Fix sloppy bound checking in `string-{ref,set!}' with the empty string.
[bpt/guile.git] / NEWS
1 Guile NEWS --- history of user-visible changes.
2 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
3 See the end for copying conditions.
4
5 Please send Guile bug reports to bug-guile@gnu.org. Note that you
6 must be subscribed to this list first, in order to successfully send a
7 report to it.
8
9 \f
10 Changes in 1.9.0:
11
12 * New modules (see the manual for details)
13
14 ** `(srfi srfi-18)', multithreading support
15 ** The `(ice-9 i18n)' module provides internationalization support
16
17 * Changes to the distribution
18
19 ** Guile now uses Gnulib as a portability aid
20
21 * Changes to the stand-alone interpreter
22 * Changes to Scheme functions and syntax
23
24 ** A new 'memoize-symbol evaluator trap has been added. This trap can
25 be used for efficiently implementing a Scheme code coverage.
26
27 ** Duplicate bindings among used modules are resolved lazily.
28 This slightly improves program startup times.
29
30 ** New thread cancellation and thread cleanup API
31 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
32
33 * Changes to the C interface
34
35 ** The GH interface (deprecated in version 1.6, 2001) was removed.
36
37 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
38
39 This makes these internal functions technically not callable from
40 application code.
41
42 ** Functions for handling `scm_option' now no longer require an argument
43 indicating length of the `scm_t_option' array.
44
45
46 \f
47 Changes in 1.8.6 (since 1.8.5)
48
49 * New features (see the manual for details)
50
51 ** New convenience function `scm_c_symbol_length ()'
52
53 ** Single stepping through code from Emacs
54
55 When you use GDS to evaluate Scheme code from Emacs, you can now use
56 `C-u' to indicate that you want to single step through that code. See
57 `Evaluating Scheme Code' in the manual for more details.
58
59 ** New "guile(1)" man page!
60
61 * Changes to the distribution
62
63 ** Automake's `AM_MAINTAINER_MODE' is no longer used
64
65 Thus, the `--enable-maintainer-mode' configure option is no longer
66 available: Guile is now always configured in "maintainer mode".
67
68 * Bugs fixed
69
70 ** `symbol->string' now returns a read-only string, as per R5RS
71 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
72 ** `guile-config link' now prints `-L$libdir' before `-lguile'
73 ** Fix memory corruption involving GOOPS' `class-redefinition'
74 ** Fix possible deadlock in `mutex-lock'
75 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
76 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
77 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
78 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
79 ** Fix misleading output from `(help rationalize)'
80 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
81 ** Fix `gcd' when called with a single, negative argument.
82 ** Fix `Stack overflow' errors seen when building on some platforms
83 ** Fix bug when `scm_with_guile ()' was called several times from the
84 same thread
85 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
86 dynamic environment of the call to `raise'
87 ** Fix potential deadlock in `make-struct'
88 ** Fix compilation problem with libltdl from Libtool 2.2.x
89 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
90
91 \f
92 Changes in 1.8.5 (since 1.8.4)
93
94 * Infrastructure changes
95
96 ** Guile repository switched from CVS to Git
97
98 The new repository can be accessed using
99 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
100 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
101
102 ** Add support for `pkg-config'
103
104 See "Autoconf Support" in the manual for details.
105
106 * New modules (see the manual for details)
107
108 ** `(srfi srfi-88)'
109
110 * New features (see the manual for details)
111
112 ** New `postfix' read option, for SRFI-88 keyword syntax
113 ** Some I/O primitives have been inlined, which improves I/O performance
114 ** New object-based traps infrastructure
115
116 This is a GOOPS-based infrastructure that builds on Guile's low-level
117 evaluator trap calls and facilitates the development of debugging
118 features like single-stepping, breakpoints, tracing and profiling.
119 See the `Traps' node of the manual for details.
120
121 ** New support for working on Guile code from within Emacs
122
123 Guile now incorporates the `GDS' library (previously distributed
124 separately) for working on Guile code from within Emacs. See the
125 `Using Guile In Emacs' node of the manual for details.
126
127 * Bugs fixed
128
129 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
130 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
131
132 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
133 would trigger an unbound variable error for `match:andmap'.
134
135 ** `(oop goops describe)' now properly provides the `describe' feature
136 ** Fixed `args-fold' from `(srfi srfi-37)'
137
138 Previously, parsing short option names of argument-less options would
139 lead to a stack overflow.
140
141 ** `(srfi srfi-35)' is now visible through `cond-expand'
142 ** Fixed type-checking for the second argument of `eval'
143 ** Fixed type-checking for SRFI-1 `partition'
144 ** Fixed `struct-ref' and `struct-set!' on "light structs"
145 ** Honor struct field access rights in GOOPS
146 ** Changed the storage strategy of source properties, which fixes a deadlock
147 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
148 ** Fixed build issue for GNU/Linux on IA64
149 ** Fixed build issues on NetBSD 1.6
150 ** Fixed build issue on Solaris 2.10 x86_64
151 ** Fixed build issue with DEC/Compaq/HP's compiler
152 ** Fixed `scm_from_complex_double' build issue on FreeBSD
153 ** Fixed `alloca' build issue on FreeBSD 6
154 ** Removed use of non-portable makefile constructs
155 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
156 ** Make sure all tests honor `$TMPDIR'
157
158 * Changes to the distribution
159
160 ** New FAQ
161
162 We've started collecting Frequently Asked Questions (FAQ), and will
163 distribute these (with answers!) in future Guile releases.
164
165 \f
166 Changes in 1.8.4 (since 1.8.3)
167
168 * Bugs fixed
169
170 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
171 ** Fixed a segmentation fault which occurred when displaying the
172 backtrace of a stack with a promise object (made by `delay') in it.
173 ** Make `accept' leave guile mode while blocking
174 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
175 ** Fixed a build problem on AIX (use of func_data identifier)
176 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
177 called with an associator proc that returns neither a pair nor #f.
178 ** Secondary threads now always return a valid module for (current-module).
179 ** Avoid MacOS build problems caused by incorrect combination of "64"
180 system and library calls.
181 ** `guile-snarf' now honors `$TMPDIR'
182 ** `guile-config compile' now reports CPPFLAGS used at compile-time
183 ** Fixed build with Sun Studio (Solaris 9)
184 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
185 uniform vectors on AIX.
186 ** Fixed a deadlock that occurs upon GC with multiple threads.
187 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
188 ** Fixed autotool-derived build problems on AIX 6.1.
189 ** Fixed NetBSD/alpha support
190 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
191
192 * New modules (see the manual for details)
193
194 ** `(srfi srfi-69)'
195
196 * Documentation fixes and improvements
197
198 ** Removed premature breakpoint documentation
199
200 The features described are not available in the series of 1.8.x
201 releases, so the documentation was misleading and has been removed.
202
203 ** More about Guile's default *random-state* variable
204
205 ** GOOPS: more about how to use `next-method'
206
207 * Changes to the distribution
208
209 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
210
211 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
212 General Public License, and the few incorrect files have now been
213 fixed to agree with the rest of the Guile distribution.
214
215 ** Removed unnecessary extra copies of COPYING*
216
217 The distribution now contains a single COPYING.LESSER at its top level.
218
219 \f
220 Changes in 1.8.3 (since 1.8.2)
221
222 * New modules (see the manual for details)
223
224 ** `(srfi srfi-35)'
225 ** `(srfi srfi-37)'
226
227 * Bugs fixed
228
229 ** The `(ice-9 slib)' module now works as expected
230 ** Expressions like "(set! 'x #t)" no longer yield a crash
231 ** Warnings about duplicate bindings now go to stderr
232 ** A memory leak in `make-socket-address' was fixed
233 ** Alignment issues (e.g., on SPARC) in network routines were fixed
234 ** A threading issue that showed up at least on NetBSD was fixed
235 ** Build problems on Solaris and IRIX fixed
236
237 * Implementation improvements
238
239 ** The reader is now faster, which reduces startup time
240 ** Procedures returned by `record-accessor' and `record-modifier' are faster
241
242
243 \f
244 Changes in 1.8.2 (since 1.8.1):
245
246 * New procedures (see the manual for details)
247
248 ** set-program-arguments
249 ** make-vtable
250
251 * Bugs fixed
252
253 ** Fractions were not `equal?' if stored in unreduced form.
254 (A subtle problem, since printing a value reduced it, making it work.)
255 ** srfi-60 `copy-bit' failed on 64-bit systems
256 ** "guile --use-srfi" option at the REPL can replace core functions
257 (Programs run with that option were ok, but in the interactive REPL
258 the core bindings got priority, preventing SRFI replacements or
259 extensions.)
260 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
261 ** `kill' on mingw throws an error for a PID other than oneself
262 ** Procedure names are attached to procedure-with-setters
263 ** Array read syntax works with negative lower bound
264 ** `array-in-bounds?' fix if an array has different lower bounds on each index
265 ** `*' returns exact 0 for "(* inexact 0)"
266 This follows what it always did for "(* 0 inexact)".
267 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
268 ** SRFI-19: `date->julian-day' did not account for timezone offset
269 ** `ttyname' no longer crashes when passed a non-tty argument
270 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
271 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
272 ** GOOPS: Fixed a bug in `method-more-specific?'
273 ** Build problems on Solaris fixed
274 ** Build problems on HP-UX IA64 fixed
275 ** Build problems on MinGW fixed
276
277 \f
278 Changes in 1.8.1 (since 1.8.0):
279
280 * LFS functions are now used to access 64-bit files on 32-bit systems.
281
282 * New procedures (see the manual for details)
283
284 ** primitive-_exit - [Scheme] the-root-module
285 ** scm_primitive__exit - [C]
286 ** make-completion-function - [Scheme] (ice-9 readline)
287 ** scm_c_locale_stringn_to_number - [C]
288 ** scm_srfi1_append_reverse [C]
289 ** scm_srfi1_append_reverse_x [C]
290 ** scm_log - [C]
291 ** scm_log10 - [C]
292 ** scm_exp - [C]
293 ** scm_sqrt - [C]
294
295 * New `(ice-9 i18n)' module (see the manual for details)
296
297 * Bugs fixed
298
299 ** Build problems have been fixed on MacOS, SunOS, and QNX.
300
301 ** `strftime' fix sign of %z timezone offset.
302
303 ** A one-dimensional array can now be 'equal?' to a vector.
304
305 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
306
307 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
308
309 ** `record-accessor' and `record-modifier' now have strict type checks.
310
311 Record accessor and modifier procedures now throw an error if the
312 record type of the record they're given is not the type expected.
313 (Previously accessors returned #f and modifiers silently did nothing).
314
315 ** It is now OK to use both autoload and use-modules on a given module.
316
317 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
318
319 Previously there was no checking on primatives like make-vector that
320 accept "one or two" arguments. Now there is.
321
322 ** The srfi-1 assoc function now calls its equality predicate properly.
323
324 Previously srfi-1 assoc would call the equality predicate with the key
325 last. According to the SRFI, the key should be first.
326
327 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
328
329 ** The array-set! procedure no longer segfaults when given a bit vector.
330
331 ** Bugs in make-shared-array have been fixed.
332
333 ** string<? and friends now follow char<? etc order on 8-bit chars.
334
335 ** The format procedure now handles inf and nan values for ~f correctly.
336
337 ** exact->inexact should no longer overflow when given certain large fractions.
338
339 ** srfi-9 accessor and modifier procedures now have strict record type checks.
340
341 This matches the srfi-9 specification.
342
343 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
344
345 Previously the (ice-9 ftw) procedures would ignore any file that had
346 the same inode number as a file they had already seen, even if that
347 file was on a different device.
348
349 \f
350 Changes in 1.8.0 (changes since the 1.6.x series):
351
352 * Changes to the distribution
353
354 ** Guile is now licensed with the GNU Lesser General Public License.
355
356 ** The manual is now licensed with the GNU Free Documentation License.
357
358 ** Guile now requires GNU MP (http://swox.com/gmp).
359
360 Guile now uses the GNU MP library for arbitrary precision arithmetic.
361
362 ** Guile now has separate private and public configuration headers.
363
364 That is, things like HAVE_STRING_H no longer leak from Guile's
365 headers.
366
367 ** Guile now provides and uses an "effective" version number.
368
369 Guile now provides scm_effective_version and effective-version
370 functions which return the "effective" version number. This is just
371 the normal full version string without the final micro-version number,
372 so the current effective-version is "1.8". The effective version
373 should remain unchanged during a stable series, and should be used for
374 items like the versioned share directory name
375 i.e. /usr/share/guile/1.8.
376
377 Providing an unchanging version number during a stable release for
378 things like the versioned share directory can be particularly
379 important for Guile "add-on" packages, since it provides a directory
380 that they can install to that won't be changed out from under them
381 with each micro release during a stable series.
382
383 ** Thread implementation has changed.
384
385 When you configure "--with-threads=null", you will get the usual
386 threading API (call-with-new-thread, make-mutex, etc), but you can't
387 actually create new threads. Also, "--with-threads=no" is now
388 equivalent to "--with-threads=null". This means that the thread API
389 is always present, although you might not be able to create new
390 threads.
391
392 When you configure "--with-threads=pthreads" or "--with-threads=yes",
393 you will get threads that are implemented with the portable POSIX
394 threads. These threads can run concurrently (unlike the previous
395 "coop" thread implementation), but need to cooperate for things like
396 the GC.
397
398 The default is "pthreads", unless your platform doesn't have pthreads,
399 in which case "null" threads are used.
400
401 See the manual for details, nodes "Initialization", "Multi-Threading",
402 "Blocking", and others.
403
404 ** There is the new notion of 'discouraged' features.
405
406 This is a milder form of deprecation.
407
408 Things that are discouraged should not be used in new code, but it is
409 OK to leave them in old code for now. When a discouraged feature is
410 used, no warning message is printed like there is for 'deprecated'
411 features. Also, things that are merely discouraged are nevertheless
412 implemented efficiently, while deprecated features can be very slow.
413
414 You can omit discouraged features from libguile by configuring it with
415 the '--disable-discouraged' option.
416
417 ** Deprecation warnings can be controlled at run-time.
418
419 (debug-enable 'warn-deprecated) switches them on and (debug-disable
420 'warn-deprecated) switches them off.
421
422 ** Support for SRFI 61, extended cond syntax for multiple values has
423 been added.
424
425 This SRFI is always available.
426
427 ** Support for require-extension, SRFI-55, has been added.
428
429 The SRFI-55 special form `require-extension' has been added. It is
430 available at startup, and provides a portable way to load Scheme
431 extensions. SRFI-55 only requires support for one type of extension,
432 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
433 13 14)).
434
435 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
436
437 The (srfi srfi-26) module is an implementation of SRFI-26 which
438 provides the `cut' and `cute' syntax. These may be used to specialize
439 parameters without currying.
440
441 ** New module (srfi srfi-31)
442
443 This is an implementation of SRFI-31 which provides a special form
444 `rec' for recursive evaluation.
445
446 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
447 been merged with the core, making their functionality always
448 available.
449
450 The modules are still available, tho, and you could use them together
451 with a renaming import, for example.
452
453 ** Guile no longer includes its own version of libltdl.
454
455 The official version is good enough now.
456
457 ** The --enable-htmldoc option has been removed from 'configure'.
458
459 Support for translating the documentation into HTML is now always
460 provided. Use 'make html'.
461
462 ** New module (ice-9 serialize):
463
464 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
465 don't trust the thread safety of most of your program, but where you
466 have some section(s) of code which you consider can run in parallel to
467 other sections. See ice-9/serialize.scm for more information.
468
469 ** The configure option '--disable-arrays' has been removed.
470
471 Support for arrays and uniform numeric arrays is now always included
472 in Guile.
473
474 * Changes to the stand-alone interpreter
475
476 ** New command line option `-L'.
477
478 This option adds a directory to the front of the load path.
479
480 ** New command line option `--no-debug'.
481
482 Specifying `--no-debug' on the command line will keep the debugging
483 evaluator turned off, even for interactive sessions.
484
485 ** User-init file ~/.guile is now loaded with the debugging evaluator.
486
487 Previously, the normal evaluator would have been used. Using the
488 debugging evaluator gives better error messages.
489
490 ** The '-e' option now 'read's its argument.
491
492 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
493 be used with '-e'. For example, you can now write a script like
494
495 #! /bin/sh
496 exec guile -e '(@ (demo) main)' -s "$0" "$@"
497 !#
498
499 (define-module (demo)
500 :export (main))
501
502 (define (main args)
503 (format #t "Demo: ~a~%" args))
504
505
506 * Changes to Scheme functions and syntax
507
508 ** Guardians have changed back to their original semantics
509
510 Guardians now behave like described in the paper by Dybvig et al. In
511 particular, they no longer make guarantees about the order in which
512 they return objects, and they can no longer be greedy.
513
514 They no longer drop cyclic data structures.
515
516 The C function scm_make_guardian has been changed incompatibly and no
517 longer takes the 'greedy_p' argument.
518
519 ** New function hashx-remove!
520
521 This function completes the set of 'hashx' functions.
522
523 ** The concept of dynamic roots has been factored into continuation
524 barriers and dynamic states.
525
526 Each thread has a current dynamic state that carries the values of the
527 fluids. You can create and copy dynamic states and use them as the
528 second argument for 'eval'. See "Fluids and Dynamic States" in the
529 manual.
530
531 To restrict the influence that captured continuations can have on the
532 control flow, you can errect continuation barriers. See "Continuation
533 Barriers" in the manual.
534
535 The function call-with-dynamic-root now essentially temporarily
536 installs a new dynamic state and errects a continuation barrier.
537
538 ** The default load path no longer includes "." at the end.
539
540 Automatically loading modules from the current directory should not
541 happen by default. If you want to allow it in a more controlled
542 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
543 variable %load-path.
544
545 ** The uniform vector and array support has been overhauled.
546
547 It now complies with SRFI-4 and the weird prototype based uniform
548 array creation has been deprecated. See the manual for more details.
549
550 Some non-compatible changes have been made:
551 - characters can no longer be stored into byte arrays.
552 - strings and bit vectors are no longer considered to be uniform numeric
553 vectors.
554 - array-rank throws an error for non-arrays instead of returning zero.
555 - array-ref does no longer accept non-arrays when no indices are given.
556
557 There is the new notion of 'generalized vectors' and corresponding
558 procedures like 'generalized-vector-ref'. Generalized vectors include
559 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
560
561 Arrays use generalized vectors as their storage, so that you still
562 have arrays of characters, bits, etc. However, uniform-array-read!
563 and uniform-array-write can no longer read/write strings and
564 bitvectors.
565
566 ** There is now support for copy-on-write substrings, mutation-sharing
567 substrings and read-only strings.
568
569 Three new procedures are related to this: substring/shared,
570 substring/copy, and substring/read-only. See the manual for more
571 information.
572
573 ** Backtraces will now highlight the value that caused the error.
574
575 By default, these values are enclosed in "{...}", such as in this
576 example:
577
578 guile> (car 'a)
579
580 Backtrace:
581 In current input:
582 1: 0* [car {a}]
583
584 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
585 <unnamed port>:1:1: Wrong type (expecting pair): a
586 ABORT: (wrong-type-arg)
587
588 The prefix and suffix used for highlighting can be set via the two new
589 printer options 'highlight-prefix' and 'highlight-suffix'. For
590 example, putting this into ~/.guile will output the bad value in bold
591 on an ANSI terminal:
592
593 (print-set! highlight-prefix "\x1b[1m")
594 (print-set! highlight-suffix "\x1b[22m")
595
596
597 ** 'gettext' support for internationalization has been added.
598
599 See the manual for details.
600
601 ** New syntax '@' and '@@':
602
603 You can now directly refer to variables exported from a module by
604 writing
605
606 (@ MODULE-NAME VARIABLE-NAME)
607
608 For example (@ (ice-9 pretty-print) pretty-print) will directly access
609 the pretty-print variable exported from the (ice-9 pretty-print)
610 module. You don't need to 'use' that module first. You can also use
611 '@' as a target of 'set!', as in (set! (@ mod var) val).
612
613 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
614 but it can also access variables that have not been exported. It is
615 intended only for kluges and temporary fixes and for debugging, not
616 for ordinary code.
617
618 ** Keyword syntax has been made more disciplined.
619
620 Previously, the name of a keyword was read as a 'token' but printed as
621 a symbol. Now, it is read as a general Scheme datum which must be a
622 symbol.
623
624 Previously:
625
626 guile> #:12
627 #:#{12}#
628 guile> #:#{12}#
629 #:#{\#{12}\#}#
630 guile> #:(a b c)
631 #:#{}#
632 ERROR: In expression (a b c):
633 Unbound variable: a
634 guile> #: foo
635 #:#{}#
636 ERROR: Unbound variable: foo
637
638 Now:
639
640 guile> #:12
641 ERROR: Wrong type (expecting symbol): 12
642 guile> #:#{12}#
643 #:#{12}#
644 guile> #:(a b c)
645 ERROR: Wrong type (expecting symbol): (a b c)
646 guile> #: foo
647 #:foo
648
649 ** The printing of symbols that might look like keywords can be
650 controlled.
651
652 The new printer option 'quote-keywordish-symbols' controls how symbols
653 are printed that have a colon as their first or last character. The
654 default now is to only quote a symbol with #{...}# when the read
655 option 'keywords' is not '#f'. Thus:
656
657 guile> (define foo (string->symbol ":foo"))
658 guile> (read-set! keywords #f)
659 guile> foo
660 :foo
661 guile> (read-set! keywords 'prefix)
662 guile> foo
663 #{:foo}#
664 guile> (print-set! quote-keywordish-symbols #f)
665 guile> foo
666 :foo
667
668 ** 'while' now provides 'break' and 'continue'
669
670 break and continue were previously bound in a while loop, but not
671 documented, and continue didn't quite work properly. The undocumented
672 parameter to break which gave a return value for the while has been
673 dropped.
674
675 ** 'call-with-current-continuation' is now also available under the name
676 'call/cc'.
677
678 ** The module system now checks for duplicate bindings.
679
680 The module system now can check for name conflicts among imported
681 bindings.
682
683 The behavior can be controlled by specifying one or more 'duplicates'
684 handlers. For example, to make Guile return an error for every name
685 collision, write:
686
687 (define-module (foo)
688 :use-module (bar)
689 :use-module (baz)
690 :duplicates check)
691
692 The new default behavior of the module system when a name collision
693 has been detected is to
694
695 1. Give priority to bindings marked as a replacement.
696 2. Issue a warning (different warning if overriding core binding).
697 3. Give priority to the last encountered binding (this corresponds to
698 the old behavior).
699
700 If you want the old behavior back without replacements or warnings you
701 can add the line:
702
703 (default-duplicate-binding-handler 'last)
704
705 to your .guile init file.
706
707 ** New define-module option: :replace
708
709 :replace works as :export, but, in addition, marks the binding as a
710 replacement.
711
712 A typical example is `format' in (ice-9 format) which is a replacement
713 for the core binding `format'.
714
715 ** Adding prefixes to imported bindings in the module system
716
717 There is now a new :use-module option :prefix. It can be used to add
718 a prefix to all imported bindings.
719
720 (define-module (foo)
721 :use-module ((bar) :prefix bar:))
722
723 will import all bindings exported from bar, but rename them by adding
724 the prefix `bar:'.
725
726 ** Conflicting generic functions can be automatically merged.
727
728 When two imported bindings conflict and they are both generic
729 functions, the two functions can now be merged automatically. This is
730 activated with the 'duplicates' handler 'merge-generics'.
731
732 ** New function: effective-version
733
734 Returns the "effective" version number. This is just the normal full
735 version string without the final micro-version number. See "Changes
736 to the distribution" above.
737
738 ** New threading functions: parallel, letpar, par-map, and friends
739
740 These are convenient ways to run calculations in parallel in new
741 threads. See "Parallel forms" in the manual for details.
742
743 ** New function 'try-mutex'.
744
745 This function will attempt to lock a mutex but will return immediately
746 instead of blocking and indicate failure.
747
748 ** Waiting on a condition variable can have a timeout.
749
750 The function 'wait-condition-variable' now takes a third, optional
751 argument that specifies the point in time where the waiting should be
752 aborted.
753
754 ** New function 'broadcast-condition-variable'.
755
756 ** New functions 'all-threads' and 'current-thread'.
757
758 ** Signals and system asyncs work better with threads.
759
760 The function 'sigaction' now takes a fourth, optional, argument that
761 specifies the thread that the handler should run in. When the
762 argument is omitted, the handler will run in the thread that called
763 'sigaction'.
764
765 Likewise, 'system-async-mark' takes a second, optional, argument that
766 specifies the thread that the async should run in. When it is
767 omitted, the async will run in the thread that called
768 'system-async-mark'.
769
770 C code can use the new functions scm_sigaction_for_thread and
771 scm_system_async_mark_for_thread to pass the new thread argument.
772
773 When a thread blocks on a mutex, a condition variable or is waiting
774 for IO to be possible, it will still execute system asyncs. This can
775 be used to interrupt such a thread by making it execute a 'throw', for
776 example.
777
778 ** The function 'system-async' is deprecated.
779
780 You can now pass any zero-argument procedure to 'system-async-mark'.
781 The function 'system-async' will just return its argument unchanged
782 now.
783
784 ** New functions 'call-with-blocked-asyncs' and
785 'call-with-unblocked-asyncs'
786
787 The expression (call-with-blocked-asyncs PROC) will call PROC and will
788 block execution of system asyncs for the current thread by one level
789 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
790 procedure and will unblock the execution of system asyncs by one
791 level for the current thread.
792
793 Only system asyncs are affected by these functions.
794
795 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
796
797 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
798 instead. Those functions are easier to use correctly and can be
799 nested.
800
801 ** New function 'unsetenv'.
802
803 ** New macro 'define-syntax-public'.
804
805 It works like 'define-syntax' and also exports the defined macro (but
806 only on top-level).
807
808 ** There is support for Infinity and NaNs.
809
810 Following PLT Scheme, Guile can now work with infinite numbers, and
811 'not-a-numbers'.
812
813 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
814 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
815 "+nan.0"). These numbers are inexact and have no exact counterpart.
816
817 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
818 sign of the dividend. The infinities are integers, and they answer #t
819 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
820 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
821
822 For example
823
824 (/ 1 0.0)
825 => +inf.0
826
827 (/ 0 0.0)
828 => +nan.0
829
830 (/ 0)
831 ERROR: Numerical overflow
832
833 Two new predicates 'inf?' and 'nan?' can be used to test for the
834 special values.
835
836 ** Inexact zero can have a sign.
837
838 Guile can now distinguish between plus and minus inexact zero, if your
839 platform supports this, too. The two zeros are equal according to
840 '=', but not according to 'eqv?'. For example
841
842 (- 0.0)
843 => -0.0
844
845 (= 0.0 (- 0.0))
846 => #t
847
848 (eqv? 0.0 (- 0.0))
849 => #f
850
851 ** Guile now has exact rationals.
852
853 Guile can now represent fractions such as 1/3 exactly. Computing with
854 them is also done exactly, of course:
855
856 (* 1/3 3/2)
857 => 1/2
858
859 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
860 for exact arguments.
861
862 For example: (floor 2) now returns an exact 2 where in the past it
863 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
864
865 ** inexact->exact no longer returns only integers.
866
867 Without exact rationals, the closest exact number was always an
868 integer, but now inexact->exact returns the fraction that is exactly
869 equal to a floating point number. For example:
870
871 (inexact->exact 1.234)
872 => 694680242521899/562949953421312
873
874 When you want the old behavior, use 'round' explicitly:
875
876 (inexact->exact (round 1.234))
877 => 1
878
879 ** New function 'rationalize'.
880
881 This function finds a simple fraction that is close to a given real
882 number. For example (and compare with inexact->exact above):
883
884 (rationalize (inexact->exact 1.234) 1/2000)
885 => 58/47
886
887 Note that, as required by R5RS, rationalize returns only then an exact
888 result when both its arguments are exact.
889
890 ** 'odd?' and 'even?' work also for inexact integers.
891
892 Previously, (odd? 1.0) would signal an error since only exact integers
893 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
894 returns #f and (odd? 1.5) signals an error.
895
896 ** Guile now has uninterned symbols.
897
898 The new function 'make-symbol' will return an uninterned symbol. This
899 is a symbol that is unique and is guaranteed to remain unique.
900 However, uninterned symbols can not yet be read back in.
901
902 Use the new function 'symbol-interned?' to check whether a symbol is
903 interned or not.
904
905 ** pretty-print has more options.
906
907 The function pretty-print from the (ice-9 pretty-print) module can now
908 also be invoked with keyword arguments that control things like
909 maximum output width. See the manual for details.
910
911 ** Variables have no longer a special behavior for `equal?'.
912
913 Previously, comparing two variables with `equal?' would recursivly
914 compare their values. This is no longer done. Variables are now only
915 `equal?' if they are `eq?'.
916
917 ** `(begin)' is now valid.
918
919 You can now use an empty `begin' form. It will yield #<unspecified>
920 when evaluated and simply be ignored in a definition context.
921
922 ** Deprecated: procedure->macro
923
924 Change your code to use 'define-macro' or r5rs macros. Also, be aware
925 that macro expansion will not be done during evaluation, but prior to
926 evaluation.
927
928 ** Soft ports now allow a `char-ready?' procedure
929
930 The vector argument to `make-soft-port' can now have a length of
931 either 5 or 6. (Previously the length had to be 5.) The optional 6th
932 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
933 that returns the number of characters that can be read immediately
934 without the soft port blocking.
935
936 ** Deprecated: undefine
937
938 There is no replacement for undefine.
939
940 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
941 have been discouraged.
942
943 They are relics from a time where a keyword like #:foo was used
944 directly as a Tcl option "-foo" and thus keywords were internally
945 stored as a symbol with a starting dash. We now store a symbol
946 without the dash.
947
948 Use symbol->keyword and keyword->symbol instead.
949
950 ** The `cheap' debug option is now obsolete
951
952 Evaluator trap calls are now unconditionally "cheap" - in other words,
953 they pass a debug object to the trap handler rather than a full
954 continuation. The trap handler code can capture a full continuation
955 by using `call-with-current-continuation' in the usual way, if it so
956 desires.
957
958 The `cheap' option is retained for now so as not to break existing
959 code which gets or sets it, but setting it now has no effect. It will
960 be removed in the next major Guile release.
961
962 ** Evaluator trap calls now support `tweaking'
963
964 `Tweaking' means that the trap handler code can modify the Scheme
965 expression that is about to be evaluated (in the case of an
966 enter-frame trap) or the value that is being returned (in the case of
967 an exit-frame trap). The trap handler code indicates that it wants to
968 do this by returning a pair whose car is the symbol 'instead and whose
969 cdr is the modified expression or return value.
970
971 * Changes to the C interface
972
973 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
974 take a 'delete' function argument.
975
976 This argument makes no sense since the delete function is used to
977 remove a pair from an alist, and this must not be configurable.
978
979 This is an incompatible change.
980
981 ** The GH interface is now subject to the deprecation mechanism
982
983 The GH interface has been deprecated for quite some time but now it is
984 actually removed from Guile when it is configured with
985 --disable-deprecated.
986
987 See the manual "Transitioning away from GH" for more information.
988
989 ** A new family of functions for converting between C values and
990 Scheme values has been added.
991
992 These functions follow a common naming scheme and are designed to be
993 easier to use, thread-safe and more future-proof than the older
994 alternatives.
995
996 - int scm_is_* (...)
997
998 These are predicates that return a C boolean: 1 or 0. Instead of
999 SCM_NFALSEP, you can now use scm_is_true, for example.
1000
1001 - <type> scm_to_<type> (SCM val, ...)
1002
1003 These are functions that convert a Scheme value into an appropriate
1004 C value. For example, you can use scm_to_int to safely convert from
1005 a SCM to an int.
1006
1007 - SCM scm_from_<type> (<type> val, ...)
1008
1009 These functions convert from a C type to a SCM value; for example,
1010 scm_from_int for ints.
1011
1012 There is a huge number of these functions, for numbers, strings,
1013 symbols, vectors, etc. They are documented in the reference manual in
1014 the API section together with the types that they apply to.
1015
1016 ** New functions for dealing with complex numbers in C have been added.
1017
1018 The new functions are scm_c_make_rectangular, scm_c_make_polar,
1019 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
1020 They work like scm_make_rectangular etc but take or return doubles
1021 directly.
1022
1023 ** The function scm_make_complex has been discouraged.
1024
1025 Use scm_c_make_rectangular instead.
1026
1027 ** The INUM macros have been deprecated.
1028
1029 A lot of code uses these macros to do general integer conversions,
1030 although the macros only work correctly with fixnums. Use the
1031 following alternatives.
1032
1033 SCM_INUMP -> scm_is_integer or similar
1034 SCM_NINUMP -> !scm_is_integer or similar
1035 SCM_MAKINUM -> scm_from_int or similar
1036 SCM_INUM -> scm_to_int or similar
1037
1038 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
1039 do the validating for you.
1040
1041 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
1042 have been discouraged.
1043
1044 Use the newer scm_to_<type> and scm_from_<type> functions instead for
1045 new code. The functions have been discouraged since they don't fit
1046 the naming scheme.
1047
1048 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
1049
1050 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
1051 evaluates its argument twice. Use scm_is_true, etc. instead for new
1052 code.
1053
1054 ** The macro SCM_EQ_P has been discouraged.
1055
1056 Use scm_is_eq for new code, which fits better into the naming
1057 conventions.
1058
1059 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
1060 been discouraged.
1061
1062 Use the function scm_is_pair or scm_is_null instead.
1063
1064 ** The functions scm_round and scm_truncate have been deprecated and
1065 are now available as scm_c_round and scm_c_truncate, respectively.
1066
1067 These functions occupy the names that scm_round_number and
1068 scm_truncate_number should have.
1069
1070 ** The functions scm_c_string2str, scm_c_substring2str, and
1071 scm_c_symbol2str have been deprecated.
1072
1073 Use scm_to_locale_stringbuf or similar instead, maybe together with
1074 scm_substring.
1075
1076 ** New functions scm_c_make_string, scm_c_string_length,
1077 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
1078 scm_c_substring_shared, scm_c_substring_copy.
1079
1080 These are like scm_make_string, scm_length, etc. but are slightly
1081 easier to use from C.
1082
1083 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
1084 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
1085
1086 They export too many assumptions about the implementation of strings
1087 and symbols that are no longer true in the presence of
1088 mutation-sharing substrings and when Guile switches to some form of
1089 Unicode.
1090
1091 When working with strings, it is often best to use the normal string
1092 functions provided by Guile, such as scm_c_string_ref,
1093 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
1094 manual since many more such functions are now provided than
1095 previously.
1096
1097 When you want to convert a SCM string to a C string, use the
1098 scm_to_locale_string function or similar instead. For symbols, use
1099 scm_symbol_to_string and then work with that string. Because of the
1100 new string representation, scm_symbol_to_string does not need to copy
1101 and is thus quite efficient.
1102
1103 ** Some string, symbol and keyword functions have been discouraged.
1104
1105 They don't fit into the uniform naming scheme and are not explicit
1106 about the character encoding.
1107
1108 Replace according to the following table:
1109
1110 scm_allocate_string -> scm_c_make_string
1111 scm_take_str -> scm_take_locale_stringn
1112 scm_take0str -> scm_take_locale_string
1113 scm_mem2string -> scm_from_locale_stringn
1114 scm_str2string -> scm_from_locale_string
1115 scm_makfrom0str -> scm_from_locale_string
1116 scm_mem2symbol -> scm_from_locale_symboln
1117 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
1118 scm_str2symbol -> scm_from_locale_symbol
1119
1120 SCM_SYMBOL_HASH -> scm_hashq
1121 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
1122
1123 scm_c_make_keyword -> scm_from_locale_keyword
1124
1125 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
1126 now also available to C code.
1127
1128 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
1129
1130 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
1131 the latter returns the true name of the keyword, not the 'dash name',
1132 as SCM_KEYWORDSYM used to do.
1133
1134 ** A new way to access arrays in a thread-safe and efficient way has
1135 been added.
1136
1137 See the manual, node "Accessing Arrays From C".
1138
1139 ** The old uniform vector and bitvector implementations have been
1140 unceremoniously removed.
1141
1142 This implementation exposed the details of the tagging system of
1143 Guile. Use the new C API explained in the manual in node "Uniform
1144 Numeric Vectors" and "Bit Vectors", respectively.
1145
1146 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
1147 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
1148 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
1149 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
1150 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
1151 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
1152 SCM_BITVEC_CLR.
1153
1154 ** The macros dealing with vectors have been deprecated.
1155
1156 Use the new functions scm_is_vector, scm_vector_elements,
1157 scm_vector_writable_elements, etc, or scm_is_simple_vector,
1158 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
1159 manual for more details.
1160
1161 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
1162 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
1163
1164 The following macros have been removed: SCM_VECTOR_BASE,
1165 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
1166 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
1167
1168 ** Some C functions and macros related to arrays have been deprecated.
1169
1170 Migrate according to the following table:
1171
1172 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
1173 scm_make_ra -> scm_make_array
1174 scm_shap2ra -> scm_make_array
1175 scm_cvref -> scm_c_generalized_vector_ref
1176 scm_ra_set_contp -> do not use
1177 scm_aind -> scm_array_handle_pos
1178 scm_raprin1 -> scm_display or scm_write
1179
1180 SCM_ARRAYP -> scm_is_array
1181 SCM_ARRAY_NDIM -> scm_c_array_rank
1182 SCM_ARRAY_DIMS -> scm_array_handle_dims
1183 SCM_ARRAY_CONTP -> do not use
1184 SCM_ARRAY_MEM -> do not use
1185 SCM_ARRAY_V -> scm_array_handle_elements or similar
1186 SCM_ARRAY_BASE -> do not use
1187
1188 ** SCM_CELL_WORD_LOC has been deprecated.
1189
1190 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
1191 to a SCM, as opposed to a pointer to a scm_t_bits.
1192
1193 This was done to allow the correct use of pointers into the Scheme
1194 heap. Previously, the heap words were of type scm_t_bits and local
1195 variables and function arguments were of type SCM, making it
1196 non-standards-conformant to have a pointer that can point to both.
1197
1198 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
1199
1200 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
1201 second and third words of double smobs. Likewise for
1202 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
1203
1204 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
1205 used to get and set the 16 exra bits in the zeroth word of a smob.
1206
1207 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
1208 accesing the first immediate word of a smob as a SCM value, and there
1209 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
1210 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
1211
1212 ** New way to deal with non-local exits and re-entries.
1213
1214 There is a new set of functions that essentially do what
1215 scm_internal_dynamic_wind does, but in a way that is more convenient
1216 for C code in some situations. Here is a quick example of how to
1217 prevent a potential memory leak:
1218
1219 void
1220 foo ()
1221 {
1222 char *mem;
1223
1224 scm_dynwind_begin (0);
1225
1226 mem = scm_malloc (100);
1227 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
1228
1229 /* MEM would leak if BAR throws an error.
1230 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
1231 */
1232
1233 bar ();
1234
1235 scm_dynwind_end ();
1236
1237 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
1238 SCM_DYNWIND_END as well.
1239 */
1240 }
1241
1242 For full documentation, see the node "Dynamic Wind" in the manual.
1243
1244 ** New function scm_dynwind_free
1245
1246 This function calls 'free' on a given pointer when a dynwind context
1247 is left. Thus the call to scm_dynwind_unwind_handler above could be
1248 replaced with simply scm_dynwind_free (mem).
1249
1250 ** New functions scm_c_call_with_blocked_asyncs and
1251 scm_c_call_with_unblocked_asyncs
1252
1253 Like scm_call_with_blocked_asyncs etc. but for C functions.
1254
1255 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
1256
1257 In addition to scm_c_call_with_blocked_asyncs you can now also use
1258 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
1259 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
1260
1261 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
1262 SCM_REALLOW_INTS have been deprecated.
1263
1264 They do no longer fulfill their original role of blocking signal
1265 delivery. Depending on what you want to achieve, replace a pair of
1266 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
1267 mutex, blocks asyncs, or both. See node "Critical Sections" in the
1268 manual.
1269
1270 ** The value 'scm_mask_ints' is no longer writable.
1271
1272 Previously, you could set scm_mask_ints directly. This is no longer
1273 possible. Use scm_c_call_with_blocked_asyncs and
1274 scm_c_call_with_unblocked_asyncs instead.
1275
1276 ** New way to temporarily set the current input, output or error ports
1277
1278 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
1279 context' (see above). <foo> is one of "input", "output" or "error".
1280
1281 ** New way to temporarily set fluids
1282
1283 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
1284 above) to temporarily set the value of a fluid.
1285
1286 ** New types scm_t_intmax and scm_t_uintmax.
1287
1288 On platforms that have them, these types are identical to intmax_t and
1289 uintmax_t, respectively. On other platforms, they are identical to
1290 the largest integer types that Guile knows about.
1291
1292 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
1293
1294 You should not have used them.
1295
1296 ** Many public #defines with generic names have been made private.
1297
1298 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
1299 private or renamed with a more suitable public name.
1300
1301 ** The macro SCM_TYP16S has been deprecated.
1302
1303 This macro is not intended for public use.
1304
1305 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
1306
1307 Use scm_is_true (scm_inexact_p (...)) instead.
1308
1309 ** The macro SCM_SLOPPY_REALP has been deprecated.
1310
1311 Use scm_is_real instead.
1312
1313 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
1314
1315 Use scm_is_complex instead.
1316
1317 ** Some preprocessor defines have been deprecated.
1318
1319 These defines indicated whether a certain feature was present in Guile
1320 or not. Going forward, assume that the features are always present.
1321
1322 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
1323 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
1324
1325 The following macros have been removed completely: MEMOIZE_LOCALS,
1326 SCM_RECKLESS, SCM_CAUTIOUS.
1327
1328 ** The preprocessor define STACK_DIRECTION has been deprecated.
1329
1330 There should be no need to know about the stack direction for ordinary
1331 programs.
1332
1333 ** New function: scm_effective_version
1334
1335 Returns the "effective" version number. This is just the normal full
1336 version string without the final micro-version number. See "Changes
1337 to the distribution" above.
1338
1339 ** The function scm_call_with_new_thread has a new prototype.
1340
1341 Instead of taking a list with the thunk and handler, these two
1342 arguments are now passed directly:
1343
1344 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
1345
1346 This is an incompatible change.
1347
1348 ** New snarfer macro SCM_DEFINE_PUBLIC.
1349
1350 This is like SCM_DEFINE, but also calls scm_c_export for the defined
1351 function in the init section.
1352
1353 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
1354
1355 ** Garbage collector rewrite.
1356
1357 The garbage collector is cleaned up a lot, and now uses lazy
1358 sweeping. This is reflected in the output of (gc-stats); since cells
1359 are being freed when they are allocated, the cells-allocated field
1360 stays roughly constant.
1361
1362 For malloc related triggers, the behavior is changed. It uses the same
1363 heuristic as the cell-triggered collections. It may be tuned with the
1364 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
1365 for minimum yield of malloc related triggers. The default is 40.
1366 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
1367 default is 200 kb.
1368
1369 Debugging operations for the freelist have been deprecated, along with
1370 the C variables that control garbage collection. The environment
1371 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
1372 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
1373
1374 For understanding the memory usage of a GUILE program, the routine
1375 gc-live-object-stats returns an alist containing the number of live
1376 objects for every type.
1377
1378
1379 ** The function scm_definedp has been renamed to scm_defined_p
1380
1381 The name scm_definedp is deprecated.
1382
1383 ** The struct scm_cell type has been renamed to scm_t_cell
1384
1385 This is in accordance to Guile's naming scheme for types. Note that
1386 the name scm_cell is now used for a function that allocates and
1387 initializes a new cell (see below).
1388
1389 ** New functions for memory management
1390
1391 A new set of functions for memory management has been added since the
1392 old way (scm_must_malloc, scm_must_free, etc) was error prone and
1393 indeed, Guile itself contained some long standing bugs that could
1394 cause aborts in long running programs.
1395
1396 The new functions are more symmetrical and do not need cooperation
1397 from smob free routines, among other improvements.
1398
1399 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
1400 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
1401 scm_gc_free, scm_gc_register_collectable_memory, and
1402 scm_gc_unregister_collectable_memory. Refer to the manual for more
1403 details and for upgrading instructions.
1404
1405 The old functions for memory management have been deprecated. They
1406 are: scm_must_malloc, scm_must_realloc, scm_must_free,
1407 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
1408
1409 ** Declarations of exported features are marked with SCM_API.
1410
1411 Every declaration of a feature that belongs to the exported Guile API
1412 has been marked by adding the macro "SCM_API" to the start of the
1413 declaration. This macro can expand into different things, the most
1414 common of which is just "extern" for Unix platforms. On Win32, it can
1415 be used to control which symbols are exported from a DLL.
1416
1417 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
1418 will expand into "__declspec (dllimport) extern", which is needed for
1419 linking to the Guile DLL in Windows.
1420
1421 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
1422 SCM_SRFI4_IMPORT, for the corresponding libraries.
1423
1424 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
1425
1426 Use the new functions scm_cell and scm_double_cell instead. The old
1427 macros had problems because with them allocation and initialization
1428 was separated and the GC could sometimes observe half initialized
1429 cells. Only careful coding by the user of SCM_NEWCELL and
1430 SCM_NEWCELL2 could make this safe and efficient.
1431
1432 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
1433
1434 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
1435 instead.
1436
1437 ** SRCBRKP has been deprecated.
1438
1439 Use scm_c_source_property_breakpoint_p instead.
1440
1441 ** Deprecated: scm_makmacro
1442
1443 Change your code to use either scm_makmmacro or to define macros in
1444 Scheme, using 'define-macro'.
1445
1446 ** New function scm_c_port_for_each.
1447
1448 This function is like scm_port_for_each but takes a pointer to a C
1449 function as the callback instead of a SCM value.
1450
1451 ** The names scm_internal_select, scm_thread_sleep, and
1452 scm_thread_usleep have been discouraged.
1453
1454 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
1455
1456 ** The GC can no longer be blocked.
1457
1458 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
1459 The GC can now run (partially) concurrently with other code and thus
1460 blocking it is not well defined.
1461
1462 ** Many definitions have been removed that were previously deprecated.
1463
1464 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
1465 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
1466 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
1467 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
1468 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
1469 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
1470 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
1471 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
1472 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
1473 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
1474 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
1475 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
1476 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
1477 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
1478 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
1479 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
1480 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
1481 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
1482 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
1483 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
1484 scm_vector_set_length_x, scm_contregs, scm_debug_info,
1485 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
1486 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
1487 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
1488 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
1489 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
1490 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
1491 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
1492 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
1493 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
1494 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
1495
1496 * Changes to bundled modules
1497
1498 ** (ice-9 debug)
1499
1500 Using the (ice-9 debug) module no longer automatically switches Guile
1501 to use the debugging evaluator. If you want to switch to the
1502 debugging evaluator (which is needed for backtrace information if you
1503 hit an error), please add an explicit "(debug-enable 'debug)" to your
1504 code just after the code to use (ice-9 debug).
1505
1506 \f
1507 Changes since Guile 1.4:
1508
1509 * Changes to the distribution
1510
1511 ** A top-level TODO file is included.
1512
1513 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
1514
1515 Guile now always uses three numbers to represent the version,
1516 i.e. "1.6.5". The first number, 1, is the major version number, the
1517 second number, 6, is the minor version number, and the third number,
1518 5, is the micro version number. Changes in major version number
1519 indicate major changes in Guile.
1520
1521 Minor version numbers that are even denote stable releases, and odd
1522 minor version numbers denote development versions (which may be
1523 unstable). The micro version number indicates a minor sub-revision of
1524 a given MAJOR.MINOR release.
1525
1526 In keeping with the new scheme, (minor-version) and scm_minor_version
1527 no longer return everything but the major version number. They now
1528 just return the minor version number. Two new functions
1529 (micro-version) and scm_micro_version have been added to report the
1530 micro version number.
1531
1532 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
1533
1534 ** New preprocessor definitions are available for checking versions.
1535
1536 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
1537 SCM_MICRO_VERSION to the appropriate integer values.
1538
1539 ** Guile now actively warns about deprecated features.
1540
1541 The new configure option `--enable-deprecated=LEVEL' and the
1542 environment variable GUILE_WARN_DEPRECATED control this mechanism.
1543 See INSTALL and README for more information.
1544
1545 ** Guile is much more likely to work on 64-bit architectures.
1546
1547 Guile now compiles and passes "make check" with only two UNRESOLVED GC
1548 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
1549 for the use of a test machine, and thanks to Stefan Jahn for ia64
1550 patches.
1551
1552 ** New functions: setitimer and getitimer.
1553
1554 These implement a fairly direct interface to the libc functions of the
1555 same name.
1556
1557 ** The #. reader extension is now disabled by default.
1558
1559 For safety reasons, #. evaluation is disabled by default. To
1560 re-enable it, set the fluid read-eval? to #t. For example:
1561
1562 (fluid-set! read-eval? #t)
1563
1564 but make sure you realize the potential security risks involved. With
1565 read-eval? enabled, reading a data file from an untrusted source can
1566 be dangerous.
1567
1568 ** New SRFI modules have been added:
1569
1570 SRFI-0 `cond-expand' is now supported in Guile, without requiring
1571 using a module.
1572
1573 (srfi srfi-1) is a library containing many useful pair- and list-processing
1574 procedures.
1575
1576 (srfi srfi-2) exports and-let*.
1577
1578 (srfi srfi-4) implements homogeneous numeric vector datatypes.
1579
1580 (srfi srfi-6) is a dummy module for now, since guile already provides
1581 all of the srfi-6 procedures by default: open-input-string,
1582 open-output-string, get-output-string.
1583
1584 (srfi srfi-8) exports receive.
1585
1586 (srfi srfi-9) exports define-record-type.
1587
1588 (srfi srfi-10) exports define-reader-ctor and implements the reader
1589 extension #,().
1590
1591 (srfi srfi-11) exports let-values and let*-values.
1592
1593 (srfi srfi-13) implements the SRFI String Library.
1594
1595 (srfi srfi-14) implements the SRFI Character-Set Library.
1596
1597 (srfi srfi-17) implements setter and getter-with-setter and redefines
1598 some accessor procedures as procedures with getters. (such as car,
1599 cdr, vector-ref etc.)
1600
1601 (srfi srfi-19) implements the SRFI Time/Date Library.
1602
1603 ** New scripts / "executable modules"
1604
1605 Subdirectory "scripts" contains Scheme modules that are packaged to
1606 also be executable as scripts. At this time, these scripts are available:
1607
1608 display-commentary
1609 doc-snarf
1610 generate-autoload
1611 punify
1612 read-scheme-source
1613 use2dot
1614
1615 See README there for more info.
1616
1617 These scripts can be invoked from the shell with the new program
1618 "guile-tools", which keeps track of installation directory for you.
1619 For example:
1620
1621 $ guile-tools display-commentary srfi/*.scm
1622
1623 guile-tools is copied to the standard $bindir on "make install".
1624
1625 ** New module (ice-9 stack-catch):
1626
1627 stack-catch is like catch, but saves the current state of the stack in
1628 the fluid the-last-stack. This fluid can be useful when using the
1629 debugger and when re-throwing an error.
1630
1631 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
1632
1633 This has been done to prevent problems on lesser operating systems
1634 that can't tolerate `*'s in file names. The exported macro continues
1635 to be named `and-let*', of course.
1636
1637 On systems that support it, there is also a compatibility module named
1638 (ice-9 and-let*). It will go away in the next release.
1639
1640 ** New modules (oop goops) etc.:
1641
1642 (oop goops)
1643 (oop goops describe)
1644 (oop goops save)
1645 (oop goops active-slot)
1646 (oop goops composite-slot)
1647
1648 The Guile Object Oriented Programming System (GOOPS) has been
1649 integrated into Guile. For further information, consult the GOOPS
1650 manual and tutorial in the `doc' directory.
1651
1652 ** New module (ice-9 rdelim).
1653
1654 This exports the following procedures which were previously defined
1655 in the default environment:
1656
1657 read-line read-line! read-delimited read-delimited! %read-delimited!
1658 %read-line write-line
1659
1660 For backwards compatibility the definitions are still imported into the
1661 default environment in this version of Guile. However you should add:
1662
1663 (use-modules (ice-9 rdelim))
1664
1665 to any program which uses the definitions, since this may change in
1666 future.
1667
1668 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
1669 can be used for similar functionality.
1670
1671 ** New module (ice-9 rw)
1672
1673 This is a subset of the (scsh rw) module from guile-scsh. Currently
1674 it defines two procedures:
1675
1676 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
1677
1678 Read characters from a port or file descriptor into a string STR.
1679 A port must have an underlying file descriptor -- a so-called
1680 fport. This procedure is scsh-compatible and can efficiently read
1681 large strings.
1682
1683 *** New function: write-string/partial str [port_or_fdes [start [end]]]
1684
1685 Write characters from a string STR to a port or file descriptor.
1686 A port must have an underlying file descriptor -- a so-called
1687 fport. This procedure is mostly compatible and can efficiently
1688 write large strings.
1689
1690 ** New module (ice-9 match)
1691
1692 This module includes Andrew K. Wright's pattern matcher. See
1693 ice-9/match.scm for brief description or
1694
1695 http://www.star-lab.com/wright/code.html
1696
1697 for complete documentation.
1698
1699 ** New module (ice-9 buffered-input)
1700
1701 This module provides procedures to construct an input port from an
1702 underlying source of input that reads and returns its input in chunks.
1703 The underlying input source is a Scheme procedure, specified by the
1704 caller, which the port invokes whenever it needs more input.
1705
1706 This is useful when building an input port whose back end is Readline
1707 or a UI element such as the GtkEntry widget.
1708
1709 ** Documentation
1710
1711 The reference and tutorial documentation that was previously
1712 distributed separately, as `guile-doc', is now included in the core
1713 Guile distribution. The documentation consists of the following
1714 manuals.
1715
1716 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
1717 to using Guile.
1718
1719 - The Guile Reference Manual (guile.texi) contains (or is intended to
1720 contain) reference documentation on all aspects of Guile.
1721
1722 - The GOOPS Manual (goops.texi) contains both tutorial-style and
1723 reference documentation for using GOOPS, Guile's Object Oriented
1724 Programming System.
1725
1726 - The Revised^5 Report on the Algorithmic Language Scheme
1727 (r5rs.texi).
1728
1729 See the README file in the `doc' directory for more details.
1730
1731 ** There are a couple of examples in the examples/ directory now.
1732
1733 * Changes to the stand-alone interpreter
1734
1735 ** New command line option `--use-srfi'
1736
1737 Using this option, SRFI modules can be loaded on startup and be
1738 available right from the beginning. This makes programming portable
1739 Scheme programs easier.
1740
1741 The option `--use-srfi' expects a comma-separated list of numbers,
1742 each representing a SRFI number to be loaded into the interpreter
1743 before starting evaluating a script file or the REPL. Additionally,
1744 the feature identifier for the loaded SRFIs is recognized by
1745 `cond-expand' when using this option.
1746
1747 Example:
1748 $ guile --use-srfi=8,13
1749 guile> (receive (x z) (values 1 2) (+ 1 2))
1750 3
1751 guile> (string-pad "bla" 20)
1752 " bla"
1753
1754 ** Guile now always starts up in the `(guile-user)' module.
1755
1756 Previously, scripts executed via the `-s' option would run in the
1757 `(guile)' module and the repl would run in the `(guile-user)' module.
1758 Now every user action takes place in the `(guile-user)' module by
1759 default.
1760
1761 * Changes to Scheme functions and syntax
1762
1763 ** Character classifiers work for non-ASCII characters.
1764
1765 The predicates `char-alphabetic?', `char-numeric?',
1766 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
1767 no longer check whether their arguments are ASCII characters.
1768 Previously, a character would only be considered alphabetic when it
1769 was also ASCII, for example.
1770
1771 ** Previously deprecated Scheme functions have been removed:
1772
1773 tag - no replacement.
1774 fseek - replaced by seek.
1775 list* - replaced by cons*.
1776
1777 ** It's now possible to create modules with controlled environments
1778
1779 Example:
1780
1781 (use-modules (ice-9 safe))
1782 (define m (make-safe-module))
1783 ;;; m will now be a module containing only a safe subset of R5RS
1784 (eval '(+ 1 2) m) --> 3
1785 (eval 'load m) --> ERROR: Unbound variable: load
1786
1787 ** Evaluation of "()", the empty list, is now an error.
1788
1789 Previously, the expression "()" evaluated to the empty list. This has
1790 been changed to signal a "missing expression" error. The correct way
1791 to write the empty list as a literal constant is to use quote: "'()".
1792
1793 ** New concept of `Guile Extensions'.
1794
1795 A Guile Extension is just a ordinary shared library that can be linked
1796 at run-time. We found it advantageous to give this simple concept a
1797 dedicated name to distinguish the issues related to shared libraries
1798 from the issues related to the module system.
1799
1800 *** New function: load-extension
1801
1802 Executing (load-extension lib init) is mostly equivalent to
1803
1804 (dynamic-call init (dynamic-link lib))
1805
1806 except when scm_register_extension has been called previously.
1807 Whenever appropriate, you should use `load-extension' instead of
1808 dynamic-link and dynamic-call.
1809
1810 *** New C function: scm_c_register_extension
1811
1812 This function registers a initialization function for use by
1813 `load-extension'. Use it when you don't want specific extensions to
1814 be loaded as shared libraries (for example on platforms that don't
1815 support dynamic linking).
1816
1817 ** Auto-loading of compiled-code modules is deprecated.
1818
1819 Guile used to be able to automatically find and link a shared
1820 library to satisfy requests for a module. For example, the module
1821 `(foo bar)' could be implemented by placing a shared library named
1822 "foo/libbar.so" (or with a different extension) in a directory on the
1823 load path of Guile.
1824
1825 This has been found to be too tricky, and is no longer supported. The
1826 shared libraries are now called "extensions". You should now write a
1827 small Scheme file that calls `load-extension' to load the shared
1828 library and initialize it explicitly.
1829
1830 The shared libraries themselves should be installed in the usual
1831 places for shared libraries, with names like "libguile-foo-bar".
1832
1833 For example, place this into a file "foo/bar.scm"
1834
1835 (define-module (foo bar))
1836
1837 (load-extension "libguile-foo-bar" "foobar_init")
1838
1839 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
1840
1841 `eval' is now R5RS, that is it takes two arguments.
1842 The second argument is an environment specifier, i.e. either
1843
1844 (scheme-report-environment 5)
1845 (null-environment 5)
1846 (interaction-environment)
1847
1848 or
1849
1850 any module.
1851
1852 ** The module system has been made more disciplined.
1853
1854 The function `eval' will save and restore the current module around
1855 the evaluation of the specified expression. While this expression is
1856 evaluated, `(current-module)' will now return the right module, which
1857 is the module specified as the second argument to `eval'.
1858
1859 A consequence of this change is that `eval' is not particularly
1860 useful when you want allow the evaluated code to change what module is
1861 designated as the current module and have this change persist from one
1862 call to `eval' to the next. The read-eval-print-loop is an example
1863 where `eval' is now inadequate. To compensate, there is a new
1864 function `primitive-eval' that does not take a module specifier and
1865 that does not save/restore the current module. You should use this
1866 function together with `set-current-module', `current-module', etc
1867 when you want to have more control over the state that is carried from
1868 one eval to the next.
1869
1870 Additionally, it has been made sure that forms that are evaluated at
1871 the top level are always evaluated with respect to the current module.
1872 Previously, subforms of top-level forms such as `begin', `case',
1873 etc. did not respect changes to the current module although these
1874 subforms are at the top-level as well.
1875
1876 To prevent strange behavior, the forms `define-module',
1877 `use-modules', `use-syntax', and `export' have been restricted to only
1878 work on the top level. The forms `define-public' and
1879 `defmacro-public' only export the new binding on the top level. They
1880 behave just like `define' and `defmacro', respectively, when they are
1881 used in a lexical environment.
1882
1883 Also, `export' will no longer silently re-export bindings imported
1884 from a used module. It will emit a `deprecation' warning and will
1885 cease to perform any re-export in the next version. If you actually
1886 want to re-export bindings, use the new `re-export' in place of
1887 `export'. The new `re-export' will not make copies of variables when
1888 rexporting them, as `export' did wrongly.
1889
1890 ** Module system now allows selection and renaming of imported bindings
1891
1892 Previously, when using `use-modules' or the `#:use-module' clause in
1893 the `define-module' form, all the bindings (association of symbols to
1894 values) for imported modules were added to the "current module" on an
1895 as-is basis. This has been changed to allow finer control through two
1896 new facilities: selection and renaming.
1897
1898 You can now select which of the imported module's bindings are to be
1899 visible in the current module by using the `:select' clause. This
1900 clause also can be used to rename individual bindings. For example:
1901
1902 ;; import all bindings no questions asked
1903 (use-modules (ice-9 common-list))
1904
1905 ;; import four bindings, renaming two of them;
1906 ;; the current module sees: every some zonk-y zonk-n
1907 (use-modules ((ice-9 common-list)
1908 :select (every some
1909 (remove-if . zonk-y)
1910 (remove-if-not . zonk-n))))
1911
1912 You can also programmatically rename all selected bindings using the
1913 `:renamer' clause, which specifies a proc that takes a symbol and
1914 returns another symbol. Because it is common practice to use a prefix,
1915 we now provide the convenience procedure `symbol-prefix-proc'. For
1916 example:
1917
1918 ;; import four bindings, renaming two of them specifically,
1919 ;; and all four w/ prefix "CL:";
1920 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
1921 (use-modules ((ice-9 common-list)
1922 :select (every some
1923 (remove-if . zonk-y)
1924 (remove-if-not . zonk-n))
1925 :renamer (symbol-prefix-proc 'CL:)))
1926
1927 ;; import four bindings, renaming two of them specifically,
1928 ;; and all four by upcasing.
1929 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
1930 (define (upcase-symbol sym)
1931 (string->symbol (string-upcase (symbol->string sym))))
1932
1933 (use-modules ((ice-9 common-list)
1934 :select (every some
1935 (remove-if . zonk-y)
1936 (remove-if-not . zonk-n))
1937 :renamer upcase-symbol))
1938
1939 Note that programmatic renaming is done *after* individual renaming.
1940 Also, the above examples show `use-modules', but the same facilities are
1941 available for the `#:use-module' clause of `define-module'.
1942
1943 See manual for more info.
1944
1945 ** The semantics of guardians have changed.
1946
1947 The changes are for the most part compatible. An important criterion
1948 was to keep the typical usage of guardians as simple as before, but to
1949 make the semantics safer and (as a result) more useful.
1950
1951 *** All objects returned from guardians are now properly alive.
1952
1953 It is now guaranteed that any object referenced by an object returned
1954 from a guardian is alive. It's now impossible for a guardian to
1955 return a "contained" object before its "containing" object.
1956
1957 One incompatible (but probably not very important) change resulting
1958 from this is that it is no longer possible to guard objects that
1959 indirectly reference themselves (i.e. are parts of cycles). If you do
1960 so accidentally, you'll get a warning.
1961
1962 *** There are now two types of guardians: greedy and sharing.
1963
1964 If you call (make-guardian #t) or just (make-guardian), you'll get a
1965 greedy guardian, and for (make-guardian #f) a sharing guardian.
1966
1967 Greedy guardians are the default because they are more "defensive".
1968 You can only greedily guard an object once. If you guard an object
1969 more than once, once in a greedy guardian and the rest of times in
1970 sharing guardians, then it is guaranteed that the object won't be
1971 returned from sharing guardians as long as it is greedily guarded
1972 and/or alive.
1973
1974 Guardians returned by calls to `make-guardian' can now take one more
1975 optional parameter, which says whether to throw an error in case an
1976 attempt is made to greedily guard an object that is already greedily
1977 guarded. The default is true, i.e. throw an error. If the parameter
1978 is false, the guardian invocation returns #t if guarding was
1979 successful and #f if it wasn't.
1980
1981 Also, since greedy guarding is, in effect, a side-effecting operation
1982 on objects, a new function is introduced: `destroy-guardian!'.
1983 Invoking this function on a guardian renders it unoperative and, if
1984 the guardian is greedy, clears the "greedily guarded" property of the
1985 objects that were guarded by it, thus undoing the side effect.
1986
1987 Note that all this hair is hardly very important, since guardian
1988 objects are usually permanent.
1989
1990 ** Continuations created by call-with-current-continuation now accept
1991 any number of arguments, as required by R5RS.
1992
1993 ** New function `issue-deprecation-warning'
1994
1995 This function is used to display the deprecation messages that are
1996 controlled by GUILE_WARN_DEPRECATION as explained in the README.
1997
1998 (define (id x)
1999 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
2000 (identity x))
2001
2002 guile> (id 1)
2003 ;; `id' is deprecated. Use `identity' instead.
2004 1
2005 guile> (id 1)
2006 1
2007
2008 ** New syntax `begin-deprecated'
2009
2010 When deprecated features are included (as determined by the configure
2011 option --enable-deprecated), `begin-deprecated' is identical to
2012 `begin'. When deprecated features are excluded, it always evaluates
2013 to `#f', ignoring the body forms.
2014
2015 ** New function `make-object-property'
2016
2017 This function returns a new `procedure with setter' P that can be used
2018 to attach a property to objects. When calling P as
2019
2020 (set! (P obj) val)
2021
2022 where `obj' is any kind of object, it attaches `val' to `obj' in such
2023 a way that it can be retrieved by calling P as
2024
2025 (P obj)
2026
2027 This function will replace procedure properties, symbol properties and
2028 source properties eventually.
2029
2030 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
2031
2032 Instead of #&optional, #&key, etc you should now use #:optional,
2033 #:key, etc. Since #:optional is a keyword, you can write it as just
2034 :optional when (read-set! keywords 'prefix) is active.
2035
2036 The old reader syntax `#&' is still supported, but deprecated. It
2037 will be removed in the next release.
2038
2039 ** New define-module option: pure
2040
2041 Tells the module system not to include any bindings from the root
2042 module.
2043
2044 Example:
2045
2046 (define-module (totally-empty-module)
2047 :pure)
2048
2049 ** New define-module option: export NAME1 ...
2050
2051 Export names NAME1 ...
2052
2053 This option is required if you want to be able to export bindings from
2054 a module which doesn't import one of `define-public' or `export'.
2055
2056 Example:
2057
2058 (define-module (foo)
2059 :pure
2060 :use-module (ice-9 r5rs)
2061 :export (bar))
2062
2063 ;;; Note that we're pure R5RS below this point!
2064
2065 (define (bar)
2066 ...)
2067
2068 ** New function: object->string OBJ
2069
2070 Return a Scheme string obtained by printing a given object.
2071
2072 ** New function: port? X
2073
2074 Returns a boolean indicating whether X is a port. Equivalent to
2075 `(or (input-port? X) (output-port? X))'.
2076
2077 ** New function: file-port?
2078
2079 Determines whether a given object is a port that is related to a file.
2080
2081 ** New function: port-for-each proc
2082
2083 Apply PROC to each port in the Guile port table in turn. The return
2084 value is unspecified. More specifically, PROC is applied exactly once
2085 to every port that exists in the system at the time PORT-FOR-EACH is
2086 invoked. Changes to the port table while PORT-FOR-EACH is running
2087 have no effect as far as PORT-FOR-EACH is concerned.
2088
2089 ** New function: dup2 oldfd newfd
2090
2091 A simple wrapper for the `dup2' system call. Copies the file
2092 descriptor OLDFD to descriptor number NEWFD, replacing the
2093 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
2094 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
2095 to move away ports which are using NEWFD. The return value is
2096 unspecified.
2097
2098 ** New function: close-fdes fd
2099
2100 A simple wrapper for the `close' system call. Close file
2101 descriptor FD, which must be an integer. Unlike close (*note
2102 close: Ports and File Descriptors.), the file descriptor will be
2103 closed even if a port is using it. The return value is
2104 unspecified.
2105
2106 ** New function: crypt password salt
2107
2108 Encrypts `password' using the standard unix password encryption
2109 algorithm.
2110
2111 ** New function: chroot path
2112
2113 Change the root directory of the running process to `path'.
2114
2115 ** New functions: getlogin, cuserid
2116
2117 Return the login name or the user name of the current effective user
2118 id, respectively.
2119
2120 ** New functions: getpriority which who, setpriority which who prio
2121
2122 Get or set the priority of the running process.
2123
2124 ** New function: getpass prompt
2125
2126 Read a password from the terminal, first displaying `prompt' and
2127 disabling echoing.
2128
2129 ** New function: flock file operation
2130
2131 Set/remove an advisory shared or exclusive lock on `file'.
2132
2133 ** New functions: sethostname name, gethostname
2134
2135 Set or get the hostname of the machine the current process is running
2136 on.
2137
2138 ** New function: mkstemp! tmpl
2139
2140 mkstemp creates a new unique file in the file system and returns a
2141 new buffered port open for reading and writing to the file. TMPL
2142 is a string specifying where the file should be created: it must
2143 end with `XXXXXX' and will be changed in place to return the name
2144 of the temporary file.
2145
2146 ** New function: open-input-string string
2147
2148 Return an input string port which delivers the characters from
2149 `string'. This procedure, together with `open-output-string' and
2150 `get-output-string' implements SRFI-6.
2151
2152 ** New function: open-output-string
2153
2154 Return an output string port which collects all data written to it.
2155 The data can then be retrieved by `get-output-string'.
2156
2157 ** New function: get-output-string
2158
2159 Return the contents of an output string port.
2160
2161 ** New function: identity
2162
2163 Return the argument.
2164
2165 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
2166 are represented in Scheme as integers with normal host byte ordering.
2167
2168 ** New function: inet-pton family address
2169
2170 Convert a printable string network address into an integer. Note that
2171 unlike the C version of this function, the result is an integer with
2172 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2173 e.g.,
2174
2175 (inet-pton AF_INET "127.0.0.1") => 2130706433
2176 (inet-pton AF_INET6 "::1") => 1
2177
2178 ** New function: inet-ntop family address
2179
2180 Convert an integer network address into a printable string. Note that
2181 unlike the C version of this function, the input is an integer with
2182 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2183 e.g.,
2184
2185 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
2186 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
2187 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
2188
2189 ** Deprecated: id
2190
2191 Use `identity' instead.
2192
2193 ** Deprecated: -1+
2194
2195 Use `1-' instead.
2196
2197 ** Deprecated: return-it
2198
2199 Do without it.
2200
2201 ** Deprecated: string-character-length
2202
2203 Use `string-length' instead.
2204
2205 ** Deprecated: flags
2206
2207 Use `logior' instead.
2208
2209 ** Deprecated: close-all-ports-except.
2210
2211 This was intended for closing ports in a child process after a fork,
2212 but it has the undesirable side effect of flushing buffers.
2213 port-for-each is more flexible.
2214
2215 ** The (ice-9 popen) module now attempts to set up file descriptors in
2216 the child process from the current Scheme ports, instead of using the
2217 current values of file descriptors 0, 1, and 2 in the parent process.
2218
2219 ** Removed function: builtin-weak-bindings
2220
2221 There is no such concept as a weak binding any more.
2222
2223 ** Removed constants: bignum-radix, scm-line-incrementors
2224
2225 ** define-method: New syntax mandatory.
2226
2227 The new method syntax is now mandatory:
2228
2229 (define-method (NAME ARG-SPEC ...) BODY ...)
2230 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
2231
2232 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
2233 REST-ARG ::= ARG-NAME
2234
2235 If you have old code using the old syntax, import
2236 (oop goops old-define-method) before (oop goops) as in:
2237
2238 (use-modules (oop goops old-define-method) (oop goops))
2239
2240 ** Deprecated function: builtin-variable
2241 Removed function: builtin-bindings
2242
2243 There is no longer a distinction between builtin or other variables.
2244 Use module system operations for all variables.
2245
2246 ** Lazy-catch handlers are no longer allowed to return.
2247
2248 That is, a call to `throw', `error', etc is now guaranteed to not
2249 return.
2250
2251 ** Bugfixes for (ice-9 getopt-long)
2252
2253 This module is now tested using test-suite/tests/getopt-long.test.
2254 The following bugs have been fixed:
2255
2256 *** Parsing for options that are specified to have `optional' args now checks
2257 if the next element is an option instead of unconditionally taking it as the
2258 option arg.
2259
2260 *** An error is now thrown for `--opt=val' when the option description
2261 does not specify `(value #t)' or `(value optional)'. This condition used to
2262 be accepted w/o error, contrary to the documentation.
2263
2264 *** The error message for unrecognized options is now more informative.
2265 It used to be "not a record", an artifact of the implementation.
2266
2267 *** The error message for `--opt' terminating the arg list (no value), when
2268 `(value #t)' is specified, is now more informative. It used to be "not enough
2269 args".
2270
2271 *** "Clumped" single-char args now preserve trailing string, use it as arg.
2272 The expansion used to be like so:
2273
2274 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
2275
2276 Note that the "5d" is dropped. Now it is like so:
2277
2278 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
2279
2280 This enables single-char options to have adjoining arguments as long as their
2281 constituent characters are not potential single-char options.
2282
2283 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
2284
2285 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
2286 property `arglist', which can be retrieved by `arity'. The result is that
2287 `arity' can give more detailed information than before:
2288
2289 Before:
2290
2291 guile> (use-modules (ice-9 optargs))
2292 guile> (define* (foo #:optional a b c) a)
2293 guile> (arity foo)
2294 0 or more arguments in `lambda*:G0'.
2295
2296 After:
2297
2298 guile> (arity foo)
2299 3 optional arguments: `a', `b' and `c'.
2300 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
2301 guile> (arity bar)
2302 2 required arguments: `a' and `b', 2 keyword arguments: `c'
2303 and `d', other keywords allowed.
2304 guile> (define* (baz a b #:optional c #:rest r) a)
2305 guile> (arity baz)
2306 2 required arguments: `a' and `b', 1 optional argument: `c',
2307 the rest in `r'.
2308
2309 * Changes to the C interface
2310
2311 ** Types have been renamed from scm_*_t to scm_t_*.
2312
2313 This has been done for POSIX sake. It reserves identifiers ending
2314 with "_t". What a concept.
2315
2316 The old names are still available with status `deprecated'.
2317
2318 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
2319
2320 ** Deprecated features have been removed.
2321
2322 *** Macros removed
2323
2324 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
2325 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
2326
2327 *** C Functions removed
2328
2329 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
2330 scm_fseek - replaced by scm_seek.
2331 gc-thunk - replaced by after-gc-hook.
2332 gh_int2scmb - replaced by gh_bool2scm.
2333 scm_tc_dblr - replaced by scm_tc16_real.
2334 scm_tc_dblc - replaced by scm_tc16_complex.
2335 scm_list_star - replaced by scm_cons_star.
2336
2337 ** Deprecated: scm_makfromstr
2338
2339 Use scm_mem2string instead.
2340
2341 ** Deprecated: scm_make_shared_substring
2342
2343 Explicit shared substrings will disappear from Guile.
2344
2345 Instead, "normal" strings will be implemented using sharing
2346 internally, combined with a copy-on-write strategy.
2347
2348 ** Deprecated: scm_read_only_string_p
2349
2350 The concept of read-only strings will disappear in next release of
2351 Guile.
2352
2353 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
2354
2355 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
2356
2357 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
2358
2359 Call a procedure with the indicated number of arguments. See "Fly
2360 Evaluation" in the manual.
2361
2362 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
2363
2364 Call a procedure with the indicated number of arguments and a list of
2365 further arguments. See "Fly Evaluation" in the manual.
2366
2367 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
2368
2369 Create a list of the given number of elements. See "List
2370 Constructors" in the manual.
2371
2372 ** Renamed function: scm_listify has been replaced by scm_list_n.
2373
2374 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
2375 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
2376
2377 Use functions scm_list_N instead.
2378
2379 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
2380
2381 Used by an application to read arbitrary number of bytes from a port.
2382 Same semantics as libc read, except that scm_c_read only returns less
2383 than SIZE bytes if at end-of-file.
2384
2385 Warning: Doesn't update port line and column counts!
2386
2387 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
2388
2389 Used by an application to write arbitrary number of bytes to an SCM
2390 port. Similar semantics as libc write. However, unlike libc
2391 write, scm_c_write writes the requested number of bytes and has no
2392 return value.
2393
2394 Warning: Doesn't update port line and column counts!
2395
2396 ** New function: scm_init_guile ()
2397
2398 In contrast to scm_boot_guile, scm_init_guile will return normally
2399 after initializing Guile. It is not available on all systems, tho.
2400
2401 ** New functions: scm_str2symbol, scm_mem2symbol
2402
2403 The function scm_str2symbol takes a const char* pointing to a zero-terminated
2404 field of characters and creates a scheme symbol object from that C string.
2405 The function scm_mem2symbol takes a const char* and a number of characters and
2406 creates a symbol from the characters in that memory area.
2407
2408 ** New functions: scm_primitive_make_property
2409 scm_primitive_property_ref
2410 scm_primitive_property_set_x
2411 scm_primitive_property_del_x
2412
2413 These functions implement a new way to deal with object properties.
2414 See libguile/properties.c for their documentation.
2415
2416 ** New function: scm_done_free (long size)
2417
2418 This function is the inverse of scm_done_malloc. Use it to report the
2419 amount of smob memory you free. The previous method, which involved
2420 calling scm_done_malloc with negative argument, was somewhat
2421 unintuitive (and is still available, of course).
2422
2423 ** New function: scm_c_memq (SCM obj, SCM list)
2424
2425 This function provides a fast C level alternative for scm_memq for the case
2426 that the list parameter is known to be a proper list. The function is a
2427 replacement for scm_sloppy_memq, but is stricter in its requirements on its
2428 list input parameter, since for anything else but a proper list the function's
2429 behaviour is undefined - it may even crash or loop endlessly. Further, for
2430 the case that the object is not found in the list, scm_c_memq returns #f which
2431 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
2432
2433 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
2434 scm_remember_upto_here
2435
2436 These functions replace the function scm_remember.
2437
2438 ** Deprecated function: scm_remember
2439
2440 Use one of the new functions scm_remember_upto_here_1,
2441 scm_remember_upto_here_2 or scm_remember_upto_here instead.
2442
2443 ** New function: scm_allocate_string
2444
2445 This function replaces the function scm_makstr.
2446
2447 ** Deprecated function: scm_makstr
2448
2449 Use the new function scm_allocate_string instead.
2450
2451 ** New global variable scm_gc_running_p introduced.
2452
2453 Use this variable to find out if garbage collection is being executed. Up to
2454 now applications have used scm_gc_heap_lock to test if garbage collection was
2455 running, which also works because of the fact that up to know only the garbage
2456 collector has set this variable. But, this is an implementation detail that
2457 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
2458 of this variable is (and has been) not fully safe anyway.
2459
2460 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
2461
2462 Use these instead of SCM_LENGTH_MAX.
2463
2464 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
2465 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
2466 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
2467
2468 Use these instead of SCM_LENGTH.
2469
2470 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
2471 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
2472 SCM_SET_BITVECTOR_LENGTH
2473
2474 Use these instead of SCM_SETLENGTH
2475
2476 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
2477 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
2478 SCM_ARRAY_MEM
2479
2480 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
2481 SCM_VELTS.
2482
2483 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
2484 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
2485 SCM_SET_VECTOR_BASE
2486
2487 Use these instead of SCM_SETCHARS.
2488
2489 ** New macro: SCM_BITVECTOR_P
2490
2491 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
2492
2493 Use instead of SCM_COERCE_SUBSTR.
2494
2495 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
2496
2497 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
2498
2499 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
2500 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
2501 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
2502 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
2503 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
2504 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
2505 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
2506 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
2507 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
2508 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
2509 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
2510 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
2511 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
2512 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
2513 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
2514
2515 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
2516 Use scm_memory_error instead of SCM_NALLOC.
2517 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
2518 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
2519 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
2520 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
2521 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
2522 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
2523 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
2524 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
2525 Use SCM_STRINGP instead of SCM_RWSTRINGP.
2526 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
2527 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
2528 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
2529 Use a type specific setter macro instead of SCM_SETLENGTH.
2530 Use a type specific setter macro instead of SCM_SETCHARS.
2531 Use a type specific length macro instead of SCM_LENGTH_MAX.
2532 Use SCM_GCMARKP instead of SCM_GC8MARKP.
2533 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
2534 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
2535 Use SCM_TYP16 instead of SCM_GCTYP16.
2536 Use SCM_CDR instead of SCM_GCCDR.
2537 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
2538 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
2539 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
2540 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
2541 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
2542 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
2543 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
2544
2545 ** Removed function: scm_struct_init
2546
2547 ** Removed variable: scm_symhash_dim
2548
2549 ** Renamed function: scm_make_cont has been replaced by
2550 scm_make_continuation, which has a different interface.
2551
2552 ** Deprecated function: scm_call_catching_errors
2553
2554 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
2555
2556 ** Deprecated function: scm_strhash
2557
2558 Use scm_string_hash instead.
2559
2560 ** Deprecated function: scm_vector_set_length_x
2561
2562 Instead, create a fresh vector of the desired size and copy the contents.
2563
2564 ** scm_gensym has changed prototype
2565
2566 scm_gensym now only takes one argument.
2567
2568 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
2569 scm_tc7_lvector
2570
2571 There is now only a single symbol type scm_tc7_symbol.
2572 The tag scm_tc7_lvector was not used anyway.
2573
2574 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
2575
2576 Use scm_make_smob_type and scm_set_smob_XXX instead.
2577
2578 ** New function scm_set_smob_apply.
2579
2580 This can be used to set an apply function to a smob type.
2581
2582 ** Deprecated function: scm_strprint_obj
2583
2584 Use scm_object_to_string instead.
2585
2586 ** Deprecated function: scm_wta
2587
2588 Use scm_wrong_type_arg, or another appropriate error signalling function
2589 instead.
2590
2591 ** Explicit support for obarrays has been deprecated.
2592
2593 Use `scm_str2symbol' and the generic hashtable functions instead.
2594
2595 ** The concept of `vcells' has been deprecated.
2596
2597 The data type `variable' is now used exclusively. `Vcells' have been
2598 a low-level concept so you are likely not affected by this change.
2599
2600 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
2601 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
2602
2603 Use scm_c_define or scm_c_lookup instead, as appropriate.
2604
2605 *** New functions: scm_c_module_lookup, scm_c_lookup,
2606 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
2607 scm_module_define, scm_define.
2608
2609 These functions work with variables instead of with vcells.
2610
2611 ** New functions for creating and defining `subr's and `gsubr's.
2612
2613 The new functions more clearly distinguish between creating a subr (or
2614 gsubr) object and adding it to the current module.
2615
2616 These new functions are available: scm_c_make_subr, scm_c_define_subr,
2617 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
2618 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
2619 scm_c_define_gsubr_with_generic.
2620
2621 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
2622 scm_make_subr_with_generic, scm_make_gsubr,
2623 scm_make_gsubr_with_generic.
2624
2625 Use the new ones from above instead.
2626
2627 ** C interface to the module system has changed.
2628
2629 While we suggest that you avoid as many explicit module system
2630 operations from C as possible for the time being, the C interface has
2631 been made more similar to the high-level Scheme module system.
2632
2633 *** New functions: scm_c_define_module, scm_c_use_module,
2634 scm_c_export, scm_c_resolve_module.
2635
2636 They mostly work like their Scheme namesakes. scm_c_define_module
2637 takes a function that is called a context where the new module is
2638 current.
2639
2640 *** Deprecated functions: scm_the_root_module, scm_make_module,
2641 scm_ensure_user_module, scm_load_scheme_module.
2642
2643 Use the new functions instead.
2644
2645 ** Renamed function: scm_internal_with_fluids becomes
2646 scm_c_with_fluids.
2647
2648 scm_internal_with_fluids is available as a deprecated function.
2649
2650 ** New function: scm_c_with_fluid.
2651
2652 Just like scm_c_with_fluids, but takes one fluid and one value instead
2653 of lists of same.
2654
2655 ** Deprecated typedefs: long_long, ulong_long.
2656
2657 They are of questionable utility and they pollute the global
2658 namespace.
2659
2660 ** Deprecated typedef: scm_sizet
2661
2662 It is of questionable utility now that Guile requires ANSI C, and is
2663 oddly named.
2664
2665 ** Deprecated typedefs: scm_port_rw_active, scm_port,
2666 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
2667 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
2668
2669 Made more compliant with the naming policy by adding a _t at the end.
2670
2671 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
2672 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
2673
2674 With the exception of the mysterious scm_2ulong2big, they are still
2675 available under new names (scm_i_mkbig etc). These functions are not
2676 intended to be used in user code. You should avoid dealing with
2677 bignums directly, and should deal with numbers in general (which can
2678 be bignums).
2679
2680 ** Change in behavior: scm_num2long, scm_num2ulong
2681
2682 The scm_num2[u]long functions don't any longer accept an inexact
2683 argument. This change in behavior is motivated by concordance with
2684 R5RS: It is more common that a primitive doesn't want to accept an
2685 inexact for an exact.
2686
2687 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
2688 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
2689 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
2690 scm_num2size.
2691
2692 These are conversion functions between the various ANSI C integral
2693 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
2694 accept an inexact argument.
2695
2696 ** New functions: scm_float2num, scm_double2num,
2697 scm_num2float, scm_num2double.
2698
2699 These are conversion functions between the two ANSI C float types and
2700 Scheme numbers.
2701
2702 ** New number validation macros:
2703 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
2704
2705 See above.
2706
2707 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
2708
2709 These are just nicer-named old scm_protect_object and
2710 scm_unprotect_object.
2711
2712 ** Deprecated functions: scm_protect_object, scm_unprotect_object
2713
2714 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
2715
2716 These functions can be used to register pointers to locations that
2717 hold SCM values.
2718
2719 ** Deprecated function: scm_create_hook.
2720
2721 Its sins are: misleading name, non-modularity and lack of general
2722 usefulness.
2723
2724 \f
2725 Changes since Guile 1.3.4:
2726
2727 * Changes to the distribution
2728
2729 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
2730
2731 We've changed the way we handle generated files in the Guile source
2732 repository. As a result, the procedure for building trees obtained
2733 from the nightly FTP snapshots or via CVS has changed:
2734 - You must have appropriate versions of autoconf, automake, and
2735 libtool installed on your system. See README for info on how to
2736 obtain these programs.
2737 - Before configuring the tree, you must first run the script
2738 `autogen.sh' at the top of the source tree.
2739
2740 The Guile repository used to contain not only source files, written by
2741 humans, but also some generated files, like configure scripts and
2742 Makefile.in files. Even though the contents of these files could be
2743 derived mechanically from other files present, we thought it would
2744 make the tree easier to build if we checked them into CVS.
2745
2746 However, this approach means that minor differences between
2747 developer's installed tools and habits affected the whole team.
2748 So we have removed the generated files from the repository, and
2749 added the autogen.sh script, which will reconstruct them
2750 appropriately.
2751
2752
2753 ** configure now has experimental options to remove support for certain
2754 features:
2755
2756 --disable-arrays omit array and uniform array support
2757 --disable-posix omit posix interfaces
2758 --disable-networking omit networking interfaces
2759 --disable-regex omit regular expression interfaces
2760
2761 These are likely to become separate modules some day.
2762
2763 ** New configure option --enable-debug-freelist
2764
2765 This enables a debugging version of SCM_NEWCELL(), and also registers
2766 an extra primitive, the setter `gc-set-debug-check-freelist!'.
2767
2768 Configure with the --enable-debug-freelist option to enable
2769 the gc-set-debug-check-freelist! primitive, and then use:
2770
2771 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
2772 (gc-set-debug-check-freelist! #f) # turn off checking
2773
2774 Checking of the freelist forces a traversal of the freelist and
2775 a garbage collection before each allocation of a cell. This can
2776 slow down the interpreter dramatically, so the setter should be used to
2777 turn on this extra processing only when necessary.
2778
2779 ** New configure option --enable-debug-malloc
2780
2781 Include code for debugging of calls to scm_must_malloc/realloc/free.
2782
2783 Checks that
2784
2785 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
2786 2. objects reallocated by scm_must_realloc has been allocated by
2787 scm_must_malloc
2788 3. reallocated objects are reallocated with the same what string
2789
2790 But, most importantly, it records the number of allocated objects of
2791 each kind. This is useful when searching for memory leaks.
2792
2793 A Guile compiled with this option provides the primitive
2794 `malloc-stats' which returns an alist with pairs of kind and the
2795 number of objects of that kind.
2796
2797 ** All includes are now referenced relative to the root directory
2798
2799 Since some users have had problems with mixups between Guile and
2800 system headers, we have decided to always refer to Guile headers via
2801 their parent directories. This essentially creates a "private name
2802 space" for Guile headers. This means that the compiler only is given
2803 -I options for the root build and root source directory.
2804
2805 ** Header files kw.h and genio.h have been removed.
2806
2807 ** The module (ice-9 getopt-gnu-style) has been removed.
2808
2809 ** New module (ice-9 documentation)
2810
2811 Implements the interface to documentation strings associated with
2812 objects.
2813
2814 ** New module (ice-9 time)
2815
2816 Provides a macro `time', which displays execution time of a given form.
2817
2818 ** New module (ice-9 history)
2819
2820 Loading this module enables value history in the repl.
2821
2822 * Changes to the stand-alone interpreter
2823
2824 ** New command line option --debug
2825
2826 Start Guile with debugging evaluator and backtraces enabled.
2827
2828 This is useful when debugging your .guile init file or scripts.
2829
2830 ** New help facility
2831
2832 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
2833 (help REGEXP) ditto for objects with names matching REGEXP (a string)
2834 (help 'NAME) gives documentation for NAME, even if it is not an object
2835 (help ,EXPR) gives documentation for object returned by EXPR
2836 (help (my module)) gives module commentary for `(my module)'
2837 (help) gives this text
2838
2839 `help' searches among bindings exported from loaded modules, while
2840 `apropos' searches among bindings visible from the "current" module.
2841
2842 Examples: (help help)
2843 (help cons)
2844 (help "output-string")
2845
2846 ** `help' and `apropos' now prints full module names
2847
2848 ** Dynamic linking now uses libltdl from the libtool package.
2849
2850 The old system dependent code for doing dynamic linking has been
2851 replaced with calls to the libltdl functions which do all the hairy
2852 details for us.
2853
2854 The major improvement is that you can now directly pass libtool
2855 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
2856 will be able to do the best shared library job you can get, via
2857 libltdl.
2858
2859 The way dynamic libraries are found has changed and is not really
2860 portable across platforms, probably. It is therefore recommended to
2861 use absolute filenames when possible.
2862
2863 If you pass a filename without an extension to `dynamic-link', it will
2864 try a few appropriate ones. Thus, the most platform ignorant way is
2865 to specify a name like "libfoo", without any directories and
2866 extensions.
2867
2868 ** Guile COOP threads are now compatible with LinuxThreads
2869
2870 Previously, COOP threading wasn't possible in applications linked with
2871 Linux POSIX threads due to their use of the stack pointer to find the
2872 thread context. This has now been fixed with a workaround which uses
2873 the pthreads to allocate the stack.
2874
2875 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
2876
2877 ** Positions of erring expression in scripts
2878
2879 With version 1.3.4, the location of the erring expression in Guile
2880 scipts is no longer automatically reported. (This should have been
2881 documented before the 1.3.4 release.)
2882
2883 You can get this information by enabling recording of positions of
2884 source expressions and running the debugging evaluator. Put this at
2885 the top of your script (or in your "site" file):
2886
2887 (read-enable 'positions)
2888 (debug-enable 'debug)
2889
2890 ** Backtraces in scripts
2891
2892 It is now possible to get backtraces in scripts.
2893
2894 Put
2895
2896 (debug-enable 'debug 'backtrace)
2897
2898 at the top of the script.
2899
2900 (The first options enables the debugging evaluator.
2901 The second enables backtraces.)
2902
2903 ** Part of module system symbol lookup now implemented in C
2904
2905 The eval closure of most modules is now implemented in C. Since this
2906 was one of the bottlenecks for loading speed, Guile now loads code
2907 substantially faster than before.
2908
2909 ** Attempting to get the value of an unbound variable now produces
2910 an exception with a key of 'unbound-variable instead of 'misc-error.
2911
2912 ** The initial default output port is now unbuffered if it's using a
2913 tty device. Previously in this situation it was line-buffered.
2914
2915 ** New hook: after-gc-hook
2916
2917 after-gc-hook takes over the role of gc-thunk. This hook is run at
2918 the first SCM_TICK after a GC. (Thus, the code is run at the same
2919 point during evaluation as signal handlers.)
2920
2921 Note that this hook should be used only for diagnostic and debugging
2922 purposes. It is not certain that it will continue to be well-defined
2923 when this hook is run in the future.
2924
2925 C programmers: Note the new C level hooks scm_before_gc_c_hook,
2926 scm_before_sweep_c_hook, scm_after_gc_c_hook.
2927
2928 ** Improvements to garbage collector
2929
2930 Guile 1.4 has a new policy for triggering heap allocation and
2931 determining the sizes of heap segments. It fixes a number of problems
2932 in the old GC.
2933
2934 1. The new policy can handle two separate pools of cells
2935 (2-word/4-word) better. (The old policy would run wild, allocating
2936 more and more memory for certain programs.)
2937
2938 2. The old code would sometimes allocate far too much heap so that the
2939 Guile process became gigantic. The new code avoids this.
2940
2941 3. The old code would sometimes allocate too little so that few cells
2942 were freed at GC so that, in turn, too much time was spent in GC.
2943
2944 4. The old code would often trigger heap allocation several times in a
2945 row. (The new scheme predicts how large the segments needs to be
2946 in order not to need further allocation.)
2947
2948 All in all, the new GC policy will make larger applications more
2949 efficient.
2950
2951 The new GC scheme also is prepared for POSIX threading. Threads can
2952 allocate private pools of cells ("clusters") with just a single
2953 function call. Allocation of single cells from such a cluster can
2954 then proceed without any need of inter-thread synchronization.
2955
2956 ** New environment variables controlling GC parameters
2957
2958 GUILE_MAX_SEGMENT_SIZE Maximal segment size
2959 (default = 2097000)
2960
2961 Allocation of 2-word cell heaps:
2962
2963 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
2964 (default = 360000)
2965
2966 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
2967 GC in percent of total heap size
2968 (default = 40)
2969
2970 Allocation of 4-word cell heaps
2971 (used for real numbers and misc other objects):
2972
2973 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
2974
2975 (See entry "Way for application to customize GC parameters" under
2976 section "Changes to the scm_ interface" below.)
2977
2978 ** Guile now implements reals using 4-word cells
2979
2980 This speeds up computation with reals. (They were earlier allocated
2981 with `malloc'.) There is still some room for optimizations, however.
2982
2983 ** Some further steps toward POSIX thread support have been taken
2984
2985 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
2986 don't have much effect any longer, and many of them will be removed in
2987 next release.
2988
2989 *** Signals
2990 are only handled at the top of the evaluator loop, immediately after
2991 I/O, and in scm_equalp.
2992
2993 *** The GC can allocate thread private pools of pairs.
2994
2995 * Changes to Scheme functions and syntax
2996
2997 ** close-input-port and close-output-port are now R5RS
2998
2999 These procedures have been turned into primitives and have R5RS behaviour.
3000
3001 ** New procedure: simple-format PORT MESSAGE ARG1 ...
3002
3003 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
3004 extended by the more sophisticated version in (ice-9 format)
3005
3006 (simple-format port message . args)
3007 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
3008 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
3009 the escapes are replaced with corresponding members of ARGS:
3010 ~A formats using `display' and ~S formats using `write'.
3011 If DESTINATION is #t, then use the `current-output-port',
3012 if DESTINATION is #f, then return a string containing the formatted text.
3013 Does not add a trailing newline."
3014
3015 ** string-ref: the second argument is no longer optional.
3016
3017 ** string, list->string: no longer accept strings in their arguments,
3018 only characters, for compatibility with R5RS.
3019
3020 ** New procedure: port-closed? PORT
3021 Returns #t if PORT is closed or #f if it is open.
3022
3023 ** Deprecated: list*
3024
3025 The list* functionality is now provided by cons* (SRFI-1 compliant)
3026
3027 ** New procedure: cons* ARG1 ARG2 ... ARGn
3028
3029 Like `list', but the last arg provides the tail of the constructed list,
3030 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
3031
3032 Requires at least one argument. If given one argument, that argument
3033 is returned as result.
3034
3035 This function is called `list*' in some other Schemes and in Common LISP.
3036
3037 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
3038
3039 ** New procedure: object-documentation OBJECT
3040
3041 Returns the documentation string associated with OBJECT. The
3042 procedure uses a caching mechanism so that subsequent lookups are
3043 faster.
3044
3045 Exported by (ice-9 documentation).
3046
3047 ** module-name now returns full names of modules
3048
3049 Previously, only the last part of the name was returned (`session' for
3050 `(ice-9 session)'). Ex: `(ice-9 session)'.
3051
3052 * Changes to the gh_ interface
3053
3054 ** Deprecated: gh_int2scmb
3055
3056 Use gh_bool2scm instead.
3057
3058 * Changes to the scm_ interface
3059
3060 ** Guile primitives now carry docstrings!
3061
3062 Thanks to Greg Badros!
3063
3064 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
3065
3066 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
3067 macros and must contain a docstring that is extracted into foo.doc using a new
3068 guile-doc-snarf script (that uses guile-doc-snarf.awk).
3069
3070 However, a major overhaul of these macros is scheduled for the next release of
3071 guile.
3072
3073 ** Guile primitives use a new technique for validation of arguments
3074
3075 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
3076 the readability of argument checking.
3077
3078 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
3079
3080 ** New macros: SCM_PACK, SCM_UNPACK
3081
3082 Compose/decompose an SCM value.
3083
3084 The SCM type is now treated as an abstract data type and may be defined as a
3085 long, a void* or as a struct, depending on the architecture and compile time
3086 options. This makes it easier to find several types of bugs, for example when
3087 SCM values are treated as integers without conversion. Values of the SCM type
3088 should be treated as "atomic" values. These macros are used when
3089 composing/decomposing an SCM value, either because you want to access
3090 individual bits, or because you want to treat it as an integer value.
3091
3092 E.g., in order to set bit 7 in an SCM value x, use the expression
3093
3094 SCM_PACK (SCM_UNPACK (x) | 0x80)
3095
3096 ** The name property of hooks is deprecated.
3097 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
3098
3099 You can emulate this feature by using object properties.
3100
3101 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
3102 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
3103 SCM_NVECTORP
3104
3105 These macros will be removed in a future release of Guile.
3106
3107 ** The following types, functions and macros from numbers.h are deprecated:
3108 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
3109 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
3110
3111 ** Port internals: the rw_random variable in the scm_port structure
3112 must be set to non-zero in any random access port. In recent Guile
3113 releases it was only set for bidirectional random-access ports.
3114
3115 ** Port internals: the seek ptob procedure is now responsible for
3116 resetting the buffers if required. The change was made so that in the
3117 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
3118 the fport and strport ptobs can avoid resetting the buffers,
3119 in particular to avoid discarding unread chars. An existing port
3120 type can be fixed by adding something like the following to the
3121 beginning of the ptob seek procedure:
3122
3123 if (pt->rw_active == SCM_PORT_READ)
3124 scm_end_input (object);
3125 else if (pt->rw_active == SCM_PORT_WRITE)
3126 ptob->flush (object);
3127
3128 although to actually avoid resetting the buffers and discard unread
3129 chars requires further hacking that depends on the characteristics
3130 of the ptob.
3131
3132 ** Deprecated functions: scm_fseek, scm_tag
3133
3134 These functions are no longer used and will be removed in a future version.
3135
3136 ** The scm_sysmissing procedure is no longer used in libguile.
3137 Unless it turns out to be unexpectedly useful to somebody, it will be
3138 removed in a future version.
3139
3140 ** The format of error message strings has changed
3141
3142 The two C procedures: scm_display_error and scm_error, as well as the
3143 primitive `scm-error', now use scm_simple_format to do their work.
3144 This means that the message strings of all code must be updated to use
3145 ~A where %s was used before, and ~S where %S was used before.
3146
3147 During the period when there still are a lot of old Guiles out there,
3148 you might want to support both old and new versions of Guile.
3149
3150 There are basically two methods to achieve this. Both methods use
3151 autoconf. Put
3152
3153 AC_CHECK_FUNCS(scm_simple_format)
3154
3155 in your configure.in.
3156
3157 Method 1: Use the string concatenation features of ANSI C's
3158 preprocessor.
3159
3160 In C:
3161
3162 #ifdef HAVE_SCM_SIMPLE_FORMAT
3163 #define FMT_S "~S"
3164 #else
3165 #define FMT_S "%S"
3166 #endif
3167
3168 Then represent each of your error messages using a preprocessor macro:
3169
3170 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
3171
3172 In Scheme:
3173
3174 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
3175 (define make-message string-append)
3176
3177 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
3178
3179 Method 2: Use the oldfmt function found in doc/oldfmt.c.
3180
3181 In C:
3182
3183 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
3184 ...);
3185
3186 In Scheme:
3187
3188 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
3189 ...)
3190
3191
3192 ** Deprecated: coop_mutex_init, coop_condition_variable_init
3193
3194 Don't use the functions coop_mutex_init and
3195 coop_condition_variable_init. They will change.
3196
3197 Use scm_mutex_init and scm_cond_init instead.
3198
3199 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
3200 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
3201 COND, as `scm_cond_wait' does, but it also bounds the duration
3202 of the wait. If COND has not been signaled before time ABSTIME,
3203 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
3204 returns the error code `ETIMEDOUT'.
3205
3206 The ABSTIME parameter specifies an absolute time, with the same
3207 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
3208 to 00:00:00 GMT, January 1, 1970.
3209
3210 ** New function: scm_cond_broadcast (scm_cond_t *COND)
3211 `scm_cond_broadcast' restarts all the threads that are waiting
3212 on the condition variable COND. Nothing happens if no threads are
3213 waiting on COND.
3214
3215 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
3216 `scm_key_create' allocates a new TSD key. The key is stored in
3217 the location pointed to by KEY. There is no limit on the number
3218 of keys allocated at a given time. The value initially associated
3219 with the returned key is `NULL' in all currently executing threads.
3220
3221 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
3222 function associated with the key. When a thread terminates,
3223 DESTR_FUNCTION is called on the value associated with the key in
3224 that thread. The DESTR_FUNCTION is not called if a key is deleted
3225 with `scm_key_delete' or a value is changed with
3226 `scm_setspecific'. The order in which destructor functions are
3227 called at thread termination time is unspecified.
3228
3229 Destructors are not yet implemented.
3230
3231 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
3232 `scm_setspecific' changes the value associated with KEY in the
3233 calling thread, storing the given POINTER instead.
3234
3235 ** New function: scm_getspecific (scm_key_t KEY)
3236 `scm_getspecific' returns the value currently associated with
3237 KEY in the calling thread.
3238
3239 ** New function: scm_key_delete (scm_key_t KEY)
3240 `scm_key_delete' deallocates a TSD key. It does not check
3241 whether non-`NULL' values are associated with that key in the
3242 currently executing threads, nor call the destructor function
3243 associated with the key.
3244
3245 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
3246
3247 Initialize a C level hook HOOK with associated HOOK_DATA and type
3248 TYPE. (See scm_c_hook_run ().)
3249
3250 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
3251
3252 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
3253 is true, add it last, otherwise first. The same FUNC can be added
3254 multiple times if FUNC_DATA differ and vice versa.
3255
3256 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
3257
3258 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
3259 function is only removed if both FUNC and FUNC_DATA matches.
3260
3261 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
3262
3263 Run hook HOOK passing DATA to the hook functions.
3264
3265 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
3266 returned is undefined.
3267
3268 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
3269 returns a non-NULL value. This value is returned as the result of
3270 scm_c_hook_run. If all functions return NULL, NULL is returned.
3271
3272 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
3273 returns a NULL value, and NULL is returned. If all functions returns
3274 a non-NULL value, the last value is returned.
3275
3276 ** New C level GC hooks
3277
3278 Five new C level hooks has been added to the garbage collector.
3279
3280 scm_before_gc_c_hook
3281 scm_after_gc_c_hook
3282
3283 are run before locking and after unlocking the heap. The system is
3284 thus in a mode where evaluation can take place. (Except that
3285 scm_before_gc_c_hook must not allocate new cells.)
3286
3287 scm_before_mark_c_hook
3288 scm_before_sweep_c_hook
3289 scm_after_sweep_c_hook
3290
3291 are run when the heap is locked. These are intended for extension of
3292 the GC in a modular fashion. Examples are the weaks and guardians
3293 modules.
3294
3295 ** Way for application to customize GC parameters
3296
3297 The application can set up other default values for the GC heap
3298 allocation parameters
3299
3300 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
3301 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
3302 GUILE_MAX_SEGMENT_SIZE,
3303
3304 by setting
3305
3306 scm_default_init_heap_size_1, scm_default_min_yield_1,
3307 scm_default_init_heap_size_2, scm_default_min_yield_2,
3308 scm_default_max_segment_size
3309
3310 respectively before callong scm_boot_guile.
3311
3312 (See entry "New environment variables ..." in section
3313 "Changes to the stand-alone interpreter" above.)
3314
3315 ** scm_protect_object/scm_unprotect_object now nest
3316
3317 This means that you can call scm_protect_object multiple times on an
3318 object and count on the object being protected until
3319 scm_unprotect_object has been call the same number of times.
3320
3321 The functions also have better time complexity.
3322
3323 Still, it is usually possible to structure the application in a way
3324 that you don't need to use these functions. For example, if you use a
3325 protected standard Guile list to keep track of live objects rather
3326 than some custom data type, objects will die a natural death when they
3327 are no longer needed.
3328
3329 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
3330
3331 Guile does not provide the float representation for inexact real numbers any
3332 more. Now, only doubles are used to represent inexact real numbers. Further,
3333 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
3334 and scm_tc16_complex, respectively.
3335
3336 ** Removed deprecated type scm_smobfuns
3337
3338 ** Removed deprecated function scm_newsmob
3339
3340 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
3341
3342 There is an ongoing discussion among the developers whether to
3343 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
3344 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
3345 until this issue has been settled.
3346
3347 ** Removed deprecated type tag scm_tc16_kw
3348
3349 ** Added type tag scm_tc16_keyword
3350
3351 (This was introduced already in release 1.3.4 but was not documented
3352 until now.)
3353
3354 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
3355
3356 * Changes to system call interfaces:
3357
3358 ** The "select" procedure now tests port buffers for the ability to
3359 provide input or accept output. Previously only the underlying file
3360 descriptors were checked.
3361
3362 ** New variable PIPE_BUF: the maximum number of bytes that can be
3363 atomically written to a pipe.
3364
3365 ** If a facility is not available on the system when Guile is
3366 compiled, the corresponding primitive procedure will not be defined.
3367 Previously it would have been defined but would throw a system-error
3368 exception if called. Exception handlers which catch this case may
3369 need minor modification: an error will be thrown with key
3370 'unbound-variable instead of 'system-error. Alternatively it's
3371 now possible to use `defined?' to check whether the facility is
3372 available.
3373
3374 ** Procedures which depend on the timezone should now give the correct
3375 result on systems which cache the TZ environment variable, even if TZ
3376 is changed without calling tzset.
3377
3378 * Changes to the networking interfaces:
3379
3380 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
3381 long integers between network and host format. For now, it's not
3382 particularly convenient to do this kind of thing, but consider:
3383
3384 (define write-network-long
3385 (lambda (value port)
3386 (let ((v (make-uniform-vector 1 1 0)))
3387 (uniform-vector-set! v 0 (htonl value))
3388 (uniform-vector-write v port))))
3389
3390 (define read-network-long
3391 (lambda (port)
3392 (let ((v (make-uniform-vector 1 1 0)))
3393 (uniform-vector-read! v port)
3394 (ntohl (uniform-vector-ref v 0)))))
3395
3396 ** If inet-aton fails, it now throws an error with key 'misc-error
3397 instead of 'system-error, since errno is not relevant.
3398
3399 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
3400 specific keys instead of 'system-error. The latter is inappropriate
3401 since errno will not have been set. The keys are:
3402 'host-not-found, 'try-again, 'no-recovery and 'no-data.
3403
3404 ** sethostent, setnetent, setprotoent, setservent: now take an
3405 optional argument STAYOPEN, which specifies whether the database
3406 remains open after a database entry is accessed randomly (e.g., using
3407 gethostbyname for the hosts database.) The default is #f. Previously
3408 #t was always used.
3409
3410 \f
3411 Changes since Guile 1.3.2:
3412
3413 * Changes to the stand-alone interpreter
3414
3415 ** Debugger
3416
3417 An initial version of the Guile debugger written by Chris Hanson has
3418 been added. The debugger is still under development but is included
3419 in the distribution anyway since it is already quite useful.
3420
3421 Type
3422
3423 (debug)
3424
3425 after an error to enter the debugger. Type `help' inside the debugger
3426 for a description of available commands.
3427
3428 If you prefer to have stack frames numbered and printed in
3429 anti-chronological order and prefer up in the stack to be down on the
3430 screen as is the case in gdb, you can put
3431
3432 (debug-enable 'backwards)
3433
3434 in your .guile startup file. (However, this means that Guile can't
3435 use indentation to indicate stack level.)
3436
3437 The debugger is autoloaded into Guile at the first use.
3438
3439 ** Further enhancements to backtraces
3440
3441 There is a new debug option `width' which controls the maximum width
3442 on the screen of printed stack frames. Fancy printing parameters
3443 ("level" and "length" as in Common LISP) are adaptively adjusted for
3444 each stack frame to give maximum information while still fitting
3445 within the bounds. If the stack frame can't be made to fit by
3446 adjusting parameters, it is simply cut off at the end. This is marked
3447 with a `$'.
3448
3449 ** Some modules are now only loaded when the repl is started
3450
3451 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
3452 regex) are now loaded into (guile-user) only if the repl has been
3453 started. The effect is that the startup time for scripts has been
3454 reduced to 30% of what it was previously.
3455
3456 Correctly written scripts load the modules they require at the top of
3457 the file and should not be affected by this change.
3458
3459 ** Hooks are now represented as smobs
3460
3461 * Changes to Scheme functions and syntax
3462
3463 ** Readline support has changed again.
3464
3465 The old (readline-activator) module is gone. Use (ice-9 readline)
3466 instead, which now contains all readline functionality. So the code
3467 to activate readline is now
3468
3469 (use-modules (ice-9 readline))
3470 (activate-readline)
3471
3472 This should work at any time, including from the guile prompt.
3473
3474 To avoid confusion about the terms of Guile's license, please only
3475 enable readline for your personal use; please don't make it the
3476 default for others. Here is why we make this rather odd-sounding
3477 request:
3478
3479 Guile is normally licensed under a weakened form of the GNU General
3480 Public License, which allows you to link code with Guile without
3481 placing that code under the GPL. This exception is important to some
3482 people.
3483
3484 However, since readline is distributed under the GNU General Public
3485 License, when you link Guile with readline, either statically or
3486 dynamically, you effectively change Guile's license to the strict GPL.
3487 Whenever you link any strictly GPL'd code into Guile, uses of Guile
3488 which are normally permitted become forbidden. This is a rather
3489 non-obvious consequence of the licensing terms.
3490
3491 So, to make sure things remain clear, please let people choose for
3492 themselves whether to link GPL'd libraries like readline with Guile.
3493
3494 ** regexp-substitute/global has changed slightly, but incompatibly.
3495
3496 If you include a function in the item list, the string of the match
3497 object it receives is the same string passed to
3498 regexp-substitute/global, not some suffix of that string.
3499 Correspondingly, the match's positions are relative to the entire
3500 string, not the suffix.
3501
3502 If the regexp can match the empty string, the way matches are chosen
3503 from the string has changed. regexp-substitute/global recognizes the
3504 same set of matches that list-matches does; see below.
3505
3506 ** New function: list-matches REGEXP STRING [FLAGS]
3507
3508 Return a list of match objects, one for every non-overlapping, maximal
3509 match of REGEXP in STRING. The matches appear in left-to-right order.
3510 list-matches only reports matches of the empty string if there are no
3511 other matches which begin on, end at, or include the empty match's
3512 position.
3513
3514 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
3515
3516 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
3517
3518 For each match of REGEXP in STRING, apply PROC to the match object,
3519 and the last value PROC returned, or INIT for the first call. Return
3520 the last value returned by PROC. We apply PROC to the matches as they
3521 appear from left to right.
3522
3523 This function recognizes matches according to the same criteria as
3524 list-matches.
3525
3526 Thus, you could define list-matches like this:
3527
3528 (define (list-matches regexp string . flags)
3529 (reverse! (apply fold-matches regexp string '() cons flags)))
3530
3531 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
3532
3533 ** Hooks
3534
3535 *** New function: hook? OBJ
3536
3537 Return #t if OBJ is a hook, otherwise #f.
3538
3539 *** New function: make-hook-with-name NAME [ARITY]
3540
3541 Return a hook with name NAME and arity ARITY. The default value for
3542 ARITY is 0. The only effect of NAME is that it will appear when the
3543 hook object is printed to ease debugging.
3544
3545 *** New function: hook-empty? HOOK
3546
3547 Return #t if HOOK doesn't contain any procedures, otherwise #f.
3548
3549 *** New function: hook->list HOOK
3550
3551 Return a list of the procedures that are called when run-hook is
3552 applied to HOOK.
3553
3554 ** `map' signals an error if its argument lists are not all the same length.
3555
3556 This is the behavior required by R5RS, so this change is really a bug
3557 fix. But it seems to affect a lot of people's code, so we're
3558 mentioning it here anyway.
3559
3560 ** Print-state handling has been made more transparent
3561
3562 Under certain circumstances, ports are represented as a port with an
3563 associated print state. Earlier, this pair was represented as a pair
3564 (see "Some magic has been added to the printer" below). It is now
3565 indistinguishable (almost; see `get-print-state') from a port on the
3566 user level.
3567
3568 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
3569
3570 Return a new port with the associated print state PRINT-STATE.
3571
3572 *** New function: get-print-state OUTPUT-PORT
3573
3574 Return the print state associated with this port if it exists,
3575 otherwise return #f.
3576
3577 *** New function: directory-stream? OBJECT
3578
3579 Returns true iff OBJECT is a directory stream --- the sort of object
3580 returned by `opendir'.
3581
3582 ** New function: using-readline?
3583
3584 Return #t if readline is in use in the current repl.
3585
3586 ** structs will be removed in 1.4
3587
3588 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
3589 and use GOOPS objects as the fundamental record type.
3590
3591 * Changes to the scm_ interface
3592
3593 ** structs will be removed in 1.4
3594
3595 The entire current struct interface (struct.c, struct.h) will be
3596 replaced in Guile 1.4. We will merge GOOPS into libguile and use
3597 GOOPS objects as the fundamental record type.
3598
3599 ** The internal representation of subr's has changed
3600
3601 Instead of giving a hint to the subr name, the CAR field of the subr
3602 now contains an index to a subr entry in scm_subr_table.
3603
3604 *** New variable: scm_subr_table
3605
3606 An array of subr entries. A subr entry contains the name, properties
3607 and documentation associated with the subr. The properties and
3608 documentation slots are not yet used.
3609
3610 ** A new scheme for "forwarding" calls to a builtin to a generic function
3611
3612 It is now possible to extend the functionality of some Guile
3613 primitives by letting them defer a call to a GOOPS generic function on
3614 argument mismatch. This means that there is no loss of efficiency in
3615 normal evaluation.
3616
3617 Example:
3618
3619 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
3620 (define-method + ((x <string>) (y <string>))
3621 (string-append x y))
3622
3623 + will still be as efficient as usual in numerical calculations, but
3624 can also be used for concatenating strings.
3625
3626 Who will be the first one to extend Guile's numerical tower to
3627 rationals? :) [OK, there a few other things to fix before this can
3628 be made in a clean way.]
3629
3630 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
3631
3632 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
3633
3634 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
3635
3636 These do the same job as SCM_PROC and SCM_PROC1, but they also define
3637 a variable GENERIC which can be used by the dispatch macros below.
3638
3639 [This is experimental code which may change soon.]
3640
3641 *** New macros for forwarding control to a generic on arg type error
3642
3643 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
3644
3645 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
3646
3647 These correspond to the scm_wta function call, and have the same
3648 behaviour until the user has called the GOOPS primitive
3649 `enable-primitive-generic!'. After that, these macros will apply the
3650 generic function GENERIC to the argument(s) instead of calling
3651 scm_wta.
3652
3653 [This is experimental code which may change soon.]
3654
3655 *** New macros for argument testing with generic dispatch
3656
3657 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
3658
3659 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
3660
3661 These correspond to the SCM_ASSERT macro, but will defer control to
3662 GENERIC on error after `enable-primitive-generic!' has been called.
3663
3664 [This is experimental code which may change soon.]
3665
3666 ** New function: SCM scm_eval_body (SCM body, SCM env)
3667
3668 Evaluates the body of a special form.
3669
3670 ** The internal representation of struct's has changed
3671
3672 Previously, four slots were allocated for the procedure(s) of entities
3673 and operators. The motivation for this representation had to do with
3674 the structure of the evaluator, the wish to support tail-recursive
3675 generic functions, and efficiency. Since the generic function
3676 dispatch mechanism has changed, there is no longer a need for such an
3677 expensive representation, and the representation has been simplified.
3678
3679 This should not make any difference for most users.
3680
3681 ** GOOPS support has been cleaned up.
3682
3683 Some code has been moved from eval.c to objects.c and code in both of
3684 these compilation units has been cleaned up and better structured.
3685
3686 *** New functions for applying generic functions
3687
3688 New function: SCM scm_apply_generic (GENERIC, ARGS)
3689 New function: SCM scm_call_generic_0 (GENERIC)
3690 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
3691 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
3692 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
3693
3694 ** Deprecated function: scm_make_named_hook
3695
3696 It is now replaced by:
3697
3698 ** New function: SCM scm_create_hook (const char *name, int arity)
3699
3700 Creates a hook in the same way as make-hook above but also
3701 binds a variable named NAME to it.
3702
3703 This is the typical way of creating a hook from C code.
3704
3705 Currently, the variable is created in the "current" module.
3706 This might change when we get the new module system.
3707
3708 [The behaviour is identical to scm_make_named_hook.]
3709
3710
3711 \f
3712 Changes since Guile 1.3:
3713
3714 * Changes to mailing lists
3715
3716 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
3717
3718 See the README file to find current addresses for all the Guile
3719 mailing lists.
3720
3721 * Changes to the distribution
3722
3723 ** Readline support is no longer included with Guile by default.
3724
3725 Based on the different license terms of Guile and Readline, we
3726 concluded that Guile should not *by default* cause the linking of
3727 Readline into an application program. Readline support is now offered
3728 as a separate module, which is linked into an application only when
3729 you explicitly specify it.
3730
3731 Although Guile is GNU software, its distribution terms add a special
3732 exception to the usual GNU General Public License (GPL). Guile's
3733 license includes a clause that allows you to link Guile with non-free
3734 programs. We add this exception so as not to put Guile at a
3735 disadvantage vis-a-vis other extensibility packages that support other
3736 languages.
3737
3738 In contrast, the GNU Readline library is distributed under the GNU
3739 General Public License pure and simple. This means that you may not
3740 link Readline, even dynamically, into an application unless it is
3741 distributed under a free software license that is compatible the GPL.
3742
3743 Because of this difference in distribution terms, an application that
3744 can use Guile may not be able to use Readline. Now users will be
3745 explicitly offered two independent decisions about the use of these
3746 two packages.
3747
3748 You can activate the readline support by issuing
3749
3750 (use-modules (readline-activator))
3751 (activate-readline)
3752
3753 from your ".guile" file, for example.
3754
3755 * Changes to the stand-alone interpreter
3756
3757 ** All builtins now print as primitives.
3758 Previously builtin procedures not belonging to the fundamental subr
3759 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
3760 Now, they print as #<primitive-procedure NAME>.
3761
3762 ** Backtraces slightly more intelligible.
3763 gsubr-apply and macro transformer application frames no longer appear
3764 in backtraces.
3765
3766 * Changes to Scheme functions and syntax
3767
3768 ** Guile now correctly handles internal defines by rewriting them into
3769 their equivalent letrec. Previously, internal defines would
3770 incrementally add to the innermost environment, without checking
3771 whether the restrictions specified in RnRS were met. This lead to the
3772 correct behaviour when these restriction actually were met, but didn't
3773 catch all illegal uses. Such an illegal use could lead to crashes of
3774 the Guile interpreter or or other unwanted results. An example of
3775 incorrect internal defines that made Guile behave erratically:
3776
3777 (let ()
3778 (define a 1)
3779 (define (b) a)
3780 (define c (1+ (b)))
3781 (define d 3)
3782
3783 (b))
3784
3785 => 2
3786
3787 The problem with this example is that the definition of `c' uses the
3788 value of `b' directly. This confuses the meoization machine of Guile
3789 so that the second call of `b' (this time in a larger environment that
3790 also contains bindings for `c' and `d') refers to the binding of `c'
3791 instead of `a'. You could also make Guile crash with a variation on
3792 this theme:
3793
3794 (define (foo flag)
3795 (define a 1)
3796 (define (b flag) (if flag a 1))
3797 (define c (1+ (b flag)))
3798 (define d 3)
3799
3800 (b #t))
3801
3802 (foo #f)
3803 (foo #t)
3804
3805 From now on, Guile will issue an `Unbound variable: b' error message
3806 for both examples.
3807
3808 ** Hooks
3809
3810 A hook contains a list of functions which should be called on
3811 particular occasions in an existing program. Hooks are used for
3812 customization.
3813
3814 A window manager might have a hook before-window-map-hook. The window
3815 manager uses the function run-hooks to call all functions stored in
3816 before-window-map-hook each time a window is mapped. The user can
3817 store functions in the hook using add-hook!.
3818
3819 In Guile, hooks are first class objects.
3820
3821 *** New function: make-hook [N_ARGS]
3822
3823 Return a hook for hook functions which can take N_ARGS arguments.
3824 The default value for N_ARGS is 0.
3825
3826 (See also scm_make_named_hook below.)
3827
3828 *** New function: add-hook! HOOK PROC [APPEND_P]
3829
3830 Put PROC at the beginning of the list of functions stored in HOOK.
3831 If APPEND_P is supplied, and non-false, put PROC at the end instead.
3832
3833 PROC must be able to take the number of arguments specified when the
3834 hook was created.
3835
3836 If PROC already exists in HOOK, then remove it first.
3837
3838 *** New function: remove-hook! HOOK PROC
3839
3840 Remove PROC from the list of functions in HOOK.
3841
3842 *** New function: reset-hook! HOOK
3843
3844 Clear the list of hook functions stored in HOOK.
3845
3846 *** New function: run-hook HOOK ARG1 ...
3847
3848 Run all hook functions stored in HOOK with arguments ARG1 ... .
3849 The number of arguments supplied must correspond to the number given
3850 when the hook was created.
3851
3852 ** The function `dynamic-link' now takes optional keyword arguments.
3853 The only keyword argument that is currently defined is `:global
3854 BOOL'. With it, you can control whether the shared library will be
3855 linked in global mode or not. In global mode, the symbols from the
3856 linked library can be used to resolve references from other
3857 dynamically linked libraries. In non-global mode, the linked
3858 library is essentially invisible and can only be accessed via
3859 `dynamic-func', etc. The default is now to link in global mode.
3860 Previously, the default has been non-global mode.
3861
3862 The `#:global' keyword is only effective on platforms that support
3863 the dlopen family of functions.
3864
3865 ** New function `provided?'
3866
3867 - Function: provided? FEATURE
3868 Return true iff FEATURE is supported by this installation of
3869 Guile. FEATURE must be a symbol naming a feature; the global
3870 variable `*features*' is a list of available features.
3871
3872 ** Changes to the module (ice-9 expect):
3873
3874 *** The expect-strings macro now matches `$' in a regular expression
3875 only at a line-break or end-of-file by default. Previously it would
3876 match the end of the string accumulated so far. The old behaviour
3877 can be obtained by setting the variable `expect-strings-exec-flags'
3878 to 0.
3879
3880 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
3881 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
3882 in a regular expression will still match before a line-break or
3883 end-of-file. The default is `regexp/noteol'.
3884
3885 *** The expect-strings macro now uses a variable
3886 `expect-strings-compile-flags' for the flags to be supplied to
3887 `make-regexp'. The default is `regexp/newline', which was previously
3888 hard-coded.
3889
3890 *** The expect macro now supplies two arguments to a match procedure:
3891 the current accumulated string and a flag to indicate whether
3892 end-of-file has been reached. Previously only the string was supplied.
3893 If end-of-file is reached, the match procedure will be called an
3894 additional time with the same accumulated string as the previous call
3895 but with the flag set.
3896
3897 ** New module (ice-9 format), implementing the Common Lisp `format' function.
3898
3899 This code, and the documentation for it that appears here, was
3900 borrowed from SLIB, with minor adaptations for Guile.
3901
3902 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
3903 An almost complete implementation of Common LISP format description
3904 according to the CL reference book `Common LISP' from Guy L.
3905 Steele, Digital Press. Backward compatible to most of the
3906 available Scheme format implementations.
3907
3908 Returns `#t', `#f' or a string; has side effect of printing
3909 according to FORMAT-STRING. If DESTINATION is `#t', the output is
3910 to the current output port and `#t' is returned. If DESTINATION
3911 is `#f', a formatted string is returned as the result of the call.
3912 NEW: If DESTINATION is a string, DESTINATION is regarded as the
3913 format string; FORMAT-STRING is then the first argument and the
3914 output is returned as a string. If DESTINATION is a number, the
3915 output is to the current error port if available by the
3916 implementation. Otherwise DESTINATION must be an output port and
3917 `#t' is returned.
3918
3919 FORMAT-STRING must be a string. In case of a formatting error
3920 format returns `#f' and prints a message on the current output or
3921 error port. Characters are output as if the string were output by
3922 the `display' function with the exception of those prefixed by a
3923 tilde (~). For a detailed description of the FORMAT-STRING syntax
3924 please consult a Common LISP format reference manual. For a test
3925 suite to verify this format implementation load `formatst.scm'.
3926 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
3927
3928 Note: `format' is not reentrant, i.e. only one `format'-call may
3929 be executed at a time.
3930
3931
3932 *** Format Specification (Format version 3.0)
3933
3934 Please consult a Common LISP format reference manual for a detailed
3935 description of the format string syntax. For a demonstration of the
3936 implemented directives see `formatst.scm'.
3937
3938 This implementation supports directive parameters and modifiers (`:'
3939 and `@' characters). Multiple parameters must be separated by a comma
3940 (`,'). Parameters can be numerical parameters (positive or negative),
3941 character parameters (prefixed by a quote character (`''), variable
3942 parameters (`v'), number of rest arguments parameter (`#'), empty and
3943 default parameters. Directive characters are case independent. The
3944 general form of a directive is:
3945
3946 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
3947
3948 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
3949
3950 *** Implemented CL Format Control Directives
3951
3952 Documentation syntax: Uppercase characters represent the
3953 corresponding control directive characters. Lowercase characters
3954 represent control directive parameter descriptions.
3955
3956 `~A'
3957 Any (print as `display' does).
3958 `~@A'
3959 left pad.
3960
3961 `~MINCOL,COLINC,MINPAD,PADCHARA'
3962 full padding.
3963
3964 `~S'
3965 S-expression (print as `write' does).
3966 `~@S'
3967 left pad.
3968
3969 `~MINCOL,COLINC,MINPAD,PADCHARS'
3970 full padding.
3971
3972 `~D'
3973 Decimal.
3974 `~@D'
3975 print number sign always.
3976
3977 `~:D'
3978 print comma separated.
3979
3980 `~MINCOL,PADCHAR,COMMACHARD'
3981 padding.
3982
3983 `~X'
3984 Hexadecimal.
3985 `~@X'
3986 print number sign always.
3987
3988 `~:X'
3989 print comma separated.
3990
3991 `~MINCOL,PADCHAR,COMMACHARX'
3992 padding.
3993
3994 `~O'
3995 Octal.
3996 `~@O'
3997 print number sign always.
3998
3999 `~:O'
4000 print comma separated.
4001
4002 `~MINCOL,PADCHAR,COMMACHARO'
4003 padding.
4004
4005 `~B'
4006 Binary.
4007 `~@B'
4008 print number sign always.
4009
4010 `~:B'
4011 print comma separated.
4012
4013 `~MINCOL,PADCHAR,COMMACHARB'
4014 padding.
4015
4016 `~NR'
4017 Radix N.
4018 `~N,MINCOL,PADCHAR,COMMACHARR'
4019 padding.
4020
4021 `~@R'
4022 print a number as a Roman numeral.
4023
4024 `~:@R'
4025 print a number as an "old fashioned" Roman numeral.
4026
4027 `~:R'
4028 print a number as an ordinal English number.
4029
4030 `~:@R'
4031 print a number as a cardinal English number.
4032
4033 `~P'
4034 Plural.
4035 `~@P'
4036 prints `y' and `ies'.
4037
4038 `~:P'
4039 as `~P but jumps 1 argument backward.'
4040
4041 `~:@P'
4042 as `~@P but jumps 1 argument backward.'
4043
4044 `~C'
4045 Character.
4046 `~@C'
4047 prints a character as the reader can understand it (i.e. `#\'
4048 prefixing).
4049
4050 `~:C'
4051 prints a character as emacs does (eg. `^C' for ASCII 03).
4052
4053 `~F'
4054 Fixed-format floating-point (prints a flonum like MMM.NNN).
4055 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
4056 `~@F'
4057 If the number is positive a plus sign is printed.
4058
4059 `~E'
4060 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
4061 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
4062 `~@E'
4063 If the number is positive a plus sign is printed.
4064
4065 `~G'
4066 General floating-point (prints a flonum either fixed or
4067 exponential).
4068 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
4069 `~@G'
4070 If the number is positive a plus sign is printed.
4071
4072 `~$'
4073 Dollars floating-point (prints a flonum in fixed with signs
4074 separated).
4075 `~DIGITS,SCALE,WIDTH,PADCHAR$'
4076 `~@$'
4077 If the number is positive a plus sign is printed.
4078
4079 `~:@$'
4080 A sign is always printed and appears before the padding.
4081
4082 `~:$'
4083 The sign appears before the padding.
4084
4085 `~%'
4086 Newline.
4087 `~N%'
4088 print N newlines.
4089
4090 `~&'
4091 print newline if not at the beginning of the output line.
4092 `~N&'
4093 prints `~&' and then N-1 newlines.
4094
4095 `~|'
4096 Page Separator.
4097 `~N|'
4098 print N page separators.
4099
4100 `~~'
4101 Tilde.
4102 `~N~'
4103 print N tildes.
4104
4105 `~'<newline>
4106 Continuation Line.
4107 `~:'<newline>
4108 newline is ignored, white space left.
4109
4110 `~@'<newline>
4111 newline is left, white space ignored.
4112
4113 `~T'
4114 Tabulation.
4115 `~@T'
4116 relative tabulation.
4117
4118 `~COLNUM,COLINCT'
4119 full tabulation.
4120
4121 `~?'
4122 Indirection (expects indirect arguments as a list).
4123 `~@?'
4124 extracts indirect arguments from format arguments.
4125
4126 `~(STR~)'
4127 Case conversion (converts by `string-downcase').
4128 `~:(STR~)'
4129 converts by `string-capitalize'.
4130
4131 `~@(STR~)'
4132 converts by `string-capitalize-first'.
4133
4134 `~:@(STR~)'
4135 converts by `string-upcase'.
4136
4137 `~*'
4138 Argument Jumping (jumps 1 argument forward).
4139 `~N*'
4140 jumps N arguments forward.
4141
4142 `~:*'
4143 jumps 1 argument backward.
4144
4145 `~N:*'
4146 jumps N arguments backward.
4147
4148 `~@*'
4149 jumps to the 0th argument.
4150
4151 `~N@*'
4152 jumps to the Nth argument (beginning from 0)
4153
4154 `~[STR0~;STR1~;...~;STRN~]'
4155 Conditional Expression (numerical clause conditional).
4156 `~N['
4157 take argument from N.
4158
4159 `~@['
4160 true test conditional.
4161
4162 `~:['
4163 if-else-then conditional.
4164
4165 `~;'
4166 clause separator.
4167
4168 `~:;'
4169 default clause follows.
4170
4171 `~{STR~}'
4172 Iteration (args come from the next argument (a list)).
4173 `~N{'
4174 at most N iterations.
4175
4176 `~:{'
4177 args from next arg (a list of lists).
4178
4179 `~@{'
4180 args from the rest of arguments.
4181
4182 `~:@{'
4183 args from the rest args (lists).
4184
4185 `~^'
4186 Up and out.
4187 `~N^'
4188 aborts if N = 0
4189
4190 `~N,M^'
4191 aborts if N = M
4192
4193 `~N,M,K^'
4194 aborts if N <= M <= K
4195
4196 *** Not Implemented CL Format Control Directives
4197
4198 `~:A'
4199 print `#f' as an empty list (see below).
4200
4201 `~:S'
4202 print `#f' as an empty list (see below).
4203
4204 `~<~>'
4205 Justification.
4206
4207 `~:^'
4208 (sorry I don't understand its semantics completely)
4209
4210 *** Extended, Replaced and Additional Control Directives
4211
4212 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
4213 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
4214 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
4215 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
4216 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
4217 COMMAWIDTH is the number of characters between two comma
4218 characters.
4219
4220 `~I'
4221 print a R4RS complex number as `~F~@Fi' with passed parameters for
4222 `~F'.
4223
4224 `~Y'
4225 Pretty print formatting of an argument for scheme code lists.
4226
4227 `~K'
4228 Same as `~?.'
4229
4230 `~!'
4231 Flushes the output if format DESTINATION is a port.
4232
4233 `~_'
4234 Print a `#\space' character
4235 `~N_'
4236 print N `#\space' characters.
4237
4238 `~/'
4239 Print a `#\tab' character
4240 `~N/'
4241 print N `#\tab' characters.
4242
4243 `~NC'
4244 Takes N as an integer representation for a character. No arguments
4245 are consumed. N is converted to a character by `integer->char'. N
4246 must be a positive decimal number.
4247
4248 `~:S'
4249 Print out readproof. Prints out internal objects represented as
4250 `#<...>' as strings `"#<...>"' so that the format output can always
4251 be processed by `read'.
4252
4253 `~:A'
4254 Print out readproof. Prints out internal objects represented as
4255 `#<...>' as strings `"#<...>"' so that the format output can always
4256 be processed by `read'.
4257
4258 `~Q'
4259 Prints information and a copyright notice on the format
4260 implementation.
4261 `~:Q'
4262 prints format version.
4263
4264 `~F, ~E, ~G, ~$'
4265 may also print number strings, i.e. passing a number as a string
4266 and format it accordingly.
4267
4268 *** Configuration Variables
4269
4270 The format module exports some configuration variables to suit the
4271 systems and users needs. There should be no modification necessary for
4272 the configuration that comes with Guile. Format detects automatically
4273 if the running scheme system implements floating point numbers and
4274 complex numbers.
4275
4276 format:symbol-case-conv
4277 Symbols are converted by `symbol->string' so the case type of the
4278 printed symbols is implementation dependent.
4279 `format:symbol-case-conv' is a one arg closure which is either
4280 `#f' (no conversion), `string-upcase', `string-downcase' or
4281 `string-capitalize'. (default `#f')
4282
4283 format:iobj-case-conv
4284 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
4285 implementation internal objects. (default `#f')
4286
4287 format:expch
4288 The character prefixing the exponent value in `~E' printing.
4289 (default `#\E')
4290
4291 *** Compatibility With Other Format Implementations
4292
4293 SLIB format 2.x:
4294 See `format.doc'.
4295
4296 SLIB format 1.4:
4297 Downward compatible except for padding support and `~A', `~S',
4298 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
4299 `printf' padding support which is completely replaced by the CL
4300 `format' padding style.
4301
4302 MIT C-Scheme 7.1:
4303 Downward compatible except for `~', which is not documented
4304 (ignores all characters inside the format string up to a newline
4305 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
4306 numerical and variable parameters and `:/@' modifiers in the CL
4307 sense).
4308
4309 Elk 1.5/2.0:
4310 Downward compatible except for `~A' and `~S' which print in
4311 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
4312 directive parameters or modifiers)).
4313
4314 Scheme->C 01nov91:
4315 Downward compatible except for an optional destination parameter:
4316 S2C accepts a format call without a destination which returns a
4317 formatted string. This is equivalent to a #f destination in S2C.
4318 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
4319 parameters or modifiers)).
4320
4321
4322 ** Changes to string-handling functions.
4323
4324 These functions were added to support the (ice-9 format) module, above.
4325
4326 *** New function: string-upcase STRING
4327 *** New function: string-downcase STRING
4328
4329 These are non-destructive versions of the existing string-upcase! and
4330 string-downcase! functions.
4331
4332 *** New function: string-capitalize! STRING
4333 *** New function: string-capitalize STRING
4334
4335 These functions convert the first letter of each word in the string to
4336 upper case. Thus:
4337
4338 (string-capitalize "howdy there")
4339 => "Howdy There"
4340
4341 As with the other functions, string-capitalize! modifies the string in
4342 place, while string-capitalize returns a modified copy of its argument.
4343
4344 *** New function: string-ci->symbol STRING
4345
4346 Return a symbol whose name is STRING, but having the same case as if
4347 the symbol had be read by `read'.
4348
4349 Guile can be configured to be sensitive or insensitive to case
4350 differences in Scheme identifiers. If Guile is case-insensitive, all
4351 symbols are converted to lower case on input. The `string-ci->symbol'
4352 function returns a symbol whose name in STRING, transformed as Guile
4353 would if STRING were input.
4354
4355 *** New function: substring-move! STRING1 START END STRING2 START
4356
4357 Copy the substring of STRING1 from START (inclusive) to END
4358 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
4359 string, and the source and destination areas may overlap; in all
4360 cases, the function behaves as if all the characters were copied
4361 simultanously.
4362
4363 *** Extended functions: substring-move-left! substring-move-right!
4364
4365 These functions now correctly copy arbitrarily overlapping substrings;
4366 they are both synonyms for substring-move!.
4367
4368
4369 ** New module (ice-9 getopt-long), with the function `getopt-long'.
4370
4371 getopt-long is a function for parsing command-line arguments in a
4372 manner consistent with other GNU programs.
4373
4374 (getopt-long ARGS GRAMMAR)
4375 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
4376
4377 ARGS should be a list of strings. Its first element should be the
4378 name of the program; subsequent elements should be the arguments
4379 that were passed to the program on the command line. The
4380 `program-arguments' procedure returns a list of this form.
4381
4382 GRAMMAR is a list of the form:
4383 ((OPTION (PROPERTY VALUE) ...) ...)
4384
4385 Each OPTION should be a symbol. `getopt-long' will accept a
4386 command-line option named `--OPTION'.
4387 Each option can have the following (PROPERTY VALUE) pairs:
4388
4389 (single-char CHAR) --- Accept `-CHAR' as a single-character
4390 equivalent to `--OPTION'. This is how to specify traditional
4391 Unix-style flags.
4392 (required? BOOL) --- If BOOL is true, the option is required.
4393 getopt-long will raise an error if it is not found in ARGS.
4394 (value BOOL) --- If BOOL is #t, the option accepts a value; if
4395 it is #f, it does not; and if it is the symbol
4396 `optional', the option may appear in ARGS with or
4397 without a value.
4398 (predicate FUNC) --- If the option accepts a value (i.e. you
4399 specified `(value #t)' for this option), then getopt
4400 will apply FUNC to the value, and throw an exception
4401 if it returns #f. FUNC should be a procedure which
4402 accepts a string and returns a boolean value; you may
4403 need to use quasiquotes to get it into GRAMMAR.
4404
4405 The (PROPERTY VALUE) pairs may occur in any order, but each
4406 property may occur only once. By default, options do not have
4407 single-character equivalents, are not required, and do not take
4408 values.
4409
4410 In ARGS, single-character options may be combined, in the usual
4411 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
4412 accepts values, then it must be the last option in the
4413 combination; the value is the next argument. So, for example, using
4414 the following grammar:
4415 ((apples (single-char #\a))
4416 (blimps (single-char #\b) (value #t))
4417 (catalexis (single-char #\c) (value #t)))
4418 the following argument lists would be acceptable:
4419 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
4420 for "blimps" and "catalexis")
4421 ("-ab" "bang" "-c" "couth") (same)
4422 ("-ac" "couth" "-b" "bang") (same)
4423 ("-abc" "couth" "bang") (an error, since `-b' is not the
4424 last option in its combination)
4425
4426 If an option's value is optional, then `getopt-long' decides
4427 whether it has a value by looking at what follows it in ARGS. If
4428 the next element is a string, and it does not appear to be an
4429 option itself, then that string is the option's value.
4430
4431 The value of a long option can appear as the next element in ARGS,
4432 or it can follow the option name, separated by an `=' character.
4433 Thus, using the same grammar as above, the following argument lists
4434 are equivalent:
4435 ("--apples" "Braeburn" "--blimps" "Goodyear")
4436 ("--apples=Braeburn" "--blimps" "Goodyear")
4437 ("--blimps" "Goodyear" "--apples=Braeburn")
4438
4439 If the option "--" appears in ARGS, argument parsing stops there;
4440 subsequent arguments are returned as ordinary arguments, even if
4441 they resemble options. So, in the argument list:
4442 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
4443 `getopt-long' will recognize the `apples' option as having the
4444 value "Granny Smith", but it will not recognize the `blimp'
4445 option; it will return the strings "--blimp" and "Goodyear" as
4446 ordinary argument strings.
4447
4448 The `getopt-long' function returns the parsed argument list as an
4449 assocation list, mapping option names --- the symbols from GRAMMAR
4450 --- onto their values, or #t if the option does not accept a value.
4451 Unused options do not appear in the alist.
4452
4453 All arguments that are not the value of any option are returned
4454 as a list, associated with the empty list.
4455
4456 `getopt-long' throws an exception if:
4457 - it finds an unrecognized option in ARGS
4458 - a required option is omitted
4459 - an option that requires an argument doesn't get one
4460 - an option that doesn't accept an argument does get one (this can
4461 only happen using the long option `--opt=value' syntax)
4462 - an option predicate fails
4463
4464 So, for example:
4465
4466 (define grammar
4467 `((lockfile-dir (required? #t)
4468 (value #t)
4469 (single-char #\k)
4470 (predicate ,file-is-directory?))
4471 (verbose (required? #f)
4472 (single-char #\v)
4473 (value #f))
4474 (x-includes (single-char #\x))
4475 (rnet-server (single-char #\y)
4476 (predicate ,string?))))
4477
4478 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
4479 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
4480 grammar)
4481 => ((() "foo1" "-fred" "foo2" "foo3")
4482 (rnet-server . "lamprod")
4483 (x-includes . "/usr/include")
4484 (lockfile-dir . "/tmp")
4485 (verbose . #t))
4486
4487 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
4488
4489 It will be removed in a few releases.
4490
4491 ** New syntax: lambda*
4492 ** New syntax: define*
4493 ** New syntax: define*-public
4494 ** New syntax: defmacro*
4495 ** New syntax: defmacro*-public
4496 Guile now supports optional arguments.
4497
4498 `lambda*', `define*', `define*-public', `defmacro*' and
4499 `defmacro*-public' are identical to the non-* versions except that
4500 they use an extended type of parameter list that has the following BNF
4501 syntax (parentheses are literal, square brackets indicate grouping,
4502 and `*', `+' and `?' have the usual meaning):
4503
4504 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
4505 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
4506 [[#&rest identifier]|[. identifier]]? ) | [identifier]
4507
4508 ext-var-decl ::= identifier | ( identifier expression )
4509
4510 The semantics are best illustrated with the following documentation
4511 and examples for `lambda*':
4512
4513 lambda* args . body
4514 lambda extended for optional and keyword arguments
4515
4516 lambda* creates a procedure that takes optional arguments. These
4517 are specified by putting them inside brackets at the end of the
4518 paramater list, but before any dotted rest argument. For example,
4519 (lambda* (a b #&optional c d . e) '())
4520 creates a procedure with fixed arguments a and b, optional arguments c
4521 and d, and rest argument e. If the optional arguments are omitted
4522 in a call, the variables for them are unbound in the procedure. This
4523 can be checked with the bound? macro.
4524
4525 lambda* can also take keyword arguments. For example, a procedure
4526 defined like this:
4527 (lambda* (#&key xyzzy larch) '())
4528 can be called with any of the argument lists (#:xyzzy 11)
4529 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
4530 are given as keywords are bound to values.
4531
4532 Optional and keyword arguments can also be given default values
4533 which they take on when they are not present in a call, by giving a
4534 two-item list in place of an optional argument, for example in:
4535 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
4536 foo is a fixed argument, bar is an optional argument with default
4537 value 42, and baz is a keyword argument with default value 73.
4538 Default value expressions are not evaluated unless they are needed
4539 and until the procedure is called.
4540
4541 lambda* now supports two more special parameter list keywords.
4542
4543 lambda*-defined procedures now throw an error by default if a
4544 keyword other than one of those specified is found in the actual
4545 passed arguments. However, specifying #&allow-other-keys
4546 immediately after the kyword argument declarations restores the
4547 previous behavior of ignoring unknown keywords. lambda* also now
4548 guarantees that if the same keyword is passed more than once, the
4549 last one passed is the one that takes effect. For example,
4550 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
4551 #:heads 37 #:tails 42 #:heads 99)
4552 would result in (99 47) being displayed.
4553
4554 #&rest is also now provided as a synonym for the dotted syntax rest
4555 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
4556 all respects to lambda*. This is provided for more similarity to DSSSL,
4557 MIT-Scheme and Kawa among others, as well as for refugees from other
4558 Lisp dialects.
4559
4560 Further documentation may be found in the optargs.scm file itself.
4561
4562 The optional argument module also exports the macros `let-optional',
4563 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
4564 are not documented here because they may be removed in the future, but
4565 full documentation is still available in optargs.scm.
4566
4567 ** New syntax: and-let*
4568 Guile now supports the `and-let*' form, described in the draft SRFI-2.
4569
4570 Syntax: (land* (<clause> ...) <body> ...)
4571 Each <clause> should have one of the following forms:
4572 (<variable> <expression>)
4573 (<expression>)
4574 <bound-variable>
4575 Each <variable> or <bound-variable> should be an identifier. Each
4576 <expression> should be a valid expression. The <body> should be a
4577 possibly empty sequence of expressions, like the <body> of a
4578 lambda form.
4579
4580 Semantics: A LAND* expression is evaluated by evaluating the
4581 <expression> or <bound-variable> of each of the <clause>s from
4582 left to right. The value of the first <expression> or
4583 <bound-variable> that evaluates to a false value is returned; the
4584 remaining <expression>s and <bound-variable>s are not evaluated.
4585 The <body> forms are evaluated iff all the <expression>s and
4586 <bound-variable>s evaluate to true values.
4587
4588 The <expression>s and the <body> are evaluated in an environment
4589 binding each <variable> of the preceding (<variable> <expression>)
4590 clauses to the value of the <expression>. Later bindings
4591 shadow earlier bindings.
4592
4593 Guile's and-let* macro was contributed by Michael Livshin.
4594
4595 ** New sorting functions
4596
4597 *** New function: sorted? SEQUENCE LESS?
4598 Returns `#t' when the sequence argument is in non-decreasing order
4599 according to LESS? (that is, there is no adjacent pair `... x y
4600 ...' for which `(less? y x)').
4601
4602 Returns `#f' when the sequence contains at least one out-of-order
4603 pair. It is an error if the sequence is neither a list nor a
4604 vector.
4605
4606 *** New function: merge LIST1 LIST2 LESS?
4607 LIST1 and LIST2 are sorted lists.
4608 Returns the sorted list of all elements in LIST1 and LIST2.
4609
4610 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
4611 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
4612 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
4613 (Here "<" should read "comes before".)
4614
4615 *** New procedure: merge! LIST1 LIST2 LESS?
4616 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
4617 the result. If the code is compiled, and LESS? constructs no new
4618 pairs, no pairs at all will be allocated. The first pair of the
4619 result will be either the first pair of LIST1 or the first pair of
4620 LIST2.
4621
4622 *** New function: sort SEQUENCE LESS?
4623 Accepts either a list or a vector, and returns a new sequence
4624 which is sorted. The new sequence is the same type as the input.
4625 Always `(sorted? (sort sequence less?) less?)'. The original
4626 sequence is not altered in any way. The new sequence shares its
4627 elements with the old one; no elements are copied.
4628
4629 *** New procedure: sort! SEQUENCE LESS
4630 Returns its sorted result in the original boxes. No new storage is
4631 allocated at all. Proper usage: (set! slist (sort! slist <))
4632
4633 *** New function: stable-sort SEQUENCE LESS?
4634 Similar to `sort' but stable. That is, if "equal" elements are
4635 ordered a < b in the original sequence, they will have the same order
4636 in the result.
4637
4638 *** New function: stable-sort! SEQUENCE LESS?
4639 Similar to `sort!' but stable.
4640 Uses temporary storage when sorting vectors.
4641
4642 *** New functions: sort-list, sort-list!
4643 Added for compatibility with scsh.
4644
4645 ** New built-in random number support
4646
4647 *** New function: random N [STATE]
4648 Accepts a positive integer or real N and returns a number of the
4649 same type between zero (inclusive) and N (exclusive). The values
4650 returned have a uniform distribution.
4651
4652 The optional argument STATE must be of the type produced by
4653 `copy-random-state' or `seed->random-state'. It defaults to the value
4654 of the variable `*random-state*'. This object is used to maintain the
4655 state of the pseudo-random-number generator and is altered as a side
4656 effect of the `random' operation.
4657
4658 *** New variable: *random-state*
4659 Holds a data structure that encodes the internal state of the
4660 random-number generator that `random' uses by default. The nature
4661 of this data structure is implementation-dependent. It may be
4662 printed out and successfully read back in, but may or may not
4663 function correctly as a random-number state object in another
4664 implementation.
4665
4666 *** New function: copy-random-state [STATE]
4667 Returns a new object of type suitable for use as the value of the
4668 variable `*random-state*' and as a second argument to `random'.
4669 If argument STATE is given, a copy of it is returned. Otherwise a
4670 copy of `*random-state*' is returned.
4671
4672 *** New function: seed->random-state SEED
4673 Returns a new object of type suitable for use as the value of the
4674 variable `*random-state*' and as a second argument to `random'.
4675 SEED is a string or a number. A new state is generated and
4676 initialized using SEED.
4677
4678 *** New function: random:uniform [STATE]
4679 Returns an uniformly distributed inexact real random number in the
4680 range between 0 and 1.
4681
4682 *** New procedure: random:solid-sphere! VECT [STATE]
4683 Fills VECT with inexact real random numbers the sum of whose
4684 squares is less than 1.0. Thinking of VECT as coordinates in
4685 space of dimension N = `(vector-length VECT)', the coordinates are
4686 uniformly distributed within the unit N-shere. The sum of the
4687 squares of the numbers is returned. VECT can be either a vector
4688 or a uniform vector of doubles.
4689
4690 *** New procedure: random:hollow-sphere! VECT [STATE]
4691 Fills VECT with inexact real random numbers the sum of whose squares
4692 is equal to 1.0. Thinking of VECT as coordinates in space of
4693 dimension n = `(vector-length VECT)', the coordinates are uniformly
4694 distributed over the surface of the unit n-shere. VECT can be either
4695 a vector or a uniform vector of doubles.
4696
4697 *** New function: random:normal [STATE]
4698 Returns an inexact real in a normal distribution with mean 0 and
4699 standard deviation 1. For a normal distribution with mean M and
4700 standard deviation D use `(+ M (* D (random:normal)))'.
4701
4702 *** New procedure: random:normal-vector! VECT [STATE]
4703 Fills VECT with inexact real random numbers which are independent and
4704 standard normally distributed (i.e., with mean 0 and variance 1).
4705 VECT can be either a vector or a uniform vector of doubles.
4706
4707 *** New function: random:exp STATE
4708 Returns an inexact real in an exponential distribution with mean 1.
4709 For an exponential distribution with mean U use (* U (random:exp)).
4710
4711 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
4712
4713 These functions now operate on numbers in the range of a C unsigned
4714 long.
4715
4716 These functions used to operate on numbers in the range of a C signed
4717 long; however, this seems inappropriate, because Guile integers don't
4718 overflow.
4719
4720 ** New function: make-guardian
4721 This is an implementation of guardians as described in
4722 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
4723 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
4724 Programming Language Design and Implementation, June 1993
4725 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
4726
4727 ** New functions: delq1!, delv1!, delete1!
4728 These procedures behave similar to delq! and friends but delete only
4729 one object if at all.
4730
4731 ** New function: unread-string STRING PORT
4732 Unread STRING to PORT, that is, push it back onto the port so that
4733 next read operation will work on the pushed back characters.
4734
4735 ** unread-char can now be called multiple times
4736 If unread-char is called multiple times, the unread characters will be
4737 read again in last-in first-out order.
4738
4739 ** the procedures uniform-array-read! and uniform-array-write! now
4740 work on any kind of port, not just ports which are open on a file.
4741
4742 ** Now 'l' in a port mode requests line buffering.
4743
4744 ** The procedure truncate-file now works on string ports as well
4745 as file ports. If the size argument is omitted, the current
4746 file position is used.
4747
4748 ** new procedure: seek PORT/FDES OFFSET WHENCE
4749 The arguments are the same as for the old fseek procedure, but it
4750 works on string ports as well as random-access file ports.
4751
4752 ** the fseek procedure now works on string ports, since it has been
4753 redefined using seek.
4754
4755 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
4756 size is not supplied.
4757
4758 ** the newline procedure no longer flushes the port if it's not
4759 line-buffered: previously it did if it was the current output port.
4760
4761 ** open-pipe and close-pipe are no longer primitive procedures, but
4762 an emulation can be obtained using `(use-modules (ice-9 popen))'.
4763
4764 ** the freopen procedure has been removed.
4765
4766 ** new procedure: drain-input PORT
4767 Drains PORT's read buffers (including any pushed-back characters)
4768 and returns the contents as a single string.
4769
4770 ** New function: map-in-order PROC LIST1 LIST2 ...
4771 Version of `map' which guarantees that the procedure is applied to the
4772 lists in serial order.
4773
4774 ** Renamed `serial-array-copy!' and `serial-array-map!' to
4775 `array-copy-in-order!' and `array-map-in-order!'. The old names are
4776 now obsolete and will go away in release 1.5.
4777
4778 ** New syntax: collect BODY1 ...
4779 Version of `begin' which returns a list of the results of the body
4780 forms instead of the result of the last body form. In contrast to
4781 `begin', `collect' allows an empty body.
4782
4783 ** New functions: read-history FILENAME, write-history FILENAME
4784 Read/write command line history from/to file. Returns #t on success
4785 and #f if an error occured.
4786
4787 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
4788
4789 These procedures return a list of definitions available in the specified
4790 argument, a relative module reference. In the case of no argument,
4791 `(current-module)' is now consulted for definitions to return, instead
4792 of simply returning #f, the former behavior.
4793
4794 ** The #/ syntax for lists is no longer supported.
4795
4796 Earlier versions of Scheme accepted this syntax, but printed a
4797 warning.
4798
4799 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
4800
4801 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
4802 modules.
4803
4804 * Changes to the gh_ interface
4805
4806 ** gh_scm2doubles
4807
4808 Now takes a second argument which is the result array. If this
4809 pointer is NULL, a new array is malloced (the old behaviour).
4810
4811 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
4812 gh_scm2shorts, gh_scm2longs, gh_scm2floats
4813
4814 New functions.
4815
4816 * Changes to the scm_ interface
4817
4818 ** Function: scm_make_named_hook (char* name, int n_args)
4819
4820 Creates a hook in the same way as make-hook above but also
4821 binds a variable named NAME to it.
4822
4823 This is the typical way of creating a hook from C code.
4824
4825 Currently, the variable is created in the "current" module. This
4826 might change when we get the new module system.
4827
4828 ** The smob interface
4829
4830 The interface for creating smobs has changed. For documentation, see
4831 data-rep.info (made from guile-core/doc/data-rep.texi).
4832
4833 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
4834
4835 >>> This function will be removed in 1.3.4. <<<
4836
4837 It is replaced by:
4838
4839 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
4840 This function adds a new smob type, named NAME, with instance size
4841 SIZE to the system. The return value is a tag that is used in
4842 creating instances of the type. If SIZE is 0, then no memory will
4843 be allocated when instances of the smob are created, and nothing
4844 will be freed by the default free function.
4845
4846 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
4847 This function sets the smob marking procedure for the smob type
4848 specified by the tag TC. TC is the tag returned by
4849 `scm_make_smob_type'.
4850
4851 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
4852 This function sets the smob freeing procedure for the smob type
4853 specified by the tag TC. TC is the tag returned by
4854 `scm_make_smob_type'.
4855
4856 *** Function: void scm_set_smob_print (tc, print)
4857
4858 - Function: void scm_set_smob_print (long tc,
4859 scm_sizet (*print) (SCM,
4860 SCM,
4861 scm_print_state *))
4862
4863 This function sets the smob printing procedure for the smob type
4864 specified by the tag TC. TC is the tag returned by
4865 `scm_make_smob_type'.
4866
4867 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
4868 This function sets the smob equality-testing predicate for the
4869 smob type specified by the tag TC. TC is the tag returned by
4870 `scm_make_smob_type'.
4871
4872 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
4873 Make VALUE contain a smob instance of the type with type code TC and
4874 smob data DATA. VALUE must be previously declared as C type `SCM'.
4875
4876 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
4877 This macro expands to a block of code that creates a smob instance
4878 of the type with type code TC and smob data DATA, and returns that
4879 `SCM' value. It should be the last piece of code in a block.
4880
4881 ** The interfaces for using I/O ports and implementing port types
4882 (ptobs) have changed significantly. The new interface is based on
4883 shared access to buffers and a new set of ptob procedures.
4884
4885 *** scm_newptob has been removed
4886
4887 It is replaced by:
4888
4889 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
4890
4891 - Function: SCM scm_make_port_type (char *type_name,
4892 int (*fill_buffer) (SCM port),
4893 void (*write_flush) (SCM port));
4894
4895 Similarly to the new smob interface, there is a set of function
4896 setters by which the user can customize the behaviour of his port
4897 type. See ports.h (scm_set_port_XXX).
4898
4899 ** scm_strport_to_string: New function: creates a new string from
4900 a string port's buffer.
4901
4902 ** Plug in interface for random number generators
4903 The variable `scm_the_rng' in random.c contains a value and three
4904 function pointers which together define the current random number
4905 generator being used by the Scheme level interface and the random
4906 number library functions.
4907
4908 The user is free to replace the default generator with the generator
4909 of his own choice.
4910
4911 *** Variable: size_t scm_the_rng.rstate_size
4912 The size of the random state type used by the current RNG
4913 measured in chars.
4914
4915 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
4916 Given the random STATE, return 32 random bits.
4917
4918 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
4919 Seed random state STATE using string S of length N.
4920
4921 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
4922 Given random state STATE, return a malloced copy.
4923
4924 ** Default RNG
4925 The default RNG is the MWC (Multiply With Carry) random number
4926 generator described by George Marsaglia at the Department of
4927 Statistics and Supercomputer Computations Research Institute, The
4928 Florida State University (http://stat.fsu.edu/~geo).
4929
4930 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
4931 passes all tests in the DIEHARD test suite
4932 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
4933 costs one multiply and one add on platforms which either supports long
4934 longs (gcc does this on most systems) or have 64 bit longs. The cost
4935 is four multiply on other systems but this can be optimized by writing
4936 scm_i_uniform32 in assembler.
4937
4938 These functions are provided through the scm_the_rng interface for use
4939 by libguile and the application.
4940
4941 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
4942 Given the random STATE, return 32 random bits.
4943 Don't use this function directly. Instead go through the plugin
4944 interface (see "Plug in interface" above).
4945
4946 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
4947 Initialize STATE using SEED of length N.
4948
4949 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
4950 Return a malloc:ed copy of STATE. This function can easily be re-used
4951 in the interfaces to other RNGs.
4952
4953 ** Random number library functions
4954 These functions use the current RNG through the scm_the_rng interface.
4955 It might be a good idea to use these functions from your C code so
4956 that only one random generator is used by all code in your program.
4957
4958 The default random state is stored in:
4959
4960 *** Variable: SCM scm_var_random_state
4961 Contains the vcell of the Scheme variable "*random-state*" which is
4962 used as default state by all random number functions in the Scheme
4963 level interface.
4964
4965 Example:
4966
4967 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
4968
4969 *** Function: scm_rstate *scm_c_default_rstate (void)
4970 This is a convenience function which returns the value of
4971 scm_var_random_state. An error message is generated if this value
4972 isn't a random state.
4973
4974 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
4975 Make a new random state from the string SEED of length LENGTH.
4976
4977 It is generally not a good idea to use multiple random states in a
4978 program. While subsequent random numbers generated from one random
4979 state are guaranteed to be reasonably independent, there is no such
4980 guarantee for numbers generated from different random states.
4981
4982 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
4983 Return 32 random bits.
4984
4985 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
4986 Return a sample from the uniform(0,1) distribution.
4987
4988 *** Function: double scm_c_normal01 (scm_rstate *STATE)
4989 Return a sample from the normal(0,1) distribution.
4990
4991 *** Function: double scm_c_exp1 (scm_rstate *STATE)
4992 Return a sample from the exp(1) distribution.
4993
4994 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
4995 Return a sample from the discrete uniform(0,M) distribution.
4996
4997 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
4998 Return a sample from the discrete uniform(0,M) distribution.
4999 M must be a bignum object. The returned value may be an INUM.
5000
5001
5002 \f
5003 Changes in Guile 1.3 (released Monday, October 19, 1998):
5004
5005 * Changes to the distribution
5006
5007 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
5008 To avoid conflicts, programs should name environment variables after
5009 themselves, except when there's a common practice establishing some
5010 other convention.
5011
5012 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
5013 giving the former precedence, and printing a warning message if the
5014 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
5015
5016 ** The header files related to multi-byte characters have been removed.
5017 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
5018 which referred to these explicitly will probably need to be rewritten,
5019 since the support for the variant string types has been removed; see
5020 below.
5021
5022 ** The header files append.h and sequences.h have been removed. These
5023 files implemented non-R4RS operations which would encourage
5024 non-portable programming style and less easy-to-read code.
5025
5026 * Changes to the stand-alone interpreter
5027
5028 ** New procedures have been added to implement a "batch mode":
5029
5030 *** Function: batch-mode?
5031
5032 Returns a boolean indicating whether the interpreter is in batch
5033 mode.
5034
5035 *** Function: set-batch-mode?! ARG
5036
5037 If ARG is true, switches the interpreter to batch mode. The `#f'
5038 case has not been implemented.
5039
5040 ** Guile now provides full command-line editing, when run interactively.
5041 To use this feature, you must have the readline library installed.
5042 The Guile build process will notice it, and automatically include
5043 support for it.
5044
5045 The readline library is available via anonymous FTP from any GNU
5046 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
5047
5048 ** the-last-stack is now a fluid.
5049
5050 * Changes to the procedure for linking libguile with your programs
5051
5052 ** You can now use the `guile-config' utility to build programs that use Guile.
5053
5054 Guile now includes a command-line utility called `guile-config', which
5055 can provide information about how to compile and link programs that
5056 use Guile.
5057
5058 *** `guile-config compile' prints any C compiler flags needed to use Guile.
5059 You should include this command's output on the command line you use
5060 to compile C or C++ code that #includes the Guile header files. It's
5061 usually just a `-I' flag to help the compiler find the Guile headers.
5062
5063
5064 *** `guile-config link' prints any linker flags necessary to link with Guile.
5065
5066 This command writes to its standard output a list of flags which you
5067 must pass to the linker to link your code against the Guile library.
5068 The flags include '-lguile' itself, any other libraries the Guile
5069 library depends upon, and any `-L' flags needed to help the linker
5070 find those libraries.
5071
5072 For example, here is a Makefile rule that builds a program named 'foo'
5073 from the object files ${FOO_OBJECTS}, and links them against Guile:
5074
5075 foo: ${FOO_OBJECTS}
5076 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
5077
5078 Previous Guile releases recommended that you use autoconf to detect
5079 which of a predefined set of libraries were present on your system.
5080 It is more robust to use `guile-config', since it records exactly which
5081 libraries the installed Guile library requires.
5082
5083 This was originally called `build-guile', but was renamed to
5084 `guile-config' before Guile 1.3 was released, to be consistent with
5085 the analogous script for the GTK+ GUI toolkit, which is called
5086 `gtk-config'.
5087
5088
5089 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
5090
5091 If you are using the GNU autoconf package to configure your program,
5092 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
5093 (described above) and gather the necessary values for use in your
5094 Makefiles.
5095
5096 The GUILE_FLAGS macro expands to configure script code which runs the
5097 `guile-config' script, to find out where Guile's header files and
5098 libraries are installed. It sets two variables, marked for
5099 substitution, as by AC_SUBST.
5100
5101 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
5102 code that uses Guile header files. This is almost always just a
5103 -I flag.
5104
5105 GUILE_LDFLAGS --- flags to pass to the linker to link a
5106 program against Guile. This includes `-lguile' for the Guile
5107 library itself, any libraries that Guile itself requires (like
5108 -lqthreads), and so on. It may also include a -L flag to tell the
5109 compiler where to find the libraries.
5110
5111 GUILE_FLAGS is defined in the file guile.m4, in the top-level
5112 directory of the Guile distribution. You can copy it into your
5113 package's aclocal.m4 file, and then use it in your configure.in file.
5114
5115 If you are using the `aclocal' program, distributed with GNU automake,
5116 to maintain your aclocal.m4 file, the Guile installation process
5117 installs guile.m4 where aclocal will find it. All you need to do is
5118 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
5119 this will copy the definition of GUILE_FLAGS into your aclocal.m4
5120 file.
5121
5122
5123 * Changes to Scheme functions and syntax
5124
5125 ** Multi-byte strings have been removed, as have multi-byte and wide
5126 ports. We felt that these were the wrong approach to
5127 internationalization support.
5128
5129 ** New function: readline [PROMPT]
5130 Read a line from the terminal, and allow the user to edit it,
5131 prompting with PROMPT. READLINE provides a large set of Emacs-like
5132 editing commands, lets the user recall previously typed lines, and
5133 works on almost every kind of terminal, including dumb terminals.
5134
5135 READLINE assumes that the cursor is at the beginning of the line when
5136 it is invoked. Thus, you can't print a prompt yourself, and then call
5137 READLINE; you need to package up your prompt as a string, pass it to
5138 the function, and let READLINE print the prompt itself. This is
5139 because READLINE needs to know the prompt's screen width.
5140
5141 For Guile to provide this function, you must have the readline
5142 library, version 2.1 or later, installed on your system. Readline is
5143 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
5144 any GNU mirror site.
5145
5146 See also ADD-HISTORY function.
5147
5148 ** New function: add-history STRING
5149 Add STRING as the most recent line in the history used by the READLINE
5150 command. READLINE does not add lines to the history itself; you must
5151 call ADD-HISTORY to make previous input available to the user.
5152
5153 ** The behavior of the read-line function has changed.
5154
5155 This function now uses standard C library functions to read the line,
5156 for speed. This means that it doesn not respect the value of
5157 scm-line-incrementors; it assumes that lines are delimited with
5158 #\newline.
5159
5160 (Note that this is read-line, the function that reads a line of text
5161 from a port, not readline, the function that reads a line from a
5162 terminal, providing full editing capabilities.)
5163
5164 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
5165
5166 This module provides some simple argument parsing. It exports one
5167 function:
5168
5169 Function: getopt-gnu-style ARG-LS
5170 Parse a list of program arguments into an alist of option
5171 descriptions.
5172
5173 Each item in the list of program arguments is examined to see if
5174 it meets the syntax of a GNU long-named option. An argument like
5175 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
5176 returned alist, where MUMBLE is a keyword object with the same
5177 name as the argument. An argument like `--MUMBLE=FROB' produces
5178 an element of the form (MUMBLE . FROB), where FROB is a string.
5179
5180 As a special case, the returned alist also contains a pair whose
5181 car is the symbol `rest'. The cdr of this pair is a list
5182 containing all the items in the argument list that are not options
5183 of the form mentioned above.
5184
5185 The argument `--' is treated specially: all items in the argument
5186 list appearing after such an argument are not examined, and are
5187 returned in the special `rest' list.
5188
5189 This function does not parse normal single-character switches.
5190 You will need to parse them out of the `rest' list yourself.
5191
5192 ** The read syntax for byte vectors and short vectors has changed.
5193
5194 Instead of #bytes(...), write #y(...).
5195
5196 Instead of #short(...), write #h(...).
5197
5198 This may seem nutty, but, like the other uniform vectors, byte vectors
5199 and short vectors want to have the same print and read syntax (and,
5200 more basic, want to have read syntax!). Changing the read syntax to
5201 use multiple characters after the hash sign breaks with the
5202 conventions used in R5RS and the conventions used for the other
5203 uniform vectors. It also introduces complexity in the current reader,
5204 both on the C and Scheme levels. (The Right solution is probably to
5205 change the syntax and prototypes for uniform vectors entirely.)
5206
5207
5208 ** The new module (ice-9 session) provides useful interactive functions.
5209
5210 *** New procedure: (apropos REGEXP OPTION ...)
5211
5212 Display a list of top-level variables whose names match REGEXP, and
5213 the modules they are imported from. Each OPTION should be one of the
5214 following symbols:
5215
5216 value --- Show the value of each matching variable.
5217 shadow --- Show bindings shadowed by subsequently imported modules.
5218 full --- Same as both `shadow' and `value'.
5219
5220 For example:
5221
5222 guile> (apropos "trace" 'full)
5223 debug: trace #<procedure trace args>
5224 debug: untrace #<procedure untrace args>
5225 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
5226 the-scm-module: before-backtrace-hook ()
5227 the-scm-module: backtrace #<primitive-procedure backtrace>
5228 the-scm-module: after-backtrace-hook ()
5229 the-scm-module: has-shown-backtrace-hint? #f
5230 guile>
5231
5232 ** There are new functions and syntax for working with macros.
5233
5234 Guile implements macros as a special object type. Any variable whose
5235 top-level binding is a macro object acts as a macro. The macro object
5236 specifies how the expression should be transformed before evaluation.
5237
5238 *** Macro objects now print in a reasonable way, resembling procedures.
5239
5240 *** New function: (macro? OBJ)
5241 True iff OBJ is a macro object.
5242
5243 *** New function: (primitive-macro? OBJ)
5244 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
5245 macro transformers, implemented in eval.c rather than Scheme code.
5246
5247 Why do we have this function?
5248 - For symmetry with procedure? and primitive-procedure?,
5249 - to allow custom print procedures to tell whether a macro is
5250 primitive, and display it differently, and
5251 - to allow compilers and user-written evaluators to distinguish
5252 builtin special forms from user-defined ones, which could be
5253 compiled.
5254
5255 *** New function: (macro-type OBJ)
5256 Return a value indicating what kind of macro OBJ is. Possible return
5257 values are:
5258
5259 The symbol `syntax' --- a macro created by procedure->syntax.
5260 The symbol `macro' --- a macro created by procedure->macro.
5261 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
5262 The boolean #f --- if OBJ is not a macro object.
5263
5264 *** New function: (macro-name MACRO)
5265 Return the name of the macro object MACRO's procedure, as returned by
5266 procedure-name.
5267
5268 *** New function: (macro-transformer MACRO)
5269 Return the transformer procedure for MACRO.
5270
5271 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
5272
5273 Specify a new macro expander to use in the current module. Each
5274 MODULE is a module name, with the same meaning as in the `use-modules'
5275 form; each named module's exported bindings are added to the current
5276 top-level environment. TRANSFORMER is an expression evaluated in the
5277 resulting environment which must yield a procedure to use as the
5278 module's eval transformer: every expression evaluated in this module
5279 is passed to this function, and the result passed to the Guile
5280 interpreter.
5281
5282 *** macro-eval! is removed. Use local-eval instead.
5283
5284 ** Some magic has been added to the printer to better handle user
5285 written printing routines (like record printers, closure printers).
5286
5287 The problem is that these user written routines must have access to
5288 the current `print-state' to be able to handle fancy things like
5289 detection of circular references. These print-states have to be
5290 passed to the builtin printing routines (display, write, etc) to
5291 properly continue the print chain.
5292
5293 We didn't want to change all existing print code so that it
5294 explicitly passes thru a print state in addition to a port. Instead,
5295 we extented the possible values that the builtin printing routines
5296 accept as a `port'. In addition to a normal port, they now also take
5297 a pair of a normal port and a print-state. Printing will go to the
5298 port and the print-state will be used to control the detection of
5299 circular references, etc. If the builtin function does not care for a
5300 print-state, it is simply ignored.
5301
5302 User written callbacks are now called with such a pair as their
5303 `port', but because every function now accepts this pair as a PORT
5304 argument, you don't have to worry about that. In fact, it is probably
5305 safest to not check for these pairs.
5306
5307 However, it is sometimes necessary to continue a print chain on a
5308 different port, for example to get a intermediate string
5309 representation of the printed value, mangle that string somehow, and
5310 then to finally print the mangled string. Use the new function
5311
5312 inherit-print-state OLD-PORT NEW-PORT
5313
5314 for this. It constructs a new `port' that prints to NEW-PORT but
5315 inherits the print-state of OLD-PORT.
5316
5317 ** struct-vtable-offset renamed to vtable-offset-user
5318
5319 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
5320
5321 ** There is now a third optional argument to make-vtable-vtable
5322 (and fourth to make-struct) when constructing new types (vtables).
5323 This argument initializes field vtable-index-printer of the vtable.
5324
5325 ** The detection of circular references has been extended to structs.
5326 That is, a structure that -- in the process of being printed -- prints
5327 itself does not lead to infinite recursion.
5328
5329 ** There is now some basic support for fluids. Please read
5330 "libguile/fluid.h" to find out more. It is accessible from Scheme with
5331 the following functions and macros:
5332
5333 Function: make-fluid
5334
5335 Create a new fluid object. Fluids are not special variables or
5336 some other extension to the semantics of Scheme, but rather
5337 ordinary Scheme objects. You can store them into variables (that
5338 are still lexically scoped, of course) or into any other place you
5339 like. Every fluid has a initial value of `#f'.
5340
5341 Function: fluid? OBJ
5342
5343 Test whether OBJ is a fluid.
5344
5345 Function: fluid-ref FLUID
5346 Function: fluid-set! FLUID VAL
5347
5348 Access/modify the fluid FLUID. Modifications are only visible
5349 within the current dynamic root (that includes threads).
5350
5351 Function: with-fluids* FLUIDS VALUES THUNK
5352
5353 FLUIDS is a list of fluids and VALUES a corresponding list of
5354 values for these fluids. Before THUNK gets called the values are
5355 installed in the fluids and the old values of the fluids are
5356 saved in the VALUES list. When the flow of control leaves THUNK
5357 or reenters it, the values get swapped again. You might think of
5358 this as a `safe-fluid-excursion'. Note that the VALUES list is
5359 modified by `with-fluids*'.
5360
5361 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
5362
5363 The same as `with-fluids*' but with a different syntax. It looks
5364 just like `let', but both FLUID and VALUE are evaluated. Remember,
5365 fluids are not special variables but ordinary objects. FLUID
5366 should evaluate to a fluid.
5367
5368 ** Changes to system call interfaces:
5369
5370 *** close-port, close-input-port and close-output-port now return a
5371 boolean instead of an `unspecified' object. #t means that the port
5372 was successfully closed, while #f means it was already closed. It is
5373 also now possible for these procedures to raise an exception if an
5374 error occurs (some errors from write can be delayed until close.)
5375
5376 *** the first argument to chmod, fcntl, ftell and fseek can now be a
5377 file descriptor.
5378
5379 *** the third argument to fcntl is now optional.
5380
5381 *** the first argument to chown can now be a file descriptor or a port.
5382
5383 *** the argument to stat can now be a port.
5384
5385 *** The following new procedures have been added (most use scsh
5386 interfaces):
5387
5388 *** procedure: close PORT/FD
5389 Similar to close-port (*note close-port: Closing Ports.), but also
5390 works on file descriptors. A side effect of closing a file
5391 descriptor is that any ports using that file descriptor are moved
5392 to a different file descriptor and have their revealed counts set
5393 to zero.
5394
5395 *** procedure: port->fdes PORT
5396 Returns the integer file descriptor underlying PORT. As a side
5397 effect the revealed count of PORT is incremented.
5398
5399 *** procedure: fdes->ports FDES
5400 Returns a list of existing ports which have FDES as an underlying
5401 file descriptor, without changing their revealed counts.
5402
5403 *** procedure: fdes->inport FDES
5404 Returns an existing input port which has FDES as its underlying
5405 file descriptor, if one exists, and increments its revealed count.
5406 Otherwise, returns a new input port with a revealed count of 1.
5407
5408 *** procedure: fdes->outport FDES
5409 Returns an existing output port which has FDES as its underlying
5410 file descriptor, if one exists, and increments its revealed count.
5411 Otherwise, returns a new output port with a revealed count of 1.
5412
5413 The next group of procedures perform a `dup2' system call, if NEWFD
5414 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
5415 duplicated can be supplied as an integer or contained in a port. The
5416 type of value returned varies depending on which procedure is used.
5417
5418 All procedures also have the side effect when performing `dup2' that
5419 any ports using NEWFD are moved to a different file descriptor and have
5420 their revealed counts set to zero.
5421
5422 *** procedure: dup->fdes PORT/FD [NEWFD]
5423 Returns an integer file descriptor.
5424
5425 *** procedure: dup->inport PORT/FD [NEWFD]
5426 Returns a new input port using the new file descriptor.
5427
5428 *** procedure: dup->outport PORT/FD [NEWFD]
5429 Returns a new output port using the new file descriptor.
5430
5431 *** procedure: dup PORT/FD [NEWFD]
5432 Returns a new port if PORT/FD is a port, with the same mode as the
5433 supplied port, otherwise returns an integer file descriptor.
5434
5435 *** procedure: dup->port PORT/FD MODE [NEWFD]
5436 Returns a new port using the new file descriptor. MODE supplies a
5437 mode string for the port (*note open-file: File Ports.).
5438
5439 *** procedure: setenv NAME VALUE
5440 Modifies the environment of the current process, which is also the
5441 default environment inherited by child processes.
5442
5443 If VALUE is `#f', then NAME is removed from the environment.
5444 Otherwise, the string NAME=VALUE is added to the environment,
5445 replacing any existing string with name matching NAME.
5446
5447 The return value is unspecified.
5448
5449 *** procedure: truncate-file OBJ SIZE
5450 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
5451 can be a string containing a file name or an integer file
5452 descriptor or port open for output on the file. The underlying
5453 system calls are `truncate' and `ftruncate'.
5454
5455 The return value is unspecified.
5456
5457 *** procedure: setvbuf PORT MODE [SIZE]
5458 Set the buffering mode for PORT. MODE can be:
5459 `_IONBF'
5460 non-buffered
5461
5462 `_IOLBF'
5463 line buffered
5464
5465 `_IOFBF'
5466 block buffered, using a newly allocated buffer of SIZE bytes.
5467 However if SIZE is zero or unspecified, the port will be made
5468 non-buffered.
5469
5470 This procedure should not be used after I/O has been performed with
5471 the port.
5472
5473 Ports are usually block buffered by default, with a default buffer
5474 size. Procedures e.g., *Note open-file: File Ports, which accept a
5475 mode string allow `0' to be added to request an unbuffered port.
5476
5477 *** procedure: fsync PORT/FD
5478 Copies any unwritten data for the specified output file descriptor
5479 to disk. If PORT/FD is a port, its buffer is flushed before the
5480 underlying file descriptor is fsync'd. The return value is
5481 unspecified.
5482
5483 *** procedure: open-fdes PATH FLAGS [MODES]
5484 Similar to `open' but returns a file descriptor instead of a port.
5485
5486 *** procedure: execle PATH ENV [ARG] ...
5487 Similar to `execl', but the environment of the new process is
5488 specified by ENV, which must be a list of strings as returned by
5489 the `environ' procedure.
5490
5491 This procedure is currently implemented using the `execve' system
5492 call, but we call it `execle' because of its Scheme calling
5493 interface.
5494
5495 *** procedure: strerror ERRNO
5496 Returns the Unix error message corresponding to ERRNO, an integer.
5497
5498 *** procedure: primitive-exit [STATUS]
5499 Terminate the current process without unwinding the Scheme stack.
5500 This is would typically be useful after a fork. The exit status
5501 is STATUS if supplied, otherwise zero.
5502
5503 *** procedure: times
5504 Returns an object with information about real and processor time.
5505 The following procedures accept such an object as an argument and
5506 return a selected component:
5507
5508 `tms:clock'
5509 The current real time, expressed as time units relative to an
5510 arbitrary base.
5511
5512 `tms:utime'
5513 The CPU time units used by the calling process.
5514
5515 `tms:stime'
5516 The CPU time units used by the system on behalf of the
5517 calling process.
5518
5519 `tms:cutime'
5520 The CPU time units used by terminated child processes of the
5521 calling process, whose status has been collected (e.g., using
5522 `waitpid').
5523
5524 `tms:cstime'
5525 Similarly, the CPU times units used by the system on behalf of
5526 terminated child processes.
5527
5528 ** Removed: list-length
5529 ** Removed: list-append, list-append!
5530 ** Removed: list-reverse, list-reverse!
5531
5532 ** array-map renamed to array-map!
5533
5534 ** serial-array-map renamed to serial-array-map!
5535
5536 ** catch doesn't take #f as first argument any longer
5537
5538 Previously, it was possible to pass #f instead of a key to `catch'.
5539 That would cause `catch' to pass a jump buffer object to the procedure
5540 passed as second argument. The procedure could then use this jump
5541 buffer objekt as an argument to throw.
5542
5543 This mechanism has been removed since its utility doesn't motivate the
5544 extra complexity it introduces.
5545
5546 ** The `#/' notation for lists now provokes a warning message from Guile.
5547 This syntax will be removed from Guile in the near future.
5548
5549 To disable the warning message, set the GUILE_HUSH environment
5550 variable to any non-empty value.
5551
5552 ** The newline character now prints as `#\newline', following the
5553 normal Scheme notation, not `#\nl'.
5554
5555 * Changes to the gh_ interface
5556
5557 ** The gh_enter function now takes care of loading the Guile startup files.
5558 gh_enter works by calling scm_boot_guile; see the remarks below.
5559
5560 ** Function: void gh_write (SCM x)
5561
5562 Write the printed representation of the scheme object x to the current
5563 output port. Corresponds to the scheme level `write'.
5564
5565 ** gh_list_length renamed to gh_length.
5566
5567 ** vector handling routines
5568
5569 Several major changes. In particular, gh_vector() now resembles
5570 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
5571 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
5572 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
5573 vector-related gh_ functions have been implemented.
5574
5575 ** pair and list routines
5576
5577 Implemented several of the R4RS pair and list functions that were
5578 missing.
5579
5580 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
5581
5582 New function. Converts double arrays back and forth between Scheme
5583 and C.
5584
5585 * Changes to the scm_ interface
5586
5587 ** The function scm_boot_guile now takes care of loading the startup files.
5588
5589 Guile's primary initialization function, scm_boot_guile, now takes
5590 care of loading `boot-9.scm', in the `ice-9' module, to initialize
5591 Guile, define the module system, and put together some standard
5592 bindings. It also loads `init.scm', which is intended to hold
5593 site-specific initialization code.
5594
5595 Since Guile cannot operate properly until boot-9.scm is loaded, there
5596 is no reason to separate loading boot-9.scm from Guile's other
5597 initialization processes.
5598
5599 This job used to be done by scm_compile_shell_switches, which didn't
5600 make much sense; in particular, it meant that people using Guile for
5601 non-shell-like applications had to jump through hoops to get Guile
5602 initialized properly.
5603
5604 ** The function scm_compile_shell_switches no longer loads the startup files.
5605 Now, Guile always loads the startup files, whenever it is initialized;
5606 see the notes above for scm_boot_guile and scm_load_startup_files.
5607
5608 ** Function: scm_load_startup_files
5609 This new function takes care of loading Guile's initialization file
5610 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
5611 this is always called by the Guile initialization process, it's
5612 probably not too useful to call this yourself, but it's there anyway.
5613
5614 ** The semantics of smob marking have changed slightly.
5615
5616 The smob marking function (the `mark' member of the scm_smobfuns
5617 structure) is no longer responsible for setting the mark bit on the
5618 smob. The generic smob handling code in the garbage collector will
5619 set this bit. The mark function need only ensure that any other
5620 objects the smob refers to get marked.
5621
5622 Note that this change means that the smob's GC8MARK bit is typically
5623 already set upon entry to the mark function. Thus, marking functions
5624 which look like this:
5625
5626 {
5627 if (SCM_GC8MARKP (ptr))
5628 return SCM_BOOL_F;
5629 SCM_SETGC8MARK (ptr);
5630 ... mark objects to which the smob refers ...
5631 }
5632
5633 are now incorrect, since they will return early, and fail to mark any
5634 other objects the smob refers to. Some code in the Guile library used
5635 to work this way.
5636
5637 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
5638
5639 If you have implemented your own I/O port type, by writing the
5640 functions required by the scm_ptobfuns and then calling scm_newptob,
5641 you will need to change your functions slightly.
5642
5643 The functions in a scm_ptobfuns structure now expect the port itself
5644 as their argument; they used to expect the `stream' member of the
5645 port's scm_port_table structure. This allows functions in an
5646 scm_ptobfuns structure to easily access the port's cell (and any flags
5647 it its CAR), and the port's scm_port_table structure.
5648
5649 Guile now passes the I/O port itself as the `port' argument in the
5650 following scm_ptobfuns functions:
5651
5652 int (*free) (SCM port);
5653 int (*fputc) (int, SCM port);
5654 int (*fputs) (char *, SCM port);
5655 scm_sizet (*fwrite) SCM_P ((char *ptr,
5656 scm_sizet size,
5657 scm_sizet nitems,
5658 SCM port));
5659 int (*fflush) (SCM port);
5660 int (*fgetc) (SCM port);
5661 int (*fclose) (SCM port);
5662
5663 The interfaces to the `mark', `print', `equalp', and `fgets' methods
5664 are unchanged.
5665
5666 If you have existing code which defines its own port types, it is easy
5667 to convert your code to the new interface; simply apply SCM_STREAM to
5668 the port argument to yield the value you code used to expect.
5669
5670 Note that since both the port and the stream have the same type in the
5671 C code --- they are both SCM values --- the C compiler will not remind
5672 you if you forget to update your scm_ptobfuns functions.
5673
5674
5675 ** Function: int scm_internal_select (int fds,
5676 SELECT_TYPE *rfds,
5677 SELECT_TYPE *wfds,
5678 SELECT_TYPE *efds,
5679 struct timeval *timeout);
5680
5681 This is a replacement for the `select' function provided by the OS.
5682 It enables I/O blocking and sleeping to happen for one cooperative
5683 thread without blocking other threads. It also avoids busy-loops in
5684 these situations. It is intended that all I/O blocking and sleeping
5685 will finally go through this function. Currently, this function is
5686 only available on systems providing `gettimeofday' and `select'.
5687
5688 ** Function: SCM scm_internal_stack_catch (SCM tag,
5689 scm_catch_body_t body,
5690 void *body_data,
5691 scm_catch_handler_t handler,
5692 void *handler_data)
5693
5694 A new sibling to the other two C level `catch' functions
5695 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
5696 the stack to be saved automatically into the variable `the-last-stack'
5697 (scm_the_last_stack_var) on error. This is necessary if you want to
5698 use advanced error reporting, such as calling scm_display_error and
5699 scm_display_backtrace. (They both take a stack object as argument.)
5700
5701 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
5702 void *body_data,
5703 scm_catch_handler_t handler,
5704 void *handler_data)
5705
5706 Spawns a new thread. It does a job similar to
5707 scm_call_with_new_thread but takes arguments more suitable when
5708 spawning threads from application C code.
5709
5710 ** The hook scm_error_callback has been removed. It was originally
5711 intended as a way for the user to install his own error handler. But
5712 that method works badly since it intervenes between throw and catch,
5713 thereby changing the semantics of expressions like (catch #t ...).
5714 The correct way to do it is to use one of the C level catch functions
5715 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
5716
5717 ** Removed functions:
5718
5719 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
5720 scm_list_reverse, scm_list_reverse_x
5721
5722 ** New macros: SCM_LISTn where n is one of the integers 0-9.
5723
5724 These can be used for pretty list creation from C. The idea is taken
5725 from Erick Gallesio's STk.
5726
5727 ** scm_array_map renamed to scm_array_map_x
5728
5729 ** mbstrings are now removed
5730
5731 This means that the type codes scm_tc7_mb_string and
5732 scm_tc7_mb_substring has been removed.
5733
5734 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
5735
5736 Since we no longer support multi-byte strings, these I/O functions
5737 have been simplified, and renamed. Here are their old names, and
5738 their new names and arguments:
5739
5740 scm_gen_putc -> void scm_putc (int c, SCM port);
5741 scm_gen_puts -> void scm_puts (char *s, SCM port);
5742 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
5743 scm_gen_getc -> void scm_getc (SCM port);
5744
5745
5746 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
5747
5748 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
5749
5750 SCM_TYP7S now masks away the bit which distinguishes substrings from
5751 strings.
5752
5753 ** scm_catch_body_t: Backward incompatible change!
5754
5755 Body functions to scm_internal_catch and friends do not any longer
5756 take a second argument. This is because it is no longer possible to
5757 pass a #f arg to catch.
5758
5759 ** Calls to scm_protect_object and scm_unprotect now nest properly.
5760
5761 The function scm_protect_object protects its argument from being freed
5762 by the garbage collector. scm_unprotect_object removes that
5763 protection.
5764
5765 These functions now nest properly. That is, for every object O, there
5766 is a counter which scm_protect_object(O) increments and
5767 scm_unprotect_object(O) decrements, if the counter is greater than
5768 zero. Every object's counter is zero when it is first created. If an
5769 object's counter is greater than zero, the garbage collector will not
5770 reclaim its storage.
5771
5772 This allows you to use scm_protect_object in your code without
5773 worrying that some other function you call will call
5774 scm_unprotect_object, and allow it to be freed. Assuming that the
5775 functions you call are well-behaved, and unprotect only those objects
5776 they protect, you can follow the same rule and have confidence that
5777 objects will be freed only at appropriate times.
5778
5779 \f
5780 Changes in Guile 1.2 (released Tuesday, June 24 1997):
5781
5782 * Changes to the distribution
5783
5784 ** Nightly snapshots are now available from ftp.red-bean.com.
5785 The old server, ftp.cyclic.com, has been relinquished to its rightful
5786 owner.
5787
5788 Nightly snapshots of the Guile development sources are now available via
5789 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
5790
5791 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
5792 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
5793
5794 ** To run Guile without installing it, the procedure has changed a bit.
5795
5796 If you used a separate build directory to compile Guile, you'll need
5797 to include the build directory in SCHEME_LOAD_PATH, as well as the
5798 source directory. See the `INSTALL' file for examples.
5799
5800 * Changes to the procedure for linking libguile with your programs
5801
5802 ** The standard Guile load path for Scheme code now includes
5803 $(datadir)/guile (usually /usr/local/share/guile). This means that
5804 you can install your own Scheme files there, and Guile will find them.
5805 (Previous versions of Guile only checked a directory whose name
5806 contained the Guile version number, so you had to re-install or move
5807 your Scheme sources each time you installed a fresh version of Guile.)
5808
5809 The load path also includes $(datadir)/guile/site; we recommend
5810 putting individual Scheme files there. If you want to install a
5811 package with multiple source files, create a directory for them under
5812 $(datadir)/guile.
5813
5814 ** Guile 1.2 will now use the Rx regular expression library, if it is
5815 installed on your system. When you are linking libguile into your own
5816 programs, this means you will have to link against -lguile, -lqt (if
5817 you configured Guile with thread support), and -lrx.
5818
5819 If you are using autoconf to generate configuration scripts for your
5820 application, the following lines should suffice to add the appropriate
5821 libraries to your link command:
5822
5823 ### Find Rx, quickthreads and libguile.
5824 AC_CHECK_LIB(rx, main)
5825 AC_CHECK_LIB(qt, main)
5826 AC_CHECK_LIB(guile, scm_shell)
5827
5828 The Guile 1.2 distribution does not contain sources for the Rx
5829 library, as Guile 1.0 did. If you want to use Rx, you'll need to
5830 retrieve it from a GNU FTP site and install it separately.
5831
5832 * Changes to Scheme functions and syntax
5833
5834 ** The dynamic linking features of Guile are now enabled by default.
5835 You can disable them by giving the `--disable-dynamic-linking' option
5836 to configure.
5837
5838 (dynamic-link FILENAME)
5839
5840 Find the object file denoted by FILENAME (a string) and link it
5841 into the running Guile application. When everything works out,
5842 return a Scheme object suitable for representing the linked object
5843 file. Otherwise an error is thrown. How object files are
5844 searched is system dependent.
5845
5846 (dynamic-object? VAL)
5847
5848 Determine whether VAL represents a dynamically linked object file.
5849
5850 (dynamic-unlink DYNOBJ)
5851
5852 Unlink the indicated object file from the application. DYNOBJ
5853 should be one of the values returned by `dynamic-link'.
5854
5855 (dynamic-func FUNCTION DYNOBJ)
5856
5857 Search the C function indicated by FUNCTION (a string or symbol)
5858 in DYNOBJ and return some Scheme object that can later be used
5859 with `dynamic-call' to actually call this function. Right now,
5860 these Scheme objects are formed by casting the address of the
5861 function to `long' and converting this number to its Scheme
5862 representation.
5863
5864 (dynamic-call FUNCTION DYNOBJ)
5865
5866 Call the C function indicated by FUNCTION and DYNOBJ. The
5867 function is passed no arguments and its return value is ignored.
5868 When FUNCTION is something returned by `dynamic-func', call that
5869 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
5870 etc.), look it up in DYNOBJ; this is equivalent to
5871
5872 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
5873
5874 Interrupts are deferred while the C function is executing (with
5875 SCM_DEFER_INTS/SCM_ALLOW_INTS).
5876
5877 (dynamic-args-call FUNCTION DYNOBJ ARGS)
5878
5879 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
5880 some arguments and return its return value. The C function is
5881 expected to take two arguments and return an `int', just like
5882 `main':
5883
5884 int c_func (int argc, char **argv);
5885
5886 ARGS must be a list of strings and is converted into an array of
5887 `char *'. The array is passed in ARGV and its size in ARGC. The
5888 return value is converted to a Scheme number and returned from the
5889 call to `dynamic-args-call'.
5890
5891 When dynamic linking is disabled or not supported on your system,
5892 the above functions throw errors, but they are still available.
5893
5894 Here is a small example that works on GNU/Linux:
5895
5896 (define libc-obj (dynamic-link "libc.so"))
5897 (dynamic-args-call 'rand libc-obj '())
5898
5899 See the file `libguile/DYNAMIC-LINKING' for additional comments.
5900
5901 ** The #/ syntax for module names is depreciated, and will be removed
5902 in a future version of Guile. Instead of
5903
5904 #/foo/bar/baz
5905
5906 instead write
5907
5908 (foo bar baz)
5909
5910 The latter syntax is more consistent with existing Lisp practice.
5911
5912 ** Guile now does fancier printing of structures. Structures are the
5913 underlying implementation for records, which in turn are used to
5914 implement modules, so all of these object now print differently and in
5915 a more informative way.
5916
5917 The Scheme printer will examine the builtin variable *struct-printer*
5918 whenever it needs to print a structure object. When this variable is
5919 not `#f' it is deemed to be a procedure and will be applied to the
5920 structure object and the output port. When *struct-printer* is `#f'
5921 or the procedure return `#f' the structure object will be printed in
5922 the boring #<struct 80458270> form.
5923
5924 This hook is used by some routines in ice-9/boot-9.scm to implement
5925 type specific printing routines. Please read the comments there about
5926 "printing structs".
5927
5928 One of the more specific uses of structs are records. The printing
5929 procedure that could be passed to MAKE-RECORD-TYPE is now actually
5930 called. It should behave like a *struct-printer* procedure (described
5931 above).
5932
5933 ** Guile now supports a new R4RS-compliant syntax for keywords. A
5934 token of the form #:NAME, where NAME has the same syntax as a Scheme
5935 symbol, is the external representation of the keyword named NAME.
5936 Keyword objects print using this syntax as well, so values containing
5937 keyword objects can be read back into Guile. When used in an
5938 expression, keywords are self-quoting objects.
5939
5940 Guile suports this read syntax, and uses this print syntax, regardless
5941 of the current setting of the `keyword' read option. The `keyword'
5942 read option only controls whether Guile recognizes the `:NAME' syntax,
5943 which is incompatible with R4RS. (R4RS says such token represent
5944 symbols.)
5945
5946 ** Guile has regular expression support again. Guile 1.0 included
5947 functions for matching regular expressions, based on the Rx library.
5948 In Guile 1.1, the Guile/Rx interface was removed to simplify the
5949 distribution, and thus Guile had no regular expression support. Guile
5950 1.2 again supports the most commonly used functions, and supports all
5951 of SCSH's regular expression functions.
5952
5953 If your system does not include a POSIX regular expression library,
5954 and you have not linked Guile with a third-party regexp library such as
5955 Rx, these functions will not be available. You can tell whether your
5956 Guile installation includes regular expression support by checking
5957 whether the `*features*' list includes the `regex' symbol.
5958
5959 *** regexp functions
5960
5961 By default, Guile supports POSIX extended regular expressions. That
5962 means that the characters `(', `)', `+' and `?' are special, and must
5963 be escaped if you wish to match the literal characters.
5964
5965 This regular expression interface was modeled after that implemented
5966 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
5967 with SCSH regular expressions.
5968
5969 **** Function: string-match PATTERN STR [START]
5970 Compile the string PATTERN into a regular expression and compare
5971 it with STR. The optional numeric argument START specifies the
5972 position of STR at which to begin matching.
5973
5974 `string-match' returns a "match structure" which describes what,
5975 if anything, was matched by the regular expression. *Note Match
5976 Structures::. If STR does not match PATTERN at all,
5977 `string-match' returns `#f'.
5978
5979 Each time `string-match' is called, it must compile its PATTERN
5980 argument into a regular expression structure. This operation is
5981 expensive, which makes `string-match' inefficient if the same regular
5982 expression is used several times (for example, in a loop). For better
5983 performance, you can compile a regular expression in advance and then
5984 match strings against the compiled regexp.
5985
5986 **** Function: make-regexp STR [FLAGS]
5987 Compile the regular expression described by STR, and return the
5988 compiled regexp structure. If STR does not describe a legal
5989 regular expression, `make-regexp' throws a
5990 `regular-expression-syntax' error.
5991
5992 FLAGS may be the bitwise-or of one or more of the following:
5993
5994 **** Constant: regexp/extended
5995 Use POSIX Extended Regular Expression syntax when interpreting
5996 STR. If not set, POSIX Basic Regular Expression syntax is used.
5997 If the FLAGS argument is omitted, we assume regexp/extended.
5998
5999 **** Constant: regexp/icase
6000 Do not differentiate case. Subsequent searches using the
6001 returned regular expression will be case insensitive.
6002
6003 **** Constant: regexp/newline
6004 Match-any-character operators don't match a newline.
6005
6006 A non-matching list ([^...]) not containing a newline matches a
6007 newline.
6008
6009 Match-beginning-of-line operator (^) matches the empty string
6010 immediately after a newline, regardless of whether the FLAGS
6011 passed to regexp-exec contain regexp/notbol.
6012
6013 Match-end-of-line operator ($) matches the empty string
6014 immediately before a newline, regardless of whether the FLAGS
6015 passed to regexp-exec contain regexp/noteol.
6016
6017 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
6018 Match the compiled regular expression REGEXP against `str'. If
6019 the optional integer START argument is provided, begin matching
6020 from that position in the string. Return a match structure
6021 describing the results of the match, or `#f' if no match could be
6022 found.
6023
6024 FLAGS may be the bitwise-or of one or more of the following:
6025
6026 **** Constant: regexp/notbol
6027 The match-beginning-of-line operator always fails to match (but
6028 see the compilation flag regexp/newline above) This flag may be
6029 used when different portions of a string are passed to
6030 regexp-exec and the beginning of the string should not be
6031 interpreted as the beginning of the line.
6032
6033 **** Constant: regexp/noteol
6034 The match-end-of-line operator always fails to match (but see the
6035 compilation flag regexp/newline above)
6036
6037 **** Function: regexp? OBJ
6038 Return `#t' if OBJ is a compiled regular expression, or `#f'
6039 otherwise.
6040
6041 Regular expressions are commonly used to find patterns in one string
6042 and replace them with the contents of another string.
6043
6044 **** Function: regexp-substitute PORT MATCH [ITEM...]
6045 Write to the output port PORT selected contents of the match
6046 structure MATCH. Each ITEM specifies what should be written, and
6047 may be one of the following arguments:
6048
6049 * A string. String arguments are written out verbatim.
6050
6051 * An integer. The submatch with that number is written.
6052
6053 * The symbol `pre'. The portion of the matched string preceding
6054 the regexp match is written.
6055
6056 * The symbol `post'. The portion of the matched string
6057 following the regexp match is written.
6058
6059 PORT may be `#f', in which case nothing is written; instead,
6060 `regexp-substitute' constructs a string from the specified ITEMs
6061 and returns that.
6062
6063 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
6064 Similar to `regexp-substitute', but can be used to perform global
6065 substitutions on STR. Instead of taking a match structure as an
6066 argument, `regexp-substitute/global' takes two string arguments: a
6067 REGEXP string describing a regular expression, and a TARGET string
6068 which should be matched against this regular expression.
6069
6070 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
6071 exceptions:
6072
6073 * A function may be supplied. When this function is called, it
6074 will be passed one argument: a match structure for a given
6075 regular expression match. It should return a string to be
6076 written out to PORT.
6077
6078 * The `post' symbol causes `regexp-substitute/global' to recurse
6079 on the unmatched portion of STR. This *must* be supplied in
6080 order to perform global search-and-replace on STR; if it is
6081 not present among the ITEMs, then `regexp-substitute/global'
6082 will return after processing a single match.
6083
6084 *** Match Structures
6085
6086 A "match structure" is the object returned by `string-match' and
6087 `regexp-exec'. It describes which portion of a string, if any, matched
6088 the given regular expression. Match structures include: a reference to
6089 the string that was checked for matches; the starting and ending
6090 positions of the regexp match; and, if the regexp included any
6091 parenthesized subexpressions, the starting and ending positions of each
6092 submatch.
6093
6094 In each of the regexp match functions described below, the `match'
6095 argument must be a match structure returned by a previous call to
6096 `string-match' or `regexp-exec'. Most of these functions return some
6097 information about the original target string that was matched against a
6098 regular expression; we will call that string TARGET for easy reference.
6099
6100 **** Function: regexp-match? OBJ
6101 Return `#t' if OBJ is a match structure returned by a previous
6102 call to `regexp-exec', or `#f' otherwise.
6103
6104 **** Function: match:substring MATCH [N]
6105 Return the portion of TARGET matched by subexpression number N.
6106 Submatch 0 (the default) represents the entire regexp match. If
6107 the regular expression as a whole matched, but the subexpression
6108 number N did not match, return `#f'.
6109
6110 **** Function: match:start MATCH [N]
6111 Return the starting position of submatch number N.
6112
6113 **** Function: match:end MATCH [N]
6114 Return the ending position of submatch number N.
6115
6116 **** Function: match:prefix MATCH
6117 Return the unmatched portion of TARGET preceding the regexp match.
6118
6119 **** Function: match:suffix MATCH
6120 Return the unmatched portion of TARGET following the regexp match.
6121
6122 **** Function: match:count MATCH
6123 Return the number of parenthesized subexpressions from MATCH.
6124 Note that the entire regular expression match itself counts as a
6125 subexpression, and failed submatches are included in the count.
6126
6127 **** Function: match:string MATCH
6128 Return the original TARGET string.
6129
6130 *** Backslash Escapes
6131
6132 Sometimes you will want a regexp to match characters like `*' or `$'
6133 exactly. For example, to check whether a particular string represents
6134 a menu entry from an Info node, it would be useful to match it against
6135 a regexp like `^* [^:]*::'. However, this won't work; because the
6136 asterisk is a metacharacter, it won't match the `*' at the beginning of
6137 the string. In this case, we want to make the first asterisk un-magic.
6138
6139 You can do this by preceding the metacharacter with a backslash
6140 character `\'. (This is also called "quoting" the metacharacter, and
6141 is known as a "backslash escape".) When Guile sees a backslash in a
6142 regular expression, it considers the following glyph to be an ordinary
6143 character, no matter what special meaning it would ordinarily have.
6144 Therefore, we can make the above example work by changing the regexp to
6145 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
6146 to match only a single asterisk in the target string.
6147
6148 Since the backslash is itself a metacharacter, you may force a
6149 regexp to match a backslash in the target string by preceding the
6150 backslash with itself. For example, to find variable references in a
6151 TeX program, you might want to find occurrences of the string `\let\'
6152 followed by any number of alphabetic characters. The regular expression
6153 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
6154 each match a single backslash in the target string.
6155
6156 **** Function: regexp-quote STR
6157 Quote each special character found in STR with a backslash, and
6158 return the resulting string.
6159
6160 *Very important:* Using backslash escapes in Guile source code (as
6161 in Emacs Lisp or C) can be tricky, because the backslash character has
6162 special meaning for the Guile reader. For example, if Guile encounters
6163 the character sequence `\n' in the middle of a string while processing
6164 Scheme code, it replaces those characters with a newline character.
6165 Similarly, the character sequence `\t' is replaced by a horizontal tab.
6166 Several of these "escape sequences" are processed by the Guile reader
6167 before your code is executed. Unrecognized escape sequences are
6168 ignored: if the characters `\*' appear in a string, they will be
6169 translated to the single character `*'.
6170
6171 This translation is obviously undesirable for regular expressions,
6172 since we want to be able to include backslashes in a string in order to
6173 escape regexp metacharacters. Therefore, to make sure that a backslash
6174 is preserved in a string in your Guile program, you must use *two*
6175 consecutive backslashes:
6176
6177 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
6178
6179 The string in this example is preprocessed by the Guile reader before
6180 any code is executed. The resulting argument to `make-regexp' is the
6181 string `^\* [^:]*', which is what we really want.
6182
6183 This also means that in order to write a regular expression that
6184 matches a single backslash character, the regular expression string in
6185 the source code must include *four* backslashes. Each consecutive pair
6186 of backslashes gets translated by the Guile reader to a single
6187 backslash, and the resulting double-backslash is interpreted by the
6188 regexp engine as matching a single backslash character. Hence:
6189
6190 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
6191
6192 The reason for the unwieldiness of this syntax is historical. Both
6193 regular expression pattern matchers and Unix string processing systems
6194 have traditionally used backslashes with the special meanings described
6195 above. The POSIX regular expression specification and ANSI C standard
6196 both require these semantics. Attempting to abandon either convention
6197 would cause other kinds of compatibility problems, possibly more severe
6198 ones. Therefore, without extending the Scheme reader to support
6199 strings with different quoting conventions (an ungainly and confusing
6200 extension when implemented in other languages), we must adhere to this
6201 cumbersome escape syntax.
6202
6203 * Changes to the gh_ interface
6204
6205 * Changes to the scm_ interface
6206
6207 * Changes to system call interfaces:
6208
6209 ** The value returned by `raise' is now unspecified. It throws an exception
6210 if an error occurs.
6211
6212 *** A new procedure `sigaction' can be used to install signal handlers
6213
6214 (sigaction signum [action] [flags])
6215
6216 signum is the signal number, which can be specified using the value
6217 of SIGINT etc.
6218
6219 If action is omitted, sigaction returns a pair: the CAR is the current
6220 signal hander, which will be either an integer with the value SIG_DFL
6221 (default action) or SIG_IGN (ignore), or the Scheme procedure which
6222 handles the signal, or #f if a non-Scheme procedure handles the
6223 signal. The CDR contains the current sigaction flags for the handler.
6224
6225 If action is provided, it is installed as the new handler for signum.
6226 action can be a Scheme procedure taking one argument, or the value of
6227 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
6228 whatever signal handler was installed before sigaction was first used.
6229 Flags can optionally be specified for the new handler (SA_RESTART is
6230 always used if the system provides it, so need not be specified.) The
6231 return value is a pair with information about the old handler as
6232 described above.
6233
6234 This interface does not provide access to the "signal blocking"
6235 facility. Maybe this is not needed, since the thread support may
6236 provide solutions to the problem of consistent access to data
6237 structures.
6238
6239 *** A new procedure `flush-all-ports' is equivalent to running
6240 `force-output' on every port open for output.
6241
6242 ** Guile now provides information on how it was built, via the new
6243 global variable, %guile-build-info. This variable records the values
6244 of the standard GNU makefile directory variables as an assocation
6245 list, mapping variable names (symbols) onto directory paths (strings).
6246 For example, to find out where the Guile link libraries were
6247 installed, you can say:
6248
6249 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
6250
6251
6252 * Changes to the scm_ interface
6253
6254 ** The new function scm_handle_by_message_noexit is just like the
6255 existing scm_handle_by_message function, except that it doesn't call
6256 exit to terminate the process. Instead, it prints a message and just
6257 returns #f. This might be a more appropriate catch-all handler for
6258 new dynamic roots and threads.
6259
6260 \f
6261 Changes in Guile 1.1 (released Friday, May 16 1997):
6262
6263 * Changes to the distribution.
6264
6265 The Guile 1.0 distribution has been split up into several smaller
6266 pieces:
6267 guile-core --- the Guile interpreter itself.
6268 guile-tcltk --- the interface between the Guile interpreter and
6269 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
6270 is a toolkit for building graphical user interfaces.
6271 guile-rgx-ctax --- the interface between Guile and the Rx regular
6272 expression matcher, and the translator for the Ctax
6273 programming language. These are packaged together because the
6274 Ctax translator uses Rx to parse Ctax source code.
6275
6276 This NEWS file describes the changes made to guile-core since the 1.0
6277 release.
6278
6279 We no longer distribute the documentation, since it was either out of
6280 date, or incomplete. As soon as we have current documentation, we
6281 will distribute it.
6282
6283
6284
6285 * Changes to the stand-alone interpreter
6286
6287 ** guile now accepts command-line arguments compatible with SCSH, Olin
6288 Shivers' Scheme Shell.
6289
6290 In general, arguments are evaluated from left to right, but there are
6291 exceptions. The following switches stop argument processing, and
6292 stash all remaining command-line arguments as the value returned by
6293 the (command-line) function.
6294 -s SCRIPT load Scheme source code from FILE, and exit
6295 -c EXPR evalute Scheme expression EXPR, and exit
6296 -- stop scanning arguments; run interactively
6297
6298 The switches below are processed as they are encountered.
6299 -l FILE load Scheme source code from FILE
6300 -e FUNCTION after reading script, apply FUNCTION to
6301 command line arguments
6302 -ds do -s script at this point
6303 --emacs enable Emacs protocol (experimental)
6304 -h, --help display this help and exit
6305 -v, --version display version information and exit
6306 \ read arguments from following script lines
6307
6308 So, for example, here is a Guile script named `ekko' (thanks, Olin)
6309 which re-implements the traditional "echo" command:
6310
6311 #!/usr/local/bin/guile -s
6312 !#
6313 (define (main args)
6314 (map (lambda (arg) (display arg) (display " "))
6315 (cdr args))
6316 (newline))
6317
6318 (main (command-line))
6319
6320 Suppose we invoke this script as follows:
6321
6322 ekko a speckled gecko
6323
6324 Through the magic of Unix script processing (triggered by the `#!'
6325 token at the top of the file), /usr/local/bin/guile receives the
6326 following list of command-line arguments:
6327
6328 ("-s" "./ekko" "a" "speckled" "gecko")
6329
6330 Unix inserts the name of the script after the argument specified on
6331 the first line of the file (in this case, "-s"), and then follows that
6332 with the arguments given to the script. Guile loads the script, which
6333 defines the `main' function, and then applies it to the list of
6334 remaining command-line arguments, ("a" "speckled" "gecko").
6335
6336 In Unix, the first line of a script file must take the following form:
6337
6338 #!INTERPRETER ARGUMENT
6339
6340 where INTERPRETER is the absolute filename of the interpreter
6341 executable, and ARGUMENT is a single command-line argument to pass to
6342 the interpreter.
6343
6344 You may only pass one argument to the interpreter, and its length is
6345 limited. These restrictions can be annoying to work around, so Guile
6346 provides a general mechanism (borrowed from, and compatible with,
6347 SCSH) for circumventing them.
6348
6349 If the ARGUMENT in a Guile script is a single backslash character,
6350 `\', Guile will open the script file, parse arguments from its second
6351 and subsequent lines, and replace the `\' with them. So, for example,
6352 here is another implementation of the `ekko' script:
6353
6354 #!/usr/local/bin/guile \
6355 -e main -s
6356 !#
6357 (define (main args)
6358 (for-each (lambda (arg) (display arg) (display " "))
6359 (cdr args))
6360 (newline))
6361
6362 If the user invokes this script as follows:
6363
6364 ekko a speckled gecko
6365
6366 Unix expands this into
6367
6368 /usr/local/bin/guile \ ekko a speckled gecko
6369
6370 When Guile sees the `\' argument, it replaces it with the arguments
6371 read from the second line of the script, producing:
6372
6373 /usr/local/bin/guile -e main -s ekko a speckled gecko
6374
6375 This tells Guile to load the `ekko' script, and apply the function
6376 `main' to the argument list ("a" "speckled" "gecko").
6377
6378 Here is how Guile parses the command-line arguments:
6379 - Each space character terminates an argument. This means that two
6380 spaces in a row introduce an empty-string argument.
6381 - The tab character is not permitted (unless you quote it with the
6382 backslash character, as described below), to avoid confusion.
6383 - The newline character terminates the sequence of arguments, and will
6384 also terminate a final non-empty argument. (However, a newline
6385 following a space will not introduce a final empty-string argument;
6386 it only terminates the argument list.)
6387 - The backslash character is the escape character. It escapes
6388 backslash, space, tab, and newline. The ANSI C escape sequences
6389 like \n and \t are also supported. These produce argument
6390 constituents; the two-character combination \n doesn't act like a
6391 terminating newline. The escape sequence \NNN for exactly three
6392 octal digits reads as the character whose ASCII code is NNN. As
6393 above, characters produced this way are argument constituents.
6394 Backslash followed by other characters is not allowed.
6395
6396 * Changes to the procedure for linking libguile with your programs
6397
6398 ** Guile now builds and installs a shared guile library, if your
6399 system support shared libraries. (It still builds a static library on
6400 all systems.) Guile automatically detects whether your system
6401 supports shared libraries. To prevent Guile from buildisg shared
6402 libraries, pass the `--disable-shared' flag to the configure script.
6403
6404 Guile takes longer to compile when it builds shared libraries, because
6405 it must compile every file twice --- once to produce position-
6406 independent object code, and once to produce normal object code.
6407
6408 ** The libthreads library has been merged into libguile.
6409
6410 To link a program against Guile, you now need only link against
6411 -lguile and -lqt; -lthreads is no longer needed. If you are using
6412 autoconf to generate configuration scripts for your application, the
6413 following lines should suffice to add the appropriate libraries to
6414 your link command:
6415
6416 ### Find quickthreads and libguile.
6417 AC_CHECK_LIB(qt, main)
6418 AC_CHECK_LIB(guile, scm_shell)
6419
6420 * Changes to Scheme functions
6421
6422 ** Guile Scheme's special syntax for keyword objects is now optional,
6423 and disabled by default.
6424
6425 The syntax variation from R4RS made it difficult to port some
6426 interesting packages to Guile. The routines which accepted keyword
6427 arguments (mostly in the module system) have been modified to also
6428 accept symbols whose names begin with `:'.
6429
6430 To change the keyword syntax, you must first import the (ice-9 debug)
6431 module:
6432 (use-modules (ice-9 debug))
6433
6434 Then you can enable the keyword syntax as follows:
6435 (read-set! keywords 'prefix)
6436
6437 To disable keyword syntax, do this:
6438 (read-set! keywords #f)
6439
6440 ** Many more primitive functions accept shared substrings as
6441 arguments. In the past, these functions required normal, mutable
6442 strings as arguments, although they never made use of this
6443 restriction.
6444
6445 ** The uniform array functions now operate on byte vectors. These
6446 functions are `array-fill!', `serial-array-copy!', `array-copy!',
6447 `serial-array-map', `array-map', `array-for-each', and
6448 `array-index-map!'.
6449
6450 ** The new functions `trace' and `untrace' implement simple debugging
6451 support for Scheme functions.
6452
6453 The `trace' function accepts any number of procedures as arguments,
6454 and tells the Guile interpreter to display each procedure's name and
6455 arguments each time the procedure is invoked. When invoked with no
6456 arguments, `trace' returns the list of procedures currently being
6457 traced.
6458
6459 The `untrace' function accepts any number of procedures as arguments,
6460 and tells the Guile interpreter not to trace them any more. When
6461 invoked with no arguments, `untrace' untraces all curretly traced
6462 procedures.
6463
6464 The tracing in Guile has an advantage over most other systems: we
6465 don't create new procedure objects, but mark the procedure objects
6466 themselves. This means that anonymous and internal procedures can be
6467 traced.
6468
6469 ** The function `assert-repl-prompt' has been renamed to
6470 `set-repl-prompt!'. It takes one argument, PROMPT.
6471 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
6472 - If PROMPT is a string, we use it as a prompt.
6473 - If PROMPT is a procedure accepting no arguments, we call it, and
6474 display the result as a prompt.
6475 - Otherwise, we display "> ".
6476
6477 ** The new function `eval-string' reads Scheme expressions from a
6478 string and evaluates them, returning the value of the last expression
6479 in the string. If the string contains no expressions, it returns an
6480 unspecified value.
6481
6482 ** The new function `thunk?' returns true iff its argument is a
6483 procedure of zero arguments.
6484
6485 ** `defined?' is now a builtin function, instead of syntax. This
6486 means that its argument should be quoted. It returns #t iff its
6487 argument is bound in the current module.
6488
6489 ** The new syntax `use-modules' allows you to add new modules to your
6490 environment without re-typing a complete `define-module' form. It
6491 accepts any number of module names as arguments, and imports their
6492 public bindings into the current module.
6493
6494 ** The new function (module-defined? NAME MODULE) returns true iff
6495 NAME, a symbol, is defined in MODULE, a module object.
6496
6497 ** The new function `builtin-bindings' creates and returns a hash
6498 table containing copies of all the root module's bindings.
6499
6500 ** The new function `builtin-weak-bindings' does the same as
6501 `builtin-bindings', but creates a doubly-weak hash table.
6502
6503 ** The `equal?' function now considers variable objects to be
6504 equivalent if they have the same name and the same value.
6505
6506 ** The new function `command-line' returns the command-line arguments
6507 given to Guile, as a list of strings.
6508
6509 When using guile as a script interpreter, `command-line' returns the
6510 script's arguments; those processed by the interpreter (like `-s' or
6511 `-c') are omitted. (In other words, you get the normal, expected
6512 behavior.) Any application that uses scm_shell to process its
6513 command-line arguments gets this behavior as well.
6514
6515 ** The new function `load-user-init' looks for a file called `.guile'
6516 in the user's home directory, and loads it if it exists. This is
6517 mostly for use by the code generated by scm_compile_shell_switches,
6518 but we thought it might also be useful in other circumstances.
6519
6520 ** The new function `log10' returns the base-10 logarithm of its
6521 argument.
6522
6523 ** Changes to I/O functions
6524
6525 *** The functions `read', `primitive-load', `read-and-eval!', and
6526 `primitive-load-path' no longer take optional arguments controlling
6527 case insensitivity and a `#' parser.
6528
6529 Case sensitivity is now controlled by a read option called
6530 `case-insensitive'. The user can add new `#' syntaxes with the
6531 `read-hash-extend' function (see below).
6532
6533 *** The new function `read-hash-extend' allows the user to change the
6534 syntax of Guile Scheme in a somewhat controlled way.
6535
6536 (read-hash-extend CHAR PROC)
6537 When parsing S-expressions, if we read a `#' character followed by
6538 the character CHAR, use PROC to parse an object from the stream.
6539 If PROC is #f, remove any parsing procedure registered for CHAR.
6540
6541 The reader applies PROC to two arguments: CHAR and an input port.
6542
6543 *** The new functions read-delimited and read-delimited! provide a
6544 general mechanism for doing delimited input on streams.
6545
6546 (read-delimited DELIMS [PORT HANDLE-DELIM])
6547 Read until we encounter one of the characters in DELIMS (a string),
6548 or end-of-file. PORT is the input port to read from; it defaults to
6549 the current input port. The HANDLE-DELIM parameter determines how
6550 the terminating character is handled; it should be one of the
6551 following symbols:
6552
6553 'trim omit delimiter from result
6554 'peek leave delimiter character in input stream
6555 'concat append delimiter character to returned value
6556 'split return a pair: (RESULT . TERMINATOR)
6557
6558 HANDLE-DELIM defaults to 'peek.
6559
6560 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
6561 A side-effecting variant of `read-delimited'.
6562
6563 The data is written into the string BUF at the indices in the
6564 half-open interval [START, END); the default interval is the whole
6565 string: START = 0 and END = (string-length BUF). The values of
6566 START and END must specify a well-defined interval in BUF, i.e.
6567 0 <= START <= END <= (string-length BUF).
6568
6569 It returns NBYTES, the number of bytes read. If the buffer filled
6570 up without a delimiter character being found, it returns #f. If the
6571 port is at EOF when the read starts, it returns the EOF object.
6572
6573 If an integer is returned (i.e., the read is successfully terminated
6574 by reading a delimiter character), then the HANDLE-DELIM parameter
6575 determines how to handle the terminating character. It is described
6576 above, and defaults to 'peek.
6577
6578 (The descriptions of these functions were borrowed from the SCSH
6579 manual, by Olin Shivers and Brian Carlstrom.)
6580
6581 *** The `%read-delimited!' function is the primitive used to implement
6582 `read-delimited' and `read-delimited!'.
6583
6584 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
6585
6586 This returns a pair of values: (TERMINATOR . NUM-READ).
6587 - TERMINATOR describes why the read was terminated. If it is a
6588 character or the eof object, then that is the value that terminated
6589 the read. If it is #f, the function filled the buffer without finding
6590 a delimiting character.
6591 - NUM-READ is the number of characters read into BUF.
6592
6593 If the read is successfully terminated by reading a delimiter
6594 character, then the gobble? parameter determines what to do with the
6595 terminating character. If true, the character is removed from the
6596 input stream; if false, the character is left in the input stream
6597 where a subsequent read operation will retrieve it. In either case,
6598 the character is also the first value returned by the procedure call.
6599
6600 (The descriptions of this function was borrowed from the SCSH manual,
6601 by Olin Shivers and Brian Carlstrom.)
6602
6603 *** The `read-line' and `read-line!' functions have changed; they now
6604 trim the terminator by default; previously they appended it to the
6605 returned string. For the old behavior, use (read-line PORT 'concat).
6606
6607 *** The functions `uniform-array-read!' and `uniform-array-write!' now
6608 take new optional START and END arguments, specifying the region of
6609 the array to read and write.
6610
6611 *** The `ungetc-char-ready?' function has been removed. We feel it's
6612 inappropriate for an interface to expose implementation details this
6613 way.
6614
6615 ** Changes to the Unix library and system call interface
6616
6617 *** The new fcntl function provides access to the Unix `fcntl' system
6618 call.
6619
6620 (fcntl PORT COMMAND VALUE)
6621 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
6622 Values for COMMAND are:
6623
6624 F_DUPFD duplicate a file descriptor
6625 F_GETFD read the descriptor's close-on-exec flag
6626 F_SETFD set the descriptor's close-on-exec flag to VALUE
6627 F_GETFL read the descriptor's flags, as set on open
6628 F_SETFL set the descriptor's flags, as set on open to VALUE
6629 F_GETOWN return the process ID of a socket's owner, for SIGIO
6630 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
6631 FD_CLOEXEC not sure what this is
6632
6633 For details, see the documentation for the fcntl system call.
6634
6635 *** The arguments to `select' have changed, for compatibility with
6636 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
6637 expected behavior. The MILLISECONDS parameter has been changed to
6638 MICROSECONDS, to more closely resemble the underlying system call.
6639 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
6640 corresponding return set will be the same.
6641
6642 *** The arguments to the `mknod' system call have changed. They are
6643 now:
6644
6645 (mknod PATH TYPE PERMS DEV)
6646 Create a new file (`node') in the file system. PATH is the name of
6647 the file to create. TYPE is the kind of file to create; it should
6648 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
6649 permission bits to give the newly created file. If TYPE is
6650 'block-special or 'char-special, DEV specifies which device the
6651 special file refers to; its interpretation depends on the kind of
6652 special file being created.
6653
6654 *** The `fork' function has been renamed to `primitive-fork', to avoid
6655 clashing with various SCSH forks.
6656
6657 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
6658 and `recvfrom!'. They no longer accept a size for a second argument;
6659 you must pass a string to hold the received value. They no longer
6660 return the buffer. Instead, `recv' returns the length of the message
6661 received, and `recvfrom' returns a pair containing the packet's length
6662 and originating address.
6663
6664 *** The file descriptor datatype has been removed, as have the
6665 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
6666 We plan to replace these functions with a SCSH-compatible interface.
6667
6668 *** The `create' function has been removed; it's just a special case
6669 of `open'.
6670
6671 *** There are new functions to break down process termination status
6672 values. In the descriptions below, STATUS is a value returned by
6673 `waitpid'.
6674
6675 (status:exit-val STATUS)
6676 If the child process exited normally, this function returns the exit
6677 code for the child process (i.e., the value passed to exit, or
6678 returned from main). If the child process did not exit normally,
6679 this function returns #f.
6680
6681 (status:stop-sig STATUS)
6682 If the child process was suspended by a signal, this function
6683 returns the signal that suspended the child. Otherwise, it returns
6684 #f.
6685
6686 (status:term-sig STATUS)
6687 If the child process terminated abnormally, this function returns
6688 the signal that terminated the child. Otherwise, this function
6689 returns false.
6690
6691 POSIX promises that exactly one of these functions will return true on
6692 a valid STATUS value.
6693
6694 These functions are compatible with SCSH.
6695
6696 *** There are new accessors and setters for the broken-out time vectors
6697 returned by `localtime', `gmtime', and that ilk. They are:
6698
6699 Component Accessor Setter
6700 ========================= ============ ============
6701 seconds tm:sec set-tm:sec
6702 minutes tm:min set-tm:min
6703 hours tm:hour set-tm:hour
6704 day of the month tm:mday set-tm:mday
6705 month tm:mon set-tm:mon
6706 year tm:year set-tm:year
6707 day of the week tm:wday set-tm:wday
6708 day in the year tm:yday set-tm:yday
6709 daylight saving time tm:isdst set-tm:isdst
6710 GMT offset, seconds tm:gmtoff set-tm:gmtoff
6711 name of time zone tm:zone set-tm:zone
6712
6713 *** There are new accessors for the vectors returned by `uname',
6714 describing the host system:
6715
6716 Component Accessor
6717 ============================================== ================
6718 name of the operating system implementation utsname:sysname
6719 network name of this machine utsname:nodename
6720 release level of the operating system utsname:release
6721 version level of the operating system utsname:version
6722 machine hardware platform utsname:machine
6723
6724 *** There are new accessors for the vectors returned by `getpw',
6725 `getpwnam', `getpwuid', and `getpwent', describing entries from the
6726 system's user database:
6727
6728 Component Accessor
6729 ====================== =================
6730 user name passwd:name
6731 user password passwd:passwd
6732 user id passwd:uid
6733 group id passwd:gid
6734 real name passwd:gecos
6735 home directory passwd:dir
6736 shell program passwd:shell
6737
6738 *** There are new accessors for the vectors returned by `getgr',
6739 `getgrnam', `getgrgid', and `getgrent', describing entries from the
6740 system's group database:
6741
6742 Component Accessor
6743 ======================= ============
6744 group name group:name
6745 group password group:passwd
6746 group id group:gid
6747 group members group:mem
6748
6749 *** There are new accessors for the vectors returned by `gethost',
6750 `gethostbyaddr', `gethostbyname', and `gethostent', describing
6751 internet hosts:
6752
6753 Component Accessor
6754 ========================= ===============
6755 official name of host hostent:name
6756 alias list hostent:aliases
6757 host address type hostent:addrtype
6758 length of address hostent:length
6759 list of addresses hostent:addr-list
6760
6761 *** There are new accessors for the vectors returned by `getnet',
6762 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
6763 networks:
6764
6765 Component Accessor
6766 ========================= ===============
6767 official name of net netent:name
6768 alias list netent:aliases
6769 net number type netent:addrtype
6770 net number netent:net
6771
6772 *** There are new accessors for the vectors returned by `getproto',
6773 `getprotobyname', `getprotobynumber', and `getprotoent', describing
6774 internet protocols:
6775
6776 Component Accessor
6777 ========================= ===============
6778 official protocol name protoent:name
6779 alias list protoent:aliases
6780 protocol number protoent:proto
6781
6782 *** There are new accessors for the vectors returned by `getserv',
6783 `getservbyname', `getservbyport', and `getservent', describing
6784 internet protocols:
6785
6786 Component Accessor
6787 ========================= ===============
6788 official service name servent:name
6789 alias list servent:aliases
6790 port number servent:port
6791 protocol to use servent:proto
6792
6793 *** There are new accessors for the sockaddr structures returned by
6794 `accept', `getsockname', `getpeername', `recvfrom!':
6795
6796 Component Accessor
6797 ======================================== ===============
6798 address format (`family') sockaddr:fam
6799 path, for file domain addresses sockaddr:path
6800 address, for internet domain addresses sockaddr:addr
6801 TCP or UDP port, for internet sockaddr:port
6802
6803 *** The `getpwent', `getgrent', `gethostent', `getnetent',
6804 `getprotoent', and `getservent' functions now return #f at the end of
6805 the user database. (They used to throw an exception.)
6806
6807 Note that calling MUMBLEent function is equivalent to calling the
6808 corresponding MUMBLE function with no arguments.
6809
6810 *** The `setpwent', `setgrent', `sethostent', `setnetent',
6811 `setprotoent', and `setservent' routines now take no arguments.
6812
6813 *** The `gethost', `getproto', `getnet', and `getserv' functions now
6814 provide more useful information when they throw an exception.
6815
6816 *** The `lnaof' function has been renamed to `inet-lnaof'.
6817
6818 *** Guile now claims to have the `current-time' feature.
6819
6820 *** The `mktime' function now takes an optional second argument ZONE,
6821 giving the time zone to use for the conversion. ZONE should be a
6822 string, in the same format as expected for the "TZ" environment variable.
6823
6824 *** The `strptime' function now returns a pair (TIME . COUNT), where
6825 TIME is the parsed time as a vector, and COUNT is the number of
6826 characters from the string left unparsed. This function used to
6827 return the remaining characters as a string.
6828
6829 *** The `gettimeofday' function has replaced the old `time+ticks' function.
6830 The return value is now (SECONDS . MICROSECONDS); the fractional
6831 component is no longer expressed in "ticks".
6832
6833 *** The `ticks/sec' constant has been removed, in light of the above change.
6834
6835 * Changes to the gh_ interface
6836
6837 ** gh_eval_str() now returns an SCM object which is the result of the
6838 evaluation
6839
6840 ** gh_scm2str() now copies the Scheme data to a caller-provided C
6841 array
6842
6843 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
6844 and returns the array
6845
6846 ** gh_scm2str0() is gone: there is no need to distinguish
6847 null-terminated from non-null-terminated, since gh_scm2newstr() allows
6848 the user to interpret the data both ways.
6849
6850 * Changes to the scm_ interface
6851
6852 ** The new function scm_symbol_value0 provides an easy way to get a
6853 symbol's value from C code:
6854
6855 SCM scm_symbol_value0 (char *NAME)
6856 Return the value of the symbol named by the null-terminated string
6857 NAME in the current module. If the symbol named NAME is unbound in
6858 the current module, return SCM_UNDEFINED.
6859
6860 ** The new function scm_sysintern0 creates new top-level variables,
6861 without assigning them a value.
6862
6863 SCM scm_sysintern0 (char *NAME)
6864 Create a new Scheme top-level variable named NAME. NAME is a
6865 null-terminated string. Return the variable's value cell.
6866
6867 ** The function scm_internal_catch is the guts of catch. It handles
6868 all the mechanics of setting up a catch target, invoking the catch
6869 body, and perhaps invoking the handler if the body does a throw.
6870
6871 The function is designed to be usable from C code, but is general
6872 enough to implement all the semantics Guile Scheme expects from throw.
6873
6874 TAG is the catch tag. Typically, this is a symbol, but this function
6875 doesn't actually care about that.
6876
6877 BODY is a pointer to a C function which runs the body of the catch;
6878 this is the code you can throw from. We call it like this:
6879 BODY (BODY_DATA, JMPBUF)
6880 where:
6881 BODY_DATA is just the BODY_DATA argument we received; we pass it
6882 through to BODY as its first argument. The caller can make
6883 BODY_DATA point to anything useful that BODY might need.
6884 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
6885 which we have just created and initialized.
6886
6887 HANDLER is a pointer to a C function to deal with a throw to TAG,
6888 should one occur. We call it like this:
6889 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
6890 where
6891 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
6892 same idea as BODY_DATA above.
6893 THROWN_TAG is the tag that the user threw to; usually this is
6894 TAG, but it could be something else if TAG was #t (i.e., a
6895 catch-all), or the user threw to a jmpbuf.
6896 THROW_ARGS is the list of arguments the user passed to the THROW
6897 function.
6898
6899 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
6900 is just a pointer we pass through to HANDLER. We don't actually
6901 use either of those pointers otherwise ourselves. The idea is
6902 that, if our caller wants to communicate something to BODY or
6903 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
6904 HANDLER can then use. Think of it as a way to make BODY and
6905 HANDLER closures, not just functions; MUMBLE_DATA points to the
6906 enclosed variables.
6907
6908 Of course, it's up to the caller to make sure that any data a
6909 MUMBLE_DATA needs is protected from GC. A common way to do this is
6910 to make MUMBLE_DATA a pointer to data stored in an automatic
6911 structure variable; since the collector must scan the stack for
6912 references anyway, this assures that any references in MUMBLE_DATA
6913 will be found.
6914
6915 ** The new function scm_internal_lazy_catch is exactly like
6916 scm_internal_catch, except:
6917
6918 - It does not unwind the stack (this is the major difference).
6919 - If handler returns, its value is returned from the throw.
6920 - BODY always receives #f as its JMPBUF argument (since there's no
6921 jmpbuf associated with a lazy catch, because we don't unwind the
6922 stack.)
6923
6924 ** scm_body_thunk is a new body function you can pass to
6925 scm_internal_catch if you want the body to be like Scheme's `catch'
6926 --- a thunk, or a function of one argument if the tag is #f.
6927
6928 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
6929 contains the Scheme procedure to invoke as the body, and the tag
6930 we're catching. If the tag is #f, then we pass JMPBUF (created by
6931 scm_internal_catch) to the body procedure; otherwise, the body gets
6932 no arguments.
6933
6934 ** scm_handle_by_proc is a new handler function you can pass to
6935 scm_internal_catch if you want the handler to act like Scheme's catch
6936 --- call a procedure with the tag and the throw arguments.
6937
6938 If the user does a throw to this catch, this function runs a handler
6939 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
6940 variable holding the Scheme procedure object to invoke. It ought to
6941 be a pointer to an automatic variable (i.e., one living on the stack),
6942 or the procedure object should be otherwise protected from GC.
6943
6944 ** scm_handle_by_message is a new handler function to use with
6945 `scm_internal_catch' if you want Guile to print a message and die.
6946 It's useful for dealing with throws to uncaught keys at the top level.
6947
6948 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
6949 message header to print; if zero, we use "guile" instead. That
6950 text is followed by a colon, then the message described by ARGS.
6951
6952 ** The return type of scm_boot_guile is now void; the function does
6953 not return a value, and indeed, never returns at all.
6954
6955 ** The new function scm_shell makes it easy for user applications to
6956 process command-line arguments in a way that is compatible with the
6957 stand-alone guile interpreter (which is in turn compatible with SCSH,
6958 the Scheme shell).
6959
6960 To use the scm_shell function, first initialize any guile modules
6961 linked into your application, and then call scm_shell with the values
6962 of ARGC and ARGV your `main' function received. scm_shell will add
6963 any SCSH-style meta-arguments from the top of the script file to the
6964 argument vector, and then process the command-line arguments. This
6965 generally means loading a script file or starting up an interactive
6966 command interpreter. For details, see "Changes to the stand-alone
6967 interpreter" above.
6968
6969 ** The new functions scm_get_meta_args and scm_count_argv help you
6970 implement the SCSH-style meta-argument, `\'.
6971
6972 char **scm_get_meta_args (int ARGC, char **ARGV)
6973 If the second element of ARGV is a string consisting of a single
6974 backslash character (i.e. "\\" in Scheme notation), open the file
6975 named by the following argument, parse arguments from it, and return
6976 the spliced command line. The returned array is terminated by a
6977 null pointer.
6978
6979 For details of argument parsing, see above, under "guile now accepts
6980 command-line arguments compatible with SCSH..."
6981
6982 int scm_count_argv (char **ARGV)
6983 Count the arguments in ARGV, assuming it is terminated by a null
6984 pointer.
6985
6986 For an example of how these functions might be used, see the source
6987 code for the function scm_shell in libguile/script.c.
6988
6989 You will usually want to use scm_shell instead of calling this
6990 function yourself.
6991
6992 ** The new function scm_compile_shell_switches turns an array of
6993 command-line arguments into Scheme code to carry out the actions they
6994 describe. Given ARGC and ARGV, it returns a Scheme expression to
6995 evaluate, and calls scm_set_program_arguments to make any remaining
6996 command-line arguments available to the Scheme code. For example,
6997 given the following arguments:
6998
6999 -e main -s ekko a speckled gecko
7000
7001 scm_set_program_arguments will return the following expression:
7002
7003 (begin (load "ekko") (main (command-line)) (quit))
7004
7005 You will usually want to use scm_shell instead of calling this
7006 function yourself.
7007
7008 ** The function scm_shell_usage prints a usage message appropriate for
7009 an interpreter that uses scm_compile_shell_switches to handle its
7010 command-line arguments.
7011
7012 void scm_shell_usage (int FATAL, char *MESSAGE)
7013 Print a usage message to the standard error output. If MESSAGE is
7014 non-zero, write it before the usage message, followed by a newline.
7015 If FATAL is non-zero, exit the process, using FATAL as the
7016 termination status. (If you want to be compatible with Guile,
7017 always use 1 as the exit status when terminating due to command-line
7018 usage problems.)
7019
7020 You will usually want to use scm_shell instead of calling this
7021 function yourself.
7022
7023 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
7024 expressions. It used to return SCM_EOL. Earth-shattering.
7025
7026 ** The macros for declaring scheme objects in C code have been
7027 rearranged slightly. They are now:
7028
7029 SCM_SYMBOL (C_NAME, SCHEME_NAME)
7030 Declare a static SCM variable named C_NAME, and initialize it to
7031 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
7032 be a C identifier, and SCHEME_NAME should be a C string.
7033
7034 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
7035 Just like SCM_SYMBOL, but make C_NAME globally visible.
7036
7037 SCM_VCELL (C_NAME, SCHEME_NAME)
7038 Create a global variable at the Scheme level named SCHEME_NAME.
7039 Declare a static SCM variable named C_NAME, and initialize it to
7040 point to the Scheme variable's value cell.
7041
7042 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
7043 Just like SCM_VCELL, but make C_NAME globally visible.
7044
7045 The `guile-snarf' script writes initialization code for these macros
7046 to its standard output, given C source code as input.
7047
7048 The SCM_GLOBAL macro is gone.
7049
7050 ** The scm_read_line and scm_read_line_x functions have been replaced
7051 by Scheme code based on the %read-delimited! procedure (known to C
7052 code as scm_read_delimited_x). See its description above for more
7053 information.
7054
7055 ** The function scm_sys_open has been renamed to scm_open. It now
7056 returns a port instead of an FD object.
7057
7058 * The dynamic linking support has changed. For more information, see
7059 libguile/DYNAMIC-LINKING.
7060
7061 \f
7062 Guile 1.0b3
7063
7064 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
7065 (Sun 5 Jan 1997):
7066
7067 * Changes to the 'guile' program:
7068
7069 ** Guile now loads some new files when it starts up. Guile first
7070 searches the load path for init.scm, and loads it if found. Then, if
7071 Guile is not being used to execute a script, and the user's home
7072 directory contains a file named `.guile', Guile loads that.
7073
7074 ** You can now use Guile as a shell script interpreter.
7075
7076 To paraphrase the SCSH manual:
7077
7078 When Unix tries to execute an executable file whose first two
7079 characters are the `#!', it treats the file not as machine code to
7080 be directly executed by the native processor, but as source code
7081 to be executed by some interpreter. The interpreter to use is
7082 specified immediately after the #! sequence on the first line of
7083 the source file. The kernel reads in the name of the interpreter,
7084 and executes that instead. It passes the interpreter the source
7085 filename as its first argument, with the original arguments
7086 following. Consult the Unix man page for the `exec' system call
7087 for more information.
7088
7089 Now you can use Guile as an interpreter, using a mechanism which is a
7090 compatible subset of that provided by SCSH.
7091
7092 Guile now recognizes a '-s' command line switch, whose argument is the
7093 name of a file of Scheme code to load. It also treats the two
7094 characters `#!' as the start of a comment, terminated by `!#'. Thus,
7095 to make a file of Scheme code directly executable by Unix, insert the
7096 following two lines at the top of the file:
7097
7098 #!/usr/local/bin/guile -s
7099 !#
7100
7101 Guile treats the argument of the `-s' command-line switch as the name
7102 of a file of Scheme code to load, and treats the sequence `#!' as the
7103 start of a block comment, terminated by `!#'.
7104
7105 For example, here's a version of 'echo' written in Scheme:
7106
7107 #!/usr/local/bin/guile -s
7108 !#
7109 (let loop ((args (cdr (program-arguments))))
7110 (if (pair? args)
7111 (begin
7112 (display (car args))
7113 (if (pair? (cdr args))
7114 (display " "))
7115 (loop (cdr args)))))
7116 (newline)
7117
7118 Why does `#!' start a block comment terminated by `!#', instead of the
7119 end of the line? That is the notation SCSH uses, and although we
7120 don't yet support the other SCSH features that motivate that choice,
7121 we would like to be backward-compatible with any existing Guile
7122 scripts once we do. Furthermore, if the path to Guile on your system
7123 is too long for your kernel, you can start the script with this
7124 horrible hack:
7125
7126 #!/bin/sh
7127 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
7128 !#
7129
7130 Note that some very old Unix systems don't support the `#!' syntax.
7131
7132
7133 ** You can now run Guile without installing it.
7134
7135 Previous versions of the interactive Guile interpreter (`guile')
7136 couldn't start up unless Guile's Scheme library had been installed;
7137 they used the value of the environment variable `SCHEME_LOAD_PATH'
7138 later on in the startup process, but not to find the startup code
7139 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
7140 code.
7141
7142 To run Guile without installing it, build it in the normal way, and
7143 then set the environment variable `SCHEME_LOAD_PATH' to a
7144 colon-separated list of directories, including the top-level directory
7145 of the Guile sources. For example, if you unpacked Guile so that the
7146 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
7147 you might say
7148
7149 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
7150
7151
7152 ** Guile's read-eval-print loop no longer prints #<unspecified>
7153 results. If the user wants to see this, she can evaluate the
7154 expression (assert-repl-print-unspecified #t), perhaps in her startup
7155 file.
7156
7157 ** Guile no longer shows backtraces by default when an error occurs;
7158 however, it does display a message saying how to get one, and how to
7159 request that they be displayed by default. After an error, evaluate
7160 (backtrace)
7161 to see a backtrace, and
7162 (debug-enable 'backtrace)
7163 to see them by default.
7164
7165
7166
7167 * Changes to Guile Scheme:
7168
7169 ** Guile now distinguishes between #f and the empty list.
7170
7171 This is for compatibility with the IEEE standard, the (possibly)
7172 upcoming Revised^5 Report on Scheme, and many extant Scheme
7173 implementations.
7174
7175 Guile used to have #f and '() denote the same object, to make Scheme's
7176 type system more compatible with Emacs Lisp's. However, the change
7177 caused too much trouble for Scheme programmers, and we found another
7178 way to reconcile Emacs Lisp with Scheme that didn't require this.
7179
7180
7181 ** Guile's delq, delv, delete functions, and their destructive
7182 counterparts, delq!, delv!, and delete!, now remove all matching
7183 elements from the list, not just the first. This matches the behavior
7184 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
7185 functions which inspired them.
7186
7187 I recognize that this change may break code in subtle ways, but it
7188 seems best to make the change before the FSF's first Guile release,
7189 rather than after.
7190
7191
7192 ** The compiled-library-path function has been deleted from libguile.
7193
7194 ** The facilities for loading Scheme source files have changed.
7195
7196 *** The variable %load-path now tells Guile which directories to search
7197 for Scheme code. Its value is a list of strings, each of which names
7198 a directory.
7199
7200 *** The variable %load-extensions now tells Guile which extensions to
7201 try appending to a filename when searching the load path. Its value
7202 is a list of strings. Its default value is ("" ".scm").
7203
7204 *** (%search-load-path FILENAME) searches the directories listed in the
7205 value of the %load-path variable for a Scheme file named FILENAME,
7206 with all the extensions listed in %load-extensions. If it finds a
7207 match, then it returns its full filename. If FILENAME is absolute, it
7208 returns it unchanged. Otherwise, it returns #f.
7209
7210 %search-load-path will not return matches that refer to directories.
7211
7212 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
7213 uses %seach-load-path to find a file named FILENAME, and loads it if
7214 it finds it. If it can't read FILENAME for any reason, it throws an
7215 error.
7216
7217 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
7218 `read' function.
7219
7220 *** load uses the same searching semantics as primitive-load.
7221
7222 *** The functions %try-load, try-load-with-path, %load, load-with-path,
7223 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
7224 path, and load-module-with-path have been deleted. The functions
7225 above should serve their purposes.
7226
7227 *** If the value of the variable %load-hook is a procedure,
7228 `primitive-load' applies its value to the name of the file being
7229 loaded (without the load path directory name prepended). If its value
7230 is #f, it is ignored. Otherwise, an error occurs.
7231
7232 This is mostly useful for printing load notification messages.
7233
7234
7235 ** The function `eval!' is no longer accessible from the scheme level.
7236 We can't allow operations which introduce glocs into the scheme level,
7237 because Guile's type system can't handle these as data. Use `eval' or
7238 `read-and-eval!' (see below) as replacement.
7239
7240 ** The new function read-and-eval! reads an expression from PORT,
7241 evaluates it, and returns the result. This is more efficient than
7242 simply calling `read' and `eval', since it is not necessary to make a
7243 copy of the expression for the evaluator to munge.
7244
7245 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
7246 for the `read' function.
7247
7248
7249 ** The function `int?' has been removed; its definition was identical
7250 to that of `integer?'.
7251
7252 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
7253 use the R4RS names for these functions.
7254
7255 ** The function object-properties no longer returns the hash handle;
7256 it simply returns the object's property list.
7257
7258 ** Many functions have been changed to throw errors, instead of
7259 returning #f on failure. The point of providing exception handling in
7260 the language is to simplify the logic of user code, but this is less
7261 useful if Guile's primitives don't throw exceptions.
7262
7263 ** The function `fileno' has been renamed from `%fileno'.
7264
7265 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
7266
7267
7268 * Changes to Guile's C interface:
7269
7270 ** The library's initialization procedure has been simplified.
7271 scm_boot_guile now has the prototype:
7272
7273 void scm_boot_guile (int ARGC,
7274 char **ARGV,
7275 void (*main_func) (),
7276 void *closure);
7277
7278 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
7279 MAIN_FUNC should do all the work of the program (initializing other
7280 packages, reading user input, etc.) before returning. When MAIN_FUNC
7281 returns, call exit (0); this function never returns. If you want some
7282 other exit value, MAIN_FUNC may call exit itself.
7283
7284 scm_boot_guile arranges for program-arguments to return the strings
7285 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
7286 scm_set_program_arguments with the final list, so Scheme code will
7287 know which arguments have been processed.
7288
7289 scm_boot_guile establishes a catch-all catch handler which prints an
7290 error message and exits the process. This means that Guile exits in a
7291 coherent way when system errors occur and the user isn't prepared to
7292 handle it. If the user doesn't like this behavior, they can establish
7293 their own universal catcher in MAIN_FUNC to shadow this one.
7294
7295 Why must the caller do all the real work from MAIN_FUNC? The garbage
7296 collector assumes that all local variables of type SCM will be above
7297 scm_boot_guile's stack frame on the stack. If you try to manipulate
7298 SCM values after this function returns, it's the luck of the draw
7299 whether the GC will be able to find the objects you allocate. So,
7300 scm_boot_guile function exits, rather than returning, to discourage
7301 people from making that mistake.
7302
7303 The IN, OUT, and ERR arguments were removed; there are other
7304 convenient ways to override these when desired.
7305
7306 The RESULT argument was deleted; this function should never return.
7307
7308 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
7309 general.
7310
7311
7312 ** Guile's header files should no longer conflict with your system's
7313 header files.
7314
7315 In order to compile code which #included <libguile.h>, previous
7316 versions of Guile required you to add a directory containing all the
7317 Guile header files to your #include path. This was a problem, since
7318 Guile's header files have names which conflict with many systems'
7319 header files.
7320
7321 Now only <libguile.h> need appear in your #include path; you must
7322 refer to all Guile's other header files as <libguile/mumble.h>.
7323 Guile's installation procedure puts libguile.h in $(includedir), and
7324 the rest in $(includedir)/libguile.
7325
7326
7327 ** Two new C functions, scm_protect_object and scm_unprotect_object,
7328 have been added to the Guile library.
7329
7330 scm_protect_object (OBJ) protects OBJ from the garbage collector.
7331 OBJ will not be freed, even if all other references are dropped,
7332 until someone does scm_unprotect_object (OBJ). Both functions
7333 return OBJ.
7334
7335 Note that calls to scm_protect_object do not nest. You can call
7336 scm_protect_object any number of times on a given object, and the
7337 next call to scm_unprotect_object will unprotect it completely.
7338
7339 Basically, scm_protect_object and scm_unprotect_object just
7340 maintain a list of references to things. Since the GC knows about
7341 this list, all objects it mentions stay alive. scm_protect_object
7342 adds its argument to the list; scm_unprotect_object remove its
7343 argument from the list.
7344
7345
7346 ** scm_eval_0str now returns the value of the last expression
7347 evaluated.
7348
7349 ** The new function scm_read_0str reads an s-expression from a
7350 null-terminated string, and returns it.
7351
7352 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
7353 to a Scheme port object.
7354
7355 ** The new function `scm_set_program_arguments' allows C code to set
7356 the value returned by the Scheme `program-arguments' function.
7357
7358 \f
7359 Older changes:
7360
7361 * Guile no longer includes sophisticated Tcl/Tk support.
7362
7363 The old Tcl/Tk support was unsatisfying to us, because it required the
7364 user to link against the Tcl library, as well as Tk and Guile. The
7365 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
7366 referring to widgets by names, rather than exporting widgets to Scheme
7367 code as a special datatype.
7368
7369 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
7370 maintainers described some very interesting changes in progress to the
7371 Tcl/Tk internals, which would facilitate clean interfaces between lone
7372 Tk and other interpreters --- even for garbage-collected languages
7373 like Scheme. They expected the new Tk to be publicly available in the
7374 fall of 1996.
7375
7376 Since it seems that Guile might soon have a new, cleaner interface to
7377 lone Tk, and that the old Guile/Tk glue code would probably need to be
7378 completely rewritten, we (Jim Blandy and Richard Stallman) have
7379 decided not to support the old code. We'll spend the time instead on
7380 a good interface to the newer Tk, as soon as it is available.
7381
7382 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
7383
7384 \f
7385 Copyright information:
7386
7387 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
7388
7389 Permission is granted to anyone to make or distribute verbatim copies
7390 of this document as received, in any medium, provided that the
7391 copyright notice and this permission notice are preserved,
7392 thus giving the recipient permission to redistribute in turn.
7393
7394 Permission is granted to distribute modified versions
7395 of this document, or of portions of it,
7396 under the above conditions, provided also that they
7397 carry prominent notices stating who last changed them.
7398
7399 \f
7400 Local variables:
7401 mode: outline
7402 paragraph-separate: "[ \f]*$"
7403 end: