1 Guile NEWS --- history of user-visible changes. -*- text -*-
2 Copyright (C) 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send Guile bug reports to bug-guile@gnu.org.
7 Changes since Guile 1.3.4:
9 * Changes to the distribution
11 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
13 We've changed the way we handle generated files in the Guile source
14 repository. As a result, the procedure for building trees obtained
15 from the nightly FTP snapshots or via CVS has changed:
16 - You must have appropriate versions of autoconf, automake, and
17 libtool installed on your system. See README for info on how to
18 obtain these programs.
19 - Before configuring the tree, you must first run the script
20 `autogen.sh' at the top of the source tree.
22 The Guile repository used to contain not only source files, written by
23 humans, but also some generated files, like configure scripts and
24 Makefile.in files. Even though the contents of these files could be
25 derived mechanically from other files present, we thought it would
26 make the tree easier to build if we checked them into CVS.
28 However, this approach means that minor differences between
29 developer's installed tools and habits affected the whole team.
30 So we have removed the generated files from the repository, and
31 added the autogen.sh script, which will reconstruct them
35 ** configure now has experimental options to remove support for certain
38 --disable-arrays omit array and uniform array support
39 --disable-posix omit posix interfaces
40 --disable-networking omit networking interfaces
41 --disable-regex omit regular expression interfaces
43 These are likely to become separate modules some day.
45 ** New configure option --enable-debug-freelist
47 This enables a debugging version of SCM_NEWCELL(), and also registers
48 an extra primitive, the setter `gc-set-debug-check-freelist!'.
50 Configure with the --enable-debug-freelist option to enable
51 the gc-set-debug-check-freelist! primitive, and then use:
53 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
54 (gc-set-debug-check-freelist! #f) # turn off checking
56 Checking of the freelist forces a traversal of the freelist and
57 a garbage collection before each allocation of a cell. This can
58 slow down the interpreter dramatically, so the setter should be used to
59 turn on this extra processing only when necessary.
61 ** New configure option --enable-debug-malloc
63 Include code for debugging of calls to scm_must_malloc/realloc/free.
67 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
68 2. objects reallocated by scm_must_realloc has been allocated by
70 3. reallocated objects are reallocated with the same what string
72 But, most importantly, it records the number of allocated objects of
73 each kind. This is useful when searching for memory leaks.
75 A Guile compiled with this option provides the primitive
76 `malloc-stats' which returns an alist with pairs of kind and the
77 number of objects of that kind.
79 ** All includes are now referenced relative to the root directory
81 Since some users have had problems with mixups between Guile and
82 system headers, we have decided to always refer to Guile headers via
83 their parent directories. This essentially creates a "private name
84 space" for Guile headers. This means that the compiler only is given
85 -I options for the root build and root source directory.
87 ** Header files kw.h and genio.h have been removed.
89 ** The module (ice-9 getopt-gnu-style) has been removed.
91 ** New module (ice-9 documentation)
93 Implements the interface to documentation strings associated with
96 * Changes to the stand-alone interpreter
98 ** New command line option --debug
100 Start Guile with debugging evaluator and backtraces enabled.
102 This is useful when debugging your .guile init file or scripts.
106 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
107 (help REGEXP) ditto for objects with names matching REGEXP (a string)
108 (help ,EXPR) gives documentation for object returned by EXPR
109 (help) gives this text
111 `help' searches among bindings exported from loaded modules, while
112 `apropos' searches among bindings visible from the "current" module.
114 Examples: (help help)
116 (help "output-string")
118 ** `help' and `apropos' now prints full module names
120 ** Dynamic linking now uses libltdl from the libtool package.
122 The old system dependent code for doing dynamic linking has been
123 replaced with calls to the libltdl functions which do all the hairy
126 The major improvement is that you can now directly pass libtool
127 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
128 will be able to do the best shared library job you can get, via
131 The way dynamic libraries are found has changed and is not really
132 portable across platforms, probably. It is therefore recommended to
133 use absolute filenames when possible.
135 If you pass a filename without an extension to `dynamic-link', it will
136 try a few appropriate ones. Thus, the most platform ignorant way is
137 to specify a name like "libfoo", without any directories and
140 ** Guile COOP threads are now compatible with LinuxThreads
142 Previously, COOP threading wasn't possible in applications linked with
143 Linux POSIX threads due to their use of the stack pointer to find the
144 thread context. This has now been fixed with a workaround which uses
145 the pthreads to allocate the stack.
147 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
149 ** Positions of erring expression in scripts
151 With version 1.3.4, the location of the erring expression in Guile
152 scipts is no longer automatically reported. (This should have been
153 documented before the 1.3.4 release.)
155 You can get this information by enabling recording of positions of
156 source expressions and running the debugging evaluator. Put this at
157 the top of your script (or in your "site" file):
159 (read-enable 'positions)
160 (debug-enable 'debug)
162 ** Backtraces in scripts
164 It is now possible to get backtraces in scripts.
168 (debug-enable 'debug 'backtrace)
170 at the top of the script.
172 (The first options enables the debugging evaluator.
173 The second enables backtraces.)
175 ** Part of module system symbol lookup now implemented in C
177 The eval closure of most modules is now implemented in C. Since this
178 was one of the bottlenecks for loading speed, Guile now loads code
179 substantially faster than before.
181 ** Attempting to get the value of an unbound variable now produces
182 an exception with a key of 'unbound-variable instead of 'misc-error.
184 ** The initial default output port is now unbuffered if it's using a
185 tty device. Previously in this situation it was line-buffered.
187 ** gc-thunk is deprecated
189 gc-thunk will be removed in next release of Guile. It has been
190 replaced by after-gc-hook.
192 ** New hook: after-gc-hook
194 after-gc-hook takes over the role of gc-thunk. This hook is run at
195 the first SCM_TICK after a GC. (Thus, the code is run at the same
196 point during evaluation as signal handlers.)
198 Note that this hook should be used only for diagnostic and debugging
199 purposes. It is not certain that it will continue to be well-defined
200 when this hook is run in the future.
202 C programmers: Note the new C level hooks scm_before_gc_c_hook,
203 scm_before_sweep_c_hook, scm_after_gc_c_hook.
205 ** Improvements to garbage collector
207 Guile 1.4 has a new policy for triggering heap allocation and
208 determining the sizes of heap segments. It fixes a number of problems
211 1. The new policy can handle two separate pools of cells
212 (2-word/4-word) better. (The old policy would run wild, allocating
213 more and more memory for certain programs.)
215 2. The old code would sometimes allocate far too much heap so that the
216 Guile process became gigantic. The new code avoids this.
218 3. The old code would sometimes allocate too little so that few cells
219 were freed at GC so that, in turn, too much time was spent in GC.
221 4. The old code would often trigger heap allocation several times in a
222 row. (The new scheme predicts how large the segments needs to be
223 in order not to need further allocation.)
225 All in all, the new GC policy will make larger applications more
228 The new GC scheme also is prepared for POSIX threading. Threads can
229 allocate private pools of cells ("clusters") with just a single
230 function call. Allocation of single cells from such a cluster can
231 then proceed without any need of inter-thread synchronization.
233 ** New environment variables controlling GC parameters
235 GUILE_MAX_SEGMENT_SIZE Maximal segment size
238 Allocation of 2-word cell heaps:
240 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
243 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
244 GC in percent of total heap size
247 Allocation of 4-word cell heaps
248 (used for real numbers and misc other objects):
250 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
252 (See entry "Way for application to customize GC parameters" under
253 section "Changes to the scm_ interface" below.)
255 ** Guile now implements reals using 4-word cells
257 This speeds up computation with reals. (They were earlier allocated
258 with `malloc'.) There is still some room for optimizations, however.
260 ** Some further steps toward POSIX thread support have been taken
262 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
263 don't have much effect any longer, and many of them will be removed in
267 are only handled at the top of the evaluator loop, immediately after
268 I/O, and in scm_equalp.
270 *** The GC can allocate thread private pools of pairs.
272 * Changes to Scheme functions and syntax
274 ** close-input-port and close-output-port are now R5RS
276 These procedures have been turned into primitives and have R5RS behaviour.
278 ** New procedure: simple-format PORT MESSAGE ARG1 ...
280 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
281 extended by the more sophisticated version in (ice-9 format)
283 (simple-format port message . args)
284 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
285 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
286 the escapes are replaced with corresponding members of ARGS:
287 ~A formats using `display' and ~S formats using `write'.
288 If DESTINATION is #t, then use the `current-output-port',
289 if DESTINATION is #f, then return a string containing the formatted text.
290 Does not add a trailing newline."
292 ** string-ref: the second argument is no longer optional.
294 ** string, list->string: no longer accept strings in their arguments,
295 only characters, for compatibility with R5RS.
297 ** New procedure: port-closed? PORT
298 Returns #t if PORT is closed or #f if it is open.
302 The list* functionality is now provided by cons* (SRFI-1 compliant)
304 ** New procedure: cons* ARG1 ARG2 ... ARGn
306 Like `list', but the last arg provides the tail of the constructed list,
307 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
309 Requires at least one argument. If given one argument, that argument
310 is returned as result.
312 This function is called `list*' in some other Schemes and in Common LISP.
314 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
316 ** New procedure: object-documentation OBJECT
318 Returns the documentation string associated with OBJECT. The
319 procedure uses a caching mechanism so that subsequent lookups are
322 Exported by (ice-9 documentation).
324 ** module-name now returns full names of modules
326 Previously, only the last part of the name was returned (`session' for
327 `(ice-9 session)'). Ex: `(ice-9 session)'.
329 * Changes to the gh_ interface
331 ** Deprecated: gh_int2scmb
333 Use gh_bool2scm instead.
335 * Changes to the scm_ interface
337 ** Guile primitives now carry docstrings!
339 Thanks to Greg Badros!
341 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
343 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
344 macros and must contain a docstring that is extracted into foo.doc using a new
345 guile-doc-snarf script (that uses guile-doc-snarf.awk).
347 However, a major overhaul of these macros is scheduled for the next release of
350 ** Guile primitives use a new technique for validation of arguments
352 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
353 the readability of argument checking.
355 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
357 ** New macros: SCM_PACK, SCM_UNPACK
359 Compose/decompose an SCM value.
361 The SCM type is now treated as an abstract data type and may be defined as a
362 long, a void* or as a struct, depending on the architecture and compile time
363 options. This makes it easier to find several types of bugs, for example when
364 SCM values are treated as integers without conversion. Values of the SCM type
365 should be treated as "atomic" values. These macros are used when
366 composing/decomposing an SCM value, either because you want to access
367 individual bits, or because you want to treat it as an integer value.
369 E.g., in order to set bit 7 in an SCM value x, use the expression
371 SCM_PACK (SCM_UNPACK (x) | 0x80)
373 ** The name property of hooks is deprecated.
374 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
376 You can emulate this feature by using object properties.
378 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
379 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
382 These macros will be removed in a future release of Guile.
384 ** The following types, functions and macros from numbers.h are deprecated:
385 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
386 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
388 Further, it is recommended not to rely on implementation details for guile's
389 current implementation of bignums. It is planned to replace this
390 implementation with gmp in the future.
392 ** Port internals: the rw_random variable in the scm_port structure
393 must be set to non-zero in any random access port. In recent Guile
394 releases it was only set for bidirectional random-access ports.
396 ** Port internals: the seek ptob procedure is now responsible for
397 resetting the buffers if required. The change was made so that in the
398 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
399 the fport and strport ptobs can avoid resetting the buffers,
400 in particular to avoid discarding unread chars. An existing port
401 type can be fixed by adding something like the following to the
402 beginning of the ptob seek procedure:
404 if (pt->rw_active == SCM_PORT_READ)
405 scm_end_input (object);
406 else if (pt->rw_active == SCM_PORT_WRITE)
407 ptob->flush (object);
409 although to actually avoid resetting the buffers and discard unread
410 chars requires further hacking that depends on the characteristics
413 ** Deprecated functions: scm_fseek, scm_tag
415 These functions are no longer used and will be removed in a future version.
417 ** The scm_sysmissing procedure is no longer used in libguile.
418 Unless it turns out to be unexpectedly useful to somebody, it will be
419 removed in a future version.
421 ** The format of error message strings has changed
423 The two C procedures: scm_display_error and scm_error, as well as the
424 primitive `scm-error', now use scm_simple_format to do their work.
425 This means that the message strings of all code must be updated to use
426 ~A where %s was used before, and ~S where %S was used before.
428 During the period when there still are a lot of old Guiles out there,
429 you might want to support both old and new versions of Guile.
431 There are basically two methods to achieve this. Both methods use
434 AC_CHECK_FUNCS(scm_simple_format)
436 in your configure.in.
438 Method 1: Use the string concatenation features of ANSI C's
443 #ifdef HAVE_SCM_SIMPLE_FORMAT
449 Then represent each of your error messages using a preprocessor macro:
451 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
455 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
456 (define make-message string-append)
458 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
460 Method 2: Use the oldfmt function found in doc/oldfmt.c.
464 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
469 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
473 ** Deprecated: coop_mutex_init, coop_condition_variable_init
475 Don't use the functions coop_mutex_init and
476 coop_condition_variable_init. They will change.
478 Use scm_mutex_init and scm_cond_init instead.
480 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
481 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
482 COND, as `scm_cond_wait' does, but it also bounds the duration
483 of the wait. If COND has not been signaled before time ABSTIME,
484 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
485 returns the error code `ETIMEDOUT'.
487 The ABSTIME parameter specifies an absolute time, with the same
488 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
489 to 00:00:00 GMT, January 1, 1970.
491 ** New function: scm_cond_broadcast (scm_cond_t *COND)
492 `scm_cond_broadcast' restarts all the threads that are waiting
493 on the condition variable COND. Nothing happens if no threads are
496 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
497 `scm_key_create' allocates a new TSD key. The key is stored in
498 the location pointed to by KEY. There is no limit on the number
499 of keys allocated at a given time. The value initially associated
500 with the returned key is `NULL' in all currently executing threads.
502 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
503 function associated with the key. When a thread terminates,
504 DESTR_FUNCTION is called on the value associated with the key in
505 that thread. The DESTR_FUNCTION is not called if a key is deleted
506 with `scm_key_delete' or a value is changed with
507 `scm_setspecific'. The order in which destructor functions are
508 called at thread termination time is unspecified.
510 Destructors are not yet implemented.
512 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
513 `scm_setspecific' changes the value associated with KEY in the
514 calling thread, storing the given POINTER instead.
516 ** New function: scm_getspecific (scm_key_t KEY)
517 `scm_getspecific' returns the value currently associated with
518 KEY in the calling thread.
520 ** New function: scm_key_delete (scm_key_t KEY)
521 `scm_key_delete' deallocates a TSD key. It does not check
522 whether non-`NULL' values are associated with that key in the
523 currently executing threads, nor call the destructor function
524 associated with the key.
526 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
528 Initialize a C level hook HOOK with associated HOOK_DATA and type
529 TYPE. (See scm_c_hook_run ().)
531 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
533 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
534 is true, add it last, otherwise first. The same FUNC can be added
535 multiple times if FUNC_DATA differ and vice versa.
537 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
539 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
540 function is only removed if both FUNC and FUNC_DATA matches.
542 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
544 Run hook HOOK passing DATA to the hook functions.
546 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
547 returned is undefined.
549 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
550 returns a non-NULL value. This value is returned as the result of
551 scm_c_hook_run. If all functions return NULL, NULL is returned.
553 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
554 returns a NULL value, and NULL is returned. If all functions returns
555 a non-NULL value, the last value is returned.
557 ** New C level GC hooks
559 Five new C level hooks has been added to the garbage collector.
564 are run before locking and after unlocking the heap. The system is
565 thus in a mode where evaluation can take place. (Except that
566 scm_before_gc_c_hook must not allocate new cells.)
568 scm_before_mark_c_hook
569 scm_before_sweep_c_hook
570 scm_after_sweep_c_hook
572 are run when the heap is locked. These are intended for extension of
573 the GC in a modular fashion. Examples are the weaks and guardians
576 ** Way for application to customize GC parameters
578 The application can set up other default values for the GC heap
579 allocation parameters
581 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
582 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
583 GUILE_MAX_SEGMENT_SIZE,
587 scm_default_init_heap_size_1, scm_default_min_yield_1,
588 scm_default_init_heap_size_2, scm_default_min_yield_2,
589 scm_default_max_segment_size
591 respectively before callong scm_boot_guile.
593 (See entry "New environment variables ..." in section
594 "Changes to the stand-alone interpreter" above.)
596 ** scm_protect_object/scm_unprotect_object now nest
598 This means that you can call scm_protect_object multiple times on an
599 object and count on the object being protected until
600 scm_unprotect_object has been call the same number of times.
602 The functions also have better time complexity.
604 Still, it is usually possible to structure the application in a way
605 that you don't need to use these functions. For example, if you use a
606 protected standard Guile list to keep track of live objects rather
607 than some custom data type, objects will die a natural death when they
608 are no longer needed.
610 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
612 Guile does not provide the float representation for inexact real numbers any
613 more. Now, only doubles are used to represent inexact real numbers. Further,
614 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
615 and scm_tc16_complex, respectively.
617 ** Removed deprecated type scm_smobfuns
619 ** Removed deprecated function scm_newsmob
621 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
623 There is an ongoing discussion among the developers whether to
624 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
625 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
626 until this issue has been settled.
628 ** Removed deprecated type tag scm_tc16_kw
630 ** Added type tag scm_tc16_keyword
632 (This was introduced already in release 1.3.4 but was not documented
635 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
637 * Changes to system call interfaces:
639 ** The "select" procedure now tests port buffers for the ability to
640 provide input or accept output. Previously only the underlying file
641 descriptors were checked.
643 ** New variable PIPE_BUF: the maximum number of bytes that can be
644 atomically written to a pipe.
646 ** If a facility is not available on the system when Guile is
647 compiled, the corresponding primitive procedure will not be defined.
648 Previously it would have been defined but would throw a system-error
649 exception if called. Exception handlers which catch this case may
650 need minor modification: an error will be thrown with key
651 'unbound-variable instead of 'system-error. Alternatively it's
652 now possible to use `defined?' to check whether the facility is
655 ** Procedures which depend on the timezone should now give the correct
656 result on systems which cache the TZ environment variable, even if TZ
657 is changed without calling tzset.
659 * Changes to the networking interfaces:
661 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
662 long integers between network and host format. For now, it's not
663 particularly convenient to do this kind of thing, but consider:
665 (define write-network-long
667 (let ((v (make-uniform-vector 1 1 0)))
668 (uniform-vector-set! v 0 (htonl value))
669 (uniform-vector-write v port))))
671 (define read-network-long
673 (let ((v (make-uniform-vector 1 1 0)))
674 (uniform-vector-read! v port)
675 (ntohl (uniform-vector-ref v 0)))))
677 ** If inet-aton fails, it now throws an error with key 'misc-error
678 instead of 'system-error, since errno is not relevant.
680 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
681 specific keys instead of 'system-error. The latter is inappropriate
682 since errno will not have been set. The keys are:
683 'host-not-found, 'try-again, 'no-recovery and 'no-data.
685 ** sethostent, setnetent, setprotoent, setservent: now take an
686 optional argument STAYOPEN, which specifies whether the database
687 remains open after a database entry is accessed randomly (e.g., using
688 gethostbyname for the hosts database.) The default is #f. Previously
692 Changes since Guile 1.3.2:
694 * Changes to the stand-alone interpreter
698 An initial version of the Guile debugger written by Chris Hanson has
699 been added. The debugger is still under development but is included
700 in the distribution anyway since it is already quite useful.
706 after an error to enter the debugger. Type `help' inside the debugger
707 for a description of available commands.
709 If you prefer to have stack frames numbered and printed in
710 anti-chronological order and prefer up in the stack to be down on the
711 screen as is the case in gdb, you can put
713 (debug-enable 'backwards)
715 in your .guile startup file. (However, this means that Guile can't
716 use indentation to indicate stack level.)
718 The debugger is autoloaded into Guile at the first use.
720 ** Further enhancements to backtraces
722 There is a new debug option `width' which controls the maximum width
723 on the screen of printed stack frames. Fancy printing parameters
724 ("level" and "length" as in Common LISP) are adaptively adjusted for
725 each stack frame to give maximum information while still fitting
726 within the bounds. If the stack frame can't be made to fit by
727 adjusting parameters, it is simply cut off at the end. This is marked
730 ** Some modules are now only loaded when the repl is started
732 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
733 regex) are now loaded into (guile-user) only if the repl has been
734 started. The effect is that the startup time for scripts has been
735 reduced to 30% of what it was previously.
737 Correctly written scripts load the modules they require at the top of
738 the file and should not be affected by this change.
740 ** Hooks are now represented as smobs
742 * Changes to Scheme functions and syntax
744 ** Readline support has changed again.
746 The old (readline-activator) module is gone. Use (ice-9 readline)
747 instead, which now contains all readline functionality. So the code
748 to activate readline is now
750 (use-modules (ice-9 readline))
753 This should work at any time, including from the guile prompt.
755 To avoid confusion about the terms of Guile's license, please only
756 enable readline for your personal use; please don't make it the
757 default for others. Here is why we make this rather odd-sounding
760 Guile is normally licensed under a weakened form of the GNU General
761 Public License, which allows you to link code with Guile without
762 placing that code under the GPL. This exception is important to some
765 However, since readline is distributed under the GNU General Public
766 License, when you link Guile with readline, either statically or
767 dynamically, you effectively change Guile's license to the strict GPL.
768 Whenever you link any strictly GPL'd code into Guile, uses of Guile
769 which are normally permitted become forbidden. This is a rather
770 non-obvious consequence of the licensing terms.
772 So, to make sure things remain clear, please let people choose for
773 themselves whether to link GPL'd libraries like readline with Guile.
775 ** regexp-substitute/global has changed slightly, but incompatibly.
777 If you include a function in the item list, the string of the match
778 object it receives is the same string passed to
779 regexp-substitute/global, not some suffix of that string.
780 Correspondingly, the match's positions are relative to the entire
781 string, not the suffix.
783 If the regexp can match the empty string, the way matches are chosen
784 from the string has changed. regexp-substitute/global recognizes the
785 same set of matches that list-matches does; see below.
787 ** New function: list-matches REGEXP STRING [FLAGS]
789 Return a list of match objects, one for every non-overlapping, maximal
790 match of REGEXP in STRING. The matches appear in left-to-right order.
791 list-matches only reports matches of the empty string if there are no
792 other matches which begin on, end at, or include the empty match's
795 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
797 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
799 For each match of REGEXP in STRING, apply PROC to the match object,
800 and the last value PROC returned, or INIT for the first call. Return
801 the last value returned by PROC. We apply PROC to the matches as they
802 appear from left to right.
804 This function recognizes matches according to the same criteria as
807 Thus, you could define list-matches like this:
809 (define (list-matches regexp string . flags)
810 (reverse! (apply fold-matches regexp string '() cons flags)))
812 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
816 *** New function: hook? OBJ
818 Return #t if OBJ is a hook, otherwise #f.
820 *** New function: make-hook-with-name NAME [ARITY]
822 Return a hook with name NAME and arity ARITY. The default value for
823 ARITY is 0. The only effect of NAME is that it will appear when the
824 hook object is printed to ease debugging.
826 *** New function: hook-empty? HOOK
828 Return #t if HOOK doesn't contain any procedures, otherwise #f.
830 *** New function: hook->list HOOK
832 Return a list of the procedures that are called when run-hook is
835 ** `map' signals an error if its argument lists are not all the same length.
837 This is the behavior required by R5RS, so this change is really a bug
838 fix. But it seems to affect a lot of people's code, so we're
839 mentioning it here anyway.
841 ** Print-state handling has been made more transparent
843 Under certain circumstances, ports are represented as a port with an
844 associated print state. Earlier, this pair was represented as a pair
845 (see "Some magic has been added to the printer" below). It is now
846 indistinguishable (almost; see `get-print-state') from a port on the
849 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
851 Return a new port with the associated print state PRINT-STATE.
853 *** New function: get-print-state OUTPUT-PORT
855 Return the print state associated with this port if it exists,
858 *** New function: directory-stream? OBJECT
860 Returns true iff OBJECT is a directory stream --- the sort of object
861 returned by `opendir'.
863 ** New function: using-readline?
865 Return #t if readline is in use in the current repl.
867 ** structs will be removed in 1.4
869 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
870 and use GOOPS objects as the fundamental record type.
872 * Changes to the scm_ interface
874 ** structs will be removed in 1.4
876 The entire current struct interface (struct.c, struct.h) will be
877 replaced in Guile 1.4. We will merge GOOPS into libguile and use
878 GOOPS objects as the fundamental record type.
880 ** The internal representation of subr's has changed
882 Instead of giving a hint to the subr name, the CAR field of the subr
883 now contains an index to a subr entry in scm_subr_table.
885 *** New variable: scm_subr_table
887 An array of subr entries. A subr entry contains the name, properties
888 and documentation associated with the subr. The properties and
889 documentation slots are not yet used.
891 ** A new scheme for "forwarding" calls to a builtin to a generic function
893 It is now possible to extend the functionality of some Guile
894 primitives by letting them defer a call to a GOOPS generic function on
895 argument mismatch. This means that there is no loss of efficiency in
900 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
901 (define-method + ((x <string>) (y <string>))
904 + will still be as efficient as usual in numerical calculations, but
905 can also be used for concatenating strings.
907 Who will be the first one to extend Guile's numerical tower to
908 rationals? :) [OK, there a few other things to fix before this can
909 be made in a clean way.]
911 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
913 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
915 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
917 These do the same job as SCM_PROC and SCM_PROC1, but they also define
918 a variable GENERIC which can be used by the dispatch macros below.
920 [This is experimental code which may change soon.]
922 *** New macros for forwarding control to a generic on arg type error
924 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
926 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
928 These correspond to the scm_wta function call, and have the same
929 behaviour until the user has called the GOOPS primitive
930 `enable-primitive-generic!'. After that, these macros will apply the
931 generic function GENERIC to the argument(s) instead of calling
934 [This is experimental code which may change soon.]
936 *** New macros for argument testing with generic dispatch
938 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
940 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
942 These correspond to the SCM_ASSERT macro, but will defer control to
943 GENERIC on error after `enable-primitive-generic!' has been called.
945 [This is experimental code which may change soon.]
947 ** New function: SCM scm_eval_body (SCM body, SCM env)
949 Evaluates the body of a special form.
951 ** The internal representation of struct's has changed
953 Previously, four slots were allocated for the procedure(s) of entities
954 and operators. The motivation for this representation had to do with
955 the structure of the evaluator, the wish to support tail-recursive
956 generic functions, and efficiency. Since the generic function
957 dispatch mechanism has changed, there is no longer a need for such an
958 expensive representation, and the representation has been simplified.
960 This should not make any difference for most users.
962 ** GOOPS support has been cleaned up.
964 Some code has been moved from eval.c to objects.c and code in both of
965 these compilation units has been cleaned up and better structured.
967 *** New functions for applying generic functions
969 New function: SCM scm_apply_generic (GENERIC, ARGS)
970 New function: SCM scm_call_generic_0 (GENERIC)
971 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
972 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
973 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
975 ** Deprecated function: scm_make_named_hook
977 It is now replaced by:
979 ** New function: SCM scm_create_hook (const char *name, int arity)
981 Creates a hook in the same way as make-hook above but also
982 binds a variable named NAME to it.
984 This is the typical way of creating a hook from C code.
986 Currently, the variable is created in the "current" module.
987 This might change when we get the new module system.
989 [The behaviour is identical to scm_make_named_hook.]
993 Changes since Guile 1.3:
995 * Changes to mailing lists
997 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
999 See the README file to find current addresses for all the Guile
1002 * Changes to the distribution
1004 ** Readline support is no longer included with Guile by default.
1006 Based on the different license terms of Guile and Readline, we
1007 concluded that Guile should not *by default* cause the linking of
1008 Readline into an application program. Readline support is now offered
1009 as a separate module, which is linked into an application only when
1010 you explicitly specify it.
1012 Although Guile is GNU software, its distribution terms add a special
1013 exception to the usual GNU General Public License (GPL). Guile's
1014 license includes a clause that allows you to link Guile with non-free
1015 programs. We add this exception so as not to put Guile at a
1016 disadvantage vis-a-vis other extensibility packages that support other
1019 In contrast, the GNU Readline library is distributed under the GNU
1020 General Public License pure and simple. This means that you may not
1021 link Readline, even dynamically, into an application unless it is
1022 distributed under a free software license that is compatible the GPL.
1024 Because of this difference in distribution terms, an application that
1025 can use Guile may not be able to use Readline. Now users will be
1026 explicitly offered two independent decisions about the use of these
1029 You can activate the readline support by issuing
1031 (use-modules (readline-activator))
1034 from your ".guile" file, for example.
1036 * Changes to the stand-alone interpreter
1038 ** All builtins now print as primitives.
1039 Previously builtin procedures not belonging to the fundamental subr
1040 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
1041 Now, they print as #<primitive-procedure NAME>.
1043 ** Backtraces slightly more intelligible.
1044 gsubr-apply and macro transformer application frames no longer appear
1047 * Changes to Scheme functions and syntax
1049 ** Guile now correctly handles internal defines by rewriting them into
1050 their equivalent letrec. Previously, internal defines would
1051 incrementally add to the innermost environment, without checking
1052 whether the restrictions specified in RnRS were met. This lead to the
1053 correct behaviour when these restriction actually were met, but didn't
1054 catch all illegal uses. Such an illegal use could lead to crashes of
1055 the Guile interpreter or or other unwanted results. An example of
1056 incorrect internal defines that made Guile behave erratically:
1068 The problem with this example is that the definition of `c' uses the
1069 value of `b' directly. This confuses the meoization machine of Guile
1070 so that the second call of `b' (this time in a larger environment that
1071 also contains bindings for `c' and `d') refers to the binding of `c'
1072 instead of `a'. You could also make Guile crash with a variation on
1077 (define (b flag) (if flag a 1))
1078 (define c (1+ (b flag)))
1086 From now on, Guile will issue an `Unbound variable: b' error message
1091 A hook contains a list of functions which should be called on
1092 particular occasions in an existing program. Hooks are used for
1095 A window manager might have a hook before-window-map-hook. The window
1096 manager uses the function run-hooks to call all functions stored in
1097 before-window-map-hook each time a window is mapped. The user can
1098 store functions in the hook using add-hook!.
1100 In Guile, hooks are first class objects.
1102 *** New function: make-hook [N_ARGS]
1104 Return a hook for hook functions which can take N_ARGS arguments.
1105 The default value for N_ARGS is 0.
1107 (See also scm_make_named_hook below.)
1109 *** New function: add-hook! HOOK PROC [APPEND_P]
1111 Put PROC at the beginning of the list of functions stored in HOOK.
1112 If APPEND_P is supplied, and non-false, put PROC at the end instead.
1114 PROC must be able to take the number of arguments specified when the
1117 If PROC already exists in HOOK, then remove it first.
1119 *** New function: remove-hook! HOOK PROC
1121 Remove PROC from the list of functions in HOOK.
1123 *** New function: reset-hook! HOOK
1125 Clear the list of hook functions stored in HOOK.
1127 *** New function: run-hook HOOK ARG1 ...
1129 Run all hook functions stored in HOOK with arguments ARG1 ... .
1130 The number of arguments supplied must correspond to the number given
1131 when the hook was created.
1133 ** The function `dynamic-link' now takes optional keyword arguments.
1134 The only keyword argument that is currently defined is `:global
1135 BOOL'. With it, you can control whether the shared library will be
1136 linked in global mode or not. In global mode, the symbols from the
1137 linked library can be used to resolve references from other
1138 dynamically linked libraries. In non-global mode, the linked
1139 library is essentially invisible and can only be accessed via
1140 `dynamic-func', etc. The default is now to link in global mode.
1141 Previously, the default has been non-global mode.
1143 The `#:global' keyword is only effective on platforms that support
1144 the dlopen family of functions.
1146 ** New function `provided?'
1148 - Function: provided? FEATURE
1149 Return true iff FEATURE is supported by this installation of
1150 Guile. FEATURE must be a symbol naming a feature; the global
1151 variable `*features*' is a list of available features.
1153 ** Changes to the module (ice-9 expect):
1155 *** The expect-strings macro now matches `$' in a regular expression
1156 only at a line-break or end-of-file by default. Previously it would
1157 match the end of the string accumulated so far. The old behaviour
1158 can be obtained by setting the variable `expect-strings-exec-flags'
1161 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
1162 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
1163 in a regular expression will still match before a line-break or
1164 end-of-file. The default is `regexp/noteol'.
1166 *** The expect-strings macro now uses a variable
1167 `expect-strings-compile-flags' for the flags to be supplied to
1168 `make-regexp'. The default is `regexp/newline', which was previously
1171 *** The expect macro now supplies two arguments to a match procedure:
1172 the current accumulated string and a flag to indicate whether
1173 end-of-file has been reached. Previously only the string was supplied.
1174 If end-of-file is reached, the match procedure will be called an
1175 additional time with the same accumulated string as the previous call
1176 but with the flag set.
1178 ** New module (ice-9 format), implementing the Common Lisp `format' function.
1180 This code, and the documentation for it that appears here, was
1181 borrowed from SLIB, with minor adaptations for Guile.
1183 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
1184 An almost complete implementation of Common LISP format description
1185 according to the CL reference book `Common LISP' from Guy L.
1186 Steele, Digital Press. Backward compatible to most of the
1187 available Scheme format implementations.
1189 Returns `#t', `#f' or a string; has side effect of printing
1190 according to FORMAT-STRING. If DESTINATION is `#t', the output is
1191 to the current output port and `#t' is returned. If DESTINATION
1192 is `#f', a formatted string is returned as the result of the call.
1193 NEW: If DESTINATION is a string, DESTINATION is regarded as the
1194 format string; FORMAT-STRING is then the first argument and the
1195 output is returned as a string. If DESTINATION is a number, the
1196 output is to the current error port if available by the
1197 implementation. Otherwise DESTINATION must be an output port and
1200 FORMAT-STRING must be a string. In case of a formatting error
1201 format returns `#f' and prints a message on the current output or
1202 error port. Characters are output as if the string were output by
1203 the `display' function with the exception of those prefixed by a
1204 tilde (~). For a detailed description of the FORMAT-STRING syntax
1205 please consult a Common LISP format reference manual. For a test
1206 suite to verify this format implementation load `formatst.scm'.
1207 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
1209 Note: `format' is not reentrant, i.e. only one `format'-call may
1210 be executed at a time.
1213 *** Format Specification (Format version 3.0)
1215 Please consult a Common LISP format reference manual for a detailed
1216 description of the format string syntax. For a demonstration of the
1217 implemented directives see `formatst.scm'.
1219 This implementation supports directive parameters and modifiers (`:'
1220 and `@' characters). Multiple parameters must be separated by a comma
1221 (`,'). Parameters can be numerical parameters (positive or negative),
1222 character parameters (prefixed by a quote character (`''), variable
1223 parameters (`v'), number of rest arguments parameter (`#'), empty and
1224 default parameters. Directive characters are case independent. The
1225 general form of a directive is:
1227 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
1229 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
1231 *** Implemented CL Format Control Directives
1233 Documentation syntax: Uppercase characters represent the
1234 corresponding control directive characters. Lowercase characters
1235 represent control directive parameter descriptions.
1238 Any (print as `display' does).
1242 `~MINCOL,COLINC,MINPAD,PADCHARA'
1246 S-expression (print as `write' does).
1250 `~MINCOL,COLINC,MINPAD,PADCHARS'
1256 print number sign always.
1259 print comma separated.
1261 `~MINCOL,PADCHAR,COMMACHARD'
1267 print number sign always.
1270 print comma separated.
1272 `~MINCOL,PADCHAR,COMMACHARX'
1278 print number sign always.
1281 print comma separated.
1283 `~MINCOL,PADCHAR,COMMACHARO'
1289 print number sign always.
1292 print comma separated.
1294 `~MINCOL,PADCHAR,COMMACHARB'
1299 `~N,MINCOL,PADCHAR,COMMACHARR'
1303 print a number as a Roman numeral.
1306 print a number as an "old fashioned" Roman numeral.
1309 print a number as an ordinal English number.
1312 print a number as a cardinal English number.
1317 prints `y' and `ies'.
1320 as `~P but jumps 1 argument backward.'
1323 as `~@P but jumps 1 argument backward.'
1328 prints a character as the reader can understand it (i.e. `#\'
1332 prints a character as emacs does (eg. `^C' for ASCII 03).
1335 Fixed-format floating-point (prints a flonum like MMM.NNN).
1336 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
1338 If the number is positive a plus sign is printed.
1341 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
1342 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
1344 If the number is positive a plus sign is printed.
1347 General floating-point (prints a flonum either fixed or
1349 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
1351 If the number is positive a plus sign is printed.
1354 Dollars floating-point (prints a flonum in fixed with signs
1356 `~DIGITS,SCALE,WIDTH,PADCHAR$'
1358 If the number is positive a plus sign is printed.
1361 A sign is always printed and appears before the padding.
1364 The sign appears before the padding.
1372 print newline if not at the beginning of the output line.
1374 prints `~&' and then N-1 newlines.
1379 print N page separators.
1389 newline is ignored, white space left.
1392 newline is left, white space ignored.
1397 relative tabulation.
1403 Indirection (expects indirect arguments as a list).
1405 extracts indirect arguments from format arguments.
1408 Case conversion (converts by `string-downcase').
1410 converts by `string-capitalize'.
1413 converts by `string-capitalize-first'.
1416 converts by `string-upcase'.
1419 Argument Jumping (jumps 1 argument forward).
1421 jumps N arguments forward.
1424 jumps 1 argument backward.
1427 jumps N arguments backward.
1430 jumps to the 0th argument.
1433 jumps to the Nth argument (beginning from 0)
1435 `~[STR0~;STR1~;...~;STRN~]'
1436 Conditional Expression (numerical clause conditional).
1438 take argument from N.
1441 true test conditional.
1444 if-else-then conditional.
1450 default clause follows.
1453 Iteration (args come from the next argument (a list)).
1455 at most N iterations.
1458 args from next arg (a list of lists).
1461 args from the rest of arguments.
1464 args from the rest args (lists).
1475 aborts if N <= M <= K
1477 *** Not Implemented CL Format Control Directives
1480 print `#f' as an empty list (see below).
1483 print `#f' as an empty list (see below).
1489 (sorry I don't understand its semantics completely)
1491 *** Extended, Replaced and Additional Control Directives
1493 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
1494 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
1495 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
1496 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
1497 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
1498 COMMAWIDTH is the number of characters between two comma
1502 print a R4RS complex number as `~F~@Fi' with passed parameters for
1506 Pretty print formatting of an argument for scheme code lists.
1512 Flushes the output if format DESTINATION is a port.
1515 Print a `#\space' character
1517 print N `#\space' characters.
1520 Print a `#\tab' character
1522 print N `#\tab' characters.
1525 Takes N as an integer representation for a character. No arguments
1526 are consumed. N is converted to a character by `integer->char'. N
1527 must be a positive decimal number.
1530 Print out readproof. Prints out internal objects represented as
1531 `#<...>' as strings `"#<...>"' so that the format output can always
1532 be processed by `read'.
1535 Print out readproof. Prints out internal objects represented as
1536 `#<...>' as strings `"#<...>"' so that the format output can always
1537 be processed by `read'.
1540 Prints information and a copyright notice on the format
1543 prints format version.
1546 may also print number strings, i.e. passing a number as a string
1547 and format it accordingly.
1549 *** Configuration Variables
1551 The format module exports some configuration variables to suit the
1552 systems and users needs. There should be no modification necessary for
1553 the configuration that comes with Guile. Format detects automatically
1554 if the running scheme system implements floating point numbers and
1557 format:symbol-case-conv
1558 Symbols are converted by `symbol->string' so the case type of the
1559 printed symbols is implementation dependent.
1560 `format:symbol-case-conv' is a one arg closure which is either
1561 `#f' (no conversion), `string-upcase', `string-downcase' or
1562 `string-capitalize'. (default `#f')
1564 format:iobj-case-conv
1565 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
1566 implementation internal objects. (default `#f')
1569 The character prefixing the exponent value in `~E' printing.
1572 *** Compatibility With Other Format Implementations
1578 Downward compatible except for padding support and `~A', `~S',
1579 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
1580 `printf' padding support which is completely replaced by the CL
1581 `format' padding style.
1584 Downward compatible except for `~', which is not documented
1585 (ignores all characters inside the format string up to a newline
1586 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
1587 numerical and variable parameters and `:/@' modifiers in the CL
1591 Downward compatible except for `~A' and `~S' which print in
1592 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
1593 directive parameters or modifiers)).
1596 Downward compatible except for an optional destination parameter:
1597 S2C accepts a format call without a destination which returns a
1598 formatted string. This is equivalent to a #f destination in S2C.
1599 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
1600 parameters or modifiers)).
1603 ** Changes to string-handling functions.
1605 These functions were added to support the (ice-9 format) module, above.
1607 *** New function: string-upcase STRING
1608 *** New function: string-downcase STRING
1610 These are non-destructive versions of the existing string-upcase! and
1611 string-downcase! functions.
1613 *** New function: string-capitalize! STRING
1614 *** New function: string-capitalize STRING
1616 These functions convert the first letter of each word in the string to
1619 (string-capitalize "howdy there")
1622 As with the other functions, string-capitalize! modifies the string in
1623 place, while string-capitalize returns a modified copy of its argument.
1625 *** New function: string-ci->symbol STRING
1627 Return a symbol whose name is STRING, but having the same case as if
1628 the symbol had be read by `read'.
1630 Guile can be configured to be sensitive or insensitive to case
1631 differences in Scheme identifiers. If Guile is case-insensitive, all
1632 symbols are converted to lower case on input. The `string-ci->symbol'
1633 function returns a symbol whose name in STRING, transformed as Guile
1634 would if STRING were input.
1636 *** New function: substring-move! STRING1 START END STRING2 START
1638 Copy the substring of STRING1 from START (inclusive) to END
1639 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
1640 string, and the source and destination areas may overlap; in all
1641 cases, the function behaves as if all the characters were copied
1644 *** Extended functions: substring-move-left! substring-move-right!
1646 These functions now correctly copy arbitrarily overlapping substrings;
1647 they are both synonyms for substring-move!.
1650 ** New module (ice-9 getopt-long), with the function `getopt-long'.
1652 getopt-long is a function for parsing command-line arguments in a
1653 manner consistent with other GNU programs.
1655 (getopt-long ARGS GRAMMAR)
1656 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
1658 ARGS should be a list of strings. Its first element should be the
1659 name of the program; subsequent elements should be the arguments
1660 that were passed to the program on the command line. The
1661 `program-arguments' procedure returns a list of this form.
1663 GRAMMAR is a list of the form:
1664 ((OPTION (PROPERTY VALUE) ...) ...)
1666 Each OPTION should be a symbol. `getopt-long' will accept a
1667 command-line option named `--OPTION'.
1668 Each option can have the following (PROPERTY VALUE) pairs:
1670 (single-char CHAR) --- Accept `-CHAR' as a single-character
1671 equivalent to `--OPTION'. This is how to specify traditional
1673 (required? BOOL) --- If BOOL is true, the option is required.
1674 getopt-long will raise an error if it is not found in ARGS.
1675 (value BOOL) --- If BOOL is #t, the option accepts a value; if
1676 it is #f, it does not; and if it is the symbol
1677 `optional', the option may appear in ARGS with or
1679 (predicate FUNC) --- If the option accepts a value (i.e. you
1680 specified `(value #t)' for this option), then getopt
1681 will apply FUNC to the value, and throw an exception
1682 if it returns #f. FUNC should be a procedure which
1683 accepts a string and returns a boolean value; you may
1684 need to use quasiquotes to get it into GRAMMAR.
1686 The (PROPERTY VALUE) pairs may occur in any order, but each
1687 property may occur only once. By default, options do not have
1688 single-character equivalents, are not required, and do not take
1691 In ARGS, single-character options may be combined, in the usual
1692 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
1693 accepts values, then it must be the last option in the
1694 combination; the value is the next argument. So, for example, using
1695 the following grammar:
1696 ((apples (single-char #\a))
1697 (blimps (single-char #\b) (value #t))
1698 (catalexis (single-char #\c) (value #t)))
1699 the following argument lists would be acceptable:
1700 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
1701 for "blimps" and "catalexis")
1702 ("-ab" "bang" "-c" "couth") (same)
1703 ("-ac" "couth" "-b" "bang") (same)
1704 ("-abc" "couth" "bang") (an error, since `-b' is not the
1705 last option in its combination)
1707 If an option's value is optional, then `getopt-long' decides
1708 whether it has a value by looking at what follows it in ARGS. If
1709 the next element is a string, and it does not appear to be an
1710 option itself, then that string is the option's value.
1712 The value of a long option can appear as the next element in ARGS,
1713 or it can follow the option name, separated by an `=' character.
1714 Thus, using the same grammar as above, the following argument lists
1716 ("--apples" "Braeburn" "--blimps" "Goodyear")
1717 ("--apples=Braeburn" "--blimps" "Goodyear")
1718 ("--blimps" "Goodyear" "--apples=Braeburn")
1720 If the option "--" appears in ARGS, argument parsing stops there;
1721 subsequent arguments are returned as ordinary arguments, even if
1722 they resemble options. So, in the argument list:
1723 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
1724 `getopt-long' will recognize the `apples' option as having the
1725 value "Granny Smith", but it will not recognize the `blimp'
1726 option; it will return the strings "--blimp" and "Goodyear" as
1727 ordinary argument strings.
1729 The `getopt-long' function returns the parsed argument list as an
1730 assocation list, mapping option names --- the symbols from GRAMMAR
1731 --- onto their values, or #t if the option does not accept a value.
1732 Unused options do not appear in the alist.
1734 All arguments that are not the value of any option are returned
1735 as a list, associated with the empty list.
1737 `getopt-long' throws an exception if:
1738 - it finds an unrecognized option in ARGS
1739 - a required option is omitted
1740 - an option that requires an argument doesn't get one
1741 - an option that doesn't accept an argument does get one (this can
1742 only happen using the long option `--opt=value' syntax)
1743 - an option predicate fails
1748 `((lockfile-dir (required? #t)
1751 (predicate ,file-is-directory?))
1752 (verbose (required? #f)
1755 (x-includes (single-char #\x))
1756 (rnet-server (single-char #\y)
1757 (predicate ,string?))))
1759 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
1760 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
1762 => ((() "foo1" "-fred" "foo2" "foo3")
1763 (rnet-server . "lamprod")
1764 (x-includes . "/usr/include")
1765 (lockfile-dir . "/tmp")
1768 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
1770 It will be removed in a few releases.
1772 ** New syntax: lambda*
1773 ** New syntax: define*
1774 ** New syntax: define*-public
1775 ** New syntax: defmacro*
1776 ** New syntax: defmacro*-public
1777 Guile now supports optional arguments.
1779 `lambda*', `define*', `define*-public', `defmacro*' and
1780 `defmacro*-public' are identical to the non-* versions except that
1781 they use an extended type of parameter list that has the following BNF
1782 syntax (parentheses are literal, square brackets indicate grouping,
1783 and `*', `+' and `?' have the usual meaning):
1785 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
1786 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
1787 [[#&rest identifier]|[. identifier]]? ) | [identifier]
1789 ext-var-decl ::= identifier | ( identifier expression )
1791 The semantics are best illustrated with the following documentation
1792 and examples for `lambda*':
1795 lambda extended for optional and keyword arguments
1797 lambda* creates a procedure that takes optional arguments. These
1798 are specified by putting them inside brackets at the end of the
1799 paramater list, but before any dotted rest argument. For example,
1800 (lambda* (a b #&optional c d . e) '())
1801 creates a procedure with fixed arguments a and b, optional arguments c
1802 and d, and rest argument e. If the optional arguments are omitted
1803 in a call, the variables for them are unbound in the procedure. This
1804 can be checked with the bound? macro.
1806 lambda* can also take keyword arguments. For example, a procedure
1808 (lambda* (#&key xyzzy larch) '())
1809 can be called with any of the argument lists (#:xyzzy 11)
1810 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
1811 are given as keywords are bound to values.
1813 Optional and keyword arguments can also be given default values
1814 which they take on when they are not present in a call, by giving a
1815 two-item list in place of an optional argument, for example in:
1816 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
1817 foo is a fixed argument, bar is an optional argument with default
1818 value 42, and baz is a keyword argument with default value 73.
1819 Default value expressions are not evaluated unless they are needed
1820 and until the procedure is called.
1822 lambda* now supports two more special parameter list keywords.
1824 lambda*-defined procedures now throw an error by default if a
1825 keyword other than one of those specified is found in the actual
1826 passed arguments. However, specifying #&allow-other-keys
1827 immediately after the kyword argument declarations restores the
1828 previous behavior of ignoring unknown keywords. lambda* also now
1829 guarantees that if the same keyword is passed more than once, the
1830 last one passed is the one that takes effect. For example,
1831 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
1832 #:heads 37 #:tails 42 #:heads 99)
1833 would result in (99 47) being displayed.
1835 #&rest is also now provided as a synonym for the dotted syntax rest
1836 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
1837 all respects to lambda*. This is provided for more similarity to DSSSL,
1838 MIT-Scheme and Kawa among others, as well as for refugees from other
1841 Further documentation may be found in the optargs.scm file itself.
1843 The optional argument module also exports the macros `let-optional',
1844 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
1845 are not documented here because they may be removed in the future, but
1846 full documentation is still available in optargs.scm.
1848 ** New syntax: and-let*
1849 Guile now supports the `and-let*' form, described in the draft SRFI-2.
1851 Syntax: (land* (<clause> ...) <body> ...)
1852 Each <clause> should have one of the following forms:
1853 (<variable> <expression>)
1856 Each <variable> or <bound-variable> should be an identifier. Each
1857 <expression> should be a valid expression. The <body> should be a
1858 possibly empty sequence of expressions, like the <body> of a
1861 Semantics: A LAND* expression is evaluated by evaluating the
1862 <expression> or <bound-variable> of each of the <clause>s from
1863 left to right. The value of the first <expression> or
1864 <bound-variable> that evaluates to a false value is returned; the
1865 remaining <expression>s and <bound-variable>s are not evaluated.
1866 The <body> forms are evaluated iff all the <expression>s and
1867 <bound-variable>s evaluate to true values.
1869 The <expression>s and the <body> are evaluated in an environment
1870 binding each <variable> of the preceding (<variable> <expression>)
1871 clauses to the value of the <expression>. Later bindings
1872 shadow earlier bindings.
1874 Guile's and-let* macro was contributed by Michael Livshin.
1876 ** New sorting functions
1878 *** New function: sorted? SEQUENCE LESS?
1879 Returns `#t' when the sequence argument is in non-decreasing order
1880 according to LESS? (that is, there is no adjacent pair `... x y
1881 ...' for which `(less? y x)').
1883 Returns `#f' when the sequence contains at least one out-of-order
1884 pair. It is an error if the sequence is neither a list nor a
1887 *** New function: merge LIST1 LIST2 LESS?
1888 LIST1 and LIST2 are sorted lists.
1889 Returns the sorted list of all elements in LIST1 and LIST2.
1891 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
1892 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
1893 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
1894 (Here "<" should read "comes before".)
1896 *** New procedure: merge! LIST1 LIST2 LESS?
1897 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
1898 the result. If the code is compiled, and LESS? constructs no new
1899 pairs, no pairs at all will be allocated. The first pair of the
1900 result will be either the first pair of LIST1 or the first pair of
1903 *** New function: sort SEQUENCE LESS?
1904 Accepts either a list or a vector, and returns a new sequence
1905 which is sorted. The new sequence is the same type as the input.
1906 Always `(sorted? (sort sequence less?) less?)'. The original
1907 sequence is not altered in any way. The new sequence shares its
1908 elements with the old one; no elements are copied.
1910 *** New procedure: sort! SEQUENCE LESS
1911 Returns its sorted result in the original boxes. No new storage is
1912 allocated at all. Proper usage: (set! slist (sort! slist <))
1914 *** New function: stable-sort SEQUENCE LESS?
1915 Similar to `sort' but stable. That is, if "equal" elements are
1916 ordered a < b in the original sequence, they will have the same order
1919 *** New function: stable-sort! SEQUENCE LESS?
1920 Similar to `sort!' but stable.
1921 Uses temporary storage when sorting vectors.
1923 *** New functions: sort-list, sort-list!
1924 Added for compatibility with scsh.
1926 ** New built-in random number support
1928 *** New function: random N [STATE]
1929 Accepts a positive integer or real N and returns a number of the
1930 same type between zero (inclusive) and N (exclusive). The values
1931 returned have a uniform distribution.
1933 The optional argument STATE must be of the type produced by
1934 `copy-random-state' or `seed->random-state'. It defaults to the value
1935 of the variable `*random-state*'. This object is used to maintain the
1936 state of the pseudo-random-number generator and is altered as a side
1937 effect of the `random' operation.
1939 *** New variable: *random-state*
1940 Holds a data structure that encodes the internal state of the
1941 random-number generator that `random' uses by default. The nature
1942 of this data structure is implementation-dependent. It may be
1943 printed out and successfully read back in, but may or may not
1944 function correctly as a random-number state object in another
1947 *** New function: copy-random-state [STATE]
1948 Returns a new object of type suitable for use as the value of the
1949 variable `*random-state*' and as a second argument to `random'.
1950 If argument STATE is given, a copy of it is returned. Otherwise a
1951 copy of `*random-state*' is returned.
1953 *** New function: seed->random-state SEED
1954 Returns a new object of type suitable for use as the value of the
1955 variable `*random-state*' and as a second argument to `random'.
1956 SEED is a string or a number. A new state is generated and
1957 initialized using SEED.
1959 *** New function: random:uniform [STATE]
1960 Returns an uniformly distributed inexact real random number in the
1961 range between 0 and 1.
1963 *** New procedure: random:solid-sphere! VECT [STATE]
1964 Fills VECT with inexact real random numbers the sum of whose
1965 squares is less than 1.0. Thinking of VECT as coordinates in
1966 space of dimension N = `(vector-length VECT)', the coordinates are
1967 uniformly distributed within the unit N-shere. The sum of the
1968 squares of the numbers is returned. VECT can be either a vector
1969 or a uniform vector of doubles.
1971 *** New procedure: random:hollow-sphere! VECT [STATE]
1972 Fills VECT with inexact real random numbers the sum of whose squares
1973 is equal to 1.0. Thinking of VECT as coordinates in space of
1974 dimension n = `(vector-length VECT)', the coordinates are uniformly
1975 distributed over the surface of the unit n-shere. VECT can be either
1976 a vector or a uniform vector of doubles.
1978 *** New function: random:normal [STATE]
1979 Returns an inexact real in a normal distribution with mean 0 and
1980 standard deviation 1. For a normal distribution with mean M and
1981 standard deviation D use `(+ M (* D (random:normal)))'.
1983 *** New procedure: random:normal-vector! VECT [STATE]
1984 Fills VECT with inexact real random numbers which are independent and
1985 standard normally distributed (i.e., with mean 0 and variance 1).
1986 VECT can be either a vector or a uniform vector of doubles.
1988 *** New function: random:exp STATE
1989 Returns an inexact real in an exponential distribution with mean 1.
1990 For an exponential distribution with mean U use (* U (random:exp)).
1992 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
1994 These functions now operate on numbers in the range of a C unsigned
1997 These functions used to operate on numbers in the range of a C signed
1998 long; however, this seems inappropriate, because Guile integers don't
2001 ** New function: make-guardian
2002 This is an implementation of guardians as described in
2003 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
2004 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
2005 Programming Language Design and Implementation, June 1993
2006 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
2008 ** New functions: delq1!, delv1!, delete1!
2009 These procedures behave similar to delq! and friends but delete only
2010 one object if at all.
2012 ** New function: unread-string STRING PORT
2013 Unread STRING to PORT, that is, push it back onto the port so that
2014 next read operation will work on the pushed back characters.
2016 ** unread-char can now be called multiple times
2017 If unread-char is called multiple times, the unread characters will be
2018 read again in last-in first-out order.
2020 ** the procedures uniform-array-read! and uniform-array-write! now
2021 work on any kind of port, not just ports which are open on a file.
2023 ** Now 'l' in a port mode requests line buffering.
2025 ** The procedure truncate-file now works on string ports as well
2026 as file ports. If the size argument is omitted, the current
2027 file position is used.
2029 ** new procedure: seek PORT/FDES OFFSET WHENCE
2030 The arguments are the same as for the old fseek procedure, but it
2031 works on string ports as well as random-access file ports.
2033 ** the fseek procedure now works on string ports, since it has been
2034 redefined using seek.
2036 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
2037 size is not supplied.
2039 ** the newline procedure no longer flushes the port if it's not
2040 line-buffered: previously it did if it was the current output port.
2042 ** open-pipe and close-pipe are no longer primitive procedures, but
2043 an emulation can be obtained using `(use-modules (ice-9 popen))'.
2045 ** the freopen procedure has been removed.
2047 ** new procedure: drain-input PORT
2048 Drains PORT's read buffers (including any pushed-back characters)
2049 and returns the contents as a single string.
2051 ** New function: map-in-order PROC LIST1 LIST2 ...
2052 Version of `map' which guarantees that the procedure is applied to the
2053 lists in serial order.
2055 ** Renamed `serial-array-copy!' and `serial-array-map!' to
2056 `array-copy-in-order!' and `array-map-in-order!'. The old names are
2057 now obsolete and will go away in release 1.5.
2059 ** New syntax: collect BODY1 ...
2060 Version of `begin' which returns a list of the results of the body
2061 forms instead of the result of the last body form. In contrast to
2062 `begin', `collect' allows an empty body.
2064 ** New functions: read-history FILENAME, write-history FILENAME
2065 Read/write command line history from/to file. Returns #t on success
2066 and #f if an error occured.
2068 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
2070 These procedures return a list of definitions available in the specified
2071 argument, a relative module reference. In the case of no argument,
2072 `(current-module)' is now consulted for definitions to return, instead
2073 of simply returning #f, the former behavior.
2075 ** The #/ syntax for lists is no longer supported.
2077 Earlier versions of Scheme accepted this syntax, but printed a
2080 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
2082 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
2085 * Changes to the gh_ interface
2089 Now takes a second argument which is the result array. If this
2090 pointer is NULL, a new array is malloced (the old behaviour).
2092 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
2093 gh_scm2shorts, gh_scm2longs, gh_scm2floats
2097 * Changes to the scm_ interface
2099 ** Function: scm_make_named_hook (char* name, int n_args)
2101 Creates a hook in the same way as make-hook above but also
2102 binds a variable named NAME to it.
2104 This is the typical way of creating a hook from C code.
2106 Currently, the variable is created in the "current" module. This
2107 might change when we get the new module system.
2109 ** The smob interface
2111 The interface for creating smobs has changed. For documentation, see
2112 data-rep.info (made from guile-core/doc/data-rep.texi).
2114 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
2116 >>> This function will be removed in 1.3.4. <<<
2120 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
2121 This function adds a new smob type, named NAME, with instance size
2122 SIZE to the system. The return value is a tag that is used in
2123 creating instances of the type. If SIZE is 0, then no memory will
2124 be allocated when instances of the smob are created, and nothing
2125 will be freed by the default free function.
2127 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
2128 This function sets the smob marking procedure for the smob type
2129 specified by the tag TC. TC is the tag returned by
2130 `scm_make_smob_type'.
2132 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
2133 This function sets the smob freeing procedure for the smob type
2134 specified by the tag TC. TC is the tag returned by
2135 `scm_make_smob_type'.
2137 *** Function: void scm_set_smob_print (tc, print)
2139 - Function: void scm_set_smob_print (long tc,
2140 scm_sizet (*print) (SCM,
2144 This function sets the smob printing procedure for the smob type
2145 specified by the tag TC. TC is the tag returned by
2146 `scm_make_smob_type'.
2148 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
2149 This function sets the smob equality-testing predicate for the
2150 smob type specified by the tag TC. TC is the tag returned by
2151 `scm_make_smob_type'.
2153 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
2154 Make VALUE contain a smob instance of the type with type code TC and
2155 smob data DATA. VALUE must be previously declared as C type `SCM'.
2157 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
2158 This macro expands to a block of code that creates a smob instance
2159 of the type with type code TC and smob data DATA, and returns that
2160 `SCM' value. It should be the last piece of code in a block.
2162 ** The interfaces for using I/O ports and implementing port types
2163 (ptobs) have changed significantly. The new interface is based on
2164 shared access to buffers and a new set of ptob procedures.
2166 *** scm_newptob has been removed
2170 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
2172 - Function: SCM scm_make_port_type (char *type_name,
2173 int (*fill_buffer) (SCM port),
2174 void (*write_flush) (SCM port));
2176 Similarly to the new smob interface, there is a set of function
2177 setters by which the user can customize the behaviour of his port
2178 type. See ports.h (scm_set_port_XXX).
2180 ** scm_strport_to_string: New function: creates a new string from
2181 a string port's buffer.
2183 ** Plug in interface for random number generators
2184 The variable `scm_the_rng' in random.c contains a value and three
2185 function pointers which together define the current random number
2186 generator being used by the Scheme level interface and the random
2187 number library functions.
2189 The user is free to replace the default generator with the generator
2192 *** Variable: size_t scm_the_rng.rstate_size
2193 The size of the random state type used by the current RNG
2196 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
2197 Given the random STATE, return 32 random bits.
2199 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
2200 Seed random state STATE using string S of length N.
2202 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
2203 Given random state STATE, return a malloced copy.
2206 The default RNG is the MWC (Multiply With Carry) random number
2207 generator described by George Marsaglia at the Department of
2208 Statistics and Supercomputer Computations Research Institute, The
2209 Florida State University (http://stat.fsu.edu/~geo).
2211 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
2212 passes all tests in the DIEHARD test suite
2213 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
2214 costs one multiply and one add on platforms which either supports long
2215 longs (gcc does this on most systems) or have 64 bit longs. The cost
2216 is four multiply on other systems but this can be optimized by writing
2217 scm_i_uniform32 in assembler.
2219 These functions are provided through the scm_the_rng interface for use
2220 by libguile and the application.
2222 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
2223 Given the random STATE, return 32 random bits.
2224 Don't use this function directly. Instead go through the plugin
2225 interface (see "Plug in interface" above).
2227 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
2228 Initialize STATE using SEED of length N.
2230 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
2231 Return a malloc:ed copy of STATE. This function can easily be re-used
2232 in the interfaces to other RNGs.
2234 ** Random number library functions
2235 These functions use the current RNG through the scm_the_rng interface.
2236 It might be a good idea to use these functions from your C code so
2237 that only one random generator is used by all code in your program.
2239 The default random state is stored in:
2241 *** Variable: SCM scm_var_random_state
2242 Contains the vcell of the Scheme variable "*random-state*" which is
2243 used as default state by all random number functions in the Scheme
2248 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
2250 *** Function: scm_rstate *scm_c_default_rstate (void)
2251 This is a convenience function which returns the value of
2252 scm_var_random_state. An error message is generated if this value
2253 isn't a random state.
2255 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
2256 Make a new random state from the string SEED of length LENGTH.
2258 It is generally not a good idea to use multiple random states in a
2259 program. While subsequent random numbers generated from one random
2260 state are guaranteed to be reasonably independent, there is no such
2261 guarantee for numbers generated from different random states.
2263 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
2264 Return 32 random bits.
2266 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
2267 Return a sample from the uniform(0,1) distribution.
2269 *** Function: double scm_c_normal01 (scm_rstate *STATE)
2270 Return a sample from the normal(0,1) distribution.
2272 *** Function: double scm_c_exp1 (scm_rstate *STATE)
2273 Return a sample from the exp(1) distribution.
2275 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
2276 Return a sample from the discrete uniform(0,M) distribution.
2278 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
2279 Return a sample from the discrete uniform(0,M) distribution.
2280 M must be a bignum object. The returned value may be an INUM.
2284 Changes in Guile 1.3 (released Monday, October 19, 1998):
2286 * Changes to the distribution
2288 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
2289 To avoid conflicts, programs should name environment variables after
2290 themselves, except when there's a common practice establishing some
2293 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
2294 giving the former precedence, and printing a warning message if the
2295 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
2297 ** The header files related to multi-byte characters have been removed.
2298 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
2299 which referred to these explicitly will probably need to be rewritten,
2300 since the support for the variant string types has been removed; see
2303 ** The header files append.h and sequences.h have been removed. These
2304 files implemented non-R4RS operations which would encourage
2305 non-portable programming style and less easy-to-read code.
2307 * Changes to the stand-alone interpreter
2309 ** New procedures have been added to implement a "batch mode":
2311 *** Function: batch-mode?
2313 Returns a boolean indicating whether the interpreter is in batch
2316 *** Function: set-batch-mode?! ARG
2318 If ARG is true, switches the interpreter to batch mode. The `#f'
2319 case has not been implemented.
2321 ** Guile now provides full command-line editing, when run interactively.
2322 To use this feature, you must have the readline library installed.
2323 The Guile build process will notice it, and automatically include
2326 The readline library is available via anonymous FTP from any GNU
2327 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
2329 ** the-last-stack is now a fluid.
2331 * Changes to the procedure for linking libguile with your programs
2333 ** You can now use the `guile-config' utility to build programs that use Guile.
2335 Guile now includes a command-line utility called `guile-config', which
2336 can provide information about how to compile and link programs that
2339 *** `guile-config compile' prints any C compiler flags needed to use Guile.
2340 You should include this command's output on the command line you use
2341 to compile C or C++ code that #includes the Guile header files. It's
2342 usually just a `-I' flag to help the compiler find the Guile headers.
2345 *** `guile-config link' prints any linker flags necessary to link with Guile.
2347 This command writes to its standard output a list of flags which you
2348 must pass to the linker to link your code against the Guile library.
2349 The flags include '-lguile' itself, any other libraries the Guile
2350 library depends upon, and any `-L' flags needed to help the linker
2351 find those libraries.
2353 For example, here is a Makefile rule that builds a program named 'foo'
2354 from the object files ${FOO_OBJECTS}, and links them against Guile:
2357 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
2359 Previous Guile releases recommended that you use autoconf to detect
2360 which of a predefined set of libraries were present on your system.
2361 It is more robust to use `guile-config', since it records exactly which
2362 libraries the installed Guile library requires.
2364 This was originally called `build-guile', but was renamed to
2365 `guile-config' before Guile 1.3 was released, to be consistent with
2366 the analogous script for the GTK+ GUI toolkit, which is called
2370 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
2372 If you are using the GNU autoconf package to configure your program,
2373 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
2374 (described above) and gather the necessary values for use in your
2377 The GUILE_FLAGS macro expands to configure script code which runs the
2378 `guile-config' script, to find out where Guile's header files and
2379 libraries are installed. It sets two variables, marked for
2380 substitution, as by AC_SUBST.
2382 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
2383 code that uses Guile header files. This is almost always just a
2386 GUILE_LDFLAGS --- flags to pass to the linker to link a
2387 program against Guile. This includes `-lguile' for the Guile
2388 library itself, any libraries that Guile itself requires (like
2389 -lqthreads), and so on. It may also include a -L flag to tell the
2390 compiler where to find the libraries.
2392 GUILE_FLAGS is defined in the file guile.m4, in the top-level
2393 directory of the Guile distribution. You can copy it into your
2394 package's aclocal.m4 file, and then use it in your configure.in file.
2396 If you are using the `aclocal' program, distributed with GNU automake,
2397 to maintain your aclocal.m4 file, the Guile installation process
2398 installs guile.m4 where aclocal will find it. All you need to do is
2399 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
2400 this will copy the definition of GUILE_FLAGS into your aclocal.m4
2404 * Changes to Scheme functions and syntax
2406 ** Multi-byte strings have been removed, as have multi-byte and wide
2407 ports. We felt that these were the wrong approach to
2408 internationalization support.
2410 ** New function: readline [PROMPT]
2411 Read a line from the terminal, and allow the user to edit it,
2412 prompting with PROMPT. READLINE provides a large set of Emacs-like
2413 editing commands, lets the user recall previously typed lines, and
2414 works on almost every kind of terminal, including dumb terminals.
2416 READLINE assumes that the cursor is at the beginning of the line when
2417 it is invoked. Thus, you can't print a prompt yourself, and then call
2418 READLINE; you need to package up your prompt as a string, pass it to
2419 the function, and let READLINE print the prompt itself. This is
2420 because READLINE needs to know the prompt's screen width.
2422 For Guile to provide this function, you must have the readline
2423 library, version 2.1 or later, installed on your system. Readline is
2424 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
2425 any GNU mirror site.
2427 See also ADD-HISTORY function.
2429 ** New function: add-history STRING
2430 Add STRING as the most recent line in the history used by the READLINE
2431 command. READLINE does not add lines to the history itself; you must
2432 call ADD-HISTORY to make previous input available to the user.
2434 ** The behavior of the read-line function has changed.
2436 This function now uses standard C library functions to read the line,
2437 for speed. This means that it doesn not respect the value of
2438 scm-line-incrementors; it assumes that lines are delimited with
2441 (Note that this is read-line, the function that reads a line of text
2442 from a port, not readline, the function that reads a line from a
2443 terminal, providing full editing capabilities.)
2445 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
2447 This module provides some simple argument parsing. It exports one
2450 Function: getopt-gnu-style ARG-LS
2451 Parse a list of program arguments into an alist of option
2454 Each item in the list of program arguments is examined to see if
2455 it meets the syntax of a GNU long-named option. An argument like
2456 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
2457 returned alist, where MUMBLE is a keyword object with the same
2458 name as the argument. An argument like `--MUMBLE=FROB' produces
2459 an element of the form (MUMBLE . FROB), where FROB is a string.
2461 As a special case, the returned alist also contains a pair whose
2462 car is the symbol `rest'. The cdr of this pair is a list
2463 containing all the items in the argument list that are not options
2464 of the form mentioned above.
2466 The argument `--' is treated specially: all items in the argument
2467 list appearing after such an argument are not examined, and are
2468 returned in the special `rest' list.
2470 This function does not parse normal single-character switches.
2471 You will need to parse them out of the `rest' list yourself.
2473 ** The read syntax for byte vectors and short vectors has changed.
2475 Instead of #bytes(...), write #y(...).
2477 Instead of #short(...), write #h(...).
2479 This may seem nutty, but, like the other uniform vectors, byte vectors
2480 and short vectors want to have the same print and read syntax (and,
2481 more basic, want to have read syntax!). Changing the read syntax to
2482 use multiple characters after the hash sign breaks with the
2483 conventions used in R5RS and the conventions used for the other
2484 uniform vectors. It also introduces complexity in the current reader,
2485 both on the C and Scheme levels. (The Right solution is probably to
2486 change the syntax and prototypes for uniform vectors entirely.)
2489 ** The new module (ice-9 session) provides useful interactive functions.
2491 *** New procedure: (apropos REGEXP OPTION ...)
2493 Display a list of top-level variables whose names match REGEXP, and
2494 the modules they are imported from. Each OPTION should be one of the
2497 value --- Show the value of each matching variable.
2498 shadow --- Show bindings shadowed by subsequently imported modules.
2499 full --- Same as both `shadow' and `value'.
2503 guile> (apropos "trace" 'full)
2504 debug: trace #<procedure trace args>
2505 debug: untrace #<procedure untrace args>
2506 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
2507 the-scm-module: before-backtrace-hook ()
2508 the-scm-module: backtrace #<primitive-procedure backtrace>
2509 the-scm-module: after-backtrace-hook ()
2510 the-scm-module: has-shown-backtrace-hint? #f
2513 ** There are new functions and syntax for working with macros.
2515 Guile implements macros as a special object type. Any variable whose
2516 top-level binding is a macro object acts as a macro. The macro object
2517 specifies how the expression should be transformed before evaluation.
2519 *** Macro objects now print in a reasonable way, resembling procedures.
2521 *** New function: (macro? OBJ)
2522 True iff OBJ is a macro object.
2524 *** New function: (primitive-macro? OBJ)
2525 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
2526 macro transformers, implemented in eval.c rather than Scheme code.
2528 Why do we have this function?
2529 - For symmetry with procedure? and primitive-procedure?,
2530 - to allow custom print procedures to tell whether a macro is
2531 primitive, and display it differently, and
2532 - to allow compilers and user-written evaluators to distinguish
2533 builtin special forms from user-defined ones, which could be
2536 *** New function: (macro-type OBJ)
2537 Return a value indicating what kind of macro OBJ is. Possible return
2540 The symbol `syntax' --- a macro created by procedure->syntax.
2541 The symbol `macro' --- a macro created by procedure->macro.
2542 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
2543 The boolean #f --- if OBJ is not a macro object.
2545 *** New function: (macro-name MACRO)
2546 Return the name of the macro object MACRO's procedure, as returned by
2549 *** New function: (macro-transformer MACRO)
2550 Return the transformer procedure for MACRO.
2552 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
2554 Specify a new macro expander to use in the current module. Each
2555 MODULE is a module name, with the same meaning as in the `use-modules'
2556 form; each named module's exported bindings are added to the current
2557 top-level environment. TRANSFORMER is an expression evaluated in the
2558 resulting environment which must yield a procedure to use as the
2559 module's eval transformer: every expression evaluated in this module
2560 is passed to this function, and the result passed to the Guile
2563 *** macro-eval! is removed. Use local-eval instead.
2565 ** Some magic has been added to the printer to better handle user
2566 written printing routines (like record printers, closure printers).
2568 The problem is that these user written routines must have access to
2569 the current `print-state' to be able to handle fancy things like
2570 detection of circular references. These print-states have to be
2571 passed to the builtin printing routines (display, write, etc) to
2572 properly continue the print chain.
2574 We didn't want to change all existing print code so that it
2575 explicitly passes thru a print state in addition to a port. Instead,
2576 we extented the possible values that the builtin printing routines
2577 accept as a `port'. In addition to a normal port, they now also take
2578 a pair of a normal port and a print-state. Printing will go to the
2579 port and the print-state will be used to control the detection of
2580 circular references, etc. If the builtin function does not care for a
2581 print-state, it is simply ignored.
2583 User written callbacks are now called with such a pair as their
2584 `port', but because every function now accepts this pair as a PORT
2585 argument, you don't have to worry about that. In fact, it is probably
2586 safest to not check for these pairs.
2588 However, it is sometimes necessary to continue a print chain on a
2589 different port, for example to get a intermediate string
2590 representation of the printed value, mangle that string somehow, and
2591 then to finally print the mangled string. Use the new function
2593 inherit-print-state OLD-PORT NEW-PORT
2595 for this. It constructs a new `port' that prints to NEW-PORT but
2596 inherits the print-state of OLD-PORT.
2598 ** struct-vtable-offset renamed to vtable-offset-user
2600 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
2602 ** There is now a fourth (optional) argument to make-vtable-vtable and
2603 make-struct when constructing new types (vtables). This argument
2604 initializes field vtable-index-printer of the vtable.
2606 ** The detection of circular references has been extended to structs.
2607 That is, a structure that -- in the process of being printed -- prints
2608 itself does not lead to infinite recursion.
2610 ** There is now some basic support for fluids. Please read
2611 "libguile/fluid.h" to find out more. It is accessible from Scheme with
2612 the following functions and macros:
2614 Function: make-fluid
2616 Create a new fluid object. Fluids are not special variables or
2617 some other extension to the semantics of Scheme, but rather
2618 ordinary Scheme objects. You can store them into variables (that
2619 are still lexically scoped, of course) or into any other place you
2620 like. Every fluid has a initial value of `#f'.
2622 Function: fluid? OBJ
2624 Test whether OBJ is a fluid.
2626 Function: fluid-ref FLUID
2627 Function: fluid-set! FLUID VAL
2629 Access/modify the fluid FLUID. Modifications are only visible
2630 within the current dynamic root (that includes threads).
2632 Function: with-fluids* FLUIDS VALUES THUNK
2634 FLUIDS is a list of fluids and VALUES a corresponding list of
2635 values for these fluids. Before THUNK gets called the values are
2636 installed in the fluids and the old values of the fluids are
2637 saved in the VALUES list. When the flow of control leaves THUNK
2638 or reenters it, the values get swapped again. You might think of
2639 this as a `safe-fluid-excursion'. Note that the VALUES list is
2640 modified by `with-fluids*'.
2642 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
2644 The same as `with-fluids*' but with a different syntax. It looks
2645 just like `let', but both FLUID and VALUE are evaluated. Remember,
2646 fluids are not special variables but ordinary objects. FLUID
2647 should evaluate to a fluid.
2649 ** Changes to system call interfaces:
2651 *** close-port, close-input-port and close-output-port now return a
2652 boolean instead of an `unspecified' object. #t means that the port
2653 was successfully closed, while #f means it was already closed. It is
2654 also now possible for these procedures to raise an exception if an
2655 error occurs (some errors from write can be delayed until close.)
2657 *** the first argument to chmod, fcntl, ftell and fseek can now be a
2660 *** the third argument to fcntl is now optional.
2662 *** the first argument to chown can now be a file descriptor or a port.
2664 *** the argument to stat can now be a port.
2666 *** The following new procedures have been added (most use scsh
2669 *** procedure: close PORT/FD
2670 Similar to close-port (*note close-port: Closing Ports.), but also
2671 works on file descriptors. A side effect of closing a file
2672 descriptor is that any ports using that file descriptor are moved
2673 to a different file descriptor and have their revealed counts set
2676 *** procedure: port->fdes PORT
2677 Returns the integer file descriptor underlying PORT. As a side
2678 effect the revealed count of PORT is incremented.
2680 *** procedure: fdes->ports FDES
2681 Returns a list of existing ports which have FDES as an underlying
2682 file descriptor, without changing their revealed counts.
2684 *** procedure: fdes->inport FDES
2685 Returns an existing input port which has FDES as its underlying
2686 file descriptor, if one exists, and increments its revealed count.
2687 Otherwise, returns a new input port with a revealed count of 1.
2689 *** procedure: fdes->outport FDES
2690 Returns an existing output port which has FDES as its underlying
2691 file descriptor, if one exists, and increments its revealed count.
2692 Otherwise, returns a new output port with a revealed count of 1.
2694 The next group of procedures perform a `dup2' system call, if NEWFD
2695 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
2696 duplicated can be supplied as an integer or contained in a port. The
2697 type of value returned varies depending on which procedure is used.
2699 All procedures also have the side effect when performing `dup2' that
2700 any ports using NEWFD are moved to a different file descriptor and have
2701 their revealed counts set to zero.
2703 *** procedure: dup->fdes PORT/FD [NEWFD]
2704 Returns an integer file descriptor.
2706 *** procedure: dup->inport PORT/FD [NEWFD]
2707 Returns a new input port using the new file descriptor.
2709 *** procedure: dup->outport PORT/FD [NEWFD]
2710 Returns a new output port using the new file descriptor.
2712 *** procedure: dup PORT/FD [NEWFD]
2713 Returns a new port if PORT/FD is a port, with the same mode as the
2714 supplied port, otherwise returns an integer file descriptor.
2716 *** procedure: dup->port PORT/FD MODE [NEWFD]
2717 Returns a new port using the new file descriptor. MODE supplies a
2718 mode string for the port (*note open-file: File Ports.).
2720 *** procedure: setenv NAME VALUE
2721 Modifies the environment of the current process, which is also the
2722 default environment inherited by child processes.
2724 If VALUE is `#f', then NAME is removed from the environment.
2725 Otherwise, the string NAME=VALUE is added to the environment,
2726 replacing any existing string with name matching NAME.
2728 The return value is unspecified.
2730 *** procedure: truncate-file OBJ SIZE
2731 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
2732 can be a string containing a file name or an integer file
2733 descriptor or port open for output on the file. The underlying
2734 system calls are `truncate' and `ftruncate'.
2736 The return value is unspecified.
2738 *** procedure: setvbuf PORT MODE [SIZE]
2739 Set the buffering mode for PORT. MODE can be:
2747 block buffered, using a newly allocated buffer of SIZE bytes.
2748 However if SIZE is zero or unspecified, the port will be made
2751 This procedure should not be used after I/O has been performed with
2754 Ports are usually block buffered by default, with a default buffer
2755 size. Procedures e.g., *Note open-file: File Ports, which accept a
2756 mode string allow `0' to be added to request an unbuffered port.
2758 *** procedure: fsync PORT/FD
2759 Copies any unwritten data for the specified output file descriptor
2760 to disk. If PORT/FD is a port, its buffer is flushed before the
2761 underlying file descriptor is fsync'd. The return value is
2764 *** procedure: open-fdes PATH FLAGS [MODES]
2765 Similar to `open' but returns a file descriptor instead of a port.
2767 *** procedure: execle PATH ENV [ARG] ...
2768 Similar to `execl', but the environment of the new process is
2769 specified by ENV, which must be a list of strings as returned by
2770 the `environ' procedure.
2772 This procedure is currently implemented using the `execve' system
2773 call, but we call it `execle' because of its Scheme calling
2776 *** procedure: strerror ERRNO
2777 Returns the Unix error message corresponding to ERRNO, an integer.
2779 *** procedure: primitive-exit [STATUS]
2780 Terminate the current process without unwinding the Scheme stack.
2781 This is would typically be useful after a fork. The exit status
2782 is STATUS if supplied, otherwise zero.
2784 *** procedure: times
2785 Returns an object with information about real and processor time.
2786 The following procedures accept such an object as an argument and
2787 return a selected component:
2790 The current real time, expressed as time units relative to an
2794 The CPU time units used by the calling process.
2797 The CPU time units used by the system on behalf of the
2801 The CPU time units used by terminated child processes of the
2802 calling process, whose status has been collected (e.g., using
2806 Similarly, the CPU times units used by the system on behalf of
2807 terminated child processes.
2809 ** Removed: list-length
2810 ** Removed: list-append, list-append!
2811 ** Removed: list-reverse, list-reverse!
2813 ** array-map renamed to array-map!
2815 ** serial-array-map renamed to serial-array-map!
2817 ** catch doesn't take #f as first argument any longer
2819 Previously, it was possible to pass #f instead of a key to `catch'.
2820 That would cause `catch' to pass a jump buffer object to the procedure
2821 passed as second argument. The procedure could then use this jump
2822 buffer objekt as an argument to throw.
2824 This mechanism has been removed since its utility doesn't motivate the
2825 extra complexity it introduces.
2827 ** The `#/' notation for lists now provokes a warning message from Guile.
2828 This syntax will be removed from Guile in the near future.
2830 To disable the warning message, set the GUILE_HUSH environment
2831 variable to any non-empty value.
2833 ** The newline character now prints as `#\newline', following the
2834 normal Scheme notation, not `#\nl'.
2836 * Changes to the gh_ interface
2838 ** The gh_enter function now takes care of loading the Guile startup files.
2839 gh_enter works by calling scm_boot_guile; see the remarks below.
2841 ** Function: void gh_write (SCM x)
2843 Write the printed representation of the scheme object x to the current
2844 output port. Corresponds to the scheme level `write'.
2846 ** gh_list_length renamed to gh_length.
2848 ** vector handling routines
2850 Several major changes. In particular, gh_vector() now resembles
2851 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
2852 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
2853 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
2854 vector-related gh_ functions have been implemented.
2856 ** pair and list routines
2858 Implemented several of the R4RS pair and list functions that were
2861 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
2863 New function. Converts double arrays back and forth between Scheme
2866 * Changes to the scm_ interface
2868 ** The function scm_boot_guile now takes care of loading the startup files.
2870 Guile's primary initialization function, scm_boot_guile, now takes
2871 care of loading `boot-9.scm', in the `ice-9' module, to initialize
2872 Guile, define the module system, and put together some standard
2873 bindings. It also loads `init.scm', which is intended to hold
2874 site-specific initialization code.
2876 Since Guile cannot operate properly until boot-9.scm is loaded, there
2877 is no reason to separate loading boot-9.scm from Guile's other
2878 initialization processes.
2880 This job used to be done by scm_compile_shell_switches, which didn't
2881 make much sense; in particular, it meant that people using Guile for
2882 non-shell-like applications had to jump through hoops to get Guile
2883 initialized properly.
2885 ** The function scm_compile_shell_switches no longer loads the startup files.
2886 Now, Guile always loads the startup files, whenever it is initialized;
2887 see the notes above for scm_boot_guile and scm_load_startup_files.
2889 ** Function: scm_load_startup_files
2890 This new function takes care of loading Guile's initialization file
2891 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
2892 this is always called by the Guile initialization process, it's
2893 probably not too useful to call this yourself, but it's there anyway.
2895 ** The semantics of smob marking have changed slightly.
2897 The smob marking function (the `mark' member of the scm_smobfuns
2898 structure) is no longer responsible for setting the mark bit on the
2899 smob. The generic smob handling code in the garbage collector will
2900 set this bit. The mark function need only ensure that any other
2901 objects the smob refers to get marked.
2903 Note that this change means that the smob's GC8MARK bit is typically
2904 already set upon entry to the mark function. Thus, marking functions
2905 which look like this:
2908 if (SCM_GC8MARKP (ptr))
2910 SCM_SETGC8MARK (ptr);
2911 ... mark objects to which the smob refers ...
2914 are now incorrect, since they will return early, and fail to mark any
2915 other objects the smob refers to. Some code in the Guile library used
2918 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
2920 If you have implemented your own I/O port type, by writing the
2921 functions required by the scm_ptobfuns and then calling scm_newptob,
2922 you will need to change your functions slightly.
2924 The functions in a scm_ptobfuns structure now expect the port itself
2925 as their argument; they used to expect the `stream' member of the
2926 port's scm_port_table structure. This allows functions in an
2927 scm_ptobfuns structure to easily access the port's cell (and any flags
2928 it its CAR), and the port's scm_port_table structure.
2930 Guile now passes the I/O port itself as the `port' argument in the
2931 following scm_ptobfuns functions:
2933 int (*free) (SCM port);
2934 int (*fputc) (int, SCM port);
2935 int (*fputs) (char *, SCM port);
2936 scm_sizet (*fwrite) SCM_P ((char *ptr,
2940 int (*fflush) (SCM port);
2941 int (*fgetc) (SCM port);
2942 int (*fclose) (SCM port);
2944 The interfaces to the `mark', `print', `equalp', and `fgets' methods
2947 If you have existing code which defines its own port types, it is easy
2948 to convert your code to the new interface; simply apply SCM_STREAM to
2949 the port argument to yield the value you code used to expect.
2951 Note that since both the port and the stream have the same type in the
2952 C code --- they are both SCM values --- the C compiler will not remind
2953 you if you forget to update your scm_ptobfuns functions.
2956 ** Function: int scm_internal_select (int fds,
2960 struct timeval *timeout);
2962 This is a replacement for the `select' function provided by the OS.
2963 It enables I/O blocking and sleeping to happen for one cooperative
2964 thread without blocking other threads. It also avoids busy-loops in
2965 these situations. It is intended that all I/O blocking and sleeping
2966 will finally go through this function. Currently, this function is
2967 only available on systems providing `gettimeofday' and `select'.
2969 ** Function: SCM scm_internal_stack_catch (SCM tag,
2970 scm_catch_body_t body,
2972 scm_catch_handler_t handler,
2975 A new sibling to the other two C level `catch' functions
2976 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
2977 the stack to be saved automatically into the variable `the-last-stack'
2978 (scm_the_last_stack_var) on error. This is necessary if you want to
2979 use advanced error reporting, such as calling scm_display_error and
2980 scm_display_backtrace. (They both take a stack object as argument.)
2982 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
2984 scm_catch_handler_t handler,
2987 Spawns a new thread. It does a job similar to
2988 scm_call_with_new_thread but takes arguments more suitable when
2989 spawning threads from application C code.
2991 ** The hook scm_error_callback has been removed. It was originally
2992 intended as a way for the user to install his own error handler. But
2993 that method works badly since it intervenes between throw and catch,
2994 thereby changing the semantics of expressions like (catch #t ...).
2995 The correct way to do it is to use one of the C level catch functions
2996 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
2998 ** Removed functions:
3000 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
3001 scm_list_reverse, scm_list_reverse_x
3003 ** New macros: SCM_LISTn where n is one of the integers 0-9.
3005 These can be used for pretty list creation from C. The idea is taken
3006 from Erick Gallesio's STk.
3008 ** scm_array_map renamed to scm_array_map_x
3010 ** mbstrings are now removed
3012 This means that the type codes scm_tc7_mb_string and
3013 scm_tc7_mb_substring has been removed.
3015 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
3017 Since we no longer support multi-byte strings, these I/O functions
3018 have been simplified, and renamed. Here are their old names, and
3019 their new names and arguments:
3021 scm_gen_putc -> void scm_putc (int c, SCM port);
3022 scm_gen_puts -> void scm_puts (char *s, SCM port);
3023 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
3024 scm_gen_getc -> void scm_getc (SCM port);
3027 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
3029 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
3031 SCM_TYP7S now masks away the bit which distinguishes substrings from
3034 ** scm_catch_body_t: Backward incompatible change!
3036 Body functions to scm_internal_catch and friends do not any longer
3037 take a second argument. This is because it is no longer possible to
3038 pass a #f arg to catch.
3040 ** Calls to scm_protect_object and scm_unprotect now nest properly.
3042 The function scm_protect_object protects its argument from being freed
3043 by the garbage collector. scm_unprotect_object removes that
3046 These functions now nest properly. That is, for every object O, there
3047 is a counter which scm_protect_object(O) increments and
3048 scm_unprotect_object(O) decrements, if the counter is greater than
3049 zero. Every object's counter is zero when it is first created. If an
3050 object's counter is greater than zero, the garbage collector will not
3051 reclaim its storage.
3053 This allows you to use scm_protect_object in your code without
3054 worrying that some other function you call will call
3055 scm_unprotect_object, and allow it to be freed. Assuming that the
3056 functions you call are well-behaved, and unprotect only those objects
3057 they protect, you can follow the same rule and have confidence that
3058 objects will be freed only at appropriate times.
3061 Changes in Guile 1.2 (released Tuesday, June 24 1997):
3063 * Changes to the distribution
3065 ** Nightly snapshots are now available from ftp.red-bean.com.
3066 The old server, ftp.cyclic.com, has been relinquished to its rightful
3069 Nightly snapshots of the Guile development sources are now available via
3070 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
3072 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
3073 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
3075 ** To run Guile without installing it, the procedure has changed a bit.
3077 If you used a separate build directory to compile Guile, you'll need
3078 to include the build directory in SCHEME_LOAD_PATH, as well as the
3079 source directory. See the `INSTALL' file for examples.
3081 * Changes to the procedure for linking libguile with your programs
3083 ** The standard Guile load path for Scheme code now includes
3084 $(datadir)/guile (usually /usr/local/share/guile). This means that
3085 you can install your own Scheme files there, and Guile will find them.
3086 (Previous versions of Guile only checked a directory whose name
3087 contained the Guile version number, so you had to re-install or move
3088 your Scheme sources each time you installed a fresh version of Guile.)
3090 The load path also includes $(datadir)/guile/site; we recommend
3091 putting individual Scheme files there. If you want to install a
3092 package with multiple source files, create a directory for them under
3095 ** Guile 1.2 will now use the Rx regular expression library, if it is
3096 installed on your system. When you are linking libguile into your own
3097 programs, this means you will have to link against -lguile, -lqt (if
3098 you configured Guile with thread support), and -lrx.
3100 If you are using autoconf to generate configuration scripts for your
3101 application, the following lines should suffice to add the appropriate
3102 libraries to your link command:
3104 ### Find Rx, quickthreads and libguile.
3105 AC_CHECK_LIB(rx, main)
3106 AC_CHECK_LIB(qt, main)
3107 AC_CHECK_LIB(guile, scm_shell)
3109 The Guile 1.2 distribution does not contain sources for the Rx
3110 library, as Guile 1.0 did. If you want to use Rx, you'll need to
3111 retrieve it from a GNU FTP site and install it separately.
3113 * Changes to Scheme functions and syntax
3115 ** The dynamic linking features of Guile are now enabled by default.
3116 You can disable them by giving the `--disable-dynamic-linking' option
3119 (dynamic-link FILENAME)
3121 Find the object file denoted by FILENAME (a string) and link it
3122 into the running Guile application. When everything works out,
3123 return a Scheme object suitable for representing the linked object
3124 file. Otherwise an error is thrown. How object files are
3125 searched is system dependent.
3127 (dynamic-object? VAL)
3129 Determine whether VAL represents a dynamically linked object file.
3131 (dynamic-unlink DYNOBJ)
3133 Unlink the indicated object file from the application. DYNOBJ
3134 should be one of the values returned by `dynamic-link'.
3136 (dynamic-func FUNCTION DYNOBJ)
3138 Search the C function indicated by FUNCTION (a string or symbol)
3139 in DYNOBJ and return some Scheme object that can later be used
3140 with `dynamic-call' to actually call this function. Right now,
3141 these Scheme objects are formed by casting the address of the
3142 function to `long' and converting this number to its Scheme
3145 (dynamic-call FUNCTION DYNOBJ)
3147 Call the C function indicated by FUNCTION and DYNOBJ. The
3148 function is passed no arguments and its return value is ignored.
3149 When FUNCTION is something returned by `dynamic-func', call that
3150 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
3151 etc.), look it up in DYNOBJ; this is equivalent to
3153 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
3155 Interrupts are deferred while the C function is executing (with
3156 SCM_DEFER_INTS/SCM_ALLOW_INTS).
3158 (dynamic-args-call FUNCTION DYNOBJ ARGS)
3160 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
3161 some arguments and return its return value. The C function is
3162 expected to take two arguments and return an `int', just like
3165 int c_func (int argc, char **argv);
3167 ARGS must be a list of strings and is converted into an array of
3168 `char *'. The array is passed in ARGV and its size in ARGC. The
3169 return value is converted to a Scheme number and returned from the
3170 call to `dynamic-args-call'.
3172 When dynamic linking is disabled or not supported on your system,
3173 the above functions throw errors, but they are still available.
3175 Here is a small example that works on GNU/Linux:
3177 (define libc-obj (dynamic-link "libc.so"))
3178 (dynamic-args-call 'rand libc-obj '())
3180 See the file `libguile/DYNAMIC-LINKING' for additional comments.
3182 ** The #/ syntax for module names is depreciated, and will be removed
3183 in a future version of Guile. Instead of
3191 The latter syntax is more consistent with existing Lisp practice.
3193 ** Guile now does fancier printing of structures. Structures are the
3194 underlying implementation for records, which in turn are used to
3195 implement modules, so all of these object now print differently and in
3196 a more informative way.
3198 The Scheme printer will examine the builtin variable *struct-printer*
3199 whenever it needs to print a structure object. When this variable is
3200 not `#f' it is deemed to be a procedure and will be applied to the
3201 structure object and the output port. When *struct-printer* is `#f'
3202 or the procedure return `#f' the structure object will be printed in
3203 the boring #<struct 80458270> form.
3205 This hook is used by some routines in ice-9/boot-9.scm to implement
3206 type specific printing routines. Please read the comments there about
3209 One of the more specific uses of structs are records. The printing
3210 procedure that could be passed to MAKE-RECORD-TYPE is now actually
3211 called. It should behave like a *struct-printer* procedure (described
3214 ** Guile now supports a new R4RS-compliant syntax for keywords. A
3215 token of the form #:NAME, where NAME has the same syntax as a Scheme
3216 symbol, is the external representation of the keyword named NAME.
3217 Keyword objects print using this syntax as well, so values containing
3218 keyword objects can be read back into Guile. When used in an
3219 expression, keywords are self-quoting objects.
3221 Guile suports this read syntax, and uses this print syntax, regardless
3222 of the current setting of the `keyword' read option. The `keyword'
3223 read option only controls whether Guile recognizes the `:NAME' syntax,
3224 which is incompatible with R4RS. (R4RS says such token represent
3227 ** Guile has regular expression support again. Guile 1.0 included
3228 functions for matching regular expressions, based on the Rx library.
3229 In Guile 1.1, the Guile/Rx interface was removed to simplify the
3230 distribution, and thus Guile had no regular expression support. Guile
3231 1.2 again supports the most commonly used functions, and supports all
3232 of SCSH's regular expression functions.
3234 If your system does not include a POSIX regular expression library,
3235 and you have not linked Guile with a third-party regexp library such as
3236 Rx, these functions will not be available. You can tell whether your
3237 Guile installation includes regular expression support by checking
3238 whether the `*features*' list includes the `regex' symbol.
3240 *** regexp functions
3242 By default, Guile supports POSIX extended regular expressions. That
3243 means that the characters `(', `)', `+' and `?' are special, and must
3244 be escaped if you wish to match the literal characters.
3246 This regular expression interface was modeled after that implemented
3247 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
3248 with SCSH regular expressions.
3250 **** Function: string-match PATTERN STR [START]
3251 Compile the string PATTERN into a regular expression and compare
3252 it with STR. The optional numeric argument START specifies the
3253 position of STR at which to begin matching.
3255 `string-match' returns a "match structure" which describes what,
3256 if anything, was matched by the regular expression. *Note Match
3257 Structures::. If STR does not match PATTERN at all,
3258 `string-match' returns `#f'.
3260 Each time `string-match' is called, it must compile its PATTERN
3261 argument into a regular expression structure. This operation is
3262 expensive, which makes `string-match' inefficient if the same regular
3263 expression is used several times (for example, in a loop). For better
3264 performance, you can compile a regular expression in advance and then
3265 match strings against the compiled regexp.
3267 **** Function: make-regexp STR [FLAGS]
3268 Compile the regular expression described by STR, and return the
3269 compiled regexp structure. If STR does not describe a legal
3270 regular expression, `make-regexp' throws a
3271 `regular-expression-syntax' error.
3273 FLAGS may be the bitwise-or of one or more of the following:
3275 **** Constant: regexp/extended
3276 Use POSIX Extended Regular Expression syntax when interpreting
3277 STR. If not set, POSIX Basic Regular Expression syntax is used.
3278 If the FLAGS argument is omitted, we assume regexp/extended.
3280 **** Constant: regexp/icase
3281 Do not differentiate case. Subsequent searches using the
3282 returned regular expression will be case insensitive.
3284 **** Constant: regexp/newline
3285 Match-any-character operators don't match a newline.
3287 A non-matching list ([^...]) not containing a newline matches a
3290 Match-beginning-of-line operator (^) matches the empty string
3291 immediately after a newline, regardless of whether the FLAGS
3292 passed to regexp-exec contain regexp/notbol.
3294 Match-end-of-line operator ($) matches the empty string
3295 immediately before a newline, regardless of whether the FLAGS
3296 passed to regexp-exec contain regexp/noteol.
3298 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
3299 Match the compiled regular expression REGEXP against `str'. If
3300 the optional integer START argument is provided, begin matching
3301 from that position in the string. Return a match structure
3302 describing the results of the match, or `#f' if no match could be
3305 FLAGS may be the bitwise-or of one or more of the following:
3307 **** Constant: regexp/notbol
3308 The match-beginning-of-line operator always fails to match (but
3309 see the compilation flag regexp/newline above) This flag may be
3310 used when different portions of a string are passed to
3311 regexp-exec and the beginning of the string should not be
3312 interpreted as the beginning of the line.
3314 **** Constant: regexp/noteol
3315 The match-end-of-line operator always fails to match (but see the
3316 compilation flag regexp/newline above)
3318 **** Function: regexp? OBJ
3319 Return `#t' if OBJ is a compiled regular expression, or `#f'
3322 Regular expressions are commonly used to find patterns in one string
3323 and replace them with the contents of another string.
3325 **** Function: regexp-substitute PORT MATCH [ITEM...]
3326 Write to the output port PORT selected contents of the match
3327 structure MATCH. Each ITEM specifies what should be written, and
3328 may be one of the following arguments:
3330 * A string. String arguments are written out verbatim.
3332 * An integer. The submatch with that number is written.
3334 * The symbol `pre'. The portion of the matched string preceding
3335 the regexp match is written.
3337 * The symbol `post'. The portion of the matched string
3338 following the regexp match is written.
3340 PORT may be `#f', in which case nothing is written; instead,
3341 `regexp-substitute' constructs a string from the specified ITEMs
3344 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
3345 Similar to `regexp-substitute', but can be used to perform global
3346 substitutions on STR. Instead of taking a match structure as an
3347 argument, `regexp-substitute/global' takes two string arguments: a
3348 REGEXP string describing a regular expression, and a TARGET string
3349 which should be matched against this regular expression.
3351 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
3354 * A function may be supplied. When this function is called, it
3355 will be passed one argument: a match structure for a given
3356 regular expression match. It should return a string to be
3357 written out to PORT.
3359 * The `post' symbol causes `regexp-substitute/global' to recurse
3360 on the unmatched portion of STR. This *must* be supplied in
3361 order to perform global search-and-replace on STR; if it is
3362 not present among the ITEMs, then `regexp-substitute/global'
3363 will return after processing a single match.
3365 *** Match Structures
3367 A "match structure" is the object returned by `string-match' and
3368 `regexp-exec'. It describes which portion of a string, if any, matched
3369 the given regular expression. Match structures include: a reference to
3370 the string that was checked for matches; the starting and ending
3371 positions of the regexp match; and, if the regexp included any
3372 parenthesized subexpressions, the starting and ending positions of each
3375 In each of the regexp match functions described below, the `match'
3376 argument must be a match structure returned by a previous call to
3377 `string-match' or `regexp-exec'. Most of these functions return some
3378 information about the original target string that was matched against a
3379 regular expression; we will call that string TARGET for easy reference.
3381 **** Function: regexp-match? OBJ
3382 Return `#t' if OBJ is a match structure returned by a previous
3383 call to `regexp-exec', or `#f' otherwise.
3385 **** Function: match:substring MATCH [N]
3386 Return the portion of TARGET matched by subexpression number N.
3387 Submatch 0 (the default) represents the entire regexp match. If
3388 the regular expression as a whole matched, but the subexpression
3389 number N did not match, return `#f'.
3391 **** Function: match:start MATCH [N]
3392 Return the starting position of submatch number N.
3394 **** Function: match:end MATCH [N]
3395 Return the ending position of submatch number N.
3397 **** Function: match:prefix MATCH
3398 Return the unmatched portion of TARGET preceding the regexp match.
3400 **** Function: match:suffix MATCH
3401 Return the unmatched portion of TARGET following the regexp match.
3403 **** Function: match:count MATCH
3404 Return the number of parenthesized subexpressions from MATCH.
3405 Note that the entire regular expression match itself counts as a
3406 subexpression, and failed submatches are included in the count.
3408 **** Function: match:string MATCH
3409 Return the original TARGET string.
3411 *** Backslash Escapes
3413 Sometimes you will want a regexp to match characters like `*' or `$'
3414 exactly. For example, to check whether a particular string represents
3415 a menu entry from an Info node, it would be useful to match it against
3416 a regexp like `^* [^:]*::'. However, this won't work; because the
3417 asterisk is a metacharacter, it won't match the `*' at the beginning of
3418 the string. In this case, we want to make the first asterisk un-magic.
3420 You can do this by preceding the metacharacter with a backslash
3421 character `\'. (This is also called "quoting" the metacharacter, and
3422 is known as a "backslash escape".) When Guile sees a backslash in a
3423 regular expression, it considers the following glyph to be an ordinary
3424 character, no matter what special meaning it would ordinarily have.
3425 Therefore, we can make the above example work by changing the regexp to
3426 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
3427 to match only a single asterisk in the target string.
3429 Since the backslash is itself a metacharacter, you may force a
3430 regexp to match a backslash in the target string by preceding the
3431 backslash with itself. For example, to find variable references in a
3432 TeX program, you might want to find occurrences of the string `\let\'
3433 followed by any number of alphabetic characters. The regular expression
3434 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
3435 each match a single backslash in the target string.
3437 **** Function: regexp-quote STR
3438 Quote each special character found in STR with a backslash, and
3439 return the resulting string.
3441 *Very important:* Using backslash escapes in Guile source code (as
3442 in Emacs Lisp or C) can be tricky, because the backslash character has
3443 special meaning for the Guile reader. For example, if Guile encounters
3444 the character sequence `\n' in the middle of a string while processing
3445 Scheme code, it replaces those characters with a newline character.
3446 Similarly, the character sequence `\t' is replaced by a horizontal tab.
3447 Several of these "escape sequences" are processed by the Guile reader
3448 before your code is executed. Unrecognized escape sequences are
3449 ignored: if the characters `\*' appear in a string, they will be
3450 translated to the single character `*'.
3452 This translation is obviously undesirable for regular expressions,
3453 since we want to be able to include backslashes in a string in order to
3454 escape regexp metacharacters. Therefore, to make sure that a backslash
3455 is preserved in a string in your Guile program, you must use *two*
3456 consecutive backslashes:
3458 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
3460 The string in this example is preprocessed by the Guile reader before
3461 any code is executed. The resulting argument to `make-regexp' is the
3462 string `^\* [^:]*', which is what we really want.
3464 This also means that in order to write a regular expression that
3465 matches a single backslash character, the regular expression string in
3466 the source code must include *four* backslashes. Each consecutive pair
3467 of backslashes gets translated by the Guile reader to a single
3468 backslash, and the resulting double-backslash is interpreted by the
3469 regexp engine as matching a single backslash character. Hence:
3471 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
3473 The reason for the unwieldiness of this syntax is historical. Both
3474 regular expression pattern matchers and Unix string processing systems
3475 have traditionally used backslashes with the special meanings described
3476 above. The POSIX regular expression specification and ANSI C standard
3477 both require these semantics. Attempting to abandon either convention
3478 would cause other kinds of compatibility problems, possibly more severe
3479 ones. Therefore, without extending the Scheme reader to support
3480 strings with different quoting conventions (an ungainly and confusing
3481 extension when implemented in other languages), we must adhere to this
3482 cumbersome escape syntax.
3484 * Changes to the gh_ interface
3486 * Changes to the scm_ interface
3488 * Changes to system call interfaces:
3490 ** The value returned by `raise' is now unspecified. It throws an exception
3493 *** A new procedure `sigaction' can be used to install signal handlers
3495 (sigaction signum [action] [flags])
3497 signum is the signal number, which can be specified using the value
3500 If action is omitted, sigaction returns a pair: the CAR is the current
3501 signal hander, which will be either an integer with the value SIG_DFL
3502 (default action) or SIG_IGN (ignore), or the Scheme procedure which
3503 handles the signal, or #f if a non-Scheme procedure handles the
3504 signal. The CDR contains the current sigaction flags for the handler.
3506 If action is provided, it is installed as the new handler for signum.
3507 action can be a Scheme procedure taking one argument, or the value of
3508 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
3509 whatever signal handler was installed before sigaction was first used.
3510 Flags can optionally be specified for the new handler (SA_RESTART is
3511 always used if the system provides it, so need not be specified.) The
3512 return value is a pair with information about the old handler as
3515 This interface does not provide access to the "signal blocking"
3516 facility. Maybe this is not needed, since the thread support may
3517 provide solutions to the problem of consistent access to data
3520 *** A new procedure `flush-all-ports' is equivalent to running
3521 `force-output' on every port open for output.
3523 ** Guile now provides information on how it was built, via the new
3524 global variable, %guile-build-info. This variable records the values
3525 of the standard GNU makefile directory variables as an assocation
3526 list, mapping variable names (symbols) onto directory paths (strings).
3527 For example, to find out where the Guile link libraries were
3528 installed, you can say:
3530 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
3533 * Changes to the scm_ interface
3535 ** The new function scm_handle_by_message_noexit is just like the
3536 existing scm_handle_by_message function, except that it doesn't call
3537 exit to terminate the process. Instead, it prints a message and just
3538 returns #f. This might be a more appropriate catch-all handler for
3539 new dynamic roots and threads.
3542 Changes in Guile 1.1 (released Friday, May 16 1997):
3544 * Changes to the distribution.
3546 The Guile 1.0 distribution has been split up into several smaller
3548 guile-core --- the Guile interpreter itself.
3549 guile-tcltk --- the interface between the Guile interpreter and
3550 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
3551 is a toolkit for building graphical user interfaces.
3552 guile-rgx-ctax --- the interface between Guile and the Rx regular
3553 expression matcher, and the translator for the Ctax
3554 programming language. These are packaged together because the
3555 Ctax translator uses Rx to parse Ctax source code.
3557 This NEWS file describes the changes made to guile-core since the 1.0
3560 We no longer distribute the documentation, since it was either out of
3561 date, or incomplete. As soon as we have current documentation, we
3566 * Changes to the stand-alone interpreter
3568 ** guile now accepts command-line arguments compatible with SCSH, Olin
3569 Shivers' Scheme Shell.
3571 In general, arguments are evaluated from left to right, but there are
3572 exceptions. The following switches stop argument processing, and
3573 stash all remaining command-line arguments as the value returned by
3574 the (command-line) function.
3575 -s SCRIPT load Scheme source code from FILE, and exit
3576 -c EXPR evalute Scheme expression EXPR, and exit
3577 -- stop scanning arguments; run interactively
3579 The switches below are processed as they are encountered.
3580 -l FILE load Scheme source code from FILE
3581 -e FUNCTION after reading script, apply FUNCTION to
3582 command line arguments
3583 -ds do -s script at this point
3584 --emacs enable Emacs protocol (experimental)
3585 -h, --help display this help and exit
3586 -v, --version display version information and exit
3587 \ read arguments from following script lines
3589 So, for example, here is a Guile script named `ekko' (thanks, Olin)
3590 which re-implements the traditional "echo" command:
3592 #!/usr/local/bin/guile -s
3595 (map (lambda (arg) (display arg) (display " "))
3599 (main (command-line))
3601 Suppose we invoke this script as follows:
3603 ekko a speckled gecko
3605 Through the magic of Unix script processing (triggered by the `#!'
3606 token at the top of the file), /usr/local/bin/guile receives the
3607 following list of command-line arguments:
3609 ("-s" "./ekko" "a" "speckled" "gecko")
3611 Unix inserts the name of the script after the argument specified on
3612 the first line of the file (in this case, "-s"), and then follows that
3613 with the arguments given to the script. Guile loads the script, which
3614 defines the `main' function, and then applies it to the list of
3615 remaining command-line arguments, ("a" "speckled" "gecko").
3617 In Unix, the first line of a script file must take the following form:
3619 #!INTERPRETER ARGUMENT
3621 where INTERPRETER is the absolute filename of the interpreter
3622 executable, and ARGUMENT is a single command-line argument to pass to
3625 You may only pass one argument to the interpreter, and its length is
3626 limited. These restrictions can be annoying to work around, so Guile
3627 provides a general mechanism (borrowed from, and compatible with,
3628 SCSH) for circumventing them.
3630 If the ARGUMENT in a Guile script is a single backslash character,
3631 `\', Guile will open the script file, parse arguments from its second
3632 and subsequent lines, and replace the `\' with them. So, for example,
3633 here is another implementation of the `ekko' script:
3635 #!/usr/local/bin/guile \
3639 (for-each (lambda (arg) (display arg) (display " "))
3643 If the user invokes this script as follows:
3645 ekko a speckled gecko
3647 Unix expands this into
3649 /usr/local/bin/guile \ ekko a speckled gecko
3651 When Guile sees the `\' argument, it replaces it with the arguments
3652 read from the second line of the script, producing:
3654 /usr/local/bin/guile -e main -s ekko a speckled gecko
3656 This tells Guile to load the `ekko' script, and apply the function
3657 `main' to the argument list ("a" "speckled" "gecko").
3659 Here is how Guile parses the command-line arguments:
3660 - Each space character terminates an argument. This means that two
3661 spaces in a row introduce an empty-string argument.
3662 - The tab character is not permitted (unless you quote it with the
3663 backslash character, as described below), to avoid confusion.
3664 - The newline character terminates the sequence of arguments, and will
3665 also terminate a final non-empty argument. (However, a newline
3666 following a space will not introduce a final empty-string argument;
3667 it only terminates the argument list.)
3668 - The backslash character is the escape character. It escapes
3669 backslash, space, tab, and newline. The ANSI C escape sequences
3670 like \n and \t are also supported. These produce argument
3671 constituents; the two-character combination \n doesn't act like a
3672 terminating newline. The escape sequence \NNN for exactly three
3673 octal digits reads as the character whose ASCII code is NNN. As
3674 above, characters produced this way are argument constituents.
3675 Backslash followed by other characters is not allowed.
3677 * Changes to the procedure for linking libguile with your programs
3679 ** Guile now builds and installs a shared guile library, if your
3680 system support shared libraries. (It still builds a static library on
3681 all systems.) Guile automatically detects whether your system
3682 supports shared libraries. To prevent Guile from buildisg shared
3683 libraries, pass the `--disable-shared' flag to the configure script.
3685 Guile takes longer to compile when it builds shared libraries, because
3686 it must compile every file twice --- once to produce position-
3687 independent object code, and once to produce normal object code.
3689 ** The libthreads library has been merged into libguile.
3691 To link a program against Guile, you now need only link against
3692 -lguile and -lqt; -lthreads is no longer needed. If you are using
3693 autoconf to generate configuration scripts for your application, the
3694 following lines should suffice to add the appropriate libraries to
3697 ### Find quickthreads and libguile.
3698 AC_CHECK_LIB(qt, main)
3699 AC_CHECK_LIB(guile, scm_shell)
3701 * Changes to Scheme functions
3703 ** Guile Scheme's special syntax for keyword objects is now optional,
3704 and disabled by default.
3706 The syntax variation from R4RS made it difficult to port some
3707 interesting packages to Guile. The routines which accepted keyword
3708 arguments (mostly in the module system) have been modified to also
3709 accept symbols whose names begin with `:'.
3711 To change the keyword syntax, you must first import the (ice-9 debug)
3713 (use-modules (ice-9 debug))
3715 Then you can enable the keyword syntax as follows:
3716 (read-set! keywords 'prefix)
3718 To disable keyword syntax, do this:
3719 (read-set! keywords #f)
3721 ** Many more primitive functions accept shared substrings as
3722 arguments. In the past, these functions required normal, mutable
3723 strings as arguments, although they never made use of this
3726 ** The uniform array functions now operate on byte vectors. These
3727 functions are `array-fill!', `serial-array-copy!', `array-copy!',
3728 `serial-array-map', `array-map', `array-for-each', and
3731 ** The new functions `trace' and `untrace' implement simple debugging
3732 support for Scheme functions.
3734 The `trace' function accepts any number of procedures as arguments,
3735 and tells the Guile interpreter to display each procedure's name and
3736 arguments each time the procedure is invoked. When invoked with no
3737 arguments, `trace' returns the list of procedures currently being
3740 The `untrace' function accepts any number of procedures as arguments,
3741 and tells the Guile interpreter not to trace them any more. When
3742 invoked with no arguments, `untrace' untraces all curretly traced
3745 The tracing in Guile has an advantage over most other systems: we
3746 don't create new procedure objects, but mark the procedure objects
3747 themselves. This means that anonymous and internal procedures can be
3750 ** The function `assert-repl-prompt' has been renamed to
3751 `set-repl-prompt!'. It takes one argument, PROMPT.
3752 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
3753 - If PROMPT is a string, we use it as a prompt.
3754 - If PROMPT is a procedure accepting no arguments, we call it, and
3755 display the result as a prompt.
3756 - Otherwise, we display "> ".
3758 ** The new function `eval-string' reads Scheme expressions from a
3759 string and evaluates them, returning the value of the last expression
3760 in the string. If the string contains no expressions, it returns an
3763 ** The new function `thunk?' returns true iff its argument is a
3764 procedure of zero arguments.
3766 ** `defined?' is now a builtin function, instead of syntax. This
3767 means that its argument should be quoted. It returns #t iff its
3768 argument is bound in the current module.
3770 ** The new syntax `use-modules' allows you to add new modules to your
3771 environment without re-typing a complete `define-module' form. It
3772 accepts any number of module names as arguments, and imports their
3773 public bindings into the current module.
3775 ** The new function (module-defined? NAME MODULE) returns true iff
3776 NAME, a symbol, is defined in MODULE, a module object.
3778 ** The new function `builtin-bindings' creates and returns a hash
3779 table containing copies of all the root module's bindings.
3781 ** The new function `builtin-weak-bindings' does the same as
3782 `builtin-bindings', but creates a doubly-weak hash table.
3784 ** The `equal?' function now considers variable objects to be
3785 equivalent if they have the same name and the same value.
3787 ** The new function `command-line' returns the command-line arguments
3788 given to Guile, as a list of strings.
3790 When using guile as a script interpreter, `command-line' returns the
3791 script's arguments; those processed by the interpreter (like `-s' or
3792 `-c') are omitted. (In other words, you get the normal, expected
3793 behavior.) Any application that uses scm_shell to process its
3794 command-line arguments gets this behavior as well.
3796 ** The new function `load-user-init' looks for a file called `.guile'
3797 in the user's home directory, and loads it if it exists. This is
3798 mostly for use by the code generated by scm_compile_shell_switches,
3799 but we thought it might also be useful in other circumstances.
3801 ** The new function `log10' returns the base-10 logarithm of its
3804 ** Changes to I/O functions
3806 *** The functions `read', `primitive-load', `read-and-eval!', and
3807 `primitive-load-path' no longer take optional arguments controlling
3808 case insensitivity and a `#' parser.
3810 Case sensitivity is now controlled by a read option called
3811 `case-insensitive'. The user can add new `#' syntaxes with the
3812 `read-hash-extend' function (see below).
3814 *** The new function `read-hash-extend' allows the user to change the
3815 syntax of Guile Scheme in a somewhat controlled way.
3817 (read-hash-extend CHAR PROC)
3818 When parsing S-expressions, if we read a `#' character followed by
3819 the character CHAR, use PROC to parse an object from the stream.
3820 If PROC is #f, remove any parsing procedure registered for CHAR.
3822 The reader applies PROC to two arguments: CHAR and an input port.
3824 *** The new functions read-delimited and read-delimited! provide a
3825 general mechanism for doing delimited input on streams.
3827 (read-delimited DELIMS [PORT HANDLE-DELIM])
3828 Read until we encounter one of the characters in DELIMS (a string),
3829 or end-of-file. PORT is the input port to read from; it defaults to
3830 the current input port. The HANDLE-DELIM parameter determines how
3831 the terminating character is handled; it should be one of the
3834 'trim omit delimiter from result
3835 'peek leave delimiter character in input stream
3836 'concat append delimiter character to returned value
3837 'split return a pair: (RESULT . TERMINATOR)
3839 HANDLE-DELIM defaults to 'peek.
3841 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
3842 A side-effecting variant of `read-delimited'.
3844 The data is written into the string BUF at the indices in the
3845 half-open interval [START, END); the default interval is the whole
3846 string: START = 0 and END = (string-length BUF). The values of
3847 START and END must specify a well-defined interval in BUF, i.e.
3848 0 <= START <= END <= (string-length BUF).
3850 It returns NBYTES, the number of bytes read. If the buffer filled
3851 up without a delimiter character being found, it returns #f. If the
3852 port is at EOF when the read starts, it returns the EOF object.
3854 If an integer is returned (i.e., the read is successfully terminated
3855 by reading a delimiter character), then the HANDLE-DELIM parameter
3856 determines how to handle the terminating character. It is described
3857 above, and defaults to 'peek.
3859 (The descriptions of these functions were borrowed from the SCSH
3860 manual, by Olin Shivers and Brian Carlstrom.)
3862 *** The `%read-delimited!' function is the primitive used to implement
3863 `read-delimited' and `read-delimited!'.
3865 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
3867 This returns a pair of values: (TERMINATOR . NUM-READ).
3868 - TERMINATOR describes why the read was terminated. If it is a
3869 character or the eof object, then that is the value that terminated
3870 the read. If it is #f, the function filled the buffer without finding
3871 a delimiting character.
3872 - NUM-READ is the number of characters read into BUF.
3874 If the read is successfully terminated by reading a delimiter
3875 character, then the gobble? parameter determines what to do with the
3876 terminating character. If true, the character is removed from the
3877 input stream; if false, the character is left in the input stream
3878 where a subsequent read operation will retrieve it. In either case,
3879 the character is also the first value returned by the procedure call.
3881 (The descriptions of this function was borrowed from the SCSH manual,
3882 by Olin Shivers and Brian Carlstrom.)
3884 *** The `read-line' and `read-line!' functions have changed; they now
3885 trim the terminator by default; previously they appended it to the
3886 returned string. For the old behavior, use (read-line PORT 'concat).
3888 *** The functions `uniform-array-read!' and `uniform-array-write!' now
3889 take new optional START and END arguments, specifying the region of
3890 the array to read and write.
3892 *** The `ungetc-char-ready?' function has been removed. We feel it's
3893 inappropriate for an interface to expose implementation details this
3896 ** Changes to the Unix library and system call interface
3898 *** The new fcntl function provides access to the Unix `fcntl' system
3901 (fcntl PORT COMMAND VALUE)
3902 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
3903 Values for COMMAND are:
3905 F_DUPFD duplicate a file descriptor
3906 F_GETFD read the descriptor's close-on-exec flag
3907 F_SETFD set the descriptor's close-on-exec flag to VALUE
3908 F_GETFL read the descriptor's flags, as set on open
3909 F_SETFL set the descriptor's flags, as set on open to VALUE
3910 F_GETOWN return the process ID of a socket's owner, for SIGIO
3911 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
3912 FD_CLOEXEC not sure what this is
3914 For details, see the documentation for the fcntl system call.
3916 *** The arguments to `select' have changed, for compatibility with
3917 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
3918 expected behavior. The MILLISECONDS parameter has been changed to
3919 MICROSECONDS, to more closely resemble the underlying system call.
3920 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
3921 corresponding return set will be the same.
3923 *** The arguments to the `mknod' system call have changed. They are
3926 (mknod PATH TYPE PERMS DEV)
3927 Create a new file (`node') in the file system. PATH is the name of
3928 the file to create. TYPE is the kind of file to create; it should
3929 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
3930 permission bits to give the newly created file. If TYPE is
3931 'block-special or 'char-special, DEV specifies which device the
3932 special file refers to; its interpretation depends on the kind of
3933 special file being created.
3935 *** The `fork' function has been renamed to `primitive-fork', to avoid
3936 clashing with various SCSH forks.
3938 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
3939 and `recvfrom!'. They no longer accept a size for a second argument;
3940 you must pass a string to hold the received value. They no longer
3941 return the buffer. Instead, `recv' returns the length of the message
3942 received, and `recvfrom' returns a pair containing the packet's length
3943 and originating address.
3945 *** The file descriptor datatype has been removed, as have the
3946 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
3947 We plan to replace these functions with a SCSH-compatible interface.
3949 *** The `create' function has been removed; it's just a special case
3952 *** There are new functions to break down process termination status
3953 values. In the descriptions below, STATUS is a value returned by
3956 (status:exit-val STATUS)
3957 If the child process exited normally, this function returns the exit
3958 code for the child process (i.e., the value passed to exit, or
3959 returned from main). If the child process did not exit normally,
3960 this function returns #f.
3962 (status:stop-sig STATUS)
3963 If the child process was suspended by a signal, this function
3964 returns the signal that suspended the child. Otherwise, it returns
3967 (status:term-sig STATUS)
3968 If the child process terminated abnormally, this function returns
3969 the signal that terminated the child. Otherwise, this function
3972 POSIX promises that exactly one of these functions will return true on
3973 a valid STATUS value.
3975 These functions are compatible with SCSH.
3977 *** There are new accessors and setters for the broken-out time vectors
3978 returned by `localtime', `gmtime', and that ilk. They are:
3980 Component Accessor Setter
3981 ========================= ============ ============
3982 seconds tm:sec set-tm:sec
3983 minutes tm:min set-tm:min
3984 hours tm:hour set-tm:hour
3985 day of the month tm:mday set-tm:mday
3986 month tm:mon set-tm:mon
3987 year tm:year set-tm:year
3988 day of the week tm:wday set-tm:wday
3989 day in the year tm:yday set-tm:yday
3990 daylight saving time tm:isdst set-tm:isdst
3991 GMT offset, seconds tm:gmtoff set-tm:gmtoff
3992 name of time zone tm:zone set-tm:zone
3994 *** There are new accessors for the vectors returned by `uname',
3995 describing the host system:
3998 ============================================== ================
3999 name of the operating system implementation utsname:sysname
4000 network name of this machine utsname:nodename
4001 release level of the operating system utsname:release
4002 version level of the operating system utsname:version
4003 machine hardware platform utsname:machine
4005 *** There are new accessors for the vectors returned by `getpw',
4006 `getpwnam', `getpwuid', and `getpwent', describing entries from the
4007 system's user database:
4010 ====================== =================
4011 user name passwd:name
4012 user password passwd:passwd
4015 real name passwd:gecos
4016 home directory passwd:dir
4017 shell program passwd:shell
4019 *** There are new accessors for the vectors returned by `getgr',
4020 `getgrnam', `getgrgid', and `getgrent', describing entries from the
4021 system's group database:
4024 ======================= ============
4025 group name group:name
4026 group password group:passwd
4028 group members group:mem
4030 *** There are new accessors for the vectors returned by `gethost',
4031 `gethostbyaddr', `gethostbyname', and `gethostent', describing
4035 ========================= ===============
4036 official name of host hostent:name
4037 alias list hostent:aliases
4038 host address type hostent:addrtype
4039 length of address hostent:length
4040 list of addresses hostent:addr-list
4042 *** There are new accessors for the vectors returned by `getnet',
4043 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
4047 ========================= ===============
4048 official name of net netent:name
4049 alias list netent:aliases
4050 net number type netent:addrtype
4051 net number netent:net
4053 *** There are new accessors for the vectors returned by `getproto',
4054 `getprotobyname', `getprotobynumber', and `getprotoent', describing
4058 ========================= ===============
4059 official protocol name protoent:name
4060 alias list protoent:aliases
4061 protocol number protoent:proto
4063 *** There are new accessors for the vectors returned by `getserv',
4064 `getservbyname', `getservbyport', and `getservent', describing
4068 ========================= ===============
4069 official service name servent:name
4070 alias list servent:aliases
4071 port number servent:port
4072 protocol to use servent:proto
4074 *** There are new accessors for the sockaddr structures returned by
4075 `accept', `getsockname', `getpeername', `recvfrom!':
4078 ======================================== ===============
4079 address format (`family') sockaddr:fam
4080 path, for file domain addresses sockaddr:path
4081 address, for internet domain addresses sockaddr:addr
4082 TCP or UDP port, for internet sockaddr:port
4084 *** The `getpwent', `getgrent', `gethostent', `getnetent',
4085 `getprotoent', and `getservent' functions now return #f at the end of
4086 the user database. (They used to throw an exception.)
4088 Note that calling MUMBLEent function is equivalent to calling the
4089 corresponding MUMBLE function with no arguments.
4091 *** The `setpwent', `setgrent', `sethostent', `setnetent',
4092 `setprotoent', and `setservent' routines now take no arguments.
4094 *** The `gethost', `getproto', `getnet', and `getserv' functions now
4095 provide more useful information when they throw an exception.
4097 *** The `lnaof' function has been renamed to `inet-lnaof'.
4099 *** Guile now claims to have the `current-time' feature.
4101 *** The `mktime' function now takes an optional second argument ZONE,
4102 giving the time zone to use for the conversion. ZONE should be a
4103 string, in the same format as expected for the "TZ" environment variable.
4105 *** The `strptime' function now returns a pair (TIME . COUNT), where
4106 TIME is the parsed time as a vector, and COUNT is the number of
4107 characters from the string left unparsed. This function used to
4108 return the remaining characters as a string.
4110 *** The `gettimeofday' function has replaced the old `time+ticks' function.
4111 The return value is now (SECONDS . MICROSECONDS); the fractional
4112 component is no longer expressed in "ticks".
4114 *** The `ticks/sec' constant has been removed, in light of the above change.
4116 * Changes to the gh_ interface
4118 ** gh_eval_str() now returns an SCM object which is the result of the
4121 ** gh_scm2str() now copies the Scheme data to a caller-provided C
4124 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
4125 and returns the array
4127 ** gh_scm2str0() is gone: there is no need to distinguish
4128 null-terminated from non-null-terminated, since gh_scm2newstr() allows
4129 the user to interpret the data both ways.
4131 * Changes to the scm_ interface
4133 ** The new function scm_symbol_value0 provides an easy way to get a
4134 symbol's value from C code:
4136 SCM scm_symbol_value0 (char *NAME)
4137 Return the value of the symbol named by the null-terminated string
4138 NAME in the current module. If the symbol named NAME is unbound in
4139 the current module, return SCM_UNDEFINED.
4141 ** The new function scm_sysintern0 creates new top-level variables,
4142 without assigning them a value.
4144 SCM scm_sysintern0 (char *NAME)
4145 Create a new Scheme top-level variable named NAME. NAME is a
4146 null-terminated string. Return the variable's value cell.
4148 ** The function scm_internal_catch is the guts of catch. It handles
4149 all the mechanics of setting up a catch target, invoking the catch
4150 body, and perhaps invoking the handler if the body does a throw.
4152 The function is designed to be usable from C code, but is general
4153 enough to implement all the semantics Guile Scheme expects from throw.
4155 TAG is the catch tag. Typically, this is a symbol, but this function
4156 doesn't actually care about that.
4158 BODY is a pointer to a C function which runs the body of the catch;
4159 this is the code you can throw from. We call it like this:
4160 BODY (BODY_DATA, JMPBUF)
4162 BODY_DATA is just the BODY_DATA argument we received; we pass it
4163 through to BODY as its first argument. The caller can make
4164 BODY_DATA point to anything useful that BODY might need.
4165 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
4166 which we have just created and initialized.
4168 HANDLER is a pointer to a C function to deal with a throw to TAG,
4169 should one occur. We call it like this:
4170 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
4172 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
4173 same idea as BODY_DATA above.
4174 THROWN_TAG is the tag that the user threw to; usually this is
4175 TAG, but it could be something else if TAG was #t (i.e., a
4176 catch-all), or the user threw to a jmpbuf.
4177 THROW_ARGS is the list of arguments the user passed to the THROW
4180 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
4181 is just a pointer we pass through to HANDLER. We don't actually
4182 use either of those pointers otherwise ourselves. The idea is
4183 that, if our caller wants to communicate something to BODY or
4184 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
4185 HANDLER can then use. Think of it as a way to make BODY and
4186 HANDLER closures, not just functions; MUMBLE_DATA points to the
4189 Of course, it's up to the caller to make sure that any data a
4190 MUMBLE_DATA needs is protected from GC. A common way to do this is
4191 to make MUMBLE_DATA a pointer to data stored in an automatic
4192 structure variable; since the collector must scan the stack for
4193 references anyway, this assures that any references in MUMBLE_DATA
4196 ** The new function scm_internal_lazy_catch is exactly like
4197 scm_internal_catch, except:
4199 - It does not unwind the stack (this is the major difference).
4200 - If handler returns, its value is returned from the throw.
4201 - BODY always receives #f as its JMPBUF argument (since there's no
4202 jmpbuf associated with a lazy catch, because we don't unwind the
4205 ** scm_body_thunk is a new body function you can pass to
4206 scm_internal_catch if you want the body to be like Scheme's `catch'
4207 --- a thunk, or a function of one argument if the tag is #f.
4209 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
4210 contains the Scheme procedure to invoke as the body, and the tag
4211 we're catching. If the tag is #f, then we pass JMPBUF (created by
4212 scm_internal_catch) to the body procedure; otherwise, the body gets
4215 ** scm_handle_by_proc is a new handler function you can pass to
4216 scm_internal_catch if you want the handler to act like Scheme's catch
4217 --- call a procedure with the tag and the throw arguments.
4219 If the user does a throw to this catch, this function runs a handler
4220 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
4221 variable holding the Scheme procedure object to invoke. It ought to
4222 be a pointer to an automatic variable (i.e., one living on the stack),
4223 or the procedure object should be otherwise protected from GC.
4225 ** scm_handle_by_message is a new handler function to use with
4226 `scm_internal_catch' if you want Guile to print a message and die.
4227 It's useful for dealing with throws to uncaught keys at the top level.
4229 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
4230 message header to print; if zero, we use "guile" instead. That
4231 text is followed by a colon, then the message described by ARGS.
4233 ** The return type of scm_boot_guile is now void; the function does
4234 not return a value, and indeed, never returns at all.
4236 ** The new function scm_shell makes it easy for user applications to
4237 process command-line arguments in a way that is compatible with the
4238 stand-alone guile interpreter (which is in turn compatible with SCSH,
4241 To use the scm_shell function, first initialize any guile modules
4242 linked into your application, and then call scm_shell with the values
4243 of ARGC and ARGV your `main' function received. scm_shell will add
4244 any SCSH-style meta-arguments from the top of the script file to the
4245 argument vector, and then process the command-line arguments. This
4246 generally means loading a script file or starting up an interactive
4247 command interpreter. For details, see "Changes to the stand-alone
4250 ** The new functions scm_get_meta_args and scm_count_argv help you
4251 implement the SCSH-style meta-argument, `\'.
4253 char **scm_get_meta_args (int ARGC, char **ARGV)
4254 If the second element of ARGV is a string consisting of a single
4255 backslash character (i.e. "\\" in Scheme notation), open the file
4256 named by the following argument, parse arguments from it, and return
4257 the spliced command line. The returned array is terminated by a
4260 For details of argument parsing, see above, under "guile now accepts
4261 command-line arguments compatible with SCSH..."
4263 int scm_count_argv (char **ARGV)
4264 Count the arguments in ARGV, assuming it is terminated by a null
4267 For an example of how these functions might be used, see the source
4268 code for the function scm_shell in libguile/script.c.
4270 You will usually want to use scm_shell instead of calling this
4273 ** The new function scm_compile_shell_switches turns an array of
4274 command-line arguments into Scheme code to carry out the actions they
4275 describe. Given ARGC and ARGV, it returns a Scheme expression to
4276 evaluate, and calls scm_set_program_arguments to make any remaining
4277 command-line arguments available to the Scheme code. For example,
4278 given the following arguments:
4280 -e main -s ekko a speckled gecko
4282 scm_set_program_arguments will return the following expression:
4284 (begin (load "ekko") (main (command-line)) (quit))
4286 You will usually want to use scm_shell instead of calling this
4289 ** The function scm_shell_usage prints a usage message appropriate for
4290 an interpreter that uses scm_compile_shell_switches to handle its
4291 command-line arguments.
4293 void scm_shell_usage (int FATAL, char *MESSAGE)
4294 Print a usage message to the standard error output. If MESSAGE is
4295 non-zero, write it before the usage message, followed by a newline.
4296 If FATAL is non-zero, exit the process, using FATAL as the
4297 termination status. (If you want to be compatible with Guile,
4298 always use 1 as the exit status when terminating due to command-line
4301 You will usually want to use scm_shell instead of calling this
4304 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
4305 expressions. It used to return SCM_EOL. Earth-shattering.
4307 ** The macros for declaring scheme objects in C code have been
4308 rearranged slightly. They are now:
4310 SCM_SYMBOL (C_NAME, SCHEME_NAME)
4311 Declare a static SCM variable named C_NAME, and initialize it to
4312 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
4313 be a C identifier, and SCHEME_NAME should be a C string.
4315 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
4316 Just like SCM_SYMBOL, but make C_NAME globally visible.
4318 SCM_VCELL (C_NAME, SCHEME_NAME)
4319 Create a global variable at the Scheme level named SCHEME_NAME.
4320 Declare a static SCM variable named C_NAME, and initialize it to
4321 point to the Scheme variable's value cell.
4323 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
4324 Just like SCM_VCELL, but make C_NAME globally visible.
4326 The `guile-snarf' script writes initialization code for these macros
4327 to its standard output, given C source code as input.
4329 The SCM_GLOBAL macro is gone.
4331 ** The scm_read_line and scm_read_line_x functions have been replaced
4332 by Scheme code based on the %read-delimited! procedure (known to C
4333 code as scm_read_delimited_x). See its description above for more
4336 ** The function scm_sys_open has been renamed to scm_open. It now
4337 returns a port instead of an FD object.
4339 * The dynamic linking support has changed. For more information, see
4340 libguile/DYNAMIC-LINKING.
4345 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
4348 * Changes to the 'guile' program:
4350 ** Guile now loads some new files when it starts up. Guile first
4351 searches the load path for init.scm, and loads it if found. Then, if
4352 Guile is not being used to execute a script, and the user's home
4353 directory contains a file named `.guile', Guile loads that.
4355 ** You can now use Guile as a shell script interpreter.
4357 To paraphrase the SCSH manual:
4359 When Unix tries to execute an executable file whose first two
4360 characters are the `#!', it treats the file not as machine code to
4361 be directly executed by the native processor, but as source code
4362 to be executed by some interpreter. The interpreter to use is
4363 specified immediately after the #! sequence on the first line of
4364 the source file. The kernel reads in the name of the interpreter,
4365 and executes that instead. It passes the interpreter the source
4366 filename as its first argument, with the original arguments
4367 following. Consult the Unix man page for the `exec' system call
4368 for more information.
4370 Now you can use Guile as an interpreter, using a mechanism which is a
4371 compatible subset of that provided by SCSH.
4373 Guile now recognizes a '-s' command line switch, whose argument is the
4374 name of a file of Scheme code to load. It also treats the two
4375 characters `#!' as the start of a comment, terminated by `!#'. Thus,
4376 to make a file of Scheme code directly executable by Unix, insert the
4377 following two lines at the top of the file:
4379 #!/usr/local/bin/guile -s
4382 Guile treats the argument of the `-s' command-line switch as the name
4383 of a file of Scheme code to load, and treats the sequence `#!' as the
4384 start of a block comment, terminated by `!#'.
4386 For example, here's a version of 'echo' written in Scheme:
4388 #!/usr/local/bin/guile -s
4390 (let loop ((args (cdr (program-arguments))))
4393 (display (car args))
4394 (if (pair? (cdr args))
4396 (loop (cdr args)))))
4399 Why does `#!' start a block comment terminated by `!#', instead of the
4400 end of the line? That is the notation SCSH uses, and although we
4401 don't yet support the other SCSH features that motivate that choice,
4402 we would like to be backward-compatible with any existing Guile
4403 scripts once we do. Furthermore, if the path to Guile on your system
4404 is too long for your kernel, you can start the script with this
4408 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
4411 Note that some very old Unix systems don't support the `#!' syntax.
4414 ** You can now run Guile without installing it.
4416 Previous versions of the interactive Guile interpreter (`guile')
4417 couldn't start up unless Guile's Scheme library had been installed;
4418 they used the value of the environment variable `SCHEME_LOAD_PATH'
4419 later on in the startup process, but not to find the startup code
4420 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
4423 To run Guile without installing it, build it in the normal way, and
4424 then set the environment variable `SCHEME_LOAD_PATH' to a
4425 colon-separated list of directories, including the top-level directory
4426 of the Guile sources. For example, if you unpacked Guile so that the
4427 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
4430 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
4433 ** Guile's read-eval-print loop no longer prints #<unspecified>
4434 results. If the user wants to see this, she can evaluate the
4435 expression (assert-repl-print-unspecified #t), perhaps in her startup
4438 ** Guile no longer shows backtraces by default when an error occurs;
4439 however, it does display a message saying how to get one, and how to
4440 request that they be displayed by default. After an error, evaluate
4442 to see a backtrace, and
4443 (debug-enable 'backtrace)
4444 to see them by default.
4448 * Changes to Guile Scheme:
4450 ** Guile now distinguishes between #f and the empty list.
4452 This is for compatibility with the IEEE standard, the (possibly)
4453 upcoming Revised^5 Report on Scheme, and many extant Scheme
4456 Guile used to have #f and '() denote the same object, to make Scheme's
4457 type system more compatible with Emacs Lisp's. However, the change
4458 caused too much trouble for Scheme programmers, and we found another
4459 way to reconcile Emacs Lisp with Scheme that didn't require this.
4462 ** Guile's delq, delv, delete functions, and their destructive
4463 counterparts, delq!, delv!, and delete!, now remove all matching
4464 elements from the list, not just the first. This matches the behavior
4465 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
4466 functions which inspired them.
4468 I recognize that this change may break code in subtle ways, but it
4469 seems best to make the change before the FSF's first Guile release,
4473 ** The compiled-library-path function has been deleted from libguile.
4475 ** The facilities for loading Scheme source files have changed.
4477 *** The variable %load-path now tells Guile which directories to search
4478 for Scheme code. Its value is a list of strings, each of which names
4481 *** The variable %load-extensions now tells Guile which extensions to
4482 try appending to a filename when searching the load path. Its value
4483 is a list of strings. Its default value is ("" ".scm").
4485 *** (%search-load-path FILENAME) searches the directories listed in the
4486 value of the %load-path variable for a Scheme file named FILENAME,
4487 with all the extensions listed in %load-extensions. If it finds a
4488 match, then it returns its full filename. If FILENAME is absolute, it
4489 returns it unchanged. Otherwise, it returns #f.
4491 %search-load-path will not return matches that refer to directories.
4493 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
4494 uses %seach-load-path to find a file named FILENAME, and loads it if
4495 it finds it. If it can't read FILENAME for any reason, it throws an
4498 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
4501 *** load uses the same searching semantics as primitive-load.
4503 *** The functions %try-load, try-load-with-path, %load, load-with-path,
4504 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
4505 path, and load-module-with-path have been deleted. The functions
4506 above should serve their purposes.
4508 *** If the value of the variable %load-hook is a procedure,
4509 `primitive-load' applies its value to the name of the file being
4510 loaded (without the load path directory name prepended). If its value
4511 is #f, it is ignored. Otherwise, an error occurs.
4513 This is mostly useful for printing load notification messages.
4516 ** The function `eval!' is no longer accessible from the scheme level.
4517 We can't allow operations which introduce glocs into the scheme level,
4518 because Guile's type system can't handle these as data. Use `eval' or
4519 `read-and-eval!' (see below) as replacement.
4521 ** The new function read-and-eval! reads an expression from PORT,
4522 evaluates it, and returns the result. This is more efficient than
4523 simply calling `read' and `eval', since it is not necessary to make a
4524 copy of the expression for the evaluator to munge.
4526 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
4527 for the `read' function.
4530 ** The function `int?' has been removed; its definition was identical
4531 to that of `integer?'.
4533 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
4534 use the R4RS names for these functions.
4536 ** The function object-properties no longer returns the hash handle;
4537 it simply returns the object's property list.
4539 ** Many functions have been changed to throw errors, instead of
4540 returning #f on failure. The point of providing exception handling in
4541 the language is to simplify the logic of user code, but this is less
4542 useful if Guile's primitives don't throw exceptions.
4544 ** The function `fileno' has been renamed from `%fileno'.
4546 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
4549 * Changes to Guile's C interface:
4551 ** The library's initialization procedure has been simplified.
4552 scm_boot_guile now has the prototype:
4554 void scm_boot_guile (int ARGC,
4556 void (*main_func) (),
4559 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
4560 MAIN_FUNC should do all the work of the program (initializing other
4561 packages, reading user input, etc.) before returning. When MAIN_FUNC
4562 returns, call exit (0); this function never returns. If you want some
4563 other exit value, MAIN_FUNC may call exit itself.
4565 scm_boot_guile arranges for program-arguments to return the strings
4566 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
4567 scm_set_program_arguments with the final list, so Scheme code will
4568 know which arguments have been processed.
4570 scm_boot_guile establishes a catch-all catch handler which prints an
4571 error message and exits the process. This means that Guile exits in a
4572 coherent way when system errors occur and the user isn't prepared to
4573 handle it. If the user doesn't like this behavior, they can establish
4574 their own universal catcher in MAIN_FUNC to shadow this one.
4576 Why must the caller do all the real work from MAIN_FUNC? The garbage
4577 collector assumes that all local variables of type SCM will be above
4578 scm_boot_guile's stack frame on the stack. If you try to manipulate
4579 SCM values after this function returns, it's the luck of the draw
4580 whether the GC will be able to find the objects you allocate. So,
4581 scm_boot_guile function exits, rather than returning, to discourage
4582 people from making that mistake.
4584 The IN, OUT, and ERR arguments were removed; there are other
4585 convenient ways to override these when desired.
4587 The RESULT argument was deleted; this function should never return.
4589 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
4593 ** Guile's header files should no longer conflict with your system's
4596 In order to compile code which #included <libguile.h>, previous
4597 versions of Guile required you to add a directory containing all the
4598 Guile header files to your #include path. This was a problem, since
4599 Guile's header files have names which conflict with many systems'
4602 Now only <libguile.h> need appear in your #include path; you must
4603 refer to all Guile's other header files as <libguile/mumble.h>.
4604 Guile's installation procedure puts libguile.h in $(includedir), and
4605 the rest in $(includedir)/libguile.
4608 ** Two new C functions, scm_protect_object and scm_unprotect_object,
4609 have been added to the Guile library.
4611 scm_protect_object (OBJ) protects OBJ from the garbage collector.
4612 OBJ will not be freed, even if all other references are dropped,
4613 until someone does scm_unprotect_object (OBJ). Both functions
4616 Note that calls to scm_protect_object do not nest. You can call
4617 scm_protect_object any number of times on a given object, and the
4618 next call to scm_unprotect_object will unprotect it completely.
4620 Basically, scm_protect_object and scm_unprotect_object just
4621 maintain a list of references to things. Since the GC knows about
4622 this list, all objects it mentions stay alive. scm_protect_object
4623 adds its argument to the list; scm_unprotect_object remove its
4624 argument from the list.
4627 ** scm_eval_0str now returns the value of the last expression
4630 ** The new function scm_read_0str reads an s-expression from a
4631 null-terminated string, and returns it.
4633 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
4634 to a Scheme port object.
4636 ** The new function `scm_set_program_arguments' allows C code to set
4637 the value returned by the Scheme `program-arguments' function.
4642 * Guile no longer includes sophisticated Tcl/Tk support.
4644 The old Tcl/Tk support was unsatisfying to us, because it required the
4645 user to link against the Tcl library, as well as Tk and Guile. The
4646 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
4647 referring to widgets by names, rather than exporting widgets to Scheme
4648 code as a special datatype.
4650 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
4651 maintainers described some very interesting changes in progress to the
4652 Tcl/Tk internals, which would facilitate clean interfaces between lone
4653 Tk and other interpreters --- even for garbage-collected languages
4654 like Scheme. They expected the new Tk to be publicly available in the
4657 Since it seems that Guile might soon have a new, cleaner interface to
4658 lone Tk, and that the old Guile/Tk glue code would probably need to be
4659 completely rewritten, we (Jim Blandy and Richard Stallman) have
4660 decided not to support the old code. We'll spend the time instead on
4661 a good interface to the newer Tk, as soon as it is available.
4663 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
4666 Copyright information:
4668 Copyright (C) 1996,1997 Free Software Foundation, Inc.
4670 Permission is granted to anyone to make or distribute verbatim copies
4671 of this document as received, in any medium, provided that the
4672 copyright notice and this permission notice are preserved,
4673 thus giving the recipient permission to redistribute in turn.
4675 Permission is granted to distribute modified versions
4676 of this document, or of portions of it,
4677 under the above conditions, provided also that they
4678 carry prominent notices stating who last changed them.
4683 paragraph-separate: "[
\f]*$"