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 1.4:
9 * Changes to the distribution
11 * Changes to the stand-alone interpreter
13 ** It's now possible to create modules with controlled environments
17 (use-modules (ice-9 safe))
18 (define m (make-safe-module))
19 ;;; m will now be a module containing only a safe subset of R5RS
20 (eval-in-module '(+ 1 2) m) --> 3
21 (eval-in-module 'load m) --> ERROR: Unbound variable: load
23 * Changes to Scheme functions and syntax
25 ** New define-module option: pure
27 Tells the module system not to include any bindings from the root
32 (define-module (totally-empty-module)
35 ** New define-module option: export NAME1 ...
37 Export names NAME1 ...
39 This option is required if you want to be able to export bindings from
40 a module which doesn't import one of `define-public' or `export'.
46 :use-module (ice-9 r5rs)
49 ;;; Note that we're pure R5RS below this point!
54 * Changes to the gh_ interface
56 * Changes to the scm_ interface
58 ** New function: scm_done_free (long size)
60 This function is the inverse of scm_done_malloc. Use it to report the
61 amount of smob memory you free. The previous method, which involved
62 calling scm_done_malloc with negative argument, was somewhat
63 unintuitive (and is still available, of course).
65 ** New global variable scm_gc_running_p introduced.
67 Use this variable to find out if garbage collection is being executed. Up to
68 now applications have used scm_gc_heap_lock to test if garbage collection was
69 running, which also works because of the fact that up to know only the garbage
70 collector has set this variable. But, this is an implementation detail that
71 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
72 of this variable is (and has been) not fully safe anyway.
74 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
75 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
76 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
77 SCM_ORD_SIG, SCM_NUM_SIGS
79 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
80 Use scm_memory_error instead of SCM_NALLOC.
82 ** Removed function: scm_struct_init
84 ** Deprecated function: scm_call_catching_errors
86 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
89 Changes since Guile 1.3.4:
91 * Changes to the distribution
93 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
95 We've changed the way we handle generated files in the Guile source
96 repository. As a result, the procedure for building trees obtained
97 from the nightly FTP snapshots or via CVS has changed:
98 - You must have appropriate versions of autoconf, automake, and
99 libtool installed on your system. See README for info on how to
100 obtain these programs.
101 - Before configuring the tree, you must first run the script
102 `autogen.sh' at the top of the source tree.
104 The Guile repository used to contain not only source files, written by
105 humans, but also some generated files, like configure scripts and
106 Makefile.in files. Even though the contents of these files could be
107 derived mechanically from other files present, we thought it would
108 make the tree easier to build if we checked them into CVS.
110 However, this approach means that minor differences between
111 developer's installed tools and habits affected the whole team.
112 So we have removed the generated files from the repository, and
113 added the autogen.sh script, which will reconstruct them
117 ** configure now has experimental options to remove support for certain
120 --disable-arrays omit array and uniform array support
121 --disable-posix omit posix interfaces
122 --disable-networking omit networking interfaces
123 --disable-regex omit regular expression interfaces
125 These are likely to become separate modules some day.
127 ** New configure option --enable-debug-freelist
129 This enables a debugging version of SCM_NEWCELL(), and also registers
130 an extra primitive, the setter `gc-set-debug-check-freelist!'.
132 Configure with the --enable-debug-freelist option to enable
133 the gc-set-debug-check-freelist! primitive, and then use:
135 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
136 (gc-set-debug-check-freelist! #f) # turn off checking
138 Checking of the freelist forces a traversal of the freelist and
139 a garbage collection before each allocation of a cell. This can
140 slow down the interpreter dramatically, so the setter should be used to
141 turn on this extra processing only when necessary.
143 ** New configure option --enable-debug-malloc
145 Include code for debugging of calls to scm_must_malloc/realloc/free.
149 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
150 2. objects reallocated by scm_must_realloc has been allocated by
152 3. reallocated objects are reallocated with the same what string
154 But, most importantly, it records the number of allocated objects of
155 each kind. This is useful when searching for memory leaks.
157 A Guile compiled with this option provides the primitive
158 `malloc-stats' which returns an alist with pairs of kind and the
159 number of objects of that kind.
161 ** All includes are now referenced relative to the root directory
163 Since some users have had problems with mixups between Guile and
164 system headers, we have decided to always refer to Guile headers via
165 their parent directories. This essentially creates a "private name
166 space" for Guile headers. This means that the compiler only is given
167 -I options for the root build and root source directory.
169 ** Header files kw.h and genio.h have been removed.
171 ** The module (ice-9 getopt-gnu-style) has been removed.
173 ** New module (ice-9 documentation)
175 Implements the interface to documentation strings associated with
178 * Changes to the stand-alone interpreter
180 ** New command line option --debug
182 Start Guile with debugging evaluator and backtraces enabled.
184 This is useful when debugging your .guile init file or scripts.
188 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
189 (help REGEXP) ditto for objects with names matching REGEXP (a string)
190 (help ,EXPR) gives documentation for object returned by EXPR
191 (help) gives this text
193 `help' searches among bindings exported from loaded modules, while
194 `apropos' searches among bindings visible from the "current" module.
196 Examples: (help help)
198 (help "output-string")
200 ** `help' and `apropos' now prints full module names
202 ** Dynamic linking now uses libltdl from the libtool package.
204 The old system dependent code for doing dynamic linking has been
205 replaced with calls to the libltdl functions which do all the hairy
208 The major improvement is that you can now directly pass libtool
209 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
210 will be able to do the best shared library job you can get, via
213 The way dynamic libraries are found has changed and is not really
214 portable across platforms, probably. It is therefore recommended to
215 use absolute filenames when possible.
217 If you pass a filename without an extension to `dynamic-link', it will
218 try a few appropriate ones. Thus, the most platform ignorant way is
219 to specify a name like "libfoo", without any directories and
222 ** Guile COOP threads are now compatible with LinuxThreads
224 Previously, COOP threading wasn't possible in applications linked with
225 Linux POSIX threads due to their use of the stack pointer to find the
226 thread context. This has now been fixed with a workaround which uses
227 the pthreads to allocate the stack.
229 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
231 ** Positions of erring expression in scripts
233 With version 1.3.4, the location of the erring expression in Guile
234 scipts is no longer automatically reported. (This should have been
235 documented before the 1.3.4 release.)
237 You can get this information by enabling recording of positions of
238 source expressions and running the debugging evaluator. Put this at
239 the top of your script (or in your "site" file):
241 (read-enable 'positions)
242 (debug-enable 'debug)
244 ** Backtraces in scripts
246 It is now possible to get backtraces in scripts.
250 (debug-enable 'debug 'backtrace)
252 at the top of the script.
254 (The first options enables the debugging evaluator.
255 The second enables backtraces.)
257 ** Part of module system symbol lookup now implemented in C
259 The eval closure of most modules is now implemented in C. Since this
260 was one of the bottlenecks for loading speed, Guile now loads code
261 substantially faster than before.
263 ** Attempting to get the value of an unbound variable now produces
264 an exception with a key of 'unbound-variable instead of 'misc-error.
266 ** The initial default output port is now unbuffered if it's using a
267 tty device. Previously in this situation it was line-buffered.
269 ** gc-thunk is deprecated
271 gc-thunk will be removed in next release of Guile. It has been
272 replaced by after-gc-hook.
274 ** New hook: after-gc-hook
276 after-gc-hook takes over the role of gc-thunk. This hook is run at
277 the first SCM_TICK after a GC. (Thus, the code is run at the same
278 point during evaluation as signal handlers.)
280 Note that this hook should be used only for diagnostic and debugging
281 purposes. It is not certain that it will continue to be well-defined
282 when this hook is run in the future.
284 C programmers: Note the new C level hooks scm_before_gc_c_hook,
285 scm_before_sweep_c_hook, scm_after_gc_c_hook.
287 ** Improvements to garbage collector
289 Guile 1.4 has a new policy for triggering heap allocation and
290 determining the sizes of heap segments. It fixes a number of problems
293 1. The new policy can handle two separate pools of cells
294 (2-word/4-word) better. (The old policy would run wild, allocating
295 more and more memory for certain programs.)
297 2. The old code would sometimes allocate far too much heap so that the
298 Guile process became gigantic. The new code avoids this.
300 3. The old code would sometimes allocate too little so that few cells
301 were freed at GC so that, in turn, too much time was spent in GC.
303 4. The old code would often trigger heap allocation several times in a
304 row. (The new scheme predicts how large the segments needs to be
305 in order not to need further allocation.)
307 All in all, the new GC policy will make larger applications more
310 The new GC scheme also is prepared for POSIX threading. Threads can
311 allocate private pools of cells ("clusters") with just a single
312 function call. Allocation of single cells from such a cluster can
313 then proceed without any need of inter-thread synchronization.
315 ** New environment variables controlling GC parameters
317 GUILE_MAX_SEGMENT_SIZE Maximal segment size
320 Allocation of 2-word cell heaps:
322 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
325 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
326 GC in percent of total heap size
329 Allocation of 4-word cell heaps
330 (used for real numbers and misc other objects):
332 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
334 (See entry "Way for application to customize GC parameters" under
335 section "Changes to the scm_ interface" below.)
337 ** Guile now implements reals using 4-word cells
339 This speeds up computation with reals. (They were earlier allocated
340 with `malloc'.) There is still some room for optimizations, however.
342 ** Some further steps toward POSIX thread support have been taken
344 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
345 don't have much effect any longer, and many of them will be removed in
349 are only handled at the top of the evaluator loop, immediately after
350 I/O, and in scm_equalp.
352 *** The GC can allocate thread private pools of pairs.
354 * Changes to Scheme functions and syntax
356 ** close-input-port and close-output-port are now R5RS
358 These procedures have been turned into primitives and have R5RS behaviour.
360 ** New procedure: simple-format PORT MESSAGE ARG1 ...
362 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
363 extended by the more sophisticated version in (ice-9 format)
365 (simple-format port message . args)
366 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
367 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
368 the escapes are replaced with corresponding members of ARGS:
369 ~A formats using `display' and ~S formats using `write'.
370 If DESTINATION is #t, then use the `current-output-port',
371 if DESTINATION is #f, then return a string containing the formatted text.
372 Does not add a trailing newline."
374 ** string-ref: the second argument is no longer optional.
376 ** string, list->string: no longer accept strings in their arguments,
377 only characters, for compatibility with R5RS.
379 ** New procedure: port-closed? PORT
380 Returns #t if PORT is closed or #f if it is open.
384 The list* functionality is now provided by cons* (SRFI-1 compliant)
386 ** New procedure: cons* ARG1 ARG2 ... ARGn
388 Like `list', but the last arg provides the tail of the constructed list,
389 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
391 Requires at least one argument. If given one argument, that argument
392 is returned as result.
394 This function is called `list*' in some other Schemes and in Common LISP.
396 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
398 ** New procedure: object-documentation OBJECT
400 Returns the documentation string associated with OBJECT. The
401 procedure uses a caching mechanism so that subsequent lookups are
404 Exported by (ice-9 documentation).
406 ** module-name now returns full names of modules
408 Previously, only the last part of the name was returned (`session' for
409 `(ice-9 session)'). Ex: `(ice-9 session)'.
411 * Changes to the gh_ interface
413 ** Deprecated: gh_int2scmb
415 Use gh_bool2scm instead.
417 * Changes to the scm_ interface
419 ** Guile primitives now carry docstrings!
421 Thanks to Greg Badros!
423 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
425 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
426 macros and must contain a docstring that is extracted into foo.doc using a new
427 guile-doc-snarf script (that uses guile-doc-snarf.awk).
429 However, a major overhaul of these macros is scheduled for the next release of
432 ** Guile primitives use a new technique for validation of arguments
434 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
435 the readability of argument checking.
437 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
439 ** New macros: SCM_PACK, SCM_UNPACK
441 Compose/decompose an SCM value.
443 The SCM type is now treated as an abstract data type and may be defined as a
444 long, a void* or as a struct, depending on the architecture and compile time
445 options. This makes it easier to find several types of bugs, for example when
446 SCM values are treated as integers without conversion. Values of the SCM type
447 should be treated as "atomic" values. These macros are used when
448 composing/decomposing an SCM value, either because you want to access
449 individual bits, or because you want to treat it as an integer value.
451 E.g., in order to set bit 7 in an SCM value x, use the expression
453 SCM_PACK (SCM_UNPACK (x) | 0x80)
455 ** The name property of hooks is deprecated.
456 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
458 You can emulate this feature by using object properties.
460 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
461 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
464 These macros will be removed in a future release of Guile.
466 ** The following types, functions and macros from numbers.h are deprecated:
467 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
468 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
470 Further, it is recommended not to rely on implementation details for guile's
471 current implementation of bignums. It is planned to replace this
472 implementation with gmp in the future.
474 ** Port internals: the rw_random variable in the scm_port structure
475 must be set to non-zero in any random access port. In recent Guile
476 releases it was only set for bidirectional random-access ports.
478 ** Port internals: the seek ptob procedure is now responsible for
479 resetting the buffers if required. The change was made so that in the
480 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
481 the fport and strport ptobs can avoid resetting the buffers,
482 in particular to avoid discarding unread chars. An existing port
483 type can be fixed by adding something like the following to the
484 beginning of the ptob seek procedure:
486 if (pt->rw_active == SCM_PORT_READ)
487 scm_end_input (object);
488 else if (pt->rw_active == SCM_PORT_WRITE)
489 ptob->flush (object);
491 although to actually avoid resetting the buffers and discard unread
492 chars requires further hacking that depends on the characteristics
495 ** Deprecated functions: scm_fseek, scm_tag
497 These functions are no longer used and will be removed in a future version.
499 ** The scm_sysmissing procedure is no longer used in libguile.
500 Unless it turns out to be unexpectedly useful to somebody, it will be
501 removed in a future version.
503 ** The format of error message strings has changed
505 The two C procedures: scm_display_error and scm_error, as well as the
506 primitive `scm-error', now use scm_simple_format to do their work.
507 This means that the message strings of all code must be updated to use
508 ~A where %s was used before, and ~S where %S was used before.
510 During the period when there still are a lot of old Guiles out there,
511 you might want to support both old and new versions of Guile.
513 There are basically two methods to achieve this. Both methods use
516 AC_CHECK_FUNCS(scm_simple_format)
518 in your configure.in.
520 Method 1: Use the string concatenation features of ANSI C's
525 #ifdef HAVE_SCM_SIMPLE_FORMAT
531 Then represent each of your error messages using a preprocessor macro:
533 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
537 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
538 (define make-message string-append)
540 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
542 Method 2: Use the oldfmt function found in doc/oldfmt.c.
546 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
551 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
555 ** Deprecated: coop_mutex_init, coop_condition_variable_init
557 Don't use the functions coop_mutex_init and
558 coop_condition_variable_init. They will change.
560 Use scm_mutex_init and scm_cond_init instead.
562 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
563 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
564 COND, as `scm_cond_wait' does, but it also bounds the duration
565 of the wait. If COND has not been signaled before time ABSTIME,
566 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
567 returns the error code `ETIMEDOUT'.
569 The ABSTIME parameter specifies an absolute time, with the same
570 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
571 to 00:00:00 GMT, January 1, 1970.
573 ** New function: scm_cond_broadcast (scm_cond_t *COND)
574 `scm_cond_broadcast' restarts all the threads that are waiting
575 on the condition variable COND. Nothing happens if no threads are
578 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
579 `scm_key_create' allocates a new TSD key. The key is stored in
580 the location pointed to by KEY. There is no limit on the number
581 of keys allocated at a given time. The value initially associated
582 with the returned key is `NULL' in all currently executing threads.
584 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
585 function associated with the key. When a thread terminates,
586 DESTR_FUNCTION is called on the value associated with the key in
587 that thread. The DESTR_FUNCTION is not called if a key is deleted
588 with `scm_key_delete' or a value is changed with
589 `scm_setspecific'. The order in which destructor functions are
590 called at thread termination time is unspecified.
592 Destructors are not yet implemented.
594 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
595 `scm_setspecific' changes the value associated with KEY in the
596 calling thread, storing the given POINTER instead.
598 ** New function: scm_getspecific (scm_key_t KEY)
599 `scm_getspecific' returns the value currently associated with
600 KEY in the calling thread.
602 ** New function: scm_key_delete (scm_key_t KEY)
603 `scm_key_delete' deallocates a TSD key. It does not check
604 whether non-`NULL' values are associated with that key in the
605 currently executing threads, nor call the destructor function
606 associated with the key.
608 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
610 Initialize a C level hook HOOK with associated HOOK_DATA and type
611 TYPE. (See scm_c_hook_run ().)
613 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
615 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
616 is true, add it last, otherwise first. The same FUNC can be added
617 multiple times if FUNC_DATA differ and vice versa.
619 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
621 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
622 function is only removed if both FUNC and FUNC_DATA matches.
624 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
626 Run hook HOOK passing DATA to the hook functions.
628 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
629 returned is undefined.
631 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
632 returns a non-NULL value. This value is returned as the result of
633 scm_c_hook_run. If all functions return NULL, NULL is returned.
635 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
636 returns a NULL value, and NULL is returned. If all functions returns
637 a non-NULL value, the last value is returned.
639 ** New C level GC hooks
641 Five new C level hooks has been added to the garbage collector.
646 are run before locking and after unlocking the heap. The system is
647 thus in a mode where evaluation can take place. (Except that
648 scm_before_gc_c_hook must not allocate new cells.)
650 scm_before_mark_c_hook
651 scm_before_sweep_c_hook
652 scm_after_sweep_c_hook
654 are run when the heap is locked. These are intended for extension of
655 the GC in a modular fashion. Examples are the weaks and guardians
658 ** Way for application to customize GC parameters
660 The application can set up other default values for the GC heap
661 allocation parameters
663 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
664 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
665 GUILE_MAX_SEGMENT_SIZE,
669 scm_default_init_heap_size_1, scm_default_min_yield_1,
670 scm_default_init_heap_size_2, scm_default_min_yield_2,
671 scm_default_max_segment_size
673 respectively before callong scm_boot_guile.
675 (See entry "New environment variables ..." in section
676 "Changes to the stand-alone interpreter" above.)
678 ** scm_protect_object/scm_unprotect_object now nest
680 This means that you can call scm_protect_object multiple times on an
681 object and count on the object being protected until
682 scm_unprotect_object has been call the same number of times.
684 The functions also have better time complexity.
686 Still, it is usually possible to structure the application in a way
687 that you don't need to use these functions. For example, if you use a
688 protected standard Guile list to keep track of live objects rather
689 than some custom data type, objects will die a natural death when they
690 are no longer needed.
692 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
694 Guile does not provide the float representation for inexact real numbers any
695 more. Now, only doubles are used to represent inexact real numbers. Further,
696 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
697 and scm_tc16_complex, respectively.
699 ** Removed deprecated type scm_smobfuns
701 ** Removed deprecated function scm_newsmob
703 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
705 There is an ongoing discussion among the developers whether to
706 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
707 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
708 until this issue has been settled.
710 ** Removed deprecated type tag scm_tc16_kw
712 ** Added type tag scm_tc16_keyword
714 (This was introduced already in release 1.3.4 but was not documented
717 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
719 * Changes to system call interfaces:
721 ** The "select" procedure now tests port buffers for the ability to
722 provide input or accept output. Previously only the underlying file
723 descriptors were checked.
725 ** New variable PIPE_BUF: the maximum number of bytes that can be
726 atomically written to a pipe.
728 ** If a facility is not available on the system when Guile is
729 compiled, the corresponding primitive procedure will not be defined.
730 Previously it would have been defined but would throw a system-error
731 exception if called. Exception handlers which catch this case may
732 need minor modification: an error will be thrown with key
733 'unbound-variable instead of 'system-error. Alternatively it's
734 now possible to use `defined?' to check whether the facility is
737 ** Procedures which depend on the timezone should now give the correct
738 result on systems which cache the TZ environment variable, even if TZ
739 is changed without calling tzset.
741 * Changes to the networking interfaces:
743 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
744 long integers between network and host format. For now, it's not
745 particularly convenient to do this kind of thing, but consider:
747 (define write-network-long
749 (let ((v (make-uniform-vector 1 1 0)))
750 (uniform-vector-set! v 0 (htonl value))
751 (uniform-vector-write v port))))
753 (define read-network-long
755 (let ((v (make-uniform-vector 1 1 0)))
756 (uniform-vector-read! v port)
757 (ntohl (uniform-vector-ref v 0)))))
759 ** If inet-aton fails, it now throws an error with key 'misc-error
760 instead of 'system-error, since errno is not relevant.
762 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
763 specific keys instead of 'system-error. The latter is inappropriate
764 since errno will not have been set. The keys are:
765 'host-not-found, 'try-again, 'no-recovery and 'no-data.
767 ** sethostent, setnetent, setprotoent, setservent: now take an
768 optional argument STAYOPEN, which specifies whether the database
769 remains open after a database entry is accessed randomly (e.g., using
770 gethostbyname for the hosts database.) The default is #f. Previously
774 Changes since Guile 1.3.2:
776 * Changes to the stand-alone interpreter
780 An initial version of the Guile debugger written by Chris Hanson has
781 been added. The debugger is still under development but is included
782 in the distribution anyway since it is already quite useful.
788 after an error to enter the debugger. Type `help' inside the debugger
789 for a description of available commands.
791 If you prefer to have stack frames numbered and printed in
792 anti-chronological order and prefer up in the stack to be down on the
793 screen as is the case in gdb, you can put
795 (debug-enable 'backwards)
797 in your .guile startup file. (However, this means that Guile can't
798 use indentation to indicate stack level.)
800 The debugger is autoloaded into Guile at the first use.
802 ** Further enhancements to backtraces
804 There is a new debug option `width' which controls the maximum width
805 on the screen of printed stack frames. Fancy printing parameters
806 ("level" and "length" as in Common LISP) are adaptively adjusted for
807 each stack frame to give maximum information while still fitting
808 within the bounds. If the stack frame can't be made to fit by
809 adjusting parameters, it is simply cut off at the end. This is marked
812 ** Some modules are now only loaded when the repl is started
814 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
815 regex) are now loaded into (guile-user) only if the repl has been
816 started. The effect is that the startup time for scripts has been
817 reduced to 30% of what it was previously.
819 Correctly written scripts load the modules they require at the top of
820 the file and should not be affected by this change.
822 ** Hooks are now represented as smobs
824 * Changes to Scheme functions and syntax
826 ** Readline support has changed again.
828 The old (readline-activator) module is gone. Use (ice-9 readline)
829 instead, which now contains all readline functionality. So the code
830 to activate readline is now
832 (use-modules (ice-9 readline))
835 This should work at any time, including from the guile prompt.
837 To avoid confusion about the terms of Guile's license, please only
838 enable readline for your personal use; please don't make it the
839 default for others. Here is why we make this rather odd-sounding
842 Guile is normally licensed under a weakened form of the GNU General
843 Public License, which allows you to link code with Guile without
844 placing that code under the GPL. This exception is important to some
847 However, since readline is distributed under the GNU General Public
848 License, when you link Guile with readline, either statically or
849 dynamically, you effectively change Guile's license to the strict GPL.
850 Whenever you link any strictly GPL'd code into Guile, uses of Guile
851 which are normally permitted become forbidden. This is a rather
852 non-obvious consequence of the licensing terms.
854 So, to make sure things remain clear, please let people choose for
855 themselves whether to link GPL'd libraries like readline with Guile.
857 ** regexp-substitute/global has changed slightly, but incompatibly.
859 If you include a function in the item list, the string of the match
860 object it receives is the same string passed to
861 regexp-substitute/global, not some suffix of that string.
862 Correspondingly, the match's positions are relative to the entire
863 string, not the suffix.
865 If the regexp can match the empty string, the way matches are chosen
866 from the string has changed. regexp-substitute/global recognizes the
867 same set of matches that list-matches does; see below.
869 ** New function: list-matches REGEXP STRING [FLAGS]
871 Return a list of match objects, one for every non-overlapping, maximal
872 match of REGEXP in STRING. The matches appear in left-to-right order.
873 list-matches only reports matches of the empty string if there are no
874 other matches which begin on, end at, or include the empty match's
877 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
879 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
881 For each match of REGEXP in STRING, apply PROC to the match object,
882 and the last value PROC returned, or INIT for the first call. Return
883 the last value returned by PROC. We apply PROC to the matches as they
884 appear from left to right.
886 This function recognizes matches according to the same criteria as
889 Thus, you could define list-matches like this:
891 (define (list-matches regexp string . flags)
892 (reverse! (apply fold-matches regexp string '() cons flags)))
894 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
898 *** New function: hook? OBJ
900 Return #t if OBJ is a hook, otherwise #f.
902 *** New function: make-hook-with-name NAME [ARITY]
904 Return a hook with name NAME and arity ARITY. The default value for
905 ARITY is 0. The only effect of NAME is that it will appear when the
906 hook object is printed to ease debugging.
908 *** New function: hook-empty? HOOK
910 Return #t if HOOK doesn't contain any procedures, otherwise #f.
912 *** New function: hook->list HOOK
914 Return a list of the procedures that are called when run-hook is
917 ** `map' signals an error if its argument lists are not all the same length.
919 This is the behavior required by R5RS, so this change is really a bug
920 fix. But it seems to affect a lot of people's code, so we're
921 mentioning it here anyway.
923 ** Print-state handling has been made more transparent
925 Under certain circumstances, ports are represented as a port with an
926 associated print state. Earlier, this pair was represented as a pair
927 (see "Some magic has been added to the printer" below). It is now
928 indistinguishable (almost; see `get-print-state') from a port on the
931 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
933 Return a new port with the associated print state PRINT-STATE.
935 *** New function: get-print-state OUTPUT-PORT
937 Return the print state associated with this port if it exists,
940 *** New function: directory-stream? OBJECT
942 Returns true iff OBJECT is a directory stream --- the sort of object
943 returned by `opendir'.
945 ** New function: using-readline?
947 Return #t if readline is in use in the current repl.
949 ** structs will be removed in 1.4
951 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
952 and use GOOPS objects as the fundamental record type.
954 * Changes to the scm_ interface
956 ** structs will be removed in 1.4
958 The entire current struct interface (struct.c, struct.h) will be
959 replaced in Guile 1.4. We will merge GOOPS into libguile and use
960 GOOPS objects as the fundamental record type.
962 ** The internal representation of subr's has changed
964 Instead of giving a hint to the subr name, the CAR field of the subr
965 now contains an index to a subr entry in scm_subr_table.
967 *** New variable: scm_subr_table
969 An array of subr entries. A subr entry contains the name, properties
970 and documentation associated with the subr. The properties and
971 documentation slots are not yet used.
973 ** A new scheme for "forwarding" calls to a builtin to a generic function
975 It is now possible to extend the functionality of some Guile
976 primitives by letting them defer a call to a GOOPS generic function on
977 argument mismatch. This means that there is no loss of efficiency in
982 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
983 (define-method + ((x <string>) (y <string>))
986 + will still be as efficient as usual in numerical calculations, but
987 can also be used for concatenating strings.
989 Who will be the first one to extend Guile's numerical tower to
990 rationals? :) [OK, there a few other things to fix before this can
991 be made in a clean way.]
993 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
995 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
997 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
999 These do the same job as SCM_PROC and SCM_PROC1, but they also define
1000 a variable GENERIC which can be used by the dispatch macros below.
1002 [This is experimental code which may change soon.]
1004 *** New macros for forwarding control to a generic on arg type error
1006 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
1008 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
1010 These correspond to the scm_wta function call, and have the same
1011 behaviour until the user has called the GOOPS primitive
1012 `enable-primitive-generic!'. After that, these macros will apply the
1013 generic function GENERIC to the argument(s) instead of calling
1016 [This is experimental code which may change soon.]
1018 *** New macros for argument testing with generic dispatch
1020 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
1022 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
1024 These correspond to the SCM_ASSERT macro, but will defer control to
1025 GENERIC on error after `enable-primitive-generic!' has been called.
1027 [This is experimental code which may change soon.]
1029 ** New function: SCM scm_eval_body (SCM body, SCM env)
1031 Evaluates the body of a special form.
1033 ** The internal representation of struct's has changed
1035 Previously, four slots were allocated for the procedure(s) of entities
1036 and operators. The motivation for this representation had to do with
1037 the structure of the evaluator, the wish to support tail-recursive
1038 generic functions, and efficiency. Since the generic function
1039 dispatch mechanism has changed, there is no longer a need for such an
1040 expensive representation, and the representation has been simplified.
1042 This should not make any difference for most users.
1044 ** GOOPS support has been cleaned up.
1046 Some code has been moved from eval.c to objects.c and code in both of
1047 these compilation units has been cleaned up and better structured.
1049 *** New functions for applying generic functions
1051 New function: SCM scm_apply_generic (GENERIC, ARGS)
1052 New function: SCM scm_call_generic_0 (GENERIC)
1053 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
1054 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
1055 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
1057 ** Deprecated function: scm_make_named_hook
1059 It is now replaced by:
1061 ** New function: SCM scm_create_hook (const char *name, int arity)
1063 Creates a hook in the same way as make-hook above but also
1064 binds a variable named NAME to it.
1066 This is the typical way of creating a hook from C code.
1068 Currently, the variable is created in the "current" module.
1069 This might change when we get the new module system.
1071 [The behaviour is identical to scm_make_named_hook.]
1075 Changes since Guile 1.3:
1077 * Changes to mailing lists
1079 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
1081 See the README file to find current addresses for all the Guile
1084 * Changes to the distribution
1086 ** Readline support is no longer included with Guile by default.
1088 Based on the different license terms of Guile and Readline, we
1089 concluded that Guile should not *by default* cause the linking of
1090 Readline into an application program. Readline support is now offered
1091 as a separate module, which is linked into an application only when
1092 you explicitly specify it.
1094 Although Guile is GNU software, its distribution terms add a special
1095 exception to the usual GNU General Public License (GPL). Guile's
1096 license includes a clause that allows you to link Guile with non-free
1097 programs. We add this exception so as not to put Guile at a
1098 disadvantage vis-a-vis other extensibility packages that support other
1101 In contrast, the GNU Readline library is distributed under the GNU
1102 General Public License pure and simple. This means that you may not
1103 link Readline, even dynamically, into an application unless it is
1104 distributed under a free software license that is compatible the GPL.
1106 Because of this difference in distribution terms, an application that
1107 can use Guile may not be able to use Readline. Now users will be
1108 explicitly offered two independent decisions about the use of these
1111 You can activate the readline support by issuing
1113 (use-modules (readline-activator))
1116 from your ".guile" file, for example.
1118 * Changes to the stand-alone interpreter
1120 ** All builtins now print as primitives.
1121 Previously builtin procedures not belonging to the fundamental subr
1122 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
1123 Now, they print as #<primitive-procedure NAME>.
1125 ** Backtraces slightly more intelligible.
1126 gsubr-apply and macro transformer application frames no longer appear
1129 * Changes to Scheme functions and syntax
1131 ** Guile now correctly handles internal defines by rewriting them into
1132 their equivalent letrec. Previously, internal defines would
1133 incrementally add to the innermost environment, without checking
1134 whether the restrictions specified in RnRS were met. This lead to the
1135 correct behaviour when these restriction actually were met, but didn't
1136 catch all illegal uses. Such an illegal use could lead to crashes of
1137 the Guile interpreter or or other unwanted results. An example of
1138 incorrect internal defines that made Guile behave erratically:
1150 The problem with this example is that the definition of `c' uses the
1151 value of `b' directly. This confuses the meoization machine of Guile
1152 so that the second call of `b' (this time in a larger environment that
1153 also contains bindings for `c' and `d') refers to the binding of `c'
1154 instead of `a'. You could also make Guile crash with a variation on
1159 (define (b flag) (if flag a 1))
1160 (define c (1+ (b flag)))
1168 From now on, Guile will issue an `Unbound variable: b' error message
1173 A hook contains a list of functions which should be called on
1174 particular occasions in an existing program. Hooks are used for
1177 A window manager might have a hook before-window-map-hook. The window
1178 manager uses the function run-hooks to call all functions stored in
1179 before-window-map-hook each time a window is mapped. The user can
1180 store functions in the hook using add-hook!.
1182 In Guile, hooks are first class objects.
1184 *** New function: make-hook [N_ARGS]
1186 Return a hook for hook functions which can take N_ARGS arguments.
1187 The default value for N_ARGS is 0.
1189 (See also scm_make_named_hook below.)
1191 *** New function: add-hook! HOOK PROC [APPEND_P]
1193 Put PROC at the beginning of the list of functions stored in HOOK.
1194 If APPEND_P is supplied, and non-false, put PROC at the end instead.
1196 PROC must be able to take the number of arguments specified when the
1199 If PROC already exists in HOOK, then remove it first.
1201 *** New function: remove-hook! HOOK PROC
1203 Remove PROC from the list of functions in HOOK.
1205 *** New function: reset-hook! HOOK
1207 Clear the list of hook functions stored in HOOK.
1209 *** New function: run-hook HOOK ARG1 ...
1211 Run all hook functions stored in HOOK with arguments ARG1 ... .
1212 The number of arguments supplied must correspond to the number given
1213 when the hook was created.
1215 ** The function `dynamic-link' now takes optional keyword arguments.
1216 The only keyword argument that is currently defined is `:global
1217 BOOL'. With it, you can control whether the shared library will be
1218 linked in global mode or not. In global mode, the symbols from the
1219 linked library can be used to resolve references from other
1220 dynamically linked libraries. In non-global mode, the linked
1221 library is essentially invisible and can only be accessed via
1222 `dynamic-func', etc. The default is now to link in global mode.
1223 Previously, the default has been non-global mode.
1225 The `#:global' keyword is only effective on platforms that support
1226 the dlopen family of functions.
1228 ** New function `provided?'
1230 - Function: provided? FEATURE
1231 Return true iff FEATURE is supported by this installation of
1232 Guile. FEATURE must be a symbol naming a feature; the global
1233 variable `*features*' is a list of available features.
1235 ** Changes to the module (ice-9 expect):
1237 *** The expect-strings macro now matches `$' in a regular expression
1238 only at a line-break or end-of-file by default. Previously it would
1239 match the end of the string accumulated so far. The old behaviour
1240 can be obtained by setting the variable `expect-strings-exec-flags'
1243 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
1244 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
1245 in a regular expression will still match before a line-break or
1246 end-of-file. The default is `regexp/noteol'.
1248 *** The expect-strings macro now uses a variable
1249 `expect-strings-compile-flags' for the flags to be supplied to
1250 `make-regexp'. The default is `regexp/newline', which was previously
1253 *** The expect macro now supplies two arguments to a match procedure:
1254 the current accumulated string and a flag to indicate whether
1255 end-of-file has been reached. Previously only the string was supplied.
1256 If end-of-file is reached, the match procedure will be called an
1257 additional time with the same accumulated string as the previous call
1258 but with the flag set.
1260 ** New module (ice-9 format), implementing the Common Lisp `format' function.
1262 This code, and the documentation for it that appears here, was
1263 borrowed from SLIB, with minor adaptations for Guile.
1265 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
1266 An almost complete implementation of Common LISP format description
1267 according to the CL reference book `Common LISP' from Guy L.
1268 Steele, Digital Press. Backward compatible to most of the
1269 available Scheme format implementations.
1271 Returns `#t', `#f' or a string; has side effect of printing
1272 according to FORMAT-STRING. If DESTINATION is `#t', the output is
1273 to the current output port and `#t' is returned. If DESTINATION
1274 is `#f', a formatted string is returned as the result of the call.
1275 NEW: If DESTINATION is a string, DESTINATION is regarded as the
1276 format string; FORMAT-STRING is then the first argument and the
1277 output is returned as a string. If DESTINATION is a number, the
1278 output is to the current error port if available by the
1279 implementation. Otherwise DESTINATION must be an output port and
1282 FORMAT-STRING must be a string. In case of a formatting error
1283 format returns `#f' and prints a message on the current output or
1284 error port. Characters are output as if the string were output by
1285 the `display' function with the exception of those prefixed by a
1286 tilde (~). For a detailed description of the FORMAT-STRING syntax
1287 please consult a Common LISP format reference manual. For a test
1288 suite to verify this format implementation load `formatst.scm'.
1289 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
1291 Note: `format' is not reentrant, i.e. only one `format'-call may
1292 be executed at a time.
1295 *** Format Specification (Format version 3.0)
1297 Please consult a Common LISP format reference manual for a detailed
1298 description of the format string syntax. For a demonstration of the
1299 implemented directives see `formatst.scm'.
1301 This implementation supports directive parameters and modifiers (`:'
1302 and `@' characters). Multiple parameters must be separated by a comma
1303 (`,'). Parameters can be numerical parameters (positive or negative),
1304 character parameters (prefixed by a quote character (`''), variable
1305 parameters (`v'), number of rest arguments parameter (`#'), empty and
1306 default parameters. Directive characters are case independent. The
1307 general form of a directive is:
1309 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
1311 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
1313 *** Implemented CL Format Control Directives
1315 Documentation syntax: Uppercase characters represent the
1316 corresponding control directive characters. Lowercase characters
1317 represent control directive parameter descriptions.
1320 Any (print as `display' does).
1324 `~MINCOL,COLINC,MINPAD,PADCHARA'
1328 S-expression (print as `write' does).
1332 `~MINCOL,COLINC,MINPAD,PADCHARS'
1338 print number sign always.
1341 print comma separated.
1343 `~MINCOL,PADCHAR,COMMACHARD'
1349 print number sign always.
1352 print comma separated.
1354 `~MINCOL,PADCHAR,COMMACHARX'
1360 print number sign always.
1363 print comma separated.
1365 `~MINCOL,PADCHAR,COMMACHARO'
1371 print number sign always.
1374 print comma separated.
1376 `~MINCOL,PADCHAR,COMMACHARB'
1381 `~N,MINCOL,PADCHAR,COMMACHARR'
1385 print a number as a Roman numeral.
1388 print a number as an "old fashioned" Roman numeral.
1391 print a number as an ordinal English number.
1394 print a number as a cardinal English number.
1399 prints `y' and `ies'.
1402 as `~P but jumps 1 argument backward.'
1405 as `~@P but jumps 1 argument backward.'
1410 prints a character as the reader can understand it (i.e. `#\'
1414 prints a character as emacs does (eg. `^C' for ASCII 03).
1417 Fixed-format floating-point (prints a flonum like MMM.NNN).
1418 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
1420 If the number is positive a plus sign is printed.
1423 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
1424 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
1426 If the number is positive a plus sign is printed.
1429 General floating-point (prints a flonum either fixed or
1431 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
1433 If the number is positive a plus sign is printed.
1436 Dollars floating-point (prints a flonum in fixed with signs
1438 `~DIGITS,SCALE,WIDTH,PADCHAR$'
1440 If the number is positive a plus sign is printed.
1443 A sign is always printed and appears before the padding.
1446 The sign appears before the padding.
1454 print newline if not at the beginning of the output line.
1456 prints `~&' and then N-1 newlines.
1461 print N page separators.
1471 newline is ignored, white space left.
1474 newline is left, white space ignored.
1479 relative tabulation.
1485 Indirection (expects indirect arguments as a list).
1487 extracts indirect arguments from format arguments.
1490 Case conversion (converts by `string-downcase').
1492 converts by `string-capitalize'.
1495 converts by `string-capitalize-first'.
1498 converts by `string-upcase'.
1501 Argument Jumping (jumps 1 argument forward).
1503 jumps N arguments forward.
1506 jumps 1 argument backward.
1509 jumps N arguments backward.
1512 jumps to the 0th argument.
1515 jumps to the Nth argument (beginning from 0)
1517 `~[STR0~;STR1~;...~;STRN~]'
1518 Conditional Expression (numerical clause conditional).
1520 take argument from N.
1523 true test conditional.
1526 if-else-then conditional.
1532 default clause follows.
1535 Iteration (args come from the next argument (a list)).
1537 at most N iterations.
1540 args from next arg (a list of lists).
1543 args from the rest of arguments.
1546 args from the rest args (lists).
1557 aborts if N <= M <= K
1559 *** Not Implemented CL Format Control Directives
1562 print `#f' as an empty list (see below).
1565 print `#f' as an empty list (see below).
1571 (sorry I don't understand its semantics completely)
1573 *** Extended, Replaced and Additional Control Directives
1575 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
1576 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
1577 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
1578 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
1579 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
1580 COMMAWIDTH is the number of characters between two comma
1584 print a R4RS complex number as `~F~@Fi' with passed parameters for
1588 Pretty print formatting of an argument for scheme code lists.
1594 Flushes the output if format DESTINATION is a port.
1597 Print a `#\space' character
1599 print N `#\space' characters.
1602 Print a `#\tab' character
1604 print N `#\tab' characters.
1607 Takes N as an integer representation for a character. No arguments
1608 are consumed. N is converted to a character by `integer->char'. N
1609 must be a positive decimal number.
1612 Print out readproof. Prints out internal objects represented as
1613 `#<...>' as strings `"#<...>"' so that the format output can always
1614 be processed by `read'.
1617 Print out readproof. Prints out internal objects represented as
1618 `#<...>' as strings `"#<...>"' so that the format output can always
1619 be processed by `read'.
1622 Prints information and a copyright notice on the format
1625 prints format version.
1628 may also print number strings, i.e. passing a number as a string
1629 and format it accordingly.
1631 *** Configuration Variables
1633 The format module exports some configuration variables to suit the
1634 systems and users needs. There should be no modification necessary for
1635 the configuration that comes with Guile. Format detects automatically
1636 if the running scheme system implements floating point numbers and
1639 format:symbol-case-conv
1640 Symbols are converted by `symbol->string' so the case type of the
1641 printed symbols is implementation dependent.
1642 `format:symbol-case-conv' is a one arg closure which is either
1643 `#f' (no conversion), `string-upcase', `string-downcase' or
1644 `string-capitalize'. (default `#f')
1646 format:iobj-case-conv
1647 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
1648 implementation internal objects. (default `#f')
1651 The character prefixing the exponent value in `~E' printing.
1654 *** Compatibility With Other Format Implementations
1660 Downward compatible except for padding support and `~A', `~S',
1661 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
1662 `printf' padding support which is completely replaced by the CL
1663 `format' padding style.
1666 Downward compatible except for `~', which is not documented
1667 (ignores all characters inside the format string up to a newline
1668 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
1669 numerical and variable parameters and `:/@' modifiers in the CL
1673 Downward compatible except for `~A' and `~S' which print in
1674 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
1675 directive parameters or modifiers)).
1678 Downward compatible except for an optional destination parameter:
1679 S2C accepts a format call without a destination which returns a
1680 formatted string. This is equivalent to a #f destination in S2C.
1681 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
1682 parameters or modifiers)).
1685 ** Changes to string-handling functions.
1687 These functions were added to support the (ice-9 format) module, above.
1689 *** New function: string-upcase STRING
1690 *** New function: string-downcase STRING
1692 These are non-destructive versions of the existing string-upcase! and
1693 string-downcase! functions.
1695 *** New function: string-capitalize! STRING
1696 *** New function: string-capitalize STRING
1698 These functions convert the first letter of each word in the string to
1701 (string-capitalize "howdy there")
1704 As with the other functions, string-capitalize! modifies the string in
1705 place, while string-capitalize returns a modified copy of its argument.
1707 *** New function: string-ci->symbol STRING
1709 Return a symbol whose name is STRING, but having the same case as if
1710 the symbol had be read by `read'.
1712 Guile can be configured to be sensitive or insensitive to case
1713 differences in Scheme identifiers. If Guile is case-insensitive, all
1714 symbols are converted to lower case on input. The `string-ci->symbol'
1715 function returns a symbol whose name in STRING, transformed as Guile
1716 would if STRING were input.
1718 *** New function: substring-move! STRING1 START END STRING2 START
1720 Copy the substring of STRING1 from START (inclusive) to END
1721 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
1722 string, and the source and destination areas may overlap; in all
1723 cases, the function behaves as if all the characters were copied
1726 *** Extended functions: substring-move-left! substring-move-right!
1728 These functions now correctly copy arbitrarily overlapping substrings;
1729 they are both synonyms for substring-move!.
1732 ** New module (ice-9 getopt-long), with the function `getopt-long'.
1734 getopt-long is a function for parsing command-line arguments in a
1735 manner consistent with other GNU programs.
1737 (getopt-long ARGS GRAMMAR)
1738 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
1740 ARGS should be a list of strings. Its first element should be the
1741 name of the program; subsequent elements should be the arguments
1742 that were passed to the program on the command line. The
1743 `program-arguments' procedure returns a list of this form.
1745 GRAMMAR is a list of the form:
1746 ((OPTION (PROPERTY VALUE) ...) ...)
1748 Each OPTION should be a symbol. `getopt-long' will accept a
1749 command-line option named `--OPTION'.
1750 Each option can have the following (PROPERTY VALUE) pairs:
1752 (single-char CHAR) --- Accept `-CHAR' as a single-character
1753 equivalent to `--OPTION'. This is how to specify traditional
1755 (required? BOOL) --- If BOOL is true, the option is required.
1756 getopt-long will raise an error if it is not found in ARGS.
1757 (value BOOL) --- If BOOL is #t, the option accepts a value; if
1758 it is #f, it does not; and if it is the symbol
1759 `optional', the option may appear in ARGS with or
1761 (predicate FUNC) --- If the option accepts a value (i.e. you
1762 specified `(value #t)' for this option), then getopt
1763 will apply FUNC to the value, and throw an exception
1764 if it returns #f. FUNC should be a procedure which
1765 accepts a string and returns a boolean value; you may
1766 need to use quasiquotes to get it into GRAMMAR.
1768 The (PROPERTY VALUE) pairs may occur in any order, but each
1769 property may occur only once. By default, options do not have
1770 single-character equivalents, are not required, and do not take
1773 In ARGS, single-character options may be combined, in the usual
1774 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
1775 accepts values, then it must be the last option in the
1776 combination; the value is the next argument. So, for example, using
1777 the following grammar:
1778 ((apples (single-char #\a))
1779 (blimps (single-char #\b) (value #t))
1780 (catalexis (single-char #\c) (value #t)))
1781 the following argument lists would be acceptable:
1782 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
1783 for "blimps" and "catalexis")
1784 ("-ab" "bang" "-c" "couth") (same)
1785 ("-ac" "couth" "-b" "bang") (same)
1786 ("-abc" "couth" "bang") (an error, since `-b' is not the
1787 last option in its combination)
1789 If an option's value is optional, then `getopt-long' decides
1790 whether it has a value by looking at what follows it in ARGS. If
1791 the next element is a string, and it does not appear to be an
1792 option itself, then that string is the option's value.
1794 The value of a long option can appear as the next element in ARGS,
1795 or it can follow the option name, separated by an `=' character.
1796 Thus, using the same grammar as above, the following argument lists
1798 ("--apples" "Braeburn" "--blimps" "Goodyear")
1799 ("--apples=Braeburn" "--blimps" "Goodyear")
1800 ("--blimps" "Goodyear" "--apples=Braeburn")
1802 If the option "--" appears in ARGS, argument parsing stops there;
1803 subsequent arguments are returned as ordinary arguments, even if
1804 they resemble options. So, in the argument list:
1805 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
1806 `getopt-long' will recognize the `apples' option as having the
1807 value "Granny Smith", but it will not recognize the `blimp'
1808 option; it will return the strings "--blimp" and "Goodyear" as
1809 ordinary argument strings.
1811 The `getopt-long' function returns the parsed argument list as an
1812 assocation list, mapping option names --- the symbols from GRAMMAR
1813 --- onto their values, or #t if the option does not accept a value.
1814 Unused options do not appear in the alist.
1816 All arguments that are not the value of any option are returned
1817 as a list, associated with the empty list.
1819 `getopt-long' throws an exception if:
1820 - it finds an unrecognized option in ARGS
1821 - a required option is omitted
1822 - an option that requires an argument doesn't get one
1823 - an option that doesn't accept an argument does get one (this can
1824 only happen using the long option `--opt=value' syntax)
1825 - an option predicate fails
1830 `((lockfile-dir (required? #t)
1833 (predicate ,file-is-directory?))
1834 (verbose (required? #f)
1837 (x-includes (single-char #\x))
1838 (rnet-server (single-char #\y)
1839 (predicate ,string?))))
1841 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
1842 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
1844 => ((() "foo1" "-fred" "foo2" "foo3")
1845 (rnet-server . "lamprod")
1846 (x-includes . "/usr/include")
1847 (lockfile-dir . "/tmp")
1850 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
1852 It will be removed in a few releases.
1854 ** New syntax: lambda*
1855 ** New syntax: define*
1856 ** New syntax: define*-public
1857 ** New syntax: defmacro*
1858 ** New syntax: defmacro*-public
1859 Guile now supports optional arguments.
1861 `lambda*', `define*', `define*-public', `defmacro*' and
1862 `defmacro*-public' are identical to the non-* versions except that
1863 they use an extended type of parameter list that has the following BNF
1864 syntax (parentheses are literal, square brackets indicate grouping,
1865 and `*', `+' and `?' have the usual meaning):
1867 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
1868 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
1869 [[#&rest identifier]|[. identifier]]? ) | [identifier]
1871 ext-var-decl ::= identifier | ( identifier expression )
1873 The semantics are best illustrated with the following documentation
1874 and examples for `lambda*':
1877 lambda extended for optional and keyword arguments
1879 lambda* creates a procedure that takes optional arguments. These
1880 are specified by putting them inside brackets at the end of the
1881 paramater list, but before any dotted rest argument. For example,
1882 (lambda* (a b #&optional c d . e) '())
1883 creates a procedure with fixed arguments a and b, optional arguments c
1884 and d, and rest argument e. If the optional arguments are omitted
1885 in a call, the variables for them are unbound in the procedure. This
1886 can be checked with the bound? macro.
1888 lambda* can also take keyword arguments. For example, a procedure
1890 (lambda* (#&key xyzzy larch) '())
1891 can be called with any of the argument lists (#:xyzzy 11)
1892 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
1893 are given as keywords are bound to values.
1895 Optional and keyword arguments can also be given default values
1896 which they take on when they are not present in a call, by giving a
1897 two-item list in place of an optional argument, for example in:
1898 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
1899 foo is a fixed argument, bar is an optional argument with default
1900 value 42, and baz is a keyword argument with default value 73.
1901 Default value expressions are not evaluated unless they are needed
1902 and until the procedure is called.
1904 lambda* now supports two more special parameter list keywords.
1906 lambda*-defined procedures now throw an error by default if a
1907 keyword other than one of those specified is found in the actual
1908 passed arguments. However, specifying #&allow-other-keys
1909 immediately after the kyword argument declarations restores the
1910 previous behavior of ignoring unknown keywords. lambda* also now
1911 guarantees that if the same keyword is passed more than once, the
1912 last one passed is the one that takes effect. For example,
1913 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
1914 #:heads 37 #:tails 42 #:heads 99)
1915 would result in (99 47) being displayed.
1917 #&rest is also now provided as a synonym for the dotted syntax rest
1918 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
1919 all respects to lambda*. This is provided for more similarity to DSSSL,
1920 MIT-Scheme and Kawa among others, as well as for refugees from other
1923 Further documentation may be found in the optargs.scm file itself.
1925 The optional argument module also exports the macros `let-optional',
1926 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
1927 are not documented here because they may be removed in the future, but
1928 full documentation is still available in optargs.scm.
1930 ** New syntax: and-let*
1931 Guile now supports the `and-let*' form, described in the draft SRFI-2.
1933 Syntax: (land* (<clause> ...) <body> ...)
1934 Each <clause> should have one of the following forms:
1935 (<variable> <expression>)
1938 Each <variable> or <bound-variable> should be an identifier. Each
1939 <expression> should be a valid expression. The <body> should be a
1940 possibly empty sequence of expressions, like the <body> of a
1943 Semantics: A LAND* expression is evaluated by evaluating the
1944 <expression> or <bound-variable> of each of the <clause>s from
1945 left to right. The value of the first <expression> or
1946 <bound-variable> that evaluates to a false value is returned; the
1947 remaining <expression>s and <bound-variable>s are not evaluated.
1948 The <body> forms are evaluated iff all the <expression>s and
1949 <bound-variable>s evaluate to true values.
1951 The <expression>s and the <body> are evaluated in an environment
1952 binding each <variable> of the preceding (<variable> <expression>)
1953 clauses to the value of the <expression>. Later bindings
1954 shadow earlier bindings.
1956 Guile's and-let* macro was contributed by Michael Livshin.
1958 ** New sorting functions
1960 *** New function: sorted? SEQUENCE LESS?
1961 Returns `#t' when the sequence argument is in non-decreasing order
1962 according to LESS? (that is, there is no adjacent pair `... x y
1963 ...' for which `(less? y x)').
1965 Returns `#f' when the sequence contains at least one out-of-order
1966 pair. It is an error if the sequence is neither a list nor a
1969 *** New function: merge LIST1 LIST2 LESS?
1970 LIST1 and LIST2 are sorted lists.
1971 Returns the sorted list of all elements in LIST1 and LIST2.
1973 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
1974 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
1975 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
1976 (Here "<" should read "comes before".)
1978 *** New procedure: merge! LIST1 LIST2 LESS?
1979 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
1980 the result. If the code is compiled, and LESS? constructs no new
1981 pairs, no pairs at all will be allocated. The first pair of the
1982 result will be either the first pair of LIST1 or the first pair of
1985 *** New function: sort SEQUENCE LESS?
1986 Accepts either a list or a vector, and returns a new sequence
1987 which is sorted. The new sequence is the same type as the input.
1988 Always `(sorted? (sort sequence less?) less?)'. The original
1989 sequence is not altered in any way. The new sequence shares its
1990 elements with the old one; no elements are copied.
1992 *** New procedure: sort! SEQUENCE LESS
1993 Returns its sorted result in the original boxes. No new storage is
1994 allocated at all. Proper usage: (set! slist (sort! slist <))
1996 *** New function: stable-sort SEQUENCE LESS?
1997 Similar to `sort' but stable. That is, if "equal" elements are
1998 ordered a < b in the original sequence, they will have the same order
2001 *** New function: stable-sort! SEQUENCE LESS?
2002 Similar to `sort!' but stable.
2003 Uses temporary storage when sorting vectors.
2005 *** New functions: sort-list, sort-list!
2006 Added for compatibility with scsh.
2008 ** New built-in random number support
2010 *** New function: random N [STATE]
2011 Accepts a positive integer or real N and returns a number of the
2012 same type between zero (inclusive) and N (exclusive). The values
2013 returned have a uniform distribution.
2015 The optional argument STATE must be of the type produced by
2016 `copy-random-state' or `seed->random-state'. It defaults to the value
2017 of the variable `*random-state*'. This object is used to maintain the
2018 state of the pseudo-random-number generator and is altered as a side
2019 effect of the `random' operation.
2021 *** New variable: *random-state*
2022 Holds a data structure that encodes the internal state of the
2023 random-number generator that `random' uses by default. The nature
2024 of this data structure is implementation-dependent. It may be
2025 printed out and successfully read back in, but may or may not
2026 function correctly as a random-number state object in another
2029 *** New function: copy-random-state [STATE]
2030 Returns a new object of type suitable for use as the value of the
2031 variable `*random-state*' and as a second argument to `random'.
2032 If argument STATE is given, a copy of it is returned. Otherwise a
2033 copy of `*random-state*' is returned.
2035 *** New function: seed->random-state SEED
2036 Returns a new object of type suitable for use as the value of the
2037 variable `*random-state*' and as a second argument to `random'.
2038 SEED is a string or a number. A new state is generated and
2039 initialized using SEED.
2041 *** New function: random:uniform [STATE]
2042 Returns an uniformly distributed inexact real random number in the
2043 range between 0 and 1.
2045 *** New procedure: random:solid-sphere! VECT [STATE]
2046 Fills VECT with inexact real random numbers the sum of whose
2047 squares is less than 1.0. Thinking of VECT as coordinates in
2048 space of dimension N = `(vector-length VECT)', the coordinates are
2049 uniformly distributed within the unit N-shere. The sum of the
2050 squares of the numbers is returned. VECT can be either a vector
2051 or a uniform vector of doubles.
2053 *** New procedure: random:hollow-sphere! VECT [STATE]
2054 Fills VECT with inexact real random numbers the sum of whose squares
2055 is equal to 1.0. Thinking of VECT as coordinates in space of
2056 dimension n = `(vector-length VECT)', the coordinates are uniformly
2057 distributed over the surface of the unit n-shere. VECT can be either
2058 a vector or a uniform vector of doubles.
2060 *** New function: random:normal [STATE]
2061 Returns an inexact real in a normal distribution with mean 0 and
2062 standard deviation 1. For a normal distribution with mean M and
2063 standard deviation D use `(+ M (* D (random:normal)))'.
2065 *** New procedure: random:normal-vector! VECT [STATE]
2066 Fills VECT with inexact real random numbers which are independent and
2067 standard normally distributed (i.e., with mean 0 and variance 1).
2068 VECT can be either a vector or a uniform vector of doubles.
2070 *** New function: random:exp STATE
2071 Returns an inexact real in an exponential distribution with mean 1.
2072 For an exponential distribution with mean U use (* U (random:exp)).
2074 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
2076 These functions now operate on numbers in the range of a C unsigned
2079 These functions used to operate on numbers in the range of a C signed
2080 long; however, this seems inappropriate, because Guile integers don't
2083 ** New function: make-guardian
2084 This is an implementation of guardians as described in
2085 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
2086 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
2087 Programming Language Design and Implementation, June 1993
2088 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
2090 ** New functions: delq1!, delv1!, delete1!
2091 These procedures behave similar to delq! and friends but delete only
2092 one object if at all.
2094 ** New function: unread-string STRING PORT
2095 Unread STRING to PORT, that is, push it back onto the port so that
2096 next read operation will work on the pushed back characters.
2098 ** unread-char can now be called multiple times
2099 If unread-char is called multiple times, the unread characters will be
2100 read again in last-in first-out order.
2102 ** the procedures uniform-array-read! and uniform-array-write! now
2103 work on any kind of port, not just ports which are open on a file.
2105 ** Now 'l' in a port mode requests line buffering.
2107 ** The procedure truncate-file now works on string ports as well
2108 as file ports. If the size argument is omitted, the current
2109 file position is used.
2111 ** new procedure: seek PORT/FDES OFFSET WHENCE
2112 The arguments are the same as for the old fseek procedure, but it
2113 works on string ports as well as random-access file ports.
2115 ** the fseek procedure now works on string ports, since it has been
2116 redefined using seek.
2118 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
2119 size is not supplied.
2121 ** the newline procedure no longer flushes the port if it's not
2122 line-buffered: previously it did if it was the current output port.
2124 ** open-pipe and close-pipe are no longer primitive procedures, but
2125 an emulation can be obtained using `(use-modules (ice-9 popen))'.
2127 ** the freopen procedure has been removed.
2129 ** new procedure: drain-input PORT
2130 Drains PORT's read buffers (including any pushed-back characters)
2131 and returns the contents as a single string.
2133 ** New function: map-in-order PROC LIST1 LIST2 ...
2134 Version of `map' which guarantees that the procedure is applied to the
2135 lists in serial order.
2137 ** Renamed `serial-array-copy!' and `serial-array-map!' to
2138 `array-copy-in-order!' and `array-map-in-order!'. The old names are
2139 now obsolete and will go away in release 1.5.
2141 ** New syntax: collect BODY1 ...
2142 Version of `begin' which returns a list of the results of the body
2143 forms instead of the result of the last body form. In contrast to
2144 `begin', `collect' allows an empty body.
2146 ** New functions: read-history FILENAME, write-history FILENAME
2147 Read/write command line history from/to file. Returns #t on success
2148 and #f if an error occured.
2150 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
2152 These procedures return a list of definitions available in the specified
2153 argument, a relative module reference. In the case of no argument,
2154 `(current-module)' is now consulted for definitions to return, instead
2155 of simply returning #f, the former behavior.
2157 ** The #/ syntax for lists is no longer supported.
2159 Earlier versions of Scheme accepted this syntax, but printed a
2162 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
2164 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
2167 * Changes to the gh_ interface
2171 Now takes a second argument which is the result array. If this
2172 pointer is NULL, a new array is malloced (the old behaviour).
2174 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
2175 gh_scm2shorts, gh_scm2longs, gh_scm2floats
2179 * Changes to the scm_ interface
2181 ** Function: scm_make_named_hook (char* name, int n_args)
2183 Creates a hook in the same way as make-hook above but also
2184 binds a variable named NAME to it.
2186 This is the typical way of creating a hook from C code.
2188 Currently, the variable is created in the "current" module. This
2189 might change when we get the new module system.
2191 ** The smob interface
2193 The interface for creating smobs has changed. For documentation, see
2194 data-rep.info (made from guile-core/doc/data-rep.texi).
2196 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
2198 >>> This function will be removed in 1.3.4. <<<
2202 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
2203 This function adds a new smob type, named NAME, with instance size
2204 SIZE to the system. The return value is a tag that is used in
2205 creating instances of the type. If SIZE is 0, then no memory will
2206 be allocated when instances of the smob are created, and nothing
2207 will be freed by the default free function.
2209 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
2210 This function sets the smob marking procedure for the smob type
2211 specified by the tag TC. TC is the tag returned by
2212 `scm_make_smob_type'.
2214 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
2215 This function sets the smob freeing procedure for the smob type
2216 specified by the tag TC. TC is the tag returned by
2217 `scm_make_smob_type'.
2219 *** Function: void scm_set_smob_print (tc, print)
2221 - Function: void scm_set_smob_print (long tc,
2222 scm_sizet (*print) (SCM,
2226 This function sets the smob printing procedure for the smob type
2227 specified by the tag TC. TC is the tag returned by
2228 `scm_make_smob_type'.
2230 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
2231 This function sets the smob equality-testing predicate for the
2232 smob type specified by the tag TC. TC is the tag returned by
2233 `scm_make_smob_type'.
2235 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
2236 Make VALUE contain a smob instance of the type with type code TC and
2237 smob data DATA. VALUE must be previously declared as C type `SCM'.
2239 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
2240 This macro expands to a block of code that creates a smob instance
2241 of the type with type code TC and smob data DATA, and returns that
2242 `SCM' value. It should be the last piece of code in a block.
2244 ** The interfaces for using I/O ports and implementing port types
2245 (ptobs) have changed significantly. The new interface is based on
2246 shared access to buffers and a new set of ptob procedures.
2248 *** scm_newptob has been removed
2252 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
2254 - Function: SCM scm_make_port_type (char *type_name,
2255 int (*fill_buffer) (SCM port),
2256 void (*write_flush) (SCM port));
2258 Similarly to the new smob interface, there is a set of function
2259 setters by which the user can customize the behaviour of his port
2260 type. See ports.h (scm_set_port_XXX).
2262 ** scm_strport_to_string: New function: creates a new string from
2263 a string port's buffer.
2265 ** Plug in interface for random number generators
2266 The variable `scm_the_rng' in random.c contains a value and three
2267 function pointers which together define the current random number
2268 generator being used by the Scheme level interface and the random
2269 number library functions.
2271 The user is free to replace the default generator with the generator
2274 *** Variable: size_t scm_the_rng.rstate_size
2275 The size of the random state type used by the current RNG
2278 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
2279 Given the random STATE, return 32 random bits.
2281 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
2282 Seed random state STATE using string S of length N.
2284 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
2285 Given random state STATE, return a malloced copy.
2288 The default RNG is the MWC (Multiply With Carry) random number
2289 generator described by George Marsaglia at the Department of
2290 Statistics and Supercomputer Computations Research Institute, The
2291 Florida State University (http://stat.fsu.edu/~geo).
2293 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
2294 passes all tests in the DIEHARD test suite
2295 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
2296 costs one multiply and one add on platforms which either supports long
2297 longs (gcc does this on most systems) or have 64 bit longs. The cost
2298 is four multiply on other systems but this can be optimized by writing
2299 scm_i_uniform32 in assembler.
2301 These functions are provided through the scm_the_rng interface for use
2302 by libguile and the application.
2304 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
2305 Given the random STATE, return 32 random bits.
2306 Don't use this function directly. Instead go through the plugin
2307 interface (see "Plug in interface" above).
2309 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
2310 Initialize STATE using SEED of length N.
2312 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
2313 Return a malloc:ed copy of STATE. This function can easily be re-used
2314 in the interfaces to other RNGs.
2316 ** Random number library functions
2317 These functions use the current RNG through the scm_the_rng interface.
2318 It might be a good idea to use these functions from your C code so
2319 that only one random generator is used by all code in your program.
2321 The default random state is stored in:
2323 *** Variable: SCM scm_var_random_state
2324 Contains the vcell of the Scheme variable "*random-state*" which is
2325 used as default state by all random number functions in the Scheme
2330 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
2332 *** Function: scm_rstate *scm_c_default_rstate (void)
2333 This is a convenience function which returns the value of
2334 scm_var_random_state. An error message is generated if this value
2335 isn't a random state.
2337 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
2338 Make a new random state from the string SEED of length LENGTH.
2340 It is generally not a good idea to use multiple random states in a
2341 program. While subsequent random numbers generated from one random
2342 state are guaranteed to be reasonably independent, there is no such
2343 guarantee for numbers generated from different random states.
2345 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
2346 Return 32 random bits.
2348 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
2349 Return a sample from the uniform(0,1) distribution.
2351 *** Function: double scm_c_normal01 (scm_rstate *STATE)
2352 Return a sample from the normal(0,1) distribution.
2354 *** Function: double scm_c_exp1 (scm_rstate *STATE)
2355 Return a sample from the exp(1) distribution.
2357 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
2358 Return a sample from the discrete uniform(0,M) distribution.
2360 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
2361 Return a sample from the discrete uniform(0,M) distribution.
2362 M must be a bignum object. The returned value may be an INUM.
2366 Changes in Guile 1.3 (released Monday, October 19, 1998):
2368 * Changes to the distribution
2370 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
2371 To avoid conflicts, programs should name environment variables after
2372 themselves, except when there's a common practice establishing some
2375 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
2376 giving the former precedence, and printing a warning message if the
2377 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
2379 ** The header files related to multi-byte characters have been removed.
2380 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
2381 which referred to these explicitly will probably need to be rewritten,
2382 since the support for the variant string types has been removed; see
2385 ** The header files append.h and sequences.h have been removed. These
2386 files implemented non-R4RS operations which would encourage
2387 non-portable programming style and less easy-to-read code.
2389 * Changes to the stand-alone interpreter
2391 ** New procedures have been added to implement a "batch mode":
2393 *** Function: batch-mode?
2395 Returns a boolean indicating whether the interpreter is in batch
2398 *** Function: set-batch-mode?! ARG
2400 If ARG is true, switches the interpreter to batch mode. The `#f'
2401 case has not been implemented.
2403 ** Guile now provides full command-line editing, when run interactively.
2404 To use this feature, you must have the readline library installed.
2405 The Guile build process will notice it, and automatically include
2408 The readline library is available via anonymous FTP from any GNU
2409 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
2411 ** the-last-stack is now a fluid.
2413 * Changes to the procedure for linking libguile with your programs
2415 ** You can now use the `guile-config' utility to build programs that use Guile.
2417 Guile now includes a command-line utility called `guile-config', which
2418 can provide information about how to compile and link programs that
2421 *** `guile-config compile' prints any C compiler flags needed to use Guile.
2422 You should include this command's output on the command line you use
2423 to compile C or C++ code that #includes the Guile header files. It's
2424 usually just a `-I' flag to help the compiler find the Guile headers.
2427 *** `guile-config link' prints any linker flags necessary to link with Guile.
2429 This command writes to its standard output a list of flags which you
2430 must pass to the linker to link your code against the Guile library.
2431 The flags include '-lguile' itself, any other libraries the Guile
2432 library depends upon, and any `-L' flags needed to help the linker
2433 find those libraries.
2435 For example, here is a Makefile rule that builds a program named 'foo'
2436 from the object files ${FOO_OBJECTS}, and links them against Guile:
2439 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
2441 Previous Guile releases recommended that you use autoconf to detect
2442 which of a predefined set of libraries were present on your system.
2443 It is more robust to use `guile-config', since it records exactly which
2444 libraries the installed Guile library requires.
2446 This was originally called `build-guile', but was renamed to
2447 `guile-config' before Guile 1.3 was released, to be consistent with
2448 the analogous script for the GTK+ GUI toolkit, which is called
2452 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
2454 If you are using the GNU autoconf package to configure your program,
2455 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
2456 (described above) and gather the necessary values for use in your
2459 The GUILE_FLAGS macro expands to configure script code which runs the
2460 `guile-config' script, to find out where Guile's header files and
2461 libraries are installed. It sets two variables, marked for
2462 substitution, as by AC_SUBST.
2464 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
2465 code that uses Guile header files. This is almost always just a
2468 GUILE_LDFLAGS --- flags to pass to the linker to link a
2469 program against Guile. This includes `-lguile' for the Guile
2470 library itself, any libraries that Guile itself requires (like
2471 -lqthreads), and so on. It may also include a -L flag to tell the
2472 compiler where to find the libraries.
2474 GUILE_FLAGS is defined in the file guile.m4, in the top-level
2475 directory of the Guile distribution. You can copy it into your
2476 package's aclocal.m4 file, and then use it in your configure.in file.
2478 If you are using the `aclocal' program, distributed with GNU automake,
2479 to maintain your aclocal.m4 file, the Guile installation process
2480 installs guile.m4 where aclocal will find it. All you need to do is
2481 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
2482 this will copy the definition of GUILE_FLAGS into your aclocal.m4
2486 * Changes to Scheme functions and syntax
2488 ** Multi-byte strings have been removed, as have multi-byte and wide
2489 ports. We felt that these were the wrong approach to
2490 internationalization support.
2492 ** New function: readline [PROMPT]
2493 Read a line from the terminal, and allow the user to edit it,
2494 prompting with PROMPT. READLINE provides a large set of Emacs-like
2495 editing commands, lets the user recall previously typed lines, and
2496 works on almost every kind of terminal, including dumb terminals.
2498 READLINE assumes that the cursor is at the beginning of the line when
2499 it is invoked. Thus, you can't print a prompt yourself, and then call
2500 READLINE; you need to package up your prompt as a string, pass it to
2501 the function, and let READLINE print the prompt itself. This is
2502 because READLINE needs to know the prompt's screen width.
2504 For Guile to provide this function, you must have the readline
2505 library, version 2.1 or later, installed on your system. Readline is
2506 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
2507 any GNU mirror site.
2509 See also ADD-HISTORY function.
2511 ** New function: add-history STRING
2512 Add STRING as the most recent line in the history used by the READLINE
2513 command. READLINE does not add lines to the history itself; you must
2514 call ADD-HISTORY to make previous input available to the user.
2516 ** The behavior of the read-line function has changed.
2518 This function now uses standard C library functions to read the line,
2519 for speed. This means that it doesn not respect the value of
2520 scm-line-incrementors; it assumes that lines are delimited with
2523 (Note that this is read-line, the function that reads a line of text
2524 from a port, not readline, the function that reads a line from a
2525 terminal, providing full editing capabilities.)
2527 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
2529 This module provides some simple argument parsing. It exports one
2532 Function: getopt-gnu-style ARG-LS
2533 Parse a list of program arguments into an alist of option
2536 Each item in the list of program arguments is examined to see if
2537 it meets the syntax of a GNU long-named option. An argument like
2538 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
2539 returned alist, where MUMBLE is a keyword object with the same
2540 name as the argument. An argument like `--MUMBLE=FROB' produces
2541 an element of the form (MUMBLE . FROB), where FROB is a string.
2543 As a special case, the returned alist also contains a pair whose
2544 car is the symbol `rest'. The cdr of this pair is a list
2545 containing all the items in the argument list that are not options
2546 of the form mentioned above.
2548 The argument `--' is treated specially: all items in the argument
2549 list appearing after such an argument are not examined, and are
2550 returned in the special `rest' list.
2552 This function does not parse normal single-character switches.
2553 You will need to parse them out of the `rest' list yourself.
2555 ** The read syntax for byte vectors and short vectors has changed.
2557 Instead of #bytes(...), write #y(...).
2559 Instead of #short(...), write #h(...).
2561 This may seem nutty, but, like the other uniform vectors, byte vectors
2562 and short vectors want to have the same print and read syntax (and,
2563 more basic, want to have read syntax!). Changing the read syntax to
2564 use multiple characters after the hash sign breaks with the
2565 conventions used in R5RS and the conventions used for the other
2566 uniform vectors. It also introduces complexity in the current reader,
2567 both on the C and Scheme levels. (The Right solution is probably to
2568 change the syntax and prototypes for uniform vectors entirely.)
2571 ** The new module (ice-9 session) provides useful interactive functions.
2573 *** New procedure: (apropos REGEXP OPTION ...)
2575 Display a list of top-level variables whose names match REGEXP, and
2576 the modules they are imported from. Each OPTION should be one of the
2579 value --- Show the value of each matching variable.
2580 shadow --- Show bindings shadowed by subsequently imported modules.
2581 full --- Same as both `shadow' and `value'.
2585 guile> (apropos "trace" 'full)
2586 debug: trace #<procedure trace args>
2587 debug: untrace #<procedure untrace args>
2588 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
2589 the-scm-module: before-backtrace-hook ()
2590 the-scm-module: backtrace #<primitive-procedure backtrace>
2591 the-scm-module: after-backtrace-hook ()
2592 the-scm-module: has-shown-backtrace-hint? #f
2595 ** There are new functions and syntax for working with macros.
2597 Guile implements macros as a special object type. Any variable whose
2598 top-level binding is a macro object acts as a macro. The macro object
2599 specifies how the expression should be transformed before evaluation.
2601 *** Macro objects now print in a reasonable way, resembling procedures.
2603 *** New function: (macro? OBJ)
2604 True iff OBJ is a macro object.
2606 *** New function: (primitive-macro? OBJ)
2607 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
2608 macro transformers, implemented in eval.c rather than Scheme code.
2610 Why do we have this function?
2611 - For symmetry with procedure? and primitive-procedure?,
2612 - to allow custom print procedures to tell whether a macro is
2613 primitive, and display it differently, and
2614 - to allow compilers and user-written evaluators to distinguish
2615 builtin special forms from user-defined ones, which could be
2618 *** New function: (macro-type OBJ)
2619 Return a value indicating what kind of macro OBJ is. Possible return
2622 The symbol `syntax' --- a macro created by procedure->syntax.
2623 The symbol `macro' --- a macro created by procedure->macro.
2624 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
2625 The boolean #f --- if OBJ is not a macro object.
2627 *** New function: (macro-name MACRO)
2628 Return the name of the macro object MACRO's procedure, as returned by
2631 *** New function: (macro-transformer MACRO)
2632 Return the transformer procedure for MACRO.
2634 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
2636 Specify a new macro expander to use in the current module. Each
2637 MODULE is a module name, with the same meaning as in the `use-modules'
2638 form; each named module's exported bindings are added to the current
2639 top-level environment. TRANSFORMER is an expression evaluated in the
2640 resulting environment which must yield a procedure to use as the
2641 module's eval transformer: every expression evaluated in this module
2642 is passed to this function, and the result passed to the Guile
2645 *** macro-eval! is removed. Use local-eval instead.
2647 ** Some magic has been added to the printer to better handle user
2648 written printing routines (like record printers, closure printers).
2650 The problem is that these user written routines must have access to
2651 the current `print-state' to be able to handle fancy things like
2652 detection of circular references. These print-states have to be
2653 passed to the builtin printing routines (display, write, etc) to
2654 properly continue the print chain.
2656 We didn't want to change all existing print code so that it
2657 explicitly passes thru a print state in addition to a port. Instead,
2658 we extented the possible values that the builtin printing routines
2659 accept as a `port'. In addition to a normal port, they now also take
2660 a pair of a normal port and a print-state. Printing will go to the
2661 port and the print-state will be used to control the detection of
2662 circular references, etc. If the builtin function does not care for a
2663 print-state, it is simply ignored.
2665 User written callbacks are now called with such a pair as their
2666 `port', but because every function now accepts this pair as a PORT
2667 argument, you don't have to worry about that. In fact, it is probably
2668 safest to not check for these pairs.
2670 However, it is sometimes necessary to continue a print chain on a
2671 different port, for example to get a intermediate string
2672 representation of the printed value, mangle that string somehow, and
2673 then to finally print the mangled string. Use the new function
2675 inherit-print-state OLD-PORT NEW-PORT
2677 for this. It constructs a new `port' that prints to NEW-PORT but
2678 inherits the print-state of OLD-PORT.
2680 ** struct-vtable-offset renamed to vtable-offset-user
2682 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
2684 ** There is now a third optional argument to make-vtable-vtable
2685 (and fourth to make-struct) when constructing new types (vtables).
2686 This argument initializes field vtable-index-printer of the vtable.
2688 ** The detection of circular references has been extended to structs.
2689 That is, a structure that -- in the process of being printed -- prints
2690 itself does not lead to infinite recursion.
2692 ** There is now some basic support for fluids. Please read
2693 "libguile/fluid.h" to find out more. It is accessible from Scheme with
2694 the following functions and macros:
2696 Function: make-fluid
2698 Create a new fluid object. Fluids are not special variables or
2699 some other extension to the semantics of Scheme, but rather
2700 ordinary Scheme objects. You can store them into variables (that
2701 are still lexically scoped, of course) or into any other place you
2702 like. Every fluid has a initial value of `#f'.
2704 Function: fluid? OBJ
2706 Test whether OBJ is a fluid.
2708 Function: fluid-ref FLUID
2709 Function: fluid-set! FLUID VAL
2711 Access/modify the fluid FLUID. Modifications are only visible
2712 within the current dynamic root (that includes threads).
2714 Function: with-fluids* FLUIDS VALUES THUNK
2716 FLUIDS is a list of fluids and VALUES a corresponding list of
2717 values for these fluids. Before THUNK gets called the values are
2718 installed in the fluids and the old values of the fluids are
2719 saved in the VALUES list. When the flow of control leaves THUNK
2720 or reenters it, the values get swapped again. You might think of
2721 this as a `safe-fluid-excursion'. Note that the VALUES list is
2722 modified by `with-fluids*'.
2724 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
2726 The same as `with-fluids*' but with a different syntax. It looks
2727 just like `let', but both FLUID and VALUE are evaluated. Remember,
2728 fluids are not special variables but ordinary objects. FLUID
2729 should evaluate to a fluid.
2731 ** Changes to system call interfaces:
2733 *** close-port, close-input-port and close-output-port now return a
2734 boolean instead of an `unspecified' object. #t means that the port
2735 was successfully closed, while #f means it was already closed. It is
2736 also now possible for these procedures to raise an exception if an
2737 error occurs (some errors from write can be delayed until close.)
2739 *** the first argument to chmod, fcntl, ftell and fseek can now be a
2742 *** the third argument to fcntl is now optional.
2744 *** the first argument to chown can now be a file descriptor or a port.
2746 *** the argument to stat can now be a port.
2748 *** The following new procedures have been added (most use scsh
2751 *** procedure: close PORT/FD
2752 Similar to close-port (*note close-port: Closing Ports.), but also
2753 works on file descriptors. A side effect of closing a file
2754 descriptor is that any ports using that file descriptor are moved
2755 to a different file descriptor and have their revealed counts set
2758 *** procedure: port->fdes PORT
2759 Returns the integer file descriptor underlying PORT. As a side
2760 effect the revealed count of PORT is incremented.
2762 *** procedure: fdes->ports FDES
2763 Returns a list of existing ports which have FDES as an underlying
2764 file descriptor, without changing their revealed counts.
2766 *** procedure: fdes->inport FDES
2767 Returns an existing input port which has FDES as its underlying
2768 file descriptor, if one exists, and increments its revealed count.
2769 Otherwise, returns a new input port with a revealed count of 1.
2771 *** procedure: fdes->outport FDES
2772 Returns an existing output port which has FDES as its underlying
2773 file descriptor, if one exists, and increments its revealed count.
2774 Otherwise, returns a new output port with a revealed count of 1.
2776 The next group of procedures perform a `dup2' system call, if NEWFD
2777 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
2778 duplicated can be supplied as an integer or contained in a port. The
2779 type of value returned varies depending on which procedure is used.
2781 All procedures also have the side effect when performing `dup2' that
2782 any ports using NEWFD are moved to a different file descriptor and have
2783 their revealed counts set to zero.
2785 *** procedure: dup->fdes PORT/FD [NEWFD]
2786 Returns an integer file descriptor.
2788 *** procedure: dup->inport PORT/FD [NEWFD]
2789 Returns a new input port using the new file descriptor.
2791 *** procedure: dup->outport PORT/FD [NEWFD]
2792 Returns a new output port using the new file descriptor.
2794 *** procedure: dup PORT/FD [NEWFD]
2795 Returns a new port if PORT/FD is a port, with the same mode as the
2796 supplied port, otherwise returns an integer file descriptor.
2798 *** procedure: dup->port PORT/FD MODE [NEWFD]
2799 Returns a new port using the new file descriptor. MODE supplies a
2800 mode string for the port (*note open-file: File Ports.).
2802 *** procedure: setenv NAME VALUE
2803 Modifies the environment of the current process, which is also the
2804 default environment inherited by child processes.
2806 If VALUE is `#f', then NAME is removed from the environment.
2807 Otherwise, the string NAME=VALUE is added to the environment,
2808 replacing any existing string with name matching NAME.
2810 The return value is unspecified.
2812 *** procedure: truncate-file OBJ SIZE
2813 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
2814 can be a string containing a file name or an integer file
2815 descriptor or port open for output on the file. The underlying
2816 system calls are `truncate' and `ftruncate'.
2818 The return value is unspecified.
2820 *** procedure: setvbuf PORT MODE [SIZE]
2821 Set the buffering mode for PORT. MODE can be:
2829 block buffered, using a newly allocated buffer of SIZE bytes.
2830 However if SIZE is zero or unspecified, the port will be made
2833 This procedure should not be used after I/O has been performed with
2836 Ports are usually block buffered by default, with a default buffer
2837 size. Procedures e.g., *Note open-file: File Ports, which accept a
2838 mode string allow `0' to be added to request an unbuffered port.
2840 *** procedure: fsync PORT/FD
2841 Copies any unwritten data for the specified output file descriptor
2842 to disk. If PORT/FD is a port, its buffer is flushed before the
2843 underlying file descriptor is fsync'd. The return value is
2846 *** procedure: open-fdes PATH FLAGS [MODES]
2847 Similar to `open' but returns a file descriptor instead of a port.
2849 *** procedure: execle PATH ENV [ARG] ...
2850 Similar to `execl', but the environment of the new process is
2851 specified by ENV, which must be a list of strings as returned by
2852 the `environ' procedure.
2854 This procedure is currently implemented using the `execve' system
2855 call, but we call it `execle' because of its Scheme calling
2858 *** procedure: strerror ERRNO
2859 Returns the Unix error message corresponding to ERRNO, an integer.
2861 *** procedure: primitive-exit [STATUS]
2862 Terminate the current process without unwinding the Scheme stack.
2863 This is would typically be useful after a fork. The exit status
2864 is STATUS if supplied, otherwise zero.
2866 *** procedure: times
2867 Returns an object with information about real and processor time.
2868 The following procedures accept such an object as an argument and
2869 return a selected component:
2872 The current real time, expressed as time units relative to an
2876 The CPU time units used by the calling process.
2879 The CPU time units used by the system on behalf of the
2883 The CPU time units used by terminated child processes of the
2884 calling process, whose status has been collected (e.g., using
2888 Similarly, the CPU times units used by the system on behalf of
2889 terminated child processes.
2891 ** Removed: list-length
2892 ** Removed: list-append, list-append!
2893 ** Removed: list-reverse, list-reverse!
2895 ** array-map renamed to array-map!
2897 ** serial-array-map renamed to serial-array-map!
2899 ** catch doesn't take #f as first argument any longer
2901 Previously, it was possible to pass #f instead of a key to `catch'.
2902 That would cause `catch' to pass a jump buffer object to the procedure
2903 passed as second argument. The procedure could then use this jump
2904 buffer objekt as an argument to throw.
2906 This mechanism has been removed since its utility doesn't motivate the
2907 extra complexity it introduces.
2909 ** The `#/' notation for lists now provokes a warning message from Guile.
2910 This syntax will be removed from Guile in the near future.
2912 To disable the warning message, set the GUILE_HUSH environment
2913 variable to any non-empty value.
2915 ** The newline character now prints as `#\newline', following the
2916 normal Scheme notation, not `#\nl'.
2918 * Changes to the gh_ interface
2920 ** The gh_enter function now takes care of loading the Guile startup files.
2921 gh_enter works by calling scm_boot_guile; see the remarks below.
2923 ** Function: void gh_write (SCM x)
2925 Write the printed representation of the scheme object x to the current
2926 output port. Corresponds to the scheme level `write'.
2928 ** gh_list_length renamed to gh_length.
2930 ** vector handling routines
2932 Several major changes. In particular, gh_vector() now resembles
2933 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
2934 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
2935 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
2936 vector-related gh_ functions have been implemented.
2938 ** pair and list routines
2940 Implemented several of the R4RS pair and list functions that were
2943 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
2945 New function. Converts double arrays back and forth between Scheme
2948 * Changes to the scm_ interface
2950 ** The function scm_boot_guile now takes care of loading the startup files.
2952 Guile's primary initialization function, scm_boot_guile, now takes
2953 care of loading `boot-9.scm', in the `ice-9' module, to initialize
2954 Guile, define the module system, and put together some standard
2955 bindings. It also loads `init.scm', which is intended to hold
2956 site-specific initialization code.
2958 Since Guile cannot operate properly until boot-9.scm is loaded, there
2959 is no reason to separate loading boot-9.scm from Guile's other
2960 initialization processes.
2962 This job used to be done by scm_compile_shell_switches, which didn't
2963 make much sense; in particular, it meant that people using Guile for
2964 non-shell-like applications had to jump through hoops to get Guile
2965 initialized properly.
2967 ** The function scm_compile_shell_switches no longer loads the startup files.
2968 Now, Guile always loads the startup files, whenever it is initialized;
2969 see the notes above for scm_boot_guile and scm_load_startup_files.
2971 ** Function: scm_load_startup_files
2972 This new function takes care of loading Guile's initialization file
2973 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
2974 this is always called by the Guile initialization process, it's
2975 probably not too useful to call this yourself, but it's there anyway.
2977 ** The semantics of smob marking have changed slightly.
2979 The smob marking function (the `mark' member of the scm_smobfuns
2980 structure) is no longer responsible for setting the mark bit on the
2981 smob. The generic smob handling code in the garbage collector will
2982 set this bit. The mark function need only ensure that any other
2983 objects the smob refers to get marked.
2985 Note that this change means that the smob's GC8MARK bit is typically
2986 already set upon entry to the mark function. Thus, marking functions
2987 which look like this:
2990 if (SCM_GC8MARKP (ptr))
2992 SCM_SETGC8MARK (ptr);
2993 ... mark objects to which the smob refers ...
2996 are now incorrect, since they will return early, and fail to mark any
2997 other objects the smob refers to. Some code in the Guile library used
3000 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
3002 If you have implemented your own I/O port type, by writing the
3003 functions required by the scm_ptobfuns and then calling scm_newptob,
3004 you will need to change your functions slightly.
3006 The functions in a scm_ptobfuns structure now expect the port itself
3007 as their argument; they used to expect the `stream' member of the
3008 port's scm_port_table structure. This allows functions in an
3009 scm_ptobfuns structure to easily access the port's cell (and any flags
3010 it its CAR), and the port's scm_port_table structure.
3012 Guile now passes the I/O port itself as the `port' argument in the
3013 following scm_ptobfuns functions:
3015 int (*free) (SCM port);
3016 int (*fputc) (int, SCM port);
3017 int (*fputs) (char *, SCM port);
3018 scm_sizet (*fwrite) SCM_P ((char *ptr,
3022 int (*fflush) (SCM port);
3023 int (*fgetc) (SCM port);
3024 int (*fclose) (SCM port);
3026 The interfaces to the `mark', `print', `equalp', and `fgets' methods
3029 If you have existing code which defines its own port types, it is easy
3030 to convert your code to the new interface; simply apply SCM_STREAM to
3031 the port argument to yield the value you code used to expect.
3033 Note that since both the port and the stream have the same type in the
3034 C code --- they are both SCM values --- the C compiler will not remind
3035 you if you forget to update your scm_ptobfuns functions.
3038 ** Function: int scm_internal_select (int fds,
3042 struct timeval *timeout);
3044 This is a replacement for the `select' function provided by the OS.
3045 It enables I/O blocking and sleeping to happen for one cooperative
3046 thread without blocking other threads. It also avoids busy-loops in
3047 these situations. It is intended that all I/O blocking and sleeping
3048 will finally go through this function. Currently, this function is
3049 only available on systems providing `gettimeofday' and `select'.
3051 ** Function: SCM scm_internal_stack_catch (SCM tag,
3052 scm_catch_body_t body,
3054 scm_catch_handler_t handler,
3057 A new sibling to the other two C level `catch' functions
3058 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
3059 the stack to be saved automatically into the variable `the-last-stack'
3060 (scm_the_last_stack_var) on error. This is necessary if you want to
3061 use advanced error reporting, such as calling scm_display_error and
3062 scm_display_backtrace. (They both take a stack object as argument.)
3064 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
3066 scm_catch_handler_t handler,
3069 Spawns a new thread. It does a job similar to
3070 scm_call_with_new_thread but takes arguments more suitable when
3071 spawning threads from application C code.
3073 ** The hook scm_error_callback has been removed. It was originally
3074 intended as a way for the user to install his own error handler. But
3075 that method works badly since it intervenes between throw and catch,
3076 thereby changing the semantics of expressions like (catch #t ...).
3077 The correct way to do it is to use one of the C level catch functions
3078 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
3080 ** Removed functions:
3082 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
3083 scm_list_reverse, scm_list_reverse_x
3085 ** New macros: SCM_LISTn where n is one of the integers 0-9.
3087 These can be used for pretty list creation from C. The idea is taken
3088 from Erick Gallesio's STk.
3090 ** scm_array_map renamed to scm_array_map_x
3092 ** mbstrings are now removed
3094 This means that the type codes scm_tc7_mb_string and
3095 scm_tc7_mb_substring has been removed.
3097 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
3099 Since we no longer support multi-byte strings, these I/O functions
3100 have been simplified, and renamed. Here are their old names, and
3101 their new names and arguments:
3103 scm_gen_putc -> void scm_putc (int c, SCM port);
3104 scm_gen_puts -> void scm_puts (char *s, SCM port);
3105 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
3106 scm_gen_getc -> void scm_getc (SCM port);
3109 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
3111 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
3113 SCM_TYP7S now masks away the bit which distinguishes substrings from
3116 ** scm_catch_body_t: Backward incompatible change!
3118 Body functions to scm_internal_catch and friends do not any longer
3119 take a second argument. This is because it is no longer possible to
3120 pass a #f arg to catch.
3122 ** Calls to scm_protect_object and scm_unprotect now nest properly.
3124 The function scm_protect_object protects its argument from being freed
3125 by the garbage collector. scm_unprotect_object removes that
3128 These functions now nest properly. That is, for every object O, there
3129 is a counter which scm_protect_object(O) increments and
3130 scm_unprotect_object(O) decrements, if the counter is greater than
3131 zero. Every object's counter is zero when it is first created. If an
3132 object's counter is greater than zero, the garbage collector will not
3133 reclaim its storage.
3135 This allows you to use scm_protect_object in your code without
3136 worrying that some other function you call will call
3137 scm_unprotect_object, and allow it to be freed. Assuming that the
3138 functions you call are well-behaved, and unprotect only those objects
3139 they protect, you can follow the same rule and have confidence that
3140 objects will be freed only at appropriate times.
3143 Changes in Guile 1.2 (released Tuesday, June 24 1997):
3145 * Changes to the distribution
3147 ** Nightly snapshots are now available from ftp.red-bean.com.
3148 The old server, ftp.cyclic.com, has been relinquished to its rightful
3151 Nightly snapshots of the Guile development sources are now available via
3152 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
3154 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
3155 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
3157 ** To run Guile without installing it, the procedure has changed a bit.
3159 If you used a separate build directory to compile Guile, you'll need
3160 to include the build directory in SCHEME_LOAD_PATH, as well as the
3161 source directory. See the `INSTALL' file for examples.
3163 * Changes to the procedure for linking libguile with your programs
3165 ** The standard Guile load path for Scheme code now includes
3166 $(datadir)/guile (usually /usr/local/share/guile). This means that
3167 you can install your own Scheme files there, and Guile will find them.
3168 (Previous versions of Guile only checked a directory whose name
3169 contained the Guile version number, so you had to re-install or move
3170 your Scheme sources each time you installed a fresh version of Guile.)
3172 The load path also includes $(datadir)/guile/site; we recommend
3173 putting individual Scheme files there. If you want to install a
3174 package with multiple source files, create a directory for them under
3177 ** Guile 1.2 will now use the Rx regular expression library, if it is
3178 installed on your system. When you are linking libguile into your own
3179 programs, this means you will have to link against -lguile, -lqt (if
3180 you configured Guile with thread support), and -lrx.
3182 If you are using autoconf to generate configuration scripts for your
3183 application, the following lines should suffice to add the appropriate
3184 libraries to your link command:
3186 ### Find Rx, quickthreads and libguile.
3187 AC_CHECK_LIB(rx, main)
3188 AC_CHECK_LIB(qt, main)
3189 AC_CHECK_LIB(guile, scm_shell)
3191 The Guile 1.2 distribution does not contain sources for the Rx
3192 library, as Guile 1.0 did. If you want to use Rx, you'll need to
3193 retrieve it from a GNU FTP site and install it separately.
3195 * Changes to Scheme functions and syntax
3197 ** The dynamic linking features of Guile are now enabled by default.
3198 You can disable them by giving the `--disable-dynamic-linking' option
3201 (dynamic-link FILENAME)
3203 Find the object file denoted by FILENAME (a string) and link it
3204 into the running Guile application. When everything works out,
3205 return a Scheme object suitable for representing the linked object
3206 file. Otherwise an error is thrown. How object files are
3207 searched is system dependent.
3209 (dynamic-object? VAL)
3211 Determine whether VAL represents a dynamically linked object file.
3213 (dynamic-unlink DYNOBJ)
3215 Unlink the indicated object file from the application. DYNOBJ
3216 should be one of the values returned by `dynamic-link'.
3218 (dynamic-func FUNCTION DYNOBJ)
3220 Search the C function indicated by FUNCTION (a string or symbol)
3221 in DYNOBJ and return some Scheme object that can later be used
3222 with `dynamic-call' to actually call this function. Right now,
3223 these Scheme objects are formed by casting the address of the
3224 function to `long' and converting this number to its Scheme
3227 (dynamic-call FUNCTION DYNOBJ)
3229 Call the C function indicated by FUNCTION and DYNOBJ. The
3230 function is passed no arguments and its return value is ignored.
3231 When FUNCTION is something returned by `dynamic-func', call that
3232 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
3233 etc.), look it up in DYNOBJ; this is equivalent to
3235 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
3237 Interrupts are deferred while the C function is executing (with
3238 SCM_DEFER_INTS/SCM_ALLOW_INTS).
3240 (dynamic-args-call FUNCTION DYNOBJ ARGS)
3242 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
3243 some arguments and return its return value. The C function is
3244 expected to take two arguments and return an `int', just like
3247 int c_func (int argc, char **argv);
3249 ARGS must be a list of strings and is converted into an array of
3250 `char *'. The array is passed in ARGV and its size in ARGC. The
3251 return value is converted to a Scheme number and returned from the
3252 call to `dynamic-args-call'.
3254 When dynamic linking is disabled or not supported on your system,
3255 the above functions throw errors, but they are still available.
3257 Here is a small example that works on GNU/Linux:
3259 (define libc-obj (dynamic-link "libc.so"))
3260 (dynamic-args-call 'rand libc-obj '())
3262 See the file `libguile/DYNAMIC-LINKING' for additional comments.
3264 ** The #/ syntax for module names is depreciated, and will be removed
3265 in a future version of Guile. Instead of
3273 The latter syntax is more consistent with existing Lisp practice.
3275 ** Guile now does fancier printing of structures. Structures are the
3276 underlying implementation for records, which in turn are used to
3277 implement modules, so all of these object now print differently and in
3278 a more informative way.
3280 The Scheme printer will examine the builtin variable *struct-printer*
3281 whenever it needs to print a structure object. When this variable is
3282 not `#f' it is deemed to be a procedure and will be applied to the
3283 structure object and the output port. When *struct-printer* is `#f'
3284 or the procedure return `#f' the structure object will be printed in
3285 the boring #<struct 80458270> form.
3287 This hook is used by some routines in ice-9/boot-9.scm to implement
3288 type specific printing routines. Please read the comments there about
3291 One of the more specific uses of structs are records. The printing
3292 procedure that could be passed to MAKE-RECORD-TYPE is now actually
3293 called. It should behave like a *struct-printer* procedure (described
3296 ** Guile now supports a new R4RS-compliant syntax for keywords. A
3297 token of the form #:NAME, where NAME has the same syntax as a Scheme
3298 symbol, is the external representation of the keyword named NAME.
3299 Keyword objects print using this syntax as well, so values containing
3300 keyword objects can be read back into Guile. When used in an
3301 expression, keywords are self-quoting objects.
3303 Guile suports this read syntax, and uses this print syntax, regardless
3304 of the current setting of the `keyword' read option. The `keyword'
3305 read option only controls whether Guile recognizes the `:NAME' syntax,
3306 which is incompatible with R4RS. (R4RS says such token represent
3309 ** Guile has regular expression support again. Guile 1.0 included
3310 functions for matching regular expressions, based on the Rx library.
3311 In Guile 1.1, the Guile/Rx interface was removed to simplify the
3312 distribution, and thus Guile had no regular expression support. Guile
3313 1.2 again supports the most commonly used functions, and supports all
3314 of SCSH's regular expression functions.
3316 If your system does not include a POSIX regular expression library,
3317 and you have not linked Guile with a third-party regexp library such as
3318 Rx, these functions will not be available. You can tell whether your
3319 Guile installation includes regular expression support by checking
3320 whether the `*features*' list includes the `regex' symbol.
3322 *** regexp functions
3324 By default, Guile supports POSIX extended regular expressions. That
3325 means that the characters `(', `)', `+' and `?' are special, and must
3326 be escaped if you wish to match the literal characters.
3328 This regular expression interface was modeled after that implemented
3329 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
3330 with SCSH regular expressions.
3332 **** Function: string-match PATTERN STR [START]
3333 Compile the string PATTERN into a regular expression and compare
3334 it with STR. The optional numeric argument START specifies the
3335 position of STR at which to begin matching.
3337 `string-match' returns a "match structure" which describes what,
3338 if anything, was matched by the regular expression. *Note Match
3339 Structures::. If STR does not match PATTERN at all,
3340 `string-match' returns `#f'.
3342 Each time `string-match' is called, it must compile its PATTERN
3343 argument into a regular expression structure. This operation is
3344 expensive, which makes `string-match' inefficient if the same regular
3345 expression is used several times (for example, in a loop). For better
3346 performance, you can compile a regular expression in advance and then
3347 match strings against the compiled regexp.
3349 **** Function: make-regexp STR [FLAGS]
3350 Compile the regular expression described by STR, and return the
3351 compiled regexp structure. If STR does not describe a legal
3352 regular expression, `make-regexp' throws a
3353 `regular-expression-syntax' error.
3355 FLAGS may be the bitwise-or of one or more of the following:
3357 **** Constant: regexp/extended
3358 Use POSIX Extended Regular Expression syntax when interpreting
3359 STR. If not set, POSIX Basic Regular Expression syntax is used.
3360 If the FLAGS argument is omitted, we assume regexp/extended.
3362 **** Constant: regexp/icase
3363 Do not differentiate case. Subsequent searches using the
3364 returned regular expression will be case insensitive.
3366 **** Constant: regexp/newline
3367 Match-any-character operators don't match a newline.
3369 A non-matching list ([^...]) not containing a newline matches a
3372 Match-beginning-of-line operator (^) matches the empty string
3373 immediately after a newline, regardless of whether the FLAGS
3374 passed to regexp-exec contain regexp/notbol.
3376 Match-end-of-line operator ($) matches the empty string
3377 immediately before a newline, regardless of whether the FLAGS
3378 passed to regexp-exec contain regexp/noteol.
3380 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
3381 Match the compiled regular expression REGEXP against `str'. If
3382 the optional integer START argument is provided, begin matching
3383 from that position in the string. Return a match structure
3384 describing the results of the match, or `#f' if no match could be
3387 FLAGS may be the bitwise-or of one or more of the following:
3389 **** Constant: regexp/notbol
3390 The match-beginning-of-line operator always fails to match (but
3391 see the compilation flag regexp/newline above) This flag may be
3392 used when different portions of a string are passed to
3393 regexp-exec and the beginning of the string should not be
3394 interpreted as the beginning of the line.
3396 **** Constant: regexp/noteol
3397 The match-end-of-line operator always fails to match (but see the
3398 compilation flag regexp/newline above)
3400 **** Function: regexp? OBJ
3401 Return `#t' if OBJ is a compiled regular expression, or `#f'
3404 Regular expressions are commonly used to find patterns in one string
3405 and replace them with the contents of another string.
3407 **** Function: regexp-substitute PORT MATCH [ITEM...]
3408 Write to the output port PORT selected contents of the match
3409 structure MATCH. Each ITEM specifies what should be written, and
3410 may be one of the following arguments:
3412 * A string. String arguments are written out verbatim.
3414 * An integer. The submatch with that number is written.
3416 * The symbol `pre'. The portion of the matched string preceding
3417 the regexp match is written.
3419 * The symbol `post'. The portion of the matched string
3420 following the regexp match is written.
3422 PORT may be `#f', in which case nothing is written; instead,
3423 `regexp-substitute' constructs a string from the specified ITEMs
3426 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
3427 Similar to `regexp-substitute', but can be used to perform global
3428 substitutions on STR. Instead of taking a match structure as an
3429 argument, `regexp-substitute/global' takes two string arguments: a
3430 REGEXP string describing a regular expression, and a TARGET string
3431 which should be matched against this regular expression.
3433 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
3436 * A function may be supplied. When this function is called, it
3437 will be passed one argument: a match structure for a given
3438 regular expression match. It should return a string to be
3439 written out to PORT.
3441 * The `post' symbol causes `regexp-substitute/global' to recurse
3442 on the unmatched portion of STR. This *must* be supplied in
3443 order to perform global search-and-replace on STR; if it is
3444 not present among the ITEMs, then `regexp-substitute/global'
3445 will return after processing a single match.
3447 *** Match Structures
3449 A "match structure" is the object returned by `string-match' and
3450 `regexp-exec'. It describes which portion of a string, if any, matched
3451 the given regular expression. Match structures include: a reference to
3452 the string that was checked for matches; the starting and ending
3453 positions of the regexp match; and, if the regexp included any
3454 parenthesized subexpressions, the starting and ending positions of each
3457 In each of the regexp match functions described below, the `match'
3458 argument must be a match structure returned by a previous call to
3459 `string-match' or `regexp-exec'. Most of these functions return some
3460 information about the original target string that was matched against a
3461 regular expression; we will call that string TARGET for easy reference.
3463 **** Function: regexp-match? OBJ
3464 Return `#t' if OBJ is a match structure returned by a previous
3465 call to `regexp-exec', or `#f' otherwise.
3467 **** Function: match:substring MATCH [N]
3468 Return the portion of TARGET matched by subexpression number N.
3469 Submatch 0 (the default) represents the entire regexp match. If
3470 the regular expression as a whole matched, but the subexpression
3471 number N did not match, return `#f'.
3473 **** Function: match:start MATCH [N]
3474 Return the starting position of submatch number N.
3476 **** Function: match:end MATCH [N]
3477 Return the ending position of submatch number N.
3479 **** Function: match:prefix MATCH
3480 Return the unmatched portion of TARGET preceding the regexp match.
3482 **** Function: match:suffix MATCH
3483 Return the unmatched portion of TARGET following the regexp match.
3485 **** Function: match:count MATCH
3486 Return the number of parenthesized subexpressions from MATCH.
3487 Note that the entire regular expression match itself counts as a
3488 subexpression, and failed submatches are included in the count.
3490 **** Function: match:string MATCH
3491 Return the original TARGET string.
3493 *** Backslash Escapes
3495 Sometimes you will want a regexp to match characters like `*' or `$'
3496 exactly. For example, to check whether a particular string represents
3497 a menu entry from an Info node, it would be useful to match it against
3498 a regexp like `^* [^:]*::'. However, this won't work; because the
3499 asterisk is a metacharacter, it won't match the `*' at the beginning of
3500 the string. In this case, we want to make the first asterisk un-magic.
3502 You can do this by preceding the metacharacter with a backslash
3503 character `\'. (This is also called "quoting" the metacharacter, and
3504 is known as a "backslash escape".) When Guile sees a backslash in a
3505 regular expression, it considers the following glyph to be an ordinary
3506 character, no matter what special meaning it would ordinarily have.
3507 Therefore, we can make the above example work by changing the regexp to
3508 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
3509 to match only a single asterisk in the target string.
3511 Since the backslash is itself a metacharacter, you may force a
3512 regexp to match a backslash in the target string by preceding the
3513 backslash with itself. For example, to find variable references in a
3514 TeX program, you might want to find occurrences of the string `\let\'
3515 followed by any number of alphabetic characters. The regular expression
3516 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
3517 each match a single backslash in the target string.
3519 **** Function: regexp-quote STR
3520 Quote each special character found in STR with a backslash, and
3521 return the resulting string.
3523 *Very important:* Using backslash escapes in Guile source code (as
3524 in Emacs Lisp or C) can be tricky, because the backslash character has
3525 special meaning for the Guile reader. For example, if Guile encounters
3526 the character sequence `\n' in the middle of a string while processing
3527 Scheme code, it replaces those characters with a newline character.
3528 Similarly, the character sequence `\t' is replaced by a horizontal tab.
3529 Several of these "escape sequences" are processed by the Guile reader
3530 before your code is executed. Unrecognized escape sequences are
3531 ignored: if the characters `\*' appear in a string, they will be
3532 translated to the single character `*'.
3534 This translation is obviously undesirable for regular expressions,
3535 since we want to be able to include backslashes in a string in order to
3536 escape regexp metacharacters. Therefore, to make sure that a backslash
3537 is preserved in a string in your Guile program, you must use *two*
3538 consecutive backslashes:
3540 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
3542 The string in this example is preprocessed by the Guile reader before
3543 any code is executed. The resulting argument to `make-regexp' is the
3544 string `^\* [^:]*', which is what we really want.
3546 This also means that in order to write a regular expression that
3547 matches a single backslash character, the regular expression string in
3548 the source code must include *four* backslashes. Each consecutive pair
3549 of backslashes gets translated by the Guile reader to a single
3550 backslash, and the resulting double-backslash is interpreted by the
3551 regexp engine as matching a single backslash character. Hence:
3553 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
3555 The reason for the unwieldiness of this syntax is historical. Both
3556 regular expression pattern matchers and Unix string processing systems
3557 have traditionally used backslashes with the special meanings described
3558 above. The POSIX regular expression specification and ANSI C standard
3559 both require these semantics. Attempting to abandon either convention
3560 would cause other kinds of compatibility problems, possibly more severe
3561 ones. Therefore, without extending the Scheme reader to support
3562 strings with different quoting conventions (an ungainly and confusing
3563 extension when implemented in other languages), we must adhere to this
3564 cumbersome escape syntax.
3566 * Changes to the gh_ interface
3568 * Changes to the scm_ interface
3570 * Changes to system call interfaces:
3572 ** The value returned by `raise' is now unspecified. It throws an exception
3575 *** A new procedure `sigaction' can be used to install signal handlers
3577 (sigaction signum [action] [flags])
3579 signum is the signal number, which can be specified using the value
3582 If action is omitted, sigaction returns a pair: the CAR is the current
3583 signal hander, which will be either an integer with the value SIG_DFL
3584 (default action) or SIG_IGN (ignore), or the Scheme procedure which
3585 handles the signal, or #f if a non-Scheme procedure handles the
3586 signal. The CDR contains the current sigaction flags for the handler.
3588 If action is provided, it is installed as the new handler for signum.
3589 action can be a Scheme procedure taking one argument, or the value of
3590 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
3591 whatever signal handler was installed before sigaction was first used.
3592 Flags can optionally be specified for the new handler (SA_RESTART is
3593 always used if the system provides it, so need not be specified.) The
3594 return value is a pair with information about the old handler as
3597 This interface does not provide access to the "signal blocking"
3598 facility. Maybe this is not needed, since the thread support may
3599 provide solutions to the problem of consistent access to data
3602 *** A new procedure `flush-all-ports' is equivalent to running
3603 `force-output' on every port open for output.
3605 ** Guile now provides information on how it was built, via the new
3606 global variable, %guile-build-info. This variable records the values
3607 of the standard GNU makefile directory variables as an assocation
3608 list, mapping variable names (symbols) onto directory paths (strings).
3609 For example, to find out where the Guile link libraries were
3610 installed, you can say:
3612 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
3615 * Changes to the scm_ interface
3617 ** The new function scm_handle_by_message_noexit is just like the
3618 existing scm_handle_by_message function, except that it doesn't call
3619 exit to terminate the process. Instead, it prints a message and just
3620 returns #f. This might be a more appropriate catch-all handler for
3621 new dynamic roots and threads.
3624 Changes in Guile 1.1 (released Friday, May 16 1997):
3626 * Changes to the distribution.
3628 The Guile 1.0 distribution has been split up into several smaller
3630 guile-core --- the Guile interpreter itself.
3631 guile-tcltk --- the interface between the Guile interpreter and
3632 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
3633 is a toolkit for building graphical user interfaces.
3634 guile-rgx-ctax --- the interface between Guile and the Rx regular
3635 expression matcher, and the translator for the Ctax
3636 programming language. These are packaged together because the
3637 Ctax translator uses Rx to parse Ctax source code.
3639 This NEWS file describes the changes made to guile-core since the 1.0
3642 We no longer distribute the documentation, since it was either out of
3643 date, or incomplete. As soon as we have current documentation, we
3648 * Changes to the stand-alone interpreter
3650 ** guile now accepts command-line arguments compatible with SCSH, Olin
3651 Shivers' Scheme Shell.
3653 In general, arguments are evaluated from left to right, but there are
3654 exceptions. The following switches stop argument processing, and
3655 stash all remaining command-line arguments as the value returned by
3656 the (command-line) function.
3657 -s SCRIPT load Scheme source code from FILE, and exit
3658 -c EXPR evalute Scheme expression EXPR, and exit
3659 -- stop scanning arguments; run interactively
3661 The switches below are processed as they are encountered.
3662 -l FILE load Scheme source code from FILE
3663 -e FUNCTION after reading script, apply FUNCTION to
3664 command line arguments
3665 -ds do -s script at this point
3666 --emacs enable Emacs protocol (experimental)
3667 -h, --help display this help and exit
3668 -v, --version display version information and exit
3669 \ read arguments from following script lines
3671 So, for example, here is a Guile script named `ekko' (thanks, Olin)
3672 which re-implements the traditional "echo" command:
3674 #!/usr/local/bin/guile -s
3677 (map (lambda (arg) (display arg) (display " "))
3681 (main (command-line))
3683 Suppose we invoke this script as follows:
3685 ekko a speckled gecko
3687 Through the magic of Unix script processing (triggered by the `#!'
3688 token at the top of the file), /usr/local/bin/guile receives the
3689 following list of command-line arguments:
3691 ("-s" "./ekko" "a" "speckled" "gecko")
3693 Unix inserts the name of the script after the argument specified on
3694 the first line of the file (in this case, "-s"), and then follows that
3695 with the arguments given to the script. Guile loads the script, which
3696 defines the `main' function, and then applies it to the list of
3697 remaining command-line arguments, ("a" "speckled" "gecko").
3699 In Unix, the first line of a script file must take the following form:
3701 #!INTERPRETER ARGUMENT
3703 where INTERPRETER is the absolute filename of the interpreter
3704 executable, and ARGUMENT is a single command-line argument to pass to
3707 You may only pass one argument to the interpreter, and its length is
3708 limited. These restrictions can be annoying to work around, so Guile
3709 provides a general mechanism (borrowed from, and compatible with,
3710 SCSH) for circumventing them.
3712 If the ARGUMENT in a Guile script is a single backslash character,
3713 `\', Guile will open the script file, parse arguments from its second
3714 and subsequent lines, and replace the `\' with them. So, for example,
3715 here is another implementation of the `ekko' script:
3717 #!/usr/local/bin/guile \
3721 (for-each (lambda (arg) (display arg) (display " "))
3725 If the user invokes this script as follows:
3727 ekko a speckled gecko
3729 Unix expands this into
3731 /usr/local/bin/guile \ ekko a speckled gecko
3733 When Guile sees the `\' argument, it replaces it with the arguments
3734 read from the second line of the script, producing:
3736 /usr/local/bin/guile -e main -s ekko a speckled gecko
3738 This tells Guile to load the `ekko' script, and apply the function
3739 `main' to the argument list ("a" "speckled" "gecko").
3741 Here is how Guile parses the command-line arguments:
3742 - Each space character terminates an argument. This means that two
3743 spaces in a row introduce an empty-string argument.
3744 - The tab character is not permitted (unless you quote it with the
3745 backslash character, as described below), to avoid confusion.
3746 - The newline character terminates the sequence of arguments, and will
3747 also terminate a final non-empty argument. (However, a newline
3748 following a space will not introduce a final empty-string argument;
3749 it only terminates the argument list.)
3750 - The backslash character is the escape character. It escapes
3751 backslash, space, tab, and newline. The ANSI C escape sequences
3752 like \n and \t are also supported. These produce argument
3753 constituents; the two-character combination \n doesn't act like a
3754 terminating newline. The escape sequence \NNN for exactly three
3755 octal digits reads as the character whose ASCII code is NNN. As
3756 above, characters produced this way are argument constituents.
3757 Backslash followed by other characters is not allowed.
3759 * Changes to the procedure for linking libguile with your programs
3761 ** Guile now builds and installs a shared guile library, if your
3762 system support shared libraries. (It still builds a static library on
3763 all systems.) Guile automatically detects whether your system
3764 supports shared libraries. To prevent Guile from buildisg shared
3765 libraries, pass the `--disable-shared' flag to the configure script.
3767 Guile takes longer to compile when it builds shared libraries, because
3768 it must compile every file twice --- once to produce position-
3769 independent object code, and once to produce normal object code.
3771 ** The libthreads library has been merged into libguile.
3773 To link a program against Guile, you now need only link against
3774 -lguile and -lqt; -lthreads is no longer needed. If you are using
3775 autoconf to generate configuration scripts for your application, the
3776 following lines should suffice to add the appropriate libraries to
3779 ### Find quickthreads and libguile.
3780 AC_CHECK_LIB(qt, main)
3781 AC_CHECK_LIB(guile, scm_shell)
3783 * Changes to Scheme functions
3785 ** Guile Scheme's special syntax for keyword objects is now optional,
3786 and disabled by default.
3788 The syntax variation from R4RS made it difficult to port some
3789 interesting packages to Guile. The routines which accepted keyword
3790 arguments (mostly in the module system) have been modified to also
3791 accept symbols whose names begin with `:'.
3793 To change the keyword syntax, you must first import the (ice-9 debug)
3795 (use-modules (ice-9 debug))
3797 Then you can enable the keyword syntax as follows:
3798 (read-set! keywords 'prefix)
3800 To disable keyword syntax, do this:
3801 (read-set! keywords #f)
3803 ** Many more primitive functions accept shared substrings as
3804 arguments. In the past, these functions required normal, mutable
3805 strings as arguments, although they never made use of this
3808 ** The uniform array functions now operate on byte vectors. These
3809 functions are `array-fill!', `serial-array-copy!', `array-copy!',
3810 `serial-array-map', `array-map', `array-for-each', and
3813 ** The new functions `trace' and `untrace' implement simple debugging
3814 support for Scheme functions.
3816 The `trace' function accepts any number of procedures as arguments,
3817 and tells the Guile interpreter to display each procedure's name and
3818 arguments each time the procedure is invoked. When invoked with no
3819 arguments, `trace' returns the list of procedures currently being
3822 The `untrace' function accepts any number of procedures as arguments,
3823 and tells the Guile interpreter not to trace them any more. When
3824 invoked with no arguments, `untrace' untraces all curretly traced
3827 The tracing in Guile has an advantage over most other systems: we
3828 don't create new procedure objects, but mark the procedure objects
3829 themselves. This means that anonymous and internal procedures can be
3832 ** The function `assert-repl-prompt' has been renamed to
3833 `set-repl-prompt!'. It takes one argument, PROMPT.
3834 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
3835 - If PROMPT is a string, we use it as a prompt.
3836 - If PROMPT is a procedure accepting no arguments, we call it, and
3837 display the result as a prompt.
3838 - Otherwise, we display "> ".
3840 ** The new function `eval-string' reads Scheme expressions from a
3841 string and evaluates them, returning the value of the last expression
3842 in the string. If the string contains no expressions, it returns an
3845 ** The new function `thunk?' returns true iff its argument is a
3846 procedure of zero arguments.
3848 ** `defined?' is now a builtin function, instead of syntax. This
3849 means that its argument should be quoted. It returns #t iff its
3850 argument is bound in the current module.
3852 ** The new syntax `use-modules' allows you to add new modules to your
3853 environment without re-typing a complete `define-module' form. It
3854 accepts any number of module names as arguments, and imports their
3855 public bindings into the current module.
3857 ** The new function (module-defined? NAME MODULE) returns true iff
3858 NAME, a symbol, is defined in MODULE, a module object.
3860 ** The new function `builtin-bindings' creates and returns a hash
3861 table containing copies of all the root module's bindings.
3863 ** The new function `builtin-weak-bindings' does the same as
3864 `builtin-bindings', but creates a doubly-weak hash table.
3866 ** The `equal?' function now considers variable objects to be
3867 equivalent if they have the same name and the same value.
3869 ** The new function `command-line' returns the command-line arguments
3870 given to Guile, as a list of strings.
3872 When using guile as a script interpreter, `command-line' returns the
3873 script's arguments; those processed by the interpreter (like `-s' or
3874 `-c') are omitted. (In other words, you get the normal, expected
3875 behavior.) Any application that uses scm_shell to process its
3876 command-line arguments gets this behavior as well.
3878 ** The new function `load-user-init' looks for a file called `.guile'
3879 in the user's home directory, and loads it if it exists. This is
3880 mostly for use by the code generated by scm_compile_shell_switches,
3881 but we thought it might also be useful in other circumstances.
3883 ** The new function `log10' returns the base-10 logarithm of its
3886 ** Changes to I/O functions
3888 *** The functions `read', `primitive-load', `read-and-eval!', and
3889 `primitive-load-path' no longer take optional arguments controlling
3890 case insensitivity and a `#' parser.
3892 Case sensitivity is now controlled by a read option called
3893 `case-insensitive'. The user can add new `#' syntaxes with the
3894 `read-hash-extend' function (see below).
3896 *** The new function `read-hash-extend' allows the user to change the
3897 syntax of Guile Scheme in a somewhat controlled way.
3899 (read-hash-extend CHAR PROC)
3900 When parsing S-expressions, if we read a `#' character followed by
3901 the character CHAR, use PROC to parse an object from the stream.
3902 If PROC is #f, remove any parsing procedure registered for CHAR.
3904 The reader applies PROC to two arguments: CHAR and an input port.
3906 *** The new functions read-delimited and read-delimited! provide a
3907 general mechanism for doing delimited input on streams.
3909 (read-delimited DELIMS [PORT HANDLE-DELIM])
3910 Read until we encounter one of the characters in DELIMS (a string),
3911 or end-of-file. PORT is the input port to read from; it defaults to
3912 the current input port. The HANDLE-DELIM parameter determines how
3913 the terminating character is handled; it should be one of the
3916 'trim omit delimiter from result
3917 'peek leave delimiter character in input stream
3918 'concat append delimiter character to returned value
3919 'split return a pair: (RESULT . TERMINATOR)
3921 HANDLE-DELIM defaults to 'peek.
3923 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
3924 A side-effecting variant of `read-delimited'.
3926 The data is written into the string BUF at the indices in the
3927 half-open interval [START, END); the default interval is the whole
3928 string: START = 0 and END = (string-length BUF). The values of
3929 START and END must specify a well-defined interval in BUF, i.e.
3930 0 <= START <= END <= (string-length BUF).
3932 It returns NBYTES, the number of bytes read. If the buffer filled
3933 up without a delimiter character being found, it returns #f. If the
3934 port is at EOF when the read starts, it returns the EOF object.
3936 If an integer is returned (i.e., the read is successfully terminated
3937 by reading a delimiter character), then the HANDLE-DELIM parameter
3938 determines how to handle the terminating character. It is described
3939 above, and defaults to 'peek.
3941 (The descriptions of these functions were borrowed from the SCSH
3942 manual, by Olin Shivers and Brian Carlstrom.)
3944 *** The `%read-delimited!' function is the primitive used to implement
3945 `read-delimited' and `read-delimited!'.
3947 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
3949 This returns a pair of values: (TERMINATOR . NUM-READ).
3950 - TERMINATOR describes why the read was terminated. If it is a
3951 character or the eof object, then that is the value that terminated
3952 the read. If it is #f, the function filled the buffer without finding
3953 a delimiting character.
3954 - NUM-READ is the number of characters read into BUF.
3956 If the read is successfully terminated by reading a delimiter
3957 character, then the gobble? parameter determines what to do with the
3958 terminating character. If true, the character is removed from the
3959 input stream; if false, the character is left in the input stream
3960 where a subsequent read operation will retrieve it. In either case,
3961 the character is also the first value returned by the procedure call.
3963 (The descriptions of this function was borrowed from the SCSH manual,
3964 by Olin Shivers and Brian Carlstrom.)
3966 *** The `read-line' and `read-line!' functions have changed; they now
3967 trim the terminator by default; previously they appended it to the
3968 returned string. For the old behavior, use (read-line PORT 'concat).
3970 *** The functions `uniform-array-read!' and `uniform-array-write!' now
3971 take new optional START and END arguments, specifying the region of
3972 the array to read and write.
3974 *** The `ungetc-char-ready?' function has been removed. We feel it's
3975 inappropriate for an interface to expose implementation details this
3978 ** Changes to the Unix library and system call interface
3980 *** The new fcntl function provides access to the Unix `fcntl' system
3983 (fcntl PORT COMMAND VALUE)
3984 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
3985 Values for COMMAND are:
3987 F_DUPFD duplicate a file descriptor
3988 F_GETFD read the descriptor's close-on-exec flag
3989 F_SETFD set the descriptor's close-on-exec flag to VALUE
3990 F_GETFL read the descriptor's flags, as set on open
3991 F_SETFL set the descriptor's flags, as set on open to VALUE
3992 F_GETOWN return the process ID of a socket's owner, for SIGIO
3993 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
3994 FD_CLOEXEC not sure what this is
3996 For details, see the documentation for the fcntl system call.
3998 *** The arguments to `select' have changed, for compatibility with
3999 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
4000 expected behavior. The MILLISECONDS parameter has been changed to
4001 MICROSECONDS, to more closely resemble the underlying system call.
4002 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
4003 corresponding return set will be the same.
4005 *** The arguments to the `mknod' system call have changed. They are
4008 (mknod PATH TYPE PERMS DEV)
4009 Create a new file (`node') in the file system. PATH is the name of
4010 the file to create. TYPE is the kind of file to create; it should
4011 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
4012 permission bits to give the newly created file. If TYPE is
4013 'block-special or 'char-special, DEV specifies which device the
4014 special file refers to; its interpretation depends on the kind of
4015 special file being created.
4017 *** The `fork' function has been renamed to `primitive-fork', to avoid
4018 clashing with various SCSH forks.
4020 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
4021 and `recvfrom!'. They no longer accept a size for a second argument;
4022 you must pass a string to hold the received value. They no longer
4023 return the buffer. Instead, `recv' returns the length of the message
4024 received, and `recvfrom' returns a pair containing the packet's length
4025 and originating address.
4027 *** The file descriptor datatype has been removed, as have the
4028 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
4029 We plan to replace these functions with a SCSH-compatible interface.
4031 *** The `create' function has been removed; it's just a special case
4034 *** There are new functions to break down process termination status
4035 values. In the descriptions below, STATUS is a value returned by
4038 (status:exit-val STATUS)
4039 If the child process exited normally, this function returns the exit
4040 code for the child process (i.e., the value passed to exit, or
4041 returned from main). If the child process did not exit normally,
4042 this function returns #f.
4044 (status:stop-sig STATUS)
4045 If the child process was suspended by a signal, this function
4046 returns the signal that suspended the child. Otherwise, it returns
4049 (status:term-sig STATUS)
4050 If the child process terminated abnormally, this function returns
4051 the signal that terminated the child. Otherwise, this function
4054 POSIX promises that exactly one of these functions will return true on
4055 a valid STATUS value.
4057 These functions are compatible with SCSH.
4059 *** There are new accessors and setters for the broken-out time vectors
4060 returned by `localtime', `gmtime', and that ilk. They are:
4062 Component Accessor Setter
4063 ========================= ============ ============
4064 seconds tm:sec set-tm:sec
4065 minutes tm:min set-tm:min
4066 hours tm:hour set-tm:hour
4067 day of the month tm:mday set-tm:mday
4068 month tm:mon set-tm:mon
4069 year tm:year set-tm:year
4070 day of the week tm:wday set-tm:wday
4071 day in the year tm:yday set-tm:yday
4072 daylight saving time tm:isdst set-tm:isdst
4073 GMT offset, seconds tm:gmtoff set-tm:gmtoff
4074 name of time zone tm:zone set-tm:zone
4076 *** There are new accessors for the vectors returned by `uname',
4077 describing the host system:
4080 ============================================== ================
4081 name of the operating system implementation utsname:sysname
4082 network name of this machine utsname:nodename
4083 release level of the operating system utsname:release
4084 version level of the operating system utsname:version
4085 machine hardware platform utsname:machine
4087 *** There are new accessors for the vectors returned by `getpw',
4088 `getpwnam', `getpwuid', and `getpwent', describing entries from the
4089 system's user database:
4092 ====================== =================
4093 user name passwd:name
4094 user password passwd:passwd
4097 real name passwd:gecos
4098 home directory passwd:dir
4099 shell program passwd:shell
4101 *** There are new accessors for the vectors returned by `getgr',
4102 `getgrnam', `getgrgid', and `getgrent', describing entries from the
4103 system's group database:
4106 ======================= ============
4107 group name group:name
4108 group password group:passwd
4110 group members group:mem
4112 *** There are new accessors for the vectors returned by `gethost',
4113 `gethostbyaddr', `gethostbyname', and `gethostent', describing
4117 ========================= ===============
4118 official name of host hostent:name
4119 alias list hostent:aliases
4120 host address type hostent:addrtype
4121 length of address hostent:length
4122 list of addresses hostent:addr-list
4124 *** There are new accessors for the vectors returned by `getnet',
4125 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
4129 ========================= ===============
4130 official name of net netent:name
4131 alias list netent:aliases
4132 net number type netent:addrtype
4133 net number netent:net
4135 *** There are new accessors for the vectors returned by `getproto',
4136 `getprotobyname', `getprotobynumber', and `getprotoent', describing
4140 ========================= ===============
4141 official protocol name protoent:name
4142 alias list protoent:aliases
4143 protocol number protoent:proto
4145 *** There are new accessors for the vectors returned by `getserv',
4146 `getservbyname', `getservbyport', and `getservent', describing
4150 ========================= ===============
4151 official service name servent:name
4152 alias list servent:aliases
4153 port number servent:port
4154 protocol to use servent:proto
4156 *** There are new accessors for the sockaddr structures returned by
4157 `accept', `getsockname', `getpeername', `recvfrom!':
4160 ======================================== ===============
4161 address format (`family') sockaddr:fam
4162 path, for file domain addresses sockaddr:path
4163 address, for internet domain addresses sockaddr:addr
4164 TCP or UDP port, for internet sockaddr:port
4166 *** The `getpwent', `getgrent', `gethostent', `getnetent',
4167 `getprotoent', and `getservent' functions now return #f at the end of
4168 the user database. (They used to throw an exception.)
4170 Note that calling MUMBLEent function is equivalent to calling the
4171 corresponding MUMBLE function with no arguments.
4173 *** The `setpwent', `setgrent', `sethostent', `setnetent',
4174 `setprotoent', and `setservent' routines now take no arguments.
4176 *** The `gethost', `getproto', `getnet', and `getserv' functions now
4177 provide more useful information when they throw an exception.
4179 *** The `lnaof' function has been renamed to `inet-lnaof'.
4181 *** Guile now claims to have the `current-time' feature.
4183 *** The `mktime' function now takes an optional second argument ZONE,
4184 giving the time zone to use for the conversion. ZONE should be a
4185 string, in the same format as expected for the "TZ" environment variable.
4187 *** The `strptime' function now returns a pair (TIME . COUNT), where
4188 TIME is the parsed time as a vector, and COUNT is the number of
4189 characters from the string left unparsed. This function used to
4190 return the remaining characters as a string.
4192 *** The `gettimeofday' function has replaced the old `time+ticks' function.
4193 The return value is now (SECONDS . MICROSECONDS); the fractional
4194 component is no longer expressed in "ticks".
4196 *** The `ticks/sec' constant has been removed, in light of the above change.
4198 * Changes to the gh_ interface
4200 ** gh_eval_str() now returns an SCM object which is the result of the
4203 ** gh_scm2str() now copies the Scheme data to a caller-provided C
4206 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
4207 and returns the array
4209 ** gh_scm2str0() is gone: there is no need to distinguish
4210 null-terminated from non-null-terminated, since gh_scm2newstr() allows
4211 the user to interpret the data both ways.
4213 * Changes to the scm_ interface
4215 ** The new function scm_symbol_value0 provides an easy way to get a
4216 symbol's value from C code:
4218 SCM scm_symbol_value0 (char *NAME)
4219 Return the value of the symbol named by the null-terminated string
4220 NAME in the current module. If the symbol named NAME is unbound in
4221 the current module, return SCM_UNDEFINED.
4223 ** The new function scm_sysintern0 creates new top-level variables,
4224 without assigning them a value.
4226 SCM scm_sysintern0 (char *NAME)
4227 Create a new Scheme top-level variable named NAME. NAME is a
4228 null-terminated string. Return the variable's value cell.
4230 ** The function scm_internal_catch is the guts of catch. It handles
4231 all the mechanics of setting up a catch target, invoking the catch
4232 body, and perhaps invoking the handler if the body does a throw.
4234 The function is designed to be usable from C code, but is general
4235 enough to implement all the semantics Guile Scheme expects from throw.
4237 TAG is the catch tag. Typically, this is a symbol, but this function
4238 doesn't actually care about that.
4240 BODY is a pointer to a C function which runs the body of the catch;
4241 this is the code you can throw from. We call it like this:
4242 BODY (BODY_DATA, JMPBUF)
4244 BODY_DATA is just the BODY_DATA argument we received; we pass it
4245 through to BODY as its first argument. The caller can make
4246 BODY_DATA point to anything useful that BODY might need.
4247 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
4248 which we have just created and initialized.
4250 HANDLER is a pointer to a C function to deal with a throw to TAG,
4251 should one occur. We call it like this:
4252 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
4254 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
4255 same idea as BODY_DATA above.
4256 THROWN_TAG is the tag that the user threw to; usually this is
4257 TAG, but it could be something else if TAG was #t (i.e., a
4258 catch-all), or the user threw to a jmpbuf.
4259 THROW_ARGS is the list of arguments the user passed to the THROW
4262 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
4263 is just a pointer we pass through to HANDLER. We don't actually
4264 use either of those pointers otherwise ourselves. The idea is
4265 that, if our caller wants to communicate something to BODY or
4266 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
4267 HANDLER can then use. Think of it as a way to make BODY and
4268 HANDLER closures, not just functions; MUMBLE_DATA points to the
4271 Of course, it's up to the caller to make sure that any data a
4272 MUMBLE_DATA needs is protected from GC. A common way to do this is
4273 to make MUMBLE_DATA a pointer to data stored in an automatic
4274 structure variable; since the collector must scan the stack for
4275 references anyway, this assures that any references in MUMBLE_DATA
4278 ** The new function scm_internal_lazy_catch is exactly like
4279 scm_internal_catch, except:
4281 - It does not unwind the stack (this is the major difference).
4282 - If handler returns, its value is returned from the throw.
4283 - BODY always receives #f as its JMPBUF argument (since there's no
4284 jmpbuf associated with a lazy catch, because we don't unwind the
4287 ** scm_body_thunk is a new body function you can pass to
4288 scm_internal_catch if you want the body to be like Scheme's `catch'
4289 --- a thunk, or a function of one argument if the tag is #f.
4291 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
4292 contains the Scheme procedure to invoke as the body, and the tag
4293 we're catching. If the tag is #f, then we pass JMPBUF (created by
4294 scm_internal_catch) to the body procedure; otherwise, the body gets
4297 ** scm_handle_by_proc is a new handler function you can pass to
4298 scm_internal_catch if you want the handler to act like Scheme's catch
4299 --- call a procedure with the tag and the throw arguments.
4301 If the user does a throw to this catch, this function runs a handler
4302 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
4303 variable holding the Scheme procedure object to invoke. It ought to
4304 be a pointer to an automatic variable (i.e., one living on the stack),
4305 or the procedure object should be otherwise protected from GC.
4307 ** scm_handle_by_message is a new handler function to use with
4308 `scm_internal_catch' if you want Guile to print a message and die.
4309 It's useful for dealing with throws to uncaught keys at the top level.
4311 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
4312 message header to print; if zero, we use "guile" instead. That
4313 text is followed by a colon, then the message described by ARGS.
4315 ** The return type of scm_boot_guile is now void; the function does
4316 not return a value, and indeed, never returns at all.
4318 ** The new function scm_shell makes it easy for user applications to
4319 process command-line arguments in a way that is compatible with the
4320 stand-alone guile interpreter (which is in turn compatible with SCSH,
4323 To use the scm_shell function, first initialize any guile modules
4324 linked into your application, and then call scm_shell with the values
4325 of ARGC and ARGV your `main' function received. scm_shell will add
4326 any SCSH-style meta-arguments from the top of the script file to the
4327 argument vector, and then process the command-line arguments. This
4328 generally means loading a script file or starting up an interactive
4329 command interpreter. For details, see "Changes to the stand-alone
4332 ** The new functions scm_get_meta_args and scm_count_argv help you
4333 implement the SCSH-style meta-argument, `\'.
4335 char **scm_get_meta_args (int ARGC, char **ARGV)
4336 If the second element of ARGV is a string consisting of a single
4337 backslash character (i.e. "\\" in Scheme notation), open the file
4338 named by the following argument, parse arguments from it, and return
4339 the spliced command line. The returned array is terminated by a
4342 For details of argument parsing, see above, under "guile now accepts
4343 command-line arguments compatible with SCSH..."
4345 int scm_count_argv (char **ARGV)
4346 Count the arguments in ARGV, assuming it is terminated by a null
4349 For an example of how these functions might be used, see the source
4350 code for the function scm_shell in libguile/script.c.
4352 You will usually want to use scm_shell instead of calling this
4355 ** The new function scm_compile_shell_switches turns an array of
4356 command-line arguments into Scheme code to carry out the actions they
4357 describe. Given ARGC and ARGV, it returns a Scheme expression to
4358 evaluate, and calls scm_set_program_arguments to make any remaining
4359 command-line arguments available to the Scheme code. For example,
4360 given the following arguments:
4362 -e main -s ekko a speckled gecko
4364 scm_set_program_arguments will return the following expression:
4366 (begin (load "ekko") (main (command-line)) (quit))
4368 You will usually want to use scm_shell instead of calling this
4371 ** The function scm_shell_usage prints a usage message appropriate for
4372 an interpreter that uses scm_compile_shell_switches to handle its
4373 command-line arguments.
4375 void scm_shell_usage (int FATAL, char *MESSAGE)
4376 Print a usage message to the standard error output. If MESSAGE is
4377 non-zero, write it before the usage message, followed by a newline.
4378 If FATAL is non-zero, exit the process, using FATAL as the
4379 termination status. (If you want to be compatible with Guile,
4380 always use 1 as the exit status when terminating due to command-line
4383 You will usually want to use scm_shell instead of calling this
4386 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
4387 expressions. It used to return SCM_EOL. Earth-shattering.
4389 ** The macros for declaring scheme objects in C code have been
4390 rearranged slightly. They are now:
4392 SCM_SYMBOL (C_NAME, SCHEME_NAME)
4393 Declare a static SCM variable named C_NAME, and initialize it to
4394 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
4395 be a C identifier, and SCHEME_NAME should be a C string.
4397 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
4398 Just like SCM_SYMBOL, but make C_NAME globally visible.
4400 SCM_VCELL (C_NAME, SCHEME_NAME)
4401 Create a global variable at the Scheme level named SCHEME_NAME.
4402 Declare a static SCM variable named C_NAME, and initialize it to
4403 point to the Scheme variable's value cell.
4405 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
4406 Just like SCM_VCELL, but make C_NAME globally visible.
4408 The `guile-snarf' script writes initialization code for these macros
4409 to its standard output, given C source code as input.
4411 The SCM_GLOBAL macro is gone.
4413 ** The scm_read_line and scm_read_line_x functions have been replaced
4414 by Scheme code based on the %read-delimited! procedure (known to C
4415 code as scm_read_delimited_x). See its description above for more
4418 ** The function scm_sys_open has been renamed to scm_open. It now
4419 returns a port instead of an FD object.
4421 * The dynamic linking support has changed. For more information, see
4422 libguile/DYNAMIC-LINKING.
4427 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
4430 * Changes to the 'guile' program:
4432 ** Guile now loads some new files when it starts up. Guile first
4433 searches the load path for init.scm, and loads it if found. Then, if
4434 Guile is not being used to execute a script, and the user's home
4435 directory contains a file named `.guile', Guile loads that.
4437 ** You can now use Guile as a shell script interpreter.
4439 To paraphrase the SCSH manual:
4441 When Unix tries to execute an executable file whose first two
4442 characters are the `#!', it treats the file not as machine code to
4443 be directly executed by the native processor, but as source code
4444 to be executed by some interpreter. The interpreter to use is
4445 specified immediately after the #! sequence on the first line of
4446 the source file. The kernel reads in the name of the interpreter,
4447 and executes that instead. It passes the interpreter the source
4448 filename as its first argument, with the original arguments
4449 following. Consult the Unix man page for the `exec' system call
4450 for more information.
4452 Now you can use Guile as an interpreter, using a mechanism which is a
4453 compatible subset of that provided by SCSH.
4455 Guile now recognizes a '-s' command line switch, whose argument is the
4456 name of a file of Scheme code to load. It also treats the two
4457 characters `#!' as the start of a comment, terminated by `!#'. Thus,
4458 to make a file of Scheme code directly executable by Unix, insert the
4459 following two lines at the top of the file:
4461 #!/usr/local/bin/guile -s
4464 Guile treats the argument of the `-s' command-line switch as the name
4465 of a file of Scheme code to load, and treats the sequence `#!' as the
4466 start of a block comment, terminated by `!#'.
4468 For example, here's a version of 'echo' written in Scheme:
4470 #!/usr/local/bin/guile -s
4472 (let loop ((args (cdr (program-arguments))))
4475 (display (car args))
4476 (if (pair? (cdr args))
4478 (loop (cdr args)))))
4481 Why does `#!' start a block comment terminated by `!#', instead of the
4482 end of the line? That is the notation SCSH uses, and although we
4483 don't yet support the other SCSH features that motivate that choice,
4484 we would like to be backward-compatible with any existing Guile
4485 scripts once we do. Furthermore, if the path to Guile on your system
4486 is too long for your kernel, you can start the script with this
4490 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
4493 Note that some very old Unix systems don't support the `#!' syntax.
4496 ** You can now run Guile without installing it.
4498 Previous versions of the interactive Guile interpreter (`guile')
4499 couldn't start up unless Guile's Scheme library had been installed;
4500 they used the value of the environment variable `SCHEME_LOAD_PATH'
4501 later on in the startup process, but not to find the startup code
4502 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
4505 To run Guile without installing it, build it in the normal way, and
4506 then set the environment variable `SCHEME_LOAD_PATH' to a
4507 colon-separated list of directories, including the top-level directory
4508 of the Guile sources. For example, if you unpacked Guile so that the
4509 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
4512 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
4515 ** Guile's read-eval-print loop no longer prints #<unspecified>
4516 results. If the user wants to see this, she can evaluate the
4517 expression (assert-repl-print-unspecified #t), perhaps in her startup
4520 ** Guile no longer shows backtraces by default when an error occurs;
4521 however, it does display a message saying how to get one, and how to
4522 request that they be displayed by default. After an error, evaluate
4524 to see a backtrace, and
4525 (debug-enable 'backtrace)
4526 to see them by default.
4530 * Changes to Guile Scheme:
4532 ** Guile now distinguishes between #f and the empty list.
4534 This is for compatibility with the IEEE standard, the (possibly)
4535 upcoming Revised^5 Report on Scheme, and many extant Scheme
4538 Guile used to have #f and '() denote the same object, to make Scheme's
4539 type system more compatible with Emacs Lisp's. However, the change
4540 caused too much trouble for Scheme programmers, and we found another
4541 way to reconcile Emacs Lisp with Scheme that didn't require this.
4544 ** Guile's delq, delv, delete functions, and their destructive
4545 counterparts, delq!, delv!, and delete!, now remove all matching
4546 elements from the list, not just the first. This matches the behavior
4547 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
4548 functions which inspired them.
4550 I recognize that this change may break code in subtle ways, but it
4551 seems best to make the change before the FSF's first Guile release,
4555 ** The compiled-library-path function has been deleted from libguile.
4557 ** The facilities for loading Scheme source files have changed.
4559 *** The variable %load-path now tells Guile which directories to search
4560 for Scheme code. Its value is a list of strings, each of which names
4563 *** The variable %load-extensions now tells Guile which extensions to
4564 try appending to a filename when searching the load path. Its value
4565 is a list of strings. Its default value is ("" ".scm").
4567 *** (%search-load-path FILENAME) searches the directories listed in the
4568 value of the %load-path variable for a Scheme file named FILENAME,
4569 with all the extensions listed in %load-extensions. If it finds a
4570 match, then it returns its full filename. If FILENAME is absolute, it
4571 returns it unchanged. Otherwise, it returns #f.
4573 %search-load-path will not return matches that refer to directories.
4575 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
4576 uses %seach-load-path to find a file named FILENAME, and loads it if
4577 it finds it. If it can't read FILENAME for any reason, it throws an
4580 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
4583 *** load uses the same searching semantics as primitive-load.
4585 *** The functions %try-load, try-load-with-path, %load, load-with-path,
4586 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
4587 path, and load-module-with-path have been deleted. The functions
4588 above should serve their purposes.
4590 *** If the value of the variable %load-hook is a procedure,
4591 `primitive-load' applies its value to the name of the file being
4592 loaded (without the load path directory name prepended). If its value
4593 is #f, it is ignored. Otherwise, an error occurs.
4595 This is mostly useful for printing load notification messages.
4598 ** The function `eval!' is no longer accessible from the scheme level.
4599 We can't allow operations which introduce glocs into the scheme level,
4600 because Guile's type system can't handle these as data. Use `eval' or
4601 `read-and-eval!' (see below) as replacement.
4603 ** The new function read-and-eval! reads an expression from PORT,
4604 evaluates it, and returns the result. This is more efficient than
4605 simply calling `read' and `eval', since it is not necessary to make a
4606 copy of the expression for the evaluator to munge.
4608 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
4609 for the `read' function.
4612 ** The function `int?' has been removed; its definition was identical
4613 to that of `integer?'.
4615 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
4616 use the R4RS names for these functions.
4618 ** The function object-properties no longer returns the hash handle;
4619 it simply returns the object's property list.
4621 ** Many functions have been changed to throw errors, instead of
4622 returning #f on failure. The point of providing exception handling in
4623 the language is to simplify the logic of user code, but this is less
4624 useful if Guile's primitives don't throw exceptions.
4626 ** The function `fileno' has been renamed from `%fileno'.
4628 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
4631 * Changes to Guile's C interface:
4633 ** The library's initialization procedure has been simplified.
4634 scm_boot_guile now has the prototype:
4636 void scm_boot_guile (int ARGC,
4638 void (*main_func) (),
4641 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
4642 MAIN_FUNC should do all the work of the program (initializing other
4643 packages, reading user input, etc.) before returning. When MAIN_FUNC
4644 returns, call exit (0); this function never returns. If you want some
4645 other exit value, MAIN_FUNC may call exit itself.
4647 scm_boot_guile arranges for program-arguments to return the strings
4648 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
4649 scm_set_program_arguments with the final list, so Scheme code will
4650 know which arguments have been processed.
4652 scm_boot_guile establishes a catch-all catch handler which prints an
4653 error message and exits the process. This means that Guile exits in a
4654 coherent way when system errors occur and the user isn't prepared to
4655 handle it. If the user doesn't like this behavior, they can establish
4656 their own universal catcher in MAIN_FUNC to shadow this one.
4658 Why must the caller do all the real work from MAIN_FUNC? The garbage
4659 collector assumes that all local variables of type SCM will be above
4660 scm_boot_guile's stack frame on the stack. If you try to manipulate
4661 SCM values after this function returns, it's the luck of the draw
4662 whether the GC will be able to find the objects you allocate. So,
4663 scm_boot_guile function exits, rather than returning, to discourage
4664 people from making that mistake.
4666 The IN, OUT, and ERR arguments were removed; there are other
4667 convenient ways to override these when desired.
4669 The RESULT argument was deleted; this function should never return.
4671 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
4675 ** Guile's header files should no longer conflict with your system's
4678 In order to compile code which #included <libguile.h>, previous
4679 versions of Guile required you to add a directory containing all the
4680 Guile header files to your #include path. This was a problem, since
4681 Guile's header files have names which conflict with many systems'
4684 Now only <libguile.h> need appear in your #include path; you must
4685 refer to all Guile's other header files as <libguile/mumble.h>.
4686 Guile's installation procedure puts libguile.h in $(includedir), and
4687 the rest in $(includedir)/libguile.
4690 ** Two new C functions, scm_protect_object and scm_unprotect_object,
4691 have been added to the Guile library.
4693 scm_protect_object (OBJ) protects OBJ from the garbage collector.
4694 OBJ will not be freed, even if all other references are dropped,
4695 until someone does scm_unprotect_object (OBJ). Both functions
4698 Note that calls to scm_protect_object do not nest. You can call
4699 scm_protect_object any number of times on a given object, and the
4700 next call to scm_unprotect_object will unprotect it completely.
4702 Basically, scm_protect_object and scm_unprotect_object just
4703 maintain a list of references to things. Since the GC knows about
4704 this list, all objects it mentions stay alive. scm_protect_object
4705 adds its argument to the list; scm_unprotect_object remove its
4706 argument from the list.
4709 ** scm_eval_0str now returns the value of the last expression
4712 ** The new function scm_read_0str reads an s-expression from a
4713 null-terminated string, and returns it.
4715 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
4716 to a Scheme port object.
4718 ** The new function `scm_set_program_arguments' allows C code to set
4719 the value returned by the Scheme `program-arguments' function.
4724 * Guile no longer includes sophisticated Tcl/Tk support.
4726 The old Tcl/Tk support was unsatisfying to us, because it required the
4727 user to link against the Tcl library, as well as Tk and Guile. The
4728 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
4729 referring to widgets by names, rather than exporting widgets to Scheme
4730 code as a special datatype.
4732 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
4733 maintainers described some very interesting changes in progress to the
4734 Tcl/Tk internals, which would facilitate clean interfaces between lone
4735 Tk and other interpreters --- even for garbage-collected languages
4736 like Scheme. They expected the new Tk to be publicly available in the
4739 Since it seems that Guile might soon have a new, cleaner interface to
4740 lone Tk, and that the old Guile/Tk glue code would probably need to be
4741 completely rewritten, we (Jim Blandy and Richard Stallman) have
4742 decided not to support the old code. We'll spend the time instead on
4743 a good interface to the newer Tk, as soon as it is available.
4745 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
4748 Copyright information:
4750 Copyright (C) 1996,1997 Free Software Foundation, Inc.
4752 Permission is granted to anyone to make or distribute verbatim copies
4753 of this document as received, in any medium, provided that the
4754 copyright notice and this permission notice are preserved,
4755 thus giving the recipient permission to redistribute in turn.
4757 Permission is granted to distribute modified versions
4758 of this document, or of portions of it,
4759 under the above conditions, provided also that they
4760 carry prominent notices stating who last changed them.
4765 paragraph-separate: "[
\f]*$"