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, 2010 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send Guile bug reports to bug-guile@gnu.org.
8 Note: During the 1.9 series, we will keep an incremental NEWS for the
9 latest prerelease, and a full NEWS corresponding to 1.8 -> 2.0.
11 Changes in 1.9.12 (since the 1.9.11 prerelease):
15 `(rnrs bytevectors)' and `(rnrs io ports)' now have version information,
16 like the rest of the modules. The `(rnrs unicode)' module is now
17 re-exported by `(rnrs)'. Top-level `import' forms may have more than one
18 clause. Warnings about duplicate bindings have been fixed, along with
19 some typos in the modules. There were a number of other bugfixes as well.
21 For the current list of known R6RS incompatibilities, see "R6RS
22 Incompatibilities" in the manual.
24 ** Documentation for standard R6RS libraries
26 See "R6RS Standard Libraries" in the manual, for more details.
28 ** Support for `letrec*'
30 Guile now supports letrec*, a form of letrec in which the identifiers
31 are bound in order. See "Local Bindings" in the manual, for more
34 ** Internal definitions now expand to `letrec*'
36 Following the R6RS, internal definitions now expand to letrec* instead
37 of letrec. The following program is invalid for R5RS, but valid for
42 (define baz (+ bar 20))
45 ;; R5RS and Guile <= 1.8:
46 (foo) => Unbound variable: bar
47 ;; R6RS and Guile >= 2.0:
50 This change should not affect correct R5RS programs, or programs written
51 in earlier Guile dialects.
53 ** Recursive debugging REPL on error
55 When Guile sees an error, instead of dropping into a special debugging
56 prompt, it will enter a recursive REPL in the dynamic context of the
57 error. See "Error Handling" in the manual, for more information.
59 A recursive REPL is the same as any other REPL, except that it
60 has been augmented with debugging information, so that one can inspect
61 the context of the error. The debugger has been integrated with the REPL
62 via a set of debugging meta-commands.
64 For example, one may access a backtrace with `,backtrace' (or
65 `,bt'). See "Interactive Debugging" in the manual, for more
68 ** Readline tab completion for arguments
70 When readline is enabled, tab completion works for arguments too, not
71 just for the operator position.
73 ** Various REPL robustness fixes
75 The REPL no longer enters the debugger when an error occurs at read-time
76 and compile-time, or when executing meta-commands. Additionally, the
77 REPL is more strict about the ports that it uses, saving the current
78 input and output ports at startup and using those ports when entering
79 recursive prompts. This allows debugging of an error within
80 e.g. `call-with-input-port'.
82 Finally, Ctrl-d can now be used to exit a recursive REPL, dropping the
83 user back to the parent REPL.
85 ** Better procedure name and source location propagation
87 There was a bug in 1.9.11 which would leave most procedures un-named;
88 this and other previous deficiencies have been fixed.
90 ** New macro: `current-source-location'
92 See FIXME in the manual, for more information.
94 ** module-filename field and accessor
96 Modules now record the file in which they are defined. This field may be
97 accessed with the new `module-filename' procedure.
99 ** call-with-error-handling / with-error-handling
103 ** Deprecate the old scm-style-repl
105 The following bindings from boot-9 are now found in `(ice-9
106 scm-style-repl)': `scm-style-repl', `error-catching-loop',
107 `error-catching-repl', `bad-throw', `scm-repl-silent'
108 `assert-repl-silence', `repl-print-unspecified',
109 `assert-repl-print-unspecified', `scm-repl-verbose',
110 `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl',
111 `default-pre-unwind-handler', `handle-system-error',
113 The following bindings have been deprecated, with no replacement:
114 `pre-unwind-handler-dispatch'.
116 The following bindings have been totally removed:
117 `before-signal-stack'.
119 Deprecated forwarding shims have been installed so that users that
120 expect these bindings in the main namespace will still work, but receive
121 a deprecation warning.
123 ** Miscellaneous other deprecations
125 `apply-to-args', `has-suffix?', `scheme-file-suffix'
126 `get-option', `for-next-option', `display-usage-report',
127 `transform-usage-lambda', `collect', `set-batch-mode?!'
129 ** All core defmacros reimplemented hygienically
131 All macros present in the default environment are now implemented with
132 syntax-case. This should not have practical ramifications, beyond those
133 listed below in "Lexical bindings introduced by hygienic macros may not
134 be referenced by nonhygienic macros".
136 ** Random generator state may be serialized to a datum
138 `random-state->datum' will serialize a random state to a datum, which
139 may be written out, read back in later, and revivified using
140 `datum->random-state'. See "Random" in the manual, for more details.
142 ** New primitive: `tmpfile'.
144 See "File System" in the manual.
146 ** Modules load within a known environment
148 It takes a few procedure calls to define a module, and those procedure
149 calls need to be in scope. Now we ensure that the current module when
150 loading a module is one that has the needed bindings, instead of relying
153 ** Remove encoding of versions into the file system
155 It used to be that, when loading a module, if the user specified a
156 version, Guile would grovel about in the filesystem to find the
157 module. This process was slow and not robust. This support has been
158 removed: modules are once more always loaded via `primitive-load-path'.
160 Module versions in the filesystem may be added again in the future, in
161 an extensible way. Contact bug-guile@gnu.org with patches.
163 ** Alex Shinn's pattern matcher for (ice-9 match).
165 Guile's copy of Andrew K. Wright's `match' library has been replaced by
166 a compatible hygienic implementation by Alex Shinn.
168 Compared to Andrew K. Wright's `match', the new `match' lacks
169 `match-define', `match:error-control', `match:set-error-control',
170 `match:error', `match:set-error', and all structure-related procedures.
172 ** Better debugging for psyntax
174 We now build psyntax-pp.go directly from psyntax.scm, which allows us to
175 preserve the original source locations and variable names. Git users
176 will also be please to note that psyntax-pp.scm is no longer
177 automatically regenerated from a normal `make' invocation.
181 Guile's build is visually quieter, due to the use of Automake 1.11's
182 AM_SILENT_RULES. Build as `make V=1' to see all of the output.
184 ** Better REPL options interface
186 The set of REPL options is now fixed and typechecked -- you can't
187 mistakenly set an option that doesn't exist, or set an value with an
188 incorrect type. The following options have been added: `prompt',
189 `compile-options', and `value-history'.
191 The new `repl-default-option-set!' procedure from `(system repl common)'
192 can set default options for future REPLs. `repl-default-prompt-set!' is
193 a convenience procedure to set default prompts.
195 ** `*unspecified*' is identifier syntax
197 `*unspecified*' is no longer a variable, so it is optimized properly by
198 the compiler, and is not `set!'-able.
200 ** `set-batch-mode?!' replaced by `ensure-batch-mode!'
202 "Batch mode" is a flag used to tell a program that it is not running
203 interactively. One usually turns it on after a fork. It may not be
204 turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!',
205 because it is a better interface, as it can only turn on batch mode, not
208 ** `(debug)' gone (fix?)
210 * module/system/vm/debug.scm (debug): Change to debug the current stack
211 instead of the last stack.
213 ** Support for settable identifier syntax
215 Following the R6RS, "variable transformers" are settable
216 identifier-syntax. See "Identifier macros" in the manual, for more
219 ** Deprecate `save-stack', `the-last-stack'
221 It used to be that the way to debug programs in Guile was to capture the
222 stack at the time of error, drop back to the REPL, then debug that
223 stack. But this approach didn't compose, was tricky to get right in the
224 presence of threads, and was not very powerful.
226 So `save-stack', `stack-saved?', and `the-last-stack' have been moved to
227 `(ice-9 save-stack)', with deprecated bindings left in the root module.
229 ** Compilation warnings at the REPL
231 By default, Guile now prints compile-time warnings for code entered at
232 the REPL. Current warnings are for unbound variables and mismatched
235 ** syntax-case treats `_' as a placeholder
237 Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches
238 anything, and binds no pattern variables. Unlike the R6RS, Guile also
239 permits `_' to be in the literals list for a pattern.
241 ** Remove old Emacs interface
243 Guile had an unused `--emacs' command line argument that was supposed to
244 help when running Guile inside Emacs. This option has been removed, and
245 the helper functions `named-module-use!' and `load-emacs-interface' have
248 ** `top-repl' has its own module
250 The `top-repl' binding, called with Guile is run interactively, is now
251 is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was
252 left in the default environment.
254 ** Value history in the REPL on by default
256 By default, the REPL will save computed values in variables like `$1',
257 `$2', and the like. There are programmatic and interactive interfaces to
258 control this. See "Value History" in the manual, for more information.
260 ** New threads are in `(guile-user)' by default, not `(guile)'
262 It used to be that a new thread entering Guile would do so in the
263 `(guile)' module, unless this was the first time Guile was initialized,
264 in which case it was `(guile-user)'. This has been fixed to have all
265 new threads unknown to Guile default to `(guile-user)'.
267 ** Backtrace improvements
269 It used to be that backtraces showed a couple of extra frames, from the
270 implementation of prompts. This has been fixed. Also, backtraces now
271 print column numbers in addition to line numbers, if possible.
273 ** `display-error' takes a frame
275 The `display-error' / `scm_display_error' helper now takes a frame as an
276 argument instead of a stack. Stacks are still supported in deprecated
277 builds. Additionally, `display-error' will again source location
278 information for the error.
280 ** Better error reporting from the VM
282 If a value of the wrong type is passed to `car', `cdr', `set-car!',
283 `set-cdr!', `apply', a struct accessor, or a bytevector accessors, the
284 error from the VM now indicates which procedure failed, instead of
285 lumping them all into the non-helpful "VM error". In addition, unbound
286 variable reporting is now more helpful.
288 ** No more `(ice-9 debug)'
290 This module had some debugging helpers that are no longer applicable to
291 the current debugging model. Importing this module will produce a
292 deprecation warning. Users should contact bug-guile for support.
294 ** No more `(system vm debug)'
296 This module is replaced by the recursive debugging REPL.
298 ** File ports handle binary mode and coding declarations
300 Binary file ports are supported again, by opening files in the
301 ISO-8859-1 encoding. If an encoding is not specified for textual ports,
302 Guile now grovels the file for a "coding:" directive, and uses that if
303 possible. See the documentation for `open-file', for more information.
305 ** R6RS character hex escapes on by default
307 However, R6RS character escapes within strings are incompatible with
308 older Guile escape sequences, so they still need the reader option
309 turned on. See "Reader Options" in the manual, for more information.
311 ** Fix random number generator on 64-bit platforms
313 There was a nasty bug on 64-bit platforms in which asking for a random
314 integer with a range between 2**32 and 2**64 caused a segfault. After
315 many embarrassing iterations, this was fixed.
317 ** Add `vhash-fold*' in `(ice-9 vlist)'.
319 See "VLists" in the manual, for more information.
321 ** Timestamps on autocompiled files checked for freshness, not equality
323 It used to be that to load a `.go' file instead of a `.scm' file, we
324 required that the timestamp of the `.go' file be equal to that of the
325 `.scm'. This has been relaxed to accept newer `.go' files, to integrate
326 better with `make' tools.
328 ** Simplifications to the `(system foreign)' API.
330 Guile's foreign function interface was simplified a great
331 deal. Interested users should see "Foreign Function Interface" in the
332 manual, for full details.
334 ** User Scheme code may be placed in a version-specific path
336 Before, there was only one way to install user Scheme code to a
337 version-specific Guile directory: install to Guile's own path,
338 e.g. /usr/share/guile/2.0. The site directory,
339 e.g. /usr/share/guile/site, was unversioned. This has been changed to
340 add a version-specific site directory, e.g. /usr/share/guile/site/2.0,
341 searched before the global site directory.
343 ** And of course, the usual collection of bugfixes
345 Interested users should see the ChangeLog for more information.
349 Changes in 1.9.x (since the 1.8.x series):
351 * New modules (see the manual for details)
353 ** `(srfi srfi-18)', more sophisticated multithreading support
354 ** `(ice-9 i18n)', internationalization support
355 ** `(rnrs bytevectors)', the R6RS bytevector API
356 ** `(rnrs io ports)', a subset of the R6RS I/O port API
357 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
358 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
359 ** `(system foreign)', foreign function interface
360 ** `(sxml match)', a pattern matcher for SXML
361 ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library
362 ** `(system vm coverage)', a line-by-line code coverage library
364 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
366 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
367 toolkit from Guile-Lib have been imported into Guile proper. See
368 "Standard Library" in the manual for more details.
370 ** Integration of lalr-scm, a parser generator
372 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
373 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
376 * Changes to the stand-alone interpreter
378 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
380 Compiled code loads much faster than Scheme source code, and runs around
381 3 or 4 times as fast, generating much less garbage in the process.
383 ** Evaluating Scheme code does not use the C stack.
385 Besides when compiling Guile itself, Guile no longer uses a recursive C
386 function as an evaluator. This obviates the need to check the C stack
387 pointer for overflow. Continuations still capture the C stack, however.
389 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
390 GUILE_SYSTEM_LOAD_COMPILED_PATH
392 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
393 for source files. It is a different path, however, because compiled
394 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
397 ** New read-eval-print loop (REPL) implementation
399 Running Guile with no arguments drops the user into the new REPL. While
400 it is self-documenting to an extent, the new REPL has not yet been
401 documented in the manual. This will be fixed before 2.0.
403 ** New reader options: `square-brackets' and `r6rs-hex-escapes'
405 The reader supports a new option (changeable via `read-options'),
406 `square-brackets', which instructs it to interpret square brackets as
407 parentheses. This option is on by default.
409 When the new `r6rs-hex-escapes' reader option is enabled, the reader
410 will recognize string escape sequences as defined in R6RS.
412 ** Function profiling and tracing at the REPL
414 The `,profile FORM' REPL meta-command can now be used to statistically
415 profile execution of a form, to see which functions are taking the most
416 time. See `,help profile' for more information.
418 Similarly, `,trace FORM' traces all function applications that occur
419 during the execution of `FORM'. See `,help trace' for more information.
423 By default, if an exception is raised at the REPL and not caught by user
424 code, Guile will drop the user into a debugger. The user may request a
425 backtrace, inspect frames, or continue raising the exception. Full
426 documentation is available from within the debugger.
428 ** New `guile-tools' commands: `compile', `disassemble'
430 Pass the `--help' command-line option to these commands for more
433 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
435 Users may now install Guile to nonstandard prefixes and just run
436 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
437 include `/path/to/lib'.
439 ** Guile's Emacs integration is now more keyboard-friendly
441 Backtraces may now be disclosed with the keyboard in addition to the
444 ** Load path change: search in version-specific paths before site paths
446 When looking for a module, Guile now searches first in Guile's
447 version-specific path (the library path), *then* in the site dir. This
448 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
449 installed. Also it should cut the number of `stat' system calls by half,
452 ** Interactive Guile follows GNU conventions
454 As recommended by the GPL, Guile now shows a brief copyright and
455 warranty disclaimer on startup, along with pointers to more information.
457 * Changes to Scheme functions and syntax
459 ** Support for R6RS libraries
461 The `library' and `import' forms from the latest Scheme report have been
462 added to Guile, in such a way that R6RS libraries share a namespace with
463 Guile modules. R6RS modules may import Guile modules, and are available
464 for Guile modules to import via use-modules and all the rest. See "R6RS
465 Libraries" in the manual for more information.
467 ** Implementations of R6RS libraries
469 Guile now has implementations for all of the libraries defined in the
470 R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS
471 Standard Libraries" in the manual for a full list of libraries.
473 ** Partial R6RS compatibility
475 Guile now has enough support for R6RS to run a reasonably large subset
478 Guile is not fully R6RS compatible. Many incompatibilities are simply
479 bugs, though some parts of Guile will remain R6RS-incompatible for the
480 foreseeable future. See "R6RS Incompatibilities" in the manual, for more
483 Please contact bug-guile@gnu.org if you have found an issue not
484 mentioned in that compatibility list.
486 ** New implementation of `primitive-eval'
488 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
489 still a C evaluator, used when building a fresh Guile to interpret the
490 compiler, so we can compile eval.scm. Thereafter all calls to
491 primitive-eval are implemented by VM-compiled code.
493 This allows all of Guile's procedures, be they interpreted or compiled,
494 to execute on the same stack, unifying multiple-value return semantics,
495 providing for proper tail recursion between interpreted and compiled
496 code, and simplifying debugging.
498 As part of this change, the evaluator no longer mutates the internal
499 representation of the code being evaluated in a thread-unsafe manner.
501 There are two negative aspects of this change, however. First, Guile
502 takes a lot longer to compile now. Also, there is less debugging
503 information available for debugging interpreted code. We hope to improve
504 both of these situations.
506 There are many changes to the internal C evalator interface, but all
507 public interfaces should be the same. See the ChangeLog for details. If
508 we have inadvertantly changed an interface that you were using, please
509 contact bug-guile@gnu.org.
511 ** Procedure removed: `the-environment'
513 This procedure was part of the interpreter's execution model, and does
514 not apply to the compiler.
516 ** No more `local-eval'
518 `local-eval' used to exist so that one could evaluate code in the
519 lexical context of a function. Since there is no way to get the lexical
520 environment any more, as that concept has no meaning for the compiler,
521 and a different meaning for the interpreter, we have removed the
524 If you think you need `local-eval', you should probably implement your
525 own metacircular evaluator. It will probably be as fast as Guile's
528 ** Scheme source files will now be compiled automatically.
530 If a compiled .go file corresponding to a .scm file is not found or is
531 not fresh, the .scm file will be compiled on the fly, and the resulting
532 .go file stored away. An advisory note will be printed on the console.
534 Note that this mechanism depends on preservation of the .scm and .go
535 modification times; if the .scm or .go files are moved after
536 installation, care should be taken to preserve their original
539 Autocompiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
540 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
541 will be created if needed.
543 To inhibit autocompilation, set the GUILE_AUTO_COMPILE environment
544 variable to 0, or pass --no-autocompile on the Guile command line.
546 ** New POSIX procedures: `getrlimit' and `setrlimit'
548 Note however that the interface of these functions is likely to change
549 in the next prerelease.
551 ** New POSIX procedure: `getsid'
553 Scheme binding for the `getsid' C library call.
555 ** New POSIX procedure: `getaddrinfo'
557 Scheme binding for the `getaddrinfo' C library function.
559 ** New procedure in `(oops goops)': `method-formals'
561 ** New procedures in (ice-9 session): `add-value-help-handler!',
562 `remove-value-help-handler!', `add-name-help-handler!'
563 `remove-name-help-handler!', `procedure-arguments'
565 The value and name help handlers provide some minimal extensibility to
566 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
567 example, to make stexinfo help documentation available. See those
568 procedures' docstrings for more information.
570 `procedure-arguments' describes the arguments that a procedure can take,
571 combining arity and formals. For example:
573 (procedure-arguments resolve-interface)
574 => ((required . (name)) (rest . args))
576 Additionally, `module-commentary' is now publically exported from
579 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
581 These procedures created primitive fexprs for the old evaluator, and are
582 no longer supported. If you feel that you need these functions, you
583 probably need to write your own metacircular evaluator (which will
584 probably be as fast as Guile's, anyway).
586 ** New language: ECMAScript
588 Guile now ships with one other high-level language supported,
589 ECMAScript. The goal is to support all of version 3.1 of the standard,
590 but not all of the libraries are there yet. This support is not yet
591 documented; ask on the mailing list if you are interested.
593 ** New language: Brainfuck
595 Brainfuck is a toy language that closely models Turing machines. Guile's
596 brainfuck compiler is meant to be an example of implementing other
597 languages. See the manual for details, or
598 http://en.wikipedia.org/wiki/Brainfuck for more information about the
599 Brainfuck language itself.
601 ** New language: Elisp
603 Guile now has an experimental Emacs Lisp compiler and runtime. You can
604 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
605 Kraft, and all bugs to bug-guile@gnu.org.
607 ** Better documentation infrastructure for macros
609 It is now possible to introspect on the type of a macro, e.g.
610 syntax-rules, identifier-syntax, etc, and extract information about that
611 macro, such as the syntax-rules patterns or the defmacro arguments.
612 `(texinfo reflection)' takes advantage of this to give better macro
615 ** Support for arbitrary procedure metadata
617 Building on its support for docstrings, Guile now supports multiple
618 docstrings, adding them to the tail of a compiled procedure's
619 properties. For example:
625 (procedure-properties foo)
626 => ((name . foo) (documentation . "one") (documentation . "two"))
628 Also, vectors of pairs are now treated as additional metadata entries:
631 #((quz . #f) (docstring . "xyzzy"))
633 (procedure-properties bar)
634 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
636 This allows arbitrary literals to be embedded as metadata in a compiled
639 ** The psyntax expander now knows how to interpret the @ and @@ special
642 ** The psyntax expander is now hygienic with respect to modules.
644 Free variables in a macro are scoped in the module that the macro was
645 defined in, not in the module the macro is used in. For example, code
648 (define-module (foo) #:export (bar))
649 (define (helper x) ...)
651 (syntax-rules () ((_ x) (helper x))))
653 (define-module (baz) #:use-module (foo))
656 It used to be you had to export `helper' from `(foo)' as well.
657 Thankfully, this has been fixed.
659 ** Complete support for version information in Guile's `module' form
661 Guile modules now have a `#:version' field. They may be loaded by
662 version as well. See "R6RS Version References", "General Information
663 about Modules", "Using Guile Modules", and "Creating Guile Modules" in
664 the manual for more information.
666 ** Support for renaming bindings on module export
668 Wherever Guile accepts a symbol as an argument to specify a binding to
669 export, it now also accepts a pair of symbols, indicating that a binding
670 should be renamed on export. See "Creating Guile Modules" in the manual
671 for more information.
673 ** New procedure: `module-export-all!'
675 This procedure exports all current and future bindings from a module.
676 Use as `(module-export-all! (current-module))'.
678 ** `eval-case' has been deprecated, and replaced by `eval-when'.
680 The semantics of `eval-when' are easier to understand. See "Eval When"
681 in the manual, for more information.
683 ** Guile is now more strict about prohibiting definitions in expression
686 Although previous versions of Guile accepted it, the following
687 expression is not valid, in R5RS or R6RS:
689 (if test (define foo 'bar) (define foo 'baz))
691 In this specific case, it would be better to do:
693 (define foo (if test 'bar 'baz))
695 It is certainly possible to circumvent this resriction with e.g.
696 `(module-define! (current-module) 'foo 'baz)'. We would appreciate
697 feedback about this change (a consequence of using psyntax as the
698 default expander), and may choose to revisit this situation before 2.0
699 in response to user feedback.
701 ** Macro expansion produces structures instead of s-expressions
703 In the olden days, macroexpanding an s-expression would yield another
704 s-expression. Though the lexical variables were renamed, expansions of
705 core forms like `if' and `begin' were still non-hygienic, as they relied
706 on the toplevel definitions of `if' et al being the conventional ones.
708 The solution is to expand to structures instead of s-expressions. There
709 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
710 etc. The expander already did this for compilation, producing Tree-IL
711 directly; it has been changed now to do so when expanding for the
714 ** Defmacros must now produce valid Scheme expressions.
716 It used to be that defmacros could unquote in Scheme values, as a way of
717 supporting partial evaluation, and avoiding some hygiene issues. For
720 (define (helper x) ...)
721 (define-macro (foo bar)
724 Assuming this macro is in the `(baz)' module, the direct translation of
727 (define (helper x) ...)
728 (define-macro (foo bar)
729 `((@@ (baz) helper) ,bar))
731 Of course, one could just use a hygienic macro instead:
735 ((_ bar) (helper bar))))
737 ** Guile's psyntax now supports docstrings and internal definitions.
739 The following Scheme is not strictly legal:
746 However its intent is fairly clear. Guile interprets "bar" to be the
747 docstring of `foo', and the definition of `baz' is still in definition
750 ** Macros need to be defined before their first use.
752 It used to be that with lazy memoization, this might work:
756 (define-macro (ref x) x)
759 But now, the body of `foo' is interpreted to mean a call to the toplevel
760 `ref' function, instead of a macro expansion. The solution is to define
761 macros before code that uses them.
763 ** Functions needed by macros at expand-time need to be present at
766 For example, this code will work at the REPL:
768 (define (double-helper x) (* x x))
769 (define-macro (double-literal x) (double-helper x))
770 (double-literal 2) => 4
772 But it will not work when a file is compiled, because the definition of
773 `double-helper' is not present at expand-time. The solution is to wrap
774 the definition of `double-helper' in `eval-when':
776 (eval-when (load compile eval)
777 (define (double-helper x) (* x x)))
778 (define-macro (double-literal x) (double-helper x))
779 (double-literal 2) => 4
781 See the documentation for eval-when for more information.
783 ** `macroexpand' produces structures, not S-expressions.
785 Given the need to maintain referential transparency, both lexically and
786 modular, the result of expanding Scheme expressions is no longer itself
787 an s-expression. If you want a human-readable approximation of the
788 result of `macroexpand', call `tree-il->scheme' from `(language
791 ** Removed function: `macroexpand-1'
793 It is unclear how to implement `macroexpand-1' with syntax-case, though
794 PLT Scheme does prove that it is possible.
796 ** New reader macros: #' #` #, #,@
798 These macros translate, respectively, to `syntax', `quasisyntax',
799 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
800 These reader macros may be overridden by `read-hash-extend'.
802 ** Incompatible change to #'
804 Guile did have a #' hash-extension, by default, which just returned the
805 subsequent datum: #'foo => foo. In the unlikely event that anyone
806 actually used this, this behavior may be reinstated via the
807 `read-hash-extend' mechanism.
809 ** Scheme expresssions may be commented out with #;
811 #; comments out an entire expression. See SRFI-62 or the R6RS for more
814 ** Prompts: Delimited, composable continuations
816 Guile now has prompts as part of its primitive language. See "Prompts"
817 in the manual, for more information.
819 Expressions entered in at the REPL, or from the command line, are
820 surrounded by a prompt with the default prompt tag.
822 ** `make-stack' with a tail-called procedural narrowing argument no longer
823 works (with compiled procedures)
825 It used to be the case that a captured stack could be narrowed to select
826 calls only up to or from a certain procedure, even if that procedure
827 already tail-called another procedure. This was because the debug
828 information from the original procedure was kept on the stack.
830 Now with the new compiler, the stack only contains active frames from
831 the current continuation. A narrow to a procedure that is not in the
832 stack will result in an empty stack. To fix this, narrow to a procedure
833 that is active in the current continuation, or narrow to a specific
834 number of stack frames.
836 ** Backtraces through compiled procedures only show procedures that are
837 active in the current continuation
839 Similarly to the previous issue, backtraces in compiled code may be
840 different from backtraces in interpreted code. There are no semantic
841 differences, however. Please mail bug-guile@gnu.org if you see any
842 deficiencies with Guile's backtraces.
844 ** syntax-rules and syntax-case macros now propagate source information
845 through to the expanded code
847 This should result in better backtraces.
849 ** The currying behavior of `define' has been removed.
851 Before, `(define ((f a) b) (* a b))' would translate to
853 (define f (lambda (a) (lambda (b) (* a b))))
855 Now a syntax error is signaled, as this syntax is not supported by
856 default. Use the `(ice-9 curried-definitions)' module to get back the
859 ** New procedure, `define!'
861 `define!' is a procedure that takes two arguments, a symbol and a value,
862 and binds the value to the symbol in the current module. It's useful to
863 programmatically make definitions in the current module, and is slightly
864 less verbose than `module-define!'.
866 ** All modules have names now
868 Before, you could have anonymous modules: modules without names. Now,
869 because of hygiene and macros, all modules have names. If a module was
870 created without a name, the first time `module-name' is called on it, a
871 fresh name will be lazily generated for it.
873 ** The module namespace is now separate from the value namespace
875 It was a little-known implementation detail of Guile's module system
876 that it was built on a single hierarchical namespace of values -- that
877 if there was a module named `(foo bar)', then in the module named
878 `(foo)' there was a binding from `bar' to the `(foo bar)' module.
880 This was a neat trick, but presented a number of problems. One problem
881 was that the bindings in a module were not apparent from the module
882 itself; perhaps the `(foo)' module had a private binding for `bar', and
883 then an external contributor defined `(foo bar)'. In the end there can
884 be only one binding, so one of the two will see the wrong thing, and
885 produce an obtuse error of unclear provenance.
887 Also, the public interface of a module was also bound in the value
888 namespace, as `%module-public-interface'. This was a hack from the early
889 days of Guile's modules.
891 Both of these warts have been fixed by the addition of fields in the
892 `module' data type. Access to modules and their interfaces from the
893 value namespace has been deprecated, and all accessors use the new
894 record accessors appropriately.
896 When Guile is built with support for deprecated code, as is the default,
897 the value namespace is still searched for modules and public interfaces,
898 and a deprecation warning is raised as appropriate.
900 Finally, to support lazy loading of modules as one used to be able to do
901 with module binder procedures, Guile now has submodule binders, called
902 if a given submodule is not found. See boot-9.scm for more information.
904 ** New procedures: module-ref-submodule, module-define-submodule,
905 nested-ref-module, nested-define-module!, local-ref-module,
908 These new accessors are like their bare variants, but operate on
909 namespaces instead of values.
911 ** The (app modules) module tree is officially deprecated
913 It used to be that one could access a module named `(foo bar)' via
914 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
915 modules)' bit was a never-used and never-documented abstraction, and has
916 been deprecated. See the following mail for a full discussion:
918 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
920 The `%app' binding is also deprecated.
922 ** Many syntax errors have different texts now
924 Syntax errors still throw to the `syntax-error' key, but the arguments
925 are often different now. Perhaps in the future, Guile will switch to
926 using standard SRFI-35 conditions.
928 ** Returning multiple values to compiled code will silently truncate the
929 values to the expected number
931 For example, the interpreter would raise an error evaluating the form,
932 `(+ (values 1 2) (values 3 4))', because it would see the operands as
933 being two compound "values" objects, to which `+' does not apply.
935 The compiler, on the other hand, receives multiple values on the stack,
936 not as a compound object. Given that it must check the number of values
937 anyway, if too many values are provided for a continuation, it chooses
938 to truncate those values, effectively evaluating `(+ 1 3)' instead.
940 The idea is that the semantics that the compiler implements is more
941 intuitive, and the use of the interpreter will fade out with time.
942 This behavior is allowed both by the R5RS and the R6RS.
944 ** Multiple values in compiled code are not represented by compound
947 This change may manifest itself in the following situation:
949 (let ((val (foo))) (do-something) val)
951 In the interpreter, if `foo' returns multiple values, multiple values
952 are produced from the `let' expression. In the compiler, those values
953 are truncated to the first value, and that first value is returned. In
954 the compiler, if `foo' returns no values, an error will be raised, while
955 the interpreter would proceed.
957 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
958 behavior is more correct, however. If you wish to preserve a potentially
959 multiply-valued return, you will need to set up a multiple-value
960 continuation, using `call-with-values'.
962 ** Defmacros are now implemented in terms of syntax-case.
964 The practical ramification of this is that the `defmacro?' predicate has
965 been removed, along with `defmacro-transformer', `macro-table',
966 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
967 `defmacro:transformer'. This is because defmacros are simply macros. If
968 any of these procedures provided useful facilities to you, we encourage
969 you to contact the Guile developers.
971 ** Hygienic macros documented as the primary syntactic extension mechanism.
973 The macro documentation was finally fleshed out with some documentation
974 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
975 expansion process. See "Macros" in the manual, for details.
977 ** psyntax is now the default expander
979 Scheme code is now expanded by default by the psyntax hygienic macro
980 expander. Expansion is performed completely before compilation or
983 Notably, syntax errors will be signalled before interpretation begins.
984 In the past, many syntax errors were only detected at runtime if the
985 code in question was memoized.
987 As part of its expansion, psyntax renames all lexically-bound
988 identifiers. Original identifier names are preserved and given to the
989 compiler, but the interpreter will see the renamed variables, e.g.,
990 `x432' instead of `x'.
992 Note that the psyntax that Guile uses is a fork, as Guile already had
993 modules before incompatible modules were added to psyntax -- about 10
994 years ago! Thus there are surely a number of bugs that have been fixed
995 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
997 ** syntax-rules and syntax-case are available by default.
999 There is no longer any need to import the `(ice-9 syncase)' module
1000 (which is now deprecated). The expander may be invoked directly via
1001 `macroexpand', though it is normally searched for via the current module
1004 Also, the helper routines for syntax-case are available in the default
1005 environment as well: `syntax->datum', `datum->syntax',
1006 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
1007 `identifier?', and `syntax-violation'. See the R6RS for documentation.
1009 ** Tail patterns in syntax-case
1011 Guile has pulled in some more recent changes from the psyntax portable
1012 syntax expander, to implement support for "tail patterns". Such patterns
1013 are supported by syntax-rules and syntax-case. This allows a syntax-case
1014 match clause to have ellipses, then a pattern at the end. For example:
1017 (syntax-rules (else)
1018 ((_ val match-clause ... (else e e* ...))
1021 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
1022 tail pattern for the else clause. Thanks to Andreas Rottmann for the
1023 patch, and Kent Dybvig for the code.
1025 ** Lexical bindings introduced by hygienic macros may not be referenced
1026 by nonhygienic macros.
1028 If a lexical binding is introduced by a hygienic macro, it may not be
1029 referenced by a nonhygienic macro. For example, this works:
1032 (define-macro (bind-x val body)
1033 `(let ((x ,val)) ,body))
1034 (define-macro (ref x)
1036 (bind-x 10 (ref x)))
1041 (define-syntax bind-x
1043 ((_ val body) (let ((x val)) body))))
1044 (define-macro (ref x)
1046 (bind-x 10 (ref x)))
1048 It is not normal to run into this situation with existing code. However,
1049 as code is ported over from defmacros to syntax-case, it is possible to
1050 run into situations like this. For example, if you have a defmacro that
1051 generates a `while' expression, the `break' bound by the `while' may not
1052 be visible within other parts of your defmacro. The solution is to port
1053 from defmacros to syntax-rules or syntax-case.
1055 ** Macros may no longer be referenced as first-class values.
1057 In the past, you could evaluate e.g. `if', and get its macro value. Now,
1058 expanding this form raises a syntax error.
1060 Macros still /exist/ as first-class values, but they must be
1061 /referenced/ via the module system, e.g. `(module-ref (current-module)
1064 ** Macros may now have docstrings.
1066 `object-documentation' from `(ice-9 documentation)' may be used to
1067 retrieve the docstring, once you have a macro value -- but see the above
1068 note about first-class macros. Docstrings are associated with the syntax
1069 transformer procedures.
1071 ** `case-lambda' is now available in the default environment.
1073 The binding in the default environment is equivalent to the one from the
1074 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
1075 to maintain compatibility with Guile 1.8 and earlier.
1077 ** Procedures may now have more than one arity.
1079 This can be the case, for example, in case-lambda procedures. The
1080 arities of compiled procedures may be accessed via procedures from the
1081 `(system vm program)' module; see "Compiled Procedures", "Optional
1082 Arguments", and "Case-lambda" in the manual.
1084 ** Deprecate arity access via (procedure-properties proc 'arity)
1086 Instead of accessing a procedure's arity as a property, use the new
1087 `procedure-minimum-arity' function, which gives the most permissive
1088 arity that the the function has, in the same format as the old arity
1091 ** `lambda*' and `define*' are now available in the default environment
1093 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
1094 compatibility purposes. No semantic change has been made (we hope).
1095 Optional and keyword arguments now dispatch via special VM operations,
1096 without the need to cons rest arguments, making them very fast.
1098 ** New function, `truncated-print', with `format' support
1100 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
1101 will ensure that the output stays within a certain width, truncating the
1102 output in what is hopefully an intelligent manner. See the manual for
1105 There is a new `format' specifier, `~@y', for doing a truncated
1106 print (as opposed to `~y', which does a pretty-print). See the `format'
1107 documentation for more details.
1109 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
1111 Guile now implements SRFI-4 vectors using bytevectors. Often when you
1112 have a numeric vector, you end up wanting to write its bytes somewhere,
1113 or have access to the underlying bytes, or read in bytes from somewhere
1114 else. Bytevectors are very good at this sort of thing. But the SRFI-4
1115 APIs are nicer to use when doing number-crunching, because they are
1116 addressed by element and not by byte.
1118 So as a compromise, Guile allows all bytevector functions to operate on
1119 numeric vectors. They address the underlying bytes in the native
1120 endianness, as one would expect.
1122 Following the same reasoning, that it's just bytes underneath, Guile
1123 also allows uniform vectors of a given type to be accessed as if they
1124 were of any type. One can fill a u32vector, and access its elements with
1125 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
1128 In this way, uniform numeric vectors may be written to and read from
1129 input/output ports using the procedures that operate on bytevectors.
1131 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
1132 inlined to the VM instructions for bytevector access.
1134 See "SRFI-4" in the manual, for more information.
1136 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
1138 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
1139 are part of the standard. Complex uniform vectors and the
1140 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
1142 Guile's default environment imports `(srfi srfi-4)', and probably should
1143 import `(srfi srfi-4 gnu)' as well.
1145 See "SRFI-4 Extensions" in the manual, for more information.
1147 ** New syntax: include-from-path.
1149 `include-from-path' is like `include', except it looks for its file in
1150 the load path. It can be used to compile other files into a file.
1152 ** New syntax: quasisyntax.
1154 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
1155 documentation for more information. Thanks to Andre van Tonder for the
1158 ** Unicode characters
1160 Unicode characters may be entered in octal format via e.g. `#\454', or
1161 created via (integer->char 300). A hex external representation will
1162 probably be introduced at some point.
1166 Internally, strings are now represented either in the `latin-1'
1167 encoding, one byte per character, or in UTF-32, with four bytes per
1168 character. Strings manage their own allocation, switching if needed.
1170 Extended characters may be written in a literal string using the
1171 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
1172 or 24-bit codepoints, respectively, or entered directly in the native
1173 encoding of the port on which the string is read.
1177 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
1179 ** Support for non-ASCII source code files
1181 The default reader now handles source code files for some of the
1182 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
1183 should have an encoding declaration near the top of the file. Also,
1184 there is a new function, `file-encoding', that scans a port for a coding
1185 declaration. See the section of the manual entitled, "Character Encoding
1188 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
1189 code. This use is now discouraged.
1191 ** Support for locale transcoding when reading from and writing to ports
1193 Ports now have an associated character encoding, and port read and write
1194 operations do conversion to and from locales automatically. Ports also
1195 have an associated strategy for how to deal with locale conversion
1198 See the documentation in the manual for the four new support functions,
1199 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
1200 and `port-conversion-strategy'.
1202 ** String and SRFI-13 functions can operate on Unicode strings
1204 ** Unicode support for SRFI-14 character sets
1206 The default character sets are no longer locale dependent and contain
1207 characters from the whole Unicode range. There is a new predefined
1208 character set, `char-set:designated', which contains all assigned
1209 Unicode characters. There is a new debugging function, `%char-set-dump'.
1211 ** Character functions operate on Unicode characters
1213 `char-upcase' and `char-downcase' use default Unicode casing rules.
1214 Character comparisons such as `char<?' and `char-ci<?' now sort based on
1215 Unicode code points.
1217 ** Global variables `scm_charnames' and `scm_charnums' are removed
1219 These variables contained the names of control characters and were
1220 used when writing characters. While these were global, they were
1221 never intended to be public API. They have been replaced with private
1224 ** EBCDIC support is removed
1226 There was an EBCDIC compile flag that altered some of the character
1227 processing. It appeared that full EBCDIC support was never completed
1228 and was unmaintained.
1230 ** Compile-time warnings
1232 Guile can warn about potentially unbound free variables. Pass the
1233 -Wunbound-variable on the `guile-tools compile' command line, or add
1234 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
1237 Guile can also warn when you pass the wrong number of arguments to a
1238 procedure, with -Warity-mismatch, or `arity-mismatch' in the
1239 `#:warnings' as above.
1241 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
1242 warn about unused local or global (top-level) variables.
1244 ** A new `memoize-symbol' evaluator trap has been added.
1246 This trap can be used for efficiently implementing a Scheme code
1249 ** Duplicate bindings among used modules are resolved lazily.
1251 This slightly improves program startup times.
1253 ** New thread cancellation and thread cleanup API
1255 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
1257 ** GOOPS dispatch in scheme
1259 As an implementation detail, GOOPS dispatch is no longer implemented by
1260 special evaluator bytecodes, but rather directly via a Scheme function
1261 associated with an applicable struct. There is some VM support for the
1262 underlying primitives, like `class-of'.
1264 This change will in the future allow users to customize generic function
1265 dispatch without incurring a performance penalty, and allow us to
1266 implement method combinations.
1268 ** Applicable struct support
1270 One may now make structs from Scheme that may be applied as procedures.
1271 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
1272 That struct will be the vtable of your applicable structs; instances of
1273 that new struct are assumed to have the procedure in their first slot.
1274 `<applicable-struct-vtable>' is like Common Lisp's
1275 `funcallable-standard-class'. Likewise there is
1276 `<applicable-struct-with-setter-vtable>', which looks for the setter in
1277 the second slot. This needs to be better documented.
1281 GOOPS had a number of concepts that were relevant to the days of Tcl,
1282 but not any more: operators and entities, mainly. These objects were
1283 never documented, and it is unlikely that they were ever used. Operators
1284 were a kind of generic specific to the Tcl support. Entities were
1285 replaced by applicable structs, mentioned above.
1287 ** New struct slot allocation: "hidden"
1289 A hidden slot is readable and writable, but will not be initialized by a
1290 call to make-struct. For example in your layout you would say "ph"
1291 instead of "pw". Hidden slots are useful for adding new slots to a
1292 vtable without breaking existing invocations to make-struct.
1294 ** eqv? not a generic
1296 One used to be able to extend `eqv?' as a primitive-generic, but no
1297 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
1298 should be able to compile to static dispatch tables, it doesn't make
1299 sense to allow extensions that would subvert this optimization.
1301 ** `inet-ntop' and `inet-pton' are always available.
1303 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
1304 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
1305 functions are deprecated.
1307 ** Fast bit operations.
1309 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
1310 have dedicated bytecodes. Guile is not just for symbolic computation,
1311 it's for number crunching too.
1313 ** Faster SRFI-9 record access
1315 SRFI-9 records are now implemented directly on top of Guile's structs,
1316 and their accessors are defined in such a way that normal call-sites
1317 inline to special VM opcodes, while still allowing for the general case
1318 (e.g. passing a record accessor to `apply').
1320 ** R6RS block comment support
1322 Guile now supports R6RS nested block comments. The start of a comment is
1323 marked with `#|', and the end with `|#'.
1325 ** `guile-2' cond-expand feature
1327 To test if your code is running under Guile 2.0 (or its alpha releases),
1328 test for the `guile-2' cond-expand feature. Like this:
1330 (cond-expand (guile-2 (eval-when (compile)
1331 ;; This must be evaluated at compile time.
1332 (fluid-set! current-reader my-reader)))
1334 ;; Earlier versions of Guile do not have a
1335 ;; separate compilation phase.
1336 (fluid-set! current-reader my-reader)))
1338 ** Fix bad interaction between `false-if-exception' and stack-call.
1340 Exceptions thrown by `false-if-exception' were erronously causing the
1341 stack to be saved, causing later errors to show the incorrectly-saved
1342 backtrace. This has been fixed.
1344 ** New global variables: %load-compiled-path, %load-compiled-extensions
1346 These are analogous to %load-path and %load-extensions.
1348 ** New fluid: `%file-port-name-canonicalization'
1350 This fluid parameterizes the file names that are associated with file
1351 ports. If %file-port-name-canonicalization is 'absolute, then file names
1352 are canonicalized to be absolute paths. If it is 'relative, then the
1353 name is canonicalized, but any prefix corresponding to a member of
1354 `%load-path' is stripped off. Otherwise the names are passed through
1357 In addition, the `compile-file' and `compile-and-load' procedures bind
1358 %file-port-name-canonicalization to their `#:canonicalization' keyword
1359 argument, which defaults to 'relative. In this way, one might compile
1360 "../module/ice-9/boot-9.scm", but the path that gets residualized into
1361 the .go is "ice-9/boot-9.scm".
1363 ** New procedure, `make-promise'
1365 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
1367 ** `defined?' may accept a module as its second argument
1369 Previously it only accepted internal structures from the evaluator.
1371 ** New entry into %guile-build-info: `ccachedir'
1373 ** Fix bug in `module-bound?'.
1375 `module-bound?' was returning true if a module did have a local
1376 variable, but one that was unbound, but another imported module bound
1377 the variable. This was an error, and was fixed.
1379 ** `(ice-9 syncase)' has been deprecated.
1381 As syntax-case is available by default, importing `(ice-9 syncase)' has
1382 no effect, and will trigger a deprecation warning.
1384 ** New readline history functions
1386 The (ice-9 readline) module now provides add-history, read-history,
1387 write-history and clear-history, which wrap the corresponding GNU
1388 History library functions.
1390 ** Removed deprecated uniform array procedures:
1391 dimensions->uniform-array, list->uniform-array, array-prototype
1393 Instead, use make-typed-array, list->typed-array, or array-type,
1396 ** Deprecated: `lazy-catch'
1398 `lazy-catch' was a form that captured the stack at the point of a
1399 `throw', but the dynamic state at the point of the `catch'. It was a bit
1400 crazy. Please change to use `catch', possibly with a throw-handler, or
1401 `with-throw-handler'.
1403 ** Deprecated `@bind' syntax
1405 `@bind' was part of an older implementation of the Emacs Lisp language,
1406 and is no longer used.
1408 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
1410 * Changes to the C interface
1412 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
1414 The semantics of `scm_gc_malloc ()' have been changed, in a
1415 backward-compatible way. A new allocation routine,
1416 `scm_gc_malloc_pointerless ()', was added.
1418 Libgc is a conservative GC, which we hope will make interaction with C
1419 code easier and less error-prone.
1421 ** New type definitions for `scm_t_intptr' and friends.
1423 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
1424 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
1425 available to C. Have fun!
1427 ** The GH interface (deprecated in version 1.6, 2001) was removed.
1429 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
1431 This makes these internal functions technically not callable from
1434 ** Functions for handling `scm_option' now no longer require an argument
1435 indicating length of the `scm_t_option' array.
1437 ** Procedures-with-setters are now implemented using applicable structs
1439 From a user's perspective this doesn't mean very much. But if, for some
1440 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
1441 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
1444 ** Remove old evaluator closures
1446 There used to be ranges of typecodes allocated to interpreted data
1447 structures, but that it no longer the case, given that interpreted
1448 procedure are now just regular VM closures. As a result, there is a
1449 newly free tc3, and a number of removed macros. See the ChangeLog for
1452 ** Primitive procedures are now VM trampoline procedures
1454 It used to be that there were something like 12 different typecodes
1455 allocated to primitive procedures, each with its own calling convention.
1456 Now there is only one, the gsubr. This may affect user code if you were
1457 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
1458 solution is to switch to use scm_c_make_gsubr. This solution works well
1459 both with the old 1.8 and and with the current 1.9 branch.
1461 Guile's old evaluator used to have special cases for applying "gsubrs",
1462 primitive procedures with specified numbers of required, optional, and
1463 rest arguments. Now, however, Guile represents gsubrs as normal VM
1464 procedures, with appropriate bytecode to parse out the correct number of
1465 arguments, including optional and rest arguments, and then with a
1466 special bytecode to apply the gsubr.
1468 This allows primitive procedures to appear on the VM stack, allowing
1469 them to be accurately counted in profiles. Also they now have more
1470 debugging information attached to them -- their number of arguments, for
1471 example. In addition, the VM can completely inline the application
1472 mechanics, allowing for faster primitive calls.
1474 However there are some changes on the C level. There is no more
1475 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
1476 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
1477 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
1478 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
1479 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
1481 Perhaps more significantly, `scm_c_make_subr',
1482 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
1483 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
1484 and there are no more subr typecodes. Use the scm_c_make_gsubr family
1487 Normal users of gsubrs should not be affected, though, as the
1488 scm_c_make_gsubr family still is the correct way to create primitive
1491 ** Remove deprecated array C interfaces
1493 Removed the deprecated array functions `scm_i_arrayp',
1494 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
1495 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
1496 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
1497 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
1499 ** Remove unused snarf macros
1501 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
1502 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
1504 ** New functions: `scm_call_n', `scm_c_run_hookn'
1506 `scm_call_n' applies to apply a function to an array of arguments.
1507 `scm_c_run_hookn' runs a hook with an array of arguments.
1509 ** Some SMOB types changed to have static typecodes
1511 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
1512 they have statically allocated tc7 typecodes.
1514 ** Preparations for changing SMOB representation
1516 If things go right, we'll be changing the SMOB representation soon. To
1517 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
1518 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
1519 in the future. Code accessing SMOBs using SCM_CELL macros was never
1520 correct, but until now things still worked. Users should be aware of
1523 ** Changed invocation mechanics of applicable SMOBs
1525 Guile's old evaluator used to have special cases for applying SMOB
1526 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
1527 trampoline procedure for it, and use the normal mechanics to apply the
1528 trampoline. This simplifies procedure application in the normal,
1531 The upshot is that the mechanics used to apply a SMOB are different from
1532 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
1533 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
1534 deprecated. Just use the scm_call_0 family of procedures.
1536 ** New C function: scm_module_public_interface
1538 This procedure corresponds to Scheme's `module-public-interface'.
1540 ** Undeprecate `scm_the_root_module ()'
1542 It's useful to be able to get the root module from C without doing a
1545 ** Inline vector allocation
1547 Instead of having vectors point out into the heap for their data, their
1548 data is now allocated inline to the vector object itself. The same is
1549 true for bytevectors, by default, though there is an indirection
1550 available which should allow for making a bytevector from an existing
1553 ** New struct constructors that don't involve making lists
1555 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
1556 constructors, respectively, for structs. You might find them useful.
1560 In Guile 1.8, there were debugging frames on the C stack. Now there is
1561 no more need to explicitly mark the stack in this way, because Guile has
1562 a VM stack that it knows how to walk, which simplifies the C API
1563 considerably. See the ChangeLog for details; the relevant interface is
1564 in libguile/stacks.h. The Scheme API has not been changed significantly.
1566 ** Removal of Guile's primitive object system.
1568 There were a number of pieces in `objects.[ch]' that tried to be a
1569 minimal object system, but were never documented, and were quickly
1570 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
1571 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
1572 from objects.h are no more. In the very unlikely case in which these
1573 were useful to you, we urge you to contact guile-devel.
1577 Actually the future is still in the state that it was, is, and ever
1578 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
1579 part of it. These files were experimental, never compiled, and would be
1580 better implemented in Scheme anyway. In the future, that is.
1582 ** Deprecate trampolines
1584 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
1585 so on. The point was to do some precomputation on the type of the
1586 procedure, then return a specialized "call" procedure. However this
1587 optimization wasn't actually an optimization, so it is now deprecated.
1588 Just use `scm_call_0', etc instead.
1590 ** Deprecated `scm_badargsp'
1592 This function is unused in Guile, but was part of its API.
1594 ** Better support for Lisp `nil'.
1596 The bit representation of `nil' has been tweaked so that it is now very
1597 efficient to check e.g. if a value is equal to Scheme's end-of-list or
1598 Lisp's nil. Additionally there are a heap of new, specific predicates
1599 like scm_is_null_or_nil.
1601 ** Better integration of Lisp `nil'.
1603 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
1604 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
1605 but when we start to integrate more with Emacs, it is possible that we
1606 break code that assumes that, for example, `(not x)' implies that `x' is
1607 `eq?' to `#f'. This is not a common assumption. Refactoring affected
1608 code to rely on properties instead of identities will improve code
1609 correctness. See "Nil" in the manual, for more details.
1611 ** Support for static allocation of strings, symbols, and subrs.
1613 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
1614 much of their associated data as static variables, reducing Guile's
1617 ** `scm_stat' has an additional argument, `exception_on_error'
1618 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
1620 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
1622 Previously they would use the `off_t' type, which is fragile since its
1623 definition depends on the application's value for `_FILE_OFFSET_BITS'.
1625 ** The `long_long' C type, deprecated in 1.8, has been removed
1627 ** Removed deprecated uniform array procedures: scm_make_uve,
1628 scm_array_prototype, scm_list_to_uniform_array,
1629 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
1630 scm_ra_set_contp, scm_aind, scm_raprin1
1632 These functions have been deprecated since early 2005.
1634 * Changes to the distribution
1636 ** Guile's license is now LGPLv3+
1638 In other words the GNU Lesser General Public License, version 3 or
1639 later (at the discretion of each person that chooses to redistribute
1642 ** GOOPS documentation folded into Guile reference manual
1644 GOOPS, Guile's object system, used to be documented in separate manuals.
1645 This content is now included in Guile's manual directly.
1647 ** `guile-config' will be deprecated in favor of `pkg-config'
1649 `guile-config' has been rewritten to get its information from
1650 `pkg-config', so this should be a transparent change. Note however that
1651 guile.m4 has yet to be modified to call pkg-config instead of
1654 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
1656 Programs that use `pkg-config' to find Guile or one of its Autoconf
1657 macros should now require `guile-2.0' instead of `guile-1.8'.
1659 ** New installation directory: $(pkglibdir)/1.9/ccache
1661 If $(libdir) is /usr/lib, for example, Guile will install its .go files
1662 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
1664 ** Parallel installability fixes
1666 Guile now installs its header files to a effective-version-specific
1667 directory, and includes the effective version (e.g. 2.0) in the library
1668 name (e.g. libguile-2.0.so).
1670 This change should be transparent to users, who should detect Guile via
1671 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
1672 parallel installs for multiple versions of Guile development
1675 ** Dynamically loadable extensions may be placed in a Guile-specific path
1677 Before, Guile only searched the system library paths for extensions
1678 (e.g. /usr/lib), which meant that the names of Guile extensions had to
1679 be globally unique. Installing them to a Guile-specific extensions
1680 directory is cleaner. Use `pkg-config --variable=extensiondir
1681 guile-2.0' to get the location of the extensions directory.
1683 ** New dependency: libgc
1685 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
1687 ** New dependency: GNU libunistring
1689 See http://www.gnu.org/software/libunistring/, for more information. Our
1690 Unicode support uses routines from libunistring.
1692 ** New dependency: libffi
1694 See http://sourceware.org/libffi/, for more information.
1698 Changes in 1.8.8 (since 1.8.7)
1702 ** Fix possible buffer overruns when parsing numbers
1703 ** Avoid clash with system setjmp/longjmp on IA64
1704 ** Fix `wrong type arg' exceptions with IPv6 addresses
1707 Changes in 1.8.7 (since 1.8.6)
1709 * New modules (see the manual for details)
1711 ** `(srfi srfi-98)', an interface to access environment variables
1715 ** Fix compilation with `--disable-deprecated'
1716 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
1717 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
1718 ** Fix build problem when scm_t_timespec is different from struct timespec
1719 ** Fix build when compiled with -Wundef -Werror
1720 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
1721 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
1722 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
1723 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
1724 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
1725 ** Fix reading of complex numbers where both parts are inexact decimals
1727 ** Allow @ macro to work with (ice-9 syncase)
1729 Previously, use of the @ macro in a module whose code is being
1730 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
1731 Now it works as you would expect (giving the value of the specified
1734 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
1737 Changes in 1.8.6 (since 1.8.5)
1739 * New features (see the manual for details)
1741 ** New convenience function `scm_c_symbol_length ()'
1743 ** Single stepping through code from Emacs
1745 When you use GDS to evaluate Scheme code from Emacs, you can now use
1746 `C-u' to indicate that you want to single step through that code. See
1747 `Evaluating Scheme Code' in the manual for more details.
1749 ** New "guile(1)" man page!
1751 * Changes to the distribution
1753 ** Automake's `AM_MAINTAINER_MODE' is no longer used
1755 Thus, the `--enable-maintainer-mode' configure option is no longer
1756 available: Guile is now always configured in "maintainer mode".
1758 ** `ChangeLog' files are no longer updated
1760 Instead, changes are detailed in the version control system's logs. See
1761 the top-level `ChangeLog' files for details.
1766 ** `symbol->string' now returns a read-only string, as per R5RS
1767 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
1768 ** `guile-config link' now prints `-L$libdir' before `-lguile'
1769 ** Fix memory corruption involving GOOPS' `class-redefinition'
1770 ** Fix possible deadlock in `mutex-lock'
1771 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
1772 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
1773 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
1774 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
1775 ** Fix misleading output from `(help rationalize)'
1776 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
1777 ** Fix `gcd' when called with a single, negative argument.
1778 ** Fix `Stack overflow' errors seen when building on some platforms
1779 ** Fix bug when `scm_with_guile ()' was called several times from the
1781 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
1782 dynamic environment of the call to `raise'
1783 ** Fix potential deadlock in `make-struct'
1784 ** Fix compilation problem with libltdl from Libtool 2.2.x
1785 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
1788 Changes in 1.8.5 (since 1.8.4)
1790 * Infrastructure changes
1792 ** Guile repository switched from CVS to Git
1794 The new repository can be accessed using
1795 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
1796 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
1798 ** Add support for `pkg-config'
1800 See "Autoconf Support" in the manual for details.
1802 * New modules (see the manual for details)
1806 * New features (see the manual for details)
1808 ** New `postfix' read option, for SRFI-88 keyword syntax
1809 ** Some I/O primitives have been inlined, which improves I/O performance
1810 ** New object-based traps infrastructure
1812 This is a GOOPS-based infrastructure that builds on Guile's low-level
1813 evaluator trap calls and facilitates the development of debugging
1814 features like single-stepping, breakpoints, tracing and profiling.
1815 See the `Traps' node of the manual for details.
1817 ** New support for working on Guile code from within Emacs
1819 Guile now incorporates the `GDS' library (previously distributed
1820 separately) for working on Guile code from within Emacs. See the
1821 `Using Guile In Emacs' node of the manual for details.
1825 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
1826 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
1828 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
1829 would trigger an unbound variable error for `match:andmap'.
1831 ** `(oop goops describe)' now properly provides the `describe' feature
1832 ** Fixed `args-fold' from `(srfi srfi-37)'
1834 Previously, parsing short option names of argument-less options would
1835 lead to a stack overflow.
1837 ** `(srfi srfi-35)' is now visible through `cond-expand'
1838 ** Fixed type-checking for the second argument of `eval'
1839 ** Fixed type-checking for SRFI-1 `partition'
1840 ** Fixed `struct-ref' and `struct-set!' on "light structs"
1841 ** Honor struct field access rights in GOOPS
1842 ** Changed the storage strategy of source properties, which fixes a deadlock
1843 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
1844 ** Fixed build issue for GNU/Linux on IA64
1845 ** Fixed build issues on NetBSD 1.6
1846 ** Fixed build issue on Solaris 2.10 x86_64
1847 ** Fixed build issue with DEC/Compaq/HP's compiler
1848 ** Fixed `scm_from_complex_double' build issue on FreeBSD
1849 ** Fixed `alloca' build issue on FreeBSD 6
1850 ** Removed use of non-portable makefile constructs
1851 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
1852 ** Make sure all tests honor `$TMPDIR'
1855 Changes in 1.8.4 (since 1.8.3)
1859 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
1860 ** Fixed a segmentation fault which occurred when displaying the
1861 backtrace of a stack with a promise object (made by `delay') in it.
1862 ** Make `accept' leave guile mode while blocking
1863 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
1864 ** Fixed a build problem on AIX (use of func_data identifier)
1865 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
1866 called with an associator proc that returns neither a pair nor #f.
1867 ** Secondary threads now always return a valid module for (current-module).
1868 ** Avoid MacOS build problems caused by incorrect combination of "64"
1869 system and library calls.
1870 ** `guile-snarf' now honors `$TMPDIR'
1871 ** `guile-config compile' now reports CPPFLAGS used at compile-time
1872 ** Fixed build with Sun Studio (Solaris 9)
1873 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
1874 uniform vectors on AIX.
1875 ** Fixed a deadlock that occurs upon GC with multiple threads.
1876 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
1877 ** Fixed autotool-derived build problems on AIX 6.1.
1878 ** Fixed NetBSD/alpha support
1879 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
1881 * New modules (see the manual for details)
1885 * Documentation fixes and improvements
1887 ** Removed premature breakpoint documentation
1889 The features described are not available in the series of 1.8.x
1890 releases, so the documentation was misleading and has been removed.
1892 ** More about Guile's default *random-state* variable
1894 ** GOOPS: more about how to use `next-method'
1896 * Changes to the distribution
1898 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
1900 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
1901 General Public License, and the few incorrect files have now been
1902 fixed to agree with the rest of the Guile distribution.
1904 ** Removed unnecessary extra copies of COPYING*
1906 The distribution now contains a single COPYING.LESSER at its top level.
1909 Changes in 1.8.3 (since 1.8.2)
1911 * New modules (see the manual for details)
1918 ** The `(ice-9 slib)' module now works as expected
1919 ** Expressions like "(set! 'x #t)" no longer yield a crash
1920 ** Warnings about duplicate bindings now go to stderr
1921 ** A memory leak in `make-socket-address' was fixed
1922 ** Alignment issues (e.g., on SPARC) in network routines were fixed
1923 ** A threading issue that showed up at least on NetBSD was fixed
1924 ** Build problems on Solaris and IRIX fixed
1926 * Implementation improvements
1928 ** The reader is now faster, which reduces startup time
1929 ** Procedures returned by `record-accessor' and `record-modifier' are faster
1932 Changes in 1.8.2 (since 1.8.1):
1934 * New procedures (see the manual for details)
1936 ** set-program-arguments
1939 * Incompatible changes
1941 ** The body of a top-level `define' no longer sees the binding being created
1943 In a top-level `define', the binding being created is no longer visible
1944 from the `define' body. This breaks code like
1945 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
1946 unbound in the body. However, such code was not R5RS-compliant anyway,
1951 ** Fractions were not `equal?' if stored in unreduced form.
1952 (A subtle problem, since printing a value reduced it, making it work.)
1953 ** srfi-60 `copy-bit' failed on 64-bit systems
1954 ** "guile --use-srfi" option at the REPL can replace core functions
1955 (Programs run with that option were ok, but in the interactive REPL
1956 the core bindings got priority, preventing SRFI replacements or
1958 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
1959 ** `kill' on mingw throws an error for a PID other than oneself
1960 ** Procedure names are attached to procedure-with-setters
1961 ** Array read syntax works with negative lower bound
1962 ** `array-in-bounds?' fix if an array has different lower bounds on each index
1963 ** `*' returns exact 0 for "(* inexact 0)"
1964 This follows what it always did for "(* 0 inexact)".
1965 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
1966 ** SRFI-19: `date->julian-day' did not account for timezone offset
1967 ** `ttyname' no longer crashes when passed a non-tty argument
1968 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
1969 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
1970 ** GOOPS: Fixed a bug in `method-more-specific?'
1971 ** Build problems on Solaris fixed
1972 ** Build problems on HP-UX IA64 fixed
1973 ** Build problems on MinGW fixed
1976 Changes in 1.8.1 (since 1.8.0):
1978 * LFS functions are now used to access 64-bit files on 32-bit systems.
1980 * New procedures (see the manual for details)
1982 ** primitive-_exit - [Scheme] the-root-module
1983 ** scm_primitive__exit - [C]
1984 ** make-completion-function - [Scheme] (ice-9 readline)
1985 ** scm_c_locale_stringn_to_number - [C]
1986 ** scm_srfi1_append_reverse [C]
1987 ** scm_srfi1_append_reverse_x [C]
1995 ** Build problems have been fixed on MacOS, SunOS, and QNX.
1997 ** `strftime' fix sign of %z timezone offset.
1999 ** A one-dimensional array can now be 'equal?' to a vector.
2001 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
2003 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
2005 ** `record-accessor' and `record-modifier' now have strict type checks.
2007 Record accessor and modifier procedures now throw an error if the
2008 record type of the record they're given is not the type expected.
2009 (Previously accessors returned #f and modifiers silently did nothing).
2011 ** It is now OK to use both autoload and use-modules on a given module.
2013 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
2015 Previously there was no checking on primatives like make-vector that
2016 accept "one or two" arguments. Now there is.
2018 ** The srfi-1 assoc function now calls its equality predicate properly.
2020 Previously srfi-1 assoc would call the equality predicate with the key
2021 last. According to the SRFI, the key should be first.
2023 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
2025 ** The array-set! procedure no longer segfaults when given a bit vector.
2027 ** Bugs in make-shared-array have been fixed.
2029 ** string<? and friends now follow char<? etc order on 8-bit chars.
2031 ** The format procedure now handles inf and nan values for ~f correctly.
2033 ** exact->inexact should no longer overflow when given certain large fractions.
2035 ** srfi-9 accessor and modifier procedures now have strict record type checks.
2037 This matches the srfi-9 specification.
2039 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
2041 Previously the (ice-9 ftw) procedures would ignore any file that had
2042 the same inode number as a file they had already seen, even if that
2043 file was on a different device.
2046 Changes in 1.8.0 (changes since the 1.6.x series):
2048 * Changes to the distribution
2050 ** Guile is now licensed with the GNU Lesser General Public License.
2052 ** The manual is now licensed with the GNU Free Documentation License.
2054 ** Guile now requires GNU MP (http://swox.com/gmp).
2056 Guile now uses the GNU MP library for arbitrary precision arithmetic.
2058 ** Guile now has separate private and public configuration headers.
2060 That is, things like HAVE_STRING_H no longer leak from Guile's
2063 ** Guile now provides and uses an "effective" version number.
2065 Guile now provides scm_effective_version and effective-version
2066 functions which return the "effective" version number. This is just
2067 the normal full version string without the final micro-version number,
2068 so the current effective-version is "1.8". The effective version
2069 should remain unchanged during a stable series, and should be used for
2070 items like the versioned share directory name
2071 i.e. /usr/share/guile/1.8.
2073 Providing an unchanging version number during a stable release for
2074 things like the versioned share directory can be particularly
2075 important for Guile "add-on" packages, since it provides a directory
2076 that they can install to that won't be changed out from under them
2077 with each micro release during a stable series.
2079 ** Thread implementation has changed.
2081 When you configure "--with-threads=null", you will get the usual
2082 threading API (call-with-new-thread, make-mutex, etc), but you can't
2083 actually create new threads. Also, "--with-threads=no" is now
2084 equivalent to "--with-threads=null". This means that the thread API
2085 is always present, although you might not be able to create new
2088 When you configure "--with-threads=pthreads" or "--with-threads=yes",
2089 you will get threads that are implemented with the portable POSIX
2090 threads. These threads can run concurrently (unlike the previous
2091 "coop" thread implementation), but need to cooperate for things like
2094 The default is "pthreads", unless your platform doesn't have pthreads,
2095 in which case "null" threads are used.
2097 See the manual for details, nodes "Initialization", "Multi-Threading",
2098 "Blocking", and others.
2100 ** There is the new notion of 'discouraged' features.
2102 This is a milder form of deprecation.
2104 Things that are discouraged should not be used in new code, but it is
2105 OK to leave them in old code for now. When a discouraged feature is
2106 used, no warning message is printed like there is for 'deprecated'
2107 features. Also, things that are merely discouraged are nevertheless
2108 implemented efficiently, while deprecated features can be very slow.
2110 You can omit discouraged features from libguile by configuring it with
2111 the '--disable-discouraged' option.
2113 ** Deprecation warnings can be controlled at run-time.
2115 (debug-enable 'warn-deprecated) switches them on and (debug-disable
2116 'warn-deprecated) switches them off.
2118 ** Support for SRFI 61, extended cond syntax for multiple values has
2121 This SRFI is always available.
2123 ** Support for require-extension, SRFI-55, has been added.
2125 The SRFI-55 special form `require-extension' has been added. It is
2126 available at startup, and provides a portable way to load Scheme
2127 extensions. SRFI-55 only requires support for one type of extension,
2128 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
2131 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
2133 The (srfi srfi-26) module is an implementation of SRFI-26 which
2134 provides the `cut' and `cute' syntax. These may be used to specialize
2135 parameters without currying.
2137 ** New module (srfi srfi-31)
2139 This is an implementation of SRFI-31 which provides a special form
2140 `rec' for recursive evaluation.
2142 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
2143 been merged with the core, making their functionality always
2146 The modules are still available, tho, and you could use them together
2147 with a renaming import, for example.
2149 ** Guile no longer includes its own version of libltdl.
2151 The official version is good enough now.
2153 ** The --enable-htmldoc option has been removed from 'configure'.
2155 Support for translating the documentation into HTML is now always
2156 provided. Use 'make html'.
2158 ** New module (ice-9 serialize):
2160 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
2161 don't trust the thread safety of most of your program, but where you
2162 have some section(s) of code which you consider can run in parallel to
2163 other sections. See ice-9/serialize.scm for more information.
2165 ** The configure option '--disable-arrays' has been removed.
2167 Support for arrays and uniform numeric arrays is now always included
2170 * Changes to the stand-alone interpreter
2172 ** New command line option `-L'.
2174 This option adds a directory to the front of the load path.
2176 ** New command line option `--no-debug'.
2178 Specifying `--no-debug' on the command line will keep the debugging
2179 evaluator turned off, even for interactive sessions.
2181 ** User-init file ~/.guile is now loaded with the debugging evaluator.
2183 Previously, the normal evaluator would have been used. Using the
2184 debugging evaluator gives better error messages.
2186 ** The '-e' option now 'read's its argument.
2188 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
2189 be used with '-e'. For example, you can now write a script like
2192 exec guile -e '(@ (demo) main)' -s "$0" "$@"
2195 (define-module (demo)
2199 (format #t "Demo: ~a~%" args))
2202 * Changes to Scheme functions and syntax
2204 ** Guardians have changed back to their original semantics
2206 Guardians now behave like described in the paper by Dybvig et al. In
2207 particular, they no longer make guarantees about the order in which
2208 they return objects, and they can no longer be greedy.
2210 They no longer drop cyclic data structures.
2212 The C function scm_make_guardian has been changed incompatibly and no
2213 longer takes the 'greedy_p' argument.
2215 ** New function hashx-remove!
2217 This function completes the set of 'hashx' functions.
2219 ** The concept of dynamic roots has been factored into continuation
2220 barriers and dynamic states.
2222 Each thread has a current dynamic state that carries the values of the
2223 fluids. You can create and copy dynamic states and use them as the
2224 second argument for 'eval'. See "Fluids and Dynamic States" in the
2227 To restrict the influence that captured continuations can have on the
2228 control flow, you can errect continuation barriers. See "Continuation
2229 Barriers" in the manual.
2231 The function call-with-dynamic-root now essentially temporarily
2232 installs a new dynamic state and errects a continuation barrier.
2234 ** The default load path no longer includes "." at the end.
2236 Automatically loading modules from the current directory should not
2237 happen by default. If you want to allow it in a more controlled
2238 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
2239 variable %load-path.
2241 ** The uniform vector and array support has been overhauled.
2243 It now complies with SRFI-4 and the weird prototype based uniform
2244 array creation has been deprecated. See the manual for more details.
2246 Some non-compatible changes have been made:
2247 - characters can no longer be stored into byte arrays.
2248 - strings and bit vectors are no longer considered to be uniform numeric
2250 - array-rank throws an error for non-arrays instead of returning zero.
2251 - array-ref does no longer accept non-arrays when no indices are given.
2253 There is the new notion of 'generalized vectors' and corresponding
2254 procedures like 'generalized-vector-ref'. Generalized vectors include
2255 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
2257 Arrays use generalized vectors as their storage, so that you still
2258 have arrays of characters, bits, etc. However, uniform-array-read!
2259 and uniform-array-write can no longer read/write strings and
2262 ** There is now support for copy-on-write substrings, mutation-sharing
2263 substrings and read-only strings.
2265 Three new procedures are related to this: substring/shared,
2266 substring/copy, and substring/read-only. See the manual for more
2269 ** Backtraces will now highlight the value that caused the error.
2271 By default, these values are enclosed in "{...}", such as in this
2280 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
2281 <unnamed port>:1:1: Wrong type (expecting pair): a
2282 ABORT: (wrong-type-arg)
2284 The prefix and suffix used for highlighting can be set via the two new
2285 printer options 'highlight-prefix' and 'highlight-suffix'. For
2286 example, putting this into ~/.guile will output the bad value in bold
2287 on an ANSI terminal:
2289 (print-set! highlight-prefix "\x1b[1m")
2290 (print-set! highlight-suffix "\x1b[22m")
2293 ** 'gettext' support for internationalization has been added.
2295 See the manual for details.
2297 ** New syntax '@' and '@@':
2299 You can now directly refer to variables exported from a module by
2302 (@ MODULE-NAME VARIABLE-NAME)
2304 For example (@ (ice-9 pretty-print) pretty-print) will directly access
2305 the pretty-print variable exported from the (ice-9 pretty-print)
2306 module. You don't need to 'use' that module first. You can also use
2307 '@' as a target of 'set!', as in (set! (@ mod var) val).
2309 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
2310 but it can also access variables that have not been exported. It is
2311 intended only for kluges and temporary fixes and for debugging, not
2314 ** Keyword syntax has been made more disciplined.
2316 Previously, the name of a keyword was read as a 'token' but printed as
2317 a symbol. Now, it is read as a general Scheme datum which must be a
2328 ERROR: In expression (a b c):
2332 ERROR: Unbound variable: foo
2337 ERROR: Wrong type (expecting symbol): 12
2341 ERROR: Wrong type (expecting symbol): (a b c)
2345 ** The printing of symbols that might look like keywords can be
2348 The new printer option 'quote-keywordish-symbols' controls how symbols
2349 are printed that have a colon as their first or last character. The
2350 default now is to only quote a symbol with #{...}# when the read
2351 option 'keywords' is not '#f'. Thus:
2353 guile> (define foo (string->symbol ":foo"))
2354 guile> (read-set! keywords #f)
2357 guile> (read-set! keywords 'prefix)
2360 guile> (print-set! quote-keywordish-symbols #f)
2364 ** 'while' now provides 'break' and 'continue'
2366 break and continue were previously bound in a while loop, but not
2367 documented, and continue didn't quite work properly. The undocumented
2368 parameter to break which gave a return value for the while has been
2371 ** 'call-with-current-continuation' is now also available under the name
2374 ** The module system now checks for duplicate bindings.
2376 The module system now can check for name conflicts among imported
2379 The behavior can be controlled by specifying one or more 'duplicates'
2380 handlers. For example, to make Guile return an error for every name
2383 (define-module (foo)
2388 The new default behavior of the module system when a name collision
2389 has been detected is to
2391 1. Give priority to bindings marked as a replacement.
2392 2. Issue a warning (different warning if overriding core binding).
2393 3. Give priority to the last encountered binding (this corresponds to
2396 If you want the old behavior back without replacements or warnings you
2399 (default-duplicate-binding-handler 'last)
2401 to your .guile init file.
2403 ** New define-module option: :replace
2405 :replace works as :export, but, in addition, marks the binding as a
2408 A typical example is `format' in (ice-9 format) which is a replacement
2409 for the core binding `format'.
2411 ** Adding prefixes to imported bindings in the module system
2413 There is now a new :use-module option :prefix. It can be used to add
2414 a prefix to all imported bindings.
2416 (define-module (foo)
2417 :use-module ((bar) :prefix bar:))
2419 will import all bindings exported from bar, but rename them by adding
2422 ** Conflicting generic functions can be automatically merged.
2424 When two imported bindings conflict and they are both generic
2425 functions, the two functions can now be merged automatically. This is
2426 activated with the 'duplicates' handler 'merge-generics'.
2428 ** New function: effective-version
2430 Returns the "effective" version number. This is just the normal full
2431 version string without the final micro-version number. See "Changes
2432 to the distribution" above.
2434 ** New threading functions: parallel, letpar, par-map, and friends
2436 These are convenient ways to run calculations in parallel in new
2437 threads. See "Parallel forms" in the manual for details.
2439 ** New function 'try-mutex'.
2441 This function will attempt to lock a mutex but will return immediately
2442 instead of blocking and indicate failure.
2444 ** Waiting on a condition variable can have a timeout.
2446 The function 'wait-condition-variable' now takes a third, optional
2447 argument that specifies the point in time where the waiting should be
2450 ** New function 'broadcast-condition-variable'.
2452 ** New functions 'all-threads' and 'current-thread'.
2454 ** Signals and system asyncs work better with threads.
2456 The function 'sigaction' now takes a fourth, optional, argument that
2457 specifies the thread that the handler should run in. When the
2458 argument is omitted, the handler will run in the thread that called
2461 Likewise, 'system-async-mark' takes a second, optional, argument that
2462 specifies the thread that the async should run in. When it is
2463 omitted, the async will run in the thread that called
2464 'system-async-mark'.
2466 C code can use the new functions scm_sigaction_for_thread and
2467 scm_system_async_mark_for_thread to pass the new thread argument.
2469 When a thread blocks on a mutex, a condition variable or is waiting
2470 for IO to be possible, it will still execute system asyncs. This can
2471 be used to interrupt such a thread by making it execute a 'throw', for
2474 ** The function 'system-async' is deprecated.
2476 You can now pass any zero-argument procedure to 'system-async-mark'.
2477 The function 'system-async' will just return its argument unchanged
2480 ** New functions 'call-with-blocked-asyncs' and
2481 'call-with-unblocked-asyncs'
2483 The expression (call-with-blocked-asyncs PROC) will call PROC and will
2484 block execution of system asyncs for the current thread by one level
2485 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
2486 procedure and will unblock the execution of system asyncs by one
2487 level for the current thread.
2489 Only system asyncs are affected by these functions.
2491 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
2493 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
2494 instead. Those functions are easier to use correctly and can be
2497 ** New function 'unsetenv'.
2499 ** New macro 'define-syntax-public'.
2501 It works like 'define-syntax' and also exports the defined macro (but
2504 ** There is support for Infinity and NaNs.
2506 Following PLT Scheme, Guile can now work with infinite numbers, and
2509 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
2510 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
2511 "+nan.0"). These numbers are inexact and have no exact counterpart.
2513 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
2514 sign of the dividend. The infinities are integers, and they answer #t
2515 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
2516 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
2527 ERROR: Numerical overflow
2529 Two new predicates 'inf?' and 'nan?' can be used to test for the
2532 ** Inexact zero can have a sign.
2534 Guile can now distinguish between plus and minus inexact zero, if your
2535 platform supports this, too. The two zeros are equal according to
2536 '=', but not according to 'eqv?'. For example
2547 ** Guile now has exact rationals.
2549 Guile can now represent fractions such as 1/3 exactly. Computing with
2550 them is also done exactly, of course:
2555 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
2556 for exact arguments.
2558 For example: (floor 2) now returns an exact 2 where in the past it
2559 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
2561 ** inexact->exact no longer returns only integers.
2563 Without exact rationals, the closest exact number was always an
2564 integer, but now inexact->exact returns the fraction that is exactly
2565 equal to a floating point number. For example:
2567 (inexact->exact 1.234)
2568 => 694680242521899/562949953421312
2570 When you want the old behavior, use 'round' explicitly:
2572 (inexact->exact (round 1.234))
2575 ** New function 'rationalize'.
2577 This function finds a simple fraction that is close to a given real
2578 number. For example (and compare with inexact->exact above):
2580 (rationalize (inexact->exact 1.234) 1/2000)
2583 Note that, as required by R5RS, rationalize returns only then an exact
2584 result when both its arguments are exact.
2586 ** 'odd?' and 'even?' work also for inexact integers.
2588 Previously, (odd? 1.0) would signal an error since only exact integers
2589 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
2590 returns #f and (odd? 1.5) signals an error.
2592 ** Guile now has uninterned symbols.
2594 The new function 'make-symbol' will return an uninterned symbol. This
2595 is a symbol that is unique and is guaranteed to remain unique.
2596 However, uninterned symbols can not yet be read back in.
2598 Use the new function 'symbol-interned?' to check whether a symbol is
2601 ** pretty-print has more options.
2603 The function pretty-print from the (ice-9 pretty-print) module can now
2604 also be invoked with keyword arguments that control things like
2605 maximum output width. See the manual for details.
2607 ** Variables have no longer a special behavior for `equal?'.
2609 Previously, comparing two variables with `equal?' would recursivly
2610 compare their values. This is no longer done. Variables are now only
2611 `equal?' if they are `eq?'.
2613 ** `(begin)' is now valid.
2615 You can now use an empty `begin' form. It will yield #<unspecified>
2616 when evaluated and simply be ignored in a definition context.
2618 ** Deprecated: procedure->macro
2620 Change your code to use 'define-macro' or r5rs macros. Also, be aware
2621 that macro expansion will not be done during evaluation, but prior to
2624 ** Soft ports now allow a `char-ready?' procedure
2626 The vector argument to `make-soft-port' can now have a length of
2627 either 5 or 6. (Previously the length had to be 5.) The optional 6th
2628 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
2629 that returns the number of characters that can be read immediately
2630 without the soft port blocking.
2632 ** Deprecated: undefine
2634 There is no replacement for undefine.
2636 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
2637 have been discouraged.
2639 They are relics from a time where a keyword like #:foo was used
2640 directly as a Tcl option "-foo" and thus keywords were internally
2641 stored as a symbol with a starting dash. We now store a symbol
2644 Use symbol->keyword and keyword->symbol instead.
2646 ** The `cheap' debug option is now obsolete
2648 Evaluator trap calls are now unconditionally "cheap" - in other words,
2649 they pass a debug object to the trap handler rather than a full
2650 continuation. The trap handler code can capture a full continuation
2651 by using `call-with-current-continuation' in the usual way, if it so
2654 The `cheap' option is retained for now so as not to break existing
2655 code which gets or sets it, but setting it now has no effect. It will
2656 be removed in the next major Guile release.
2658 ** Evaluator trap calls now support `tweaking'
2660 `Tweaking' means that the trap handler code can modify the Scheme
2661 expression that is about to be evaluated (in the case of an
2662 enter-frame trap) or the value that is being returned (in the case of
2663 an exit-frame trap). The trap handler code indicates that it wants to
2664 do this by returning a pair whose car is the symbol 'instead and whose
2665 cdr is the modified expression or return value.
2667 * Changes to the C interface
2669 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
2670 take a 'delete' function argument.
2672 This argument makes no sense since the delete function is used to
2673 remove a pair from an alist, and this must not be configurable.
2675 This is an incompatible change.
2677 ** The GH interface is now subject to the deprecation mechanism
2679 The GH interface has been deprecated for quite some time but now it is
2680 actually removed from Guile when it is configured with
2681 --disable-deprecated.
2683 See the manual "Transitioning away from GH" for more information.
2685 ** A new family of functions for converting between C values and
2686 Scheme values has been added.
2688 These functions follow a common naming scheme and are designed to be
2689 easier to use, thread-safe and more future-proof than the older
2692 - int scm_is_* (...)
2694 These are predicates that return a C boolean: 1 or 0. Instead of
2695 SCM_NFALSEP, you can now use scm_is_true, for example.
2697 - <type> scm_to_<type> (SCM val, ...)
2699 These are functions that convert a Scheme value into an appropriate
2700 C value. For example, you can use scm_to_int to safely convert from
2703 - SCM scm_from_<type> (<type> val, ...)
2705 These functions convert from a C type to a SCM value; for example,
2706 scm_from_int for ints.
2708 There is a huge number of these functions, for numbers, strings,
2709 symbols, vectors, etc. They are documented in the reference manual in
2710 the API section together with the types that they apply to.
2712 ** New functions for dealing with complex numbers in C have been added.
2714 The new functions are scm_c_make_rectangular, scm_c_make_polar,
2715 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
2716 They work like scm_make_rectangular etc but take or return doubles
2719 ** The function scm_make_complex has been discouraged.
2721 Use scm_c_make_rectangular instead.
2723 ** The INUM macros have been deprecated.
2725 A lot of code uses these macros to do general integer conversions,
2726 although the macros only work correctly with fixnums. Use the
2727 following alternatives.
2729 SCM_INUMP -> scm_is_integer or similar
2730 SCM_NINUMP -> !scm_is_integer or similar
2731 SCM_MAKINUM -> scm_from_int or similar
2732 SCM_INUM -> scm_to_int or similar
2734 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
2735 do the validating for you.
2737 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
2738 have been discouraged.
2740 Use the newer scm_to_<type> and scm_from_<type> functions instead for
2741 new code. The functions have been discouraged since they don't fit
2744 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
2746 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
2747 evaluates its argument twice. Use scm_is_true, etc. instead for new
2750 ** The macro SCM_EQ_P has been discouraged.
2752 Use scm_is_eq for new code, which fits better into the naming
2755 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
2758 Use the function scm_is_pair or scm_is_null instead.
2760 ** The functions scm_round and scm_truncate have been deprecated and
2761 are now available as scm_c_round and scm_c_truncate, respectively.
2763 These functions occupy the names that scm_round_number and
2764 scm_truncate_number should have.
2766 ** The functions scm_c_string2str, scm_c_substring2str, and
2767 scm_c_symbol2str have been deprecated.
2769 Use scm_to_locale_stringbuf or similar instead, maybe together with
2772 ** New functions scm_c_make_string, scm_c_string_length,
2773 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
2774 scm_c_substring_shared, scm_c_substring_copy.
2776 These are like scm_make_string, scm_length, etc. but are slightly
2777 easier to use from C.
2779 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
2780 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
2782 They export too many assumptions about the implementation of strings
2783 and symbols that are no longer true in the presence of
2784 mutation-sharing substrings and when Guile switches to some form of
2787 When working with strings, it is often best to use the normal string
2788 functions provided by Guile, such as scm_c_string_ref,
2789 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
2790 manual since many more such functions are now provided than
2793 When you want to convert a SCM string to a C string, use the
2794 scm_to_locale_string function or similar instead. For symbols, use
2795 scm_symbol_to_string and then work with that string. Because of the
2796 new string representation, scm_symbol_to_string does not need to copy
2797 and is thus quite efficient.
2799 ** Some string, symbol and keyword functions have been discouraged.
2801 They don't fit into the uniform naming scheme and are not explicit
2802 about the character encoding.
2804 Replace according to the following table:
2806 scm_allocate_string -> scm_c_make_string
2807 scm_take_str -> scm_take_locale_stringn
2808 scm_take0str -> scm_take_locale_string
2809 scm_mem2string -> scm_from_locale_stringn
2810 scm_str2string -> scm_from_locale_string
2811 scm_makfrom0str -> scm_from_locale_string
2812 scm_mem2symbol -> scm_from_locale_symboln
2813 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
2814 scm_str2symbol -> scm_from_locale_symbol
2816 SCM_SYMBOL_HASH -> scm_hashq
2817 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
2819 scm_c_make_keyword -> scm_from_locale_keyword
2821 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
2822 now also available to C code.
2824 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
2826 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
2827 the latter returns the true name of the keyword, not the 'dash name',
2828 as SCM_KEYWORDSYM used to do.
2830 ** A new way to access arrays in a thread-safe and efficient way has
2833 See the manual, node "Accessing Arrays From C".
2835 ** The old uniform vector and bitvector implementations have been
2836 unceremoniously removed.
2838 This implementation exposed the details of the tagging system of
2839 Guile. Use the new C API explained in the manual in node "Uniform
2840 Numeric Vectors" and "Bit Vectors", respectively.
2842 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
2843 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
2844 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
2845 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
2846 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
2847 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
2850 ** The macros dealing with vectors have been deprecated.
2852 Use the new functions scm_is_vector, scm_vector_elements,
2853 scm_vector_writable_elements, etc, or scm_is_simple_vector,
2854 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
2855 manual for more details.
2857 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
2858 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
2860 The following macros have been removed: SCM_VECTOR_BASE,
2861 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
2862 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
2864 ** Some C functions and macros related to arrays have been deprecated.
2866 Migrate according to the following table:
2868 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
2869 scm_make_ra -> scm_make_array
2870 scm_shap2ra -> scm_make_array
2871 scm_cvref -> scm_c_generalized_vector_ref
2872 scm_ra_set_contp -> do not use
2873 scm_aind -> scm_array_handle_pos
2874 scm_raprin1 -> scm_display or scm_write
2876 SCM_ARRAYP -> scm_is_array
2877 SCM_ARRAY_NDIM -> scm_c_array_rank
2878 SCM_ARRAY_DIMS -> scm_array_handle_dims
2879 SCM_ARRAY_CONTP -> do not use
2880 SCM_ARRAY_MEM -> do not use
2881 SCM_ARRAY_V -> scm_array_handle_elements or similar
2882 SCM_ARRAY_BASE -> do not use
2884 ** SCM_CELL_WORD_LOC has been deprecated.
2886 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
2887 to a SCM, as opposed to a pointer to a scm_t_bits.
2889 This was done to allow the correct use of pointers into the Scheme
2890 heap. Previously, the heap words were of type scm_t_bits and local
2891 variables and function arguments were of type SCM, making it
2892 non-standards-conformant to have a pointer that can point to both.
2894 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
2896 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
2897 second and third words of double smobs. Likewise for
2898 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
2900 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
2901 used to get and set the 16 exra bits in the zeroth word of a smob.
2903 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
2904 accesing the first immediate word of a smob as a SCM value, and there
2905 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
2906 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
2908 ** New way to deal with non-local exits and re-entries.
2910 There is a new set of functions that essentially do what
2911 scm_internal_dynamic_wind does, but in a way that is more convenient
2912 for C code in some situations. Here is a quick example of how to
2913 prevent a potential memory leak:
2920 scm_dynwind_begin (0);
2922 mem = scm_malloc (100);
2923 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
2925 /* MEM would leak if BAR throws an error.
2926 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
2933 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
2934 SCM_DYNWIND_END as well.
2938 For full documentation, see the node "Dynamic Wind" in the manual.
2940 ** New function scm_dynwind_free
2942 This function calls 'free' on a given pointer when a dynwind context
2943 is left. Thus the call to scm_dynwind_unwind_handler above could be
2944 replaced with simply scm_dynwind_free (mem).
2946 ** New functions scm_c_call_with_blocked_asyncs and
2947 scm_c_call_with_unblocked_asyncs
2949 Like scm_call_with_blocked_asyncs etc. but for C functions.
2951 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
2953 In addition to scm_c_call_with_blocked_asyncs you can now also use
2954 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
2955 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
2957 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
2958 SCM_REALLOW_INTS have been deprecated.
2960 They do no longer fulfill their original role of blocking signal
2961 delivery. Depending on what you want to achieve, replace a pair of
2962 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
2963 mutex, blocks asyncs, or both. See node "Critical Sections" in the
2966 ** The value 'scm_mask_ints' is no longer writable.
2968 Previously, you could set scm_mask_ints directly. This is no longer
2969 possible. Use scm_c_call_with_blocked_asyncs and
2970 scm_c_call_with_unblocked_asyncs instead.
2972 ** New way to temporarily set the current input, output or error ports
2974 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
2975 context' (see above). <foo> is one of "input", "output" or "error".
2977 ** New way to temporarily set fluids
2979 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
2980 above) to temporarily set the value of a fluid.
2982 ** New types scm_t_intmax and scm_t_uintmax.
2984 On platforms that have them, these types are identical to intmax_t and
2985 uintmax_t, respectively. On other platforms, they are identical to
2986 the largest integer types that Guile knows about.
2988 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
2990 You should not have used them.
2992 ** Many public #defines with generic names have been made private.
2994 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
2995 private or renamed with a more suitable public name.
2997 ** The macro SCM_TYP16S has been deprecated.
2999 This macro is not intended for public use.
3001 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
3003 Use scm_is_true (scm_inexact_p (...)) instead.
3005 ** The macro SCM_SLOPPY_REALP has been deprecated.
3007 Use scm_is_real instead.
3009 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
3011 Use scm_is_complex instead.
3013 ** Some preprocessor defines have been deprecated.
3015 These defines indicated whether a certain feature was present in Guile
3016 or not. Going forward, assume that the features are always present.
3018 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
3019 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
3021 The following macros have been removed completely: MEMOIZE_LOCALS,
3022 SCM_RECKLESS, SCM_CAUTIOUS.
3024 ** The preprocessor define STACK_DIRECTION has been deprecated.
3026 There should be no need to know about the stack direction for ordinary
3029 ** New function: scm_effective_version
3031 Returns the "effective" version number. This is just the normal full
3032 version string without the final micro-version number. See "Changes
3033 to the distribution" above.
3035 ** The function scm_call_with_new_thread has a new prototype.
3037 Instead of taking a list with the thunk and handler, these two
3038 arguments are now passed directly:
3040 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
3042 This is an incompatible change.
3044 ** New snarfer macro SCM_DEFINE_PUBLIC.
3046 This is like SCM_DEFINE, but also calls scm_c_export for the defined
3047 function in the init section.
3049 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
3051 ** Garbage collector rewrite.
3053 The garbage collector is cleaned up a lot, and now uses lazy
3054 sweeping. This is reflected in the output of (gc-stats); since cells
3055 are being freed when they are allocated, the cells-allocated field
3056 stays roughly constant.
3058 For malloc related triggers, the behavior is changed. It uses the same
3059 heuristic as the cell-triggered collections. It may be tuned with the
3060 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
3061 for minimum yield of malloc related triggers. The default is 40.
3062 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
3065 Debugging operations for the freelist have been deprecated, along with
3066 the C variables that control garbage collection. The environment
3067 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
3068 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
3070 For understanding the memory usage of a GUILE program, the routine
3071 gc-live-object-stats returns an alist containing the number of live
3072 objects for every type.
3075 ** The function scm_definedp has been renamed to scm_defined_p
3077 The name scm_definedp is deprecated.
3079 ** The struct scm_cell type has been renamed to scm_t_cell
3081 This is in accordance to Guile's naming scheme for types. Note that
3082 the name scm_cell is now used for a function that allocates and
3083 initializes a new cell (see below).
3085 ** New functions for memory management
3087 A new set of functions for memory management has been added since the
3088 old way (scm_must_malloc, scm_must_free, etc) was error prone and
3089 indeed, Guile itself contained some long standing bugs that could
3090 cause aborts in long running programs.
3092 The new functions are more symmetrical and do not need cooperation
3093 from smob free routines, among other improvements.
3095 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
3096 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
3097 scm_gc_free, scm_gc_register_collectable_memory, and
3098 scm_gc_unregister_collectable_memory. Refer to the manual for more
3099 details and for upgrading instructions.
3101 The old functions for memory management have been deprecated. They
3102 are: scm_must_malloc, scm_must_realloc, scm_must_free,
3103 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
3105 ** Declarations of exported features are marked with SCM_API.
3107 Every declaration of a feature that belongs to the exported Guile API
3108 has been marked by adding the macro "SCM_API" to the start of the
3109 declaration. This macro can expand into different things, the most
3110 common of which is just "extern" for Unix platforms. On Win32, it can
3111 be used to control which symbols are exported from a DLL.
3113 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
3114 will expand into "__declspec (dllimport) extern", which is needed for
3115 linking to the Guile DLL in Windows.
3117 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
3118 SCM_SRFI4_IMPORT, for the corresponding libraries.
3120 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
3122 Use the new functions scm_cell and scm_double_cell instead. The old
3123 macros had problems because with them allocation and initialization
3124 was separated and the GC could sometimes observe half initialized
3125 cells. Only careful coding by the user of SCM_NEWCELL and
3126 SCM_NEWCELL2 could make this safe and efficient.
3128 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
3130 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
3133 ** SRCBRKP has been deprecated.
3135 Use scm_c_source_property_breakpoint_p instead.
3137 ** Deprecated: scm_makmacro
3139 Change your code to use either scm_makmmacro or to define macros in
3140 Scheme, using 'define-macro'.
3142 ** New function scm_c_port_for_each.
3144 This function is like scm_port_for_each but takes a pointer to a C
3145 function as the callback instead of a SCM value.
3147 ** The names scm_internal_select, scm_thread_sleep, and
3148 scm_thread_usleep have been discouraged.
3150 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
3152 ** The GC can no longer be blocked.
3154 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
3155 The GC can now run (partially) concurrently with other code and thus
3156 blocking it is not well defined.
3158 ** Many definitions have been removed that were previously deprecated.
3160 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
3161 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
3162 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
3163 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
3164 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
3165 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
3166 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
3167 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
3168 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
3169 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
3170 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
3171 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
3172 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
3173 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
3174 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
3175 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
3176 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
3177 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
3178 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
3179 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
3180 scm_vector_set_length_x, scm_contregs, scm_debug_info,
3181 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
3182 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
3183 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
3184 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
3185 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
3186 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
3187 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
3188 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
3189 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
3190 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
3192 * Changes to bundled modules
3196 Using the (ice-9 debug) module no longer automatically switches Guile
3197 to use the debugging evaluator. If you want to switch to the
3198 debugging evaluator (which is needed for backtrace information if you
3199 hit an error), please add an explicit "(debug-enable 'debug)" to your
3200 code just after the code to use (ice-9 debug).
3203 Changes since Guile 1.4:
3205 * Changes to the distribution
3207 ** A top-level TODO file is included.
3209 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
3211 Guile now always uses three numbers to represent the version,
3212 i.e. "1.6.5". The first number, 1, is the major version number, the
3213 second number, 6, is the minor version number, and the third number,
3214 5, is the micro version number. Changes in major version number
3215 indicate major changes in Guile.
3217 Minor version numbers that are even denote stable releases, and odd
3218 minor version numbers denote development versions (which may be
3219 unstable). The micro version number indicates a minor sub-revision of
3220 a given MAJOR.MINOR release.
3222 In keeping with the new scheme, (minor-version) and scm_minor_version
3223 no longer return everything but the major version number. They now
3224 just return the minor version number. Two new functions
3225 (micro-version) and scm_micro_version have been added to report the
3226 micro version number.
3228 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
3230 ** New preprocessor definitions are available for checking versions.
3232 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
3233 SCM_MICRO_VERSION to the appropriate integer values.
3235 ** Guile now actively warns about deprecated features.
3237 The new configure option `--enable-deprecated=LEVEL' and the
3238 environment variable GUILE_WARN_DEPRECATED control this mechanism.
3239 See INSTALL and README for more information.
3241 ** Guile is much more likely to work on 64-bit architectures.
3243 Guile now compiles and passes "make check" with only two UNRESOLVED GC
3244 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
3245 for the use of a test machine, and thanks to Stefan Jahn for ia64
3248 ** New functions: setitimer and getitimer.
3250 These implement a fairly direct interface to the libc functions of the
3253 ** The #. reader extension is now disabled by default.
3255 For safety reasons, #. evaluation is disabled by default. To
3256 re-enable it, set the fluid read-eval? to #t. For example:
3258 (fluid-set! read-eval? #t)
3260 but make sure you realize the potential security risks involved. With
3261 read-eval? enabled, reading a data file from an untrusted source can
3264 ** New SRFI modules have been added:
3266 SRFI-0 `cond-expand' is now supported in Guile, without requiring
3269 (srfi srfi-1) is a library containing many useful pair- and list-processing
3272 (srfi srfi-2) exports and-let*.
3274 (srfi srfi-4) implements homogeneous numeric vector datatypes.
3276 (srfi srfi-6) is a dummy module for now, since guile already provides
3277 all of the srfi-6 procedures by default: open-input-string,
3278 open-output-string, get-output-string.
3280 (srfi srfi-8) exports receive.
3282 (srfi srfi-9) exports define-record-type.
3284 (srfi srfi-10) exports define-reader-ctor and implements the reader
3287 (srfi srfi-11) exports let-values and let*-values.
3289 (srfi srfi-13) implements the SRFI String Library.
3291 (srfi srfi-14) implements the SRFI Character-Set Library.
3293 (srfi srfi-17) implements setter and getter-with-setter and redefines
3294 some accessor procedures as procedures with getters. (such as car,
3295 cdr, vector-ref etc.)
3297 (srfi srfi-19) implements the SRFI Time/Date Library.
3299 ** New scripts / "executable modules"
3301 Subdirectory "scripts" contains Scheme modules that are packaged to
3302 also be executable as scripts. At this time, these scripts are available:
3311 See README there for more info.
3313 These scripts can be invoked from the shell with the new program
3314 "guile-tools", which keeps track of installation directory for you.
3317 $ guile-tools display-commentary srfi/*.scm
3319 guile-tools is copied to the standard $bindir on "make install".
3321 ** New module (ice-9 stack-catch):
3323 stack-catch is like catch, but saves the current state of the stack in
3324 the fluid the-last-stack. This fluid can be useful when using the
3325 debugger and when re-throwing an error.
3327 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
3329 This has been done to prevent problems on lesser operating systems
3330 that can't tolerate `*'s in file names. The exported macro continues
3331 to be named `and-let*', of course.
3333 On systems that support it, there is also a compatibility module named
3334 (ice-9 and-let*). It will go away in the next release.
3336 ** New modules (oop goops) etc.:
3339 (oop goops describe)
3341 (oop goops active-slot)
3342 (oop goops composite-slot)
3344 The Guile Object Oriented Programming System (GOOPS) has been
3345 integrated into Guile. For further information, consult the GOOPS
3346 manual and tutorial in the `doc' directory.
3348 ** New module (ice-9 rdelim).
3350 This exports the following procedures which were previously defined
3351 in the default environment:
3353 read-line read-line! read-delimited read-delimited! %read-delimited!
3354 %read-line write-line
3356 For backwards compatibility the definitions are still imported into the
3357 default environment in this version of Guile. However you should add:
3359 (use-modules (ice-9 rdelim))
3361 to any program which uses the definitions, since this may change in
3364 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
3365 can be used for similar functionality.
3367 ** New module (ice-9 rw)
3369 This is a subset of the (scsh rw) module from guile-scsh. Currently
3370 it defines two procedures:
3372 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
3374 Read characters from a port or file descriptor into a string STR.
3375 A port must have an underlying file descriptor -- a so-called
3376 fport. This procedure is scsh-compatible and can efficiently read
3379 *** New function: write-string/partial str [port_or_fdes [start [end]]]
3381 Write characters from a string STR to a port or file descriptor.
3382 A port must have an underlying file descriptor -- a so-called
3383 fport. This procedure is mostly compatible and can efficiently
3384 write large strings.
3386 ** New module (ice-9 match)
3388 This module includes Andrew K. Wright's pattern matcher. See
3389 ice-9/match.scm for brief description or
3391 http://www.star-lab.com/wright/code.html
3393 for complete documentation.
3395 ** New module (ice-9 buffered-input)
3397 This module provides procedures to construct an input port from an
3398 underlying source of input that reads and returns its input in chunks.
3399 The underlying input source is a Scheme procedure, specified by the
3400 caller, which the port invokes whenever it needs more input.
3402 This is useful when building an input port whose back end is Readline
3403 or a UI element such as the GtkEntry widget.
3407 The reference and tutorial documentation that was previously
3408 distributed separately, as `guile-doc', is now included in the core
3409 Guile distribution. The documentation consists of the following
3412 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
3415 - The Guile Reference Manual (guile.texi) contains (or is intended to
3416 contain) reference documentation on all aspects of Guile.
3418 - The GOOPS Manual (goops.texi) contains both tutorial-style and
3419 reference documentation for using GOOPS, Guile's Object Oriented
3422 - The Revised^5 Report on the Algorithmic Language Scheme
3425 See the README file in the `doc' directory for more details.
3427 ** There are a couple of examples in the examples/ directory now.
3429 * Changes to the stand-alone interpreter
3431 ** New command line option `--use-srfi'
3433 Using this option, SRFI modules can be loaded on startup and be
3434 available right from the beginning. This makes programming portable
3435 Scheme programs easier.
3437 The option `--use-srfi' expects a comma-separated list of numbers,
3438 each representing a SRFI number to be loaded into the interpreter
3439 before starting evaluating a script file or the REPL. Additionally,
3440 the feature identifier for the loaded SRFIs is recognized by
3441 `cond-expand' when using this option.
3444 $ guile --use-srfi=8,13
3445 guile> (receive (x z) (values 1 2) (+ 1 2))
3447 guile> (string-pad "bla" 20)
3450 ** Guile now always starts up in the `(guile-user)' module.
3452 Previously, scripts executed via the `-s' option would run in the
3453 `(guile)' module and the repl would run in the `(guile-user)' module.
3454 Now every user action takes place in the `(guile-user)' module by
3457 * Changes to Scheme functions and syntax
3459 ** Character classifiers work for non-ASCII characters.
3461 The predicates `char-alphabetic?', `char-numeric?',
3462 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
3463 no longer check whether their arguments are ASCII characters.
3464 Previously, a character would only be considered alphabetic when it
3465 was also ASCII, for example.
3467 ** Previously deprecated Scheme functions have been removed:
3469 tag - no replacement.
3470 fseek - replaced by seek.
3471 list* - replaced by cons*.
3473 ** It's now possible to create modules with controlled environments
3477 (use-modules (ice-9 safe))
3478 (define m (make-safe-module))
3479 ;;; m will now be a module containing only a safe subset of R5RS
3480 (eval '(+ 1 2) m) --> 3
3481 (eval 'load m) --> ERROR: Unbound variable: load
3483 ** Evaluation of "()", the empty list, is now an error.
3485 Previously, the expression "()" evaluated to the empty list. This has
3486 been changed to signal a "missing expression" error. The correct way
3487 to write the empty list as a literal constant is to use quote: "'()".
3489 ** New concept of `Guile Extensions'.
3491 A Guile Extension is just a ordinary shared library that can be linked
3492 at run-time. We found it advantageous to give this simple concept a
3493 dedicated name to distinguish the issues related to shared libraries
3494 from the issues related to the module system.
3496 *** New function: load-extension
3498 Executing (load-extension lib init) is mostly equivalent to
3500 (dynamic-call init (dynamic-link lib))
3502 except when scm_register_extension has been called previously.
3503 Whenever appropriate, you should use `load-extension' instead of
3504 dynamic-link and dynamic-call.
3506 *** New C function: scm_c_register_extension
3508 This function registers a initialization function for use by
3509 `load-extension'. Use it when you don't want specific extensions to
3510 be loaded as shared libraries (for example on platforms that don't
3511 support dynamic linking).
3513 ** Auto-loading of compiled-code modules is deprecated.
3515 Guile used to be able to automatically find and link a shared
3516 library to satisfy requests for a module. For example, the module
3517 `(foo bar)' could be implemented by placing a shared library named
3518 "foo/libbar.so" (or with a different extension) in a directory on the
3521 This has been found to be too tricky, and is no longer supported. The
3522 shared libraries are now called "extensions". You should now write a
3523 small Scheme file that calls `load-extension' to load the shared
3524 library and initialize it explicitly.
3526 The shared libraries themselves should be installed in the usual
3527 places for shared libraries, with names like "libguile-foo-bar".
3529 For example, place this into a file "foo/bar.scm"
3531 (define-module (foo bar))
3533 (load-extension "libguile-foo-bar" "foobar_init")
3535 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
3537 `eval' is now R5RS, that is it takes two arguments.
3538 The second argument is an environment specifier, i.e. either
3540 (scheme-report-environment 5)
3541 (null-environment 5)
3542 (interaction-environment)
3548 ** The module system has been made more disciplined.
3550 The function `eval' will save and restore the current module around
3551 the evaluation of the specified expression. While this expression is
3552 evaluated, `(current-module)' will now return the right module, which
3553 is the module specified as the second argument to `eval'.
3555 A consequence of this change is that `eval' is not particularly
3556 useful when you want allow the evaluated code to change what module is
3557 designated as the current module and have this change persist from one
3558 call to `eval' to the next. The read-eval-print-loop is an example
3559 where `eval' is now inadequate. To compensate, there is a new
3560 function `primitive-eval' that does not take a module specifier and
3561 that does not save/restore the current module. You should use this
3562 function together with `set-current-module', `current-module', etc
3563 when you want to have more control over the state that is carried from
3564 one eval to the next.
3566 Additionally, it has been made sure that forms that are evaluated at
3567 the top level are always evaluated with respect to the current module.
3568 Previously, subforms of top-level forms such as `begin', `case',
3569 etc. did not respect changes to the current module although these
3570 subforms are at the top-level as well.
3572 To prevent strange behavior, the forms `define-module',
3573 `use-modules', `use-syntax', and `export' have been restricted to only
3574 work on the top level. The forms `define-public' and
3575 `defmacro-public' only export the new binding on the top level. They
3576 behave just like `define' and `defmacro', respectively, when they are
3577 used in a lexical environment.
3579 Also, `export' will no longer silently re-export bindings imported
3580 from a used module. It will emit a `deprecation' warning and will
3581 cease to perform any re-export in the next version. If you actually
3582 want to re-export bindings, use the new `re-export' in place of
3583 `export'. The new `re-export' will not make copies of variables when
3584 rexporting them, as `export' did wrongly.
3586 ** Module system now allows selection and renaming of imported bindings
3588 Previously, when using `use-modules' or the `#:use-module' clause in
3589 the `define-module' form, all the bindings (association of symbols to
3590 values) for imported modules were added to the "current module" on an
3591 as-is basis. This has been changed to allow finer control through two
3592 new facilities: selection and renaming.
3594 You can now select which of the imported module's bindings are to be
3595 visible in the current module by using the `:select' clause. This
3596 clause also can be used to rename individual bindings. For example:
3598 ;; import all bindings no questions asked
3599 (use-modules (ice-9 common-list))
3601 ;; import four bindings, renaming two of them;
3602 ;; the current module sees: every some zonk-y zonk-n
3603 (use-modules ((ice-9 common-list)
3605 (remove-if . zonk-y)
3606 (remove-if-not . zonk-n))))
3608 You can also programmatically rename all selected bindings using the
3609 `:renamer' clause, which specifies a proc that takes a symbol and
3610 returns another symbol. Because it is common practice to use a prefix,
3611 we now provide the convenience procedure `symbol-prefix-proc'. For
3614 ;; import four bindings, renaming two of them specifically,
3615 ;; and all four w/ prefix "CL:";
3616 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
3617 (use-modules ((ice-9 common-list)
3619 (remove-if . zonk-y)
3620 (remove-if-not . zonk-n))
3621 :renamer (symbol-prefix-proc 'CL:)))
3623 ;; import four bindings, renaming two of them specifically,
3624 ;; and all four by upcasing.
3625 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
3626 (define (upcase-symbol sym)
3627 (string->symbol (string-upcase (symbol->string sym))))
3629 (use-modules ((ice-9 common-list)
3631 (remove-if . zonk-y)
3632 (remove-if-not . zonk-n))
3633 :renamer upcase-symbol))
3635 Note that programmatic renaming is done *after* individual renaming.
3636 Also, the above examples show `use-modules', but the same facilities are
3637 available for the `#:use-module' clause of `define-module'.
3639 See manual for more info.
3641 ** The semantics of guardians have changed.
3643 The changes are for the most part compatible. An important criterion
3644 was to keep the typical usage of guardians as simple as before, but to
3645 make the semantics safer and (as a result) more useful.
3647 *** All objects returned from guardians are now properly alive.
3649 It is now guaranteed that any object referenced by an object returned
3650 from a guardian is alive. It's now impossible for a guardian to
3651 return a "contained" object before its "containing" object.
3653 One incompatible (but probably not very important) change resulting
3654 from this is that it is no longer possible to guard objects that
3655 indirectly reference themselves (i.e. are parts of cycles). If you do
3656 so accidentally, you'll get a warning.
3658 *** There are now two types of guardians: greedy and sharing.
3660 If you call (make-guardian #t) or just (make-guardian), you'll get a
3661 greedy guardian, and for (make-guardian #f) a sharing guardian.
3663 Greedy guardians are the default because they are more "defensive".
3664 You can only greedily guard an object once. If you guard an object
3665 more than once, once in a greedy guardian and the rest of times in
3666 sharing guardians, then it is guaranteed that the object won't be
3667 returned from sharing guardians as long as it is greedily guarded
3670 Guardians returned by calls to `make-guardian' can now take one more
3671 optional parameter, which says whether to throw an error in case an
3672 attempt is made to greedily guard an object that is already greedily
3673 guarded. The default is true, i.e. throw an error. If the parameter
3674 is false, the guardian invocation returns #t if guarding was
3675 successful and #f if it wasn't.
3677 Also, since greedy guarding is, in effect, a side-effecting operation
3678 on objects, a new function is introduced: `destroy-guardian!'.
3679 Invoking this function on a guardian renders it unoperative and, if
3680 the guardian is greedy, clears the "greedily guarded" property of the
3681 objects that were guarded by it, thus undoing the side effect.
3683 Note that all this hair is hardly very important, since guardian
3684 objects are usually permanent.
3686 ** Continuations created by call-with-current-continuation now accept
3687 any number of arguments, as required by R5RS.
3689 ** New function `issue-deprecation-warning'
3691 This function is used to display the deprecation messages that are
3692 controlled by GUILE_WARN_DEPRECATION as explained in the README.
3695 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
3699 ;; `id' is deprecated. Use `identity' instead.
3704 ** New syntax `begin-deprecated'
3706 When deprecated features are included (as determined by the configure
3707 option --enable-deprecated), `begin-deprecated' is identical to
3708 `begin'. When deprecated features are excluded, it always evaluates
3709 to `#f', ignoring the body forms.
3711 ** New function `make-object-property'
3713 This function returns a new `procedure with setter' P that can be used
3714 to attach a property to objects. When calling P as
3718 where `obj' is any kind of object, it attaches `val' to `obj' in such
3719 a way that it can be retrieved by calling P as
3723 This function will replace procedure properties, symbol properties and
3724 source properties eventually.
3726 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
3728 Instead of #&optional, #&key, etc you should now use #:optional,
3729 #:key, etc. Since #:optional is a keyword, you can write it as just
3730 :optional when (read-set! keywords 'prefix) is active.
3732 The old reader syntax `#&' is still supported, but deprecated. It
3733 will be removed in the next release.
3735 ** New define-module option: pure
3737 Tells the module system not to include any bindings from the root
3742 (define-module (totally-empty-module)
3745 ** New define-module option: export NAME1 ...
3747 Export names NAME1 ...
3749 This option is required if you want to be able to export bindings from
3750 a module which doesn't import one of `define-public' or `export'.
3754 (define-module (foo)
3756 :use-module (ice-9 r5rs)
3759 ;;; Note that we're pure R5RS below this point!
3764 ** New function: object->string OBJ
3766 Return a Scheme string obtained by printing a given object.
3768 ** New function: port? X
3770 Returns a boolean indicating whether X is a port. Equivalent to
3771 `(or (input-port? X) (output-port? X))'.
3773 ** New function: file-port?
3775 Determines whether a given object is a port that is related to a file.
3777 ** New function: port-for-each proc
3779 Apply PROC to each port in the Guile port table in turn. The return
3780 value is unspecified. More specifically, PROC is applied exactly once
3781 to every port that exists in the system at the time PORT-FOR-EACH is
3782 invoked. Changes to the port table while PORT-FOR-EACH is running
3783 have no effect as far as PORT-FOR-EACH is concerned.
3785 ** New function: dup2 oldfd newfd
3787 A simple wrapper for the `dup2' system call. Copies the file
3788 descriptor OLDFD to descriptor number NEWFD, replacing the
3789 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
3790 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
3791 to move away ports which are using NEWFD. The return value is
3794 ** New function: close-fdes fd
3796 A simple wrapper for the `close' system call. Close file
3797 descriptor FD, which must be an integer. Unlike close (*note
3798 close: Ports and File Descriptors.), the file descriptor will be
3799 closed even if a port is using it. The return value is
3802 ** New function: crypt password salt
3804 Encrypts `password' using the standard unix password encryption
3807 ** New function: chroot path
3809 Change the root directory of the running process to `path'.
3811 ** New functions: getlogin, cuserid
3813 Return the login name or the user name of the current effective user
3816 ** New functions: getpriority which who, setpriority which who prio
3818 Get or set the priority of the running process.
3820 ** New function: getpass prompt
3822 Read a password from the terminal, first displaying `prompt' and
3825 ** New function: flock file operation
3827 Set/remove an advisory shared or exclusive lock on `file'.
3829 ** New functions: sethostname name, gethostname
3831 Set or get the hostname of the machine the current process is running
3834 ** New function: mkstemp! tmpl
3836 mkstemp creates a new unique file in the file system and returns a
3837 new buffered port open for reading and writing to the file. TMPL
3838 is a string specifying where the file should be created: it must
3839 end with `XXXXXX' and will be changed in place to return the name
3840 of the temporary file.
3842 ** New function: open-input-string string
3844 Return an input string port which delivers the characters from
3845 `string'. This procedure, together with `open-output-string' and
3846 `get-output-string' implements SRFI-6.
3848 ** New function: open-output-string
3850 Return an output string port which collects all data written to it.
3851 The data can then be retrieved by `get-output-string'.
3853 ** New function: get-output-string
3855 Return the contents of an output string port.
3857 ** New function: identity
3859 Return the argument.
3861 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
3862 are represented in Scheme as integers with normal host byte ordering.
3864 ** New function: inet-pton family address
3866 Convert a printable string network address into an integer. Note that
3867 unlike the C version of this function, the result is an integer with
3868 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3871 (inet-pton AF_INET "127.0.0.1") => 2130706433
3872 (inet-pton AF_INET6 "::1") => 1
3874 ** New function: inet-ntop family address
3876 Convert an integer network address into a printable string. Note that
3877 unlike the C version of this function, the input is an integer with
3878 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3881 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
3882 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
3883 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
3887 Use `identity' instead.
3893 ** Deprecated: return-it
3897 ** Deprecated: string-character-length
3899 Use `string-length' instead.
3901 ** Deprecated: flags
3903 Use `logior' instead.
3905 ** Deprecated: close-all-ports-except.
3907 This was intended for closing ports in a child process after a fork,
3908 but it has the undesirable side effect of flushing buffers.
3909 port-for-each is more flexible.
3911 ** The (ice-9 popen) module now attempts to set up file descriptors in
3912 the child process from the current Scheme ports, instead of using the
3913 current values of file descriptors 0, 1, and 2 in the parent process.
3915 ** Removed function: builtin-weak-bindings
3917 There is no such concept as a weak binding any more.
3919 ** Removed constants: bignum-radix, scm-line-incrementors
3921 ** define-method: New syntax mandatory.
3923 The new method syntax is now mandatory:
3925 (define-method (NAME ARG-SPEC ...) BODY ...)
3926 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
3928 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
3929 REST-ARG ::= ARG-NAME
3931 If you have old code using the old syntax, import
3932 (oop goops old-define-method) before (oop goops) as in:
3934 (use-modules (oop goops old-define-method) (oop goops))
3936 ** Deprecated function: builtin-variable
3937 Removed function: builtin-bindings
3939 There is no longer a distinction between builtin or other variables.
3940 Use module system operations for all variables.
3942 ** Lazy-catch handlers are no longer allowed to return.
3944 That is, a call to `throw', `error', etc is now guaranteed to not
3947 ** Bugfixes for (ice-9 getopt-long)
3949 This module is now tested using test-suite/tests/getopt-long.test.
3950 The following bugs have been fixed:
3952 *** Parsing for options that are specified to have `optional' args now checks
3953 if the next element is an option instead of unconditionally taking it as the
3956 *** An error is now thrown for `--opt=val' when the option description
3957 does not specify `(value #t)' or `(value optional)'. This condition used to
3958 be accepted w/o error, contrary to the documentation.
3960 *** The error message for unrecognized options is now more informative.
3961 It used to be "not a record", an artifact of the implementation.
3963 *** The error message for `--opt' terminating the arg list (no value), when
3964 `(value #t)' is specified, is now more informative. It used to be "not enough
3967 *** "Clumped" single-char args now preserve trailing string, use it as arg.
3968 The expansion used to be like so:
3970 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
3972 Note that the "5d" is dropped. Now it is like so:
3974 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
3976 This enables single-char options to have adjoining arguments as long as their
3977 constituent characters are not potential single-char options.
3979 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
3981 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
3982 property `arglist', which can be retrieved by `arity'. The result is that
3983 `arity' can give more detailed information than before:
3987 guile> (use-modules (ice-9 optargs))
3988 guile> (define* (foo #:optional a b c) a)
3990 0 or more arguments in `lambda*:G0'.
3995 3 optional arguments: `a', `b' and `c'.
3996 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
3998 2 required arguments: `a' and `b', 2 keyword arguments: `c'
3999 and `d', other keywords allowed.
4000 guile> (define* (baz a b #:optional c #:rest r) a)
4002 2 required arguments: `a' and `b', 1 optional argument: `c',
4005 * Changes to the C interface
4007 ** Types have been renamed from scm_*_t to scm_t_*.
4009 This has been done for POSIX sake. It reserves identifiers ending
4010 with "_t". What a concept.
4012 The old names are still available with status `deprecated'.
4014 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
4016 ** Deprecated features have been removed.
4020 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
4021 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
4023 *** C Functions removed
4025 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
4026 scm_fseek - replaced by scm_seek.
4027 gc-thunk - replaced by after-gc-hook.
4028 gh_int2scmb - replaced by gh_bool2scm.
4029 scm_tc_dblr - replaced by scm_tc16_real.
4030 scm_tc_dblc - replaced by scm_tc16_complex.
4031 scm_list_star - replaced by scm_cons_star.
4033 ** Deprecated: scm_makfromstr
4035 Use scm_mem2string instead.
4037 ** Deprecated: scm_make_shared_substring
4039 Explicit shared substrings will disappear from Guile.
4041 Instead, "normal" strings will be implemented using sharing
4042 internally, combined with a copy-on-write strategy.
4044 ** Deprecated: scm_read_only_string_p
4046 The concept of read-only strings will disappear in next release of
4049 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
4051 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
4053 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
4055 Call a procedure with the indicated number of arguments. See "Fly
4056 Evaluation" in the manual.
4058 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
4060 Call a procedure with the indicated number of arguments and a list of
4061 further arguments. See "Fly Evaluation" in the manual.
4063 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
4065 Create a list of the given number of elements. See "List
4066 Constructors" in the manual.
4068 ** Renamed function: scm_listify has been replaced by scm_list_n.
4070 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
4071 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
4073 Use functions scm_list_N instead.
4075 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
4077 Used by an application to read arbitrary number of bytes from a port.
4078 Same semantics as libc read, except that scm_c_read only returns less
4079 than SIZE bytes if at end-of-file.
4081 Warning: Doesn't update port line and column counts!
4083 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
4085 Used by an application to write arbitrary number of bytes to an SCM
4086 port. Similar semantics as libc write. However, unlike libc
4087 write, scm_c_write writes the requested number of bytes and has no
4090 Warning: Doesn't update port line and column counts!
4092 ** New function: scm_init_guile ()
4094 In contrast to scm_boot_guile, scm_init_guile will return normally
4095 after initializing Guile. It is not available on all systems, tho.
4097 ** New functions: scm_str2symbol, scm_mem2symbol
4099 The function scm_str2symbol takes a const char* pointing to a zero-terminated
4100 field of characters and creates a scheme symbol object from that C string.
4101 The function scm_mem2symbol takes a const char* and a number of characters and
4102 creates a symbol from the characters in that memory area.
4104 ** New functions: scm_primitive_make_property
4105 scm_primitive_property_ref
4106 scm_primitive_property_set_x
4107 scm_primitive_property_del_x
4109 These functions implement a new way to deal with object properties.
4110 See libguile/properties.c for their documentation.
4112 ** New function: scm_done_free (long size)
4114 This function is the inverse of scm_done_malloc. Use it to report the
4115 amount of smob memory you free. The previous method, which involved
4116 calling scm_done_malloc with negative argument, was somewhat
4117 unintuitive (and is still available, of course).
4119 ** New function: scm_c_memq (SCM obj, SCM list)
4121 This function provides a fast C level alternative for scm_memq for the case
4122 that the list parameter is known to be a proper list. The function is a
4123 replacement for scm_sloppy_memq, but is stricter in its requirements on its
4124 list input parameter, since for anything else but a proper list the function's
4125 behaviour is undefined - it may even crash or loop endlessly. Further, for
4126 the case that the object is not found in the list, scm_c_memq returns #f which
4127 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
4129 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
4130 scm_remember_upto_here
4132 These functions replace the function scm_remember.
4134 ** Deprecated function: scm_remember
4136 Use one of the new functions scm_remember_upto_here_1,
4137 scm_remember_upto_here_2 or scm_remember_upto_here instead.
4139 ** New function: scm_allocate_string
4141 This function replaces the function scm_makstr.
4143 ** Deprecated function: scm_makstr
4145 Use the new function scm_allocate_string instead.
4147 ** New global variable scm_gc_running_p introduced.
4149 Use this variable to find out if garbage collection is being executed. Up to
4150 now applications have used scm_gc_heap_lock to test if garbage collection was
4151 running, which also works because of the fact that up to know only the garbage
4152 collector has set this variable. But, this is an implementation detail that
4153 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
4154 of this variable is (and has been) not fully safe anyway.
4156 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
4158 Use these instead of SCM_LENGTH_MAX.
4160 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
4161 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
4162 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
4164 Use these instead of SCM_LENGTH.
4166 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
4167 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
4168 SCM_SET_BITVECTOR_LENGTH
4170 Use these instead of SCM_SETLENGTH
4172 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
4173 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
4176 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
4179 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
4180 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
4183 Use these instead of SCM_SETCHARS.
4185 ** New macro: SCM_BITVECTOR_P
4187 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
4189 Use instead of SCM_COERCE_SUBSTR.
4191 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
4193 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
4195 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
4196 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
4197 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
4198 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
4199 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
4200 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
4201 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
4202 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
4203 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
4204 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
4205 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
4206 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
4207 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
4208 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
4209 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
4211 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
4212 Use scm_memory_error instead of SCM_NALLOC.
4213 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
4214 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
4215 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
4216 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
4217 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
4218 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
4219 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
4220 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
4221 Use SCM_STRINGP instead of SCM_RWSTRINGP.
4222 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
4223 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
4224 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
4225 Use a type specific setter macro instead of SCM_SETLENGTH.
4226 Use a type specific setter macro instead of SCM_SETCHARS.
4227 Use a type specific length macro instead of SCM_LENGTH_MAX.
4228 Use SCM_GCMARKP instead of SCM_GC8MARKP.
4229 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
4230 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
4231 Use SCM_TYP16 instead of SCM_GCTYP16.
4232 Use SCM_CDR instead of SCM_GCCDR.
4233 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
4234 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
4235 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
4236 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
4237 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
4238 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
4239 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
4241 ** Removed function: scm_struct_init
4243 ** Removed variable: scm_symhash_dim
4245 ** Renamed function: scm_make_cont has been replaced by
4246 scm_make_continuation, which has a different interface.
4248 ** Deprecated function: scm_call_catching_errors
4250 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
4252 ** Deprecated function: scm_strhash
4254 Use scm_string_hash instead.
4256 ** Deprecated function: scm_vector_set_length_x
4258 Instead, create a fresh vector of the desired size and copy the contents.
4260 ** scm_gensym has changed prototype
4262 scm_gensym now only takes one argument.
4264 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
4267 There is now only a single symbol type scm_tc7_symbol.
4268 The tag scm_tc7_lvector was not used anyway.
4270 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
4272 Use scm_make_smob_type and scm_set_smob_XXX instead.
4274 ** New function scm_set_smob_apply.
4276 This can be used to set an apply function to a smob type.
4278 ** Deprecated function: scm_strprint_obj
4280 Use scm_object_to_string instead.
4282 ** Deprecated function: scm_wta
4284 Use scm_wrong_type_arg, or another appropriate error signalling function
4287 ** Explicit support for obarrays has been deprecated.
4289 Use `scm_str2symbol' and the generic hashtable functions instead.
4291 ** The concept of `vcells' has been deprecated.
4293 The data type `variable' is now used exclusively. `Vcells' have been
4294 a low-level concept so you are likely not affected by this change.
4296 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
4297 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
4299 Use scm_c_define or scm_c_lookup instead, as appropriate.
4301 *** New functions: scm_c_module_lookup, scm_c_lookup,
4302 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
4303 scm_module_define, scm_define.
4305 These functions work with variables instead of with vcells.
4307 ** New functions for creating and defining `subr's and `gsubr's.
4309 The new functions more clearly distinguish between creating a subr (or
4310 gsubr) object and adding it to the current module.
4312 These new functions are available: scm_c_make_subr, scm_c_define_subr,
4313 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
4314 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
4315 scm_c_define_gsubr_with_generic.
4317 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
4318 scm_make_subr_with_generic, scm_make_gsubr,
4319 scm_make_gsubr_with_generic.
4321 Use the new ones from above instead.
4323 ** C interface to the module system has changed.
4325 While we suggest that you avoid as many explicit module system
4326 operations from C as possible for the time being, the C interface has
4327 been made more similar to the high-level Scheme module system.
4329 *** New functions: scm_c_define_module, scm_c_use_module,
4330 scm_c_export, scm_c_resolve_module.
4332 They mostly work like their Scheme namesakes. scm_c_define_module
4333 takes a function that is called a context where the new module is
4336 *** Deprecated functions: scm_the_root_module, scm_make_module,
4337 scm_ensure_user_module, scm_load_scheme_module.
4339 Use the new functions instead.
4341 ** Renamed function: scm_internal_with_fluids becomes
4344 scm_internal_with_fluids is available as a deprecated function.
4346 ** New function: scm_c_with_fluid.
4348 Just like scm_c_with_fluids, but takes one fluid and one value instead
4351 ** Deprecated typedefs: long_long, ulong_long.
4353 They are of questionable utility and they pollute the global
4356 ** Deprecated typedef: scm_sizet
4358 It is of questionable utility now that Guile requires ANSI C, and is
4361 ** Deprecated typedefs: scm_port_rw_active, scm_port,
4362 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
4363 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
4365 Made more compliant with the naming policy by adding a _t at the end.
4367 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
4368 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
4370 With the exception of the mysterious scm_2ulong2big, they are still
4371 available under new names (scm_i_mkbig etc). These functions are not
4372 intended to be used in user code. You should avoid dealing with
4373 bignums directly, and should deal with numbers in general (which can
4376 ** Change in behavior: scm_num2long, scm_num2ulong
4378 The scm_num2[u]long functions don't any longer accept an inexact
4379 argument. This change in behavior is motivated by concordance with
4380 R5RS: It is more common that a primitive doesn't want to accept an
4381 inexact for an exact.
4383 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
4384 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
4385 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
4388 These are conversion functions between the various ANSI C integral
4389 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
4390 accept an inexact argument.
4392 ** New functions: scm_float2num, scm_double2num,
4393 scm_num2float, scm_num2double.
4395 These are conversion functions between the two ANSI C float types and
4398 ** New number validation macros:
4399 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
4403 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
4405 These are just nicer-named old scm_protect_object and
4406 scm_unprotect_object.
4408 ** Deprecated functions: scm_protect_object, scm_unprotect_object
4410 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
4412 These functions can be used to register pointers to locations that
4415 ** Deprecated function: scm_create_hook.
4417 Its sins are: misleading name, non-modularity and lack of general
4421 Changes since Guile 1.3.4:
4423 * Changes to the distribution
4425 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
4427 We've changed the way we handle generated files in the Guile source
4428 repository. As a result, the procedure for building trees obtained
4429 from the nightly FTP snapshots or via CVS has changed:
4430 - You must have appropriate versions of autoconf, automake, and
4431 libtool installed on your system. See README for info on how to
4432 obtain these programs.
4433 - Before configuring the tree, you must first run the script
4434 `autogen.sh' at the top of the source tree.
4436 The Guile repository used to contain not only source files, written by
4437 humans, but also some generated files, like configure scripts and
4438 Makefile.in files. Even though the contents of these files could be
4439 derived mechanically from other files present, we thought it would
4440 make the tree easier to build if we checked them into CVS.
4442 However, this approach means that minor differences between
4443 developer's installed tools and habits affected the whole team.
4444 So we have removed the generated files from the repository, and
4445 added the autogen.sh script, which will reconstruct them
4449 ** configure now has experimental options to remove support for certain
4452 --disable-arrays omit array and uniform array support
4453 --disable-posix omit posix interfaces
4454 --disable-networking omit networking interfaces
4455 --disable-regex omit regular expression interfaces
4457 These are likely to become separate modules some day.
4459 ** New configure option --enable-debug-freelist
4461 This enables a debugging version of SCM_NEWCELL(), and also registers
4462 an extra primitive, the setter `gc-set-debug-check-freelist!'.
4464 Configure with the --enable-debug-freelist option to enable
4465 the gc-set-debug-check-freelist! primitive, and then use:
4467 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
4468 (gc-set-debug-check-freelist! #f) # turn off checking
4470 Checking of the freelist forces a traversal of the freelist and
4471 a garbage collection before each allocation of a cell. This can
4472 slow down the interpreter dramatically, so the setter should be used to
4473 turn on this extra processing only when necessary.
4475 ** New configure option --enable-debug-malloc
4477 Include code for debugging of calls to scm_must_malloc/realloc/free.
4481 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
4482 2. objects reallocated by scm_must_realloc has been allocated by
4484 3. reallocated objects are reallocated with the same what string
4486 But, most importantly, it records the number of allocated objects of
4487 each kind. This is useful when searching for memory leaks.
4489 A Guile compiled with this option provides the primitive
4490 `malloc-stats' which returns an alist with pairs of kind and the
4491 number of objects of that kind.
4493 ** All includes are now referenced relative to the root directory
4495 Since some users have had problems with mixups between Guile and
4496 system headers, we have decided to always refer to Guile headers via
4497 their parent directories. This essentially creates a "private name
4498 space" for Guile headers. This means that the compiler only is given
4499 -I options for the root build and root source directory.
4501 ** Header files kw.h and genio.h have been removed.
4503 ** The module (ice-9 getopt-gnu-style) has been removed.
4505 ** New module (ice-9 documentation)
4507 Implements the interface to documentation strings associated with
4510 ** New module (ice-9 time)
4512 Provides a macro `time', which displays execution time of a given form.
4514 ** New module (ice-9 history)
4516 Loading this module enables value history in the repl.
4518 * Changes to the stand-alone interpreter
4520 ** New command line option --debug
4522 Start Guile with debugging evaluator and backtraces enabled.
4524 This is useful when debugging your .guile init file or scripts.
4526 ** New help facility
4528 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
4529 (help REGEXP) ditto for objects with names matching REGEXP (a string)
4530 (help 'NAME) gives documentation for NAME, even if it is not an object
4531 (help ,EXPR) gives documentation for object returned by EXPR
4532 (help (my module)) gives module commentary for `(my module)'
4533 (help) gives this text
4535 `help' searches among bindings exported from loaded modules, while
4536 `apropos' searches among bindings visible from the "current" module.
4538 Examples: (help help)
4540 (help "output-string")
4542 ** `help' and `apropos' now prints full module names
4544 ** Dynamic linking now uses libltdl from the libtool package.
4546 The old system dependent code for doing dynamic linking has been
4547 replaced with calls to the libltdl functions which do all the hairy
4550 The major improvement is that you can now directly pass libtool
4551 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
4552 will be able to do the best shared library job you can get, via
4555 The way dynamic libraries are found has changed and is not really
4556 portable across platforms, probably. It is therefore recommended to
4557 use absolute filenames when possible.
4559 If you pass a filename without an extension to `dynamic-link', it will
4560 try a few appropriate ones. Thus, the most platform ignorant way is
4561 to specify a name like "libfoo", without any directories and
4564 ** Guile COOP threads are now compatible with LinuxThreads
4566 Previously, COOP threading wasn't possible in applications linked with
4567 Linux POSIX threads due to their use of the stack pointer to find the
4568 thread context. This has now been fixed with a workaround which uses
4569 the pthreads to allocate the stack.
4571 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
4573 ** Positions of erring expression in scripts
4575 With version 1.3.4, the location of the erring expression in Guile
4576 scipts is no longer automatically reported. (This should have been
4577 documented before the 1.3.4 release.)
4579 You can get this information by enabling recording of positions of
4580 source expressions and running the debugging evaluator. Put this at
4581 the top of your script (or in your "site" file):
4583 (read-enable 'positions)
4584 (debug-enable 'debug)
4586 ** Backtraces in scripts
4588 It is now possible to get backtraces in scripts.
4592 (debug-enable 'debug 'backtrace)
4594 at the top of the script.
4596 (The first options enables the debugging evaluator.
4597 The second enables backtraces.)
4599 ** Part of module system symbol lookup now implemented in C
4601 The eval closure of most modules is now implemented in C. Since this
4602 was one of the bottlenecks for loading speed, Guile now loads code
4603 substantially faster than before.
4605 ** Attempting to get the value of an unbound variable now produces
4606 an exception with a key of 'unbound-variable instead of 'misc-error.
4608 ** The initial default output port is now unbuffered if it's using a
4609 tty device. Previously in this situation it was line-buffered.
4611 ** New hook: after-gc-hook
4613 after-gc-hook takes over the role of gc-thunk. This hook is run at
4614 the first SCM_TICK after a GC. (Thus, the code is run at the same
4615 point during evaluation as signal handlers.)
4617 Note that this hook should be used only for diagnostic and debugging
4618 purposes. It is not certain that it will continue to be well-defined
4619 when this hook is run in the future.
4621 C programmers: Note the new C level hooks scm_before_gc_c_hook,
4622 scm_before_sweep_c_hook, scm_after_gc_c_hook.
4624 ** Improvements to garbage collector
4626 Guile 1.4 has a new policy for triggering heap allocation and
4627 determining the sizes of heap segments. It fixes a number of problems
4630 1. The new policy can handle two separate pools of cells
4631 (2-word/4-word) better. (The old policy would run wild, allocating
4632 more and more memory for certain programs.)
4634 2. The old code would sometimes allocate far too much heap so that the
4635 Guile process became gigantic. The new code avoids this.
4637 3. The old code would sometimes allocate too little so that few cells
4638 were freed at GC so that, in turn, too much time was spent in GC.
4640 4. The old code would often trigger heap allocation several times in a
4641 row. (The new scheme predicts how large the segments needs to be
4642 in order not to need further allocation.)
4644 All in all, the new GC policy will make larger applications more
4647 The new GC scheme also is prepared for POSIX threading. Threads can
4648 allocate private pools of cells ("clusters") with just a single
4649 function call. Allocation of single cells from such a cluster can
4650 then proceed without any need of inter-thread synchronization.
4652 ** New environment variables controlling GC parameters
4654 GUILE_MAX_SEGMENT_SIZE Maximal segment size
4657 Allocation of 2-word cell heaps:
4659 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
4662 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
4663 GC in percent of total heap size
4666 Allocation of 4-word cell heaps
4667 (used for real numbers and misc other objects):
4669 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
4671 (See entry "Way for application to customize GC parameters" under
4672 section "Changes to the scm_ interface" below.)
4674 ** Guile now implements reals using 4-word cells
4676 This speeds up computation with reals. (They were earlier allocated
4677 with `malloc'.) There is still some room for optimizations, however.
4679 ** Some further steps toward POSIX thread support have been taken
4681 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
4682 don't have much effect any longer, and many of them will be removed in
4686 are only handled at the top of the evaluator loop, immediately after
4687 I/O, and in scm_equalp.
4689 *** The GC can allocate thread private pools of pairs.
4691 * Changes to Scheme functions and syntax
4693 ** close-input-port and close-output-port are now R5RS
4695 These procedures have been turned into primitives and have R5RS behaviour.
4697 ** New procedure: simple-format PORT MESSAGE ARG1 ...
4699 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
4700 extended by the more sophisticated version in (ice-9 format)
4702 (simple-format port message . args)
4703 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
4704 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
4705 the escapes are replaced with corresponding members of ARGS:
4706 ~A formats using `display' and ~S formats using `write'.
4707 If DESTINATION is #t, then use the `current-output-port',
4708 if DESTINATION is #f, then return a string containing the formatted text.
4709 Does not add a trailing newline."
4711 ** string-ref: the second argument is no longer optional.
4713 ** string, list->string: no longer accept strings in their arguments,
4714 only characters, for compatibility with R5RS.
4716 ** New procedure: port-closed? PORT
4717 Returns #t if PORT is closed or #f if it is open.
4719 ** Deprecated: list*
4721 The list* functionality is now provided by cons* (SRFI-1 compliant)
4723 ** New procedure: cons* ARG1 ARG2 ... ARGn
4725 Like `list', but the last arg provides the tail of the constructed list,
4726 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
4728 Requires at least one argument. If given one argument, that argument
4729 is returned as result.
4731 This function is called `list*' in some other Schemes and in Common LISP.
4733 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
4735 ** New procedure: object-documentation OBJECT
4737 Returns the documentation string associated with OBJECT. The
4738 procedure uses a caching mechanism so that subsequent lookups are
4741 Exported by (ice-9 documentation).
4743 ** module-name now returns full names of modules
4745 Previously, only the last part of the name was returned (`session' for
4746 `(ice-9 session)'). Ex: `(ice-9 session)'.
4748 * Changes to the gh_ interface
4750 ** Deprecated: gh_int2scmb
4752 Use gh_bool2scm instead.
4754 * Changes to the scm_ interface
4756 ** Guile primitives now carry docstrings!
4758 Thanks to Greg Badros!
4760 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4762 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4763 macros and must contain a docstring that is extracted into foo.doc using a new
4764 guile-doc-snarf script (that uses guile-doc-snarf.awk).
4766 However, a major overhaul of these macros is scheduled for the next release of
4769 ** Guile primitives use a new technique for validation of arguments
4771 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
4772 the readability of argument checking.
4774 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
4776 ** New macros: SCM_PACK, SCM_UNPACK
4778 Compose/decompose an SCM value.
4780 The SCM type is now treated as an abstract data type and may be defined as a
4781 long, a void* or as a struct, depending on the architecture and compile time
4782 options. This makes it easier to find several types of bugs, for example when
4783 SCM values are treated as integers without conversion. Values of the SCM type
4784 should be treated as "atomic" values. These macros are used when
4785 composing/decomposing an SCM value, either because you want to access
4786 individual bits, or because you want to treat it as an integer value.
4788 E.g., in order to set bit 7 in an SCM value x, use the expression
4790 SCM_PACK (SCM_UNPACK (x) | 0x80)
4792 ** The name property of hooks is deprecated.
4793 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
4795 You can emulate this feature by using object properties.
4797 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
4798 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
4801 These macros will be removed in a future release of Guile.
4803 ** The following types, functions and macros from numbers.h are deprecated:
4804 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
4805 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
4807 ** Port internals: the rw_random variable in the scm_port structure
4808 must be set to non-zero in any random access port. In recent Guile
4809 releases it was only set for bidirectional random-access ports.
4811 ** Port internals: the seek ptob procedure is now responsible for
4812 resetting the buffers if required. The change was made so that in the
4813 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
4814 the fport and strport ptobs can avoid resetting the buffers,
4815 in particular to avoid discarding unread chars. An existing port
4816 type can be fixed by adding something like the following to the
4817 beginning of the ptob seek procedure:
4819 if (pt->rw_active == SCM_PORT_READ)
4820 scm_end_input (object);
4821 else if (pt->rw_active == SCM_PORT_WRITE)
4822 ptob->flush (object);
4824 although to actually avoid resetting the buffers and discard unread
4825 chars requires further hacking that depends on the characteristics
4828 ** Deprecated functions: scm_fseek, scm_tag
4830 These functions are no longer used and will be removed in a future version.
4832 ** The scm_sysmissing procedure is no longer used in libguile.
4833 Unless it turns out to be unexpectedly useful to somebody, it will be
4834 removed in a future version.
4836 ** The format of error message strings has changed
4838 The two C procedures: scm_display_error and scm_error, as well as the
4839 primitive `scm-error', now use scm_simple_format to do their work.
4840 This means that the message strings of all code must be updated to use
4841 ~A where %s was used before, and ~S where %S was used before.
4843 During the period when there still are a lot of old Guiles out there,
4844 you might want to support both old and new versions of Guile.
4846 There are basically two methods to achieve this. Both methods use
4849 AC_CHECK_FUNCS(scm_simple_format)
4851 in your configure.in.
4853 Method 1: Use the string concatenation features of ANSI C's
4858 #ifdef HAVE_SCM_SIMPLE_FORMAT
4864 Then represent each of your error messages using a preprocessor macro:
4866 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
4870 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
4871 (define make-message string-append)
4873 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
4875 Method 2: Use the oldfmt function found in doc/oldfmt.c.
4879 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
4884 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
4888 ** Deprecated: coop_mutex_init, coop_condition_variable_init
4890 Don't use the functions coop_mutex_init and
4891 coop_condition_variable_init. They will change.
4893 Use scm_mutex_init and scm_cond_init instead.
4895 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
4896 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
4897 COND, as `scm_cond_wait' does, but it also bounds the duration
4898 of the wait. If COND has not been signaled before time ABSTIME,
4899 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
4900 returns the error code `ETIMEDOUT'.
4902 The ABSTIME parameter specifies an absolute time, with the same
4903 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
4904 to 00:00:00 GMT, January 1, 1970.
4906 ** New function: scm_cond_broadcast (scm_cond_t *COND)
4907 `scm_cond_broadcast' restarts all the threads that are waiting
4908 on the condition variable COND. Nothing happens if no threads are
4911 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
4912 `scm_key_create' allocates a new TSD key. The key is stored in
4913 the location pointed to by KEY. There is no limit on the number
4914 of keys allocated at a given time. The value initially associated
4915 with the returned key is `NULL' in all currently executing threads.
4917 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
4918 function associated with the key. When a thread terminates,
4919 DESTR_FUNCTION is called on the value associated with the key in
4920 that thread. The DESTR_FUNCTION is not called if a key is deleted
4921 with `scm_key_delete' or a value is changed with
4922 `scm_setspecific'. The order in which destructor functions are
4923 called at thread termination time is unspecified.
4925 Destructors are not yet implemented.
4927 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
4928 `scm_setspecific' changes the value associated with KEY in the
4929 calling thread, storing the given POINTER instead.
4931 ** New function: scm_getspecific (scm_key_t KEY)
4932 `scm_getspecific' returns the value currently associated with
4933 KEY in the calling thread.
4935 ** New function: scm_key_delete (scm_key_t KEY)
4936 `scm_key_delete' deallocates a TSD key. It does not check
4937 whether non-`NULL' values are associated with that key in the
4938 currently executing threads, nor call the destructor function
4939 associated with the key.
4941 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
4943 Initialize a C level hook HOOK with associated HOOK_DATA and type
4944 TYPE. (See scm_c_hook_run ().)
4946 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
4948 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
4949 is true, add it last, otherwise first. The same FUNC can be added
4950 multiple times if FUNC_DATA differ and vice versa.
4952 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
4954 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
4955 function is only removed if both FUNC and FUNC_DATA matches.
4957 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
4959 Run hook HOOK passing DATA to the hook functions.
4961 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
4962 returned is undefined.
4964 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
4965 returns a non-NULL value. This value is returned as the result of
4966 scm_c_hook_run. If all functions return NULL, NULL is returned.
4968 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
4969 returns a NULL value, and NULL is returned. If all functions returns
4970 a non-NULL value, the last value is returned.
4972 ** New C level GC hooks
4974 Five new C level hooks has been added to the garbage collector.
4976 scm_before_gc_c_hook
4979 are run before locking and after unlocking the heap. The system is
4980 thus in a mode where evaluation can take place. (Except that
4981 scm_before_gc_c_hook must not allocate new cells.)
4983 scm_before_mark_c_hook
4984 scm_before_sweep_c_hook
4985 scm_after_sweep_c_hook
4987 are run when the heap is locked. These are intended for extension of
4988 the GC in a modular fashion. Examples are the weaks and guardians
4991 ** Way for application to customize GC parameters
4993 The application can set up other default values for the GC heap
4994 allocation parameters
4996 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
4997 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
4998 GUILE_MAX_SEGMENT_SIZE,
5002 scm_default_init_heap_size_1, scm_default_min_yield_1,
5003 scm_default_init_heap_size_2, scm_default_min_yield_2,
5004 scm_default_max_segment_size
5006 respectively before callong scm_boot_guile.
5008 (See entry "New environment variables ..." in section
5009 "Changes to the stand-alone interpreter" above.)
5011 ** scm_protect_object/scm_unprotect_object now nest
5013 This means that you can call scm_protect_object multiple times on an
5014 object and count on the object being protected until
5015 scm_unprotect_object has been call the same number of times.
5017 The functions also have better time complexity.
5019 Still, it is usually possible to structure the application in a way
5020 that you don't need to use these functions. For example, if you use a
5021 protected standard Guile list to keep track of live objects rather
5022 than some custom data type, objects will die a natural death when they
5023 are no longer needed.
5025 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
5027 Guile does not provide the float representation for inexact real numbers any
5028 more. Now, only doubles are used to represent inexact real numbers. Further,
5029 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
5030 and scm_tc16_complex, respectively.
5032 ** Removed deprecated type scm_smobfuns
5034 ** Removed deprecated function scm_newsmob
5036 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
5038 There is an ongoing discussion among the developers whether to
5039 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
5040 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
5041 until this issue has been settled.
5043 ** Removed deprecated type tag scm_tc16_kw
5045 ** Added type tag scm_tc16_keyword
5047 (This was introduced already in release 1.3.4 but was not documented
5050 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
5052 * Changes to system call interfaces:
5054 ** The "select" procedure now tests port buffers for the ability to
5055 provide input or accept output. Previously only the underlying file
5056 descriptors were checked.
5058 ** New variable PIPE_BUF: the maximum number of bytes that can be
5059 atomically written to a pipe.
5061 ** If a facility is not available on the system when Guile is
5062 compiled, the corresponding primitive procedure will not be defined.
5063 Previously it would have been defined but would throw a system-error
5064 exception if called. Exception handlers which catch this case may
5065 need minor modification: an error will be thrown with key
5066 'unbound-variable instead of 'system-error. Alternatively it's
5067 now possible to use `defined?' to check whether the facility is
5070 ** Procedures which depend on the timezone should now give the correct
5071 result on systems which cache the TZ environment variable, even if TZ
5072 is changed without calling tzset.
5074 * Changes to the networking interfaces:
5076 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
5077 long integers between network and host format. For now, it's not
5078 particularly convenient to do this kind of thing, but consider:
5080 (define write-network-long
5081 (lambda (value port)
5082 (let ((v (make-uniform-vector 1 1 0)))
5083 (uniform-vector-set! v 0 (htonl value))
5084 (uniform-vector-write v port))))
5086 (define read-network-long
5088 (let ((v (make-uniform-vector 1 1 0)))
5089 (uniform-vector-read! v port)
5090 (ntohl (uniform-vector-ref v 0)))))
5092 ** If inet-aton fails, it now throws an error with key 'misc-error
5093 instead of 'system-error, since errno is not relevant.
5095 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
5096 specific keys instead of 'system-error. The latter is inappropriate
5097 since errno will not have been set. The keys are:
5098 'host-not-found, 'try-again, 'no-recovery and 'no-data.
5100 ** sethostent, setnetent, setprotoent, setservent: now take an
5101 optional argument STAYOPEN, which specifies whether the database
5102 remains open after a database entry is accessed randomly (e.g., using
5103 gethostbyname for the hosts database.) The default is #f. Previously
5107 Changes since Guile 1.3.2:
5109 * Changes to the stand-alone interpreter
5113 An initial version of the Guile debugger written by Chris Hanson has
5114 been added. The debugger is still under development but is included
5115 in the distribution anyway since it is already quite useful.
5121 after an error to enter the debugger. Type `help' inside the debugger
5122 for a description of available commands.
5124 If you prefer to have stack frames numbered and printed in
5125 anti-chronological order and prefer up in the stack to be down on the
5126 screen as is the case in gdb, you can put
5128 (debug-enable 'backwards)
5130 in your .guile startup file. (However, this means that Guile can't
5131 use indentation to indicate stack level.)
5133 The debugger is autoloaded into Guile at the first use.
5135 ** Further enhancements to backtraces
5137 There is a new debug option `width' which controls the maximum width
5138 on the screen of printed stack frames. Fancy printing parameters
5139 ("level" and "length" as in Common LISP) are adaptively adjusted for
5140 each stack frame to give maximum information while still fitting
5141 within the bounds. If the stack frame can't be made to fit by
5142 adjusting parameters, it is simply cut off at the end. This is marked
5145 ** Some modules are now only loaded when the repl is started
5147 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
5148 regex) are now loaded into (guile-user) only if the repl has been
5149 started. The effect is that the startup time for scripts has been
5150 reduced to 30% of what it was previously.
5152 Correctly written scripts load the modules they require at the top of
5153 the file and should not be affected by this change.
5155 ** Hooks are now represented as smobs
5157 * Changes to Scheme functions and syntax
5159 ** Readline support has changed again.
5161 The old (readline-activator) module is gone. Use (ice-9 readline)
5162 instead, which now contains all readline functionality. So the code
5163 to activate readline is now
5165 (use-modules (ice-9 readline))
5168 This should work at any time, including from the guile prompt.
5170 To avoid confusion about the terms of Guile's license, please only
5171 enable readline for your personal use; please don't make it the
5172 default for others. Here is why we make this rather odd-sounding
5175 Guile is normally licensed under a weakened form of the GNU General
5176 Public License, which allows you to link code with Guile without
5177 placing that code under the GPL. This exception is important to some
5180 However, since readline is distributed under the GNU General Public
5181 License, when you link Guile with readline, either statically or
5182 dynamically, you effectively change Guile's license to the strict GPL.
5183 Whenever you link any strictly GPL'd code into Guile, uses of Guile
5184 which are normally permitted become forbidden. This is a rather
5185 non-obvious consequence of the licensing terms.
5187 So, to make sure things remain clear, please let people choose for
5188 themselves whether to link GPL'd libraries like readline with Guile.
5190 ** regexp-substitute/global has changed slightly, but incompatibly.
5192 If you include a function in the item list, the string of the match
5193 object it receives is the same string passed to
5194 regexp-substitute/global, not some suffix of that string.
5195 Correspondingly, the match's positions are relative to the entire
5196 string, not the suffix.
5198 If the regexp can match the empty string, the way matches are chosen
5199 from the string has changed. regexp-substitute/global recognizes the
5200 same set of matches that list-matches does; see below.
5202 ** New function: list-matches REGEXP STRING [FLAGS]
5204 Return a list of match objects, one for every non-overlapping, maximal
5205 match of REGEXP in STRING. The matches appear in left-to-right order.
5206 list-matches only reports matches of the empty string if there are no
5207 other matches which begin on, end at, or include the empty match's
5210 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5212 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
5214 For each match of REGEXP in STRING, apply PROC to the match object,
5215 and the last value PROC returned, or INIT for the first call. Return
5216 the last value returned by PROC. We apply PROC to the matches as they
5217 appear from left to right.
5219 This function recognizes matches according to the same criteria as
5222 Thus, you could define list-matches like this:
5224 (define (list-matches regexp string . flags)
5225 (reverse! (apply fold-matches regexp string '() cons flags)))
5227 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5231 *** New function: hook? OBJ
5233 Return #t if OBJ is a hook, otherwise #f.
5235 *** New function: make-hook-with-name NAME [ARITY]
5237 Return a hook with name NAME and arity ARITY. The default value for
5238 ARITY is 0. The only effect of NAME is that it will appear when the
5239 hook object is printed to ease debugging.
5241 *** New function: hook-empty? HOOK
5243 Return #t if HOOK doesn't contain any procedures, otherwise #f.
5245 *** New function: hook->list HOOK
5247 Return a list of the procedures that are called when run-hook is
5250 ** `map' signals an error if its argument lists are not all the same length.
5252 This is the behavior required by R5RS, so this change is really a bug
5253 fix. But it seems to affect a lot of people's code, so we're
5254 mentioning it here anyway.
5256 ** Print-state handling has been made more transparent
5258 Under certain circumstances, ports are represented as a port with an
5259 associated print state. Earlier, this pair was represented as a pair
5260 (see "Some magic has been added to the printer" below). It is now
5261 indistinguishable (almost; see `get-print-state') from a port on the
5264 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
5266 Return a new port with the associated print state PRINT-STATE.
5268 *** New function: get-print-state OUTPUT-PORT
5270 Return the print state associated with this port if it exists,
5271 otherwise return #f.
5273 *** New function: directory-stream? OBJECT
5275 Returns true iff OBJECT is a directory stream --- the sort of object
5276 returned by `opendir'.
5278 ** New function: using-readline?
5280 Return #t if readline is in use in the current repl.
5282 ** structs will be removed in 1.4
5284 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
5285 and use GOOPS objects as the fundamental record type.
5287 * Changes to the scm_ interface
5289 ** structs will be removed in 1.4
5291 The entire current struct interface (struct.c, struct.h) will be
5292 replaced in Guile 1.4. We will merge GOOPS into libguile and use
5293 GOOPS objects as the fundamental record type.
5295 ** The internal representation of subr's has changed
5297 Instead of giving a hint to the subr name, the CAR field of the subr
5298 now contains an index to a subr entry in scm_subr_table.
5300 *** New variable: scm_subr_table
5302 An array of subr entries. A subr entry contains the name, properties
5303 and documentation associated with the subr. The properties and
5304 documentation slots are not yet used.
5306 ** A new scheme for "forwarding" calls to a builtin to a generic function
5308 It is now possible to extend the functionality of some Guile
5309 primitives by letting them defer a call to a GOOPS generic function on
5310 argument mismatch. This means that there is no loss of efficiency in
5315 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
5316 (define-method + ((x <string>) (y <string>))
5317 (string-append x y))
5319 + will still be as efficient as usual in numerical calculations, but
5320 can also be used for concatenating strings.
5322 Who will be the first one to extend Guile's numerical tower to
5323 rationals? :) [OK, there a few other things to fix before this can
5324 be made in a clean way.]
5326 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
5328 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
5330 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
5332 These do the same job as SCM_PROC and SCM_PROC1, but they also define
5333 a variable GENERIC which can be used by the dispatch macros below.
5335 [This is experimental code which may change soon.]
5337 *** New macros for forwarding control to a generic on arg type error
5339 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
5341 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
5343 These correspond to the scm_wta function call, and have the same
5344 behaviour until the user has called the GOOPS primitive
5345 `enable-primitive-generic!'. After that, these macros will apply the
5346 generic function GENERIC to the argument(s) instead of calling
5349 [This is experimental code which may change soon.]
5351 *** New macros for argument testing with generic dispatch
5353 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
5355 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
5357 These correspond to the SCM_ASSERT macro, but will defer control to
5358 GENERIC on error after `enable-primitive-generic!' has been called.
5360 [This is experimental code which may change soon.]
5362 ** New function: SCM scm_eval_body (SCM body, SCM env)
5364 Evaluates the body of a special form.
5366 ** The internal representation of struct's has changed
5368 Previously, four slots were allocated for the procedure(s) of entities
5369 and operators. The motivation for this representation had to do with
5370 the structure of the evaluator, the wish to support tail-recursive
5371 generic functions, and efficiency. Since the generic function
5372 dispatch mechanism has changed, there is no longer a need for such an
5373 expensive representation, and the representation has been simplified.
5375 This should not make any difference for most users.
5377 ** GOOPS support has been cleaned up.
5379 Some code has been moved from eval.c to objects.c and code in both of
5380 these compilation units has been cleaned up and better structured.
5382 *** New functions for applying generic functions
5384 New function: SCM scm_apply_generic (GENERIC, ARGS)
5385 New function: SCM scm_call_generic_0 (GENERIC)
5386 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
5387 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
5388 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
5390 ** Deprecated function: scm_make_named_hook
5392 It is now replaced by:
5394 ** New function: SCM scm_create_hook (const char *name, int arity)
5396 Creates a hook in the same way as make-hook above but also
5397 binds a variable named NAME to it.
5399 This is the typical way of creating a hook from C code.
5401 Currently, the variable is created in the "current" module.
5402 This might change when we get the new module system.
5404 [The behaviour is identical to scm_make_named_hook.]
5408 Changes since Guile 1.3:
5410 * Changes to mailing lists
5412 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
5414 See the README file to find current addresses for all the Guile
5417 * Changes to the distribution
5419 ** Readline support is no longer included with Guile by default.
5421 Based on the different license terms of Guile and Readline, we
5422 concluded that Guile should not *by default* cause the linking of
5423 Readline into an application program. Readline support is now offered
5424 as a separate module, which is linked into an application only when
5425 you explicitly specify it.
5427 Although Guile is GNU software, its distribution terms add a special
5428 exception to the usual GNU General Public License (GPL). Guile's
5429 license includes a clause that allows you to link Guile with non-free
5430 programs. We add this exception so as not to put Guile at a
5431 disadvantage vis-a-vis other extensibility packages that support other
5434 In contrast, the GNU Readline library is distributed under the GNU
5435 General Public License pure and simple. This means that you may not
5436 link Readline, even dynamically, into an application unless it is
5437 distributed under a free software license that is compatible the GPL.
5439 Because of this difference in distribution terms, an application that
5440 can use Guile may not be able to use Readline. Now users will be
5441 explicitly offered two independent decisions about the use of these
5444 You can activate the readline support by issuing
5446 (use-modules (readline-activator))
5449 from your ".guile" file, for example.
5451 * Changes to the stand-alone interpreter
5453 ** All builtins now print as primitives.
5454 Previously builtin procedures not belonging to the fundamental subr
5455 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
5456 Now, they print as #<primitive-procedure NAME>.
5458 ** Backtraces slightly more intelligible.
5459 gsubr-apply and macro transformer application frames no longer appear
5462 * Changes to Scheme functions and syntax
5464 ** Guile now correctly handles internal defines by rewriting them into
5465 their equivalent letrec. Previously, internal defines would
5466 incrementally add to the innermost environment, without checking
5467 whether the restrictions specified in RnRS were met. This lead to the
5468 correct behaviour when these restriction actually were met, but didn't
5469 catch all illegal uses. Such an illegal use could lead to crashes of
5470 the Guile interpreter or or other unwanted results. An example of
5471 incorrect internal defines that made Guile behave erratically:
5483 The problem with this example is that the definition of `c' uses the
5484 value of `b' directly. This confuses the meoization machine of Guile
5485 so that the second call of `b' (this time in a larger environment that
5486 also contains bindings for `c' and `d') refers to the binding of `c'
5487 instead of `a'. You could also make Guile crash with a variation on
5492 (define (b flag) (if flag a 1))
5493 (define c (1+ (b flag)))
5501 From now on, Guile will issue an `Unbound variable: b' error message
5506 A hook contains a list of functions which should be called on
5507 particular occasions in an existing program. Hooks are used for
5510 A window manager might have a hook before-window-map-hook. The window
5511 manager uses the function run-hooks to call all functions stored in
5512 before-window-map-hook each time a window is mapped. The user can
5513 store functions in the hook using add-hook!.
5515 In Guile, hooks are first class objects.
5517 *** New function: make-hook [N_ARGS]
5519 Return a hook for hook functions which can take N_ARGS arguments.
5520 The default value for N_ARGS is 0.
5522 (See also scm_make_named_hook below.)
5524 *** New function: add-hook! HOOK PROC [APPEND_P]
5526 Put PROC at the beginning of the list of functions stored in HOOK.
5527 If APPEND_P is supplied, and non-false, put PROC at the end instead.
5529 PROC must be able to take the number of arguments specified when the
5532 If PROC already exists in HOOK, then remove it first.
5534 *** New function: remove-hook! HOOK PROC
5536 Remove PROC from the list of functions in HOOK.
5538 *** New function: reset-hook! HOOK
5540 Clear the list of hook functions stored in HOOK.
5542 *** New function: run-hook HOOK ARG1 ...
5544 Run all hook functions stored in HOOK with arguments ARG1 ... .
5545 The number of arguments supplied must correspond to the number given
5546 when the hook was created.
5548 ** The function `dynamic-link' now takes optional keyword arguments.
5549 The only keyword argument that is currently defined is `:global
5550 BOOL'. With it, you can control whether the shared library will be
5551 linked in global mode or not. In global mode, the symbols from the
5552 linked library can be used to resolve references from other
5553 dynamically linked libraries. In non-global mode, the linked
5554 library is essentially invisible and can only be accessed via
5555 `dynamic-func', etc. The default is now to link in global mode.
5556 Previously, the default has been non-global mode.
5558 The `#:global' keyword is only effective on platforms that support
5559 the dlopen family of functions.
5561 ** New function `provided?'
5563 - Function: provided? FEATURE
5564 Return true iff FEATURE is supported by this installation of
5565 Guile. FEATURE must be a symbol naming a feature; the global
5566 variable `*features*' is a list of available features.
5568 ** Changes to the module (ice-9 expect):
5570 *** The expect-strings macro now matches `$' in a regular expression
5571 only at a line-break or end-of-file by default. Previously it would
5572 match the end of the string accumulated so far. The old behaviour
5573 can be obtained by setting the variable `expect-strings-exec-flags'
5576 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
5577 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
5578 in a regular expression will still match before a line-break or
5579 end-of-file. The default is `regexp/noteol'.
5581 *** The expect-strings macro now uses a variable
5582 `expect-strings-compile-flags' for the flags to be supplied to
5583 `make-regexp'. The default is `regexp/newline', which was previously
5586 *** The expect macro now supplies two arguments to a match procedure:
5587 the current accumulated string and a flag to indicate whether
5588 end-of-file has been reached. Previously only the string was supplied.
5589 If end-of-file is reached, the match procedure will be called an
5590 additional time with the same accumulated string as the previous call
5591 but with the flag set.
5593 ** New module (ice-9 format), implementing the Common Lisp `format' function.
5595 This code, and the documentation for it that appears here, was
5596 borrowed from SLIB, with minor adaptations for Guile.
5598 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
5599 An almost complete implementation of Common LISP format description
5600 according to the CL reference book `Common LISP' from Guy L.
5601 Steele, Digital Press. Backward compatible to most of the
5602 available Scheme format implementations.
5604 Returns `#t', `#f' or a string; has side effect of printing
5605 according to FORMAT-STRING. If DESTINATION is `#t', the output is
5606 to the current output port and `#t' is returned. If DESTINATION
5607 is `#f', a formatted string is returned as the result of the call.
5608 NEW: If DESTINATION is a string, DESTINATION is regarded as the
5609 format string; FORMAT-STRING is then the first argument and the
5610 output is returned as a string. If DESTINATION is a number, the
5611 output is to the current error port if available by the
5612 implementation. Otherwise DESTINATION must be an output port and
5615 FORMAT-STRING must be a string. In case of a formatting error
5616 format returns `#f' and prints a message on the current output or
5617 error port. Characters are output as if the string were output by
5618 the `display' function with the exception of those prefixed by a
5619 tilde (~). For a detailed description of the FORMAT-STRING syntax
5620 please consult a Common LISP format reference manual. For a test
5621 suite to verify this format implementation load `formatst.scm'.
5622 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
5624 Note: `format' is not reentrant, i.e. only one `format'-call may
5625 be executed at a time.
5628 *** Format Specification (Format version 3.0)
5630 Please consult a Common LISP format reference manual for a detailed
5631 description of the format string syntax. For a demonstration of the
5632 implemented directives see `formatst.scm'.
5634 This implementation supports directive parameters and modifiers (`:'
5635 and `@' characters). Multiple parameters must be separated by a comma
5636 (`,'). Parameters can be numerical parameters (positive or negative),
5637 character parameters (prefixed by a quote character (`''), variable
5638 parameters (`v'), number of rest arguments parameter (`#'), empty and
5639 default parameters. Directive characters are case independent. The
5640 general form of a directive is:
5642 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
5644 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
5646 *** Implemented CL Format Control Directives
5648 Documentation syntax: Uppercase characters represent the
5649 corresponding control directive characters. Lowercase characters
5650 represent control directive parameter descriptions.
5653 Any (print as `display' does).
5657 `~MINCOL,COLINC,MINPAD,PADCHARA'
5661 S-expression (print as `write' does).
5665 `~MINCOL,COLINC,MINPAD,PADCHARS'
5671 print number sign always.
5674 print comma separated.
5676 `~MINCOL,PADCHAR,COMMACHARD'
5682 print number sign always.
5685 print comma separated.
5687 `~MINCOL,PADCHAR,COMMACHARX'
5693 print number sign always.
5696 print comma separated.
5698 `~MINCOL,PADCHAR,COMMACHARO'
5704 print number sign always.
5707 print comma separated.
5709 `~MINCOL,PADCHAR,COMMACHARB'
5714 `~N,MINCOL,PADCHAR,COMMACHARR'
5718 print a number as a Roman numeral.
5721 print a number as an "old fashioned" Roman numeral.
5724 print a number as an ordinal English number.
5727 print a number as a cardinal English number.
5732 prints `y' and `ies'.
5735 as `~P but jumps 1 argument backward.'
5738 as `~@P but jumps 1 argument backward.'
5743 prints a character as the reader can understand it (i.e. `#\'
5747 prints a character as emacs does (eg. `^C' for ASCII 03).
5750 Fixed-format floating-point (prints a flonum like MMM.NNN).
5751 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
5753 If the number is positive a plus sign is printed.
5756 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
5757 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
5759 If the number is positive a plus sign is printed.
5762 General floating-point (prints a flonum either fixed or
5764 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
5766 If the number is positive a plus sign is printed.
5769 Dollars floating-point (prints a flonum in fixed with signs
5771 `~DIGITS,SCALE,WIDTH,PADCHAR$'
5773 If the number is positive a plus sign is printed.
5776 A sign is always printed and appears before the padding.
5779 The sign appears before the padding.
5787 print newline if not at the beginning of the output line.
5789 prints `~&' and then N-1 newlines.
5794 print N page separators.
5804 newline is ignored, white space left.
5807 newline is left, white space ignored.
5812 relative tabulation.
5818 Indirection (expects indirect arguments as a list).
5820 extracts indirect arguments from format arguments.
5823 Case conversion (converts by `string-downcase').
5825 converts by `string-capitalize'.
5828 converts by `string-capitalize-first'.
5831 converts by `string-upcase'.
5834 Argument Jumping (jumps 1 argument forward).
5836 jumps N arguments forward.
5839 jumps 1 argument backward.
5842 jumps N arguments backward.
5845 jumps to the 0th argument.
5848 jumps to the Nth argument (beginning from 0)
5850 `~[STR0~;STR1~;...~;STRN~]'
5851 Conditional Expression (numerical clause conditional).
5853 take argument from N.
5856 true test conditional.
5859 if-else-then conditional.
5865 default clause follows.
5868 Iteration (args come from the next argument (a list)).
5870 at most N iterations.
5873 args from next arg (a list of lists).
5876 args from the rest of arguments.
5879 args from the rest args (lists).
5890 aborts if N <= M <= K
5892 *** Not Implemented CL Format Control Directives
5895 print `#f' as an empty list (see below).
5898 print `#f' as an empty list (see below).
5904 (sorry I don't understand its semantics completely)
5906 *** Extended, Replaced and Additional Control Directives
5908 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
5909 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
5910 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
5911 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
5912 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
5913 COMMAWIDTH is the number of characters between two comma
5917 print a R4RS complex number as `~F~@Fi' with passed parameters for
5921 Pretty print formatting of an argument for scheme code lists.
5927 Flushes the output if format DESTINATION is a port.
5930 Print a `#\space' character
5932 print N `#\space' characters.
5935 Print a `#\tab' character
5937 print N `#\tab' characters.
5940 Takes N as an integer representation for a character. No arguments
5941 are consumed. N is converted to a character by `integer->char'. N
5942 must be a positive decimal number.
5945 Print out readproof. Prints out internal objects represented as
5946 `#<...>' as strings `"#<...>"' so that the format output can always
5947 be processed by `read'.
5950 Print out readproof. Prints out internal objects represented as
5951 `#<...>' as strings `"#<...>"' so that the format output can always
5952 be processed by `read'.
5955 Prints information and a copyright notice on the format
5958 prints format version.
5961 may also print number strings, i.e. passing a number as a string
5962 and format it accordingly.
5964 *** Configuration Variables
5966 The format module exports some configuration variables to suit the
5967 systems and users needs. There should be no modification necessary for
5968 the configuration that comes with Guile. Format detects automatically
5969 if the running scheme system implements floating point numbers and
5972 format:symbol-case-conv
5973 Symbols are converted by `symbol->string' so the case type of the
5974 printed symbols is implementation dependent.
5975 `format:symbol-case-conv' is a one arg closure which is either
5976 `#f' (no conversion), `string-upcase', `string-downcase' or
5977 `string-capitalize'. (default `#f')
5979 format:iobj-case-conv
5980 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
5981 implementation internal objects. (default `#f')
5984 The character prefixing the exponent value in `~E' printing.
5987 *** Compatibility With Other Format Implementations
5993 Downward compatible except for padding support and `~A', `~S',
5994 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
5995 `printf' padding support which is completely replaced by the CL
5996 `format' padding style.
5999 Downward compatible except for `~', which is not documented
6000 (ignores all characters inside the format string up to a newline
6001 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
6002 numerical and variable parameters and `:/@' modifiers in the CL
6006 Downward compatible except for `~A' and `~S' which print in
6007 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
6008 directive parameters or modifiers)).
6011 Downward compatible except for an optional destination parameter:
6012 S2C accepts a format call without a destination which returns a
6013 formatted string. This is equivalent to a #f destination in S2C.
6014 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
6015 parameters or modifiers)).
6018 ** Changes to string-handling functions.
6020 These functions were added to support the (ice-9 format) module, above.
6022 *** New function: string-upcase STRING
6023 *** New function: string-downcase STRING
6025 These are non-destructive versions of the existing string-upcase! and
6026 string-downcase! functions.
6028 *** New function: string-capitalize! STRING
6029 *** New function: string-capitalize STRING
6031 These functions convert the first letter of each word in the string to
6034 (string-capitalize "howdy there")
6037 As with the other functions, string-capitalize! modifies the string in
6038 place, while string-capitalize returns a modified copy of its argument.
6040 *** New function: string-ci->symbol STRING
6042 Return a symbol whose name is STRING, but having the same case as if
6043 the symbol had be read by `read'.
6045 Guile can be configured to be sensitive or insensitive to case
6046 differences in Scheme identifiers. If Guile is case-insensitive, all
6047 symbols are converted to lower case on input. The `string-ci->symbol'
6048 function returns a symbol whose name in STRING, transformed as Guile
6049 would if STRING were input.
6051 *** New function: substring-move! STRING1 START END STRING2 START
6053 Copy the substring of STRING1 from START (inclusive) to END
6054 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
6055 string, and the source and destination areas may overlap; in all
6056 cases, the function behaves as if all the characters were copied
6059 *** Extended functions: substring-move-left! substring-move-right!
6061 These functions now correctly copy arbitrarily overlapping substrings;
6062 they are both synonyms for substring-move!.
6065 ** New module (ice-9 getopt-long), with the function `getopt-long'.
6067 getopt-long is a function for parsing command-line arguments in a
6068 manner consistent with other GNU programs.
6070 (getopt-long ARGS GRAMMAR)
6071 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
6073 ARGS should be a list of strings. Its first element should be the
6074 name of the program; subsequent elements should be the arguments
6075 that were passed to the program on the command line. The
6076 `program-arguments' procedure returns a list of this form.
6078 GRAMMAR is a list of the form:
6079 ((OPTION (PROPERTY VALUE) ...) ...)
6081 Each OPTION should be a symbol. `getopt-long' will accept a
6082 command-line option named `--OPTION'.
6083 Each option can have the following (PROPERTY VALUE) pairs:
6085 (single-char CHAR) --- Accept `-CHAR' as a single-character
6086 equivalent to `--OPTION'. This is how to specify traditional
6088 (required? BOOL) --- If BOOL is true, the option is required.
6089 getopt-long will raise an error if it is not found in ARGS.
6090 (value BOOL) --- If BOOL is #t, the option accepts a value; if
6091 it is #f, it does not; and if it is the symbol
6092 `optional', the option may appear in ARGS with or
6094 (predicate FUNC) --- If the option accepts a value (i.e. you
6095 specified `(value #t)' for this option), then getopt
6096 will apply FUNC to the value, and throw an exception
6097 if it returns #f. FUNC should be a procedure which
6098 accepts a string and returns a boolean value; you may
6099 need to use quasiquotes to get it into GRAMMAR.
6101 The (PROPERTY VALUE) pairs may occur in any order, but each
6102 property may occur only once. By default, options do not have
6103 single-character equivalents, are not required, and do not take
6106 In ARGS, single-character options may be combined, in the usual
6107 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
6108 accepts values, then it must be the last option in the
6109 combination; the value is the next argument. So, for example, using
6110 the following grammar:
6111 ((apples (single-char #\a))
6112 (blimps (single-char #\b) (value #t))
6113 (catalexis (single-char #\c) (value #t)))
6114 the following argument lists would be acceptable:
6115 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
6116 for "blimps" and "catalexis")
6117 ("-ab" "bang" "-c" "couth") (same)
6118 ("-ac" "couth" "-b" "bang") (same)
6119 ("-abc" "couth" "bang") (an error, since `-b' is not the
6120 last option in its combination)
6122 If an option's value is optional, then `getopt-long' decides
6123 whether it has a value by looking at what follows it in ARGS. If
6124 the next element is a string, and it does not appear to be an
6125 option itself, then that string is the option's value.
6127 The value of a long option can appear as the next element in ARGS,
6128 or it can follow the option name, separated by an `=' character.
6129 Thus, using the same grammar as above, the following argument lists
6131 ("--apples" "Braeburn" "--blimps" "Goodyear")
6132 ("--apples=Braeburn" "--blimps" "Goodyear")
6133 ("--blimps" "Goodyear" "--apples=Braeburn")
6135 If the option "--" appears in ARGS, argument parsing stops there;
6136 subsequent arguments are returned as ordinary arguments, even if
6137 they resemble options. So, in the argument list:
6138 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
6139 `getopt-long' will recognize the `apples' option as having the
6140 value "Granny Smith", but it will not recognize the `blimp'
6141 option; it will return the strings "--blimp" and "Goodyear" as
6142 ordinary argument strings.
6144 The `getopt-long' function returns the parsed argument list as an
6145 assocation list, mapping option names --- the symbols from GRAMMAR
6146 --- onto their values, or #t if the option does not accept a value.
6147 Unused options do not appear in the alist.
6149 All arguments that are not the value of any option are returned
6150 as a list, associated with the empty list.
6152 `getopt-long' throws an exception if:
6153 - it finds an unrecognized option in ARGS
6154 - a required option is omitted
6155 - an option that requires an argument doesn't get one
6156 - an option that doesn't accept an argument does get one (this can
6157 only happen using the long option `--opt=value' syntax)
6158 - an option predicate fails
6163 `((lockfile-dir (required? #t)
6166 (predicate ,file-is-directory?))
6167 (verbose (required? #f)
6170 (x-includes (single-char #\x))
6171 (rnet-server (single-char #\y)
6172 (predicate ,string?))))
6174 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
6175 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
6177 => ((() "foo1" "-fred" "foo2" "foo3")
6178 (rnet-server . "lamprod")
6179 (x-includes . "/usr/include")
6180 (lockfile-dir . "/tmp")
6183 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
6185 It will be removed in a few releases.
6187 ** New syntax: lambda*
6188 ** New syntax: define*
6189 ** New syntax: define*-public
6190 ** New syntax: defmacro*
6191 ** New syntax: defmacro*-public
6192 Guile now supports optional arguments.
6194 `lambda*', `define*', `define*-public', `defmacro*' and
6195 `defmacro*-public' are identical to the non-* versions except that
6196 they use an extended type of parameter list that has the following BNF
6197 syntax (parentheses are literal, square brackets indicate grouping,
6198 and `*', `+' and `?' have the usual meaning):
6200 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
6201 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
6202 [[#&rest identifier]|[. identifier]]? ) | [identifier]
6204 ext-var-decl ::= identifier | ( identifier expression )
6206 The semantics are best illustrated with the following documentation
6207 and examples for `lambda*':
6210 lambda extended for optional and keyword arguments
6212 lambda* creates a procedure that takes optional arguments. These
6213 are specified by putting them inside brackets at the end of the
6214 paramater list, but before any dotted rest argument. For example,
6215 (lambda* (a b #&optional c d . e) '())
6216 creates a procedure with fixed arguments a and b, optional arguments c
6217 and d, and rest argument e. If the optional arguments are omitted
6218 in a call, the variables for them are unbound in the procedure. This
6219 can be checked with the bound? macro.
6221 lambda* can also take keyword arguments. For example, a procedure
6223 (lambda* (#&key xyzzy larch) '())
6224 can be called with any of the argument lists (#:xyzzy 11)
6225 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
6226 are given as keywords are bound to values.
6228 Optional and keyword arguments can also be given default values
6229 which they take on when they are not present in a call, by giving a
6230 two-item list in place of an optional argument, for example in:
6231 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
6232 foo is a fixed argument, bar is an optional argument with default
6233 value 42, and baz is a keyword argument with default value 73.
6234 Default value expressions are not evaluated unless they are needed
6235 and until the procedure is called.
6237 lambda* now supports two more special parameter list keywords.
6239 lambda*-defined procedures now throw an error by default if a
6240 keyword other than one of those specified is found in the actual
6241 passed arguments. However, specifying #&allow-other-keys
6242 immediately after the kyword argument declarations restores the
6243 previous behavior of ignoring unknown keywords. lambda* also now
6244 guarantees that if the same keyword is passed more than once, the
6245 last one passed is the one that takes effect. For example,
6246 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
6247 #:heads 37 #:tails 42 #:heads 99)
6248 would result in (99 47) being displayed.
6250 #&rest is also now provided as a synonym for the dotted syntax rest
6251 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
6252 all respects to lambda*. This is provided for more similarity to DSSSL,
6253 MIT-Scheme and Kawa among others, as well as for refugees from other
6256 Further documentation may be found in the optargs.scm file itself.
6258 The optional argument module also exports the macros `let-optional',
6259 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
6260 are not documented here because they may be removed in the future, but
6261 full documentation is still available in optargs.scm.
6263 ** New syntax: and-let*
6264 Guile now supports the `and-let*' form, described in the draft SRFI-2.
6266 Syntax: (land* (<clause> ...) <body> ...)
6267 Each <clause> should have one of the following forms:
6268 (<variable> <expression>)
6271 Each <variable> or <bound-variable> should be an identifier. Each
6272 <expression> should be a valid expression. The <body> should be a
6273 possibly empty sequence of expressions, like the <body> of a
6276 Semantics: A LAND* expression is evaluated by evaluating the
6277 <expression> or <bound-variable> of each of the <clause>s from
6278 left to right. The value of the first <expression> or
6279 <bound-variable> that evaluates to a false value is returned; the
6280 remaining <expression>s and <bound-variable>s are not evaluated.
6281 The <body> forms are evaluated iff all the <expression>s and
6282 <bound-variable>s evaluate to true values.
6284 The <expression>s and the <body> are evaluated in an environment
6285 binding each <variable> of the preceding (<variable> <expression>)
6286 clauses to the value of the <expression>. Later bindings
6287 shadow earlier bindings.
6289 Guile's and-let* macro was contributed by Michael Livshin.
6291 ** New sorting functions
6293 *** New function: sorted? SEQUENCE LESS?
6294 Returns `#t' when the sequence argument is in non-decreasing order
6295 according to LESS? (that is, there is no adjacent pair `... x y
6296 ...' for which `(less? y x)').
6298 Returns `#f' when the sequence contains at least one out-of-order
6299 pair. It is an error if the sequence is neither a list nor a
6302 *** New function: merge LIST1 LIST2 LESS?
6303 LIST1 and LIST2 are sorted lists.
6304 Returns the sorted list of all elements in LIST1 and LIST2.
6306 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
6307 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
6308 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
6309 (Here "<" should read "comes before".)
6311 *** New procedure: merge! LIST1 LIST2 LESS?
6312 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
6313 the result. If the code is compiled, and LESS? constructs no new
6314 pairs, no pairs at all will be allocated. The first pair of the
6315 result will be either the first pair of LIST1 or the first pair of
6318 *** New function: sort SEQUENCE LESS?
6319 Accepts either a list or a vector, and returns a new sequence
6320 which is sorted. The new sequence is the same type as the input.
6321 Always `(sorted? (sort sequence less?) less?)'. The original
6322 sequence is not altered in any way. The new sequence shares its
6323 elements with the old one; no elements are copied.
6325 *** New procedure: sort! SEQUENCE LESS
6326 Returns its sorted result in the original boxes. No new storage is
6327 allocated at all. Proper usage: (set! slist (sort! slist <))
6329 *** New function: stable-sort SEQUENCE LESS?
6330 Similar to `sort' but stable. That is, if "equal" elements are
6331 ordered a < b in the original sequence, they will have the same order
6334 *** New function: stable-sort! SEQUENCE LESS?
6335 Similar to `sort!' but stable.
6336 Uses temporary storage when sorting vectors.
6338 *** New functions: sort-list, sort-list!
6339 Added for compatibility with scsh.
6341 ** New built-in random number support
6343 *** New function: random N [STATE]
6344 Accepts a positive integer or real N and returns a number of the
6345 same type between zero (inclusive) and N (exclusive). The values
6346 returned have a uniform distribution.
6348 The optional argument STATE must be of the type produced by
6349 `copy-random-state' or `seed->random-state'. It defaults to the value
6350 of the variable `*random-state*'. This object is used to maintain the
6351 state of the pseudo-random-number generator and is altered as a side
6352 effect of the `random' operation.
6354 *** New variable: *random-state*
6355 Holds a data structure that encodes the internal state of the
6356 random-number generator that `random' uses by default. The nature
6357 of this data structure is implementation-dependent. It may be
6358 printed out and successfully read back in, but may or may not
6359 function correctly as a random-number state object in another
6362 *** New function: copy-random-state [STATE]
6363 Returns a new object of type suitable for use as the value of the
6364 variable `*random-state*' and as a second argument to `random'.
6365 If argument STATE is given, a copy of it is returned. Otherwise a
6366 copy of `*random-state*' is returned.
6368 *** New function: seed->random-state SEED
6369 Returns a new object of type suitable for use as the value of the
6370 variable `*random-state*' and as a second argument to `random'.
6371 SEED is a string or a number. A new state is generated and
6372 initialized using SEED.
6374 *** New function: random:uniform [STATE]
6375 Returns an uniformly distributed inexact real random number in the
6376 range between 0 and 1.
6378 *** New procedure: random:solid-sphere! VECT [STATE]
6379 Fills VECT with inexact real random numbers the sum of whose
6380 squares is less than 1.0. Thinking of VECT as coordinates in
6381 space of dimension N = `(vector-length VECT)', the coordinates are
6382 uniformly distributed within the unit N-shere. The sum of the
6383 squares of the numbers is returned. VECT can be either a vector
6384 or a uniform vector of doubles.
6386 *** New procedure: random:hollow-sphere! VECT [STATE]
6387 Fills VECT with inexact real random numbers the sum of whose squares
6388 is equal to 1.0. Thinking of VECT as coordinates in space of
6389 dimension n = `(vector-length VECT)', the coordinates are uniformly
6390 distributed over the surface of the unit n-shere. VECT can be either
6391 a vector or a uniform vector of doubles.
6393 *** New function: random:normal [STATE]
6394 Returns an inexact real in a normal distribution with mean 0 and
6395 standard deviation 1. For a normal distribution with mean M and
6396 standard deviation D use `(+ M (* D (random:normal)))'.
6398 *** New procedure: random:normal-vector! VECT [STATE]
6399 Fills VECT with inexact real random numbers which are independent and
6400 standard normally distributed (i.e., with mean 0 and variance 1).
6401 VECT can be either a vector or a uniform vector of doubles.
6403 *** New function: random:exp STATE
6404 Returns an inexact real in an exponential distribution with mean 1.
6405 For an exponential distribution with mean U use (* U (random:exp)).
6407 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
6409 These functions now operate on numbers in the range of a C unsigned
6412 These functions used to operate on numbers in the range of a C signed
6413 long; however, this seems inappropriate, because Guile integers don't
6416 ** New function: make-guardian
6417 This is an implementation of guardians as described in
6418 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
6419 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
6420 Programming Language Design and Implementation, June 1993
6421 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
6423 ** New functions: delq1!, delv1!, delete1!
6424 These procedures behave similar to delq! and friends but delete only
6425 one object if at all.
6427 ** New function: unread-string STRING PORT
6428 Unread STRING to PORT, that is, push it back onto the port so that
6429 next read operation will work on the pushed back characters.
6431 ** unread-char can now be called multiple times
6432 If unread-char is called multiple times, the unread characters will be
6433 read again in last-in first-out order.
6435 ** the procedures uniform-array-read! and uniform-array-write! now
6436 work on any kind of port, not just ports which are open on a file.
6438 ** Now 'l' in a port mode requests line buffering.
6440 ** The procedure truncate-file now works on string ports as well
6441 as file ports. If the size argument is omitted, the current
6442 file position is used.
6444 ** new procedure: seek PORT/FDES OFFSET WHENCE
6445 The arguments are the same as for the old fseek procedure, but it
6446 works on string ports as well as random-access file ports.
6448 ** the fseek procedure now works on string ports, since it has been
6449 redefined using seek.
6451 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
6452 size is not supplied.
6454 ** the newline procedure no longer flushes the port if it's not
6455 line-buffered: previously it did if it was the current output port.
6457 ** open-pipe and close-pipe are no longer primitive procedures, but
6458 an emulation can be obtained using `(use-modules (ice-9 popen))'.
6460 ** the freopen procedure has been removed.
6462 ** new procedure: drain-input PORT
6463 Drains PORT's read buffers (including any pushed-back characters)
6464 and returns the contents as a single string.
6466 ** New function: map-in-order PROC LIST1 LIST2 ...
6467 Version of `map' which guarantees that the procedure is applied to the
6468 lists in serial order.
6470 ** Renamed `serial-array-copy!' and `serial-array-map!' to
6471 `array-copy-in-order!' and `array-map-in-order!'. The old names are
6472 now obsolete and will go away in release 1.5.
6474 ** New syntax: collect BODY1 ...
6475 Version of `begin' which returns a list of the results of the body
6476 forms instead of the result of the last body form. In contrast to
6477 `begin', `collect' allows an empty body.
6479 ** New functions: read-history FILENAME, write-history FILENAME
6480 Read/write command line history from/to file. Returns #t on success
6481 and #f if an error occured.
6483 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
6485 These procedures return a list of definitions available in the specified
6486 argument, a relative module reference. In the case of no argument,
6487 `(current-module)' is now consulted for definitions to return, instead
6488 of simply returning #f, the former behavior.
6490 ** The #/ syntax for lists is no longer supported.
6492 Earlier versions of Scheme accepted this syntax, but printed a
6495 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
6497 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
6500 * Changes to the gh_ interface
6504 Now takes a second argument which is the result array. If this
6505 pointer is NULL, a new array is malloced (the old behaviour).
6507 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
6508 gh_scm2shorts, gh_scm2longs, gh_scm2floats
6512 * Changes to the scm_ interface
6514 ** Function: scm_make_named_hook (char* name, int n_args)
6516 Creates a hook in the same way as make-hook above but also
6517 binds a variable named NAME to it.
6519 This is the typical way of creating a hook from C code.
6521 Currently, the variable is created in the "current" module. This
6522 might change when we get the new module system.
6524 ** The smob interface
6526 The interface for creating smobs has changed. For documentation, see
6527 data-rep.info (made from guile-core/doc/data-rep.texi).
6529 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
6531 >>> This function will be removed in 1.3.4. <<<
6535 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
6536 This function adds a new smob type, named NAME, with instance size
6537 SIZE to the system. The return value is a tag that is used in
6538 creating instances of the type. If SIZE is 0, then no memory will
6539 be allocated when instances of the smob are created, and nothing
6540 will be freed by the default free function.
6542 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
6543 This function sets the smob marking procedure for the smob type
6544 specified by the tag TC. TC is the tag returned by
6545 `scm_make_smob_type'.
6547 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
6548 This function sets the smob freeing procedure for the smob type
6549 specified by the tag TC. TC is the tag returned by
6550 `scm_make_smob_type'.
6552 *** Function: void scm_set_smob_print (tc, print)
6554 - Function: void scm_set_smob_print (long tc,
6555 scm_sizet (*print) (SCM,
6559 This function sets the smob printing procedure for the smob type
6560 specified by the tag TC. TC is the tag returned by
6561 `scm_make_smob_type'.
6563 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
6564 This function sets the smob equality-testing predicate for the
6565 smob type specified by the tag TC. TC is the tag returned by
6566 `scm_make_smob_type'.
6568 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
6569 Make VALUE contain a smob instance of the type with type code TC and
6570 smob data DATA. VALUE must be previously declared as C type `SCM'.
6572 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
6573 This macro expands to a block of code that creates a smob instance
6574 of the type with type code TC and smob data DATA, and returns that
6575 `SCM' value. It should be the last piece of code in a block.
6577 ** The interfaces for using I/O ports and implementing port types
6578 (ptobs) have changed significantly. The new interface is based on
6579 shared access to buffers and a new set of ptob procedures.
6581 *** scm_newptob has been removed
6585 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
6587 - Function: SCM scm_make_port_type (char *type_name,
6588 int (*fill_buffer) (SCM port),
6589 void (*write_flush) (SCM port));
6591 Similarly to the new smob interface, there is a set of function
6592 setters by which the user can customize the behaviour of his port
6593 type. See ports.h (scm_set_port_XXX).
6595 ** scm_strport_to_string: New function: creates a new string from
6596 a string port's buffer.
6598 ** Plug in interface for random number generators
6599 The variable `scm_the_rng' in random.c contains a value and three
6600 function pointers which together define the current random number
6601 generator being used by the Scheme level interface and the random
6602 number library functions.
6604 The user is free to replace the default generator with the generator
6607 *** Variable: size_t scm_the_rng.rstate_size
6608 The size of the random state type used by the current RNG
6611 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
6612 Given the random STATE, return 32 random bits.
6614 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
6615 Seed random state STATE using string S of length N.
6617 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
6618 Given random state STATE, return a malloced copy.
6621 The default RNG is the MWC (Multiply With Carry) random number
6622 generator described by George Marsaglia at the Department of
6623 Statistics and Supercomputer Computations Research Institute, The
6624 Florida State University (http://stat.fsu.edu/~geo).
6626 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
6627 passes all tests in the DIEHARD test suite
6628 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
6629 costs one multiply and one add on platforms which either supports long
6630 longs (gcc does this on most systems) or have 64 bit longs. The cost
6631 is four multiply on other systems but this can be optimized by writing
6632 scm_i_uniform32 in assembler.
6634 These functions are provided through the scm_the_rng interface for use
6635 by libguile and the application.
6637 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
6638 Given the random STATE, return 32 random bits.
6639 Don't use this function directly. Instead go through the plugin
6640 interface (see "Plug in interface" above).
6642 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
6643 Initialize STATE using SEED of length N.
6645 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
6646 Return a malloc:ed copy of STATE. This function can easily be re-used
6647 in the interfaces to other RNGs.
6649 ** Random number library functions
6650 These functions use the current RNG through the scm_the_rng interface.
6651 It might be a good idea to use these functions from your C code so
6652 that only one random generator is used by all code in your program.
6654 The default random state is stored in:
6656 *** Variable: SCM scm_var_random_state
6657 Contains the vcell of the Scheme variable "*random-state*" which is
6658 used as default state by all random number functions in the Scheme
6663 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
6665 *** Function: scm_rstate *scm_c_default_rstate (void)
6666 This is a convenience function which returns the value of
6667 scm_var_random_state. An error message is generated if this value
6668 isn't a random state.
6670 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
6671 Make a new random state from the string SEED of length LENGTH.
6673 It is generally not a good idea to use multiple random states in a
6674 program. While subsequent random numbers generated from one random
6675 state are guaranteed to be reasonably independent, there is no such
6676 guarantee for numbers generated from different random states.
6678 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
6679 Return 32 random bits.
6681 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
6682 Return a sample from the uniform(0,1) distribution.
6684 *** Function: double scm_c_normal01 (scm_rstate *STATE)
6685 Return a sample from the normal(0,1) distribution.
6687 *** Function: double scm_c_exp1 (scm_rstate *STATE)
6688 Return a sample from the exp(1) distribution.
6690 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
6691 Return a sample from the discrete uniform(0,M) distribution.
6693 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
6694 Return a sample from the discrete uniform(0,M) distribution.
6695 M must be a bignum object. The returned value may be an INUM.
6699 Changes in Guile 1.3 (released Monday, October 19, 1998):
6701 * Changes to the distribution
6703 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
6704 To avoid conflicts, programs should name environment variables after
6705 themselves, except when there's a common practice establishing some
6708 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
6709 giving the former precedence, and printing a warning message if the
6710 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
6712 ** The header files related to multi-byte characters have been removed.
6713 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
6714 which referred to these explicitly will probably need to be rewritten,
6715 since the support for the variant string types has been removed; see
6718 ** The header files append.h and sequences.h have been removed. These
6719 files implemented non-R4RS operations which would encourage
6720 non-portable programming style and less easy-to-read code.
6722 * Changes to the stand-alone interpreter
6724 ** New procedures have been added to implement a "batch mode":
6726 *** Function: batch-mode?
6728 Returns a boolean indicating whether the interpreter is in batch
6731 *** Function: set-batch-mode?! ARG
6733 If ARG is true, switches the interpreter to batch mode. The `#f'
6734 case has not been implemented.
6736 ** Guile now provides full command-line editing, when run interactively.
6737 To use this feature, you must have the readline library installed.
6738 The Guile build process will notice it, and automatically include
6741 The readline library is available via anonymous FTP from any GNU
6742 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
6744 ** the-last-stack is now a fluid.
6746 * Changes to the procedure for linking libguile with your programs
6748 ** You can now use the `guile-config' utility to build programs that use Guile.
6750 Guile now includes a command-line utility called `guile-config', which
6751 can provide information about how to compile and link programs that
6754 *** `guile-config compile' prints any C compiler flags needed to use Guile.
6755 You should include this command's output on the command line you use
6756 to compile C or C++ code that #includes the Guile header files. It's
6757 usually just a `-I' flag to help the compiler find the Guile headers.
6760 *** `guile-config link' prints any linker flags necessary to link with Guile.
6762 This command writes to its standard output a list of flags which you
6763 must pass to the linker to link your code against the Guile library.
6764 The flags include '-lguile' itself, any other libraries the Guile
6765 library depends upon, and any `-L' flags needed to help the linker
6766 find those libraries.
6768 For example, here is a Makefile rule that builds a program named 'foo'
6769 from the object files ${FOO_OBJECTS}, and links them against Guile:
6772 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
6774 Previous Guile releases recommended that you use autoconf to detect
6775 which of a predefined set of libraries were present on your system.
6776 It is more robust to use `guile-config', since it records exactly which
6777 libraries the installed Guile library requires.
6779 This was originally called `build-guile', but was renamed to
6780 `guile-config' before Guile 1.3 was released, to be consistent with
6781 the analogous script for the GTK+ GUI toolkit, which is called
6785 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
6787 If you are using the GNU autoconf package to configure your program,
6788 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
6789 (described above) and gather the necessary values for use in your
6792 The GUILE_FLAGS macro expands to configure script code which runs the
6793 `guile-config' script, to find out where Guile's header files and
6794 libraries are installed. It sets two variables, marked for
6795 substitution, as by AC_SUBST.
6797 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
6798 code that uses Guile header files. This is almost always just a
6801 GUILE_LDFLAGS --- flags to pass to the linker to link a
6802 program against Guile. This includes `-lguile' for the Guile
6803 library itself, any libraries that Guile itself requires (like
6804 -lqthreads), and so on. It may also include a -L flag to tell the
6805 compiler where to find the libraries.
6807 GUILE_FLAGS is defined in the file guile.m4, in the top-level
6808 directory of the Guile distribution. You can copy it into your
6809 package's aclocal.m4 file, and then use it in your configure.in file.
6811 If you are using the `aclocal' program, distributed with GNU automake,
6812 to maintain your aclocal.m4 file, the Guile installation process
6813 installs guile.m4 where aclocal will find it. All you need to do is
6814 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
6815 this will copy the definition of GUILE_FLAGS into your aclocal.m4
6819 * Changes to Scheme functions and syntax
6821 ** Multi-byte strings have been removed, as have multi-byte and wide
6822 ports. We felt that these were the wrong approach to
6823 internationalization support.
6825 ** New function: readline [PROMPT]
6826 Read a line from the terminal, and allow the user to edit it,
6827 prompting with PROMPT. READLINE provides a large set of Emacs-like
6828 editing commands, lets the user recall previously typed lines, and
6829 works on almost every kind of terminal, including dumb terminals.
6831 READLINE assumes that the cursor is at the beginning of the line when
6832 it is invoked. Thus, you can't print a prompt yourself, and then call
6833 READLINE; you need to package up your prompt as a string, pass it to
6834 the function, and let READLINE print the prompt itself. This is
6835 because READLINE needs to know the prompt's screen width.
6837 For Guile to provide this function, you must have the readline
6838 library, version 2.1 or later, installed on your system. Readline is
6839 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
6840 any GNU mirror site.
6842 See also ADD-HISTORY function.
6844 ** New function: add-history STRING
6845 Add STRING as the most recent line in the history used by the READLINE
6846 command. READLINE does not add lines to the history itself; you must
6847 call ADD-HISTORY to make previous input available to the user.
6849 ** The behavior of the read-line function has changed.
6851 This function now uses standard C library functions to read the line,
6852 for speed. This means that it doesn not respect the value of
6853 scm-line-incrementors; it assumes that lines are delimited with
6856 (Note that this is read-line, the function that reads a line of text
6857 from a port, not readline, the function that reads a line from a
6858 terminal, providing full editing capabilities.)
6860 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
6862 This module provides some simple argument parsing. It exports one
6865 Function: getopt-gnu-style ARG-LS
6866 Parse a list of program arguments into an alist of option
6869 Each item in the list of program arguments is examined to see if
6870 it meets the syntax of a GNU long-named option. An argument like
6871 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
6872 returned alist, where MUMBLE is a keyword object with the same
6873 name as the argument. An argument like `--MUMBLE=FROB' produces
6874 an element of the form (MUMBLE . FROB), where FROB is a string.
6876 As a special case, the returned alist also contains a pair whose
6877 car is the symbol `rest'. The cdr of this pair is a list
6878 containing all the items in the argument list that are not options
6879 of the form mentioned above.
6881 The argument `--' is treated specially: all items in the argument
6882 list appearing after such an argument are not examined, and are
6883 returned in the special `rest' list.
6885 This function does not parse normal single-character switches.
6886 You will need to parse them out of the `rest' list yourself.
6888 ** The read syntax for byte vectors and short vectors has changed.
6890 Instead of #bytes(...), write #y(...).
6892 Instead of #short(...), write #h(...).
6894 This may seem nutty, but, like the other uniform vectors, byte vectors
6895 and short vectors want to have the same print and read syntax (and,
6896 more basic, want to have read syntax!). Changing the read syntax to
6897 use multiple characters after the hash sign breaks with the
6898 conventions used in R5RS and the conventions used for the other
6899 uniform vectors. It also introduces complexity in the current reader,
6900 both on the C and Scheme levels. (The Right solution is probably to
6901 change the syntax and prototypes for uniform vectors entirely.)
6904 ** The new module (ice-9 session) provides useful interactive functions.
6906 *** New procedure: (apropos REGEXP OPTION ...)
6908 Display a list of top-level variables whose names match REGEXP, and
6909 the modules they are imported from. Each OPTION should be one of the
6912 value --- Show the value of each matching variable.
6913 shadow --- Show bindings shadowed by subsequently imported modules.
6914 full --- Same as both `shadow' and `value'.
6918 guile> (apropos "trace" 'full)
6919 debug: trace #<procedure trace args>
6920 debug: untrace #<procedure untrace args>
6921 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
6922 the-scm-module: before-backtrace-hook ()
6923 the-scm-module: backtrace #<primitive-procedure backtrace>
6924 the-scm-module: after-backtrace-hook ()
6925 the-scm-module: has-shown-backtrace-hint? #f
6928 ** There are new functions and syntax for working with macros.
6930 Guile implements macros as a special object type. Any variable whose
6931 top-level binding is a macro object acts as a macro. The macro object
6932 specifies how the expression should be transformed before evaluation.
6934 *** Macro objects now print in a reasonable way, resembling procedures.
6936 *** New function: (macro? OBJ)
6937 True iff OBJ is a macro object.
6939 *** New function: (primitive-macro? OBJ)
6940 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
6941 macro transformers, implemented in eval.c rather than Scheme code.
6943 Why do we have this function?
6944 - For symmetry with procedure? and primitive-procedure?,
6945 - to allow custom print procedures to tell whether a macro is
6946 primitive, and display it differently, and
6947 - to allow compilers and user-written evaluators to distinguish
6948 builtin special forms from user-defined ones, which could be
6951 *** New function: (macro-type OBJ)
6952 Return a value indicating what kind of macro OBJ is. Possible return
6955 The symbol `syntax' --- a macro created by procedure->syntax.
6956 The symbol `macro' --- a macro created by procedure->macro.
6957 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
6958 The boolean #f --- if OBJ is not a macro object.
6960 *** New function: (macro-name MACRO)
6961 Return the name of the macro object MACRO's procedure, as returned by
6964 *** New function: (macro-transformer MACRO)
6965 Return the transformer procedure for MACRO.
6967 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
6969 Specify a new macro expander to use in the current module. Each
6970 MODULE is a module name, with the same meaning as in the `use-modules'
6971 form; each named module's exported bindings are added to the current
6972 top-level environment. TRANSFORMER is an expression evaluated in the
6973 resulting environment which must yield a procedure to use as the
6974 module's eval transformer: every expression evaluated in this module
6975 is passed to this function, and the result passed to the Guile
6978 *** macro-eval! is removed. Use local-eval instead.
6980 ** Some magic has been added to the printer to better handle user
6981 written printing routines (like record printers, closure printers).
6983 The problem is that these user written routines must have access to
6984 the current `print-state' to be able to handle fancy things like
6985 detection of circular references. These print-states have to be
6986 passed to the builtin printing routines (display, write, etc) to
6987 properly continue the print chain.
6989 We didn't want to change all existing print code so that it
6990 explicitly passes thru a print state in addition to a port. Instead,
6991 we extented the possible values that the builtin printing routines
6992 accept as a `port'. In addition to a normal port, they now also take
6993 a pair of a normal port and a print-state. Printing will go to the
6994 port and the print-state will be used to control the detection of
6995 circular references, etc. If the builtin function does not care for a
6996 print-state, it is simply ignored.
6998 User written callbacks are now called with such a pair as their
6999 `port', but because every function now accepts this pair as a PORT
7000 argument, you don't have to worry about that. In fact, it is probably
7001 safest to not check for these pairs.
7003 However, it is sometimes necessary to continue a print chain on a
7004 different port, for example to get a intermediate string
7005 representation of the printed value, mangle that string somehow, and
7006 then to finally print the mangled string. Use the new function
7008 inherit-print-state OLD-PORT NEW-PORT
7010 for this. It constructs a new `port' that prints to NEW-PORT but
7011 inherits the print-state of OLD-PORT.
7013 ** struct-vtable-offset renamed to vtable-offset-user
7015 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
7017 ** There is now a third optional argument to make-vtable-vtable
7018 (and fourth to make-struct) when constructing new types (vtables).
7019 This argument initializes field vtable-index-printer of the vtable.
7021 ** The detection of circular references has been extended to structs.
7022 That is, a structure that -- in the process of being printed -- prints
7023 itself does not lead to infinite recursion.
7025 ** There is now some basic support for fluids. Please read
7026 "libguile/fluid.h" to find out more. It is accessible from Scheme with
7027 the following functions and macros:
7029 Function: make-fluid
7031 Create a new fluid object. Fluids are not special variables or
7032 some other extension to the semantics of Scheme, but rather
7033 ordinary Scheme objects. You can store them into variables (that
7034 are still lexically scoped, of course) or into any other place you
7035 like. Every fluid has a initial value of `#f'.
7037 Function: fluid? OBJ
7039 Test whether OBJ is a fluid.
7041 Function: fluid-ref FLUID
7042 Function: fluid-set! FLUID VAL
7044 Access/modify the fluid FLUID. Modifications are only visible
7045 within the current dynamic root (that includes threads).
7047 Function: with-fluids* FLUIDS VALUES THUNK
7049 FLUIDS is a list of fluids and VALUES a corresponding list of
7050 values for these fluids. Before THUNK gets called the values are
7051 installed in the fluids and the old values of the fluids are
7052 saved in the VALUES list. When the flow of control leaves THUNK
7053 or reenters it, the values get swapped again. You might think of
7054 this as a `safe-fluid-excursion'. Note that the VALUES list is
7055 modified by `with-fluids*'.
7057 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
7059 The same as `with-fluids*' but with a different syntax. It looks
7060 just like `let', but both FLUID and VALUE are evaluated. Remember,
7061 fluids are not special variables but ordinary objects. FLUID
7062 should evaluate to a fluid.
7064 ** Changes to system call interfaces:
7066 *** close-port, close-input-port and close-output-port now return a
7067 boolean instead of an `unspecified' object. #t means that the port
7068 was successfully closed, while #f means it was already closed. It is
7069 also now possible for these procedures to raise an exception if an
7070 error occurs (some errors from write can be delayed until close.)
7072 *** the first argument to chmod, fcntl, ftell and fseek can now be a
7075 *** the third argument to fcntl is now optional.
7077 *** the first argument to chown can now be a file descriptor or a port.
7079 *** the argument to stat can now be a port.
7081 *** The following new procedures have been added (most use scsh
7084 *** procedure: close PORT/FD
7085 Similar to close-port (*note close-port: Closing Ports.), but also
7086 works on file descriptors. A side effect of closing a file
7087 descriptor is that any ports using that file descriptor are moved
7088 to a different file descriptor and have their revealed counts set
7091 *** procedure: port->fdes PORT
7092 Returns the integer file descriptor underlying PORT. As a side
7093 effect the revealed count of PORT is incremented.
7095 *** procedure: fdes->ports FDES
7096 Returns a list of existing ports which have FDES as an underlying
7097 file descriptor, without changing their revealed counts.
7099 *** procedure: fdes->inport FDES
7100 Returns an existing input port which has FDES as its underlying
7101 file descriptor, if one exists, and increments its revealed count.
7102 Otherwise, returns a new input port with a revealed count of 1.
7104 *** procedure: fdes->outport FDES
7105 Returns an existing output port which has FDES as its underlying
7106 file descriptor, if one exists, and increments its revealed count.
7107 Otherwise, returns a new output port with a revealed count of 1.
7109 The next group of procedures perform a `dup2' system call, if NEWFD
7110 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
7111 duplicated can be supplied as an integer or contained in a port. The
7112 type of value returned varies depending on which procedure is used.
7114 All procedures also have the side effect when performing `dup2' that
7115 any ports using NEWFD are moved to a different file descriptor and have
7116 their revealed counts set to zero.
7118 *** procedure: dup->fdes PORT/FD [NEWFD]
7119 Returns an integer file descriptor.
7121 *** procedure: dup->inport PORT/FD [NEWFD]
7122 Returns a new input port using the new file descriptor.
7124 *** procedure: dup->outport PORT/FD [NEWFD]
7125 Returns a new output port using the new file descriptor.
7127 *** procedure: dup PORT/FD [NEWFD]
7128 Returns a new port if PORT/FD is a port, with the same mode as the
7129 supplied port, otherwise returns an integer file descriptor.
7131 *** procedure: dup->port PORT/FD MODE [NEWFD]
7132 Returns a new port using the new file descriptor. MODE supplies a
7133 mode string for the port (*note open-file: File Ports.).
7135 *** procedure: setenv NAME VALUE
7136 Modifies the environment of the current process, which is also the
7137 default environment inherited by child processes.
7139 If VALUE is `#f', then NAME is removed from the environment.
7140 Otherwise, the string NAME=VALUE is added to the environment,
7141 replacing any existing string with name matching NAME.
7143 The return value is unspecified.
7145 *** procedure: truncate-file OBJ SIZE
7146 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
7147 can be a string containing a file name or an integer file
7148 descriptor or port open for output on the file. The underlying
7149 system calls are `truncate' and `ftruncate'.
7151 The return value is unspecified.
7153 *** procedure: setvbuf PORT MODE [SIZE]
7154 Set the buffering mode for PORT. MODE can be:
7162 block buffered, using a newly allocated buffer of SIZE bytes.
7163 However if SIZE is zero or unspecified, the port will be made
7166 This procedure should not be used after I/O has been performed with
7169 Ports are usually block buffered by default, with a default buffer
7170 size. Procedures e.g., *Note open-file: File Ports, which accept a
7171 mode string allow `0' to be added to request an unbuffered port.
7173 *** procedure: fsync PORT/FD
7174 Copies any unwritten data for the specified output file descriptor
7175 to disk. If PORT/FD is a port, its buffer is flushed before the
7176 underlying file descriptor is fsync'd. The return value is
7179 *** procedure: open-fdes PATH FLAGS [MODES]
7180 Similar to `open' but returns a file descriptor instead of a port.
7182 *** procedure: execle PATH ENV [ARG] ...
7183 Similar to `execl', but the environment of the new process is
7184 specified by ENV, which must be a list of strings as returned by
7185 the `environ' procedure.
7187 This procedure is currently implemented using the `execve' system
7188 call, but we call it `execle' because of its Scheme calling
7191 *** procedure: strerror ERRNO
7192 Returns the Unix error message corresponding to ERRNO, an integer.
7194 *** procedure: primitive-exit [STATUS]
7195 Terminate the current process without unwinding the Scheme stack.
7196 This is would typically be useful after a fork. The exit status
7197 is STATUS if supplied, otherwise zero.
7199 *** procedure: times
7200 Returns an object with information about real and processor time.
7201 The following procedures accept such an object as an argument and
7202 return a selected component:
7205 The current real time, expressed as time units relative to an
7209 The CPU time units used by the calling process.
7212 The CPU time units used by the system on behalf of the
7216 The CPU time units used by terminated child processes of the
7217 calling process, whose status has been collected (e.g., using
7221 Similarly, the CPU times units used by the system on behalf of
7222 terminated child processes.
7224 ** Removed: list-length
7225 ** Removed: list-append, list-append!
7226 ** Removed: list-reverse, list-reverse!
7228 ** array-map renamed to array-map!
7230 ** serial-array-map renamed to serial-array-map!
7232 ** catch doesn't take #f as first argument any longer
7234 Previously, it was possible to pass #f instead of a key to `catch'.
7235 That would cause `catch' to pass a jump buffer object to the procedure
7236 passed as second argument. The procedure could then use this jump
7237 buffer objekt as an argument to throw.
7239 This mechanism has been removed since its utility doesn't motivate the
7240 extra complexity it introduces.
7242 ** The `#/' notation for lists now provokes a warning message from Guile.
7243 This syntax will be removed from Guile in the near future.
7245 To disable the warning message, set the GUILE_HUSH environment
7246 variable to any non-empty value.
7248 ** The newline character now prints as `#\newline', following the
7249 normal Scheme notation, not `#\nl'.
7251 * Changes to the gh_ interface
7253 ** The gh_enter function now takes care of loading the Guile startup files.
7254 gh_enter works by calling scm_boot_guile; see the remarks below.
7256 ** Function: void gh_write (SCM x)
7258 Write the printed representation of the scheme object x to the current
7259 output port. Corresponds to the scheme level `write'.
7261 ** gh_list_length renamed to gh_length.
7263 ** vector handling routines
7265 Several major changes. In particular, gh_vector() now resembles
7266 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
7267 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
7268 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
7269 vector-related gh_ functions have been implemented.
7271 ** pair and list routines
7273 Implemented several of the R4RS pair and list functions that were
7276 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
7278 New function. Converts double arrays back and forth between Scheme
7281 * Changes to the scm_ interface
7283 ** The function scm_boot_guile now takes care of loading the startup files.
7285 Guile's primary initialization function, scm_boot_guile, now takes
7286 care of loading `boot-9.scm', in the `ice-9' module, to initialize
7287 Guile, define the module system, and put together some standard
7288 bindings. It also loads `init.scm', which is intended to hold
7289 site-specific initialization code.
7291 Since Guile cannot operate properly until boot-9.scm is loaded, there
7292 is no reason to separate loading boot-9.scm from Guile's other
7293 initialization processes.
7295 This job used to be done by scm_compile_shell_switches, which didn't
7296 make much sense; in particular, it meant that people using Guile for
7297 non-shell-like applications had to jump through hoops to get Guile
7298 initialized properly.
7300 ** The function scm_compile_shell_switches no longer loads the startup files.
7301 Now, Guile always loads the startup files, whenever it is initialized;
7302 see the notes above for scm_boot_guile and scm_load_startup_files.
7304 ** Function: scm_load_startup_files
7305 This new function takes care of loading Guile's initialization file
7306 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
7307 this is always called by the Guile initialization process, it's
7308 probably not too useful to call this yourself, but it's there anyway.
7310 ** The semantics of smob marking have changed slightly.
7312 The smob marking function (the `mark' member of the scm_smobfuns
7313 structure) is no longer responsible for setting the mark bit on the
7314 smob. The generic smob handling code in the garbage collector will
7315 set this bit. The mark function need only ensure that any other
7316 objects the smob refers to get marked.
7318 Note that this change means that the smob's GC8MARK bit is typically
7319 already set upon entry to the mark function. Thus, marking functions
7320 which look like this:
7323 if (SCM_GC8MARKP (ptr))
7325 SCM_SETGC8MARK (ptr);
7326 ... mark objects to which the smob refers ...
7329 are now incorrect, since they will return early, and fail to mark any
7330 other objects the smob refers to. Some code in the Guile library used
7333 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
7335 If you have implemented your own I/O port type, by writing the
7336 functions required by the scm_ptobfuns and then calling scm_newptob,
7337 you will need to change your functions slightly.
7339 The functions in a scm_ptobfuns structure now expect the port itself
7340 as their argument; they used to expect the `stream' member of the
7341 port's scm_port_table structure. This allows functions in an
7342 scm_ptobfuns structure to easily access the port's cell (and any flags
7343 it its CAR), and the port's scm_port_table structure.
7345 Guile now passes the I/O port itself as the `port' argument in the
7346 following scm_ptobfuns functions:
7348 int (*free) (SCM port);
7349 int (*fputc) (int, SCM port);
7350 int (*fputs) (char *, SCM port);
7351 scm_sizet (*fwrite) SCM_P ((char *ptr,
7355 int (*fflush) (SCM port);
7356 int (*fgetc) (SCM port);
7357 int (*fclose) (SCM port);
7359 The interfaces to the `mark', `print', `equalp', and `fgets' methods
7362 If you have existing code which defines its own port types, it is easy
7363 to convert your code to the new interface; simply apply SCM_STREAM to
7364 the port argument to yield the value you code used to expect.
7366 Note that since both the port and the stream have the same type in the
7367 C code --- they are both SCM values --- the C compiler will not remind
7368 you if you forget to update your scm_ptobfuns functions.
7371 ** Function: int scm_internal_select (int fds,
7375 struct timeval *timeout);
7377 This is a replacement for the `select' function provided by the OS.
7378 It enables I/O blocking and sleeping to happen for one cooperative
7379 thread without blocking other threads. It also avoids busy-loops in
7380 these situations. It is intended that all I/O blocking and sleeping
7381 will finally go through this function. Currently, this function is
7382 only available on systems providing `gettimeofday' and `select'.
7384 ** Function: SCM scm_internal_stack_catch (SCM tag,
7385 scm_catch_body_t body,
7387 scm_catch_handler_t handler,
7390 A new sibling to the other two C level `catch' functions
7391 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
7392 the stack to be saved automatically into the variable `the-last-stack'
7393 (scm_the_last_stack_var) on error. This is necessary if you want to
7394 use advanced error reporting, such as calling scm_display_error and
7395 scm_display_backtrace. (They both take a stack object as argument.)
7397 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
7399 scm_catch_handler_t handler,
7402 Spawns a new thread. It does a job similar to
7403 scm_call_with_new_thread but takes arguments more suitable when
7404 spawning threads from application C code.
7406 ** The hook scm_error_callback has been removed. It was originally
7407 intended as a way for the user to install his own error handler. But
7408 that method works badly since it intervenes between throw and catch,
7409 thereby changing the semantics of expressions like (catch #t ...).
7410 The correct way to do it is to use one of the C level catch functions
7411 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
7413 ** Removed functions:
7415 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
7416 scm_list_reverse, scm_list_reverse_x
7418 ** New macros: SCM_LISTn where n is one of the integers 0-9.
7420 These can be used for pretty list creation from C. The idea is taken
7421 from Erick Gallesio's STk.
7423 ** scm_array_map renamed to scm_array_map_x
7425 ** mbstrings are now removed
7427 This means that the type codes scm_tc7_mb_string and
7428 scm_tc7_mb_substring has been removed.
7430 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
7432 Since we no longer support multi-byte strings, these I/O functions
7433 have been simplified, and renamed. Here are their old names, and
7434 their new names and arguments:
7436 scm_gen_putc -> void scm_putc (int c, SCM port);
7437 scm_gen_puts -> void scm_puts (char *s, SCM port);
7438 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
7439 scm_gen_getc -> void scm_getc (SCM port);
7442 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
7444 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
7446 SCM_TYP7S now masks away the bit which distinguishes substrings from
7449 ** scm_catch_body_t: Backward incompatible change!
7451 Body functions to scm_internal_catch and friends do not any longer
7452 take a second argument. This is because it is no longer possible to
7453 pass a #f arg to catch.
7455 ** Calls to scm_protect_object and scm_unprotect now nest properly.
7457 The function scm_protect_object protects its argument from being freed
7458 by the garbage collector. scm_unprotect_object removes that
7461 These functions now nest properly. That is, for every object O, there
7462 is a counter which scm_protect_object(O) increments and
7463 scm_unprotect_object(O) decrements, if the counter is greater than
7464 zero. Every object's counter is zero when it is first created. If an
7465 object's counter is greater than zero, the garbage collector will not
7466 reclaim its storage.
7468 This allows you to use scm_protect_object in your code without
7469 worrying that some other function you call will call
7470 scm_unprotect_object, and allow it to be freed. Assuming that the
7471 functions you call are well-behaved, and unprotect only those objects
7472 they protect, you can follow the same rule and have confidence that
7473 objects will be freed only at appropriate times.
7476 Changes in Guile 1.2 (released Tuesday, June 24 1997):
7478 * Changes to the distribution
7480 ** Nightly snapshots are now available from ftp.red-bean.com.
7481 The old server, ftp.cyclic.com, has been relinquished to its rightful
7484 Nightly snapshots of the Guile development sources are now available via
7485 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
7487 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
7488 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
7490 ** To run Guile without installing it, the procedure has changed a bit.
7492 If you used a separate build directory to compile Guile, you'll need
7493 to include the build directory in SCHEME_LOAD_PATH, as well as the
7494 source directory. See the `INSTALL' file for examples.
7496 * Changes to the procedure for linking libguile with your programs
7498 ** The standard Guile load path for Scheme code now includes
7499 $(datadir)/guile (usually /usr/local/share/guile). This means that
7500 you can install your own Scheme files there, and Guile will find them.
7501 (Previous versions of Guile only checked a directory whose name
7502 contained the Guile version number, so you had to re-install or move
7503 your Scheme sources each time you installed a fresh version of Guile.)
7505 The load path also includes $(datadir)/guile/site; we recommend
7506 putting individual Scheme files there. If you want to install a
7507 package with multiple source files, create a directory for them under
7510 ** Guile 1.2 will now use the Rx regular expression library, if it is
7511 installed on your system. When you are linking libguile into your own
7512 programs, this means you will have to link against -lguile, -lqt (if
7513 you configured Guile with thread support), and -lrx.
7515 If you are using autoconf to generate configuration scripts for your
7516 application, the following lines should suffice to add the appropriate
7517 libraries to your link command:
7519 ### Find Rx, quickthreads and libguile.
7520 AC_CHECK_LIB(rx, main)
7521 AC_CHECK_LIB(qt, main)
7522 AC_CHECK_LIB(guile, scm_shell)
7524 The Guile 1.2 distribution does not contain sources for the Rx
7525 library, as Guile 1.0 did. If you want to use Rx, you'll need to
7526 retrieve it from a GNU FTP site and install it separately.
7528 * Changes to Scheme functions and syntax
7530 ** The dynamic linking features of Guile are now enabled by default.
7531 You can disable them by giving the `--disable-dynamic-linking' option
7534 (dynamic-link FILENAME)
7536 Find the object file denoted by FILENAME (a string) and link it
7537 into the running Guile application. When everything works out,
7538 return a Scheme object suitable for representing the linked object
7539 file. Otherwise an error is thrown. How object files are
7540 searched is system dependent.
7542 (dynamic-object? VAL)
7544 Determine whether VAL represents a dynamically linked object file.
7546 (dynamic-unlink DYNOBJ)
7548 Unlink the indicated object file from the application. DYNOBJ
7549 should be one of the values returned by `dynamic-link'.
7551 (dynamic-func FUNCTION DYNOBJ)
7553 Search the C function indicated by FUNCTION (a string or symbol)
7554 in DYNOBJ and return some Scheme object that can later be used
7555 with `dynamic-call' to actually call this function. Right now,
7556 these Scheme objects are formed by casting the address of the
7557 function to `long' and converting this number to its Scheme
7560 (dynamic-call FUNCTION DYNOBJ)
7562 Call the C function indicated by FUNCTION and DYNOBJ. The
7563 function is passed no arguments and its return value is ignored.
7564 When FUNCTION is something returned by `dynamic-func', call that
7565 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
7566 etc.), look it up in DYNOBJ; this is equivalent to
7568 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
7570 Interrupts are deferred while the C function is executing (with
7571 SCM_DEFER_INTS/SCM_ALLOW_INTS).
7573 (dynamic-args-call FUNCTION DYNOBJ ARGS)
7575 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
7576 some arguments and return its return value. The C function is
7577 expected to take two arguments and return an `int', just like
7580 int c_func (int argc, char **argv);
7582 ARGS must be a list of strings and is converted into an array of
7583 `char *'. The array is passed in ARGV and its size in ARGC. The
7584 return value is converted to a Scheme number and returned from the
7585 call to `dynamic-args-call'.
7587 When dynamic linking is disabled or not supported on your system,
7588 the above functions throw errors, but they are still available.
7590 Here is a small example that works on GNU/Linux:
7592 (define libc-obj (dynamic-link "libc.so"))
7593 (dynamic-args-call 'rand libc-obj '())
7595 See the file `libguile/DYNAMIC-LINKING' for additional comments.
7597 ** The #/ syntax for module names is depreciated, and will be removed
7598 in a future version of Guile. Instead of
7606 The latter syntax is more consistent with existing Lisp practice.
7608 ** Guile now does fancier printing of structures. Structures are the
7609 underlying implementation for records, which in turn are used to
7610 implement modules, so all of these object now print differently and in
7611 a more informative way.
7613 The Scheme printer will examine the builtin variable *struct-printer*
7614 whenever it needs to print a structure object. When this variable is
7615 not `#f' it is deemed to be a procedure and will be applied to the
7616 structure object and the output port. When *struct-printer* is `#f'
7617 or the procedure return `#f' the structure object will be printed in
7618 the boring #<struct 80458270> form.
7620 This hook is used by some routines in ice-9/boot-9.scm to implement
7621 type specific printing routines. Please read the comments there about
7624 One of the more specific uses of structs are records. The printing
7625 procedure that could be passed to MAKE-RECORD-TYPE is now actually
7626 called. It should behave like a *struct-printer* procedure (described
7629 ** Guile now supports a new R4RS-compliant syntax for keywords. A
7630 token of the form #:NAME, where NAME has the same syntax as a Scheme
7631 symbol, is the external representation of the keyword named NAME.
7632 Keyword objects print using this syntax as well, so values containing
7633 keyword objects can be read back into Guile. When used in an
7634 expression, keywords are self-quoting objects.
7636 Guile suports this read syntax, and uses this print syntax, regardless
7637 of the current setting of the `keyword' read option. The `keyword'
7638 read option only controls whether Guile recognizes the `:NAME' syntax,
7639 which is incompatible with R4RS. (R4RS says such token represent
7642 ** Guile has regular expression support again. Guile 1.0 included
7643 functions for matching regular expressions, based on the Rx library.
7644 In Guile 1.1, the Guile/Rx interface was removed to simplify the
7645 distribution, and thus Guile had no regular expression support. Guile
7646 1.2 again supports the most commonly used functions, and supports all
7647 of SCSH's regular expression functions.
7649 If your system does not include a POSIX regular expression library,
7650 and you have not linked Guile with a third-party regexp library such as
7651 Rx, these functions will not be available. You can tell whether your
7652 Guile installation includes regular expression support by checking
7653 whether the `*features*' list includes the `regex' symbol.
7655 *** regexp functions
7657 By default, Guile supports POSIX extended regular expressions. That
7658 means that the characters `(', `)', `+' and `?' are special, and must
7659 be escaped if you wish to match the literal characters.
7661 This regular expression interface was modeled after that implemented
7662 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
7663 with SCSH regular expressions.
7665 **** Function: string-match PATTERN STR [START]
7666 Compile the string PATTERN into a regular expression and compare
7667 it with STR. The optional numeric argument START specifies the
7668 position of STR at which to begin matching.
7670 `string-match' returns a "match structure" which describes what,
7671 if anything, was matched by the regular expression. *Note Match
7672 Structures::. If STR does not match PATTERN at all,
7673 `string-match' returns `#f'.
7675 Each time `string-match' is called, it must compile its PATTERN
7676 argument into a regular expression structure. This operation is
7677 expensive, which makes `string-match' inefficient if the same regular
7678 expression is used several times (for example, in a loop). For better
7679 performance, you can compile a regular expression in advance and then
7680 match strings against the compiled regexp.
7682 **** Function: make-regexp STR [FLAGS]
7683 Compile the regular expression described by STR, and return the
7684 compiled regexp structure. If STR does not describe a legal
7685 regular expression, `make-regexp' throws a
7686 `regular-expression-syntax' error.
7688 FLAGS may be the bitwise-or of one or more of the following:
7690 **** Constant: regexp/extended
7691 Use POSIX Extended Regular Expression syntax when interpreting
7692 STR. If not set, POSIX Basic Regular Expression syntax is used.
7693 If the FLAGS argument is omitted, we assume regexp/extended.
7695 **** Constant: regexp/icase
7696 Do not differentiate case. Subsequent searches using the
7697 returned regular expression will be case insensitive.
7699 **** Constant: regexp/newline
7700 Match-any-character operators don't match a newline.
7702 A non-matching list ([^...]) not containing a newline matches a
7705 Match-beginning-of-line operator (^) matches the empty string
7706 immediately after a newline, regardless of whether the FLAGS
7707 passed to regexp-exec contain regexp/notbol.
7709 Match-end-of-line operator ($) matches the empty string
7710 immediately before a newline, regardless of whether the FLAGS
7711 passed to regexp-exec contain regexp/noteol.
7713 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
7714 Match the compiled regular expression REGEXP against `str'. If
7715 the optional integer START argument is provided, begin matching
7716 from that position in the string. Return a match structure
7717 describing the results of the match, or `#f' if no match could be
7720 FLAGS may be the bitwise-or of one or more of the following:
7722 **** Constant: regexp/notbol
7723 The match-beginning-of-line operator always fails to match (but
7724 see the compilation flag regexp/newline above) This flag may be
7725 used when different portions of a string are passed to
7726 regexp-exec and the beginning of the string should not be
7727 interpreted as the beginning of the line.
7729 **** Constant: regexp/noteol
7730 The match-end-of-line operator always fails to match (but see the
7731 compilation flag regexp/newline above)
7733 **** Function: regexp? OBJ
7734 Return `#t' if OBJ is a compiled regular expression, or `#f'
7737 Regular expressions are commonly used to find patterns in one string
7738 and replace them with the contents of another string.
7740 **** Function: regexp-substitute PORT MATCH [ITEM...]
7741 Write to the output port PORT selected contents of the match
7742 structure MATCH. Each ITEM specifies what should be written, and
7743 may be one of the following arguments:
7745 * A string. String arguments are written out verbatim.
7747 * An integer. The submatch with that number is written.
7749 * The symbol `pre'. The portion of the matched string preceding
7750 the regexp match is written.
7752 * The symbol `post'. The portion of the matched string
7753 following the regexp match is written.
7755 PORT may be `#f', in which case nothing is written; instead,
7756 `regexp-substitute' constructs a string from the specified ITEMs
7759 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
7760 Similar to `regexp-substitute', but can be used to perform global
7761 substitutions on STR. Instead of taking a match structure as an
7762 argument, `regexp-substitute/global' takes two string arguments: a
7763 REGEXP string describing a regular expression, and a TARGET string
7764 which should be matched against this regular expression.
7766 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
7769 * A function may be supplied. When this function is called, it
7770 will be passed one argument: a match structure for a given
7771 regular expression match. It should return a string to be
7772 written out to PORT.
7774 * The `post' symbol causes `regexp-substitute/global' to recurse
7775 on the unmatched portion of STR. This *must* be supplied in
7776 order to perform global search-and-replace on STR; if it is
7777 not present among the ITEMs, then `regexp-substitute/global'
7778 will return after processing a single match.
7780 *** Match Structures
7782 A "match structure" is the object returned by `string-match' and
7783 `regexp-exec'. It describes which portion of a string, if any, matched
7784 the given regular expression. Match structures include: a reference to
7785 the string that was checked for matches; the starting and ending
7786 positions of the regexp match; and, if the regexp included any
7787 parenthesized subexpressions, the starting and ending positions of each
7790 In each of the regexp match functions described below, the `match'
7791 argument must be a match structure returned by a previous call to
7792 `string-match' or `regexp-exec'. Most of these functions return some
7793 information about the original target string that was matched against a
7794 regular expression; we will call that string TARGET for easy reference.
7796 **** Function: regexp-match? OBJ
7797 Return `#t' if OBJ is a match structure returned by a previous
7798 call to `regexp-exec', or `#f' otherwise.
7800 **** Function: match:substring MATCH [N]
7801 Return the portion of TARGET matched by subexpression number N.
7802 Submatch 0 (the default) represents the entire regexp match. If
7803 the regular expression as a whole matched, but the subexpression
7804 number N did not match, return `#f'.
7806 **** Function: match:start MATCH [N]
7807 Return the starting position of submatch number N.
7809 **** Function: match:end MATCH [N]
7810 Return the ending position of submatch number N.
7812 **** Function: match:prefix MATCH
7813 Return the unmatched portion of TARGET preceding the regexp match.
7815 **** Function: match:suffix MATCH
7816 Return the unmatched portion of TARGET following the regexp match.
7818 **** Function: match:count MATCH
7819 Return the number of parenthesized subexpressions from MATCH.
7820 Note that the entire regular expression match itself counts as a
7821 subexpression, and failed submatches are included in the count.
7823 **** Function: match:string MATCH
7824 Return the original TARGET string.
7826 *** Backslash Escapes
7828 Sometimes you will want a regexp to match characters like `*' or `$'
7829 exactly. For example, to check whether a particular string represents
7830 a menu entry from an Info node, it would be useful to match it against
7831 a regexp like `^* [^:]*::'. However, this won't work; because the
7832 asterisk is a metacharacter, it won't match the `*' at the beginning of
7833 the string. In this case, we want to make the first asterisk un-magic.
7835 You can do this by preceding the metacharacter with a backslash
7836 character `\'. (This is also called "quoting" the metacharacter, and
7837 is known as a "backslash escape".) When Guile sees a backslash in a
7838 regular expression, it considers the following glyph to be an ordinary
7839 character, no matter what special meaning it would ordinarily have.
7840 Therefore, we can make the above example work by changing the regexp to
7841 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
7842 to match only a single asterisk in the target string.
7844 Since the backslash is itself a metacharacter, you may force a
7845 regexp to match a backslash in the target string by preceding the
7846 backslash with itself. For example, to find variable references in a
7847 TeX program, you might want to find occurrences of the string `\let\'
7848 followed by any number of alphabetic characters. The regular expression
7849 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
7850 each match a single backslash in the target string.
7852 **** Function: regexp-quote STR
7853 Quote each special character found in STR with a backslash, and
7854 return the resulting string.
7856 *Very important:* Using backslash escapes in Guile source code (as
7857 in Emacs Lisp or C) can be tricky, because the backslash character has
7858 special meaning for the Guile reader. For example, if Guile encounters
7859 the character sequence `\n' in the middle of a string while processing
7860 Scheme code, it replaces those characters with a newline character.
7861 Similarly, the character sequence `\t' is replaced by a horizontal tab.
7862 Several of these "escape sequences" are processed by the Guile reader
7863 before your code is executed. Unrecognized escape sequences are
7864 ignored: if the characters `\*' appear in a string, they will be
7865 translated to the single character `*'.
7867 This translation is obviously undesirable for regular expressions,
7868 since we want to be able to include backslashes in a string in order to
7869 escape regexp metacharacters. Therefore, to make sure that a backslash
7870 is preserved in a string in your Guile program, you must use *two*
7871 consecutive backslashes:
7873 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
7875 The string in this example is preprocessed by the Guile reader before
7876 any code is executed. The resulting argument to `make-regexp' is the
7877 string `^\* [^:]*', which is what we really want.
7879 This also means that in order to write a regular expression that
7880 matches a single backslash character, the regular expression string in
7881 the source code must include *four* backslashes. Each consecutive pair
7882 of backslashes gets translated by the Guile reader to a single
7883 backslash, and the resulting double-backslash is interpreted by the
7884 regexp engine as matching a single backslash character. Hence:
7886 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
7888 The reason for the unwieldiness of this syntax is historical. Both
7889 regular expression pattern matchers and Unix string processing systems
7890 have traditionally used backslashes with the special meanings described
7891 above. The POSIX regular expression specification and ANSI C standard
7892 both require these semantics. Attempting to abandon either convention
7893 would cause other kinds of compatibility problems, possibly more severe
7894 ones. Therefore, without extending the Scheme reader to support
7895 strings with different quoting conventions (an ungainly and confusing
7896 extension when implemented in other languages), we must adhere to this
7897 cumbersome escape syntax.
7899 * Changes to the gh_ interface
7901 * Changes to the scm_ interface
7903 * Changes to system call interfaces:
7905 ** The value returned by `raise' is now unspecified. It throws an exception
7908 *** A new procedure `sigaction' can be used to install signal handlers
7910 (sigaction signum [action] [flags])
7912 signum is the signal number, which can be specified using the value
7915 If action is omitted, sigaction returns a pair: the CAR is the current
7916 signal hander, which will be either an integer with the value SIG_DFL
7917 (default action) or SIG_IGN (ignore), or the Scheme procedure which
7918 handles the signal, or #f if a non-Scheme procedure handles the
7919 signal. The CDR contains the current sigaction flags for the handler.
7921 If action is provided, it is installed as the new handler for signum.
7922 action can be a Scheme procedure taking one argument, or the value of
7923 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
7924 whatever signal handler was installed before sigaction was first used.
7925 Flags can optionally be specified for the new handler (SA_RESTART is
7926 always used if the system provides it, so need not be specified.) The
7927 return value is a pair with information about the old handler as
7930 This interface does not provide access to the "signal blocking"
7931 facility. Maybe this is not needed, since the thread support may
7932 provide solutions to the problem of consistent access to data
7935 *** A new procedure `flush-all-ports' is equivalent to running
7936 `force-output' on every port open for output.
7938 ** Guile now provides information on how it was built, via the new
7939 global variable, %guile-build-info. This variable records the values
7940 of the standard GNU makefile directory variables as an assocation
7941 list, mapping variable names (symbols) onto directory paths (strings).
7942 For example, to find out where the Guile link libraries were
7943 installed, you can say:
7945 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
7948 * Changes to the scm_ interface
7950 ** The new function scm_handle_by_message_noexit is just like the
7951 existing scm_handle_by_message function, except that it doesn't call
7952 exit to terminate the process. Instead, it prints a message and just
7953 returns #f. This might be a more appropriate catch-all handler for
7954 new dynamic roots and threads.
7957 Changes in Guile 1.1 (released Friday, May 16 1997):
7959 * Changes to the distribution.
7961 The Guile 1.0 distribution has been split up into several smaller
7963 guile-core --- the Guile interpreter itself.
7964 guile-tcltk --- the interface between the Guile interpreter and
7965 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
7966 is a toolkit for building graphical user interfaces.
7967 guile-rgx-ctax --- the interface between Guile and the Rx regular
7968 expression matcher, and the translator for the Ctax
7969 programming language. These are packaged together because the
7970 Ctax translator uses Rx to parse Ctax source code.
7972 This NEWS file describes the changes made to guile-core since the 1.0
7975 We no longer distribute the documentation, since it was either out of
7976 date, or incomplete. As soon as we have current documentation, we
7981 * Changes to the stand-alone interpreter
7983 ** guile now accepts command-line arguments compatible with SCSH, Olin
7984 Shivers' Scheme Shell.
7986 In general, arguments are evaluated from left to right, but there are
7987 exceptions. The following switches stop argument processing, and
7988 stash all remaining command-line arguments as the value returned by
7989 the (command-line) function.
7990 -s SCRIPT load Scheme source code from FILE, and exit
7991 -c EXPR evalute Scheme expression EXPR, and exit
7992 -- stop scanning arguments; run interactively
7994 The switches below are processed as they are encountered.
7995 -l FILE load Scheme source code from FILE
7996 -e FUNCTION after reading script, apply FUNCTION to
7997 command line arguments
7998 -ds do -s script at this point
7999 --emacs enable Emacs protocol (experimental)
8000 -h, --help display this help and exit
8001 -v, --version display version information and exit
8002 \ read arguments from following script lines
8004 So, for example, here is a Guile script named `ekko' (thanks, Olin)
8005 which re-implements the traditional "echo" command:
8007 #!/usr/local/bin/guile -s
8010 (map (lambda (arg) (display arg) (display " "))
8014 (main (command-line))
8016 Suppose we invoke this script as follows:
8018 ekko a speckled gecko
8020 Through the magic of Unix script processing (triggered by the `#!'
8021 token at the top of the file), /usr/local/bin/guile receives the
8022 following list of command-line arguments:
8024 ("-s" "./ekko" "a" "speckled" "gecko")
8026 Unix inserts the name of the script after the argument specified on
8027 the first line of the file (in this case, "-s"), and then follows that
8028 with the arguments given to the script. Guile loads the script, which
8029 defines the `main' function, and then applies it to the list of
8030 remaining command-line arguments, ("a" "speckled" "gecko").
8032 In Unix, the first line of a script file must take the following form:
8034 #!INTERPRETER ARGUMENT
8036 where INTERPRETER is the absolute filename of the interpreter
8037 executable, and ARGUMENT is a single command-line argument to pass to
8040 You may only pass one argument to the interpreter, and its length is
8041 limited. These restrictions can be annoying to work around, so Guile
8042 provides a general mechanism (borrowed from, and compatible with,
8043 SCSH) for circumventing them.
8045 If the ARGUMENT in a Guile script is a single backslash character,
8046 `\', Guile will open the script file, parse arguments from its second
8047 and subsequent lines, and replace the `\' with them. So, for example,
8048 here is another implementation of the `ekko' script:
8050 #!/usr/local/bin/guile \
8054 (for-each (lambda (arg) (display arg) (display " "))
8058 If the user invokes this script as follows:
8060 ekko a speckled gecko
8062 Unix expands this into
8064 /usr/local/bin/guile \ ekko a speckled gecko
8066 When Guile sees the `\' argument, it replaces it with the arguments
8067 read from the second line of the script, producing:
8069 /usr/local/bin/guile -e main -s ekko a speckled gecko
8071 This tells Guile to load the `ekko' script, and apply the function
8072 `main' to the argument list ("a" "speckled" "gecko").
8074 Here is how Guile parses the command-line arguments:
8075 - Each space character terminates an argument. This means that two
8076 spaces in a row introduce an empty-string argument.
8077 - The tab character is not permitted (unless you quote it with the
8078 backslash character, as described below), to avoid confusion.
8079 - The newline character terminates the sequence of arguments, and will
8080 also terminate a final non-empty argument. (However, a newline
8081 following a space will not introduce a final empty-string argument;
8082 it only terminates the argument list.)
8083 - The backslash character is the escape character. It escapes
8084 backslash, space, tab, and newline. The ANSI C escape sequences
8085 like \n and \t are also supported. These produce argument
8086 constituents; the two-character combination \n doesn't act like a
8087 terminating newline. The escape sequence \NNN for exactly three
8088 octal digits reads as the character whose ASCII code is NNN. As
8089 above, characters produced this way are argument constituents.
8090 Backslash followed by other characters is not allowed.
8092 * Changes to the procedure for linking libguile with your programs
8094 ** Guile now builds and installs a shared guile library, if your
8095 system support shared libraries. (It still builds a static library on
8096 all systems.) Guile automatically detects whether your system
8097 supports shared libraries. To prevent Guile from buildisg shared
8098 libraries, pass the `--disable-shared' flag to the configure script.
8100 Guile takes longer to compile when it builds shared libraries, because
8101 it must compile every file twice --- once to produce position-
8102 independent object code, and once to produce normal object code.
8104 ** The libthreads library has been merged into libguile.
8106 To link a program against Guile, you now need only link against
8107 -lguile and -lqt; -lthreads is no longer needed. If you are using
8108 autoconf to generate configuration scripts for your application, the
8109 following lines should suffice to add the appropriate libraries to
8112 ### Find quickthreads and libguile.
8113 AC_CHECK_LIB(qt, main)
8114 AC_CHECK_LIB(guile, scm_shell)
8116 * Changes to Scheme functions
8118 ** Guile Scheme's special syntax for keyword objects is now optional,
8119 and disabled by default.
8121 The syntax variation from R4RS made it difficult to port some
8122 interesting packages to Guile. The routines which accepted keyword
8123 arguments (mostly in the module system) have been modified to also
8124 accept symbols whose names begin with `:'.
8126 To change the keyword syntax, you must first import the (ice-9 debug)
8128 (use-modules (ice-9 debug))
8130 Then you can enable the keyword syntax as follows:
8131 (read-set! keywords 'prefix)
8133 To disable keyword syntax, do this:
8134 (read-set! keywords #f)
8136 ** Many more primitive functions accept shared substrings as
8137 arguments. In the past, these functions required normal, mutable
8138 strings as arguments, although they never made use of this
8141 ** The uniform array functions now operate on byte vectors. These
8142 functions are `array-fill!', `serial-array-copy!', `array-copy!',
8143 `serial-array-map', `array-map', `array-for-each', and
8146 ** The new functions `trace' and `untrace' implement simple debugging
8147 support for Scheme functions.
8149 The `trace' function accepts any number of procedures as arguments,
8150 and tells the Guile interpreter to display each procedure's name and
8151 arguments each time the procedure is invoked. When invoked with no
8152 arguments, `trace' returns the list of procedures currently being
8155 The `untrace' function accepts any number of procedures as arguments,
8156 and tells the Guile interpreter not to trace them any more. When
8157 invoked with no arguments, `untrace' untraces all curretly traced
8160 The tracing in Guile has an advantage over most other systems: we
8161 don't create new procedure objects, but mark the procedure objects
8162 themselves. This means that anonymous and internal procedures can be
8165 ** The function `assert-repl-prompt' has been renamed to
8166 `set-repl-prompt!'. It takes one argument, PROMPT.
8167 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
8168 - If PROMPT is a string, we use it as a prompt.
8169 - If PROMPT is a procedure accepting no arguments, we call it, and
8170 display the result as a prompt.
8171 - Otherwise, we display "> ".
8173 ** The new function `eval-string' reads Scheme expressions from a
8174 string and evaluates them, returning the value of the last expression
8175 in the string. If the string contains no expressions, it returns an
8178 ** The new function `thunk?' returns true iff its argument is a
8179 procedure of zero arguments.
8181 ** `defined?' is now a builtin function, instead of syntax. This
8182 means that its argument should be quoted. It returns #t iff its
8183 argument is bound in the current module.
8185 ** The new syntax `use-modules' allows you to add new modules to your
8186 environment without re-typing a complete `define-module' form. It
8187 accepts any number of module names as arguments, and imports their
8188 public bindings into the current module.
8190 ** The new function (module-defined? NAME MODULE) returns true iff
8191 NAME, a symbol, is defined in MODULE, a module object.
8193 ** The new function `builtin-bindings' creates and returns a hash
8194 table containing copies of all the root module's bindings.
8196 ** The new function `builtin-weak-bindings' does the same as
8197 `builtin-bindings', but creates a doubly-weak hash table.
8199 ** The `equal?' function now considers variable objects to be
8200 equivalent if they have the same name and the same value.
8202 ** The new function `command-line' returns the command-line arguments
8203 given to Guile, as a list of strings.
8205 When using guile as a script interpreter, `command-line' returns the
8206 script's arguments; those processed by the interpreter (like `-s' or
8207 `-c') are omitted. (In other words, you get the normal, expected
8208 behavior.) Any application that uses scm_shell to process its
8209 command-line arguments gets this behavior as well.
8211 ** The new function `load-user-init' looks for a file called `.guile'
8212 in the user's home directory, and loads it if it exists. This is
8213 mostly for use by the code generated by scm_compile_shell_switches,
8214 but we thought it might also be useful in other circumstances.
8216 ** The new function `log10' returns the base-10 logarithm of its
8219 ** Changes to I/O functions
8221 *** The functions `read', `primitive-load', `read-and-eval!', and
8222 `primitive-load-path' no longer take optional arguments controlling
8223 case insensitivity and a `#' parser.
8225 Case sensitivity is now controlled by a read option called
8226 `case-insensitive'. The user can add new `#' syntaxes with the
8227 `read-hash-extend' function (see below).
8229 *** The new function `read-hash-extend' allows the user to change the
8230 syntax of Guile Scheme in a somewhat controlled way.
8232 (read-hash-extend CHAR PROC)
8233 When parsing S-expressions, if we read a `#' character followed by
8234 the character CHAR, use PROC to parse an object from the stream.
8235 If PROC is #f, remove any parsing procedure registered for CHAR.
8237 The reader applies PROC to two arguments: CHAR and an input port.
8239 *** The new functions read-delimited and read-delimited! provide a
8240 general mechanism for doing delimited input on streams.
8242 (read-delimited DELIMS [PORT HANDLE-DELIM])
8243 Read until we encounter one of the characters in DELIMS (a string),
8244 or end-of-file. PORT is the input port to read from; it defaults to
8245 the current input port. The HANDLE-DELIM parameter determines how
8246 the terminating character is handled; it should be one of the
8249 'trim omit delimiter from result
8250 'peek leave delimiter character in input stream
8251 'concat append delimiter character to returned value
8252 'split return a pair: (RESULT . TERMINATOR)
8254 HANDLE-DELIM defaults to 'peek.
8256 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
8257 A side-effecting variant of `read-delimited'.
8259 The data is written into the string BUF at the indices in the
8260 half-open interval [START, END); the default interval is the whole
8261 string: START = 0 and END = (string-length BUF). The values of
8262 START and END must specify a well-defined interval in BUF, i.e.
8263 0 <= START <= END <= (string-length BUF).
8265 It returns NBYTES, the number of bytes read. If the buffer filled
8266 up without a delimiter character being found, it returns #f. If the
8267 port is at EOF when the read starts, it returns the EOF object.
8269 If an integer is returned (i.e., the read is successfully terminated
8270 by reading a delimiter character), then the HANDLE-DELIM parameter
8271 determines how to handle the terminating character. It is described
8272 above, and defaults to 'peek.
8274 (The descriptions of these functions were borrowed from the SCSH
8275 manual, by Olin Shivers and Brian Carlstrom.)
8277 *** The `%read-delimited!' function is the primitive used to implement
8278 `read-delimited' and `read-delimited!'.
8280 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
8282 This returns a pair of values: (TERMINATOR . NUM-READ).
8283 - TERMINATOR describes why the read was terminated. If it is a
8284 character or the eof object, then that is the value that terminated
8285 the read. If it is #f, the function filled the buffer without finding
8286 a delimiting character.
8287 - NUM-READ is the number of characters read into BUF.
8289 If the read is successfully terminated by reading a delimiter
8290 character, then the gobble? parameter determines what to do with the
8291 terminating character. If true, the character is removed from the
8292 input stream; if false, the character is left in the input stream
8293 where a subsequent read operation will retrieve it. In either case,
8294 the character is also the first value returned by the procedure call.
8296 (The descriptions of this function was borrowed from the SCSH manual,
8297 by Olin Shivers and Brian Carlstrom.)
8299 *** The `read-line' and `read-line!' functions have changed; they now
8300 trim the terminator by default; previously they appended it to the
8301 returned string. For the old behavior, use (read-line PORT 'concat).
8303 *** The functions `uniform-array-read!' and `uniform-array-write!' now
8304 take new optional START and END arguments, specifying the region of
8305 the array to read and write.
8307 *** The `ungetc-char-ready?' function has been removed. We feel it's
8308 inappropriate for an interface to expose implementation details this
8311 ** Changes to the Unix library and system call interface
8313 *** The new fcntl function provides access to the Unix `fcntl' system
8316 (fcntl PORT COMMAND VALUE)
8317 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
8318 Values for COMMAND are:
8320 F_DUPFD duplicate a file descriptor
8321 F_GETFD read the descriptor's close-on-exec flag
8322 F_SETFD set the descriptor's close-on-exec flag to VALUE
8323 F_GETFL read the descriptor's flags, as set on open
8324 F_SETFL set the descriptor's flags, as set on open to VALUE
8325 F_GETOWN return the process ID of a socket's owner, for SIGIO
8326 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
8327 FD_CLOEXEC not sure what this is
8329 For details, see the documentation for the fcntl system call.
8331 *** The arguments to `select' have changed, for compatibility with
8332 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
8333 expected behavior. The MILLISECONDS parameter has been changed to
8334 MICROSECONDS, to more closely resemble the underlying system call.
8335 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
8336 corresponding return set will be the same.
8338 *** The arguments to the `mknod' system call have changed. They are
8341 (mknod PATH TYPE PERMS DEV)
8342 Create a new file (`node') in the file system. PATH is the name of
8343 the file to create. TYPE is the kind of file to create; it should
8344 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
8345 permission bits to give the newly created file. If TYPE is
8346 'block-special or 'char-special, DEV specifies which device the
8347 special file refers to; its interpretation depends on the kind of
8348 special file being created.
8350 *** The `fork' function has been renamed to `primitive-fork', to avoid
8351 clashing with various SCSH forks.
8353 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
8354 and `recvfrom!'. They no longer accept a size for a second argument;
8355 you must pass a string to hold the received value. They no longer
8356 return the buffer. Instead, `recv' returns the length of the message
8357 received, and `recvfrom' returns a pair containing the packet's length
8358 and originating address.
8360 *** The file descriptor datatype has been removed, as have the
8361 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
8362 We plan to replace these functions with a SCSH-compatible interface.
8364 *** The `create' function has been removed; it's just a special case
8367 *** There are new functions to break down process termination status
8368 values. In the descriptions below, STATUS is a value returned by
8371 (status:exit-val STATUS)
8372 If the child process exited normally, this function returns the exit
8373 code for the child process (i.e., the value passed to exit, or
8374 returned from main). If the child process did not exit normally,
8375 this function returns #f.
8377 (status:stop-sig STATUS)
8378 If the child process was suspended by a signal, this function
8379 returns the signal that suspended the child. Otherwise, it returns
8382 (status:term-sig STATUS)
8383 If the child process terminated abnormally, this function returns
8384 the signal that terminated the child. Otherwise, this function
8387 POSIX promises that exactly one of these functions will return true on
8388 a valid STATUS value.
8390 These functions are compatible with SCSH.
8392 *** There are new accessors and setters for the broken-out time vectors
8393 returned by `localtime', `gmtime', and that ilk. They are:
8395 Component Accessor Setter
8396 ========================= ============ ============
8397 seconds tm:sec set-tm:sec
8398 minutes tm:min set-tm:min
8399 hours tm:hour set-tm:hour
8400 day of the month tm:mday set-tm:mday
8401 month tm:mon set-tm:mon
8402 year tm:year set-tm:year
8403 day of the week tm:wday set-tm:wday
8404 day in the year tm:yday set-tm:yday
8405 daylight saving time tm:isdst set-tm:isdst
8406 GMT offset, seconds tm:gmtoff set-tm:gmtoff
8407 name of time zone tm:zone set-tm:zone
8409 *** There are new accessors for the vectors returned by `uname',
8410 describing the host system:
8413 ============================================== ================
8414 name of the operating system implementation utsname:sysname
8415 network name of this machine utsname:nodename
8416 release level of the operating system utsname:release
8417 version level of the operating system utsname:version
8418 machine hardware platform utsname:machine
8420 *** There are new accessors for the vectors returned by `getpw',
8421 `getpwnam', `getpwuid', and `getpwent', describing entries from the
8422 system's user database:
8425 ====================== =================
8426 user name passwd:name
8427 user password passwd:passwd
8430 real name passwd:gecos
8431 home directory passwd:dir
8432 shell program passwd:shell
8434 *** There are new accessors for the vectors returned by `getgr',
8435 `getgrnam', `getgrgid', and `getgrent', describing entries from the
8436 system's group database:
8439 ======================= ============
8440 group name group:name
8441 group password group:passwd
8443 group members group:mem
8445 *** There are new accessors for the vectors returned by `gethost',
8446 `gethostbyaddr', `gethostbyname', and `gethostent', describing
8450 ========================= ===============
8451 official name of host hostent:name
8452 alias list hostent:aliases
8453 host address type hostent:addrtype
8454 length of address hostent:length
8455 list of addresses hostent:addr-list
8457 *** There are new accessors for the vectors returned by `getnet',
8458 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
8462 ========================= ===============
8463 official name of net netent:name
8464 alias list netent:aliases
8465 net number type netent:addrtype
8466 net number netent:net
8468 *** There are new accessors for the vectors returned by `getproto',
8469 `getprotobyname', `getprotobynumber', and `getprotoent', describing
8473 ========================= ===============
8474 official protocol name protoent:name
8475 alias list protoent:aliases
8476 protocol number protoent:proto
8478 *** There are new accessors for the vectors returned by `getserv',
8479 `getservbyname', `getservbyport', and `getservent', describing
8483 ========================= ===============
8484 official service name servent:name
8485 alias list servent:aliases
8486 port number servent:port
8487 protocol to use servent:proto
8489 *** There are new accessors for the sockaddr structures returned by
8490 `accept', `getsockname', `getpeername', `recvfrom!':
8493 ======================================== ===============
8494 address format (`family') sockaddr:fam
8495 path, for file domain addresses sockaddr:path
8496 address, for internet domain addresses sockaddr:addr
8497 TCP or UDP port, for internet sockaddr:port
8499 *** The `getpwent', `getgrent', `gethostent', `getnetent',
8500 `getprotoent', and `getservent' functions now return #f at the end of
8501 the user database. (They used to throw an exception.)
8503 Note that calling MUMBLEent function is equivalent to calling the
8504 corresponding MUMBLE function with no arguments.
8506 *** The `setpwent', `setgrent', `sethostent', `setnetent',
8507 `setprotoent', and `setservent' routines now take no arguments.
8509 *** The `gethost', `getproto', `getnet', and `getserv' functions now
8510 provide more useful information when they throw an exception.
8512 *** The `lnaof' function has been renamed to `inet-lnaof'.
8514 *** Guile now claims to have the `current-time' feature.
8516 *** The `mktime' function now takes an optional second argument ZONE,
8517 giving the time zone to use for the conversion. ZONE should be a
8518 string, in the same format as expected for the "TZ" environment variable.
8520 *** The `strptime' function now returns a pair (TIME . COUNT), where
8521 TIME is the parsed time as a vector, and COUNT is the number of
8522 characters from the string left unparsed. This function used to
8523 return the remaining characters as a string.
8525 *** The `gettimeofday' function has replaced the old `time+ticks' function.
8526 The return value is now (SECONDS . MICROSECONDS); the fractional
8527 component is no longer expressed in "ticks".
8529 *** The `ticks/sec' constant has been removed, in light of the above change.
8531 * Changes to the gh_ interface
8533 ** gh_eval_str() now returns an SCM object which is the result of the
8536 ** gh_scm2str() now copies the Scheme data to a caller-provided C
8539 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
8540 and returns the array
8542 ** gh_scm2str0() is gone: there is no need to distinguish
8543 null-terminated from non-null-terminated, since gh_scm2newstr() allows
8544 the user to interpret the data both ways.
8546 * Changes to the scm_ interface
8548 ** The new function scm_symbol_value0 provides an easy way to get a
8549 symbol's value from C code:
8551 SCM scm_symbol_value0 (char *NAME)
8552 Return the value of the symbol named by the null-terminated string
8553 NAME in the current module. If the symbol named NAME is unbound in
8554 the current module, return SCM_UNDEFINED.
8556 ** The new function scm_sysintern0 creates new top-level variables,
8557 without assigning them a value.
8559 SCM scm_sysintern0 (char *NAME)
8560 Create a new Scheme top-level variable named NAME. NAME is a
8561 null-terminated string. Return the variable's value cell.
8563 ** The function scm_internal_catch is the guts of catch. It handles
8564 all the mechanics of setting up a catch target, invoking the catch
8565 body, and perhaps invoking the handler if the body does a throw.
8567 The function is designed to be usable from C code, but is general
8568 enough to implement all the semantics Guile Scheme expects from throw.
8570 TAG is the catch tag. Typically, this is a symbol, but this function
8571 doesn't actually care about that.
8573 BODY is a pointer to a C function which runs the body of the catch;
8574 this is the code you can throw from. We call it like this:
8575 BODY (BODY_DATA, JMPBUF)
8577 BODY_DATA is just the BODY_DATA argument we received; we pass it
8578 through to BODY as its first argument. The caller can make
8579 BODY_DATA point to anything useful that BODY might need.
8580 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
8581 which we have just created and initialized.
8583 HANDLER is a pointer to a C function to deal with a throw to TAG,
8584 should one occur. We call it like this:
8585 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
8587 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
8588 same idea as BODY_DATA above.
8589 THROWN_TAG is the tag that the user threw to; usually this is
8590 TAG, but it could be something else if TAG was #t (i.e., a
8591 catch-all), or the user threw to a jmpbuf.
8592 THROW_ARGS is the list of arguments the user passed to the THROW
8595 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
8596 is just a pointer we pass through to HANDLER. We don't actually
8597 use either of those pointers otherwise ourselves. The idea is
8598 that, if our caller wants to communicate something to BODY or
8599 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
8600 HANDLER can then use. Think of it as a way to make BODY and
8601 HANDLER closures, not just functions; MUMBLE_DATA points to the
8604 Of course, it's up to the caller to make sure that any data a
8605 MUMBLE_DATA needs is protected from GC. A common way to do this is
8606 to make MUMBLE_DATA a pointer to data stored in an automatic
8607 structure variable; since the collector must scan the stack for
8608 references anyway, this assures that any references in MUMBLE_DATA
8611 ** The new function scm_internal_lazy_catch is exactly like
8612 scm_internal_catch, except:
8614 - It does not unwind the stack (this is the major difference).
8615 - If handler returns, its value is returned from the throw.
8616 - BODY always receives #f as its JMPBUF argument (since there's no
8617 jmpbuf associated with a lazy catch, because we don't unwind the
8620 ** scm_body_thunk is a new body function you can pass to
8621 scm_internal_catch if you want the body to be like Scheme's `catch'
8622 --- a thunk, or a function of one argument if the tag is #f.
8624 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
8625 contains the Scheme procedure to invoke as the body, and the tag
8626 we're catching. If the tag is #f, then we pass JMPBUF (created by
8627 scm_internal_catch) to the body procedure; otherwise, the body gets
8630 ** scm_handle_by_proc is a new handler function you can pass to
8631 scm_internal_catch if you want the handler to act like Scheme's catch
8632 --- call a procedure with the tag and the throw arguments.
8634 If the user does a throw to this catch, this function runs a handler
8635 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
8636 variable holding the Scheme procedure object to invoke. It ought to
8637 be a pointer to an automatic variable (i.e., one living on the stack),
8638 or the procedure object should be otherwise protected from GC.
8640 ** scm_handle_by_message is a new handler function to use with
8641 `scm_internal_catch' if you want Guile to print a message and die.
8642 It's useful for dealing with throws to uncaught keys at the top level.
8644 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
8645 message header to print; if zero, we use "guile" instead. That
8646 text is followed by a colon, then the message described by ARGS.
8648 ** The return type of scm_boot_guile is now void; the function does
8649 not return a value, and indeed, never returns at all.
8651 ** The new function scm_shell makes it easy for user applications to
8652 process command-line arguments in a way that is compatible with the
8653 stand-alone guile interpreter (which is in turn compatible with SCSH,
8656 To use the scm_shell function, first initialize any guile modules
8657 linked into your application, and then call scm_shell with the values
8658 of ARGC and ARGV your `main' function received. scm_shell will add
8659 any SCSH-style meta-arguments from the top of the script file to the
8660 argument vector, and then process the command-line arguments. This
8661 generally means loading a script file or starting up an interactive
8662 command interpreter. For details, see "Changes to the stand-alone
8665 ** The new functions scm_get_meta_args and scm_count_argv help you
8666 implement the SCSH-style meta-argument, `\'.
8668 char **scm_get_meta_args (int ARGC, char **ARGV)
8669 If the second element of ARGV is a string consisting of a single
8670 backslash character (i.e. "\\" in Scheme notation), open the file
8671 named by the following argument, parse arguments from it, and return
8672 the spliced command line. The returned array is terminated by a
8675 For details of argument parsing, see above, under "guile now accepts
8676 command-line arguments compatible with SCSH..."
8678 int scm_count_argv (char **ARGV)
8679 Count the arguments in ARGV, assuming it is terminated by a null
8682 For an example of how these functions might be used, see the source
8683 code for the function scm_shell in libguile/script.c.
8685 You will usually want to use scm_shell instead of calling this
8688 ** The new function scm_compile_shell_switches turns an array of
8689 command-line arguments into Scheme code to carry out the actions they
8690 describe. Given ARGC and ARGV, it returns a Scheme expression to
8691 evaluate, and calls scm_set_program_arguments to make any remaining
8692 command-line arguments available to the Scheme code. For example,
8693 given the following arguments:
8695 -e main -s ekko a speckled gecko
8697 scm_set_program_arguments will return the following expression:
8699 (begin (load "ekko") (main (command-line)) (quit))
8701 You will usually want to use scm_shell instead of calling this
8704 ** The function scm_shell_usage prints a usage message appropriate for
8705 an interpreter that uses scm_compile_shell_switches to handle its
8706 command-line arguments.
8708 void scm_shell_usage (int FATAL, char *MESSAGE)
8709 Print a usage message to the standard error output. If MESSAGE is
8710 non-zero, write it before the usage message, followed by a newline.
8711 If FATAL is non-zero, exit the process, using FATAL as the
8712 termination status. (If you want to be compatible with Guile,
8713 always use 1 as the exit status when terminating due to command-line
8716 You will usually want to use scm_shell instead of calling this
8719 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
8720 expressions. It used to return SCM_EOL. Earth-shattering.
8722 ** The macros for declaring scheme objects in C code have been
8723 rearranged slightly. They are now:
8725 SCM_SYMBOL (C_NAME, SCHEME_NAME)
8726 Declare a static SCM variable named C_NAME, and initialize it to
8727 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
8728 be a C identifier, and SCHEME_NAME should be a C string.
8730 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
8731 Just like SCM_SYMBOL, but make C_NAME globally visible.
8733 SCM_VCELL (C_NAME, SCHEME_NAME)
8734 Create a global variable at the Scheme level named SCHEME_NAME.
8735 Declare a static SCM variable named C_NAME, and initialize it to
8736 point to the Scheme variable's value cell.
8738 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
8739 Just like SCM_VCELL, but make C_NAME globally visible.
8741 The `guile-snarf' script writes initialization code for these macros
8742 to its standard output, given C source code as input.
8744 The SCM_GLOBAL macro is gone.
8746 ** The scm_read_line and scm_read_line_x functions have been replaced
8747 by Scheme code based on the %read-delimited! procedure (known to C
8748 code as scm_read_delimited_x). See its description above for more
8751 ** The function scm_sys_open has been renamed to scm_open. It now
8752 returns a port instead of an FD object.
8754 * The dynamic linking support has changed. For more information, see
8755 libguile/DYNAMIC-LINKING.
8760 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
8763 * Changes to the 'guile' program:
8765 ** Guile now loads some new files when it starts up. Guile first
8766 searches the load path for init.scm, and loads it if found. Then, if
8767 Guile is not being used to execute a script, and the user's home
8768 directory contains a file named `.guile', Guile loads that.
8770 ** You can now use Guile as a shell script interpreter.
8772 To paraphrase the SCSH manual:
8774 When Unix tries to execute an executable file whose first two
8775 characters are the `#!', it treats the file not as machine code to
8776 be directly executed by the native processor, but as source code
8777 to be executed by some interpreter. The interpreter to use is
8778 specified immediately after the #! sequence on the first line of
8779 the source file. The kernel reads in the name of the interpreter,
8780 and executes that instead. It passes the interpreter the source
8781 filename as its first argument, with the original arguments
8782 following. Consult the Unix man page for the `exec' system call
8783 for more information.
8785 Now you can use Guile as an interpreter, using a mechanism which is a
8786 compatible subset of that provided by SCSH.
8788 Guile now recognizes a '-s' command line switch, whose argument is the
8789 name of a file of Scheme code to load. It also treats the two
8790 characters `#!' as the start of a comment, terminated by `!#'. Thus,
8791 to make a file of Scheme code directly executable by Unix, insert the
8792 following two lines at the top of the file:
8794 #!/usr/local/bin/guile -s
8797 Guile treats the argument of the `-s' command-line switch as the name
8798 of a file of Scheme code to load, and treats the sequence `#!' as the
8799 start of a block comment, terminated by `!#'.
8801 For example, here's a version of 'echo' written in Scheme:
8803 #!/usr/local/bin/guile -s
8805 (let loop ((args (cdr (program-arguments))))
8808 (display (car args))
8809 (if (pair? (cdr args))
8811 (loop (cdr args)))))
8814 Why does `#!' start a block comment terminated by `!#', instead of the
8815 end of the line? That is the notation SCSH uses, and although we
8816 don't yet support the other SCSH features that motivate that choice,
8817 we would like to be backward-compatible with any existing Guile
8818 scripts once we do. Furthermore, if the path to Guile on your system
8819 is too long for your kernel, you can start the script with this
8823 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
8826 Note that some very old Unix systems don't support the `#!' syntax.
8829 ** You can now run Guile without installing it.
8831 Previous versions of the interactive Guile interpreter (`guile')
8832 couldn't start up unless Guile's Scheme library had been installed;
8833 they used the value of the environment variable `SCHEME_LOAD_PATH'
8834 later on in the startup process, but not to find the startup code
8835 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
8838 To run Guile without installing it, build it in the normal way, and
8839 then set the environment variable `SCHEME_LOAD_PATH' to a
8840 colon-separated list of directories, including the top-level directory
8841 of the Guile sources. For example, if you unpacked Guile so that the
8842 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
8845 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
8848 ** Guile's read-eval-print loop no longer prints #<unspecified>
8849 results. If the user wants to see this, she can evaluate the
8850 expression (assert-repl-print-unspecified #t), perhaps in her startup
8853 ** Guile no longer shows backtraces by default when an error occurs;
8854 however, it does display a message saying how to get one, and how to
8855 request that they be displayed by default. After an error, evaluate
8857 to see a backtrace, and
8858 (debug-enable 'backtrace)
8859 to see them by default.
8863 * Changes to Guile Scheme:
8865 ** Guile now distinguishes between #f and the empty list.
8867 This is for compatibility with the IEEE standard, the (possibly)
8868 upcoming Revised^5 Report on Scheme, and many extant Scheme
8871 Guile used to have #f and '() denote the same object, to make Scheme's
8872 type system more compatible with Emacs Lisp's. However, the change
8873 caused too much trouble for Scheme programmers, and we found another
8874 way to reconcile Emacs Lisp with Scheme that didn't require this.
8877 ** Guile's delq, delv, delete functions, and their destructive
8878 counterparts, delq!, delv!, and delete!, now remove all matching
8879 elements from the list, not just the first. This matches the behavior
8880 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
8881 functions which inspired them.
8883 I recognize that this change may break code in subtle ways, but it
8884 seems best to make the change before the FSF's first Guile release,
8888 ** The compiled-library-path function has been deleted from libguile.
8890 ** The facilities for loading Scheme source files have changed.
8892 *** The variable %load-path now tells Guile which directories to search
8893 for Scheme code. Its value is a list of strings, each of which names
8896 *** The variable %load-extensions now tells Guile which extensions to
8897 try appending to a filename when searching the load path. Its value
8898 is a list of strings. Its default value is ("" ".scm").
8900 *** (%search-load-path FILENAME) searches the directories listed in the
8901 value of the %load-path variable for a Scheme file named FILENAME,
8902 with all the extensions listed in %load-extensions. If it finds a
8903 match, then it returns its full filename. If FILENAME is absolute, it
8904 returns it unchanged. Otherwise, it returns #f.
8906 %search-load-path will not return matches that refer to directories.
8908 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
8909 uses %seach-load-path to find a file named FILENAME, and loads it if
8910 it finds it. If it can't read FILENAME for any reason, it throws an
8913 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
8916 *** load uses the same searching semantics as primitive-load.
8918 *** The functions %try-load, try-load-with-path, %load, load-with-path,
8919 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
8920 path, and load-module-with-path have been deleted. The functions
8921 above should serve their purposes.
8923 *** If the value of the variable %load-hook is a procedure,
8924 `primitive-load' applies its value to the name of the file being
8925 loaded (without the load path directory name prepended). If its value
8926 is #f, it is ignored. Otherwise, an error occurs.
8928 This is mostly useful for printing load notification messages.
8931 ** The function `eval!' is no longer accessible from the scheme level.
8932 We can't allow operations which introduce glocs into the scheme level,
8933 because Guile's type system can't handle these as data. Use `eval' or
8934 `read-and-eval!' (see below) as replacement.
8936 ** The new function read-and-eval! reads an expression from PORT,
8937 evaluates it, and returns the result. This is more efficient than
8938 simply calling `read' and `eval', since it is not necessary to make a
8939 copy of the expression for the evaluator to munge.
8941 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
8942 for the `read' function.
8945 ** The function `int?' has been removed; its definition was identical
8946 to that of `integer?'.
8948 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
8949 use the R4RS names for these functions.
8951 ** The function object-properties no longer returns the hash handle;
8952 it simply returns the object's property list.
8954 ** Many functions have been changed to throw errors, instead of
8955 returning #f on failure. The point of providing exception handling in
8956 the language is to simplify the logic of user code, but this is less
8957 useful if Guile's primitives don't throw exceptions.
8959 ** The function `fileno' has been renamed from `%fileno'.
8961 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
8964 * Changes to Guile's C interface:
8966 ** The library's initialization procedure has been simplified.
8967 scm_boot_guile now has the prototype:
8969 void scm_boot_guile (int ARGC,
8971 void (*main_func) (),
8974 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
8975 MAIN_FUNC should do all the work of the program (initializing other
8976 packages, reading user input, etc.) before returning. When MAIN_FUNC
8977 returns, call exit (0); this function never returns. If you want some
8978 other exit value, MAIN_FUNC may call exit itself.
8980 scm_boot_guile arranges for program-arguments to return the strings
8981 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
8982 scm_set_program_arguments with the final list, so Scheme code will
8983 know which arguments have been processed.
8985 scm_boot_guile establishes a catch-all catch handler which prints an
8986 error message and exits the process. This means that Guile exits in a
8987 coherent way when system errors occur and the user isn't prepared to
8988 handle it. If the user doesn't like this behavior, they can establish
8989 their own universal catcher in MAIN_FUNC to shadow this one.
8991 Why must the caller do all the real work from MAIN_FUNC? The garbage
8992 collector assumes that all local variables of type SCM will be above
8993 scm_boot_guile's stack frame on the stack. If you try to manipulate
8994 SCM values after this function returns, it's the luck of the draw
8995 whether the GC will be able to find the objects you allocate. So,
8996 scm_boot_guile function exits, rather than returning, to discourage
8997 people from making that mistake.
8999 The IN, OUT, and ERR arguments were removed; there are other
9000 convenient ways to override these when desired.
9002 The RESULT argument was deleted; this function should never return.
9004 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
9008 ** Guile's header files should no longer conflict with your system's
9011 In order to compile code which #included <libguile.h>, previous
9012 versions of Guile required you to add a directory containing all the
9013 Guile header files to your #include path. This was a problem, since
9014 Guile's header files have names which conflict with many systems'
9017 Now only <libguile.h> need appear in your #include path; you must
9018 refer to all Guile's other header files as <libguile/mumble.h>.
9019 Guile's installation procedure puts libguile.h in $(includedir), and
9020 the rest in $(includedir)/libguile.
9023 ** Two new C functions, scm_protect_object and scm_unprotect_object,
9024 have been added to the Guile library.
9026 scm_protect_object (OBJ) protects OBJ from the garbage collector.
9027 OBJ will not be freed, even if all other references are dropped,
9028 until someone does scm_unprotect_object (OBJ). Both functions
9031 Note that calls to scm_protect_object do not nest. You can call
9032 scm_protect_object any number of times on a given object, and the
9033 next call to scm_unprotect_object will unprotect it completely.
9035 Basically, scm_protect_object and scm_unprotect_object just
9036 maintain a list of references to things. Since the GC knows about
9037 this list, all objects it mentions stay alive. scm_protect_object
9038 adds its argument to the list; scm_unprotect_object remove its
9039 argument from the list.
9042 ** scm_eval_0str now returns the value of the last expression
9045 ** The new function scm_read_0str reads an s-expression from a
9046 null-terminated string, and returns it.
9048 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
9049 to a Scheme port object.
9051 ** The new function `scm_set_program_arguments' allows C code to set
9052 the value returned by the Scheme `program-arguments' function.
9057 * Guile no longer includes sophisticated Tcl/Tk support.
9059 The old Tcl/Tk support was unsatisfying to us, because it required the
9060 user to link against the Tcl library, as well as Tk and Guile. The
9061 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
9062 referring to widgets by names, rather than exporting widgets to Scheme
9063 code as a special datatype.
9065 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
9066 maintainers described some very interesting changes in progress to the
9067 Tcl/Tk internals, which would facilitate clean interfaces between lone
9068 Tk and other interpreters --- even for garbage-collected languages
9069 like Scheme. They expected the new Tk to be publicly available in the
9072 Since it seems that Guile might soon have a new, cleaner interface to
9073 lone Tk, and that the old Guile/Tk glue code would probably need to be
9074 completely rewritten, we (Jim Blandy and Richard Stallman) have
9075 decided not to support the old code. We'll spend the time instead on
9076 a good interface to the newer Tk, as soon as it is available.
9078 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
9081 Copyright information:
9083 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
9085 Permission is granted to anyone to make or distribute verbatim copies
9086 of this document as received, in any medium, provided that the
9087 copyright notice and this permission notice are preserved,
9088 thus giving the recipient permission to redistribute in turn.
9090 Permission is granted to distribute modified versions
9091 of this document, or of portions of it,
9092 under the above conditions, provided also that they
9093 carry prominent notices stating who last changed them.
9098 paragraph-separate: "[
\f]*$"