1 Guile NEWS --- history of user-visible changes.
2 Copyright (C) 1996-2012 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send Guile bug reports to bug-guile@gnu.org.
8 Changes in 2.0.7 (since 2.0.6):
12 ** SRFI-105 curly infix expressions are supported
14 Curly infix expressions as described at
15 http://srfi.schemers.org/srfi-105/srfi-105.html are now supported by
16 Guile's reader. This allows users to write things like {a * {b + c}}
17 instead of (* a (+ b c)). SRFI-105 support is enabled by using the
18 `#!curly-infix' directive in source code, or the `curly-infix' reader
19 option. See the manual for details.
21 ** Reader options may now be per-port
23 Historically, `read-options' and related procedures would manipulate
24 global options, affecting the `read' procedure for all threads, and all
25 current uses of `read'.
27 Guile can now associate `read' options with specific ports, allowing
28 different ports to use different options. For instance, the
29 `#!fold-case' and `#!no-fold-case' reader directives have been
30 implemented, and their effect is to modify the current read options of
31 the current port only; similarly for `#!curly-infix'. Thus, it is
32 possible, for instance, to have one port reading case-sensitive code,
33 while another port reads case-insensitive code.
35 ** Futures may now be nested
37 Futures may now be nested: a future can itself spawn and then `touch'
38 other futures. In addition, any thread that touches a future that has
39 not completed now processes other futures while waiting for the touched
40 future to completed. This allows all threads to be kept busy, and was
41 made possible by the use of delimited continuations (see the manual for
44 Consequently, `par-map' has been rewritten and can now use all cores.
46 ** `load-in-vicinity' search for `.go' files in `%load-compiled-path'
48 Previously, `load-in-vicinity' would look for compiled files in the
49 auto-compilation cache, but not in `%load-compiled-path'. This is now
50 fixed. This affects `load', and the `-l' command-line flag. See
51 <http://bugs.gnu.org/12519> for details.
53 ** Extension search order fixed, and LD_LIBRARY_PATH preserved
55 Up to 2.0.6, Guile would modify the `LD_LIBRARY_PATH' environment
56 variable (or whichever is relevant for the host OS) to insert its own
57 default extension directories in the search path (using GNU libltdl
58 facilities was not possible here.) This approach was problematic in two
61 First, the `LD_LIBRARY_PATH' modification would be visible to
62 sub-processes, and would also affect future calls to `dlopen', which
63 could lead to subtle bugs in the application or sub-processes. Second,
64 when the installation prefix is /usr, the `LD_LIBRARY_PATH' modification
65 would typically end up inserting /usr/lib before /usr/local/lib in the
66 search path, which is often the opposite of system-wide settings such as
69 Both issues have now been fixed.
71 ** `make-vtable-vtable' is now deprecated
73 Programs should instead use `make-vtable' and `<standard-vtable>'.
75 ** The `-Wduplicate-case-datum' and `-Wbad-case-datum' are enabled
77 These recently introduced warnings have been documented and are now
78 enabled by default when auto-compiling.
80 ** Optimize calls to `equal?' or `eqv?' with a constant argument
82 The compiler simplifies calls to `equal?' or `eqv?' with a constant
83 argument to use `eq?' instead, when applicable.
87 ** SRFI-9 records now documented under "Compound Data Types"
89 The documentation of SRFI-9 record types has been moved in the "Compound
90 Data Types", next to Guile's other record APIs. A new section
91 introduces the various record APIs, and describes the trade-offs they
92 make. These changes were made in an attempt to better guide users
93 through the maze of records API, and to recommend SRFI-9 as the main
96 The documentation of Guile's raw `struct' API has also been improved.
98 ** (ice-9 and-let-star) and (ice-9 curried-definitions) now documented
100 These modules were missing from the manual.
104 ** New "functional record setters" as a GNU extension of SRFI-9
106 The (srfi srfi-9 gnu) module now provides three new macros to deal with
107 "updates" of immutable records: `define-immutable-record-type',
108 `set-field', and `set-fields'.
110 The first one allows record type "functional setters" to be defined;
111 such setters keep the record unchanged, and instead return a new record
112 with only one different field. The remaining macros provide the same
113 functionality, and also optimize updates of multiple or nested fields.
114 See the manual for details.
116 ** web: New `http-get*', `response-body-port', and `text-content-type?'
119 These procedures return a port from which to read the response's body.
120 Unlike `http-get' and `read-response-body', they allow the body to be
121 processed incrementally instead of being stored entirely in memory.
123 The `text-content-type?' predicate allows users to determine whether the
124 content type of a response is textual.
126 See the manual for details.
128 ** `string-split' accepts character sets and predicates
130 The `string-split' procedure can now be given a SRFI-14 character set or
131 a predicate, instead of just a character.
133 ** R6RS SRFI support --- FIXME
135 5d7c55b R6RS srfi library names should ignore first identifier after the :n
136 acc1d8e Preserve additional R6RS library name components after srfi :n
138 ** `define-public' is no a longer curried definition by default
140 The (ice-9 curried-definitions) should be used for such uses. See the
145 ** Remove reference to `scm_init_popen' when `fork' is unavailable
147 This fixes a MinGW build issue (http://bugs.gnu.org/12477).
149 ** Fix race between installing `guild' and the `guile-tools' symlink
153 ** Procedures returned by `eval' now have docstrings
154 (http://bugs.gnu.org/12173)
155 ** web client: correctly handle uri-query, etc. in relative URI headers
156 (http://bugs.gnu.org/12827)
157 ** Fix docs for R6RS `hashtable-copy'
158 ** R6RS `string-for-each' now accepts multiple string arguments
159 ** Fix out-of-range error in the compiler's CSE pass
160 (http://bugs.gnu.org/12883)
161 ** Add missing R6RS `open-file-input/output-port' procedure
162 ** Futures: Avoid creating the worker pool more than once
163 ** Fix invalid assertion about mutex ownership in threads.c
164 (http://bugs.gnu.org/12719)
165 ** Have `SCM_NUM2FLOAT' and `SCM_NUM2DOUBLE' use `scm_to_double'
166 ** The `scandir' procedure now uses `lstat' instead of `stat'
167 ** Fix `generalized-vector->list' indexing bug with shared arrays
168 (http://bugs.gnu.org/12465)
169 ** web: Change `http-get' to try all the addresses for the given URI
170 ** Implement `hash' for structs
171 (http://lists.gnu.org/archive/html/guile-devel/2012-10/msg00031.html)
172 ** `read' now adds source properties for data types beyond pairs
173 ** Improve error reporting in `append!'
174 ** In fold-matches, set regexp/notbol unless matching string start
175 ** Don't stat(2) and access(2) the .go location before using it
176 ** SRFI-19: use zero padding for hours in ISO 8601 format, not blanks
177 ** web: Fix uri-encoding for strings with no unreserved chars, and octets 0-15
178 ** More robust texinfo alias handling
179 ** Optimize `format' and `simple-format'
180 (http://bugs.gnu.org/12033)
181 ** Angle of -0.0 is pi, not zero
184 Changes in 2.0.6 (since 2.0.5):
188 ** New optimization pass: common subexpression elimination (CSE)
190 Guile's optimizer will now run a CSE pass after partial evaluation.
191 This pass propagates static information about branches taken, bound
192 lexicals, and effects from an expression's dominators. It can replace
193 common subexpressions with their boolean values (potentially enabling
194 dead code elimination), equivalent bound lexicals, or it can elide them
195 entirely, depending on the context in which they are executed. This
196 pass is especially useful in removing duplicate type checks, such as
197 those produced by SRFI-9 record accessors.
199 ** Improvements to the partial evaluator
201 Peval can now hoist tests that are common to both branches of a
202 conditional into the test. This can help with long chains of
203 conditionals, such as those generated by the `match' macro. Peval can
204 now do simple beta-reductions of procedures with rest arguments. It
205 also avoids residualizing degenerate lexical aliases, even when full
206 inlining is not possible. Finally, peval now uses the effects analysis
207 introduced for the CSE pass. More precise effects analysis allows peval
210 ** Run finalizers asynchronously in asyncs
212 Finalizers are now run asynchronously, via an async. See Asyncs in the
213 manual. This allows Guile and user code to safely allocate memory while
216 ** Update SRFI-14 character sets to Unicode 6.1
218 Note that this update causes the Latin-1 characters `§' and `¶' to be
219 reclassified as punctuation. They were previously considered to be part
220 of `char-set:symbol'.
222 ** Better source information for datums
224 When the `positions' reader option is on, as it is by default, Guile's
225 reader will record source information for more kinds of datums.
227 ** Improved error and warning messages
229 `syntax-violation' errors now prefer `subform' for source info, with
230 `form' as fallback. Syntactic errors in `cond' and `case' now produce
231 better errors. `case' can now warn on duplicate datums, or datums that
232 cannot be usefully compared with `eqv?'. `-Warity-mismatch' now handles
233 applicable structs. `-Wformat' is more robust in the presence of
234 `gettext'. Finally, various exceptions thrown by the Web modules now
235 define appropriate exception printers.
237 ** A few important bug fixes in the HTTP modules.
239 Guile's web server framework now checks if an application returns a body
240 where it is not permitted, for example in response to a HEAD request,
241 and warn or truncate the response as appropriate. Bad requests now
242 cause a 400 Bad Request response to be printed before closing the port.
243 Finally, some date-printing and URL-parsing bugs were fixed.
245 ** Pretty-print improvements
247 When Guile needs to pretty-print Tree-IL, it will try to reconstruct
248 `cond', `or`, and other derived syntax forms from the primitive tree-IL
249 forms. It also uses the original names instead of the fresh unique
250 names, when it is unambiguous to do so. This can be seen in the output
251 of REPL commands like `,optimize'.
253 Also, the `pretty-print' procedure has a new keyword argument,
256 ** Fix memory leak involving applicable SMOBs
258 At some point in the 1.9.x series, Guile began leaking any applicable
259 SMOB that was actually applied. (There was a weak-key map from SMOB to
260 trampoline functions, where the value had a strong reference on the
261 key.) This has been fixed. There was much rejoicing!
263 ** Support for HTTP/1.1 chunked transfer coding
265 See "Transfer Codings" in the manual, for more.
267 ** Micro-optimizations
269 A pile of micro-optimizations: the `string-trim' function when called
270 with `char-set:whitespace'; the `(web http)' parsers; SMOB application;
271 conversion of raw UTF-8 and UTF-32 data to and from SCM strings; vlists
272 and vhashes; `read' when processing string literals.
274 ** Incompatible change to `scandir'
276 As was the original intention, `scandir' now runs the `select?'
277 procedure on all items, including subdirectories and the `.' and `..'
278 entries. It receives the basename of the file in question instead of
279 the full name. We apologize for this incompatible change to this
280 function introduced in the 2.0.4 release.
284 The manual has been made much more consistent in its naming conventions
285 with regards to formal parameters of functions. Thanks to Bake Timmons.
289 ** New C function: `scm_to_pointer'
290 ** New C inline functions: `scm_new_smob', `scm_new_double_smob'
291 ** (ice-9 format): Add ~h specifier for localized number output.
292 ** (web response): New procedure: `response-must-not-include-body?'
293 ** New predicate: 'supports-source-properties?'
294 ** New C helpers: `scm_c_values', `scm_c_nvalues'
295 ** Newly public inline C function: `scm_unget_byte'
296 ** (language tree-il): New functions: `tree-il=?', `tree-il-hash'
297 ** New fluid: `%default-port-conversion-strategy'
298 ** New syntax: `=>' within `case'
299 ** (web http): `make-chunked-input-port', `make-chunked-output-port'
300 ** (web http): `declare-opaque-header!'
302 Search the manual for these identifiers, for more information.
306 ** `close-io-port' deprecated
310 ** `scm_sym2var' deprecated
312 In most cases, replace with `scm_lookup' or `scm_module_variable'. Use
313 `scm_define' or `scm_module_ensure_local_variable' if the second
314 argument is nonzero. See "Accessing Modules from C" in the manual, for
317 ** Lookup closures deprecated
319 These were never documented. See "Module System Reflection" in the
320 manual for replacements.
324 ** Fix compilation against uninstalled Guile on non-GNU platforms.
325 ** Fix `SCM_I_ERROR' definition for MinGW without networking.
326 ** Fix compilation with the Sun C compiler.
327 ** Fix check for `clock_gettime' on OpenBSD and some other systems.
328 ** Fix build with --enable-debug-malloc.
329 ** Honor $(program_transform_name) for the `guile-tools' symlink.
330 ** Fix cross-compilation of GOOPS-using code.
334 ** Fix use of unitialized stat buffer in search-path of absolute paths.
335 ** Avoid calling `freelocale' with a NULL argument.
336 ** Work around erroneous tr_TR locale in Darwin 8 in tests.
337 ** Fix `getaddrinfo' test for Darwin 8.
338 ** Use Gnulib's `regex' module for better regex portability.
339 ** `source-properties' and friends work on any object
340 ** Rewrite open-process in C, for robustness related to threads and fork
341 ** Fix <TAG>vector-length when applied to other uniform vector types
342 ** Fix escape-only prompt optimization (was disabled previously)
343 ** Fix a segfault when /dev/urandom is not accessible
344 ** Fix flush on soft ports, so that it actually runs.
345 ** Better compatibility of SRFI-9 records with core records
346 ** Fix and clarify documentation of `sorted?'.
347 ** Fix IEEE-754 endianness conversion in bytevectors.
348 ** Correct thunk check in the `wind' instruction.
349 ** Add @acronym support to texinfo modules
350 ** Fix docbook->texi for <ulink> without URL
351 ** Fix `setvbuf' to leave the line/column number unchanged.
352 ** Add missing public declaration for `scm_take_from_input_buffers'.
353 ** Fix relative file name canonicalization with empty %LOAD-PATH entries.
354 ** Import newer (ice-9 match) from Chibi-Scheme.
355 ** Fix unbound variables and unbound values in ECMAScript runtime.
356 ** Make SRFI-6 string ports Unicode-capable.
359 Changes in 2.0.5 (since 2.0.4):
361 This release fixes the binary interface information (SONAME) of
362 libguile, which was incorrect in 2.0.4. It does not contain other
366 Changes in 2.0.4 (since 2.0.3):
370 ** Better debuggability for interpreted procedures.
372 Guile 2.0 came with a great debugging experience for compiled
373 procedures, but the story for interpreted procedures was terrible. Now,
374 at least, interpreted procedures have names, and the `arity' procedure
375 property is always correct (or, as correct as it can be, in the presence
378 ** Support for cross-compilation.
380 One can now use a native Guile to cross-compile `.go' files for a
381 different architecture. See the documentation for `--target' in the
382 "Compilation" section of the manual, for information on how to use the
383 cross-compiler. See the "Cross building Guile" section of the README,
384 for more on how to cross-compile Guile itself.
386 ** The return of `local-eval'.
388 Back by popular demand, `the-environment' and `local-eval' allow the
389 user to capture a lexical environment, and then evaluate arbitrary
390 expressions in that context. There is also a new `local-compile'
391 command. See "Local Evaluation" in the manual, for more. Special
392 thanks to Mark Weaver for an initial implementation of this feature.
394 ** Fluids can now have default values.
396 Fluids are used for dynamic and thread-local binding. They have always
397 inherited their values from the context or thread that created them.
398 However, there was a case in which a new thread would enter Guile, and
399 the default values of all the fluids would be `#f' for that thread.
401 This has now been fixed so that `make-fluid' has an optional default
402 value for fluids in unrelated dynamic roots, which defaults to `#f'.
404 ** Garbage collector tuning.
406 The garbage collector has now been tuned to run more often under some
409 *** Unmanaged allocation
411 The new `scm_gc_register_allocation' function will notify the collector
412 of unmanaged allocation. This will cause the collector to run sooner.
413 Guile's `scm_malloc', `scm_calloc', and `scm_realloc' unmanaged
414 allocators eventually call this function. This leads to better
415 performance under steady-state unmanaged allocation.
417 *** Transient allocation
419 When the collector runs, it will try to record the total memory
420 footprint of a process, if the platform supports this information. If
421 the memory footprint is growing, the collector will run more frequently.
422 This reduces the increase of the resident size of a process in response
423 to a transient increase in allocation.
425 *** Management of threads, bignums
427 Creating a thread will allocate a fair amount of memory. Guile now does
428 some GC work (using `GC_collect_a_little') when allocating a thread.
429 This leads to a better memory footprint when creating many short-lived
432 Similarly, bignums can occupy a lot of memory. Guile now offers hooks
433 to enable custom GMP allocators that end up calling
434 `scm_gc_register_allocation'. These allocators are enabled by default
435 when running Guile from the command-line. To enable them in libraries,
436 set the `scm_install_gmp_memory_functions' variable to a nonzero value
437 before loading Guile.
439 ** SRFI-39 parameters are available by default.
441 Guile now includes support for parameters, as defined by SRFI-39, in the
442 default environment. See "Parameters" in the manual, for more
443 information. `current-input-port', `current-output-port', and
444 `current-error-port' are now parameters.
446 ** Add `current-warning-port'.
448 Guile now outputs warnings on a separate port, `current-warning-port',
449 initialized to the value that `current-error-port' has on startup.
451 ** Syntax parameters.
453 Following Racket's lead, Guile now supports syntax parameters. See
454 "Syntax parameters" in the manual, for more.
456 Also see Barzilay, Culpepper, and Flatt's 2011 SFP workshop paper,
457 "Keeping it Clean with syntax-parameterize".
459 ** Parse command-line arguments from the locale encoding.
461 Guile now attempts to parse command-line arguments using the user's
462 locale. However for backwards compatibility with other 2.0.x releases,
463 it does so without actually calling `setlocale'. Please report any bugs
464 in this facility to bug-guile@gnu.org.
466 ** One-armed conditionals: `when' and `unless'
468 Guile finally has `when' and `unless' in the default environment. Use
469 them whenever you would use an `if' with only one branch. See
470 "Conditionals" in the manual, for more.
472 ** `current-filename', `add-to-load-path'
474 There is a new form, `(current-filename)', which expands out to the
475 source file in which it occurs. Combined with the new
476 `add-to-load-path', this allows simple scripts to easily add nearby
477 directories to the load path. See "Load Paths" in the manual, for more.
479 ** `random-state-from-platform'
481 This procedure initializes a random seed using good random sources
482 available on your platform, such as /dev/urandom. See "Random Number
483 Generation" in the manual, for more.
485 ** Warn about unsupported `simple-format' options.
487 The `-Wformat' compilation option now reports unsupported format options
488 passed to `simple-format'.
492 Besides the sections already mentioned, the following manual sections
493 are new in this release: "Modules and the File System", "Module System
494 Reflection", "Syntax Transformer Helpers", and "Local Inclusion".
498 ** (ice-9 session): `apropos-hook'
499 ** New print option: `escape-newlines', defaults to #t.
500 ** (ice-9 ftw): `file-system-fold', `file-system-tree', `scandir'
501 ** `scm_c_value_ref': access to multiple returned values from C
502 ** scm_call (a varargs version), scm_call_7, scm_call_8, scm_call_9
503 ** Some new syntax helpers in (system syntax)
505 Search the manual for these identifiers and modules, for more.
509 ** FreeBSD build fixes.
510 ** OpenBSD compilation fixes.
511 ** Solaris 2.10 test suite fixes.
512 ** IA64 compilation fix.
513 ** MinGW build fixes.
514 ** Work around instruction reordering on SPARC and HPPA in the VM.
515 ** Gnulib updates: added `dirfd', `setenv' modules.
519 ** Add a deprecated alias for $expt.
520 ** Add an exception printer for `getaddrinfo-error'.
521 ** Add deprecated shim for `scm_display_error' with stack as first argument.
522 ** Add warnings for unsupported `simple-format' options.
523 ** Allow overlapping regions to be passed to `bytevector-copy!'.
524 ** Better function prologue disassembly
525 ** Compiler: fix miscompilation of (values foo ...) in some contexts.
526 ** Compiler: fix serialization of #nil-terminated lists.
527 ** Compiler: allow values bound in non-tail let expressions to be collected.
528 ** Deprecate SCM_ASRTGO.
529 ** Document invalidity of (begin) as expression; add back-compat shim.
530 ** Don't leak file descriptors when mmaping objcode.
531 ** Empty substrings no longer reference the original stringbuf.
532 ** FFI: Fix `set-pointer-finalizer!' to leave the type cell unchanged.
533 ** FFI: Hold a weak reference to the CIF made by `procedure->pointer'.
534 ** FFI: Hold a weak reference to the procedure passed to `procedure->pointer'.
535 ** FFI: Properly unpack small integer return values in closure call.
536 ** Fix R6RS `fold-left' so the accumulator is the first argument.
537 ** Fix bit-set*! bug from 2005.
538 ** Fix bug in `make-repl' when `lang' is actually a <language>.
539 ** Fix bugs related to mutation, the null string, and shared substrings.
540 ** Fix <dynwind> serialization.
541 ** Fix erroneous check in `set-procedure-properties!'.
542 ** Fix generalized-vector-{ref,set!} for slices.
543 ** Fix error messages involving definition forms.
544 ** Fix primitive-eval to return #<unspecified> for definitions.
545 ** HTTP: Extend handling of "Cache-Control" header.
546 ** HTTP: Fix qstring writing of cache-extension values
547 ** HTTP: Fix validators for various list-style headers.
548 ** HTTP: Permit non-date values for Expires header.
549 ** HTTP: `write-request-line' writes absolute paths, not absolute URIs.
550 ** Hack the port-column of current-output-port after printing a prompt.
551 ** Make sure `regexp-quote' tests use Unicode-capable string ports.
552 ** Peval: Fix bugs in the new optimizer.
553 ** Statistically unique marks and labels, for robust hygiene across sessions.
554 ** Web: Allow URIs with empty authorities, like "file:///etc/hosts".
555 ** `,language' at REPL sets the current-language fluid.
556 ** `primitive-load' returns the value(s) of the last expression.
557 ** `scm_from_stringn' always returns unique strings.
558 ** `scm_i_substring_copy' tries to narrow the substring.
559 ** i18n: Fix gc_malloc/free mismatch on non-GNU systems.
562 Changes in 2.0.3 (since 2.0.2):
566 ** Guile has a new optimizer, `peval'.
568 `Peval' is a partial evaluator that performs constant folding, dead code
569 elimination, copy propagation, and inlining. By default it runs on
570 every piece of code that Guile compiles, to fold computations that can
571 happen at compile-time, so they don't have to happen at runtime.
573 If we did our job right, the only impact you would see would be your
574 programs getting faster. But if you notice slowdowns or bloated code,
575 please send a mail to bug-guile@gnu.org with details.
577 Thanks to William R. Cook, Oscar Waddell, and Kent Dybvig for inspiring
578 peval and its implementation.
580 You can see what peval does on a given piece of code by running the new
581 `,optimize' REPL meta-command, and comparing it to the output of
582 `,expand'. See "Compile Commands" in the manual, for more.
584 ** Fewer calls to `stat'.
586 Guile now stats only the .go file and the .scm file when loading a fresh
591 ** New module: `(web client)', a simple synchronous web client.
593 See "Web Client" in the manual, for more.
595 ** Users can now install compiled `.go' files.
597 See "Installing Site Packages" in the manual.
599 ** Remove Front-Cover and Back-Cover text from the manual.
601 The manual is still under the GNU Free Documentation License, but no
602 longer has any invariant sections.
604 ** More helpful `guild help'.
606 `guild' is Guile's multi-tool, for use in shell scripting. Now it has a
607 nicer interface for querying the set of existing commands, and getting
608 help on those commands. Try it out and see!
610 ** New macro: `define-syntax-rule'
612 `define-syntax-rule' is a shorthand to make a `syntax-rules' macro with
613 one clause. See "Syntax Rules" in the manual, for more.
615 ** The `,time' REPL meta-command now has more precision.
617 The output of this command now has microsecond precision, instead of
618 10-millisecond precision.
620 ** `(ice-9 match)' can now match records.
622 See "Pattern Matching" in the manual, for more on matching records.
624 ** New module: `(language tree-il debug)'.
626 This module provides a tree-il verifier. This is useful for people that
627 generate tree-il, usually as part of a language compiler.
629 ** New functions: `scm_is_exact', `scm_is_inexact'.
631 These provide a nice C interface for Scheme's `exact?' and `inexact?',
636 See the git log (or the ChangeLog) for more details on these bugs.
638 ** Fix order of importing modules and resolving duplicates handlers.
639 ** Fix a number of bugs involving extended (merged) generics.
640 ** Fix invocation of merge-generics duplicate handler.
641 ** Fix write beyond array end in arrays.c.
642 ** Fix read beyond end of hashtable size array in hashtab.c.
643 ** (web http): Locale-independent parsing and serialization of dates.
644 ** Ensure presence of Host header in HTTP/1.1 requests.
645 ** Fix take-right and drop-right for improper lists.
646 ** Fix leak in get_current_locale().
647 ** Fix recursive define-inlinable expansions.
648 ** Check that srfi-1 procedure arguments are procedures.
649 ** Fix r6rs `map' for multiple returns.
650 ** Fix scm_tmpfile leak on POSIX platforms.
651 ** Fix a couple of leaks (objcode->bytecode, make-boot-program).
652 ** Fix guile-lib back-compatibility for module-stexi-documentation.
653 ** Fix --listen option to allow other ports.
654 ** Fix scm_to_latin1_stringn for substrings.
655 ** Fix compilation of untyped arrays of rank not 1.
656 ** Fix unparse-tree-il of <dynset>.
657 ** Fix reading of #||||#.
658 ** Fix segfault in GOOPS when class fields are redefined.
659 ** Prefer poll(2) over select(2) to allow file descriptors above FD_SETSIZE.
662 Changes in 2.0.2 (since 2.0.1):
666 ** `guile-tools' renamed to `guild'
668 The new name is shorter. Its intended future use is for a CPAN-like
669 system for Guile wizards and journeyfolk to band together to share code;
670 hence the name. `guile-tools' is provided as a backward-compatible
671 symbolic link. See "Using Guile Tools" in the manual, for more.
673 ** New control operators: `shift' and `reset'
675 See "Shift and Reset" in the manual, for more information.
677 ** `while' as an expression
679 Previously the return value of `while' was unspecified. Now its
680 values are specified both in the case of normal termination, and via
681 termination by invoking `break', possibly with arguments. See "while
682 do" in the manual for more.
684 ** Disallow access to handles of weak hash tables
686 `hash-get-handle' and `hash-create-handle!' are no longer permitted to
687 be called on weak hash tables, because the fields in a weak handle could
688 be nulled out by the garbage collector at any time, but yet they are
689 otherwise indistinguishable from pairs. Use `hash-ref' and `hash-set!'
692 ** More precision for `get-internal-run-time', `get-internal-real-time'
694 On 64-bit systems which support POSIX clocks, Guile's internal timing
695 procedures offer nanosecond resolution instead of the 10-millisecond
696 resolution previously available. 32-bit systems now use 1-millisecond
699 ** Guile now measures time spent in GC
701 `gc-stats' now returns a meaningful value for `gc-time-taken'.
705 The statprof profiler now exports a `gcprof' procedure, driven by the
706 `after-gc-hook', to see which parts of your program are causing GC. Let
707 us know if you find it useful.
709 ** `map', `for-each' and some others now implemented in Scheme
711 We would not mention this in NEWS, as it is not a user-visible change,
712 if it were not for one thing: `map' and `for-each' are no longer
713 primitive generics. Instead they are normal bindings, which can be
714 wrapped by normal generics. This fixes some modularity issues between
715 core `map', SRFI-1 `map', and GOOPS.
717 Also it's pretty cool that we can do this without a performance impact.
719 ** Add `scm_peek_byte_or_eof'.
721 This helper is like `scm_peek_char_or_eof', but for bytes instead of
724 ** Implement #:stop-at-first-non-option option for getopt-long
726 See "getopt-long Reference" in the manual, for more information.
728 ** Improve R6RS conformance for conditions in the I/O libraries
730 The `(rnrs io simple)' module now raises the correct R6RS conditions in
731 error cases. `(rnrs io ports)' is also more correct now, though it is
732 still a work in progress.
734 ** All deprecated routines emit warnings
736 A few deprecated routines were lacking deprecation warnings. This has
741 ** Constants in compiled code now share state better
743 Constants with shared state, like `("foo")' and `"foo"', now share state
744 as much as possible, in the entire compilation unit. This cuts compiled
745 `.go' file sizes in half, generally, and speeds startup.
747 ** VLists: optimize `vlist-fold-right', and add `vhash-fold-right'
749 These procedures are now twice as fast as they were.
751 ** UTF-8 ports to bypass `iconv' entirely
753 This reduces memory usage in a very common case.
757 The compiler is now about 40% faster. (Note that this is only the case
758 once the compiler is itself compiled, so the build still takes as long
763 Some assertions that were mostly useful for sanity-checks on the
764 bytecode compiler are now off for both "regular" and "debug" engines.
765 This together with a fix to cache a TLS access and some other tweaks
766 improve the VM's performance by about 20%.
768 ** SRFI-1 list-set optimizations
770 lset-adjoin and lset-union now have fast paths for eq? sets.
772 ** `memq', `memv' optimizations
774 These procedures are now at least twice as fast than in 2.0.1.
778 ** Deprecate scm_whash API
780 `scm_whash_get_handle', `SCM_WHASHFOUNDP', `SCM_WHASHREF',
781 `SCM_WHASHSET', `scm_whash_create_handle', `scm_whash_lookup', and
782 `scm_whash_insert' are now deprecated. Use the normal hash table API
785 ** Deprecate scm_struct_table
787 `SCM_STRUCT_TABLE_NAME', `SCM_SET_STRUCT_TABLE_NAME',
788 `SCM_STRUCT_TABLE_CLASS', `SCM_SET_STRUCT_TABLE_CLASS',
789 `scm_struct_table', and `scm_struct_create_handle' are now deprecated.
790 These routines formed part of the internals of the map between structs
793 ** Deprecate scm_internal_dynamic_wind
795 The `scm_t_inner' type and `scm_internal_dynamic_wind' are deprecated,
796 as the `scm_dynwind' API is better, and this API encourages users to
797 stuff SCM values into pointers.
799 ** Deprecate scm_immutable_cell, scm_immutable_double_cell
801 These routines are deprecated, as the GC_STUBBORN API doesn't do
806 Andreas Rottman kindly transcribed the missing parts of the `(rnrs io
807 ports)' documentation from the R6RS documentation. Thanks Andreas!
811 ** Fix double-loading of script in -ds case
812 ** -x error message fix
813 ** iconveh-related cross-compilation fixes
814 ** Fix small integer return value packing on big endian machines.
815 ** Fix hash-set! in weak-value table from non-immediate to immediate
816 ** Fix call-with-input-file & relatives for multiple values
817 ** Fix `hash' for inf and nan
818 ** Fix libguile internal type errors caught by typing-strictness==2
819 ** Fix compile error in MinGW fstat socket detection
820 ** Fix generation of auto-compiled file names on MinGW
821 ** Fix multithreaded access to internal hash tables
822 ** Emit a 1-based line number in error messages
823 ** Fix define-module ordering
824 ** Fix several POSIX functions to use the locale encoding
825 ** Add type and range checks to the complex generalized vector accessors
826 ** Fix unaligned accesses for bytevectors of complex numbers
828 ** Fix erroneous VM stack overflow for canceled threads
831 Changes in 2.0.1 (since 2.0.0):
835 ** guile.m4 supports linking with rpath
837 The GUILE_FLAGS macro now sets GUILE_LIBS and GUILE_LTLIBS, which
838 include appropriate directives to the linker to include libguile-2.0.so
839 in the runtime library lookup path.
841 ** `begin' expands macros in its body before other expressions
843 This enables support for programs like the following:
848 (or (= x 0) (odd? (- x 1)))))
851 ((odd? x) (not (even? x)))))
854 ** REPL reader usability enhancements
856 The REPL now flushes input after a read error, which should prevent one
857 error from causing other errors. The REPL also now interprets comments
860 ** REPL output has configurable width
862 The REPL now defaults to output with the current terminal's width, in
863 columns. See "Debug Commands" in the manual for more information on
866 ** Better C access to the module system
868 Guile now has convenient C accessors to look up variables or values in
869 modules and their public interfaces. See `scm_c_public_ref' and friends
870 in "Accessing Modules from C" in the manual.
872 ** Added `scm_call_5', `scm_call_6'
874 See "Fly Evaluation" in the manual.
876 ** Added `scm_from_latin1_keyword', `scm_from_utf8_keyword'
878 See "Keyword Procedures" in the manual, for more. Note that
879 `scm_from_locale_keyword' should not be used when the name is a C string
882 ** R6RS unicode and string I/O work
884 Added efficient implementations of `get-string-n' and `get-string-n!'
885 for binary ports. Exported `current-input-port', `current-output-port'
886 and `current-error-port' from `(rnrs io ports)', and enhanced support
889 ** Added `pointer->scm', `scm->pointer' to `(system foreign)'
891 These procedure are useful if one needs to pass and receive SCM values
892 to and from foreign functions. See "Foreign Variables" in the manual,
895 ** Added `heap-allocated-since-gc' to `(gc-stats)'
897 Also fixed the long-standing bug in the REPL `,stat' command.
899 ** Add `on-error' REPL option
901 This option controls what happens when an error occurs at the REPL, and
902 defaults to `debug', indicating that Guile should enter the debugger.
903 Other values include `report', which will simply print a backtrace
904 without entering the debugger. See "System Commands" in the manual.
906 ** Enforce immutability of string literals
908 Attempting to mutate a string literal now causes a runtime error.
910 ** Fix pthread redirection
912 Guile 2.0.0 shipped with headers that, if configured with pthread
913 support, would re-define `pthread_create', `pthread_join', and other API
914 to redirect to the BDW-GC wrappers, `GC_pthread_create', etc. This was
915 unintended, and not necessary: because threads must enter Guile with
916 `scm_with_guile', Guile can handle thread registration itself, without
917 needing to make the GC aware of all threads. This oversight has been
920 ** `with-continuation-barrier' now unwinds on `quit'
922 A throw to `quit' in a continuation barrier will cause Guile to exit.
923 Before, it would do so before unwinding to the barrier, which would
924 prevent cleanup handlers from running. This has been fixed so that it
925 exits only after unwinding.
927 ** `string->pointer' and `pointer->string' have optional encoding arg
929 This allows users of the FFI to more easily deal in strings with
930 particular (non-locale) encodings, like "utf-8". See "Void Pointers and
931 Byte Access" in the manual, for more.
933 ** R6RS fixnum arithmetic optimizations
935 R6RS fixnum operations are are still slower than generic arithmetic,
938 ** New procedure: `define-inlinable'
940 See "Inlinable Procedures" in the manual, for more.
942 ** New procedure: `exact-integer-sqrt'
944 See "Integer Operations" in the manual, for more.
946 ** "Extended read syntax" for symbols parses better
948 In #{foo}# symbols, backslashes are now treated as escapes, as the
949 symbol-printing code intended. Additionally, "\x" within #{foo}# is now
950 interpreted as starting an R6RS hex escape. This is backward compatible
951 because the symbol printer would never produce a "\x" before. The
952 printer also works better too.
954 ** Added `--fresh-auto-compile' option
956 This allows a user to invalidate the auto-compilation cache. It's
957 usually not needed. See "Compilation" in the manual, for a discussion.
961 ** GOOPS documentation updates
965 Thanks to Mark Harig for improvements to guile.1.
967 ** SRFI-23 documented
969 The humble `error' SRFI now has an entry in the manual.
973 ** `(ice-9 binary-ports)': "R6RS I/O Ports", in the manual
974 ** `(ice-9 eval-string)': "Fly Evaluation", in the manual
975 ** `(ice-9 command-line)', not documented yet
979 ** Fixed `iconv_t' memory leak on close-port
980 ** Fixed some leaks with weak hash tables
981 ** Export `vhash-delq' and `vhash-delv' from `(ice-9 vlist)'
982 ** `after-gc-hook' works again
983 ** `define-record-type' now allowed in nested contexts
984 ** `exact-integer-sqrt' now handles large integers correctly
985 ** Fixed C extension examples in manual
986 ** `vhash-delete' honors HASH argument
987 ** Make `locale-digit-grouping' more robust
988 ** Default exception printer robustness fixes
989 ** Fix presence of non-I CPPFLAGS in `guile-2.0.pc'
990 ** `read' updates line/column numbers when reading SCSH block comments
991 ** Fix imports of multiple custom interfaces of same module
992 ** Fix encoding scanning for non-seekable ports
993 ** Fix `setter' when called with a non-setter generic
994 ** Fix f32 and f64 bytevectors to not accept rationals
995 ** Fix description of the R6RS `finite?' in manual
996 ** Quotient, remainder and modulo accept inexact integers again
997 ** Fix `continue' within `while' to take zero arguments
998 ** Fix alignment for structures in FFI
999 ** Fix port-filename of stdin, stdout, stderr to match the docs
1000 ** Fix weak hash table-related bug in `define-wrapped-pointer-type'
1001 ** Fix partial continuation application with pending procedure calls
1002 ** scm_{to,from}_locale_string use current locale, not current ports
1003 ** Fix thread cleanup, by using a pthread_key destructor
1004 ** Fix `quit' at the REPL
1005 ** Fix a failure to sync regs in vm bytevector ops
1006 ** Fix (texinfo reflection) to handle nested structures like syntax patterns
1007 ** Fix stexi->html double translation
1008 ** Fix tree-il->scheme fix for <prompt>
1009 ** Fix compilation of <prompt> in <fix> in single-value context
1010 ** Fix race condition in ensure-writable-dir
1011 ** Fix error message on ,disassemble "non-procedure"
1012 ** Fix prompt and abort with the boot evaluator
1013 ** Fix `procedure->pointer' for functions returning `void'
1014 ** Fix error reporting in dynamic-pointer
1015 ** Fix problems detecting coding: in block comments
1016 ** Fix duplicate load-path and load-compiled-path in compilation environment
1017 ** Add fallback read(2) suppport for .go files if mmap(2) unavailable
1018 ** Fix c32vector-set!, c64vector-set!
1019 ** Fix mistakenly deprecated read syntax for uniform complex vectors
1020 ** Fix parsing of exact numbers with negative exponents
1021 ** Ignore SIGPIPE in (system repl server)
1022 ** Fix optional second arg to R6RS log function
1023 ** Fix R6RS `assert' to return true value.
1024 ** Fix fencepost error when seeking in bytevector input ports
1025 ** Gracefully handle `setlocale' errors when starting the REPL
1026 ** Improve support of the `--disable-posix' configure option
1027 ** Make sure R6RS binary ports pass `binary-port?' regardless of the locale
1028 ** Gracefully handle unterminated UTF-8 sequences instead of hitting an `assert'
1032 Changes in 2.0.0 (changes since the 1.8.x series):
1034 * New modules (see the manual for details)
1036 ** `(srfi srfi-18)', more sophisticated multithreading support
1037 ** `(srfi srfi-27)', sources of random bits
1038 ** `(srfi srfi-38)', External Representation for Data With Shared Structure
1039 ** `(srfi srfi-42)', eager comprehensions
1040 ** `(srfi srfi-45)', primitives for expressing iterative lazy algorithms
1041 ** `(srfi srfi-67)', compare procedures
1042 ** `(ice-9 i18n)', internationalization support
1043 ** `(ice-9 futures)', fine-grain parallelism
1044 ** `(rnrs bytevectors)', the R6RS bytevector API
1045 ** `(rnrs io ports)', a subset of the R6RS I/O port API
1046 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
1047 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
1048 ** `(system foreign)', foreign function interface
1049 ** `(sxml match)', a pattern matcher for SXML
1050 ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library
1051 ** `(system vm coverage)', a line-by-line code coverage library
1052 ** `(web uri)', URI data type, parser, and unparser
1053 ** `(web http)', HTTP header parsers and unparsers
1054 ** `(web request)', HTTP request data type, reader, and writer
1055 ** `(web response)', HTTP response data type, reader, and writer
1056 ** `(web server)', Generic HTTP server
1057 ** `(ice-9 poll)', a poll wrapper
1058 ** `(web server http)', HTTP-over-TCP web server implementation
1060 ** Replaced `(ice-9 match)' with Alex Shinn's compatible, hygienic matcher.
1062 Guile's copy of Andrew K. Wright's `match' library has been replaced by
1063 a compatible hygienic implementation by Alex Shinn. It is now
1064 documented, see "Pattern Matching" in the manual.
1066 Compared to Andrew K. Wright's `match', the new `match' lacks
1067 `match-define', `match:error-control', `match:set-error-control',
1068 `match:error', `match:set-error', and all structure-related procedures.
1070 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
1072 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
1073 toolkit from Guile-Lib have been imported into Guile proper. See
1074 "Standard Library" in the manual for more details.
1076 ** Integration of lalr-scm, a parser generator
1078 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
1079 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
1082 * Changes to the stand-alone interpreter
1084 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
1086 Compiled code loads much faster than Scheme source code, and runs around
1087 3 or 4 times as fast, generating much less garbage in the process.
1089 ** Evaluating Scheme code does not use the C stack.
1091 Besides when compiling Guile itself, Guile no longer uses a recursive C
1092 function as an evaluator. This obviates the need to check the C stack
1093 pointer for overflow. Continuations still capture the C stack, however.
1095 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
1096 GUILE_SYSTEM_LOAD_COMPILED_PATH
1098 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
1099 for source files. It is a different path, however, because compiled
1100 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
1103 ** New read-eval-print loop (REPL) implementation
1105 Running Guile with no arguments drops the user into the new REPL. See
1106 "Using Guile Interactively" in the manual, for more information.
1108 ** Remove old Emacs interface
1110 Guile had an unused `--emacs' command line argument that was supposed to
1111 help when running Guile inside Emacs. This option has been removed, and
1112 the helper functions `named-module-use!' and `load-emacs-interface' have
1115 ** Add `(system repl server)' module and `--listen' command-line argument
1117 The `(system repl server)' module exposes procedures to listen on
1118 sockets for connections, and serve REPLs to those clients. The --listen
1119 command-line argument allows any Guile program to thus be remotely
1122 See "Invoking Guile" for more information on `--listen'.
1124 ** Command line additions
1126 The guile binary now supports a new switch "-x", which can be used to
1127 extend the list of filename extensions tried when loading files
1130 ** New reader options: `square-brackets', `r6rs-hex-escapes',
1131 `hungry-eol-escapes'
1133 The reader supports a new option (changeable via `read-options'),
1134 `square-brackets', which instructs it to interpret square brackets as
1135 parentheses. This option is on by default.
1137 When the new `r6rs-hex-escapes' reader option is enabled, the reader
1138 will recognize string escape sequences as defined in R6RS. R6RS string
1139 escape sequences are incompatible with Guile's existing escapes, though,
1140 so this option is off by default.
1142 Additionally, Guile follows the R6RS newline escaping rules when the
1143 `hungry-eol-escapes' option is enabled.
1145 See "String Syntax" in the manual, for more information.
1147 ** Function profiling and tracing at the REPL
1149 The `,profile FORM' REPL meta-command can now be used to statistically
1150 profile execution of a form, to see which functions are taking the most
1151 time. See `,help profile' for more information.
1153 Similarly, `,trace FORM' traces all function applications that occur
1154 during the execution of `FORM'. See `,help trace' for more information.
1156 ** Recursive debugging REPL on error
1158 When Guile sees an error at the REPL, instead of saving the stack, Guile
1159 will directly enter a recursive REPL in the dynamic context of the
1160 error. See "Error Handling" in the manual, for more information.
1162 A recursive REPL is the same as any other REPL, except that it
1163 has been augmented with debugging information, so that one can inspect
1164 the context of the error. The debugger has been integrated with the REPL
1165 via a set of debugging meta-commands.
1167 For example, one may access a backtrace with `,backtrace' (or
1168 `,bt'). See "Interactive Debugging" in the manual, for more
1171 ** New `guile-tools' commands: `compile', `disassemble'
1173 Pass the `--help' command-line option to these commands for more
1176 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
1178 Users may now install Guile to nonstandard prefixes and just run
1179 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
1180 include `/path/to/lib'.
1182 ** Guile's Emacs integration is now more keyboard-friendly
1184 Backtraces may now be disclosed with the keyboard in addition to the
1187 ** Load path change: search in version-specific paths before site paths
1189 When looking for a module, Guile now searches first in Guile's
1190 version-specific path (the library path), *then* in the site dir. This
1191 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
1192 installed. Also it should cut the number of `stat' system calls by half,
1195 ** Value history in the REPL on by default
1197 By default, the REPL will save computed values in variables like `$1',
1198 `$2', and the like. There are programmatic and interactive interfaces to
1199 control this. See "Value History" in the manual, for more information.
1201 ** Readline tab completion for arguments
1203 When readline is enabled, tab completion works for arguments too, not
1204 just for the operator position.
1206 ** Expression-oriented readline history
1208 Guile's readline history now tries to operate on expressions instead of
1209 input lines. Let us know what you think!
1211 ** Interactive Guile follows GNU conventions
1213 As recommended by the GPL, Guile now shows a brief copyright and
1214 warranty disclaimer on startup, along with pointers to more information.
1216 * Changes to Scheme functions and syntax
1218 ** Support for R6RS libraries
1220 The `library' and `import' forms from the latest Scheme report have been
1221 added to Guile, in such a way that R6RS libraries share a namespace with
1222 Guile modules. R6RS modules may import Guile modules, and are available
1223 for Guile modules to import via use-modules and all the rest. See "R6RS
1224 Libraries" in the manual for more information.
1226 ** Implementations of R6RS libraries
1228 Guile now has implementations for all of the libraries defined in the
1229 R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS
1230 Standard Libraries" in the manual for a full list of libraries.
1232 ** Partial R6RS compatibility
1234 Guile now has enough support for R6RS to run a reasonably large subset
1237 Guile is not fully R6RS compatible. Many incompatibilities are simply
1238 bugs, though some parts of Guile will remain R6RS-incompatible for the
1239 foreseeable future. See "R6RS Incompatibilities" in the manual, for more
1242 Please contact bug-guile@gnu.org if you have found an issue not
1243 mentioned in that compatibility list.
1245 ** New implementation of `primitive-eval'
1247 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
1248 still a C evaluator, used when building a fresh Guile to interpret the
1249 compiler, so we can compile eval.scm. Thereafter all calls to
1250 primitive-eval are implemented by VM-compiled code.
1252 This allows all of Guile's procedures, be they interpreted or compiled,
1253 to execute on the same stack, unifying multiple-value return semantics,
1254 providing for proper tail recursion between interpreted and compiled
1255 code, and simplifying debugging.
1257 As part of this change, the evaluator no longer mutates the internal
1258 representation of the code being evaluated in a thread-unsafe manner.
1260 There are two negative aspects of this change, however. First, Guile
1261 takes a lot longer to compile now. Also, there is less debugging
1262 information available for debugging interpreted code. We hope to improve
1263 both of these situations.
1265 There are many changes to the internal C evalator interface, but all
1266 public interfaces should be the same. See the ChangeLog for details. If
1267 we have inadvertantly changed an interface that you were using, please
1268 contact bug-guile@gnu.org.
1270 ** Procedure removed: `the-environment'
1272 This procedure was part of the interpreter's execution model, and does
1273 not apply to the compiler.
1275 ** No more `local-eval'
1277 `local-eval' used to exist so that one could evaluate code in the
1278 lexical context of a function. Since there is no way to get the lexical
1279 environment any more, as that concept has no meaning for the compiler,
1280 and a different meaning for the interpreter, we have removed the
1283 If you think you need `local-eval', you should probably implement your
1284 own metacircular evaluator. It will probably be as fast as Guile's
1287 ** Scheme source files will now be compiled automatically.
1289 If a compiled .go file corresponding to a .scm file is not found or is
1290 not fresh, the .scm file will be compiled on the fly, and the resulting
1291 .go file stored away. An advisory note will be printed on the console.
1293 Note that this mechanism depends on the timestamp of the .go file being
1294 newer than that of the .scm file; if the .scm or .go files are moved
1295 after installation, care should be taken to preserve their original
1298 Auto-compiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
1299 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
1300 will be created if needed.
1302 To inhibit automatic compilation, set the GUILE_AUTO_COMPILE environment
1303 variable to 0, or pass --no-auto-compile on the Guile command line.
1305 ** New POSIX procedures: `getrlimit' and `setrlimit'
1307 Note however that the interface of these functions is likely to change
1308 in the next prerelease.
1310 ** New POSIX procedure: `getsid'
1312 Scheme binding for the `getsid' C library call.
1314 ** New POSIX procedure: `getaddrinfo'
1316 Scheme binding for the `getaddrinfo' C library function.
1318 ** Multicast socket options
1320 Support was added for the IP_MULTICAST_TTL and IP_MULTICAST_IF socket
1321 options. See "Network Sockets and Communication" in the manual, for
1324 ** `recv!', `recvfrom!', `send', `sendto' now deal in bytevectors
1326 These socket procedures now take bytevectors as arguments, instead of
1327 strings. There is some deprecated string support, however.
1329 ** New GNU procedures: `setaffinity' and `getaffinity'.
1331 See "Processes" in the manual, for more information.
1333 ** New procedures: `compose', `negate', and `const'
1335 See "Higher-Order Functions" in the manual, for more information.
1337 ** New procedure in `(oops goops)': `method-formals'
1339 ** New procedures in (ice-9 session): `add-value-help-handler!',
1340 `remove-value-help-handler!', `add-name-help-handler!'
1341 `remove-name-help-handler!', `procedure-arguments'
1343 The value and name help handlers provide some minimal extensibility to
1344 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
1345 example, to make stexinfo help documentation available. See those
1346 procedures' docstrings for more information.
1348 `procedure-arguments' describes the arguments that a procedure can take,
1349 combining arity and formals. For example:
1351 (procedure-arguments resolve-interface)
1352 => ((required . (name)) (rest . args))
1354 Additionally, `module-commentary' is now publically exported from
1357 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
1359 These procedures created primitive fexprs for the old evaluator, and are
1360 no longer supported. If you feel that you need these functions, you
1361 probably need to write your own metacircular evaluator (which will
1362 probably be as fast as Guile's, anyway).
1364 ** New language: ECMAScript
1366 Guile now ships with one other high-level language supported,
1367 ECMAScript. The goal is to support all of version 3.1 of the standard,
1368 but not all of the libraries are there yet. This support is not yet
1369 documented; ask on the mailing list if you are interested.
1371 ** New language: Brainfuck
1373 Brainfuck is a toy language that closely models Turing machines. Guile's
1374 brainfuck compiler is meant to be an example of implementing other
1375 languages. See the manual for details, or
1376 http://en.wikipedia.org/wiki/Brainfuck for more information about the
1377 Brainfuck language itself.
1379 ** New language: Elisp
1381 Guile now has an experimental Emacs Lisp compiler and runtime. You can
1382 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
1383 Kraft and Brian Templeton, and all bugs to bug-guile@gnu.org.
1385 ** Better documentation infrastructure for macros
1387 It is now possible to introspect on the type of a macro, e.g.
1388 syntax-rules, identifier-syntax, etc, and extract information about that
1389 macro, such as the syntax-rules patterns or the defmacro arguments.
1390 `(texinfo reflection)' takes advantage of this to give better macro
1393 ** Support for arbitrary procedure metadata
1395 Building on its support for docstrings, Guile now supports multiple
1396 docstrings, adding them to the tail of a compiled procedure's
1397 properties. For example:
1403 (procedure-properties foo)
1404 => ((name . foo) (documentation . "one") (documentation . "two"))
1406 Also, vectors of pairs are now treated as additional metadata entries:
1409 #((quz . #f) (docstring . "xyzzy"))
1411 (procedure-properties bar)
1412 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
1414 This allows arbitrary literals to be embedded as metadata in a compiled
1417 ** The psyntax expander now knows how to interpret the @ and @@ special
1420 ** The psyntax expander is now hygienic with respect to modules.
1422 Free variables in a macro are scoped in the module that the macro was
1423 defined in, not in the module the macro is used in. For example, code
1424 like this works now:
1426 (define-module (foo) #:export (bar))
1427 (define (helper x) ...)
1429 (syntax-rules () ((_ x) (helper x))))
1431 (define-module (baz) #:use-module (foo))
1434 It used to be you had to export `helper' from `(foo)' as well.
1435 Thankfully, this has been fixed.
1437 ** Support for version information in Guile's `module' form
1439 Guile modules now have a `#:version' field. See "R6RS Version
1440 References", "General Information about Modules", "Using Guile Modules",
1441 and "Creating Guile Modules" in the manual for more information.
1443 ** Support for renaming bindings on module export
1445 Wherever Guile accepts a symbol as an argument to specify a binding to
1446 export, it now also accepts a pair of symbols, indicating that a binding
1447 should be renamed on export. See "Creating Guile Modules" in the manual
1448 for more information.
1450 ** New procedure: `module-export-all!'
1452 This procedure exports all current and future bindings from a module.
1453 Use as `(module-export-all! (current-module))'.
1455 ** New procedure `reload-module', and `,reload' REPL command
1457 See "Module System Reflection" and "Module Commands" in the manual, for
1460 ** `eval-case' has been deprecated, and replaced by `eval-when'.
1462 The semantics of `eval-when' are easier to understand. See "Eval When"
1463 in the manual, for more information.
1465 ** Guile is now more strict about prohibiting definitions in expression
1468 Although previous versions of Guile accepted it, the following
1469 expression is not valid, in R5RS or R6RS:
1471 (if test (define foo 'bar) (define foo 'baz))
1473 In this specific case, it would be better to do:
1475 (define foo (if test 'bar 'baz))
1477 It is possible to circumvent this restriction with e.g.
1478 `(module-define! (current-module) 'foo 'baz)'. Contact the list if you
1481 ** Support for `letrec*'
1483 Guile now supports `letrec*', a recursive lexical binding operator in
1484 which the identifiers are bound in order. See "Local Bindings" in the
1485 manual, for more details.
1487 ** Internal definitions now expand to `letrec*'
1489 Following the R6RS, internal definitions now expand to letrec* instead
1490 of letrec. The following program is invalid for R5RS, but valid for
1495 (define baz (+ bar 20))
1498 ;; R5RS and Guile <= 1.8:
1499 (foo) => Unbound variable: bar
1500 ;; R6RS and Guile >= 2.0:
1503 This change should not affect correct R5RS programs, or programs written
1504 in earlier Guile dialects.
1506 ** Macro expansion produces structures instead of s-expressions
1508 In the olden days, macroexpanding an s-expression would yield another
1509 s-expression. Though the lexical variables were renamed, expansions of
1510 core forms like `if' and `begin' were still non-hygienic, as they relied
1511 on the toplevel definitions of `if' et al being the conventional ones.
1513 The solution is to expand to structures instead of s-expressions. There
1514 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
1515 etc. The expander already did this for compilation, producing Tree-IL
1516 directly; it has been changed now to do so when expanding for the
1519 ** Defmacros must now produce valid Scheme expressions.
1521 It used to be that defmacros could unquote in Scheme values, as a way of
1522 supporting partial evaluation, and avoiding some hygiene issues. For
1525 (define (helper x) ...)
1526 (define-macro (foo bar)
1529 Assuming this macro is in the `(baz)' module, the direct translation of
1532 (define (helper x) ...)
1533 (define-macro (foo bar)
1534 `((@@ (baz) helper) ,bar))
1536 Of course, one could just use a hygienic macro instead:
1540 ((_ bar) (helper bar))))
1542 ** Guile's psyntax now supports docstrings and internal definitions.
1544 The following Scheme is not strictly legal:
1551 However its intent is fairly clear. Guile interprets "bar" to be the
1552 docstring of `foo', and the definition of `baz' is still in definition
1555 ** Support for settable identifier syntax
1557 Following the R6RS, "variable transformers" are settable
1558 identifier-syntax. See "Identifier macros" in the manual, for more
1561 ** syntax-case treats `_' as a placeholder
1563 Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches
1564 anything, and binds no pattern variables. Unlike the R6RS, Guile also
1565 permits `_' to be in the literals list for a pattern.
1567 ** Macros need to be defined before their first use.
1569 It used to be that with lazy memoization, this might work:
1573 (define-macro (ref x) x)
1576 But now, the body of `foo' is interpreted to mean a call to the toplevel
1577 `ref' function, instead of a macro expansion. The solution is to define
1578 macros before code that uses them.
1580 ** Functions needed by macros at expand-time need to be present at
1583 For example, this code will work at the REPL:
1585 (define (double-helper x) (* x x))
1586 (define-macro (double-literal x) (double-helper x))
1587 (double-literal 2) => 4
1589 But it will not work when a file is compiled, because the definition of
1590 `double-helper' is not present at expand-time. The solution is to wrap
1591 the definition of `double-helper' in `eval-when':
1593 (eval-when (load compile eval)
1594 (define (double-helper x) (* x x)))
1595 (define-macro (double-literal x) (double-helper x))
1596 (double-literal 2) => 4
1598 See the documentation for eval-when for more information.
1600 ** `macroexpand' produces structures, not S-expressions.
1602 Given the need to maintain referential transparency, both lexically and
1603 modular, the result of expanding Scheme expressions is no longer itself
1604 an s-expression. If you want a human-readable approximation of the
1605 result of `macroexpand', call `tree-il->scheme' from `(language
1608 ** Removed function: `macroexpand-1'
1610 It is unclear how to implement `macroexpand-1' with syntax-case, though
1611 PLT Scheme does prove that it is possible.
1613 ** New reader macros: #' #` #, #,@
1615 These macros translate, respectively, to `syntax', `quasisyntax',
1616 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
1617 These reader macros may be overridden by `read-hash-extend'.
1619 ** Incompatible change to #'
1621 Guile did have a #' hash-extension, by default, which just returned the
1622 subsequent datum: #'foo => foo. In the unlikely event that anyone
1623 actually used this, this behavior may be reinstated via the
1624 `read-hash-extend' mechanism.
1626 ** `unquote' and `unquote-splicing' accept multiple expressions
1628 As per the R6RS, these syntax operators can now accept any number of
1629 expressions to unquote.
1631 ** Scheme expresssions may be commented out with #;
1633 #; comments out an entire expression. See SRFI-62 or the R6RS for more
1636 ** Prompts: Delimited, composable continuations
1638 Guile now has prompts as part of its primitive language. See "Prompts"
1639 in the manual, for more information.
1641 Expressions entered in at the REPL, or from the command line, are
1642 surrounded by a prompt with the default prompt tag.
1644 ** `make-stack' with a tail-called procedural narrowing argument no longer
1645 works (with compiled procedures)
1647 It used to be the case that a captured stack could be narrowed to select
1648 calls only up to or from a certain procedure, even if that procedure
1649 already tail-called another procedure. This was because the debug
1650 information from the original procedure was kept on the stack.
1652 Now with the new compiler, the stack only contains active frames from
1653 the current continuation. A narrow to a procedure that is not in the
1654 stack will result in an empty stack. To fix this, narrow to a procedure
1655 that is active in the current continuation, or narrow to a specific
1656 number of stack frames.
1658 ** Backtraces through compiled procedures only show procedures that are
1659 active in the current continuation
1661 Similarly to the previous issue, backtraces in compiled code may be
1662 different from backtraces in interpreted code. There are no semantic
1663 differences, however. Please mail bug-guile@gnu.org if you see any
1664 deficiencies with Guile's backtraces.
1666 ** `positions' reader option enabled by default
1668 This change allows primitive-load without --auto-compile to also
1669 propagate source information through the expander, for better errors and
1670 to let macros know their source locations. The compiler was already
1671 turning it on anyway.
1673 ** New macro: `current-source-location'
1675 The macro returns the current source location (to be documented).
1677 ** syntax-rules and syntax-case macros now propagate source information
1678 through to the expanded code
1680 This should result in better backtraces.
1682 ** The currying behavior of `define' has been removed.
1684 Before, `(define ((f a) b) (* a b))' would translate to
1686 (define f (lambda (a) (lambda (b) (* a b))))
1688 Now a syntax error is signaled, as this syntax is not supported by
1689 default. Use the `(ice-9 curried-definitions)' module to get back the
1692 ** New procedure, `define!'
1694 `define!' is a procedure that takes two arguments, a symbol and a value,
1695 and binds the value to the symbol in the current module. It's useful to
1696 programmatically make definitions in the current module, and is slightly
1697 less verbose than `module-define!'.
1699 ** All modules have names now
1701 Before, you could have anonymous modules: modules without names. Now,
1702 because of hygiene and macros, all modules have names. If a module was
1703 created without a name, the first time `module-name' is called on it, a
1704 fresh name will be lazily generated for it.
1706 ** The module namespace is now separate from the value namespace
1708 It was a little-known implementation detail of Guile's module system
1709 that it was built on a single hierarchical namespace of values -- that
1710 if there was a module named `(foo bar)', then in the module named
1711 `(foo)' there was a binding from `bar' to the `(foo bar)' module.
1713 This was a neat trick, but presented a number of problems. One problem
1714 was that the bindings in a module were not apparent from the module
1715 itself; perhaps the `(foo)' module had a private binding for `bar', and
1716 then an external contributor defined `(foo bar)'. In the end there can
1717 be only one binding, so one of the two will see the wrong thing, and
1718 produce an obtuse error of unclear provenance.
1720 Also, the public interface of a module was also bound in the value
1721 namespace, as `%module-public-interface'. This was a hack from the early
1722 days of Guile's modules.
1724 Both of these warts have been fixed by the addition of fields in the
1725 `module' data type. Access to modules and their interfaces from the
1726 value namespace has been deprecated, and all accessors use the new
1727 record accessors appropriately.
1729 When Guile is built with support for deprecated code, as is the default,
1730 the value namespace is still searched for modules and public interfaces,
1731 and a deprecation warning is raised as appropriate.
1733 Finally, to support lazy loading of modules as one used to be able to do
1734 with module binder procedures, Guile now has submodule binders, called
1735 if a given submodule is not found. See boot-9.scm for more information.
1737 ** New procedures: module-ref-submodule, module-define-submodule,
1738 nested-ref-module, nested-define-module!, local-ref-module,
1741 These new accessors are like their bare variants, but operate on
1742 namespaces instead of values.
1744 ** The (app modules) module tree is officially deprecated
1746 It used to be that one could access a module named `(foo bar)' via
1747 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
1748 modules)' bit was a never-used and never-documented abstraction, and has
1749 been deprecated. See the following mail for a full discussion:
1751 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
1753 The `%app' binding is also deprecated.
1755 ** `module-filename' field and accessor
1757 Modules now record the file in which they are defined. This field may be
1758 accessed with the new `module-filename' procedure.
1760 ** Modules load within a known environment
1762 It takes a few procedure calls to define a module, and those procedure
1763 calls need to be in scope. Now we ensure that the current module when
1764 loading a module is one that has the needed bindings, instead of relying
1767 ** `load' is a macro (!) that resolves paths relative to source file dir
1769 The familiar Schem `load' procedure is now a macro that captures the
1770 name of the source file being expanded, and dispatches to the new
1771 `load-in-vicinity'. Referencing `load' by bare name returns a closure
1772 that embeds the current source file name.
1774 This fix allows `load' of relative paths to be resolved with respect to
1775 the location of the file that calls `load'.
1777 ** Many syntax errors have different texts now
1779 Syntax errors still throw to the `syntax-error' key, but the arguments
1780 are often different now. Perhaps in the future, Guile will switch to
1781 using standard SRFI-35 conditions.
1783 ** Returning multiple values to compiled code will silently truncate the
1784 values to the expected number
1786 For example, the interpreter would raise an error evaluating the form,
1787 `(+ (values 1 2) (values 3 4))', because it would see the operands as
1788 being two compound "values" objects, to which `+' does not apply.
1790 The compiler, on the other hand, receives multiple values on the stack,
1791 not as a compound object. Given that it must check the number of values
1792 anyway, if too many values are provided for a continuation, it chooses
1793 to truncate those values, effectively evaluating `(+ 1 3)' instead.
1795 The idea is that the semantics that the compiler implements is more
1796 intuitive, and the use of the interpreter will fade out with time.
1797 This behavior is allowed both by the R5RS and the R6RS.
1799 ** Multiple values in compiled code are not represented by compound
1802 This change may manifest itself in the following situation:
1804 (let ((val (foo))) (do-something) val)
1806 In the interpreter, if `foo' returns multiple values, multiple values
1807 are produced from the `let' expression. In the compiler, those values
1808 are truncated to the first value, and that first value is returned. In
1809 the compiler, if `foo' returns no values, an error will be raised, while
1810 the interpreter would proceed.
1812 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
1813 behavior is more correct, however. If you wish to preserve a potentially
1814 multiply-valued return, you will need to set up a multiple-value
1815 continuation, using `call-with-values'.
1817 ** Defmacros are now implemented in terms of syntax-case.
1819 The practical ramification of this is that the `defmacro?' predicate has
1820 been removed, along with `defmacro-transformer', `macro-table',
1821 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
1822 `defmacro:transformer'. This is because defmacros are simply macros. If
1823 any of these procedures provided useful facilities to you, we encourage
1824 you to contact the Guile developers.
1826 ** Hygienic macros documented as the primary syntactic extension mechanism.
1828 The macro documentation was finally fleshed out with some documentation
1829 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
1830 expansion process. See "Macros" in the manual, for details.
1832 ** psyntax is now the default expander
1834 Scheme code is now expanded by default by the psyntax hygienic macro
1835 expander. Expansion is performed completely before compilation or
1838 Notably, syntax errors will be signalled before interpretation begins.
1839 In the past, many syntax errors were only detected at runtime if the
1840 code in question was memoized.
1842 As part of its expansion, psyntax renames all lexically-bound
1843 identifiers. Original identifier names are preserved and given to the
1844 compiler, but the interpreter will see the renamed variables, e.g.,
1845 `x432' instead of `x'.
1847 Note that the psyntax that Guile uses is a fork, as Guile already had
1848 modules before incompatible modules were added to psyntax -- about 10
1849 years ago! Thus there are surely a number of bugs that have been fixed
1850 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
1852 ** syntax-rules and syntax-case are available by default.
1854 There is no longer any need to import the `(ice-9 syncase)' module
1855 (which is now deprecated). The expander may be invoked directly via
1856 `macroexpand', though it is normally searched for via the current module
1859 Also, the helper routines for syntax-case are available in the default
1860 environment as well: `syntax->datum', `datum->syntax',
1861 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
1862 `identifier?', and `syntax-violation'. See the R6RS for documentation.
1864 ** Tail patterns in syntax-case
1866 Guile has pulled in some more recent changes from the psyntax portable
1867 syntax expander, to implement support for "tail patterns". Such patterns
1868 are supported by syntax-rules and syntax-case. This allows a syntax-case
1869 match clause to have ellipses, then a pattern at the end. For example:
1872 (syntax-rules (else)
1873 ((_ val match-clause ... (else e e* ...))
1876 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
1877 tail pattern for the else clause. Thanks to Andreas Rottmann for the
1878 patch, and Kent Dybvig for the code.
1880 ** Lexical bindings introduced by hygienic macros may not be referenced
1881 by nonhygienic macros.
1883 If a lexical binding is introduced by a hygienic macro, it may not be
1884 referenced by a nonhygienic macro. For example, this works:
1887 (define-macro (bind-x val body)
1888 `(let ((x ,val)) ,body))
1889 (define-macro (ref x)
1891 (bind-x 10 (ref x)))
1896 (define-syntax bind-x
1898 ((_ val body) (let ((x val)) body))))
1899 (define-macro (ref x)
1901 (bind-x 10 (ref x)))
1903 It is not normal to run into this situation with existing code. However,
1904 if you have defmacros that expand to hygienic macros, it is possible to
1905 run into situations like this. For example, if you have a defmacro that
1906 generates a `while' expression, the `break' bound by the `while' may not
1907 be visible within other parts of your defmacro. The solution is to port
1908 from defmacros to syntax-rules or syntax-case.
1910 ** Macros may no longer be referenced as first-class values.
1912 In the past, you could evaluate e.g. `if', and get its macro value. Now,
1913 expanding this form raises a syntax error.
1915 Macros still /exist/ as first-class values, but they must be
1916 /referenced/ via the module system, e.g. `(module-ref (current-module)
1919 ** Macros may now have docstrings.
1921 `object-documentation' from `(ice-9 documentation)' may be used to
1922 retrieve the docstring, once you have a macro value -- but see the above
1923 note about first-class macros. Docstrings are associated with the syntax
1924 transformer procedures.
1926 ** `case-lambda' is now available in the default environment.
1928 The binding in the default environment is equivalent to the one from the
1929 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
1930 to maintain compatibility with Guile 1.8 and earlier.
1932 ** Procedures may now have more than one arity.
1934 This can be the case, for example, in case-lambda procedures. The
1935 arities of compiled procedures may be accessed via procedures from the
1936 `(system vm program)' module; see "Compiled Procedures", "Optional
1937 Arguments", and "Case-lambda" in the manual.
1939 ** Deprecate arity access via (procedure-properties proc 'arity)
1941 Instead of accessing a procedure's arity as a property, use the new
1942 `procedure-minimum-arity' function, which gives the most permissive
1943 arity that the function has, in the same format as the old arity
1946 ** `lambda*' and `define*' are now available in the default environment
1948 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
1949 compatibility purposes. No semantic change has been made (we hope).
1950 Optional and keyword arguments now dispatch via special VM operations,
1951 without the need to cons rest arguments, making them very fast.
1953 ** New syntax: define-once
1955 `define-once' is like Lisp's `defvar': it creates a toplevel binding,
1956 but only if one does not exist already.
1958 ** New function, `truncated-print', with `format' support
1960 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
1961 will ensure that the output stays within a certain width, truncating the
1962 output in what is hopefully an intelligent manner. See the manual for
1965 There is a new `format' specifier, `~@y', for doing a truncated
1966 print (as opposed to `~y', which does a pretty-print). See the `format'
1967 documentation for more details.
1969 ** Better pretty-printing
1971 Indentation recognizes more special forms, like `syntax-case', and read
1972 macros like `quote' are printed better.
1974 ** Passing a number as the destination of `format' is deprecated
1976 The `format' procedure in `(ice-9 format)' now emits a deprecation
1977 warning if a number is passed as its first argument.
1979 Also, it used to be that you could omit passing a port to `format', in
1980 some cases. This still works, but has been formally deprecated.
1982 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
1984 Guile now implements SRFI-4 vectors using bytevectors. Often when you
1985 have a numeric vector, you end up wanting to write its bytes somewhere,
1986 or have access to the underlying bytes, or read in bytes from somewhere
1987 else. Bytevectors are very good at this sort of thing. But the SRFI-4
1988 APIs are nicer to use when doing number-crunching, because they are
1989 addressed by element and not by byte.
1991 So as a compromise, Guile allows all bytevector functions to operate on
1992 numeric vectors. They address the underlying bytes in the native
1993 endianness, as one would expect.
1995 Following the same reasoning, that it's just bytes underneath, Guile
1996 also allows uniform vectors of a given type to be accessed as if they
1997 were of any type. One can fill a u32vector, and access its elements with
1998 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
2001 In this way, uniform numeric vectors may be written to and read from
2002 input/output ports using the procedures that operate on bytevectors.
2004 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
2005 inlined to the VM instructions for bytevector access.
2007 See "SRFI-4" in the manual, for more information.
2009 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
2011 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
2012 are part of the standard. Complex uniform vectors and the
2013 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
2015 Guile's default environment imports `(srfi srfi-4)', and probably should
2016 import `(srfi srfi-4 gnu)' as well.
2018 See "SRFI-4 Extensions" in the manual, for more information.
2020 ** New syntax: include-from-path.
2022 `include-from-path' is like `include', except it looks for its file in
2023 the load path. It can be used to compile other files into a file.
2025 ** New syntax: quasisyntax.
2027 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
2028 documentation for more information. Thanks to Andre van Tonder for the
2031 ** `*unspecified*' is identifier syntax
2033 `*unspecified*' is no longer a variable, so it is optimized properly by
2034 the compiler, and is not `set!'-able.
2036 ** Changes and bugfixes in numerics code
2038 *** Added six new sets of fast quotient and remainder operators
2040 Added six new sets of fast quotient and remainder operator pairs with
2041 different semantics than the R5RS operators. They support not only
2042 integers, but all reals, including exact rationals and inexact
2043 floating point numbers.
2045 These procedures accept two real numbers N and D, where the divisor D
2046 must be non-zero. Each set of operators computes an integer quotient
2047 Q and a real remainder R such that N = Q*D + R and |R| < |D|. They
2048 differ only in how N/D is rounded to produce Q.
2050 `euclidean-quotient' returns the integer Q and `euclidean-remainder'
2051 returns the real R such that N = Q*D + R and 0 <= R < |D|. `euclidean/'
2052 returns both Q and R, and is more efficient than computing each
2053 separately. Note that when D > 0, `euclidean-quotient' returns
2054 floor(N/D), and when D < 0 it returns ceiling(N/D).
2056 `centered-quotient', `centered-remainder', and `centered/' are similar
2057 except that the range of remainders is -abs(D/2) <= R < abs(D/2), and
2058 `centered-quotient' rounds N/D to the nearest integer. Note that these
2059 operators are equivalent to the R6RS integer division operators `div',
2060 `mod', `div-and-mod', `div0', `mod0', and `div0-and-mod0'.
2062 `floor-quotient' and `floor-remainder' compute Q and R, respectively,
2063 where Q has been rounded toward negative infinity. `floor/' returns
2064 both Q and R, and is more efficient than computing each separately.
2065 Note that when applied to integers, `floor-remainder' is equivalent to
2066 the R5RS integer-only `modulo' operator. `ceiling-quotient',
2067 `ceiling-remainder', and `ceiling/' are similar except that Q is
2068 rounded toward positive infinity.
2070 For `truncate-quotient', `truncate-remainder', and `truncate/', Q is
2071 rounded toward zero. Note that when applied to integers,
2072 `truncate-quotient' and `truncate-remainder' are equivalent to the
2073 R5RS integer-only operators `quotient' and `remainder'.
2075 For `round-quotient', `round-remainder', and `round/', Q is rounded to
2076 the nearest integer, with ties going to the nearest even integer.
2078 *** Complex number changes
2080 Guile is now able to represent non-real complex numbers whose
2081 imaginary part is an _inexact_ zero (0.0 or -0.0), per R6RS.
2082 Previously, such numbers were immediately changed into inexact reals.
2084 (real? 0.0+0.0i) now returns #f, per R6RS, although (zero? 0.0+0.0i)
2085 still returns #t, per R6RS. (= 0 0.0+0.0i) and (= 0.0 0.0+0.0i) are
2086 #t, but the same comparisons using `eqv?' or `equal?' are #f.
2088 Like other non-real numbers, these complex numbers with inexact zero
2089 imaginary part will raise exceptions is passed to procedures requiring
2090 reals, such as `<', `>', `<=', `>=', `min', `max', `positive?',
2091 `negative?', `inf?', `nan?', `finite?', etc.
2093 **** `make-rectangular' changes
2095 scm_make_rectangular `make-rectangular' now returns a real number only
2096 if the imaginary part is an _exact_ 0. Previously, it would return a
2097 real number if the imaginary part was an inexact zero.
2099 scm_c_make_rectangular now always returns a non-real complex number,
2100 even if the imaginary part is zero. Previously, it would return a
2101 real number if the imaginary part was zero.
2103 **** `make-polar' changes
2105 scm_make_polar `make-polar' now returns a real number only if the
2106 angle or magnitude is an _exact_ 0. If the magnitude is an exact 0,
2107 it now returns an exact 0. Previously, it would return a real
2108 number if the imaginary part was an inexact zero.
2110 scm_c_make_polar now always returns a non-real complex number, even if
2111 the imaginary part is 0.0. Previously, it would return a real number
2112 if the imaginary part was 0.0.
2114 **** `imag-part' changes
2116 scm_imag_part `imag-part' now returns an exact 0 if applied to an
2117 inexact real number. Previously it returned an inexact zero in this
2120 *** `eqv?' and `equal?' now compare numbers equivalently
2122 scm_equal_p `equal?' now behaves equivalently to scm_eqv_p `eqv?' for
2123 numeric values, per R5RS. Previously, equal? worked differently,
2124 e.g. `(equal? 0.0 -0.0)' returned #t but `(eqv? 0.0 -0.0)' returned #f,
2125 and `(equal? +nan.0 +nan.0)' returned #f but `(eqv? +nan.0 +nan.0)'
2128 *** `(equal? +nan.0 +nan.0)' now returns #t
2130 Previously, `(equal? +nan.0 +nan.0)' returned #f, although
2131 `(let ((x +nan.0)) (equal? x x))' and `(eqv? +nan.0 +nan.0)'
2132 both returned #t. R5RS requires that `equal?' behave like
2133 `eqv?' when comparing numbers.
2135 *** Change in handling products `*' involving exact 0
2137 scm_product `*' now handles exact 0 differently. A product containing
2138 an exact 0 now returns an exact 0 if and only if the other arguments
2139 are all exact. An inexact zero is returned if and only if the other
2140 arguments are all finite but not all exact. If an infinite or NaN
2141 value is present, a NaN value is returned. Previously, any product
2142 containing an exact 0 yielded an exact 0, regardless of the other
2145 *** `expt' and `integer-expt' changes when the base is 0
2147 While `(expt 0 0)' is still 1, and `(expt 0 N)' for N > 0 is still
2148 zero, `(expt 0 N)' for N < 0 is now a NaN value, and likewise for
2149 integer-expt. This is more correct, and conforming to R6RS, but seems
2150 to be incompatible with R5RS, which would return 0 for all non-zero
2153 *** `expt' and `integer-expt' are more generic, less strict
2155 When raising to an exact non-negative integer exponent, `expt' and
2156 `integer-expt' are now able to exponentiate any object that can be
2157 multiplied using `*'. They can also raise an object to an exact
2158 negative integer power if its reciprocal can be taken using `/'.
2159 In order to allow this, the type of the first argument is no longer
2160 checked when raising to an exact integer power. If the exponent is 0
2161 or 1, the first parameter is not manipulated at all, and need not
2162 even support multiplication.
2164 *** Infinities are no longer integers, nor rationals
2166 scm_integer_p `integer?' and scm_rational_p `rational?' now return #f
2167 for infinities, per R6RS. Previously they returned #t for real
2168 infinities. The real infinities and NaNs are still considered real by
2169 scm_real `real?' however, per R6RS.
2171 *** NaNs are no longer rationals
2173 scm_rational_p `rational?' now returns #f for NaN values, per R6RS.
2174 Previously it returned #t for real NaN values. They are still
2175 considered real by scm_real `real?' however, per R6RS.
2177 *** `inf?' and `nan?' now throw exceptions for non-reals
2179 The domain of `inf?' and `nan?' is the real numbers. Guile now signals
2180 an error when a non-real number or non-number is passed to these
2181 procedures. (Note that NaNs _are_ considered numbers by scheme, despite
2184 *** `rationalize' bugfixes and changes
2186 Fixed bugs in scm_rationalize `rationalize'. Previously, it returned
2187 exact integers unmodified, although that was incorrect if the epsilon
2188 was at least 1 or inexact, e.g. (rationalize 4 1) should return 3 per
2189 R5RS and R6RS, but previously it returned 4. It also now handles
2190 cases involving infinities and NaNs properly, per R6RS.
2192 *** Trigonometric functions now return exact numbers in some cases
2194 scm_sin `sin', scm_cos `cos', scm_tan `tan', scm_asin `asin', scm_acos
2195 `acos', scm_atan `atan', scm_sinh `sinh', scm_cosh `cosh', scm_tanh
2196 `tanh', scm_sys_asinh `asinh', scm_sys_acosh `acosh', and
2197 scm_sys_atanh `atanh' now return exact results in some cases.
2199 *** New procedure: `finite?'
2201 Add scm_finite_p `finite?' from R6RS to guile core, which returns #t
2202 if and only if its argument is neither infinite nor a NaN. Note that
2203 this is not the same as (not (inf? x)) or (not (infinite? x)), since
2204 NaNs are neither finite nor infinite.
2206 *** Improved exactness handling for complex number parsing
2208 When parsing non-real complex numbers, exactness specifiers are now
2209 applied to each component, as is done in PLT Scheme. For complex
2210 numbers written in rectangular form, exactness specifiers are applied
2211 to the real and imaginary parts before calling scm_make_rectangular.
2212 For complex numbers written in polar form, exactness specifiers are
2213 applied to the magnitude and angle before calling scm_make_polar.
2215 Previously, exactness specifiers were applied to the number as a whole
2216 _after_ calling scm_make_rectangular or scm_make_polar.
2218 For example, (string->number "#i5.0+0i") now does the equivalent of:
2220 (make-rectangular (exact->inexact 5.0) (exact->inexact 0))
2222 which yields 5.0+0.0i. Previously it did the equivalent of:
2224 (exact->inexact (make-rectangular 5.0 0))
2228 ** Unicode characters
2230 Unicode characters may be entered in octal format via e.g. `#\454', or
2231 created via (integer->char 300). A hex external representation will
2232 probably be introduced at some point.
2236 Internally, strings are now represented either in the `latin-1'
2237 encoding, one byte per character, or in UTF-32, with four bytes per
2238 character. Strings manage their own allocation, switching if needed.
2240 Extended characters may be written in a literal string using the
2241 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
2242 or 24-bit codepoints, respectively, or entered directly in the native
2243 encoding of the port on which the string is read.
2247 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
2249 ** Support for non-ASCII source code files
2251 The default reader now handles source code files for some of the
2252 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
2253 should have an encoding declaration near the top of the file. Also,
2254 there is a new function, `file-encoding', that scans a port for a coding
2255 declaration. See the section of the manual entitled, "Character Encoding
2258 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
2259 code. This use is now discouraged. Binary input and output is
2260 currently supported by opening ports in the ISO-8859-1 locale.
2262 ** Source files default to UTF-8.
2264 If source files do not specify their encoding via a `coding:' block,
2265 the default encoding is UTF-8, instead of being taken from the current
2268 ** Interactive Guile installs the current locale.
2270 Instead of leaving the user in the "C" locale, running the Guile REPL
2271 installs the current locale. [FIXME xref?]
2273 ** Support for locale transcoding when reading from and writing to ports
2275 Ports now have an associated character encoding, and port read and write
2276 operations do conversion to and from locales automatically. Ports also
2277 have an associated strategy for how to deal with locale conversion
2280 See the documentation in the manual for the four new support functions,
2281 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
2282 and `port-conversion-strategy'.
2284 ** String and SRFI-13 functions can operate on Unicode strings
2286 ** Unicode support for SRFI-14 character sets
2288 The default character sets are no longer locale dependent and contain
2289 characters from the whole Unicode range. There is a new predefined
2290 character set, `char-set:designated', which contains all assigned
2291 Unicode characters. There is a new debugging function, `%char-set-dump'.
2293 ** Character functions operate on Unicode characters
2295 `char-upcase' and `char-downcase' use default Unicode casing rules.
2296 Character comparisons such as `char<?' and `char-ci<?' now sort based on
2297 Unicode code points.
2299 ** Global variables `scm_charnames' and `scm_charnums' are removed
2301 These variables contained the names of control characters and were
2302 used when writing characters. While these were global, they were
2303 never intended to be public API. They have been replaced with private
2306 ** EBCDIC support is removed
2308 There was an EBCDIC compile flag that altered some of the character
2309 processing. It appeared that full EBCDIC support was never completed
2310 and was unmaintained.
2312 ** Compile-time warnings
2314 Guile can warn about potentially unbound free variables. Pass the
2315 -Wunbound-variable on the `guile-tools compile' command line, or add
2316 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
2317 invocation. Warnings are also enabled by default for expressions entered
2320 Guile can also warn when you pass the wrong number of arguments to a
2321 procedure, with -Warity-mismatch, or `arity-mismatch' in the
2322 `#:warnings' as above.
2324 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
2325 warn about unused local or global (top-level) variables, and `-Wformat',
2326 to check for various errors related to the `format' procedure.
2328 ** A new `memoize-symbol' evaluator trap has been added.
2330 This trap can be used for efficiently implementing a Scheme code
2333 ** Duplicate bindings among used modules are resolved lazily.
2335 This slightly improves program startup times.
2337 ** New thread cancellation and thread cleanup API
2339 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
2341 ** New threads are in `(guile-user)' by default, not `(guile)'
2343 It used to be that a new thread entering Guile would do so in the
2344 `(guile)' module, unless this was the first time Guile was initialized,
2345 in which case it was `(guile-user)'. This has been fixed to have all
2346 new threads unknown to Guile default to `(guile-user)'.
2348 ** New helpers: `print-exception', `set-exception-printer!'
2350 These functions implement an extensible exception printer. Guile
2351 registers printers for all of the exceptions it throws. Users may add
2352 their own printers. There is also `scm_print_exception', for use by C
2353 programs. Pleasantly, this allows SRFI-35 and R6RS exceptions to be
2354 printed appropriately.
2356 ** GOOPS dispatch in scheme
2358 As an implementation detail, GOOPS dispatch is no longer implemented by
2359 special evaluator bytecodes, but rather directly via a Scheme function
2360 associated with an applicable struct. There is some VM support for the
2361 underlying primitives, like `class-of'.
2363 This change will in the future allow users to customize generic function
2364 dispatch without incurring a performance penalty, and allow us to
2365 implement method combinations.
2367 ** Applicable struct support
2369 One may now make structs from Scheme that may be applied as procedures.
2370 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
2371 That struct will be the vtable of your applicable structs; instances of
2372 that new struct are assumed to have the procedure in their first slot.
2373 `<applicable-struct-vtable>' is like Common Lisp's
2374 `funcallable-standard-class'. Likewise there is
2375 `<applicable-struct-with-setter-vtable>', which looks for the setter in
2376 the second slot. This needs to be better documented.
2380 GOOPS had a number of concepts that were relevant to the days of Tcl,
2381 but not any more: operators and entities, mainly. These objects were
2382 never documented, and it is unlikely that they were ever used. Operators
2383 were a kind of generic specific to the Tcl support. Entities were
2384 replaced by applicable structs, mentioned above.
2386 ** New struct slot allocation: "hidden"
2388 A hidden slot is readable and writable, but will not be initialized by a
2389 call to make-struct. For example in your layout you would say "ph"
2390 instead of "pw". Hidden slots are useful for adding new slots to a
2391 vtable without breaking existing invocations to make-struct.
2393 ** eqv? not a generic
2395 One used to be able to extend `eqv?' as a primitive-generic, but no
2396 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
2397 should be able to compile to static dispatch tables, it doesn't make
2398 sense to allow extensions that would subvert this optimization.
2400 ** `inet-ntop' and `inet-pton' are always available.
2402 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
2403 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
2404 functions are deprecated.
2406 ** `getopt-long' parsing errors throw to `quit', not `misc-error'
2408 This change should inhibit backtraces on argument parsing errors.
2409 `getopt-long' has been modified to print out the error that it throws
2412 ** New primitive: `tmpfile'.
2414 See "File System" in the manual.
2416 ** Random generator state may be serialized to a datum
2418 `random-state->datum' will serialize a random state to a datum, which
2419 may be written out, read back in later, and revivified using
2420 `datum->random-state'. See "Random" in the manual, for more details.
2422 ** Fix random number generator on 64-bit platforms
2424 There was a nasty bug on 64-bit platforms in which asking for a random
2425 integer with a range between 2**32 and 2**64 caused a segfault. After
2426 many embarrassing iterations, this was fixed.
2428 ** Fast bit operations.
2430 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
2431 have dedicated bytecodes. Guile is not just for symbolic computation,
2432 it's for number crunching too.
2434 ** Faster SRFI-9 record access
2436 SRFI-9 records are now implemented directly on top of Guile's structs,
2437 and their accessors are defined in such a way that normal call-sites
2438 inline to special VM opcodes, while still allowing for the general case
2439 (e.g. passing a record accessor to `apply').
2441 ** R6RS block comment support
2443 Guile now supports R6RS nested block comments. The start of a comment is
2444 marked with `#|', and the end with `|#'.
2446 ** `guile-2' cond-expand feature
2448 To test if your code is running under Guile 2.0 (or its alpha releases),
2449 test for the `guile-2' cond-expand feature. Like this:
2451 (cond-expand (guile-2 (eval-when (compile)
2452 ;; This must be evaluated at compile time.
2453 (fluid-set! current-reader my-reader)))
2455 ;; Earlier versions of Guile do not have a
2456 ;; separate compilation phase.
2457 (fluid-set! current-reader my-reader)))
2459 ** New global variables: %load-compiled-path, %load-compiled-extensions
2461 These are analogous to %load-path and %load-extensions.
2463 ** New fluid: `%file-port-name-canonicalization'
2465 This fluid parameterizes the file names that are associated with file
2466 ports. If %file-port-name-canonicalization is 'absolute, then file names
2467 are canonicalized to be absolute paths. If it is 'relative, then the
2468 name is canonicalized, but any prefix corresponding to a member of
2469 `%load-path' is stripped off. Otherwise the names are passed through
2472 In addition, the `compile-file' and `compile-and-load' procedures bind
2473 %file-port-name-canonicalization to their `#:canonicalization' keyword
2474 argument, which defaults to 'relative. In this way, one might compile
2475 "../module/ice-9/boot-9.scm", but the path that gets residualized into
2476 the .go is "ice-9/boot-9.scm".
2478 ** New procedure, `make-promise'
2480 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
2482 ** `defined?' may accept a module as its second argument
2484 Previously it only accepted internal structures from the evaluator.
2486 ** New entry into %guile-build-info: `ccachedir'
2488 ** Fix bug in `module-bound?'.
2490 `module-bound?' was returning true if a module did have a local
2491 variable, but one that was unbound, but another imported module bound
2492 the variable. This was an error, and was fixed.
2494 ** `(ice-9 syncase)' has been deprecated.
2496 As syntax-case is available by default, importing `(ice-9 syncase)' has
2497 no effect, and will trigger a deprecation warning.
2499 ** New readline history functions
2501 The (ice-9 readline) module now provides add-history, read-history,
2502 write-history and clear-history, which wrap the corresponding GNU
2503 History library functions.
2505 ** Removed deprecated uniform array procedures:
2506 dimensions->uniform-array, list->uniform-array, array-prototype
2508 Instead, use make-typed-array, list->typed-array, or array-type,
2511 ** Deprecate the old `scm-style-repl'
2513 The following bindings from boot-9 are now found in `(ice-9
2514 scm-style-repl)': `scm-style-repl', `error-catching-loop',
2515 `error-catching-repl', `bad-throw', `scm-repl-silent'
2516 `assert-repl-silence', `repl-print-unspecified',
2517 `assert-repl-print-unspecified', `scm-repl-verbose',
2518 `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl',
2519 `default-pre-unwind-handler', `handle-system-error',
2521 The following bindings have been deprecated, with no replacement:
2522 `pre-unwind-handler-dispatch'.
2524 The following bindings have been totally removed:
2525 `before-signal-stack'.
2527 Deprecated forwarding shims have been installed so that users that
2528 expect these bindings in the main namespace will still work, but receive
2529 a deprecation warning.
2531 ** `set-batch-mode?!' replaced by `ensure-batch-mode!'
2533 "Batch mode" is a flag used to tell a program that it is not running
2534 interactively. One usually turns it on after a fork. It may not be
2535 turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!',
2536 because it is a better interface, as it can only turn on batch mode, not
2539 ** Deprecate `save-stack', `the-last-stack'
2541 It used to be that the way to debug programs in Guile was to capture the
2542 stack at the time of error, drop back to the REPL, then debug that
2543 stack. But this approach didn't compose, was tricky to get right in the
2544 presence of threads, and was not very powerful.
2546 So `save-stack', `stack-saved?', and `the-last-stack' have been moved to
2547 `(ice-9 save-stack)', with deprecated bindings left in the root module.
2549 ** `top-repl' has its own module
2551 The `top-repl' binding, called with Guile is run interactively, is now
2552 is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was
2553 left in the default environment.
2555 ** `display-error' takes a frame
2557 The `display-error' / `scm_display_error' helper now takes a frame as an
2558 argument instead of a stack. Stacks are still supported in deprecated
2559 builds. Additionally, `display-error' will again source location
2560 information for the error.
2562 ** No more `(ice-9 debug)'
2564 This module had some debugging helpers that are no longer applicable to
2565 the current debugging model. Importing this module will produce a
2566 deprecation warning. Users should contact bug-guile for support.
2568 ** Remove obsolete debug-options
2570 Removed `breakpoints', `trace', `procnames', `indent', `frames',
2571 `maxdepth', and `debug' debug-options.
2573 ** `backtrace' debug option on by default
2575 Given that Guile 2.0 can always give you a backtrace, backtraces are now
2578 ** `turn-on-debugging' deprecated
2580 ** Remove obsolete print-options
2582 The `source' and `closure-hook' print options are obsolete, and have
2585 ** Remove obsolete read-options
2587 The "elisp-strings" and "elisp-vectors" read options were unused and
2588 obsolete, so they have been removed.
2590 ** Remove eval-options and trap-options
2592 Eval-options and trap-options are obsolete with the new VM and
2595 ** Remove (ice-9 debugger) and (ice-9 debugging)
2597 See "Traps" and "Interactive Debugging" in the manual, for information
2598 on their replacements.
2600 ** Remove the GDS Emacs integration
2602 See "Using Guile in Emacs" in the manual, for info on how we think you
2603 should use Guile with Emacs.
2605 ** Deprecated: `lazy-catch'
2607 `lazy-catch' was a form that captured the stack at the point of a
2608 `throw', but the dynamic state at the point of the `catch'. It was a bit
2609 crazy. Please change to use `catch', possibly with a throw-handler, or
2610 `with-throw-handler'.
2612 ** Deprecated: primitive properties
2614 The `primitive-make-property', `primitive-property-set!',
2615 `primitive-property-ref', and `primitive-property-del!' procedures were
2616 crufty and only used to implement object properties, which has a new,
2617 threadsafe implementation. Use object properties or weak hash tables
2620 ** Deprecated `@bind' syntax
2622 `@bind' was part of an older implementation of the Emacs Lisp language,
2623 and is no longer used.
2625 ** Miscellaneous other deprecations
2627 `cuserid' has been deprecated, as it only returns 8 bytes of a user's
2628 login. Use `(passwd:name (getpwuid (geteuid)))' instead.
2630 Additionally, the procedures `apply-to-args', `has-suffix?', `scheme-file-suffix'
2631 `get-option', `for-next-option', `display-usage-report',
2632 `transform-usage-lambda', `collect', and `set-batch-mode?!' have all
2635 ** Add support for unbound fluids
2637 See `make-unbound-fluid', `fluid-unset!', and `fluid-bound?' in the
2640 ** Add `variable-unset!'
2642 See "Variables" in the manual, for more details.
2644 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
2646 * Changes to the C interface
2648 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
2650 The semantics of `scm_gc_malloc ()' have been changed, in a
2651 backward-compatible way. A new allocation routine,
2652 `scm_gc_malloc_pointerless ()', was added.
2654 Libgc is a conservative GC, which we hope will make interaction with C
2655 code easier and less error-prone.
2657 ** New procedures: `scm_to_stringn', `scm_from_stringn'
2658 ** New procedures: scm_{to,from}_{utf8,latin1}_symbol{n,}
2659 ** New procedures: scm_{to,from}_{utf8,utf32,latin1}_string{n,}
2661 These new procedures convert to and from string representations in
2662 particular encodings.
2664 Users should continue to use locale encoding for user input, user
2665 output, or interacting with the C library.
2667 Use the Latin-1 functions for ASCII, and for literals in source code.
2669 Use UTF-8 functions for interaction with modern libraries which deal in
2670 UTF-8, and UTF-32 for interaction with utf32-using libraries.
2672 Otherwise, use scm_to_stringn or scm_from_stringn with a specific
2675 ** New type definitions for `scm_t_intptr' and friends.
2677 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
2678 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
2679 available to C. Have fun!
2681 ** The GH interface (deprecated in version 1.6, 2001) was removed.
2683 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
2685 This makes these internal functions technically not callable from
2688 ** Functions for handling `scm_option' now no longer require an argument
2689 indicating length of the `scm_t_option' array.
2691 ** Procedures-with-setters are now implemented using applicable structs
2693 From a user's perspective this doesn't mean very much. But if, for some
2694 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
2695 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
2698 ** Remove old evaluator closures
2700 There used to be ranges of typecodes allocated to interpreted data
2701 structures, but that it no longer the case, given that interpreted
2702 procedure are now just regular VM closures. As a result, there is a
2703 newly free tc3, and a number of removed macros. See the ChangeLog for
2706 ** Primitive procedures are now VM trampoline procedures
2708 It used to be that there were something like 12 different typecodes
2709 allocated to primitive procedures, each with its own calling convention.
2710 Now there is only one, the gsubr. This may affect user code if you were
2711 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
2712 solution is to switch to use scm_c_make_gsubr. This solution works well
2713 both with the old 1.8 and with the current 1.9 branch.
2715 Guile's old evaluator used to have special cases for applying "gsubrs",
2716 primitive procedures with specified numbers of required, optional, and
2717 rest arguments. Now, however, Guile represents gsubrs as normal VM
2718 procedures, with appropriate bytecode to parse out the correct number of
2719 arguments, including optional and rest arguments, and then with a
2720 special bytecode to apply the gsubr.
2722 This allows primitive procedures to appear on the VM stack, allowing
2723 them to be accurately counted in profiles. Also they now have more
2724 debugging information attached to them -- their number of arguments, for
2725 example. In addition, the VM can completely inline the application
2726 mechanics, allowing for faster primitive calls.
2728 However there are some changes on the C level. There is no more
2729 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
2730 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
2731 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
2732 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
2733 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
2735 Perhaps more significantly, `scm_c_make_subr',
2736 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
2737 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
2738 and there are no more subr typecodes. Use the scm_c_make_gsubr family
2741 Normal users of gsubrs should not be affected, though, as the
2742 scm_c_make_gsubr family still is the correct way to create primitive
2745 ** Remove deprecated array C interfaces
2747 Removed the deprecated array functions `scm_i_arrayp',
2748 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
2749 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
2750 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
2751 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
2753 ** Remove unused snarf macros
2755 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
2756 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
2758 ** New functions: `scm_call_n', `scm_c_run_hookn'
2760 `scm_call_n' applies to apply a function to an array of arguments.
2761 `scm_c_run_hookn' runs a hook with an array of arguments.
2763 ** Some SMOB types changed to have static typecodes
2765 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
2766 they have statically allocated tc7 typecodes.
2768 ** Preparations for changing SMOB representation
2770 If things go right, we'll be changing the SMOB representation soon. To
2771 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
2772 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
2773 in the future. Code accessing SMOBs using SCM_CELL macros was never
2774 correct, but until now things still worked. Users should be aware of
2777 ** Changed invocation mechanics of applicable SMOBs
2779 Guile's old evaluator used to have special cases for applying SMOB
2780 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
2781 trampoline procedure for it, and use the normal mechanics to apply the
2782 trampoline. This simplifies procedure application in the normal,
2785 The upshot is that the mechanics used to apply a SMOB are different from
2786 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
2787 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
2788 deprecated. Just use the scm_call_0 family of procedures.
2790 ** Removed support shlibs for SRFIs 1, 4, 13, 14, and 60
2792 Though these SRFI support libraries did expose API, they encoded a
2793 strange version string into their library names. That version was never
2794 programmatically exported, so there was no way people could use the
2797 This was a fortunate oversight, as it allows us to remove the need for
2798 extra, needless shared libraries --- the C support code for SRFIs 4, 13,
2799 and 14 was already in core --- and allow us to incrementally return the
2800 SRFI implementation to Scheme.
2802 ** New C function: scm_module_public_interface
2804 This procedure corresponds to Scheme's `module-public-interface'.
2806 ** Undeprecate `scm_the_root_module ()'
2808 It's useful to be able to get the root module from C without doing a
2811 ** Inline vector allocation
2813 Instead of having vectors point out into the heap for their data, their
2814 data is now allocated inline to the vector object itself. The same is
2815 true for bytevectors, by default, though there is an indirection
2816 available which should allow for making a bytevector from an existing
2819 ** New struct constructors that don't involve making lists
2821 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
2822 constructors, respectively, for structs. You might find them useful.
2826 In Guile 1.8, there were debugging frames on the C stack. Now there is
2827 no more need to explicitly mark the stack in this way, because Guile has
2828 a VM stack that it knows how to walk, which simplifies the C API
2829 considerably. See the ChangeLog for details; the relevant interface is
2830 in libguile/stacks.h. The Scheme API has not been changed significantly.
2832 ** Removal of Guile's primitive object system.
2834 There were a number of pieces in `objects.[ch]' that tried to be a
2835 minimal object system, but were never documented, and were quickly
2836 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
2837 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
2838 from objects.h are no more. In the very unlikely case in which these
2839 were useful to you, we urge you to contact guile-devel.
2843 Actually the future is still in the state that it was, is, and ever
2844 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
2845 part of it. These files were experimental, never compiled, and would be
2846 better implemented in Scheme anyway. In the future, that is.
2848 ** Deprecate trampolines
2850 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
2851 so on. The point was to do some precomputation on the type of the
2852 procedure, then return a specialized "call" procedure. However this
2853 optimization wasn't actually an optimization, so it is now deprecated.
2854 Just use `scm_call_0', etc instead.
2856 ** Deprecated `scm_badargsp'
2858 This function is unused in Guile, but was part of its API.
2860 ** Better support for Lisp `nil'.
2862 The bit representation of `nil' has been tweaked so that it is now very
2863 efficient to check e.g. if a value is equal to Scheme's end-of-list or
2864 Lisp's nil. Additionally there are a heap of new, specific predicates
2865 like scm_is_null_or_nil.
2867 ** Better integration of Lisp `nil'.
2869 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
2870 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
2871 but when we start to integrate more with Emacs, it is possible that we
2872 break code that assumes that, for example, `(not x)' implies that `x' is
2873 `eq?' to `#f'. This is not a common assumption. Refactoring affected
2874 code to rely on properties instead of identities will improve code
2875 correctness. See "Nil" in the manual, for more details.
2877 ** Support for static allocation of strings, symbols, and subrs.
2879 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
2880 much of their associated data as static variables, reducing Guile's
2883 ** `scm_stat' has an additional argument, `exception_on_error'
2884 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
2886 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
2888 Previously they would use the `off_t' type, which is fragile since its
2889 definition depends on the application's value for `_FILE_OFFSET_BITS'.
2891 ** The `long_long' C type, deprecated in 1.8, has been removed
2893 ** Removed deprecated uniform array procedures: scm_make_uve,
2894 scm_array_prototype, scm_list_to_uniform_array,
2895 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
2896 scm_ra_set_contp, scm_aind, scm_raprin1
2898 These functions have been deprecated since early 2005.
2900 * Changes to the distribution
2902 ** Guile's license is now LGPLv3+
2904 In other words the GNU Lesser General Public License, version 3 or
2905 later (at the discretion of each person that chooses to redistribute
2910 Guile's build is visually quieter, due to the use of Automake 1.11's
2911 AM_SILENT_RULES. Build as `make V=1' to see all of the output.
2913 ** GOOPS documentation folded into Guile reference manual
2915 GOOPS, Guile's object system, used to be documented in separate manuals.
2916 This content is now included in Guile's manual directly.
2918 ** `guile-config' will be deprecated in favor of `pkg-config'
2920 `guile-config' has been rewritten to get its information from
2921 `pkg-config', so this should be a transparent change. Note however that
2922 guile.m4 has yet to be modified to call pkg-config instead of
2925 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
2927 Programs that use `pkg-config' to find Guile or one of its Autoconf
2928 macros should now require `guile-2.0' instead of `guile-1.8'.
2930 ** New installation directory: $(pkglibdir)/1.9/ccache
2932 If $(libdir) is /usr/lib, for example, Guile will install its .go files
2933 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
2935 ** Parallel installability fixes
2937 Guile now installs its header files to a effective-version-specific
2938 directory, and includes the effective version (e.g. 2.0) in the library
2939 name (e.g. libguile-2.0.so).
2941 This change should be transparent to users, who should detect Guile via
2942 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
2943 parallel installs for multiple versions of Guile development
2946 ** Dynamically loadable extensions may be placed in a Guile-specific path
2948 Before, Guile only searched the system library paths for extensions
2949 (e.g. /usr/lib), which meant that the names of Guile extensions had to
2950 be globally unique. Installing them to a Guile-specific extensions
2951 directory is cleaner. Use `pkg-config --variable=extensiondir
2952 guile-2.0' to get the location of the extensions directory.
2954 ** User Scheme code may be placed in a version-specific path
2956 Before, there was only one way to install user Scheme code to a
2957 version-specific Guile directory: install to Guile's own path,
2958 e.g. /usr/share/guile/2.0. The site directory,
2959 e.g. /usr/share/guile/site, was unversioned. This has been changed to
2960 add a version-specific site directory, e.g. /usr/share/guile/site/2.0,
2961 searched before the global site directory.
2963 ** New dependency: libgc
2965 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
2967 ** New dependency: GNU libunistring
2969 See http://www.gnu.org/software/libunistring/, for more information. Our
2970 Unicode support uses routines from libunistring.
2972 ** New dependency: libffi
2974 See http://sourceware.org/libffi/, for more information.
2978 Changes in 1.8.8 (since 1.8.7)
2982 ** Fix possible buffer overruns when parsing numbers
2983 ** Avoid clash with system setjmp/longjmp on IA64
2984 ** Fix `wrong type arg' exceptions with IPv6 addresses
2987 Changes in 1.8.7 (since 1.8.6)
2989 * New modules (see the manual for details)
2991 ** `(srfi srfi-98)', an interface to access environment variables
2995 ** Fix compilation with `--disable-deprecated'
2996 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
2997 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
2998 ** Fix build problem when scm_t_timespec is different from struct timespec
2999 ** Fix build when compiled with -Wundef -Werror
3000 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
3001 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
3002 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
3003 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
3004 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
3005 ** Fix reading of complex numbers where both parts are inexact decimals
3007 ** Allow @ macro to work with (ice-9 syncase)
3009 Previously, use of the @ macro in a module whose code is being
3010 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
3011 Now it works as you would expect (giving the value of the specified
3014 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
3017 Changes in 1.8.6 (since 1.8.5)
3019 * New features (see the manual for details)
3021 ** New convenience function `scm_c_symbol_length ()'
3023 ** Single stepping through code from Emacs
3025 When you use GDS to evaluate Scheme code from Emacs, you can now use
3026 `C-u' to indicate that you want to single step through that code. See
3027 `Evaluating Scheme Code' in the manual for more details.
3029 ** New "guile(1)" man page!
3031 * Changes to the distribution
3033 ** Automake's `AM_MAINTAINER_MODE' is no longer used
3035 Thus, the `--enable-maintainer-mode' configure option is no longer
3036 available: Guile is now always configured in "maintainer mode".
3038 ** `ChangeLog' files are no longer updated
3040 Instead, changes are detailed in the version control system's logs. See
3041 the top-level `ChangeLog' files for details.
3046 ** `symbol->string' now returns a read-only string, as per R5RS
3047 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
3048 ** `guile-config link' now prints `-L$libdir' before `-lguile'
3049 ** Fix memory corruption involving GOOPS' `class-redefinition'
3050 ** Fix possible deadlock in `mutex-lock'
3051 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
3052 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
3053 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
3054 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
3055 ** Fix misleading output from `(help rationalize)'
3056 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
3057 ** Fix `gcd' when called with a single, negative argument.
3058 ** Fix `Stack overflow' errors seen when building on some platforms
3059 ** Fix bug when `scm_with_guile ()' was called several times from the
3061 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
3062 dynamic environment of the call to `raise'
3063 ** Fix potential deadlock in `make-struct'
3064 ** Fix compilation problem with libltdl from Libtool 2.2.x
3065 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
3068 Changes in 1.8.5 (since 1.8.4)
3070 * Infrastructure changes
3072 ** Guile repository switched from CVS to Git
3074 The new repository can be accessed using
3075 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
3076 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
3078 ** Add support for `pkg-config'
3080 See "Autoconf Support" in the manual for details.
3082 * New modules (see the manual for details)
3086 * New features (see the manual for details)
3088 ** New `postfix' read option, for SRFI-88 keyword syntax
3089 ** Some I/O primitives have been inlined, which improves I/O performance
3090 ** New object-based traps infrastructure
3092 This is a GOOPS-based infrastructure that builds on Guile's low-level
3093 evaluator trap calls and facilitates the development of debugging
3094 features like single-stepping, breakpoints, tracing and profiling.
3095 See the `Traps' node of the manual for details.
3097 ** New support for working on Guile code from within Emacs
3099 Guile now incorporates the `GDS' library (previously distributed
3100 separately) for working on Guile code from within Emacs. See the
3101 `Using Guile In Emacs' node of the manual for details.
3105 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
3106 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
3108 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
3109 would trigger an unbound variable error for `match:andmap'.
3111 ** `(oop goops describe)' now properly provides the `describe' feature
3112 ** Fixed `args-fold' from `(srfi srfi-37)'
3114 Previously, parsing short option names of argument-less options would
3115 lead to a stack overflow.
3117 ** `(srfi srfi-35)' is now visible through `cond-expand'
3118 ** Fixed type-checking for the second argument of `eval'
3119 ** Fixed type-checking for SRFI-1 `partition'
3120 ** Fixed `struct-ref' and `struct-set!' on "light structs"
3121 ** Honor struct field access rights in GOOPS
3122 ** Changed the storage strategy of source properties, which fixes a deadlock
3123 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
3124 ** Fixed build issue for GNU/Linux on IA64
3125 ** Fixed build issues on NetBSD 1.6
3126 ** Fixed build issue on Solaris 2.10 x86_64
3127 ** Fixed build issue with DEC/Compaq/HP's compiler
3128 ** Fixed `scm_from_complex_double' build issue on FreeBSD
3129 ** Fixed `alloca' build issue on FreeBSD 6
3130 ** Removed use of non-portable makefile constructs
3131 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
3132 ** Make sure all tests honor `$TMPDIR'
3135 Changes in 1.8.4 (since 1.8.3)
3139 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
3140 ** Fixed a segmentation fault which occurred when displaying the
3141 backtrace of a stack with a promise object (made by `delay') in it.
3142 ** Make `accept' leave guile mode while blocking
3143 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
3144 ** Fixed a build problem on AIX (use of func_data identifier)
3145 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
3146 called with an associator proc that returns neither a pair nor #f.
3147 ** Secondary threads now always return a valid module for (current-module).
3148 ** Avoid MacOS build problems caused by incorrect combination of "64"
3149 system and library calls.
3150 ** `guile-snarf' now honors `$TMPDIR'
3151 ** `guile-config compile' now reports CPPFLAGS used at compile-time
3152 ** Fixed build with Sun Studio (Solaris 9)
3153 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
3154 uniform vectors on AIX.
3155 ** Fixed a deadlock that occurs upon GC with multiple threads.
3156 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
3157 ** Fixed autotool-derived build problems on AIX 6.1.
3158 ** Fixed NetBSD/alpha support
3159 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
3161 * New modules (see the manual for details)
3165 * Documentation fixes and improvements
3167 ** Removed premature breakpoint documentation
3169 The features described are not available in the series of 1.8.x
3170 releases, so the documentation was misleading and has been removed.
3172 ** More about Guile's default *random-state* variable
3174 ** GOOPS: more about how to use `next-method'
3176 * Changes to the distribution
3178 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
3180 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
3181 General Public License, and the few incorrect files have now been
3182 fixed to agree with the rest of the Guile distribution.
3184 ** Removed unnecessary extra copies of COPYING*
3186 The distribution now contains a single COPYING.LESSER at its top level.
3189 Changes in 1.8.3 (since 1.8.2)
3191 * New modules (see the manual for details)
3198 ** The `(ice-9 slib)' module now works as expected
3199 ** Expressions like "(set! 'x #t)" no longer yield a crash
3200 ** Warnings about duplicate bindings now go to stderr
3201 ** A memory leak in `make-socket-address' was fixed
3202 ** Alignment issues (e.g., on SPARC) in network routines were fixed
3203 ** A threading issue that showed up at least on NetBSD was fixed
3204 ** Build problems on Solaris and IRIX fixed
3206 * Implementation improvements
3208 ** The reader is now faster, which reduces startup time
3209 ** Procedures returned by `record-accessor' and `record-modifier' are faster
3212 Changes in 1.8.2 (since 1.8.1):
3214 * New procedures (see the manual for details)
3216 ** set-program-arguments
3219 * Incompatible changes
3221 ** The body of a top-level `define' no longer sees the binding being created
3223 In a top-level `define', the binding being created is no longer visible
3224 from the `define' body. This breaks code like
3225 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
3226 unbound in the body. However, such code was not R5RS-compliant anyway,
3231 ** Fractions were not `equal?' if stored in unreduced form.
3232 (A subtle problem, since printing a value reduced it, making it work.)
3233 ** srfi-60 `copy-bit' failed on 64-bit systems
3234 ** "guile --use-srfi" option at the REPL can replace core functions
3235 (Programs run with that option were ok, but in the interactive REPL
3236 the core bindings got priority, preventing SRFI replacements or
3238 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
3239 ** `kill' on mingw throws an error for a PID other than oneself
3240 ** Procedure names are attached to procedure-with-setters
3241 ** Array read syntax works with negative lower bound
3242 ** `array-in-bounds?' fix if an array has different lower bounds on each index
3243 ** `*' returns exact 0 for "(* inexact 0)"
3244 This follows what it always did for "(* 0 inexact)".
3245 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
3246 ** SRFI-19: `date->julian-day' did not account for timezone offset
3247 ** `ttyname' no longer crashes when passed a non-tty argument
3248 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
3249 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
3250 ** GOOPS: Fixed a bug in `method-more-specific?'
3251 ** Build problems on Solaris fixed
3252 ** Build problems on HP-UX IA64 fixed
3253 ** Build problems on MinGW fixed
3256 Changes in 1.8.1 (since 1.8.0):
3258 * LFS functions are now used to access 64-bit files on 32-bit systems.
3260 * New procedures (see the manual for details)
3262 ** primitive-_exit - [Scheme] the-root-module
3263 ** scm_primitive__exit - [C]
3264 ** make-completion-function - [Scheme] (ice-9 readline)
3265 ** scm_c_locale_stringn_to_number - [C]
3266 ** scm_srfi1_append_reverse [C]
3267 ** scm_srfi1_append_reverse_x [C]
3275 ** Build problems have been fixed on MacOS, SunOS, and QNX.
3277 ** `strftime' fix sign of %z timezone offset.
3279 ** A one-dimensional array can now be 'equal?' to a vector.
3281 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
3283 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
3285 ** `record-accessor' and `record-modifier' now have strict type checks.
3287 Record accessor and modifier procedures now throw an error if the
3288 record type of the record they're given is not the type expected.
3289 (Previously accessors returned #f and modifiers silently did nothing).
3291 ** It is now OK to use both autoload and use-modules on a given module.
3293 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
3295 Previously there was no checking on primatives like make-vector that
3296 accept "one or two" arguments. Now there is.
3298 ** The srfi-1 assoc function now calls its equality predicate properly.
3300 Previously srfi-1 assoc would call the equality predicate with the key
3301 last. According to the SRFI, the key should be first.
3303 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
3305 ** The array-set! procedure no longer segfaults when given a bit vector.
3307 ** Bugs in make-shared-array have been fixed.
3309 ** string<? and friends now follow char<? etc order on 8-bit chars.
3311 ** The format procedure now handles inf and nan values for ~f correctly.
3313 ** exact->inexact should no longer overflow when given certain large fractions.
3315 ** srfi-9 accessor and modifier procedures now have strict record type checks.
3317 This matches the srfi-9 specification.
3319 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
3321 Previously the (ice-9 ftw) procedures would ignore any file that had
3322 the same inode number as a file they had already seen, even if that
3323 file was on a different device.
3326 Changes in 1.8.0 (changes since the 1.6.x series):
3328 * Changes to the distribution
3330 ** Guile is now licensed with the GNU Lesser General Public License.
3332 ** The manual is now licensed with the GNU Free Documentation License.
3334 ** Guile now requires GNU MP (http://swox.com/gmp).
3336 Guile now uses the GNU MP library for arbitrary precision arithmetic.
3338 ** Guile now has separate private and public configuration headers.
3340 That is, things like HAVE_STRING_H no longer leak from Guile's
3343 ** Guile now provides and uses an "effective" version number.
3345 Guile now provides scm_effective_version and effective-version
3346 functions which return the "effective" version number. This is just
3347 the normal full version string without the final micro-version number,
3348 so the current effective-version is "1.8". The effective version
3349 should remain unchanged during a stable series, and should be used for
3350 items like the versioned share directory name
3351 i.e. /usr/share/guile/1.8.
3353 Providing an unchanging version number during a stable release for
3354 things like the versioned share directory can be particularly
3355 important for Guile "add-on" packages, since it provides a directory
3356 that they can install to that won't be changed out from under them
3357 with each micro release during a stable series.
3359 ** Thread implementation has changed.
3361 When you configure "--with-threads=null", you will get the usual
3362 threading API (call-with-new-thread, make-mutex, etc), but you can't
3363 actually create new threads. Also, "--with-threads=no" is now
3364 equivalent to "--with-threads=null". This means that the thread API
3365 is always present, although you might not be able to create new
3368 When you configure "--with-threads=pthreads" or "--with-threads=yes",
3369 you will get threads that are implemented with the portable POSIX
3370 threads. These threads can run concurrently (unlike the previous
3371 "coop" thread implementation), but need to cooperate for things like
3374 The default is "pthreads", unless your platform doesn't have pthreads,
3375 in which case "null" threads are used.
3377 See the manual for details, nodes "Initialization", "Multi-Threading",
3378 "Blocking", and others.
3380 ** There is the new notion of 'discouraged' features.
3382 This is a milder form of deprecation.
3384 Things that are discouraged should not be used in new code, but it is
3385 OK to leave them in old code for now. When a discouraged feature is
3386 used, no warning message is printed like there is for 'deprecated'
3387 features. Also, things that are merely discouraged are nevertheless
3388 implemented efficiently, while deprecated features can be very slow.
3390 You can omit discouraged features from libguile by configuring it with
3391 the '--disable-discouraged' option.
3393 ** Deprecation warnings can be controlled at run-time.
3395 (debug-enable 'warn-deprecated) switches them on and (debug-disable
3396 'warn-deprecated) switches them off.
3398 ** Support for SRFI 61, extended cond syntax for multiple values has
3401 This SRFI is always available.
3403 ** Support for require-extension, SRFI-55, has been added.
3405 The SRFI-55 special form `require-extension' has been added. It is
3406 available at startup, and provides a portable way to load Scheme
3407 extensions. SRFI-55 only requires support for one type of extension,
3408 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
3411 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
3413 The (srfi srfi-26) module is an implementation of SRFI-26 which
3414 provides the `cut' and `cute' syntax. These may be used to specialize
3415 parameters without currying.
3417 ** New module (srfi srfi-31)
3419 This is an implementation of SRFI-31 which provides a special form
3420 `rec' for recursive evaluation.
3422 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
3423 been merged with the core, making their functionality always
3426 The modules are still available, tho, and you could use them together
3427 with a renaming import, for example.
3429 ** Guile no longer includes its own version of libltdl.
3431 The official version is good enough now.
3433 ** The --enable-htmldoc option has been removed from 'configure'.
3435 Support for translating the documentation into HTML is now always
3436 provided. Use 'make html'.
3438 ** New module (ice-9 serialize):
3440 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
3441 don't trust the thread safety of most of your program, but where you
3442 have some section(s) of code which you consider can run in parallel to
3443 other sections. See ice-9/serialize.scm for more information.
3445 ** The configure option '--disable-arrays' has been removed.
3447 Support for arrays and uniform numeric arrays is now always included
3450 * Changes to the stand-alone interpreter
3452 ** New command line option `-L'.
3454 This option adds a directory to the front of the load path.
3456 ** New command line option `--no-debug'.
3458 Specifying `--no-debug' on the command line will keep the debugging
3459 evaluator turned off, even for interactive sessions.
3461 ** User-init file ~/.guile is now loaded with the debugging evaluator.
3463 Previously, the normal evaluator would have been used. Using the
3464 debugging evaluator gives better error messages.
3466 ** The '-e' option now 'read's its argument.
3468 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
3469 be used with '-e'. For example, you can now write a script like
3472 exec guile -e '(@ (demo) main)' -s "$0" "$@"
3475 (define-module (demo)
3479 (format #t "Demo: ~a~%" args))
3482 * Changes to Scheme functions and syntax
3484 ** Guardians have changed back to their original semantics
3486 Guardians now behave like described in the paper by Dybvig et al. In
3487 particular, they no longer make guarantees about the order in which
3488 they return objects, and they can no longer be greedy.
3490 They no longer drop cyclic data structures.
3492 The C function scm_make_guardian has been changed incompatibly and no
3493 longer takes the 'greedy_p' argument.
3495 ** New function hashx-remove!
3497 This function completes the set of 'hashx' functions.
3499 ** The concept of dynamic roots has been factored into continuation
3500 barriers and dynamic states.
3502 Each thread has a current dynamic state that carries the values of the
3503 fluids. You can create and copy dynamic states and use them as the
3504 second argument for 'eval'. See "Fluids and Dynamic States" in the
3507 To restrict the influence that captured continuations can have on the
3508 control flow, you can errect continuation barriers. See "Continuation
3509 Barriers" in the manual.
3511 The function call-with-dynamic-root now essentially temporarily
3512 installs a new dynamic state and errects a continuation barrier.
3514 ** The default load path no longer includes "." at the end.
3516 Automatically loading modules from the current directory should not
3517 happen by default. If you want to allow it in a more controlled
3518 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
3519 variable %load-path.
3521 ** The uniform vector and array support has been overhauled.
3523 It now complies with SRFI-4 and the weird prototype based uniform
3524 array creation has been deprecated. See the manual for more details.
3526 Some non-compatible changes have been made:
3527 - characters can no longer be stored into byte arrays.
3528 - strings and bit vectors are no longer considered to be uniform numeric
3530 - array-rank throws an error for non-arrays instead of returning zero.
3531 - array-ref does no longer accept non-arrays when no indices are given.
3533 There is the new notion of 'generalized vectors' and corresponding
3534 procedures like 'generalized-vector-ref'. Generalized vectors include
3535 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
3537 Arrays use generalized vectors as their storage, so that you still
3538 have arrays of characters, bits, etc. However, uniform-array-read!
3539 and uniform-array-write can no longer read/write strings and
3542 ** There is now support for copy-on-write substrings, mutation-sharing
3543 substrings and read-only strings.
3545 Three new procedures are related to this: substring/shared,
3546 substring/copy, and substring/read-only. See the manual for more
3549 ** Backtraces will now highlight the value that caused the error.
3551 By default, these values are enclosed in "{...}", such as in this
3560 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
3561 <unnamed port>:1:1: Wrong type (expecting pair): a
3562 ABORT: (wrong-type-arg)
3564 The prefix and suffix used for highlighting can be set via the two new
3565 printer options 'highlight-prefix' and 'highlight-suffix'. For
3566 example, putting this into ~/.guile will output the bad value in bold
3567 on an ANSI terminal:
3569 (print-set! highlight-prefix "\x1b[1m")
3570 (print-set! highlight-suffix "\x1b[22m")
3573 ** 'gettext' support for internationalization has been added.
3575 See the manual for details.
3577 ** New syntax '@' and '@@':
3579 You can now directly refer to variables exported from a module by
3582 (@ MODULE-NAME VARIABLE-NAME)
3584 For example (@ (ice-9 pretty-print) pretty-print) will directly access
3585 the pretty-print variable exported from the (ice-9 pretty-print)
3586 module. You don't need to 'use' that module first. You can also use
3587 '@' as a target of 'set!', as in (set! (@ mod var) val).
3589 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
3590 but it can also access variables that have not been exported. It is
3591 intended only for kluges and temporary fixes and for debugging, not
3594 ** Keyword syntax has been made more disciplined.
3596 Previously, the name of a keyword was read as a 'token' but printed as
3597 a symbol. Now, it is read as a general Scheme datum which must be a
3608 ERROR: In expression (a b c):
3612 ERROR: Unbound variable: foo
3617 ERROR: Wrong type (expecting symbol): 12
3621 ERROR: Wrong type (expecting symbol): (a b c)
3625 ** The printing of symbols that might look like keywords can be
3628 The new printer option 'quote-keywordish-symbols' controls how symbols
3629 are printed that have a colon as their first or last character. The
3630 default now is to only quote a symbol with #{...}# when the read
3631 option 'keywords' is not '#f'. Thus:
3633 guile> (define foo (string->symbol ":foo"))
3634 guile> (read-set! keywords #f)
3637 guile> (read-set! keywords 'prefix)
3640 guile> (print-set! quote-keywordish-symbols #f)
3644 ** 'while' now provides 'break' and 'continue'
3646 break and continue were previously bound in a while loop, but not
3647 documented, and continue didn't quite work properly. The undocumented
3648 parameter to break which gave a return value for the while has been
3651 ** 'call-with-current-continuation' is now also available under the name
3654 ** The module system now checks for duplicate bindings.
3656 The module system now can check for name conflicts among imported
3659 The behavior can be controlled by specifying one or more 'duplicates'
3660 handlers. For example, to make Guile return an error for every name
3663 (define-module (foo)
3668 The new default behavior of the module system when a name collision
3669 has been detected is to
3671 1. Give priority to bindings marked as a replacement.
3672 2. Issue a warning (different warning if overriding core binding).
3673 3. Give priority to the last encountered binding (this corresponds to
3676 If you want the old behavior back without replacements or warnings you
3679 (default-duplicate-binding-handler 'last)
3681 to your .guile init file.
3683 ** New define-module option: :replace
3685 :replace works as :export, but, in addition, marks the binding as a
3688 A typical example is `format' in (ice-9 format) which is a replacement
3689 for the core binding `format'.
3691 ** Adding prefixes to imported bindings in the module system
3693 There is now a new :use-module option :prefix. It can be used to add
3694 a prefix to all imported bindings.
3696 (define-module (foo)
3697 :use-module ((bar) :prefix bar:))
3699 will import all bindings exported from bar, but rename them by adding
3702 ** Conflicting generic functions can be automatically merged.
3704 When two imported bindings conflict and they are both generic
3705 functions, the two functions can now be merged automatically. This is
3706 activated with the 'duplicates' handler 'merge-generics'.
3708 ** New function: effective-version
3710 Returns the "effective" version number. This is just the normal full
3711 version string without the final micro-version number. See "Changes
3712 to the distribution" above.
3714 ** New threading functions: parallel, letpar, par-map, and friends
3716 These are convenient ways to run calculations in parallel in new
3717 threads. See "Parallel forms" in the manual for details.
3719 ** New function 'try-mutex'.
3721 This function will attempt to lock a mutex but will return immediately
3722 instead of blocking and indicate failure.
3724 ** Waiting on a condition variable can have a timeout.
3726 The function 'wait-condition-variable' now takes a third, optional
3727 argument that specifies the point in time where the waiting should be
3730 ** New function 'broadcast-condition-variable'.
3732 ** New functions 'all-threads' and 'current-thread'.
3734 ** Signals and system asyncs work better with threads.
3736 The function 'sigaction' now takes a fourth, optional, argument that
3737 specifies the thread that the handler should run in. When the
3738 argument is omitted, the handler will run in the thread that called
3741 Likewise, 'system-async-mark' takes a second, optional, argument that
3742 specifies the thread that the async should run in. When it is
3743 omitted, the async will run in the thread that called
3744 'system-async-mark'.
3746 C code can use the new functions scm_sigaction_for_thread and
3747 scm_system_async_mark_for_thread to pass the new thread argument.
3749 When a thread blocks on a mutex, a condition variable or is waiting
3750 for IO to be possible, it will still execute system asyncs. This can
3751 be used to interrupt such a thread by making it execute a 'throw', for
3754 ** The function 'system-async' is deprecated.
3756 You can now pass any zero-argument procedure to 'system-async-mark'.
3757 The function 'system-async' will just return its argument unchanged
3760 ** New functions 'call-with-blocked-asyncs' and
3761 'call-with-unblocked-asyncs'
3763 The expression (call-with-blocked-asyncs PROC) will call PROC and will
3764 block execution of system asyncs for the current thread by one level
3765 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
3766 procedure and will unblock the execution of system asyncs by one
3767 level for the current thread.
3769 Only system asyncs are affected by these functions.
3771 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
3773 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
3774 instead. Those functions are easier to use correctly and can be
3777 ** New function 'unsetenv'.
3779 ** New macro 'define-syntax-public'.
3781 It works like 'define-syntax' and also exports the defined macro (but
3784 ** There is support for Infinity and NaNs.
3786 Following PLT Scheme, Guile can now work with infinite numbers, and
3789 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
3790 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
3791 "+nan.0"). These numbers are inexact and have no exact counterpart.
3793 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
3794 sign of the dividend. The infinities are integers, and they answer #t
3795 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
3796 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
3807 ERROR: Numerical overflow
3809 Two new predicates 'inf?' and 'nan?' can be used to test for the
3812 ** Inexact zero can have a sign.
3814 Guile can now distinguish between plus and minus inexact zero, if your
3815 platform supports this, too. The two zeros are equal according to
3816 '=', but not according to 'eqv?'. For example
3827 ** Guile now has exact rationals.
3829 Guile can now represent fractions such as 1/3 exactly. Computing with
3830 them is also done exactly, of course:
3835 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
3836 for exact arguments.
3838 For example: (floor 2) now returns an exact 2 where in the past it
3839 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
3841 ** inexact->exact no longer returns only integers.
3843 Without exact rationals, the closest exact number was always an
3844 integer, but now inexact->exact returns the fraction that is exactly
3845 equal to a floating point number. For example:
3847 (inexact->exact 1.234)
3848 => 694680242521899/562949953421312
3850 When you want the old behavior, use 'round' explicitly:
3852 (inexact->exact (round 1.234))
3855 ** New function 'rationalize'.
3857 This function finds a simple fraction that is close to a given real
3858 number. For example (and compare with inexact->exact above):
3860 (rationalize (inexact->exact 1.234) 1/2000)
3863 Note that, as required by R5RS, rationalize returns only then an exact
3864 result when both its arguments are exact.
3866 ** 'odd?' and 'even?' work also for inexact integers.
3868 Previously, (odd? 1.0) would signal an error since only exact integers
3869 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
3870 returns #f and (odd? 1.5) signals an error.
3872 ** Guile now has uninterned symbols.
3874 The new function 'make-symbol' will return an uninterned symbol. This
3875 is a symbol that is unique and is guaranteed to remain unique.
3876 However, uninterned symbols can not yet be read back in.
3878 Use the new function 'symbol-interned?' to check whether a symbol is
3881 ** pretty-print has more options.
3883 The function pretty-print from the (ice-9 pretty-print) module can now
3884 also be invoked with keyword arguments that control things like
3885 maximum output width. See the manual for details.
3887 ** Variables have no longer a special behavior for `equal?'.
3889 Previously, comparing two variables with `equal?' would recursivly
3890 compare their values. This is no longer done. Variables are now only
3891 `equal?' if they are `eq?'.
3893 ** `(begin)' is now valid.
3895 You can now use an empty `begin' form. It will yield #<unspecified>
3896 when evaluated and simply be ignored in a definition context.
3898 ** Deprecated: procedure->macro
3900 Change your code to use 'define-macro' or r5rs macros. Also, be aware
3901 that macro expansion will not be done during evaluation, but prior to
3904 ** Soft ports now allow a `char-ready?' procedure
3906 The vector argument to `make-soft-port' can now have a length of
3907 either 5 or 6. (Previously the length had to be 5.) The optional 6th
3908 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
3909 that returns the number of characters that can be read immediately
3910 without the soft port blocking.
3912 ** Deprecated: undefine
3914 There is no replacement for undefine.
3916 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
3917 have been discouraged.
3919 They are relics from a time where a keyword like #:foo was used
3920 directly as a Tcl option "-foo" and thus keywords were internally
3921 stored as a symbol with a starting dash. We now store a symbol
3924 Use symbol->keyword and keyword->symbol instead.
3926 ** The `cheap' debug option is now obsolete
3928 Evaluator trap calls are now unconditionally "cheap" - in other words,
3929 they pass a debug object to the trap handler rather than a full
3930 continuation. The trap handler code can capture a full continuation
3931 by using `call-with-current-continuation' in the usual way, if it so
3934 The `cheap' option is retained for now so as not to break existing
3935 code which gets or sets it, but setting it now has no effect. It will
3936 be removed in the next major Guile release.
3938 ** Evaluator trap calls now support `tweaking'
3940 `Tweaking' means that the trap handler code can modify the Scheme
3941 expression that is about to be evaluated (in the case of an
3942 enter-frame trap) or the value that is being returned (in the case of
3943 an exit-frame trap). The trap handler code indicates that it wants to
3944 do this by returning a pair whose car is the symbol 'instead and whose
3945 cdr is the modified expression or return value.
3947 * Changes to the C interface
3949 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
3950 take a 'delete' function argument.
3952 This argument makes no sense since the delete function is used to
3953 remove a pair from an alist, and this must not be configurable.
3955 This is an incompatible change.
3957 ** The GH interface is now subject to the deprecation mechanism
3959 The GH interface has been deprecated for quite some time but now it is
3960 actually removed from Guile when it is configured with
3961 --disable-deprecated.
3963 See the manual "Transitioning away from GH" for more information.
3965 ** A new family of functions for converting between C values and
3966 Scheme values has been added.
3968 These functions follow a common naming scheme and are designed to be
3969 easier to use, thread-safe and more future-proof than the older
3972 - int scm_is_* (...)
3974 These are predicates that return a C boolean: 1 or 0. Instead of
3975 SCM_NFALSEP, you can now use scm_is_true, for example.
3977 - <type> scm_to_<type> (SCM val, ...)
3979 These are functions that convert a Scheme value into an appropriate
3980 C value. For example, you can use scm_to_int to safely convert from
3983 - SCM scm_from_<type> (<type> val, ...)
3985 These functions convert from a C type to a SCM value; for example,
3986 scm_from_int for ints.
3988 There is a huge number of these functions, for numbers, strings,
3989 symbols, vectors, etc. They are documented in the reference manual in
3990 the API section together with the types that they apply to.
3992 ** New functions for dealing with complex numbers in C have been added.
3994 The new functions are scm_c_make_rectangular, scm_c_make_polar,
3995 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
3996 They work like scm_make_rectangular etc but take or return doubles
3999 ** The function scm_make_complex has been discouraged.
4001 Use scm_c_make_rectangular instead.
4003 ** The INUM macros have been deprecated.
4005 A lot of code uses these macros to do general integer conversions,
4006 although the macros only work correctly with fixnums. Use the
4007 following alternatives.
4009 SCM_INUMP -> scm_is_integer or similar
4010 SCM_NINUMP -> !scm_is_integer or similar
4011 SCM_MAKINUM -> scm_from_int or similar
4012 SCM_INUM -> scm_to_int or similar
4014 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
4015 do the validating for you.
4017 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
4018 have been discouraged.
4020 Use the newer scm_to_<type> and scm_from_<type> functions instead for
4021 new code. The functions have been discouraged since they don't fit
4024 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
4026 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
4027 evaluates its argument twice. Use scm_is_true, etc. instead for new
4030 ** The macro SCM_EQ_P has been discouraged.
4032 Use scm_is_eq for new code, which fits better into the naming
4035 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
4038 Use the function scm_is_pair or scm_is_null instead.
4040 ** The functions scm_round and scm_truncate have been deprecated and
4041 are now available as scm_c_round and scm_c_truncate, respectively.
4043 These functions occupy the names that scm_round_number and
4044 scm_truncate_number should have.
4046 ** The functions scm_c_string2str, scm_c_substring2str, and
4047 scm_c_symbol2str have been deprecated.
4049 Use scm_to_locale_stringbuf or similar instead, maybe together with
4052 ** New functions scm_c_make_string, scm_c_string_length,
4053 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
4054 scm_c_substring_shared, scm_c_substring_copy.
4056 These are like scm_make_string, scm_length, etc. but are slightly
4057 easier to use from C.
4059 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
4060 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
4062 They export too many assumptions about the implementation of strings
4063 and symbols that are no longer true in the presence of
4064 mutation-sharing substrings and when Guile switches to some form of
4067 When working with strings, it is often best to use the normal string
4068 functions provided by Guile, such as scm_c_string_ref,
4069 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
4070 manual since many more such functions are now provided than
4073 When you want to convert a SCM string to a C string, use the
4074 scm_to_locale_string function or similar instead. For symbols, use
4075 scm_symbol_to_string and then work with that string. Because of the
4076 new string representation, scm_symbol_to_string does not need to copy
4077 and is thus quite efficient.
4079 ** Some string, symbol and keyword functions have been discouraged.
4081 They don't fit into the uniform naming scheme and are not explicit
4082 about the character encoding.
4084 Replace according to the following table:
4086 scm_allocate_string -> scm_c_make_string
4087 scm_take_str -> scm_take_locale_stringn
4088 scm_take0str -> scm_take_locale_string
4089 scm_mem2string -> scm_from_locale_stringn
4090 scm_str2string -> scm_from_locale_string
4091 scm_makfrom0str -> scm_from_locale_string
4092 scm_mem2symbol -> scm_from_locale_symboln
4093 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
4094 scm_str2symbol -> scm_from_locale_symbol
4096 SCM_SYMBOL_HASH -> scm_hashq
4097 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
4099 scm_c_make_keyword -> scm_from_locale_keyword
4101 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
4102 now also available to C code.
4104 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
4106 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
4107 the latter returns the true name of the keyword, not the 'dash name',
4108 as SCM_KEYWORDSYM used to do.
4110 ** A new way to access arrays in a thread-safe and efficient way has
4113 See the manual, node "Accessing Arrays From C".
4115 ** The old uniform vector and bitvector implementations have been
4116 unceremoniously removed.
4118 This implementation exposed the details of the tagging system of
4119 Guile. Use the new C API explained in the manual in node "Uniform
4120 Numeric Vectors" and "Bit Vectors", respectively.
4122 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
4123 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
4124 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
4125 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
4126 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
4127 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
4130 ** The macros dealing with vectors have been deprecated.
4132 Use the new functions scm_is_vector, scm_vector_elements,
4133 scm_vector_writable_elements, etc, or scm_is_simple_vector,
4134 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
4135 manual for more details.
4137 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
4138 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
4140 The following macros have been removed: SCM_VECTOR_BASE,
4141 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
4142 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
4144 ** Some C functions and macros related to arrays have been deprecated.
4146 Migrate according to the following table:
4148 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
4149 scm_make_ra -> scm_make_array
4150 scm_shap2ra -> scm_make_array
4151 scm_cvref -> scm_c_generalized_vector_ref
4152 scm_ra_set_contp -> do not use
4153 scm_aind -> scm_array_handle_pos
4154 scm_raprin1 -> scm_display or scm_write
4156 SCM_ARRAYP -> scm_is_array
4157 SCM_ARRAY_NDIM -> scm_c_array_rank
4158 SCM_ARRAY_DIMS -> scm_array_handle_dims
4159 SCM_ARRAY_CONTP -> do not use
4160 SCM_ARRAY_MEM -> do not use
4161 SCM_ARRAY_V -> scm_array_handle_elements or similar
4162 SCM_ARRAY_BASE -> do not use
4164 ** SCM_CELL_WORD_LOC has been deprecated.
4166 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
4167 to a SCM, as opposed to a pointer to a scm_t_bits.
4169 This was done to allow the correct use of pointers into the Scheme
4170 heap. Previously, the heap words were of type scm_t_bits and local
4171 variables and function arguments were of type SCM, making it
4172 non-standards-conformant to have a pointer that can point to both.
4174 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
4176 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
4177 second and third words of double smobs. Likewise for
4178 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
4180 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
4181 used to get and set the 16 exra bits in the zeroth word of a smob.
4183 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
4184 accesing the first immediate word of a smob as a SCM value, and there
4185 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
4186 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
4188 ** New way to deal with non-local exits and re-entries.
4190 There is a new set of functions that essentially do what
4191 scm_internal_dynamic_wind does, but in a way that is more convenient
4192 for C code in some situations. Here is a quick example of how to
4193 prevent a potential memory leak:
4200 scm_dynwind_begin (0);
4202 mem = scm_malloc (100);
4203 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
4205 /* MEM would leak if BAR throws an error.
4206 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
4213 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
4214 SCM_DYNWIND_END as well.
4218 For full documentation, see the node "Dynamic Wind" in the manual.
4220 ** New function scm_dynwind_free
4222 This function calls 'free' on a given pointer when a dynwind context
4223 is left. Thus the call to scm_dynwind_unwind_handler above could be
4224 replaced with simply scm_dynwind_free (mem).
4226 ** New functions scm_c_call_with_blocked_asyncs and
4227 scm_c_call_with_unblocked_asyncs
4229 Like scm_call_with_blocked_asyncs etc. but for C functions.
4231 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
4233 In addition to scm_c_call_with_blocked_asyncs you can now also use
4234 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
4235 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
4237 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
4238 SCM_REALLOW_INTS have been deprecated.
4240 They do no longer fulfill their original role of blocking signal
4241 delivery. Depending on what you want to achieve, replace a pair of
4242 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
4243 mutex, blocks asyncs, or both. See node "Critical Sections" in the
4246 ** The value 'scm_mask_ints' is no longer writable.
4248 Previously, you could set scm_mask_ints directly. This is no longer
4249 possible. Use scm_c_call_with_blocked_asyncs and
4250 scm_c_call_with_unblocked_asyncs instead.
4252 ** New way to temporarily set the current input, output or error ports
4254 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
4255 context' (see above). <foo> is one of "input", "output" or "error".
4257 ** New way to temporarily set fluids
4259 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
4260 above) to temporarily set the value of a fluid.
4262 ** New types scm_t_intmax and scm_t_uintmax.
4264 On platforms that have them, these types are identical to intmax_t and
4265 uintmax_t, respectively. On other platforms, they are identical to
4266 the largest integer types that Guile knows about.
4268 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
4270 You should not have used them.
4272 ** Many public #defines with generic names have been made private.
4274 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
4275 private or renamed with a more suitable public name.
4277 ** The macro SCM_TYP16S has been deprecated.
4279 This macro is not intended for public use.
4281 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
4283 Use scm_is_true (scm_inexact_p (...)) instead.
4285 ** The macro SCM_SLOPPY_REALP has been deprecated.
4287 Use scm_is_real instead.
4289 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
4291 Use scm_is_complex instead.
4293 ** Some preprocessor defines have been deprecated.
4295 These defines indicated whether a certain feature was present in Guile
4296 or not. Going forward, assume that the features are always present.
4298 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
4299 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
4301 The following macros have been removed completely: MEMOIZE_LOCALS,
4302 SCM_RECKLESS, SCM_CAUTIOUS.
4304 ** The preprocessor define STACK_DIRECTION has been deprecated.
4306 There should be no need to know about the stack direction for ordinary
4309 ** New function: scm_effective_version
4311 Returns the "effective" version number. This is just the normal full
4312 version string without the final micro-version number. See "Changes
4313 to the distribution" above.
4315 ** The function scm_call_with_new_thread has a new prototype.
4317 Instead of taking a list with the thunk and handler, these two
4318 arguments are now passed directly:
4320 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
4322 This is an incompatible change.
4324 ** New snarfer macro SCM_DEFINE_PUBLIC.
4326 This is like SCM_DEFINE, but also calls scm_c_export for the defined
4327 function in the init section.
4329 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
4331 ** Garbage collector rewrite.
4333 The garbage collector is cleaned up a lot, and now uses lazy
4334 sweeping. This is reflected in the output of (gc-stats); since cells
4335 are being freed when they are allocated, the cells-allocated field
4336 stays roughly constant.
4338 For malloc related triggers, the behavior is changed. It uses the same
4339 heuristic as the cell-triggered collections. It may be tuned with the
4340 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
4341 for minimum yield of malloc related triggers. The default is 40.
4342 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
4345 Debugging operations for the freelist have been deprecated, along with
4346 the C variables that control garbage collection. The environment
4347 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
4348 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
4350 For understanding the memory usage of a GUILE program, the routine
4351 gc-live-object-stats returns an alist containing the number of live
4352 objects for every type.
4355 ** The function scm_definedp has been renamed to scm_defined_p
4357 The name scm_definedp is deprecated.
4359 ** The struct scm_cell type has been renamed to scm_t_cell
4361 This is in accordance to Guile's naming scheme for types. Note that
4362 the name scm_cell is now used for a function that allocates and
4363 initializes a new cell (see below).
4365 ** New functions for memory management
4367 A new set of functions for memory management has been added since the
4368 old way (scm_must_malloc, scm_must_free, etc) was error prone and
4369 indeed, Guile itself contained some long standing bugs that could
4370 cause aborts in long running programs.
4372 The new functions are more symmetrical and do not need cooperation
4373 from smob free routines, among other improvements.
4375 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
4376 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
4377 scm_gc_free, scm_gc_register_collectable_memory, and
4378 scm_gc_unregister_collectable_memory. Refer to the manual for more
4379 details and for upgrading instructions.
4381 The old functions for memory management have been deprecated. They
4382 are: scm_must_malloc, scm_must_realloc, scm_must_free,
4383 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
4385 ** Declarations of exported features are marked with SCM_API.
4387 Every declaration of a feature that belongs to the exported Guile API
4388 has been marked by adding the macro "SCM_API" to the start of the
4389 declaration. This macro can expand into different things, the most
4390 common of which is just "extern" for Unix platforms. On Win32, it can
4391 be used to control which symbols are exported from a DLL.
4393 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
4394 will expand into "__declspec (dllimport) extern", which is needed for
4395 linking to the Guile DLL in Windows.
4397 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
4398 SCM_SRFI4_IMPORT, for the corresponding libraries.
4400 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
4402 Use the new functions scm_cell and scm_double_cell instead. The old
4403 macros had problems because with them allocation and initialization
4404 was separated and the GC could sometimes observe half initialized
4405 cells. Only careful coding by the user of SCM_NEWCELL and
4406 SCM_NEWCELL2 could make this safe and efficient.
4408 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
4410 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
4413 ** SRCBRKP has been deprecated.
4415 Use scm_c_source_property_breakpoint_p instead.
4417 ** Deprecated: scm_makmacro
4419 Change your code to use either scm_makmmacro or to define macros in
4420 Scheme, using 'define-macro'.
4422 ** New function scm_c_port_for_each.
4424 This function is like scm_port_for_each but takes a pointer to a C
4425 function as the callback instead of a SCM value.
4427 ** The names scm_internal_select, scm_thread_sleep, and
4428 scm_thread_usleep have been discouraged.
4430 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
4432 ** The GC can no longer be blocked.
4434 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
4435 The GC can now run (partially) concurrently with other code and thus
4436 blocking it is not well defined.
4438 ** Many definitions have been removed that were previously deprecated.
4440 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
4441 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
4442 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
4443 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
4444 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
4445 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
4446 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
4447 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
4448 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
4449 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
4450 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
4451 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
4452 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
4453 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
4454 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
4455 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
4456 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
4457 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
4458 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
4459 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
4460 scm_vector_set_length_x, scm_contregs, scm_debug_info,
4461 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
4462 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
4463 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
4464 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
4465 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
4466 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
4467 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
4468 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
4469 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
4470 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
4472 * Changes to bundled modules
4476 Using the (ice-9 debug) module no longer automatically switches Guile
4477 to use the debugging evaluator. If you want to switch to the
4478 debugging evaluator (which is needed for backtrace information if you
4479 hit an error), please add an explicit "(debug-enable 'debug)" to your
4480 code just after the code to use (ice-9 debug).
4483 Changes since Guile 1.4:
4485 * Changes to the distribution
4487 ** A top-level TODO file is included.
4489 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
4491 Guile now always uses three numbers to represent the version,
4492 i.e. "1.6.5". The first number, 1, is the major version number, the
4493 second number, 6, is the minor version number, and the third number,
4494 5, is the micro version number. Changes in major version number
4495 indicate major changes in Guile.
4497 Minor version numbers that are even denote stable releases, and odd
4498 minor version numbers denote development versions (which may be
4499 unstable). The micro version number indicates a minor sub-revision of
4500 a given MAJOR.MINOR release.
4502 In keeping with the new scheme, (minor-version) and scm_minor_version
4503 no longer return everything but the major version number. They now
4504 just return the minor version number. Two new functions
4505 (micro-version) and scm_micro_version have been added to report the
4506 micro version number.
4508 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
4510 ** New preprocessor definitions are available for checking versions.
4512 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
4513 SCM_MICRO_VERSION to the appropriate integer values.
4515 ** Guile now actively warns about deprecated features.
4517 The new configure option `--enable-deprecated=LEVEL' and the
4518 environment variable GUILE_WARN_DEPRECATED control this mechanism.
4519 See INSTALL and README for more information.
4521 ** Guile is much more likely to work on 64-bit architectures.
4523 Guile now compiles and passes "make check" with only two UNRESOLVED GC
4524 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
4525 for the use of a test machine, and thanks to Stefan Jahn for ia64
4528 ** New functions: setitimer and getitimer.
4530 These implement a fairly direct interface to the libc functions of the
4533 ** The #. reader extension is now disabled by default.
4535 For safety reasons, #. evaluation is disabled by default. To
4536 re-enable it, set the fluid read-eval? to #t. For example:
4538 (fluid-set! read-eval? #t)
4540 but make sure you realize the potential security risks involved. With
4541 read-eval? enabled, reading a data file from an untrusted source can
4544 ** New SRFI modules have been added:
4546 SRFI-0 `cond-expand' is now supported in Guile, without requiring
4549 (srfi srfi-1) is a library containing many useful pair- and list-processing
4552 (srfi srfi-2) exports and-let*.
4554 (srfi srfi-4) implements homogeneous numeric vector datatypes.
4556 (srfi srfi-6) is a dummy module for now, since guile already provides
4557 all of the srfi-6 procedures by default: open-input-string,
4558 open-output-string, get-output-string.
4560 (srfi srfi-8) exports receive.
4562 (srfi srfi-9) exports define-record-type.
4564 (srfi srfi-10) exports define-reader-ctor and implements the reader
4567 (srfi srfi-11) exports let-values and let*-values.
4569 (srfi srfi-13) implements the SRFI String Library.
4571 (srfi srfi-14) implements the SRFI Character-Set Library.
4573 (srfi srfi-17) implements setter and getter-with-setter and redefines
4574 some accessor procedures as procedures with getters. (such as car,
4575 cdr, vector-ref etc.)
4577 (srfi srfi-19) implements the SRFI Time/Date Library.
4579 ** New scripts / "executable modules"
4581 Subdirectory "scripts" contains Scheme modules that are packaged to
4582 also be executable as scripts. At this time, these scripts are available:
4591 See README there for more info.
4593 These scripts can be invoked from the shell with the new program
4594 "guile-tools", which keeps track of installation directory for you.
4597 $ guile-tools display-commentary srfi/*.scm
4599 guile-tools is copied to the standard $bindir on "make install".
4601 ** New module (ice-9 stack-catch):
4603 stack-catch is like catch, but saves the current state of the stack in
4604 the fluid the-last-stack. This fluid can be useful when using the
4605 debugger and when re-throwing an error.
4607 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
4609 This has been done to prevent problems on lesser operating systems
4610 that can't tolerate `*'s in file names. The exported macro continues
4611 to be named `and-let*', of course.
4613 On systems that support it, there is also a compatibility module named
4614 (ice-9 and-let*). It will go away in the next release.
4616 ** New modules (oop goops) etc.:
4619 (oop goops describe)
4621 (oop goops active-slot)
4622 (oop goops composite-slot)
4624 The Guile Object Oriented Programming System (GOOPS) has been
4625 integrated into Guile. For further information, consult the GOOPS
4626 manual and tutorial in the `doc' directory.
4628 ** New module (ice-9 rdelim).
4630 This exports the following procedures which were previously defined
4631 in the default environment:
4633 read-line read-line! read-delimited read-delimited! %read-delimited!
4634 %read-line write-line
4636 For backwards compatibility the definitions are still imported into the
4637 default environment in this version of Guile. However you should add:
4639 (use-modules (ice-9 rdelim))
4641 to any program which uses the definitions, since this may change in
4644 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
4645 can be used for similar functionality.
4647 ** New module (ice-9 rw)
4649 This is a subset of the (scsh rw) module from guile-scsh. Currently
4650 it defines two procedures:
4652 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
4654 Read characters from a port or file descriptor into a string STR.
4655 A port must have an underlying file descriptor -- a so-called
4656 fport. This procedure is scsh-compatible and can efficiently read
4659 *** New function: write-string/partial str [port_or_fdes [start [end]]]
4661 Write characters from a string STR to a port or file descriptor.
4662 A port must have an underlying file descriptor -- a so-called
4663 fport. This procedure is mostly compatible and can efficiently
4664 write large strings.
4666 ** New module (ice-9 match)
4668 This module includes Andrew K. Wright's pattern matcher. See
4669 ice-9/match.scm for brief description or
4671 http://www.star-lab.com/wright/code.html
4673 for complete documentation.
4675 ** New module (ice-9 buffered-input)
4677 This module provides procedures to construct an input port from an
4678 underlying source of input that reads and returns its input in chunks.
4679 The underlying input source is a Scheme procedure, specified by the
4680 caller, which the port invokes whenever it needs more input.
4682 This is useful when building an input port whose back end is Readline
4683 or a UI element such as the GtkEntry widget.
4687 The reference and tutorial documentation that was previously
4688 distributed separately, as `guile-doc', is now included in the core
4689 Guile distribution. The documentation consists of the following
4692 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
4695 - The Guile Reference Manual (guile.texi) contains (or is intended to
4696 contain) reference documentation on all aspects of Guile.
4698 - The GOOPS Manual (goops.texi) contains both tutorial-style and
4699 reference documentation for using GOOPS, Guile's Object Oriented
4702 - The Revised^5 Report on the Algorithmic Language Scheme
4705 See the README file in the `doc' directory for more details.
4707 ** There are a couple of examples in the examples/ directory now.
4709 * Changes to the stand-alone interpreter
4711 ** New command line option `--use-srfi'
4713 Using this option, SRFI modules can be loaded on startup and be
4714 available right from the beginning. This makes programming portable
4715 Scheme programs easier.
4717 The option `--use-srfi' expects a comma-separated list of numbers,
4718 each representing a SRFI number to be loaded into the interpreter
4719 before starting evaluating a script file or the REPL. Additionally,
4720 the feature identifier for the loaded SRFIs is recognized by
4721 `cond-expand' when using this option.
4724 $ guile --use-srfi=8,13
4725 guile> (receive (x z) (values 1 2) (+ 1 2))
4727 guile> (string-pad "bla" 20)
4730 ** Guile now always starts up in the `(guile-user)' module.
4732 Previously, scripts executed via the `-s' option would run in the
4733 `(guile)' module and the repl would run in the `(guile-user)' module.
4734 Now every user action takes place in the `(guile-user)' module by
4737 * Changes to Scheme functions and syntax
4739 ** Character classifiers work for non-ASCII characters.
4741 The predicates `char-alphabetic?', `char-numeric?',
4742 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
4743 no longer check whether their arguments are ASCII characters.
4744 Previously, a character would only be considered alphabetic when it
4745 was also ASCII, for example.
4747 ** Previously deprecated Scheme functions have been removed:
4749 tag - no replacement.
4750 fseek - replaced by seek.
4751 list* - replaced by cons*.
4753 ** It's now possible to create modules with controlled environments
4757 (use-modules (ice-9 safe))
4758 (define m (make-safe-module))
4759 ;;; m will now be a module containing only a safe subset of R5RS
4760 (eval '(+ 1 2) m) --> 3
4761 (eval 'load m) --> ERROR: Unbound variable: load
4763 ** Evaluation of "()", the empty list, is now an error.
4765 Previously, the expression "()" evaluated to the empty list. This has
4766 been changed to signal a "missing expression" error. The correct way
4767 to write the empty list as a literal constant is to use quote: "'()".
4769 ** New concept of `Guile Extensions'.
4771 A Guile Extension is just a ordinary shared library that can be linked
4772 at run-time. We found it advantageous to give this simple concept a
4773 dedicated name to distinguish the issues related to shared libraries
4774 from the issues related to the module system.
4776 *** New function: load-extension
4778 Executing (load-extension lib init) is mostly equivalent to
4780 (dynamic-call init (dynamic-link lib))
4782 except when scm_register_extension has been called previously.
4783 Whenever appropriate, you should use `load-extension' instead of
4784 dynamic-link and dynamic-call.
4786 *** New C function: scm_c_register_extension
4788 This function registers a initialization function for use by
4789 `load-extension'. Use it when you don't want specific extensions to
4790 be loaded as shared libraries (for example on platforms that don't
4791 support dynamic linking).
4793 ** Auto-loading of compiled-code modules is deprecated.
4795 Guile used to be able to automatically find and link a shared
4796 library to satisfy requests for a module. For example, the module
4797 `(foo bar)' could be implemented by placing a shared library named
4798 "foo/libbar.so" (or with a different extension) in a directory on the
4801 This has been found to be too tricky, and is no longer supported. The
4802 shared libraries are now called "extensions". You should now write a
4803 small Scheme file that calls `load-extension' to load the shared
4804 library and initialize it explicitly.
4806 The shared libraries themselves should be installed in the usual
4807 places for shared libraries, with names like "libguile-foo-bar".
4809 For example, place this into a file "foo/bar.scm"
4811 (define-module (foo bar))
4813 (load-extension "libguile-foo-bar" "foobar_init")
4815 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
4817 `eval' is now R5RS, that is it takes two arguments.
4818 The second argument is an environment specifier, i.e. either
4820 (scheme-report-environment 5)
4821 (null-environment 5)
4822 (interaction-environment)
4828 ** The module system has been made more disciplined.
4830 The function `eval' will save and restore the current module around
4831 the evaluation of the specified expression. While this expression is
4832 evaluated, `(current-module)' will now return the right module, which
4833 is the module specified as the second argument to `eval'.
4835 A consequence of this change is that `eval' is not particularly
4836 useful when you want allow the evaluated code to change what module is
4837 designated as the current module and have this change persist from one
4838 call to `eval' to the next. The read-eval-print-loop is an example
4839 where `eval' is now inadequate. To compensate, there is a new
4840 function `primitive-eval' that does not take a module specifier and
4841 that does not save/restore the current module. You should use this
4842 function together with `set-current-module', `current-module', etc
4843 when you want to have more control over the state that is carried from
4844 one eval to the next.
4846 Additionally, it has been made sure that forms that are evaluated at
4847 the top level are always evaluated with respect to the current module.
4848 Previously, subforms of top-level forms such as `begin', `case',
4849 etc. did not respect changes to the current module although these
4850 subforms are at the top-level as well.
4852 To prevent strange behavior, the forms `define-module',
4853 `use-modules', `use-syntax', and `export' have been restricted to only
4854 work on the top level. The forms `define-public' and
4855 `defmacro-public' only export the new binding on the top level. They
4856 behave just like `define' and `defmacro', respectively, when they are
4857 used in a lexical environment.
4859 Also, `export' will no longer silently re-export bindings imported
4860 from a used module. It will emit a `deprecation' warning and will
4861 cease to perform any re-export in the next version. If you actually
4862 want to re-export bindings, use the new `re-export' in place of
4863 `export'. The new `re-export' will not make copies of variables when
4864 rexporting them, as `export' did wrongly.
4866 ** Module system now allows selection and renaming of imported bindings
4868 Previously, when using `use-modules' or the `#:use-module' clause in
4869 the `define-module' form, all the bindings (association of symbols to
4870 values) for imported modules were added to the "current module" on an
4871 as-is basis. This has been changed to allow finer control through two
4872 new facilities: selection and renaming.
4874 You can now select which of the imported module's bindings are to be
4875 visible in the current module by using the `:select' clause. This
4876 clause also can be used to rename individual bindings. For example:
4878 ;; import all bindings no questions asked
4879 (use-modules (ice-9 common-list))
4881 ;; import four bindings, renaming two of them;
4882 ;; the current module sees: every some zonk-y zonk-n
4883 (use-modules ((ice-9 common-list)
4885 (remove-if . zonk-y)
4886 (remove-if-not . zonk-n))))
4888 You can also programmatically rename all selected bindings using the
4889 `:renamer' clause, which specifies a proc that takes a symbol and
4890 returns another symbol. Because it is common practice to use a prefix,
4891 we now provide the convenience procedure `symbol-prefix-proc'. For
4894 ;; import four bindings, renaming two of them specifically,
4895 ;; and all four w/ prefix "CL:";
4896 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
4897 (use-modules ((ice-9 common-list)
4899 (remove-if . zonk-y)
4900 (remove-if-not . zonk-n))
4901 :renamer (symbol-prefix-proc 'CL:)))
4903 ;; import four bindings, renaming two of them specifically,
4904 ;; and all four by upcasing.
4905 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
4906 (define (upcase-symbol sym)
4907 (string->symbol (string-upcase (symbol->string sym))))
4909 (use-modules ((ice-9 common-list)
4911 (remove-if . zonk-y)
4912 (remove-if-not . zonk-n))
4913 :renamer upcase-symbol))
4915 Note that programmatic renaming is done *after* individual renaming.
4916 Also, the above examples show `use-modules', but the same facilities are
4917 available for the `#:use-module' clause of `define-module'.
4919 See manual for more info.
4921 ** The semantics of guardians have changed.
4923 The changes are for the most part compatible. An important criterion
4924 was to keep the typical usage of guardians as simple as before, but to
4925 make the semantics safer and (as a result) more useful.
4927 *** All objects returned from guardians are now properly alive.
4929 It is now guaranteed that any object referenced by an object returned
4930 from a guardian is alive. It's now impossible for a guardian to
4931 return a "contained" object before its "containing" object.
4933 One incompatible (but probably not very important) change resulting
4934 from this is that it is no longer possible to guard objects that
4935 indirectly reference themselves (i.e. are parts of cycles). If you do
4936 so accidentally, you'll get a warning.
4938 *** There are now two types of guardians: greedy and sharing.
4940 If you call (make-guardian #t) or just (make-guardian), you'll get a
4941 greedy guardian, and for (make-guardian #f) a sharing guardian.
4943 Greedy guardians are the default because they are more "defensive".
4944 You can only greedily guard an object once. If you guard an object
4945 more than once, once in a greedy guardian and the rest of times in
4946 sharing guardians, then it is guaranteed that the object won't be
4947 returned from sharing guardians as long as it is greedily guarded
4950 Guardians returned by calls to `make-guardian' can now take one more
4951 optional parameter, which says whether to throw an error in case an
4952 attempt is made to greedily guard an object that is already greedily
4953 guarded. The default is true, i.e. throw an error. If the parameter
4954 is false, the guardian invocation returns #t if guarding was
4955 successful and #f if it wasn't.
4957 Also, since greedy guarding is, in effect, a side-effecting operation
4958 on objects, a new function is introduced: `destroy-guardian!'.
4959 Invoking this function on a guardian renders it unoperative and, if
4960 the guardian is greedy, clears the "greedily guarded" property of the
4961 objects that were guarded by it, thus undoing the side effect.
4963 Note that all this hair is hardly very important, since guardian
4964 objects are usually permanent.
4966 ** Continuations created by call-with-current-continuation now accept
4967 any number of arguments, as required by R5RS.
4969 ** New function `issue-deprecation-warning'
4971 This function is used to display the deprecation messages that are
4972 controlled by GUILE_WARN_DEPRECATION as explained in the README.
4975 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
4979 ;; `id' is deprecated. Use `identity' instead.
4984 ** New syntax `begin-deprecated'
4986 When deprecated features are included (as determined by the configure
4987 option --enable-deprecated), `begin-deprecated' is identical to
4988 `begin'. When deprecated features are excluded, it always evaluates
4989 to `#f', ignoring the body forms.
4991 ** New function `make-object-property'
4993 This function returns a new `procedure with setter' P that can be used
4994 to attach a property to objects. When calling P as
4998 where `obj' is any kind of object, it attaches `val' to `obj' in such
4999 a way that it can be retrieved by calling P as
5003 This function will replace procedure properties, symbol properties and
5004 source properties eventually.
5006 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
5008 Instead of #&optional, #&key, etc you should now use #:optional,
5009 #:key, etc. Since #:optional is a keyword, you can write it as just
5010 :optional when (read-set! keywords 'prefix) is active.
5012 The old reader syntax `#&' is still supported, but deprecated. It
5013 will be removed in the next release.
5015 ** New define-module option: pure
5017 Tells the module system not to include any bindings from the root
5022 (define-module (totally-empty-module)
5025 ** New define-module option: export NAME1 ...
5027 Export names NAME1 ...
5029 This option is required if you want to be able to export bindings from
5030 a module which doesn't import one of `define-public' or `export'.
5034 (define-module (foo)
5036 :use-module (ice-9 r5rs)
5039 ;;; Note that we're pure R5RS below this point!
5044 ** New function: object->string OBJ
5046 Return a Scheme string obtained by printing a given object.
5048 ** New function: port? X
5050 Returns a boolean indicating whether X is a port. Equivalent to
5051 `(or (input-port? X) (output-port? X))'.
5053 ** New function: file-port?
5055 Determines whether a given object is a port that is related to a file.
5057 ** New function: port-for-each proc
5059 Apply PROC to each port in the Guile port table in turn. The return
5060 value is unspecified. More specifically, PROC is applied exactly once
5061 to every port that exists in the system at the time PORT-FOR-EACH is
5062 invoked. Changes to the port table while PORT-FOR-EACH is running
5063 have no effect as far as PORT-FOR-EACH is concerned.
5065 ** New function: dup2 oldfd newfd
5067 A simple wrapper for the `dup2' system call. Copies the file
5068 descriptor OLDFD to descriptor number NEWFD, replacing the
5069 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
5070 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
5071 to move away ports which are using NEWFD. The return value is
5074 ** New function: close-fdes fd
5076 A simple wrapper for the `close' system call. Close file
5077 descriptor FD, which must be an integer. Unlike close (*note
5078 close: Ports and File Descriptors.), the file descriptor will be
5079 closed even if a port is using it. The return value is
5082 ** New function: crypt password salt
5084 Encrypts `password' using the standard unix password encryption
5087 ** New function: chroot path
5089 Change the root directory of the running process to `path'.
5091 ** New functions: getlogin, cuserid
5093 Return the login name or the user name of the current effective user
5096 ** New functions: getpriority which who, setpriority which who prio
5098 Get or set the priority of the running process.
5100 ** New function: getpass prompt
5102 Read a password from the terminal, first displaying `prompt' and
5105 ** New function: flock file operation
5107 Set/remove an advisory shared or exclusive lock on `file'.
5109 ** New functions: sethostname name, gethostname
5111 Set or get the hostname of the machine the current process is running
5114 ** New function: mkstemp! tmpl
5116 mkstemp creates a new unique file in the file system and returns a
5117 new buffered port open for reading and writing to the file. TMPL
5118 is a string specifying where the file should be created: it must
5119 end with `XXXXXX' and will be changed in place to return the name
5120 of the temporary file.
5122 ** New function: open-input-string string
5124 Return an input string port which delivers the characters from
5125 `string'. This procedure, together with `open-output-string' and
5126 `get-output-string' implements SRFI-6.
5128 ** New function: open-output-string
5130 Return an output string port which collects all data written to it.
5131 The data can then be retrieved by `get-output-string'.
5133 ** New function: get-output-string
5135 Return the contents of an output string port.
5137 ** New function: identity
5139 Return the argument.
5141 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
5142 are represented in Scheme as integers with normal host byte ordering.
5144 ** New function: inet-pton family address
5146 Convert a printable string network address into an integer. Note that
5147 unlike the C version of this function, the result is an integer with
5148 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
5151 (inet-pton AF_INET "127.0.0.1") => 2130706433
5152 (inet-pton AF_INET6 "::1") => 1
5154 ** New function: inet-ntop family address
5156 Convert an integer network address into a printable string. Note that
5157 unlike the C version of this function, the input is an integer with
5158 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
5161 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
5162 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
5163 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
5167 Use `identity' instead.
5173 ** Deprecated: return-it
5177 ** Deprecated: string-character-length
5179 Use `string-length' instead.
5181 ** Deprecated: flags
5183 Use `logior' instead.
5185 ** Deprecated: close-all-ports-except.
5187 This was intended for closing ports in a child process after a fork,
5188 but it has the undesirable side effect of flushing buffers.
5189 port-for-each is more flexible.
5191 ** The (ice-9 popen) module now attempts to set up file descriptors in
5192 the child process from the current Scheme ports, instead of using the
5193 current values of file descriptors 0, 1, and 2 in the parent process.
5195 ** Removed function: builtin-weak-bindings
5197 There is no such concept as a weak binding any more.
5199 ** Removed constants: bignum-radix, scm-line-incrementors
5201 ** define-method: New syntax mandatory.
5203 The new method syntax is now mandatory:
5205 (define-method (NAME ARG-SPEC ...) BODY ...)
5206 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
5208 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
5209 REST-ARG ::= ARG-NAME
5211 If you have old code using the old syntax, import
5212 (oop goops old-define-method) before (oop goops) as in:
5214 (use-modules (oop goops old-define-method) (oop goops))
5216 ** Deprecated function: builtin-variable
5217 Removed function: builtin-bindings
5219 There is no longer a distinction between builtin or other variables.
5220 Use module system operations for all variables.
5222 ** Lazy-catch handlers are no longer allowed to return.
5224 That is, a call to `throw', `error', etc is now guaranteed to not
5227 ** Bugfixes for (ice-9 getopt-long)
5229 This module is now tested using test-suite/tests/getopt-long.test.
5230 The following bugs have been fixed:
5232 *** Parsing for options that are specified to have `optional' args now checks
5233 if the next element is an option instead of unconditionally taking it as the
5236 *** An error is now thrown for `--opt=val' when the option description
5237 does not specify `(value #t)' or `(value optional)'. This condition used to
5238 be accepted w/o error, contrary to the documentation.
5240 *** The error message for unrecognized options is now more informative.
5241 It used to be "not a record", an artifact of the implementation.
5243 *** The error message for `--opt' terminating the arg list (no value), when
5244 `(value #t)' is specified, is now more informative. It used to be "not enough
5247 *** "Clumped" single-char args now preserve trailing string, use it as arg.
5248 The expansion used to be like so:
5250 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
5252 Note that the "5d" is dropped. Now it is like so:
5254 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
5256 This enables single-char options to have adjoining arguments as long as their
5257 constituent characters are not potential single-char options.
5259 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
5261 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
5262 property `arglist', which can be retrieved by `arity'. The result is that
5263 `arity' can give more detailed information than before:
5267 guile> (use-modules (ice-9 optargs))
5268 guile> (define* (foo #:optional a b c) a)
5270 0 or more arguments in `lambda*:G0'.
5275 3 optional arguments: `a', `b' and `c'.
5276 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
5278 2 required arguments: `a' and `b', 2 keyword arguments: `c'
5279 and `d', other keywords allowed.
5280 guile> (define* (baz a b #:optional c #:rest r) a)
5282 2 required arguments: `a' and `b', 1 optional argument: `c',
5285 * Changes to the C interface
5287 ** Types have been renamed from scm_*_t to scm_t_*.
5289 This has been done for POSIX sake. It reserves identifiers ending
5290 with "_t". What a concept.
5292 The old names are still available with status `deprecated'.
5294 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
5296 ** Deprecated features have been removed.
5300 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
5301 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
5303 *** C Functions removed
5305 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
5306 scm_fseek - replaced by scm_seek.
5307 gc-thunk - replaced by after-gc-hook.
5308 gh_int2scmb - replaced by gh_bool2scm.
5309 scm_tc_dblr - replaced by scm_tc16_real.
5310 scm_tc_dblc - replaced by scm_tc16_complex.
5311 scm_list_star - replaced by scm_cons_star.
5313 ** Deprecated: scm_makfromstr
5315 Use scm_mem2string instead.
5317 ** Deprecated: scm_make_shared_substring
5319 Explicit shared substrings will disappear from Guile.
5321 Instead, "normal" strings will be implemented using sharing
5322 internally, combined with a copy-on-write strategy.
5324 ** Deprecated: scm_read_only_string_p
5326 The concept of read-only strings will disappear in next release of
5329 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
5331 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
5333 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
5335 Call a procedure with the indicated number of arguments. See "Fly
5336 Evaluation" in the manual.
5338 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
5340 Call a procedure with the indicated number of arguments and a list of
5341 further arguments. See "Fly Evaluation" in the manual.
5343 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
5345 Create a list of the given number of elements. See "List
5346 Constructors" in the manual.
5348 ** Renamed function: scm_listify has been replaced by scm_list_n.
5350 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
5351 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
5353 Use functions scm_list_N instead.
5355 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
5357 Used by an application to read arbitrary number of bytes from a port.
5358 Same semantics as libc read, except that scm_c_read only returns less
5359 than SIZE bytes if at end-of-file.
5361 Warning: Doesn't update port line and column counts!
5363 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
5365 Used by an application to write arbitrary number of bytes to an SCM
5366 port. Similar semantics as libc write. However, unlike libc
5367 write, scm_c_write writes the requested number of bytes and has no
5370 Warning: Doesn't update port line and column counts!
5372 ** New function: scm_init_guile ()
5374 In contrast to scm_boot_guile, scm_init_guile will return normally
5375 after initializing Guile. It is not available on all systems, tho.
5377 ** New functions: scm_str2symbol, scm_mem2symbol
5379 The function scm_str2symbol takes a const char* pointing to a zero-terminated
5380 field of characters and creates a scheme symbol object from that C string.
5381 The function scm_mem2symbol takes a const char* and a number of characters and
5382 creates a symbol from the characters in that memory area.
5384 ** New functions: scm_primitive_make_property
5385 scm_primitive_property_ref
5386 scm_primitive_property_set_x
5387 scm_primitive_property_del_x
5389 These functions implement a new way to deal with object properties.
5390 See libguile/properties.c for their documentation.
5392 ** New function: scm_done_free (long size)
5394 This function is the inverse of scm_done_malloc. Use it to report the
5395 amount of smob memory you free. The previous method, which involved
5396 calling scm_done_malloc with negative argument, was somewhat
5397 unintuitive (and is still available, of course).
5399 ** New function: scm_c_memq (SCM obj, SCM list)
5401 This function provides a fast C level alternative for scm_memq for the case
5402 that the list parameter is known to be a proper list. The function is a
5403 replacement for scm_sloppy_memq, but is stricter in its requirements on its
5404 list input parameter, since for anything else but a proper list the function's
5405 behaviour is undefined - it may even crash or loop endlessly. Further, for
5406 the case that the object is not found in the list, scm_c_memq returns #f which
5407 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
5409 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
5410 scm_remember_upto_here
5412 These functions replace the function scm_remember.
5414 ** Deprecated function: scm_remember
5416 Use one of the new functions scm_remember_upto_here_1,
5417 scm_remember_upto_here_2 or scm_remember_upto_here instead.
5419 ** New function: scm_allocate_string
5421 This function replaces the function scm_makstr.
5423 ** Deprecated function: scm_makstr
5425 Use the new function scm_allocate_string instead.
5427 ** New global variable scm_gc_running_p introduced.
5429 Use this variable to find out if garbage collection is being executed. Up to
5430 now applications have used scm_gc_heap_lock to test if garbage collection was
5431 running, which also works because of the fact that up to know only the garbage
5432 collector has set this variable. But, this is an implementation detail that
5433 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
5434 of this variable is (and has been) not fully safe anyway.
5436 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
5438 Use these instead of SCM_LENGTH_MAX.
5440 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
5441 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
5442 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
5444 Use these instead of SCM_LENGTH.
5446 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
5447 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
5448 SCM_SET_BITVECTOR_LENGTH
5450 Use these instead of SCM_SETLENGTH
5452 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
5453 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
5456 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
5459 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
5460 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
5463 Use these instead of SCM_SETCHARS.
5465 ** New macro: SCM_BITVECTOR_P
5467 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
5469 Use instead of SCM_COERCE_SUBSTR.
5471 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
5473 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
5475 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
5476 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
5477 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
5478 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
5479 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
5480 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
5481 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
5482 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
5483 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
5484 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
5485 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
5486 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
5487 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
5488 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
5489 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
5491 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
5492 Use scm_memory_error instead of SCM_NALLOC.
5493 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
5494 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
5495 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
5496 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
5497 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
5498 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
5499 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
5500 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
5501 Use SCM_STRINGP instead of SCM_RWSTRINGP.
5502 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
5503 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
5504 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
5505 Use a type specific setter macro instead of SCM_SETLENGTH.
5506 Use a type specific setter macro instead of SCM_SETCHARS.
5507 Use a type specific length macro instead of SCM_LENGTH_MAX.
5508 Use SCM_GCMARKP instead of SCM_GC8MARKP.
5509 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
5510 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
5511 Use SCM_TYP16 instead of SCM_GCTYP16.
5512 Use SCM_CDR instead of SCM_GCCDR.
5513 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
5514 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
5515 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
5516 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
5517 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
5518 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
5519 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
5521 ** Removed function: scm_struct_init
5523 ** Removed variable: scm_symhash_dim
5525 ** Renamed function: scm_make_cont has been replaced by
5526 scm_make_continuation, which has a different interface.
5528 ** Deprecated function: scm_call_catching_errors
5530 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
5532 ** Deprecated function: scm_strhash
5534 Use scm_string_hash instead.
5536 ** Deprecated function: scm_vector_set_length_x
5538 Instead, create a fresh vector of the desired size and copy the contents.
5540 ** scm_gensym has changed prototype
5542 scm_gensym now only takes one argument.
5544 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
5547 There is now only a single symbol type scm_tc7_symbol.
5548 The tag scm_tc7_lvector was not used anyway.
5550 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
5552 Use scm_make_smob_type and scm_set_smob_XXX instead.
5554 ** New function scm_set_smob_apply.
5556 This can be used to set an apply function to a smob type.
5558 ** Deprecated function: scm_strprint_obj
5560 Use scm_object_to_string instead.
5562 ** Deprecated function: scm_wta
5564 Use scm_wrong_type_arg, or another appropriate error signalling function
5567 ** Explicit support for obarrays has been deprecated.
5569 Use `scm_str2symbol' and the generic hashtable functions instead.
5571 ** The concept of `vcells' has been deprecated.
5573 The data type `variable' is now used exclusively. `Vcells' have been
5574 a low-level concept so you are likely not affected by this change.
5576 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
5577 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
5579 Use scm_c_define or scm_c_lookup instead, as appropriate.
5581 *** New functions: scm_c_module_lookup, scm_c_lookup,
5582 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
5583 scm_module_define, scm_define.
5585 These functions work with variables instead of with vcells.
5587 ** New functions for creating and defining `subr's and `gsubr's.
5589 The new functions more clearly distinguish between creating a subr (or
5590 gsubr) object and adding it to the current module.
5592 These new functions are available: scm_c_make_subr, scm_c_define_subr,
5593 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
5594 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
5595 scm_c_define_gsubr_with_generic.
5597 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
5598 scm_make_subr_with_generic, scm_make_gsubr,
5599 scm_make_gsubr_with_generic.
5601 Use the new ones from above instead.
5603 ** C interface to the module system has changed.
5605 While we suggest that you avoid as many explicit module system
5606 operations from C as possible for the time being, the C interface has
5607 been made more similar to the high-level Scheme module system.
5609 *** New functions: scm_c_define_module, scm_c_use_module,
5610 scm_c_export, scm_c_resolve_module.
5612 They mostly work like their Scheme namesakes. scm_c_define_module
5613 takes a function that is called a context where the new module is
5616 *** Deprecated functions: scm_the_root_module, scm_make_module,
5617 scm_ensure_user_module, scm_load_scheme_module.
5619 Use the new functions instead.
5621 ** Renamed function: scm_internal_with_fluids becomes
5624 scm_internal_with_fluids is available as a deprecated function.
5626 ** New function: scm_c_with_fluid.
5628 Just like scm_c_with_fluids, but takes one fluid and one value instead
5631 ** Deprecated typedefs: long_long, ulong_long.
5633 They are of questionable utility and they pollute the global
5636 ** Deprecated typedef: scm_sizet
5638 It is of questionable utility now that Guile requires ANSI C, and is
5641 ** Deprecated typedefs: scm_port_rw_active, scm_port,
5642 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
5643 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
5645 Made more compliant with the naming policy by adding a _t at the end.
5647 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
5648 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
5650 With the exception of the mysterious scm_2ulong2big, they are still
5651 available under new names (scm_i_mkbig etc). These functions are not
5652 intended to be used in user code. You should avoid dealing with
5653 bignums directly, and should deal with numbers in general (which can
5656 ** Change in behavior: scm_num2long, scm_num2ulong
5658 The scm_num2[u]long functions don't any longer accept an inexact
5659 argument. This change in behavior is motivated by concordance with
5660 R5RS: It is more common that a primitive doesn't want to accept an
5661 inexact for an exact.
5663 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
5664 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
5665 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
5668 These are conversion functions between the various ANSI C integral
5669 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
5670 accept an inexact argument.
5672 ** New functions: scm_float2num, scm_double2num,
5673 scm_num2float, scm_num2double.
5675 These are conversion functions between the two ANSI C float types and
5678 ** New number validation macros:
5679 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
5683 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
5685 These are just nicer-named old scm_protect_object and
5686 scm_unprotect_object.
5688 ** Deprecated functions: scm_protect_object, scm_unprotect_object
5690 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
5692 These functions can be used to register pointers to locations that
5695 ** Deprecated function: scm_create_hook.
5697 Its sins are: misleading name, non-modularity and lack of general
5701 Changes since Guile 1.3.4:
5703 * Changes to the distribution
5705 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
5707 We've changed the way we handle generated files in the Guile source
5708 repository. As a result, the procedure for building trees obtained
5709 from the nightly FTP snapshots or via CVS has changed:
5710 - You must have appropriate versions of autoconf, automake, and
5711 libtool installed on your system. See README for info on how to
5712 obtain these programs.
5713 - Before configuring the tree, you must first run the script
5714 `autogen.sh' at the top of the source tree.
5716 The Guile repository used to contain not only source files, written by
5717 humans, but also some generated files, like configure scripts and
5718 Makefile.in files. Even though the contents of these files could be
5719 derived mechanically from other files present, we thought it would
5720 make the tree easier to build if we checked them into CVS.
5722 However, this approach means that minor differences between
5723 developer's installed tools and habits affected the whole team.
5724 So we have removed the generated files from the repository, and
5725 added the autogen.sh script, which will reconstruct them
5729 ** configure now has experimental options to remove support for certain
5732 --disable-arrays omit array and uniform array support
5733 --disable-posix omit posix interfaces
5734 --disable-networking omit networking interfaces
5735 --disable-regex omit regular expression interfaces
5737 These are likely to become separate modules some day.
5739 ** New configure option --enable-debug-freelist
5741 This enables a debugging version of SCM_NEWCELL(), and also registers
5742 an extra primitive, the setter `gc-set-debug-check-freelist!'.
5744 Configure with the --enable-debug-freelist option to enable
5745 the gc-set-debug-check-freelist! primitive, and then use:
5747 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
5748 (gc-set-debug-check-freelist! #f) # turn off checking
5750 Checking of the freelist forces a traversal of the freelist and
5751 a garbage collection before each allocation of a cell. This can
5752 slow down the interpreter dramatically, so the setter should be used to
5753 turn on this extra processing only when necessary.
5755 ** New configure option --enable-debug-malloc
5757 Include code for debugging of calls to scm_must_malloc/realloc/free.
5761 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
5762 2. objects reallocated by scm_must_realloc has been allocated by
5764 3. reallocated objects are reallocated with the same what string
5766 But, most importantly, it records the number of allocated objects of
5767 each kind. This is useful when searching for memory leaks.
5769 A Guile compiled with this option provides the primitive
5770 `malloc-stats' which returns an alist with pairs of kind and the
5771 number of objects of that kind.
5773 ** All includes are now referenced relative to the root directory
5775 Since some users have had problems with mixups between Guile and
5776 system headers, we have decided to always refer to Guile headers via
5777 their parent directories. This essentially creates a "private name
5778 space" for Guile headers. This means that the compiler only is given
5779 -I options for the root build and root source directory.
5781 ** Header files kw.h and genio.h have been removed.
5783 ** The module (ice-9 getopt-gnu-style) has been removed.
5785 ** New module (ice-9 documentation)
5787 Implements the interface to documentation strings associated with
5790 ** New module (ice-9 time)
5792 Provides a macro `time', which displays execution time of a given form.
5794 ** New module (ice-9 history)
5796 Loading this module enables value history in the repl.
5798 * Changes to the stand-alone interpreter
5800 ** New command line option --debug
5802 Start Guile with debugging evaluator and backtraces enabled.
5804 This is useful when debugging your .guile init file or scripts.
5806 ** New help facility
5808 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
5809 (help REGEXP) ditto for objects with names matching REGEXP (a string)
5810 (help 'NAME) gives documentation for NAME, even if it is not an object
5811 (help ,EXPR) gives documentation for object returned by EXPR
5812 (help (my module)) gives module commentary for `(my module)'
5813 (help) gives this text
5815 `help' searches among bindings exported from loaded modules, while
5816 `apropos' searches among bindings visible from the "current" module.
5818 Examples: (help help)
5820 (help "output-string")
5822 ** `help' and `apropos' now prints full module names
5824 ** Dynamic linking now uses libltdl from the libtool package.
5826 The old system dependent code for doing dynamic linking has been
5827 replaced with calls to the libltdl functions which do all the hairy
5830 The major improvement is that you can now directly pass libtool
5831 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
5832 will be able to do the best shared library job you can get, via
5835 The way dynamic libraries are found has changed and is not really
5836 portable across platforms, probably. It is therefore recommended to
5837 use absolute filenames when possible.
5839 If you pass a filename without an extension to `dynamic-link', it will
5840 try a few appropriate ones. Thus, the most platform ignorant way is
5841 to specify a name like "libfoo", without any directories and
5844 ** Guile COOP threads are now compatible with LinuxThreads
5846 Previously, COOP threading wasn't possible in applications linked with
5847 Linux POSIX threads due to their use of the stack pointer to find the
5848 thread context. This has now been fixed with a workaround which uses
5849 the pthreads to allocate the stack.
5851 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
5853 ** Positions of erring expression in scripts
5855 With version 1.3.4, the location of the erring expression in Guile
5856 scipts is no longer automatically reported. (This should have been
5857 documented before the 1.3.4 release.)
5859 You can get this information by enabling recording of positions of
5860 source expressions and running the debugging evaluator. Put this at
5861 the top of your script (or in your "site" file):
5863 (read-enable 'positions)
5864 (debug-enable 'debug)
5866 ** Backtraces in scripts
5868 It is now possible to get backtraces in scripts.
5872 (debug-enable 'debug 'backtrace)
5874 at the top of the script.
5876 (The first options enables the debugging evaluator.
5877 The second enables backtraces.)
5879 ** Part of module system symbol lookup now implemented in C
5881 The eval closure of most modules is now implemented in C. Since this
5882 was one of the bottlenecks for loading speed, Guile now loads code
5883 substantially faster than before.
5885 ** Attempting to get the value of an unbound variable now produces
5886 an exception with a key of 'unbound-variable instead of 'misc-error.
5888 ** The initial default output port is now unbuffered if it's using a
5889 tty device. Previously in this situation it was line-buffered.
5891 ** New hook: after-gc-hook
5893 after-gc-hook takes over the role of gc-thunk. This hook is run at
5894 the first SCM_TICK after a GC. (Thus, the code is run at the same
5895 point during evaluation as signal handlers.)
5897 Note that this hook should be used only for diagnostic and debugging
5898 purposes. It is not certain that it will continue to be well-defined
5899 when this hook is run in the future.
5901 C programmers: Note the new C level hooks scm_before_gc_c_hook,
5902 scm_before_sweep_c_hook, scm_after_gc_c_hook.
5904 ** Improvements to garbage collector
5906 Guile 1.4 has a new policy for triggering heap allocation and
5907 determining the sizes of heap segments. It fixes a number of problems
5910 1. The new policy can handle two separate pools of cells
5911 (2-word/4-word) better. (The old policy would run wild, allocating
5912 more and more memory for certain programs.)
5914 2. The old code would sometimes allocate far too much heap so that the
5915 Guile process became gigantic. The new code avoids this.
5917 3. The old code would sometimes allocate too little so that few cells
5918 were freed at GC so that, in turn, too much time was spent in GC.
5920 4. The old code would often trigger heap allocation several times in a
5921 row. (The new scheme predicts how large the segments needs to be
5922 in order not to need further allocation.)
5924 All in all, the new GC policy will make larger applications more
5927 The new GC scheme also is prepared for POSIX threading. Threads can
5928 allocate private pools of cells ("clusters") with just a single
5929 function call. Allocation of single cells from such a cluster can
5930 then proceed without any need of inter-thread synchronization.
5932 ** New environment variables controlling GC parameters
5934 GUILE_MAX_SEGMENT_SIZE Maximal segment size
5937 Allocation of 2-word cell heaps:
5939 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
5942 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
5943 GC in percent of total heap size
5946 Allocation of 4-word cell heaps
5947 (used for real numbers and misc other objects):
5949 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
5951 (See entry "Way for application to customize GC parameters" under
5952 section "Changes to the scm_ interface" below.)
5954 ** Guile now implements reals using 4-word cells
5956 This speeds up computation with reals. (They were earlier allocated
5957 with `malloc'.) There is still some room for optimizations, however.
5959 ** Some further steps toward POSIX thread support have been taken
5961 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
5962 don't have much effect any longer, and many of them will be removed in
5966 are only handled at the top of the evaluator loop, immediately after
5967 I/O, and in scm_equalp.
5969 *** The GC can allocate thread private pools of pairs.
5971 * Changes to Scheme functions and syntax
5973 ** close-input-port and close-output-port are now R5RS
5975 These procedures have been turned into primitives and have R5RS behaviour.
5977 ** New procedure: simple-format PORT MESSAGE ARG1 ...
5979 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
5980 extended by the more sophisticated version in (ice-9 format)
5982 (simple-format port message . args)
5983 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
5984 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
5985 the escapes are replaced with corresponding members of ARGS:
5986 ~A formats using `display' and ~S formats using `write'.
5987 If DESTINATION is #t, then use the `current-output-port',
5988 if DESTINATION is #f, then return a string containing the formatted text.
5989 Does not add a trailing newline."
5991 ** string-ref: the second argument is no longer optional.
5993 ** string, list->string: no longer accept strings in their arguments,
5994 only characters, for compatibility with R5RS.
5996 ** New procedure: port-closed? PORT
5997 Returns #t if PORT is closed or #f if it is open.
5999 ** Deprecated: list*
6001 The list* functionality is now provided by cons* (SRFI-1 compliant)
6003 ** New procedure: cons* ARG1 ARG2 ... ARGn
6005 Like `list', but the last arg provides the tail of the constructed list,
6006 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
6008 Requires at least one argument. If given one argument, that argument
6009 is returned as result.
6011 This function is called `list*' in some other Schemes and in Common LISP.
6013 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
6015 ** New procedure: object-documentation OBJECT
6017 Returns the documentation string associated with OBJECT. The
6018 procedure uses a caching mechanism so that subsequent lookups are
6021 Exported by (ice-9 documentation).
6023 ** module-name now returns full names of modules
6025 Previously, only the last part of the name was returned (`session' for
6026 `(ice-9 session)'). Ex: `(ice-9 session)'.
6028 * Changes to the gh_ interface
6030 ** Deprecated: gh_int2scmb
6032 Use gh_bool2scm instead.
6034 * Changes to the scm_ interface
6036 ** Guile primitives now carry docstrings!
6038 Thanks to Greg Badros!
6040 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
6042 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
6043 macros and must contain a docstring that is extracted into foo.doc using a new
6044 guile-doc-snarf script (that uses guile-doc-snarf.awk).
6046 However, a major overhaul of these macros is scheduled for the next release of
6049 ** Guile primitives use a new technique for validation of arguments
6051 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
6052 the readability of argument checking.
6054 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
6056 ** New macros: SCM_PACK, SCM_UNPACK
6058 Compose/decompose an SCM value.
6060 The SCM type is now treated as an abstract data type and may be defined as a
6061 long, a void* or as a struct, depending on the architecture and compile time
6062 options. This makes it easier to find several types of bugs, for example when
6063 SCM values are treated as integers without conversion. Values of the SCM type
6064 should be treated as "atomic" values. These macros are used when
6065 composing/decomposing an SCM value, either because you want to access
6066 individual bits, or because you want to treat it as an integer value.
6068 E.g., in order to set bit 7 in an SCM value x, use the expression
6070 SCM_PACK (SCM_UNPACK (x) | 0x80)
6072 ** The name property of hooks is deprecated.
6073 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
6075 You can emulate this feature by using object properties.
6077 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
6078 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
6081 These macros will be removed in a future release of Guile.
6083 ** The following types, functions and macros from numbers.h are deprecated:
6084 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
6085 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
6087 ** Port internals: the rw_random variable in the scm_port structure
6088 must be set to non-zero in any random access port. In recent Guile
6089 releases it was only set for bidirectional random-access ports.
6091 ** Port internals: the seek ptob procedure is now responsible for
6092 resetting the buffers if required. The change was made so that in the
6093 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
6094 the fport and strport ptobs can avoid resetting the buffers,
6095 in particular to avoid discarding unread chars. An existing port
6096 type can be fixed by adding something like the following to the
6097 beginning of the ptob seek procedure:
6099 if (pt->rw_active == SCM_PORT_READ)
6100 scm_end_input (object);
6101 else if (pt->rw_active == SCM_PORT_WRITE)
6102 ptob->flush (object);
6104 although to actually avoid resetting the buffers and discard unread
6105 chars requires further hacking that depends on the characteristics
6108 ** Deprecated functions: scm_fseek, scm_tag
6110 These functions are no longer used and will be removed in a future version.
6112 ** The scm_sysmissing procedure is no longer used in libguile.
6113 Unless it turns out to be unexpectedly useful to somebody, it will be
6114 removed in a future version.
6116 ** The format of error message strings has changed
6118 The two C procedures: scm_display_error and scm_error, as well as the
6119 primitive `scm-error', now use scm_simple_format to do their work.
6120 This means that the message strings of all code must be updated to use
6121 ~A where %s was used before, and ~S where %S was used before.
6123 During the period when there still are a lot of old Guiles out there,
6124 you might want to support both old and new versions of Guile.
6126 There are basically two methods to achieve this. Both methods use
6129 AC_CHECK_FUNCS(scm_simple_format)
6131 in your configure.in.
6133 Method 1: Use the string concatenation features of ANSI C's
6138 #ifdef HAVE_SCM_SIMPLE_FORMAT
6144 Then represent each of your error messages using a preprocessor macro:
6146 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
6150 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
6151 (define make-message string-append)
6153 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
6155 Method 2: Use the oldfmt function found in doc/oldfmt.c.
6159 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
6164 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
6168 ** Deprecated: coop_mutex_init, coop_condition_variable_init
6170 Don't use the functions coop_mutex_init and
6171 coop_condition_variable_init. They will change.
6173 Use scm_mutex_init and scm_cond_init instead.
6175 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
6176 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
6177 COND, as `scm_cond_wait' does, but it also bounds the duration
6178 of the wait. If COND has not been signaled before time ABSTIME,
6179 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
6180 returns the error code `ETIMEDOUT'.
6182 The ABSTIME parameter specifies an absolute time, with the same
6183 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
6184 to 00:00:00 GMT, January 1, 1970.
6186 ** New function: scm_cond_broadcast (scm_cond_t *COND)
6187 `scm_cond_broadcast' restarts all the threads that are waiting
6188 on the condition variable COND. Nothing happens if no threads are
6191 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
6192 `scm_key_create' allocates a new TSD key. The key is stored in
6193 the location pointed to by KEY. There is no limit on the number
6194 of keys allocated at a given time. The value initially associated
6195 with the returned key is `NULL' in all currently executing threads.
6197 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
6198 function associated with the key. When a thread terminates,
6199 DESTR_FUNCTION is called on the value associated with the key in
6200 that thread. The DESTR_FUNCTION is not called if a key is deleted
6201 with `scm_key_delete' or a value is changed with
6202 `scm_setspecific'. The order in which destructor functions are
6203 called at thread termination time is unspecified.
6205 Destructors are not yet implemented.
6207 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
6208 `scm_setspecific' changes the value associated with KEY in the
6209 calling thread, storing the given POINTER instead.
6211 ** New function: scm_getspecific (scm_key_t KEY)
6212 `scm_getspecific' returns the value currently associated with
6213 KEY in the calling thread.
6215 ** New function: scm_key_delete (scm_key_t KEY)
6216 `scm_key_delete' deallocates a TSD key. It does not check
6217 whether non-`NULL' values are associated with that key in the
6218 currently executing threads, nor call the destructor function
6219 associated with the key.
6221 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
6223 Initialize a C level hook HOOK with associated HOOK_DATA and type
6224 TYPE. (See scm_c_hook_run ().)
6226 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
6228 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
6229 is true, add it last, otherwise first. The same FUNC can be added
6230 multiple times if FUNC_DATA differ and vice versa.
6232 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
6234 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
6235 function is only removed if both FUNC and FUNC_DATA matches.
6237 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
6239 Run hook HOOK passing DATA to the hook functions.
6241 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
6242 returned is undefined.
6244 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
6245 returns a non-NULL value. This value is returned as the result of
6246 scm_c_hook_run. If all functions return NULL, NULL is returned.
6248 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
6249 returns a NULL value, and NULL is returned. If all functions returns
6250 a non-NULL value, the last value is returned.
6252 ** New C level GC hooks
6254 Five new C level hooks has been added to the garbage collector.
6256 scm_before_gc_c_hook
6259 are run before locking and after unlocking the heap. The system is
6260 thus in a mode where evaluation can take place. (Except that
6261 scm_before_gc_c_hook must not allocate new cells.)
6263 scm_before_mark_c_hook
6264 scm_before_sweep_c_hook
6265 scm_after_sweep_c_hook
6267 are run when the heap is locked. These are intended for extension of
6268 the GC in a modular fashion. Examples are the weaks and guardians
6271 ** Way for application to customize GC parameters
6273 The application can set up other default values for the GC heap
6274 allocation parameters
6276 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
6277 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
6278 GUILE_MAX_SEGMENT_SIZE,
6282 scm_default_init_heap_size_1, scm_default_min_yield_1,
6283 scm_default_init_heap_size_2, scm_default_min_yield_2,
6284 scm_default_max_segment_size
6286 respectively before callong scm_boot_guile.
6288 (See entry "New environment variables ..." in section
6289 "Changes to the stand-alone interpreter" above.)
6291 ** scm_protect_object/scm_unprotect_object now nest
6293 This means that you can call scm_protect_object multiple times on an
6294 object and count on the object being protected until
6295 scm_unprotect_object has been call the same number of times.
6297 The functions also have better time complexity.
6299 Still, it is usually possible to structure the application in a way
6300 that you don't need to use these functions. For example, if you use a
6301 protected standard Guile list to keep track of live objects rather
6302 than some custom data type, objects will die a natural death when they
6303 are no longer needed.
6305 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
6307 Guile does not provide the float representation for inexact real numbers any
6308 more. Now, only doubles are used to represent inexact real numbers. Further,
6309 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
6310 and scm_tc16_complex, respectively.
6312 ** Removed deprecated type scm_smobfuns
6314 ** Removed deprecated function scm_newsmob
6316 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
6318 There is an ongoing discussion among the developers whether to
6319 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
6320 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
6321 until this issue has been settled.
6323 ** Removed deprecated type tag scm_tc16_kw
6325 ** Added type tag scm_tc16_keyword
6327 (This was introduced already in release 1.3.4 but was not documented
6330 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
6332 * Changes to system call interfaces:
6334 ** The "select" procedure now tests port buffers for the ability to
6335 provide input or accept output. Previously only the underlying file
6336 descriptors were checked.
6338 ** New variable PIPE_BUF: the maximum number of bytes that can be
6339 atomically written to a pipe.
6341 ** If a facility is not available on the system when Guile is
6342 compiled, the corresponding primitive procedure will not be defined.
6343 Previously it would have been defined but would throw a system-error
6344 exception if called. Exception handlers which catch this case may
6345 need minor modification: an error will be thrown with key
6346 'unbound-variable instead of 'system-error. Alternatively it's
6347 now possible to use `defined?' to check whether the facility is
6350 ** Procedures which depend on the timezone should now give the correct
6351 result on systems which cache the TZ environment variable, even if TZ
6352 is changed without calling tzset.
6354 * Changes to the networking interfaces:
6356 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
6357 long integers between network and host format. For now, it's not
6358 particularly convenient to do this kind of thing, but consider:
6360 (define write-network-long
6361 (lambda (value port)
6362 (let ((v (make-uniform-vector 1 1 0)))
6363 (uniform-vector-set! v 0 (htonl value))
6364 (uniform-vector-write v port))))
6366 (define read-network-long
6368 (let ((v (make-uniform-vector 1 1 0)))
6369 (uniform-vector-read! v port)
6370 (ntohl (uniform-vector-ref v 0)))))
6372 ** If inet-aton fails, it now throws an error with key 'misc-error
6373 instead of 'system-error, since errno is not relevant.
6375 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
6376 specific keys instead of 'system-error. The latter is inappropriate
6377 since errno will not have been set. The keys are:
6378 'host-not-found, 'try-again, 'no-recovery and 'no-data.
6380 ** sethostent, setnetent, setprotoent, setservent: now take an
6381 optional argument STAYOPEN, which specifies whether the database
6382 remains open after a database entry is accessed randomly (e.g., using
6383 gethostbyname for the hosts database.) The default is #f. Previously
6387 Changes since Guile 1.3.2:
6389 * Changes to the stand-alone interpreter
6393 An initial version of the Guile debugger written by Chris Hanson has
6394 been added. The debugger is still under development but is included
6395 in the distribution anyway since it is already quite useful.
6401 after an error to enter the debugger. Type `help' inside the debugger
6402 for a description of available commands.
6404 If you prefer to have stack frames numbered and printed in
6405 anti-chronological order and prefer up in the stack to be down on the
6406 screen as is the case in gdb, you can put
6408 (debug-enable 'backwards)
6410 in your .guile startup file. (However, this means that Guile can't
6411 use indentation to indicate stack level.)
6413 The debugger is autoloaded into Guile at the first use.
6415 ** Further enhancements to backtraces
6417 There is a new debug option `width' which controls the maximum width
6418 on the screen of printed stack frames. Fancy printing parameters
6419 ("level" and "length" as in Common LISP) are adaptively adjusted for
6420 each stack frame to give maximum information while still fitting
6421 within the bounds. If the stack frame can't be made to fit by
6422 adjusting parameters, it is simply cut off at the end. This is marked
6425 ** Some modules are now only loaded when the repl is started
6427 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
6428 regex) are now loaded into (guile-user) only if the repl has been
6429 started. The effect is that the startup time for scripts has been
6430 reduced to 30% of what it was previously.
6432 Correctly written scripts load the modules they require at the top of
6433 the file and should not be affected by this change.
6435 ** Hooks are now represented as smobs
6437 * Changes to Scheme functions and syntax
6439 ** Readline support has changed again.
6441 The old (readline-activator) module is gone. Use (ice-9 readline)
6442 instead, which now contains all readline functionality. So the code
6443 to activate readline is now
6445 (use-modules (ice-9 readline))
6448 This should work at any time, including from the guile prompt.
6450 To avoid confusion about the terms of Guile's license, please only
6451 enable readline for your personal use; please don't make it the
6452 default for others. Here is why we make this rather odd-sounding
6455 Guile is normally licensed under a weakened form of the GNU General
6456 Public License, which allows you to link code with Guile without
6457 placing that code under the GPL. This exception is important to some
6460 However, since readline is distributed under the GNU General Public
6461 License, when you link Guile with readline, either statically or
6462 dynamically, you effectively change Guile's license to the strict GPL.
6463 Whenever you link any strictly GPL'd code into Guile, uses of Guile
6464 which are normally permitted become forbidden. This is a rather
6465 non-obvious consequence of the licensing terms.
6467 So, to make sure things remain clear, please let people choose for
6468 themselves whether to link GPL'd libraries like readline with Guile.
6470 ** regexp-substitute/global has changed slightly, but incompatibly.
6472 If you include a function in the item list, the string of the match
6473 object it receives is the same string passed to
6474 regexp-substitute/global, not some suffix of that string.
6475 Correspondingly, the match's positions are relative to the entire
6476 string, not the suffix.
6478 If the regexp can match the empty string, the way matches are chosen
6479 from the string has changed. regexp-substitute/global recognizes the
6480 same set of matches that list-matches does; see below.
6482 ** New function: list-matches REGEXP STRING [FLAGS]
6484 Return a list of match objects, one for every non-overlapping, maximal
6485 match of REGEXP in STRING. The matches appear in left-to-right order.
6486 list-matches only reports matches of the empty string if there are no
6487 other matches which begin on, end at, or include the empty match's
6490 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6492 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
6494 For each match of REGEXP in STRING, apply PROC to the match object,
6495 and the last value PROC returned, or INIT for the first call. Return
6496 the last value returned by PROC. We apply PROC to the matches as they
6497 appear from left to right.
6499 This function recognizes matches according to the same criteria as
6502 Thus, you could define list-matches like this:
6504 (define (list-matches regexp string . flags)
6505 (reverse! (apply fold-matches regexp string '() cons flags)))
6507 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6511 *** New function: hook? OBJ
6513 Return #t if OBJ is a hook, otherwise #f.
6515 *** New function: make-hook-with-name NAME [ARITY]
6517 Return a hook with name NAME and arity ARITY. The default value for
6518 ARITY is 0. The only effect of NAME is that it will appear when the
6519 hook object is printed to ease debugging.
6521 *** New function: hook-empty? HOOK
6523 Return #t if HOOK doesn't contain any procedures, otherwise #f.
6525 *** New function: hook->list HOOK
6527 Return a list of the procedures that are called when run-hook is
6530 ** `map' signals an error if its argument lists are not all the same length.
6532 This is the behavior required by R5RS, so this change is really a bug
6533 fix. But it seems to affect a lot of people's code, so we're
6534 mentioning it here anyway.
6536 ** Print-state handling has been made more transparent
6538 Under certain circumstances, ports are represented as a port with an
6539 associated print state. Earlier, this pair was represented as a pair
6540 (see "Some magic has been added to the printer" below). It is now
6541 indistinguishable (almost; see `get-print-state') from a port on the
6544 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
6546 Return a new port with the associated print state PRINT-STATE.
6548 *** New function: get-print-state OUTPUT-PORT
6550 Return the print state associated with this port if it exists,
6551 otherwise return #f.
6553 *** New function: directory-stream? OBJECT
6555 Returns true iff OBJECT is a directory stream --- the sort of object
6556 returned by `opendir'.
6558 ** New function: using-readline?
6560 Return #t if readline is in use in the current repl.
6562 ** structs will be removed in 1.4
6564 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
6565 and use GOOPS objects as the fundamental record type.
6567 * Changes to the scm_ interface
6569 ** structs will be removed in 1.4
6571 The entire current struct interface (struct.c, struct.h) will be
6572 replaced in Guile 1.4. We will merge GOOPS into libguile and use
6573 GOOPS objects as the fundamental record type.
6575 ** The internal representation of subr's has changed
6577 Instead of giving a hint to the subr name, the CAR field of the subr
6578 now contains an index to a subr entry in scm_subr_table.
6580 *** New variable: scm_subr_table
6582 An array of subr entries. A subr entry contains the name, properties
6583 and documentation associated with the subr. The properties and
6584 documentation slots are not yet used.
6586 ** A new scheme for "forwarding" calls to a builtin to a generic function
6588 It is now possible to extend the functionality of some Guile
6589 primitives by letting them defer a call to a GOOPS generic function on
6590 argument mismatch. This means that there is no loss of efficiency in
6595 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
6596 (define-method + ((x <string>) (y <string>))
6597 (string-append x y))
6599 + will still be as efficient as usual in numerical calculations, but
6600 can also be used for concatenating strings.
6602 Who will be the first one to extend Guile's numerical tower to
6603 rationals? :) [OK, there a few other things to fix before this can
6604 be made in a clean way.]
6606 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
6608 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
6610 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
6612 These do the same job as SCM_PROC and SCM_PROC1, but they also define
6613 a variable GENERIC which can be used by the dispatch macros below.
6615 [This is experimental code which may change soon.]
6617 *** New macros for forwarding control to a generic on arg type error
6619 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
6621 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
6623 These correspond to the scm_wta function call, and have the same
6624 behaviour until the user has called the GOOPS primitive
6625 `enable-primitive-generic!'. After that, these macros will apply the
6626 generic function GENERIC to the argument(s) instead of calling
6629 [This is experimental code which may change soon.]
6631 *** New macros for argument testing with generic dispatch
6633 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
6635 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
6637 These correspond to the SCM_ASSERT macro, but will defer control to
6638 GENERIC on error after `enable-primitive-generic!' has been called.
6640 [This is experimental code which may change soon.]
6642 ** New function: SCM scm_eval_body (SCM body, SCM env)
6644 Evaluates the body of a special form.
6646 ** The internal representation of struct's has changed
6648 Previously, four slots were allocated for the procedure(s) of entities
6649 and operators. The motivation for this representation had to do with
6650 the structure of the evaluator, the wish to support tail-recursive
6651 generic functions, and efficiency. Since the generic function
6652 dispatch mechanism has changed, there is no longer a need for such an
6653 expensive representation, and the representation has been simplified.
6655 This should not make any difference for most users.
6657 ** GOOPS support has been cleaned up.
6659 Some code has been moved from eval.c to objects.c and code in both of
6660 these compilation units has been cleaned up and better structured.
6662 *** New functions for applying generic functions
6664 New function: SCM scm_apply_generic (GENERIC, ARGS)
6665 New function: SCM scm_call_generic_0 (GENERIC)
6666 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
6667 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
6668 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
6670 ** Deprecated function: scm_make_named_hook
6672 It is now replaced by:
6674 ** New function: SCM scm_create_hook (const char *name, int arity)
6676 Creates a hook in the same way as make-hook above but also
6677 binds a variable named NAME to it.
6679 This is the typical way of creating a hook from C code.
6681 Currently, the variable is created in the "current" module.
6682 This might change when we get the new module system.
6684 [The behaviour is identical to scm_make_named_hook.]
6688 Changes since Guile 1.3:
6690 * Changes to mailing lists
6692 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
6694 See the README file to find current addresses for all the Guile
6697 * Changes to the distribution
6699 ** Readline support is no longer included with Guile by default.
6701 Based on the different license terms of Guile and Readline, we
6702 concluded that Guile should not *by default* cause the linking of
6703 Readline into an application program. Readline support is now offered
6704 as a separate module, which is linked into an application only when
6705 you explicitly specify it.
6707 Although Guile is GNU software, its distribution terms add a special
6708 exception to the usual GNU General Public License (GPL). Guile's
6709 license includes a clause that allows you to link Guile with non-free
6710 programs. We add this exception so as not to put Guile at a
6711 disadvantage vis-a-vis other extensibility packages that support other
6714 In contrast, the GNU Readline library is distributed under the GNU
6715 General Public License pure and simple. This means that you may not
6716 link Readline, even dynamically, into an application unless it is
6717 distributed under a free software license that is compatible the GPL.
6719 Because of this difference in distribution terms, an application that
6720 can use Guile may not be able to use Readline. Now users will be
6721 explicitly offered two independent decisions about the use of these
6724 You can activate the readline support by issuing
6726 (use-modules (readline-activator))
6729 from your ".guile" file, for example.
6731 * Changes to the stand-alone interpreter
6733 ** All builtins now print as primitives.
6734 Previously builtin procedures not belonging to the fundamental subr
6735 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
6736 Now, they print as #<primitive-procedure NAME>.
6738 ** Backtraces slightly more intelligible.
6739 gsubr-apply and macro transformer application frames no longer appear
6742 * Changes to Scheme functions and syntax
6744 ** Guile now correctly handles internal defines by rewriting them into
6745 their equivalent letrec. Previously, internal defines would
6746 incrementally add to the innermost environment, without checking
6747 whether the restrictions specified in RnRS were met. This lead to the
6748 correct behaviour when these restriction actually were met, but didn't
6749 catch all illegal uses. Such an illegal use could lead to crashes of
6750 the Guile interpreter or other unwanted results. An example of
6751 incorrect internal defines that made Guile behave erratically:
6763 The problem with this example is that the definition of `c' uses the
6764 value of `b' directly. This confuses the meoization machine of Guile
6765 so that the second call of `b' (this time in a larger environment that
6766 also contains bindings for `c' and `d') refers to the binding of `c'
6767 instead of `a'. You could also make Guile crash with a variation on
6772 (define (b flag) (if flag a 1))
6773 (define c (1+ (b flag)))
6781 From now on, Guile will issue an `Unbound variable: b' error message
6786 A hook contains a list of functions which should be called on
6787 particular occasions in an existing program. Hooks are used for
6790 A window manager might have a hook before-window-map-hook. The window
6791 manager uses the function run-hooks to call all functions stored in
6792 before-window-map-hook each time a window is mapped. The user can
6793 store functions in the hook using add-hook!.
6795 In Guile, hooks are first class objects.
6797 *** New function: make-hook [N_ARGS]
6799 Return a hook for hook functions which can take N_ARGS arguments.
6800 The default value for N_ARGS is 0.
6802 (See also scm_make_named_hook below.)
6804 *** New function: add-hook! HOOK PROC [APPEND_P]
6806 Put PROC at the beginning of the list of functions stored in HOOK.
6807 If APPEND_P is supplied, and non-false, put PROC at the end instead.
6809 PROC must be able to take the number of arguments specified when the
6812 If PROC already exists in HOOK, then remove it first.
6814 *** New function: remove-hook! HOOK PROC
6816 Remove PROC from the list of functions in HOOK.
6818 *** New function: reset-hook! HOOK
6820 Clear the list of hook functions stored in HOOK.
6822 *** New function: run-hook HOOK ARG1 ...
6824 Run all hook functions stored in HOOK with arguments ARG1 ... .
6825 The number of arguments supplied must correspond to the number given
6826 when the hook was created.
6828 ** The function `dynamic-link' now takes optional keyword arguments.
6829 The only keyword argument that is currently defined is `:global
6830 BOOL'. With it, you can control whether the shared library will be
6831 linked in global mode or not. In global mode, the symbols from the
6832 linked library can be used to resolve references from other
6833 dynamically linked libraries. In non-global mode, the linked
6834 library is essentially invisible and can only be accessed via
6835 `dynamic-func', etc. The default is now to link in global mode.
6836 Previously, the default has been non-global mode.
6838 The `#:global' keyword is only effective on platforms that support
6839 the dlopen family of functions.
6841 ** New function `provided?'
6843 - Function: provided? FEATURE
6844 Return true iff FEATURE is supported by this installation of
6845 Guile. FEATURE must be a symbol naming a feature; the global
6846 variable `*features*' is a list of available features.
6848 ** Changes to the module (ice-9 expect):
6850 *** The expect-strings macro now matches `$' in a regular expression
6851 only at a line-break or end-of-file by default. Previously it would
6852 match the end of the string accumulated so far. The old behaviour
6853 can be obtained by setting the variable `expect-strings-exec-flags'
6856 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
6857 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
6858 in a regular expression will still match before a line-break or
6859 end-of-file. The default is `regexp/noteol'.
6861 *** The expect-strings macro now uses a variable
6862 `expect-strings-compile-flags' for the flags to be supplied to
6863 `make-regexp'. The default is `regexp/newline', which was previously
6866 *** The expect macro now supplies two arguments to a match procedure:
6867 the current accumulated string and a flag to indicate whether
6868 end-of-file has been reached. Previously only the string was supplied.
6869 If end-of-file is reached, the match procedure will be called an
6870 additional time with the same accumulated string as the previous call
6871 but with the flag set.
6873 ** New module (ice-9 format), implementing the Common Lisp `format' function.
6875 This code, and the documentation for it that appears here, was
6876 borrowed from SLIB, with minor adaptations for Guile.
6878 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
6879 An almost complete implementation of Common LISP format description
6880 according to the CL reference book `Common LISP' from Guy L.
6881 Steele, Digital Press. Backward compatible to most of the
6882 available Scheme format implementations.
6884 Returns `#t', `#f' or a string; has side effect of printing
6885 according to FORMAT-STRING. If DESTINATION is `#t', the output is
6886 to the current output port and `#t' is returned. If DESTINATION
6887 is `#f', a formatted string is returned as the result of the call.
6888 NEW: If DESTINATION is a string, DESTINATION is regarded as the
6889 format string; FORMAT-STRING is then the first argument and the
6890 output is returned as a string. If DESTINATION is a number, the
6891 output is to the current error port if available by the
6892 implementation. Otherwise DESTINATION must be an output port and
6895 FORMAT-STRING must be a string. In case of a formatting error
6896 format returns `#f' and prints a message on the current output or
6897 error port. Characters are output as if the string were output by
6898 the `display' function with the exception of those prefixed by a
6899 tilde (~). For a detailed description of the FORMAT-STRING syntax
6900 please consult a Common LISP format reference manual. For a test
6901 suite to verify this format implementation load `formatst.scm'.
6902 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
6904 Note: `format' is not reentrant, i.e. only one `format'-call may
6905 be executed at a time.
6908 *** Format Specification (Format version 3.0)
6910 Please consult a Common LISP format reference manual for a detailed
6911 description of the format string syntax. For a demonstration of the
6912 implemented directives see `formatst.scm'.
6914 This implementation supports directive parameters and modifiers (`:'
6915 and `@' characters). Multiple parameters must be separated by a comma
6916 (`,'). Parameters can be numerical parameters (positive or negative),
6917 character parameters (prefixed by a quote character (`''), variable
6918 parameters (`v'), number of rest arguments parameter (`#'), empty and
6919 default parameters. Directive characters are case independent. The
6920 general form of a directive is:
6922 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
6924 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
6926 *** Implemented CL Format Control Directives
6928 Documentation syntax: Uppercase characters represent the
6929 corresponding control directive characters. Lowercase characters
6930 represent control directive parameter descriptions.
6933 Any (print as `display' does).
6937 `~MINCOL,COLINC,MINPAD,PADCHARA'
6941 S-expression (print as `write' does).
6945 `~MINCOL,COLINC,MINPAD,PADCHARS'
6951 print number sign always.
6954 print comma separated.
6956 `~MINCOL,PADCHAR,COMMACHARD'
6962 print number sign always.
6965 print comma separated.
6967 `~MINCOL,PADCHAR,COMMACHARX'
6973 print number sign always.
6976 print comma separated.
6978 `~MINCOL,PADCHAR,COMMACHARO'
6984 print number sign always.
6987 print comma separated.
6989 `~MINCOL,PADCHAR,COMMACHARB'
6994 `~N,MINCOL,PADCHAR,COMMACHARR'
6998 print a number as a Roman numeral.
7001 print a number as an "old fashioned" Roman numeral.
7004 print a number as an ordinal English number.
7007 print a number as a cardinal English number.
7012 prints `y' and `ies'.
7015 as `~P but jumps 1 argument backward.'
7018 as `~@P but jumps 1 argument backward.'
7023 prints a character as the reader can understand it (i.e. `#\'
7027 prints a character as emacs does (eg. `^C' for ASCII 03).
7030 Fixed-format floating-point (prints a flonum like MMM.NNN).
7031 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
7033 If the number is positive a plus sign is printed.
7036 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
7037 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
7039 If the number is positive a plus sign is printed.
7042 General floating-point (prints a flonum either fixed or
7044 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
7046 If the number is positive a plus sign is printed.
7049 Dollars floating-point (prints a flonum in fixed with signs
7051 `~DIGITS,SCALE,WIDTH,PADCHAR$'
7053 If the number is positive a plus sign is printed.
7056 A sign is always printed and appears before the padding.
7059 The sign appears before the padding.
7067 print newline if not at the beginning of the output line.
7069 prints `~&' and then N-1 newlines.
7074 print N page separators.
7084 newline is ignored, white space left.
7087 newline is left, white space ignored.
7092 relative tabulation.
7098 Indirection (expects indirect arguments as a list).
7100 extracts indirect arguments from format arguments.
7103 Case conversion (converts by `string-downcase').
7105 converts by `string-capitalize'.
7108 converts by `string-capitalize-first'.
7111 converts by `string-upcase'.
7114 Argument Jumping (jumps 1 argument forward).
7116 jumps N arguments forward.
7119 jumps 1 argument backward.
7122 jumps N arguments backward.
7125 jumps to the 0th argument.
7128 jumps to the Nth argument (beginning from 0)
7130 `~[STR0~;STR1~;...~;STRN~]'
7131 Conditional Expression (numerical clause conditional).
7133 take argument from N.
7136 true test conditional.
7139 if-else-then conditional.
7145 default clause follows.
7148 Iteration (args come from the next argument (a list)).
7150 at most N iterations.
7153 args from next arg (a list of lists).
7156 args from the rest of arguments.
7159 args from the rest args (lists).
7170 aborts if N <= M <= K
7172 *** Not Implemented CL Format Control Directives
7175 print `#f' as an empty list (see below).
7178 print `#f' as an empty list (see below).
7184 (sorry I don't understand its semantics completely)
7186 *** Extended, Replaced and Additional Control Directives
7188 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
7189 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
7190 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
7191 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
7192 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
7193 COMMAWIDTH is the number of characters between two comma
7197 print a R4RS complex number as `~F~@Fi' with passed parameters for
7201 Pretty print formatting of an argument for scheme code lists.
7207 Flushes the output if format DESTINATION is a port.
7210 Print a `#\space' character
7212 print N `#\space' characters.
7215 Print a `#\tab' character
7217 print N `#\tab' characters.
7220 Takes N as an integer representation for a character. No arguments
7221 are consumed. N is converted to a character by `integer->char'. N
7222 must be a positive decimal number.
7225 Print out readproof. Prints out internal objects represented as
7226 `#<...>' as strings `"#<...>"' so that the format output can always
7227 be processed by `read'.
7230 Print out readproof. Prints out internal objects represented as
7231 `#<...>' as strings `"#<...>"' so that the format output can always
7232 be processed by `read'.
7235 Prints information and a copyright notice on the format
7238 prints format version.
7241 may also print number strings, i.e. passing a number as a string
7242 and format it accordingly.
7244 *** Configuration Variables
7246 The format module exports some configuration variables to suit the
7247 systems and users needs. There should be no modification necessary for
7248 the configuration that comes with Guile. Format detects automatically
7249 if the running scheme system implements floating point numbers and
7252 format:symbol-case-conv
7253 Symbols are converted by `symbol->string' so the case type of the
7254 printed symbols is implementation dependent.
7255 `format:symbol-case-conv' is a one arg closure which is either
7256 `#f' (no conversion), `string-upcase', `string-downcase' or
7257 `string-capitalize'. (default `#f')
7259 format:iobj-case-conv
7260 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
7261 implementation internal objects. (default `#f')
7264 The character prefixing the exponent value in `~E' printing.
7267 *** Compatibility With Other Format Implementations
7273 Downward compatible except for padding support and `~A', `~S',
7274 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
7275 `printf' padding support which is completely replaced by the CL
7276 `format' padding style.
7279 Downward compatible except for `~', which is not documented
7280 (ignores all characters inside the format string up to a newline
7281 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
7282 numerical and variable parameters and `:/@' modifiers in the CL
7286 Downward compatible except for `~A' and `~S' which print in
7287 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
7288 directive parameters or modifiers)).
7291 Downward compatible except for an optional destination parameter:
7292 S2C accepts a format call without a destination which returns a
7293 formatted string. This is equivalent to a #f destination in S2C.
7294 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
7295 parameters or modifiers)).
7298 ** Changes to string-handling functions.
7300 These functions were added to support the (ice-9 format) module, above.
7302 *** New function: string-upcase STRING
7303 *** New function: string-downcase STRING
7305 These are non-destructive versions of the existing string-upcase! and
7306 string-downcase! functions.
7308 *** New function: string-capitalize! STRING
7309 *** New function: string-capitalize STRING
7311 These functions convert the first letter of each word in the string to
7314 (string-capitalize "howdy there")
7317 As with the other functions, string-capitalize! modifies the string in
7318 place, while string-capitalize returns a modified copy of its argument.
7320 *** New function: string-ci->symbol STRING
7322 Return a symbol whose name is STRING, but having the same case as if
7323 the symbol had be read by `read'.
7325 Guile can be configured to be sensitive or insensitive to case
7326 differences in Scheme identifiers. If Guile is case-insensitive, all
7327 symbols are converted to lower case on input. The `string-ci->symbol'
7328 function returns a symbol whose name in STRING, transformed as Guile
7329 would if STRING were input.
7331 *** New function: substring-move! STRING1 START END STRING2 START
7333 Copy the substring of STRING1 from START (inclusive) to END
7334 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
7335 string, and the source and destination areas may overlap; in all
7336 cases, the function behaves as if all the characters were copied
7339 *** Extended functions: substring-move-left! substring-move-right!
7341 These functions now correctly copy arbitrarily overlapping substrings;
7342 they are both synonyms for substring-move!.
7345 ** New module (ice-9 getopt-long), with the function `getopt-long'.
7347 getopt-long is a function for parsing command-line arguments in a
7348 manner consistent with other GNU programs.
7350 (getopt-long ARGS GRAMMAR)
7351 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
7353 ARGS should be a list of strings. Its first element should be the
7354 name of the program; subsequent elements should be the arguments
7355 that were passed to the program on the command line. The
7356 `program-arguments' procedure returns a list of this form.
7358 GRAMMAR is a list of the form:
7359 ((OPTION (PROPERTY VALUE) ...) ...)
7361 Each OPTION should be a symbol. `getopt-long' will accept a
7362 command-line option named `--OPTION'.
7363 Each option can have the following (PROPERTY VALUE) pairs:
7365 (single-char CHAR) --- Accept `-CHAR' as a single-character
7366 equivalent to `--OPTION'. This is how to specify traditional
7368 (required? BOOL) --- If BOOL is true, the option is required.
7369 getopt-long will raise an error if it is not found in ARGS.
7370 (value BOOL) --- If BOOL is #t, the option accepts a value; if
7371 it is #f, it does not; and if it is the symbol
7372 `optional', the option may appear in ARGS with or
7374 (predicate FUNC) --- If the option accepts a value (i.e. you
7375 specified `(value #t)' for this option), then getopt
7376 will apply FUNC to the value, and throw an exception
7377 if it returns #f. FUNC should be a procedure which
7378 accepts a string and returns a boolean value; you may
7379 need to use quasiquotes to get it into GRAMMAR.
7381 The (PROPERTY VALUE) pairs may occur in any order, but each
7382 property may occur only once. By default, options do not have
7383 single-character equivalents, are not required, and do not take
7386 In ARGS, single-character options may be combined, in the usual
7387 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
7388 accepts values, then it must be the last option in the
7389 combination; the value is the next argument. So, for example, using
7390 the following grammar:
7391 ((apples (single-char #\a))
7392 (blimps (single-char #\b) (value #t))
7393 (catalexis (single-char #\c) (value #t)))
7394 the following argument lists would be acceptable:
7395 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
7396 for "blimps" and "catalexis")
7397 ("-ab" "bang" "-c" "couth") (same)
7398 ("-ac" "couth" "-b" "bang") (same)
7399 ("-abc" "couth" "bang") (an error, since `-b' is not the
7400 last option in its combination)
7402 If an option's value is optional, then `getopt-long' decides
7403 whether it has a value by looking at what follows it in ARGS. If
7404 the next element is a string, and it does not appear to be an
7405 option itself, then that string is the option's value.
7407 The value of a long option can appear as the next element in ARGS,
7408 or it can follow the option name, separated by an `=' character.
7409 Thus, using the same grammar as above, the following argument lists
7411 ("--apples" "Braeburn" "--blimps" "Goodyear")
7412 ("--apples=Braeburn" "--blimps" "Goodyear")
7413 ("--blimps" "Goodyear" "--apples=Braeburn")
7415 If the option "--" appears in ARGS, argument parsing stops there;
7416 subsequent arguments are returned as ordinary arguments, even if
7417 they resemble options. So, in the argument list:
7418 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
7419 `getopt-long' will recognize the `apples' option as having the
7420 value "Granny Smith", but it will not recognize the `blimp'
7421 option; it will return the strings "--blimp" and "Goodyear" as
7422 ordinary argument strings.
7424 The `getopt-long' function returns the parsed argument list as an
7425 assocation list, mapping option names --- the symbols from GRAMMAR
7426 --- onto their values, or #t if the option does not accept a value.
7427 Unused options do not appear in the alist.
7429 All arguments that are not the value of any option are returned
7430 as a list, associated with the empty list.
7432 `getopt-long' throws an exception if:
7433 - it finds an unrecognized option in ARGS
7434 - a required option is omitted
7435 - an option that requires an argument doesn't get one
7436 - an option that doesn't accept an argument does get one (this can
7437 only happen using the long option `--opt=value' syntax)
7438 - an option predicate fails
7443 `((lockfile-dir (required? #t)
7446 (predicate ,file-is-directory?))
7447 (verbose (required? #f)
7450 (x-includes (single-char #\x))
7451 (rnet-server (single-char #\y)
7452 (predicate ,string?))))
7454 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
7455 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
7457 => ((() "foo1" "-fred" "foo2" "foo3")
7458 (rnet-server . "lamprod")
7459 (x-includes . "/usr/include")
7460 (lockfile-dir . "/tmp")
7463 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
7465 It will be removed in a few releases.
7467 ** New syntax: lambda*
7468 ** New syntax: define*
7469 ** New syntax: define*-public
7470 ** New syntax: defmacro*
7471 ** New syntax: defmacro*-public
7472 Guile now supports optional arguments.
7474 `lambda*', `define*', `define*-public', `defmacro*' and
7475 `defmacro*-public' are identical to the non-* versions except that
7476 they use an extended type of parameter list that has the following BNF
7477 syntax (parentheses are literal, square brackets indicate grouping,
7478 and `*', `+' and `?' have the usual meaning):
7480 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
7481 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
7482 [[#&rest identifier]|[. identifier]]? ) | [identifier]
7484 ext-var-decl ::= identifier | ( identifier expression )
7486 The semantics are best illustrated with the following documentation
7487 and examples for `lambda*':
7490 lambda extended for optional and keyword arguments
7492 lambda* creates a procedure that takes optional arguments. These
7493 are specified by putting them inside brackets at the end of the
7494 paramater list, but before any dotted rest argument. For example,
7495 (lambda* (a b #&optional c d . e) '())
7496 creates a procedure with fixed arguments a and b, optional arguments c
7497 and d, and rest argument e. If the optional arguments are omitted
7498 in a call, the variables for them are unbound in the procedure. This
7499 can be checked with the bound? macro.
7501 lambda* can also take keyword arguments. For example, a procedure
7503 (lambda* (#&key xyzzy larch) '())
7504 can be called with any of the argument lists (#:xyzzy 11)
7505 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
7506 are given as keywords are bound to values.
7508 Optional and keyword arguments can also be given default values
7509 which they take on when they are not present in a call, by giving a
7510 two-item list in place of an optional argument, for example in:
7511 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
7512 foo is a fixed argument, bar is an optional argument with default
7513 value 42, and baz is a keyword argument with default value 73.
7514 Default value expressions are not evaluated unless they are needed
7515 and until the procedure is called.
7517 lambda* now supports two more special parameter list keywords.
7519 lambda*-defined procedures now throw an error by default if a
7520 keyword other than one of those specified is found in the actual
7521 passed arguments. However, specifying #&allow-other-keys
7522 immediately after the kyword argument declarations restores the
7523 previous behavior of ignoring unknown keywords. lambda* also now
7524 guarantees that if the same keyword is passed more than once, the
7525 last one passed is the one that takes effect. For example,
7526 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
7527 #:heads 37 #:tails 42 #:heads 99)
7528 would result in (99 47) being displayed.
7530 #&rest is also now provided as a synonym for the dotted syntax rest
7531 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
7532 all respects to lambda*. This is provided for more similarity to DSSSL,
7533 MIT-Scheme and Kawa among others, as well as for refugees from other
7536 Further documentation may be found in the optargs.scm file itself.
7538 The optional argument module also exports the macros `let-optional',
7539 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
7540 are not documented here because they may be removed in the future, but
7541 full documentation is still available in optargs.scm.
7543 ** New syntax: and-let*
7544 Guile now supports the `and-let*' form, described in the draft SRFI-2.
7546 Syntax: (land* (<clause> ...) <body> ...)
7547 Each <clause> should have one of the following forms:
7548 (<variable> <expression>)
7551 Each <variable> or <bound-variable> should be an identifier. Each
7552 <expression> should be a valid expression. The <body> should be a
7553 possibly empty sequence of expressions, like the <body> of a
7556 Semantics: A LAND* expression is evaluated by evaluating the
7557 <expression> or <bound-variable> of each of the <clause>s from
7558 left to right. The value of the first <expression> or
7559 <bound-variable> that evaluates to a false value is returned; the
7560 remaining <expression>s and <bound-variable>s are not evaluated.
7561 The <body> forms are evaluated iff all the <expression>s and
7562 <bound-variable>s evaluate to true values.
7564 The <expression>s and the <body> are evaluated in an environment
7565 binding each <variable> of the preceding (<variable> <expression>)
7566 clauses to the value of the <expression>. Later bindings
7567 shadow earlier bindings.
7569 Guile's and-let* macro was contributed by Michael Livshin.
7571 ** New sorting functions
7573 *** New function: sorted? SEQUENCE LESS?
7574 Returns `#t' when the sequence argument is in non-decreasing order
7575 according to LESS? (that is, there is no adjacent pair `... x y
7576 ...' for which `(less? y x)').
7578 Returns `#f' when the sequence contains at least one out-of-order
7579 pair. It is an error if the sequence is neither a list nor a
7582 *** New function: merge LIST1 LIST2 LESS?
7583 LIST1 and LIST2 are sorted lists.
7584 Returns the sorted list of all elements in LIST1 and LIST2.
7586 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
7587 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
7588 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
7589 (Here "<" should read "comes before".)
7591 *** New procedure: merge! LIST1 LIST2 LESS?
7592 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
7593 the result. If the code is compiled, and LESS? constructs no new
7594 pairs, no pairs at all will be allocated. The first pair of the
7595 result will be either the first pair of LIST1 or the first pair of
7598 *** New function: sort SEQUENCE LESS?
7599 Accepts either a list or a vector, and returns a new sequence
7600 which is sorted. The new sequence is the same type as the input.
7601 Always `(sorted? (sort sequence less?) less?)'. The original
7602 sequence is not altered in any way. The new sequence shares its
7603 elements with the old one; no elements are copied.
7605 *** New procedure: sort! SEQUENCE LESS
7606 Returns its sorted result in the original boxes. No new storage is
7607 allocated at all. Proper usage: (set! slist (sort! slist <))
7609 *** New function: stable-sort SEQUENCE LESS?
7610 Similar to `sort' but stable. That is, if "equal" elements are
7611 ordered a < b in the original sequence, they will have the same order
7614 *** New function: stable-sort! SEQUENCE LESS?
7615 Similar to `sort!' but stable.
7616 Uses temporary storage when sorting vectors.
7618 *** New functions: sort-list, sort-list!
7619 Added for compatibility with scsh.
7621 ** New built-in random number support
7623 *** New function: random N [STATE]
7624 Accepts a positive integer or real N and returns a number of the
7625 same type between zero (inclusive) and N (exclusive). The values
7626 returned have a uniform distribution.
7628 The optional argument STATE must be of the type produced by
7629 `copy-random-state' or `seed->random-state'. It defaults to the value
7630 of the variable `*random-state*'. This object is used to maintain the
7631 state of the pseudo-random-number generator and is altered as a side
7632 effect of the `random' operation.
7634 *** New variable: *random-state*
7635 Holds a data structure that encodes the internal state of the
7636 random-number generator that `random' uses by default. The nature
7637 of this data structure is implementation-dependent. It may be
7638 printed out and successfully read back in, but may or may not
7639 function correctly as a random-number state object in another
7642 *** New function: copy-random-state [STATE]
7643 Returns a new object of type suitable for use as the value of the
7644 variable `*random-state*' and as a second argument to `random'.
7645 If argument STATE is given, a copy of it is returned. Otherwise a
7646 copy of `*random-state*' is returned.
7648 *** New function: seed->random-state SEED
7649 Returns a new object of type suitable for use as the value of the
7650 variable `*random-state*' and as a second argument to `random'.
7651 SEED is a string or a number. A new state is generated and
7652 initialized using SEED.
7654 *** New function: random:uniform [STATE]
7655 Returns an uniformly distributed inexact real random number in the
7656 range between 0 and 1.
7658 *** New procedure: random:solid-sphere! VECT [STATE]
7659 Fills VECT with inexact real random numbers the sum of whose
7660 squares is less than 1.0. Thinking of VECT as coordinates in
7661 space of dimension N = `(vector-length VECT)', the coordinates are
7662 uniformly distributed within the unit N-shere. The sum of the
7663 squares of the numbers is returned. VECT can be either a vector
7664 or a uniform vector of doubles.
7666 *** New procedure: random:hollow-sphere! VECT [STATE]
7667 Fills VECT with inexact real random numbers the sum of whose squares
7668 is equal to 1.0. Thinking of VECT as coordinates in space of
7669 dimension n = `(vector-length VECT)', the coordinates are uniformly
7670 distributed over the surface of the unit n-shere. VECT can be either
7671 a vector or a uniform vector of doubles.
7673 *** New function: random:normal [STATE]
7674 Returns an inexact real in a normal distribution with mean 0 and
7675 standard deviation 1. For a normal distribution with mean M and
7676 standard deviation D use `(+ M (* D (random:normal)))'.
7678 *** New procedure: random:normal-vector! VECT [STATE]
7679 Fills VECT with inexact real random numbers which are independent and
7680 standard normally distributed (i.e., with mean 0 and variance 1).
7681 VECT can be either a vector or a uniform vector of doubles.
7683 *** New function: random:exp STATE
7684 Returns an inexact real in an exponential distribution with mean 1.
7685 For an exponential distribution with mean U use (* U (random:exp)).
7687 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
7689 These functions now operate on numbers in the range of a C unsigned
7692 These functions used to operate on numbers in the range of a C signed
7693 long; however, this seems inappropriate, because Guile integers don't
7696 ** New function: make-guardian
7697 This is an implementation of guardians as described in
7698 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
7699 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
7700 Programming Language Design and Implementation, June 1993
7701 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
7703 ** New functions: delq1!, delv1!, delete1!
7704 These procedures behave similar to delq! and friends but delete only
7705 one object if at all.
7707 ** New function: unread-string STRING PORT
7708 Unread STRING to PORT, that is, push it back onto the port so that
7709 next read operation will work on the pushed back characters.
7711 ** unread-char can now be called multiple times
7712 If unread-char is called multiple times, the unread characters will be
7713 read again in last-in first-out order.
7715 ** the procedures uniform-array-read! and uniform-array-write! now
7716 work on any kind of port, not just ports which are open on a file.
7718 ** Now 'l' in a port mode requests line buffering.
7720 ** The procedure truncate-file now works on string ports as well
7721 as file ports. If the size argument is omitted, the current
7722 file position is used.
7724 ** new procedure: seek PORT/FDES OFFSET WHENCE
7725 The arguments are the same as for the old fseek procedure, but it
7726 works on string ports as well as random-access file ports.
7728 ** the fseek procedure now works on string ports, since it has been
7729 redefined using seek.
7731 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
7732 size is not supplied.
7734 ** the newline procedure no longer flushes the port if it's not
7735 line-buffered: previously it did if it was the current output port.
7737 ** open-pipe and close-pipe are no longer primitive procedures, but
7738 an emulation can be obtained using `(use-modules (ice-9 popen))'.
7740 ** the freopen procedure has been removed.
7742 ** new procedure: drain-input PORT
7743 Drains PORT's read buffers (including any pushed-back characters)
7744 and returns the contents as a single string.
7746 ** New function: map-in-order PROC LIST1 LIST2 ...
7747 Version of `map' which guarantees that the procedure is applied to the
7748 lists in serial order.
7750 ** Renamed `serial-array-copy!' and `serial-array-map!' to
7751 `array-copy-in-order!' and `array-map-in-order!'. The old names are
7752 now obsolete and will go away in release 1.5.
7754 ** New syntax: collect BODY1 ...
7755 Version of `begin' which returns a list of the results of the body
7756 forms instead of the result of the last body form. In contrast to
7757 `begin', `collect' allows an empty body.
7759 ** New functions: read-history FILENAME, write-history FILENAME
7760 Read/write command line history from/to file. Returns #t on success
7761 and #f if an error occured.
7763 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
7765 These procedures return a list of definitions available in the specified
7766 argument, a relative module reference. In the case of no argument,
7767 `(current-module)' is now consulted for definitions to return, instead
7768 of simply returning #f, the former behavior.
7770 ** The #/ syntax for lists is no longer supported.
7772 Earlier versions of Scheme accepted this syntax, but printed a
7775 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
7777 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
7780 * Changes to the gh_ interface
7784 Now takes a second argument which is the result array. If this
7785 pointer is NULL, a new array is malloced (the old behaviour).
7787 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
7788 gh_scm2shorts, gh_scm2longs, gh_scm2floats
7792 * Changes to the scm_ interface
7794 ** Function: scm_make_named_hook (char* name, int n_args)
7796 Creates a hook in the same way as make-hook above but also
7797 binds a variable named NAME to it.
7799 This is the typical way of creating a hook from C code.
7801 Currently, the variable is created in the "current" module. This
7802 might change when we get the new module system.
7804 ** The smob interface
7806 The interface for creating smobs has changed. For documentation, see
7807 data-rep.info (made from guile-core/doc/data-rep.texi).
7809 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
7811 >>> This function will be removed in 1.3.4. <<<
7815 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
7816 This function adds a new smob type, named NAME, with instance size
7817 SIZE to the system. The return value is a tag that is used in
7818 creating instances of the type. If SIZE is 0, then no memory will
7819 be allocated when instances of the smob are created, and nothing
7820 will be freed by the default free function.
7822 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
7823 This function sets the smob marking procedure for the smob type
7824 specified by the tag TC. TC is the tag returned by
7825 `scm_make_smob_type'.
7827 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
7828 This function sets the smob freeing procedure for the smob type
7829 specified by the tag TC. TC is the tag returned by
7830 `scm_make_smob_type'.
7832 *** Function: void scm_set_smob_print (tc, print)
7834 - Function: void scm_set_smob_print (long tc,
7835 scm_sizet (*print) (SCM,
7839 This function sets the smob printing procedure for the smob type
7840 specified by the tag TC. TC is the tag returned by
7841 `scm_make_smob_type'.
7843 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
7844 This function sets the smob equality-testing predicate for the
7845 smob type specified by the tag TC. TC is the tag returned by
7846 `scm_make_smob_type'.
7848 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
7849 Make VALUE contain a smob instance of the type with type code TC and
7850 smob data DATA. VALUE must be previously declared as C type `SCM'.
7852 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
7853 This macro expands to a block of code that creates a smob instance
7854 of the type with type code TC and smob data DATA, and returns that
7855 `SCM' value. It should be the last piece of code in a block.
7857 ** The interfaces for using I/O ports and implementing port types
7858 (ptobs) have changed significantly. The new interface is based on
7859 shared access to buffers and a new set of ptob procedures.
7861 *** scm_newptob has been removed
7865 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
7867 - Function: SCM scm_make_port_type (char *type_name,
7868 int (*fill_buffer) (SCM port),
7869 void (*write_flush) (SCM port));
7871 Similarly to the new smob interface, there is a set of function
7872 setters by which the user can customize the behaviour of his port
7873 type. See ports.h (scm_set_port_XXX).
7875 ** scm_strport_to_string: New function: creates a new string from
7876 a string port's buffer.
7878 ** Plug in interface for random number generators
7879 The variable `scm_the_rng' in random.c contains a value and three
7880 function pointers which together define the current random number
7881 generator being used by the Scheme level interface and the random
7882 number library functions.
7884 The user is free to replace the default generator with the generator
7887 *** Variable: size_t scm_the_rng.rstate_size
7888 The size of the random state type used by the current RNG
7891 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
7892 Given the random STATE, return 32 random bits.
7894 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
7895 Seed random state STATE using string S of length N.
7897 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
7898 Given random state STATE, return a malloced copy.
7901 The default RNG is the MWC (Multiply With Carry) random number
7902 generator described by George Marsaglia at the Department of
7903 Statistics and Supercomputer Computations Research Institute, The
7904 Florida State University (http://stat.fsu.edu/~geo).
7906 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
7907 passes all tests in the DIEHARD test suite
7908 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
7909 costs one multiply and one add on platforms which either supports long
7910 longs (gcc does this on most systems) or have 64 bit longs. The cost
7911 is four multiply on other systems but this can be optimized by writing
7912 scm_i_uniform32 in assembler.
7914 These functions are provided through the scm_the_rng interface for use
7915 by libguile and the application.
7917 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
7918 Given the random STATE, return 32 random bits.
7919 Don't use this function directly. Instead go through the plugin
7920 interface (see "Plug in interface" above).
7922 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
7923 Initialize STATE using SEED of length N.
7925 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
7926 Return a malloc:ed copy of STATE. This function can easily be re-used
7927 in the interfaces to other RNGs.
7929 ** Random number library functions
7930 These functions use the current RNG through the scm_the_rng interface.
7931 It might be a good idea to use these functions from your C code so
7932 that only one random generator is used by all code in your program.
7934 The default random state is stored in:
7936 *** Variable: SCM scm_var_random_state
7937 Contains the vcell of the Scheme variable "*random-state*" which is
7938 used as default state by all random number functions in the Scheme
7943 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
7945 *** Function: scm_rstate *scm_c_default_rstate (void)
7946 This is a convenience function which returns the value of
7947 scm_var_random_state. An error message is generated if this value
7948 isn't a random state.
7950 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
7951 Make a new random state from the string SEED of length LENGTH.
7953 It is generally not a good idea to use multiple random states in a
7954 program. While subsequent random numbers generated from one random
7955 state are guaranteed to be reasonably independent, there is no such
7956 guarantee for numbers generated from different random states.
7958 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
7959 Return 32 random bits.
7961 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
7962 Return a sample from the uniform(0,1) distribution.
7964 *** Function: double scm_c_normal01 (scm_rstate *STATE)
7965 Return a sample from the normal(0,1) distribution.
7967 *** Function: double scm_c_exp1 (scm_rstate *STATE)
7968 Return a sample from the exp(1) distribution.
7970 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
7971 Return a sample from the discrete uniform(0,M) distribution.
7973 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
7974 Return a sample from the discrete uniform(0,M) distribution.
7975 M must be a bignum object. The returned value may be an INUM.
7979 Changes in Guile 1.3 (released Monday, October 19, 1998):
7981 * Changes to the distribution
7983 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
7984 To avoid conflicts, programs should name environment variables after
7985 themselves, except when there's a common practice establishing some
7988 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
7989 giving the former precedence, and printing a warning message if the
7990 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
7992 ** The header files related to multi-byte characters have been removed.
7993 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
7994 which referred to these explicitly will probably need to be rewritten,
7995 since the support for the variant string types has been removed; see
7998 ** The header files append.h and sequences.h have been removed. These
7999 files implemented non-R4RS operations which would encourage
8000 non-portable programming style and less easy-to-read code.
8002 * Changes to the stand-alone interpreter
8004 ** New procedures have been added to implement a "batch mode":
8006 *** Function: batch-mode?
8008 Returns a boolean indicating whether the interpreter is in batch
8011 *** Function: set-batch-mode?! ARG
8013 If ARG is true, switches the interpreter to batch mode. The `#f'
8014 case has not been implemented.
8016 ** Guile now provides full command-line editing, when run interactively.
8017 To use this feature, you must have the readline library installed.
8018 The Guile build process will notice it, and automatically include
8021 The readline library is available via anonymous FTP from any GNU
8022 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
8024 ** the-last-stack is now a fluid.
8026 * Changes to the procedure for linking libguile with your programs
8028 ** You can now use the `guile-config' utility to build programs that use Guile.
8030 Guile now includes a command-line utility called `guile-config', which
8031 can provide information about how to compile and link programs that
8034 *** `guile-config compile' prints any C compiler flags needed to use Guile.
8035 You should include this command's output on the command line you use
8036 to compile C or C++ code that #includes the Guile header files. It's
8037 usually just a `-I' flag to help the compiler find the Guile headers.
8040 *** `guile-config link' prints any linker flags necessary to link with Guile.
8042 This command writes to its standard output a list of flags which you
8043 must pass to the linker to link your code against the Guile library.
8044 The flags include '-lguile' itself, any other libraries the Guile
8045 library depends upon, and any `-L' flags needed to help the linker
8046 find those libraries.
8048 For example, here is a Makefile rule that builds a program named 'foo'
8049 from the object files ${FOO_OBJECTS}, and links them against Guile:
8052 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
8054 Previous Guile releases recommended that you use autoconf to detect
8055 which of a predefined set of libraries were present on your system.
8056 It is more robust to use `guile-config', since it records exactly which
8057 libraries the installed Guile library requires.
8059 This was originally called `build-guile', but was renamed to
8060 `guile-config' before Guile 1.3 was released, to be consistent with
8061 the analogous script for the GTK+ GUI toolkit, which is called
8065 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
8067 If you are using the GNU autoconf package to configure your program,
8068 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
8069 (described above) and gather the necessary values for use in your
8072 The GUILE_FLAGS macro expands to configure script code which runs the
8073 `guile-config' script, to find out where Guile's header files and
8074 libraries are installed. It sets two variables, marked for
8075 substitution, as by AC_SUBST.
8077 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
8078 code that uses Guile header files. This is almost always just a
8081 GUILE_LDFLAGS --- flags to pass to the linker to link a
8082 program against Guile. This includes `-lguile' for the Guile
8083 library itself, any libraries that Guile itself requires (like
8084 -lqthreads), and so on. It may also include a -L flag to tell the
8085 compiler where to find the libraries.
8087 GUILE_FLAGS is defined in the file guile.m4, in the top-level
8088 directory of the Guile distribution. You can copy it into your
8089 package's aclocal.m4 file, and then use it in your configure.in file.
8091 If you are using the `aclocal' program, distributed with GNU automake,
8092 to maintain your aclocal.m4 file, the Guile installation process
8093 installs guile.m4 where aclocal will find it. All you need to do is
8094 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
8095 this will copy the definition of GUILE_FLAGS into your aclocal.m4
8099 * Changes to Scheme functions and syntax
8101 ** Multi-byte strings have been removed, as have multi-byte and wide
8102 ports. We felt that these were the wrong approach to
8103 internationalization support.
8105 ** New function: readline [PROMPT]
8106 Read a line from the terminal, and allow the user to edit it,
8107 prompting with PROMPT. READLINE provides a large set of Emacs-like
8108 editing commands, lets the user recall previously typed lines, and
8109 works on almost every kind of terminal, including dumb terminals.
8111 READLINE assumes that the cursor is at the beginning of the line when
8112 it is invoked. Thus, you can't print a prompt yourself, and then call
8113 READLINE; you need to package up your prompt as a string, pass it to
8114 the function, and let READLINE print the prompt itself. This is
8115 because READLINE needs to know the prompt's screen width.
8117 For Guile to provide this function, you must have the readline
8118 library, version 2.1 or later, installed on your system. Readline is
8119 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
8120 any GNU mirror site.
8122 See also ADD-HISTORY function.
8124 ** New function: add-history STRING
8125 Add STRING as the most recent line in the history used by the READLINE
8126 command. READLINE does not add lines to the history itself; you must
8127 call ADD-HISTORY to make previous input available to the user.
8129 ** The behavior of the read-line function has changed.
8131 This function now uses standard C library functions to read the line,
8132 for speed. This means that it doesn not respect the value of
8133 scm-line-incrementors; it assumes that lines are delimited with
8136 (Note that this is read-line, the function that reads a line of text
8137 from a port, not readline, the function that reads a line from a
8138 terminal, providing full editing capabilities.)
8140 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
8142 This module provides some simple argument parsing. It exports one
8145 Function: getopt-gnu-style ARG-LS
8146 Parse a list of program arguments into an alist of option
8149 Each item in the list of program arguments is examined to see if
8150 it meets the syntax of a GNU long-named option. An argument like
8151 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
8152 returned alist, where MUMBLE is a keyword object with the same
8153 name as the argument. An argument like `--MUMBLE=FROB' produces
8154 an element of the form (MUMBLE . FROB), where FROB is a string.
8156 As a special case, the returned alist also contains a pair whose
8157 car is the symbol `rest'. The cdr of this pair is a list
8158 containing all the items in the argument list that are not options
8159 of the form mentioned above.
8161 The argument `--' is treated specially: all items in the argument
8162 list appearing after such an argument are not examined, and are
8163 returned in the special `rest' list.
8165 This function does not parse normal single-character switches.
8166 You will need to parse them out of the `rest' list yourself.
8168 ** The read syntax for byte vectors and short vectors has changed.
8170 Instead of #bytes(...), write #y(...).
8172 Instead of #short(...), write #h(...).
8174 This may seem nutty, but, like the other uniform vectors, byte vectors
8175 and short vectors want to have the same print and read syntax (and,
8176 more basic, want to have read syntax!). Changing the read syntax to
8177 use multiple characters after the hash sign breaks with the
8178 conventions used in R5RS and the conventions used for the other
8179 uniform vectors. It also introduces complexity in the current reader,
8180 both on the C and Scheme levels. (The Right solution is probably to
8181 change the syntax and prototypes for uniform vectors entirely.)
8184 ** The new module (ice-9 session) provides useful interactive functions.
8186 *** New procedure: (apropos REGEXP OPTION ...)
8188 Display a list of top-level variables whose names match REGEXP, and
8189 the modules they are imported from. Each OPTION should be one of the
8192 value --- Show the value of each matching variable.
8193 shadow --- Show bindings shadowed by subsequently imported modules.
8194 full --- Same as both `shadow' and `value'.
8198 guile> (apropos "trace" 'full)
8199 debug: trace #<procedure trace args>
8200 debug: untrace #<procedure untrace args>
8201 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
8202 the-scm-module: before-backtrace-hook ()
8203 the-scm-module: backtrace #<primitive-procedure backtrace>
8204 the-scm-module: after-backtrace-hook ()
8205 the-scm-module: has-shown-backtrace-hint? #f
8208 ** There are new functions and syntax for working with macros.
8210 Guile implements macros as a special object type. Any variable whose
8211 top-level binding is a macro object acts as a macro. The macro object
8212 specifies how the expression should be transformed before evaluation.
8214 *** Macro objects now print in a reasonable way, resembling procedures.
8216 *** New function: (macro? OBJ)
8217 True iff OBJ is a macro object.
8219 *** New function: (primitive-macro? OBJ)
8220 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
8221 macro transformers, implemented in eval.c rather than Scheme code.
8223 Why do we have this function?
8224 - For symmetry with procedure? and primitive-procedure?,
8225 - to allow custom print procedures to tell whether a macro is
8226 primitive, and display it differently, and
8227 - to allow compilers and user-written evaluators to distinguish
8228 builtin special forms from user-defined ones, which could be
8231 *** New function: (macro-type OBJ)
8232 Return a value indicating what kind of macro OBJ is. Possible return
8235 The symbol `syntax' --- a macro created by procedure->syntax.
8236 The symbol `macro' --- a macro created by procedure->macro.
8237 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
8238 The boolean #f --- if OBJ is not a macro object.
8240 *** New function: (macro-name MACRO)
8241 Return the name of the macro object MACRO's procedure, as returned by
8244 *** New function: (macro-transformer MACRO)
8245 Return the transformer procedure for MACRO.
8247 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
8249 Specify a new macro expander to use in the current module. Each
8250 MODULE is a module name, with the same meaning as in the `use-modules'
8251 form; each named module's exported bindings are added to the current
8252 top-level environment. TRANSFORMER is an expression evaluated in the
8253 resulting environment which must yield a procedure to use as the
8254 module's eval transformer: every expression evaluated in this module
8255 is passed to this function, and the result passed to the Guile
8258 *** macro-eval! is removed. Use local-eval instead.
8260 ** Some magic has been added to the printer to better handle user
8261 written printing routines (like record printers, closure printers).
8263 The problem is that these user written routines must have access to
8264 the current `print-state' to be able to handle fancy things like
8265 detection of circular references. These print-states have to be
8266 passed to the builtin printing routines (display, write, etc) to
8267 properly continue the print chain.
8269 We didn't want to change all existing print code so that it
8270 explicitly passes thru a print state in addition to a port. Instead,
8271 we extented the possible values that the builtin printing routines
8272 accept as a `port'. In addition to a normal port, they now also take
8273 a pair of a normal port and a print-state. Printing will go to the
8274 port and the print-state will be used to control the detection of
8275 circular references, etc. If the builtin function does not care for a
8276 print-state, it is simply ignored.
8278 User written callbacks are now called with such a pair as their
8279 `port', but because every function now accepts this pair as a PORT
8280 argument, you don't have to worry about that. In fact, it is probably
8281 safest to not check for these pairs.
8283 However, it is sometimes necessary to continue a print chain on a
8284 different port, for example to get a intermediate string
8285 representation of the printed value, mangle that string somehow, and
8286 then to finally print the mangled string. Use the new function
8288 inherit-print-state OLD-PORT NEW-PORT
8290 for this. It constructs a new `port' that prints to NEW-PORT but
8291 inherits the print-state of OLD-PORT.
8293 ** struct-vtable-offset renamed to vtable-offset-user
8295 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
8297 ** There is now a third optional argument to make-vtable-vtable
8298 (and fourth to make-struct) when constructing new types (vtables).
8299 This argument initializes field vtable-index-printer of the vtable.
8301 ** The detection of circular references has been extended to structs.
8302 That is, a structure that -- in the process of being printed -- prints
8303 itself does not lead to infinite recursion.
8305 ** There is now some basic support for fluids. Please read
8306 "libguile/fluid.h" to find out more. It is accessible from Scheme with
8307 the following functions and macros:
8309 Function: make-fluid
8311 Create a new fluid object. Fluids are not special variables or
8312 some other extension to the semantics of Scheme, but rather
8313 ordinary Scheme objects. You can store them into variables (that
8314 are still lexically scoped, of course) or into any other place you
8315 like. Every fluid has a initial value of `#f'.
8317 Function: fluid? OBJ
8319 Test whether OBJ is a fluid.
8321 Function: fluid-ref FLUID
8322 Function: fluid-set! FLUID VAL
8324 Access/modify the fluid FLUID. Modifications are only visible
8325 within the current dynamic root (that includes threads).
8327 Function: with-fluids* FLUIDS VALUES THUNK
8329 FLUIDS is a list of fluids and VALUES a corresponding list of
8330 values for these fluids. Before THUNK gets called the values are
8331 installed in the fluids and the old values of the fluids are
8332 saved in the VALUES list. When the flow of control leaves THUNK
8333 or reenters it, the values get swapped again. You might think of
8334 this as a `safe-fluid-excursion'. Note that the VALUES list is
8335 modified by `with-fluids*'.
8337 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
8339 The same as `with-fluids*' but with a different syntax. It looks
8340 just like `let', but both FLUID and VALUE are evaluated. Remember,
8341 fluids are not special variables but ordinary objects. FLUID
8342 should evaluate to a fluid.
8344 ** Changes to system call interfaces:
8346 *** close-port, close-input-port and close-output-port now return a
8347 boolean instead of an `unspecified' object. #t means that the port
8348 was successfully closed, while #f means it was already closed. It is
8349 also now possible for these procedures to raise an exception if an
8350 error occurs (some errors from write can be delayed until close.)
8352 *** the first argument to chmod, fcntl, ftell and fseek can now be a
8355 *** the third argument to fcntl is now optional.
8357 *** the first argument to chown can now be a file descriptor or a port.
8359 *** the argument to stat can now be a port.
8361 *** The following new procedures have been added (most use scsh
8364 *** procedure: close PORT/FD
8365 Similar to close-port (*note close-port: Closing Ports.), but also
8366 works on file descriptors. A side effect of closing a file
8367 descriptor is that any ports using that file descriptor are moved
8368 to a different file descriptor and have their revealed counts set
8371 *** procedure: port->fdes PORT
8372 Returns the integer file descriptor underlying PORT. As a side
8373 effect the revealed count of PORT is incremented.
8375 *** procedure: fdes->ports FDES
8376 Returns a list of existing ports which have FDES as an underlying
8377 file descriptor, without changing their revealed counts.
8379 *** procedure: fdes->inport FDES
8380 Returns an existing input port which has FDES as its underlying
8381 file descriptor, if one exists, and increments its revealed count.
8382 Otherwise, returns a new input port with a revealed count of 1.
8384 *** procedure: fdes->outport FDES
8385 Returns an existing output port which has FDES as its underlying
8386 file descriptor, if one exists, and increments its revealed count.
8387 Otherwise, returns a new output port with a revealed count of 1.
8389 The next group of procedures perform a `dup2' system call, if NEWFD
8390 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
8391 duplicated can be supplied as an integer or contained in a port. The
8392 type of value returned varies depending on which procedure is used.
8394 All procedures also have the side effect when performing `dup2' that
8395 any ports using NEWFD are moved to a different file descriptor and have
8396 their revealed counts set to zero.
8398 *** procedure: dup->fdes PORT/FD [NEWFD]
8399 Returns an integer file descriptor.
8401 *** procedure: dup->inport PORT/FD [NEWFD]
8402 Returns a new input port using the new file descriptor.
8404 *** procedure: dup->outport PORT/FD [NEWFD]
8405 Returns a new output port using the new file descriptor.
8407 *** procedure: dup PORT/FD [NEWFD]
8408 Returns a new port if PORT/FD is a port, with the same mode as the
8409 supplied port, otherwise returns an integer file descriptor.
8411 *** procedure: dup->port PORT/FD MODE [NEWFD]
8412 Returns a new port using the new file descriptor. MODE supplies a
8413 mode string for the port (*note open-file: File Ports.).
8415 *** procedure: setenv NAME VALUE
8416 Modifies the environment of the current process, which is also the
8417 default environment inherited by child processes.
8419 If VALUE is `#f', then NAME is removed from the environment.
8420 Otherwise, the string NAME=VALUE is added to the environment,
8421 replacing any existing string with name matching NAME.
8423 The return value is unspecified.
8425 *** procedure: truncate-file OBJ SIZE
8426 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
8427 can be a string containing a file name or an integer file
8428 descriptor or port open for output on the file. The underlying
8429 system calls are `truncate' and `ftruncate'.
8431 The return value is unspecified.
8433 *** procedure: setvbuf PORT MODE [SIZE]
8434 Set the buffering mode for PORT. MODE can be:
8442 block buffered, using a newly allocated buffer of SIZE bytes.
8443 However if SIZE is zero or unspecified, the port will be made
8446 This procedure should not be used after I/O has been performed with
8449 Ports are usually block buffered by default, with a default buffer
8450 size. Procedures e.g., *Note open-file: File Ports, which accept a
8451 mode string allow `0' to be added to request an unbuffered port.
8453 *** procedure: fsync PORT/FD
8454 Copies any unwritten data for the specified output file descriptor
8455 to disk. If PORT/FD is a port, its buffer is flushed before the
8456 underlying file descriptor is fsync'd. The return value is
8459 *** procedure: open-fdes PATH FLAGS [MODES]
8460 Similar to `open' but returns a file descriptor instead of a port.
8462 *** procedure: execle PATH ENV [ARG] ...
8463 Similar to `execl', but the environment of the new process is
8464 specified by ENV, which must be a list of strings as returned by
8465 the `environ' procedure.
8467 This procedure is currently implemented using the `execve' system
8468 call, but we call it `execle' because of its Scheme calling
8471 *** procedure: strerror ERRNO
8472 Returns the Unix error message corresponding to ERRNO, an integer.
8474 *** procedure: primitive-exit [STATUS]
8475 Terminate the current process without unwinding the Scheme stack.
8476 This is would typically be useful after a fork. The exit status
8477 is STATUS if supplied, otherwise zero.
8479 *** procedure: times
8480 Returns an object with information about real and processor time.
8481 The following procedures accept such an object as an argument and
8482 return a selected component:
8485 The current real time, expressed as time units relative to an
8489 The CPU time units used by the calling process.
8492 The CPU time units used by the system on behalf of the
8496 The CPU time units used by terminated child processes of the
8497 calling process, whose status has been collected (e.g., using
8501 Similarly, the CPU times units used by the system on behalf of
8502 terminated child processes.
8504 ** Removed: list-length
8505 ** Removed: list-append, list-append!
8506 ** Removed: list-reverse, list-reverse!
8508 ** array-map renamed to array-map!
8510 ** serial-array-map renamed to serial-array-map!
8512 ** catch doesn't take #f as first argument any longer
8514 Previously, it was possible to pass #f instead of a key to `catch'.
8515 That would cause `catch' to pass a jump buffer object to the procedure
8516 passed as second argument. The procedure could then use this jump
8517 buffer objekt as an argument to throw.
8519 This mechanism has been removed since its utility doesn't motivate the
8520 extra complexity it introduces.
8522 ** The `#/' notation for lists now provokes a warning message from Guile.
8523 This syntax will be removed from Guile in the near future.
8525 To disable the warning message, set the GUILE_HUSH environment
8526 variable to any non-empty value.
8528 ** The newline character now prints as `#\newline', following the
8529 normal Scheme notation, not `#\nl'.
8531 * Changes to the gh_ interface
8533 ** The gh_enter function now takes care of loading the Guile startup files.
8534 gh_enter works by calling scm_boot_guile; see the remarks below.
8536 ** Function: void gh_write (SCM x)
8538 Write the printed representation of the scheme object x to the current
8539 output port. Corresponds to the scheme level `write'.
8541 ** gh_list_length renamed to gh_length.
8543 ** vector handling routines
8545 Several major changes. In particular, gh_vector() now resembles
8546 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
8547 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
8548 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
8549 vector-related gh_ functions have been implemented.
8551 ** pair and list routines
8553 Implemented several of the R4RS pair and list functions that were
8556 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
8558 New function. Converts double arrays back and forth between Scheme
8561 * Changes to the scm_ interface
8563 ** The function scm_boot_guile now takes care of loading the startup files.
8565 Guile's primary initialization function, scm_boot_guile, now takes
8566 care of loading `boot-9.scm', in the `ice-9' module, to initialize
8567 Guile, define the module system, and put together some standard
8568 bindings. It also loads `init.scm', which is intended to hold
8569 site-specific initialization code.
8571 Since Guile cannot operate properly until boot-9.scm is loaded, there
8572 is no reason to separate loading boot-9.scm from Guile's other
8573 initialization processes.
8575 This job used to be done by scm_compile_shell_switches, which didn't
8576 make much sense; in particular, it meant that people using Guile for
8577 non-shell-like applications had to jump through hoops to get Guile
8578 initialized properly.
8580 ** The function scm_compile_shell_switches no longer loads the startup files.
8581 Now, Guile always loads the startup files, whenever it is initialized;
8582 see the notes above for scm_boot_guile and scm_load_startup_files.
8584 ** Function: scm_load_startup_files
8585 This new function takes care of loading Guile's initialization file
8586 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
8587 this is always called by the Guile initialization process, it's
8588 probably not too useful to call this yourself, but it's there anyway.
8590 ** The semantics of smob marking have changed slightly.
8592 The smob marking function (the `mark' member of the scm_smobfuns
8593 structure) is no longer responsible for setting the mark bit on the
8594 smob. The generic smob handling code in the garbage collector will
8595 set this bit. The mark function need only ensure that any other
8596 objects the smob refers to get marked.
8598 Note that this change means that the smob's GC8MARK bit is typically
8599 already set upon entry to the mark function. Thus, marking functions
8600 which look like this:
8603 if (SCM_GC8MARKP (ptr))
8605 SCM_SETGC8MARK (ptr);
8606 ... mark objects to which the smob refers ...
8609 are now incorrect, since they will return early, and fail to mark any
8610 other objects the smob refers to. Some code in the Guile library used
8613 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
8615 If you have implemented your own I/O port type, by writing the
8616 functions required by the scm_ptobfuns and then calling scm_newptob,
8617 you will need to change your functions slightly.
8619 The functions in a scm_ptobfuns structure now expect the port itself
8620 as their argument; they used to expect the `stream' member of the
8621 port's scm_port_table structure. This allows functions in an
8622 scm_ptobfuns structure to easily access the port's cell (and any flags
8623 it its CAR), and the port's scm_port_table structure.
8625 Guile now passes the I/O port itself as the `port' argument in the
8626 following scm_ptobfuns functions:
8628 int (*free) (SCM port);
8629 int (*fputc) (int, SCM port);
8630 int (*fputs) (char *, SCM port);
8631 scm_sizet (*fwrite) SCM_P ((char *ptr,
8635 int (*fflush) (SCM port);
8636 int (*fgetc) (SCM port);
8637 int (*fclose) (SCM port);
8639 The interfaces to the `mark', `print', `equalp', and `fgets' methods
8642 If you have existing code which defines its own port types, it is easy
8643 to convert your code to the new interface; simply apply SCM_STREAM to
8644 the port argument to yield the value you code used to expect.
8646 Note that since both the port and the stream have the same type in the
8647 C code --- they are both SCM values --- the C compiler will not remind
8648 you if you forget to update your scm_ptobfuns functions.
8651 ** Function: int scm_internal_select (int fds,
8655 struct timeval *timeout);
8657 This is a replacement for the `select' function provided by the OS.
8658 It enables I/O blocking and sleeping to happen for one cooperative
8659 thread without blocking other threads. It also avoids busy-loops in
8660 these situations. It is intended that all I/O blocking and sleeping
8661 will finally go through this function. Currently, this function is
8662 only available on systems providing `gettimeofday' and `select'.
8664 ** Function: SCM scm_internal_stack_catch (SCM tag,
8665 scm_catch_body_t body,
8667 scm_catch_handler_t handler,
8670 A new sibling to the other two C level `catch' functions
8671 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
8672 the stack to be saved automatically into the variable `the-last-stack'
8673 (scm_the_last_stack_var) on error. This is necessary if you want to
8674 use advanced error reporting, such as calling scm_display_error and
8675 scm_display_backtrace. (They both take a stack object as argument.)
8677 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
8679 scm_catch_handler_t handler,
8682 Spawns a new thread. It does a job similar to
8683 scm_call_with_new_thread but takes arguments more suitable when
8684 spawning threads from application C code.
8686 ** The hook scm_error_callback has been removed. It was originally
8687 intended as a way for the user to install his own error handler. But
8688 that method works badly since it intervenes between throw and catch,
8689 thereby changing the semantics of expressions like (catch #t ...).
8690 The correct way to do it is to use one of the C level catch functions
8691 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
8693 ** Removed functions:
8695 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
8696 scm_list_reverse, scm_list_reverse_x
8698 ** New macros: SCM_LISTn where n is one of the integers 0-9.
8700 These can be used for pretty list creation from C. The idea is taken
8701 from Erick Gallesio's STk.
8703 ** scm_array_map renamed to scm_array_map_x
8705 ** mbstrings are now removed
8707 This means that the type codes scm_tc7_mb_string and
8708 scm_tc7_mb_substring has been removed.
8710 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
8712 Since we no longer support multi-byte strings, these I/O functions
8713 have been simplified, and renamed. Here are their old names, and
8714 their new names and arguments:
8716 scm_gen_putc -> void scm_putc (int c, SCM port);
8717 scm_gen_puts -> void scm_puts (char *s, SCM port);
8718 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
8719 scm_gen_getc -> void scm_getc (SCM port);
8722 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
8724 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
8726 SCM_TYP7S now masks away the bit which distinguishes substrings from
8729 ** scm_catch_body_t: Backward incompatible change!
8731 Body functions to scm_internal_catch and friends do not any longer
8732 take a second argument. This is because it is no longer possible to
8733 pass a #f arg to catch.
8735 ** Calls to scm_protect_object and scm_unprotect now nest properly.
8737 The function scm_protect_object protects its argument from being freed
8738 by the garbage collector. scm_unprotect_object removes that
8741 These functions now nest properly. That is, for every object O, there
8742 is a counter which scm_protect_object(O) increments and
8743 scm_unprotect_object(O) decrements, if the counter is greater than
8744 zero. Every object's counter is zero when it is first created. If an
8745 object's counter is greater than zero, the garbage collector will not
8746 reclaim its storage.
8748 This allows you to use scm_protect_object in your code without
8749 worrying that some other function you call will call
8750 scm_unprotect_object, and allow it to be freed. Assuming that the
8751 functions you call are well-behaved, and unprotect only those objects
8752 they protect, you can follow the same rule and have confidence that
8753 objects will be freed only at appropriate times.
8756 Changes in Guile 1.2 (released Tuesday, June 24 1997):
8758 * Changes to the distribution
8760 ** Nightly snapshots are now available from ftp.red-bean.com.
8761 The old server, ftp.cyclic.com, has been relinquished to its rightful
8764 Nightly snapshots of the Guile development sources are now available via
8765 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
8767 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
8768 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
8770 ** To run Guile without installing it, the procedure has changed a bit.
8772 If you used a separate build directory to compile Guile, you'll need
8773 to include the build directory in SCHEME_LOAD_PATH, as well as the
8774 source directory. See the `INSTALL' file for examples.
8776 * Changes to the procedure for linking libguile with your programs
8778 ** The standard Guile load path for Scheme code now includes
8779 $(datadir)/guile (usually /usr/local/share/guile). This means that
8780 you can install your own Scheme files there, and Guile will find them.
8781 (Previous versions of Guile only checked a directory whose name
8782 contained the Guile version number, so you had to re-install or move
8783 your Scheme sources each time you installed a fresh version of Guile.)
8785 The load path also includes $(datadir)/guile/site; we recommend
8786 putting individual Scheme files there. If you want to install a
8787 package with multiple source files, create a directory for them under
8790 ** Guile 1.2 will now use the Rx regular expression library, if it is
8791 installed on your system. When you are linking libguile into your own
8792 programs, this means you will have to link against -lguile, -lqt (if
8793 you configured Guile with thread support), and -lrx.
8795 If you are using autoconf to generate configuration scripts for your
8796 application, the following lines should suffice to add the appropriate
8797 libraries to your link command:
8799 ### Find Rx, quickthreads and libguile.
8800 AC_CHECK_LIB(rx, main)
8801 AC_CHECK_LIB(qt, main)
8802 AC_CHECK_LIB(guile, scm_shell)
8804 The Guile 1.2 distribution does not contain sources for the Rx
8805 library, as Guile 1.0 did. If you want to use Rx, you'll need to
8806 retrieve it from a GNU FTP site and install it separately.
8808 * Changes to Scheme functions and syntax
8810 ** The dynamic linking features of Guile are now enabled by default.
8811 You can disable them by giving the `--disable-dynamic-linking' option
8814 (dynamic-link FILENAME)
8816 Find the object file denoted by FILENAME (a string) and link it
8817 into the running Guile application. When everything works out,
8818 return a Scheme object suitable for representing the linked object
8819 file. Otherwise an error is thrown. How object files are
8820 searched is system dependent.
8822 (dynamic-object? VAL)
8824 Determine whether VAL represents a dynamically linked object file.
8826 (dynamic-unlink DYNOBJ)
8828 Unlink the indicated object file from the application. DYNOBJ
8829 should be one of the values returned by `dynamic-link'.
8831 (dynamic-func FUNCTION DYNOBJ)
8833 Search the C function indicated by FUNCTION (a string or symbol)
8834 in DYNOBJ and return some Scheme object that can later be used
8835 with `dynamic-call' to actually call this function. Right now,
8836 these Scheme objects are formed by casting the address of the
8837 function to `long' and converting this number to its Scheme
8840 (dynamic-call FUNCTION DYNOBJ)
8842 Call the C function indicated by FUNCTION and DYNOBJ. The
8843 function is passed no arguments and its return value is ignored.
8844 When FUNCTION is something returned by `dynamic-func', call that
8845 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
8846 etc.), look it up in DYNOBJ; this is equivalent to
8848 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
8850 Interrupts are deferred while the C function is executing (with
8851 SCM_DEFER_INTS/SCM_ALLOW_INTS).
8853 (dynamic-args-call FUNCTION DYNOBJ ARGS)
8855 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
8856 some arguments and return its return value. The C function is
8857 expected to take two arguments and return an `int', just like
8860 int c_func (int argc, char **argv);
8862 ARGS must be a list of strings and is converted into an array of
8863 `char *'. The array is passed in ARGV and its size in ARGC. The
8864 return value is converted to a Scheme number and returned from the
8865 call to `dynamic-args-call'.
8867 When dynamic linking is disabled or not supported on your system,
8868 the above functions throw errors, but they are still available.
8870 Here is a small example that works on GNU/Linux:
8872 (define libc-obj (dynamic-link "libc.so"))
8873 (dynamic-args-call 'rand libc-obj '())
8875 See the file `libguile/DYNAMIC-LINKING' for additional comments.
8877 ** The #/ syntax for module names is depreciated, and will be removed
8878 in a future version of Guile. Instead of
8886 The latter syntax is more consistent with existing Lisp practice.
8888 ** Guile now does fancier printing of structures. Structures are the
8889 underlying implementation for records, which in turn are used to
8890 implement modules, so all of these object now print differently and in
8891 a more informative way.
8893 The Scheme printer will examine the builtin variable *struct-printer*
8894 whenever it needs to print a structure object. When this variable is
8895 not `#f' it is deemed to be a procedure and will be applied to the
8896 structure object and the output port. When *struct-printer* is `#f'
8897 or the procedure return `#f' the structure object will be printed in
8898 the boring #<struct 80458270> form.
8900 This hook is used by some routines in ice-9/boot-9.scm to implement
8901 type specific printing routines. Please read the comments there about
8904 One of the more specific uses of structs are records. The printing
8905 procedure that could be passed to MAKE-RECORD-TYPE is now actually
8906 called. It should behave like a *struct-printer* procedure (described
8909 ** Guile now supports a new R4RS-compliant syntax for keywords. A
8910 token of the form #:NAME, where NAME has the same syntax as a Scheme
8911 symbol, is the external representation of the keyword named NAME.
8912 Keyword objects print using this syntax as well, so values containing
8913 keyword objects can be read back into Guile. When used in an
8914 expression, keywords are self-quoting objects.
8916 Guile suports this read syntax, and uses this print syntax, regardless
8917 of the current setting of the `keyword' read option. The `keyword'
8918 read option only controls whether Guile recognizes the `:NAME' syntax,
8919 which is incompatible with R4RS. (R4RS says such token represent
8922 ** Guile has regular expression support again. Guile 1.0 included
8923 functions for matching regular expressions, based on the Rx library.
8924 In Guile 1.1, the Guile/Rx interface was removed to simplify the
8925 distribution, and thus Guile had no regular expression support. Guile
8926 1.2 again supports the most commonly used functions, and supports all
8927 of SCSH's regular expression functions.
8929 If your system does not include a POSIX regular expression library,
8930 and you have not linked Guile with a third-party regexp library such as
8931 Rx, these functions will not be available. You can tell whether your
8932 Guile installation includes regular expression support by checking
8933 whether the `*features*' list includes the `regex' symbol.
8935 *** regexp functions
8937 By default, Guile supports POSIX extended regular expressions. That
8938 means that the characters `(', `)', `+' and `?' are special, and must
8939 be escaped if you wish to match the literal characters.
8941 This regular expression interface was modeled after that implemented
8942 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
8943 with SCSH regular expressions.
8945 **** Function: string-match PATTERN STR [START]
8946 Compile the string PATTERN into a regular expression and compare
8947 it with STR. The optional numeric argument START specifies the
8948 position of STR at which to begin matching.
8950 `string-match' returns a "match structure" which describes what,
8951 if anything, was matched by the regular expression. *Note Match
8952 Structures::. If STR does not match PATTERN at all,
8953 `string-match' returns `#f'.
8955 Each time `string-match' is called, it must compile its PATTERN
8956 argument into a regular expression structure. This operation is
8957 expensive, which makes `string-match' inefficient if the same regular
8958 expression is used several times (for example, in a loop). For better
8959 performance, you can compile a regular expression in advance and then
8960 match strings against the compiled regexp.
8962 **** Function: make-regexp STR [FLAGS]
8963 Compile the regular expression described by STR, and return the
8964 compiled regexp structure. If STR does not describe a legal
8965 regular expression, `make-regexp' throws a
8966 `regular-expression-syntax' error.
8968 FLAGS may be the bitwise-or of one or more of the following:
8970 **** Constant: regexp/extended
8971 Use POSIX Extended Regular Expression syntax when interpreting
8972 STR. If not set, POSIX Basic Regular Expression syntax is used.
8973 If the FLAGS argument is omitted, we assume regexp/extended.
8975 **** Constant: regexp/icase
8976 Do not differentiate case. Subsequent searches using the
8977 returned regular expression will be case insensitive.
8979 **** Constant: regexp/newline
8980 Match-any-character operators don't match a newline.
8982 A non-matching list ([^...]) not containing a newline matches a
8985 Match-beginning-of-line operator (^) matches the empty string
8986 immediately after a newline, regardless of whether the FLAGS
8987 passed to regexp-exec contain regexp/notbol.
8989 Match-end-of-line operator ($) matches the empty string
8990 immediately before a newline, regardless of whether the FLAGS
8991 passed to regexp-exec contain regexp/noteol.
8993 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
8994 Match the compiled regular expression REGEXP against `str'. If
8995 the optional integer START argument is provided, begin matching
8996 from that position in the string. Return a match structure
8997 describing the results of the match, or `#f' if no match could be
9000 FLAGS may be the bitwise-or of one or more of the following:
9002 **** Constant: regexp/notbol
9003 The match-beginning-of-line operator always fails to match (but
9004 see the compilation flag regexp/newline above) This flag may be
9005 used when different portions of a string are passed to
9006 regexp-exec and the beginning of the string should not be
9007 interpreted as the beginning of the line.
9009 **** Constant: regexp/noteol
9010 The match-end-of-line operator always fails to match (but see the
9011 compilation flag regexp/newline above)
9013 **** Function: regexp? OBJ
9014 Return `#t' if OBJ is a compiled regular expression, or `#f'
9017 Regular expressions are commonly used to find patterns in one string
9018 and replace them with the contents of another string.
9020 **** Function: regexp-substitute PORT MATCH [ITEM...]
9021 Write to the output port PORT selected contents of the match
9022 structure MATCH. Each ITEM specifies what should be written, and
9023 may be one of the following arguments:
9025 * A string. String arguments are written out verbatim.
9027 * An integer. The submatch with that number is written.
9029 * The symbol `pre'. The portion of the matched string preceding
9030 the regexp match is written.
9032 * The symbol `post'. The portion of the matched string
9033 following the regexp match is written.
9035 PORT may be `#f', in which case nothing is written; instead,
9036 `regexp-substitute' constructs a string from the specified ITEMs
9039 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
9040 Similar to `regexp-substitute', but can be used to perform global
9041 substitutions on STR. Instead of taking a match structure as an
9042 argument, `regexp-substitute/global' takes two string arguments: a
9043 REGEXP string describing a regular expression, and a TARGET string
9044 which should be matched against this regular expression.
9046 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
9049 * A function may be supplied. When this function is called, it
9050 will be passed one argument: a match structure for a given
9051 regular expression match. It should return a string to be
9052 written out to PORT.
9054 * The `post' symbol causes `regexp-substitute/global' to recurse
9055 on the unmatched portion of STR. This *must* be supplied in
9056 order to perform global search-and-replace on STR; if it is
9057 not present among the ITEMs, then `regexp-substitute/global'
9058 will return after processing a single match.
9060 *** Match Structures
9062 A "match structure" is the object returned by `string-match' and
9063 `regexp-exec'. It describes which portion of a string, if any, matched
9064 the given regular expression. Match structures include: a reference to
9065 the string that was checked for matches; the starting and ending
9066 positions of the regexp match; and, if the regexp included any
9067 parenthesized subexpressions, the starting and ending positions of each
9070 In each of the regexp match functions described below, the `match'
9071 argument must be a match structure returned by a previous call to
9072 `string-match' or `regexp-exec'. Most of these functions return some
9073 information about the original target string that was matched against a
9074 regular expression; we will call that string TARGET for easy reference.
9076 **** Function: regexp-match? OBJ
9077 Return `#t' if OBJ is a match structure returned by a previous
9078 call to `regexp-exec', or `#f' otherwise.
9080 **** Function: match:substring MATCH [N]
9081 Return the portion of TARGET matched by subexpression number N.
9082 Submatch 0 (the default) represents the entire regexp match. If
9083 the regular expression as a whole matched, but the subexpression
9084 number N did not match, return `#f'.
9086 **** Function: match:start MATCH [N]
9087 Return the starting position of submatch number N.
9089 **** Function: match:end MATCH [N]
9090 Return the ending position of submatch number N.
9092 **** Function: match:prefix MATCH
9093 Return the unmatched portion of TARGET preceding the regexp match.
9095 **** Function: match:suffix MATCH
9096 Return the unmatched portion of TARGET following the regexp match.
9098 **** Function: match:count MATCH
9099 Return the number of parenthesized subexpressions from MATCH.
9100 Note that the entire regular expression match itself counts as a
9101 subexpression, and failed submatches are included in the count.
9103 **** Function: match:string MATCH
9104 Return the original TARGET string.
9106 *** Backslash Escapes
9108 Sometimes you will want a regexp to match characters like `*' or `$'
9109 exactly. For example, to check whether a particular string represents
9110 a menu entry from an Info node, it would be useful to match it against
9111 a regexp like `^* [^:]*::'. However, this won't work; because the
9112 asterisk is a metacharacter, it won't match the `*' at the beginning of
9113 the string. In this case, we want to make the first asterisk un-magic.
9115 You can do this by preceding the metacharacter with a backslash
9116 character `\'. (This is also called "quoting" the metacharacter, and
9117 is known as a "backslash escape".) When Guile sees a backslash in a
9118 regular expression, it considers the following glyph to be an ordinary
9119 character, no matter what special meaning it would ordinarily have.
9120 Therefore, we can make the above example work by changing the regexp to
9121 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
9122 to match only a single asterisk in the target string.
9124 Since the backslash is itself a metacharacter, you may force a
9125 regexp to match a backslash in the target string by preceding the
9126 backslash with itself. For example, to find variable references in a
9127 TeX program, you might want to find occurrences of the string `\let\'
9128 followed by any number of alphabetic characters. The regular expression
9129 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
9130 each match a single backslash in the target string.
9132 **** Function: regexp-quote STR
9133 Quote each special character found in STR with a backslash, and
9134 return the resulting string.
9136 *Very important:* Using backslash escapes in Guile source code (as
9137 in Emacs Lisp or C) can be tricky, because the backslash character has
9138 special meaning for the Guile reader. For example, if Guile encounters
9139 the character sequence `\n' in the middle of a string while processing
9140 Scheme code, it replaces those characters with a newline character.
9141 Similarly, the character sequence `\t' is replaced by a horizontal tab.
9142 Several of these "escape sequences" are processed by the Guile reader
9143 before your code is executed. Unrecognized escape sequences are
9144 ignored: if the characters `\*' appear in a string, they will be
9145 translated to the single character `*'.
9147 This translation is obviously undesirable for regular expressions,
9148 since we want to be able to include backslashes in a string in order to
9149 escape regexp metacharacters. Therefore, to make sure that a backslash
9150 is preserved in a string in your Guile program, you must use *two*
9151 consecutive backslashes:
9153 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
9155 The string in this example is preprocessed by the Guile reader before
9156 any code is executed. The resulting argument to `make-regexp' is the
9157 string `^\* [^:]*', which is what we really want.
9159 This also means that in order to write a regular expression that
9160 matches a single backslash character, the regular expression string in
9161 the source code must include *four* backslashes. Each consecutive pair
9162 of backslashes gets translated by the Guile reader to a single
9163 backslash, and the resulting double-backslash is interpreted by the
9164 regexp engine as matching a single backslash character. Hence:
9166 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
9168 The reason for the unwieldiness of this syntax is historical. Both
9169 regular expression pattern matchers and Unix string processing systems
9170 have traditionally used backslashes with the special meanings described
9171 above. The POSIX regular expression specification and ANSI C standard
9172 both require these semantics. Attempting to abandon either convention
9173 would cause other kinds of compatibility problems, possibly more severe
9174 ones. Therefore, without extending the Scheme reader to support
9175 strings with different quoting conventions (an ungainly and confusing
9176 extension when implemented in other languages), we must adhere to this
9177 cumbersome escape syntax.
9179 * Changes to the gh_ interface
9181 * Changes to the scm_ interface
9183 * Changes to system call interfaces:
9185 ** The value returned by `raise' is now unspecified. It throws an exception
9188 *** A new procedure `sigaction' can be used to install signal handlers
9190 (sigaction signum [action] [flags])
9192 signum is the signal number, which can be specified using the value
9195 If action is omitted, sigaction returns a pair: the CAR is the current
9196 signal hander, which will be either an integer with the value SIG_DFL
9197 (default action) or SIG_IGN (ignore), or the Scheme procedure which
9198 handles the signal, or #f if a non-Scheme procedure handles the
9199 signal. The CDR contains the current sigaction flags for the handler.
9201 If action is provided, it is installed as the new handler for signum.
9202 action can be a Scheme procedure taking one argument, or the value of
9203 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
9204 whatever signal handler was installed before sigaction was first used.
9205 Flags can optionally be specified for the new handler (SA_RESTART is
9206 always used if the system provides it, so need not be specified.) The
9207 return value is a pair with information about the old handler as
9210 This interface does not provide access to the "signal blocking"
9211 facility. Maybe this is not needed, since the thread support may
9212 provide solutions to the problem of consistent access to data
9215 *** A new procedure `flush-all-ports' is equivalent to running
9216 `force-output' on every port open for output.
9218 ** Guile now provides information on how it was built, via the new
9219 global variable, %guile-build-info. This variable records the values
9220 of the standard GNU makefile directory variables as an assocation
9221 list, mapping variable names (symbols) onto directory paths (strings).
9222 For example, to find out where the Guile link libraries were
9223 installed, you can say:
9225 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
9228 * Changes to the scm_ interface
9230 ** The new function scm_handle_by_message_noexit is just like the
9231 existing scm_handle_by_message function, except that it doesn't call
9232 exit to terminate the process. Instead, it prints a message and just
9233 returns #f. This might be a more appropriate catch-all handler for
9234 new dynamic roots and threads.
9237 Changes in Guile 1.1 (released Friday, May 16 1997):
9239 * Changes to the distribution.
9241 The Guile 1.0 distribution has been split up into several smaller
9243 guile-core --- the Guile interpreter itself.
9244 guile-tcltk --- the interface between the Guile interpreter and
9245 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
9246 is a toolkit for building graphical user interfaces.
9247 guile-rgx-ctax --- the interface between Guile and the Rx regular
9248 expression matcher, and the translator for the Ctax
9249 programming language. These are packaged together because the
9250 Ctax translator uses Rx to parse Ctax source code.
9252 This NEWS file describes the changes made to guile-core since the 1.0
9255 We no longer distribute the documentation, since it was either out of
9256 date, or incomplete. As soon as we have current documentation, we
9261 * Changes to the stand-alone interpreter
9263 ** guile now accepts command-line arguments compatible with SCSH, Olin
9264 Shivers' Scheme Shell.
9266 In general, arguments are evaluated from left to right, but there are
9267 exceptions. The following switches stop argument processing, and
9268 stash all remaining command-line arguments as the value returned by
9269 the (command-line) function.
9270 -s SCRIPT load Scheme source code from FILE, and exit
9271 -c EXPR evalute Scheme expression EXPR, and exit
9272 -- stop scanning arguments; run interactively
9274 The switches below are processed as they are encountered.
9275 -l FILE load Scheme source code from FILE
9276 -e FUNCTION after reading script, apply FUNCTION to
9277 command line arguments
9278 -ds do -s script at this point
9279 --emacs enable Emacs protocol (experimental)
9280 -h, --help display this help and exit
9281 -v, --version display version information and exit
9282 \ read arguments from following script lines
9284 So, for example, here is a Guile script named `ekko' (thanks, Olin)
9285 which re-implements the traditional "echo" command:
9287 #!/usr/local/bin/guile -s
9290 (map (lambda (arg) (display arg) (display " "))
9294 (main (command-line))
9296 Suppose we invoke this script as follows:
9298 ekko a speckled gecko
9300 Through the magic of Unix script processing (triggered by the `#!'
9301 token at the top of the file), /usr/local/bin/guile receives the
9302 following list of command-line arguments:
9304 ("-s" "./ekko" "a" "speckled" "gecko")
9306 Unix inserts the name of the script after the argument specified on
9307 the first line of the file (in this case, "-s"), and then follows that
9308 with the arguments given to the script. Guile loads the script, which
9309 defines the `main' function, and then applies it to the list of
9310 remaining command-line arguments, ("a" "speckled" "gecko").
9312 In Unix, the first line of a script file must take the following form:
9314 #!INTERPRETER ARGUMENT
9316 where INTERPRETER is the absolute filename of the interpreter
9317 executable, and ARGUMENT is a single command-line argument to pass to
9320 You may only pass one argument to the interpreter, and its length is
9321 limited. These restrictions can be annoying to work around, so Guile
9322 provides a general mechanism (borrowed from, and compatible with,
9323 SCSH) for circumventing them.
9325 If the ARGUMENT in a Guile script is a single backslash character,
9326 `\', Guile will open the script file, parse arguments from its second
9327 and subsequent lines, and replace the `\' with them. So, for example,
9328 here is another implementation of the `ekko' script:
9330 #!/usr/local/bin/guile \
9334 (for-each (lambda (arg) (display arg) (display " "))
9338 If the user invokes this script as follows:
9340 ekko a speckled gecko
9342 Unix expands this into
9344 /usr/local/bin/guile \ ekko a speckled gecko
9346 When Guile sees the `\' argument, it replaces it with the arguments
9347 read from the second line of the script, producing:
9349 /usr/local/bin/guile -e main -s ekko a speckled gecko
9351 This tells Guile to load the `ekko' script, and apply the function
9352 `main' to the argument list ("a" "speckled" "gecko").
9354 Here is how Guile parses the command-line arguments:
9355 - Each space character terminates an argument. This means that two
9356 spaces in a row introduce an empty-string argument.
9357 - The tab character is not permitted (unless you quote it with the
9358 backslash character, as described below), to avoid confusion.
9359 - The newline character terminates the sequence of arguments, and will
9360 also terminate a final non-empty argument. (However, a newline
9361 following a space will not introduce a final empty-string argument;
9362 it only terminates the argument list.)
9363 - The backslash character is the escape character. It escapes
9364 backslash, space, tab, and newline. The ANSI C escape sequences
9365 like \n and \t are also supported. These produce argument
9366 constituents; the two-character combination \n doesn't act like a
9367 terminating newline. The escape sequence \NNN for exactly three
9368 octal digits reads as the character whose ASCII code is NNN. As
9369 above, characters produced this way are argument constituents.
9370 Backslash followed by other characters is not allowed.
9372 * Changes to the procedure for linking libguile with your programs
9374 ** Guile now builds and installs a shared guile library, if your
9375 system support shared libraries. (It still builds a static library on
9376 all systems.) Guile automatically detects whether your system
9377 supports shared libraries. To prevent Guile from buildisg shared
9378 libraries, pass the `--disable-shared' flag to the configure script.
9380 Guile takes longer to compile when it builds shared libraries, because
9381 it must compile every file twice --- once to produce position-
9382 independent object code, and once to produce normal object code.
9384 ** The libthreads library has been merged into libguile.
9386 To link a program against Guile, you now need only link against
9387 -lguile and -lqt; -lthreads is no longer needed. If you are using
9388 autoconf to generate configuration scripts for your application, the
9389 following lines should suffice to add the appropriate libraries to
9392 ### Find quickthreads and libguile.
9393 AC_CHECK_LIB(qt, main)
9394 AC_CHECK_LIB(guile, scm_shell)
9396 * Changes to Scheme functions
9398 ** Guile Scheme's special syntax for keyword objects is now optional,
9399 and disabled by default.
9401 The syntax variation from R4RS made it difficult to port some
9402 interesting packages to Guile. The routines which accepted keyword
9403 arguments (mostly in the module system) have been modified to also
9404 accept symbols whose names begin with `:'.
9406 To change the keyword syntax, you must first import the (ice-9 debug)
9408 (use-modules (ice-9 debug))
9410 Then you can enable the keyword syntax as follows:
9411 (read-set! keywords 'prefix)
9413 To disable keyword syntax, do this:
9414 (read-set! keywords #f)
9416 ** Many more primitive functions accept shared substrings as
9417 arguments. In the past, these functions required normal, mutable
9418 strings as arguments, although they never made use of this
9421 ** The uniform array functions now operate on byte vectors. These
9422 functions are `array-fill!', `serial-array-copy!', `array-copy!',
9423 `serial-array-map', `array-map', `array-for-each', and
9426 ** The new functions `trace' and `untrace' implement simple debugging
9427 support for Scheme functions.
9429 The `trace' function accepts any number of procedures as arguments,
9430 and tells the Guile interpreter to display each procedure's name and
9431 arguments each time the procedure is invoked. When invoked with no
9432 arguments, `trace' returns the list of procedures currently being
9435 The `untrace' function accepts any number of procedures as arguments,
9436 and tells the Guile interpreter not to trace them any more. When
9437 invoked with no arguments, `untrace' untraces all curretly traced
9440 The tracing in Guile has an advantage over most other systems: we
9441 don't create new procedure objects, but mark the procedure objects
9442 themselves. This means that anonymous and internal procedures can be
9445 ** The function `assert-repl-prompt' has been renamed to
9446 `set-repl-prompt!'. It takes one argument, PROMPT.
9447 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
9448 - If PROMPT is a string, we use it as a prompt.
9449 - If PROMPT is a procedure accepting no arguments, we call it, and
9450 display the result as a prompt.
9451 - Otherwise, we display "> ".
9453 ** The new function `eval-string' reads Scheme expressions from a
9454 string and evaluates them, returning the value of the last expression
9455 in the string. If the string contains no expressions, it returns an
9458 ** The new function `thunk?' returns true iff its argument is a
9459 procedure of zero arguments.
9461 ** `defined?' is now a builtin function, instead of syntax. This
9462 means that its argument should be quoted. It returns #t iff its
9463 argument is bound in the current module.
9465 ** The new syntax `use-modules' allows you to add new modules to your
9466 environment without re-typing a complete `define-module' form. It
9467 accepts any number of module names as arguments, and imports their
9468 public bindings into the current module.
9470 ** The new function (module-defined? NAME MODULE) returns true iff
9471 NAME, a symbol, is defined in MODULE, a module object.
9473 ** The new function `builtin-bindings' creates and returns a hash
9474 table containing copies of all the root module's bindings.
9476 ** The new function `builtin-weak-bindings' does the same as
9477 `builtin-bindings', but creates a doubly-weak hash table.
9479 ** The `equal?' function now considers variable objects to be
9480 equivalent if they have the same name and the same value.
9482 ** The new function `command-line' returns the command-line arguments
9483 given to Guile, as a list of strings.
9485 When using guile as a script interpreter, `command-line' returns the
9486 script's arguments; those processed by the interpreter (like `-s' or
9487 `-c') are omitted. (In other words, you get the normal, expected
9488 behavior.) Any application that uses scm_shell to process its
9489 command-line arguments gets this behavior as well.
9491 ** The new function `load-user-init' looks for a file called `.guile'
9492 in the user's home directory, and loads it if it exists. This is
9493 mostly for use by the code generated by scm_compile_shell_switches,
9494 but we thought it might also be useful in other circumstances.
9496 ** The new function `log10' returns the base-10 logarithm of its
9499 ** Changes to I/O functions
9501 *** The functions `read', `primitive-load', `read-and-eval!', and
9502 `primitive-load-path' no longer take optional arguments controlling
9503 case insensitivity and a `#' parser.
9505 Case sensitivity is now controlled by a read option called
9506 `case-insensitive'. The user can add new `#' syntaxes with the
9507 `read-hash-extend' function (see below).
9509 *** The new function `read-hash-extend' allows the user to change the
9510 syntax of Guile Scheme in a somewhat controlled way.
9512 (read-hash-extend CHAR PROC)
9513 When parsing S-expressions, if we read a `#' character followed by
9514 the character CHAR, use PROC to parse an object from the stream.
9515 If PROC is #f, remove any parsing procedure registered for CHAR.
9517 The reader applies PROC to two arguments: CHAR and an input port.
9519 *** The new functions read-delimited and read-delimited! provide a
9520 general mechanism for doing delimited input on streams.
9522 (read-delimited DELIMS [PORT HANDLE-DELIM])
9523 Read until we encounter one of the characters in DELIMS (a string),
9524 or end-of-file. PORT is the input port to read from; it defaults to
9525 the current input port. The HANDLE-DELIM parameter determines how
9526 the terminating character is handled; it should be one of the
9529 'trim omit delimiter from result
9530 'peek leave delimiter character in input stream
9531 'concat append delimiter character to returned value
9532 'split return a pair: (RESULT . TERMINATOR)
9534 HANDLE-DELIM defaults to 'peek.
9536 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
9537 A side-effecting variant of `read-delimited'.
9539 The data is written into the string BUF at the indices in the
9540 half-open interval [START, END); the default interval is the whole
9541 string: START = 0 and END = (string-length BUF). The values of
9542 START and END must specify a well-defined interval in BUF, i.e.
9543 0 <= START <= END <= (string-length BUF).
9545 It returns NBYTES, the number of bytes read. If the buffer filled
9546 up without a delimiter character being found, it returns #f. If the
9547 port is at EOF when the read starts, it returns the EOF object.
9549 If an integer is returned (i.e., the read is successfully terminated
9550 by reading a delimiter character), then the HANDLE-DELIM parameter
9551 determines how to handle the terminating character. It is described
9552 above, and defaults to 'peek.
9554 (The descriptions of these functions were borrowed from the SCSH
9555 manual, by Olin Shivers and Brian Carlstrom.)
9557 *** The `%read-delimited!' function is the primitive used to implement
9558 `read-delimited' and `read-delimited!'.
9560 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
9562 This returns a pair of values: (TERMINATOR . NUM-READ).
9563 - TERMINATOR describes why the read was terminated. If it is a
9564 character or the eof object, then that is the value that terminated
9565 the read. If it is #f, the function filled the buffer without finding
9566 a delimiting character.
9567 - NUM-READ is the number of characters read into BUF.
9569 If the read is successfully terminated by reading a delimiter
9570 character, then the gobble? parameter determines what to do with the
9571 terminating character. If true, the character is removed from the
9572 input stream; if false, the character is left in the input stream
9573 where a subsequent read operation will retrieve it. In either case,
9574 the character is also the first value returned by the procedure call.
9576 (The descriptions of this function was borrowed from the SCSH manual,
9577 by Olin Shivers and Brian Carlstrom.)
9579 *** The `read-line' and `read-line!' functions have changed; they now
9580 trim the terminator by default; previously they appended it to the
9581 returned string. For the old behavior, use (read-line PORT 'concat).
9583 *** The functions `uniform-array-read!' and `uniform-array-write!' now
9584 take new optional START and END arguments, specifying the region of
9585 the array to read and write.
9587 *** The `ungetc-char-ready?' function has been removed. We feel it's
9588 inappropriate for an interface to expose implementation details this
9591 ** Changes to the Unix library and system call interface
9593 *** The new fcntl function provides access to the Unix `fcntl' system
9596 (fcntl PORT COMMAND VALUE)
9597 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
9598 Values for COMMAND are:
9600 F_DUPFD duplicate a file descriptor
9601 F_GETFD read the descriptor's close-on-exec flag
9602 F_SETFD set the descriptor's close-on-exec flag to VALUE
9603 F_GETFL read the descriptor's flags, as set on open
9604 F_SETFL set the descriptor's flags, as set on open to VALUE
9605 F_GETOWN return the process ID of a socket's owner, for SIGIO
9606 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
9607 FD_CLOEXEC not sure what this is
9609 For details, see the documentation for the fcntl system call.
9611 *** The arguments to `select' have changed, for compatibility with
9612 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
9613 expected behavior. The MILLISECONDS parameter has been changed to
9614 MICROSECONDS, to more closely resemble the underlying system call.
9615 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
9616 corresponding return set will be the same.
9618 *** The arguments to the `mknod' system call have changed. They are
9621 (mknod PATH TYPE PERMS DEV)
9622 Create a new file (`node') in the file system. PATH is the name of
9623 the file to create. TYPE is the kind of file to create; it should
9624 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
9625 permission bits to give the newly created file. If TYPE is
9626 'block-special or 'char-special, DEV specifies which device the
9627 special file refers to; its interpretation depends on the kind of
9628 special file being created.
9630 *** The `fork' function has been renamed to `primitive-fork', to avoid
9631 clashing with various SCSH forks.
9633 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
9634 and `recvfrom!'. They no longer accept a size for a second argument;
9635 you must pass a string to hold the received value. They no longer
9636 return the buffer. Instead, `recv' returns the length of the message
9637 received, and `recvfrom' returns a pair containing the packet's length
9638 and originating address.
9640 *** The file descriptor datatype has been removed, as have the
9641 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
9642 We plan to replace these functions with a SCSH-compatible interface.
9644 *** The `create' function has been removed; it's just a special case
9647 *** There are new functions to break down process termination status
9648 values. In the descriptions below, STATUS is a value returned by
9651 (status:exit-val STATUS)
9652 If the child process exited normally, this function returns the exit
9653 code for the child process (i.e., the value passed to exit, or
9654 returned from main). If the child process did not exit normally,
9655 this function returns #f.
9657 (status:stop-sig STATUS)
9658 If the child process was suspended by a signal, this function
9659 returns the signal that suspended the child. Otherwise, it returns
9662 (status:term-sig STATUS)
9663 If the child process terminated abnormally, this function returns
9664 the signal that terminated the child. Otherwise, this function
9667 POSIX promises that exactly one of these functions will return true on
9668 a valid STATUS value.
9670 These functions are compatible with SCSH.
9672 *** There are new accessors and setters for the broken-out time vectors
9673 returned by `localtime', `gmtime', and that ilk. They are:
9675 Component Accessor Setter
9676 ========================= ============ ============
9677 seconds tm:sec set-tm:sec
9678 minutes tm:min set-tm:min
9679 hours tm:hour set-tm:hour
9680 day of the month tm:mday set-tm:mday
9681 month tm:mon set-tm:mon
9682 year tm:year set-tm:year
9683 day of the week tm:wday set-tm:wday
9684 day in the year tm:yday set-tm:yday
9685 daylight saving time tm:isdst set-tm:isdst
9686 GMT offset, seconds tm:gmtoff set-tm:gmtoff
9687 name of time zone tm:zone set-tm:zone
9689 *** There are new accessors for the vectors returned by `uname',
9690 describing the host system:
9693 ============================================== ================
9694 name of the operating system implementation utsname:sysname
9695 network name of this machine utsname:nodename
9696 release level of the operating system utsname:release
9697 version level of the operating system utsname:version
9698 machine hardware platform utsname:machine
9700 *** There are new accessors for the vectors returned by `getpw',
9701 `getpwnam', `getpwuid', and `getpwent', describing entries from the
9702 system's user database:
9705 ====================== =================
9706 user name passwd:name
9707 user password passwd:passwd
9710 real name passwd:gecos
9711 home directory passwd:dir
9712 shell program passwd:shell
9714 *** There are new accessors for the vectors returned by `getgr',
9715 `getgrnam', `getgrgid', and `getgrent', describing entries from the
9716 system's group database:
9719 ======================= ============
9720 group name group:name
9721 group password group:passwd
9723 group members group:mem
9725 *** There are new accessors for the vectors returned by `gethost',
9726 `gethostbyaddr', `gethostbyname', and `gethostent', describing
9730 ========================= ===============
9731 official name of host hostent:name
9732 alias list hostent:aliases
9733 host address type hostent:addrtype
9734 length of address hostent:length
9735 list of addresses hostent:addr-list
9737 *** There are new accessors for the vectors returned by `getnet',
9738 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
9742 ========================= ===============
9743 official name of net netent:name
9744 alias list netent:aliases
9745 net number type netent:addrtype
9746 net number netent:net
9748 *** There are new accessors for the vectors returned by `getproto',
9749 `getprotobyname', `getprotobynumber', and `getprotoent', describing
9753 ========================= ===============
9754 official protocol name protoent:name
9755 alias list protoent:aliases
9756 protocol number protoent:proto
9758 *** There are new accessors for the vectors returned by `getserv',
9759 `getservbyname', `getservbyport', and `getservent', describing
9763 ========================= ===============
9764 official service name servent:name
9765 alias list servent:aliases
9766 port number servent:port
9767 protocol to use servent:proto
9769 *** There are new accessors for the sockaddr structures returned by
9770 `accept', `getsockname', `getpeername', `recvfrom!':
9773 ======================================== ===============
9774 address format (`family') sockaddr:fam
9775 path, for file domain addresses sockaddr:path
9776 address, for internet domain addresses sockaddr:addr
9777 TCP or UDP port, for internet sockaddr:port
9779 *** The `getpwent', `getgrent', `gethostent', `getnetent',
9780 `getprotoent', and `getservent' functions now return #f at the end of
9781 the user database. (They used to throw an exception.)
9783 Note that calling MUMBLEent function is equivalent to calling the
9784 corresponding MUMBLE function with no arguments.
9786 *** The `setpwent', `setgrent', `sethostent', `setnetent',
9787 `setprotoent', and `setservent' routines now take no arguments.
9789 *** The `gethost', `getproto', `getnet', and `getserv' functions now
9790 provide more useful information when they throw an exception.
9792 *** The `lnaof' function has been renamed to `inet-lnaof'.
9794 *** Guile now claims to have the `current-time' feature.
9796 *** The `mktime' function now takes an optional second argument ZONE,
9797 giving the time zone to use for the conversion. ZONE should be a
9798 string, in the same format as expected for the "TZ" environment variable.
9800 *** The `strptime' function now returns a pair (TIME . COUNT), where
9801 TIME is the parsed time as a vector, and COUNT is the number of
9802 characters from the string left unparsed. This function used to
9803 return the remaining characters as a string.
9805 *** The `gettimeofday' function has replaced the old `time+ticks' function.
9806 The return value is now (SECONDS . MICROSECONDS); the fractional
9807 component is no longer expressed in "ticks".
9809 *** The `ticks/sec' constant has been removed, in light of the above change.
9811 * Changes to the gh_ interface
9813 ** gh_eval_str() now returns an SCM object which is the result of the
9816 ** gh_scm2str() now copies the Scheme data to a caller-provided C
9819 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
9820 and returns the array
9822 ** gh_scm2str0() is gone: there is no need to distinguish
9823 null-terminated from non-null-terminated, since gh_scm2newstr() allows
9824 the user to interpret the data both ways.
9826 * Changes to the scm_ interface
9828 ** The new function scm_symbol_value0 provides an easy way to get a
9829 symbol's value from C code:
9831 SCM scm_symbol_value0 (char *NAME)
9832 Return the value of the symbol named by the null-terminated string
9833 NAME in the current module. If the symbol named NAME is unbound in
9834 the current module, return SCM_UNDEFINED.
9836 ** The new function scm_sysintern0 creates new top-level variables,
9837 without assigning them a value.
9839 SCM scm_sysintern0 (char *NAME)
9840 Create a new Scheme top-level variable named NAME. NAME is a
9841 null-terminated string. Return the variable's value cell.
9843 ** The function scm_internal_catch is the guts of catch. It handles
9844 all the mechanics of setting up a catch target, invoking the catch
9845 body, and perhaps invoking the handler if the body does a throw.
9847 The function is designed to be usable from C code, but is general
9848 enough to implement all the semantics Guile Scheme expects from throw.
9850 TAG is the catch tag. Typically, this is a symbol, but this function
9851 doesn't actually care about that.
9853 BODY is a pointer to a C function which runs the body of the catch;
9854 this is the code you can throw from. We call it like this:
9855 BODY (BODY_DATA, JMPBUF)
9857 BODY_DATA is just the BODY_DATA argument we received; we pass it
9858 through to BODY as its first argument. The caller can make
9859 BODY_DATA point to anything useful that BODY might need.
9860 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
9861 which we have just created and initialized.
9863 HANDLER is a pointer to a C function to deal with a throw to TAG,
9864 should one occur. We call it like this:
9865 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
9867 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
9868 same idea as BODY_DATA above.
9869 THROWN_TAG is the tag that the user threw to; usually this is
9870 TAG, but it could be something else if TAG was #t (i.e., a
9871 catch-all), or the user threw to a jmpbuf.
9872 THROW_ARGS is the list of arguments the user passed to the THROW
9875 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
9876 is just a pointer we pass through to HANDLER. We don't actually
9877 use either of those pointers otherwise ourselves. The idea is
9878 that, if our caller wants to communicate something to BODY or
9879 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
9880 HANDLER can then use. Think of it as a way to make BODY and
9881 HANDLER closures, not just functions; MUMBLE_DATA points to the
9884 Of course, it's up to the caller to make sure that any data a
9885 MUMBLE_DATA needs is protected from GC. A common way to do this is
9886 to make MUMBLE_DATA a pointer to data stored in an automatic
9887 structure variable; since the collector must scan the stack for
9888 references anyway, this assures that any references in MUMBLE_DATA
9891 ** The new function scm_internal_lazy_catch is exactly like
9892 scm_internal_catch, except:
9894 - It does not unwind the stack (this is the major difference).
9895 - If handler returns, its value is returned from the throw.
9896 - BODY always receives #f as its JMPBUF argument (since there's no
9897 jmpbuf associated with a lazy catch, because we don't unwind the
9900 ** scm_body_thunk is a new body function you can pass to
9901 scm_internal_catch if you want the body to be like Scheme's `catch'
9902 --- a thunk, or a function of one argument if the tag is #f.
9904 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
9905 contains the Scheme procedure to invoke as the body, and the tag
9906 we're catching. If the tag is #f, then we pass JMPBUF (created by
9907 scm_internal_catch) to the body procedure; otherwise, the body gets
9910 ** scm_handle_by_proc is a new handler function you can pass to
9911 scm_internal_catch if you want the handler to act like Scheme's catch
9912 --- call a procedure with the tag and the throw arguments.
9914 If the user does a throw to this catch, this function runs a handler
9915 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
9916 variable holding the Scheme procedure object to invoke. It ought to
9917 be a pointer to an automatic variable (i.e., one living on the stack),
9918 or the procedure object should be otherwise protected from GC.
9920 ** scm_handle_by_message is a new handler function to use with
9921 `scm_internal_catch' if you want Guile to print a message and die.
9922 It's useful for dealing with throws to uncaught keys at the top level.
9924 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
9925 message header to print; if zero, we use "guile" instead. That
9926 text is followed by a colon, then the message described by ARGS.
9928 ** The return type of scm_boot_guile is now void; the function does
9929 not return a value, and indeed, never returns at all.
9931 ** The new function scm_shell makes it easy for user applications to
9932 process command-line arguments in a way that is compatible with the
9933 stand-alone guile interpreter (which is in turn compatible with SCSH,
9936 To use the scm_shell function, first initialize any guile modules
9937 linked into your application, and then call scm_shell with the values
9938 of ARGC and ARGV your `main' function received. scm_shell will add
9939 any SCSH-style meta-arguments from the top of the script file to the
9940 argument vector, and then process the command-line arguments. This
9941 generally means loading a script file or starting up an interactive
9942 command interpreter. For details, see "Changes to the stand-alone
9945 ** The new functions scm_get_meta_args and scm_count_argv help you
9946 implement the SCSH-style meta-argument, `\'.
9948 char **scm_get_meta_args (int ARGC, char **ARGV)
9949 If the second element of ARGV is a string consisting of a single
9950 backslash character (i.e. "\\" in Scheme notation), open the file
9951 named by the following argument, parse arguments from it, and return
9952 the spliced command line. The returned array is terminated by a
9955 For details of argument parsing, see above, under "guile now accepts
9956 command-line arguments compatible with SCSH..."
9958 int scm_count_argv (char **ARGV)
9959 Count the arguments in ARGV, assuming it is terminated by a null
9962 For an example of how these functions might be used, see the source
9963 code for the function scm_shell in libguile/script.c.
9965 You will usually want to use scm_shell instead of calling this
9968 ** The new function scm_compile_shell_switches turns an array of
9969 command-line arguments into Scheme code to carry out the actions they
9970 describe. Given ARGC and ARGV, it returns a Scheme expression to
9971 evaluate, and calls scm_set_program_arguments to make any remaining
9972 command-line arguments available to the Scheme code. For example,
9973 given the following arguments:
9975 -e main -s ekko a speckled gecko
9977 scm_set_program_arguments will return the following expression:
9979 (begin (load "ekko") (main (command-line)) (quit))
9981 You will usually want to use scm_shell instead of calling this
9984 ** The function scm_shell_usage prints a usage message appropriate for
9985 an interpreter that uses scm_compile_shell_switches to handle its
9986 command-line arguments.
9988 void scm_shell_usage (int FATAL, char *MESSAGE)
9989 Print a usage message to the standard error output. If MESSAGE is
9990 non-zero, write it before the usage message, followed by a newline.
9991 If FATAL is non-zero, exit the process, using FATAL as the
9992 termination status. (If you want to be compatible with Guile,
9993 always use 1 as the exit status when terminating due to command-line
9996 You will usually want to use scm_shell instead of calling this
9999 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
10000 expressions. It used to return SCM_EOL. Earth-shattering.
10002 ** The macros for declaring scheme objects in C code have been
10003 rearranged slightly. They are now:
10005 SCM_SYMBOL (C_NAME, SCHEME_NAME)
10006 Declare a static SCM variable named C_NAME, and initialize it to
10007 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
10008 be a C identifier, and SCHEME_NAME should be a C string.
10010 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
10011 Just like SCM_SYMBOL, but make C_NAME globally visible.
10013 SCM_VCELL (C_NAME, SCHEME_NAME)
10014 Create a global variable at the Scheme level named SCHEME_NAME.
10015 Declare a static SCM variable named C_NAME, and initialize it to
10016 point to the Scheme variable's value cell.
10018 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
10019 Just like SCM_VCELL, but make C_NAME globally visible.
10021 The `guile-snarf' script writes initialization code for these macros
10022 to its standard output, given C source code as input.
10024 The SCM_GLOBAL macro is gone.
10026 ** The scm_read_line and scm_read_line_x functions have been replaced
10027 by Scheme code based on the %read-delimited! procedure (known to C
10028 code as scm_read_delimited_x). See its description above for more
10031 ** The function scm_sys_open has been renamed to scm_open. It now
10032 returns a port instead of an FD object.
10034 * The dynamic linking support has changed. For more information, see
10035 libguile/DYNAMIC-LINKING.
10040 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
10043 * Changes to the 'guile' program:
10045 ** Guile now loads some new files when it starts up. Guile first
10046 searches the load path for init.scm, and loads it if found. Then, if
10047 Guile is not being used to execute a script, and the user's home
10048 directory contains a file named `.guile', Guile loads that.
10050 ** You can now use Guile as a shell script interpreter.
10052 To paraphrase the SCSH manual:
10054 When Unix tries to execute an executable file whose first two
10055 characters are the `#!', it treats the file not as machine code to
10056 be directly executed by the native processor, but as source code
10057 to be executed by some interpreter. The interpreter to use is
10058 specified immediately after the #! sequence on the first line of
10059 the source file. The kernel reads in the name of the interpreter,
10060 and executes that instead. It passes the interpreter the source
10061 filename as its first argument, with the original arguments
10062 following. Consult the Unix man page for the `exec' system call
10063 for more information.
10065 Now you can use Guile as an interpreter, using a mechanism which is a
10066 compatible subset of that provided by SCSH.
10068 Guile now recognizes a '-s' command line switch, whose argument is the
10069 name of a file of Scheme code to load. It also treats the two
10070 characters `#!' as the start of a comment, terminated by `!#'. Thus,
10071 to make a file of Scheme code directly executable by Unix, insert the
10072 following two lines at the top of the file:
10074 #!/usr/local/bin/guile -s
10077 Guile treats the argument of the `-s' command-line switch as the name
10078 of a file of Scheme code to load, and treats the sequence `#!' as the
10079 start of a block comment, terminated by `!#'.
10081 For example, here's a version of 'echo' written in Scheme:
10083 #!/usr/local/bin/guile -s
10085 (let loop ((args (cdr (program-arguments))))
10088 (display (car args))
10089 (if (pair? (cdr args))
10091 (loop (cdr args)))))
10094 Why does `#!' start a block comment terminated by `!#', instead of the
10095 end of the line? That is the notation SCSH uses, and although we
10096 don't yet support the other SCSH features that motivate that choice,
10097 we would like to be backward-compatible with any existing Guile
10098 scripts once we do. Furthermore, if the path to Guile on your system
10099 is too long for your kernel, you can start the script with this
10103 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
10106 Note that some very old Unix systems don't support the `#!' syntax.
10109 ** You can now run Guile without installing it.
10111 Previous versions of the interactive Guile interpreter (`guile')
10112 couldn't start up unless Guile's Scheme library had been installed;
10113 they used the value of the environment variable `SCHEME_LOAD_PATH'
10114 later on in the startup process, but not to find the startup code
10115 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
10118 To run Guile without installing it, build it in the normal way, and
10119 then set the environment variable `SCHEME_LOAD_PATH' to a
10120 colon-separated list of directories, including the top-level directory
10121 of the Guile sources. For example, if you unpacked Guile so that the
10122 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
10125 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
10128 ** Guile's read-eval-print loop no longer prints #<unspecified>
10129 results. If the user wants to see this, she can evaluate the
10130 expression (assert-repl-print-unspecified #t), perhaps in her startup
10133 ** Guile no longer shows backtraces by default when an error occurs;
10134 however, it does display a message saying how to get one, and how to
10135 request that they be displayed by default. After an error, evaluate
10137 to see a backtrace, and
10138 (debug-enable 'backtrace)
10139 to see them by default.
10143 * Changes to Guile Scheme:
10145 ** Guile now distinguishes between #f and the empty list.
10147 This is for compatibility with the IEEE standard, the (possibly)
10148 upcoming Revised^5 Report on Scheme, and many extant Scheme
10151 Guile used to have #f and '() denote the same object, to make Scheme's
10152 type system more compatible with Emacs Lisp's. However, the change
10153 caused too much trouble for Scheme programmers, and we found another
10154 way to reconcile Emacs Lisp with Scheme that didn't require this.
10157 ** Guile's delq, delv, delete functions, and their destructive
10158 counterparts, delq!, delv!, and delete!, now remove all matching
10159 elements from the list, not just the first. This matches the behavior
10160 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
10161 functions which inspired them.
10163 I recognize that this change may break code in subtle ways, but it
10164 seems best to make the change before the FSF's first Guile release,
10168 ** The compiled-library-path function has been deleted from libguile.
10170 ** The facilities for loading Scheme source files have changed.
10172 *** The variable %load-path now tells Guile which directories to search
10173 for Scheme code. Its value is a list of strings, each of which names
10176 *** The variable %load-extensions now tells Guile which extensions to
10177 try appending to a filename when searching the load path. Its value
10178 is a list of strings. Its default value is ("" ".scm").
10180 *** (%search-load-path FILENAME) searches the directories listed in the
10181 value of the %load-path variable for a Scheme file named FILENAME,
10182 with all the extensions listed in %load-extensions. If it finds a
10183 match, then it returns its full filename. If FILENAME is absolute, it
10184 returns it unchanged. Otherwise, it returns #f.
10186 %search-load-path will not return matches that refer to directories.
10188 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
10189 uses %seach-load-path to find a file named FILENAME, and loads it if
10190 it finds it. If it can't read FILENAME for any reason, it throws an
10193 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
10196 *** load uses the same searching semantics as primitive-load.
10198 *** The functions %try-load, try-load-with-path, %load, load-with-path,
10199 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
10200 path, and load-module-with-path have been deleted. The functions
10201 above should serve their purposes.
10203 *** If the value of the variable %load-hook is a procedure,
10204 `primitive-load' applies its value to the name of the file being
10205 loaded (without the load path directory name prepended). If its value
10206 is #f, it is ignored. Otherwise, an error occurs.
10208 This is mostly useful for printing load notification messages.
10211 ** The function `eval!' is no longer accessible from the scheme level.
10212 We can't allow operations which introduce glocs into the scheme level,
10213 because Guile's type system can't handle these as data. Use `eval' or
10214 `read-and-eval!' (see below) as replacement.
10216 ** The new function read-and-eval! reads an expression from PORT,
10217 evaluates it, and returns the result. This is more efficient than
10218 simply calling `read' and `eval', since it is not necessary to make a
10219 copy of the expression for the evaluator to munge.
10221 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
10222 for the `read' function.
10225 ** The function `int?' has been removed; its definition was identical
10226 to that of `integer?'.
10228 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
10229 use the R4RS names for these functions.
10231 ** The function object-properties no longer returns the hash handle;
10232 it simply returns the object's property list.
10234 ** Many functions have been changed to throw errors, instead of
10235 returning #f on failure. The point of providing exception handling in
10236 the language is to simplify the logic of user code, but this is less
10237 useful if Guile's primitives don't throw exceptions.
10239 ** The function `fileno' has been renamed from `%fileno'.
10241 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
10244 * Changes to Guile's C interface:
10246 ** The library's initialization procedure has been simplified.
10247 scm_boot_guile now has the prototype:
10249 void scm_boot_guile (int ARGC,
10251 void (*main_func) (),
10254 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
10255 MAIN_FUNC should do all the work of the program (initializing other
10256 packages, reading user input, etc.) before returning. When MAIN_FUNC
10257 returns, call exit (0); this function never returns. If you want some
10258 other exit value, MAIN_FUNC may call exit itself.
10260 scm_boot_guile arranges for program-arguments to return the strings
10261 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
10262 scm_set_program_arguments with the final list, so Scheme code will
10263 know which arguments have been processed.
10265 scm_boot_guile establishes a catch-all catch handler which prints an
10266 error message and exits the process. This means that Guile exits in a
10267 coherent way when system errors occur and the user isn't prepared to
10268 handle it. If the user doesn't like this behavior, they can establish
10269 their own universal catcher in MAIN_FUNC to shadow this one.
10271 Why must the caller do all the real work from MAIN_FUNC? The garbage
10272 collector assumes that all local variables of type SCM will be above
10273 scm_boot_guile's stack frame on the stack. If you try to manipulate
10274 SCM values after this function returns, it's the luck of the draw
10275 whether the GC will be able to find the objects you allocate. So,
10276 scm_boot_guile function exits, rather than returning, to discourage
10277 people from making that mistake.
10279 The IN, OUT, and ERR arguments were removed; there are other
10280 convenient ways to override these when desired.
10282 The RESULT argument was deleted; this function should never return.
10284 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
10288 ** Guile's header files should no longer conflict with your system's
10291 In order to compile code which #included <libguile.h>, previous
10292 versions of Guile required you to add a directory containing all the
10293 Guile header files to your #include path. This was a problem, since
10294 Guile's header files have names which conflict with many systems'
10297 Now only <libguile.h> need appear in your #include path; you must
10298 refer to all Guile's other header files as <libguile/mumble.h>.
10299 Guile's installation procedure puts libguile.h in $(includedir), and
10300 the rest in $(includedir)/libguile.
10303 ** Two new C functions, scm_protect_object and scm_unprotect_object,
10304 have been added to the Guile library.
10306 scm_protect_object (OBJ) protects OBJ from the garbage collector.
10307 OBJ will not be freed, even if all other references are dropped,
10308 until someone does scm_unprotect_object (OBJ). Both functions
10311 Note that calls to scm_protect_object do not nest. You can call
10312 scm_protect_object any number of times on a given object, and the
10313 next call to scm_unprotect_object will unprotect it completely.
10315 Basically, scm_protect_object and scm_unprotect_object just
10316 maintain a list of references to things. Since the GC knows about
10317 this list, all objects it mentions stay alive. scm_protect_object
10318 adds its argument to the list; scm_unprotect_object remove its
10319 argument from the list.
10322 ** scm_eval_0str now returns the value of the last expression
10325 ** The new function scm_read_0str reads an s-expression from a
10326 null-terminated string, and returns it.
10328 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
10329 to a Scheme port object.
10331 ** The new function `scm_set_program_arguments' allows C code to set
10332 the value returned by the Scheme `program-arguments' function.
10337 * Guile no longer includes sophisticated Tcl/Tk support.
10339 The old Tcl/Tk support was unsatisfying to us, because it required the
10340 user to link against the Tcl library, as well as Tk and Guile. The
10341 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
10342 referring to widgets by names, rather than exporting widgets to Scheme
10343 code as a special datatype.
10345 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
10346 maintainers described some very interesting changes in progress to the
10347 Tcl/Tk internals, which would facilitate clean interfaces between lone
10348 Tk and other interpreters --- even for garbage-collected languages
10349 like Scheme. They expected the new Tk to be publicly available in the
10352 Since it seems that Guile might soon have a new, cleaner interface to
10353 lone Tk, and that the old Guile/Tk glue code would probably need to be
10354 completely rewritten, we (Jim Blandy and Richard Stallman) have
10355 decided not to support the old code. We'll spend the time instead on
10356 a good interface to the newer Tk, as soon as it is available.
10358 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
10361 Copyright information:
10363 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
10365 Permission is granted to anyone to make or distribute verbatim copies
10366 of this document as received, in any medium, provided that the
10367 copyright notice and this permission notice are preserved,
10368 thus giving the recipient permission to redistribute in turn.
10370 Permission is granted to distribute modified versions
10371 of this document, or of portions of it,
10372 under the above conditions, provided also that they
10373 carry prominent notices stating who last changed them.
10378 paragraph-separate: "[
\f]*$"