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.
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
12 * New modules (see the manual for details)
14 ** The `(ice-9 i18n)' module provides internationalization support
16 * Changes to the distribution
18 ** Guile now uses Gnulib as a portability aid
20 * Changes to the stand-alone interpreter
21 * Changes to Scheme functions and syntax
23 ** A new 'memoize-symbol evaluator trap has been added. This trap can
24 be used for efficiently implementing a Scheme code coverage.
26 ** Duplicate bindings among used modules are resolved lazily.
27 This slightly improves program startup times.
29 ** New thread cancellation and thread cleanup API
30 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
32 * Changes to the C interface
34 ** Functions for handling `scm_option' now no longer require an argument
35 indicating length of the `scm_t_option' array.
39 Changes in 1.8.4 (since 1.8.3)
43 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
44 ** Fixed a segmentation fault which occurred when displaying the
45 backtrace of a stack with a promise object (made by `delay') in it.
46 ** Make `accept' leave guile mode while blocking
47 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
48 ** Fixed a build problem on AIX (use of func_data identifier)
49 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
50 called with an associator proc that returns neither a pair nor #f.
51 ** Secondary threads now always return a valid module for (current-module).
52 ** Avoid MacOS build problems caused by incorrect combination of "64"
53 system and library calls.
54 ** Fixed build with Sun Studio (Solaris 9)
55 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
56 uniform vectors on AIX.
57 ** Fixed a deadlock that occurs upon GC with multiple threads.
58 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
60 * New modules (see the manual for details)
64 * Changes to the distribution
66 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
68 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
69 General Public License, and the few incorrect files have now been
70 fixed to agree with the rest of the Guile distribution.
72 ** Removed unnecessary extra copies of COPYING*
74 The distribution now contains a single COPYING.LESSER at its top level.
77 Changes in 1.8.3 (since 1.8.2)
79 * New modules (see the manual for details)
86 ** The `(ice-9 slib)' module now works as expected
87 ** Expressions like "(set! 'x #t)" no longer yield a crash
88 ** Warnings about duplicate bindings now go to stderr
89 ** A memory leak in `make-socket-address' was fixed
90 ** Alignment issues (e.g., on SPARC) in network routines were fixed
91 ** A threading issue that showed up at least on NetBSD was fixed
92 ** Build problems on Solaris and IRIX fixed
94 * Implementation improvements
96 ** The reader is now faster, which reduces startup time
97 ** Procedures returned by `record-accessor' and `record-modifier' are faster
101 Changes in 1.8.2 (since 1.8.1):
103 * New procedures (see the manual for details)
105 ** set-program-arguments
110 ** Fractions were not `equal?' if stored in unreduced form.
111 (A subtle problem, since printing a value reduced it, making it work.)
112 ** srfi-60 `copy-bit' failed on 64-bit systems
113 ** "guile --use-srfi" option at the REPL can replace core functions
114 (Programs run with that option were ok, but in the interactive REPL
115 the core bindings got priority, preventing SRFI replacements or
117 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
118 ** `kill' on mingw throws an error for a PID other than oneself
119 ** Procedure names are attached to procedure-with-setters
120 ** Array read syntax works with negative lower bound
121 ** `array-in-bounds?' fix if an array has different lower bounds on each index
122 ** `*' returns exact 0 for "(* inexact 0)"
123 This follows what it always did for "(* 0 inexact)".
124 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
125 ** SRFI-19: `date->julian-day' did not account for timezone offset
126 ** `ttyname' no longer crashes when passed a non-tty argument
127 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
128 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
129 ** GOOPS: Fixed a bug in `method-more-specific?'
130 ** Build problems on Solaris fixed
131 ** Build problems on HP-UX IA64 fixed
132 ** Build problems on MinGW fixed
135 Changes in 1.8.1 (since 1.8.0):
137 * LFS functions are now used to access 64-bit files on 32-bit systems.
139 * New procedures (see the manual for details)
141 ** primitive-_exit - [Scheme] the-root-module
142 ** scm_primitive__exit - [C]
143 ** make-completion-function - [Scheme] (ice-9 readline)
144 ** scm_c_locale_stringn_to_number - [C]
145 ** scm_srfi1_append_reverse [C]
146 ** scm_srfi1_append_reverse_x [C]
152 * New `(ice-9 i18n)' module (see the manual for details)
156 ** Build problems have been fixed on MacOS, SunOS, and QNX.
158 ** `strftime' fix sign of %z timezone offset.
160 ** A one-dimensional array can now be 'equal?' to a vector.
162 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
164 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
166 ** `record-accessor' and `record-modifier' now have strict type checks.
168 Record accessor and modifier procedures now throw an error if the
169 record type of the record they're given is not the type expected.
170 (Previously accessors returned #f and modifiers silently did nothing).
172 ** It is now OK to use both autoload and use-modules on a given module.
174 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
176 Previously there was no checking on primatives like make-vector that
177 accept "one or two" arguments. Now there is.
179 ** The srfi-1 assoc function now calls its equality predicate properly.
181 Previously srfi-1 assoc would call the equality predicate with the key
182 last. According to the SRFI, the key should be first.
184 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
186 ** The array-set! procedure no longer segfaults when given a bit vector.
188 ** Bugs in make-shared-array have been fixed.
190 ** string<? and friends now follow char<? etc order on 8-bit chars.
192 ** The format procedure now handles inf and nan values for ~f correctly.
194 ** exact->inexact should no longer overflow when given certain large fractions.
196 ** srfi-9 accessor and modifier procedures now have strict record type checks.
198 This matches the srfi-9 specification.
200 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
202 Previously the (ice-9 ftw) procedures would ignore any file that had
203 the same inode number as a file they had already seen, even if that
204 file was on a different device.
207 Changes in 1.8.0 (changes since the 1.6.x series):
209 * Changes to the distribution
211 ** Guile is now licensed with the GNU Lesser General Public License.
213 ** The manual is now licensed with the GNU Free Documentation License.
215 ** Guile now requires GNU MP (http://swox.com/gmp).
217 Guile now uses the GNU MP library for arbitrary precision arithmetic.
219 ** Guile now has separate private and public configuration headers.
221 That is, things like HAVE_STRING_H no longer leak from Guile's
224 ** Guile now provides and uses an "effective" version number.
226 Guile now provides scm_effective_version and effective-version
227 functions which return the "effective" version number. This is just
228 the normal full version string without the final micro-version number,
229 so the current effective-version is "1.8". The effective version
230 should remain unchanged during a stable series, and should be used for
231 items like the versioned share directory name
232 i.e. /usr/share/guile/1.8.
234 Providing an unchanging version number during a stable release for
235 things like the versioned share directory can be particularly
236 important for Guile "add-on" packages, since it provides a directory
237 that they can install to that won't be changed out from under them
238 with each micro release during a stable series.
240 ** Thread implementation has changed.
242 When you configure "--with-threads=null", you will get the usual
243 threading API (call-with-new-thread, make-mutex, etc), but you can't
244 actually create new threads. Also, "--with-threads=no" is now
245 equivalent to "--with-threads=null". This means that the thread API
246 is always present, although you might not be able to create new
249 When you configure "--with-threads=pthreads" or "--with-threads=yes",
250 you will get threads that are implemented with the portable POSIX
251 threads. These threads can run concurrently (unlike the previous
252 "coop" thread implementation), but need to cooperate for things like
255 The default is "pthreads", unless your platform doesn't have pthreads,
256 in which case "null" threads are used.
258 See the manual for details, nodes "Initialization", "Multi-Threading",
259 "Blocking", and others.
261 ** There is the new notion of 'discouraged' features.
263 This is a milder form of deprecation.
265 Things that are discouraged should not be used in new code, but it is
266 OK to leave them in old code for now. When a discouraged feature is
267 used, no warning message is printed like there is for 'deprecated'
268 features. Also, things that are merely discouraged are nevertheless
269 implemented efficiently, while deprecated features can be very slow.
271 You can omit discouraged features from libguile by configuring it with
272 the '--disable-discouraged' option.
274 ** Deprecation warnings can be controlled at run-time.
276 (debug-enable 'warn-deprecated) switches them on and (debug-disable
277 'warn-deprecated) switches them off.
279 ** Support for SRFI 61, extended cond syntax for multiple values has
282 This SRFI is always available.
284 ** Support for require-extension, SRFI-55, has been added.
286 The SRFI-55 special form `require-extension' has been added. It is
287 available at startup, and provides a portable way to load Scheme
288 extensions. SRFI-55 only requires support for one type of extension,
289 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
292 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
294 The (srfi srfi-26) module is an implementation of SRFI-26 which
295 provides the `cut' and `cute' syntax. These may be used to specialize
296 parameters without currying.
298 ** New module (srfi srfi-31)
300 This is an implementation of SRFI-31 which provides a special form
301 `rec' for recursive evaluation.
303 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
304 been merged with the core, making their functionality always
307 The modules are still available, tho, and you could use them together
308 with a renaming import, for example.
310 ** Guile no longer includes its own version of libltdl.
312 The official version is good enough now.
314 ** The --enable-htmldoc option has been removed from 'configure'.
316 Support for translating the documentation into HTML is now always
317 provided. Use 'make html'.
319 ** New module (ice-9 serialize):
321 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
322 don't trust the thread safety of most of your program, but where you
323 have some section(s) of code which you consider can run in parallel to
324 other sections. See ice-9/serialize.scm for more information.
326 ** The configure option '--disable-arrays' has been removed.
328 Support for arrays and uniform numeric arrays is now always included
331 * Changes to the stand-alone interpreter
333 ** New command line option `-L'.
335 This option adds a directory to the front of the load path.
337 ** New command line option `--no-debug'.
339 Specifying `--no-debug' on the command line will keep the debugging
340 evaluator turned off, even for interactive sessions.
342 ** User-init file ~/.guile is now loaded with the debugging evaluator.
344 Previously, the normal evaluator would have been used. Using the
345 debugging evaluator gives better error messages.
347 ** The '-e' option now 'read's its argument.
349 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
350 be used with '-e'. For example, you can now write a script like
353 exec guile -e '(@ (demo) main)' -s "$0" "$@"
356 (define-module (demo)
360 (format #t "Demo: ~a~%" args))
363 * Changes to Scheme functions and syntax
365 ** Guardians have changed back to their original semantics
367 Guardians now behave like described in the paper by Dybvig et al. In
368 particular, they no longer make guarantees about the order in which
369 they return objects, and they can no longer be greedy.
371 They no longer drop cyclic data structures.
373 The C function scm_make_guardian has been changed incompatibly and no
374 longer takes the 'greedy_p' argument.
376 ** New function hashx-remove!
378 This function completes the set of 'hashx' functions.
380 ** The concept of dynamic roots has been factored into continuation
381 barriers and dynamic states.
383 Each thread has a current dynamic state that carries the values of the
384 fluids. You can create and copy dynamic states and use them as the
385 second argument for 'eval'. See "Fluids and Dynamic States" in the
388 To restrict the influence that captured continuations can have on the
389 control flow, you can errect continuation barriers. See "Continuation
390 Barriers" in the manual.
392 The function call-with-dynamic-root now essentially temporarily
393 installs a new dynamic state and errects a continuation barrier.
395 ** The default load path no longer includes "." at the end.
397 Automatically loading modules from the current directory should not
398 happen by default. If you want to allow it in a more controlled
399 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
402 ** The uniform vector and array support has been overhauled.
404 It now complies with SRFI-4 and the weird prototype based uniform
405 array creation has been deprecated. See the manual for more details.
407 Some non-compatible changes have been made:
408 - characters can no longer be stored into byte arrays.
409 - strings and bit vectors are no longer considered to be uniform numeric
411 - array-rank throws an error for non-arrays instead of returning zero.
412 - array-ref does no longer accept non-arrays when no indices are given.
414 There is the new notion of 'generalized vectors' and corresponding
415 procedures like 'generalized-vector-ref'. Generalized vectors include
416 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
418 Arrays use generalized vectors as their storage, so that you still
419 have arrays of characters, bits, etc. However, uniform-array-read!
420 and uniform-array-write can no longer read/write strings and
423 ** There is now support for copy-on-write substrings, mutation-sharing
424 substrings and read-only strings.
426 Three new procedures are related to this: substring/shared,
427 substring/copy, and substring/read-only. See the manual for more
430 ** Backtraces will now highlight the value that caused the error.
432 By default, these values are enclosed in "{...}", such as in this
441 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
442 <unnamed port>:1:1: Wrong type (expecting pair): a
443 ABORT: (wrong-type-arg)
445 The prefix and suffix used for highlighting can be set via the two new
446 printer options 'highlight-prefix' and 'highlight-suffix'. For
447 example, putting this into ~/.guile will output the bad value in bold
450 (print-set! highlight-prefix "\x1b[1m")
451 (print-set! highlight-suffix "\x1b[22m")
454 ** 'gettext' support for internationalization has been added.
456 See the manual for details.
458 ** New syntax '@' and '@@':
460 You can now directly refer to variables exported from a module by
463 (@ MODULE-NAME VARIABLE-NAME)
465 For example (@ (ice-9 pretty-print) pretty-print) will directly access
466 the pretty-print variable exported from the (ice-9 pretty-print)
467 module. You don't need to 'use' that module first. You can also use
468 '@' as a target of 'set!', as in (set! (@ mod var) val).
470 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
471 but it can also access variables that have not been exported. It is
472 intended only for kluges and temporary fixes and for debugging, not
475 ** Keyword syntax has been made more disciplined.
477 Previously, the name of a keyword was read as a 'token' but printed as
478 a symbol. Now, it is read as a general Scheme datum which must be a
489 ERROR: In expression (a b c):
493 ERROR: Unbound variable: foo
498 ERROR: Wrong type (expecting symbol): 12
502 ERROR: Wrong type (expecting symbol): (a b c)
506 ** The printing of symbols that might look like keywords can be
509 The new printer option 'quote-keywordish-symbols' controls how symbols
510 are printed that have a colon as their first or last character. The
511 default now is to only quote a symbol with #{...}# when the read
512 option 'keywords' is not '#f'. Thus:
514 guile> (define foo (string->symbol ":foo"))
515 guile> (read-set! keywords #f)
518 guile> (read-set! keywords 'prefix)
521 guile> (print-set! quote-keywordish-symbols #f)
525 ** 'while' now provides 'break' and 'continue'
527 break and continue were previously bound in a while loop, but not
528 documented, and continue didn't quite work properly. The undocumented
529 parameter to break which gave a return value for the while has been
532 ** 'call-with-current-continuation' is now also available under the name
535 ** The module system now checks for duplicate bindings.
537 The module system now can check for name conflicts among imported
540 The behavior can be controlled by specifying one or more 'duplicates'
541 handlers. For example, to make Guile return an error for every name
549 The new default behavior of the module system when a name collision
550 has been detected is to
552 1. Give priority to bindings marked as a replacement.
553 2. Issue a warning (different warning if overriding core binding).
554 3. Give priority to the last encountered binding (this corresponds to
557 If you want the old behavior back without replacements or warnings you
560 (default-duplicate-binding-handler 'last)
562 to your .guile init file.
564 ** New define-module option: :replace
566 :replace works as :export, but, in addition, marks the binding as a
569 A typical example is `format' in (ice-9 format) which is a replacement
570 for the core binding `format'.
572 ** Adding prefixes to imported bindings in the module system
574 There is now a new :use-module option :prefix. It can be used to add
575 a prefix to all imported bindings.
578 :use-module ((bar) :prefix bar:))
580 will import all bindings exported from bar, but rename them by adding
583 ** Conflicting generic functions can be automatically merged.
585 When two imported bindings conflict and they are both generic
586 functions, the two functions can now be merged automatically. This is
587 activated with the 'duplicates' handler 'merge-generics'.
589 ** New function: effective-version
591 Returns the "effective" version number. This is just the normal full
592 version string without the final micro-version number. See "Changes
593 to the distribution" above.
595 ** New threading functions: parallel, letpar, par-map, and friends
597 These are convenient ways to run calculations in parallel in new
598 threads. See "Parallel forms" in the manual for details.
600 ** New function 'try-mutex'.
602 This function will attempt to lock a mutex but will return immediately
603 instead of blocking and indicate failure.
605 ** Waiting on a condition variable can have a timeout.
607 The function 'wait-condition-variable' now takes a third, optional
608 argument that specifies the point in time where the waiting should be
611 ** New function 'broadcast-condition-variable'.
613 ** New functions 'all-threads' and 'current-thread'.
615 ** Signals and system asyncs work better with threads.
617 The function 'sigaction' now takes a fourth, optional, argument that
618 specifies the thread that the handler should run in. When the
619 argument is omitted, the handler will run in the thread that called
622 Likewise, 'system-async-mark' takes a second, optional, argument that
623 specifies the thread that the async should run in. When it is
624 omitted, the async will run in the thread that called
627 C code can use the new functions scm_sigaction_for_thread and
628 scm_system_async_mark_for_thread to pass the new thread argument.
630 When a thread blocks on a mutex, a condition variable or is waiting
631 for IO to be possible, it will still execute system asyncs. This can
632 be used to interrupt such a thread by making it execute a 'throw', for
635 ** The function 'system-async' is deprecated.
637 You can now pass any zero-argument procedure to 'system-async-mark'.
638 The function 'system-async' will just return its argument unchanged
641 ** New functions 'call-with-blocked-asyncs' and
642 'call-with-unblocked-asyncs'
644 The expression (call-with-blocked-asyncs PROC) will call PROC and will
645 block execution of system asyncs for the current thread by one level
646 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
647 procedure and will unblock the execution of system asyncs by one
648 level for the current thread.
650 Only system asyncs are affected by these functions.
652 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
654 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
655 instead. Those functions are easier to use correctly and can be
658 ** New function 'unsetenv'.
660 ** New macro 'define-syntax-public'.
662 It works like 'define-syntax' and also exports the defined macro (but
665 ** There is support for Infinity and NaNs.
667 Following PLT Scheme, Guile can now work with infinite numbers, and
670 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
671 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
672 "+nan.0"). These numbers are inexact and have no exact counterpart.
674 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
675 sign of the dividend. The infinities are integers, and they answer #t
676 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
677 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
688 ERROR: Numerical overflow
690 Two new predicates 'inf?' and 'nan?' can be used to test for the
693 ** Inexact zero can have a sign.
695 Guile can now distinguish between plus and minus inexact zero, if your
696 platform supports this, too. The two zeros are equal according to
697 '=', but not according to 'eqv?'. For example
708 ** Guile now has exact rationals.
710 Guile can now represent fractions such as 1/3 exactly. Computing with
711 them is also done exactly, of course:
716 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
719 For example: (floor 2) now returns an exact 2 where in the past it
720 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
722 ** inexact->exact no longer returns only integers.
724 Without exact rationals, the closest exact number was always an
725 integer, but now inexact->exact returns the fraction that is exactly
726 equal to a floating point number. For example:
728 (inexact->exact 1.234)
729 => 694680242521899/562949953421312
731 When you want the old behavior, use 'round' explicitly:
733 (inexact->exact (round 1.234))
736 ** New function 'rationalize'.
738 This function finds a simple fraction that is close to a given real
739 number. For example (and compare with inexact->exact above):
741 (rationalize (inexact->exact 1.234) 1/2000)
744 Note that, as required by R5RS, rationalize returns only then an exact
745 result when both its arguments are exact.
747 ** 'odd?' and 'even?' work also for inexact integers.
749 Previously, (odd? 1.0) would signal an error since only exact integers
750 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
751 returns #f and (odd? 1.5) signals an error.
753 ** Guile now has uninterned symbols.
755 The new function 'make-symbol' will return an uninterned symbol. This
756 is a symbol that is unique and is guaranteed to remain unique.
757 However, uninterned symbols can not yet be read back in.
759 Use the new function 'symbol-interned?' to check whether a symbol is
762 ** pretty-print has more options.
764 The function pretty-print from the (ice-9 pretty-print) module can now
765 also be invoked with keyword arguments that control things like
766 maximum output width. See the manual for details.
768 ** Variables have no longer a special behavior for `equal?'.
770 Previously, comparing two variables with `equal?' would recursivly
771 compare their values. This is no longer done. Variables are now only
772 `equal?' if they are `eq?'.
774 ** `(begin)' is now valid.
776 You can now use an empty `begin' form. It will yield #<unspecified>
777 when evaluated and simply be ignored in a definition context.
779 ** Deprecated: procedure->macro
781 Change your code to use 'define-macro' or r5rs macros. Also, be aware
782 that macro expansion will not be done during evaluation, but prior to
785 ** Soft ports now allow a `char-ready?' procedure
787 The vector argument to `make-soft-port' can now have a length of
788 either 5 or 6. (Previously the length had to be 5.) The optional 6th
789 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
790 that returns the number of characters that can be read immediately
791 without the soft port blocking.
793 ** Deprecated: undefine
795 There is no replacement for undefine.
797 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
798 have been discouraged.
800 They are relics from a time where a keyword like #:foo was used
801 directly as a Tcl option "-foo" and thus keywords were internally
802 stored as a symbol with a starting dash. We now store a symbol
805 Use symbol->keyword and keyword->symbol instead.
807 ** The `cheap' debug option is now obsolete
809 Evaluator trap calls are now unconditionally "cheap" - in other words,
810 they pass a debug object to the trap handler rather than a full
811 continuation. The trap handler code can capture a full continuation
812 by using `call-with-current-continuation' in the usual way, if it so
815 The `cheap' option is retained for now so as not to break existing
816 code which gets or sets it, but setting it now has no effect. It will
817 be removed in the next major Guile release.
819 ** Evaluator trap calls now support `tweaking'
821 `Tweaking' means that the trap handler code can modify the Scheme
822 expression that is about to be evaluated (in the case of an
823 enter-frame trap) or the value that is being returned (in the case of
824 an exit-frame trap). The trap handler code indicates that it wants to
825 do this by returning a pair whose car is the symbol 'instead and whose
826 cdr is the modified expression or return value.
828 * Changes to the C interface
830 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
831 take a 'delete' function argument.
833 This argument makes no sense since the delete function is used to
834 remove a pair from an alist, and this must not be configurable.
836 This is an incompatible change.
838 ** The GH interface is now subject to the deprecation mechanism
840 The GH interface has been deprecated for quite some time but now it is
841 actually removed from Guile when it is configured with
842 --disable-deprecated.
844 See the manual "Transitioning away from GH" for more information.
846 ** A new family of functions for converting between C values and
847 Scheme values has been added.
849 These functions follow a common naming scheme and are designed to be
850 easier to use, thread-safe and more future-proof than the older
855 These are predicates that return a C boolean: 1 or 0. Instead of
856 SCM_NFALSEP, you can now use scm_is_true, for example.
858 - <type> scm_to_<type> (SCM val, ...)
860 These are functions that convert a Scheme value into an appropriate
861 C value. For example, you can use scm_to_int to safely convert from
864 - SCM scm_from_<type> (<type> val, ...)
866 These functions convert from a C type to a SCM value; for example,
867 scm_from_int for ints.
869 There is a huge number of these functions, for numbers, strings,
870 symbols, vectors, etc. They are documented in the reference manual in
871 the API section together with the types that they apply to.
873 ** New functions for dealing with complex numbers in C have been added.
875 The new functions are scm_c_make_rectangular, scm_c_make_polar,
876 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
877 They work like scm_make_rectangular etc but take or return doubles
880 ** The function scm_make_complex has been discouraged.
882 Use scm_c_make_rectangular instead.
884 ** The INUM macros have been deprecated.
886 A lot of code uses these macros to do general integer conversions,
887 although the macros only work correctly with fixnums. Use the
888 following alternatives.
890 SCM_INUMP -> scm_is_integer or similar
891 SCM_NINUMP -> !scm_is_integer or similar
892 SCM_MAKINUM -> scm_from_int or similar
893 SCM_INUM -> scm_to_int or similar
895 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
896 do the validating for you.
898 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
899 have been discouraged.
901 Use the newer scm_to_<type> and scm_from_<type> functions instead for
902 new code. The functions have been discouraged since they don't fit
905 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
907 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
908 evaluates its argument twice. Use scm_is_true, etc. instead for new
911 ** The macro SCM_EQ_P has been discouraged.
913 Use scm_is_eq for new code, which fits better into the naming
916 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
919 Use the function scm_is_pair or scm_is_null instead.
921 ** The functions scm_round and scm_truncate have been deprecated and
922 are now available as scm_c_round and scm_c_truncate, respectively.
924 These functions occupy the names that scm_round_number and
925 scm_truncate_number should have.
927 ** The functions scm_c_string2str, scm_c_substring2str, and
928 scm_c_symbol2str have been deprecated.
930 Use scm_to_locale_stringbuf or similar instead, maybe together with
933 ** New functions scm_c_make_string, scm_c_string_length,
934 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
935 scm_c_substring_shared, scm_c_substring_copy.
937 These are like scm_make_string, scm_length, etc. but are slightly
938 easier to use from C.
940 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
941 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
943 They export too many assumptions about the implementation of strings
944 and symbols that are no longer true in the presence of
945 mutation-sharing substrings and when Guile switches to some form of
948 When working with strings, it is often best to use the normal string
949 functions provided by Guile, such as scm_c_string_ref,
950 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
951 manual since many more such functions are now provided than
954 When you want to convert a SCM string to a C string, use the
955 scm_to_locale_string function or similar instead. For symbols, use
956 scm_symbol_to_string and then work with that string. Because of the
957 new string representation, scm_symbol_to_string does not need to copy
958 and is thus quite efficient.
960 ** Some string, symbol and keyword functions have been discouraged.
962 They don't fit into the uniform naming scheme and are not explicit
963 about the character encoding.
965 Replace according to the following table:
967 scm_allocate_string -> scm_c_make_string
968 scm_take_str -> scm_take_locale_stringn
969 scm_take0str -> scm_take_locale_string
970 scm_mem2string -> scm_from_locale_stringn
971 scm_str2string -> scm_from_locale_string
972 scm_makfrom0str -> scm_from_locale_string
973 scm_mem2symbol -> scm_from_locale_symboln
974 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
975 scm_str2symbol -> scm_from_locale_symbol
977 SCM_SYMBOL_HASH -> scm_hashq
978 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
980 scm_c_make_keyword -> scm_from_locale_keyword
982 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
983 now also available to C code.
985 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
987 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
988 the latter returns the true name of the keyword, not the 'dash name',
989 as SCM_KEYWORDSYM used to do.
991 ** A new way to access arrays in a thread-safe and efficient way has
994 See the manual, node "Accessing Arrays From C".
996 ** The old uniform vector and bitvector implementations have been
997 unceremoniously removed.
999 This implementation exposed the details of the tagging system of
1000 Guile. Use the new C API explained in the manual in node "Uniform
1001 Numeric Vectors" and "Bit Vectors", respectively.
1003 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
1004 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
1005 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
1006 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
1007 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
1008 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
1011 ** The macros dealing with vectors have been deprecated.
1013 Use the new functions scm_is_vector, scm_vector_elements,
1014 scm_vector_writable_elements, etc, or scm_is_simple_vector,
1015 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
1016 manual for more details.
1018 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
1019 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
1021 The following macros have been removed: SCM_VECTOR_BASE,
1022 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
1023 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
1025 ** Some C functions and macros related to arrays have been deprecated.
1027 Migrate according to the following table:
1029 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
1030 scm_make_ra -> scm_make_array
1031 scm_shap2ra -> scm_make_array
1032 scm_cvref -> scm_c_generalized_vector_ref
1033 scm_ra_set_contp -> do not use
1034 scm_aind -> scm_array_handle_pos
1035 scm_raprin1 -> scm_display or scm_write
1037 SCM_ARRAYP -> scm_is_array
1038 SCM_ARRAY_NDIM -> scm_c_array_rank
1039 SCM_ARRAY_DIMS -> scm_array_handle_dims
1040 SCM_ARRAY_CONTP -> do not use
1041 SCM_ARRAY_MEM -> do not use
1042 SCM_ARRAY_V -> scm_array_handle_elements or similar
1043 SCM_ARRAY_BASE -> do not use
1045 ** SCM_CELL_WORD_LOC has been deprecated.
1047 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
1048 to a SCM, as opposed to a pointer to a scm_t_bits.
1050 This was done to allow the correct use of pointers into the Scheme
1051 heap. Previously, the heap words were of type scm_t_bits and local
1052 variables and function arguments were of type SCM, making it
1053 non-standards-conformant to have a pointer that can point to both.
1055 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
1057 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
1058 second and third words of double smobs. Likewise for
1059 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
1061 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
1062 used to get and set the 16 exra bits in the zeroth word of a smob.
1064 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
1065 accesing the first immediate word of a smob as a SCM value, and there
1066 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
1067 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
1069 ** New way to deal with non-local exits and re-entries.
1071 There is a new set of functions that essentially do what
1072 scm_internal_dynamic_wind does, but in a way that is more convenient
1073 for C code in some situations. Here is a quick example of how to
1074 prevent a potential memory leak:
1081 scm_dynwind_begin (0);
1083 mem = scm_malloc (100);
1084 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
1086 /* MEM would leak if BAR throws an error.
1087 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
1094 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
1095 SCM_DYNWIND_END as well.
1099 For full documentation, see the node "Dynamic Wind" in the manual.
1101 ** New function scm_dynwind_free
1103 This function calls 'free' on a given pointer when a dynwind context
1104 is left. Thus the call to scm_dynwind_unwind_handler above could be
1105 replaced with simply scm_dynwind_free (mem).
1107 ** New functions scm_c_call_with_blocked_asyncs and
1108 scm_c_call_with_unblocked_asyncs
1110 Like scm_call_with_blocked_asyncs etc. but for C functions.
1112 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
1114 In addition to scm_c_call_with_blocked_asyncs you can now also use
1115 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
1116 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
1118 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
1119 SCM_REALLOW_INTS have been deprecated.
1121 They do no longer fulfill their original role of blocking signal
1122 delivery. Depending on what you want to achieve, replace a pair of
1123 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
1124 mutex, blocks asyncs, or both. See node "Critical Sections" in the
1127 ** The value 'scm_mask_ints' is no longer writable.
1129 Previously, you could set scm_mask_ints directly. This is no longer
1130 possible. Use scm_c_call_with_blocked_asyncs and
1131 scm_c_call_with_unblocked_asyncs instead.
1133 ** New way to temporarily set the current input, output or error ports
1135 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
1136 context' (see above). <foo> is one of "input", "output" or "error".
1138 ** New way to temporarily set fluids
1140 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
1141 above) to temporarily set the value of a fluid.
1143 ** New types scm_t_intmax and scm_t_uintmax.
1145 On platforms that have them, these types are identical to intmax_t and
1146 uintmax_t, respectively. On other platforms, they are identical to
1147 the largest integer types that Guile knows about.
1149 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
1151 You should not have used them.
1153 ** Many public #defines with generic names have been made private.
1155 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
1156 private or renamed with a more suitable public name.
1158 ** The macro SCM_TYP16S has been deprecated.
1160 This macro is not intended for public use.
1162 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
1164 Use scm_is_true (scm_inexact_p (...)) instead.
1166 ** The macro SCM_SLOPPY_REALP has been deprecated.
1168 Use scm_is_real instead.
1170 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
1172 Use scm_is_complex instead.
1174 ** Some preprocessor defines have been deprecated.
1176 These defines indicated whether a certain feature was present in Guile
1177 or not. Going forward, assume that the features are always present.
1179 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
1180 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
1182 The following macros have been removed completely: MEMOIZE_LOCALS,
1183 SCM_RECKLESS, SCM_CAUTIOUS.
1185 ** The preprocessor define STACK_DIRECTION has been deprecated.
1187 There should be no need to know about the stack direction for ordinary
1190 ** New function: scm_effective_version
1192 Returns the "effective" version number. This is just the normal full
1193 version string without the final micro-version number. See "Changes
1194 to the distribution" above.
1196 ** The function scm_call_with_new_thread has a new prototype.
1198 Instead of taking a list with the thunk and handler, these two
1199 arguments are now passed directly:
1201 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
1203 This is an incompatible change.
1205 ** New snarfer macro SCM_DEFINE_PUBLIC.
1207 This is like SCM_DEFINE, but also calls scm_c_export for the defined
1208 function in the init section.
1210 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
1212 ** Garbage collector rewrite.
1214 The garbage collector is cleaned up a lot, and now uses lazy
1215 sweeping. This is reflected in the output of (gc-stats); since cells
1216 are being freed when they are allocated, the cells-allocated field
1217 stays roughly constant.
1219 For malloc related triggers, the behavior is changed. It uses the same
1220 heuristic as the cell-triggered collections. It may be tuned with the
1221 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
1222 for minimum yield of malloc related triggers. The default is 40.
1223 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
1226 Debugging operations for the freelist have been deprecated, along with
1227 the C variables that control garbage collection. The environment
1228 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
1229 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
1231 For understanding the memory usage of a GUILE program, the routine
1232 gc-live-object-stats returns an alist containing the number of live
1233 objects for every type.
1236 ** The function scm_definedp has been renamed to scm_defined_p
1238 The name scm_definedp is deprecated.
1240 ** The struct scm_cell type has been renamed to scm_t_cell
1242 This is in accordance to Guile's naming scheme for types. Note that
1243 the name scm_cell is now used for a function that allocates and
1244 initializes a new cell (see below).
1246 ** New functions for memory management
1248 A new set of functions for memory management has been added since the
1249 old way (scm_must_malloc, scm_must_free, etc) was error prone and
1250 indeed, Guile itself contained some long standing bugs that could
1251 cause aborts in long running programs.
1253 The new functions are more symmetrical and do not need cooperation
1254 from smob free routines, among other improvements.
1256 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
1257 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
1258 scm_gc_free, scm_gc_register_collectable_memory, and
1259 scm_gc_unregister_collectable_memory. Refer to the manual for more
1260 details and for upgrading instructions.
1262 The old functions for memory management have been deprecated. They
1263 are: scm_must_malloc, scm_must_realloc, scm_must_free,
1264 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
1266 ** Declarations of exported features are marked with SCM_API.
1268 Every declaration of a feature that belongs to the exported Guile API
1269 has been marked by adding the macro "SCM_API" to the start of the
1270 declaration. This macro can expand into different things, the most
1271 common of which is just "extern" for Unix platforms. On Win32, it can
1272 be used to control which symbols are exported from a DLL.
1274 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
1275 will expand into "__declspec (dllimport) extern", which is needed for
1276 linking to the Guile DLL in Windows.
1278 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
1279 SCM_SRFI4_IMPORT, for the corresponding libraries.
1281 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
1283 Use the new functions scm_cell and scm_double_cell instead. The old
1284 macros had problems because with them allocation and initialization
1285 was separated and the GC could sometimes observe half initialized
1286 cells. Only careful coding by the user of SCM_NEWCELL and
1287 SCM_NEWCELL2 could make this safe and efficient.
1289 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
1291 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
1294 ** SRCBRKP has been deprecated.
1296 Use scm_c_source_property_breakpoint_p instead.
1298 ** Deprecated: scm_makmacro
1300 Change your code to use either scm_makmmacro or to define macros in
1301 Scheme, using 'define-macro'.
1303 ** New function scm_c_port_for_each.
1305 This function is like scm_port_for_each but takes a pointer to a C
1306 function as the callback instead of a SCM value.
1308 ** The names scm_internal_select, scm_thread_sleep, and
1309 scm_thread_usleep have been discouraged.
1311 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
1313 ** The GC can no longer be blocked.
1315 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
1316 The GC can now run (partially) concurrently with other code and thus
1317 blocking it is not well defined.
1319 ** Many definitions have been removed that were previously deprecated.
1321 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
1322 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
1323 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
1324 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
1325 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
1326 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
1327 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
1328 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
1329 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
1330 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
1331 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
1332 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
1333 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
1334 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
1335 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
1336 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
1337 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
1338 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
1339 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
1340 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
1341 scm_vector_set_length_x, scm_contregs, scm_debug_info,
1342 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
1343 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
1344 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
1345 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
1346 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
1347 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
1348 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
1349 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
1350 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
1351 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
1353 * Changes to bundled modules
1357 Using the (ice-9 debug) module no longer automatically switches Guile
1358 to use the debugging evaluator. If you want to switch to the
1359 debugging evaluator (which is needed for backtrace information if you
1360 hit an error), please add an explicit "(debug-enable 'debug)" to your
1361 code just after the code to use (ice-9 debug).
1364 Changes since Guile 1.4:
1366 * Changes to the distribution
1368 ** A top-level TODO file is included.
1370 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
1372 Guile now always uses three numbers to represent the version,
1373 i.e. "1.6.5". The first number, 1, is the major version number, the
1374 second number, 6, is the minor version number, and the third number,
1375 5, is the micro version number. Changes in major version number
1376 indicate major changes in Guile.
1378 Minor version numbers that are even denote stable releases, and odd
1379 minor version numbers denote development versions (which may be
1380 unstable). The micro version number indicates a minor sub-revision of
1381 a given MAJOR.MINOR release.
1383 In keeping with the new scheme, (minor-version) and scm_minor_version
1384 no longer return everything but the major version number. They now
1385 just return the minor version number. Two new functions
1386 (micro-version) and scm_micro_version have been added to report the
1387 micro version number.
1389 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
1391 ** New preprocessor definitions are available for checking versions.
1393 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
1394 SCM_MICRO_VERSION to the appropriate integer values.
1396 ** Guile now actively warns about deprecated features.
1398 The new configure option `--enable-deprecated=LEVEL' and the
1399 environment variable GUILE_WARN_DEPRECATED control this mechanism.
1400 See INSTALL and README for more information.
1402 ** Guile is much more likely to work on 64-bit architectures.
1404 Guile now compiles and passes "make check" with only two UNRESOLVED GC
1405 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
1406 for the use of a test machine, and thanks to Stefan Jahn for ia64
1409 ** New functions: setitimer and getitimer.
1411 These implement a fairly direct interface to the libc functions of the
1414 ** The #. reader extension is now disabled by default.
1416 For safety reasons, #. evaluation is disabled by default. To
1417 re-enable it, set the fluid read-eval? to #t. For example:
1419 (fluid-set! read-eval? #t)
1421 but make sure you realize the potential security risks involved. With
1422 read-eval? enabled, reading a data file from an untrusted source can
1425 ** New SRFI modules have been added:
1427 SRFI-0 `cond-expand' is now supported in Guile, without requiring
1430 (srfi srfi-1) is a library containing many useful pair- and list-processing
1433 (srfi srfi-2) exports and-let*.
1435 (srfi srfi-4) implements homogeneous numeric vector datatypes.
1437 (srfi srfi-6) is a dummy module for now, since guile already provides
1438 all of the srfi-6 procedures by default: open-input-string,
1439 open-output-string, get-output-string.
1441 (srfi srfi-8) exports receive.
1443 (srfi srfi-9) exports define-record-type.
1445 (srfi srfi-10) exports define-reader-ctor and implements the reader
1448 (srfi srfi-11) exports let-values and let*-values.
1450 (srfi srfi-13) implements the SRFI String Library.
1452 (srfi srfi-14) implements the SRFI Character-Set Library.
1454 (srfi srfi-17) implements setter and getter-with-setter and redefines
1455 some accessor procedures as procedures with getters. (such as car,
1456 cdr, vector-ref etc.)
1458 (srfi srfi-19) implements the SRFI Time/Date Library.
1460 ** New scripts / "executable modules"
1462 Subdirectory "scripts" contains Scheme modules that are packaged to
1463 also be executable as scripts. At this time, these scripts are available:
1472 See README there for more info.
1474 These scripts can be invoked from the shell with the new program
1475 "guile-tools", which keeps track of installation directory for you.
1478 $ guile-tools display-commentary srfi/*.scm
1480 guile-tools is copied to the standard $bindir on "make install".
1482 ** New module (ice-9 stack-catch):
1484 stack-catch is like catch, but saves the current state of the stack in
1485 the fluid the-last-stack. This fluid can be useful when using the
1486 debugger and when re-throwing an error.
1488 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
1490 This has been done to prevent problems on lesser operating systems
1491 that can't tolerate `*'s in file names. The exported macro continues
1492 to be named `and-let*', of course.
1494 On systems that support it, there is also a compatibility module named
1495 (ice-9 and-let*). It will go away in the next release.
1497 ** New modules (oop goops) etc.:
1500 (oop goops describe)
1502 (oop goops active-slot)
1503 (oop goops composite-slot)
1505 The Guile Object Oriented Programming System (GOOPS) has been
1506 integrated into Guile. For further information, consult the GOOPS
1507 manual and tutorial in the `doc' directory.
1509 ** New module (ice-9 rdelim).
1511 This exports the following procedures which were previously defined
1512 in the default environment:
1514 read-line read-line! read-delimited read-delimited! %read-delimited!
1515 %read-line write-line
1517 For backwards compatibility the definitions are still imported into the
1518 default environment in this version of Guile. However you should add:
1520 (use-modules (ice-9 rdelim))
1522 to any program which uses the definitions, since this may change in
1525 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
1526 can be used for similar functionality.
1528 ** New module (ice-9 rw)
1530 This is a subset of the (scsh rw) module from guile-scsh. Currently
1531 it defines two procedures:
1533 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
1535 Read characters from a port or file descriptor into a string STR.
1536 A port must have an underlying file descriptor -- a so-called
1537 fport. This procedure is scsh-compatible and can efficiently read
1540 *** New function: write-string/partial str [port_or_fdes [start [end]]]
1542 Write characters from a string STR to a port or file descriptor.
1543 A port must have an underlying file descriptor -- a so-called
1544 fport. This procedure is mostly compatible and can efficiently
1545 write large strings.
1547 ** New module (ice-9 match)
1549 This module includes Andrew K. Wright's pattern matcher. See
1550 ice-9/match.scm for brief description or
1552 http://www.star-lab.com/wright/code.html
1554 for complete documentation.
1556 ** New module (ice-9 buffered-input)
1558 This module provides procedures to construct an input port from an
1559 underlying source of input that reads and returns its input in chunks.
1560 The underlying input source is a Scheme procedure, specified by the
1561 caller, which the port invokes whenever it needs more input.
1563 This is useful when building an input port whose back end is Readline
1564 or a UI element such as the GtkEntry widget.
1568 The reference and tutorial documentation that was previously
1569 distributed separately, as `guile-doc', is now included in the core
1570 Guile distribution. The documentation consists of the following
1573 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
1576 - The Guile Reference Manual (guile.texi) contains (or is intended to
1577 contain) reference documentation on all aspects of Guile.
1579 - The GOOPS Manual (goops.texi) contains both tutorial-style and
1580 reference documentation for using GOOPS, Guile's Object Oriented
1583 - The Revised^5 Report on the Algorithmic Language Scheme
1586 See the README file in the `doc' directory for more details.
1588 ** There are a couple of examples in the examples/ directory now.
1590 * Changes to the stand-alone interpreter
1592 ** New command line option `--use-srfi'
1594 Using this option, SRFI modules can be loaded on startup and be
1595 available right from the beginning. This makes programming portable
1596 Scheme programs easier.
1598 The option `--use-srfi' expects a comma-separated list of numbers,
1599 each representing a SRFI number to be loaded into the interpreter
1600 before starting evaluating a script file or the REPL. Additionally,
1601 the feature identifier for the loaded SRFIs is recognized by
1602 `cond-expand' when using this option.
1605 $ guile --use-srfi=8,13
1606 guile> (receive (x z) (values 1 2) (+ 1 2))
1608 guile> (string-pad "bla" 20)
1611 ** Guile now always starts up in the `(guile-user)' module.
1613 Previously, scripts executed via the `-s' option would run in the
1614 `(guile)' module and the repl would run in the `(guile-user)' module.
1615 Now every user action takes place in the `(guile-user)' module by
1618 * Changes to Scheme functions and syntax
1620 ** Character classifiers work for non-ASCII characters.
1622 The predicates `char-alphabetic?', `char-numeric?',
1623 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
1624 no longer check whether their arguments are ASCII characters.
1625 Previously, a character would only be considered alphabetic when it
1626 was also ASCII, for example.
1628 ** Previously deprecated Scheme functions have been removed:
1630 tag - no replacement.
1631 fseek - replaced by seek.
1632 list* - replaced by cons*.
1634 ** It's now possible to create modules with controlled environments
1638 (use-modules (ice-9 safe))
1639 (define m (make-safe-module))
1640 ;;; m will now be a module containing only a safe subset of R5RS
1641 (eval '(+ 1 2) m) --> 3
1642 (eval 'load m) --> ERROR: Unbound variable: load
1644 ** Evaluation of "()", the empty list, is now an error.
1646 Previously, the expression "()" evaluated to the empty list. This has
1647 been changed to signal a "missing expression" error. The correct way
1648 to write the empty list as a literal constant is to use quote: "'()".
1650 ** New concept of `Guile Extensions'.
1652 A Guile Extension is just a ordinary shared library that can be linked
1653 at run-time. We found it advantageous to give this simple concept a
1654 dedicated name to distinguish the issues related to shared libraries
1655 from the issues related to the module system.
1657 *** New function: load-extension
1659 Executing (load-extension lib init) is mostly equivalent to
1661 (dynamic-call init (dynamic-link lib))
1663 except when scm_register_extension has been called previously.
1664 Whenever appropriate, you should use `load-extension' instead of
1665 dynamic-link and dynamic-call.
1667 *** New C function: scm_c_register_extension
1669 This function registers a initialization function for use by
1670 `load-extension'. Use it when you don't want specific extensions to
1671 be loaded as shared libraries (for example on platforms that don't
1672 support dynamic linking).
1674 ** Auto-loading of compiled-code modules is deprecated.
1676 Guile used to be able to automatically find and link a shared
1677 library to satisfy requests for a module. For example, the module
1678 `(foo bar)' could be implemented by placing a shared library named
1679 "foo/libbar.so" (or with a different extension) in a directory on the
1682 This has been found to be too tricky, and is no longer supported. The
1683 shared libraries are now called "extensions". You should now write a
1684 small Scheme file that calls `load-extension' to load the shared
1685 library and initialize it explicitly.
1687 The shared libraries themselves should be installed in the usual
1688 places for shared libraries, with names like "libguile-foo-bar".
1690 For example, place this into a file "foo/bar.scm"
1692 (define-module (foo bar))
1694 (load-extension "libguile-foo-bar" "foobar_init")
1696 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
1698 `eval' is now R5RS, that is it takes two arguments.
1699 The second argument is an environment specifier, i.e. either
1701 (scheme-report-environment 5)
1702 (null-environment 5)
1703 (interaction-environment)
1709 ** The module system has been made more disciplined.
1711 The function `eval' will save and restore the current module around
1712 the evaluation of the specified expression. While this expression is
1713 evaluated, `(current-module)' will now return the right module, which
1714 is the module specified as the second argument to `eval'.
1716 A consequence of this change is that `eval' is not particularly
1717 useful when you want allow the evaluated code to change what module is
1718 designated as the current module and have this change persist from one
1719 call to `eval' to the next. The read-eval-print-loop is an example
1720 where `eval' is now inadequate. To compensate, there is a new
1721 function `primitive-eval' that does not take a module specifier and
1722 that does not save/restore the current module. You should use this
1723 function together with `set-current-module', `current-module', etc
1724 when you want to have more control over the state that is carried from
1725 one eval to the next.
1727 Additionally, it has been made sure that forms that are evaluated at
1728 the top level are always evaluated with respect to the current module.
1729 Previously, subforms of top-level forms such as `begin', `case',
1730 etc. did not respect changes to the current module although these
1731 subforms are at the top-level as well.
1733 To prevent strange behavior, the forms `define-module',
1734 `use-modules', `use-syntax', and `export' have been restricted to only
1735 work on the top level. The forms `define-public' and
1736 `defmacro-public' only export the new binding on the top level. They
1737 behave just like `define' and `defmacro', respectively, when they are
1738 used in a lexical environment.
1740 Also, `export' will no longer silently re-export bindings imported
1741 from a used module. It will emit a `deprecation' warning and will
1742 cease to perform any re-export in the next version. If you actually
1743 want to re-export bindings, use the new `re-export' in place of
1744 `export'. The new `re-export' will not make copies of variables when
1745 rexporting them, as `export' did wrongly.
1747 ** Module system now allows selection and renaming of imported bindings
1749 Previously, when using `use-modules' or the `#:use-module' clause in
1750 the `define-module' form, all the bindings (association of symbols to
1751 values) for imported modules were added to the "current module" on an
1752 as-is basis. This has been changed to allow finer control through two
1753 new facilities: selection and renaming.
1755 You can now select which of the imported module's bindings are to be
1756 visible in the current module by using the `:select' clause. This
1757 clause also can be used to rename individual bindings. For example:
1759 ;; import all bindings no questions asked
1760 (use-modules (ice-9 common-list))
1762 ;; import four bindings, renaming two of them;
1763 ;; the current module sees: every some zonk-y zonk-n
1764 (use-modules ((ice-9 common-list)
1766 (remove-if . zonk-y)
1767 (remove-if-not . zonk-n))))
1769 You can also programmatically rename all selected bindings using the
1770 `:renamer' clause, which specifies a proc that takes a symbol and
1771 returns another symbol. Because it is common practice to use a prefix,
1772 we now provide the convenience procedure `symbol-prefix-proc'. For
1775 ;; import four bindings, renaming two of them specifically,
1776 ;; and all four w/ prefix "CL:";
1777 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
1778 (use-modules ((ice-9 common-list)
1780 (remove-if . zonk-y)
1781 (remove-if-not . zonk-n))
1782 :renamer (symbol-prefix-proc 'CL:)))
1784 ;; import four bindings, renaming two of them specifically,
1785 ;; and all four by upcasing.
1786 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
1787 (define (upcase-symbol sym)
1788 (string->symbol (string-upcase (symbol->string sym))))
1790 (use-modules ((ice-9 common-list)
1792 (remove-if . zonk-y)
1793 (remove-if-not . zonk-n))
1794 :renamer upcase-symbol))
1796 Note that programmatic renaming is done *after* individual renaming.
1797 Also, the above examples show `use-modules', but the same facilities are
1798 available for the `#:use-module' clause of `define-module'.
1800 See manual for more info.
1802 ** The semantics of guardians have changed.
1804 The changes are for the most part compatible. An important criterion
1805 was to keep the typical usage of guardians as simple as before, but to
1806 make the semantics safer and (as a result) more useful.
1808 *** All objects returned from guardians are now properly alive.
1810 It is now guaranteed that any object referenced by an object returned
1811 from a guardian is alive. It's now impossible for a guardian to
1812 return a "contained" object before its "containing" object.
1814 One incompatible (but probably not very important) change resulting
1815 from this is that it is no longer possible to guard objects that
1816 indirectly reference themselves (i.e. are parts of cycles). If you do
1817 so accidentally, you'll get a warning.
1819 *** There are now two types of guardians: greedy and sharing.
1821 If you call (make-guardian #t) or just (make-guardian), you'll get a
1822 greedy guardian, and for (make-guardian #f) a sharing guardian.
1824 Greedy guardians are the default because they are more "defensive".
1825 You can only greedily guard an object once. If you guard an object
1826 more than once, once in a greedy guardian and the rest of times in
1827 sharing guardians, then it is guaranteed that the object won't be
1828 returned from sharing guardians as long as it is greedily guarded
1831 Guardians returned by calls to `make-guardian' can now take one more
1832 optional parameter, which says whether to throw an error in case an
1833 attempt is made to greedily guard an object that is already greedily
1834 guarded. The default is true, i.e. throw an error. If the parameter
1835 is false, the guardian invocation returns #t if guarding was
1836 successful and #f if it wasn't.
1838 Also, since greedy guarding is, in effect, a side-effecting operation
1839 on objects, a new function is introduced: `destroy-guardian!'.
1840 Invoking this function on a guardian renders it unoperative and, if
1841 the guardian is greedy, clears the "greedily guarded" property of the
1842 objects that were guarded by it, thus undoing the side effect.
1844 Note that all this hair is hardly very important, since guardian
1845 objects are usually permanent.
1847 ** Continuations created by call-with-current-continuation now accept
1848 any number of arguments, as required by R5RS.
1850 ** New function `issue-deprecation-warning'
1852 This function is used to display the deprecation messages that are
1853 controlled by GUILE_WARN_DEPRECATION as explained in the README.
1856 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
1860 ;; `id' is deprecated. Use `identity' instead.
1865 ** New syntax `begin-deprecated'
1867 When deprecated features are included (as determined by the configure
1868 option --enable-deprecated), `begin-deprecated' is identical to
1869 `begin'. When deprecated features are excluded, it always evaluates
1870 to `#f', ignoring the body forms.
1872 ** New function `make-object-property'
1874 This function returns a new `procedure with setter' P that can be used
1875 to attach a property to objects. When calling P as
1879 where `obj' is any kind of object, it attaches `val' to `obj' in such
1880 a way that it can be retrieved by calling P as
1884 This function will replace procedure properties, symbol properties and
1885 source properties eventually.
1887 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
1889 Instead of #&optional, #&key, etc you should now use #:optional,
1890 #:key, etc. Since #:optional is a keyword, you can write it as just
1891 :optional when (read-set! keywords 'prefix) is active.
1893 The old reader syntax `#&' is still supported, but deprecated. It
1894 will be removed in the next release.
1896 ** New define-module option: pure
1898 Tells the module system not to include any bindings from the root
1903 (define-module (totally-empty-module)
1906 ** New define-module option: export NAME1 ...
1908 Export names NAME1 ...
1910 This option is required if you want to be able to export bindings from
1911 a module which doesn't import one of `define-public' or `export'.
1915 (define-module (foo)
1917 :use-module (ice-9 r5rs)
1920 ;;; Note that we're pure R5RS below this point!
1925 ** New function: object->string OBJ
1927 Return a Scheme string obtained by printing a given object.
1929 ** New function: port? X
1931 Returns a boolean indicating whether X is a port. Equivalent to
1932 `(or (input-port? X) (output-port? X))'.
1934 ** New function: file-port?
1936 Determines whether a given object is a port that is related to a file.
1938 ** New function: port-for-each proc
1940 Apply PROC to each port in the Guile port table in turn. The return
1941 value is unspecified. More specifically, PROC is applied exactly once
1942 to every port that exists in the system at the time PORT-FOR-EACH is
1943 invoked. Changes to the port table while PORT-FOR-EACH is running
1944 have no effect as far as PORT-FOR-EACH is concerned.
1946 ** New function: dup2 oldfd newfd
1948 A simple wrapper for the `dup2' system call. Copies the file
1949 descriptor OLDFD to descriptor number NEWFD, replacing the
1950 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
1951 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
1952 to move away ports which are using NEWFD. The return value is
1955 ** New function: close-fdes fd
1957 A simple wrapper for the `close' system call. Close file
1958 descriptor FD, which must be an integer. Unlike close (*note
1959 close: Ports and File Descriptors.), the file descriptor will be
1960 closed even if a port is using it. The return value is
1963 ** New function: crypt password salt
1965 Encrypts `password' using the standard unix password encryption
1968 ** New function: chroot path
1970 Change the root directory of the running process to `path'.
1972 ** New functions: getlogin, cuserid
1974 Return the login name or the user name of the current effective user
1977 ** New functions: getpriority which who, setpriority which who prio
1979 Get or set the priority of the running process.
1981 ** New function: getpass prompt
1983 Read a password from the terminal, first displaying `prompt' and
1986 ** New function: flock file operation
1988 Set/remove an advisory shared or exclusive lock on `file'.
1990 ** New functions: sethostname name, gethostname
1992 Set or get the hostname of the machine the current process is running
1995 ** New function: mkstemp! tmpl
1997 mkstemp creates a new unique file in the file system and returns a
1998 new buffered port open for reading and writing to the file. TMPL
1999 is a string specifying where the file should be created: it must
2000 end with `XXXXXX' and will be changed in place to return the name
2001 of the temporary file.
2003 ** New function: open-input-string string
2005 Return an input string port which delivers the characters from
2006 `string'. This procedure, together with `open-output-string' and
2007 `get-output-string' implements SRFI-6.
2009 ** New function: open-output-string
2011 Return an output string port which collects all data written to it.
2012 The data can then be retrieved by `get-output-string'.
2014 ** New function: get-output-string
2016 Return the contents of an output string port.
2018 ** New function: identity
2020 Return the argument.
2022 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
2023 are represented in Scheme as integers with normal host byte ordering.
2025 ** New function: inet-pton family address
2027 Convert a printable string network address into an integer. Note that
2028 unlike the C version of this function, the result is an integer with
2029 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2032 (inet-pton AF_INET "127.0.0.1") => 2130706433
2033 (inet-pton AF_INET6 "::1") => 1
2035 ** New function: inet-ntop family address
2037 Convert an integer network address into a printable string. Note that
2038 unlike the C version of this function, the input is an integer with
2039 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
2042 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
2043 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
2044 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
2048 Use `identity' instead.
2054 ** Deprecated: return-it
2058 ** Deprecated: string-character-length
2060 Use `string-length' instead.
2062 ** Deprecated: flags
2064 Use `logior' instead.
2066 ** Deprecated: close-all-ports-except.
2068 This was intended for closing ports in a child process after a fork,
2069 but it has the undesirable side effect of flushing buffers.
2070 port-for-each is more flexible.
2072 ** The (ice-9 popen) module now attempts to set up file descriptors in
2073 the child process from the current Scheme ports, instead of using the
2074 current values of file descriptors 0, 1, and 2 in the parent process.
2076 ** Removed function: builtin-weak-bindings
2078 There is no such concept as a weak binding any more.
2080 ** Removed constants: bignum-radix, scm-line-incrementors
2082 ** define-method: New syntax mandatory.
2084 The new method syntax is now mandatory:
2086 (define-method (NAME ARG-SPEC ...) BODY ...)
2087 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
2089 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
2090 REST-ARG ::= ARG-NAME
2092 If you have old code using the old syntax, import
2093 (oop goops old-define-method) before (oop goops) as in:
2095 (use-modules (oop goops old-define-method) (oop goops))
2097 ** Deprecated function: builtin-variable
2098 Removed function: builtin-bindings
2100 There is no longer a distinction between builtin or other variables.
2101 Use module system operations for all variables.
2103 ** Lazy-catch handlers are no longer allowed to return.
2105 That is, a call to `throw', `error', etc is now guaranteed to not
2108 ** Bugfixes for (ice-9 getopt-long)
2110 This module is now tested using test-suite/tests/getopt-long.test.
2111 The following bugs have been fixed:
2113 *** Parsing for options that are specified to have `optional' args now checks
2114 if the next element is an option instead of unconditionally taking it as the
2117 *** An error is now thrown for `--opt=val' when the option description
2118 does not specify `(value #t)' or `(value optional)'. This condition used to
2119 be accepted w/o error, contrary to the documentation.
2121 *** The error message for unrecognized options is now more informative.
2122 It used to be "not a record", an artifact of the implementation.
2124 *** The error message for `--opt' terminating the arg list (no value), when
2125 `(value #t)' is specified, is now more informative. It used to be "not enough
2128 *** "Clumped" single-char args now preserve trailing string, use it as arg.
2129 The expansion used to be like so:
2131 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
2133 Note that the "5d" is dropped. Now it is like so:
2135 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
2137 This enables single-char options to have adjoining arguments as long as their
2138 constituent characters are not potential single-char options.
2140 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
2142 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
2143 property `arglist', which can be retrieved by `arity'. The result is that
2144 `arity' can give more detailed information than before:
2148 guile> (use-modules (ice-9 optargs))
2149 guile> (define* (foo #:optional a b c) a)
2151 0 or more arguments in `lambda*:G0'.
2156 3 optional arguments: `a', `b' and `c'.
2157 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
2159 2 required arguments: `a' and `b', 2 keyword arguments: `c'
2160 and `d', other keywords allowed.
2161 guile> (define* (baz a b #:optional c #:rest r) a)
2163 2 required arguments: `a' and `b', 1 optional argument: `c',
2166 * Changes to the C interface
2168 ** Types have been renamed from scm_*_t to scm_t_*.
2170 This has been done for POSIX sake. It reserves identifiers ending
2171 with "_t". What a concept.
2173 The old names are still available with status `deprecated'.
2175 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
2177 ** Deprecated features have been removed.
2181 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
2182 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
2184 *** C Functions removed
2186 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
2187 scm_fseek - replaced by scm_seek.
2188 gc-thunk - replaced by after-gc-hook.
2189 gh_int2scmb - replaced by gh_bool2scm.
2190 scm_tc_dblr - replaced by scm_tc16_real.
2191 scm_tc_dblc - replaced by scm_tc16_complex.
2192 scm_list_star - replaced by scm_cons_star.
2194 ** Deprecated: scm_makfromstr
2196 Use scm_mem2string instead.
2198 ** Deprecated: scm_make_shared_substring
2200 Explicit shared substrings will disappear from Guile.
2202 Instead, "normal" strings will be implemented using sharing
2203 internally, combined with a copy-on-write strategy.
2205 ** Deprecated: scm_read_only_string_p
2207 The concept of read-only strings will disappear in next release of
2210 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
2212 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
2214 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
2216 Call a procedure with the indicated number of arguments. See "Fly
2217 Evaluation" in the manual.
2219 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
2221 Call a procedure with the indicated number of arguments and a list of
2222 further arguments. See "Fly Evaluation" in the manual.
2224 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
2226 Create a list of the given number of elements. See "List
2227 Constructors" in the manual.
2229 ** Renamed function: scm_listify has been replaced by scm_list_n.
2231 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
2232 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
2234 Use functions scm_list_N instead.
2236 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
2238 Used by an application to read arbitrary number of bytes from a port.
2239 Same semantics as libc read, except that scm_c_read only returns less
2240 than SIZE bytes if at end-of-file.
2242 Warning: Doesn't update port line and column counts!
2244 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
2246 Used by an application to write arbitrary number of bytes to an SCM
2247 port. Similar semantics as libc write. However, unlike libc
2248 write, scm_c_write writes the requested number of bytes and has no
2251 Warning: Doesn't update port line and column counts!
2253 ** New function: scm_init_guile ()
2255 In contrast to scm_boot_guile, scm_init_guile will return normally
2256 after initializing Guile. It is not available on all systems, tho.
2258 ** New functions: scm_str2symbol, scm_mem2symbol
2260 The function scm_str2symbol takes a const char* pointing to a zero-terminated
2261 field of characters and creates a scheme symbol object from that C string.
2262 The function scm_mem2symbol takes a const char* and a number of characters and
2263 creates a symbol from the characters in that memory area.
2265 ** New functions: scm_primitive_make_property
2266 scm_primitive_property_ref
2267 scm_primitive_property_set_x
2268 scm_primitive_property_del_x
2270 These functions implement a new way to deal with object properties.
2271 See libguile/properties.c for their documentation.
2273 ** New function: scm_done_free (long size)
2275 This function is the inverse of scm_done_malloc. Use it to report the
2276 amount of smob memory you free. The previous method, which involved
2277 calling scm_done_malloc with negative argument, was somewhat
2278 unintuitive (and is still available, of course).
2280 ** New function: scm_c_memq (SCM obj, SCM list)
2282 This function provides a fast C level alternative for scm_memq for the case
2283 that the list parameter is known to be a proper list. The function is a
2284 replacement for scm_sloppy_memq, but is stricter in its requirements on its
2285 list input parameter, since for anything else but a proper list the function's
2286 behaviour is undefined - it may even crash or loop endlessly. Further, for
2287 the case that the object is not found in the list, scm_c_memq returns #f which
2288 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
2290 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
2291 scm_remember_upto_here
2293 These functions replace the function scm_remember.
2295 ** Deprecated function: scm_remember
2297 Use one of the new functions scm_remember_upto_here_1,
2298 scm_remember_upto_here_2 or scm_remember_upto_here instead.
2300 ** New function: scm_allocate_string
2302 This function replaces the function scm_makstr.
2304 ** Deprecated function: scm_makstr
2306 Use the new function scm_allocate_string instead.
2308 ** New global variable scm_gc_running_p introduced.
2310 Use this variable to find out if garbage collection is being executed. Up to
2311 now applications have used scm_gc_heap_lock to test if garbage collection was
2312 running, which also works because of the fact that up to know only the garbage
2313 collector has set this variable. But, this is an implementation detail that
2314 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
2315 of this variable is (and has been) not fully safe anyway.
2317 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
2319 Use these instead of SCM_LENGTH_MAX.
2321 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
2322 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
2323 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
2325 Use these instead of SCM_LENGTH.
2327 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
2328 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
2329 SCM_SET_BITVECTOR_LENGTH
2331 Use these instead of SCM_SETLENGTH
2333 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
2334 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
2337 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
2340 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
2341 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
2344 Use these instead of SCM_SETCHARS.
2346 ** New macro: SCM_BITVECTOR_P
2348 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
2350 Use instead of SCM_COERCE_SUBSTR.
2352 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
2354 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
2356 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
2357 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
2358 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
2359 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
2360 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
2361 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
2362 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
2363 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
2364 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
2365 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
2366 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
2367 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
2368 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
2369 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
2370 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
2372 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
2373 Use scm_memory_error instead of SCM_NALLOC.
2374 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
2375 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
2376 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
2377 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
2378 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
2379 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
2380 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
2381 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
2382 Use SCM_STRINGP instead of SCM_RWSTRINGP.
2383 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
2384 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
2385 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
2386 Use a type specific setter macro instead of SCM_SETLENGTH.
2387 Use a type specific setter macro instead of SCM_SETCHARS.
2388 Use a type specific length macro instead of SCM_LENGTH_MAX.
2389 Use SCM_GCMARKP instead of SCM_GC8MARKP.
2390 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
2391 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
2392 Use SCM_TYP16 instead of SCM_GCTYP16.
2393 Use SCM_CDR instead of SCM_GCCDR.
2394 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
2395 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
2396 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
2397 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
2398 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
2399 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
2400 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
2402 ** Removed function: scm_struct_init
2404 ** Removed variable: scm_symhash_dim
2406 ** Renamed function: scm_make_cont has been replaced by
2407 scm_make_continuation, which has a different interface.
2409 ** Deprecated function: scm_call_catching_errors
2411 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
2413 ** Deprecated function: scm_strhash
2415 Use scm_string_hash instead.
2417 ** Deprecated function: scm_vector_set_length_x
2419 Instead, create a fresh vector of the desired size and copy the contents.
2421 ** scm_gensym has changed prototype
2423 scm_gensym now only takes one argument.
2425 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
2428 There is now only a single symbol type scm_tc7_symbol.
2429 The tag scm_tc7_lvector was not used anyway.
2431 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
2433 Use scm_make_smob_type and scm_set_smob_XXX instead.
2435 ** New function scm_set_smob_apply.
2437 This can be used to set an apply function to a smob type.
2439 ** Deprecated function: scm_strprint_obj
2441 Use scm_object_to_string instead.
2443 ** Deprecated function: scm_wta
2445 Use scm_wrong_type_arg, or another appropriate error signalling function
2448 ** Explicit support for obarrays has been deprecated.
2450 Use `scm_str2symbol' and the generic hashtable functions instead.
2452 ** The concept of `vcells' has been deprecated.
2454 The data type `variable' is now used exclusively. `Vcells' have been
2455 a low-level concept so you are likely not affected by this change.
2457 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
2458 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
2460 Use scm_c_define or scm_c_lookup instead, as appropriate.
2462 *** New functions: scm_c_module_lookup, scm_c_lookup,
2463 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
2464 scm_module_define, scm_define.
2466 These functions work with variables instead of with vcells.
2468 ** New functions for creating and defining `subr's and `gsubr's.
2470 The new functions more clearly distinguish between creating a subr (or
2471 gsubr) object and adding it to the current module.
2473 These new functions are available: scm_c_make_subr, scm_c_define_subr,
2474 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
2475 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
2476 scm_c_define_gsubr_with_generic.
2478 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
2479 scm_make_subr_with_generic, scm_make_gsubr,
2480 scm_make_gsubr_with_generic.
2482 Use the new ones from above instead.
2484 ** C interface to the module system has changed.
2486 While we suggest that you avoid as many explicit module system
2487 operations from C as possible for the time being, the C interface has
2488 been made more similar to the high-level Scheme module system.
2490 *** New functions: scm_c_define_module, scm_c_use_module,
2491 scm_c_export, scm_c_resolve_module.
2493 They mostly work like their Scheme namesakes. scm_c_define_module
2494 takes a function that is called a context where the new module is
2497 *** Deprecated functions: scm_the_root_module, scm_make_module,
2498 scm_ensure_user_module, scm_load_scheme_module.
2500 Use the new functions instead.
2502 ** Renamed function: scm_internal_with_fluids becomes
2505 scm_internal_with_fluids is available as a deprecated function.
2507 ** New function: scm_c_with_fluid.
2509 Just like scm_c_with_fluids, but takes one fluid and one value instead
2512 ** Deprecated typedefs: long_long, ulong_long.
2514 They are of questionable utility and they pollute the global
2517 ** Deprecated typedef: scm_sizet
2519 It is of questionable utility now that Guile requires ANSI C, and is
2522 ** Deprecated typedefs: scm_port_rw_active, scm_port,
2523 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
2524 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
2526 Made more compliant with the naming policy by adding a _t at the end.
2528 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
2529 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
2531 With the exception of the mysterious scm_2ulong2big, they are still
2532 available under new names (scm_i_mkbig etc). These functions are not
2533 intended to be used in user code. You should avoid dealing with
2534 bignums directly, and should deal with numbers in general (which can
2537 ** Change in behavior: scm_num2long, scm_num2ulong
2539 The scm_num2[u]long functions don't any longer accept an inexact
2540 argument. This change in behavior is motivated by concordance with
2541 R5RS: It is more common that a primitive doesn't want to accept an
2542 inexact for an exact.
2544 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
2545 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
2546 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
2549 These are conversion functions between the various ANSI C integral
2550 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
2551 accept an inexact argument.
2553 ** New functions: scm_float2num, scm_double2num,
2554 scm_num2float, scm_num2double.
2556 These are conversion functions between the two ANSI C float types and
2559 ** New number validation macros:
2560 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
2564 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
2566 These are just nicer-named old scm_protect_object and
2567 scm_unprotect_object.
2569 ** Deprecated functions: scm_protect_object, scm_unprotect_object
2571 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
2573 These functions can be used to register pointers to locations that
2576 ** Deprecated function: scm_create_hook.
2578 Its sins are: misleading name, non-modularity and lack of general
2582 Changes since Guile 1.3.4:
2584 * Changes to the distribution
2586 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
2588 We've changed the way we handle generated files in the Guile source
2589 repository. As a result, the procedure for building trees obtained
2590 from the nightly FTP snapshots or via CVS has changed:
2591 - You must have appropriate versions of autoconf, automake, and
2592 libtool installed on your system. See README for info on how to
2593 obtain these programs.
2594 - Before configuring the tree, you must first run the script
2595 `autogen.sh' at the top of the source tree.
2597 The Guile repository used to contain not only source files, written by
2598 humans, but also some generated files, like configure scripts and
2599 Makefile.in files. Even though the contents of these files could be
2600 derived mechanically from other files present, we thought it would
2601 make the tree easier to build if we checked them into CVS.
2603 However, this approach means that minor differences between
2604 developer's installed tools and habits affected the whole team.
2605 So we have removed the generated files from the repository, and
2606 added the autogen.sh script, which will reconstruct them
2610 ** configure now has experimental options to remove support for certain
2613 --disable-arrays omit array and uniform array support
2614 --disable-posix omit posix interfaces
2615 --disable-networking omit networking interfaces
2616 --disable-regex omit regular expression interfaces
2618 These are likely to become separate modules some day.
2620 ** New configure option --enable-debug-freelist
2622 This enables a debugging version of SCM_NEWCELL(), and also registers
2623 an extra primitive, the setter `gc-set-debug-check-freelist!'.
2625 Configure with the --enable-debug-freelist option to enable
2626 the gc-set-debug-check-freelist! primitive, and then use:
2628 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
2629 (gc-set-debug-check-freelist! #f) # turn off checking
2631 Checking of the freelist forces a traversal of the freelist and
2632 a garbage collection before each allocation of a cell. This can
2633 slow down the interpreter dramatically, so the setter should be used to
2634 turn on this extra processing only when necessary.
2636 ** New configure option --enable-debug-malloc
2638 Include code for debugging of calls to scm_must_malloc/realloc/free.
2642 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
2643 2. objects reallocated by scm_must_realloc has been allocated by
2645 3. reallocated objects are reallocated with the same what string
2647 But, most importantly, it records the number of allocated objects of
2648 each kind. This is useful when searching for memory leaks.
2650 A Guile compiled with this option provides the primitive
2651 `malloc-stats' which returns an alist with pairs of kind and the
2652 number of objects of that kind.
2654 ** All includes are now referenced relative to the root directory
2656 Since some users have had problems with mixups between Guile and
2657 system headers, we have decided to always refer to Guile headers via
2658 their parent directories. This essentially creates a "private name
2659 space" for Guile headers. This means that the compiler only is given
2660 -I options for the root build and root source directory.
2662 ** Header files kw.h and genio.h have been removed.
2664 ** The module (ice-9 getopt-gnu-style) has been removed.
2666 ** New module (ice-9 documentation)
2668 Implements the interface to documentation strings associated with
2671 ** New module (ice-9 time)
2673 Provides a macro `time', which displays execution time of a given form.
2675 ** New module (ice-9 history)
2677 Loading this module enables value history in the repl.
2679 * Changes to the stand-alone interpreter
2681 ** New command line option --debug
2683 Start Guile with debugging evaluator and backtraces enabled.
2685 This is useful when debugging your .guile init file or scripts.
2687 ** New help facility
2689 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
2690 (help REGEXP) ditto for objects with names matching REGEXP (a string)
2691 (help 'NAME) gives documentation for NAME, even if it is not an object
2692 (help ,EXPR) gives documentation for object returned by EXPR
2693 (help (my module)) gives module commentary for `(my module)'
2694 (help) gives this text
2696 `help' searches among bindings exported from loaded modules, while
2697 `apropos' searches among bindings visible from the "current" module.
2699 Examples: (help help)
2701 (help "output-string")
2703 ** `help' and `apropos' now prints full module names
2705 ** Dynamic linking now uses libltdl from the libtool package.
2707 The old system dependent code for doing dynamic linking has been
2708 replaced with calls to the libltdl functions which do all the hairy
2711 The major improvement is that you can now directly pass libtool
2712 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
2713 will be able to do the best shared library job you can get, via
2716 The way dynamic libraries are found has changed and is not really
2717 portable across platforms, probably. It is therefore recommended to
2718 use absolute filenames when possible.
2720 If you pass a filename without an extension to `dynamic-link', it will
2721 try a few appropriate ones. Thus, the most platform ignorant way is
2722 to specify a name like "libfoo", without any directories and
2725 ** Guile COOP threads are now compatible with LinuxThreads
2727 Previously, COOP threading wasn't possible in applications linked with
2728 Linux POSIX threads due to their use of the stack pointer to find the
2729 thread context. This has now been fixed with a workaround which uses
2730 the pthreads to allocate the stack.
2732 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
2734 ** Positions of erring expression in scripts
2736 With version 1.3.4, the location of the erring expression in Guile
2737 scipts is no longer automatically reported. (This should have been
2738 documented before the 1.3.4 release.)
2740 You can get this information by enabling recording of positions of
2741 source expressions and running the debugging evaluator. Put this at
2742 the top of your script (or in your "site" file):
2744 (read-enable 'positions)
2745 (debug-enable 'debug)
2747 ** Backtraces in scripts
2749 It is now possible to get backtraces in scripts.
2753 (debug-enable 'debug 'backtrace)
2755 at the top of the script.
2757 (The first options enables the debugging evaluator.
2758 The second enables backtraces.)
2760 ** Part of module system symbol lookup now implemented in C
2762 The eval closure of most modules is now implemented in C. Since this
2763 was one of the bottlenecks for loading speed, Guile now loads code
2764 substantially faster than before.
2766 ** Attempting to get the value of an unbound variable now produces
2767 an exception with a key of 'unbound-variable instead of 'misc-error.
2769 ** The initial default output port is now unbuffered if it's using a
2770 tty device. Previously in this situation it was line-buffered.
2772 ** New hook: after-gc-hook
2774 after-gc-hook takes over the role of gc-thunk. This hook is run at
2775 the first SCM_TICK after a GC. (Thus, the code is run at the same
2776 point during evaluation as signal handlers.)
2778 Note that this hook should be used only for diagnostic and debugging
2779 purposes. It is not certain that it will continue to be well-defined
2780 when this hook is run in the future.
2782 C programmers: Note the new C level hooks scm_before_gc_c_hook,
2783 scm_before_sweep_c_hook, scm_after_gc_c_hook.
2785 ** Improvements to garbage collector
2787 Guile 1.4 has a new policy for triggering heap allocation and
2788 determining the sizes of heap segments. It fixes a number of problems
2791 1. The new policy can handle two separate pools of cells
2792 (2-word/4-word) better. (The old policy would run wild, allocating
2793 more and more memory for certain programs.)
2795 2. The old code would sometimes allocate far too much heap so that the
2796 Guile process became gigantic. The new code avoids this.
2798 3. The old code would sometimes allocate too little so that few cells
2799 were freed at GC so that, in turn, too much time was spent in GC.
2801 4. The old code would often trigger heap allocation several times in a
2802 row. (The new scheme predicts how large the segments needs to be
2803 in order not to need further allocation.)
2805 All in all, the new GC policy will make larger applications more
2808 The new GC scheme also is prepared for POSIX threading. Threads can
2809 allocate private pools of cells ("clusters") with just a single
2810 function call. Allocation of single cells from such a cluster can
2811 then proceed without any need of inter-thread synchronization.
2813 ** New environment variables controlling GC parameters
2815 GUILE_MAX_SEGMENT_SIZE Maximal segment size
2818 Allocation of 2-word cell heaps:
2820 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
2823 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
2824 GC in percent of total heap size
2827 Allocation of 4-word cell heaps
2828 (used for real numbers and misc other objects):
2830 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
2832 (See entry "Way for application to customize GC parameters" under
2833 section "Changes to the scm_ interface" below.)
2835 ** Guile now implements reals using 4-word cells
2837 This speeds up computation with reals. (They were earlier allocated
2838 with `malloc'.) There is still some room for optimizations, however.
2840 ** Some further steps toward POSIX thread support have been taken
2842 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
2843 don't have much effect any longer, and many of them will be removed in
2847 are only handled at the top of the evaluator loop, immediately after
2848 I/O, and in scm_equalp.
2850 *** The GC can allocate thread private pools of pairs.
2852 * Changes to Scheme functions and syntax
2854 ** close-input-port and close-output-port are now R5RS
2856 These procedures have been turned into primitives and have R5RS behaviour.
2858 ** New procedure: simple-format PORT MESSAGE ARG1 ...
2860 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
2861 extended by the more sophisticated version in (ice-9 format)
2863 (simple-format port message . args)
2864 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
2865 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
2866 the escapes are replaced with corresponding members of ARGS:
2867 ~A formats using `display' and ~S formats using `write'.
2868 If DESTINATION is #t, then use the `current-output-port',
2869 if DESTINATION is #f, then return a string containing the formatted text.
2870 Does not add a trailing newline."
2872 ** string-ref: the second argument is no longer optional.
2874 ** string, list->string: no longer accept strings in their arguments,
2875 only characters, for compatibility with R5RS.
2877 ** New procedure: port-closed? PORT
2878 Returns #t if PORT is closed or #f if it is open.
2880 ** Deprecated: list*
2882 The list* functionality is now provided by cons* (SRFI-1 compliant)
2884 ** New procedure: cons* ARG1 ARG2 ... ARGn
2886 Like `list', but the last arg provides the tail of the constructed list,
2887 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
2889 Requires at least one argument. If given one argument, that argument
2890 is returned as result.
2892 This function is called `list*' in some other Schemes and in Common LISP.
2894 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
2896 ** New procedure: object-documentation OBJECT
2898 Returns the documentation string associated with OBJECT. The
2899 procedure uses a caching mechanism so that subsequent lookups are
2902 Exported by (ice-9 documentation).
2904 ** module-name now returns full names of modules
2906 Previously, only the last part of the name was returned (`session' for
2907 `(ice-9 session)'). Ex: `(ice-9 session)'.
2909 * Changes to the gh_ interface
2911 ** Deprecated: gh_int2scmb
2913 Use gh_bool2scm instead.
2915 * Changes to the scm_ interface
2917 ** Guile primitives now carry docstrings!
2919 Thanks to Greg Badros!
2921 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
2923 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
2924 macros and must contain a docstring that is extracted into foo.doc using a new
2925 guile-doc-snarf script (that uses guile-doc-snarf.awk).
2927 However, a major overhaul of these macros is scheduled for the next release of
2930 ** Guile primitives use a new technique for validation of arguments
2932 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
2933 the readability of argument checking.
2935 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
2937 ** New macros: SCM_PACK, SCM_UNPACK
2939 Compose/decompose an SCM value.
2941 The SCM type is now treated as an abstract data type and may be defined as a
2942 long, a void* or as a struct, depending on the architecture and compile time
2943 options. This makes it easier to find several types of bugs, for example when
2944 SCM values are treated as integers without conversion. Values of the SCM type
2945 should be treated as "atomic" values. These macros are used when
2946 composing/decomposing an SCM value, either because you want to access
2947 individual bits, or because you want to treat it as an integer value.
2949 E.g., in order to set bit 7 in an SCM value x, use the expression
2951 SCM_PACK (SCM_UNPACK (x) | 0x80)
2953 ** The name property of hooks is deprecated.
2954 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
2956 You can emulate this feature by using object properties.
2958 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
2959 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
2962 These macros will be removed in a future release of Guile.
2964 ** The following types, functions and macros from numbers.h are deprecated:
2965 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
2966 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
2968 ** Port internals: the rw_random variable in the scm_port structure
2969 must be set to non-zero in any random access port. In recent Guile
2970 releases it was only set for bidirectional random-access ports.
2972 ** Port internals: the seek ptob procedure is now responsible for
2973 resetting the buffers if required. The change was made so that in the
2974 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
2975 the fport and strport ptobs can avoid resetting the buffers,
2976 in particular to avoid discarding unread chars. An existing port
2977 type can be fixed by adding something like the following to the
2978 beginning of the ptob seek procedure:
2980 if (pt->rw_active == SCM_PORT_READ)
2981 scm_end_input (object);
2982 else if (pt->rw_active == SCM_PORT_WRITE)
2983 ptob->flush (object);
2985 although to actually avoid resetting the buffers and discard unread
2986 chars requires further hacking that depends on the characteristics
2989 ** Deprecated functions: scm_fseek, scm_tag
2991 These functions are no longer used and will be removed in a future version.
2993 ** The scm_sysmissing procedure is no longer used in libguile.
2994 Unless it turns out to be unexpectedly useful to somebody, it will be
2995 removed in a future version.
2997 ** The format of error message strings has changed
2999 The two C procedures: scm_display_error and scm_error, as well as the
3000 primitive `scm-error', now use scm_simple_format to do their work.
3001 This means that the message strings of all code must be updated to use
3002 ~A where %s was used before, and ~S where %S was used before.
3004 During the period when there still are a lot of old Guiles out there,
3005 you might want to support both old and new versions of Guile.
3007 There are basically two methods to achieve this. Both methods use
3010 AC_CHECK_FUNCS(scm_simple_format)
3012 in your configure.in.
3014 Method 1: Use the string concatenation features of ANSI C's
3019 #ifdef HAVE_SCM_SIMPLE_FORMAT
3025 Then represent each of your error messages using a preprocessor macro:
3027 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
3031 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
3032 (define make-message string-append)
3034 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
3036 Method 2: Use the oldfmt function found in doc/oldfmt.c.
3040 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
3045 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
3049 ** Deprecated: coop_mutex_init, coop_condition_variable_init
3051 Don't use the functions coop_mutex_init and
3052 coop_condition_variable_init. They will change.
3054 Use scm_mutex_init and scm_cond_init instead.
3056 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
3057 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
3058 COND, as `scm_cond_wait' does, but it also bounds the duration
3059 of the wait. If COND has not been signaled before time ABSTIME,
3060 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
3061 returns the error code `ETIMEDOUT'.
3063 The ABSTIME parameter specifies an absolute time, with the same
3064 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
3065 to 00:00:00 GMT, January 1, 1970.
3067 ** New function: scm_cond_broadcast (scm_cond_t *COND)
3068 `scm_cond_broadcast' restarts all the threads that are waiting
3069 on the condition variable COND. Nothing happens if no threads are
3072 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
3073 `scm_key_create' allocates a new TSD key. The key is stored in
3074 the location pointed to by KEY. There is no limit on the number
3075 of keys allocated at a given time. The value initially associated
3076 with the returned key is `NULL' in all currently executing threads.
3078 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
3079 function associated with the key. When a thread terminates,
3080 DESTR_FUNCTION is called on the value associated with the key in
3081 that thread. The DESTR_FUNCTION is not called if a key is deleted
3082 with `scm_key_delete' or a value is changed with
3083 `scm_setspecific'. The order in which destructor functions are
3084 called at thread termination time is unspecified.
3086 Destructors are not yet implemented.
3088 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
3089 `scm_setspecific' changes the value associated with KEY in the
3090 calling thread, storing the given POINTER instead.
3092 ** New function: scm_getspecific (scm_key_t KEY)
3093 `scm_getspecific' returns the value currently associated with
3094 KEY in the calling thread.
3096 ** New function: scm_key_delete (scm_key_t KEY)
3097 `scm_key_delete' deallocates a TSD key. It does not check
3098 whether non-`NULL' values are associated with that key in the
3099 currently executing threads, nor call the destructor function
3100 associated with the key.
3102 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
3104 Initialize a C level hook HOOK with associated HOOK_DATA and type
3105 TYPE. (See scm_c_hook_run ().)
3107 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
3109 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
3110 is true, add it last, otherwise first. The same FUNC can be added
3111 multiple times if FUNC_DATA differ and vice versa.
3113 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
3115 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
3116 function is only removed if both FUNC and FUNC_DATA matches.
3118 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
3120 Run hook HOOK passing DATA to the hook functions.
3122 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
3123 returned is undefined.
3125 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
3126 returns a non-NULL value. This value is returned as the result of
3127 scm_c_hook_run. If all functions return NULL, NULL is returned.
3129 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
3130 returns a NULL value, and NULL is returned. If all functions returns
3131 a non-NULL value, the last value is returned.
3133 ** New C level GC hooks
3135 Five new C level hooks has been added to the garbage collector.
3137 scm_before_gc_c_hook
3140 are run before locking and after unlocking the heap. The system is
3141 thus in a mode where evaluation can take place. (Except that
3142 scm_before_gc_c_hook must not allocate new cells.)
3144 scm_before_mark_c_hook
3145 scm_before_sweep_c_hook
3146 scm_after_sweep_c_hook
3148 are run when the heap is locked. These are intended for extension of
3149 the GC in a modular fashion. Examples are the weaks and guardians
3152 ** Way for application to customize GC parameters
3154 The application can set up other default values for the GC heap
3155 allocation parameters
3157 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
3158 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
3159 GUILE_MAX_SEGMENT_SIZE,
3163 scm_default_init_heap_size_1, scm_default_min_yield_1,
3164 scm_default_init_heap_size_2, scm_default_min_yield_2,
3165 scm_default_max_segment_size
3167 respectively before callong scm_boot_guile.
3169 (See entry "New environment variables ..." in section
3170 "Changes to the stand-alone interpreter" above.)
3172 ** scm_protect_object/scm_unprotect_object now nest
3174 This means that you can call scm_protect_object multiple times on an
3175 object and count on the object being protected until
3176 scm_unprotect_object has been call the same number of times.
3178 The functions also have better time complexity.
3180 Still, it is usually possible to structure the application in a way
3181 that you don't need to use these functions. For example, if you use a
3182 protected standard Guile list to keep track of live objects rather
3183 than some custom data type, objects will die a natural death when they
3184 are no longer needed.
3186 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
3188 Guile does not provide the float representation for inexact real numbers any
3189 more. Now, only doubles are used to represent inexact real numbers. Further,
3190 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
3191 and scm_tc16_complex, respectively.
3193 ** Removed deprecated type scm_smobfuns
3195 ** Removed deprecated function scm_newsmob
3197 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
3199 There is an ongoing discussion among the developers whether to
3200 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
3201 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
3202 until this issue has been settled.
3204 ** Removed deprecated type tag scm_tc16_kw
3206 ** Added type tag scm_tc16_keyword
3208 (This was introduced already in release 1.3.4 but was not documented
3211 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
3213 * Changes to system call interfaces:
3215 ** The "select" procedure now tests port buffers for the ability to
3216 provide input or accept output. Previously only the underlying file
3217 descriptors were checked.
3219 ** New variable PIPE_BUF: the maximum number of bytes that can be
3220 atomically written to a pipe.
3222 ** If a facility is not available on the system when Guile is
3223 compiled, the corresponding primitive procedure will not be defined.
3224 Previously it would have been defined but would throw a system-error
3225 exception if called. Exception handlers which catch this case may
3226 need minor modification: an error will be thrown with key
3227 'unbound-variable instead of 'system-error. Alternatively it's
3228 now possible to use `defined?' to check whether the facility is
3231 ** Procedures which depend on the timezone should now give the correct
3232 result on systems which cache the TZ environment variable, even if TZ
3233 is changed without calling tzset.
3235 * Changes to the networking interfaces:
3237 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
3238 long integers between network and host format. For now, it's not
3239 particularly convenient to do this kind of thing, but consider:
3241 (define write-network-long
3242 (lambda (value port)
3243 (let ((v (make-uniform-vector 1 1 0)))
3244 (uniform-vector-set! v 0 (htonl value))
3245 (uniform-vector-write v port))))
3247 (define read-network-long
3249 (let ((v (make-uniform-vector 1 1 0)))
3250 (uniform-vector-read! v port)
3251 (ntohl (uniform-vector-ref v 0)))))
3253 ** If inet-aton fails, it now throws an error with key 'misc-error
3254 instead of 'system-error, since errno is not relevant.
3256 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
3257 specific keys instead of 'system-error. The latter is inappropriate
3258 since errno will not have been set. The keys are:
3259 'host-not-found, 'try-again, 'no-recovery and 'no-data.
3261 ** sethostent, setnetent, setprotoent, setservent: now take an
3262 optional argument STAYOPEN, which specifies whether the database
3263 remains open after a database entry is accessed randomly (e.g., using
3264 gethostbyname for the hosts database.) The default is #f. Previously
3268 Changes since Guile 1.3.2:
3270 * Changes to the stand-alone interpreter
3274 An initial version of the Guile debugger written by Chris Hanson has
3275 been added. The debugger is still under development but is included
3276 in the distribution anyway since it is already quite useful.
3282 after an error to enter the debugger. Type `help' inside the debugger
3283 for a description of available commands.
3285 If you prefer to have stack frames numbered and printed in
3286 anti-chronological order and prefer up in the stack to be down on the
3287 screen as is the case in gdb, you can put
3289 (debug-enable 'backwards)
3291 in your .guile startup file. (However, this means that Guile can't
3292 use indentation to indicate stack level.)
3294 The debugger is autoloaded into Guile at the first use.
3296 ** Further enhancements to backtraces
3298 There is a new debug option `width' which controls the maximum width
3299 on the screen of printed stack frames. Fancy printing parameters
3300 ("level" and "length" as in Common LISP) are adaptively adjusted for
3301 each stack frame to give maximum information while still fitting
3302 within the bounds. If the stack frame can't be made to fit by
3303 adjusting parameters, it is simply cut off at the end. This is marked
3306 ** Some modules are now only loaded when the repl is started
3308 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
3309 regex) are now loaded into (guile-user) only if the repl has been
3310 started. The effect is that the startup time for scripts has been
3311 reduced to 30% of what it was previously.
3313 Correctly written scripts load the modules they require at the top of
3314 the file and should not be affected by this change.
3316 ** Hooks are now represented as smobs
3318 * Changes to Scheme functions and syntax
3320 ** Readline support has changed again.
3322 The old (readline-activator) module is gone. Use (ice-9 readline)
3323 instead, which now contains all readline functionality. So the code
3324 to activate readline is now
3326 (use-modules (ice-9 readline))
3329 This should work at any time, including from the guile prompt.
3331 To avoid confusion about the terms of Guile's license, please only
3332 enable readline for your personal use; please don't make it the
3333 default for others. Here is why we make this rather odd-sounding
3336 Guile is normally licensed under a weakened form of the GNU General
3337 Public License, which allows you to link code with Guile without
3338 placing that code under the GPL. This exception is important to some
3341 However, since readline is distributed under the GNU General Public
3342 License, when you link Guile with readline, either statically or
3343 dynamically, you effectively change Guile's license to the strict GPL.
3344 Whenever you link any strictly GPL'd code into Guile, uses of Guile
3345 which are normally permitted become forbidden. This is a rather
3346 non-obvious consequence of the licensing terms.
3348 So, to make sure things remain clear, please let people choose for
3349 themselves whether to link GPL'd libraries like readline with Guile.
3351 ** regexp-substitute/global has changed slightly, but incompatibly.
3353 If you include a function in the item list, the string of the match
3354 object it receives is the same string passed to
3355 regexp-substitute/global, not some suffix of that string.
3356 Correspondingly, the match's positions are relative to the entire
3357 string, not the suffix.
3359 If the regexp can match the empty string, the way matches are chosen
3360 from the string has changed. regexp-substitute/global recognizes the
3361 same set of matches that list-matches does; see below.
3363 ** New function: list-matches REGEXP STRING [FLAGS]
3365 Return a list of match objects, one for every non-overlapping, maximal
3366 match of REGEXP in STRING. The matches appear in left-to-right order.
3367 list-matches only reports matches of the empty string if there are no
3368 other matches which begin on, end at, or include the empty match's
3371 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
3373 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
3375 For each match of REGEXP in STRING, apply PROC to the match object,
3376 and the last value PROC returned, or INIT for the first call. Return
3377 the last value returned by PROC. We apply PROC to the matches as they
3378 appear from left to right.
3380 This function recognizes matches according to the same criteria as
3383 Thus, you could define list-matches like this:
3385 (define (list-matches regexp string . flags)
3386 (reverse! (apply fold-matches regexp string '() cons flags)))
3388 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
3392 *** New function: hook? OBJ
3394 Return #t if OBJ is a hook, otherwise #f.
3396 *** New function: make-hook-with-name NAME [ARITY]
3398 Return a hook with name NAME and arity ARITY. The default value for
3399 ARITY is 0. The only effect of NAME is that it will appear when the
3400 hook object is printed to ease debugging.
3402 *** New function: hook-empty? HOOK
3404 Return #t if HOOK doesn't contain any procedures, otherwise #f.
3406 *** New function: hook->list HOOK
3408 Return a list of the procedures that are called when run-hook is
3411 ** `map' signals an error if its argument lists are not all the same length.
3413 This is the behavior required by R5RS, so this change is really a bug
3414 fix. But it seems to affect a lot of people's code, so we're
3415 mentioning it here anyway.
3417 ** Print-state handling has been made more transparent
3419 Under certain circumstances, ports are represented as a port with an
3420 associated print state. Earlier, this pair was represented as a pair
3421 (see "Some magic has been added to the printer" below). It is now
3422 indistinguishable (almost; see `get-print-state') from a port on the
3425 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
3427 Return a new port with the associated print state PRINT-STATE.
3429 *** New function: get-print-state OUTPUT-PORT
3431 Return the print state associated with this port if it exists,
3432 otherwise return #f.
3434 *** New function: directory-stream? OBJECT
3436 Returns true iff OBJECT is a directory stream --- the sort of object
3437 returned by `opendir'.
3439 ** New function: using-readline?
3441 Return #t if readline is in use in the current repl.
3443 ** structs will be removed in 1.4
3445 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
3446 and use GOOPS objects as the fundamental record type.
3448 * Changes to the scm_ interface
3450 ** structs will be removed in 1.4
3452 The entire current struct interface (struct.c, struct.h) will be
3453 replaced in Guile 1.4. We will merge GOOPS into libguile and use
3454 GOOPS objects as the fundamental record type.
3456 ** The internal representation of subr's has changed
3458 Instead of giving a hint to the subr name, the CAR field of the subr
3459 now contains an index to a subr entry in scm_subr_table.
3461 *** New variable: scm_subr_table
3463 An array of subr entries. A subr entry contains the name, properties
3464 and documentation associated with the subr. The properties and
3465 documentation slots are not yet used.
3467 ** A new scheme for "forwarding" calls to a builtin to a generic function
3469 It is now possible to extend the functionality of some Guile
3470 primitives by letting them defer a call to a GOOPS generic function on
3471 argument mismatch. This means that there is no loss of efficiency in
3476 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
3477 (define-method + ((x <string>) (y <string>))
3478 (string-append x y))
3480 + will still be as efficient as usual in numerical calculations, but
3481 can also be used for concatenating strings.
3483 Who will be the first one to extend Guile's numerical tower to
3484 rationals? :) [OK, there a few other things to fix before this can
3485 be made in a clean way.]
3487 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
3489 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
3491 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
3493 These do the same job as SCM_PROC and SCM_PROC1, but they also define
3494 a variable GENERIC which can be used by the dispatch macros below.
3496 [This is experimental code which may change soon.]
3498 *** New macros for forwarding control to a generic on arg type error
3500 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
3502 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
3504 These correspond to the scm_wta function call, and have the same
3505 behaviour until the user has called the GOOPS primitive
3506 `enable-primitive-generic!'. After that, these macros will apply the
3507 generic function GENERIC to the argument(s) instead of calling
3510 [This is experimental code which may change soon.]
3512 *** New macros for argument testing with generic dispatch
3514 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
3516 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
3518 These correspond to the SCM_ASSERT macro, but will defer control to
3519 GENERIC on error after `enable-primitive-generic!' has been called.
3521 [This is experimental code which may change soon.]
3523 ** New function: SCM scm_eval_body (SCM body, SCM env)
3525 Evaluates the body of a special form.
3527 ** The internal representation of struct's has changed
3529 Previously, four slots were allocated for the procedure(s) of entities
3530 and operators. The motivation for this representation had to do with
3531 the structure of the evaluator, the wish to support tail-recursive
3532 generic functions, and efficiency. Since the generic function
3533 dispatch mechanism has changed, there is no longer a need for such an
3534 expensive representation, and the representation has been simplified.
3536 This should not make any difference for most users.
3538 ** GOOPS support has been cleaned up.
3540 Some code has been moved from eval.c to objects.c and code in both of
3541 these compilation units has been cleaned up and better structured.
3543 *** New functions for applying generic functions
3545 New function: SCM scm_apply_generic (GENERIC, ARGS)
3546 New function: SCM scm_call_generic_0 (GENERIC)
3547 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
3548 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
3549 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
3551 ** Deprecated function: scm_make_named_hook
3553 It is now replaced by:
3555 ** New function: SCM scm_create_hook (const char *name, int arity)
3557 Creates a hook in the same way as make-hook above but also
3558 binds a variable named NAME to it.
3560 This is the typical way of creating a hook from C code.
3562 Currently, the variable is created in the "current" module.
3563 This might change when we get the new module system.
3565 [The behaviour is identical to scm_make_named_hook.]
3569 Changes since Guile 1.3:
3571 * Changes to mailing lists
3573 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
3575 See the README file to find current addresses for all the Guile
3578 * Changes to the distribution
3580 ** Readline support is no longer included with Guile by default.
3582 Based on the different license terms of Guile and Readline, we
3583 concluded that Guile should not *by default* cause the linking of
3584 Readline into an application program. Readline support is now offered
3585 as a separate module, which is linked into an application only when
3586 you explicitly specify it.
3588 Although Guile is GNU software, its distribution terms add a special
3589 exception to the usual GNU General Public License (GPL). Guile's
3590 license includes a clause that allows you to link Guile with non-free
3591 programs. We add this exception so as not to put Guile at a
3592 disadvantage vis-a-vis other extensibility packages that support other
3595 In contrast, the GNU Readline library is distributed under the GNU
3596 General Public License pure and simple. This means that you may not
3597 link Readline, even dynamically, into an application unless it is
3598 distributed under a free software license that is compatible the GPL.
3600 Because of this difference in distribution terms, an application that
3601 can use Guile may not be able to use Readline. Now users will be
3602 explicitly offered two independent decisions about the use of these
3605 You can activate the readline support by issuing
3607 (use-modules (readline-activator))
3610 from your ".guile" file, for example.
3612 * Changes to the stand-alone interpreter
3614 ** All builtins now print as primitives.
3615 Previously builtin procedures not belonging to the fundamental subr
3616 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
3617 Now, they print as #<primitive-procedure NAME>.
3619 ** Backtraces slightly more intelligible.
3620 gsubr-apply and macro transformer application frames no longer appear
3623 * Changes to Scheme functions and syntax
3625 ** Guile now correctly handles internal defines by rewriting them into
3626 their equivalent letrec. Previously, internal defines would
3627 incrementally add to the innermost environment, without checking
3628 whether the restrictions specified in RnRS were met. This lead to the
3629 correct behaviour when these restriction actually were met, but didn't
3630 catch all illegal uses. Such an illegal use could lead to crashes of
3631 the Guile interpreter or or other unwanted results. An example of
3632 incorrect internal defines that made Guile behave erratically:
3644 The problem with this example is that the definition of `c' uses the
3645 value of `b' directly. This confuses the meoization machine of Guile
3646 so that the second call of `b' (this time in a larger environment that
3647 also contains bindings for `c' and `d') refers to the binding of `c'
3648 instead of `a'. You could also make Guile crash with a variation on
3653 (define (b flag) (if flag a 1))
3654 (define c (1+ (b flag)))
3662 From now on, Guile will issue an `Unbound variable: b' error message
3667 A hook contains a list of functions which should be called on
3668 particular occasions in an existing program. Hooks are used for
3671 A window manager might have a hook before-window-map-hook. The window
3672 manager uses the function run-hooks to call all functions stored in
3673 before-window-map-hook each time a window is mapped. The user can
3674 store functions in the hook using add-hook!.
3676 In Guile, hooks are first class objects.
3678 *** New function: make-hook [N_ARGS]
3680 Return a hook for hook functions which can take N_ARGS arguments.
3681 The default value for N_ARGS is 0.
3683 (See also scm_make_named_hook below.)
3685 *** New function: add-hook! HOOK PROC [APPEND_P]
3687 Put PROC at the beginning of the list of functions stored in HOOK.
3688 If APPEND_P is supplied, and non-false, put PROC at the end instead.
3690 PROC must be able to take the number of arguments specified when the
3693 If PROC already exists in HOOK, then remove it first.
3695 *** New function: remove-hook! HOOK PROC
3697 Remove PROC from the list of functions in HOOK.
3699 *** New function: reset-hook! HOOK
3701 Clear the list of hook functions stored in HOOK.
3703 *** New function: run-hook HOOK ARG1 ...
3705 Run all hook functions stored in HOOK with arguments ARG1 ... .
3706 The number of arguments supplied must correspond to the number given
3707 when the hook was created.
3709 ** The function `dynamic-link' now takes optional keyword arguments.
3710 The only keyword argument that is currently defined is `:global
3711 BOOL'. With it, you can control whether the shared library will be
3712 linked in global mode or not. In global mode, the symbols from the
3713 linked library can be used to resolve references from other
3714 dynamically linked libraries. In non-global mode, the linked
3715 library is essentially invisible and can only be accessed via
3716 `dynamic-func', etc. The default is now to link in global mode.
3717 Previously, the default has been non-global mode.
3719 The `#:global' keyword is only effective on platforms that support
3720 the dlopen family of functions.
3722 ** New function `provided?'
3724 - Function: provided? FEATURE
3725 Return true iff FEATURE is supported by this installation of
3726 Guile. FEATURE must be a symbol naming a feature; the global
3727 variable `*features*' is a list of available features.
3729 ** Changes to the module (ice-9 expect):
3731 *** The expect-strings macro now matches `$' in a regular expression
3732 only at a line-break or end-of-file by default. Previously it would
3733 match the end of the string accumulated so far. The old behaviour
3734 can be obtained by setting the variable `expect-strings-exec-flags'
3737 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
3738 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
3739 in a regular expression will still match before a line-break or
3740 end-of-file. The default is `regexp/noteol'.
3742 *** The expect-strings macro now uses a variable
3743 `expect-strings-compile-flags' for the flags to be supplied to
3744 `make-regexp'. The default is `regexp/newline', which was previously
3747 *** The expect macro now supplies two arguments to a match procedure:
3748 the current accumulated string and a flag to indicate whether
3749 end-of-file has been reached. Previously only the string was supplied.
3750 If end-of-file is reached, the match procedure will be called an
3751 additional time with the same accumulated string as the previous call
3752 but with the flag set.
3754 ** New module (ice-9 format), implementing the Common Lisp `format' function.
3756 This code, and the documentation for it that appears here, was
3757 borrowed from SLIB, with minor adaptations for Guile.
3759 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
3760 An almost complete implementation of Common LISP format description
3761 according to the CL reference book `Common LISP' from Guy L.
3762 Steele, Digital Press. Backward compatible to most of the
3763 available Scheme format implementations.
3765 Returns `#t', `#f' or a string; has side effect of printing
3766 according to FORMAT-STRING. If DESTINATION is `#t', the output is
3767 to the current output port and `#t' is returned. If DESTINATION
3768 is `#f', a formatted string is returned as the result of the call.
3769 NEW: If DESTINATION is a string, DESTINATION is regarded as the
3770 format string; FORMAT-STRING is then the first argument and the
3771 output is returned as a string. If DESTINATION is a number, the
3772 output is to the current error port if available by the
3773 implementation. Otherwise DESTINATION must be an output port and
3776 FORMAT-STRING must be a string. In case of a formatting error
3777 format returns `#f' and prints a message on the current output or
3778 error port. Characters are output as if the string were output by
3779 the `display' function with the exception of those prefixed by a
3780 tilde (~). For a detailed description of the FORMAT-STRING syntax
3781 please consult a Common LISP format reference manual. For a test
3782 suite to verify this format implementation load `formatst.scm'.
3783 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
3785 Note: `format' is not reentrant, i.e. only one `format'-call may
3786 be executed at a time.
3789 *** Format Specification (Format version 3.0)
3791 Please consult a Common LISP format reference manual for a detailed
3792 description of the format string syntax. For a demonstration of the
3793 implemented directives see `formatst.scm'.
3795 This implementation supports directive parameters and modifiers (`:'
3796 and `@' characters). Multiple parameters must be separated by a comma
3797 (`,'). Parameters can be numerical parameters (positive or negative),
3798 character parameters (prefixed by a quote character (`''), variable
3799 parameters (`v'), number of rest arguments parameter (`#'), empty and
3800 default parameters. Directive characters are case independent. The
3801 general form of a directive is:
3803 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
3805 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
3807 *** Implemented CL Format Control Directives
3809 Documentation syntax: Uppercase characters represent the
3810 corresponding control directive characters. Lowercase characters
3811 represent control directive parameter descriptions.
3814 Any (print as `display' does).
3818 `~MINCOL,COLINC,MINPAD,PADCHARA'
3822 S-expression (print as `write' does).
3826 `~MINCOL,COLINC,MINPAD,PADCHARS'
3832 print number sign always.
3835 print comma separated.
3837 `~MINCOL,PADCHAR,COMMACHARD'
3843 print number sign always.
3846 print comma separated.
3848 `~MINCOL,PADCHAR,COMMACHARX'
3854 print number sign always.
3857 print comma separated.
3859 `~MINCOL,PADCHAR,COMMACHARO'
3865 print number sign always.
3868 print comma separated.
3870 `~MINCOL,PADCHAR,COMMACHARB'
3875 `~N,MINCOL,PADCHAR,COMMACHARR'
3879 print a number as a Roman numeral.
3882 print a number as an "old fashioned" Roman numeral.
3885 print a number as an ordinal English number.
3888 print a number as a cardinal English number.
3893 prints `y' and `ies'.
3896 as `~P but jumps 1 argument backward.'
3899 as `~@P but jumps 1 argument backward.'
3904 prints a character as the reader can understand it (i.e. `#\'
3908 prints a character as emacs does (eg. `^C' for ASCII 03).
3911 Fixed-format floating-point (prints a flonum like MMM.NNN).
3912 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
3914 If the number is positive a plus sign is printed.
3917 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
3918 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
3920 If the number is positive a plus sign is printed.
3923 General floating-point (prints a flonum either fixed or
3925 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
3927 If the number is positive a plus sign is printed.
3930 Dollars floating-point (prints a flonum in fixed with signs
3932 `~DIGITS,SCALE,WIDTH,PADCHAR$'
3934 If the number is positive a plus sign is printed.
3937 A sign is always printed and appears before the padding.
3940 The sign appears before the padding.
3948 print newline if not at the beginning of the output line.
3950 prints `~&' and then N-1 newlines.
3955 print N page separators.
3965 newline is ignored, white space left.
3968 newline is left, white space ignored.
3973 relative tabulation.
3979 Indirection (expects indirect arguments as a list).
3981 extracts indirect arguments from format arguments.
3984 Case conversion (converts by `string-downcase').
3986 converts by `string-capitalize'.
3989 converts by `string-capitalize-first'.
3992 converts by `string-upcase'.
3995 Argument Jumping (jumps 1 argument forward).
3997 jumps N arguments forward.
4000 jumps 1 argument backward.
4003 jumps N arguments backward.
4006 jumps to the 0th argument.
4009 jumps to the Nth argument (beginning from 0)
4011 `~[STR0~;STR1~;...~;STRN~]'
4012 Conditional Expression (numerical clause conditional).
4014 take argument from N.
4017 true test conditional.
4020 if-else-then conditional.
4026 default clause follows.
4029 Iteration (args come from the next argument (a list)).
4031 at most N iterations.
4034 args from next arg (a list of lists).
4037 args from the rest of arguments.
4040 args from the rest args (lists).
4051 aborts if N <= M <= K
4053 *** Not Implemented CL Format Control Directives
4056 print `#f' as an empty list (see below).
4059 print `#f' as an empty list (see below).
4065 (sorry I don't understand its semantics completely)
4067 *** Extended, Replaced and Additional Control Directives
4069 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
4070 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
4071 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
4072 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
4073 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
4074 COMMAWIDTH is the number of characters between two comma
4078 print a R4RS complex number as `~F~@Fi' with passed parameters for
4082 Pretty print formatting of an argument for scheme code lists.
4088 Flushes the output if format DESTINATION is a port.
4091 Print a `#\space' character
4093 print N `#\space' characters.
4096 Print a `#\tab' character
4098 print N `#\tab' characters.
4101 Takes N as an integer representation for a character. No arguments
4102 are consumed. N is converted to a character by `integer->char'. N
4103 must be a positive decimal number.
4106 Print out readproof. Prints out internal objects represented as
4107 `#<...>' as strings `"#<...>"' so that the format output can always
4108 be processed by `read'.
4111 Print out readproof. Prints out internal objects represented as
4112 `#<...>' as strings `"#<...>"' so that the format output can always
4113 be processed by `read'.
4116 Prints information and a copyright notice on the format
4119 prints format version.
4122 may also print number strings, i.e. passing a number as a string
4123 and format it accordingly.
4125 *** Configuration Variables
4127 The format module exports some configuration variables to suit the
4128 systems and users needs. There should be no modification necessary for
4129 the configuration that comes with Guile. Format detects automatically
4130 if the running scheme system implements floating point numbers and
4133 format:symbol-case-conv
4134 Symbols are converted by `symbol->string' so the case type of the
4135 printed symbols is implementation dependent.
4136 `format:symbol-case-conv' is a one arg closure which is either
4137 `#f' (no conversion), `string-upcase', `string-downcase' or
4138 `string-capitalize'. (default `#f')
4140 format:iobj-case-conv
4141 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
4142 implementation internal objects. (default `#f')
4145 The character prefixing the exponent value in `~E' printing.
4148 *** Compatibility With Other Format Implementations
4154 Downward compatible except for padding support and `~A', `~S',
4155 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
4156 `printf' padding support which is completely replaced by the CL
4157 `format' padding style.
4160 Downward compatible except for `~', which is not documented
4161 (ignores all characters inside the format string up to a newline
4162 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
4163 numerical and variable parameters and `:/@' modifiers in the CL
4167 Downward compatible except for `~A' and `~S' which print in
4168 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
4169 directive parameters or modifiers)).
4172 Downward compatible except for an optional destination parameter:
4173 S2C accepts a format call without a destination which returns a
4174 formatted string. This is equivalent to a #f destination in S2C.
4175 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
4176 parameters or modifiers)).
4179 ** Changes to string-handling functions.
4181 These functions were added to support the (ice-9 format) module, above.
4183 *** New function: string-upcase STRING
4184 *** New function: string-downcase STRING
4186 These are non-destructive versions of the existing string-upcase! and
4187 string-downcase! functions.
4189 *** New function: string-capitalize! STRING
4190 *** New function: string-capitalize STRING
4192 These functions convert the first letter of each word in the string to
4195 (string-capitalize "howdy there")
4198 As with the other functions, string-capitalize! modifies the string in
4199 place, while string-capitalize returns a modified copy of its argument.
4201 *** New function: string-ci->symbol STRING
4203 Return a symbol whose name is STRING, but having the same case as if
4204 the symbol had be read by `read'.
4206 Guile can be configured to be sensitive or insensitive to case
4207 differences in Scheme identifiers. If Guile is case-insensitive, all
4208 symbols are converted to lower case on input. The `string-ci->symbol'
4209 function returns a symbol whose name in STRING, transformed as Guile
4210 would if STRING were input.
4212 *** New function: substring-move! STRING1 START END STRING2 START
4214 Copy the substring of STRING1 from START (inclusive) to END
4215 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
4216 string, and the source and destination areas may overlap; in all
4217 cases, the function behaves as if all the characters were copied
4220 *** Extended functions: substring-move-left! substring-move-right!
4222 These functions now correctly copy arbitrarily overlapping substrings;
4223 they are both synonyms for substring-move!.
4226 ** New module (ice-9 getopt-long), with the function `getopt-long'.
4228 getopt-long is a function for parsing command-line arguments in a
4229 manner consistent with other GNU programs.
4231 (getopt-long ARGS GRAMMAR)
4232 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
4234 ARGS should be a list of strings. Its first element should be the
4235 name of the program; subsequent elements should be the arguments
4236 that were passed to the program on the command line. The
4237 `program-arguments' procedure returns a list of this form.
4239 GRAMMAR is a list of the form:
4240 ((OPTION (PROPERTY VALUE) ...) ...)
4242 Each OPTION should be a symbol. `getopt-long' will accept a
4243 command-line option named `--OPTION'.
4244 Each option can have the following (PROPERTY VALUE) pairs:
4246 (single-char CHAR) --- Accept `-CHAR' as a single-character
4247 equivalent to `--OPTION'. This is how to specify traditional
4249 (required? BOOL) --- If BOOL is true, the option is required.
4250 getopt-long will raise an error if it is not found in ARGS.
4251 (value BOOL) --- If BOOL is #t, the option accepts a value; if
4252 it is #f, it does not; and if it is the symbol
4253 `optional', the option may appear in ARGS with or
4255 (predicate FUNC) --- If the option accepts a value (i.e. you
4256 specified `(value #t)' for this option), then getopt
4257 will apply FUNC to the value, and throw an exception
4258 if it returns #f. FUNC should be a procedure which
4259 accepts a string and returns a boolean value; you may
4260 need to use quasiquotes to get it into GRAMMAR.
4262 The (PROPERTY VALUE) pairs may occur in any order, but each
4263 property may occur only once. By default, options do not have
4264 single-character equivalents, are not required, and do not take
4267 In ARGS, single-character options may be combined, in the usual
4268 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
4269 accepts values, then it must be the last option in the
4270 combination; the value is the next argument. So, for example, using
4271 the following grammar:
4272 ((apples (single-char #\a))
4273 (blimps (single-char #\b) (value #t))
4274 (catalexis (single-char #\c) (value #t)))
4275 the following argument lists would be acceptable:
4276 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
4277 for "blimps" and "catalexis")
4278 ("-ab" "bang" "-c" "couth") (same)
4279 ("-ac" "couth" "-b" "bang") (same)
4280 ("-abc" "couth" "bang") (an error, since `-b' is not the
4281 last option in its combination)
4283 If an option's value is optional, then `getopt-long' decides
4284 whether it has a value by looking at what follows it in ARGS. If
4285 the next element is a string, and it does not appear to be an
4286 option itself, then that string is the option's value.
4288 The value of a long option can appear as the next element in ARGS,
4289 or it can follow the option name, separated by an `=' character.
4290 Thus, using the same grammar as above, the following argument lists
4292 ("--apples" "Braeburn" "--blimps" "Goodyear")
4293 ("--apples=Braeburn" "--blimps" "Goodyear")
4294 ("--blimps" "Goodyear" "--apples=Braeburn")
4296 If the option "--" appears in ARGS, argument parsing stops there;
4297 subsequent arguments are returned as ordinary arguments, even if
4298 they resemble options. So, in the argument list:
4299 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
4300 `getopt-long' will recognize the `apples' option as having the
4301 value "Granny Smith", but it will not recognize the `blimp'
4302 option; it will return the strings "--blimp" and "Goodyear" as
4303 ordinary argument strings.
4305 The `getopt-long' function returns the parsed argument list as an
4306 assocation list, mapping option names --- the symbols from GRAMMAR
4307 --- onto their values, or #t if the option does not accept a value.
4308 Unused options do not appear in the alist.
4310 All arguments that are not the value of any option are returned
4311 as a list, associated with the empty list.
4313 `getopt-long' throws an exception if:
4314 - it finds an unrecognized option in ARGS
4315 - a required option is omitted
4316 - an option that requires an argument doesn't get one
4317 - an option that doesn't accept an argument does get one (this can
4318 only happen using the long option `--opt=value' syntax)
4319 - an option predicate fails
4324 `((lockfile-dir (required? #t)
4327 (predicate ,file-is-directory?))
4328 (verbose (required? #f)
4331 (x-includes (single-char #\x))
4332 (rnet-server (single-char #\y)
4333 (predicate ,string?))))
4335 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
4336 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
4338 => ((() "foo1" "-fred" "foo2" "foo3")
4339 (rnet-server . "lamprod")
4340 (x-includes . "/usr/include")
4341 (lockfile-dir . "/tmp")
4344 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
4346 It will be removed in a few releases.
4348 ** New syntax: lambda*
4349 ** New syntax: define*
4350 ** New syntax: define*-public
4351 ** New syntax: defmacro*
4352 ** New syntax: defmacro*-public
4353 Guile now supports optional arguments.
4355 `lambda*', `define*', `define*-public', `defmacro*' and
4356 `defmacro*-public' are identical to the non-* versions except that
4357 they use an extended type of parameter list that has the following BNF
4358 syntax (parentheses are literal, square brackets indicate grouping,
4359 and `*', `+' and `?' have the usual meaning):
4361 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
4362 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
4363 [[#&rest identifier]|[. identifier]]? ) | [identifier]
4365 ext-var-decl ::= identifier | ( identifier expression )
4367 The semantics are best illustrated with the following documentation
4368 and examples for `lambda*':
4371 lambda extended for optional and keyword arguments
4373 lambda* creates a procedure that takes optional arguments. These
4374 are specified by putting them inside brackets at the end of the
4375 paramater list, but before any dotted rest argument. For example,
4376 (lambda* (a b #&optional c d . e) '())
4377 creates a procedure with fixed arguments a and b, optional arguments c
4378 and d, and rest argument e. If the optional arguments are omitted
4379 in a call, the variables for them are unbound in the procedure. This
4380 can be checked with the bound? macro.
4382 lambda* can also take keyword arguments. For example, a procedure
4384 (lambda* (#&key xyzzy larch) '())
4385 can be called with any of the argument lists (#:xyzzy 11)
4386 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
4387 are given as keywords are bound to values.
4389 Optional and keyword arguments can also be given default values
4390 which they take on when they are not present in a call, by giving a
4391 two-item list in place of an optional argument, for example in:
4392 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
4393 foo is a fixed argument, bar is an optional argument with default
4394 value 42, and baz is a keyword argument with default value 73.
4395 Default value expressions are not evaluated unless they are needed
4396 and until the procedure is called.
4398 lambda* now supports two more special parameter list keywords.
4400 lambda*-defined procedures now throw an error by default if a
4401 keyword other than one of those specified is found in the actual
4402 passed arguments. However, specifying #&allow-other-keys
4403 immediately after the kyword argument declarations restores the
4404 previous behavior of ignoring unknown keywords. lambda* also now
4405 guarantees that if the same keyword is passed more than once, the
4406 last one passed is the one that takes effect. For example,
4407 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
4408 #:heads 37 #:tails 42 #:heads 99)
4409 would result in (99 47) being displayed.
4411 #&rest is also now provided as a synonym for the dotted syntax rest
4412 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
4413 all respects to lambda*. This is provided for more similarity to DSSSL,
4414 MIT-Scheme and Kawa among others, as well as for refugees from other
4417 Further documentation may be found in the optargs.scm file itself.
4419 The optional argument module also exports the macros `let-optional',
4420 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
4421 are not documented here because they may be removed in the future, but
4422 full documentation is still available in optargs.scm.
4424 ** New syntax: and-let*
4425 Guile now supports the `and-let*' form, described in the draft SRFI-2.
4427 Syntax: (land* (<clause> ...) <body> ...)
4428 Each <clause> should have one of the following forms:
4429 (<variable> <expression>)
4432 Each <variable> or <bound-variable> should be an identifier. Each
4433 <expression> should be a valid expression. The <body> should be a
4434 possibly empty sequence of expressions, like the <body> of a
4437 Semantics: A LAND* expression is evaluated by evaluating the
4438 <expression> or <bound-variable> of each of the <clause>s from
4439 left to right. The value of the first <expression> or
4440 <bound-variable> that evaluates to a false value is returned; the
4441 remaining <expression>s and <bound-variable>s are not evaluated.
4442 The <body> forms are evaluated iff all the <expression>s and
4443 <bound-variable>s evaluate to true values.
4445 The <expression>s and the <body> are evaluated in an environment
4446 binding each <variable> of the preceding (<variable> <expression>)
4447 clauses to the value of the <expression>. Later bindings
4448 shadow earlier bindings.
4450 Guile's and-let* macro was contributed by Michael Livshin.
4452 ** New sorting functions
4454 *** New function: sorted? SEQUENCE LESS?
4455 Returns `#t' when the sequence argument is in non-decreasing order
4456 according to LESS? (that is, there is no adjacent pair `... x y
4457 ...' for which `(less? y x)').
4459 Returns `#f' when the sequence contains at least one out-of-order
4460 pair. It is an error if the sequence is neither a list nor a
4463 *** New function: merge LIST1 LIST2 LESS?
4464 LIST1 and LIST2 are sorted lists.
4465 Returns the sorted list of all elements in LIST1 and LIST2.
4467 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
4468 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
4469 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
4470 (Here "<" should read "comes before".)
4472 *** New procedure: merge! LIST1 LIST2 LESS?
4473 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
4474 the result. If the code is compiled, and LESS? constructs no new
4475 pairs, no pairs at all will be allocated. The first pair of the
4476 result will be either the first pair of LIST1 or the first pair of
4479 *** New function: sort SEQUENCE LESS?
4480 Accepts either a list or a vector, and returns a new sequence
4481 which is sorted. The new sequence is the same type as the input.
4482 Always `(sorted? (sort sequence less?) less?)'. The original
4483 sequence is not altered in any way. The new sequence shares its
4484 elements with the old one; no elements are copied.
4486 *** New procedure: sort! SEQUENCE LESS
4487 Returns its sorted result in the original boxes. No new storage is
4488 allocated at all. Proper usage: (set! slist (sort! slist <))
4490 *** New function: stable-sort SEQUENCE LESS?
4491 Similar to `sort' but stable. That is, if "equal" elements are
4492 ordered a < b in the original sequence, they will have the same order
4495 *** New function: stable-sort! SEQUENCE LESS?
4496 Similar to `sort!' but stable.
4497 Uses temporary storage when sorting vectors.
4499 *** New functions: sort-list, sort-list!
4500 Added for compatibility with scsh.
4502 ** New built-in random number support
4504 *** New function: random N [STATE]
4505 Accepts a positive integer or real N and returns a number of the
4506 same type between zero (inclusive) and N (exclusive). The values
4507 returned have a uniform distribution.
4509 The optional argument STATE must be of the type produced by
4510 `copy-random-state' or `seed->random-state'. It defaults to the value
4511 of the variable `*random-state*'. This object is used to maintain the
4512 state of the pseudo-random-number generator and is altered as a side
4513 effect of the `random' operation.
4515 *** New variable: *random-state*
4516 Holds a data structure that encodes the internal state of the
4517 random-number generator that `random' uses by default. The nature
4518 of this data structure is implementation-dependent. It may be
4519 printed out and successfully read back in, but may or may not
4520 function correctly as a random-number state object in another
4523 *** New function: copy-random-state [STATE]
4524 Returns a new object of type suitable for use as the value of the
4525 variable `*random-state*' and as a second argument to `random'.
4526 If argument STATE is given, a copy of it is returned. Otherwise a
4527 copy of `*random-state*' is returned.
4529 *** New function: seed->random-state SEED
4530 Returns a new object of type suitable for use as the value of the
4531 variable `*random-state*' and as a second argument to `random'.
4532 SEED is a string or a number. A new state is generated and
4533 initialized using SEED.
4535 *** New function: random:uniform [STATE]
4536 Returns an uniformly distributed inexact real random number in the
4537 range between 0 and 1.
4539 *** New procedure: random:solid-sphere! VECT [STATE]
4540 Fills VECT with inexact real random numbers the sum of whose
4541 squares is less than 1.0. Thinking of VECT as coordinates in
4542 space of dimension N = `(vector-length VECT)', the coordinates are
4543 uniformly distributed within the unit N-shere. The sum of the
4544 squares of the numbers is returned. VECT can be either a vector
4545 or a uniform vector of doubles.
4547 *** New procedure: random:hollow-sphere! VECT [STATE]
4548 Fills VECT with inexact real random numbers the sum of whose squares
4549 is equal to 1.0. Thinking of VECT as coordinates in space of
4550 dimension n = `(vector-length VECT)', the coordinates are uniformly
4551 distributed over the surface of the unit n-shere. VECT can be either
4552 a vector or a uniform vector of doubles.
4554 *** New function: random:normal [STATE]
4555 Returns an inexact real in a normal distribution with mean 0 and
4556 standard deviation 1. For a normal distribution with mean M and
4557 standard deviation D use `(+ M (* D (random:normal)))'.
4559 *** New procedure: random:normal-vector! VECT [STATE]
4560 Fills VECT with inexact real random numbers which are independent and
4561 standard normally distributed (i.e., with mean 0 and variance 1).
4562 VECT can be either a vector or a uniform vector of doubles.
4564 *** New function: random:exp STATE
4565 Returns an inexact real in an exponential distribution with mean 1.
4566 For an exponential distribution with mean U use (* U (random:exp)).
4568 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
4570 These functions now operate on numbers in the range of a C unsigned
4573 These functions used to operate on numbers in the range of a C signed
4574 long; however, this seems inappropriate, because Guile integers don't
4577 ** New function: make-guardian
4578 This is an implementation of guardians as described in
4579 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
4580 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
4581 Programming Language Design and Implementation, June 1993
4582 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
4584 ** New functions: delq1!, delv1!, delete1!
4585 These procedures behave similar to delq! and friends but delete only
4586 one object if at all.
4588 ** New function: unread-string STRING PORT
4589 Unread STRING to PORT, that is, push it back onto the port so that
4590 next read operation will work on the pushed back characters.
4592 ** unread-char can now be called multiple times
4593 If unread-char is called multiple times, the unread characters will be
4594 read again in last-in first-out order.
4596 ** the procedures uniform-array-read! and uniform-array-write! now
4597 work on any kind of port, not just ports which are open on a file.
4599 ** Now 'l' in a port mode requests line buffering.
4601 ** The procedure truncate-file now works on string ports as well
4602 as file ports. If the size argument is omitted, the current
4603 file position is used.
4605 ** new procedure: seek PORT/FDES OFFSET WHENCE
4606 The arguments are the same as for the old fseek procedure, but it
4607 works on string ports as well as random-access file ports.
4609 ** the fseek procedure now works on string ports, since it has been
4610 redefined using seek.
4612 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
4613 size is not supplied.
4615 ** the newline procedure no longer flushes the port if it's not
4616 line-buffered: previously it did if it was the current output port.
4618 ** open-pipe and close-pipe are no longer primitive procedures, but
4619 an emulation can be obtained using `(use-modules (ice-9 popen))'.
4621 ** the freopen procedure has been removed.
4623 ** new procedure: drain-input PORT
4624 Drains PORT's read buffers (including any pushed-back characters)
4625 and returns the contents as a single string.
4627 ** New function: map-in-order PROC LIST1 LIST2 ...
4628 Version of `map' which guarantees that the procedure is applied to the
4629 lists in serial order.
4631 ** Renamed `serial-array-copy!' and `serial-array-map!' to
4632 `array-copy-in-order!' and `array-map-in-order!'. The old names are
4633 now obsolete and will go away in release 1.5.
4635 ** New syntax: collect BODY1 ...
4636 Version of `begin' which returns a list of the results of the body
4637 forms instead of the result of the last body form. In contrast to
4638 `begin', `collect' allows an empty body.
4640 ** New functions: read-history FILENAME, write-history FILENAME
4641 Read/write command line history from/to file. Returns #t on success
4642 and #f if an error occured.
4644 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
4646 These procedures return a list of definitions available in the specified
4647 argument, a relative module reference. In the case of no argument,
4648 `(current-module)' is now consulted for definitions to return, instead
4649 of simply returning #f, the former behavior.
4651 ** The #/ syntax for lists is no longer supported.
4653 Earlier versions of Scheme accepted this syntax, but printed a
4656 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
4658 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
4661 * Changes to the gh_ interface
4665 Now takes a second argument which is the result array. If this
4666 pointer is NULL, a new array is malloced (the old behaviour).
4668 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
4669 gh_scm2shorts, gh_scm2longs, gh_scm2floats
4673 * Changes to the scm_ interface
4675 ** Function: scm_make_named_hook (char* name, int n_args)
4677 Creates a hook in the same way as make-hook above but also
4678 binds a variable named NAME to it.
4680 This is the typical way of creating a hook from C code.
4682 Currently, the variable is created in the "current" module. This
4683 might change when we get the new module system.
4685 ** The smob interface
4687 The interface for creating smobs has changed. For documentation, see
4688 data-rep.info (made from guile-core/doc/data-rep.texi).
4690 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
4692 >>> This function will be removed in 1.3.4. <<<
4696 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
4697 This function adds a new smob type, named NAME, with instance size
4698 SIZE to the system. The return value is a tag that is used in
4699 creating instances of the type. If SIZE is 0, then no memory will
4700 be allocated when instances of the smob are created, and nothing
4701 will be freed by the default free function.
4703 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
4704 This function sets the smob marking procedure for the smob type
4705 specified by the tag TC. TC is the tag returned by
4706 `scm_make_smob_type'.
4708 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
4709 This function sets the smob freeing procedure for the smob type
4710 specified by the tag TC. TC is the tag returned by
4711 `scm_make_smob_type'.
4713 *** Function: void scm_set_smob_print (tc, print)
4715 - Function: void scm_set_smob_print (long tc,
4716 scm_sizet (*print) (SCM,
4720 This function sets the smob printing procedure for the smob type
4721 specified by the tag TC. TC is the tag returned by
4722 `scm_make_smob_type'.
4724 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
4725 This function sets the smob equality-testing predicate for the
4726 smob type specified by the tag TC. TC is the tag returned by
4727 `scm_make_smob_type'.
4729 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
4730 Make VALUE contain a smob instance of the type with type code TC and
4731 smob data DATA. VALUE must be previously declared as C type `SCM'.
4733 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
4734 This macro expands to a block of code that creates a smob instance
4735 of the type with type code TC and smob data DATA, and returns that
4736 `SCM' value. It should be the last piece of code in a block.
4738 ** The interfaces for using I/O ports and implementing port types
4739 (ptobs) have changed significantly. The new interface is based on
4740 shared access to buffers and a new set of ptob procedures.
4742 *** scm_newptob has been removed
4746 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
4748 - Function: SCM scm_make_port_type (char *type_name,
4749 int (*fill_buffer) (SCM port),
4750 void (*write_flush) (SCM port));
4752 Similarly to the new smob interface, there is a set of function
4753 setters by which the user can customize the behaviour of his port
4754 type. See ports.h (scm_set_port_XXX).
4756 ** scm_strport_to_string: New function: creates a new string from
4757 a string port's buffer.
4759 ** Plug in interface for random number generators
4760 The variable `scm_the_rng' in random.c contains a value and three
4761 function pointers which together define the current random number
4762 generator being used by the Scheme level interface and the random
4763 number library functions.
4765 The user is free to replace the default generator with the generator
4768 *** Variable: size_t scm_the_rng.rstate_size
4769 The size of the random state type used by the current RNG
4772 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
4773 Given the random STATE, return 32 random bits.
4775 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
4776 Seed random state STATE using string S of length N.
4778 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
4779 Given random state STATE, return a malloced copy.
4782 The default RNG is the MWC (Multiply With Carry) random number
4783 generator described by George Marsaglia at the Department of
4784 Statistics and Supercomputer Computations Research Institute, The
4785 Florida State University (http://stat.fsu.edu/~geo).
4787 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
4788 passes all tests in the DIEHARD test suite
4789 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
4790 costs one multiply and one add on platforms which either supports long
4791 longs (gcc does this on most systems) or have 64 bit longs. The cost
4792 is four multiply on other systems but this can be optimized by writing
4793 scm_i_uniform32 in assembler.
4795 These functions are provided through the scm_the_rng interface for use
4796 by libguile and the application.
4798 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
4799 Given the random STATE, return 32 random bits.
4800 Don't use this function directly. Instead go through the plugin
4801 interface (see "Plug in interface" above).
4803 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
4804 Initialize STATE using SEED of length N.
4806 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
4807 Return a malloc:ed copy of STATE. This function can easily be re-used
4808 in the interfaces to other RNGs.
4810 ** Random number library functions
4811 These functions use the current RNG through the scm_the_rng interface.
4812 It might be a good idea to use these functions from your C code so
4813 that only one random generator is used by all code in your program.
4815 The default random state is stored in:
4817 *** Variable: SCM scm_var_random_state
4818 Contains the vcell of the Scheme variable "*random-state*" which is
4819 used as default state by all random number functions in the Scheme
4824 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
4826 *** Function: scm_rstate *scm_c_default_rstate (void)
4827 This is a convenience function which returns the value of
4828 scm_var_random_state. An error message is generated if this value
4829 isn't a random state.
4831 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
4832 Make a new random state from the string SEED of length LENGTH.
4834 It is generally not a good idea to use multiple random states in a
4835 program. While subsequent random numbers generated from one random
4836 state are guaranteed to be reasonably independent, there is no such
4837 guarantee for numbers generated from different random states.
4839 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
4840 Return 32 random bits.
4842 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
4843 Return a sample from the uniform(0,1) distribution.
4845 *** Function: double scm_c_normal01 (scm_rstate *STATE)
4846 Return a sample from the normal(0,1) distribution.
4848 *** Function: double scm_c_exp1 (scm_rstate *STATE)
4849 Return a sample from the exp(1) distribution.
4851 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
4852 Return a sample from the discrete uniform(0,M) distribution.
4854 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
4855 Return a sample from the discrete uniform(0,M) distribution.
4856 M must be a bignum object. The returned value may be an INUM.
4860 Changes in Guile 1.3 (released Monday, October 19, 1998):
4862 * Changes to the distribution
4864 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
4865 To avoid conflicts, programs should name environment variables after
4866 themselves, except when there's a common practice establishing some
4869 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
4870 giving the former precedence, and printing a warning message if the
4871 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
4873 ** The header files related to multi-byte characters have been removed.
4874 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
4875 which referred to these explicitly will probably need to be rewritten,
4876 since the support for the variant string types has been removed; see
4879 ** The header files append.h and sequences.h have been removed. These
4880 files implemented non-R4RS operations which would encourage
4881 non-portable programming style and less easy-to-read code.
4883 * Changes to the stand-alone interpreter
4885 ** New procedures have been added to implement a "batch mode":
4887 *** Function: batch-mode?
4889 Returns a boolean indicating whether the interpreter is in batch
4892 *** Function: set-batch-mode?! ARG
4894 If ARG is true, switches the interpreter to batch mode. The `#f'
4895 case has not been implemented.
4897 ** Guile now provides full command-line editing, when run interactively.
4898 To use this feature, you must have the readline library installed.
4899 The Guile build process will notice it, and automatically include
4902 The readline library is available via anonymous FTP from any GNU
4903 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
4905 ** the-last-stack is now a fluid.
4907 * Changes to the procedure for linking libguile with your programs
4909 ** You can now use the `guile-config' utility to build programs that use Guile.
4911 Guile now includes a command-line utility called `guile-config', which
4912 can provide information about how to compile and link programs that
4915 *** `guile-config compile' prints any C compiler flags needed to use Guile.
4916 You should include this command's output on the command line you use
4917 to compile C or C++ code that #includes the Guile header files. It's
4918 usually just a `-I' flag to help the compiler find the Guile headers.
4921 *** `guile-config link' prints any linker flags necessary to link with Guile.
4923 This command writes to its standard output a list of flags which you
4924 must pass to the linker to link your code against the Guile library.
4925 The flags include '-lguile' itself, any other libraries the Guile
4926 library depends upon, and any `-L' flags needed to help the linker
4927 find those libraries.
4929 For example, here is a Makefile rule that builds a program named 'foo'
4930 from the object files ${FOO_OBJECTS}, and links them against Guile:
4933 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
4935 Previous Guile releases recommended that you use autoconf to detect
4936 which of a predefined set of libraries were present on your system.
4937 It is more robust to use `guile-config', since it records exactly which
4938 libraries the installed Guile library requires.
4940 This was originally called `build-guile', but was renamed to
4941 `guile-config' before Guile 1.3 was released, to be consistent with
4942 the analogous script for the GTK+ GUI toolkit, which is called
4946 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
4948 If you are using the GNU autoconf package to configure your program,
4949 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
4950 (described above) and gather the necessary values for use in your
4953 The GUILE_FLAGS macro expands to configure script code which runs the
4954 `guile-config' script, to find out where Guile's header files and
4955 libraries are installed. It sets two variables, marked for
4956 substitution, as by AC_SUBST.
4958 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
4959 code that uses Guile header files. This is almost always just a
4962 GUILE_LDFLAGS --- flags to pass to the linker to link a
4963 program against Guile. This includes `-lguile' for the Guile
4964 library itself, any libraries that Guile itself requires (like
4965 -lqthreads), and so on. It may also include a -L flag to tell the
4966 compiler where to find the libraries.
4968 GUILE_FLAGS is defined in the file guile.m4, in the top-level
4969 directory of the Guile distribution. You can copy it into your
4970 package's aclocal.m4 file, and then use it in your configure.in file.
4972 If you are using the `aclocal' program, distributed with GNU automake,
4973 to maintain your aclocal.m4 file, the Guile installation process
4974 installs guile.m4 where aclocal will find it. All you need to do is
4975 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
4976 this will copy the definition of GUILE_FLAGS into your aclocal.m4
4980 * Changes to Scheme functions and syntax
4982 ** Multi-byte strings have been removed, as have multi-byte and wide
4983 ports. We felt that these were the wrong approach to
4984 internationalization support.
4986 ** New function: readline [PROMPT]
4987 Read a line from the terminal, and allow the user to edit it,
4988 prompting with PROMPT. READLINE provides a large set of Emacs-like
4989 editing commands, lets the user recall previously typed lines, and
4990 works on almost every kind of terminal, including dumb terminals.
4992 READLINE assumes that the cursor is at the beginning of the line when
4993 it is invoked. Thus, you can't print a prompt yourself, and then call
4994 READLINE; you need to package up your prompt as a string, pass it to
4995 the function, and let READLINE print the prompt itself. This is
4996 because READLINE needs to know the prompt's screen width.
4998 For Guile to provide this function, you must have the readline
4999 library, version 2.1 or later, installed on your system. Readline is
5000 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
5001 any GNU mirror site.
5003 See also ADD-HISTORY function.
5005 ** New function: add-history STRING
5006 Add STRING as the most recent line in the history used by the READLINE
5007 command. READLINE does not add lines to the history itself; you must
5008 call ADD-HISTORY to make previous input available to the user.
5010 ** The behavior of the read-line function has changed.
5012 This function now uses standard C library functions to read the line,
5013 for speed. This means that it doesn not respect the value of
5014 scm-line-incrementors; it assumes that lines are delimited with
5017 (Note that this is read-line, the function that reads a line of text
5018 from a port, not readline, the function that reads a line from a
5019 terminal, providing full editing capabilities.)
5021 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
5023 This module provides some simple argument parsing. It exports one
5026 Function: getopt-gnu-style ARG-LS
5027 Parse a list of program arguments into an alist of option
5030 Each item in the list of program arguments is examined to see if
5031 it meets the syntax of a GNU long-named option. An argument like
5032 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
5033 returned alist, where MUMBLE is a keyword object with the same
5034 name as the argument. An argument like `--MUMBLE=FROB' produces
5035 an element of the form (MUMBLE . FROB), where FROB is a string.
5037 As a special case, the returned alist also contains a pair whose
5038 car is the symbol `rest'. The cdr of this pair is a list
5039 containing all the items in the argument list that are not options
5040 of the form mentioned above.
5042 The argument `--' is treated specially: all items in the argument
5043 list appearing after such an argument are not examined, and are
5044 returned in the special `rest' list.
5046 This function does not parse normal single-character switches.
5047 You will need to parse them out of the `rest' list yourself.
5049 ** The read syntax for byte vectors and short vectors has changed.
5051 Instead of #bytes(...), write #y(...).
5053 Instead of #short(...), write #h(...).
5055 This may seem nutty, but, like the other uniform vectors, byte vectors
5056 and short vectors want to have the same print and read syntax (and,
5057 more basic, want to have read syntax!). Changing the read syntax to
5058 use multiple characters after the hash sign breaks with the
5059 conventions used in R5RS and the conventions used for the other
5060 uniform vectors. It also introduces complexity in the current reader,
5061 both on the C and Scheme levels. (The Right solution is probably to
5062 change the syntax and prototypes for uniform vectors entirely.)
5065 ** The new module (ice-9 session) provides useful interactive functions.
5067 *** New procedure: (apropos REGEXP OPTION ...)
5069 Display a list of top-level variables whose names match REGEXP, and
5070 the modules they are imported from. Each OPTION should be one of the
5073 value --- Show the value of each matching variable.
5074 shadow --- Show bindings shadowed by subsequently imported modules.
5075 full --- Same as both `shadow' and `value'.
5079 guile> (apropos "trace" 'full)
5080 debug: trace #<procedure trace args>
5081 debug: untrace #<procedure untrace args>
5082 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
5083 the-scm-module: before-backtrace-hook ()
5084 the-scm-module: backtrace #<primitive-procedure backtrace>
5085 the-scm-module: after-backtrace-hook ()
5086 the-scm-module: has-shown-backtrace-hint? #f
5089 ** There are new functions and syntax for working with macros.
5091 Guile implements macros as a special object type. Any variable whose
5092 top-level binding is a macro object acts as a macro. The macro object
5093 specifies how the expression should be transformed before evaluation.
5095 *** Macro objects now print in a reasonable way, resembling procedures.
5097 *** New function: (macro? OBJ)
5098 True iff OBJ is a macro object.
5100 *** New function: (primitive-macro? OBJ)
5101 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
5102 macro transformers, implemented in eval.c rather than Scheme code.
5104 Why do we have this function?
5105 - For symmetry with procedure? and primitive-procedure?,
5106 - to allow custom print procedures to tell whether a macro is
5107 primitive, and display it differently, and
5108 - to allow compilers and user-written evaluators to distinguish
5109 builtin special forms from user-defined ones, which could be
5112 *** New function: (macro-type OBJ)
5113 Return a value indicating what kind of macro OBJ is. Possible return
5116 The symbol `syntax' --- a macro created by procedure->syntax.
5117 The symbol `macro' --- a macro created by procedure->macro.
5118 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
5119 The boolean #f --- if OBJ is not a macro object.
5121 *** New function: (macro-name MACRO)
5122 Return the name of the macro object MACRO's procedure, as returned by
5125 *** New function: (macro-transformer MACRO)
5126 Return the transformer procedure for MACRO.
5128 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
5130 Specify a new macro expander to use in the current module. Each
5131 MODULE is a module name, with the same meaning as in the `use-modules'
5132 form; each named module's exported bindings are added to the current
5133 top-level environment. TRANSFORMER is an expression evaluated in the
5134 resulting environment which must yield a procedure to use as the
5135 module's eval transformer: every expression evaluated in this module
5136 is passed to this function, and the result passed to the Guile
5139 *** macro-eval! is removed. Use local-eval instead.
5141 ** Some magic has been added to the printer to better handle user
5142 written printing routines (like record printers, closure printers).
5144 The problem is that these user written routines must have access to
5145 the current `print-state' to be able to handle fancy things like
5146 detection of circular references. These print-states have to be
5147 passed to the builtin printing routines (display, write, etc) to
5148 properly continue the print chain.
5150 We didn't want to change all existing print code so that it
5151 explicitly passes thru a print state in addition to a port. Instead,
5152 we extented the possible values that the builtin printing routines
5153 accept as a `port'. In addition to a normal port, they now also take
5154 a pair of a normal port and a print-state. Printing will go to the
5155 port and the print-state will be used to control the detection of
5156 circular references, etc. If the builtin function does not care for a
5157 print-state, it is simply ignored.
5159 User written callbacks are now called with such a pair as their
5160 `port', but because every function now accepts this pair as a PORT
5161 argument, you don't have to worry about that. In fact, it is probably
5162 safest to not check for these pairs.
5164 However, it is sometimes necessary to continue a print chain on a
5165 different port, for example to get a intermediate string
5166 representation of the printed value, mangle that string somehow, and
5167 then to finally print the mangled string. Use the new function
5169 inherit-print-state OLD-PORT NEW-PORT
5171 for this. It constructs a new `port' that prints to NEW-PORT but
5172 inherits the print-state of OLD-PORT.
5174 ** struct-vtable-offset renamed to vtable-offset-user
5176 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
5178 ** There is now a third optional argument to make-vtable-vtable
5179 (and fourth to make-struct) when constructing new types (vtables).
5180 This argument initializes field vtable-index-printer of the vtable.
5182 ** The detection of circular references has been extended to structs.
5183 That is, a structure that -- in the process of being printed -- prints
5184 itself does not lead to infinite recursion.
5186 ** There is now some basic support for fluids. Please read
5187 "libguile/fluid.h" to find out more. It is accessible from Scheme with
5188 the following functions and macros:
5190 Function: make-fluid
5192 Create a new fluid object. Fluids are not special variables or
5193 some other extension to the semantics of Scheme, but rather
5194 ordinary Scheme objects. You can store them into variables (that
5195 are still lexically scoped, of course) or into any other place you
5196 like. Every fluid has a initial value of `#f'.
5198 Function: fluid? OBJ
5200 Test whether OBJ is a fluid.
5202 Function: fluid-ref FLUID
5203 Function: fluid-set! FLUID VAL
5205 Access/modify the fluid FLUID. Modifications are only visible
5206 within the current dynamic root (that includes threads).
5208 Function: with-fluids* FLUIDS VALUES THUNK
5210 FLUIDS is a list of fluids and VALUES a corresponding list of
5211 values for these fluids. Before THUNK gets called the values are
5212 installed in the fluids and the old values of the fluids are
5213 saved in the VALUES list. When the flow of control leaves THUNK
5214 or reenters it, the values get swapped again. You might think of
5215 this as a `safe-fluid-excursion'. Note that the VALUES list is
5216 modified by `with-fluids*'.
5218 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
5220 The same as `with-fluids*' but with a different syntax. It looks
5221 just like `let', but both FLUID and VALUE are evaluated. Remember,
5222 fluids are not special variables but ordinary objects. FLUID
5223 should evaluate to a fluid.
5225 ** Changes to system call interfaces:
5227 *** close-port, close-input-port and close-output-port now return a
5228 boolean instead of an `unspecified' object. #t means that the port
5229 was successfully closed, while #f means it was already closed. It is
5230 also now possible for these procedures to raise an exception if an
5231 error occurs (some errors from write can be delayed until close.)
5233 *** the first argument to chmod, fcntl, ftell and fseek can now be a
5236 *** the third argument to fcntl is now optional.
5238 *** the first argument to chown can now be a file descriptor or a port.
5240 *** the argument to stat can now be a port.
5242 *** The following new procedures have been added (most use scsh
5245 *** procedure: close PORT/FD
5246 Similar to close-port (*note close-port: Closing Ports.), but also
5247 works on file descriptors. A side effect of closing a file
5248 descriptor is that any ports using that file descriptor are moved
5249 to a different file descriptor and have their revealed counts set
5252 *** procedure: port->fdes PORT
5253 Returns the integer file descriptor underlying PORT. As a side
5254 effect the revealed count of PORT is incremented.
5256 *** procedure: fdes->ports FDES
5257 Returns a list of existing ports which have FDES as an underlying
5258 file descriptor, without changing their revealed counts.
5260 *** procedure: fdes->inport FDES
5261 Returns an existing input port which has FDES as its underlying
5262 file descriptor, if one exists, and increments its revealed count.
5263 Otherwise, returns a new input port with a revealed count of 1.
5265 *** procedure: fdes->outport FDES
5266 Returns an existing output port which has FDES as its underlying
5267 file descriptor, if one exists, and increments its revealed count.
5268 Otherwise, returns a new output port with a revealed count of 1.
5270 The next group of procedures perform a `dup2' system call, if NEWFD
5271 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
5272 duplicated can be supplied as an integer or contained in a port. The
5273 type of value returned varies depending on which procedure is used.
5275 All procedures also have the side effect when performing `dup2' that
5276 any ports using NEWFD are moved to a different file descriptor and have
5277 their revealed counts set to zero.
5279 *** procedure: dup->fdes PORT/FD [NEWFD]
5280 Returns an integer file descriptor.
5282 *** procedure: dup->inport PORT/FD [NEWFD]
5283 Returns a new input port using the new file descriptor.
5285 *** procedure: dup->outport PORT/FD [NEWFD]
5286 Returns a new output port using the new file descriptor.
5288 *** procedure: dup PORT/FD [NEWFD]
5289 Returns a new port if PORT/FD is a port, with the same mode as the
5290 supplied port, otherwise returns an integer file descriptor.
5292 *** procedure: dup->port PORT/FD MODE [NEWFD]
5293 Returns a new port using the new file descriptor. MODE supplies a
5294 mode string for the port (*note open-file: File Ports.).
5296 *** procedure: setenv NAME VALUE
5297 Modifies the environment of the current process, which is also the
5298 default environment inherited by child processes.
5300 If VALUE is `#f', then NAME is removed from the environment.
5301 Otherwise, the string NAME=VALUE is added to the environment,
5302 replacing any existing string with name matching NAME.
5304 The return value is unspecified.
5306 *** procedure: truncate-file OBJ SIZE
5307 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
5308 can be a string containing a file name or an integer file
5309 descriptor or port open for output on the file. The underlying
5310 system calls are `truncate' and `ftruncate'.
5312 The return value is unspecified.
5314 *** procedure: setvbuf PORT MODE [SIZE]
5315 Set the buffering mode for PORT. MODE can be:
5323 block buffered, using a newly allocated buffer of SIZE bytes.
5324 However if SIZE is zero or unspecified, the port will be made
5327 This procedure should not be used after I/O has been performed with
5330 Ports are usually block buffered by default, with a default buffer
5331 size. Procedures e.g., *Note open-file: File Ports, which accept a
5332 mode string allow `0' to be added to request an unbuffered port.
5334 *** procedure: fsync PORT/FD
5335 Copies any unwritten data for the specified output file descriptor
5336 to disk. If PORT/FD is a port, its buffer is flushed before the
5337 underlying file descriptor is fsync'd. The return value is
5340 *** procedure: open-fdes PATH FLAGS [MODES]
5341 Similar to `open' but returns a file descriptor instead of a port.
5343 *** procedure: execle PATH ENV [ARG] ...
5344 Similar to `execl', but the environment of the new process is
5345 specified by ENV, which must be a list of strings as returned by
5346 the `environ' procedure.
5348 This procedure is currently implemented using the `execve' system
5349 call, but we call it `execle' because of its Scheme calling
5352 *** procedure: strerror ERRNO
5353 Returns the Unix error message corresponding to ERRNO, an integer.
5355 *** procedure: primitive-exit [STATUS]
5356 Terminate the current process without unwinding the Scheme stack.
5357 This is would typically be useful after a fork. The exit status
5358 is STATUS if supplied, otherwise zero.
5360 *** procedure: times
5361 Returns an object with information about real and processor time.
5362 The following procedures accept such an object as an argument and
5363 return a selected component:
5366 The current real time, expressed as time units relative to an
5370 The CPU time units used by the calling process.
5373 The CPU time units used by the system on behalf of the
5377 The CPU time units used by terminated child processes of the
5378 calling process, whose status has been collected (e.g., using
5382 Similarly, the CPU times units used by the system on behalf of
5383 terminated child processes.
5385 ** Removed: list-length
5386 ** Removed: list-append, list-append!
5387 ** Removed: list-reverse, list-reverse!
5389 ** array-map renamed to array-map!
5391 ** serial-array-map renamed to serial-array-map!
5393 ** catch doesn't take #f as first argument any longer
5395 Previously, it was possible to pass #f instead of a key to `catch'.
5396 That would cause `catch' to pass a jump buffer object to the procedure
5397 passed as second argument. The procedure could then use this jump
5398 buffer objekt as an argument to throw.
5400 This mechanism has been removed since its utility doesn't motivate the
5401 extra complexity it introduces.
5403 ** The `#/' notation for lists now provokes a warning message from Guile.
5404 This syntax will be removed from Guile in the near future.
5406 To disable the warning message, set the GUILE_HUSH environment
5407 variable to any non-empty value.
5409 ** The newline character now prints as `#\newline', following the
5410 normal Scheme notation, not `#\nl'.
5412 * Changes to the gh_ interface
5414 ** The gh_enter function now takes care of loading the Guile startup files.
5415 gh_enter works by calling scm_boot_guile; see the remarks below.
5417 ** Function: void gh_write (SCM x)
5419 Write the printed representation of the scheme object x to the current
5420 output port. Corresponds to the scheme level `write'.
5422 ** gh_list_length renamed to gh_length.
5424 ** vector handling routines
5426 Several major changes. In particular, gh_vector() now resembles
5427 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
5428 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
5429 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
5430 vector-related gh_ functions have been implemented.
5432 ** pair and list routines
5434 Implemented several of the R4RS pair and list functions that were
5437 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
5439 New function. Converts double arrays back and forth between Scheme
5442 * Changes to the scm_ interface
5444 ** The function scm_boot_guile now takes care of loading the startup files.
5446 Guile's primary initialization function, scm_boot_guile, now takes
5447 care of loading `boot-9.scm', in the `ice-9' module, to initialize
5448 Guile, define the module system, and put together some standard
5449 bindings. It also loads `init.scm', which is intended to hold
5450 site-specific initialization code.
5452 Since Guile cannot operate properly until boot-9.scm is loaded, there
5453 is no reason to separate loading boot-9.scm from Guile's other
5454 initialization processes.
5456 This job used to be done by scm_compile_shell_switches, which didn't
5457 make much sense; in particular, it meant that people using Guile for
5458 non-shell-like applications had to jump through hoops to get Guile
5459 initialized properly.
5461 ** The function scm_compile_shell_switches no longer loads the startup files.
5462 Now, Guile always loads the startup files, whenever it is initialized;
5463 see the notes above for scm_boot_guile and scm_load_startup_files.
5465 ** Function: scm_load_startup_files
5466 This new function takes care of loading Guile's initialization file
5467 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
5468 this is always called by the Guile initialization process, it's
5469 probably not too useful to call this yourself, but it's there anyway.
5471 ** The semantics of smob marking have changed slightly.
5473 The smob marking function (the `mark' member of the scm_smobfuns
5474 structure) is no longer responsible for setting the mark bit on the
5475 smob. The generic smob handling code in the garbage collector will
5476 set this bit. The mark function need only ensure that any other
5477 objects the smob refers to get marked.
5479 Note that this change means that the smob's GC8MARK bit is typically
5480 already set upon entry to the mark function. Thus, marking functions
5481 which look like this:
5484 if (SCM_GC8MARKP (ptr))
5486 SCM_SETGC8MARK (ptr);
5487 ... mark objects to which the smob refers ...
5490 are now incorrect, since they will return early, and fail to mark any
5491 other objects the smob refers to. Some code in the Guile library used
5494 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
5496 If you have implemented your own I/O port type, by writing the
5497 functions required by the scm_ptobfuns and then calling scm_newptob,
5498 you will need to change your functions slightly.
5500 The functions in a scm_ptobfuns structure now expect the port itself
5501 as their argument; they used to expect the `stream' member of the
5502 port's scm_port_table structure. This allows functions in an
5503 scm_ptobfuns structure to easily access the port's cell (and any flags
5504 it its CAR), and the port's scm_port_table structure.
5506 Guile now passes the I/O port itself as the `port' argument in the
5507 following scm_ptobfuns functions:
5509 int (*free) (SCM port);
5510 int (*fputc) (int, SCM port);
5511 int (*fputs) (char *, SCM port);
5512 scm_sizet (*fwrite) SCM_P ((char *ptr,
5516 int (*fflush) (SCM port);
5517 int (*fgetc) (SCM port);
5518 int (*fclose) (SCM port);
5520 The interfaces to the `mark', `print', `equalp', and `fgets' methods
5523 If you have existing code which defines its own port types, it is easy
5524 to convert your code to the new interface; simply apply SCM_STREAM to
5525 the port argument to yield the value you code used to expect.
5527 Note that since both the port and the stream have the same type in the
5528 C code --- they are both SCM values --- the C compiler will not remind
5529 you if you forget to update your scm_ptobfuns functions.
5532 ** Function: int scm_internal_select (int fds,
5536 struct timeval *timeout);
5538 This is a replacement for the `select' function provided by the OS.
5539 It enables I/O blocking and sleeping to happen for one cooperative
5540 thread without blocking other threads. It also avoids busy-loops in
5541 these situations. It is intended that all I/O blocking and sleeping
5542 will finally go through this function. Currently, this function is
5543 only available on systems providing `gettimeofday' and `select'.
5545 ** Function: SCM scm_internal_stack_catch (SCM tag,
5546 scm_catch_body_t body,
5548 scm_catch_handler_t handler,
5551 A new sibling to the other two C level `catch' functions
5552 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
5553 the stack to be saved automatically into the variable `the-last-stack'
5554 (scm_the_last_stack_var) on error. This is necessary if you want to
5555 use advanced error reporting, such as calling scm_display_error and
5556 scm_display_backtrace. (They both take a stack object as argument.)
5558 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
5560 scm_catch_handler_t handler,
5563 Spawns a new thread. It does a job similar to
5564 scm_call_with_new_thread but takes arguments more suitable when
5565 spawning threads from application C code.
5567 ** The hook scm_error_callback has been removed. It was originally
5568 intended as a way for the user to install his own error handler. But
5569 that method works badly since it intervenes between throw and catch,
5570 thereby changing the semantics of expressions like (catch #t ...).
5571 The correct way to do it is to use one of the C level catch functions
5572 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
5574 ** Removed functions:
5576 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
5577 scm_list_reverse, scm_list_reverse_x
5579 ** New macros: SCM_LISTn where n is one of the integers 0-9.
5581 These can be used for pretty list creation from C. The idea is taken
5582 from Erick Gallesio's STk.
5584 ** scm_array_map renamed to scm_array_map_x
5586 ** mbstrings are now removed
5588 This means that the type codes scm_tc7_mb_string and
5589 scm_tc7_mb_substring has been removed.
5591 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
5593 Since we no longer support multi-byte strings, these I/O functions
5594 have been simplified, and renamed. Here are their old names, and
5595 their new names and arguments:
5597 scm_gen_putc -> void scm_putc (int c, SCM port);
5598 scm_gen_puts -> void scm_puts (char *s, SCM port);
5599 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
5600 scm_gen_getc -> void scm_getc (SCM port);
5603 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
5605 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
5607 SCM_TYP7S now masks away the bit which distinguishes substrings from
5610 ** scm_catch_body_t: Backward incompatible change!
5612 Body functions to scm_internal_catch and friends do not any longer
5613 take a second argument. This is because it is no longer possible to
5614 pass a #f arg to catch.
5616 ** Calls to scm_protect_object and scm_unprotect now nest properly.
5618 The function scm_protect_object protects its argument from being freed
5619 by the garbage collector. scm_unprotect_object removes that
5622 These functions now nest properly. That is, for every object O, there
5623 is a counter which scm_protect_object(O) increments and
5624 scm_unprotect_object(O) decrements, if the counter is greater than
5625 zero. Every object's counter is zero when it is first created. If an
5626 object's counter is greater than zero, the garbage collector will not
5627 reclaim its storage.
5629 This allows you to use scm_protect_object in your code without
5630 worrying that some other function you call will call
5631 scm_unprotect_object, and allow it to be freed. Assuming that the
5632 functions you call are well-behaved, and unprotect only those objects
5633 they protect, you can follow the same rule and have confidence that
5634 objects will be freed only at appropriate times.
5637 Changes in Guile 1.2 (released Tuesday, June 24 1997):
5639 * Changes to the distribution
5641 ** Nightly snapshots are now available from ftp.red-bean.com.
5642 The old server, ftp.cyclic.com, has been relinquished to its rightful
5645 Nightly snapshots of the Guile development sources are now available via
5646 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
5648 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
5649 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
5651 ** To run Guile without installing it, the procedure has changed a bit.
5653 If you used a separate build directory to compile Guile, you'll need
5654 to include the build directory in SCHEME_LOAD_PATH, as well as the
5655 source directory. See the `INSTALL' file for examples.
5657 * Changes to the procedure for linking libguile with your programs
5659 ** The standard Guile load path for Scheme code now includes
5660 $(datadir)/guile (usually /usr/local/share/guile). This means that
5661 you can install your own Scheme files there, and Guile will find them.
5662 (Previous versions of Guile only checked a directory whose name
5663 contained the Guile version number, so you had to re-install or move
5664 your Scheme sources each time you installed a fresh version of Guile.)
5666 The load path also includes $(datadir)/guile/site; we recommend
5667 putting individual Scheme files there. If you want to install a
5668 package with multiple source files, create a directory for them under
5671 ** Guile 1.2 will now use the Rx regular expression library, if it is
5672 installed on your system. When you are linking libguile into your own
5673 programs, this means you will have to link against -lguile, -lqt (if
5674 you configured Guile with thread support), and -lrx.
5676 If you are using autoconf to generate configuration scripts for your
5677 application, the following lines should suffice to add the appropriate
5678 libraries to your link command:
5680 ### Find Rx, quickthreads and libguile.
5681 AC_CHECK_LIB(rx, main)
5682 AC_CHECK_LIB(qt, main)
5683 AC_CHECK_LIB(guile, scm_shell)
5685 The Guile 1.2 distribution does not contain sources for the Rx
5686 library, as Guile 1.0 did. If you want to use Rx, you'll need to
5687 retrieve it from a GNU FTP site and install it separately.
5689 * Changes to Scheme functions and syntax
5691 ** The dynamic linking features of Guile are now enabled by default.
5692 You can disable them by giving the `--disable-dynamic-linking' option
5695 (dynamic-link FILENAME)
5697 Find the object file denoted by FILENAME (a string) and link it
5698 into the running Guile application. When everything works out,
5699 return a Scheme object suitable for representing the linked object
5700 file. Otherwise an error is thrown. How object files are
5701 searched is system dependent.
5703 (dynamic-object? VAL)
5705 Determine whether VAL represents a dynamically linked object file.
5707 (dynamic-unlink DYNOBJ)
5709 Unlink the indicated object file from the application. DYNOBJ
5710 should be one of the values returned by `dynamic-link'.
5712 (dynamic-func FUNCTION DYNOBJ)
5714 Search the C function indicated by FUNCTION (a string or symbol)
5715 in DYNOBJ and return some Scheme object that can later be used
5716 with `dynamic-call' to actually call this function. Right now,
5717 these Scheme objects are formed by casting the address of the
5718 function to `long' and converting this number to its Scheme
5721 (dynamic-call FUNCTION DYNOBJ)
5723 Call the C function indicated by FUNCTION and DYNOBJ. The
5724 function is passed no arguments and its return value is ignored.
5725 When FUNCTION is something returned by `dynamic-func', call that
5726 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
5727 etc.), look it up in DYNOBJ; this is equivalent to
5729 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
5731 Interrupts are deferred while the C function is executing (with
5732 SCM_DEFER_INTS/SCM_ALLOW_INTS).
5734 (dynamic-args-call FUNCTION DYNOBJ ARGS)
5736 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
5737 some arguments and return its return value. The C function is
5738 expected to take two arguments and return an `int', just like
5741 int c_func (int argc, char **argv);
5743 ARGS must be a list of strings and is converted into an array of
5744 `char *'. The array is passed in ARGV and its size in ARGC. The
5745 return value is converted to a Scheme number and returned from the
5746 call to `dynamic-args-call'.
5748 When dynamic linking is disabled or not supported on your system,
5749 the above functions throw errors, but they are still available.
5751 Here is a small example that works on GNU/Linux:
5753 (define libc-obj (dynamic-link "libc.so"))
5754 (dynamic-args-call 'rand libc-obj '())
5756 See the file `libguile/DYNAMIC-LINKING' for additional comments.
5758 ** The #/ syntax for module names is depreciated, and will be removed
5759 in a future version of Guile. Instead of
5767 The latter syntax is more consistent with existing Lisp practice.
5769 ** Guile now does fancier printing of structures. Structures are the
5770 underlying implementation for records, which in turn are used to
5771 implement modules, so all of these object now print differently and in
5772 a more informative way.
5774 The Scheme printer will examine the builtin variable *struct-printer*
5775 whenever it needs to print a structure object. When this variable is
5776 not `#f' it is deemed to be a procedure and will be applied to the
5777 structure object and the output port. When *struct-printer* is `#f'
5778 or the procedure return `#f' the structure object will be printed in
5779 the boring #<struct 80458270> form.
5781 This hook is used by some routines in ice-9/boot-9.scm to implement
5782 type specific printing routines. Please read the comments there about
5785 One of the more specific uses of structs are records. The printing
5786 procedure that could be passed to MAKE-RECORD-TYPE is now actually
5787 called. It should behave like a *struct-printer* procedure (described
5790 ** Guile now supports a new R4RS-compliant syntax for keywords. A
5791 token of the form #:NAME, where NAME has the same syntax as a Scheme
5792 symbol, is the external representation of the keyword named NAME.
5793 Keyword objects print using this syntax as well, so values containing
5794 keyword objects can be read back into Guile. When used in an
5795 expression, keywords are self-quoting objects.
5797 Guile suports this read syntax, and uses this print syntax, regardless
5798 of the current setting of the `keyword' read option. The `keyword'
5799 read option only controls whether Guile recognizes the `:NAME' syntax,
5800 which is incompatible with R4RS. (R4RS says such token represent
5803 ** Guile has regular expression support again. Guile 1.0 included
5804 functions for matching regular expressions, based on the Rx library.
5805 In Guile 1.1, the Guile/Rx interface was removed to simplify the
5806 distribution, and thus Guile had no regular expression support. Guile
5807 1.2 again supports the most commonly used functions, and supports all
5808 of SCSH's regular expression functions.
5810 If your system does not include a POSIX regular expression library,
5811 and you have not linked Guile with a third-party regexp library such as
5812 Rx, these functions will not be available. You can tell whether your
5813 Guile installation includes regular expression support by checking
5814 whether the `*features*' list includes the `regex' symbol.
5816 *** regexp functions
5818 By default, Guile supports POSIX extended regular expressions. That
5819 means that the characters `(', `)', `+' and `?' are special, and must
5820 be escaped if you wish to match the literal characters.
5822 This regular expression interface was modeled after that implemented
5823 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
5824 with SCSH regular expressions.
5826 **** Function: string-match PATTERN STR [START]
5827 Compile the string PATTERN into a regular expression and compare
5828 it with STR. The optional numeric argument START specifies the
5829 position of STR at which to begin matching.
5831 `string-match' returns a "match structure" which describes what,
5832 if anything, was matched by the regular expression. *Note Match
5833 Structures::. If STR does not match PATTERN at all,
5834 `string-match' returns `#f'.
5836 Each time `string-match' is called, it must compile its PATTERN
5837 argument into a regular expression structure. This operation is
5838 expensive, which makes `string-match' inefficient if the same regular
5839 expression is used several times (for example, in a loop). For better
5840 performance, you can compile a regular expression in advance and then
5841 match strings against the compiled regexp.
5843 **** Function: make-regexp STR [FLAGS]
5844 Compile the regular expression described by STR, and return the
5845 compiled regexp structure. If STR does not describe a legal
5846 regular expression, `make-regexp' throws a
5847 `regular-expression-syntax' error.
5849 FLAGS may be the bitwise-or of one or more of the following:
5851 **** Constant: regexp/extended
5852 Use POSIX Extended Regular Expression syntax when interpreting
5853 STR. If not set, POSIX Basic Regular Expression syntax is used.
5854 If the FLAGS argument is omitted, we assume regexp/extended.
5856 **** Constant: regexp/icase
5857 Do not differentiate case. Subsequent searches using the
5858 returned regular expression will be case insensitive.
5860 **** Constant: regexp/newline
5861 Match-any-character operators don't match a newline.
5863 A non-matching list ([^...]) not containing a newline matches a
5866 Match-beginning-of-line operator (^) matches the empty string
5867 immediately after a newline, regardless of whether the FLAGS
5868 passed to regexp-exec contain regexp/notbol.
5870 Match-end-of-line operator ($) matches the empty string
5871 immediately before a newline, regardless of whether the FLAGS
5872 passed to regexp-exec contain regexp/noteol.
5874 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
5875 Match the compiled regular expression REGEXP against `str'. If
5876 the optional integer START argument is provided, begin matching
5877 from that position in the string. Return a match structure
5878 describing the results of the match, or `#f' if no match could be
5881 FLAGS may be the bitwise-or of one or more of the following:
5883 **** Constant: regexp/notbol
5884 The match-beginning-of-line operator always fails to match (but
5885 see the compilation flag regexp/newline above) This flag may be
5886 used when different portions of a string are passed to
5887 regexp-exec and the beginning of the string should not be
5888 interpreted as the beginning of the line.
5890 **** Constant: regexp/noteol
5891 The match-end-of-line operator always fails to match (but see the
5892 compilation flag regexp/newline above)
5894 **** Function: regexp? OBJ
5895 Return `#t' if OBJ is a compiled regular expression, or `#f'
5898 Regular expressions are commonly used to find patterns in one string
5899 and replace them with the contents of another string.
5901 **** Function: regexp-substitute PORT MATCH [ITEM...]
5902 Write to the output port PORT selected contents of the match
5903 structure MATCH. Each ITEM specifies what should be written, and
5904 may be one of the following arguments:
5906 * A string. String arguments are written out verbatim.
5908 * An integer. The submatch with that number is written.
5910 * The symbol `pre'. The portion of the matched string preceding
5911 the regexp match is written.
5913 * The symbol `post'. The portion of the matched string
5914 following the regexp match is written.
5916 PORT may be `#f', in which case nothing is written; instead,
5917 `regexp-substitute' constructs a string from the specified ITEMs
5920 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
5921 Similar to `regexp-substitute', but can be used to perform global
5922 substitutions on STR. Instead of taking a match structure as an
5923 argument, `regexp-substitute/global' takes two string arguments: a
5924 REGEXP string describing a regular expression, and a TARGET string
5925 which should be matched against this regular expression.
5927 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
5930 * A function may be supplied. When this function is called, it
5931 will be passed one argument: a match structure for a given
5932 regular expression match. It should return a string to be
5933 written out to PORT.
5935 * The `post' symbol causes `regexp-substitute/global' to recurse
5936 on the unmatched portion of STR. This *must* be supplied in
5937 order to perform global search-and-replace on STR; if it is
5938 not present among the ITEMs, then `regexp-substitute/global'
5939 will return after processing a single match.
5941 *** Match Structures
5943 A "match structure" is the object returned by `string-match' and
5944 `regexp-exec'. It describes which portion of a string, if any, matched
5945 the given regular expression. Match structures include: a reference to
5946 the string that was checked for matches; the starting and ending
5947 positions of the regexp match; and, if the regexp included any
5948 parenthesized subexpressions, the starting and ending positions of each
5951 In each of the regexp match functions described below, the `match'
5952 argument must be a match structure returned by a previous call to
5953 `string-match' or `regexp-exec'. Most of these functions return some
5954 information about the original target string that was matched against a
5955 regular expression; we will call that string TARGET for easy reference.
5957 **** Function: regexp-match? OBJ
5958 Return `#t' if OBJ is a match structure returned by a previous
5959 call to `regexp-exec', or `#f' otherwise.
5961 **** Function: match:substring MATCH [N]
5962 Return the portion of TARGET matched by subexpression number N.
5963 Submatch 0 (the default) represents the entire regexp match. If
5964 the regular expression as a whole matched, but the subexpression
5965 number N did not match, return `#f'.
5967 **** Function: match:start MATCH [N]
5968 Return the starting position of submatch number N.
5970 **** Function: match:end MATCH [N]
5971 Return the ending position of submatch number N.
5973 **** Function: match:prefix MATCH
5974 Return the unmatched portion of TARGET preceding the regexp match.
5976 **** Function: match:suffix MATCH
5977 Return the unmatched portion of TARGET following the regexp match.
5979 **** Function: match:count MATCH
5980 Return the number of parenthesized subexpressions from MATCH.
5981 Note that the entire regular expression match itself counts as a
5982 subexpression, and failed submatches are included in the count.
5984 **** Function: match:string MATCH
5985 Return the original TARGET string.
5987 *** Backslash Escapes
5989 Sometimes you will want a regexp to match characters like `*' or `$'
5990 exactly. For example, to check whether a particular string represents
5991 a menu entry from an Info node, it would be useful to match it against
5992 a regexp like `^* [^:]*::'. However, this won't work; because the
5993 asterisk is a metacharacter, it won't match the `*' at the beginning of
5994 the string. In this case, we want to make the first asterisk un-magic.
5996 You can do this by preceding the metacharacter with a backslash
5997 character `\'. (This is also called "quoting" the metacharacter, and
5998 is known as a "backslash escape".) When Guile sees a backslash in a
5999 regular expression, it considers the following glyph to be an ordinary
6000 character, no matter what special meaning it would ordinarily have.
6001 Therefore, we can make the above example work by changing the regexp to
6002 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
6003 to match only a single asterisk in the target string.
6005 Since the backslash is itself a metacharacter, you may force a
6006 regexp to match a backslash in the target string by preceding the
6007 backslash with itself. For example, to find variable references in a
6008 TeX program, you might want to find occurrences of the string `\let\'
6009 followed by any number of alphabetic characters. The regular expression
6010 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
6011 each match a single backslash in the target string.
6013 **** Function: regexp-quote STR
6014 Quote each special character found in STR with a backslash, and
6015 return the resulting string.
6017 *Very important:* Using backslash escapes in Guile source code (as
6018 in Emacs Lisp or C) can be tricky, because the backslash character has
6019 special meaning for the Guile reader. For example, if Guile encounters
6020 the character sequence `\n' in the middle of a string while processing
6021 Scheme code, it replaces those characters with a newline character.
6022 Similarly, the character sequence `\t' is replaced by a horizontal tab.
6023 Several of these "escape sequences" are processed by the Guile reader
6024 before your code is executed. Unrecognized escape sequences are
6025 ignored: if the characters `\*' appear in a string, they will be
6026 translated to the single character `*'.
6028 This translation is obviously undesirable for regular expressions,
6029 since we want to be able to include backslashes in a string in order to
6030 escape regexp metacharacters. Therefore, to make sure that a backslash
6031 is preserved in a string in your Guile program, you must use *two*
6032 consecutive backslashes:
6034 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
6036 The string in this example is preprocessed by the Guile reader before
6037 any code is executed. The resulting argument to `make-regexp' is the
6038 string `^\* [^:]*', which is what we really want.
6040 This also means that in order to write a regular expression that
6041 matches a single backslash character, the regular expression string in
6042 the source code must include *four* backslashes. Each consecutive pair
6043 of backslashes gets translated by the Guile reader to a single
6044 backslash, and the resulting double-backslash is interpreted by the
6045 regexp engine as matching a single backslash character. Hence:
6047 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
6049 The reason for the unwieldiness of this syntax is historical. Both
6050 regular expression pattern matchers and Unix string processing systems
6051 have traditionally used backslashes with the special meanings described
6052 above. The POSIX regular expression specification and ANSI C standard
6053 both require these semantics. Attempting to abandon either convention
6054 would cause other kinds of compatibility problems, possibly more severe
6055 ones. Therefore, without extending the Scheme reader to support
6056 strings with different quoting conventions (an ungainly and confusing
6057 extension when implemented in other languages), we must adhere to this
6058 cumbersome escape syntax.
6060 * Changes to the gh_ interface
6062 * Changes to the scm_ interface
6064 * Changes to system call interfaces:
6066 ** The value returned by `raise' is now unspecified. It throws an exception
6069 *** A new procedure `sigaction' can be used to install signal handlers
6071 (sigaction signum [action] [flags])
6073 signum is the signal number, which can be specified using the value
6076 If action is omitted, sigaction returns a pair: the CAR is the current
6077 signal hander, which will be either an integer with the value SIG_DFL
6078 (default action) or SIG_IGN (ignore), or the Scheme procedure which
6079 handles the signal, or #f if a non-Scheme procedure handles the
6080 signal. The CDR contains the current sigaction flags for the handler.
6082 If action is provided, it is installed as the new handler for signum.
6083 action can be a Scheme procedure taking one argument, or the value of
6084 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
6085 whatever signal handler was installed before sigaction was first used.
6086 Flags can optionally be specified for the new handler (SA_RESTART is
6087 always used if the system provides it, so need not be specified.) The
6088 return value is a pair with information about the old handler as
6091 This interface does not provide access to the "signal blocking"
6092 facility. Maybe this is not needed, since the thread support may
6093 provide solutions to the problem of consistent access to data
6096 *** A new procedure `flush-all-ports' is equivalent to running
6097 `force-output' on every port open for output.
6099 ** Guile now provides information on how it was built, via the new
6100 global variable, %guile-build-info. This variable records the values
6101 of the standard GNU makefile directory variables as an assocation
6102 list, mapping variable names (symbols) onto directory paths (strings).
6103 For example, to find out where the Guile link libraries were
6104 installed, you can say:
6106 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
6109 * Changes to the scm_ interface
6111 ** The new function scm_handle_by_message_noexit is just like the
6112 existing scm_handle_by_message function, except that it doesn't call
6113 exit to terminate the process. Instead, it prints a message and just
6114 returns #f. This might be a more appropriate catch-all handler for
6115 new dynamic roots and threads.
6118 Changes in Guile 1.1 (released Friday, May 16 1997):
6120 * Changes to the distribution.
6122 The Guile 1.0 distribution has been split up into several smaller
6124 guile-core --- the Guile interpreter itself.
6125 guile-tcltk --- the interface between the Guile interpreter and
6126 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
6127 is a toolkit for building graphical user interfaces.
6128 guile-rgx-ctax --- the interface between Guile and the Rx regular
6129 expression matcher, and the translator for the Ctax
6130 programming language. These are packaged together because the
6131 Ctax translator uses Rx to parse Ctax source code.
6133 This NEWS file describes the changes made to guile-core since the 1.0
6136 We no longer distribute the documentation, since it was either out of
6137 date, or incomplete. As soon as we have current documentation, we
6142 * Changes to the stand-alone interpreter
6144 ** guile now accepts command-line arguments compatible with SCSH, Olin
6145 Shivers' Scheme Shell.
6147 In general, arguments are evaluated from left to right, but there are
6148 exceptions. The following switches stop argument processing, and
6149 stash all remaining command-line arguments as the value returned by
6150 the (command-line) function.
6151 -s SCRIPT load Scheme source code from FILE, and exit
6152 -c EXPR evalute Scheme expression EXPR, and exit
6153 -- stop scanning arguments; run interactively
6155 The switches below are processed as they are encountered.
6156 -l FILE load Scheme source code from FILE
6157 -e FUNCTION after reading script, apply FUNCTION to
6158 command line arguments
6159 -ds do -s script at this point
6160 --emacs enable Emacs protocol (experimental)
6161 -h, --help display this help and exit
6162 -v, --version display version information and exit
6163 \ read arguments from following script lines
6165 So, for example, here is a Guile script named `ekko' (thanks, Olin)
6166 which re-implements the traditional "echo" command:
6168 #!/usr/local/bin/guile -s
6171 (map (lambda (arg) (display arg) (display " "))
6175 (main (command-line))
6177 Suppose we invoke this script as follows:
6179 ekko a speckled gecko
6181 Through the magic of Unix script processing (triggered by the `#!'
6182 token at the top of the file), /usr/local/bin/guile receives the
6183 following list of command-line arguments:
6185 ("-s" "./ekko" "a" "speckled" "gecko")
6187 Unix inserts the name of the script after the argument specified on
6188 the first line of the file (in this case, "-s"), and then follows that
6189 with the arguments given to the script. Guile loads the script, which
6190 defines the `main' function, and then applies it to the list of
6191 remaining command-line arguments, ("a" "speckled" "gecko").
6193 In Unix, the first line of a script file must take the following form:
6195 #!INTERPRETER ARGUMENT
6197 where INTERPRETER is the absolute filename of the interpreter
6198 executable, and ARGUMENT is a single command-line argument to pass to
6201 You may only pass one argument to the interpreter, and its length is
6202 limited. These restrictions can be annoying to work around, so Guile
6203 provides a general mechanism (borrowed from, and compatible with,
6204 SCSH) for circumventing them.
6206 If the ARGUMENT in a Guile script is a single backslash character,
6207 `\', Guile will open the script file, parse arguments from its second
6208 and subsequent lines, and replace the `\' with them. So, for example,
6209 here is another implementation of the `ekko' script:
6211 #!/usr/local/bin/guile \
6215 (for-each (lambda (arg) (display arg) (display " "))
6219 If the user invokes this script as follows:
6221 ekko a speckled gecko
6223 Unix expands this into
6225 /usr/local/bin/guile \ ekko a speckled gecko
6227 When Guile sees the `\' argument, it replaces it with the arguments
6228 read from the second line of the script, producing:
6230 /usr/local/bin/guile -e main -s ekko a speckled gecko
6232 This tells Guile to load the `ekko' script, and apply the function
6233 `main' to the argument list ("a" "speckled" "gecko").
6235 Here is how Guile parses the command-line arguments:
6236 - Each space character terminates an argument. This means that two
6237 spaces in a row introduce an empty-string argument.
6238 - The tab character is not permitted (unless you quote it with the
6239 backslash character, as described below), to avoid confusion.
6240 - The newline character terminates the sequence of arguments, and will
6241 also terminate a final non-empty argument. (However, a newline
6242 following a space will not introduce a final empty-string argument;
6243 it only terminates the argument list.)
6244 - The backslash character is the escape character. It escapes
6245 backslash, space, tab, and newline. The ANSI C escape sequences
6246 like \n and \t are also supported. These produce argument
6247 constituents; the two-character combination \n doesn't act like a
6248 terminating newline. The escape sequence \NNN for exactly three
6249 octal digits reads as the character whose ASCII code is NNN. As
6250 above, characters produced this way are argument constituents.
6251 Backslash followed by other characters is not allowed.
6253 * Changes to the procedure for linking libguile with your programs
6255 ** Guile now builds and installs a shared guile library, if your
6256 system support shared libraries. (It still builds a static library on
6257 all systems.) Guile automatically detects whether your system
6258 supports shared libraries. To prevent Guile from buildisg shared
6259 libraries, pass the `--disable-shared' flag to the configure script.
6261 Guile takes longer to compile when it builds shared libraries, because
6262 it must compile every file twice --- once to produce position-
6263 independent object code, and once to produce normal object code.
6265 ** The libthreads library has been merged into libguile.
6267 To link a program against Guile, you now need only link against
6268 -lguile and -lqt; -lthreads is no longer needed. If you are using
6269 autoconf to generate configuration scripts for your application, the
6270 following lines should suffice to add the appropriate libraries to
6273 ### Find quickthreads and libguile.
6274 AC_CHECK_LIB(qt, main)
6275 AC_CHECK_LIB(guile, scm_shell)
6277 * Changes to Scheme functions
6279 ** Guile Scheme's special syntax for keyword objects is now optional,
6280 and disabled by default.
6282 The syntax variation from R4RS made it difficult to port some
6283 interesting packages to Guile. The routines which accepted keyword
6284 arguments (mostly in the module system) have been modified to also
6285 accept symbols whose names begin with `:'.
6287 To change the keyword syntax, you must first import the (ice-9 debug)
6289 (use-modules (ice-9 debug))
6291 Then you can enable the keyword syntax as follows:
6292 (read-set! keywords 'prefix)
6294 To disable keyword syntax, do this:
6295 (read-set! keywords #f)
6297 ** Many more primitive functions accept shared substrings as
6298 arguments. In the past, these functions required normal, mutable
6299 strings as arguments, although they never made use of this
6302 ** The uniform array functions now operate on byte vectors. These
6303 functions are `array-fill!', `serial-array-copy!', `array-copy!',
6304 `serial-array-map', `array-map', `array-for-each', and
6307 ** The new functions `trace' and `untrace' implement simple debugging
6308 support for Scheme functions.
6310 The `trace' function accepts any number of procedures as arguments,
6311 and tells the Guile interpreter to display each procedure's name and
6312 arguments each time the procedure is invoked. When invoked with no
6313 arguments, `trace' returns the list of procedures currently being
6316 The `untrace' function accepts any number of procedures as arguments,
6317 and tells the Guile interpreter not to trace them any more. When
6318 invoked with no arguments, `untrace' untraces all curretly traced
6321 The tracing in Guile has an advantage over most other systems: we
6322 don't create new procedure objects, but mark the procedure objects
6323 themselves. This means that anonymous and internal procedures can be
6326 ** The function `assert-repl-prompt' has been renamed to
6327 `set-repl-prompt!'. It takes one argument, PROMPT.
6328 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
6329 - If PROMPT is a string, we use it as a prompt.
6330 - If PROMPT is a procedure accepting no arguments, we call it, and
6331 display the result as a prompt.
6332 - Otherwise, we display "> ".
6334 ** The new function `eval-string' reads Scheme expressions from a
6335 string and evaluates them, returning the value of the last expression
6336 in the string. If the string contains no expressions, it returns an
6339 ** The new function `thunk?' returns true iff its argument is a
6340 procedure of zero arguments.
6342 ** `defined?' is now a builtin function, instead of syntax. This
6343 means that its argument should be quoted. It returns #t iff its
6344 argument is bound in the current module.
6346 ** The new syntax `use-modules' allows you to add new modules to your
6347 environment without re-typing a complete `define-module' form. It
6348 accepts any number of module names as arguments, and imports their
6349 public bindings into the current module.
6351 ** The new function (module-defined? NAME MODULE) returns true iff
6352 NAME, a symbol, is defined in MODULE, a module object.
6354 ** The new function `builtin-bindings' creates and returns a hash
6355 table containing copies of all the root module's bindings.
6357 ** The new function `builtin-weak-bindings' does the same as
6358 `builtin-bindings', but creates a doubly-weak hash table.
6360 ** The `equal?' function now considers variable objects to be
6361 equivalent if they have the same name and the same value.
6363 ** The new function `command-line' returns the command-line arguments
6364 given to Guile, as a list of strings.
6366 When using guile as a script interpreter, `command-line' returns the
6367 script's arguments; those processed by the interpreter (like `-s' or
6368 `-c') are omitted. (In other words, you get the normal, expected
6369 behavior.) Any application that uses scm_shell to process its
6370 command-line arguments gets this behavior as well.
6372 ** The new function `load-user-init' looks for a file called `.guile'
6373 in the user's home directory, and loads it if it exists. This is
6374 mostly for use by the code generated by scm_compile_shell_switches,
6375 but we thought it might also be useful in other circumstances.
6377 ** The new function `log10' returns the base-10 logarithm of its
6380 ** Changes to I/O functions
6382 *** The functions `read', `primitive-load', `read-and-eval!', and
6383 `primitive-load-path' no longer take optional arguments controlling
6384 case insensitivity and a `#' parser.
6386 Case sensitivity is now controlled by a read option called
6387 `case-insensitive'. The user can add new `#' syntaxes with the
6388 `read-hash-extend' function (see below).
6390 *** The new function `read-hash-extend' allows the user to change the
6391 syntax of Guile Scheme in a somewhat controlled way.
6393 (read-hash-extend CHAR PROC)
6394 When parsing S-expressions, if we read a `#' character followed by
6395 the character CHAR, use PROC to parse an object from the stream.
6396 If PROC is #f, remove any parsing procedure registered for CHAR.
6398 The reader applies PROC to two arguments: CHAR and an input port.
6400 *** The new functions read-delimited and read-delimited! provide a
6401 general mechanism for doing delimited input on streams.
6403 (read-delimited DELIMS [PORT HANDLE-DELIM])
6404 Read until we encounter one of the characters in DELIMS (a string),
6405 or end-of-file. PORT is the input port to read from; it defaults to
6406 the current input port. The HANDLE-DELIM parameter determines how
6407 the terminating character is handled; it should be one of the
6410 'trim omit delimiter from result
6411 'peek leave delimiter character in input stream
6412 'concat append delimiter character to returned value
6413 'split return a pair: (RESULT . TERMINATOR)
6415 HANDLE-DELIM defaults to 'peek.
6417 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
6418 A side-effecting variant of `read-delimited'.
6420 The data is written into the string BUF at the indices in the
6421 half-open interval [START, END); the default interval is the whole
6422 string: START = 0 and END = (string-length BUF). The values of
6423 START and END must specify a well-defined interval in BUF, i.e.
6424 0 <= START <= END <= (string-length BUF).
6426 It returns NBYTES, the number of bytes read. If the buffer filled
6427 up without a delimiter character being found, it returns #f. If the
6428 port is at EOF when the read starts, it returns the EOF object.
6430 If an integer is returned (i.e., the read is successfully terminated
6431 by reading a delimiter character), then the HANDLE-DELIM parameter
6432 determines how to handle the terminating character. It is described
6433 above, and defaults to 'peek.
6435 (The descriptions of these functions were borrowed from the SCSH
6436 manual, by Olin Shivers and Brian Carlstrom.)
6438 *** The `%read-delimited!' function is the primitive used to implement
6439 `read-delimited' and `read-delimited!'.
6441 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
6443 This returns a pair of values: (TERMINATOR . NUM-READ).
6444 - TERMINATOR describes why the read was terminated. If it is a
6445 character or the eof object, then that is the value that terminated
6446 the read. If it is #f, the function filled the buffer without finding
6447 a delimiting character.
6448 - NUM-READ is the number of characters read into BUF.
6450 If the read is successfully terminated by reading a delimiter
6451 character, then the gobble? parameter determines what to do with the
6452 terminating character. If true, the character is removed from the
6453 input stream; if false, the character is left in the input stream
6454 where a subsequent read operation will retrieve it. In either case,
6455 the character is also the first value returned by the procedure call.
6457 (The descriptions of this function was borrowed from the SCSH manual,
6458 by Olin Shivers and Brian Carlstrom.)
6460 *** The `read-line' and `read-line!' functions have changed; they now
6461 trim the terminator by default; previously they appended it to the
6462 returned string. For the old behavior, use (read-line PORT 'concat).
6464 *** The functions `uniform-array-read!' and `uniform-array-write!' now
6465 take new optional START and END arguments, specifying the region of
6466 the array to read and write.
6468 *** The `ungetc-char-ready?' function has been removed. We feel it's
6469 inappropriate for an interface to expose implementation details this
6472 ** Changes to the Unix library and system call interface
6474 *** The new fcntl function provides access to the Unix `fcntl' system
6477 (fcntl PORT COMMAND VALUE)
6478 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
6479 Values for COMMAND are:
6481 F_DUPFD duplicate a file descriptor
6482 F_GETFD read the descriptor's close-on-exec flag
6483 F_SETFD set the descriptor's close-on-exec flag to VALUE
6484 F_GETFL read the descriptor's flags, as set on open
6485 F_SETFL set the descriptor's flags, as set on open to VALUE
6486 F_GETOWN return the process ID of a socket's owner, for SIGIO
6487 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
6488 FD_CLOEXEC not sure what this is
6490 For details, see the documentation for the fcntl system call.
6492 *** The arguments to `select' have changed, for compatibility with
6493 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
6494 expected behavior. The MILLISECONDS parameter has been changed to
6495 MICROSECONDS, to more closely resemble the underlying system call.
6496 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
6497 corresponding return set will be the same.
6499 *** The arguments to the `mknod' system call have changed. They are
6502 (mknod PATH TYPE PERMS DEV)
6503 Create a new file (`node') in the file system. PATH is the name of
6504 the file to create. TYPE is the kind of file to create; it should
6505 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
6506 permission bits to give the newly created file. If TYPE is
6507 'block-special or 'char-special, DEV specifies which device the
6508 special file refers to; its interpretation depends on the kind of
6509 special file being created.
6511 *** The `fork' function has been renamed to `primitive-fork', to avoid
6512 clashing with various SCSH forks.
6514 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
6515 and `recvfrom!'. They no longer accept a size for a second argument;
6516 you must pass a string to hold the received value. They no longer
6517 return the buffer. Instead, `recv' returns the length of the message
6518 received, and `recvfrom' returns a pair containing the packet's length
6519 and originating address.
6521 *** The file descriptor datatype has been removed, as have the
6522 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
6523 We plan to replace these functions with a SCSH-compatible interface.
6525 *** The `create' function has been removed; it's just a special case
6528 *** There are new functions to break down process termination status
6529 values. In the descriptions below, STATUS is a value returned by
6532 (status:exit-val STATUS)
6533 If the child process exited normally, this function returns the exit
6534 code for the child process (i.e., the value passed to exit, or
6535 returned from main). If the child process did not exit normally,
6536 this function returns #f.
6538 (status:stop-sig STATUS)
6539 If the child process was suspended by a signal, this function
6540 returns the signal that suspended the child. Otherwise, it returns
6543 (status:term-sig STATUS)
6544 If the child process terminated abnormally, this function returns
6545 the signal that terminated the child. Otherwise, this function
6548 POSIX promises that exactly one of these functions will return true on
6549 a valid STATUS value.
6551 These functions are compatible with SCSH.
6553 *** There are new accessors and setters for the broken-out time vectors
6554 returned by `localtime', `gmtime', and that ilk. They are:
6556 Component Accessor Setter
6557 ========================= ============ ============
6558 seconds tm:sec set-tm:sec
6559 minutes tm:min set-tm:min
6560 hours tm:hour set-tm:hour
6561 day of the month tm:mday set-tm:mday
6562 month tm:mon set-tm:mon
6563 year tm:year set-tm:year
6564 day of the week tm:wday set-tm:wday
6565 day in the year tm:yday set-tm:yday
6566 daylight saving time tm:isdst set-tm:isdst
6567 GMT offset, seconds tm:gmtoff set-tm:gmtoff
6568 name of time zone tm:zone set-tm:zone
6570 *** There are new accessors for the vectors returned by `uname',
6571 describing the host system:
6574 ============================================== ================
6575 name of the operating system implementation utsname:sysname
6576 network name of this machine utsname:nodename
6577 release level of the operating system utsname:release
6578 version level of the operating system utsname:version
6579 machine hardware platform utsname:machine
6581 *** There are new accessors for the vectors returned by `getpw',
6582 `getpwnam', `getpwuid', and `getpwent', describing entries from the
6583 system's user database:
6586 ====================== =================
6587 user name passwd:name
6588 user password passwd:passwd
6591 real name passwd:gecos
6592 home directory passwd:dir
6593 shell program passwd:shell
6595 *** There are new accessors for the vectors returned by `getgr',
6596 `getgrnam', `getgrgid', and `getgrent', describing entries from the
6597 system's group database:
6600 ======================= ============
6601 group name group:name
6602 group password group:passwd
6604 group members group:mem
6606 *** There are new accessors for the vectors returned by `gethost',
6607 `gethostbyaddr', `gethostbyname', and `gethostent', describing
6611 ========================= ===============
6612 official name of host hostent:name
6613 alias list hostent:aliases
6614 host address type hostent:addrtype
6615 length of address hostent:length
6616 list of addresses hostent:addr-list
6618 *** There are new accessors for the vectors returned by `getnet',
6619 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
6623 ========================= ===============
6624 official name of net netent:name
6625 alias list netent:aliases
6626 net number type netent:addrtype
6627 net number netent:net
6629 *** There are new accessors for the vectors returned by `getproto',
6630 `getprotobyname', `getprotobynumber', and `getprotoent', describing
6634 ========================= ===============
6635 official protocol name protoent:name
6636 alias list protoent:aliases
6637 protocol number protoent:proto
6639 *** There are new accessors for the vectors returned by `getserv',
6640 `getservbyname', `getservbyport', and `getservent', describing
6644 ========================= ===============
6645 official service name servent:name
6646 alias list servent:aliases
6647 port number servent:port
6648 protocol to use servent:proto
6650 *** There are new accessors for the sockaddr structures returned by
6651 `accept', `getsockname', `getpeername', `recvfrom!':
6654 ======================================== ===============
6655 address format (`family') sockaddr:fam
6656 path, for file domain addresses sockaddr:path
6657 address, for internet domain addresses sockaddr:addr
6658 TCP or UDP port, for internet sockaddr:port
6660 *** The `getpwent', `getgrent', `gethostent', `getnetent',
6661 `getprotoent', and `getservent' functions now return #f at the end of
6662 the user database. (They used to throw an exception.)
6664 Note that calling MUMBLEent function is equivalent to calling the
6665 corresponding MUMBLE function with no arguments.
6667 *** The `setpwent', `setgrent', `sethostent', `setnetent',
6668 `setprotoent', and `setservent' routines now take no arguments.
6670 *** The `gethost', `getproto', `getnet', and `getserv' functions now
6671 provide more useful information when they throw an exception.
6673 *** The `lnaof' function has been renamed to `inet-lnaof'.
6675 *** Guile now claims to have the `current-time' feature.
6677 *** The `mktime' function now takes an optional second argument ZONE,
6678 giving the time zone to use for the conversion. ZONE should be a
6679 string, in the same format as expected for the "TZ" environment variable.
6681 *** The `strptime' function now returns a pair (TIME . COUNT), where
6682 TIME is the parsed time as a vector, and COUNT is the number of
6683 characters from the string left unparsed. This function used to
6684 return the remaining characters as a string.
6686 *** The `gettimeofday' function has replaced the old `time+ticks' function.
6687 The return value is now (SECONDS . MICROSECONDS); the fractional
6688 component is no longer expressed in "ticks".
6690 *** The `ticks/sec' constant has been removed, in light of the above change.
6692 * Changes to the gh_ interface
6694 ** gh_eval_str() now returns an SCM object which is the result of the
6697 ** gh_scm2str() now copies the Scheme data to a caller-provided C
6700 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
6701 and returns the array
6703 ** gh_scm2str0() is gone: there is no need to distinguish
6704 null-terminated from non-null-terminated, since gh_scm2newstr() allows
6705 the user to interpret the data both ways.
6707 * Changes to the scm_ interface
6709 ** The new function scm_symbol_value0 provides an easy way to get a
6710 symbol's value from C code:
6712 SCM scm_symbol_value0 (char *NAME)
6713 Return the value of the symbol named by the null-terminated string
6714 NAME in the current module. If the symbol named NAME is unbound in
6715 the current module, return SCM_UNDEFINED.
6717 ** The new function scm_sysintern0 creates new top-level variables,
6718 without assigning them a value.
6720 SCM scm_sysintern0 (char *NAME)
6721 Create a new Scheme top-level variable named NAME. NAME is a
6722 null-terminated string. Return the variable's value cell.
6724 ** The function scm_internal_catch is the guts of catch. It handles
6725 all the mechanics of setting up a catch target, invoking the catch
6726 body, and perhaps invoking the handler if the body does a throw.
6728 The function is designed to be usable from C code, but is general
6729 enough to implement all the semantics Guile Scheme expects from throw.
6731 TAG is the catch tag. Typically, this is a symbol, but this function
6732 doesn't actually care about that.
6734 BODY is a pointer to a C function which runs the body of the catch;
6735 this is the code you can throw from. We call it like this:
6736 BODY (BODY_DATA, JMPBUF)
6738 BODY_DATA is just the BODY_DATA argument we received; we pass it
6739 through to BODY as its first argument. The caller can make
6740 BODY_DATA point to anything useful that BODY might need.
6741 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
6742 which we have just created and initialized.
6744 HANDLER is a pointer to a C function to deal with a throw to TAG,
6745 should one occur. We call it like this:
6746 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
6748 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
6749 same idea as BODY_DATA above.
6750 THROWN_TAG is the tag that the user threw to; usually this is
6751 TAG, but it could be something else if TAG was #t (i.e., a
6752 catch-all), or the user threw to a jmpbuf.
6753 THROW_ARGS is the list of arguments the user passed to the THROW
6756 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
6757 is just a pointer we pass through to HANDLER. We don't actually
6758 use either of those pointers otherwise ourselves. The idea is
6759 that, if our caller wants to communicate something to BODY or
6760 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
6761 HANDLER can then use. Think of it as a way to make BODY and
6762 HANDLER closures, not just functions; MUMBLE_DATA points to the
6765 Of course, it's up to the caller to make sure that any data a
6766 MUMBLE_DATA needs is protected from GC. A common way to do this is
6767 to make MUMBLE_DATA a pointer to data stored in an automatic
6768 structure variable; since the collector must scan the stack for
6769 references anyway, this assures that any references in MUMBLE_DATA
6772 ** The new function scm_internal_lazy_catch is exactly like
6773 scm_internal_catch, except:
6775 - It does not unwind the stack (this is the major difference).
6776 - If handler returns, its value is returned from the throw.
6777 - BODY always receives #f as its JMPBUF argument (since there's no
6778 jmpbuf associated with a lazy catch, because we don't unwind the
6781 ** scm_body_thunk is a new body function you can pass to
6782 scm_internal_catch if you want the body to be like Scheme's `catch'
6783 --- a thunk, or a function of one argument if the tag is #f.
6785 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
6786 contains the Scheme procedure to invoke as the body, and the tag
6787 we're catching. If the tag is #f, then we pass JMPBUF (created by
6788 scm_internal_catch) to the body procedure; otherwise, the body gets
6791 ** scm_handle_by_proc is a new handler function you can pass to
6792 scm_internal_catch if you want the handler to act like Scheme's catch
6793 --- call a procedure with the tag and the throw arguments.
6795 If the user does a throw to this catch, this function runs a handler
6796 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
6797 variable holding the Scheme procedure object to invoke. It ought to
6798 be a pointer to an automatic variable (i.e., one living on the stack),
6799 or the procedure object should be otherwise protected from GC.
6801 ** scm_handle_by_message is a new handler function to use with
6802 `scm_internal_catch' if you want Guile to print a message and die.
6803 It's useful for dealing with throws to uncaught keys at the top level.
6805 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
6806 message header to print; if zero, we use "guile" instead. That
6807 text is followed by a colon, then the message described by ARGS.
6809 ** The return type of scm_boot_guile is now void; the function does
6810 not return a value, and indeed, never returns at all.
6812 ** The new function scm_shell makes it easy for user applications to
6813 process command-line arguments in a way that is compatible with the
6814 stand-alone guile interpreter (which is in turn compatible with SCSH,
6817 To use the scm_shell function, first initialize any guile modules
6818 linked into your application, and then call scm_shell with the values
6819 of ARGC and ARGV your `main' function received. scm_shell will add
6820 any SCSH-style meta-arguments from the top of the script file to the
6821 argument vector, and then process the command-line arguments. This
6822 generally means loading a script file or starting up an interactive
6823 command interpreter. For details, see "Changes to the stand-alone
6826 ** The new functions scm_get_meta_args and scm_count_argv help you
6827 implement the SCSH-style meta-argument, `\'.
6829 char **scm_get_meta_args (int ARGC, char **ARGV)
6830 If the second element of ARGV is a string consisting of a single
6831 backslash character (i.e. "\\" in Scheme notation), open the file
6832 named by the following argument, parse arguments from it, and return
6833 the spliced command line. The returned array is terminated by a
6836 For details of argument parsing, see above, under "guile now accepts
6837 command-line arguments compatible with SCSH..."
6839 int scm_count_argv (char **ARGV)
6840 Count the arguments in ARGV, assuming it is terminated by a null
6843 For an example of how these functions might be used, see the source
6844 code for the function scm_shell in libguile/script.c.
6846 You will usually want to use scm_shell instead of calling this
6849 ** The new function scm_compile_shell_switches turns an array of
6850 command-line arguments into Scheme code to carry out the actions they
6851 describe. Given ARGC and ARGV, it returns a Scheme expression to
6852 evaluate, and calls scm_set_program_arguments to make any remaining
6853 command-line arguments available to the Scheme code. For example,
6854 given the following arguments:
6856 -e main -s ekko a speckled gecko
6858 scm_set_program_arguments will return the following expression:
6860 (begin (load "ekko") (main (command-line)) (quit))
6862 You will usually want to use scm_shell instead of calling this
6865 ** The function scm_shell_usage prints a usage message appropriate for
6866 an interpreter that uses scm_compile_shell_switches to handle its
6867 command-line arguments.
6869 void scm_shell_usage (int FATAL, char *MESSAGE)
6870 Print a usage message to the standard error output. If MESSAGE is
6871 non-zero, write it before the usage message, followed by a newline.
6872 If FATAL is non-zero, exit the process, using FATAL as the
6873 termination status. (If you want to be compatible with Guile,
6874 always use 1 as the exit status when terminating due to command-line
6877 You will usually want to use scm_shell instead of calling this
6880 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
6881 expressions. It used to return SCM_EOL. Earth-shattering.
6883 ** The macros for declaring scheme objects in C code have been
6884 rearranged slightly. They are now:
6886 SCM_SYMBOL (C_NAME, SCHEME_NAME)
6887 Declare a static SCM variable named C_NAME, and initialize it to
6888 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
6889 be a C identifier, and SCHEME_NAME should be a C string.
6891 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
6892 Just like SCM_SYMBOL, but make C_NAME globally visible.
6894 SCM_VCELL (C_NAME, SCHEME_NAME)
6895 Create a global variable at the Scheme level named SCHEME_NAME.
6896 Declare a static SCM variable named C_NAME, and initialize it to
6897 point to the Scheme variable's value cell.
6899 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
6900 Just like SCM_VCELL, but make C_NAME globally visible.
6902 The `guile-snarf' script writes initialization code for these macros
6903 to its standard output, given C source code as input.
6905 The SCM_GLOBAL macro is gone.
6907 ** The scm_read_line and scm_read_line_x functions have been replaced
6908 by Scheme code based on the %read-delimited! procedure (known to C
6909 code as scm_read_delimited_x). See its description above for more
6912 ** The function scm_sys_open has been renamed to scm_open. It now
6913 returns a port instead of an FD object.
6915 * The dynamic linking support has changed. For more information, see
6916 libguile/DYNAMIC-LINKING.
6921 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
6924 * Changes to the 'guile' program:
6926 ** Guile now loads some new files when it starts up. Guile first
6927 searches the load path for init.scm, and loads it if found. Then, if
6928 Guile is not being used to execute a script, and the user's home
6929 directory contains a file named `.guile', Guile loads that.
6931 ** You can now use Guile as a shell script interpreter.
6933 To paraphrase the SCSH manual:
6935 When Unix tries to execute an executable file whose first two
6936 characters are the `#!', it treats the file not as machine code to
6937 be directly executed by the native processor, but as source code
6938 to be executed by some interpreter. The interpreter to use is
6939 specified immediately after the #! sequence on the first line of
6940 the source file. The kernel reads in the name of the interpreter,
6941 and executes that instead. It passes the interpreter the source
6942 filename as its first argument, with the original arguments
6943 following. Consult the Unix man page for the `exec' system call
6944 for more information.
6946 Now you can use Guile as an interpreter, using a mechanism which is a
6947 compatible subset of that provided by SCSH.
6949 Guile now recognizes a '-s' command line switch, whose argument is the
6950 name of a file of Scheme code to load. It also treats the two
6951 characters `#!' as the start of a comment, terminated by `!#'. Thus,
6952 to make a file of Scheme code directly executable by Unix, insert the
6953 following two lines at the top of the file:
6955 #!/usr/local/bin/guile -s
6958 Guile treats the argument of the `-s' command-line switch as the name
6959 of a file of Scheme code to load, and treats the sequence `#!' as the
6960 start of a block comment, terminated by `!#'.
6962 For example, here's a version of 'echo' written in Scheme:
6964 #!/usr/local/bin/guile -s
6966 (let loop ((args (cdr (program-arguments))))
6969 (display (car args))
6970 (if (pair? (cdr args))
6972 (loop (cdr args)))))
6975 Why does `#!' start a block comment terminated by `!#', instead of the
6976 end of the line? That is the notation SCSH uses, and although we
6977 don't yet support the other SCSH features that motivate that choice,
6978 we would like to be backward-compatible with any existing Guile
6979 scripts once we do. Furthermore, if the path to Guile on your system
6980 is too long for your kernel, you can start the script with this
6984 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
6987 Note that some very old Unix systems don't support the `#!' syntax.
6990 ** You can now run Guile without installing it.
6992 Previous versions of the interactive Guile interpreter (`guile')
6993 couldn't start up unless Guile's Scheme library had been installed;
6994 they used the value of the environment variable `SCHEME_LOAD_PATH'
6995 later on in the startup process, but not to find the startup code
6996 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
6999 To run Guile without installing it, build it in the normal way, and
7000 then set the environment variable `SCHEME_LOAD_PATH' to a
7001 colon-separated list of directories, including the top-level directory
7002 of the Guile sources. For example, if you unpacked Guile so that the
7003 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
7006 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
7009 ** Guile's read-eval-print loop no longer prints #<unspecified>
7010 results. If the user wants to see this, she can evaluate the
7011 expression (assert-repl-print-unspecified #t), perhaps in her startup
7014 ** Guile no longer shows backtraces by default when an error occurs;
7015 however, it does display a message saying how to get one, and how to
7016 request that they be displayed by default. After an error, evaluate
7018 to see a backtrace, and
7019 (debug-enable 'backtrace)
7020 to see them by default.
7024 * Changes to Guile Scheme:
7026 ** Guile now distinguishes between #f and the empty list.
7028 This is for compatibility with the IEEE standard, the (possibly)
7029 upcoming Revised^5 Report on Scheme, and many extant Scheme
7032 Guile used to have #f and '() denote the same object, to make Scheme's
7033 type system more compatible with Emacs Lisp's. However, the change
7034 caused too much trouble for Scheme programmers, and we found another
7035 way to reconcile Emacs Lisp with Scheme that didn't require this.
7038 ** Guile's delq, delv, delete functions, and their destructive
7039 counterparts, delq!, delv!, and delete!, now remove all matching
7040 elements from the list, not just the first. This matches the behavior
7041 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
7042 functions which inspired them.
7044 I recognize that this change may break code in subtle ways, but it
7045 seems best to make the change before the FSF's first Guile release,
7049 ** The compiled-library-path function has been deleted from libguile.
7051 ** The facilities for loading Scheme source files have changed.
7053 *** The variable %load-path now tells Guile which directories to search
7054 for Scheme code. Its value is a list of strings, each of which names
7057 *** The variable %load-extensions now tells Guile which extensions to
7058 try appending to a filename when searching the load path. Its value
7059 is a list of strings. Its default value is ("" ".scm").
7061 *** (%search-load-path FILENAME) searches the directories listed in the
7062 value of the %load-path variable for a Scheme file named FILENAME,
7063 with all the extensions listed in %load-extensions. If it finds a
7064 match, then it returns its full filename. If FILENAME is absolute, it
7065 returns it unchanged. Otherwise, it returns #f.
7067 %search-load-path will not return matches that refer to directories.
7069 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
7070 uses %seach-load-path to find a file named FILENAME, and loads it if
7071 it finds it. If it can't read FILENAME for any reason, it throws an
7074 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
7077 *** load uses the same searching semantics as primitive-load.
7079 *** The functions %try-load, try-load-with-path, %load, load-with-path,
7080 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
7081 path, and load-module-with-path have been deleted. The functions
7082 above should serve their purposes.
7084 *** If the value of the variable %load-hook is a procedure,
7085 `primitive-load' applies its value to the name of the file being
7086 loaded (without the load path directory name prepended). If its value
7087 is #f, it is ignored. Otherwise, an error occurs.
7089 This is mostly useful for printing load notification messages.
7092 ** The function `eval!' is no longer accessible from the scheme level.
7093 We can't allow operations which introduce glocs into the scheme level,
7094 because Guile's type system can't handle these as data. Use `eval' or
7095 `read-and-eval!' (see below) as replacement.
7097 ** The new function read-and-eval! reads an expression from PORT,
7098 evaluates it, and returns the result. This is more efficient than
7099 simply calling `read' and `eval', since it is not necessary to make a
7100 copy of the expression for the evaluator to munge.
7102 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
7103 for the `read' function.
7106 ** The function `int?' has been removed; its definition was identical
7107 to that of `integer?'.
7109 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
7110 use the R4RS names for these functions.
7112 ** The function object-properties no longer returns the hash handle;
7113 it simply returns the object's property list.
7115 ** Many functions have been changed to throw errors, instead of
7116 returning #f on failure. The point of providing exception handling in
7117 the language is to simplify the logic of user code, but this is less
7118 useful if Guile's primitives don't throw exceptions.
7120 ** The function `fileno' has been renamed from `%fileno'.
7122 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
7125 * Changes to Guile's C interface:
7127 ** The library's initialization procedure has been simplified.
7128 scm_boot_guile now has the prototype:
7130 void scm_boot_guile (int ARGC,
7132 void (*main_func) (),
7135 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
7136 MAIN_FUNC should do all the work of the program (initializing other
7137 packages, reading user input, etc.) before returning. When MAIN_FUNC
7138 returns, call exit (0); this function never returns. If you want some
7139 other exit value, MAIN_FUNC may call exit itself.
7141 scm_boot_guile arranges for program-arguments to return the strings
7142 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
7143 scm_set_program_arguments with the final list, so Scheme code will
7144 know which arguments have been processed.
7146 scm_boot_guile establishes a catch-all catch handler which prints an
7147 error message and exits the process. This means that Guile exits in a
7148 coherent way when system errors occur and the user isn't prepared to
7149 handle it. If the user doesn't like this behavior, they can establish
7150 their own universal catcher in MAIN_FUNC to shadow this one.
7152 Why must the caller do all the real work from MAIN_FUNC? The garbage
7153 collector assumes that all local variables of type SCM will be above
7154 scm_boot_guile's stack frame on the stack. If you try to manipulate
7155 SCM values after this function returns, it's the luck of the draw
7156 whether the GC will be able to find the objects you allocate. So,
7157 scm_boot_guile function exits, rather than returning, to discourage
7158 people from making that mistake.
7160 The IN, OUT, and ERR arguments were removed; there are other
7161 convenient ways to override these when desired.
7163 The RESULT argument was deleted; this function should never return.
7165 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
7169 ** Guile's header files should no longer conflict with your system's
7172 In order to compile code which #included <libguile.h>, previous
7173 versions of Guile required you to add a directory containing all the
7174 Guile header files to your #include path. This was a problem, since
7175 Guile's header files have names which conflict with many systems'
7178 Now only <libguile.h> need appear in your #include path; you must
7179 refer to all Guile's other header files as <libguile/mumble.h>.
7180 Guile's installation procedure puts libguile.h in $(includedir), and
7181 the rest in $(includedir)/libguile.
7184 ** Two new C functions, scm_protect_object and scm_unprotect_object,
7185 have been added to the Guile library.
7187 scm_protect_object (OBJ) protects OBJ from the garbage collector.
7188 OBJ will not be freed, even if all other references are dropped,
7189 until someone does scm_unprotect_object (OBJ). Both functions
7192 Note that calls to scm_protect_object do not nest. You can call
7193 scm_protect_object any number of times on a given object, and the
7194 next call to scm_unprotect_object will unprotect it completely.
7196 Basically, scm_protect_object and scm_unprotect_object just
7197 maintain a list of references to things. Since the GC knows about
7198 this list, all objects it mentions stay alive. scm_protect_object
7199 adds its argument to the list; scm_unprotect_object remove its
7200 argument from the list.
7203 ** scm_eval_0str now returns the value of the last expression
7206 ** The new function scm_read_0str reads an s-expression from a
7207 null-terminated string, and returns it.
7209 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
7210 to a Scheme port object.
7212 ** The new function `scm_set_program_arguments' allows C code to set
7213 the value returned by the Scheme `program-arguments' function.
7218 * Guile no longer includes sophisticated Tcl/Tk support.
7220 The old Tcl/Tk support was unsatisfying to us, because it required the
7221 user to link against the Tcl library, as well as Tk and Guile. The
7222 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
7223 referring to widgets by names, rather than exporting widgets to Scheme
7224 code as a special datatype.
7226 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
7227 maintainers described some very interesting changes in progress to the
7228 Tcl/Tk internals, which would facilitate clean interfaces between lone
7229 Tk and other interpreters --- even for garbage-collected languages
7230 like Scheme. They expected the new Tk to be publicly available in the
7233 Since it seems that Guile might soon have a new, cleaner interface to
7234 lone Tk, and that the old Guile/Tk glue code would probably need to be
7235 completely rewritten, we (Jim Blandy and Richard Stallman) have
7236 decided not to support the old code. We'll spend the time instead on
7237 a good interface to the newer Tk, as soon as it is available.
7239 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
7242 Copyright information:
7244 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
7246 Permission is granted to anyone to make or distribute verbatim copies
7247 of this document as received, in any medium, provided that the
7248 copyright notice and this permission notice are preserved,
7249 thus giving the recipient permission to redistribute in turn.
7251 Permission is granted to distribute modified versions
7252 of this document, or of portions of it,
7253 under the above conditions, provided also that they
7254 carry prominent notices stating who last changed them.
7259 paragraph-separate: "[
\f]*$"