to the stand-alone interpreter
1011 ** New command line option `-L'.
1013 This option adds a directory to the front of the load path.
1015 ** New command line option `--no-debug'.
1017 Specifying `--no-debug' on the command line will keep the debugging
1018 evaluator turned off, even for interactive sessions.
1020 ** User-init file ~/.guile is now loaded with the debugging evaluator.
1022 Previously, the normal evaluator would have been used. Using the
1023 debugging evaluator gives better error messages.
1025 ** The '-e' option now 'read's its argument.
1027 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
1028 be used with '-e'. For example, you can now write a script like
1031 exec guile -e '(@ (demo) main)' -s "$0" "$@"
1034 (define-module (demo)
1038 (format #t "Demo: ~a~%" args))
1041 * Changes to Scheme functions and syntax
1043 ** Guardians have changed back to their original semantics
1045 Guardians now behave like described in the paper by Dybvig et al. In
1046 particular, they no longer make guarantees about the order in which
1047 they return objects, and they can no longer be greedy.
1049 They no longer drop cyclic data structures.
1051 The C function scm_make_guardian has been changed incompatibly and no
1052 longer takes the 'greedy_p' argument.
1054 ** New function hashx-remove!
1056 This function completes the set of 'hashx' functions.
1058 ** The concept of dynamic roots has been factored into continuation
1059 barriers and dynamic states.
1061 Each thread has a current dynamic state that carries the values of the
1062 fluids. You can create and copy dynamic states and use them as the
1063 second argument for 'eval'. See "Fluids and Dynamic States" in the
1066 To restrict the influence that captured continuations can have on the
1067 control flow, you can errect continuation barriers. See "Continuation
1068 Barriers" in the manual.
1070 The function call-with-dynamic-root now essentially temporarily
1071 installs a new dynamic state and errects a continuation barrier.
1073 ** The default load path no longer includes "." at the end.
1075 Automatically loading modules from the current directory should not
1076 happen by default. If you want to allow it in a more controlled
1077 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
1078 variable %load-path.
1080 ** The uniform vector and array support has been overhauled.
1082 It now complies with SRFI-4 and the weird prototype based uniform
1083 array creation has been deprecated. See the manual for more details.
1085 Some non-compatible changes have been made:
1086 - characters can no longer be stored into byte arrays.
1087 - strings and bit vectors are no longer considered to be uniform numeric
1089 - array-rank throws an error for non-arrays instead of returning zero.
1090 - array-ref does no longer accept non-arrays when no indices are given.
1092 There is the new notion of 'generalized vectors' and corresponding
1093 procedures like 'generalized-vector-ref'. Generalized vectors include
1094 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
1096 Arrays use generalized vectors as their storage, so that you still
1097 have arrays of characters, bits, etc. However, uniform-array-read!
1098 and uniform-array-write can no longer read/write strings and
1101 ** There is now support for copy-on-write substrings, mutation-sharing
1102 substrings and read-only strings.
1104 Three new procedures are related to this: substring/shared,
1105 substring/copy, and substring/read-only. See the manual for more
1108 ** Backtraces will now highlight the value that caused the error.
1110 By default, these values are enclosed in "{...}", such as in this
1119 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
1120 <unnamed port>:1:1: Wrong type (expecting pair): a
1121 ABORT: (wrong-type-arg)
1123 The prefix and suffix used for highlighting can be set via the two new
1124 printer options 'highlight-prefix' and 'highlight-suffix'. For
1125 example, putting this into ~/.guile will output the bad value in bold
1126 on an ANSI terminal:
1128 (print-set! highlight-prefix "\x1b[1m")
1129 (print-set! highlight-suffix "\x1b[22m")
1132 ** 'gettext' support for internationalization has been added.
1134 See the manual for details.
1136 ** New syntax '@' and '@@':
1138 You can now directly refer to variables exported from a module by
1141 (@ MODULE-NAME VARIABLE-NAME)
1143 For example (@ (ice-9 pretty-print) pretty-print) will directly access
1144 the pretty-print variable exported from the (ice-9 pretty-print)
1145 module. You don't need to 'use' that module first. You can also use
1146 '@' as a target of 'set!', as in (set! (@ mod var) val).
1148 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
1149 but it can also access variables that have not been exported. It is
1150 intended only for kluges and temporary fixes and for debugging, not
1153 ** Keyword syntax has been made more disciplined.
1155 Previously, the name of a keyword was read as a 'token' but printed as
1156 a symbol. Now, it is read as a general Scheme datum which must be a
1167 ERROR: In expression (a b c):
1171 ERROR: Unbound variable: foo
1176 ERROR: Wrong type (expecting symbol): 12
1180 ERROR: Wrong type (expecting symbol): (a b c)
1184 ** The printing of symbols that might look like keywords can be
1187 The new printer option 'quote-keywordish-symbols' controls how symbols
1188 are printed that have a colon as their first or last character. The
1189 default now is to only quote a symbol with #{...}# when the read
1190 option 'keywords' is not '#f'. Thus:
1192 guile> (define foo (string->symbol ":foo"))
1193 guile> (read-set! keywords #f)
1196 guile> (read-set! keywords 'prefix)
1199 guile> (print-set! quote-keywordish-symbols #f)
1203 ** 'while' now provides 'break' and 'continue'
1205 break and continue were previously bound in a while loop, but not
1206 documented, and continue didn't quite work properly. The undocumented
1207 parameter to break which gave a return value for the while has been
1210 ** 'call-with-current-continuation' is now also available under the name
1213 ** The module system now checks for duplicate bindings.
1215 The module system now can check for name conflicts among imported
1218 The behavior can be controlled by specifying one or more 'duplicates'
1219 handlers. For example, to make Guile return an error for every name
1222 (define-module (foo)
1227 The new default behavior of the module system when a name collision
1228 has been detected is to
1230 1. Give priority to bindings marked as a replacement.
1231 2. Issue a warning (different warning if overriding core binding).
1232 3. Give priority to the last encountered binding (this corresponds to
1235 If you want the old behavior back without replacements or warnings you
1238 (default-duplicate-binding-handler 'last)
1240 to your .guile init file.
1242 ** New define-module option: :replace
1244 :replace works as :export, but, in addition, marks the binding as a
1247 A typical example is `format' in (ice-9 format) which is a replacement
1248 for the core binding `format'.
1250 ** Adding prefixes to imported bindings in the module system
1252 There is now a new :use-module option :prefix. It can be used to add
1253 a prefix to all imported bindings.
1255 (define-module (foo)
1256 :use-module ((bar) :prefix bar:))
1258 will import all bindings exported from bar, but rename them by adding
1261 ** Conflicting generic functions can be automatically merged.
1263 When two imported bindings conflict and they are both generic
1264 functions, the two functions can now be merged automatically. This is
1265 activated with the 'duplicates' handler 'merge-generics'.
1267 ** New function: effective-version
1269 Returns the "effective" version number. This is just the normal full
1270 version string without the final micro-version number. See "Changes
1271 to the distribution" above.
1273 ** New threading functions: parallel, letpar, par-map, and friends
1275 These are convenient ways to run calculations in parallel in new
1276 threads. See "Parallel forms" in the manual for details.
1278 ** New function 'try-mutex'.
1280 This function will attempt to lock a mutex but will return immediately
1281 instead of blocking and indicate failure.
1283 ** Waiting on a condition variable can have a timeout.
1285 The function 'wait-condition-variable' now takes a third, optional
1286 argument that specifies the point in time where the waiting should be
1289 ** New function 'broadcast-condition-variable'.
1291 ** New functions 'all-threads' and 'current-thread'.
1293 ** Signals and system asyncs work better with threads.
1295 The function 'sigaction' now takes a fourth, optional, argument that
1296 specifies the thread that the handler should run in. When the
1297 argument is omitted, the handler will run in the thread that called
1300 Likewise, 'system-async-mark' takes a second, optional, argument that
1301 specifies the thread that the async should run in. When it is
1302 omitted, the async will run in the thread that called
1303 'system-async-mark'.
1305 C code can use the new functions scm_sigaction_for_thread and
1306 scm_system_async_mark_for_thread to pass the new thread argument.
1308 When a thread blocks on a mutex, a condition variable or is waiting
1309 for IO to be possible, it will still execute system asyncs. This can
1310 be used to interrupt such a thread by making it execute a 'throw', for
1313 ** The function 'system-async' is deprecated.
1315 You can now pass any zero-argument procedure to 'system-async-mark'.
1316 The function 'system-async' will just return its argument unchanged
1319 ** New functions 'call-with-blocked-asyncs' and
1320 'call-with-unblocked-asyncs'
1322 The expression (call-with-blocked-asyncs PROC) will call PROC and will
1323 block execution of system asyncs for the current thread by one level
1324 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
1325 procedure and will unblock the execution of system asyncs by one
1326 level for the current thread.
1328 Only system asyncs are affected by these functions.
1330 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
1332 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
1333 instead. Those functions are easier to use correctly and can be
1336 ** New function 'unsetenv'.
1338 ** New macro 'define-syntax-public'.
1340 It works like 'define-syntax' and also exports the defined macro (but
1343 ** There is support for Infinity and NaNs.
1345 Following PLT Scheme, Guile can now work with infinite numbers, and
1348 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
1349 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
1350 "+nan.0"). These numbers are inexact and have no exact counterpart.
1352 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
1353 sign of the dividend. The infinities are integers, and they answer #t
1354 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
1355 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
1366 ERROR: Numerical overflow
1368 Two new predicates 'inf?' and 'nan?' can be used to test for the
1371 ** Inexact zero can have a sign.
1373 Guile can now distinguish between plus and minus inexact zero, if your
1374 platform supports this, too. The two zeros are equal according to
1375 '=', but not according to 'eqv?'. For example
1386 ** Guile now has exact rationals.
1388 Guile can now represent fractions such as 1/3 exactly. Computing with
1389 them is also done exactly, of course:
1394 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
1395 for exact arguments.
1397 For example: (floor 2) now returns an exact 2 where in the past it
1398 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
1400 ** inexact->exact no longer returns only integers.
1402 Without exact rationals, the closest exact number was always an
1403 integer, but now inexact->exact returns the fraction that is exactly
1404 equal to a floating point number. For example:
1406 (inexact->exact 1.234)
1407 => 694680242521899/562949953421312
1409 When you want the old behavior, use 'round' explicitly:
1411 (inexact->exact (round 1.234))
1414 ** New function 'rationalize'.
1416 This function finds a simple fraction that is close to a given real
1417 number. For example (and compare with inexact->exact above):
1419 (rationalize (inexact->exact 1.234) 1/2000)
1422 Note that, as required by R5RS, rationalize returns only then an exact
1423 result when both its arguments are exact.
1425 ** 'odd?' and 'even?' work also for inexact integers.
1427 Previously, (odd? 1.0) would signal an error since only exact integers
1428 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
1429 returns #f and (odd? 1.5) signals an error.
1431 ** Guile now has uninterned symbols.
1433 The new function 'make-symbol' will return an uninterned symbol. This
1434 is a symbol that is unique and is guaranteed to remain unique.
1435 However, uninterned symbols can not yet be read back in.
1437 Use the new function 'symbol-interned?' to check whether a symbol is
1440 ** pretty-print has more options.
1442 The function pretty-print from the (ice-9 pretty-print) module can now
1443 also be invoked with keyword arguments that control things like
1444 maximum output width. See the manual for details.
1446 ** Variables have no longer a special behavior for `equal?'.
1448 Previously, comparing two variables with `equal?' would recursivly
1449 compare their values. This is no longer done. Variables are now only
1450 `equal?' if they are `eq?'.
1452 ** `(begin)' is now valid.
1454 You can now use an empty `begin' form. It will yield #<unspecified>
1455 when evaluated and simply be ignored in a definition context.
1457 ** Deprecated: procedure->macro
1459 Change your code to use 'define-macro' or r5rs macros. Also, be aware
1460 that macro expansion will not be done during evaluation, but prior to
1463 ** Soft ports now allow a `char-ready?' procedure
1465 The vector argument to `make-soft-port' can now have a length of
1466 either 5 or 6. (Previously the length had to be 5.) The optional 6th
1467 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
1468 that returns the number of characters that can be read immediately
1469 without the soft port blocking.
1471 ** Deprecated: undefine
1473 There is no replacement for undefine.
1475 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
1476 have been discouraged.
1478 They are relics from a time where a keyword like #:foo was used
1479 directly as a Tcl option "-foo" and thus keywords were internally
1480 stored as a symbol with a starting dash. We now store a symbol
1483 Use symbol->keyword and keyword->symbol instead.
1485 ** The `cheap' debug option is now obsolete
1487 Evaluator trap calls are now unconditionally "cheap" - in other words,
1488 they pass a debug object to the trap handler rather than a full
1489 continuation. The trap handler code can capture a full continuation
1490 by using `call-with-current-continuation' in the usual way, if it so
1493 The `cheap' option is retained for now so as not to break existing
1494 code which gets or sets it, but setting it now has no effect. It will
1495 be removed in the next major Guile release.
1497 ** Evaluator trap calls now support `tweaking'
1499 `Tweaking' means that the trap handler code can modify the Scheme
1500 expression that is about to be evaluated (in the case of an
1501 enter-frame trap) or the value that is being returned (in the case of
1502 an exit-frame trap). The trap handler code indicates that it wants to
1503 do this by returning a pair whose car is the symbol 'instead and whose
1504 cdr is the modified expression or return value.
1506 * Changes to the C interface
1508 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
1509 take a 'delete' function argument.
1511 This argument makes no sense since the delete function is used to
1512 remove a pair from an alist, and this must not be configurable.
1514 This is an incompatible change.
1516 ** The GH interface is now subject to the deprecation mechanism
1518 The GH interface has been deprecated for quite some time but now it is
1519 actually removed from Guile when it is configured with
1520 --disable-deprecated.
1522 See the manual "Transitioning away from GH" for more information.
1524 ** A new family of functions for converting between C values and
1525 Scheme values has been added.
1527 These functions follow a common naming scheme and are designed to be
1528 easier to use, thread-safe and more future-proof than the older
1531 - int scm_is_* (...)
1533 These are predicates that return a C boolean: 1 or 0. Instead of
1534 SCM_NFALSEP, you can now use scm_is_true, for example.
1536 - <type> scm_to_<type> (SCM val, ...)
1538 These are functions that convert a Scheme value into an appropriate
1539 C value. For example, you can use scm_to_int to safely convert from
1542 - SCM scm_from_<type> (<type> val, ...)
1544 These functions convert from a C type to a SCM value; for example,
1545 scm_from_int for ints.
1547 There is a huge number of these functions, for numbers, strings,
1548 symbols, vectors, etc. They are documented in the reference manual in
1549 the API section together with the types that they apply to.
1551 ** New functions for dealing with complex numbers in C have been added.
1553 The new functions are scm_c_make_rectangular, scm_c_make_polar,
1554 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
1555 They work like scm_make_rectangular etc but take or return doubles
1558 ** The function scm_make_complex has been discouraged.
1560 Use scm_c_make_rectangular instead.
1562 ** The INUM macros have been deprecated.
1564 A lot of code uses these macros to do general integer conversions,
1565 although the macros only work correctly with fixnums. Use the
1566 following alternatives.
1568 SCM_INUMP -> scm_is_integer or similar
1569 SCM_NINUMP -> !scm_is_integer or similar
1570 SCM_MAKINUM -> scm_from_int or similar
1571 SCM_INUM -> scm_to_int or similar
1573 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
1574 do the validating for you.
1576 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
1577 have been discouraged.
1579 Use the newer scm_to_<type> and scm_from_<type> functions instead for
1580 new code. The functions have been discouraged since they don't fit
1583 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
1585 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
1586 evaluates its argument twice. Use scm_is_true, etc. instead for new
1589 ** The macro SCM_EQ_P has been discouraged.
1591 Use scm_is_eq for new code, which fits better into the naming
1594 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
1597 Use the function scm_is_pair or scm_is_null instead.
1599 ** The functions scm_round and scm_truncate have been deprecated and
1600 are now available as scm_c_round and scm_c_truncate, respectively.
1602 These functions occupy the names that scm_round_number and
1603 scm_truncate_number should have.
1605 ** The functions scm_c_string2str, scm_c_substring2str, and
1606 scm_c_symbol2str have been deprecated.
1608 Use scm_to_locale_stringbuf or similar instead, maybe together with
1611 ** New functions scm_c_make_string, scm_c_string_length,
1612 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
1613 scm_c_substring_shared, scm_c_substring_copy.
1615 These are like scm_make_string, scm_length, etc. but are slightly
1616 easier to use from C.
1618 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
1619 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
1621 They export too many assumptions about the implementation of strings
1622 and symbols that are no longer true in the presence of
1623 mutation-sharing substrings and when Guile switches to some form of
1626 When working with strings, it is often best to use the normal string
1627 functions provided by Guile, such as scm_c_string_ref,
1628 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
1629 manual since many more such functions are now provided than
1632 When you want to convert a SCM string to a C string, use the
1633 scm_to_locale_string function or similar instead. For symbols, use
1634 scm_symbol_to_string and then work with that string. Because of the
1635 new string representation, scm_symbol_to_string does not need to copy
1636 and is thus quite efficient.
1638 ** Some string, symbol and keyword functions have been discouraged.
1640 They don't fit into the uniform naming scheme and are not explicit
1641 about the character encoding.
1643 Replace according to the following table:
1645 scm_allocate_string -> scm_c_make_string
1646 scm_take_str -> scm_take_locale_stringn
1647 scm_take0str -> scm_take_locale_string
1648 scm_mem2string -> scm_from_locale_stringn
1649 scm_str2string -> scm_from_locale_string
1650 scm_makfrom0str -> scm_from_locale_string
1651 scm_mem2symbol -> scm_from_locale_symboln
1652 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
1653 scm_str2symbol -> scm_from_locale_symbol
1655 SCM_SYMBOL_HASH -> scm_hashq
1656 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
1658 scm_c_make_keyword -> scm_from_locale_keyword
1660 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
1661 now also available to C code.
1663 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
1665 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
1666 the latter returns the true name of the keyword, not the 'dash name',
1667 as SCM_KEYWORDSYM used to do.
1669 ** A new way to access arrays in a thread-safe and efficient way has
1672 See the manual, node "Accessing Arrays From C".
1674 ** The old uniform vector and bitvector implementations have been
1675 unceremoniously removed.
1677 This implementation exposed the details of the tagging system of
1678 Guile. Use the new C API explained in the manual in node "Uniform
1679 Numeric Vectors" and "Bit Vectors", respectively.
1681 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
1682 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
1683 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
1684 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
1685 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
1686 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
1689 ** The macros dealing with vectors have been deprecated.
1691 Use the new functions scm_is_vector, scm_vector_elements,
1692 scm_vector_writable_elements, etc, or scm_is_simple_vector,
1693 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
1694 manual for more details.
1696 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
1697 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
1699 The following macros have been removed: SCM_VECTOR_BASE,
1700 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
1701 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
1703 ** Some C functions and macros related to arrays have been deprecated.
1705 Migrate according to the following table:
1707 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
1708 scm_make_ra -> scm_make_array
1709 scm_shap2ra -> scm_make_array
1710 scm_cvref -> scm_c_generalized_vector_ref
1711 scm_ra_set_contp -> do not use
1712 scm_aind -> scm_array_handle_pos
1713 scm_raprin1 -> scm_display or scm_write
1715 SCM_ARRAYP -> scm_is_array
1716 SCM_ARRAY_NDIM -> scm_c_array_rank
1717 SCM_ARRAY_DIMS -> scm_array_handle_dims
1718 SCM_ARRAY_CONTP -> do not use
1719 SCM_ARRAY_MEM -> do not use
1720 SCM_ARRAY_V -> scm_array_handle_elements or similar
1721 SCM_ARRAY_BASE -> do not use
1723 ** SCM_CELL_WORD_LOC has been deprecated.
1725 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
1726 to a SCM, as opposed to a pointer to a scm_t_bits.
1728 This was done to allow the correct use of pointers into the Scheme
1729 heap. Previously, the heap words were of type scm_t_bits and local
1730 variables and function arguments were of type SCM, making it
1731 non-standards-conformant to have a pointer that can point to both.
1733 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
1735 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
1736 second and third words of double smobs. Likewise for
1737 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
1739 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
1740 used to get and set the 16 exra bits in the zeroth word of a smob.
1742 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
1743 accesing the first immediate word of a smob as a SCM value, and there
1744 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
1745 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
1747 ** New way to deal with non-local exits and re-entries.
1749 There is a new set of functions that essentially do what
1750 scm_internal_dynamic_wind does, but in a way that is more convenient
1751 for C code in some situations. Here is a quick example of how to
1752 prevent a potential memory leak:
1759 scm_dynwind_begin (0);
1761 mem = scm_malloc (100);
1762 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
1764 /* MEM would leak if BAR throws an error.
1765 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
1772 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
1773 SCM_DYNWIND_END as well.
1777 For full documentation, see the node "Dynamic Wind" in the manual.
1779 ** New function scm_dynwind_free
1781 This function calls 'free' on a given pointer when a dynwind context
1782 is left. Thus the call to scm_dynwind_unwind_handler above could be
1783 replaced with simply scm_dynwind_free (mem).
1785 ** New functions scm_c_call_with_blocked_asyncs and
1786 scm_c_call_with_unblocked_asyncs
1788 Like scm_call_with_blocked_asyncs etc. but for C functions.
1790 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
1792 In addition to scm_c_call_with_blocked_asyncs you can now also use
1793 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
1794 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
1796 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
1797 SCM_REALLOW_INTS have been deprecated.
1799 They do no longer fulfill their original role of blocking signal
1800 delivery. Depending on what you want to achieve, replace a pair of
1801 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
1802 mutex, blocks asyncs, or both. See node "Critical Sections" in the
1805 ** The value 'scm_mask_ints' is no longer writable.
1807 Previously, you could set scm_mask_ints directly. This is no longer
1808 possible. Use scm_c_call_with_blocked_asyncs and
1809 scm_c_call_with_unblocked_asyncs instead.
1811 ** New way to temporarily set the current input, output or error ports
1813 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
1814 context' (see above). <foo> is one of "input", "output" or "error".
1816 ** New way to temporarily set fluids
1818 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
1819 above) to temporarily set the value of a fluid.
1821 ** New types scm_t_intmax and scm_t_uintmax.
1823 On platforms that have them, these types are identical to intmax_t and
1824 uintmax_t, respectively. On other platforms, they are identical to
1825 the largest integer types that Guile knows about.
1827 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
1829 You should not have used them.
1831 ** Many public #defines with generic names have been made private.
1833 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
1834 private or renamed with a more suitable public name.
1836 ** The macro SCM_TYP16S has been deprecated.
1838 This macro is not intended for public use.
1840 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
1842 Use scm_is_true (scm_inexact_p (...)) instead.
1844 ** The macro SCM_SLOPPY_REALP has been deprecated.
1846 Use scm_is_real instead.
1848 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
1850 Use scm_is_complex instead.
1852 ** Some preprocessor defines have been deprecated.
1854 These defines indicated whether a certain feature was present in Guile
1855 or not. Going forward, assume that the features are always present.
1857 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
1858 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
1860 The following macros have been removed completely: MEMOIZE_LOCALS,
1861 SCM_RECKLESS, SCM_CAUTIOUS.
1863 ** The preprocessor define STACK_DIRECTION has been deprecated.
1865 There should be no need to know about the stack direction for ordinary
1868 ** New function: scm_effective_version
1870 Returns the "effective" version number. This is just the normal full
1871 version string without the final micro-version number. See "Changes
1872 to the distribution" above.
1874 ** The function scm_call_with_new_thread has a new prototype.
1876 Instead of taking a list with the thunk and handler, these two
1877 arguments are now passed directly:
1879 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
1881 This is an incompatible change.
1883 ** New snarfer macro SCM_DEFINE_PUBLIC.
1885 This is like SCM_DEFINE, but also calls scm_c_export for the defined
1886 function in the init section.
1888 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
1890 ** Garbage collector rewrite.
1892 The garbage collector is cleaned up a lot, and now uses lazy
1893 sweeping. This is reflected in the output of (gc-stats); since cells
1894 are being freed when they are allocated, the cells-allocated field
1895 stays roughly constant.
1897 For malloc related triggers, the behavior is changed. It uses the same
1898 heuristic as the cell-triggered collections. It may be tuned with the
1899 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
1900 for minimum yield of malloc related triggers. The default is 40.
1901 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
1904 Debugging operations for the freelist have been deprecated, along with
1905 the C variables that control garbage collection. The environment
1906 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
1907 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
1909 For understanding the memory usage of a GUILE program, the routine
1910 gc-live-object-stats returns an alist containing the number of live
1911 objects for every type.
1914 ** The function scm_definedp has been renamed to scm_defined_p
1916 The name scm_definedp is deprecated.
1918 ** The struct scm_cell type has been renamed to scm_t_cell
1920 This is in accordance to Guile's naming scheme for types. Note that
1921 the name scm_cell is now used for a function that allocates and
1922 initializes a new cell (see below).
1924 ** New functions for memory management
1926 A new set of functions for memory management has been added since the
1927 old way (scm_must_malloc, scm_must_free, etc) was error prone and
1928 indeed, Guile itself contained some long standing bugs that could
1929 cause aborts in long running programs.
1931 The new functions are more symmetrical and do not need cooperation
1932 from smob free routines, among other improvements.
1934 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
1935 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
1936 scm_gc_free, scm_gc_register_collectable_memory, and
1937 scm_gc_unregister_collectable_memory. Refer to the manual for more
1938 details and for upgrading instructions.
1940 The old functions for memory management have been deprecated. They
1941 are: scm_must_malloc, scm_must_realloc, scm_must_free,
1942 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
1944 ** Declarations of exported features are marked with SCM_API.
1946 Every declaration of a feature that belongs to the exported Guile API
1947 has been marked by adding the macro "SCM_API" to the start of the
1948 declaration. This macro can expand into different things, the most
1949 common of which is just "extern" for Unix platforms. On Win32, it can
1950 be used to control which symbols are exported from a DLL.
1952 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
1953 will expand into "__declspec (dllimport) extern", which is needed for
1954 linking to the Guile DLL in Windows.
1956 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
1957 SCM_SRFI4_IMPORT, for the corresponding libraries.
1959 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
1961 Use the new functions scm_cell and scm_double_cell instead. The old
1962 macros had problems because with them allocation and initialization
1963 was separated and the GC could sometimes observe half initialized
1964 cells. Only careful coding by the user of SCM_NEWCELL and
1965 SCM_NEWCELL2 could make this safe and efficient.
1967 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
1969 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
1972 ** SRCBRKP has been deprecated.
1974 Use scm_c_source_property_breakpoint_p instead.
1976 ** Deprecated: scm_makmacro
1978 Change your code to use either scm_makmmacro or to define macros in
1979 Scheme, using 'define-macro'.
1981 ** New function scm_c_port_for_each.
1983 This function is like scm_port_for_each but takes a pointer to a C
1984 function as the callback instead of a SCM value.
1986 ** The names scm_internal_select, scm_thread_sleep, and
1987 scm_thread_usleep have been discouraged.
1989 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
1991 ** The GC can no longer be blocked.
1993 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
1994 The GC can now run (partially) concurrently with other code and thus
1995 blocking it is not well defined.
1997 ** Many definitions have been removed that were previously deprecated.
1999 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
2000 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
2001 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
2002 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
2003 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
2004 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
2005 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
2006 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
2007 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
2008 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
2009 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
2010 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
2011 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
2012 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
2013 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
2014 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
2015 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
2016 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
2017 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
2018 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
2019 scm_vector_set_length_x, scm_contregs, scm_debug_info,
2020 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
2021 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
2022 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
2023 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
2024 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
2025 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
2026 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
2027 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
2028 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
2029 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
2031 * Changes to bundled modules
2035 Using the (ice-9 debug) module no longer automatically switches Guile
2036 to use the debugging evaluator. If you want to switch to the
2037 debugging evaluator (which is needed for backtrace information if you
2038 hit an error), please add an explicit "(debug-enable 'debug)" to your
2039 code just after the code to use (ice-9 debug).
2042 Changes since Guile 1.4:
2044 * Changes to the distribution
2046 ** A top-level TODO file is included.
2048 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
2050 Guile now always uses three numbers to represent the version,
2051 i.e. "1.6.5". The first number, 1, is the major version number, the
2052 second number, 6, is the minor version number, and the third number,
2053 5, is the micro version number. Changes in major version number
2054 indicate major changes in Guile.
2056 Minor version numbers that are even denote stable releases, and odd
2057 minor version numbers denote development versions (which may be
2058 unstable). The micro version number indicates a minor sub-revision of
2059 a given MAJOR.MINOR release.
2061 In keeping with the new scheme, (minor-version) and scm_minor_version
2062 no longer return everything but the major version number. They now
2063 just return the minor version number. Two new functions
2064 (micro-version) and scm_micro_version have been added to report the
2065 micro version number.
2067 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
2069 ** New preprocessor definitions are available for checking versions.
2071 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
2072 SCM_MICRO_VERSION to the appropriate integer values.
2074 ** Guile now actively warns about deprecated features.
2076 The new configure option `--enable-deprecated=LEVEL' and the
2077 environment variable GUILE_WARN_DEPRECATED control this mechanism.
2078 See INSTALL and README for more information.
2080 ** Guile is much more likely to work on 64-bit architectures.
2082 Guile now compiles and passes "make check" with only two UNRESOLVED GC
2083 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
2084 for the use of a test machine, and thanks to Stefan Jahn for ia64
2087 ** New functions: setitimer and getitimer.
2089 These implement a fairly direct interface to the libc functions of the
2092 ** The #. reader extension is now disabled by default.
2094 For safety reasons, #. evaluation is disabled by default. To
2095 re-enable it, set the fluid read-eval? to #t. For example:
2097 (fluid-set! read-eval? #t)
2099 but make sure you realize the potential security risks involved. With
2100 read-eval? enabled, reading a data file from an untrusted source can
2103 ** New SRFI modules have been added:
2105 SRFI-0 `cond-expand' is now supported in Guile, without requiring
2108 (srfi srfi-1) is a library containing many useful pair- and list-processing
2111 (srfi srfi-2) exports and-let*.
2113 (srfi srfi-4) implements homogeneous numeric vector datatypes.
2115 (srfi srfi-6) is a dummy module for now, since guile already provides
2116 all of the srfi-6 procedures by default: open-input-string,
2117 open-output-string, get-output-string.
2119 (srfi srfi-8) exports receive.
2121 (srfi srfi-9) exports define-record-type.
2123 (srfi srfi-10) exports define-reader-ctor and implements the reader
2126 (srfi srfi-11) exports let-values and let*-values.
2128 (srfi srfi-13) implements the SRFI String Library.
2130 (srfi srfi-14) implements the SRFI Character-Set Library.
2132 (srfi srfi-17) implements setter and getter-with-setter and redefines
2133 some accessor procedures as procedures with getters. (such as car,
2134 cdr, vector-ref etc.)
2136 (srfi srfi-19) implements the SRFI Time/Date Library.
2138 ** New scripts / "executable modules"
2140 Subdirectory "scripts" contains Scheme modules that are packaged to
2141 also be executable as scripts. At this time, these scripts are available:
2150 See README there for more info.
2152 These scripts can be invoked from the shell with the new program
2153 "guile-tools", which keeps track of installation directory for you.
2156 $ guile-tools display-commentary srfi/*.scm
2158 guile-tools is copied to the standard $bindir on "make install".
2160 ** New module (ice-9 stack-catch):
2162 stack-catch is like catch, but saves the current state of the stack in
2163 the fluid the-last-stack. This fluid can be useful when using the
2164 debugger and when re-throwing an error.
2166 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
2168 This has been done to prevent problems on lesser operating systems
2169 that can't tolerate `*'s in file names. The exported macro continues
2170 to be named `and-let*', of course.
2172 On systems that support it, there is also a compatibility module named
2173 (ice-9 and-let*). It will go away in the next release.
2175 ** New modules (oop goops) etc.:
2178 (oop goops describe)
2180 (oop goops active-slot)
2181 (oop goops composite-slot)
2183 The Guile Object Oriented Programming System (GOOPS) has been
2184 integrated into Guile. For further information, consult the GOOPS
2185 manual and tutorial in the `doc' directory.
2187 ** New module (ice-9 rdelim).
2189 This exports the following procedures which were previously defined
2190 in the default environment:
2192 read-line read-line! read-delimited read-delimited! %read-delimited!
2193 %read-line write-line
2195 For backwards compatibility the definitions are still imported into the
2196 default environment in this version of Guile. However you should add:
2198 (use-modules (ice-9 rdelim))
2200 to any program which uses the definitions, since this may change in
2203 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
2204 can be used for similar functionality.
2206 ** New module (ice-9 rw)
2208 This is a subset of the (scsh rw) module from guile-scsh. Currently
2209 it defines two procedures:
2211 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
2213 Read characters from a port or file descriptor into a string STR.
2214 A port must have an underlying file descriptor -- a so-called
2215 fport. This procedure is scsh-compatible and can efficiently read
2218 *** New function: write-string/partial str [port_or_fdes [start [end]]]
2220 Write characters from a string STR to a port or file descriptor.
2221 A port must have an underlying file descriptor -- a so-called
2222 fport. This procedure is mostly compatible and can efficiently
2223 write large strings.
2225 ** New module (ice-9 match)
2227 This module includes Andrew K. Wright's pattern matcher. See
2228 ice-9/match.scm for brief description or
2230 http://www.star-lab.com/wright/code.html
2232 for complete documentation.
2234 ** New module (ice-9 buffered-input)
2236 This module provides procedures to construct an input port from an
2237 underlying source of input that reads and returns its input in chunks.
2238 The underlying input source is a Scheme procedure, specified by the
2239 caller, which the port invokes whenever it needs more input.
2241 This is useful when building an input port whose back end is Readline
2242 or a UI element such as the GtkEntry widget.
2246 The reference and tutorial documentation that was previously
2247 distributed separately, as `guile-doc', is now included in the core
2248 Guile distribution. The documentation consists of the following
2251 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
2254 - The Guile Reference Manual (guile.texi) contains (or is intended to
2255 contain) reference documentation on all aspects of Guile.
2257 - The GOOPS Manual (goops.texi) contains both tutorial-style and
2258 reference documentation for using GOOPS, Guile's Object Oriented
2261 - The Revised^5 Report on the Algorithmic Language Scheme
2264 See the README file in the `doc' directory for more details.
2266 ** There are a couple of examples in the examples/ directory now.
2268 * Changes to the stand-alone interpreter
2270 ** New command line option `--use-srfi'
2272 Using this option, SRFI modules can be loaded on startup and be
2273 available right from the beginning. This makes programming portable
2274 Scheme programs easier.
2276 The option `--use-srfi' expects a comma-separated list of numbers,
2277 each representing a SRFI number to be loaded into the interpreter
2278 before starting evaluating a script file or the REPL. Additionally,
2279 the feature identifier for the loaded SRFIs is recognized by
2280 `cond-expand' when using this option.
2283 $ guile --use-srfi=8,13
2284 guile> (receive (x z) (values 1 2) (+ 1 2))
2286 guile> (string-pad "bla" 20)
2289 ** Guile now always starts up in the `(guile-user)' module.
2291 Previously, scripts executed via the `-s' option would run in the
2292 `(guile)' module and the repl would run in the `(guile-user)' module.
2293 Now every user action takes place in the `(guile-user)' module by
2296 * Changes to Scheme functions and syntax
2298 ** Character classifiers work for non-ASCII characters.
2300 The predicates `char-alphabetic?', `char-numeric?',
2301 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
2302 no longer check whether their arguments are ASCII characters.
2303 Previously, a character would only be considered alphabetic when it
2304 was also ASCII, for example.
2306 ** Previously deprecated Scheme functions have been removed:
2308 tag - no replacement.
2309 fseek - replaced by seek.
2310 list* - replaced by cons*.
2312 ** It's now possible to create modules with controlled environments
2316 (use-modules (ice-9 safe))
2317 (define m (make-safe-module))
2318 ;;; m will now be a module containing only a safe subset of R5RS
2319 (eval '(+ 1 2) m) --> 3
2320 (eval 'load m) --> ERROR: Unbound variable: load
2322 ** Evaluation of "()", the empty list, is now an error.
2324 Previously, the expression "()" evaluated to the empty list. This has
2325 been changed to signal a "missing expression" error. The correct way
2326 to write the empty list as a literal constant is to use quote: "'()".
2328 ** New concept of `Guile Extensions'.
2330 A Guile Extension is just a ordinary shared library that can be linked
2331 at run-time. We found it advantageous to give this simple concept a
2332 dedicated name to distinguish the issues related to shared libraries
2333 from the issues related to the module system.
2335 *** New function: load-extension
2337 Executing (load-extension lib init) is mostly equivalent to
2339 (dynamic-call init (dynamic-link lib))
2341 except when scm_register_extension has been called previously.
2342 Whenever appropriate, you should use `load-extension' instead of
2343 dynamic-link and dynamic-call.
2345 *** New C function: scm_c_register_extension
2347 This function registers a initialization function for use by
2348 `load-extension'. Use it when you don't want specific extensions to
2349 be loaded as shared libraries (for example on platforms that don't
2350 support dynamic linking).
2352 ** Auto-loading of compiled-code modules is deprecated.
2354 Guile used to be able to automatically find and link a shared
2355 library to satisfy requests for a module. For example, the module
2356 `(foo bar)' could be implemented by placing a shared library named
2357 "foo/libbar.so" (or with a different extension) in a directory on the
2360 This has been found to be too tricky, and is no longer supported. The
2361 shared libraries are now called "extensions". You should now write a
2362 small Scheme file that calls `load-extension' to load the shared
2363 library and initialize it explicitly.
2365 The shared libraries themselves should be installed in the usual
2366 places for shared libraries, with names like "libguile-foo-bar".
2368 For example, place this into a file "foo/bar.scm"
2370 (define-module (foo bar))
2372 (load-extension "libguile-foo-bar" "foobar_init")
2374 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
2376 `eval' is now R5RS, that is it takes two arguments.
2377 The second argument is an environment specifier, i.e. either
2379 (scheme-report-environment 5)
2380 (null-environment 5)
2381 (interaction-environment)
2387 ** The module system has been made more disciplined.
2389 The function `eval' will save and restore the current module around
2390 the evaluation of the specified expression. While this expression is
2391 evaluated, `(current-module)' will now return the right module, which
2392 is the module specified as the second argument to `eval'.
2394 A consequence of this change is that `eval' is not particularly
2395 useful when you want allow the evaluated code to change what module is
2396 designated as the current module and have this change persist from one
2397 call to `eval' to the next. The read-eval-print-loop is an example
2398 where `eval' is now inadequate. To compensate, there is a new
2399 function `primitive-eval' that does not take a module specifier and
2400 that does not save/restore the current module. You should use this
2401 function together with `set-current-module', `current-module', etc
2402 when you want to have more control over the state that is carried from
2403 one eval to the next.
2405 Additionally, it has been made sure that forms that are evaluated at
2406 the top level are always evaluated with respect to the current module.
2407 Previously, subforms of top-level forms such as `begin', `case',
2408 etc. did not respect changes to the current module although these
2409 subforms are at the top-level as well.
2411 To prevent strange behavior, the forms `define-module',
2412 `use-modules', `use-syntax', and `export' have been restricted to only
2413 work on the top level. The forms `define-public' and
2414 `defmacro-public' only export the new binding on the top level. They
2415 behave just like `define' and `defmacro', respectively, when they are
2416 used in a lexical environment.
2418 Also, `export' will no longer silently re-export bindings imported
2419 from a used module. It will emit a `deprecation' warning and will
2420 cease to perform any re-export in the next version. If you actually
2421 want to re-export bindings, use the new `re-export' in place of
2422 `export'. The new `re-export' will not make copies of variables when
2423 rexporting them, as `export' did wrongly.
2425 ** Module system now allows selection and renaming of imported bindings
2427 Previously, when using `use-modules' or the `#:use-module' clause in
2428 the `define-module' form, all the bindings (association of symbols to
2429 values) for imported modules were added to the "current module" on an
2430 as-is basis. This has been changed to allow finer control through two
2431 new facilities: selection and renaming.
2433 You can now select which of the imported module's bindings are to be
2434 visible in the current module by using the `:select' clause. This
2435 clause also can be used to rename individual bindings. For example:
2437 ;; import all bindings no questions asked
2438 (use-modules (ice-9 common-list))
2440 ;; import four bindings, renaming two of them;
2441 ;; the current module sees: every some zonk-y zonk-n
2442 (use-modules ((ice-9 common-list)
2444 (remove-if . zonk-y)
2445 (remove-if-not . zonk-n))))
2447 You can also programmatically rename all selected bindings using the
2448 `:renamer' clause, which specifies a proc that takes a symbol and
2449 returns another symbol. Because it is common practice to use a prefix,
2450 we now provide the convenience procedure `symbol-prefix-proc'. For
2453 ;; import four bindings, renaming two of them specifically,
2454 ;; and all four w/ prefix "CL:";
2455 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
2456 (use-modules ((ice-9 common-list)
2458 (remove-if . zonk-y)
2459 (remove-if-not . zonk-n))
2460 :renamer (symbol-prefix-proc 'CL:)))
2462 ;; import four bindings, renaming two of them specifically,
2463 ;; and all four by upcasing.
2464 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
2465 (define (upcase-symbol sym)
2466 (string->symbol (string-upcase (symbol->string sym))))
2468 (use-modules ((ice-9 common-list)
2470 (remove-if . zonk-y)
2471 (remove-if-not . zonk-n))
2472 :renamer upcase-symbol))
2474 Note that programmatic renaming is done *after* individual renaming.
2475 Also, the above examples show `use-modules', but the same facilities are
2476 available for the `#:use-module' clause of `define-module'.
2478 See manual for more info.
2480 ** The semantics of guardians have changed.
2482 The changes are for the most part compatible. An important criterion
2483 was to keep the typical usage of guardians as simple as before, but to
2484 make the semantics safer and (as a result) more useful.
2486 *** All objects returned from guardians are now properly alive.
2488 It is now guaranteed that any object referenced by an object returned
2489 from a guardian is alive. It's now impossible for a guardian to
2490 return a "contained" object before its "containing" object.
2492 One incompatible (but probably not very important) change resulting
2493 from this is that it is no longer possible to guard objects that
2494 indirectly reference themselves (i.e. are parts of cycles). If you do
2495 so accidentally, you'll get a warning.
2497 *** There are now two types of guardians: greedy and sharing.
2499 If you call (make-guardian #t) or just (make-guardian), you'll get a
2500 greedy guardian, and for (make-guardian #f) a sharing guardian.
2502 Greedy guardians are the default because they are more "defensive".
2503 You can only greedily guard an object once. If you guard an object
2504 more than once, once in a greedy guardian and the rest of times in
2505 sharing guardians, then it is guaranteed that the object won't be
2506 returned from sharing guardians as long as it is greedily guarded
2509 Guardians returned by calls to `make-guardian' can now take one more
2510 optional parameter, which says whether to throw an error in case an
2511 attempt is made to greedily guard an object that is already greedily
2512 guarded. The default is true, i.e. throw an error. If the parameter
2513 is false, the guardian invocation returns #t if guarding was
2514 successful and #f if it wasn't.
2516 Also, since greedy guarding is, in effect, a side-effecting operation
2517 on objects, a new function is introduced: `destroy-guardian!'.
2518 Invoking this function on a guardian renders it unoperative and, if
2519 the guardian is greedy, clears the "greedily guarded" property of the
2520 objects that were guarded by it, thus undoing the side effect.
2522 Note that all this hair is hardly very important, since guardian
2523 objects are usually permanent.
2525 ** Continuations created by call-with-current-continuation now accept
2526 any number of arguments, as required by R5RS.
2528 ** New function `issue-deprecation-warning'
2530 This function is used to display the deprecation messages that are
2531 controlled by GUILE_WARN_DEPRECATION as explained in the README.
2534 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
2538 ;; `id' is deprecated. Use `identity' instead.
2543 ** New syntax `begin-deprecated'
2545 When deprecated features are included (as determined by the configure
2546 option --enable-deprecated), `begin-deprecated' is identical to
2547 `begin'. When deprecated features are excluded, it always evaluates
2548 to `#f', ignoring the body forms.
2550 ** New function `make-object-property'
2552 This function returns a new `procedure with setter' P that can be used
2553 to attach a property to objects. When calling P as
2557 where `obj' is any kind of object, it attaches `val' to `obj' in such
2558 a way that it can be retrieved by calling P as
2562 This function will replace procedure properties, symbol properties and
2563 source properties eventually.
2565 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
2567 Instead of #&optional, #&key, etc you should now use #:optional,
2568 #:key, etc. Since #:optional is a keyword, you can write it as just
2569 :optional when (read-set! keywords 'prefix) is active.
2571 The old reader syntax `#&' is still supported, but deprecated. It
2572 will be removed in the next release.
2574 ** New define-module option: pure
2576 Tells the module system not to include any bindings from the root
2581 (define-module (totally-empty-module)
2584 ** New define-module option: export NAME1 ...
2586 Export names NAME1 ...
2588 This option is required if you want to be able to export bindings from
2589 a module which doesn't import one of `define-public' or `export'.
2593 (define-module (foo)
2595 :use-module (ice-9 r5rs)
2598 ;;; Note that we're pure R5RS below this point!
2603 ** New function: object->string OBJ
2605 Return a Scheme string obtained by printing a given object.
2607 ** New function: port? X
2609 Returns a boolean indicating whether X is a port. Equivalent to
2610 `(or (input-port? X) (output-port? X))'.
2612 ** New function: file-port?
2614 Determines whether a given object is a port that is related to a file.
2616 ** New function: port-for-each proc
2618 Apply PROC to each port in the Guile port table in turn. The return
2619 value is unspecified. More specifically, PROC is applied exactly once
2620 to every port that exists in the system at the time PORT-FOR-EACH is
2621 invoked. Changes to the port table while PORT-FOR-EACH is running
2622 have no effect as far as PORT-FOR-EACH is concerned.
2624 ** New function: dup2 oldfd newfd
2626 A simple wrapper for the `dup2' system call. Copies the file
2627 descriptor OLDFD to descriptor number NEWFD, replacing the
2628 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
2629 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
2630 to move away ports which are using NEWFD. The return value is
2633 ** New function: close-fdes fd
2635 A simple wrapper for the `close' system call. Close file
2636 descriptor FD, which must be an integer. Unlike close (*note
2637 close: Ports and File Descriptors.), the file descriptor will be
2638 closed even if a port is using it. The return value is
2641 ** New function: crypt password salt
2643 Encrypts `password' using the standard unix password encryption
2646 ** New function: chroot path
2648 Change the root directory of the running process to `path'.
2650 ** New functions: getlogin, cuserid
2652 Return the login name or the user name of the current effective user
2655 ** New functions: getpriority which who, setpriority which who prio
2657 Get or set the priority of the running process.
2659 ** New function: getpass prompt
2661 Read a password from the terminal, first displaying `prompt' and
2664 ** New function: flock file operation
2666 Set/remove an advisory shared or exclusive lock on `file'.
2668 ** New functions: sethostname name, gethostname
2670 Set or get the hostname of the machine the current process is running
2673 ** New function: mkstemp! tmpl
2675 mkstemp creates a new unique file in the file system and returns a
2676 new buffered port open for reading and writing to the file. TMPL
2677 is a string specifying where the file should be created: it must
2678 end with `XXXXXX' and will be changed in place to return the name
2679 of the temporary file.
2681 ** New function: open-input-string string
2683 Return an input string port which delivers the characters from
2684 `string'. This procedure, together with `open-output-string' and
2685 `get-output-string' implements SRFI-6.
2687 ** New function: open-output-string
2689 Return an output string port which collects all data written to it.
2690 The data can then be retrieved by `get-output-string'.
2692 ** New function: get-output-string
2694 Return the contents of an output string port.
2696 ** New function: identity
2698 Return the argument.
2700 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
2701 are represented in Scheme as integers with normal host byte ordering.
2703 ** New function: inet-pton family address
2705 Convert a printable string network address into an integer. Note that
2706 unlike the C version of this function, the result is an integer with
2707 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2710 (inet-pton AF_INET "127.0.0.1") => 2130706433
2711 (inet-pton AF_INET6 "::1") => 1
2713 ** New function: inet-ntop family address
2715 Convert an integer network address into a printable string. Note that
2716 unlike the C version of this function, the input is an integer with
2717 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2720 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
2721 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
2722 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
2726 Use `identity' instead.
2732 ** Deprecated: return-it
2736 ** Deprecated: string-character-length
2738 Use `string-length' instead.
2740 ** Deprecated: flags
2742 Use `logior' instead.
2744 ** Deprecated: close-all-ports-except.
2746 This was intended for closing ports in a child process after a fork,
2747 but it has the undesirable side effect of flushing buffers.
2748 port-for-each is more flexible.
2750 ** The (ice-9 popen) module now attempts to set up file descriptors in
2751 the child process from the current Scheme ports, instead of using the
2752 current values of file descriptors 0, 1, and 2 in the parent process.
2754 ** Removed function: builtin-weak-bindings
2756 There is no such concept as a weak binding any more.
2758 ** Removed constants: bignum-radix, scm-line-incrementors
2760 ** define-method: New syntax mandatory.
2762 The new method syntax is now mandatory:
2764 (define-method (NAME ARG-SPEC ...) BODY ...)
2765 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
2767 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
2768 REST-ARG ::= ARG-NAME
2770 If you have old code using the old syntax, import
2771 (oop goops old-define-method) before (oop goops) as in:
2773 (use-modules (oop goops old-define-method) (oop goops))
2775 ** Deprecated function: builtin-variable
2776 Removed function: builtin-bindings
2778 There is no longer a distinction between builtin or other variables.
2779 Use module system operations for all variables.
2781 ** Lazy-catch handlers are no longer allowed to return.
2783 That is, a call to `throw', `error', etc is now guaranteed to not
2786 ** Bugfixes for (ice-9 getopt-long)
2788 This module is now tested using test-suite/tests/getopt-long.test.
2789 The following bugs have been fixed:
2791 *** Parsing for options that are specified to have `optional' args now checks
2792 if the next element is an option instead of unconditionally taking it as the
2795 *** An error is now thrown for `--opt=val' when the option description
2796 does not specify `(value #t)' or `(value optional)'. This condition used to
2797 be accepted w/o error, contrary to the documentation.
2799 *** The error message for unrecognized options is now more informative.
2800 It used to be "not a record", an artifact of the implementation.
2802 *** The error message for `--opt' terminating the arg list (no value), when
2803 `(value #t)' is specified, is now more informative. It used to be "not enough
2806 *** "Clumped" single-char args now preserve trailing string, use it as arg.
2807 The expansion used to be like so:
2809 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
2811 Note that the "5d" is dropped. Now it is like so:
2813 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
2815 This enables single-char options to have adjoining arguments as long as their
2816 constituent characters are not potential single-char options.
2818 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
2820 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
2821 property `arglist', which can be retrieved by `arity'. The result is that
2822 `arity' can give more detailed information than before:
2826 guile> (use-modules (ice-9 optargs))
2827 guile> (define* (foo #:optional a b c) a)
2829 0 or more arguments in `lambda*:G0'.
2834 3 optional arguments: `a', `b' and `c'.
2835 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
2837 2 required arguments: `a' and `b', 2 keyword arguments: `c'
2838 and `d', other keywords allowed.
2839 guile> (define* (baz a b #:optional c #:rest r) a)
2841 2 required arguments: `a' and `b', 1 optional argument: `c',
2844 * Changes to the C interface
2846 ** Types have been renamed from scm_*_t to scm_t_*.
2848 This has been done for POSIX sake. It reserves identifiers ending
2849 with "_t". What a concept.
2851 The old names are still available with status `deprecated'.
2853 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
2855 ** Deprecated features have been removed.
2859 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
2860 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
2862 *** C Functions removed
2864 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
2865 scm_fseek - replaced by scm_seek.
2866 gc-thunk - replaced by after-gc-hook.
2867 gh_int2scmb - replaced by gh_bool2scm.
2868 scm_tc_dblr - replaced by scm_tc16_real.
2869 scm_tc_dblc - replaced by scm_tc16_complex.
2870 scm_list_star - replaced by scm_cons_star.
2872 ** Deprecated: scm_makfromstr
2874 Use scm_mem2string instead.
2876 ** Deprecated: scm_make_shared_substring
2878 Explicit shared substrings will disappear from Guile.
2880 Instead, "normal" strings will be implemented using sharing
2881 internally, combined with a copy-on-write strategy.
2883 ** Deprecated: scm_read_only_string_p
2885 The concept of read-only strings will disappear in next release of
2888 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
2890 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
2892 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
2894 Call a procedure with the indicated number of arguments. See "Fly
2895 Evaluation" in the manual.
2897 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
2899 Call a procedure with the indicated number of arguments and a list of
2900 further arguments. See "Fly Evaluation" in the manual.
2902 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
2904 Create a list of the given number of elements. See "List
2905 Constructors" in the manual.
2907 ** Renamed function: scm_listify has been replaced by scm_list_n.
2909 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
2910 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
2912 Use functions scm_list_N instead.
2914 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
2916 Used by an application to read arbitrary number of bytes from a port.
2917 Same semantics as libc read, except that scm_c_read only returns less
2918 than SIZE bytes if at end-of-file.
2920 Warning: Doesn't update port line and column counts!
2922 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
2924 Used by an application to write arbitrary number of bytes to an SCM
2925 port. Similar semantics as libc write. However, unlike libc
2926 write, scm_c_write writes the requested number of bytes and has no
2929 Warning: Doesn't update port line and column counts!
2931 ** New function: scm_init_guile ()
2933 In contrast to scm_boot_guile, scm_init_guile will return normally
2934 after initializing Guile. It is not available on all systems, tho.
2936 ** New functions: scm_str2symbol, scm_mem2symbol
2938 The function scm_str2symbol takes a const char* pointing to a zero-terminated
2939 field of characters and creates a scheme symbol object from that C string.
2940 The function scm_mem2symbol takes a const char* and a number of characters and
2941 creates a symbol from the characters in that memory area.
2943 ** New functions: scm_primitive_make_property
2944 scm_primitive_property_ref
2945 scm_primitive_property_set_x
2946 scm_primitive_property_del_x
2948 These functions implement a new way to deal with object properties.
2949 See libguile/properties.c for their documentation.
2951 ** New function: scm_done_free (long size)
2953 This function is the inverse of scm_done_malloc. Use it to report the
2954 amount of smob memory you free. The previous method, which involved
2955 calling scm_done_malloc with negative argument, was somewhat
2956 unintuitive (and is still available, of course).
2958 ** New function: scm_c_memq (SCM obj, SCM list)
2960 This function provides a fast C level alternative for scm_memq for the case
2961 that the list parameter is known to be a proper list. The function is a
2962 replacement for scm_sloppy_memq, but is stricter in its requirements on its
2963 list input parameter, since for anything else but a proper list the function's
2964 behaviour is undefined - it may even crash or loop endlessly. Further, for
2965 the case that the object is not found in the list, scm_c_memq returns #f which
2966 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
2968 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
2969 scm_remember_upto_here
2971 These functions replace the function scm_remember.
2973 ** Deprecated function: scm_remember
2975 Use one of the new functions scm_remember_upto_here_1,
2976 scm_remember_upto_here_2 or scm_remember_upto_here instead.
2978 ** New function: scm_allocate_string
2980 This function replaces the function scm_makstr.
2982 ** Deprecated function: scm_makstr
2984 Use the new function scm_allocate_string instead.
2986 ** New global variable scm_gc_running_p introduced.
2988 Use this variable to find out if garbage collection is being executed. Up to
2989 now applications have used scm_gc_heap_lock to test if garbage collection was
2990 running, which also works because of the fact that up to know only the garbage
2991 collector has set this variable. But, this is an implementation detail that
2992 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
2993 of this variable is (and has been) not fully safe anyway.
2995 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
2997 Use these instead of SCM_LENGTH_MAX.
2999 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
3000 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
3001 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
3003 Use these instead of SCM_LENGTH.
3005 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
3006 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
3007 SCM_SET_BITVECTOR_LENGTH
3009 Use these instead of SCM_SETLENGTH
3011 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
3012 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
3015 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
3018 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
3019 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
3022 Use these instead of SCM_SETCHARS.
3024 ** New macro: SCM_BITVECTOR_P
3026 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
3028 Use instead of SCM_COERCE_SUBSTR.
3030 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
3032 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
3034 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
3035 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
3036 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
3037 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
3038 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
3039 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
3040 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
3041 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
3042 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
3043 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
3044 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
3045 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
3046 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
3047 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
3048 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
3050 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
3051 Use scm_memory_error instead of SCM_NALLOC.
3052 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
3053 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
3054 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
3055 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
3056 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
3057 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
3058 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
3059 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
3060 Use SCM_STRINGP instead of SCM_RWSTRINGP.
3061 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
3062 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
3063 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
3064 Use a type specific setter macro instead of SCM_SETLENGTH.
3065 Use a type specific setter macro instead of SCM_SETCHARS.
3066 Use a type specific length macro instead of SCM_LENGTH_MAX.
3067 Use SCM_GCMARKP instead of SCM_GC8MARKP.
3068 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
3069 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
3070 Use SCM_TYP16 instead of SCM_GCTYP16.
3071 Use SCM_CDR instead of SCM_GCCDR.
3072 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
3073 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
3074 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
3075 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
3076 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
3077 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
3078 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
3080 ** Removed function: scm_struct_init
3082 ** Removed variable: scm_symhash_dim
3084 ** Renamed function: scm_make_cont has been replaced by
3085 scm_make_continuation, which has a different interface.
3087 ** Deprecated function: scm_call_catching_errors
3089 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
3091 ** Deprecated function: scm_strhash
3093 Use scm_string_hash instead.
3095 ** Deprecated function: scm_vector_set_length_x
3097 Instead, create a fresh vector of the desired size and copy the contents.
3099 ** scm_gensym has changed prototype
3101 scm_gensym now only takes one argument.
3103 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
3106 There is now only a single symbol type scm_tc7_symbol.
3107 The tag scm_tc7_lvector was not used anyway.
3109 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
3111 Use scm_make_smob_type and scm_set_smob_XXX instead.
3113 ** New function scm_set_smob_apply.
3115 This can be used to set an apply function to a smob type.
3117 ** Deprecated function: scm_strprint_obj
3119 Use scm_object_to_string instead.
3121 ** Deprecated function: scm_wta
3123 Use scm_wrong_type_arg, or another appropriate error signalling function
3126 ** Explicit support for obarrays has been deprecated.
3128 Use `scm_str2symbol' and the generic hashtable functions instead.
3130 ** The concept of `vcells' has been deprecated.
3132 The data type `variable' is now used exclusively. `Vcells' have been
3133 a low-level concept so you are likely not affected by this change.
3135 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
3136 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
3138 Use scm_c_define or scm_c_lookup instead, as appropriate.
3140 *** New functions: scm_c_module_lookup, scm_c_lookup,
3141 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
3142 scm_module_define, scm_define.
3144 These functions work with variables instead of with vcells.
3146 ** New functions for creating and defining `subr's and `gsubr's.
3148 The new functions more clearly distinguish between creating a subr (or
3149 gsubr) object and adding it to the current module.
3151 These new functions are available: scm_c_make_subr, scm_c_define_subr,
3152 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
3153 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
3154 scm_c_define_gsubr_with_generic.
3156 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
3157 scm_make_subr_with_generic, scm_make_gsubr,
3158 scm_make_gsubr_with_generic.
3160 Use the new ones from above instead.
3162 ** C interface to the module system has changed.
3164 While we suggest that you avoid as many explicit module system
3165 operations from C as possible for the time being, the C interface has
3166 been made more similar to the high-level Scheme module system.
3168 *** New functions: scm_c_define_module, scm_c_use_module,
3169 scm_c_export, scm_c_resolve_module.
3171 They mostly work like their Scheme namesakes. scm_c_define_module
3172 takes a function that is called a context where the new module is
3175 *** Deprecated functions: scm_the_root_module, scm_make_module,
3176 scm_ensure_user_module, scm_load_scheme_module.
3178 Use the new functions instead.
3180 ** Renamed function: scm_internal_with_fluids becomes
3183 scm_internal_with_fluids is available as a deprecated function.
3185 ** New function: scm_c_with_fluid.
3187 Just like scm_c_with_fluids, but takes one fluid and one value instead
3190 ** Deprecated typedefs: long_long, ulong_long.
3192 They are of questionable utility and they pollute the global
3195 ** Deprecated typedef: scm_sizet
3197 It is of questionable utility now that Guile requires ANSI C, and is
3200 ** Deprecated typedefs: scm_port_rw_active, scm_port,
3201 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
3202 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
3204 Made more compliant with the naming policy by adding a _t at the end.
3206 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
3207 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
3209 With the exception of the mysterious scm_2ulong2big, they are still
3210 available under new names (scm_i_mkbig etc). These functions are not
3211 intended to be used in user code. You should avoid dealing with
3212 bignums directly, and should deal with numbers in general (which can
3215 ** Change in behavior: scm_num2long, scm_num2ulong
3217 The scm_num2[u]long functions don't any longer accept an inexact
3218 argument. This change in behavior is motivated by concordance with
3219 R5RS: It is more common that a primitive doesn't want to accept an
3220 inexact for an exact.
3222 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
3223 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
3224 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
3227 These are conversion functions between the various ANSI C integral
3228 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
3229 accept an inexact argument.
3231 ** New functions: scm_float2num, scm_double2num,
3232 scm_num2float, scm_num2double.
3234 These are conversion functions between the two ANSI C float types and
3237 ** New number validation macros:
3238 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
3242 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
3244 These are just nicer-named old scm_protect_object and
3245 scm_unprotect_object.
3247 ** Deprecated functions: scm_protect_object, scm_unprotect_object
3249 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
3251 These functions can be used to register pointers to locations that
3254 ** Deprecated function: scm_create_hook.
3256 Its sins are: misleading name, non-modularity and lack of general
3260 Changes since Guile 1.3.4:
3262 * Changes to the distribution
3264 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
3266 We've changed the way we handle generated files in the Guile source
3267 repository. As a result, the procedure for building trees obtained
3268 from the nightly FTP snapshots or via CVS has changed:
3269 - You must have appropriate versions of autoconf, automake, and
3270 libtool installed on your system. See README for info on how to
3271 obtain these programs.
3272 - Before configuring the tree, you must first run the script
3273 `autogen.sh' at the top of the source tree.
3275 The Guile repository used to contain not only source files, written by
3276 humans, but also some generated files, like configure scripts and
3277 Makefile.in files. Even though the contents of these files could be
3278 derived mechanically from other files present, we thought it would
3279 make the tree easier to build if we checked them into CVS.
3281 However, this approach means that minor differences between
3282 developer's installed tools and habits affected the whole team.
3283 So we have removed the generated files from the repository, and
3284 added the autogen.sh script, which will reconstruct them
3288 ** configure now has experimental options to remove support for certain
3291 --disable-arrays omit array and uniform array support
3292 --disable-posix omit posix interfaces
3293 --disable-networking omit networking interfaces
3294 --disable-regex omit regular expression interfaces
3296 These are likely to become separate modules some day.
3298 ** New configure option --enable-debug-freelist
3300 This enables a debugging version of SCM_NEWCELL(), and also registers
3301 an extra primitive, the setter `gc-set-debug-check-freelist!'.
3303 Configure with the --enable-debug-freelist option to enable
3304 the gc-set-debug-check-freelist! primitive, and then use:
3306 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
3307 (gc-set-debug-check-freelist! #f) # turn off checking
3309 Checking of the freelist forces a traversal of the freelist and
3310 a garbage collection before each allocation of a cell. This can
3311 slow down the interpreter dramatically, so the setter should be used to
3312 turn on this extra processing only when necessary.
3314 ** New configure option --enable-debug-malloc
3316 Include code for debugging of calls to scm_must_malloc/realloc/free.
3320 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
3321 2. objects reallocated by scm_must_realloc has been allocated by
3323 3. reallocated objects are reallocated with the same what string
3325 But, most importantly, it records the number of allocated objects of
3326 each kind. This is useful when searching for memory leaks.
3328 A Guile compiled with this option provides the primitive
3329 `malloc-stats' which returns an alist with pairs of kind and the
3330 number of objects of that kind.
3332 ** All includes are now referenced relative to the root directory
3334 Since some users have had problems with mixups between Guile and
3335 system headers, we have decided to always refer to Guile headers via
3336 their parent directories. This essentially creates a "private name
3337 space" for Guile headers. This means that the compiler only is given
3338 -I options for the root build and root source directory.
3340 ** Header files kw.h and genio.h have been removed.
3342 ** The module (ice-9 getopt-gnu-style) has been removed.
3344 ** New module (ice-9 documentation)
3346 Implements the interface to documentation strings associated with
3349 ** New module (ice-9 time)
3351 Provides a macro `time', which displays execution time of a given form.
3353 ** New module (ice-9 history)
3355 Loading this module enables value history in the repl.
3357 * Changes to the stand-alone interpreter
3359 ** New command line option --debug
3361 Start Guile with debugging evaluator and backtraces enabled.
3363 This is useful when debugging your .guile init file or scripts.
3365 ** New help facility
3367 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
3368 (help REGEXP) ditto for objects with names matching REGEXP (a string)
3369 (help 'NAME) gives documentation for NAME, even if it is not an object
3370 (help ,EXPR) gives documentation for object returned by EXPR
3371 (help (my module)) gives module commentary for `(my module)'
3372 (help) gives this text
3374 `help' searches among bindings exported from loaded modules, while
3375 `apropos' searches among bindings visible from the "current" module.
3377 Examples: (help help)
3379 (help "output-string")
3381 ** `help' and `apropos' now prints full module names
3383 ** Dynamic linking now uses libltdl from the libtool package.
3385 The old system dependent code for doing dynamic linking has been
3386 replaced with calls to the libltdl functions which do all the hairy
3389 The major improvement is that you can now directly pass libtool
3390 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
3391 will be able to do the best shared library job you can get, via
3394 The way dynamic libraries are found has changed and is not really
3395 portable across platforms, probably. It is therefore recommended to
3396 use absolute filenames when possible.
3398 If you pass a filename without an extension to `dynamic-link', it will
3399 try a few appropriate ones. Thus, the most platform ignorant way is
3400 to specify a name like "libfoo", without any directories and
3403 ** Guile COOP threads are now compatible with LinuxThreads
3405 Previously, COOP threading wasn't possible in applications linked with
3406 Linux POSIX threads due to their use of the stack pointer to find the
3407 thread context. This has now been fixed with a workaround which uses
3408 the pthreads to allocate the stack.
3410 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
3412 ** Positions of erring expression in scripts
3414 With version 1.3.4, the location of the erring expression in Guile
3415 scipts is no longer automatically reported. (This should have been
3416 documented before the 1.3.4 release.)
3418 You can get this information by enabling recording of positions of
3419 source expressions and running the debugging evaluator. Put this at
3420 the top of your script (or in your "site" file):
3422 (read-enable 'positions)
3423 (debug-enable 'debug)
3425 ** Backtraces in scripts
3427 It is now possible to get backtraces in scripts.
3431 (debug-enable 'debug 'backtrace)
3433 at the top of the script.
3435 (The first options enables the debugging evaluator.
3436 The second enables backtraces.)
3438 ** Part of module system symbol lookup now implemented in C
3440 The eval closure of most modules is now implemented in C. Since this
3441 was one of the bottlenecks for loading speed, Guile now loads code
3442 substantially faster than before.
3444 ** Attempting to get the value of an unbound variable now produces
3445 an exception with a key of 'unbound-variable instead of 'misc-error.
3447 ** The initial default output port is now unbuffered if it's using a
3448 tty device. Previously in this situation it was line-buffered.
3450 ** New hook: after-gc-hook
3452 after-gc-hook takes over the role of gc-thunk. This hook is run at
3453 the first SCM_TICK after a GC. (Thus, the code is run at the same
3454 point during evaluation as signal handlers.)
3456 Note that this hook should be used only for diagnostic and debugging
3457 purposes. It is not certain that it will continue to be well-defined
3458 when this hook is run in the future.
3460 C programmers: Note the new C level hooks scm_before_gc_c_hook,
3461 scm_before_sweep_c_hook, scm_after_gc_c_hook.
3463 ** Improvements to garbage collector
3465 Guile 1.4 has a new policy for triggering heap allocation and
3466 determining the sizes of heap segments. It fixes a number of problems
3469 1. The new policy can handle two separate pools of cells
3470 (2-word/4-word) better. (The old policy would run wild, allocating
3471 more and more memory for certain programs.)
3473 2. The old code would sometimes allocate far too much heap so that the
3474 Guile process became gigantic. The new code avoids this.
3476 3. The old code would sometimes allocate too little so that few cells
3477 were freed at GC so that, in turn, too much time was spent in GC.
3479 4. The old code would often trigger heap allocation several times in a
3480 row. (The new scheme predicts how large the segments needs to be
3481 in order not to need further allocation.)
3483 All in all, the new GC policy will make larger applications more
3486 The new GC scheme also is prepared for POSIX threading. Threads can
3487 allocate private pools of cells ("clusters") with just a single
3488 function call. Allocation of single cells from such a cluster can
3489 then proceed without any need of inter-thread synchronization.
3491 ** New environment variables controlling GC parameters
3493 GUILE_MAX_SEGMENT_SIZE Maximal segment size
3496 Allocation of 2-word cell heaps:
3498 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
3501 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
3502 GC in percent of total heap size
3505 Allocation of 4-word cell heaps
3506 (used for real numbers and misc other objects):
3508 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
3510 (See entry "Way for application to customize GC parameters" under
3511 section "Changes to the scm_ interface" below.)
3513 ** Guile now implements reals using 4-word cells
3515 This speeds up computation with reals. (They were earlier allocated
3516 with `malloc'.) There is still some room for optimizations, however.
3518 ** Some further steps toward POSIX thread support have been taken
3520 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
3521 don't have much effect any longer, and many of them will be removed in
3525 are only handled at the top of the evaluator loop, immediately after
3526 I/O, and in scm_equalp.
3528 *** The GC can allocate thread private pools of pairs.
3530 * Changes to Scheme functions and syntax
3532 ** close-input-port and close-output-port are now R5RS
3534 These procedures have been turned into primitives and have R5RS behaviour.
3536 ** New procedure: simple-format PORT MESSAGE ARG1 ...
3538 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
3539 extended by the more sophisticated version in (ice-9 format)
3541 (simple-format port message . args)
3542 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
3543 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
3544 the escapes are replaced with corresponding members of ARGS:
3545 ~A formats using `display' and ~S formats using `write'.
3546 If DESTINATION is #t, then use the `current-output-port',
3547 if DESTINATION is #f, then return a string containing the formatted text.
3548 Does not add a trailing newline."
3550 ** string-ref: the second argument is no longer optional.
3552 ** string, list->string: no longer accept strings in their arguments,
3553 only characters, for compatibility with R5RS.
3555 ** New procedure: port-closed? PORT
3556 Returns #t if PORT is closed or #f if it is open.
3558 ** Deprecated: list*
3560 The list* functionality is now provided by cons* (SRFI-1 compliant)
3562 ** New procedure: cons* ARG1 ARG2 ... ARGn
3564 Like `list', but the last arg provides the tail of the constructed list,
3565 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
3567 Requires at least one argument. If given one argument, that argument
3568 is returned as result.
3570 This function is called `list*' in some other Schemes and in Common LISP.
3572 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
3574 ** New procedure: object-documentation OBJECT
3576 Returns the documentation string associated with OBJECT. The
3577 procedure uses a caching mechanism so that subsequent lookups are
3580 Exported by (ice-9 documentation).
3582 ** module-name now returns full names of modules
3584 Previously, only the last part of the name was returned (`session' for
3585 `(ice-9 session)'). Ex: `(ice-9 session)'.
3587 * Changes to the gh_ interface
3589 ** Deprecated: gh_int2scmb
3591 Use gh_bool2scm instead.
3593 * Changes to the scm_ interface
3595 ** Guile primitives now carry docstrings!
3597 Thanks to Greg Badros!
3599 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
3601 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
3602 macros and must contain a docstring that is extracted into foo.doc using a new
3603 guile-doc-snarf script (that uses guile-doc-snarf.awk).
3605 However, a major overhaul of these macros is scheduled for the next release of
3608 ** Guile primitives use a new technique for validation of arguments
3610 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
3611 the readability of argument checking.
3613 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
3615 ** New macros: SCM_PACK, SCM_UNPACK
3617 Compose/decompose an SCM value.
3619 The SCM type is now treated as an abstract data type and may be defined as a
3620 long, a void* or as a struct, depending on the architecture and compile time
3621 options. This makes it easier to find several types of bugs, for example when
3622 SCM values are treated as integers without conversion. Values of the SCM type
3623 should be treated as "atomic" values. These macros are used when
3624 composing/decomposing an SCM value, either because you want to access
3625 individual bits, or because you want to treat it as an integer value.
3627 E.g., in order to set bit 7 in an SCM value x, use the expression
3629 SCM_PACK (SCM_UNPACK (x) | 0x80)
3631 ** The name property of hooks is deprecated.
3632 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
3634 You can emulate this feature by using object properties.
3636 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
3637 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
3640 These macros will be removed in a future release of Guile.
3642 ** The following types, functions and macros from numbers.h are deprecated:
3643 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
3644 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
3646 ** Port internals: the rw_random variable in the scm_port structure
3647 must be set to non-zero in any random access port. In recent Guile
3648 releases it was only set for bidirectional random-access ports.
3650 ** Port internals: the seek ptob procedure is now responsible for
3651 resetting the buffers if required. The change was made so that in the
3652 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
3653 the fport and strport ptobs can avoid resetting the buffers,
3654 in particular to avoid discarding unread chars. An existing port
3655 type can be fixed by adding something like the following to the
3656 beginning of the ptob seek procedure:
3658 if (pt->rw_active == SCM_PORT_READ)
3659 scm_end_input (object);
3660 else if (pt->rw_active == SCM_PORT_WRITE)
3661 ptob->flush (object);
3663 although to actually avoid resetting the buffers and discard unread
3664 chars requires further hacking that depends on the characteristics
3667 ** Deprecated functions: scm_fseek, scm_tag
3669 These functions are no longer used and will be removed in a future version.
3671 ** The scm_sysmissing procedure is no longer used in libguile.
3672 Unless it turns out to be unexpectedly useful to somebody, it will be
3673 removed in a future version.
3675 ** The format of error message strings has changed
3677 The two C procedures: scm_display_error and scm_error, as well as the
3678 primitive `scm-error', now use scm_simple_format to do their work.
3679 This means that the message strings of all code must be updated to use
3680 ~A where %s was used before, and ~S where %S was used before.
3682 During the period when there still are a lot of old Guiles out there,
3683 you might want to support both old and new versions of Guile.
3685 There are basically two methods to achieve this. Both methods use
3688 AC_CHECK_FUNCS(scm_simple_format)
3690 in your configure.in.
3692 Method 1: Use the string concatenation features of ANSI C's
3697 #ifdef HAVE_SCM_SIMPLE_FORMAT
3703 Then represent each of your error messages using a preprocessor macro:
3705 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
3709 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
3710 (define make-message string-append)
3712 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
3714 Method 2: Use the oldfmt function found in doc/oldfmt.c.
3718 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
3723 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
3727 ** Deprecated: coop_mutex_init, coop_condition_variable_init
3729 Don't use the functions coop_mutex_init and
3730 coop_condition_variable_init. They will change.
3732 Use scm_mutex_init and scm_cond_init instead.
3734 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
3735 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
3736 COND, as `scm_cond_wait' does, but it also bounds the duration
3737 of the wait. If COND has not been signaled before time ABSTIME,
3738 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
3739 returns the error code `ETIMEDOUT'.
3741 The ABSTIME parameter specifies an absolute time, with the same
3742 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
3743 to 00:00:00 GMT, January 1, 1970.
3745 ** New function: scm_cond_broadcast (scm_cond_t *COND)
3746 `scm_cond_broadcast' restarts all the threads that are waiting
3747 on the condition variable COND. Nothing happens if no threads are
3750 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
3751 `scm_key_create' allocates a new TSD key. The key is stored in
3752 the location pointed to by KEY. There is no limit on the number
3753 of keys allocated at a given time. The value initially associated
3754 with the returned key is `NULL' in all currently executing threads.
3756 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
3757 function associated with the key. When a thread terminates,
3758 DESTR_FUNCTION is called on the value associated with the key in
3759 that thread. The DESTR_FUNCTION is not called if a key is deleted
3760 with `scm_key_delete' or a value is changed with
3761 `scm_setspecific'. The order in which destructor functions are
3762 called at thread termination time is unspecified.
3764 Destructors are not yet implemented.
3766 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
3767 `scm_setspecific' changes the value associated with KEY in the
3768 calling thread, storing the given POINTER instead.
3770 ** New function: scm_getspecific (scm_key_t KEY)
3771 `scm_getspecific' returns the value currently associated with
3772 KEY in the calling thread.
3774 ** New function: scm_key_delete (scm_key_t KEY)
3775 `scm_key_delete' deallocates a TSD key. It does not check
3776 whether non-`NULL' values are associated with that key in the
3777 currently executing threads, nor call the destructor function
3778 associated with the key.
3780 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
3782 Initialize a C level hook HOOK with associated HOOK_DATA and type
3783 TYPE. (See scm_c_hook_run ().)
3785 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
3787 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
3788 is true, add it last, otherwise first. The same FUNC can be added
3789 multiple times if FUNC_DATA differ and vice versa.
3791 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
3793 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
3794 function is only removed if both FUNC and FUNC_DATA matches.
3796 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
3798 Run hook HOOK passing DATA to the hook functions.
3800 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
3801 returned is undefined.
3803 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
3804 returns a non-NULL value. This value is returned as the result of
3805 scm_c_hook_run. If all functions return NULL, NULL is returned.
3807 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
3808 returns a NULL value, and NULL is returned. If all functions returns
3809 a non-NULL value, the last value is returned.
3811 ** New C level GC hooks
3813 Five new C level hooks has been added to the garbage collector.
3815 scm_before_gc_c_hook
3818 are run before locking and after unlocking the heap. The system is
3819 thus in a mode where evaluation can take place. (Except that
3820 scm_before_gc_c_hook must not allocate new cells.)
3822 scm_before_mark_c_hook
3823 scm_before_sweep_c_hook
3824 scm_after_sweep_c_hook
3826 are run when the heap is locked. These are intended for extension of
3827 the GC in a modular fashion. Examples are the weaks and guardians
3830 ** Way for application to customize GC parameters
3832 The application can set up other default values for the GC heap
3833 allocation parameters
3835 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
3836 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
3837 GUILE_MAX_SEGMENT_SIZE,
3841 scm_default_init_heap_size_1, scm_default_min_yield_1,
3842 scm_default_init_heap_size_2, scm_default_min_yield_2,
3843 scm_default_max_segment_size
3845 respectively before callong scm_boot_guile.
3847 (See entry "New environment variables ..." in section
3848 "Changes to the stand-alone interpreter" above.)
3850 ** scm_protect_object/scm_unprotect_object now nest
3852 This means that you can call scm_protect_object multiple times on an
3853 object and count on the object being protected until
3854 scm_unprotect_object has been call the same number of times.
3856 The functions also have better time complexity.
3858 Still, it is usually possible to structure the application in a way
3859 that you don't need to use these functions. For example, if you use a
3860 protected standard Guile list to keep track of live objects rather
3861 than some custom data type, objects will die a natural death when they
3862 are no longer needed.
3864 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
3866 Guile does not provide the float representation for inexact real numbers any
3867 more. Now, only doubles are used to represent inexact real numbers. Further,
3868 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
3869 and scm_tc16_complex, respectively.
3871 ** Removed deprecated type scm_smobfuns
3873 ** Removed deprecated function scm_newsmob
3875 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
3877 There is an ongoing discussion among the developers whether to
3878 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
3879 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
3880 until this issue has been settled.
3882 ** Removed deprecated type tag scm_tc16_kw
3884 ** Added type tag scm_tc16_keyword
3886 (This was introduced already in release 1.3.4 but was not documented
3889 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
3891 * Changes to system call interfaces:
3893 ** The "select" procedure now tests port buffers for the ability to
3894 provide input or accept output. Previously only the underlying file
3895 descriptors were checked.
3897 ** New variable PIPE_BUF: the maximum number of bytes that can be
3898 atomically written to a pipe.
3900 ** If a facility is not available on the system when Guile is
3901 compiled, the corresponding primitive procedure will not be defined.
3902 Previously it would have been defined but would throw a system-error
3903 exception if called. Exception handlers which catch this case may
3904 need minor modification: an error will be thrown with key
3905 'unbound-variable instead of 'system-error. Alternatively it's
3906 now possible to use `defined?' to check whether the facility is
3909 ** Procedures which depend on the timezone should now give the correct
3910 result on systems which cache the TZ environment variable, even if TZ
3911 is changed without calling tzset.
3913 * Changes to the networking interfaces:
3915 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
3916 long integers between network and host format. For now, it's not
3917 particularly convenient to do this kind of thing, but consider:
3919 (define write-network-long
3920 (lambda (value port)
3921 (let ((v (make-uniform-vector 1 1 0)))
3922 (uniform-vector-set! v 0 (htonl value))
3923 (uniform-vector-write v port))))
3925 (define read-network-long
3927 (let ((v (make-uniform-vector 1 1 0)))
3928 (uniform-vector-read! v port)
3929 (ntohl (uniform-vector-ref v 0)))))
3931 ** If inet-aton fails, it now throws an error with key 'misc-error
3932 instead of 'system-error, since errno is not relevant.
3934 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
3935 specific keys instead of 'system-error. The latter is inappropriate
3936 since errno will not have been set. The keys are:
3937 'host-not-found, 'try-again, 'no-recovery and 'no-data.
3939 ** sethostent, setnetent, setprotoent, setservent: now take an
3940 optional argument STAYOPEN, which specifies whether the database
3941 remains open after a database entry is accessed randomly (e.g., using
3942 gethostbyname for the hosts database.) The default is #f. Previously
3946 Changes since Guile 1.3.2:
3948 * Changes to the stand-alone interpreter
3952 An initial version of the Guile debugger written by Chris Hanson has
3953 been added. The debugger is still under development but is included
3954 in the distribution anyway since it is already quite useful.
3960 after an error to enter the debugger. Type `help' inside the debugger
3961 for a description of available commands.
3963 If you prefer to have stack frames numbered and printed in
3964 anti-chronological order and prefer up in the stack to be down on the
3965 screen as is the case in gdb, you can put
3967 (debug-enable 'backwards)
3969 in your .guile startup file. (However, this means that Guile can't
3970 use indentation to indicate stack level.)
3972 The debugger is autoloaded into Guile at the first use.
3974 ** Further enhancements to backtraces
3976 There is a new debug option `width' which controls the maximum width
3977 on the screen of printed stack frames. Fancy printing parameters
3978 ("level" and "length" as in Common LISP) are adaptively adjusted for
3979 each stack frame to give maximum information while still fitting
3980 within the bounds. If the stack frame can't be made to fit by
3981 adjusting parameters, it is simply cut off at the end. This is marked
3984 ** Some modules are now only loaded when the repl is started
3986 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
3987 regex) are now loaded into (guile-user) only if the repl has been
3988 started. The effect is that the startup time for scripts has been
3989 reduced to 30% of what it was previously.
3991 Correctly written scripts load the modules they require at the top of
3992 the file and should not be affected by this change.
3994 ** Hooks are now represented as smobs
3996 * Changes to Scheme functions and syntax
3998 ** Readline support has changed again.
4000 The old (readline-activator) module is gone. Use (ice-9 readline)
4001 instead, which now contains all readline functionality. So the code
4002 to activate readline is now
4004 (use-modules (ice-9 readline))
4007 This should work at any time, including from the guile prompt.
4009 To avoid confusion about the terms of Guile's license, please only
4010 enable readline for your personal use; please don't make it the
4011 default for others. Here is why we make this rather odd-sounding
4014 Guile is normally licensed under a weakened form of the GNU General
4015 Public License, which allows you to link code with Guile without
4016 placing that code under the GPL. This exception is important to some
4019 However, since readline is distributed under the GNU General Public
4020 License, when you link Guile with readline, either statically or
4021 dynamically, you effectively change Guile's license to the strict GPL.
4022 Whenever you link any strictly GPL'd code into Guile, uses of Guile
4023 which are normally permitted become forbidden. This is a rather
4024 non-obvious consequence of the licensing terms.
4026 So, to make sure things remain clear, please let people choose for
4027 themselves whether to link GPL'd libraries like readline with Guile.
4029 ** regexp-substitute/global has changed slightly, but incompatibly.
4031 If you include a function in the item list, the string of the match
4032 object it receives is the same string passed to
4033 regexp-substitute/global, not some suffix of that string.
4034 Correspondingly, the match's positions are relative to the entire
4035 string, not the suffix.
4037 If the regexp can match the empty string, the way matches are chosen
4038 from the string has changed. regexp-substitute/global recognizes the
4039 same set of matches that list-matches does; see below.
4041 ** New function: list-matches REGEXP STRING [FLAGS]
4043 Return a list of match objects, one for every non-overlapping, maximal
4044 match of REGEXP in STRING. The matches appear in left-to-right order.
4045 list-matches only reports matches of the empty string if there are no
4046 other matches which begin on, end at, or include the empty match's
4049 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
4051 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
4053 For each match of REGEXP in STRING, apply PROC to the match object,
4054 and the last value PROC returned, or INIT for the first call. Return
4055 the last value returned by PROC. We apply PROC to the matches as they
4056 appear from left to right.
4058 This function recognizes matches according to the same criteria as
4061 Thus, you could define list-matches like this:
4063 (define (list-matches regexp string . flags)
4064 (reverse! (apply fold-matches regexp string '() cons flags)))
4066 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
4070 *** New function: hook? OBJ
4072 Return #t if OBJ is a hook, otherwise #f.
4074 *** New function: make-hook-with-name NAME [ARITY]
4076 Return a hook with name NAME and arity ARITY. The default value for
4077 ARITY is 0. The only effect of NAME is that it will appear when the
4078 hook object is printed to ease debugging.
4080 *** New function: hook-empty? HOOK
4082 Return #t if HOOK doesn't contain any procedures, otherwise #f.
4084 *** New function: hook->list HOOK
4086 Return a list of the procedures that are called when run-hook is
4089 ** `map' signals an error if its argument lists are not all the same length.
4091 This is the behavior required by R5RS, so this change is really a bug
4092 fix. But it seems to affect a lot of people's code, so we're
4093 mentioning it here anyway.
4095 ** Print-state handling has been made more transparent
4097 Under certain circumstances, ports are represented as a port with an
4098 associated print state. Earlier, this pair was represented as a pair
4099 (see "Some magic has been added to the printer" below). It is now
4100 indistinguishable (almost; see `get-print-state') from a port on the
4103 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
4105 Return a new port with the associated print state PRINT-STATE.
4107 *** New function: get-print-state OUTPUT-PORT
4109 Return the print state associated with this port if it exists,
4110 otherwise return #f.
4112 *** New function: directory-stream? OBJECT
4114 Returns true iff OBJECT is a directory stream --- the sort of object
4115 returned by `opendir'.
4117 ** New function: using-readline?
4119 Return #t if readline is in use in the current repl.
4121 ** structs will be removed in 1.4
4123 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
4124 and use GOOPS objects as the fundamental record type.
4126 * Changes to the scm_ interface
4128 ** structs will be removed in 1.4
4130 The entire current struct interface (struct.c, struct.h) will be
4131 replaced in Guile 1.4. We will merge GOOPS into libguile and use
4132 GOOPS objects as the fundamental record type.
4134 ** The internal representation of subr's has changed
4136 Instead of giving a hint to the subr name, the CAR field of the subr
4137 now contains an index to a subr entry in scm_subr_table.
4139 *** New variable: scm_subr_table
4141 An array of subr entries. A subr entry contains the name, properties
4142 and documentation associated with the subr. The properties and
4143 documentation slots are not yet used.
4145 ** A new scheme for "forwarding" calls to a builtin to a generic function
4147 It is now possible to extend the functionality of some Guile
4148 primitives by letting them defer a call to a GOOPS generic function on
4149 argument mismatch. This means that there is no loss of efficiency in
4154 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
4155 (define-method + ((x <string>) (y <string>))
4156 (string-append x y))
4158 + will still be as efficient as usual in numerical calculations, but
4159 can also be used for concatenating strings.
4161 Who will be the first one to extend Guile's numerical tower to
4162 rationals? :) [OK, there a few other things to fix before this can
4163 be made in a clean way.]
4165 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
4167 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
4169 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
4171 These do the same job as SCM_PROC and SCM_PROC1, but they also define
4172 a variable GENERIC which can be used by the dispatch macros below.
4174 [This is experimental code which may change soon.]
4176 *** New macros for forwarding control to a generic on arg type error
4178 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
4180 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
4182 These correspond to the scm_wta function call, and have the same
4183 behaviour until the user has called the GOOPS primitive
4184 `enable-primitive-generic!'. After that, these macros will apply the
4185 generic function GENERIC to the argument(s) instead of calling
4188 [This is experimental code which may change soon.]
4190 *** New macros for argument testing with generic dispatch
4192 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
4194 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
4196 These correspond to the SCM_ASSERT macro, but will defer control to
4197 GENERIC on error after `enable-primitive-generic!' has been called.
4199 [This is experimental code which may change soon.]
4201 ** New function: SCM scm_eval_body (SCM body, SCM env)
4203 Evaluates the body of a special form.
4205 ** The internal representation of struct's has changed
4207 Previously, four slots were allocated for the procedure(s) of entities
4208 and operators. The motivation for this representation had to do with
4209 the structure of the evaluator, the wish to support tail-recursive
4210 generic functions, and efficiency. Since the generic function
4211 dispatch mechanism has changed, there is no longer a need for such an
4212 expensive representation, and the representation has been simplified.
4214 This should not make any difference for most users.
4216 ** GOOPS support has been cleaned up.
4218 Some code has been moved from eval.c to objects.c and code in both of
4219 these compilation units has been cleaned up and better structured.
4221 *** New functions for applying generic functions
4223 New function: SCM scm_apply_generic (GENERIC, ARGS)
4224 New function: SCM scm_call_generic_0 (GENERIC)
4225 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
4226 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
4227 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
4229 ** Deprecated function: scm_make_named_hook
4231 It is now replaced by:
4233 ** New function: SCM scm_create_hook (const char *name, int arity)
4235 Creates a hook in the same way as make-hook above but also
4236 binds a variable named NAME to it.
4238 This is the typical way of creating a hook from C code.
4240 Currently, the variable is created in the "current" module.
4241 This might change when we get the new module system.
4243 [The behaviour is identical to scm_make_named_hook.]
4247 Changes since Guile 1.3:
4249 * Changes to mailing lists
4251 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
4253 See the README file to find current addresses for all the Guile
4256 * Changes to the distribution
4258 ** Readline support is no longer included with Guile by default.
4260 Based on the different license terms of Guile and Readline, we
4261 concluded that Guile should not *by default* cause the linking of
4262 Readline into an application program. Readline support is now offered
4263 as a separate module, which is linked into an application only when
4264 you explicitly specify it.
4266 Although Guile is GNU software, its distribution terms add a special
4267 exception to the usual GNU General Public License (GPL). Guile's
4268 license includes a clause that allows you to link Guile with non-free
4269 programs. We add this exception so as not to put Guile at a
4270 disadvantage vis-a-vis other extensibility packages that support other
4273 In contrast, the GNU Readline library is distributed under the GNU
4274 General Public License pure and simple. This means that you may not
4275 link Readline, even dynamically, into an application unless it is
4276 distributed under a free software license that is compatible the GPL.
4278 Because of this difference in distribution terms, an application that
4279 can use Guile may not be able to use Readline. Now users will be
4280 explicitly offered two independent decisions about the use of these
4283 You can activate the readline support by issuing
4285 (use-modules (readline-activator))
4288 from your ".guile" file, for example.
4290 * Changes to the stand-alone interpreter
4292 ** All builtins now print as primitives.
4293 Previously builtin procedures not belonging to the fundamental subr
4294 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
4295 Now, they print as #<primitive-procedure NAME>.
4297 ** Backtraces slightly more intelligible.
4298 gsubr-apply and macro transformer application frames no longer appear
4301 * Changes to Scheme functions and syntax
4303 ** Guile now correctly handles internal defines by rewriting them into
4304 their equivalent letrec. Previously, internal defines would
4305 incrementally add to the innermost environment, without checking
4306 whether the restrictions specified in RnRS were met. This lead to the
4307 correct behaviour when these restriction actually were met, but didn't
4308 catch all illegal uses. Such an illegal use could lead to crashes of
4309 the Guile interpreter or or other unwanted results. An example of
4310 incorrect internal defines that made Guile behave erratically:
4322 The problem with this example is that the definition of `c' uses the
4323 value of `b' directly. This confuses the meoization machine of Guile
4324 so that the second call of `b' (this time in a larger environment that
4325 also contains bindings for `c' and `d') refers to the binding of `c'
4326 instead of `a'. You could also make Guile crash with a variation on
4331 (define (b flag) (if flag a 1))
4332 (define c (1+ (b flag)))
4340 From now on, Guile will issue an `Unbound variable: b' error message
4345 A hook contains a list of functions which should be called on
4346 particular occasions in an existing program. Hooks are used for
4349 A window manager might have a hook before-window-map-hook. The window
4350 manager uses the function run-hooks to call all functions stored in
4351 before-window-map-hook each time a window is mapped. The user can
4352 store functions in the hook using add-hook!.
4354 In Guile, hooks are first class objects.
4356 *** New function: make-hook [N_ARGS]
4358 Return a hook for hook functions which can take N_ARGS arguments.
4359 The default value for N_ARGS is 0.
4361 (See also scm_make_named_hook below.)
4363 *** New function: add-hook! HOOK PROC [APPEND_P]
4365 Put PROC at the beginning of the list of functions stored in HOOK.
4366 If APPEND_P is supplied, and non-false, put PROC at the end instead.
4368 PROC must be able to take the number of arguments specified when the
4371 If PROC already exists in HOOK, then remove it first.
4373 *** New function: remove-hook! HOOK PROC
4375 Remove PROC from the list of functions in HOOK.
4377 *** New function: reset-hook! HOOK
4379 Clear the list of hook functions stored in HOOK.
4381 *** New function: run-hook HOOK ARG1 ...
4383 Run all hook functions stored in HOOK with arguments ARG1 ... .
4384 The number of arguments supplied must correspond to the number given
4385 when the hook was created.
4387 ** The function `dynamic-link' now takes optional keyword arguments.
4388 The only keyword argument that is currently defined is `:global
4389 BOOL'. With it, you can control whether the shared library will be
4390 linked in global mode or not. In global mode, the symbols from the
4391 linked library can be used to resolve references from other
4392 dynamically linked libraries. In non-global mode, the linked
4393 library is essentially invisible and can only be accessed via
4394 `dynamic-func', etc. The default is now to link in global mode.
4395 Previously, the default has been non-global mode.
4397 The `#:global' keyword is only effective on platforms that support
4398 the dlopen family of functions.
4400 ** New function `provided?'
4402 - Function: provided? FEATURE
4403 Return true iff FEATURE is supported by this installation of
4404 Guile. FEATURE must be a symbol naming a feature; the global
4405 variable `*features*' is a list of available features.
4407 ** Changes to the module (ice-9 expect):
4409 *** The expect-strings macro now matches `$' in a regular expression
4410 only at a line-break or end-of-file by default. Previously it would
4411 match the end of the string accumulated so far. The old behaviour
4412 can be obtained by setting the variable `expect-strings-exec-flags'
4415 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
4416 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
4417 in a regular expression will still match before a line-break or
4418 end-of-file. The default is `regexp/noteol'.
4420 *** The expect-strings macro now uses a variable
4421 `expect-strings-compile-flags' for the flags to be supplied to
4422 `make-regexp'. The default is `regexp/newline', which was previously
4425 *** The expect macro now supplies two arguments to a match procedure:
4426 the current accumulated string and a flag to indicate whether
4427 end-of-file has been reached. Previously only the string was supplied.
4428 If end-of-file is reached, the match procedure will be called an
4429 additional time with the same accumulated string as the previous call
4430 but with the flag set.
4432 ** New module (ice-9 format), implementing the Common Lisp `format' function.
4434 This code, and the documentation for it that appears here, was
4435 borrowed from SLIB, with minor adaptations for Guile.
4437 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
4438 An almost complete implementation of Common LISP format description
4439 according to the CL reference book `Common LISP' from Guy L.
4440 Steele, Digital Press. Backward compatible to most of the
4441 available Scheme format implementations.
4443 Returns `#t', `#f' or a string; has side effect of printing
4444 according to FORMAT-STRING. If DESTINATION is `#t', the output is
4445 to the current output port and `#t' is returned. If DESTINATION
4446 is `#f', a formatted string is returned as the result of the call.
4447 NEW: If DESTINATION is a string, DESTINATION is regarded as the
4448 format string; FORMAT-STRING is then the first argument and the
4449 output is returned as a string. If DESTINATION is a number, the
4450 output is to the current error port if available by the
4451 implementation. Otherwise DESTINATION must be an output port and
4454 FORMAT-STRING must be a string. In case of a formatting error
4455 format returns `#f' and prints a message on the current output or
4456 error port. Characters are output as if the string were output by
4457 the `display' function with the exception of those prefixed by a
4458 tilde (~). For a detailed description of the FORMAT-STRING syntax
4459 please consult a Common LISP format reference manual. For a test
4460 suite to verify this format implementation load `formatst.scm'.
4461 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
4463 Note: `format' is not reentrant, i.e. only one `format'-call may
4464 be executed at a time.
4467 *** Format Specification (Format version 3.0)
4469 Please consult a Common LISP format reference manual for a detailed
4470 description of the format string syntax. For a demonstration of the
4471 implemented directives see `formatst.scm'.
4473 This implementation supports directive parameters and modifiers (`:'
4474 and `@' characters). Multiple parameters must be separated by a comma
4475 (`,'). Parameters can be numerical parameters (positive or negative),
4476 character parameters (prefixed by a quote character (`''), variable
4477 parameters (`v'), number of rest arguments parameter (`#'), empty and
4478 default parameters. Directive characters are case independent. The
4479 general form of a directive is:
4481 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
4483 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
4485 *** Implemented CL Format Control Directives
4487 Documentation syntax: Uppercase characters represent the
4488 corresponding control directive characters. Lowercase characters
4489 represent control directive parameter descriptions.
4492 Any (print as `display' does).
4496 `~MINCOL,COLINC,MINPAD,PADCHARA'
4500 S-expression (print as `write' does).
4504 `~MINCOL,COLINC,MINPAD,PADCHARS'
4510 print number sign always.
4513 print comma separated.
4515 `~MINCOL,PADCHAR,COMMACHARD'
4521 print number sign always.
4524 print comma separated.
4526 `~MINCOL,PADCHAR,COMMACHARX'
4532 print number sign always.
4535 print comma separated.
4537 `~MINCOL,PADCHAR,COMMACHARO'
4543 print number sign always.
4546 print comma separated.
4548 `~MINCOL,PADCHAR,COMMACHARB'
4553 `~N,MINCOL,PADCHAR,COMMACHARR'
4557 print a number as a Roman numeral.
4560 print a number as an "old fashioned" Roman numeral.
4563 print a number as an ordinal English number.
4566 print a number as a cardinal English number.
4571 prints `y' and `ies'.
4574 as `~P but jumps 1 argument backward.'
4577 as `~@P but jumps 1 argument backward.'
4582 prints a character as the reader can understand it (i.e. `#\'
4586 prints a character as emacs does (eg. `^C' for ASCII 03).
4589 Fixed-format floating-point (prints a flonum like MMM.NNN).
4590 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
4592 If the number is positive a plus sign is printed.
4595 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
4596 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
4598 If the number is positive a plus sign is printed.
4601 General floating-point (prints a flonum either fixed or
4603 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
4605 If the number is positive a plus sign is printed.
4608 Dollars floating-point (prints a flonum in fixed with signs
4610 `~DIGITS,SCALE,WIDTH,PADCHAR$'
4612 If the number is positive a plus sign is printed.
4615 A sign is always printed and appears before the padding.
4618 The sign appears before the padding.
4626 print newline if not at the beginning of the output line.
4628 prints `~&' and then N-1 newlines.
4633 print N page separators.
4643 newline is ignored, white space left.
4646 newline is left, white space ignored.
4651 relative tabulation.
4657 Indirection (expects indirect arguments as a list).
4659 extracts indirect arguments from format arguments.
4662 Case conversion (converts by `string-downcase').
4664 converts by `string-capitalize'.
4667 converts by `string-capitalize-first'.
4670 converts by `string-upcase'.
4673 Argument Jumping (jumps 1 argument forward).
4675 jumps N arguments forward.
4678 jumps 1 argument backward.
4681 jumps N arguments backward.
4684 jumps to the 0th argument.
4687 jumps to the Nth argument (beginning from 0)
4689 `~[STR0~;STR1~;...~;STRN~]'
4690 Conditional Expression (numerical clause conditional).
4692 take argument from N.
4695 true test conditional.
4698 if-else-then conditional.
4704 default clause follows.
4707 Iteration (args come from the next argument (a list)).
4709 at most N iterations.
4712 args from next arg (a list of lists).
4715 args from the rest of arguments.
4718 args from the rest args (lists).
4729 aborts if N <= M <= K
4731 *** Not Implemented CL Format Control Directives
4734 print `#f' as an empty list (see below).
4737 print `#f' as an empty list (see below).
4743 (sorry I don't understand its semantics completely)
4745 *** Extended, Replaced and Additional Control Directives
4747 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
4748 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
4749 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
4750 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
4751 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
4752 COMMAWIDTH is the number of characters between two comma
4756 print a R4RS complex number as `~F~@Fi' with passed parameters for
4760 Pretty print formatting of an argument for scheme code lists.
4766 Flushes the output if format DESTINATION is a port.
4769 Print a `#\space' character
4771 print N `#\space' characters.
4774 Print a `#\tab' character
4776 print N `#\tab' characters.
4779 Takes N as an integer representation for a character. No arguments
4780 are consumed. N is converted to a character by `integer->char'. N
4781 must be a positive decimal number.
4784 Print out readproof. Prints out internal objects represented as
4785 `#<...>' as strings `"#<...>"' so that the format output can always
4786 be processed by `read'.
4789 Print out readproof. Prints out internal objects represented as
4790 `#<...>' as strings `"#<...>"' so that the format output can always
4791 be processed by `read'.
4794 Prints information and a copyright notice on the format
4797 prints format version.
4800 may also print number strings, i.e. passing a number as a string
4801 and format it accordingly.
4803 *** Configuration Variables
4805 The format module exports some configuration variables to suit the
4806 systems and users needs. There should be no modification necessary for
4807 the configuration that comes with Guile. Format detects automatically
4808 if the running scheme system implements floating point numbers and
4811 format:symbol-case-conv
4812 Symbols are converted by `symbol->string' so the case type of the
4813 printed symbols is implementation dependent.
4814 `format:symbol-case-conv' is a one arg closure which is either
4815 `#f' (no conversion), `string-upcase', `string-downcase' or
4816 `string-capitalize'. (default `#f')
4818 format:iobj-case-conv
4819 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
4820 implementation internal objects. (default `#f')
4823 The character prefixing the exponent value in `~E' printing.
4826 *** Compatibility With Other Format Implementations
4832 Downward compatible except for padding support and `~A', `~S',
4833 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
4834 `printf' padding support which is completely replaced by the CL
4835 `format' padding style.
4838 Downward compatible except for `~', which is not documented
4839 (ignores all characters inside the format string up to a newline
4840 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
4841 numerical and variable parameters and `:/@' modifiers in the CL
4845 Downward compatible except for `~A' and `~S' which print in
4846 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
4847 directive parameters or modifiers)).
4850 Downward compatible except for an optional destination parameter:
4851 S2C accepts a format call without a destination which returns a
4852 formatted string. This is equivalent to a #f destination in S2C.
4853 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
4854 parameters or modifiers)).
4857 ** Changes to string-handling functions.
4859 These functions were added to support the (ice-9 format) module, above.
4861 *** New function: string-upcase STRING
4862 *** New function: string-downcase STRING
4864 These are non-destructive versions of the existing string-upcase! and
4865 string-downcase! functions.
4867 *** New function: string-capitalize! STRING
4868 *** New function: string-capitalize STRING
4870 These functions convert the first letter of each word in the string to
4873 (string-capitalize "howdy there")
4876 As with the other functions, string-capitalize! modifies the string in
4877 place, while string-capitalize returns a modified copy of its argument.
4879 *** New function: string-ci->symbol STRING
4881 Return a symbol whose name is STRING, but having the same case as if
4882 the symbol had be read by `read'.
4884 Guile can be configured to be sensitive or insensitive to case
4885 differences in Scheme identifiers. If Guile is case-insensitive, all
4886 symbols are converted to lower case on input. The `string-ci->symbol'
4887 function returns a symbol whose name in STRING, transformed as Guile
4888 would if STRING were input.
4890 *** New function: substring-move! STRING1 START END STRING2 START
4892 Copy the substring of STRING1 from START (inclusive) to END
4893 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
4894 string, and the source and destination areas may overlap; in all
4895 cases, the function behaves as if all the characters were copied
4898 *** Extended functions: substring-move-left! substring-move-right!
4900 These functions now correctly copy arbitrarily overlapping substrings;
4901 they are both synonyms for substring-move!.
4904 ** New module (ice-9 getopt-long), with the function `getopt-long'.
4906 getopt-long is a function for parsing command-line arguments in a
4907 manner consistent with other GNU programs.
4909 (getopt-long ARGS GRAMMAR)
4910 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
4912 ARGS should be a list of strings. Its first element should be the
4913 name of the program; subsequent elements should be the arguments
4914 that were passed to the program on the command line. The
4915 `program-arguments' procedure returns a list of this form.
4917 GRAMMAR is a list of the form:
4918 ((OPTION (PROPERTY VALUE) ...) ...)
4920 Each OPTION should be a symbol. `getopt-long' will accept a
4921 command-line option named `--OPTION'.
4922 Each option can have the following (PROPERTY VALUE) pairs:
4924 (single-char CHAR) --- Accept `-CHAR' as a single-character
4925 equivalent to `--OPTION'. This is how to specify traditional
4927 (required? BOOL) --- If BOOL is true, the option is required.
4928 getopt-long will raise an error if it is not found in ARGS.
4929 (value BOOL) --- If BOOL is #t, the option accepts a value; if
4930 it is #f, it does not; and if it is the symbol
4931 `optional', the option may appear in ARGS with or
4933 (predicate FUNC) --- If the option accepts a value (i.e. you
4934 specified `(value #t)' for this option), then getopt
4935 will apply FUNC to the value, and throw an exception
4936 if it returns #f. FUNC should be a procedure which
4937 accepts a string and returns a boolean value; you may
4938 need to use quasiquotes to get it into GRAMMAR.
4940 The (PROPERTY VALUE) pairs may occur in any order, but each
4941 property may occur only once. By default, options do not have
4942 single-character equivalents, are not required, and do not take
4945 In ARGS, single-character options may be combined, in the usual
4946 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
4947 accepts values, then it must be the last option in the
4948 combination; the value is the next argument. So, for example, using
4949 the following grammar:
4950 ((apples (single-char #\a))
4951 (blimps (single-char #\b) (value #t))
4952 (catalexis (single-char #\c) (value #t)))
4953 the following argument lists would be acceptable:
4954 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
4955 for "blimps" and "catalexis")
4956 ("-ab" "bang" "-c" "couth") (same)
4957 ("-ac" "couth" "-b" "bang") (same)
4958 ("-abc" "couth" "bang") (an error, since `-b' is not the
4959 last option in its combination)
4961 If an option's value is optional, then `getopt-long' decides
4962 whether it has a value by looking at what follows it in ARGS. If
4963 the next element is a string, and it does not appear to be an
4964 option itself, then that string is the option's value.
4966 The value of a long option can appear as the next element in ARGS,
4967 or it can follow the option name, separated by an `=' character.
4968 Thus, using the same grammar as above, the following argument lists
4970 ("--apples" "Braeburn" "--blimps" "Goodyear")
4971 ("--apples=Braeburn" "--blimps" "Goodyear")
4972 ("--blimps" "Goodyear" "--apples=Braeburn")
4974 If the option "--" appears in ARGS, argument parsing stops there;
4975 subsequent arguments are returned as ordinary arguments, even if
4976 they resemble options. So, in the argument list:
4977 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
4978 `getopt-long' will recognize the `apples' option as having the
4979 value "Granny Smith", but it will not recognize the `blimp'
4980 option; it will return the strings "--blimp" and "Goodyear" as
4981 ordinary argument strings.
4983 The `getopt-long' function returns the parsed argument list as an
4984 assocation list, mapping option names --- the symbols from GRAMMAR
4985 --- onto their values, or #t if the option does not accept a value.
4986 Unused options do not appear in the alist.
4988 All arguments that are not the value of any option are returned
4989 as a list, associated with the empty list.
4991 `getopt-long' throws an exception if:
4992 - it finds an unrecognized option in ARGS
4993 - a required option is omitted
4994 - an option that requires an argument doesn't get one
4995 - an option that doesn't accept an argument does get one (this can
4996 only happen using the long option `--opt=value' syntax)
4997 - an option predicate fails
5002 `((lockfile-dir (required? #t)
5005 (predicate ,file-is-directory?))
5006 (verbose (required? #f)
5009 (x-includes (single-char #\x))
5010 (rnet-server (single-char #\y)
5011 (predicate ,string?))))
5013 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
5014 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
5016 => ((() "foo1" "-fred" "foo2" "foo3")
5017 (rnet-server . "lamprod")
5018 (x-includes . "/usr/include")
5019 (lockfile-dir . "/tmp")
5022 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
5024 It will be removed in a few releases.
5026 ** New syntax: lambda*
5027 ** New syntax: define*
5028 ** New syntax: define*-public
5029 ** New syntax: defmacro*
5030 ** New syntax: defmacro*-public
5031 Guile now supports optional arguments.
5033 `lambda*', `define*', `define*-public', `defmacro*' and
5034 `defmacro*-public' are identical to the non-* versions except that
5035 they use an extended type of parameter list that has the following BNF
5036 syntax (parentheses are literal, square brackets indicate grouping,
5037 and `*', `+' and `?' have the usual meaning):
5039 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
5040 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
5041 [[#&rest identifier]|[. identifier]]? ) | [identifier]
5043 ext-var-decl ::= identifier | ( identifier expression )
5045 The semantics are best illustrated with the following documentation
5046 and examples for `lambda*':
5049 lambda extended for optional and keyword arguments
5051 lambda* creates a procedure that takes optional arguments. These
5052 are specified by putting them inside brackets at the end of the
5053 paramater list, but before any dotted rest argument. For example,
5054 (lambda* (a b #&optional c d . e) '())
5055 creates a procedure with fixed arguments a and b, optional arguments c
5056 and d, and rest argument e. If the optional arguments are omitted
5057 in a call, the variables for them are unbound in the procedure. This
5058 can be checked with the bound? macro.
5060 lambda* can also take keyword arguments. For example, a procedure
5062 (lambda* (#&key xyzzy larch) '())
5063 can be called with any of the argument lists (#:xyzzy 11)
5064 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
5065 are given as keywords are bound to values.
5067 Optional and keyword arguments can also be given default values
5068 which they take on when they are not present in a call, by giving a
5069 two-item list in place of an optional argument, for example in:
5070 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
5071 foo is a fixed argument, bar is an optional argument with default
5072 value 42, and baz is a keyword argument with default value 73.
5073 Default value expressions are not evaluated unless they are needed
5074 and until the procedure is called.
5076 lambda* now supports two more special parameter list keywords.
5078 lambda*-defined procedures now throw an error by default if a
5079 keyword other than one of those specified is found in the actual
5080 passed arguments. However, specifying #&allow-other-keys
5081 immediately after the kyword argument declarations restores the
5082 previous behavior of ignoring unknown keywords. lambda* also now
5083 guarantees that if the same keyword is passed more than once, the
5084 last one passed is the one that takes effect. For example,
5085 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
5086 #:heads 37 #:tails 42 #:heads 99)
5087 would result in (99 47) being displayed.
5089 #&rest is also now provided as a synonym for the dotted syntax rest
5090 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
5091 all respects to lambda*. This is provided for more similarity to DSSSL,
5092 MIT-Scheme and Kawa among others, as well as for refugees from other
5095 Further documentation may be found in the optargs.scm file itself.
5097 The optional argument module also exports the macros `let-optional',
5098 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
5099 are not documented here because they may be removed in the future, but
5100 full documentation is still available in optargs.scm.
5102 ** New syntax: and-let*
5103 Guile now supports the `and-let*' form, described in the draft SRFI-2.
5105 Syntax: (land* (<clause> ...) <body> ...)
5106 Each <clause> should have one of the following forms:
5107 (<variable> <expression>)
5110 Each <variable> or <bound-variable> should be an identifier. Each
5111 <expression> should be a valid expression. The <body> should be a
5112 possibly empty sequence of expressions, like the <body> of a
5115 Semantics: A LAND* expression is evaluated by evaluating the
5116 <expression> or <bound-variable> of each of the <clause>s from
5117 left to right. The value of the first <expression> or
5118 <bound-variable> that evaluates to a false value is returned; the
5119 remaining <expression>s and <bound-variable>s are not evaluated.
5120 The <body> forms are evaluated iff all the <expression>s and
5121 <bound-variable>s evaluate to true values.
5123 The <expression>s and the <body> are evaluated in an environment
5124 binding each <variable> of the preceding (<variable> <expression>)
5125 clauses to the value of the <expression>. Later bindings
5126 shadow earlier bindings.
5128 Guile's and-let* macro was contributed by Michael Livshin.
5130 ** New sorting functions
5132 *** New function: sorted? SEQUENCE LESS?
5133 Returns `#t' when the sequence argument is in non-decreasing order
5134 according to LESS? (that is, there is no adjacent pair `... x y
5135 ...' for which `(less? y x)').
5137 Returns `#f' when the sequence contains at least one out-of-order
5138 pair. It is an error if the sequence is neither a list nor a
5141 *** New function: merge LIST1 LIST2 LESS?
5142 LIST1 and LIST2 are sorted lists.
5143 Returns the sorted list of all elements in LIST1 and LIST2.
5145 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
5146 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
5147 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
5148 (Here "<" should read "comes before".)
5150 *** New procedure: merge! LIST1 LIST2 LESS?
5151 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
5152 the result. If the code is compiled, and LESS? constructs no new
5153 pairs, no pairs at all will be allocated. The first pair of the
5154 result will be either the first pair of LIST1 or the first pair of
5157 *** New function: sort SEQUENCE LESS?
5158 Accepts either a list or a vector, and returns a new sequence
5159 which is sorted. The new sequence is the same type as the input.
5160 Always `(sorted? (sort sequence less?) less?)'. The original
5161 sequence is not altered in any way. The new sequence shares its
5162 elements with the old one; no elements are copied.
5164 *** New procedure: sort! SEQUENCE LESS
5165 Returns its sorted result in the original boxes. No new storage is
5166 allocated at all. Proper usage: (set! slist (sort! slist <))
5168 *** New function: stable-sort SEQUENCE LESS?
5169 Similar to `sort' but stable. That is, if "equal" elements are
5170 ordered a < b in the original sequence, they will have the same order
5173 *** New function: stable-sort! SEQUENCE LESS?
5174 Similar to `sort!' but stable.
5175 Uses temporary storage when sorting vectors.
5177 *** New functions: sort-list, sort-list!
5178 Added for compatibility with scsh.
5180 ** New built-in random number support
5182 *** New function: random N [STATE]
5183 Accepts a positive integer or real N and returns a number of the
5184 same type between zero (inclusive) and N (exclusive). The values
5185 returned have a uniform distribution.
5187 The optional argument STATE must be of the type produced by
5188 `copy-random-state' or `seed->random-state'. It defaults to the value
5189 of the variable `*random-state*'. This object is used to maintain the
5190 state of the pseudo-random-number generator and is altered as a side
5191 effect of the `random' operation.
5193 *** New variable: *random-state*
5194 Holds a data structure that encodes the internal state of the
5195 random-number generator that `random' uses by default. The nature
5196 of this data structure is implementation-dependent. It may be
5197 printed out and successfully read back in, but may or may not
5198 function correctly as a random-number state object in another
5201 *** New function: copy-random-state [STATE]
5202 Returns a new object of type suitable for use as the value of the
5203 variable `*random-state*' and as a second argument to `random'.
5204 If argument STATE is given, a copy of it is returned. Otherwise a
5205 copy of `*random-state*' is returned.
5207 *** New function: seed->random-state SEED
5208 Returns a new object of type suitable for use as the value of the
5209 variable `*random-state*' and as a second argument to `random'.
5210 SEED is a string or a number. A new state is generated and
5211 initialized using SEED.
5213 *** New function: random:uniform [STATE]
5214 Returns an uniformly distributed inexact real random number in the
5215 range between 0 and 1.
5217 *** New procedure: random:solid-sphere! VECT [STATE]
5218 Fills VECT with inexact real random numbers the sum of whose
5219 squares is less than 1.0. Thinking of VECT as coordinates in
5220 space of dimension N = `(vector-length VECT)', the coordinates are
5221 uniformly distributed within the unit N-shere. The sum of the
5222 squares of the numbers is returned. VECT can be either a vector
5223 or a uniform vector of doubles.
5225 *** New procedure: random:hollow-sphere! VECT [STATE]
5226 Fills VECT with inexact real random numbers the sum of whose squares
5227 is equal to 1.0. Thinking of VECT as coordinates in space of
5228 dimension n = `(vector-length VECT)', the coordinates are uniformly
5229 distributed over the surface of the unit n-shere. VECT can be either
5230 a vector or a uniform vector of doubles.
5232 *** New function: random:normal [STATE]
5233 Returns an inexact real in a normal distribution with mean 0 and
5234 standard deviation 1. For a normal distribution with mean M and
5235 standard deviation D use `(+ M (* D (random:normal)))'.
5237 *** New procedure: random:normal-vector! VECT [STATE]
5238 Fills VECT with inexact real random numbers which are independent and
5239 standard normally distributed (i.e., with mean 0 and variance 1).
5240 VECT can be either a vector or a uniform vector of doubles.
5242 *** New function: random:exp STATE
5243 Returns an inexact real in an exponential distribution with mean 1.
5244 For an exponential distribution with mean U use (* U (random:exp)).
5246 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
5248 These functions now operate on numbers in the range of a C unsigned
5251 These functions used to operate on numbers in the range of a C signed
5252 long; however, this seems inappropriate, because Guile integers don't
5255 ** New function: make-guardian
5256 This is an implementation of guardians as described in
5257 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
5258 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
5259 Programming Language Design and Implementation, June 1993
5260 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
5262 ** New functions: delq1!, delv1!, delete1!
5263 These procedures behave similar to delq! and friends but delete only
5264 one object if at all.
5266 ** New function: unread-string STRING PORT
5267 Unread STRING to PORT, that is, push it back onto the port so that
5268 next read operation will work on the pushed back characters.
5270 ** unread-char can now be called multiple times
5271 If unread-char is called multiple times, the unread characters will be
5272 read again in last-in first-out order.
5274 ** the procedures uniform-array-read! and uniform-array-write! now
5275 work on any kind of port, not just ports which are open on a file.
5277 ** Now 'l' in a port mode requests line buffering.
5279 ** The procedure truncate-file now works on string ports as well
5280 as file ports. If the size argument is omitted, the current
5281 file position is used.
5283 ** new procedure: seek PORT/FDES OFFSET WHENCE
5284 The arguments are the same as for the old fseek procedure, but it
5285 works on string ports as well as random-access file ports.
5287 ** the fseek procedure now works on string ports, since it has been
5288 redefined using seek.
5290 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
5291 size is not supplied.
5293 ** the newline procedure no longer flushes the port if it's not
5294 line-buffered: previously it did if it was the current output port.
5296 ** open-pipe and close-pipe are no longer primitive procedures, but
5297 an emulation can be obtained using `(use-modules (ice-9 popen))'.
5299 ** the freopen procedure has been removed.
5301 ** new procedure: drain-input PORT
5302 Drains PORT's read buffers (including any pushed-back characters)
5303 and returns the contents as a single string.
5305 ** New function: map-in-order PROC LIST1 LIST2 ...
5306 Version of `map' which guarantees that the procedure is applied to the
5307 lists in serial order.
5309 ** Renamed `serial-array-copy!' and `serial-array-map!' to
5310 `array-copy-in-order!' and `array-map-in-order!'. The old names are
5311 now obsolete and will go away in release 1.5.
5313 ** New syntax: collect BODY1 ...
5314 Version of `begin' which returns a list of the results of the body
5315 forms instead of the result of the last body form. In contrast to
5316 `begin', `collect' allows an empty body.
5318 ** New functions: read-history FILENAME, write-history FILENAME
5319 Read/write command line history from/to file. Returns #t on success
5320 and #f if an error occured.
5322 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
5324 These procedures return a list of definitions available in the specified
5325 argument, a relative module reference. In the case of no argument,
5326 `(current-module)' is now consulted for definitions to return, instead
5327 of simply returning #f, the former behavior.
5329 ** The #/ syntax for lists is no longer supported.
5331 Earlier versions of Scheme accepted this syntax, but printed a
5334 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
5336 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
5339 * Changes to the gh_ interface
5343 Now takes a second argument which is the result array. If this
5344 pointer is NULL, a new array is malloced (the old behaviour).
5346 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
5347 gh_scm2shorts, gh_scm2longs, gh_scm2floats
5351 * Changes to the scm_ interface
5353 ** Function: scm_make_named_hook (char* name, int n_args)
5355 Creates a hook in the same way as make-hook above but also
5356 binds a variable named NAME to it.
5358 This is the typical way of creating a hook from C code.
5360 Currently, the variable is created in the "current" module. This
5361 might change when we get the new module system.
5363 ** The smob interface
5365 The interface for creating smobs has changed. For documentation, see
5366 data-rep.info (made from guile-core/doc/data-rep.texi).
5368 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
5370 >>> This function will be removed in 1.3.4. <<<
5374 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
5375 This function adds a new smob type, named NAME, with instance size
5376 SIZE to the system. The return value is a tag that is used in
5377 creating instances of the type. If SIZE is 0, then no memory will
5378 be allocated when instances of the smob are created, and nothing
5379 will be freed by the default free function.
5381 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
5382 This function sets the smob marking procedure for the smob type
5383 specified by the tag TC. TC is the tag returned by
5384 `scm_make_smob_type'.
5386 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
5387 This function sets the smob freeing procedure for the smob type
5388 specified by the tag TC. TC is the tag returned by
5389 `scm_make_smob_type'.
5391 *** Function: void scm_set_smob_print (tc, print)
5393 - Function: void scm_set_smob_print (long tc,
5394 scm_sizet (*print) (SCM,
5398 This function sets the smob printing procedure for the smob type
5399 specified by the tag TC. TC is the tag returned by
5400 `scm_make_smob_type'.
5402 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
5403 This function sets the smob equality-testing predicate for the
5404 smob type specified by the tag TC. TC is the tag returned by
5405 `scm_make_smob_type'.
5407 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
5408 Make VALUE contain a smob instance of the type with type code TC and
5409 smob data DATA. VALUE must be previously declared as C type `SCM'.
5411 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
5412 This macro expands to a block of code that creates a smob instance
5413 of the type with type code TC and smob data DATA, and returns that
5414 `SCM' value. It should be the last piece of code in a block.
5416 ** The interfaces for using I/O ports and implementing port types
5417 (ptobs) have changed significantly. The new interface is based on
5418 shared access to buffers and a new set of ptob procedures.
5420 *** scm_newptob has been removed
5424 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
5426 - Function: SCM scm_make_port_type (char *type_name,
5427 int (*fill_buffer) (SCM port),
5428 void (*write_flush) (SCM port));
5430 Similarly to the new smob interface, there is a set of function
5431 setters by which the user can customize the behaviour of his port
5432 type. See ports.h (scm_set_port_XXX).
5434 ** scm_strport_to_string: New function: creates a new string from
5435 a string port's buffer.
5437 ** Plug in interface for random number generators
5438 The variable `scm_the_rng' in random.c contains a value and three
5439 function pointers which together define the current random number
5440 generator being used by the Scheme level interface and the random
5441 number library functions.
5443 The user is free to replace the default generator with the generator
5446 *** Variable: size_t scm_the_rng.rstate_size
5447 The size of the random state type used by the current RNG
5450 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
5451 Given the random STATE, return 32 random bits.
5453 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
5454 Seed random state STATE using string S of length N.
5456 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
5457 Given random state STATE, return a malloced copy.
5460 The default RNG is the MWC (Multiply With Carry) random number
5461 generator described by George Marsaglia at the Department of
5462 Statistics and Supercomputer Computations Research Institute, The
5463 Florida State University (http://stat.fsu.edu/~geo).
5465 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
5466 passes all tests in the DIEHARD test suite
5467 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
5468 costs one multiply and one add on platforms which either supports long
5469 longs (gcc does this on most systems) or have 64 bit longs. The cost
5470 is four multiply on other systems but this can be optimized by writing
5471 scm_i_uniform32 in assembler.
5473 These functions are provided through the scm_the_rng interface for use
5474 by libguile and the application.
5476 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
5477 Given the random STATE, return 32 random bits.
5478 Don't use this function directly. Instead go through the plugin
5479 interface (see "Plug in interface" above).
5481 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
5482 Initialize STATE using SEED of length N.
5484 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
5485 Return a malloc:ed copy of STATE. This function can easily be re-used
5486 in the interfaces to other RNGs.
5488 ** Random number library functions
5489 These functions use the current RNG through the scm_the_rng interface.
5490 It might be a good idea to use these functions from your C code so
5491 that only one random generator is used by all code in your program.
5493 The default random state is stored in:
5495 *** Variable: SCM scm_var_random_state
5496 Contains the vcell of the Scheme variable "*random-state*" which is
5497 used as default state by all random number functions in the Scheme
5502 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
5504 *** Function: scm_rstate *scm_c_default_rstate (void)
5505 This is a convenience function which returns the value of
5506 scm_var_random_state. An error message is generated if this value
5507 isn't a random state.
5509 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
5510 Make a new random state from the string SEED of length LENGTH.
5512 It is generally not a good idea to use multiple random states in a
5513 program. While subsequent random numbers generated from one random
5514 state are guaranteed to be reasonably independent, there is no such
5515 guarantee for numbers generated from different random states.
5517 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
5518 Return 32 random bits.
5520 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
5521 Return a sample from the uniform(0,1) distribution.
5523 *** Function: double scm_c_normal01 (scm_rstate *STATE)
5524 Return a sample from the normal(0,1) distribution.
5526 *** Function: double scm_c_exp1 (scm_rstate *STATE)
5527 Return a sample from the exp(1) distribution.
5529 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
5530 Return a sample from the discrete uniform(0,M) distribution.
5532 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
5533 Return a sample from the discrete uniform(0,M) distribution.
5534 M must be a bignum object. The returned value may be an INUM.
5538 Changes in Guile 1.3 (released Monday, October 19, 1998):
5540 * Changes to the distribution
5542 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
5543 To avoid conflicts, programs should name environment variables after
5544 themselves, except when there's a common practice establishing some
5547 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
5548 giving the former precedence, and printing a warning message if the
5549 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
5551 ** The header files related to multi-byte characters have been removed.
5552 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
5553 which referred to these explicitly will probably need to be rewritten,
5554 since the support for the variant string types has been removed; see
5557 ** The header files append.h and sequences.h have been removed. These
5558 files implemented non-R4RS operations which would encourage
5559 non-portable programming style and less easy-to-read code.
5561 * Changes to the stand-alone interpreter
5563 ** New procedures have been added to implement a "batch mode":
5565 *** Function: batch-mode?
5567 Returns a boolean indicating whether the interpreter is in batch
5570 *** Function: set-batch-mode?! ARG
5572 If ARG is true, switches the interpreter to batch mode. The `#f'
5573 case has not been implemented.
5575 ** Guile now provides full command-line editing, when run interactively.
5576 To use this feature, you must have the readline library installed.
5577 The Guile build process will notice it, and automatically include
5580 The readline library is available via anonymous FTP from any GNU
5581 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
5583 ** the-last-stack is now a fluid.
5585 * Changes to the procedure for linking libguile with your programs
5587 ** You can now use the `guile-config' utility to build programs that use Guile.
5589 Guile now includes a command-line utility called `guile-config', which
5590 can provide information about how to compile and link programs that
5593 *** `guile-config compile' prints any C compiler flags needed to use Guile.
5594 You should include this command's output on the command line you use
5595 to compile C or C++ code that #includes the Guile header files. It's
5596 usually just a `-I' flag to help the compiler find the Guile headers.
5599 *** `guile-config link' prints any linker flags necessary to link with Guile.
5601 This command writes to its standard output a list of flags which you
5602 must pass to the linker to link your code against the Guile library.
5603 The flags include '-lguile' itself, any other libraries the Guile
5604 library depends upon, and any `-L' flags needed to help the linker
5605 find those libraries.
5607 For example, here is a Makefile rule that builds a program named 'foo'
5608 from the object files ${FOO_OBJECTS}, and links them against Guile:
5611 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
5613 Previous Guile releases recommended that you use autoconf to detect
5614 which of a predefined set of libraries were present on your system.
5615 It is more robust to use `guile-config', since it records exactly which
5616 libraries the installed Guile library requires.
5618 This was originally called `build-guile', but was renamed to
5619 `guile-config' before Guile 1.3 was released, to be consistent with
5620 the analogous script for the GTK+ GUI toolkit, which is called
5624 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
5626 If you are using the GNU autoconf package to configure your program,
5627 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
5628 (described above) and gather the necessary values for use in your
5631 The GUILE_FLAGS macro expands to configure script code which runs the
5632 `guile-config' script, to find out where Guile's header files and
5633 libraries are installed. It sets two variables, marked for
5634 substitution, as by AC_SUBST.
5636 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
5637 code that uses Guile header files. This is almost always just a
5640 GUILE_LDFLAGS --- flags to pass to the linker to link a
5641 program against Guile. This includes `-lguile' for the Guile
5642 library itself, any libraries that Guile itself requires (like
5643 -lqthreads), and so on. It may also include a -L flag to tell the
5644 compiler where to find the libraries.
5646 GUILE_FLAGS is defined in the file guile.m4, in the top-level
5647 directory of the Guile distribution. You can copy it into your
5648 package's aclocal.m4 file, and then use it in your configure.in file.
5650 If you are using the `aclocal' program, distributed with GNU automake,
5651 to maintain your aclocal.m4 file, the Guile installation process
5652 installs guile.m4 where aclocal will find it. All you need to do is
5653 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
5654 this will copy the definition of GUILE_FLAGS into your aclocal.m4
5658 * Changes to Scheme functions and syntax
5660 ** Multi-byte strings have been removed, as have multi-byte and wide
5661 ports. We felt that these were the wrong approach to
5662 internationalization support.
5664 ** New function: readline [PROMPT]
5665 Read a line from the terminal, and allow the user to edit it,
5666 prompting with PROMPT. READLINE provides a large set of Emacs-like
5667 editing commands, lets the user recall previously typed lines, and
5668 works on almost every kind of terminal, including dumb terminals.
5670 READLINE assumes that the cursor is at the beginning of the line when
5671 it is invoked. Thus, you can't print a prompt yourself, and then call
5672 READLINE; you need to package up your prompt as a string, pass it to
5673 the function, and let READLINE print the prompt itself. This is
5674 because READLINE needs to know the prompt's screen width.
5676 For Guile to provide this function, you must have the readline
5677 library, version 2.1 or later, installed on your system. Readline is
5678 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
5679 any GNU mirror site.
5681 See also ADD-HISTORY function.
5683 ** New function: add-history STRING
5684 Add STRING as the most recent line in the history used by the READLINE
5685 command. READLINE does not add lines to the history itself; you must
5686 call ADD-HISTORY to make previous input available to the user.
5688 ** The behavior of the read-line function has changed.
5690 This function now uses standard C library functions to read the line,
5691 for speed. This means that it doesn not respect the value of
5692 scm-line-incrementors; it assumes that lines are delimited with
5695 (Note that this is read-line, the function that reads a line of text
5696 from a port, not readline, the function that reads a line from a
5697 terminal, providing full editing capabilities.)
5699 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
5701 This module provides some simple argument parsing. It exports one
5704 Function: getopt-gnu-style ARG-LS
5705 Parse a list of program arguments into an alist of option
5708 Each item in the list of program arguments is examined to see if
5709 it meets the syntax of a GNU long-named option. An argument like
5710 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
5711 returned alist, where MUMBLE is a keyword object with the same
5712 name as the argument. An argument like `--MUMBLE=FROB' produces
5713 an element of the form (MUMBLE . FROB), where FROB is a string.
5715 As a special case, the returned alist also contains a pair whose
5716 car is the symbol `rest'. The cdr of this pair is a list
5717 containing all the items in the argument list that are not options
5718 of the form mentioned above.
5720 The argument `--' is treated specially: all items in the argument
5721 list appearing after such an argument are not examined, and are
5722 returned in the special `rest' list.
5724 This function does not parse normal single-character switches.
5725 You will need to parse them out of the `rest' list yourself.
5727 ** The read syntax for byte vectors and short vectors has changed.
5729 Instead of #bytes(...), write #y(...).
5731 Instead of #short(...), write #h(...).
5733 This may seem nutty, but, like the other uniform vectors, byte vectors
5734 and short vectors want to have the same print and read syntax (and,
5735 more basic, want to have read syntax!). Changing the read syntax to
5736 use multiple characters after the hash sign breaks with the
5737 conventions used in R5RS and the conventions used for the other
5738 uniform vectors. It also introduces complexity in the current reader,
5739 both on the C and Scheme levels. (The Right solution is probably to
5740 change the syntax and prototypes for uniform vectors entirely.)
5743 ** The new module (ice-9 session) provides useful interactive functions.
5745 *** New procedure: (apropos REGEXP OPTION ...)
5747 Display a list of top-level variables whose names match REGEXP, and
5748 the modules they are imported from. Each OPTION should be one of the
5751 value --- Show the value of each matching variable.
5752 shadow --- Show bindings shadowed by subsequently imported modules.
5753 full --- Same as both `shadow' and `value'.
5757 guile> (apropos "trace" 'full)
5758 debug: trace #<procedure trace args>
5759 debug: untrace #<procedure untrace args>
5760 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
5761 the-scm-module: before-backtrace-hook ()
5762 the-scm-module: backtrace #<primitive-procedure backtrace>
5763 the-scm-module: after-backtrace-hook ()
5764 the-scm-module: has-shown-backtrace-hint? #f
5767 ** There are new functions and syntax for working with macros.
5769 Guile implements macros as a special object type. Any variable whose
5770 top-level binding is a macro object acts as a macro. The macro object
5771 specifies how the expression should be transformed before evaluation.
5773 *** Macro objects now print in a reasonable way, resembling procedures.
5775 *** New function: (macro? OBJ)
5776 True iff OBJ is a macro object.
5778 *** New function: (primitive-macro? OBJ)
5779 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
5780 macro transformers, implemented in eval.c rather than Scheme code.
5782 Why do we have this function?
5783 - For symmetry with procedure? and primitive-procedure?,
5784 - to allow custom print procedures to tell whether a macro is
5785 primitive, and display it differently, and
5786 - to allow compilers and user-written evaluators to distinguish
5787 builtin special forms from user-defined ones, which could be
5790 *** New function: (macro-type OBJ)
5791 Return a value indicating what kind of macro OBJ is. Possible return
5794 The symbol `syntax' --- a macro created by procedure->syntax.
5795 The symbol `macro' --- a macro created by procedure->macro.
5796 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
5797 The boolean #f --- if OBJ is not a macro object.
5799 *** New function: (macro-name MACRO)
5800 Return the name of the macro object MACRO's procedure, as returned by
5803 *** New function: (macro-transformer MACRO)
5804 Return the transformer procedure for MACRO.
5806 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
5808 Specify a new macro expander to use in the current module. Each
5809 MODULE is a module name, with the same meaning as in the `use-modules'
5810 form; each named module's exported bindings are added to the current
5811 top-level environment. TRANSFORMER is an expression evaluated in the
5812 resulting environment which must yield a procedure to use as the
5813 module's eval transformer: every expression evaluated in this module
5814 is passed to this function, and the result passed to the Guile
5817 *** macro-eval! is removed. Use local-eval instead.
5819 ** Some magic has been added to the printer to better handle user
5820 written printing routines (like record printers, closure printers).
5822 The problem is that these user written routines must have access to
5823 the current `print-state' to be able to handle fancy things like
5824 detection of circular references. These print-states have to be
5825 passed to the builtin printing routines (display, write, etc) to
5826 properly continue the print chain.
5828 We didn't want to change all existing print code so that it
5829 explicitly passes thru a print state in addition to a port. Instead,
5830 we extented the possible values that the builtin printing routines
5831 accept as a `port'. In addition to a normal port, they now also take
5832 a pair of a normal port and a print-state. Printing will go to the
5833 port and the print-state will be used to control the detection of
5834 circular references, etc. If the builtin function does not care for a
5835 print-state, it is simply ignored.
5837 User written callbacks are now called with such a pair as their
5838 `port', but because every function now accepts this pair as a PORT
5839 argument, you don't have to worry about that. In fact, it is probably
5840 safest to not check for these pairs.
5842 However, it is sometimes necessary to continue a print chain on a
5843 different port, for example to get a intermediate string
5844 representation of the printed value, mangle that string somehow, and
5845 then to finally print the mangled string. Use the new function
5847 inherit-print-state OLD-PORT NEW-PORT
5849 for this. It constructs a new `port' that prints to NEW-PORT but
5850 inherits the print-state of OLD-PORT.
5852 ** struct-vtable-offset renamed to vtable-offset-user
5854 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
5856 ** There is now a third optional argument to make-vtable-vtable
5857 (and fourth to make-struct) when constructing new types (vtables).
5858 This argument initializes field vtable-index-printer of the vtable.
5860 ** The detection of circular references has been extended to structs.
5861 That is, a structure that -- in the process of being printed -- prints
5862 itself does not lead to infinite recursion.
5864 ** There is now some basic support for fluids. Please read
5865 "libguile/fluid.h" to find out more. It is accessible from Scheme with
5866 the following functions and macros:
5868 Function: make-fluid
5870 Create a new fluid object. Fluids are not special variables or
5871 some other extension to the semantics of Scheme, but rather
5872 ordinary Scheme objects. You can store them into variables (that
5873 are still lexically scoped, of course) or into any other place you
5874 like. Every fluid has a initial value of `#f'.
5876 Function: fluid? OBJ
5878 Test whether OBJ is a fluid.
5880 Function: fluid-ref FLUID
5881 Function: fluid-set! FLUID VAL
5883 Access/modify the fluid FLUID. Modifications are only visible
5884 within the current dynamic root (that includes threads).
5886 Function: with-fluids* FLUIDS VALUES THUNK
5888 FLUIDS is a list of fluids and VALUES a corresponding list of
5889 values for these fluids. Before THUNK gets called the values are
5890 installed in the fluids and the old values of the fluids are
5891 saved in the VALUES list. When the flow of control leaves THUNK
5892 or reenters it, the values get swapped again. You might think of
5893 this as a `safe-fluid-excursion'. Note that the VALUES list is
5894 modified by `with-fluids*'.
5896 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
5898 The same as `with-fluids*' but with a different syntax. It looks
5899 just like `let', but both FLUID and VALUE are evaluated. Remember,
5900 fluids are not special variables but ordinary objects. FLUID
5901 should evaluate to a fluid.
5903 ** Changes to system call interfaces:
5905 *** close-port, close-input-port and close-output-port now return a
5906 boolean instead of an `unspecified' object. #t means that the port
5907 was successfully closed, while #f means it was already closed. It is
5908 also now possible for these procedures to raise an exception if an
5909 error occurs (some errors from write can be delayed until close.)
5911 *** the first argument to chmod, fcntl, ftell and fseek can now be a
5914 *** the third argument to fcntl is now optional.
5916 *** the first argument to chown can now be a file descriptor or a port.
5918 *** the argument to stat can now be a port.
5920 *** The following new procedures have been added (most use scsh
5923 *** procedure: close PORT/FD
5924 Similar to close-port (*note close-port: Closing Ports.), but also
5925 works on file descriptors. A side effect of closing a file
5926 descriptor is that any ports using that file descriptor are moved
5927 to a different file descriptor and have their revealed counts set
5930 *** procedure: port->fdes PORT
5931 Returns the integer file descriptor underlying PORT. As a side
5932 effect the revealed count of PORT is incremented.
5934 *** procedure: fdes->ports FDES
5935 Returns a list of existing ports which have FDES as an underlying
5936 file descriptor, without changing their revealed counts.
5938 *** procedure: fdes->inport FDES
5939 Returns an existing input port which has FDES as its underlying
5940 file descriptor, if one exists, and increments its revealed count.
5941 Otherwise, returns a new input port with a revealed count of 1.
5943 *** procedure: fdes->outport FDES
5944 Returns an existing output port which has FDES as its underlying
5945 file descriptor, if one exists, and increments its revealed count.
5946 Otherwise, returns a new output port with a revealed count of 1.
5948 The next group of procedures perform a `dup2' system call, if NEWFD
5949 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
5950 duplicated can be supplied as an integer or contained in a port. The
5951 type of value returned varies depending on which procedure is used.
5953 All procedures also have the side effect when performing `dup2' that
5954 any ports using NEWFD are moved to a different file descriptor and have
5955 their revealed counts set to zero.
5957 *** procedure: dup->fdes PORT/FD [NEWFD]
5958 Returns an integer file descriptor.
5960 *** procedure: dup->inport PORT/FD [NEWFD]
5961 Returns a new input port using the new file descriptor.
5963 *** procedure: dup->outport PORT/FD [NEWFD]
5964 Returns a new output port using the new file descriptor.
5966 *** procedure: dup PORT/FD [NEWFD]
5967 Returns a new port if PORT/FD is a port, with the same mode as the
5968 supplied port, otherwise returns an integer file descriptor.
5970 *** procedure: dup->port PORT/FD MODE [NEWFD]
5971 Returns a new port using the new file descriptor. MODE supplies a
5972 mode string for the port (*note open-file: File Ports.).
5974 *** procedure: setenv NAME VALUE
5975 Modifies the environment of the current process, which is also the
5976 default environment inherited by child processes.
5978 If VALUE is `#f', then NAME is removed from the environment.
5979 Otherwise, the string NAME=VALUE is added to the environment,
5980 replacing any existing string with name matching NAME.
5982 The return value is unspecified.
5984 *** procedure: truncate-file OBJ SIZE
5985 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
5986 can be a string containing a file name or an integer file
5987 descriptor or port open for output on the file. The underlying
5988 system calls are `truncate' and `ftruncate'.
5990 The return value is unspecified.
5992 *** procedure: setvbuf PORT MODE [SIZE]
5993 Set the buffering mode for PORT. MODE can be:
6001 block buffered, using a newly allocated buffer of SIZE bytes.
6002 However if SIZE is zero or unspecified, the port will be made
6005 This procedure should not be used after I/O has been performed with
6008 Ports are usually block buffered by default, with a default buffer
6009 size. Procedures e.g., *Note open-file: File Ports, which accept a
6010 mode string allow `0' to be added to request an unbuffered port.
6012 *** procedure: fsync PORT/FD
6013 Copies any unwritten data for the specified output file descriptor
6014 to disk. If PORT/FD is a port, its buffer is flushed before the
6015 underlying file descriptor is fsync'd. The return value is
6018 *** procedure: open-fdes PATH FLAGS [MODES]
6019 Similar to `open' but returns a file descriptor instead of a port.
6021 *** procedure: execle PATH ENV [ARG] ...
6022 Similar to `execl', but the environment of the new process is
6023 specified by ENV, which must be a list of strings as returned by
6024 the `environ' procedure.
6026 This procedure is currently implemented using the `execve' system
6027 call, but we call it `execle' because of its Scheme calling
6030 *** procedure: strerror ERRNO
6031 Returns the Unix error message corresponding to ERRNO, an integer.
6033 *** procedure: primitive-exit [STATUS]
6034 Terminate the current process without unwinding the Scheme stack.
6035 This is would typically be useful after a fork. The exit status
6036 is STATUS if supplied, otherwise zero.
6038 *** procedure: times
6039 Returns an object with information about real and processor time.
6040 The following procedures accept such an object as an argument and
6041 return a selected component:
6044 The current real time, expressed as time units relative to an
6048 The CPU time units used by the calling process.
6051 The CPU time units used by the system on behalf of the
6055 The CPU time units used by terminated child processes of the
6056 calling process, whose status has been collected (e.g., using
6060 Similarly, the CPU times units used by the system on behalf of
6061 terminated child processes.
6063 ** Removed: list-length
6064 ** Removed: list-append, list-append!
6065 ** Removed: list-reverse, list-reverse!
6067 ** array-map renamed to array-map!
6069 ** serial-array-map renamed to serial-array-map!
6071 ** catch doesn't take #f as first argument any longer
6073 Previously, it was possible to pass #f instead of a key to `catch'.
6074 That would cause `catch' to pass a jump buffer object to the procedure
6075 passed as second argument. The procedure could then use this jump
6076 buffer objekt as an argument to throw.
6078 This mechanism has been removed since its utility doesn't motivate the
6079 extra complexity it introduces.
6081 ** The `#/' notation for lists now provokes a warning message from Guile.
6082 This syntax will be removed from Guile in the near future.
6084 To disable the warning message, set the GUILE_HUSH environment
6085 variable to any non-empty value.
6087 ** The newline character now prints as `#\newline', following the
6088 normal Scheme notation, not `#\nl'.
6090 * Changes to the gh_ interface
6092 ** The gh_enter function now takes care of loading the Guile startup files.
6093 gh_enter works by calling scm_boot_guile; see the remarks below.
6095 ** Function: void gh_write (SCM x)
6097 Write the printed representation of the scheme object x to the current
6098 output port. Corresponds to the scheme level `write'.
6100 ** gh_list_length renamed to gh_length.
6102 ** vector handling routines
6104 Several major changes. In particular, gh_vector() now resembles
6105 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
6106 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
6107 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
6108 vector-related gh_ functions have been implemented.
6110 ** pair and list routines
6112 Implemented several of the R4RS pair and list functions that were
6115 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
6117 New function. Converts double arrays back and forth between Scheme
6120 * Changes to the scm_ interface
6122 ** The function scm_boot_guile now takes care of loading the startup files.
6124 Guile's primary initialization function, scm_boot_guile, now takes
6125 care of loading `boot-9.scm', in the `ice-9' module, to initialize
6126 Guile, define the module system, and put together some standard
6127 bindings. It also loads `init.scm', which is intended to hold
6128 site-specific initialization code.
6130 Since Guile cannot operate properly until boot-9.scm is loaded, there
6131 is no reason to separate loading boot-9.scm from Guile's other
6132 initialization processes.
6134 This job used to be done by scm_compile_shell_switches, which didn't
6135 make much sense; in particular, it meant that people using Guile for
6136 non-shell-like applications had to jump through hoops to get Guile
6137 initialized properly.
6139 ** The function scm_compile_shell_switches no longer loads the startup files.
6140 Now, Guile always loads the startup files, whenever it is initialized;
6141 see the notes above for scm_boot_guile and scm_load_startup_files.
6143 ** Function: scm_load_startup_files
6144 This new function takes care of loading Guile's initialization file
6145 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
6146 this is always called by the Guile initialization process, it's
6147 probably not too useful to call this yourself, but it's there anyway.
6149 ** The semantics of smob marking have changed slightly.
6151 The smob marking function (the `mark' member of the scm_smobfuns
6152 structure) is no longer responsible for setting the mark bit on the
6153 smob. The generic smob handling code in the garbage collector will
6154 set this bit. The mark function need only ensure that any other
6155 objects the smob refers to get marked.
6157 Note that this change means that the smob's GC8MARK bit is typically
6158 already set upon entry to the mark function. Thus, marking functions
6159 which look like this:
6162 if (SCM_GC8MARKP (ptr))
6164 SCM_SETGC8MARK (ptr);
6165 ... mark objects to which the smob refers ...
6168 are now incorrect, since they will return early, and fail to mark any
6169 other objects the smob refers to. Some code in the Guile library used
6172 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
6174 If you have implemented your own I/O port type, by writing the
6175 functions required by the scm_ptobfuns and then calling scm_newptob,
6176 you will need to change your functions slightly.
6178 The functions in a scm_ptobfuns structure now expect the port itself
6179 as their argument; they used to expect the `stream' member of the
6180 port's scm_port_table structure. This allows functions in an
6181 scm_ptobfuns structure to easily access the port's cell (and any flags
6182 it its CAR), and the port's scm_port_table structure.
6184 Guile now passes the I/O port itself as the `port' argument in the
6185 following scm_ptobfuns functions:
6187 int (*free) (SCM port);
6188 int (*fputc) (int, SCM port);
6189 int (*fputs) (char *, SCM port);
6190 scm_sizet (*fwrite) SCM_P ((char *ptr,
6194 int (*fflush) (SCM port);
6195 int (*fgetc) (SCM port);
6196 int (*fclose) (SCM port);
6198 The interfaces to the `mark', `print', `equalp', and `fgets' methods
6201 If you have existing code which defines its own port types, it is easy
6202 to convert your code to the new interface; simply apply SCM_STREAM to
6203 the port argument to yield the value you code used to expect.
6205 Note that since both the port and the stream have the same type in the
6206 C code --- they are both SCM values --- the C compiler will not remind
6207 you if you forget to update your scm_ptobfuns functions.
6210 ** Function: int scm_internal_select (int fds,
6214 struct timeval *timeout);
6216 This is a replacement for the `select' function provided by the OS.
6217 It enables I/O blocking and sleeping to happen for one cooperative
6218 thread without blocking other threads. It also avoids busy-loops in
6219 these situations. It is intended that all I/O blocking and sleeping
6220 will finally go through this function. Currently, this function is
6221 only available on systems providing `gettimeofday' and `select'.
6223 ** Function: SCM scm_internal_stack_catch (SCM tag,
6224 scm_catch_body_t body,
6226 scm_catch_handler_t handler,
6229 A new sibling to the other two C level `catch' functions
6230 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
6231 the stack to be saved automatically into the variable `the-last-stack'
6232 (scm_the_last_stack_var) on error. This is necessary if you want to
6233 use advanced error reporting, such as calling scm_display_error and
6234 scm_display_backtrace. (They both take a stack object as argument.)
6236 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
6238 scm_catch_handler_t handler,
6241 Spawns a new thread. It does a job similar to
6242 scm_call_with_new_thread but takes arguments more suitable when
6243 spawning threads from application C code.
6245 ** The hook scm_error_callback has been removed. It was originally
6246 intended as a way for the user to install his own error handler. But
6247 that method works badly since it intervenes between throw and catch,
6248 thereby changing the semantics of expressions like (catch #t ...).
6249 The correct way to do it is to use one of the C level catch functions
6250 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
6252 ** Removed functions:
6254 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
6255 scm_list_reverse, scm_list_reverse_x
6257 ** New macros: SCM_LISTn where n is one of the integers 0-9.
6259 These can be used for pretty list creation from C. The idea is taken
6260 from Erick Gallesio's STk.
6262 ** scm_array_map renamed to scm_array_map_x
6264 ** mbstrings are now removed
6266 This means that the type codes scm_tc7_mb_string and
6267 scm_tc7_mb_substring has been removed.
6269 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
6271 Since we no longer support multi-byte strings, these I/O functions
6272 have been simplified, and renamed. Here are their old names, and
6273 their new names and arguments:
6275 scm_gen_putc -> void scm_putc (int c, SCM port);
6276 scm_gen_puts -> void scm_puts (char *s, SCM port);
6277 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
6278 scm_gen_getc -> void scm_getc (SCM port);
6281 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
6283 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
6285 SCM_TYP7S now masks away the bit which distinguishes substrings from
6288 ** scm_catch_body_t: Backward incompatible change!
6290 Body functions to scm_internal_catch and friends do not any longer
6291 take a second argument. This is because it is no longer possible to
6292 pass a #f arg to catch.
6294 ** Calls to scm_protect_object and scm_unprotect now nest properly.
6296 The function scm_protect_object protects its argument from being freed
6297 by the garbage collector. scm_unprotect_object removes that
6300 These functions now nest properly. That is, for every object O, there
6301 is a counter which scm_protect_object(O) increments and
6302 scm_unprotect_object(O) decrements, if the counter is greater than
6303 zero. Every object's counter is zero when it is first created. If an
6304 object's counter is greater than zero, the garbage collector will not
6305 reclaim its storage.
6307 This allows you to use scm_protect_object in your code without
6308 worrying that some other function you call will call
6309 scm_unprotect_object, and allow it to be freed. Assuming that the
6310 functions you call are well-behaved, and unprotect only those objects
6311 they protect, you can follow the same rule and have confidence that
6312 objects will be freed only at appropriate times.
6315 Changes in Guile 1.2 (released Tuesday, June 24 1997):
6317 * Changes to the distribution
6319 ** Nightly snapshots are now available from ftp.red-bean.com.
6320 The old server, ftp.cyclic.com, has been relinquished to its rightful
6323 Nightly snapshots of the Guile development sources are now available via
6324 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
6326 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
6327 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
6329 ** To run Guile without installing it, the procedure has changed a bit.
6331 If you used a separate build directory to compile Guile, you'll need
6332 to include the build directory in SCHEME_LOAD_PATH, as well as the
6333 source directory. See the `INSTALL' file for examples.
6335 * Changes to the procedure for linking libguile with your programs
6337 ** The standard Guile load path for Scheme code now includes
6338 $(datadir)/guile (usually /usr/local/share/guile). This means that
6339 you can install your own Scheme files there, and Guile will find them.
6340 (Previous versions of Guile only checked a directory whose name
6341 contained the Guile version number, so you had to re-install or move
6342 your Scheme sources each time you installed a fresh version of Guile.)
6344 The load path also includes $(datadir)/guile/site; we recommend
6345 putting individual Scheme files there. If you want to install a
6346 package with multiple source files, create a directory for them under
6349 ** Guile 1.2 will now use the Rx regular expression library, if it is
6350 installed on your system. When you are linking libguile into your own
6351 programs, this means you will have to link against -lguile, -lqt (if
6352 you configured Guile with thread support), and -lrx.
6354 If you are using autoconf to generate configuration scripts for your
6355 application, the following lines should suffice to add the appropriate
6356 libraries to your link command:
6358 ### Find Rx, quickthreads and libguile.
6359 AC_CHECK_LIB(rx, main)
6360 AC_CHECK_LIB(qt, main)
6361 AC_CHECK_LIB(guile, scm_shell)
6363 The Guile 1.2 distribution does not contain sources for the Rx
6364 library, as Guile 1.0 did. If you want to use Rx, you'll need to
6365 retrieve it from a GNU FTP site and install it separately.
6367 * Changes to Scheme functions and syntax
6369 ** The dynamic linking features of Guile are now enabled by default.
6370 You can disable them by giving the `--disable-dynamic-linking' option
6373 (dynamic-link FILENAME)
6375 Find the object file denoted by FILENAME (a string) and link it
6376 into the running Guile application. When everything works out,
6377 return a Scheme object suitable for representing the linked object
6378 file. Otherwise an error is thrown. How object files are
6379 searched is system dependent.
6381 (dynamic-object? VAL)
6383 Determine whether VAL represents a dynamically linked object file.
6385 (dynamic-unlink DYNOBJ)
6387 Unlink the indicated object file from the application. DYNOBJ
6388 should be one of the values returned by `dynamic-link'.
6390 (dynamic-func FUNCTION DYNOBJ)
6392 Search the C function indicated by FUNCTION (a string or symbol)
6393 in DYNOBJ and return some Scheme object that can later be used
6394 with `dynamic-call' to actually call this function. Right now,
6395 these Scheme objects are formed by casting the address of the
6396 function to `long' and converting this number to its Scheme
6399 (dynamic-call FUNCTION DYNOBJ)
6401 Call the C function indicated by FUNCTION and DYNOBJ. The
6402 function is passed no arguments and its return value is ignored.
6403 When FUNCTION is something returned by `dynamic-func', call that
6404 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
6405 etc.), look it up in DYNOBJ; this is equivalent to
6407 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
6409 Interrupts are deferred while the C function is executing (with
6410 SCM_DEFER_INTS/SCM_ALLOW_INTS).
6412 (dynamic-args-call FUNCTION DYNOBJ ARGS)
6414 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
6415 some arguments and return its return value. The C function is
6416 expected to take two arguments and return an `int', just like
6419 int c_func (int argc, char **argv);
6421 ARGS must be a list of strings and is converted into an array of
6422 `char *'. The array is passed in ARGV and its size in ARGC. The
6423 return value is converted to a Scheme number and returned from the
6424 call to `dynamic-args-call'.
6426 When dynamic linking is disabled or not supported on your system,
6427 the above functions throw errors, but they are still available.
6429 Here is a small example that works on GNU/Linux:
6431 (define libc-obj (dynamic-link "libc.so"))
6432 (dynamic-args-call 'rand libc-obj '())
6434 See the file `libguile/DYNAMIC-LINKING' for additional comments.
6436 ** The #/ syntax for module names is depreciated, and will be removed
6437 in a future version of Guile. Instead of
6445 The latter syntax is more consistent with existing Lisp practice.
6447 ** Guile now does fancier printing of structures. Structures are the
6448 underlying implementation for records, which in turn are used to
6449 implement modules, so all of these object now print differently and in
6450 a more informative way.
6452 The Scheme printer will examine the builtin variable *struct-printer*
6453 whenever it needs to print a structure object. When this variable is
6454 not `#f' it is deemed to be a procedure and will be applied to the
6455 structure object and the output port. When *struct-printer* is `#f'
6456 or the procedure return `#f' the structure object will be printed in
6457 the boring #<struct 80458270> form.
6459 This hook is used by some routines in ice-9/boot-9.scm to implement
6460 type specific printing routines. Please read the comments there about
6463 One of the more specific uses of structs are records. The printing
6464 procedure that could be passed to MAKE-RECORD-TYPE is now actually
6465 called. It should behave like a *struct-printer* procedure (described
6468 ** Guile now supports a new R4RS-compliant syntax for keywords. A
6469 token of the form #:NAME, where NAME has the same syntax as a Scheme
6470 symbol, is the external representation of the keyword named NAME.
6471 Keyword objects print using this syntax as well, so values containing
6472 keyword objects can be read back into Guile. When used in an
6473 expression, keywords are self-quoting objects.
6475 Guile suports this read syntax, and uses this print syntax, regardless
6476 of the current setting of the `keyword' read option. The `keyword'
6477 read option only controls whether Guile recognizes the `:NAME' syntax,
6478 which is incompatible with R4RS. (R4RS says such token represent
6481 ** Guile has regular expression support again. Guile 1.0 included
6482 functions for matching regular expressions, based on the Rx library.
6483 In Guile 1.1, the Guile/Rx interface was removed to simplify the
6484 distribution, and thus Guile had no regular expression support. Guile
6485 1.2 again supports the most commonly used functions, and supports all
6486 of SCSH's regular expression functions.
6488 If your system does not include a POSIX regular expression library,
6489 and you have not linked Guile with a third-party regexp library such as
6490 Rx, these functions will not be available. You can tell whether your
6491 Guile installation includes regular expression support by checking
6492 whether the `*features*' list includes the `regex' symbol.
6494 *** regexp functions
6496 By default, Guile supports POSIX extended regular expressions. That
6497 means that the characters `(', `)', `+' and `?' are special, and must
6498 be escaped if you wish to match the literal characters.
6500 This regular expression interface was modeled after that implemented
6501 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
6502 with SCSH regular expressions.
6504 **** Function: string-match PATTERN STR [START]
6505 Compile the string PATTERN into a regular expression and compare
6506 it with STR. The optional numeric argument START specifies the
6507 position of STR at which to begin matching.
6509 `string-match' returns a "match structure" which describes what,
6510 if anything, was matched by the regular expression. *Note Match
6511 Structures::. If STR does not match PATTERN at all,
6512 `string-match' returns `#f'.
6514 Each time `string-match' is called, it must compile its PATTERN
6515 argument into a regular expression structure. This operation is
6516 expensive, which makes `string-match' inefficient if the same regular
6517 expression is used several times (for example, in a loop). For better
6518 performance, you can compile a regular expression in advance and then
6519 match strings against the compiled regexp.
6521 **** Function: make-regexp STR [FLAGS]
6522 Compile the regular expression described by STR, and return the
6523 compiled regexp structure. If STR does not describe a legal
6524 regular expression, `make-regexp' throws a
6525 `regular-expression-syntax' error.
6527 FLAGS may be the bitwise-or of one or more of the following:
6529 **** Constant: regexp/extended
6530 Use POSIX Extended Regular Expression syntax when interpreting
6531 STR. If not set, POSIX Basic Regular Expression syntax is used.
6532 If the FLAGS argument is omitted, we assume regexp/extended.
6534 **** Constant: regexp/icase
6535 Do not differentiate case. Subsequent searches using the
6536 returned regular expression will be case insensitive.
6538 **** Constant: regexp/newline
6539 Match-any-character operators don't match a newline.
6541 A non-matching list ([^...]) not containing a newline matches a
6544 Match-beginning-of-line operator (^) matches the empty string
6545 immediately after a newline, regardless of whether the FLAGS
6546 passed to regexp-exec contain regexp/notbol.
6548 Match-end-of-line operator ($) matches the empty string
6549 immediately before a newline, regardless of whether the FLAGS
6550 passed to regexp-exec contain regexp/noteol.
6552 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
6553 Match the compiled regular expression REGEXP against `str'. If
6554 the optional integer START argument is provided, begin matching
6555 from that position in the string. Return a match structure
6556 describing the results of the match, or `#f' if no match could be
6559 FLAGS may be the bitwise-or of one or more of the following:
6561 **** Constant: regexp/notbol
6562 The match-beginning-of-line operator always fails to match (but
6563 see the compilation flag regexp/newline above) This flag may be
6564 used when different portions of a string are passed to
6565 regexp-exec and the beginning of the string should not be
6566 interpreted as the beginning of the line.
6568 **** Constant: regexp/noteol
6569 The match-end-of-line operator always fails to match (but see the
6570 compilation flag regexp/newline above)
6572 **** Function: regexp? OBJ
6573 Return `#t' if OBJ is a compiled regular expression, or `#f'
6576 Regular expressions are commonly used to find patterns in one string
6577 and replace them with the contents of another string.
6579 **** Function: regexp-substitute PORT MATCH [ITEM...]
6580 Write to the output port PORT selected contents of the match
6581 structure MATCH. Each ITEM specifies what should be written, and
6582 may be one of the following arguments:
6584 * A string. String arguments are written out verbatim.
6586 * An integer. The submatch with that number is written.
6588 * The symbol `pre'. The portion of the matched string preceding
6589 the regexp match is written.
6591 * The symbol `post'. The portion of the matched string
6592 following the regexp match is written.
6594 PORT may be `#f', in which case nothing is written; instead,
6595 `regexp-substitute' constructs a string from the specified ITEMs
6598 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
6599 Similar to `regexp-substitute', but can be used to perform global
6600 substitutions on STR. Instead of taking a match structure as an
6601 argument, `regexp-substitute/global' takes two string arguments: a
6602 REGEXP string describing a regular expression, and a TARGET string
6603 which should be matched against this regular expression.
6605 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
6608 * A function may be supplied. When this function is called, it
6609 will be passed one argument: a match structure for a given
6610 regular expression match. It should return a string to be
6611 written out to PORT.
6613 * The `post' symbol causes `regexp-substitute/global' to recurse
6614 on the unmatched portion of STR. This *must* be supplied in
6615 order to perform global search-and-replace on STR; if it is
6616 not present among the ITEMs, then `regexp-substitute/global'
6617 will return after processing a single match.
6619 *** Match Structures
6621 A "match structure" is the object returned by `string-match' and
6622 `regexp-exec'. It describes which portion of a string, if any, matched
6623 the given regular expression. Match structures include: a reference to
6624 the string that was checked for matches; the starting and ending
6625 positions of the regexp match; and, if the regexp included any
6626 parenthesized subexpressions, the starting and ending positions of each
6629 In each of the regexp match functions described below, the `match'
6630 argument must be a match structure returned by a previous call to
6631 `string-match' or `regexp-exec'. Most of these functions return some
6632 information about the original target string that was matched against a
6633 regular expression; we will call that string TARGET for easy reference.
6635 **** Function: regexp-match? OBJ
6636 Return `#t' if OBJ is a match structure returned by a previous
6637 call to `regexp-exec', or `#f' otherwise.
6639 **** Function: match:substring MATCH [N]
6640 Return the portion of TARGET matched by subexpression number N.
6641 Submatch 0 (the default) represents the entire regexp match. If
6642 the regular expression as a whole matched, but the subexpression
6643 number N did not match, return `#f'.
6645 **** Function: match:start MATCH [N]
6646 Return the starting position of submatch number N.
6648 **** Function: match:end MATCH [N]
6649 Return the ending position of submatch number N.
6651 **** Function: match:prefix MATCH
6652 Return the unmatched portion of TARGET preceding the regexp match.
6654 **** Function: match:suffix MATCH
6655 Return the unmatched portion of TARGET following the regexp match.
6657 **** Function: match:count MATCH
6658 Return the number of parenthesized subexpressions from MATCH.
6659 Note that the entire regular expression match itself counts as a
6660 subexpression, and failed submatches are included in the count.
6662 **** Function: match:string MATCH
6663 Return the original TARGET string.
6665 *** Backslash Escapes
6667 Sometimes you will want a regexp to match characters like `*' or `$'
6668 exactly. For example, to check whether a particular string represents
6669 a menu entry from an Info node, it would be useful to match it against
6670 a regexp like `^* [^:]*::'. However, this won't work; because the
6671 asterisk is a metacharacter, it won't match the `*' at the beginning of
6672 the string. In this case, we want to make the first asterisk un-magic.
6674 You can do this by preceding the metacharacter with a backslash
6675 character `\'. (This is also called "quoting" the metacharacter, and
6676 is known as a "backslash escape".) When Guile sees a backslash in a
6677 regular expression, it considers the following glyph to be an ordinary
6678 character, no matter what special meaning it would ordinarily have.
6679 Therefore, we can make the above example work by changing the regexp to
6680 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
6681 to match only a single asterisk in the target string.
6683 Since the backslash is itself a metacharacter, you may force a
6684 regexp to match a backslash in the target string by preceding the
6685 backslash with itself. For example, to find variable references in a
6686 TeX program, you might want to find occurrences of the string `\let\'
6687 followed by any number of alphabetic characters. The regular expression
6688 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
6689 each match a single backslash in the target string.
6691 **** Function: regexp-quote STR
6692 Quote each special character found in STR with a backslash, and
6693 return the resulting string.
6695 *Very important:* Using backslash escapes in Guile source code (as
6696 in Emacs Lisp or C) can be tricky, because the backslash character has
6697 special meaning for the Guile reader. For example, if Guile encounters
6698 the character sequence `\n' in the middle of a string while processing
6699 Scheme code, it replaces those characters with a newline character.
6700 Similarly, the character sequence `\t' is replaced by a horizontal tab.
6701 Several of these "escape sequences" are processed by the Guile reader
6702 before your code is executed. Unrecognized escape sequences are
6703 ignored: if the characters `\*' appear in a string, they will be
6704 translated to the single character `*'.
6706 This translation is obviously undesirable for regular expressions,
6707 since we want to be able to include backslashes in a string in order to
6708 escape regexp metacharacters. Therefore, to make sure that a backslash
6709 is preserved in a string in your Guile program, you must use *two*
6710 consecutive backslashes:
6712 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
6714 The string in this example is preprocessed by the Guile reader before
6715 any code is executed. The resulting argument to `make-regexp' is the
6716 string `^\* [^:]*', which is what we really want.
6718 This also means that in order to write a regular expression that
6719 matches a single backslash character, the regular expression string in
6720 the source code must include *four* backslashes. Each consecutive pair
6721 of backslashes gets translated by the Guile reader to a single
6722 backslash, and the resulting double-backslash is interpreted by the
6723 regexp engine as matching a single backslash character. Hence:
6725 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
6727 The reason for the unwieldiness of this syntax is historical. Both
6728 regular expression pattern matchers and Unix string processing systems
6729 have traditionally used backslashes with the special meanings described
6730 above. The POSIX regular expression specification and ANSI C standard
6731 both require these semantics. Attempting to abandon either convention
6732 would cause other kinds of compatibility problems, possibly more severe
6733 ones. Therefore, without extending the Scheme reader to support
6734 strings with different quoting conventions (an ungainly and confusing
6735 extension when implemented in other languages), we must adhere to this
6736 cumbersome escape syntax.
6738 * Changes to the gh_ interface
6740 * Changes to the scm_ interface
6742 * Changes to system call interfaces:
6744 ** The value returned by `raise' is now unspecified. It throws an exception
6747 *** A new procedure `sigaction' can be used to install signal handlers
6749 (sigaction signum [action] [flags])
6751 signum is the signal number, which can be specified using the value
6754 If action is omitted, sigaction returns a pair: the CAR is the current
6755 signal hander, which will be either an integer with the value SIG_DFL
6756 (default action) or SIG_IGN (ignore), or the Scheme procedure which
6757 handles the signal, or #f if a non-Scheme procedure handles the
6758 signal. The CDR contains the current sigaction flags for the handler.
6760 If action is provided, it is installed as the new handler for signum.
6761 action can be a Scheme procedure taking one argument, or the value of
6762 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
6763 whatever signal handler was installed before sigaction was first used.
6764 Flags can optionally be specified for the new handler (SA_RESTART is
6765 always used if the system provides it, so need not be specified.) The
6766 return value is a pair with information about the old handler as
6769 This interface does not provide access to the "signal blocking"
6770 facility. Maybe this is not needed, since the thread support may
6771 provide solutions to the problem of consistent access to data
6774 *** A new procedure `flush-all-ports' is equivalent to running
6775 `force-output' on every port open for output.
6777 ** Guile now provides information on how it was built, via the new
6778 global variable, %guile-build-info. This variable records the values
6779 of the standard GNU makefile directory variables as an assocation
6780 list, mapping variable names (symbols) onto directory paths (strings).
6781 For example, to find out where the Guile link libraries were
6782 installed, you can say:
6784 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
6787 * Changes to the scm_ interface
6789 ** The new function scm_handle_by_message_noexit is just like the
6790 existing scm_handle_by_message function, except that it doesn't call
6791 exit to terminate the process. Instead, it prints a message and just
6792 returns #f. This might be a more appropriate catch-all handler for
6793 new dynamic roots and threads.
6796 Changes in Guile 1.1 (released Friday, May 16 1997):
6798 * Changes to the distribution.
6800 The Guile 1.0 distribution has been split up into several smaller
6802 guile-core --- the Guile interpreter itself.
6803 guile-tcltk --- the interface between the Guile interpreter and
6804 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
6805 is a toolkit for building graphical user interfaces.
6806 guile-rgx-ctax --- the interface between Guile and the Rx regular
6807 expression matcher, and the translator for the Ctax
6808 programming language. These are packaged together because the
6809 Ctax translator uses Rx to parse Ctax source code.
6811 This NEWS file describes the changes made to guile-core since the 1.0
6814 We no longer distribute the documentation, since it was either out of
6815 date, or incomplete. As soon as we have current documentation, we
6820 * Changes to the stand-alone interpreter
6822 ** guile now accepts command-line arguments compatible with SCSH, Olin
6823 Shivers' Scheme Shell.
6825 In general, arguments are evaluated from left to right, but there are
6826 exceptions. The following switches stop argument processing, and
6827 stash all remaining command-line arguments as the value returned by
6828 the (command-line) function.
6829 -s SCRIPT load Scheme source code from FILE, and exit
6830 -c EXPR evalute Scheme expression EXPR, and exit
6831 -- stop scanning arguments; run interactively
6833 The switches below are processed as they are encountered.
6834 -l FILE load Scheme source code from FILE
6835 -e FUNCTION after reading script, apply FUNCTION to
6836 command line arguments
6837 -ds do -s script at this point
6838 --emacs enable Emacs protocol (experimental)
6839 -h, --help display this help and exit
6840 -v, --version display version information and exit
6841 \ read arguments from following script lines
6843 So, for example, here is a Guile script named `ekko' (thanks, Olin)
6844 which re-implements the traditional "echo" command:
6846 #!/usr/local/bin/guile -s
6849 (map (lambda (arg) (display arg) (display " "))
6853 (main (command-line))
6855 Suppose we invoke this script as follows:
6857 ekko a speckled gecko
6859 Through the magic of Unix script processing (triggered by the `#!'
6860 token at the top of the file), /usr/local/bin/guile receives the
6861 following list of command-line arguments:
6863 ("-s" "./ekko" "a" "speckled" "gecko")
6865 Unix inserts the name of the script after the argument specified on
6866 the first line of the file (in this case, "-s"), and then follows that
6867 with the arguments given to the script. Guile loads the script, which
6868 defines the `main' function, and then applies it to the list of
6869 remaining command-line arguments, ("a" "speckled" "gecko").
6871 In Unix, the first line of a script file must take the following form:
6873 #!INTERPRETER ARGUMENT
6875 where INTERPRETER is the absolute filename of the interpreter
6876 executable, and ARGUMENT is a single command-line argument to pass to
6879 You may only pass one argument to the interpreter, and its length is
6880 limited. These restrictions can be annoying to work around, so Guile
6881 provides a general mechanism (borrowed from, and compatible with,
6882 SCSH) for circumventing them.
6884 If the ARGUMENT in a Guile script is a single backslash character,
6885 `\', Guile will open the script file, parse arguments from its second
6886 and subsequent lines, and replace the `\' with them. So, for example,
6887 here is another implementation of the `ekko' script:
6889 #!/usr/local/bin/guile \
6893 (for-each (lambda (arg) (display arg) (display " "))
6897 If the user invokes this script as follows:
6899 ekko a speckled gecko
6901 Unix expands this into
6903 /usr/local/bin/guile \ ekko a speckled gecko
6905 When Guile sees the `\' argument, it replaces it with the arguments
6906 read from the second line of the script, producing:
6908 /usr/local/bin/guile -e main -s ekko a speckled gecko
6910 This tells Guile to load the `ekko' script, and apply the function
6911 `main' to the argument list ("a" "speckled" "gecko").
6913 Here is how Guile parses the command-line arguments:
6914 - Each space character terminates an argument. This means that two
6915 spaces in a row introduce an empty-string argument.
6916 - The tab character is not permitted (unless you quote it with the
6917 backslash character, as described below), to avoid confusion.
6918 - The newline character terminates the sequence of arguments, and will
6919 also terminate a final non-empty argument. (However, a newline
6920 following a space will not introduce a final empty-string argument;
6921 it only terminates the argument list.)
6922 - The backslash character is the escape character. It escapes
6923 backslash, space, tab, and newline. The ANSI C escape sequences
6924 like \n and \t are also supported. These produce argument
6925 constituents; the two-character combination \n doesn't act like a
6926 terminating newline. The escape sequence \NNN for exactly three
6927 octal digits reads as the character whose ASCII code is NNN. As
6928 above, characters produced this way are argument constituents.
6929 Backslash followed by other characters is not allowed.
6931 * Changes to the procedure for linking libguile with your programs
6933 ** Guile now builds and installs a shared guile library, if your
6934 system support shared libraries. (It still builds a static library on
6935 all systems.) Guile automatically detects whether your system
6936 supports shared libraries. To prevent Guile from buildisg shared
6937 libraries, pass the `--disable-shared' flag to the configure script.
6939 Guile takes longer to compile when it builds shared libraries, because
6940 it must compile every file twice --- once to produce position-
6941 independent object code, and once to produce normal object code.
6943 ** The libthreads library has been merged into libguile.
6945 To link a program against Guile, you now need only link against
6946 -lguile and -lqt; -lthreads is no longer needed. If you are using
6947 autoconf to generate configuration scripts for your application, the
6948 following lines should suffice to add the appropriate libraries to
6951 ### Find quickthreads and libguile.
6952 AC_CHECK_LIB(qt, main)
6953 AC_CHECK_LIB(guile, scm_shell)
6955 * Changes to Scheme functions
6957 ** Guile Scheme's special syntax for keyword objects is now optional,
6958 and disabled by default.
6960 The syntax variation from R4RS made it difficult to port some
6961 interesting packages to Guile. The routines which accepted keyword
6962 arguments (mostly in the module system) have been modified to also
6963 accept symbols whose names begin with `:'.
6965 To change the keyword syntax, you must first import the (ice-9 debug)
6967 (use-modules (ice-9 debug))
6969 Then you can enable the keyword syntax as follows:
6970 (read-set! keywords 'prefix)
6972 To disable keyword syntax, do this:
6973 (read-set! keywords #f)
6975 ** Many more primitive functions accept shared substrings as
6976 arguments. In the past, these functions required normal, mutable
6977 strings as arguments, although they never made use of this
6980 ** The uniform array functions now operate on byte vectors. These
6981 functions are `array-fill!', `serial-array-copy!', `array-copy!',
6982 `serial-array-map', `array-map', `array-for-each', and
6985 ** The new functions `trace' and `untrace' implement simple debugging
6986 support for Scheme functions.
6988 The `trace' function accepts any number of procedures as arguments,
6989 and tells the Guile interpreter to display each procedure's name and
6990 arguments each time the procedure is invoked. When invoked with no
6991 arguments, `trace' returns the list of procedures currently being
6994 The `untrace' function accepts any number of procedures as arguments,
6995 and tells the Guile interpreter not to trace them any more. When
6996 invoked with no arguments, `untrace' untraces all curretly traced
6999 The tracing in Guile has an advantage over most other systems: we
7000 don't create new procedure objects, but mark the procedure objects
7001 themselves. This means that anonymous and internal procedures can be
7004 ** The function `assert-repl-prompt' has been renamed to
7005 `set-repl-prompt!'. It takes one argument, PROMPT.
7006 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
7007 - If PROMPT is a string, we use it as a prompt.
7008 - If PROMPT is a procedure accepting no arguments, we call it, and
7009 display the result as a prompt.
7010 - Otherwise, we display "> ".
7012 ** The new function `eval-string' reads Scheme expressions from a
7013 string and evaluates them, returning the value of the last expression
7014 in the string. If the string contains no expressions, it returns an
7017 ** The new function `thunk?' returns true iff its argument is a
7018 procedure of zero arguments.
7020 ** `defined?' is now a builtin function, instead of syntax. This
7021 means that its argument should be quoted. It returns #t iff its
7022 argument is bound in the current module.
7024 ** The new syntax `use-modules' allows you to add new modules to your
7025 environment without re-typing a complete `define-module' form. It
7026 accepts any number of module names as arguments, and imports their
7027 public bindings into the current module.
7029 ** The new function (module-defined? NAME MODULE) returns true iff
7030 NAME, a symbol, is defined in MODULE, a module object.
7032 ** The new function `builtin-bindings' creates and returns a hash
7033 table containing copies of all the root module's bindings.
7035 ** The new function `builtin-weak-bindings' does the same as
7036 `builtin-bindings', but creates a doubly-weak hash table.
7038 ** The `equal?' function now considers variable objects to be
7039 equivalent if they have the same name and the same value.
7041 ** The new function `command-line' returns the command-line arguments
7042 given to Guile, as a list of strings.
7044 When using guile as a script interpreter, `command-line' returns the
7045 script's arguments; those processed by the interpreter (like `-s' or
7046 `-c') are omitted. (In other words, you get the normal, expected
7047 behavior.) Any application that uses scm_shell to process its
7048 command-line arguments gets this behavior as well.
7050 ** The new function `load-user-init' looks for a file called `.guile'
7051 in the user's home directory, and loads it if it exists. This is
7052 mostly for use by the code generated by scm_compile_shell_switches,
7053 but we thought it might also be useful in other circumstances.
7055 ** The new function `log10' returns the base-10 logarithm of its
7058 ** Changes to I/O functions
7060 *** The functions `read', `primitive-load', `read-and-eval!', and
7061 `primitive-load-path' no longer take optional arguments controlling
7062 case insensitivity and a `#' parser.
7064 Case sensitivity is now controlled by a read option called
7065 `case-insensitive'. The user can add new `#' syntaxes with the
7066 `read-hash-extend' function (see below).
7068 *** The new function `read-hash-extend' allows the user to change the
7069 syntax of Guile Scheme in a somewhat controlled way.
7071 (read-hash-extend CHAR PROC)
7072 When parsing S-expressions, if we read a `#' character followed by
7073 the character CHAR, use PROC to parse an object from the stream.
7074 If PROC is #f, remove any parsing procedure registered for CHAR.
7076 The reader applies PROC to two arguments: CHAR and an input port.
7078 *** The new functions read-delimited and read-delimited! provide a
7079 general mechanism for doing delimited input on streams.
7081 (read-delimited DELIMS [PORT HANDLE-DELIM])
7082 Read until we encounter one of the characters in DELIMS (a string),
7083 or end-of-file. PORT is the input port to read from; it defaults to
7084 the current input port. The HANDLE-DELIM parameter determines how
7085 the terminating character is handled; it should be one of the
7088 'trim omit delimiter from result
7089 'peek leave delimiter character in input stream
7090 'concat append delimiter character to returned value
7091 'split return a pair: (RESULT . TERMINATOR)
7093 HANDLE-DELIM defaults to 'peek.
7095 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
7096 A side-effecting variant of `read-delimited'.
7098 The data is written into the string BUF at the indices in the
7099 half-open interval [START, END); the default interval is the whole
7100 string: START = 0 and END = (string-length BUF). The values of
7101 START and END must specify a well-defined interval in BUF, i.e.
7102 0 <= START <= END <= (string-length BUF).
7104 It returns NBYTES, the number of bytes read. If the buffer filled
7105 up without a delimiter character being found, it returns #f. If the
7106 port is at EOF when the read starts, it returns the EOF object.
7108 If an integer is returned (i.e., the read is successfully terminated
7109 by reading a delimiter character), then the HANDLE-DELIM parameter
7110 determines how to handle the terminating character. It is described
7111 above, and defaults to 'peek.
7113 (The descriptions of these functions were borrowed from the SCSH
7114 manual, by Olin Shivers and Brian Carlstrom.)
7116 *** The `%read-delimited!' function is the primitive used to implement
7117 `read-delimited' and `read-delimited!'.
7119 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
7121 This returns a pair of values: (TERMINATOR . NUM-READ).
7122 - TERMINATOR describes why the read was terminated. If it is a
7123 character or the eof object, then that is the value that terminated
7124 the read. If it is #f, the function filled the buffer without finding
7125 a delimiting character.
7126 - NUM-READ is the number of characters read into BUF.
7128 If the read is successfully terminated by reading a delimiter
7129 character, then the gobble? parameter determines what to do with the
7130 terminating character. If true, the character is removed from the
7131 input stream; if false, the character is left in the input stream
7132 where a subsequent read operation will retrieve it. In either case,
7133 the character is also the first value returned by the procedure call.
7135 (The descriptions of this function was borrowed from the SCSH manual,
7136 by Olin Shivers and Brian Carlstrom.)
7138 *** The `read-line' and `read-line!' functions have changed; they now
7139 trim the terminator by default; previously they appended it to the
7140 returned string. For the old behavior, use (read-line PORT 'concat).
7142 *** The functions `uniform-array-read!' and `uniform-array-write!' now
7143 take new optional START and END arguments, specifying the region of
7144 the array to read and write.
7146 *** The `ungetc-char-ready?' function has been removed. We feel it's
7147 inappropriate for an interface to expose implementation details this
7150 ** Changes to the Unix library and system call interface
7152 *** The new fcntl function provides access to the Unix `fcntl' system
7155 (fcntl PORT COMMAND VALUE)
7156 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
7157 Values for COMMAND are:
7159 F_DUPFD duplicate a file descriptor
7160 F_GETFD read the descriptor's close-on-exec flag
7161 F_SETFD set the descriptor's close-on-exec flag to VALUE
7162 F_GETFL read the descriptor's flags, as set on open
7163 F_SETFL set the descriptor's flags, as set on open to VALUE
7164 F_GETOWN return the process ID of a socket's owner, for SIGIO
7165 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
7166 FD_CLOEXEC not sure what this is
7168 For details, see the documentation for the fcntl system call.
7170 *** The arguments to `select' have changed, for compatibility with
7171 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
7172 expected behavior. The MILLISECONDS parameter has been changed to
7173 MICROSECONDS, to more closely resemble the underlying system call.
7174 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
7175 corresponding return set will be the same.
7177 *** The arguments to the `mknod' system call have changed. They are
7180 (mknod PATH TYPE PERMS DEV)
7181 Create a new file (`node') in the file system. PATH is the name of
7182 the file to create. TYPE is the kind of file to create; it should
7183 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
7184 permission bits to give the newly created file. If TYPE is
7185 'block-special or 'char-special, DEV specifies which device the
7186 special file refers to; its interpretation depends on the kind of
7187 special file being created.
7189 *** The `fork' function has been renamed to `primitive-fork', to avoid
7190 clashing with various SCSH forks.
7192 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
7193 and `recvfrom!'. They no longer accept a size for a second argument;
7194 you must pass a string to hold the received value. They no longer
7195 return the buffer. Instead, `recv' returns the length of the message
7196 received, and `recvfrom' returns a pair containing the packet's length
7197 and originating address.
7199 *** The file descriptor datatype has been removed, as have the
7200 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
7201 We plan to replace these functions with a SCSH-compatible interface.
7203 *** The `create' function has been removed; it's just a special case
7206 *** There are new functions to break down process termination status
7207 values. In the descriptions below, STATUS is a value returned by
7210 (status:exit-val STATUS)
7211 If the child process exited normally, this function returns the exit
7212 code for the child process (i.e., the value passed to exit, or
7213 returned from main). If the child process did not exit normally,
7214 this function returns #f.
7216 (status:stop-sig STATUS)
7217 If the child process was suspended by a signal, this function
7218 returns the signal that suspended the child. Otherwise, it returns
7221 (status:term-sig STATUS)
7222 If the child process terminated abnormally, this function returns
7223 the signal that terminated the child. Otherwise, this function
7226 POSIX promises that exactly one of these functions will return true on
7227 a valid STATUS value.
7229 These functions are compatible with SCSH.
7231 *** There are new accessors and setters for the broken-out time vectors
7232 returned by `localtime', `gmtime', and that ilk. They are:
7234 Component Accessor Setter
7235 ========================= ============ ============
7236 seconds tm:sec set-tm:sec
7237 minutes tm:min set-tm:min
7238 hours tm:hour set-tm:hour
7239 day of the month tm:mday set-tm:mday
7240 month tm:mon set-tm:mon
7241 year tm:year set-tm:year
7242 day of the week tm:wday set-tm:wday
7243 day in the year tm:yday set-tm:yday
7244 daylight saving time tm:isdst set-tm:isdst
7245 GMT offset, seconds tm:gmtoff set-tm:gmtoff
7246 name of time zone tm:zone set-tm:zone
7248 *** There are new accessors for the vectors returned by `uname',
7249 describing the host system:
7252 ============================================== ================
7253 name of the operating system implementation utsname:sysname
7254 network name of this machine utsname:nodename
7255 release level of the operating system utsname:release
7256 version level of the operating system utsname:version
7257 machine hardware platform utsname:machine
7259 *** There are new accessors for the vectors returned by `getpw',
7260 `getpwnam', `getpwuid', and `getpwent', describing entries from the
7261 system's user database:
7264 ====================== =================
7265 user name passwd:name
7266 user password passwd:passwd
7269 real name passwd:gecos
7270 home directory passwd:dir
7271 shell program passwd:shell
7273 *** There are new accessors for the vectors returned by `getgr',
7274 `getgrnam', `getgrgid', and `getgrent', describing entries from the
7275 system's group database:
7278 ======================= ============
7279 group name group:name
7280 group password group:passwd
7282 group members group:mem
7284 *** There are new accessors for the vectors returned by `gethost',
7285 `gethostbyaddr', `gethostbyname', and `gethostent', describing
7289 ========================= ===============
7290 official name of host hostent:name
7291 alias list hostent:aliases
7292 host address type hostent:addrtype
7293 length of address hostent:length
7294 list of addresses hostent:addr-list
7296 *** There are new accessors for the vectors returned by `getnet',
7297 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
7301 ========================= ===============
7302 official name of net netent:name
7303 alias list netent:aliases
7304 net number type netent:addrtype
7305 net number netent:net
7307 *** There are new accessors for the vectors returned by `getproto',
7308 `getprotobyname', `getprotobynumber', and `getprotoent', describing
7312 ========================= ===============
7313 official protocol name protoent:name
7314 alias list protoent:aliases
7315 protocol number protoent:proto
7317 *** There are new accessors for the vectors returned by `getserv',
7318 `getservbyname', `getservbyport', and `getservent', describing
7322 ========================= ===============
7323 official service name servent:name
7324 alias list servent:aliases
7325 port number servent:port
7326 protocol to use servent:proto
7328 *** There are new accessors for the sockaddr structures returned by
7329 `accept', `getsockname', `getpeername', `recvfrom!':
7332 ======================================== ===============
7333 address format (`family') sockaddr:fam
7334 path, for file domain addresses sockaddr:path
7335 address, for internet domain addresses sockaddr:addr
7336 TCP or UDP port, for internet sockaddr:port
7338 *** The `getpwent', `getgrent', `gethostent', `getnetent',
7339 `getprotoent', and `getservent' functions now return #f at the end of
7340 the user database. (They used to throw an exception.)
7342 Note that calling MUMBLEent function is equivalent to calling the
7343 corresponding MUMBLE function with no arguments.
7345 *** The `setpwent', `setgrent', `sethostent', `setnetent',
7346 `setprotoent', and `setservent' routines now take no arguments.
7348 *** The `gethost', `getproto', `getnet', and `getserv' functions now
7349 provide more useful information when they throw an exception.
7351 *** The `lnaof' function has been renamed to `inet-lnaof'.
7353 *** Guile now claims to have the `current-time' feature.
7355 *** The `mktime' function now takes an optional second argument ZONE,
7356 giving the time zone to use for the conversion. ZONE should be a
7357 string, in the same format as expected for the "TZ" environment variable.
7359 *** The `strptime' function now returns a pair (TIME . COUNT), where
7360 TIME is the parsed time as a vector, and COUNT is the number of
7361 characters from the string left unparsed. This function used to
7362 return the remaining characters as a string.
7364 *** The `gettimeofday' function has replaced the old `time+ticks' function.
7365 The return value is now (SECONDS . MICROSECONDS); the fractional
7366 component is no longer expressed in "ticks".
7368 *** The `ticks/sec' constant has been removed, in light of the above change.
7370 * Changes to the gh_ interface
7372 ** gh_eval_str() now returns an SCM object which is the result of the
7375 ** gh_scm2str() now copies the Scheme data to a caller-provided C
7378 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
7379 and returns the array
7381 ** gh_scm2str0() is gone: there is no need to distinguish
7382 null-terminated from non-null-terminated, since gh_scm2newstr() allows
7383 the user to interpret the data both ways.
7385 * Changes to the scm_ interface
7387 ** The new function scm_symbol_value0 provides an easy way to get a
7388 symbol's value from C code:
7390 SCM scm_symbol_value0 (char *NAME)
7391 Return the value of the symbol named by the null-terminated string
7392 NAME in the current module. If the symbol named NAME is unbound in
7393 the current module, return SCM_UNDEFINED.
7395 ** The new function scm_sysintern0 creates new top-level variables,
7396 without assigning them a value.
7398 SCM scm_sysintern0 (char *NAME)
7399 Create a new Scheme top-level variable named NAME. NAME is a
7400 null-terminated string. Return the variable's value cell.
7402 ** The function scm_internal_catch is the guts of catch. It handles
7403 all the mechanics of setting up a catch target, invoking the catch
7404 body, and perhaps invoking the handler if the body does a throw.
7406 The function is designed to be usable from C code, but is general
7407 enough to implement all the semantics Guile Scheme expects from throw.
7409 TAG is the catch tag. Typically, this is a symbol, but this function
7410 doesn't actually care about that.
7412 BODY is a pointer to a C function which runs the body of the catch;
7413 this is the code you can throw from. We call it like this:
7414 BODY (BODY_DATA, JMPBUF)
7416 BODY_DATA is just the BODY_DATA argument we received; we pass it
7417 through to BODY as its first argument. The caller can make
7418 BODY_DATA point to anything useful that BODY might need.
7419 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
7420 which we have just created and initialized.
7422 HANDLER is a pointer to a C function to deal with a throw to TAG,
7423 should one occur. We call it like this:
7424 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
7426 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
7427 same idea as BODY_DATA above.
7428 THROWN_TAG is the tag that the user threw to; usually this is
7429 TAG, but it could be something else if TAG was #t (i.e., a
7430 catch-all), or the user threw to a jmpbuf.
7431 THROW_ARGS is the list of arguments the user passed to the THROW
7434 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
7435 is just a pointer we pass through to HANDLER. We don't actually
7436 use either of those pointers otherwise ourselves. The idea is
7437 that, if our caller wants to communicate something to BODY or
7438 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
7439 HANDLER can then use. Think of it as a way to make BODY and
7440 HANDLER closures, not just functions; MUMBLE_DATA points to the
7443 Of course, it's up to the caller to make sure that any data a
7444 MUMBLE_DATA needs is protected from GC. A common way to do this is
7445 to make MUMBLE_DATA a pointer to data stored in an automatic
7446 structure variable; since the collector must scan the stack for
7447 references anyway, this assures that any references in MUMBLE_DATA
7450 ** The new function scm_internal_lazy_catch is exactly like
7451 scm_internal_catch, except:
7453 - It does not unwind the stack (this is the major difference).
7454 - If handler returns, its value is returned from the throw.
7455 - BODY always receives #f as its JMPBUF argument (since there's no
7456 jmpbuf associated with a lazy catch, because we don't unwind the
7459 ** scm_body_thunk is a new body function you can pass to
7460 scm_internal_catch if you want the body to be like Scheme's `catch'
7461 --- a thunk, or a function of one argument if the tag is #f.
7463 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
7464 contains the Scheme procedure to invoke as the body, and the tag
7465 we're catching. If the tag is #f, then we pass JMPBUF (created by
7466 scm_internal_catch) to the body procedure; otherwise, the body gets
7469 ** scm_handle_by_proc is a new handler function you can pass to
7470 scm_internal_catch if you want the handler to act like Scheme's catch
7471 --- call a procedure with the tag and the throw arguments.
7473 If the user does a throw to this catch, this function runs a handler
7474 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
7475 variable holding the Scheme procedure object to invoke. It ought to
7476 be a pointer to an automatic variable (i.e., one living on the stack),
7477 or the procedure object should be otherwise protected from GC.
7479 ** scm_handle_by_message is a new handler function to use with
7480 `scm_internal_catch' if you want Guile to print a message and die.
7481 It's useful for dealing with throws to uncaught keys at the top level.
7483 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
7484 message header to print; if zero, we use "guile" instead. That
7485 text is followed by a colon, then the message described by ARGS.
7487 ** The return type of scm_boot_guile is now void; the function does
7488 not return a value, and indeed, never returns at all.
7490 ** The new function scm_shell makes it easy for user applications to
7491 process command-line arguments in a way that is compatible with the
7492 stand-alone guile interpreter (which is in turn compatible with SCSH,
7495 To use the scm_shell function, first initialize any guile modules
7496 linked into your application, and then call scm_shell with the values
7497 of ARGC and ARGV your `main' function received. scm_shell will add
7498 any SCSH-style meta-arguments from the top of the script file to the
7499 argument vector, and then process the command-line arguments. This
7500 generally means loading a script file or starting up an interactive
7501 command interpreter. For details, see "Changes to the stand-alone
7504 ** The new functions scm_get_meta_args and scm_count_argv help you
7505 implement the SCSH-style meta-argument, `\'.
7507 char **scm_get_meta_args (int ARGC, char **ARGV)
7508 If the second element of ARGV is a string consisting of a single
7509 backslash character (i.e. "\\" in Scheme notation), open the file
7510 named by the following argument, parse arguments from it, and return
7511 the spliced command line. The returned array is terminated by a
7514 For details of argument parsing, see above, under "guile now accepts
7515 command-line arguments compatible with SCSH..."
7517 int scm_count_argv (char **ARGV)
7518 Count the arguments in ARGV, assuming it is terminated by a null
7521 For an example of how these functions might be used, see the source
7522 code for the function scm_shell in libguile/script.c.
7524 You will usually want to use scm_shell instead of calling this
7527 ** The new function scm_compile_shell_switches turns an array of
7528 command-line arguments into Scheme code to carry out the actions they
7529 describe. Given ARGC and ARGV, it returns a Scheme expression to
7530 evaluate, and calls scm_set_program_arguments to make any remaining
7531 command-line arguments available to the Scheme code. For example,
7532 given the following arguments:
7534 -e main -s ekko a speckled gecko
7536 scm_set_program_arguments will return the following expression:
7538 (begin (load "ekko") (main (command-line)) (quit))
7540 You will usually want to use scm_shell instead of calling this
7543 ** The function scm_shell_usage prints a usage message appropriate for
7544 an interpreter that uses scm_compile_shell_switches to handle its
7545 command-line arguments.
7547 void scm_shell_usage (int FATAL, char *MESSAGE)
7548 Print a usage message to the standard error output. If MESSAGE is
7549 non-zero, write it before the usage message, followed by a newline.
7550 If FATAL is non-zero, exit the process, using FATAL as the
7551 termination status. (If you want to be compatible with Guile,
7552 always use 1 as the exit status when terminating due to command-line
7555 You will usually want to use scm_shell instead of calling this
7558 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
7559 expressions. It used to return SCM_EOL. Earth-shattering.
7561 ** The macros for declaring scheme objects in C code have been
7562 rearranged slightly. They are now:
7564 SCM_SYMBOL (C_NAME, SCHEME_NAME)
7565 Declare a static SCM variable named C_NAME, and initialize it to
7566 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
7567 be a C identifier, and SCHEME_NAME should be a C string.
7569 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
7570 Just like SCM_SYMBOL, but make C_NAME globally visible.
7572 SCM_VCELL (C_NAME, SCHEME_NAME)
7573 Create a global variable at the Scheme level named SCHEME_NAME.
7574 Declare a static SCM variable named C_NAME, and initialize it to
7575 point to the Scheme variable's value cell.
7577 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
7578 Just like SCM_VCELL, but make C_NAME globally visible.
7580 The `guile-snarf' script writes initialization code for these macros
7581 to its standard output, given C source code as input.
7583 The SCM_GLOBAL macro is gone.
7585 ** The scm_read_line and scm_read_line_x functions have been replaced
7586 by Scheme code based on the %read-delimited! procedure (known to C
7587 code as scm_read_delimited_x). See its description above for more
7590 ** The function scm_sys_open has been renamed to scm_open. It now
7591 returns a port instead of an FD object.
7593 * The dynamic linking support has changed. For more information, see
7594 libguile/DYNAMIC-LINKING.
7599 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
7602 * Changes to the 'guile' program:
7604 ** Guile now loads some new files when it starts up. Guile first
7605 searches the load path for init.scm, and loads it if found. Then, if
7606 Guile is not being used to execute a script, and the user's home
7607 directory contains a file named `.guile', Guile loads that.
7609 ** You can now use Guile as a shell script interpreter.
7611 To paraphrase the SCSH manual:
7613 When Unix tries to execute an executable file whose first two
7614 characters are the `#!', it treats the file not as machine code to
7615 be directly executed by the native processor, but as source code
7616 to be executed by some interpreter. The interpreter to use is
7617 specified immediately after the #! sequence on the first line of
7618 the source file. The kernel reads in the name of the interpreter,
7619 and executes that instead. It passes the interpreter the source
7620 filename as its first argument, with the original arguments
7621 following. Consult the Unix man page for the `exec' system call
7622 for more information.
7624 Now you can use Guile as an interpreter, using a mechanism which is a
7625 compatible subset of that provided by SCSH.
7627 Guile now recognizes a '-s' command line switch, whose argument is the
7628 name of a file of Scheme code to load. It also treats the two
7629 characters `#!' as the start of a comment, terminated by `!#'. Thus,
7630 to make a file of Scheme code directly executable by Unix, insert the
7631 following two lines at the top of the file:
7633 #!/usr/local/bin/guile -s
7636 Guile treats the argument of the `-s' command-line switch as the name
7637 of a file of Scheme code to load, and treats the sequence `#!' as the
7638 start of a block comment, terminated by `!#'.
7640 For example, here's a version of 'echo' written in Scheme:
7642 #!/usr/local/bin/guile -s
7644 (let loop ((args (cdr (program-arguments))))
7647 (display (car args))
7648 (if (pair? (cdr args))
7650 (loop (cdr args)))))
7653 Why does `#!' start a block comment terminated by `!#', instead of the
7654 end of the line? That is the notation SCSH uses, and although we
7655 don't yet support the other SCSH features that motivate that choice,
7656 we would like to be backward-compatible with any existing Guile
7657 scripts once we do. Furthermore, if the path to Guile on your system
7658 is too long for your kernel, you can start the script with this
7662 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
7665 Note that some very old Unix systems don't support the `#!' syntax.
7668 ** You can now run Guile without installing it.
7670 Previous versions of the interactive Guile interpreter (`guile')
7671 couldn't start up unless Guile's Scheme library had been installed;
7672 they used the value of the environment variable `SCHEME_LOAD_PATH'
7673 later on in the startup process, but not to find the startup code
7674 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
7677 To run Guile without installing it, build it in the normal way, and
7678 then set the environment variable `SCHEME_LOAD_PATH' to a
7679 colon-separated list of directories, including the top-level directory
7680 of the Guile sources. For example, if you unpacked Guile so that the
7681 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
7684 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
7687 ** Guile's read-eval-print loop no longer prints #<unspecified>
7688 results. If the user wants to see this, she can evaluate the
7689 expression (assert-repl-print-unspecified #t), perhaps in her startup
7692 ** Guile no longer shows backtraces by default when an error occurs;
7693 however, it does display a message saying how to get one, and how to
7694 request that they be displayed by default. After an error, evaluate
7696 to see a backtrace, and
7697 (debug-enable 'backtrace)
7698 to see them by default.
7702 * Changes to Guile Scheme:
7704 ** Guile now distinguishes between #f and the empty list.
7706 This is for compatibility with the IEEE standard, the (possibly)
7707 upcoming Revised^5 Report on Scheme, and many extant Scheme
7710 Guile used to have #f and '() denote the same object, to make Scheme's
7711 type system more compatible with Emacs Lisp's. However, the change
7712 caused too much trouble for Scheme programmers, and we found another
7713 way to reconcile Emacs Lisp with Scheme that didn't require this.
7716 ** Guile's delq, delv, delete functions, and their destructive
7717 counterparts, delq!, delv!, and delete!, now remove all matching
7718 elements from the list, not just the first. This matches the behavior
7719 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
7720 functions which inspired them.
7722 I recognize that this change may break code in subtle ways, but it
7723 seems best to make the change before the FSF's first Guile release,
7727 ** The compiled-library-path function has been deleted from libguile.
7729 ** The facilities for loading Scheme source files have changed.
7731 *** The variable %load-path now tells Guile which directories to search
7732 for Scheme code. Its value is a list of strings, each of which names
7735 *** The variable %load-extensions now tells Guile which extensions to
7736 try appending to a filename when searching the load path. Its value
7737 is a list of strings. Its default value is ("" ".scm").
7739 *** (%search-load-path FILENAME) searches the directories listed in the
7740 value of the %load-path variable for a Scheme file named FILENAME,
7741 with all the extensions listed in %load-extensions. If it finds a
7742 match, then it returns its full filename. If FILENAME is absolute, it
7743 returns it unchanged. Otherwise, it returns #f.
7745 %search-load-path will not return matches that refer to directories.
7747 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
7748 uses %seach-load-path to find a file named FILENAME, and loads it if
7749 it finds it. If it can't read FILENAME for any reason, it throws an
7752 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
7755 *** load uses the same searching semantics as primitive-load.
7757 *** The functions %try-load, try-load-with-path, %load, load-with-path,
7758 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
7759 path, and load-module-with-path have been deleted. The functions
7760 above should serve their purposes.
7762 *** If the value of the variable %load-hook is a procedure,
7763 `primitive-load' applies its value to the name of the file being
7764 loaded (without the load path directory name prepended). If its value
7765 is #f, it is ignored. Otherwise, an error occurs.
7767 This is mostly useful for printing load notification messages.
7770 ** The function `eval!' is no longer accessible from the scheme level.
7771 We can't allow operations which introduce glocs into the scheme level,
7772 because Guile's type system can't handle these as data. Use `eval' or
7773 `read-and-eval!' (see below) as replacement.
7775 ** The new function read-and-eval! reads an expression from PORT,
7776 evaluates it, and returns the result. This is more efficient than
7777 simply calling `read' and `eval', since it is not necessary to make a
7778 copy of the expression for the evaluator to munge.
7780 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
7781 for the `read' function.
7784 ** The function `int?' has been removed; its definition was identical
7785 to that of `integer?'.
7787 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
7788 use the R4RS names for these functions.
7790 ** The function object-properties no longer returns the hash handle;
7791 it simply returns the object's property list.
7793 ** Many functions have been changed to throw errors, instead of
7794 returning #f on failure. The point of providing exception handling in
7795 the language is to simplify the logic of user code, but this is less
7796 useful if Guile's primitives don't throw exceptions.
7798 ** The function `fileno' has been renamed from `%fileno'.
7800 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
7803 * Changes to Guile's C interface:
7805 ** The library's initialization procedure has been simplified.
7806 scm_boot_guile now has the prototype:
7808 void scm_boot_guile (int ARGC,
7810 void (*main_func) (),
7813 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
7814 MAIN_FUNC should do all the work of the program (initializing other
7815 packages, reading user input, etc.) before returning. When MAIN_FUNC
7816 returns, call exit (0); this function never returns. If you want some
7817 other exit value, MAIN_FUNC may call exit itself.
7819 scm_boot_guile arranges for program-arguments to return the strings
7820 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
7821 scm_set_program_arguments with the final list, so Scheme code will
7822 know which arguments have been processed.
7824 scm_boot_guile establishes a catch-all catch handler which prints an
7825 error message and exits the process. This means that Guile exits in a
7826 coherent way when system errors occur and the user isn't prepared to
7827 handle it. If the user doesn't like this behavior, they can establish
7828 their own universal catcher in MAIN_FUNC to shadow this one.
7830 Why must the caller do all the real work from MAIN_FUNC? The garbage
7831 collector assumes that all local variables of type SCM will be above
7832 scm_boot_guile's stack frame on the stack. If you try to manipulate
7833 SCM values after this function returns, it's the luck of the draw
7834 whether the GC will be able to find the objects you allocate. So,
7835 scm_boot_guile function exits, rather than returning, to discourage
7836 people from making that mistake.
7838 The IN, OUT, and ERR arguments were removed; there are other
7839 convenient ways to override these when desired.
7841 The RESULT argument was deleted; this function should never return.
7843 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
7847 ** Guile's header files should no longer conflict with your system's
7850 In order to compile code which #included <libguile.h>, previous
7851 versions of Guile required you to add a directory containing all the
7852 Guile header files to your #include path. This was a problem, since
7853 Guile's header files have names which conflict with many systems'
7856 Now only <libguile.h> need appear in your #include path; you must
7857 refer to all Guile's other header files as <libguile/mumble.h>.
7858 Guile's installation procedure puts libguile.h in $(includedir), and
7859 the rest in $(includedir)/libguile.
7862 ** Two new C functions, scm_protect_object and scm_unprotect_object,
7863 have been added to the Guile library.
7865 scm_protect_object (OBJ) protects OBJ from the garbage collector.
7866 OBJ will not be freed, even if all other references are dropped,
7867 until someone does scm_unprotect_object (OBJ). Both functions
7870 Note that calls to scm_protect_object do not nest. You can call
7871 scm_protect_object any number of times on a given object, and the
7872 next call to scm_unprotect_object will unprotect it completely.
7874 Basically, scm_protect_object and scm_unprotect_object just
7875 maintain a list of references to things. Since the GC knows about
7876 this list, all objects it mentions stay alive. scm_protect_object
7877 adds its argument to the list; scm_unprotect_object remove its
7878 argument from the list.
7881 ** scm_eval_0str now returns the value of the last expression
7884 ** The new function scm_read_0str reads an s-expression from a
7885 null-terminated string, and returns it.
7887 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
7888 to a Scheme port object.
7890 ** The new function `scm_set_program_arguments' allows C code to set
7891 the value returned by the Scheme `program-arguments' function.
7896 * Guile no longer includes sophisticated Tcl/Tk support.
7898 The old Tcl/Tk support was unsatisfying to us, because it required the
7899 user to link against the Tcl library, as well as Tk and Guile. The
7900 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
7901 referring to widgets by names, rather than exporting widgets to Scheme
7902 code as a special datatype.
7904 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
7905 maintainers described some very interesting changes in progress to the
7906 Tcl/Tk internals, which would facilitate clean interfaces between lone
7907 Tk and other interpreters --- even for garbage-collected languages
7908 like Scheme. They expected the new Tk to be publicly available in the
7911 Since it seems that Guile might soon have a new, cleaner interface to
7912 lone Tk, and that the old Guile/Tk glue code would probably need to be
7913 completely rewritten, we (Jim Blandy and Richard Stallman) have
7914 decided not to support the old code. We'll spend the time instead on
7915 a good interface to the newer Tk, as soon as it is available.
7917 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
7920 Copyright information:
7922 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
7924 Permission is granted to anyone to make or distribute verbatim copies
7925 of this document as received, in any medium, provided that the
7926 copyright notice and this permission notice are preserved,
7927 thus giving the recipient permission to redistribute in turn.
7929 Permission is granted to distribute modified versions
7930 of this document, or of portions of it,
7931 under the above conditions, provided also that they
7932 carry prominent notices stating who last changed them.
7937 paragraph-separate: "[
\f]*$"