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 2.0.4 (since 2.0.3):
12 ** Better debuggability for interpreted procedures.
14 Guile 2.0 came with a great debugging experience for compiled
15 procedures, but the story for interpreted procedures was terrible. Now,
16 at least, interpreted procedures have names, and the `arity' procedure
17 property is always correct (or, as correct as it can be, in the presence
20 ** Support for cross-compilation.
22 One can now use a native Guile to cross-compile `.go' files for a
23 different architecture. See the documentation for `--target' in the
24 "Compilation" section of the manual, for information on how to use the
25 cross-compiler. See the "Cross building Guile" section of the README,
26 for more on how to cross-compile Guile itself.
28 ** The return of `local-eval'.
30 Back by popular demand, `the-environment' and `local-eval' allow the
31 user to capture a lexical environment, and then evaluate arbitrary
32 expressions in that context. There is also a new `local-compile'
33 command. See "Local Evaluation" in the manual, for more. Special
34 thanks to Mark Weaver for an initial implementation of this feature.
36 ** Fluids can now have default values.
38 Fluids are used for dynamic and thread-local binding. They have always
39 inherited their values from the context or thread that created them.
40 However, there was a case in which a new thread would enter Guile, and
41 the default values of all the fluids would be `#f' for that thread.
43 This has now been fixed so that `make-fluid' has an optional default
44 value for fluids in unrelated dynamic roots, which defaults to `#f'.
46 ** Garbage collector tuning.
48 The garbage collector has now been tuned to run more often under some
51 *** Unmanaged allocation
53 The new `scm_gc_register_allocation' function will notify the collector
54 of unmanaged allocation. This will cause the collector to run sooner.
55 Guile's `scm_malloc', `scm_calloc', and `scm_realloc' unmanaged
56 allocators eventually call this function. This leads to better
57 performance under steady-state unmanaged allocation.
59 *** Transient allocation
61 When the collector runs, it will try to record the total memory
62 footprint of a process, if the platform supports this information. If
63 the memory footprint is growing, the collector will run more frequently.
64 This reduces the increase of the resident size of a process in response
65 to a transient increase in allocation.
67 *** Management of threads, bignums
69 Creating a thread will allocate a fair amount of memory. Guile now does
70 some GC work (using `GC_collect_a_little') when allocating a thread.
71 This leads to a better memory footprint when creating many short-lived
74 Similarly, bignums can occupy a lot of memory. Guile now offers hooks
75 to enable custom GMP allocators that end up calling
76 `scm_gc_register_allocation'. These allocators are enabled by default
77 when running Guile from the command-line. To enable them in libraries,
78 set the `scm_install_gmp_memory_functions' variable to a nonzero value
81 ** SRFI-39 parameters are available by default.
83 Guile now includes support for parameters, as defined by SRFI-39, in the
84 default environment. See "Parameters" in the manual, for more
85 information. `current-input-port', `current-output-port', and
86 `current-error-port' are now parameters.
88 ** Add `current-warning-port'.
90 Guile now outputs warnings on a separate port, `current-warning-port',
91 initialized to the value that `current-error-port' has on startup.
95 Following Racket's lead, Guile now supports syntax parameters. See
96 "Syntax parameters" in the manual, for more.
98 Also see Barzilay, Culpepper, and Flatt's 2011 SFP workshop paper,
99 "Keeping it Clean with syntax-parameterize".
101 ** Parse command-line arguments from the locale encoding.
103 Guile now attempts to parse command-line arguments using the user's
104 locale. However for backwards compatibility with other 2.0.x releases,
105 it does so without actually calling `setlocale'. Please report any bugs
106 in this facility to bug-guile@gnu.org.
108 ** One-armed conditionals: `when' and `unless'
110 Guile finally has `when' and `unless' in the default environment. Use
111 them whenever you would use an `if' with only one branch. See
112 "Conditionals" in the manual, for more.
114 ** `current-filename', `add-to-load-path'
116 There is a new form, `(current-filename)', which expands out to the
117 source file in which it occurs. Combined with the new
118 `add-to-load-path', this allows simple scripts to easily add nearby
119 directories to the load path. See "Load Paths" in the manual, for more.
121 ** `random-state-from-platform'
123 This procedure initializes a random seed using good random sources
124 available on your platform, such as /dev/urandom. See "Random Number
125 Generation" in the manual, for more.
129 Besides the sections already mentioned, the following manual sections
130 are new in this release: "Modules and the File System", "Module System
131 Reflection", "Syntax Transformer Helpers", and "Local Inclusion".
135 ** (ice-9 session): `apropos-hook'
136 ** New print option: `escape-newlines', defaults to #t.
137 ** (ice-9 ftw): `file-system-fold', `file-system-tree', `scandir'
138 ** `scm_c_value_ref': access to multiple returned values from C
139 ** scm_call (a varargs version), scm_call_7, scm_call_8, scm_call_9
140 ** Some new syntax helpers in (system syntax)
142 Search the manual for these identifiers and modules, for more.
146 ** FreeBSD build fixes.
147 ** OpenBSD compilation fixes.
148 ** Solaris 2.10 test suite fixes.
149 ** IA64 compilation fix.
150 ** MinGW build fixes.
151 ** Work around instruction reordering on SPARC and HPPA in the VM.
152 ** Gnulib updates: added `dirfd', `setenv' modules.
156 ** Add a deprecated alias for $expt.
157 ** Add an exception printer for `getaddrinfo-error'.
158 ** Add deprecated shim for `scm_display_error' with stack as first argument.
159 ** Add warnings for unsupported `simple-format' options.
160 ** Allow overlapping regions to be passed to `bytevector-copy!'.
161 ** Avoid calling `u32_conv_from_encoding' on the null string.
162 ** Better function prologue disassembly
163 ** Compiler: fix miscompilation of (values foo ...) in some contexts.
164 ** Compiler: fix serialization of #nil-terminated lists.
165 ** Compiler: allow values bound in non-tail let expressions to be collected.
166 ** Deprecate SCM_ASRTGO.
167 ** Document invalidity of (begin) as expression; add back-compat shim.
168 ** Don't leak file descriptors when mmaping objcode.
169 ** Empty substrings no longer reference the original stringbuf.
170 ** FFI: Fix `set-pointer-finalizer!' to leave the type cell unchanged.
171 ** FFI: Fix signed/unsigned pointer mismatches in implementation.
172 ** FFI: Hold a weak reference to the CIF made by `procedure->pointer'.
173 ** FFI: Hold a weak reference to the procedure passed to `procedure->pointer'.
174 ** FFI: Properly unpack small integer return values in closure call.
175 ** Fix R6RS `fold-left' so the accumulator is the first argument.
176 ** Fix `validate-target' in (system base target).
177 ** Fix bit-set*! bug from 2005.
178 ** Fix bug in `make-repl' when `lang' is actually a <language>.
179 ** Fix bugs related to mutation, the null string, and shared substrings.
180 ** Fix <dynwind> serialization.
181 ** Fix erroneous check in `set-procedure-properties!'.
182 ** Fix generalized-vector-{ref,set!} for slices.
183 ** Fix error messages involving definition forms.
184 ** Fix primitive-eval to return #<unspecified> for definitions.
185 ** HTTP: Extend handling of "Cache-Control" header.
186 ** HTTP: Fix qstring writing of cache-extension values
187 ** HTTP: Fix validators for various list-style headers.
188 ** HTTP: Permit non-date values for Expires header.
189 ** HTTP: `write-request-line' writes absolute paths, not absolute URIs.
190 ** Hack the port-column of current-output-port after printing a prompt.
191 ** Have `cpu-word-size' error out on unknown CPUs; add support for MIPSEL.
192 ** Make sure `regexp-quote' tests use Unicode-capable string ports.
193 ** Peval: Fix bugs in the new optimizer.
194 ** Peval: fold (values FOO) to FOO in more cases
195 ** Statistically unique marks and labels, for robust hygiene across sessions.
196 ** Web: Allow URIs with empty authorities, like "file:///etc/hosts".
197 ** `,language' at REPL sets the current-language fluid.
198 ** `primitive-load' returns the value(s) of the last expression.
199 ** `scm_from_stringn' always returns unique strings.
200 ** `scm_i_substring_copy' tries to narrow the substring.
201 ** guile-readline: Clean `.go' files.
202 ** i18n: Fix gc_malloc/free mismatch on non-GNU systems.
204 Changes in 2.0.3 (since 2.0.2):
208 ** Guile has a new optimizer, `peval'.
210 `Peval' is a partial evaluator that performs constant folding, dead code
211 elimination, copy propagation, and inlining. By default it runs on
212 every piece of code that Guile compiles, to fold computations that can
213 happen at compile-time, so they don't have to happen at runtime.
215 If we did our job right, the only impact you would see would be your
216 programs getting faster. But if you notice slowdowns or bloated code,
217 please send a mail to bug-guile@gnu.org with details.
219 Thanks to William R. Cook, Oscar Waddell, and Kent Dybvig for inspiring
220 peval and its implementation.
222 You can see what peval does on a given piece of code by running the new
223 `,optimize' REPL meta-command, and comparing it to the output of
224 `,expand'. See "Compile Commands" in the manual, for more.
226 ** Fewer calls to `stat'.
228 Guile now stats only the .go file and the .scm file when loading a fresh
233 ** New module: `(web client)', a simple synchronous web client.
235 See "Web Client" in the manual, for more.
237 ** Users can now install compiled `.go' files.
239 See "Installing Site Packages" in the manual.
241 ** Remove Front-Cover and Back-Cover text from the manual.
243 The manual is still under the GNU Free Documentation License, but no
244 longer has any invariant sections.
246 ** More helpful `guild help'.
248 `guild' is Guile's multi-tool, for use in shell scripting. Now it has a
249 nicer interface for querying the set of existing commands, and getting
250 help on those commands. Try it out and see!
252 ** New macro: `define-syntax-rule'
254 `define-syntax-rule' is a shorthand to make a `syntax-rules' macro with
255 one clause. See "Syntax Rules" in the manual, for more.
257 ** The `,time' REPL meta-command now has more precision.
259 The output of this command now has microsecond precision, instead of
260 10-millisecond precision.
262 ** `(ice-9 match)' can now match records.
264 See "Pattern Matching" in the manual, for more on matching records.
266 ** New module: `(language tree-il debug)'.
268 This module provides a tree-il verifier. This is useful for people that
269 generate tree-il, usually as part of a language compiler.
271 ** New functions: `scm_is_exact', `scm_is_inexact'.
273 These provide a nice C interface for Scheme's `exact?' and `inexact?',
278 See the git log (or the ChangeLog) for more details on these bugs.
280 ** Fix order of importing modules and resolving duplicates handlers.
281 ** Fix a number of bugs involving extended (merged) generics.
282 ** Fix invocation of merge-generics duplicate handler.
283 ** Fix write beyond array end in arrays.c.
284 ** Fix read beyond end of hashtable size array in hashtab.c.
285 ** (web http): Locale-independent parsing and serialization of dates.
286 ** Ensure presence of Host header in HTTP/1.1 requests.
287 ** Fix take-right and drop-right for improper lists.
288 ** Fix leak in get_current_locale().
289 ** Fix recursive define-inlinable expansions.
290 ** Check that srfi-1 procedure arguments are procedures.
291 ** Fix r6rs `map' for multiple returns.
292 ** Fix scm_tmpfile leak on POSIX platforms.
293 ** Fix a couple of leaks (objcode->bytecode, make-boot-program).
294 ** Fix guile-lib back-compatibility for module-stexi-documentation.
295 ** Fix --listen option to allow other ports.
296 ** Fix scm_to_latin1_stringn for substrings.
297 ** Fix compilation of untyped arrays of rank not 1.
298 ** Fix unparse-tree-il of <dynset>.
299 ** Fix reading of #||||#.
300 ** Fix segfault in GOOPS when class fields are redefined.
301 ** Prefer poll(2) over select(2) to allow file descriptors above FD_SETSIZE.
304 Changes in 2.0.2 (since 2.0.1):
308 ** `guile-tools' renamed to `guild'
310 The new name is shorter. Its intended future use is for a CPAN-like
311 system for Guile wizards and journeyfolk to band together to share code;
312 hence the name. `guile-tools' is provided as a backward-compatible
313 symbolic link. See "Using Guile Tools" in the manual, for more.
315 ** New control operators: `shift' and `reset'
317 See "Shift and Reset" in the manual, for more information.
319 ** `while' as an expression
321 Previously the return value of `while' was unspecified. Now its
322 values are specified both in the case of normal termination, and via
323 termination by invoking `break', possibly with arguments. See "while
324 do" in the manual for more.
326 ** Disallow access to handles of weak hash tables
328 `hash-get-handle' and `hash-create-handle!' are no longer permitted to
329 be called on weak hash tables, because the fields in a weak handle could
330 be nulled out by the garbage collector at any time, but yet they are
331 otherwise indistinguishable from pairs. Use `hash-ref' and `hash-set!'
334 ** More precision for `get-internal-run-time', `get-internal-real-time'
336 On 64-bit systems which support POSIX clocks, Guile's internal timing
337 procedures offer nanosecond resolution instead of the 10-millisecond
338 resolution previously available. 32-bit systems now use 1-millisecond
341 ** Guile now measures time spent in GC
343 `gc-stats' now returns a meaningful value for `gc-time-taken'.
347 The statprof profiler now exports a `gcprof' procedure, driven by the
348 `after-gc-hook', to see which parts of your program are causing GC. Let
349 us know if you find it useful.
351 ** `map', `for-each' and some others now implemented in Scheme
353 We would not mention this in NEWS, as it is not a user-visible change,
354 if it were not for one thing: `map' and `for-each' are no longer
355 primitive generics. Instead they are normal bindings, which can be
356 wrapped by normal generics. This fixes some modularity issues between
357 core `map', SRFI-1 `map', and GOOPS.
359 Also it's pretty cool that we can do this without a performance impact.
361 ** Add `scm_peek_byte_or_eof'.
363 This helper is like `scm_peek_char_or_eof', but for bytes instead of
366 ** Implement #:stop-at-first-non-option option for getopt-long
368 See "getopt-long Reference" in the manual, for more information.
370 ** Improve R6RS conformance for conditions in the I/O libraries
372 The `(rnrs io simple)' module now raises the correct R6RS conditions in
373 error cases. `(rnrs io ports)' is also more correct now, though it is
374 still a work in progress.
376 ** All deprecated routines emit warnings
378 A few deprecated routines were lacking deprecation warnings. This has
383 ** Constants in compiled code now share state better
385 Constants with shared state, like `("foo")' and `"foo"', now share state
386 as much as possible, in the entire compilation unit. This cuts compiled
387 `.go' file sizes in half, generally, and speeds startup.
389 ** VLists: optimize `vlist-fold-right', and add `vhash-fold-right'
391 These procedures are now twice as fast as they were.
393 ** UTF-8 ports to bypass `iconv' entirely
395 This reduces memory usage in a very common case.
399 The compiler is now about 40% faster. (Note that this is only the case
400 once the compiler is itself compiled, so the build still takes as long
405 Some assertions that were mostly useful for sanity-checks on the
406 bytecode compiler are now off for both "regular" and "debug" engines.
407 This together with a fix to cache a TLS access and some other tweaks
408 improve the VM's performance by about 20%.
410 ** SRFI-1 list-set optimizations
412 lset-adjoin and lset-union now have fast paths for eq? sets.
414 ** `memq', `memv' optimizations
416 These procedures are now at least twice as fast than in 2.0.1.
420 ** Deprecate scm_whash API
422 `scm_whash_get_handle', `SCM_WHASHFOUNDP', `SCM_WHASHREF',
423 `SCM_WHASHSET', `scm_whash_create_handle', `scm_whash_lookup', and
424 `scm_whash_insert' are now deprecated. Use the normal hash table API
427 ** Deprecate scm_struct_table
429 `SCM_STRUCT_TABLE_NAME', `SCM_SET_STRUCT_TABLE_NAME',
430 `SCM_STRUCT_TABLE_CLASS', `SCM_SET_STRUCT_TABLE_CLASS',
431 `scm_struct_table', and `scm_struct_create_handle' are now deprecated.
432 These routines formed part of the internals of the map between structs
435 ** Deprecate scm_internal_dynamic_wind
437 The `scm_t_inner' type and `scm_internal_dynamic_wind' are deprecated,
438 as the `scm_dynwind' API is better, and this API encourages users to
439 stuff SCM values into pointers.
441 ** Deprecate scm_immutable_cell, scm_immutable_double_cell
443 These routines are deprecated, as the GC_STUBBORN API doesn't do
448 Andreas Rottman kindly transcribed the missing parts of the `(rnrs io
449 ports)' documentation from the R6RS documentation. Thanks Andreas!
453 ** Fix double-loading of script in -ds case
454 ** -x error message fix
455 ** iconveh-related cross-compilation fixes
456 ** Fix small integer return value packing on big endian machines.
457 ** Fix hash-set! in weak-value table from non-immediate to immediate
458 ** Fix call-with-input-file & relatives for multiple values
459 ** Fix `hash' for inf and nan
460 ** Fix libguile internal type errors caught by typing-strictness==2
461 ** Fix compile error in MinGW fstat socket detection
462 ** Fix generation of auto-compiled file names on MinGW
463 ** Fix multithreaded access to internal hash tables
464 ** Emit a 1-based line number in error messages
465 ** Fix define-module ordering
466 ** Fix several POSIX functions to use the locale encoding
467 ** Add type and range checks to the complex generalized vector accessors
468 ** Fix unaligned accesses for bytevectors of complex numbers
470 ** Fix erroneous VM stack overflow for canceled threads
473 Changes in 2.0.1 (since 2.0.0):
477 ** guile.m4 supports linking with rpath
479 The GUILE_FLAGS macro now sets GUILE_LIBS and GUILE_LTLIBS, which
480 include appropriate directives to the linker to include libguile-2.0.so
481 in the runtime library lookup path.
483 ** `begin' expands macros in its body before other expressions
485 This enables support for programs like the following:
490 (or (= x 0) (odd? (- x 1)))))
493 ((odd? x) (not (even? x)))))
496 ** REPL reader usability enhancements
498 The REPL now flushes input after a read error, which should prevent one
499 error from causing other errors. The REPL also now interprets comments
502 ** REPL output has configurable width
504 The REPL now defaults to output with the current terminal's width, in
505 columns. See "Debug Commands" in the manual for more information on
508 ** Better C access to the module system
510 Guile now has convenient C accessors to look up variables or values in
511 modules and their public interfaces. See `scm_c_public_ref' and friends
512 in "Accessing Modules from C" in the manual.
514 ** Added `scm_call_5', `scm_call_6'
516 See "Fly Evaluation" in the manual.
518 ** Added `scm_from_latin1_keyword', `scm_from_utf8_keyword'
520 See "Keyword Procedures" in the manual, for more. Note that
521 `scm_from_locale_keyword' should not be used when the name is a C string
524 ** R6RS unicode and string I/O work
526 Added efficient implementations of `get-string-n' and `get-string-n!'
527 for binary ports. Exported `current-input-port', `current-output-port'
528 and `current-error-port' from `(rnrs io ports)', and enhanced support
531 ** Added `pointer->scm', `scm->pointer' to `(system foreign)'
533 These procedure are useful if one needs to pass and receive SCM values
534 to and from foreign functions. See "Foreign Variables" in the manual,
537 ** Added `heap-allocated-since-gc' to `(gc-stats)'
539 Also fixed the long-standing bug in the REPL `,stat' command.
541 ** Add `on-error' REPL option
543 This option controls what happens when an error occurs at the REPL, and
544 defaults to `debug', indicating that Guile should enter the debugger.
545 Other values include `report', which will simply print a backtrace
546 without entering the debugger. See "System Commands" in the manual.
548 ** Enforce immutability of string literals
550 Attempting to mutate a string literal now causes a runtime error.
552 ** Fix pthread redirection
554 Guile 2.0.0 shipped with headers that, if configured with pthread
555 support, would re-define `pthread_create', `pthread_join', and other API
556 to redirect to the BDW-GC wrappers, `GC_pthread_create', etc. This was
557 unintended, and not necessary: because threads must enter Guile with
558 `scm_with_guile', Guile can handle thread registration itself, without
559 needing to make the GC aware of all threads. This oversight has been
562 ** `with-continuation-barrier' now unwinds on `quit'
564 A throw to `quit' in a continuation barrier will cause Guile to exit.
565 Before, it would do so before unwinding to the barrier, which would
566 prevent cleanup handlers from running. This has been fixed so that it
567 exits only after unwinding.
569 ** `string->pointer' and `pointer->string' have optional encoding arg
571 This allows users of the FFI to more easily deal in strings with
572 particular (non-locale) encodings, like "utf-8". See "Void Pointers and
573 Byte Access" in the manual, for more.
575 ** R6RS fixnum arithmetic optimizations
577 R6RS fixnum operations are are still slower than generic arithmetic,
580 ** New procedure: `define-inlinable'
582 See "Inlinable Procedures" in the manual, for more.
584 ** New procedure: `exact-integer-sqrt'
586 See "Integer Operations" in the manual, for more.
588 ** "Extended read syntax" for symbols parses better
590 In #{foo}# symbols, backslashes are now treated as escapes, as the
591 symbol-printing code intended. Additionally, "\x" within #{foo}# is now
592 interpreted as starting an R6RS hex escape. This is backward compatible
593 because the symbol printer would never produce a "\x" before. The
594 printer also works better too.
596 ** Added `--fresh-auto-compile' option
598 This allows a user to invalidate the auto-compilation cache. It's
599 usually not needed. See "Compilation" in the manual, for a discussion.
603 ** GOOPS documentation updates
607 Thanks to Mark Harig for improvements to guile.1.
609 ** SRFI-23 documented
611 The humble `error' SRFI now has an entry in the manual.
615 ** `(ice-9 binary-ports)': "R6RS I/O Ports", in the manual
616 ** `(ice-9 eval-string)': "Fly Evaluation", in the manual
617 ** `(ice-9 command-line)', not documented yet
621 ** Fixed `iconv_t' memory leak on close-port
622 ** Fixed some leaks with weak hash tables
623 ** Export `vhash-delq' and `vhash-delv' from `(ice-9 vlist)'
624 ** `after-gc-hook' works again
625 ** `define-record-type' now allowed in nested contexts
626 ** `exact-integer-sqrt' now handles large integers correctly
627 ** Fixed C extension examples in manual
628 ** `vhash-delete' honors HASH argument
629 ** Make `locale-digit-grouping' more robust
630 ** Default exception printer robustness fixes
631 ** Fix presence of non-I CPPFLAGS in `guile-2.0.pc'
632 ** `read' updates line/column numbers when reading SCSH block comments
633 ** Fix imports of multiple custom interfaces of same module
634 ** Fix encoding scanning for non-seekable ports
635 ** Fix `setter' when called with a non-setter generic
636 ** Fix f32 and f64 bytevectors to not accept rationals
637 ** Fix description of the R6RS `finite?' in manual
638 ** Quotient, remainder and modulo accept inexact integers again
639 ** Fix `continue' within `while' to take zero arguments
640 ** Fix alignment for structures in FFI
641 ** Fix port-filename of stdin, stdout, stderr to match the docs
642 ** Fix weak hash table-related bug in `define-wrapped-pointer-type'
643 ** Fix partial continuation application with pending procedure calls
644 ** scm_{to,from}_locale_string use current locale, not current ports
645 ** Fix thread cleanup, by using a pthread_key destructor
646 ** Fix `quit' at the REPL
647 ** Fix a failure to sync regs in vm bytevector ops
648 ** Fix (texinfo reflection) to handle nested structures like syntax patterns
649 ** Fix stexi->html double translation
650 ** Fix tree-il->scheme fix for <prompt>
651 ** Fix compilation of <prompt> in <fix> in single-value context
652 ** Fix race condition in ensure-writable-dir
653 ** Fix error message on ,disassemble "non-procedure"
654 ** Fix prompt and abort with the boot evaluator
655 ** Fix `procedure->pointer' for functions returning `void'
656 ** Fix error reporting in dynamic-pointer
657 ** Fix problems detecting coding: in block comments
658 ** Fix duplicate load-path and load-compiled-path in compilation environment
659 ** Add fallback read(2) suppport for .go files if mmap(2) unavailable
660 ** Fix c32vector-set!, c64vector-set!
661 ** Fix mistakenly deprecated read syntax for uniform complex vectors
662 ** Fix parsing of exact numbers with negative exponents
663 ** Ignore SIGPIPE in (system repl server)
664 ** Fix optional second arg to R6RS log function
665 ** Fix R6RS `assert' to return true value.
666 ** Fix fencepost error when seeking in bytevector input ports
667 ** Gracefully handle `setlocale' errors when starting the REPL
668 ** Improve support of the `--disable-posix' configure option
669 ** Make sure R6RS binary ports pass `binary-port?' regardless of the locale
670 ** Gracefully handle unterminated UTF-8 sequences instead of hitting an `assert'
674 Changes in 2.0.0 (changes since the 1.8.x series):
676 * New modules (see the manual for details)
678 ** `(srfi srfi-18)', more sophisticated multithreading support
679 ** `(srfi srfi-27)', sources of random bits
680 ** `(srfi srfi-38)', External Representation for Data With Shared Structure
681 ** `(srfi srfi-42)', eager comprehensions
682 ** `(srfi srfi-45)', primitives for expressing iterative lazy algorithms
683 ** `(srfi srfi-67)', compare procedures
684 ** `(ice-9 i18n)', internationalization support
685 ** `(ice-9 futures)', fine-grain parallelism
686 ** `(rnrs bytevectors)', the R6RS bytevector API
687 ** `(rnrs io ports)', a subset of the R6RS I/O port API
688 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
689 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
690 ** `(system foreign)', foreign function interface
691 ** `(sxml match)', a pattern matcher for SXML
692 ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library
693 ** `(system vm coverage)', a line-by-line code coverage library
694 ** `(web uri)', URI data type, parser, and unparser
695 ** `(web http)', HTTP header parsers and unparsers
696 ** `(web request)', HTTP request data type, reader, and writer
697 ** `(web response)', HTTP response data type, reader, and writer
698 ** `(web server)', Generic HTTP server
699 ** `(ice-9 poll)', a poll wrapper
700 ** `(web server http)', HTTP-over-TCP web server implementation
702 ** Replaced `(ice-9 match)' with Alex Shinn's compatible, hygienic matcher.
704 Guile's copy of Andrew K. Wright's `match' library has been replaced by
705 a compatible hygienic implementation by Alex Shinn. It is now
706 documented, see "Pattern Matching" in the manual.
708 Compared to Andrew K. Wright's `match', the new `match' lacks
709 `match-define', `match:error-control', `match:set-error-control',
710 `match:error', `match:set-error', and all structure-related procedures.
712 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
714 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
715 toolkit from Guile-Lib have been imported into Guile proper. See
716 "Standard Library" in the manual for more details.
718 ** Integration of lalr-scm, a parser generator
720 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
721 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
724 * Changes to the stand-alone interpreter
726 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
728 Compiled code loads much faster than Scheme source code, and runs around
729 3 or 4 times as fast, generating much less garbage in the process.
731 ** Evaluating Scheme code does not use the C stack.
733 Besides when compiling Guile itself, Guile no longer uses a recursive C
734 function as an evaluator. This obviates the need to check the C stack
735 pointer for overflow. Continuations still capture the C stack, however.
737 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
738 GUILE_SYSTEM_LOAD_COMPILED_PATH
740 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
741 for source files. It is a different path, however, because compiled
742 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
745 ** New read-eval-print loop (REPL) implementation
747 Running Guile with no arguments drops the user into the new REPL. See
748 "Using Guile Interactively" in the manual, for more information.
750 ** Remove old Emacs interface
752 Guile had an unused `--emacs' command line argument that was supposed to
753 help when running Guile inside Emacs. This option has been removed, and
754 the helper functions `named-module-use!' and `load-emacs-interface' have
757 ** Add `(system repl server)' module and `--listen' command-line argument
759 The `(system repl server)' module exposes procedures to listen on
760 sockets for connections, and serve REPLs to those clients. The --listen
761 command-line argument allows any Guile program to thus be remotely
764 See "Invoking Guile" for more information on `--listen'.
766 ** Command line additions
768 The guile binary now supports a new switch "-x", which can be used to
769 extend the list of filename extensions tried when loading files
772 ** New reader options: `square-brackets', `r6rs-hex-escapes',
775 The reader supports a new option (changeable via `read-options'),
776 `square-brackets', which instructs it to interpret square brackets as
777 parentheses. This option is on by default.
779 When the new `r6rs-hex-escapes' reader option is enabled, the reader
780 will recognize string escape sequences as defined in R6RS. R6RS string
781 escape sequences are incompatible with Guile's existing escapes, though,
782 so this option is off by default.
784 Additionally, Guile follows the R6RS newline escaping rules when the
785 `hungry-eol-escapes' option is enabled.
787 See "String Syntax" in the manual, for more information.
789 ** Function profiling and tracing at the REPL
791 The `,profile FORM' REPL meta-command can now be used to statistically
792 profile execution of a form, to see which functions are taking the most
793 time. See `,help profile' for more information.
795 Similarly, `,trace FORM' traces all function applications that occur
796 during the execution of `FORM'. See `,help trace' for more information.
798 ** Recursive debugging REPL on error
800 When Guile sees an error at the REPL, instead of saving the stack, Guile
801 will directly enter a recursive REPL in the dynamic context of the
802 error. See "Error Handling" in the manual, for more information.
804 A recursive REPL is the same as any other REPL, except that it
805 has been augmented with debugging information, so that one can inspect
806 the context of the error. The debugger has been integrated with the REPL
807 via a set of debugging meta-commands.
809 For example, one may access a backtrace with `,backtrace' (or
810 `,bt'). See "Interactive Debugging" in the manual, for more
813 ** New `guile-tools' commands: `compile', `disassemble'
815 Pass the `--help' command-line option to these commands for more
818 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
820 Users may now install Guile to nonstandard prefixes and just run
821 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
822 include `/path/to/lib'.
824 ** Guile's Emacs integration is now more keyboard-friendly
826 Backtraces may now be disclosed with the keyboard in addition to the
829 ** Load path change: search in version-specific paths before site paths
831 When looking for a module, Guile now searches first in Guile's
832 version-specific path (the library path), *then* in the site dir. This
833 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
834 installed. Also it should cut the number of `stat' system calls by half,
837 ** Value history in the REPL on by default
839 By default, the REPL will save computed values in variables like `$1',
840 `$2', and the like. There are programmatic and interactive interfaces to
841 control this. See "Value History" in the manual, for more information.
843 ** Readline tab completion for arguments
845 When readline is enabled, tab completion works for arguments too, not
846 just for the operator position.
848 ** Expression-oriented readline history
850 Guile's readline history now tries to operate on expressions instead of
851 input lines. Let us know what you think!
853 ** Interactive Guile follows GNU conventions
855 As recommended by the GPL, Guile now shows a brief copyright and
856 warranty disclaimer on startup, along with pointers to more information.
858 * Changes to Scheme functions and syntax
860 ** Support for R6RS libraries
862 The `library' and `import' forms from the latest Scheme report have been
863 added to Guile, in such a way that R6RS libraries share a namespace with
864 Guile modules. R6RS modules may import Guile modules, and are available
865 for Guile modules to import via use-modules and all the rest. See "R6RS
866 Libraries" in the manual for more information.
868 ** Implementations of R6RS libraries
870 Guile now has implementations for all of the libraries defined in the
871 R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS
872 Standard Libraries" in the manual for a full list of libraries.
874 ** Partial R6RS compatibility
876 Guile now has enough support for R6RS to run a reasonably large subset
879 Guile is not fully R6RS compatible. Many incompatibilities are simply
880 bugs, though some parts of Guile will remain R6RS-incompatible for the
881 foreseeable future. See "R6RS Incompatibilities" in the manual, for more
884 Please contact bug-guile@gnu.org if you have found an issue not
885 mentioned in that compatibility list.
887 ** New implementation of `primitive-eval'
889 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
890 still a C evaluator, used when building a fresh Guile to interpret the
891 compiler, so we can compile eval.scm. Thereafter all calls to
892 primitive-eval are implemented by VM-compiled code.
894 This allows all of Guile's procedures, be they interpreted or compiled,
895 to execute on the same stack, unifying multiple-value return semantics,
896 providing for proper tail recursion between interpreted and compiled
897 code, and simplifying debugging.
899 As part of this change, the evaluator no longer mutates the internal
900 representation of the code being evaluated in a thread-unsafe manner.
902 There are two negative aspects of this change, however. First, Guile
903 takes a lot longer to compile now. Also, there is less debugging
904 information available for debugging interpreted code. We hope to improve
905 both of these situations.
907 There are many changes to the internal C evalator interface, but all
908 public interfaces should be the same. See the ChangeLog for details. If
909 we have inadvertantly changed an interface that you were using, please
910 contact bug-guile@gnu.org.
912 ** Procedure removed: `the-environment'
914 This procedure was part of the interpreter's execution model, and does
915 not apply to the compiler.
917 ** No more `local-eval'
919 `local-eval' used to exist so that one could evaluate code in the
920 lexical context of a function. Since there is no way to get the lexical
921 environment any more, as that concept has no meaning for the compiler,
922 and a different meaning for the interpreter, we have removed the
925 If you think you need `local-eval', you should probably implement your
926 own metacircular evaluator. It will probably be as fast as Guile's
929 ** Scheme source files will now be compiled automatically.
931 If a compiled .go file corresponding to a .scm file is not found or is
932 not fresh, the .scm file will be compiled on the fly, and the resulting
933 .go file stored away. An advisory note will be printed on the console.
935 Note that this mechanism depends on the timestamp of the .go file being
936 newer than that of the .scm file; if the .scm or .go files are moved
937 after installation, care should be taken to preserve their original
940 Auto-compiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
941 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
942 will be created if needed.
944 To inhibit automatic compilation, set the GUILE_AUTO_COMPILE environment
945 variable to 0, or pass --no-auto-compile on the Guile command line.
947 ** New POSIX procedures: `getrlimit' and `setrlimit'
949 Note however that the interface of these functions is likely to change
950 in the next prerelease.
952 ** New POSIX procedure: `getsid'
954 Scheme binding for the `getsid' C library call.
956 ** New POSIX procedure: `getaddrinfo'
958 Scheme binding for the `getaddrinfo' C library function.
960 ** Multicast socket options
962 Support was added for the IP_MULTICAST_TTL and IP_MULTICAST_IF socket
963 options. See "Network Sockets and Communication" in the manual, for
966 ** `recv!', `recvfrom!', `send', `sendto' now deal in bytevectors
968 These socket procedures now take bytevectors as arguments, instead of
969 strings. There is some deprecated string support, however.
971 ** New GNU procedures: `setaffinity' and `getaffinity'.
973 See "Processes" in the manual, for more information.
975 ** New procedures: `compose', `negate', and `const'
977 See "Higher-Order Functions" in the manual, for more information.
979 ** New procedure in `(oops goops)': `method-formals'
981 ** New procedures in (ice-9 session): `add-value-help-handler!',
982 `remove-value-help-handler!', `add-name-help-handler!'
983 `remove-name-help-handler!', `procedure-arguments'
985 The value and name help handlers provide some minimal extensibility to
986 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
987 example, to make stexinfo help documentation available. See those
988 procedures' docstrings for more information.
990 `procedure-arguments' describes the arguments that a procedure can take,
991 combining arity and formals. For example:
993 (procedure-arguments resolve-interface)
994 => ((required . (name)) (rest . args))
996 Additionally, `module-commentary' is now publically exported from
999 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
1001 These procedures created primitive fexprs for the old evaluator, and are
1002 no longer supported. If you feel that you need these functions, you
1003 probably need to write your own metacircular evaluator (which will
1004 probably be as fast as Guile's, anyway).
1006 ** New language: ECMAScript
1008 Guile now ships with one other high-level language supported,
1009 ECMAScript. The goal is to support all of version 3.1 of the standard,
1010 but not all of the libraries are there yet. This support is not yet
1011 documented; ask on the mailing list if you are interested.
1013 ** New language: Brainfuck
1015 Brainfuck is a toy language that closely models Turing machines. Guile's
1016 brainfuck compiler is meant to be an example of implementing other
1017 languages. See the manual for details, or
1018 http://en.wikipedia.org/wiki/Brainfuck for more information about the
1019 Brainfuck language itself.
1021 ** New language: Elisp
1023 Guile now has an experimental Emacs Lisp compiler and runtime. You can
1024 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
1025 Kraft and Brian Templeton, and all bugs to bug-guile@gnu.org.
1027 ** Better documentation infrastructure for macros
1029 It is now possible to introspect on the type of a macro, e.g.
1030 syntax-rules, identifier-syntax, etc, and extract information about that
1031 macro, such as the syntax-rules patterns or the defmacro arguments.
1032 `(texinfo reflection)' takes advantage of this to give better macro
1035 ** Support for arbitrary procedure metadata
1037 Building on its support for docstrings, Guile now supports multiple
1038 docstrings, adding them to the tail of a compiled procedure's
1039 properties. For example:
1045 (procedure-properties foo)
1046 => ((name . foo) (documentation . "one") (documentation . "two"))
1048 Also, vectors of pairs are now treated as additional metadata entries:
1051 #((quz . #f) (docstring . "xyzzy"))
1053 (procedure-properties bar)
1054 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
1056 This allows arbitrary literals to be embedded as metadata in a compiled
1059 ** The psyntax expander now knows how to interpret the @ and @@ special
1062 ** The psyntax expander is now hygienic with respect to modules.
1064 Free variables in a macro are scoped in the module that the macro was
1065 defined in, not in the module the macro is used in. For example, code
1066 like this works now:
1068 (define-module (foo) #:export (bar))
1069 (define (helper x) ...)
1071 (syntax-rules () ((_ x) (helper x))))
1073 (define-module (baz) #:use-module (foo))
1076 It used to be you had to export `helper' from `(foo)' as well.
1077 Thankfully, this has been fixed.
1079 ** Support for version information in Guile's `module' form
1081 Guile modules now have a `#:version' field. See "R6RS Version
1082 References", "General Information about Modules", "Using Guile Modules",
1083 and "Creating Guile Modules" in the manual for more information.
1085 ** Support for renaming bindings on module export
1087 Wherever Guile accepts a symbol as an argument to specify a binding to
1088 export, it now also accepts a pair of symbols, indicating that a binding
1089 should be renamed on export. See "Creating Guile Modules" in the manual
1090 for more information.
1092 ** New procedure: `module-export-all!'
1094 This procedure exports all current and future bindings from a module.
1095 Use as `(module-export-all! (current-module))'.
1097 ** New procedure `reload-module', and `,reload' REPL command
1099 See "Module System Reflection" and "Module Commands" in the manual, for
1102 ** `eval-case' has been deprecated, and replaced by `eval-when'.
1104 The semantics of `eval-when' are easier to understand. See "Eval When"
1105 in the manual, for more information.
1107 ** Guile is now more strict about prohibiting definitions in expression
1110 Although previous versions of Guile accepted it, the following
1111 expression is not valid, in R5RS or R6RS:
1113 (if test (define foo 'bar) (define foo 'baz))
1115 In this specific case, it would be better to do:
1117 (define foo (if test 'bar 'baz))
1119 It is possible to circumvent this restriction with e.g.
1120 `(module-define! (current-module) 'foo 'baz)'. Contact the list if you
1123 ** Support for `letrec*'
1125 Guile now supports `letrec*', a recursive lexical binding operator in
1126 which the identifiers are bound in order. See "Local Bindings" in the
1127 manual, for more details.
1129 ** Internal definitions now expand to `letrec*'
1131 Following the R6RS, internal definitions now expand to letrec* instead
1132 of letrec. The following program is invalid for R5RS, but valid for
1137 (define baz (+ bar 20))
1140 ;; R5RS and Guile <= 1.8:
1141 (foo) => Unbound variable: bar
1142 ;; R6RS and Guile >= 2.0:
1145 This change should not affect correct R5RS programs, or programs written
1146 in earlier Guile dialects.
1148 ** Macro expansion produces structures instead of s-expressions
1150 In the olden days, macroexpanding an s-expression would yield another
1151 s-expression. Though the lexical variables were renamed, expansions of
1152 core forms like `if' and `begin' were still non-hygienic, as they relied
1153 on the toplevel definitions of `if' et al being the conventional ones.
1155 The solution is to expand to structures instead of s-expressions. There
1156 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
1157 etc. The expander already did this for compilation, producing Tree-IL
1158 directly; it has been changed now to do so when expanding for the
1161 ** Defmacros must now produce valid Scheme expressions.
1163 It used to be that defmacros could unquote in Scheme values, as a way of
1164 supporting partial evaluation, and avoiding some hygiene issues. For
1167 (define (helper x) ...)
1168 (define-macro (foo bar)
1171 Assuming this macro is in the `(baz)' module, the direct translation of
1174 (define (helper x) ...)
1175 (define-macro (foo bar)
1176 `((@@ (baz) helper) ,bar))
1178 Of course, one could just use a hygienic macro instead:
1182 ((_ bar) (helper bar))))
1184 ** Guile's psyntax now supports docstrings and internal definitions.
1186 The following Scheme is not strictly legal:
1193 However its intent is fairly clear. Guile interprets "bar" to be the
1194 docstring of `foo', and the definition of `baz' is still in definition
1197 ** Support for settable identifier syntax
1199 Following the R6RS, "variable transformers" are settable
1200 identifier-syntax. See "Identifier macros" in the manual, for more
1203 ** syntax-case treats `_' as a placeholder
1205 Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches
1206 anything, and binds no pattern variables. Unlike the R6RS, Guile also
1207 permits `_' to be in the literals list for a pattern.
1209 ** Macros need to be defined before their first use.
1211 It used to be that with lazy memoization, this might work:
1215 (define-macro (ref x) x)
1218 But now, the body of `foo' is interpreted to mean a call to the toplevel
1219 `ref' function, instead of a macro expansion. The solution is to define
1220 macros before code that uses them.
1222 ** Functions needed by macros at expand-time need to be present at
1225 For example, this code will work at the REPL:
1227 (define (double-helper x) (* x x))
1228 (define-macro (double-literal x) (double-helper x))
1229 (double-literal 2) => 4
1231 But it will not work when a file is compiled, because the definition of
1232 `double-helper' is not present at expand-time. The solution is to wrap
1233 the definition of `double-helper' in `eval-when':
1235 (eval-when (load compile eval)
1236 (define (double-helper x) (* x x)))
1237 (define-macro (double-literal x) (double-helper x))
1238 (double-literal 2) => 4
1240 See the documentation for eval-when for more information.
1242 ** `macroexpand' produces structures, not S-expressions.
1244 Given the need to maintain referential transparency, both lexically and
1245 modular, the result of expanding Scheme expressions is no longer itself
1246 an s-expression. If you want a human-readable approximation of the
1247 result of `macroexpand', call `tree-il->scheme' from `(language
1250 ** Removed function: `macroexpand-1'
1252 It is unclear how to implement `macroexpand-1' with syntax-case, though
1253 PLT Scheme does prove that it is possible.
1255 ** New reader macros: #' #` #, #,@
1257 These macros translate, respectively, to `syntax', `quasisyntax',
1258 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
1259 These reader macros may be overridden by `read-hash-extend'.
1261 ** Incompatible change to #'
1263 Guile did have a #' hash-extension, by default, which just returned the
1264 subsequent datum: #'foo => foo. In the unlikely event that anyone
1265 actually used this, this behavior may be reinstated via the
1266 `read-hash-extend' mechanism.
1268 ** `unquote' and `unquote-splicing' accept multiple expressions
1270 As per the R6RS, these syntax operators can now accept any number of
1271 expressions to unquote.
1273 ** Scheme expresssions may be commented out with #;
1275 #; comments out an entire expression. See SRFI-62 or the R6RS for more
1278 ** Prompts: Delimited, composable continuations
1280 Guile now has prompts as part of its primitive language. See "Prompts"
1281 in the manual, for more information.
1283 Expressions entered in at the REPL, or from the command line, are
1284 surrounded by a prompt with the default prompt tag.
1286 ** `make-stack' with a tail-called procedural narrowing argument no longer
1287 works (with compiled procedures)
1289 It used to be the case that a captured stack could be narrowed to select
1290 calls only up to or from a certain procedure, even if that procedure
1291 already tail-called another procedure. This was because the debug
1292 information from the original procedure was kept on the stack.
1294 Now with the new compiler, the stack only contains active frames from
1295 the current continuation. A narrow to a procedure that is not in the
1296 stack will result in an empty stack. To fix this, narrow to a procedure
1297 that is active in the current continuation, or narrow to a specific
1298 number of stack frames.
1300 ** Backtraces through compiled procedures only show procedures that are
1301 active in the current continuation
1303 Similarly to the previous issue, backtraces in compiled code may be
1304 different from backtraces in interpreted code. There are no semantic
1305 differences, however. Please mail bug-guile@gnu.org if you see any
1306 deficiencies with Guile's backtraces.
1308 ** `positions' reader option enabled by default
1310 This change allows primitive-load without --auto-compile to also
1311 propagate source information through the expander, for better errors and
1312 to let macros know their source locations. The compiler was already
1313 turning it on anyway.
1315 ** New macro: `current-source-location'
1317 The macro returns the current source location (to be documented).
1319 ** syntax-rules and syntax-case macros now propagate source information
1320 through to the expanded code
1322 This should result in better backtraces.
1324 ** The currying behavior of `define' has been removed.
1326 Before, `(define ((f a) b) (* a b))' would translate to
1328 (define f (lambda (a) (lambda (b) (* a b))))
1330 Now a syntax error is signaled, as this syntax is not supported by
1331 default. Use the `(ice-9 curried-definitions)' module to get back the
1334 ** New procedure, `define!'
1336 `define!' is a procedure that takes two arguments, a symbol and a value,
1337 and binds the value to the symbol in the current module. It's useful to
1338 programmatically make definitions in the current module, and is slightly
1339 less verbose than `module-define!'.
1341 ** All modules have names now
1343 Before, you could have anonymous modules: modules without names. Now,
1344 because of hygiene and macros, all modules have names. If a module was
1345 created without a name, the first time `module-name' is called on it, a
1346 fresh name will be lazily generated for it.
1348 ** The module namespace is now separate from the value namespace
1350 It was a little-known implementation detail of Guile's module system
1351 that it was built on a single hierarchical namespace of values -- that
1352 if there was a module named `(foo bar)', then in the module named
1353 `(foo)' there was a binding from `bar' to the `(foo bar)' module.
1355 This was a neat trick, but presented a number of problems. One problem
1356 was that the bindings in a module were not apparent from the module
1357 itself; perhaps the `(foo)' module had a private binding for `bar', and
1358 then an external contributor defined `(foo bar)'. In the end there can
1359 be only one binding, so one of the two will see the wrong thing, and
1360 produce an obtuse error of unclear provenance.
1362 Also, the public interface of a module was also bound in the value
1363 namespace, as `%module-public-interface'. This was a hack from the early
1364 days of Guile's modules.
1366 Both of these warts have been fixed by the addition of fields in the
1367 `module' data type. Access to modules and their interfaces from the
1368 value namespace has been deprecated, and all accessors use the new
1369 record accessors appropriately.
1371 When Guile is built with support for deprecated code, as is the default,
1372 the value namespace is still searched for modules and public interfaces,
1373 and a deprecation warning is raised as appropriate.
1375 Finally, to support lazy loading of modules as one used to be able to do
1376 with module binder procedures, Guile now has submodule binders, called
1377 if a given submodule is not found. See boot-9.scm for more information.
1379 ** New procedures: module-ref-submodule, module-define-submodule,
1380 nested-ref-module, nested-define-module!, local-ref-module,
1383 These new accessors are like their bare variants, but operate on
1384 namespaces instead of values.
1386 ** The (app modules) module tree is officially deprecated
1388 It used to be that one could access a module named `(foo bar)' via
1389 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
1390 modules)' bit was a never-used and never-documented abstraction, and has
1391 been deprecated. See the following mail for a full discussion:
1393 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
1395 The `%app' binding is also deprecated.
1397 ** `module-filename' field and accessor
1399 Modules now record the file in which they are defined. This field may be
1400 accessed with the new `module-filename' procedure.
1402 ** Modules load within a known environment
1404 It takes a few procedure calls to define a module, and those procedure
1405 calls need to be in scope. Now we ensure that the current module when
1406 loading a module is one that has the needed bindings, instead of relying
1409 ** `load' is a macro (!) that resolves paths relative to source file dir
1411 The familiar Schem `load' procedure is now a macro that captures the
1412 name of the source file being expanded, and dispatches to the new
1413 `load-in-vicinity'. Referencing `load' by bare name returns a closure
1414 that embeds the current source file name.
1416 This fix allows `load' of relative paths to be resolved with respect to
1417 the location of the file that calls `load'.
1419 ** Many syntax errors have different texts now
1421 Syntax errors still throw to the `syntax-error' key, but the arguments
1422 are often different now. Perhaps in the future, Guile will switch to
1423 using standard SRFI-35 conditions.
1425 ** Returning multiple values to compiled code will silently truncate the
1426 values to the expected number
1428 For example, the interpreter would raise an error evaluating the form,
1429 `(+ (values 1 2) (values 3 4))', because it would see the operands as
1430 being two compound "values" objects, to which `+' does not apply.
1432 The compiler, on the other hand, receives multiple values on the stack,
1433 not as a compound object. Given that it must check the number of values
1434 anyway, if too many values are provided for a continuation, it chooses
1435 to truncate those values, effectively evaluating `(+ 1 3)' instead.
1437 The idea is that the semantics that the compiler implements is more
1438 intuitive, and the use of the interpreter will fade out with time.
1439 This behavior is allowed both by the R5RS and the R6RS.
1441 ** Multiple values in compiled code are not represented by compound
1444 This change may manifest itself in the following situation:
1446 (let ((val (foo))) (do-something) val)
1448 In the interpreter, if `foo' returns multiple values, multiple values
1449 are produced from the `let' expression. In the compiler, those values
1450 are truncated to the first value, and that first value is returned. In
1451 the compiler, if `foo' returns no values, an error will be raised, while
1452 the interpreter would proceed.
1454 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
1455 behavior is more correct, however. If you wish to preserve a potentially
1456 multiply-valued return, you will need to set up a multiple-value
1457 continuation, using `call-with-values'.
1459 ** Defmacros are now implemented in terms of syntax-case.
1461 The practical ramification of this is that the `defmacro?' predicate has
1462 been removed, along with `defmacro-transformer', `macro-table',
1463 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
1464 `defmacro:transformer'. This is because defmacros are simply macros. If
1465 any of these procedures provided useful facilities to you, we encourage
1466 you to contact the Guile developers.
1468 ** Hygienic macros documented as the primary syntactic extension mechanism.
1470 The macro documentation was finally fleshed out with some documentation
1471 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
1472 expansion process. See "Macros" in the manual, for details.
1474 ** psyntax is now the default expander
1476 Scheme code is now expanded by default by the psyntax hygienic macro
1477 expander. Expansion is performed completely before compilation or
1480 Notably, syntax errors will be signalled before interpretation begins.
1481 In the past, many syntax errors were only detected at runtime if the
1482 code in question was memoized.
1484 As part of its expansion, psyntax renames all lexically-bound
1485 identifiers. Original identifier names are preserved and given to the
1486 compiler, but the interpreter will see the renamed variables, e.g.,
1487 `x432' instead of `x'.
1489 Note that the psyntax that Guile uses is a fork, as Guile already had
1490 modules before incompatible modules were added to psyntax -- about 10
1491 years ago! Thus there are surely a number of bugs that have been fixed
1492 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
1494 ** syntax-rules and syntax-case are available by default.
1496 There is no longer any need to import the `(ice-9 syncase)' module
1497 (which is now deprecated). The expander may be invoked directly via
1498 `macroexpand', though it is normally searched for via the current module
1501 Also, the helper routines for syntax-case are available in the default
1502 environment as well: `syntax->datum', `datum->syntax',
1503 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
1504 `identifier?', and `syntax-violation'. See the R6RS for documentation.
1506 ** Tail patterns in syntax-case
1508 Guile has pulled in some more recent changes from the psyntax portable
1509 syntax expander, to implement support for "tail patterns". Such patterns
1510 are supported by syntax-rules and syntax-case. This allows a syntax-case
1511 match clause to have ellipses, then a pattern at the end. For example:
1514 (syntax-rules (else)
1515 ((_ val match-clause ... (else e e* ...))
1518 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
1519 tail pattern for the else clause. Thanks to Andreas Rottmann for the
1520 patch, and Kent Dybvig for the code.
1522 ** Lexical bindings introduced by hygienic macros may not be referenced
1523 by nonhygienic macros.
1525 If a lexical binding is introduced by a hygienic macro, it may not be
1526 referenced by a nonhygienic macro. For example, this works:
1529 (define-macro (bind-x val body)
1530 `(let ((x ,val)) ,body))
1531 (define-macro (ref x)
1533 (bind-x 10 (ref x)))
1538 (define-syntax bind-x
1540 ((_ val body) (let ((x val)) body))))
1541 (define-macro (ref x)
1543 (bind-x 10 (ref x)))
1545 It is not normal to run into this situation with existing code. However,
1546 if you have defmacros that expand to hygienic macros, it is possible to
1547 run into situations like this. For example, if you have a defmacro that
1548 generates a `while' expression, the `break' bound by the `while' may not
1549 be visible within other parts of your defmacro. The solution is to port
1550 from defmacros to syntax-rules or syntax-case.
1552 ** Macros may no longer be referenced as first-class values.
1554 In the past, you could evaluate e.g. `if', and get its macro value. Now,
1555 expanding this form raises a syntax error.
1557 Macros still /exist/ as first-class values, but they must be
1558 /referenced/ via the module system, e.g. `(module-ref (current-module)
1561 ** Macros may now have docstrings.
1563 `object-documentation' from `(ice-9 documentation)' may be used to
1564 retrieve the docstring, once you have a macro value -- but see the above
1565 note about first-class macros. Docstrings are associated with the syntax
1566 transformer procedures.
1568 ** `case-lambda' is now available in the default environment.
1570 The binding in the default environment is equivalent to the one from the
1571 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
1572 to maintain compatibility with Guile 1.8 and earlier.
1574 ** Procedures may now have more than one arity.
1576 This can be the case, for example, in case-lambda procedures. The
1577 arities of compiled procedures may be accessed via procedures from the
1578 `(system vm program)' module; see "Compiled Procedures", "Optional
1579 Arguments", and "Case-lambda" in the manual.
1581 ** Deprecate arity access via (procedure-properties proc 'arity)
1583 Instead of accessing a procedure's arity as a property, use the new
1584 `procedure-minimum-arity' function, which gives the most permissive
1585 arity that the function has, in the same format as the old arity
1588 ** `lambda*' and `define*' are now available in the default environment
1590 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
1591 compatibility purposes. No semantic change has been made (we hope).
1592 Optional and keyword arguments now dispatch via special VM operations,
1593 without the need to cons rest arguments, making them very fast.
1595 ** New syntax: define-once
1597 `define-once' is like Lisp's `defvar': it creates a toplevel binding,
1598 but only if one does not exist already.
1600 ** New function, `truncated-print', with `format' support
1602 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
1603 will ensure that the output stays within a certain width, truncating the
1604 output in what is hopefully an intelligent manner. See the manual for
1607 There is a new `format' specifier, `~@y', for doing a truncated
1608 print (as opposed to `~y', which does a pretty-print). See the `format'
1609 documentation for more details.
1611 ** Better pretty-printing
1613 Indentation recognizes more special forms, like `syntax-case', and read
1614 macros like `quote' are printed better.
1616 ** Passing a number as the destination of `format' is deprecated
1618 The `format' procedure in `(ice-9 format)' now emits a deprecation
1619 warning if a number is passed as its first argument.
1621 Also, it used to be that you could omit passing a port to `format', in
1622 some cases. This still works, but has been formally deprecated.
1624 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
1626 Guile now implements SRFI-4 vectors using bytevectors. Often when you
1627 have a numeric vector, you end up wanting to write its bytes somewhere,
1628 or have access to the underlying bytes, or read in bytes from somewhere
1629 else. Bytevectors are very good at this sort of thing. But the SRFI-4
1630 APIs are nicer to use when doing number-crunching, because they are
1631 addressed by element and not by byte.
1633 So as a compromise, Guile allows all bytevector functions to operate on
1634 numeric vectors. They address the underlying bytes in the native
1635 endianness, as one would expect.
1637 Following the same reasoning, that it's just bytes underneath, Guile
1638 also allows uniform vectors of a given type to be accessed as if they
1639 were of any type. One can fill a u32vector, and access its elements with
1640 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
1643 In this way, uniform numeric vectors may be written to and read from
1644 input/output ports using the procedures that operate on bytevectors.
1646 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
1647 inlined to the VM instructions for bytevector access.
1649 See "SRFI-4" in the manual, for more information.
1651 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
1653 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
1654 are part of the standard. Complex uniform vectors and the
1655 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
1657 Guile's default environment imports `(srfi srfi-4)', and probably should
1658 import `(srfi srfi-4 gnu)' as well.
1660 See "SRFI-4 Extensions" in the manual, for more information.
1662 ** New syntax: include-from-path.
1664 `include-from-path' is like `include', except it looks for its file in
1665 the load path. It can be used to compile other files into a file.
1667 ** New syntax: quasisyntax.
1669 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
1670 documentation for more information. Thanks to Andre van Tonder for the
1673 ** `*unspecified*' is identifier syntax
1675 `*unspecified*' is no longer a variable, so it is optimized properly by
1676 the compiler, and is not `set!'-able.
1678 ** Changes and bugfixes in numerics code
1680 *** Added six new sets of fast quotient and remainder operators
1682 Added six new sets of fast quotient and remainder operator pairs with
1683 different semantics than the R5RS operators. They support not only
1684 integers, but all reals, including exact rationals and inexact
1685 floating point numbers.
1687 These procedures accept two real numbers N and D, where the divisor D
1688 must be non-zero. Each set of operators computes an integer quotient
1689 Q and a real remainder R such that N = Q*D + R and |R| < |D|. They
1690 differ only in how N/D is rounded to produce Q.
1692 `euclidean-quotient' returns the integer Q and `euclidean-remainder'
1693 returns the real R such that N = Q*D + R and 0 <= R < |D|. `euclidean/'
1694 returns both Q and R, and is more efficient than computing each
1695 separately. Note that when D > 0, `euclidean-quotient' returns
1696 floor(N/D), and when D < 0 it returns ceiling(N/D).
1698 `centered-quotient', `centered-remainder', and `centered/' are similar
1699 except that the range of remainders is -abs(D/2) <= R < abs(D/2), and
1700 `centered-quotient' rounds N/D to the nearest integer. Note that these
1701 operators are equivalent to the R6RS integer division operators `div',
1702 `mod', `div-and-mod', `div0', `mod0', and `div0-and-mod0'.
1704 `floor-quotient' and `floor-remainder' compute Q and R, respectively,
1705 where Q has been rounded toward negative infinity. `floor/' returns
1706 both Q and R, and is more efficient than computing each separately.
1707 Note that when applied to integers, `floor-remainder' is equivalent to
1708 the R5RS integer-only `modulo' operator. `ceiling-quotient',
1709 `ceiling-remainder', and `ceiling/' are similar except that Q is
1710 rounded toward positive infinity.
1712 For `truncate-quotient', `truncate-remainder', and `truncate/', Q is
1713 rounded toward zero. Note that when applied to integers,
1714 `truncate-quotient' and `truncate-remainder' are equivalent to the
1715 R5RS integer-only operators `quotient' and `remainder'.
1717 For `round-quotient', `round-remainder', and `round/', Q is rounded to
1718 the nearest integer, with ties going to the nearest even integer.
1720 *** Complex number changes
1722 Guile is now able to represent non-real complex numbers whose
1723 imaginary part is an _inexact_ zero (0.0 or -0.0), per R6RS.
1724 Previously, such numbers were immediately changed into inexact reals.
1726 (real? 0.0+0.0i) now returns #f, per R6RS, although (zero? 0.0+0.0i)
1727 still returns #t, per R6RS. (= 0 0.0+0.0i) and (= 0.0 0.0+0.0i) are
1728 #t, but the same comparisons using `eqv?' or `equal?' are #f.
1730 Like other non-real numbers, these complex numbers with inexact zero
1731 imaginary part will raise exceptions is passed to procedures requiring
1732 reals, such as `<', `>', `<=', `>=', `min', `max', `positive?',
1733 `negative?', `inf?', `nan?', `finite?', etc.
1735 **** `make-rectangular' changes
1737 scm_make_rectangular `make-rectangular' now returns a real number only
1738 if the imaginary part is an _exact_ 0. Previously, it would return a
1739 real number if the imaginary part was an inexact zero.
1741 scm_c_make_rectangular now always returns a non-real complex number,
1742 even if the imaginary part is zero. Previously, it would return a
1743 real number if the imaginary part was zero.
1745 **** `make-polar' changes
1747 scm_make_polar `make-polar' now returns a real number only if the
1748 angle or magnitude is an _exact_ 0. If the magnitude is an exact 0,
1749 it now returns an exact 0. Previously, it would return a real
1750 number if the imaginary part was an inexact zero.
1752 scm_c_make_polar now always returns a non-real complex number, even if
1753 the imaginary part is 0.0. Previously, it would return a real number
1754 if the imaginary part was 0.0.
1756 **** `imag-part' changes
1758 scm_imag_part `imag-part' now returns an exact 0 if applied to an
1759 inexact real number. Previously it returned an inexact zero in this
1762 *** `eqv?' and `equal?' now compare numbers equivalently
1764 scm_equal_p `equal?' now behaves equivalently to scm_eqv_p `eqv?' for
1765 numeric values, per R5RS. Previously, equal? worked differently,
1766 e.g. `(equal? 0.0 -0.0)' returned #t but `(eqv? 0.0 -0.0)' returned #f,
1767 and `(equal? +nan.0 +nan.0)' returned #f but `(eqv? +nan.0 +nan.0)'
1770 *** `(equal? +nan.0 +nan.0)' now returns #t
1772 Previously, `(equal? +nan.0 +nan.0)' returned #f, although
1773 `(let ((x +nan.0)) (equal? x x))' and `(eqv? +nan.0 +nan.0)'
1774 both returned #t. R5RS requires that `equal?' behave like
1775 `eqv?' when comparing numbers.
1777 *** Change in handling products `*' involving exact 0
1779 scm_product `*' now handles exact 0 differently. A product containing
1780 an exact 0 now returns an exact 0 if and only if the other arguments
1781 are all exact. An inexact zero is returned if and only if the other
1782 arguments are all finite but not all exact. If an infinite or NaN
1783 value is present, a NaN value is returned. Previously, any product
1784 containing an exact 0 yielded an exact 0, regardless of the other
1787 *** `expt' and `integer-expt' changes when the base is 0
1789 While `(expt 0 0)' is still 1, and `(expt 0 N)' for N > 0 is still
1790 zero, `(expt 0 N)' for N < 0 is now a NaN value, and likewise for
1791 integer-expt. This is more correct, and conforming to R6RS, but seems
1792 to be incompatible with R5RS, which would return 0 for all non-zero
1795 *** `expt' and `integer-expt' are more generic, less strict
1797 When raising to an exact non-negative integer exponent, `expt' and
1798 `integer-expt' are now able to exponentiate any object that can be
1799 multiplied using `*'. They can also raise an object to an exact
1800 negative integer power if its reciprocal can be taken using `/'.
1801 In order to allow this, the type of the first argument is no longer
1802 checked when raising to an exact integer power. If the exponent is 0
1803 or 1, the first parameter is not manipulated at all, and need not
1804 even support multiplication.
1806 *** Infinities are no longer integers, nor rationals
1808 scm_integer_p `integer?' and scm_rational_p `rational?' now return #f
1809 for infinities, per R6RS. Previously they returned #t for real
1810 infinities. The real infinities and NaNs are still considered real by
1811 scm_real `real?' however, per R6RS.
1813 *** NaNs are no longer rationals
1815 scm_rational_p `rational?' now returns #f for NaN values, per R6RS.
1816 Previously it returned #t for real NaN values. They are still
1817 considered real by scm_real `real?' however, per R6RS.
1819 *** `inf?' and `nan?' now throw exceptions for non-reals
1821 The domain of `inf?' and `nan?' is the real numbers. Guile now signals
1822 an error when a non-real number or non-number is passed to these
1823 procedures. (Note that NaNs _are_ considered numbers by scheme, despite
1826 *** `rationalize' bugfixes and changes
1828 Fixed bugs in scm_rationalize `rationalize'. Previously, it returned
1829 exact integers unmodified, although that was incorrect if the epsilon
1830 was at least 1 or inexact, e.g. (rationalize 4 1) should return 3 per
1831 R5RS and R6RS, but previously it returned 4. It also now handles
1832 cases involving infinities and NaNs properly, per R6RS.
1834 *** Trigonometric functions now return exact numbers in some cases
1836 scm_sin `sin', scm_cos `cos', scm_tan `tan', scm_asin `asin', scm_acos
1837 `acos', scm_atan `atan', scm_sinh `sinh', scm_cosh `cosh', scm_tanh
1838 `tanh', scm_sys_asinh `asinh', scm_sys_acosh `acosh', and
1839 scm_sys_atanh `atanh' now return exact results in some cases.
1841 *** New procedure: `finite?'
1843 Add scm_finite_p `finite?' from R6RS to guile core, which returns #t
1844 if and only if its argument is neither infinite nor a NaN. Note that
1845 this is not the same as (not (inf? x)) or (not (infinite? x)), since
1846 NaNs are neither finite nor infinite.
1848 *** Improved exactness handling for complex number parsing
1850 When parsing non-real complex numbers, exactness specifiers are now
1851 applied to each component, as is done in PLT Scheme. For complex
1852 numbers written in rectangular form, exactness specifiers are applied
1853 to the real and imaginary parts before calling scm_make_rectangular.
1854 For complex numbers written in polar form, exactness specifiers are
1855 applied to the magnitude and angle before calling scm_make_polar.
1857 Previously, exactness specifiers were applied to the number as a whole
1858 _after_ calling scm_make_rectangular or scm_make_polar.
1860 For example, (string->number "#i5.0+0i") now does the equivalent of:
1862 (make-rectangular (exact->inexact 5.0) (exact->inexact 0))
1864 which yields 5.0+0.0i. Previously it did the equivalent of:
1866 (exact->inexact (make-rectangular 5.0 0))
1870 ** Unicode characters
1872 Unicode characters may be entered in octal format via e.g. `#\454', or
1873 created via (integer->char 300). A hex external representation will
1874 probably be introduced at some point.
1878 Internally, strings are now represented either in the `latin-1'
1879 encoding, one byte per character, or in UTF-32, with four bytes per
1880 character. Strings manage their own allocation, switching if needed.
1882 Extended characters may be written in a literal string using the
1883 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
1884 or 24-bit codepoints, respectively, or entered directly in the native
1885 encoding of the port on which the string is read.
1889 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
1891 ** Support for non-ASCII source code files
1893 The default reader now handles source code files for some of the
1894 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
1895 should have an encoding declaration near the top of the file. Also,
1896 there is a new function, `file-encoding', that scans a port for a coding
1897 declaration. See the section of the manual entitled, "Character Encoding
1900 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
1901 code. This use is now discouraged. Binary input and output is
1902 currently supported by opening ports in the ISO-8859-1 locale.
1904 ** Source files default to UTF-8.
1906 If source files do not specify their encoding via a `coding:' block,
1907 the default encoding is UTF-8, instead of being taken from the current
1910 ** Interactive Guile installs the current locale.
1912 Instead of leaving the user in the "C" locale, running the Guile REPL
1913 installs the current locale. [FIXME xref?]
1915 ** Support for locale transcoding when reading from and writing to ports
1917 Ports now have an associated character encoding, and port read and write
1918 operations do conversion to and from locales automatically. Ports also
1919 have an associated strategy for how to deal with locale conversion
1922 See the documentation in the manual for the four new support functions,
1923 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
1924 and `port-conversion-strategy'.
1926 ** String and SRFI-13 functions can operate on Unicode strings
1928 ** Unicode support for SRFI-14 character sets
1930 The default character sets are no longer locale dependent and contain
1931 characters from the whole Unicode range. There is a new predefined
1932 character set, `char-set:designated', which contains all assigned
1933 Unicode characters. There is a new debugging function, `%char-set-dump'.
1935 ** Character functions operate on Unicode characters
1937 `char-upcase' and `char-downcase' use default Unicode casing rules.
1938 Character comparisons such as `char<?' and `char-ci<?' now sort based on
1939 Unicode code points.
1941 ** Global variables `scm_charnames' and `scm_charnums' are removed
1943 These variables contained the names of control characters and were
1944 used when writing characters. While these were global, they were
1945 never intended to be public API. They have been replaced with private
1948 ** EBCDIC support is removed
1950 There was an EBCDIC compile flag that altered some of the character
1951 processing. It appeared that full EBCDIC support was never completed
1952 and was unmaintained.
1954 ** Compile-time warnings
1956 Guile can warn about potentially unbound free variables. Pass the
1957 -Wunbound-variable on the `guile-tools compile' command line, or add
1958 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
1959 invocation. Warnings are also enabled by default for expressions entered
1962 Guile can also warn when you pass the wrong number of arguments to a
1963 procedure, with -Warity-mismatch, or `arity-mismatch' in the
1964 `#:warnings' as above.
1966 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
1967 warn about unused local or global (top-level) variables, and `-Wformat',
1968 to check for various errors related to the `format' procedure.
1970 ** A new `memoize-symbol' evaluator trap has been added.
1972 This trap can be used for efficiently implementing a Scheme code
1975 ** Duplicate bindings among used modules are resolved lazily.
1977 This slightly improves program startup times.
1979 ** New thread cancellation and thread cleanup API
1981 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
1983 ** New threads are in `(guile-user)' by default, not `(guile)'
1985 It used to be that a new thread entering Guile would do so in the
1986 `(guile)' module, unless this was the first time Guile was initialized,
1987 in which case it was `(guile-user)'. This has been fixed to have all
1988 new threads unknown to Guile default to `(guile-user)'.
1990 ** New helpers: `print-exception', `set-exception-printer!'
1992 These functions implement an extensible exception printer. Guile
1993 registers printers for all of the exceptions it throws. Users may add
1994 their own printers. There is also `scm_print_exception', for use by C
1995 programs. Pleasantly, this allows SRFI-35 and R6RS exceptions to be
1996 printed appropriately.
1998 ** GOOPS dispatch in scheme
2000 As an implementation detail, GOOPS dispatch is no longer implemented by
2001 special evaluator bytecodes, but rather directly via a Scheme function
2002 associated with an applicable struct. There is some VM support for the
2003 underlying primitives, like `class-of'.
2005 This change will in the future allow users to customize generic function
2006 dispatch without incurring a performance penalty, and allow us to
2007 implement method combinations.
2009 ** Applicable struct support
2011 One may now make structs from Scheme that may be applied as procedures.
2012 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
2013 That struct will be the vtable of your applicable structs; instances of
2014 that new struct are assumed to have the procedure in their first slot.
2015 `<applicable-struct-vtable>' is like Common Lisp's
2016 `funcallable-standard-class'. Likewise there is
2017 `<applicable-struct-with-setter-vtable>', which looks for the setter in
2018 the second slot. This needs to be better documented.
2022 GOOPS had a number of concepts that were relevant to the days of Tcl,
2023 but not any more: operators and entities, mainly. These objects were
2024 never documented, and it is unlikely that they were ever used. Operators
2025 were a kind of generic specific to the Tcl support. Entities were
2026 replaced by applicable structs, mentioned above.
2028 ** New struct slot allocation: "hidden"
2030 A hidden slot is readable and writable, but will not be initialized by a
2031 call to make-struct. For example in your layout you would say "ph"
2032 instead of "pw". Hidden slots are useful for adding new slots to a
2033 vtable without breaking existing invocations to make-struct.
2035 ** eqv? not a generic
2037 One used to be able to extend `eqv?' as a primitive-generic, but no
2038 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
2039 should be able to compile to static dispatch tables, it doesn't make
2040 sense to allow extensions that would subvert this optimization.
2042 ** `inet-ntop' and `inet-pton' are always available.
2044 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
2045 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
2046 functions are deprecated.
2048 ** `getopt-long' parsing errors throw to `quit', not `misc-error'
2050 This change should inhibit backtraces on argument parsing errors.
2051 `getopt-long' has been modified to print out the error that it throws
2054 ** New primitive: `tmpfile'.
2056 See "File System" in the manual.
2058 ** Random generator state may be serialized to a datum
2060 `random-state->datum' will serialize a random state to a datum, which
2061 may be written out, read back in later, and revivified using
2062 `datum->random-state'. See "Random" in the manual, for more details.
2064 ** Fix random number generator on 64-bit platforms
2066 There was a nasty bug on 64-bit platforms in which asking for a random
2067 integer with a range between 2**32 and 2**64 caused a segfault. After
2068 many embarrassing iterations, this was fixed.
2070 ** Fast bit operations.
2072 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
2073 have dedicated bytecodes. Guile is not just for symbolic computation,
2074 it's for number crunching too.
2076 ** Faster SRFI-9 record access
2078 SRFI-9 records are now implemented directly on top of Guile's structs,
2079 and their accessors are defined in such a way that normal call-sites
2080 inline to special VM opcodes, while still allowing for the general case
2081 (e.g. passing a record accessor to `apply').
2083 ** R6RS block comment support
2085 Guile now supports R6RS nested block comments. The start of a comment is
2086 marked with `#|', and the end with `|#'.
2088 ** `guile-2' cond-expand feature
2090 To test if your code is running under Guile 2.0 (or its alpha releases),
2091 test for the `guile-2' cond-expand feature. Like this:
2093 (cond-expand (guile-2 (eval-when (compile)
2094 ;; This must be evaluated at compile time.
2095 (fluid-set! current-reader my-reader)))
2097 ;; Earlier versions of Guile do not have a
2098 ;; separate compilation phase.
2099 (fluid-set! current-reader my-reader)))
2101 ** New global variables: %load-compiled-path, %load-compiled-extensions
2103 These are analogous to %load-path and %load-extensions.
2105 ** New fluid: `%file-port-name-canonicalization'
2107 This fluid parameterizes the file names that are associated with file
2108 ports. If %file-port-name-canonicalization is 'absolute, then file names
2109 are canonicalized to be absolute paths. If it is 'relative, then the
2110 name is canonicalized, but any prefix corresponding to a member of
2111 `%load-path' is stripped off. Otherwise the names are passed through
2114 In addition, the `compile-file' and `compile-and-load' procedures bind
2115 %file-port-name-canonicalization to their `#:canonicalization' keyword
2116 argument, which defaults to 'relative. In this way, one might compile
2117 "../module/ice-9/boot-9.scm", but the path that gets residualized into
2118 the .go is "ice-9/boot-9.scm".
2120 ** New procedure, `make-promise'
2122 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
2124 ** `defined?' may accept a module as its second argument
2126 Previously it only accepted internal structures from the evaluator.
2128 ** New entry into %guile-build-info: `ccachedir'
2130 ** Fix bug in `module-bound?'.
2132 `module-bound?' was returning true if a module did have a local
2133 variable, but one that was unbound, but another imported module bound
2134 the variable. This was an error, and was fixed.
2136 ** `(ice-9 syncase)' has been deprecated.
2138 As syntax-case is available by default, importing `(ice-9 syncase)' has
2139 no effect, and will trigger a deprecation warning.
2141 ** New readline history functions
2143 The (ice-9 readline) module now provides add-history, read-history,
2144 write-history and clear-history, which wrap the corresponding GNU
2145 History library functions.
2147 ** Removed deprecated uniform array procedures:
2148 dimensions->uniform-array, list->uniform-array, array-prototype
2150 Instead, use make-typed-array, list->typed-array, or array-type,
2153 ** Deprecate the old `scm-style-repl'
2155 The following bindings from boot-9 are now found in `(ice-9
2156 scm-style-repl)': `scm-style-repl', `error-catching-loop',
2157 `error-catching-repl', `bad-throw', `scm-repl-silent'
2158 `assert-repl-silence', `repl-print-unspecified',
2159 `assert-repl-print-unspecified', `scm-repl-verbose',
2160 `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl',
2161 `default-pre-unwind-handler', `handle-system-error',
2163 The following bindings have been deprecated, with no replacement:
2164 `pre-unwind-handler-dispatch'.
2166 The following bindings have been totally removed:
2167 `before-signal-stack'.
2169 Deprecated forwarding shims have been installed so that users that
2170 expect these bindings in the main namespace will still work, but receive
2171 a deprecation warning.
2173 ** `set-batch-mode?!' replaced by `ensure-batch-mode!'
2175 "Batch mode" is a flag used to tell a program that it is not running
2176 interactively. One usually turns it on after a fork. It may not be
2177 turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!',
2178 because it is a better interface, as it can only turn on batch mode, not
2181 ** Deprecate `save-stack', `the-last-stack'
2183 It used to be that the way to debug programs in Guile was to capture the
2184 stack at the time of error, drop back to the REPL, then debug that
2185 stack. But this approach didn't compose, was tricky to get right in the
2186 presence of threads, and was not very powerful.
2188 So `save-stack', `stack-saved?', and `the-last-stack' have been moved to
2189 `(ice-9 save-stack)', with deprecated bindings left in the root module.
2191 ** `top-repl' has its own module
2193 The `top-repl' binding, called with Guile is run interactively, is now
2194 is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was
2195 left in the default environment.
2197 ** `display-error' takes a frame
2199 The `display-error' / `scm_display_error' helper now takes a frame as an
2200 argument instead of a stack. Stacks are still supported in deprecated
2201 builds. Additionally, `display-error' will again source location
2202 information for the error.
2204 ** No more `(ice-9 debug)'
2206 This module had some debugging helpers that are no longer applicable to
2207 the current debugging model. Importing this module will produce a
2208 deprecation warning. Users should contact bug-guile for support.
2210 ** Remove obsolete debug-options
2212 Removed `breakpoints', `trace', `procnames', `indent', `frames',
2213 `maxdepth', and `debug' debug-options.
2215 ** `backtrace' debug option on by default
2217 Given that Guile 2.0 can always give you a backtrace, backtraces are now
2220 ** `turn-on-debugging' deprecated
2222 ** Remove obsolete print-options
2224 The `source' and `closure-hook' print options are obsolete, and have
2227 ** Remove obsolete read-options
2229 The "elisp-strings" and "elisp-vectors" read options were unused and
2230 obsolete, so they have been removed.
2232 ** Remove eval-options and trap-options
2234 Eval-options and trap-options are obsolete with the new VM and
2237 ** Remove (ice-9 debugger) and (ice-9 debugging)
2239 See "Traps" and "Interactive Debugging" in the manual, for information
2240 on their replacements.
2242 ** Remove the GDS Emacs integration
2244 See "Using Guile in Emacs" in the manual, for info on how we think you
2245 should use Guile with Emacs.
2247 ** Deprecated: `lazy-catch'
2249 `lazy-catch' was a form that captured the stack at the point of a
2250 `throw', but the dynamic state at the point of the `catch'. It was a bit
2251 crazy. Please change to use `catch', possibly with a throw-handler, or
2252 `with-throw-handler'.
2254 ** Deprecated: primitive properties
2256 The `primitive-make-property', `primitive-property-set!',
2257 `primitive-property-ref', and `primitive-property-del!' procedures were
2258 crufty and only used to implement object properties, which has a new,
2259 threadsafe implementation. Use object properties or weak hash tables
2262 ** Deprecated `@bind' syntax
2264 `@bind' was part of an older implementation of the Emacs Lisp language,
2265 and is no longer used.
2267 ** Miscellaneous other deprecations
2269 `cuserid' has been deprecated, as it only returns 8 bytes of a user's
2270 login. Use `(passwd:name (getpwuid (geteuid)))' instead.
2272 Additionally, the procedures `apply-to-args', `has-suffix?', `scheme-file-suffix'
2273 `get-option', `for-next-option', `display-usage-report',
2274 `transform-usage-lambda', `collect', and `set-batch-mode?!' have all
2277 ** Add support for unbound fluids
2279 See `make-unbound-fluid', `fluid-unset!', and `fluid-bound?' in the
2282 ** Add `variable-unset!'
2284 See "Variables" in the manual, for more details.
2286 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
2288 * Changes to the C interface
2290 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
2292 The semantics of `scm_gc_malloc ()' have been changed, in a
2293 backward-compatible way. A new allocation routine,
2294 `scm_gc_malloc_pointerless ()', was added.
2296 Libgc is a conservative GC, which we hope will make interaction with C
2297 code easier and less error-prone.
2299 ** New procedures: `scm_to_stringn', `scm_from_stringn'
2300 ** New procedures: scm_{to,from}_{utf8,latin1}_symbol{n,}
2301 ** New procedures: scm_{to,from}_{utf8,utf32,latin1}_string{n,}
2303 These new procedures convert to and from string representations in
2304 particular encodings.
2306 Users should continue to use locale encoding for user input, user
2307 output, or interacting with the C library.
2309 Use the Latin-1 functions for ASCII, and for literals in source code.
2311 Use UTF-8 functions for interaction with modern libraries which deal in
2312 UTF-8, and UTF-32 for interaction with utf32-using libraries.
2314 Otherwise, use scm_to_stringn or scm_from_stringn with a specific
2317 ** New type definitions for `scm_t_intptr' and friends.
2319 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
2320 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
2321 available to C. Have fun!
2323 ** The GH interface (deprecated in version 1.6, 2001) was removed.
2325 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
2327 This makes these internal functions technically not callable from
2330 ** Functions for handling `scm_option' now no longer require an argument
2331 indicating length of the `scm_t_option' array.
2333 ** Procedures-with-setters are now implemented using applicable structs
2335 From a user's perspective this doesn't mean very much. But if, for some
2336 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
2337 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
2340 ** Remove old evaluator closures
2342 There used to be ranges of typecodes allocated to interpreted data
2343 structures, but that it no longer the case, given that interpreted
2344 procedure are now just regular VM closures. As a result, there is a
2345 newly free tc3, and a number of removed macros. See the ChangeLog for
2348 ** Primitive procedures are now VM trampoline procedures
2350 It used to be that there were something like 12 different typecodes
2351 allocated to primitive procedures, each with its own calling convention.
2352 Now there is only one, the gsubr. This may affect user code if you were
2353 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
2354 solution is to switch to use scm_c_make_gsubr. This solution works well
2355 both with the old 1.8 and with the current 1.9 branch.
2357 Guile's old evaluator used to have special cases for applying "gsubrs",
2358 primitive procedures with specified numbers of required, optional, and
2359 rest arguments. Now, however, Guile represents gsubrs as normal VM
2360 procedures, with appropriate bytecode to parse out the correct number of
2361 arguments, including optional and rest arguments, and then with a
2362 special bytecode to apply the gsubr.
2364 This allows primitive procedures to appear on the VM stack, allowing
2365 them to be accurately counted in profiles. Also they now have more
2366 debugging information attached to them -- their number of arguments, for
2367 example. In addition, the VM can completely inline the application
2368 mechanics, allowing for faster primitive calls.
2370 However there are some changes on the C level. There is no more
2371 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
2372 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
2373 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
2374 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
2375 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
2377 Perhaps more significantly, `scm_c_make_subr',
2378 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
2379 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
2380 and there are no more subr typecodes. Use the scm_c_make_gsubr family
2383 Normal users of gsubrs should not be affected, though, as the
2384 scm_c_make_gsubr family still is the correct way to create primitive
2387 ** Remove deprecated array C interfaces
2389 Removed the deprecated array functions `scm_i_arrayp',
2390 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
2391 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
2392 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
2393 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
2395 ** Remove unused snarf macros
2397 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
2398 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
2400 ** New functions: `scm_call_n', `scm_c_run_hookn'
2402 `scm_call_n' applies to apply a function to an array of arguments.
2403 `scm_c_run_hookn' runs a hook with an array of arguments.
2405 ** Some SMOB types changed to have static typecodes
2407 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
2408 they have statically allocated tc7 typecodes.
2410 ** Preparations for changing SMOB representation
2412 If things go right, we'll be changing the SMOB representation soon. To
2413 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
2414 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
2415 in the future. Code accessing SMOBs using SCM_CELL macros was never
2416 correct, but until now things still worked. Users should be aware of
2419 ** Changed invocation mechanics of applicable SMOBs
2421 Guile's old evaluator used to have special cases for applying SMOB
2422 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
2423 trampoline procedure for it, and use the normal mechanics to apply the
2424 trampoline. This simplifies procedure application in the normal,
2427 The upshot is that the mechanics used to apply a SMOB are different from
2428 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
2429 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
2430 deprecated. Just use the scm_call_0 family of procedures.
2432 ** Removed support shlibs for SRFIs 1, 4, 13, 14, and 60
2434 Though these SRFI support libraries did expose API, they encoded a
2435 strange version string into their library names. That version was never
2436 programmatically exported, so there was no way people could use the
2439 This was a fortunate oversight, as it allows us to remove the need for
2440 extra, needless shared libraries --- the C support code for SRFIs 4, 13,
2441 and 14 was already in core --- and allow us to incrementally return the
2442 SRFI implementation to Scheme.
2444 ** New C function: scm_module_public_interface
2446 This procedure corresponds to Scheme's `module-public-interface'.
2448 ** Undeprecate `scm_the_root_module ()'
2450 It's useful to be able to get the root module from C without doing a
2453 ** Inline vector allocation
2455 Instead of having vectors point out into the heap for their data, their
2456 data is now allocated inline to the vector object itself. The same is
2457 true for bytevectors, by default, though there is an indirection
2458 available which should allow for making a bytevector from an existing
2461 ** New struct constructors that don't involve making lists
2463 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
2464 constructors, respectively, for structs. You might find them useful.
2468 In Guile 1.8, there were debugging frames on the C stack. Now there is
2469 no more need to explicitly mark the stack in this way, because Guile has
2470 a VM stack that it knows how to walk, which simplifies the C API
2471 considerably. See the ChangeLog for details; the relevant interface is
2472 in libguile/stacks.h. The Scheme API has not been changed significantly.
2474 ** Removal of Guile's primitive object system.
2476 There were a number of pieces in `objects.[ch]' that tried to be a
2477 minimal object system, but were never documented, and were quickly
2478 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
2479 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
2480 from objects.h are no more. In the very unlikely case in which these
2481 were useful to you, we urge you to contact guile-devel.
2485 Actually the future is still in the state that it was, is, and ever
2486 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
2487 part of it. These files were experimental, never compiled, and would be
2488 better implemented in Scheme anyway. In the future, that is.
2490 ** Deprecate trampolines
2492 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
2493 so on. The point was to do some precomputation on the type of the
2494 procedure, then return a specialized "call" procedure. However this
2495 optimization wasn't actually an optimization, so it is now deprecated.
2496 Just use `scm_call_0', etc instead.
2498 ** Deprecated `scm_badargsp'
2500 This function is unused in Guile, but was part of its API.
2502 ** Better support for Lisp `nil'.
2504 The bit representation of `nil' has been tweaked so that it is now very
2505 efficient to check e.g. if a value is equal to Scheme's end-of-list or
2506 Lisp's nil. Additionally there are a heap of new, specific predicates
2507 like scm_is_null_or_nil.
2509 ** Better integration of Lisp `nil'.
2511 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
2512 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
2513 but when we start to integrate more with Emacs, it is possible that we
2514 break code that assumes that, for example, `(not x)' implies that `x' is
2515 `eq?' to `#f'. This is not a common assumption. Refactoring affected
2516 code to rely on properties instead of identities will improve code
2517 correctness. See "Nil" in the manual, for more details.
2519 ** Support for static allocation of strings, symbols, and subrs.
2521 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
2522 much of their associated data as static variables, reducing Guile's
2525 ** `scm_stat' has an additional argument, `exception_on_error'
2526 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
2528 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
2530 Previously they would use the `off_t' type, which is fragile since its
2531 definition depends on the application's value for `_FILE_OFFSET_BITS'.
2533 ** The `long_long' C type, deprecated in 1.8, has been removed
2535 ** Removed deprecated uniform array procedures: scm_make_uve,
2536 scm_array_prototype, scm_list_to_uniform_array,
2537 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
2538 scm_ra_set_contp, scm_aind, scm_raprin1
2540 These functions have been deprecated since early 2005.
2542 * Changes to the distribution
2544 ** Guile's license is now LGPLv3+
2546 In other words the GNU Lesser General Public License, version 3 or
2547 later (at the discretion of each person that chooses to redistribute
2552 Guile's build is visually quieter, due to the use of Automake 1.11's
2553 AM_SILENT_RULES. Build as `make V=1' to see all of the output.
2555 ** GOOPS documentation folded into Guile reference manual
2557 GOOPS, Guile's object system, used to be documented in separate manuals.
2558 This content is now included in Guile's manual directly.
2560 ** `guile-config' will be deprecated in favor of `pkg-config'
2562 `guile-config' has been rewritten to get its information from
2563 `pkg-config', so this should be a transparent change. Note however that
2564 guile.m4 has yet to be modified to call pkg-config instead of
2567 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
2569 Programs that use `pkg-config' to find Guile or one of its Autoconf
2570 macros should now require `guile-2.0' instead of `guile-1.8'.
2572 ** New installation directory: $(pkglibdir)/1.9/ccache
2574 If $(libdir) is /usr/lib, for example, Guile will install its .go files
2575 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
2577 ** Parallel installability fixes
2579 Guile now installs its header files to a effective-version-specific
2580 directory, and includes the effective version (e.g. 2.0) in the library
2581 name (e.g. libguile-2.0.so).
2583 This change should be transparent to users, who should detect Guile via
2584 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
2585 parallel installs for multiple versions of Guile development
2588 ** Dynamically loadable extensions may be placed in a Guile-specific path
2590 Before, Guile only searched the system library paths for extensions
2591 (e.g. /usr/lib), which meant that the names of Guile extensions had to
2592 be globally unique. Installing them to a Guile-specific extensions
2593 directory is cleaner. Use `pkg-config --variable=extensiondir
2594 guile-2.0' to get the location of the extensions directory.
2596 ** User Scheme code may be placed in a version-specific path
2598 Before, there was only one way to install user Scheme code to a
2599 version-specific Guile directory: install to Guile's own path,
2600 e.g. /usr/share/guile/2.0. The site directory,
2601 e.g. /usr/share/guile/site, was unversioned. This has been changed to
2602 add a version-specific site directory, e.g. /usr/share/guile/site/2.0,
2603 searched before the global site directory.
2605 ** New dependency: libgc
2607 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
2609 ** New dependency: GNU libunistring
2611 See http://www.gnu.org/software/libunistring/, for more information. Our
2612 Unicode support uses routines from libunistring.
2614 ** New dependency: libffi
2616 See http://sourceware.org/libffi/, for more information.
2620 Changes in 1.8.8 (since 1.8.7)
2624 ** Fix possible buffer overruns when parsing numbers
2625 ** Avoid clash with system setjmp/longjmp on IA64
2626 ** Fix `wrong type arg' exceptions with IPv6 addresses
2629 Changes in 1.8.7 (since 1.8.6)
2631 * New modules (see the manual for details)
2633 ** `(srfi srfi-98)', an interface to access environment variables
2637 ** Fix compilation with `--disable-deprecated'
2638 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
2639 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
2640 ** Fix build problem when scm_t_timespec is different from struct timespec
2641 ** Fix build when compiled with -Wundef -Werror
2642 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
2643 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
2644 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
2645 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
2646 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
2647 ** Fix reading of complex numbers where both parts are inexact decimals
2649 ** Allow @ macro to work with (ice-9 syncase)
2651 Previously, use of the @ macro in a module whose code is being
2652 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
2653 Now it works as you would expect (giving the value of the specified
2656 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
2659 Changes in 1.8.6 (since 1.8.5)
2661 * New features (see the manual for details)
2663 ** New convenience function `scm_c_symbol_length ()'
2665 ** Single stepping through code from Emacs
2667 When you use GDS to evaluate Scheme code from Emacs, you can now use
2668 `C-u' to indicate that you want to single step through that code. See
2669 `Evaluating Scheme Code' in the manual for more details.
2671 ** New "guile(1)" man page!
2673 * Changes to the distribution
2675 ** Automake's `AM_MAINTAINER_MODE' is no longer used
2677 Thus, the `--enable-maintainer-mode' configure option is no longer
2678 available: Guile is now always configured in "maintainer mode".
2680 ** `ChangeLog' files are no longer updated
2682 Instead, changes are detailed in the version control system's logs. See
2683 the top-level `ChangeLog' files for details.
2688 ** `symbol->string' now returns a read-only string, as per R5RS
2689 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
2690 ** `guile-config link' now prints `-L$libdir' before `-lguile'
2691 ** Fix memory corruption involving GOOPS' `class-redefinition'
2692 ** Fix possible deadlock in `mutex-lock'
2693 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
2694 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
2695 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
2696 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
2697 ** Fix misleading output from `(help rationalize)'
2698 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
2699 ** Fix `gcd' when called with a single, negative argument.
2700 ** Fix `Stack overflow' errors seen when building on some platforms
2701 ** Fix bug when `scm_with_guile ()' was called several times from the
2703 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
2704 dynamic environment of the call to `raise'
2705 ** Fix potential deadlock in `make-struct'
2706 ** Fix compilation problem with libltdl from Libtool 2.2.x
2707 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
2710 Changes in 1.8.5 (since 1.8.4)
2712 * Infrastructure changes
2714 ** Guile repository switched from CVS to Git
2716 The new repository can be accessed using
2717 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
2718 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
2720 ** Add support for `pkg-config'
2722 See "Autoconf Support" in the manual for details.
2724 * New modules (see the manual for details)
2728 * New features (see the manual for details)
2730 ** New `postfix' read option, for SRFI-88 keyword syntax
2731 ** Some I/O primitives have been inlined, which improves I/O performance
2732 ** New object-based traps infrastructure
2734 This is a GOOPS-based infrastructure that builds on Guile's low-level
2735 evaluator trap calls and facilitates the development of debugging
2736 features like single-stepping, breakpoints, tracing and profiling.
2737 See the `Traps' node of the manual for details.
2739 ** New support for working on Guile code from within Emacs
2741 Guile now incorporates the `GDS' library (previously distributed
2742 separately) for working on Guile code from within Emacs. See the
2743 `Using Guile In Emacs' node of the manual for details.
2747 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
2748 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
2750 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
2751 would trigger an unbound variable error for `match:andmap'.
2753 ** `(oop goops describe)' now properly provides the `describe' feature
2754 ** Fixed `args-fold' from `(srfi srfi-37)'
2756 Previously, parsing short option names of argument-less options would
2757 lead to a stack overflow.
2759 ** `(srfi srfi-35)' is now visible through `cond-expand'
2760 ** Fixed type-checking for the second argument of `eval'
2761 ** Fixed type-checking for SRFI-1 `partition'
2762 ** Fixed `struct-ref' and `struct-set!' on "light structs"
2763 ** Honor struct field access rights in GOOPS
2764 ** Changed the storage strategy of source properties, which fixes a deadlock
2765 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
2766 ** Fixed build issue for GNU/Linux on IA64
2767 ** Fixed build issues on NetBSD 1.6
2768 ** Fixed build issue on Solaris 2.10 x86_64
2769 ** Fixed build issue with DEC/Compaq/HP's compiler
2770 ** Fixed `scm_from_complex_double' build issue on FreeBSD
2771 ** Fixed `alloca' build issue on FreeBSD 6
2772 ** Removed use of non-portable makefile constructs
2773 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
2774 ** Make sure all tests honor `$TMPDIR'
2777 Changes in 1.8.4 (since 1.8.3)
2781 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
2782 ** Fixed a segmentation fault which occurred when displaying the
2783 backtrace of a stack with a promise object (made by `delay') in it.
2784 ** Make `accept' leave guile mode while blocking
2785 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
2786 ** Fixed a build problem on AIX (use of func_data identifier)
2787 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
2788 called with an associator proc that returns neither a pair nor #f.
2789 ** Secondary threads now always return a valid module for (current-module).
2790 ** Avoid MacOS build problems caused by incorrect combination of "64"
2791 system and library calls.
2792 ** `guile-snarf' now honors `$TMPDIR'
2793 ** `guile-config compile' now reports CPPFLAGS used at compile-time
2794 ** Fixed build with Sun Studio (Solaris 9)
2795 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
2796 uniform vectors on AIX.
2797 ** Fixed a deadlock that occurs upon GC with multiple threads.
2798 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
2799 ** Fixed autotool-derived build problems on AIX 6.1.
2800 ** Fixed NetBSD/alpha support
2801 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
2803 * New modules (see the manual for details)
2807 * Documentation fixes and improvements
2809 ** Removed premature breakpoint documentation
2811 The features described are not available in the series of 1.8.x
2812 releases, so the documentation was misleading and has been removed.
2814 ** More about Guile's default *random-state* variable
2816 ** GOOPS: more about how to use `next-method'
2818 * Changes to the distribution
2820 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
2822 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
2823 General Public License, and the few incorrect files have now been
2824 fixed to agree with the rest of the Guile distribution.
2826 ** Removed unnecessary extra copies of COPYING*
2828 The distribution now contains a single COPYING.LESSER at its top level.
2831 Changes in 1.8.3 (since 1.8.2)
2833 * New modules (see the manual for details)
2840 ** The `(ice-9 slib)' module now works as expected
2841 ** Expressions like "(set! 'x #t)" no longer yield a crash
2842 ** Warnings about duplicate bindings now go to stderr
2843 ** A memory leak in `make-socket-address' was fixed
2844 ** Alignment issues (e.g., on SPARC) in network routines were fixed
2845 ** A threading issue that showed up at least on NetBSD was fixed
2846 ** Build problems on Solaris and IRIX fixed
2848 * Implementation improvements
2850 ** The reader is now faster, which reduces startup time
2851 ** Procedures returned by `record-accessor' and `record-modifier' are faster
2854 Changes in 1.8.2 (since 1.8.1):
2856 * New procedures (see the manual for details)
2858 ** set-program-arguments
2861 * Incompatible changes
2863 ** The body of a top-level `define' no longer sees the binding being created
2865 In a top-level `define', the binding being created is no longer visible
2866 from the `define' body. This breaks code like
2867 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
2868 unbound in the body. However, such code was not R5RS-compliant anyway,
2873 ** Fractions were not `equal?' if stored in unreduced form.
2874 (A subtle problem, since printing a value reduced it, making it work.)
2875 ** srfi-60 `copy-bit' failed on 64-bit systems
2876 ** "guile --use-srfi" option at the REPL can replace core functions
2877 (Programs run with that option were ok, but in the interactive REPL
2878 the core bindings got priority, preventing SRFI replacements or
2880 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
2881 ** `kill' on mingw throws an error for a PID other than oneself
2882 ** Procedure names are attached to procedure-with-setters
2883 ** Array read syntax works with negative lower bound
2884 ** `array-in-bounds?' fix if an array has different lower bounds on each index
2885 ** `*' returns exact 0 for "(* inexact 0)"
2886 This follows what it always did for "(* 0 inexact)".
2887 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
2888 ** SRFI-19: `date->julian-day' did not account for timezone offset
2889 ** `ttyname' no longer crashes when passed a non-tty argument
2890 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
2891 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
2892 ** GOOPS: Fixed a bug in `method-more-specific?'
2893 ** Build problems on Solaris fixed
2894 ** Build problems on HP-UX IA64 fixed
2895 ** Build problems on MinGW fixed
2898 Changes in 1.8.1 (since 1.8.0):
2900 * LFS functions are now used to access 64-bit files on 32-bit systems.
2902 * New procedures (see the manual for details)
2904 ** primitive-_exit - [Scheme] the-root-module
2905 ** scm_primitive__exit - [C]
2906 ** make-completion-function - [Scheme] (ice-9 readline)
2907 ** scm_c_locale_stringn_to_number - [C]
2908 ** scm_srfi1_append_reverse [C]
2909 ** scm_srfi1_append_reverse_x [C]
2917 ** Build problems have been fixed on MacOS, SunOS, and QNX.
2919 ** `strftime' fix sign of %z timezone offset.
2921 ** A one-dimensional array can now be 'equal?' to a vector.
2923 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
2925 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
2927 ** `record-accessor' and `record-modifier' now have strict type checks.
2929 Record accessor and modifier procedures now throw an error if the
2930 record type of the record they're given is not the type expected.
2931 (Previously accessors returned #f and modifiers silently did nothing).
2933 ** It is now OK to use both autoload and use-modules on a given module.
2935 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
2937 Previously there was no checking on primatives like make-vector that
2938 accept "one or two" arguments. Now there is.
2940 ** The srfi-1 assoc function now calls its equality predicate properly.
2942 Previously srfi-1 assoc would call the equality predicate with the key
2943 last. According to the SRFI, the key should be first.
2945 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
2947 ** The array-set! procedure no longer segfaults when given a bit vector.
2949 ** Bugs in make-shared-array have been fixed.
2951 ** string<? and friends now follow char<? etc order on 8-bit chars.
2953 ** The format procedure now handles inf and nan values for ~f correctly.
2955 ** exact->inexact should no longer overflow when given certain large fractions.
2957 ** srfi-9 accessor and modifier procedures now have strict record type checks.
2959 This matches the srfi-9 specification.
2961 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
2963 Previously the (ice-9 ftw) procedures would ignore any file that had
2964 the same inode number as a file they had already seen, even if that
2965 file was on a different device.
2968 Changes in 1.8.0 (changes since the 1.6.x series):
2970 * Changes to the distribution
2972 ** Guile is now licensed with the GNU Lesser General Public License.
2974 ** The manual is now licensed with the GNU Free Documentation License.
2976 ** Guile now requires GNU MP (http://swox.com/gmp).
2978 Guile now uses the GNU MP library for arbitrary precision arithmetic.
2980 ** Guile now has separate private and public configuration headers.
2982 That is, things like HAVE_STRING_H no longer leak from Guile's
2985 ** Guile now provides and uses an "effective" version number.
2987 Guile now provides scm_effective_version and effective-version
2988 functions which return the "effective" version number. This is just
2989 the normal full version string without the final micro-version number,
2990 so the current effective-version is "1.8". The effective version
2991 should remain unchanged during a stable series, and should be used for
2992 items like the versioned share directory name
2993 i.e. /usr/share/guile/1.8.
2995 Providing an unchanging version number during a stable release for
2996 things like the versioned share directory can be particularly
2997 important for Guile "add-on" packages, since it provides a directory
2998 that they can install to that won't be changed out from under them
2999 with each micro release during a stable series.
3001 ** Thread implementation has changed.
3003 When you configure "--with-threads=null", you will get the usual
3004 threading API (call-with-new-thread, make-mutex, etc), but you can't
3005 actually create new threads. Also, "--with-threads=no" is now
3006 equivalent to "--with-threads=null". This means that the thread API
3007 is always present, although you might not be able to create new
3010 When you configure "--with-threads=pthreads" or "--with-threads=yes",
3011 you will get threads that are implemented with the portable POSIX
3012 threads. These threads can run concurrently (unlike the previous
3013 "coop" thread implementation), but need to cooperate for things like
3016 The default is "pthreads", unless your platform doesn't have pthreads,
3017 in which case "null" threads are used.
3019 See the manual for details, nodes "Initialization", "Multi-Threading",
3020 "Blocking", and others.
3022 ** There is the new notion of 'discouraged' features.
3024 This is a milder form of deprecation.
3026 Things that are discouraged should not be used in new code, but it is
3027 OK to leave them in old code for now. When a discouraged feature is
3028 used, no warning message is printed like there is for 'deprecated'
3029 features. Also, things that are merely discouraged are nevertheless
3030 implemented efficiently, while deprecated features can be very slow.
3032 You can omit discouraged features from libguile by configuring it with
3033 the '--disable-discouraged' option.
3035 ** Deprecation warnings can be controlled at run-time.
3037 (debug-enable 'warn-deprecated) switches them on and (debug-disable
3038 'warn-deprecated) switches them off.
3040 ** Support for SRFI 61, extended cond syntax for multiple values has
3043 This SRFI is always available.
3045 ** Support for require-extension, SRFI-55, has been added.
3047 The SRFI-55 special form `require-extension' has been added. It is
3048 available at startup, and provides a portable way to load Scheme
3049 extensions. SRFI-55 only requires support for one type of extension,
3050 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
3053 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
3055 The (srfi srfi-26) module is an implementation of SRFI-26 which
3056 provides the `cut' and `cute' syntax. These may be used to specialize
3057 parameters without currying.
3059 ** New module (srfi srfi-31)
3061 This is an implementation of SRFI-31 which provides a special form
3062 `rec' for recursive evaluation.
3064 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
3065 been merged with the core, making their functionality always
3068 The modules are still available, tho, and you could use them together
3069 with a renaming import, for example.
3071 ** Guile no longer includes its own version of libltdl.
3073 The official version is good enough now.
3075 ** The --enable-htmldoc option has been removed from 'configure'.
3077 Support for translating the documentation into HTML is now always
3078 provided. Use 'make html'.
3080 ** New module (ice-9 serialize):
3082 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
3083 don't trust the thread safety of most of your program, but where you
3084 have some section(s) of code which you consider can run in parallel to
3085 other sections. See ice-9/serialize.scm for more information.
3087 ** The configure option '--disable-arrays' has been removed.
3089 Support for arrays and uniform numeric arrays is now always included
3092 * Changes to the stand-alone interpreter
3094 ** New command line option `-L'.
3096 This option adds a directory to the front of the load path.
3098 ** New command line option `--no-debug'.
3100 Specifying `--no-debug' on the command line will keep the debugging
3101 evaluator turned off, even for interactive sessions.
3103 ** User-init file ~/.guile is now loaded with the debugging evaluator.
3105 Previously, the normal evaluator would have been used. Using the
3106 debugging evaluator gives better error messages.
3108 ** The '-e' option now 'read's its argument.
3110 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
3111 be used with '-e'. For example, you can now write a script like
3114 exec guile -e '(@ (demo) main)' -s "$0" "$@"
3117 (define-module (demo)
3121 (format #t "Demo: ~a~%" args))
3124 * Changes to Scheme functions and syntax
3126 ** Guardians have changed back to their original semantics
3128 Guardians now behave like described in the paper by Dybvig et al. In
3129 particular, they no longer make guarantees about the order in which
3130 they return objects, and they can no longer be greedy.
3132 They no longer drop cyclic data structures.
3134 The C function scm_make_guardian has been changed incompatibly and no
3135 longer takes the 'greedy_p' argument.
3137 ** New function hashx-remove!
3139 This function completes the set of 'hashx' functions.
3141 ** The concept of dynamic roots has been factored into continuation
3142 barriers and dynamic states.
3144 Each thread has a current dynamic state that carries the values of the
3145 fluids. You can create and copy dynamic states and use them as the
3146 second argument for 'eval'. See "Fluids and Dynamic States" in the
3149 To restrict the influence that captured continuations can have on the
3150 control flow, you can errect continuation barriers. See "Continuation
3151 Barriers" in the manual.
3153 The function call-with-dynamic-root now essentially temporarily
3154 installs a new dynamic state and errects a continuation barrier.
3156 ** The default load path no longer includes "." at the end.
3158 Automatically loading modules from the current directory should not
3159 happen by default. If you want to allow it in a more controlled
3160 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
3161 variable %load-path.
3163 ** The uniform vector and array support has been overhauled.
3165 It now complies with SRFI-4 and the weird prototype based uniform
3166 array creation has been deprecated. See the manual for more details.
3168 Some non-compatible changes have been made:
3169 - characters can no longer be stored into byte arrays.
3170 - strings and bit vectors are no longer considered to be uniform numeric
3172 - array-rank throws an error for non-arrays instead of returning zero.
3173 - array-ref does no longer accept non-arrays when no indices are given.
3175 There is the new notion of 'generalized vectors' and corresponding
3176 procedures like 'generalized-vector-ref'. Generalized vectors include
3177 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
3179 Arrays use generalized vectors as their storage, so that you still
3180 have arrays of characters, bits, etc. However, uniform-array-read!
3181 and uniform-array-write can no longer read/write strings and
3184 ** There is now support for copy-on-write substrings, mutation-sharing
3185 substrings and read-only strings.
3187 Three new procedures are related to this: substring/shared,
3188 substring/copy, and substring/read-only. See the manual for more
3191 ** Backtraces will now highlight the value that caused the error.
3193 By default, these values are enclosed in "{...}", such as in this
3202 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
3203 <unnamed port>:1:1: Wrong type (expecting pair): a
3204 ABORT: (wrong-type-arg)
3206 The prefix and suffix used for highlighting can be set via the two new
3207 printer options 'highlight-prefix' and 'highlight-suffix'. For
3208 example, putting this into ~/.guile will output the bad value in bold
3209 on an ANSI terminal:
3211 (print-set! highlight-prefix "\x1b[1m")
3212 (print-set! highlight-suffix "\x1b[22m")
3215 ** 'gettext' support for internationalization has been added.
3217 See the manual for details.
3219 ** New syntax '@' and '@@':
3221 You can now directly refer to variables exported from a module by
3224 (@ MODULE-NAME VARIABLE-NAME)
3226 For example (@ (ice-9 pretty-print) pretty-print) will directly access
3227 the pretty-print variable exported from the (ice-9 pretty-print)
3228 module. You don't need to 'use' that module first. You can also use
3229 '@' as a target of 'set!', as in (set! (@ mod var) val).
3231 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
3232 but it can also access variables that have not been exported. It is
3233 intended only for kluges and temporary fixes and for debugging, not
3236 ** Keyword syntax has been made more disciplined.
3238 Previously, the name of a keyword was read as a 'token' but printed as
3239 a symbol. Now, it is read as a general Scheme datum which must be a
3250 ERROR: In expression (a b c):
3254 ERROR: Unbound variable: foo
3259 ERROR: Wrong type (expecting symbol): 12
3263 ERROR: Wrong type (expecting symbol): (a b c)
3267 ** The printing of symbols that might look like keywords can be
3270 The new printer option 'quote-keywordish-symbols' controls how symbols
3271 are printed that have a colon as their first or last character. The
3272 default now is to only quote a symbol with #{...}# when the read
3273 option 'keywords' is not '#f'. Thus:
3275 guile> (define foo (string->symbol ":foo"))
3276 guile> (read-set! keywords #f)
3279 guile> (read-set! keywords 'prefix)
3282 guile> (print-set! quote-keywordish-symbols #f)
3286 ** 'while' now provides 'break' and 'continue'
3288 break and continue were previously bound in a while loop, but not
3289 documented, and continue didn't quite work properly. The undocumented
3290 parameter to break which gave a return value for the while has been
3293 ** 'call-with-current-continuation' is now also available under the name
3296 ** The module system now checks for duplicate bindings.
3298 The module system now can check for name conflicts among imported
3301 The behavior can be controlled by specifying one or more 'duplicates'
3302 handlers. For example, to make Guile return an error for every name
3305 (define-module (foo)
3310 The new default behavior of the module system when a name collision
3311 has been detected is to
3313 1. Give priority to bindings marked as a replacement.
3314 2. Issue a warning (different warning if overriding core binding).
3315 3. Give priority to the last encountered binding (this corresponds to
3318 If you want the old behavior back without replacements or warnings you
3321 (default-duplicate-binding-handler 'last)
3323 to your .guile init file.
3325 ** New define-module option: :replace
3327 :replace works as :export, but, in addition, marks the binding as a
3330 A typical example is `format' in (ice-9 format) which is a replacement
3331 for the core binding `format'.
3333 ** Adding prefixes to imported bindings in the module system
3335 There is now a new :use-module option :prefix. It can be used to add
3336 a prefix to all imported bindings.
3338 (define-module (foo)
3339 :use-module ((bar) :prefix bar:))
3341 will import all bindings exported from bar, but rename them by adding
3344 ** Conflicting generic functions can be automatically merged.
3346 When two imported bindings conflict and they are both generic
3347 functions, the two functions can now be merged automatically. This is
3348 activated with the 'duplicates' handler 'merge-generics'.
3350 ** New function: effective-version
3352 Returns the "effective" version number. This is just the normal full
3353 version string without the final micro-version number. See "Changes
3354 to the distribution" above.
3356 ** New threading functions: parallel, letpar, par-map, and friends
3358 These are convenient ways to run calculations in parallel in new
3359 threads. See "Parallel forms" in the manual for details.
3361 ** New function 'try-mutex'.
3363 This function will attempt to lock a mutex but will return immediately
3364 instead of blocking and indicate failure.
3366 ** Waiting on a condition variable can have a timeout.
3368 The function 'wait-condition-variable' now takes a third, optional
3369 argument that specifies the point in time where the waiting should be
3372 ** New function 'broadcast-condition-variable'.
3374 ** New functions 'all-threads' and 'current-thread'.
3376 ** Signals and system asyncs work better with threads.
3378 The function 'sigaction' now takes a fourth, optional, argument that
3379 specifies the thread that the handler should run in. When the
3380 argument is omitted, the handler will run in the thread that called
3383 Likewise, 'system-async-mark' takes a second, optional, argument that
3384 specifies the thread that the async should run in. When it is
3385 omitted, the async will run in the thread that called
3386 'system-async-mark'.
3388 C code can use the new functions scm_sigaction_for_thread and
3389 scm_system_async_mark_for_thread to pass the new thread argument.
3391 When a thread blocks on a mutex, a condition variable or is waiting
3392 for IO to be possible, it will still execute system asyncs. This can
3393 be used to interrupt such a thread by making it execute a 'throw', for
3396 ** The function 'system-async' is deprecated.
3398 You can now pass any zero-argument procedure to 'system-async-mark'.
3399 The function 'system-async' will just return its argument unchanged
3402 ** New functions 'call-with-blocked-asyncs' and
3403 'call-with-unblocked-asyncs'
3405 The expression (call-with-blocked-asyncs PROC) will call PROC and will
3406 block execution of system asyncs for the current thread by one level
3407 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
3408 procedure and will unblock the execution of system asyncs by one
3409 level for the current thread.
3411 Only system asyncs are affected by these functions.
3413 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
3415 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
3416 instead. Those functions are easier to use correctly and can be
3419 ** New function 'unsetenv'.
3421 ** New macro 'define-syntax-public'.
3423 It works like 'define-syntax' and also exports the defined macro (but
3426 ** There is support for Infinity and NaNs.
3428 Following PLT Scheme, Guile can now work with infinite numbers, and
3431 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
3432 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
3433 "+nan.0"). These numbers are inexact and have no exact counterpart.
3435 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
3436 sign of the dividend. The infinities are integers, and they answer #t
3437 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
3438 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
3449 ERROR: Numerical overflow
3451 Two new predicates 'inf?' and 'nan?' can be used to test for the
3454 ** Inexact zero can have a sign.
3456 Guile can now distinguish between plus and minus inexact zero, if your
3457 platform supports this, too. The two zeros are equal according to
3458 '=', but not according to 'eqv?'. For example
3469 ** Guile now has exact rationals.
3471 Guile can now represent fractions such as 1/3 exactly. Computing with
3472 them is also done exactly, of course:
3477 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
3478 for exact arguments.
3480 For example: (floor 2) now returns an exact 2 where in the past it
3481 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
3483 ** inexact->exact no longer returns only integers.
3485 Without exact rationals, the closest exact number was always an
3486 integer, but now inexact->exact returns the fraction that is exactly
3487 equal to a floating point number. For example:
3489 (inexact->exact 1.234)
3490 => 694680242521899/562949953421312
3492 When you want the old behavior, use 'round' explicitly:
3494 (inexact->exact (round 1.234))
3497 ** New function 'rationalize'.
3499 This function finds a simple fraction that is close to a given real
3500 number. For example (and compare with inexact->exact above):
3502 (rationalize (inexact->exact 1.234) 1/2000)
3505 Note that, as required by R5RS, rationalize returns only then an exact
3506 result when both its arguments are exact.
3508 ** 'odd?' and 'even?' work also for inexact integers.
3510 Previously, (odd? 1.0) would signal an error since only exact integers
3511 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
3512 returns #f and (odd? 1.5) signals an error.
3514 ** Guile now has uninterned symbols.
3516 The new function 'make-symbol' will return an uninterned symbol. This
3517 is a symbol that is unique and is guaranteed to remain unique.
3518 However, uninterned symbols can not yet be read back in.
3520 Use the new function 'symbol-interned?' to check whether a symbol is
3523 ** pretty-print has more options.
3525 The function pretty-print from the (ice-9 pretty-print) module can now
3526 also be invoked with keyword arguments that control things like
3527 maximum output width. See the manual for details.
3529 ** Variables have no longer a special behavior for `equal?'.
3531 Previously, comparing two variables with `equal?' would recursivly
3532 compare their values. This is no longer done. Variables are now only
3533 `equal?' if they are `eq?'.
3535 ** `(begin)' is now valid.
3537 You can now use an empty `begin' form. It will yield #<unspecified>
3538 when evaluated and simply be ignored in a definition context.
3540 ** Deprecated: procedure->macro
3542 Change your code to use 'define-macro' or r5rs macros. Also, be aware
3543 that macro expansion will not be done during evaluation, but prior to
3546 ** Soft ports now allow a `char-ready?' procedure
3548 The vector argument to `make-soft-port' can now have a length of
3549 either 5 or 6. (Previously the length had to be 5.) The optional 6th
3550 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
3551 that returns the number of characters that can be read immediately
3552 without the soft port blocking.
3554 ** Deprecated: undefine
3556 There is no replacement for undefine.
3558 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
3559 have been discouraged.
3561 They are relics from a time where a keyword like #:foo was used
3562 directly as a Tcl option "-foo" and thus keywords were internally
3563 stored as a symbol with a starting dash. We now store a symbol
3566 Use symbol->keyword and keyword->symbol instead.
3568 ** The `cheap' debug option is now obsolete
3570 Evaluator trap calls are now unconditionally "cheap" - in other words,
3571 they pass a debug object to the trap handler rather than a full
3572 continuation. The trap handler code can capture a full continuation
3573 by using `call-with-current-continuation' in the usual way, if it so
3576 The `cheap' option is retained for now so as not to break existing
3577 code which gets or sets it, but setting it now has no effect. It will
3578 be removed in the next major Guile release.
3580 ** Evaluator trap calls now support `tweaking'
3582 `Tweaking' means that the trap handler code can modify the Scheme
3583 expression that is about to be evaluated (in the case of an
3584 enter-frame trap) or the value that is being returned (in the case of
3585 an exit-frame trap). The trap handler code indicates that it wants to
3586 do this by returning a pair whose car is the symbol 'instead and whose
3587 cdr is the modified expression or return value.
3589 * Changes to the C interface
3591 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
3592 take a 'delete' function argument.
3594 This argument makes no sense since the delete function is used to
3595 remove a pair from an alist, and this must not be configurable.
3597 This is an incompatible change.
3599 ** The GH interface is now subject to the deprecation mechanism
3601 The GH interface has been deprecated for quite some time but now it is
3602 actually removed from Guile when it is configured with
3603 --disable-deprecated.
3605 See the manual "Transitioning away from GH" for more information.
3607 ** A new family of functions for converting between C values and
3608 Scheme values has been added.
3610 These functions follow a common naming scheme and are designed to be
3611 easier to use, thread-safe and more future-proof than the older
3614 - int scm_is_* (...)
3616 These are predicates that return a C boolean: 1 or 0. Instead of
3617 SCM_NFALSEP, you can now use scm_is_true, for example.
3619 - <type> scm_to_<type> (SCM val, ...)
3621 These are functions that convert a Scheme value into an appropriate
3622 C value. For example, you can use scm_to_int to safely convert from
3625 - SCM scm_from_<type> (<type> val, ...)
3627 These functions convert from a C type to a SCM value; for example,
3628 scm_from_int for ints.
3630 There is a huge number of these functions, for numbers, strings,
3631 symbols, vectors, etc. They are documented in the reference manual in
3632 the API section together with the types that they apply to.
3634 ** New functions for dealing with complex numbers in C have been added.
3636 The new functions are scm_c_make_rectangular, scm_c_make_polar,
3637 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
3638 They work like scm_make_rectangular etc but take or return doubles
3641 ** The function scm_make_complex has been discouraged.
3643 Use scm_c_make_rectangular instead.
3645 ** The INUM macros have been deprecated.
3647 A lot of code uses these macros to do general integer conversions,
3648 although the macros only work correctly with fixnums. Use the
3649 following alternatives.
3651 SCM_INUMP -> scm_is_integer or similar
3652 SCM_NINUMP -> !scm_is_integer or similar
3653 SCM_MAKINUM -> scm_from_int or similar
3654 SCM_INUM -> scm_to_int or similar
3656 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
3657 do the validating for you.
3659 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
3660 have been discouraged.
3662 Use the newer scm_to_<type> and scm_from_<type> functions instead for
3663 new code. The functions have been discouraged since they don't fit
3666 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
3668 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
3669 evaluates its argument twice. Use scm_is_true, etc. instead for new
3672 ** The macro SCM_EQ_P has been discouraged.
3674 Use scm_is_eq for new code, which fits better into the naming
3677 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
3680 Use the function scm_is_pair or scm_is_null instead.
3682 ** The functions scm_round and scm_truncate have been deprecated and
3683 are now available as scm_c_round and scm_c_truncate, respectively.
3685 These functions occupy the names that scm_round_number and
3686 scm_truncate_number should have.
3688 ** The functions scm_c_string2str, scm_c_substring2str, and
3689 scm_c_symbol2str have been deprecated.
3691 Use scm_to_locale_stringbuf or similar instead, maybe together with
3694 ** New functions scm_c_make_string, scm_c_string_length,
3695 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
3696 scm_c_substring_shared, scm_c_substring_copy.
3698 These are like scm_make_string, scm_length, etc. but are slightly
3699 easier to use from C.
3701 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
3702 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
3704 They export too many assumptions about the implementation of strings
3705 and symbols that are no longer true in the presence of
3706 mutation-sharing substrings and when Guile switches to some form of
3709 When working with strings, it is often best to use the normal string
3710 functions provided by Guile, such as scm_c_string_ref,
3711 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
3712 manual since many more such functions are now provided than
3715 When you want to convert a SCM string to a C string, use the
3716 scm_to_locale_string function or similar instead. For symbols, use
3717 scm_symbol_to_string and then work with that string. Because of the
3718 new string representation, scm_symbol_to_string does not need to copy
3719 and is thus quite efficient.
3721 ** Some string, symbol and keyword functions have been discouraged.
3723 They don't fit into the uniform naming scheme and are not explicit
3724 about the character encoding.
3726 Replace according to the following table:
3728 scm_allocate_string -> scm_c_make_string
3729 scm_take_str -> scm_take_locale_stringn
3730 scm_take0str -> scm_take_locale_string
3731 scm_mem2string -> scm_from_locale_stringn
3732 scm_str2string -> scm_from_locale_string
3733 scm_makfrom0str -> scm_from_locale_string
3734 scm_mem2symbol -> scm_from_locale_symboln
3735 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
3736 scm_str2symbol -> scm_from_locale_symbol
3738 SCM_SYMBOL_HASH -> scm_hashq
3739 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
3741 scm_c_make_keyword -> scm_from_locale_keyword
3743 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
3744 now also available to C code.
3746 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
3748 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
3749 the latter returns the true name of the keyword, not the 'dash name',
3750 as SCM_KEYWORDSYM used to do.
3752 ** A new way to access arrays in a thread-safe and efficient way has
3755 See the manual, node "Accessing Arrays From C".
3757 ** The old uniform vector and bitvector implementations have been
3758 unceremoniously removed.
3760 This implementation exposed the details of the tagging system of
3761 Guile. Use the new C API explained in the manual in node "Uniform
3762 Numeric Vectors" and "Bit Vectors", respectively.
3764 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
3765 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
3766 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
3767 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
3768 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
3769 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
3772 ** The macros dealing with vectors have been deprecated.
3774 Use the new functions scm_is_vector, scm_vector_elements,
3775 scm_vector_writable_elements, etc, or scm_is_simple_vector,
3776 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
3777 manual for more details.
3779 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
3780 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
3782 The following macros have been removed: SCM_VECTOR_BASE,
3783 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
3784 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
3786 ** Some C functions and macros related to arrays have been deprecated.
3788 Migrate according to the following table:
3790 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
3791 scm_make_ra -> scm_make_array
3792 scm_shap2ra -> scm_make_array
3793 scm_cvref -> scm_c_generalized_vector_ref
3794 scm_ra_set_contp -> do not use
3795 scm_aind -> scm_array_handle_pos
3796 scm_raprin1 -> scm_display or scm_write
3798 SCM_ARRAYP -> scm_is_array
3799 SCM_ARRAY_NDIM -> scm_c_array_rank
3800 SCM_ARRAY_DIMS -> scm_array_handle_dims
3801 SCM_ARRAY_CONTP -> do not use
3802 SCM_ARRAY_MEM -> do not use
3803 SCM_ARRAY_V -> scm_array_handle_elements or similar
3804 SCM_ARRAY_BASE -> do not use
3806 ** SCM_CELL_WORD_LOC has been deprecated.
3808 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
3809 to a SCM, as opposed to a pointer to a scm_t_bits.
3811 This was done to allow the correct use of pointers into the Scheme
3812 heap. Previously, the heap words were of type scm_t_bits and local
3813 variables and function arguments were of type SCM, making it
3814 non-standards-conformant to have a pointer that can point to both.
3816 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
3818 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
3819 second and third words of double smobs. Likewise for
3820 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
3822 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
3823 used to get and set the 16 exra bits in the zeroth word of a smob.
3825 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
3826 accesing the first immediate word of a smob as a SCM value, and there
3827 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
3828 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
3830 ** New way to deal with non-local exits and re-entries.
3832 There is a new set of functions that essentially do what
3833 scm_internal_dynamic_wind does, but in a way that is more convenient
3834 for C code in some situations. Here is a quick example of how to
3835 prevent a potential memory leak:
3842 scm_dynwind_begin (0);
3844 mem = scm_malloc (100);
3845 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
3847 /* MEM would leak if BAR throws an error.
3848 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
3855 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
3856 SCM_DYNWIND_END as well.
3860 For full documentation, see the node "Dynamic Wind" in the manual.
3862 ** New function scm_dynwind_free
3864 This function calls 'free' on a given pointer when a dynwind context
3865 is left. Thus the call to scm_dynwind_unwind_handler above could be
3866 replaced with simply scm_dynwind_free (mem).
3868 ** New functions scm_c_call_with_blocked_asyncs and
3869 scm_c_call_with_unblocked_asyncs
3871 Like scm_call_with_blocked_asyncs etc. but for C functions.
3873 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
3875 In addition to scm_c_call_with_blocked_asyncs you can now also use
3876 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
3877 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
3879 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
3880 SCM_REALLOW_INTS have been deprecated.
3882 They do no longer fulfill their original role of blocking signal
3883 delivery. Depending on what you want to achieve, replace a pair of
3884 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
3885 mutex, blocks asyncs, or both. See node "Critical Sections" in the
3888 ** The value 'scm_mask_ints' is no longer writable.
3890 Previously, you could set scm_mask_ints directly. This is no longer
3891 possible. Use scm_c_call_with_blocked_asyncs and
3892 scm_c_call_with_unblocked_asyncs instead.
3894 ** New way to temporarily set the current input, output or error ports
3896 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
3897 context' (see above). <foo> is one of "input", "output" or "error".
3899 ** New way to temporarily set fluids
3901 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
3902 above) to temporarily set the value of a fluid.
3904 ** New types scm_t_intmax and scm_t_uintmax.
3906 On platforms that have them, these types are identical to intmax_t and
3907 uintmax_t, respectively. On other platforms, they are identical to
3908 the largest integer types that Guile knows about.
3910 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
3912 You should not have used them.
3914 ** Many public #defines with generic names have been made private.
3916 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
3917 private or renamed with a more suitable public name.
3919 ** The macro SCM_TYP16S has been deprecated.
3921 This macro is not intended for public use.
3923 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
3925 Use scm_is_true (scm_inexact_p (...)) instead.
3927 ** The macro SCM_SLOPPY_REALP has been deprecated.
3929 Use scm_is_real instead.
3931 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
3933 Use scm_is_complex instead.
3935 ** Some preprocessor defines have been deprecated.
3937 These defines indicated whether a certain feature was present in Guile
3938 or not. Going forward, assume that the features are always present.
3940 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
3941 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
3943 The following macros have been removed completely: MEMOIZE_LOCALS,
3944 SCM_RECKLESS, SCM_CAUTIOUS.
3946 ** The preprocessor define STACK_DIRECTION has been deprecated.
3948 There should be no need to know about the stack direction for ordinary
3951 ** New function: scm_effective_version
3953 Returns the "effective" version number. This is just the normal full
3954 version string without the final micro-version number. See "Changes
3955 to the distribution" above.
3957 ** The function scm_call_with_new_thread has a new prototype.
3959 Instead of taking a list with the thunk and handler, these two
3960 arguments are now passed directly:
3962 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
3964 This is an incompatible change.
3966 ** New snarfer macro SCM_DEFINE_PUBLIC.
3968 This is like SCM_DEFINE, but also calls scm_c_export for the defined
3969 function in the init section.
3971 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
3973 ** Garbage collector rewrite.
3975 The garbage collector is cleaned up a lot, and now uses lazy
3976 sweeping. This is reflected in the output of (gc-stats); since cells
3977 are being freed when they are allocated, the cells-allocated field
3978 stays roughly constant.
3980 For malloc related triggers, the behavior is changed. It uses the same
3981 heuristic as the cell-triggered collections. It may be tuned with the
3982 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
3983 for minimum yield of malloc related triggers. The default is 40.
3984 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
3987 Debugging operations for the freelist have been deprecated, along with
3988 the C variables that control garbage collection. The environment
3989 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
3990 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
3992 For understanding the memory usage of a GUILE program, the routine
3993 gc-live-object-stats returns an alist containing the number of live
3994 objects for every type.
3997 ** The function scm_definedp has been renamed to scm_defined_p
3999 The name scm_definedp is deprecated.
4001 ** The struct scm_cell type has been renamed to scm_t_cell
4003 This is in accordance to Guile's naming scheme for types. Note that
4004 the name scm_cell is now used for a function that allocates and
4005 initializes a new cell (see below).
4007 ** New functions for memory management
4009 A new set of functions for memory management has been added since the
4010 old way (scm_must_malloc, scm_must_free, etc) was error prone and
4011 indeed, Guile itself contained some long standing bugs that could
4012 cause aborts in long running programs.
4014 The new functions are more symmetrical and do not need cooperation
4015 from smob free routines, among other improvements.
4017 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
4018 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
4019 scm_gc_free, scm_gc_register_collectable_memory, and
4020 scm_gc_unregister_collectable_memory. Refer to the manual for more
4021 details and for upgrading instructions.
4023 The old functions for memory management have been deprecated. They
4024 are: scm_must_malloc, scm_must_realloc, scm_must_free,
4025 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
4027 ** Declarations of exported features are marked with SCM_API.
4029 Every declaration of a feature that belongs to the exported Guile API
4030 has been marked by adding the macro "SCM_API" to the start of the
4031 declaration. This macro can expand into different things, the most
4032 common of which is just "extern" for Unix platforms. On Win32, it can
4033 be used to control which symbols are exported from a DLL.
4035 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
4036 will expand into "__declspec (dllimport) extern", which is needed for
4037 linking to the Guile DLL in Windows.
4039 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
4040 SCM_SRFI4_IMPORT, for the corresponding libraries.
4042 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
4044 Use the new functions scm_cell and scm_double_cell instead. The old
4045 macros had problems because with them allocation and initialization
4046 was separated and the GC could sometimes observe half initialized
4047 cells. Only careful coding by the user of SCM_NEWCELL and
4048 SCM_NEWCELL2 could make this safe and efficient.
4050 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
4052 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
4055 ** SRCBRKP has been deprecated.
4057 Use scm_c_source_property_breakpoint_p instead.
4059 ** Deprecated: scm_makmacro
4061 Change your code to use either scm_makmmacro or to define macros in
4062 Scheme, using 'define-macro'.
4064 ** New function scm_c_port_for_each.
4066 This function is like scm_port_for_each but takes a pointer to a C
4067 function as the callback instead of a SCM value.
4069 ** The names scm_internal_select, scm_thread_sleep, and
4070 scm_thread_usleep have been discouraged.
4072 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
4074 ** The GC can no longer be blocked.
4076 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
4077 The GC can now run (partially) concurrently with other code and thus
4078 blocking it is not well defined.
4080 ** Many definitions have been removed that were previously deprecated.
4082 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
4083 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
4084 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
4085 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
4086 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
4087 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
4088 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
4089 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
4090 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
4091 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
4092 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
4093 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
4094 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
4095 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
4096 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
4097 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
4098 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
4099 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
4100 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
4101 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
4102 scm_vector_set_length_x, scm_contregs, scm_debug_info,
4103 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
4104 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
4105 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
4106 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
4107 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
4108 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
4109 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
4110 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
4111 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
4112 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
4114 * Changes to bundled modules
4118 Using the (ice-9 debug) module no longer automatically switches Guile
4119 to use the debugging evaluator. If you want to switch to the
4120 debugging evaluator (which is needed for backtrace information if you
4121 hit an error), please add an explicit "(debug-enable 'debug)" to your
4122 code just after the code to use (ice-9 debug).
4125 Changes since Guile 1.4:
4127 * Changes to the distribution
4129 ** A top-level TODO file is included.
4131 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
4133 Guile now always uses three numbers to represent the version,
4134 i.e. "1.6.5". The first number, 1, is the major version number, the
4135 second number, 6, is the minor version number, and the third number,
4136 5, is the micro version number. Changes in major version number
4137 indicate major changes in Guile.
4139 Minor version numbers that are even denote stable releases, and odd
4140 minor version numbers denote development versions (which may be
4141 unstable). The micro version number indicates a minor sub-revision of
4142 a given MAJOR.MINOR release.
4144 In keeping with the new scheme, (minor-version) and scm_minor_version
4145 no longer return everything but the major version number. They now
4146 just return the minor version number. Two new functions
4147 (micro-version) and scm_micro_version have been added to report the
4148 micro version number.
4150 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
4152 ** New preprocessor definitions are available for checking versions.
4154 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
4155 SCM_MICRO_VERSION to the appropriate integer values.
4157 ** Guile now actively warns about deprecated features.
4159 The new configure option `--enable-deprecated=LEVEL' and the
4160 environment variable GUILE_WARN_DEPRECATED control this mechanism.
4161 See INSTALL and README for more information.
4163 ** Guile is much more likely to work on 64-bit architectures.
4165 Guile now compiles and passes "make check" with only two UNRESOLVED GC
4166 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
4167 for the use of a test machine, and thanks to Stefan Jahn for ia64
4170 ** New functions: setitimer and getitimer.
4172 These implement a fairly direct interface to the libc functions of the
4175 ** The #. reader extension is now disabled by default.
4177 For safety reasons, #. evaluation is disabled by default. To
4178 re-enable it, set the fluid read-eval? to #t. For example:
4180 (fluid-set! read-eval? #t)
4182 but make sure you realize the potential security risks involved. With
4183 read-eval? enabled, reading a data file from an untrusted source can
4186 ** New SRFI modules have been added:
4188 SRFI-0 `cond-expand' is now supported in Guile, without requiring
4191 (srfi srfi-1) is a library containing many useful pair- and list-processing
4194 (srfi srfi-2) exports and-let*.
4196 (srfi srfi-4) implements homogeneous numeric vector datatypes.
4198 (srfi srfi-6) is a dummy module for now, since guile already provides
4199 all of the srfi-6 procedures by default: open-input-string,
4200 open-output-string, get-output-string.
4202 (srfi srfi-8) exports receive.
4204 (srfi srfi-9) exports define-record-type.
4206 (srfi srfi-10) exports define-reader-ctor and implements the reader
4209 (srfi srfi-11) exports let-values and let*-values.
4211 (srfi srfi-13) implements the SRFI String Library.
4213 (srfi srfi-14) implements the SRFI Character-Set Library.
4215 (srfi srfi-17) implements setter and getter-with-setter and redefines
4216 some accessor procedures as procedures with getters. (such as car,
4217 cdr, vector-ref etc.)
4219 (srfi srfi-19) implements the SRFI Time/Date Library.
4221 ** New scripts / "executable modules"
4223 Subdirectory "scripts" contains Scheme modules that are packaged to
4224 also be executable as scripts. At this time, these scripts are available:
4233 See README there for more info.
4235 These scripts can be invoked from the shell with the new program
4236 "guile-tools", which keeps track of installation directory for you.
4239 $ guile-tools display-commentary srfi/*.scm
4241 guile-tools is copied to the standard $bindir on "make install".
4243 ** New module (ice-9 stack-catch):
4245 stack-catch is like catch, but saves the current state of the stack in
4246 the fluid the-last-stack. This fluid can be useful when using the
4247 debugger and when re-throwing an error.
4249 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
4251 This has been done to prevent problems on lesser operating systems
4252 that can't tolerate `*'s in file names. The exported macro continues
4253 to be named `and-let*', of course.
4255 On systems that support it, there is also a compatibility module named
4256 (ice-9 and-let*). It will go away in the next release.
4258 ** New modules (oop goops) etc.:
4261 (oop goops describe)
4263 (oop goops active-slot)
4264 (oop goops composite-slot)
4266 The Guile Object Oriented Programming System (GOOPS) has been
4267 integrated into Guile. For further information, consult the GOOPS
4268 manual and tutorial in the `doc' directory.
4270 ** New module (ice-9 rdelim).
4272 This exports the following procedures which were previously defined
4273 in the default environment:
4275 read-line read-line! read-delimited read-delimited! %read-delimited!
4276 %read-line write-line
4278 For backwards compatibility the definitions are still imported into the
4279 default environment in this version of Guile. However you should add:
4281 (use-modules (ice-9 rdelim))
4283 to any program which uses the definitions, since this may change in
4286 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
4287 can be used for similar functionality.
4289 ** New module (ice-9 rw)
4291 This is a subset of the (scsh rw) module from guile-scsh. Currently
4292 it defines two procedures:
4294 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
4296 Read characters from a port or file descriptor into a string STR.
4297 A port must have an underlying file descriptor -- a so-called
4298 fport. This procedure is scsh-compatible and can efficiently read
4301 *** New function: write-string/partial str [port_or_fdes [start [end]]]
4303 Write characters from a string STR to a port or file descriptor.
4304 A port must have an underlying file descriptor -- a so-called
4305 fport. This procedure is mostly compatible and can efficiently
4306 write large strings.
4308 ** New module (ice-9 match)
4310 This module includes Andrew K. Wright's pattern matcher. See
4311 ice-9/match.scm for brief description or
4313 http://www.star-lab.com/wright/code.html
4315 for complete documentation.
4317 ** New module (ice-9 buffered-input)
4319 This module provides procedures to construct an input port from an
4320 underlying source of input that reads and returns its input in chunks.
4321 The underlying input source is a Scheme procedure, specified by the
4322 caller, which the port invokes whenever it needs more input.
4324 This is useful when building an input port whose back end is Readline
4325 or a UI element such as the GtkEntry widget.
4329 The reference and tutorial documentation that was previously
4330 distributed separately, as `guile-doc', is now included in the core
4331 Guile distribution. The documentation consists of the following
4334 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
4337 - The Guile Reference Manual (guile.texi) contains (or is intended to
4338 contain) reference documentation on all aspects of Guile.
4340 - The GOOPS Manual (goops.texi) contains both tutorial-style and
4341 reference documentation for using GOOPS, Guile's Object Oriented
4344 - The Revised^5 Report on the Algorithmic Language Scheme
4347 See the README file in the `doc' directory for more details.
4349 ** There are a couple of examples in the examples/ directory now.
4351 * Changes to the stand-alone interpreter
4353 ** New command line option `--use-srfi'
4355 Using this option, SRFI modules can be loaded on startup and be
4356 available right from the beginning. This makes programming portable
4357 Scheme programs easier.
4359 The option `--use-srfi' expects a comma-separated list of numbers,
4360 each representing a SRFI number to be loaded into the interpreter
4361 before starting evaluating a script file or the REPL. Additionally,
4362 the feature identifier for the loaded SRFIs is recognized by
4363 `cond-expand' when using this option.
4366 $ guile --use-srfi=8,13
4367 guile> (receive (x z) (values 1 2) (+ 1 2))
4369 guile> (string-pad "bla" 20)
4372 ** Guile now always starts up in the `(guile-user)' module.
4374 Previously, scripts executed via the `-s' option would run in the
4375 `(guile)' module and the repl would run in the `(guile-user)' module.
4376 Now every user action takes place in the `(guile-user)' module by
4379 * Changes to Scheme functions and syntax
4381 ** Character classifiers work for non-ASCII characters.
4383 The predicates `char-alphabetic?', `char-numeric?',
4384 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
4385 no longer check whether their arguments are ASCII characters.
4386 Previously, a character would only be considered alphabetic when it
4387 was also ASCII, for example.
4389 ** Previously deprecated Scheme functions have been removed:
4391 tag - no replacement.
4392 fseek - replaced by seek.
4393 list* - replaced by cons*.
4395 ** It's now possible to create modules with controlled environments
4399 (use-modules (ice-9 safe))
4400 (define m (make-safe-module))
4401 ;;; m will now be a module containing only a safe subset of R5RS
4402 (eval '(+ 1 2) m) --> 3
4403 (eval 'load m) --> ERROR: Unbound variable: load
4405 ** Evaluation of "()", the empty list, is now an error.
4407 Previously, the expression "()" evaluated to the empty list. This has
4408 been changed to signal a "missing expression" error. The correct way
4409 to write the empty list as a literal constant is to use quote: "'()".
4411 ** New concept of `Guile Extensions'.
4413 A Guile Extension is just a ordinary shared library that can be linked
4414 at run-time. We found it advantageous to give this simple concept a
4415 dedicated name to distinguish the issues related to shared libraries
4416 from the issues related to the module system.
4418 *** New function: load-extension
4420 Executing (load-extension lib init) is mostly equivalent to
4422 (dynamic-call init (dynamic-link lib))
4424 except when scm_register_extension has been called previously.
4425 Whenever appropriate, you should use `load-extension' instead of
4426 dynamic-link and dynamic-call.
4428 *** New C function: scm_c_register_extension
4430 This function registers a initialization function for use by
4431 `load-extension'. Use it when you don't want specific extensions to
4432 be loaded as shared libraries (for example on platforms that don't
4433 support dynamic linking).
4435 ** Auto-loading of compiled-code modules is deprecated.
4437 Guile used to be able to automatically find and link a shared
4438 library to satisfy requests for a module. For example, the module
4439 `(foo bar)' could be implemented by placing a shared library named
4440 "foo/libbar.so" (or with a different extension) in a directory on the
4443 This has been found to be too tricky, and is no longer supported. The
4444 shared libraries are now called "extensions". You should now write a
4445 small Scheme file that calls `load-extension' to load the shared
4446 library and initialize it explicitly.
4448 The shared libraries themselves should be installed in the usual
4449 places for shared libraries, with names like "libguile-foo-bar".
4451 For example, place this into a file "foo/bar.scm"
4453 (define-module (foo bar))
4455 (load-extension "libguile-foo-bar" "foobar_init")
4457 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
4459 `eval' is now R5RS, that is it takes two arguments.
4460 The second argument is an environment specifier, i.e. either
4462 (scheme-report-environment 5)
4463 (null-environment 5)
4464 (interaction-environment)
4470 ** The module system has been made more disciplined.
4472 The function `eval' will save and restore the current module around
4473 the evaluation of the specified expression. While this expression is
4474 evaluated, `(current-module)' will now return the right module, which
4475 is the module specified as the second argument to `eval'.
4477 A consequence of this change is that `eval' is not particularly
4478 useful when you want allow the evaluated code to change what module is
4479 designated as the current module and have this change persist from one
4480 call to `eval' to the next. The read-eval-print-loop is an example
4481 where `eval' is now inadequate. To compensate, there is a new
4482 function `primitive-eval' that does not take a module specifier and
4483 that does not save/restore the current module. You should use this
4484 function together with `set-current-module', `current-module', etc
4485 when you want to have more control over the state that is carried from
4486 one eval to the next.
4488 Additionally, it has been made sure that forms that are evaluated at
4489 the top level are always evaluated with respect to the current module.
4490 Previously, subforms of top-level forms such as `begin', `case',
4491 etc. did not respect changes to the current module although these
4492 subforms are at the top-level as well.
4494 To prevent strange behavior, the forms `define-module',
4495 `use-modules', `use-syntax', and `export' have been restricted to only
4496 work on the top level. The forms `define-public' and
4497 `defmacro-public' only export the new binding on the top level. They
4498 behave just like `define' and `defmacro', respectively, when they are
4499 used in a lexical environment.
4501 Also, `export' will no longer silently re-export bindings imported
4502 from a used module. It will emit a `deprecation' warning and will
4503 cease to perform any re-export in the next version. If you actually
4504 want to re-export bindings, use the new `re-export' in place of
4505 `export'. The new `re-export' will not make copies of variables when
4506 rexporting them, as `export' did wrongly.
4508 ** Module system now allows selection and renaming of imported bindings
4510 Previously, when using `use-modules' or the `#:use-module' clause in
4511 the `define-module' form, all the bindings (association of symbols to
4512 values) for imported modules were added to the "current module" on an
4513 as-is basis. This has been changed to allow finer control through two
4514 new facilities: selection and renaming.
4516 You can now select which of the imported module's bindings are to be
4517 visible in the current module by using the `:select' clause. This
4518 clause also can be used to rename individual bindings. For example:
4520 ;; import all bindings no questions asked
4521 (use-modules (ice-9 common-list))
4523 ;; import four bindings, renaming two of them;
4524 ;; the current module sees: every some zonk-y zonk-n
4525 (use-modules ((ice-9 common-list)
4527 (remove-if . zonk-y)
4528 (remove-if-not . zonk-n))))
4530 You can also programmatically rename all selected bindings using the
4531 `:renamer' clause, which specifies a proc that takes a symbol and
4532 returns another symbol. Because it is common practice to use a prefix,
4533 we now provide the convenience procedure `symbol-prefix-proc'. For
4536 ;; import four bindings, renaming two of them specifically,
4537 ;; and all four w/ prefix "CL:";
4538 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
4539 (use-modules ((ice-9 common-list)
4541 (remove-if . zonk-y)
4542 (remove-if-not . zonk-n))
4543 :renamer (symbol-prefix-proc 'CL:)))
4545 ;; import four bindings, renaming two of them specifically,
4546 ;; and all four by upcasing.
4547 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
4548 (define (upcase-symbol sym)
4549 (string->symbol (string-upcase (symbol->string sym))))
4551 (use-modules ((ice-9 common-list)
4553 (remove-if . zonk-y)
4554 (remove-if-not . zonk-n))
4555 :renamer upcase-symbol))
4557 Note that programmatic renaming is done *after* individual renaming.
4558 Also, the above examples show `use-modules', but the same facilities are
4559 available for the `#:use-module' clause of `define-module'.
4561 See manual for more info.
4563 ** The semantics of guardians have changed.
4565 The changes are for the most part compatible. An important criterion
4566 was to keep the typical usage of guardians as simple as before, but to
4567 make the semantics safer and (as a result) more useful.
4569 *** All objects returned from guardians are now properly alive.
4571 It is now guaranteed that any object referenced by an object returned
4572 from a guardian is alive. It's now impossible for a guardian to
4573 return a "contained" object before its "containing" object.
4575 One incompatible (but probably not very important) change resulting
4576 from this is that it is no longer possible to guard objects that
4577 indirectly reference themselves (i.e. are parts of cycles). If you do
4578 so accidentally, you'll get a warning.
4580 *** There are now two types of guardians: greedy and sharing.
4582 If you call (make-guardian #t) or just (make-guardian), you'll get a
4583 greedy guardian, and for (make-guardian #f) a sharing guardian.
4585 Greedy guardians are the default because they are more "defensive".
4586 You can only greedily guard an object once. If you guard an object
4587 more than once, once in a greedy guardian and the rest of times in
4588 sharing guardians, then it is guaranteed that the object won't be
4589 returned from sharing guardians as long as it is greedily guarded
4592 Guardians returned by calls to `make-guardian' can now take one more
4593 optional parameter, which says whether to throw an error in case an
4594 attempt is made to greedily guard an object that is already greedily
4595 guarded. The default is true, i.e. throw an error. If the parameter
4596 is false, the guardian invocation returns #t if guarding was
4597 successful and #f if it wasn't.
4599 Also, since greedy guarding is, in effect, a side-effecting operation
4600 on objects, a new function is introduced: `destroy-guardian!'.
4601 Invoking this function on a guardian renders it unoperative and, if
4602 the guardian is greedy, clears the "greedily guarded" property of the
4603 objects that were guarded by it, thus undoing the side effect.
4605 Note that all this hair is hardly very important, since guardian
4606 objects are usually permanent.
4608 ** Continuations created by call-with-current-continuation now accept
4609 any number of arguments, as required by R5RS.
4611 ** New function `issue-deprecation-warning'
4613 This function is used to display the deprecation messages that are
4614 controlled by GUILE_WARN_DEPRECATION as explained in the README.
4617 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
4621 ;; `id' is deprecated. Use `identity' instead.
4626 ** New syntax `begin-deprecated'
4628 When deprecated features are included (as determined by the configure
4629 option --enable-deprecated), `begin-deprecated' is identical to
4630 `begin'. When deprecated features are excluded, it always evaluates
4631 to `#f', ignoring the body forms.
4633 ** New function `make-object-property'
4635 This function returns a new `procedure with setter' P that can be used
4636 to attach a property to objects. When calling P as
4640 where `obj' is any kind of object, it attaches `val' to `obj' in such
4641 a way that it can be retrieved by calling P as
4645 This function will replace procedure properties, symbol properties and
4646 source properties eventually.
4648 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
4650 Instead of #&optional, #&key, etc you should now use #:optional,
4651 #:key, etc. Since #:optional is a keyword, you can write it as just
4652 :optional when (read-set! keywords 'prefix) is active.
4654 The old reader syntax `#&' is still supported, but deprecated. It
4655 will be removed in the next release.
4657 ** New define-module option: pure
4659 Tells the module system not to include any bindings from the root
4664 (define-module (totally-empty-module)
4667 ** New define-module option: export NAME1 ...
4669 Export names NAME1 ...
4671 This option is required if you want to be able to export bindings from
4672 a module which doesn't import one of `define-public' or `export'.
4676 (define-module (foo)
4678 :use-module (ice-9 r5rs)
4681 ;;; Note that we're pure R5RS below this point!
4686 ** New function: object->string OBJ
4688 Return a Scheme string obtained by printing a given object.
4690 ** New function: port? X
4692 Returns a boolean indicating whether X is a port. Equivalent to
4693 `(or (input-port? X) (output-port? X))'.
4695 ** New function: file-port?
4697 Determines whether a given object is a port that is related to a file.
4699 ** New function: port-for-each proc
4701 Apply PROC to each port in the Guile port table in turn. The return
4702 value is unspecified. More specifically, PROC is applied exactly once
4703 to every port that exists in the system at the time PORT-FOR-EACH is
4704 invoked. Changes to the port table while PORT-FOR-EACH is running
4705 have no effect as far as PORT-FOR-EACH is concerned.
4707 ** New function: dup2 oldfd newfd
4709 A simple wrapper for the `dup2' system call. Copies the file
4710 descriptor OLDFD to descriptor number NEWFD, replacing the
4711 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
4712 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
4713 to move away ports which are using NEWFD. The return value is
4716 ** New function: close-fdes fd
4718 A simple wrapper for the `close' system call. Close file
4719 descriptor FD, which must be an integer. Unlike close (*note
4720 close: Ports and File Descriptors.), the file descriptor will be
4721 closed even if a port is using it. The return value is
4724 ** New function: crypt password salt
4726 Encrypts `password' using the standard unix password encryption
4729 ** New function: chroot path
4731 Change the root directory of the running process to `path'.
4733 ** New functions: getlogin, cuserid
4735 Return the login name or the user name of the current effective user
4738 ** New functions: getpriority which who, setpriority which who prio
4740 Get or set the priority of the running process.
4742 ** New function: getpass prompt
4744 Read a password from the terminal, first displaying `prompt' and
4747 ** New function: flock file operation
4749 Set/remove an advisory shared or exclusive lock on `file'.
4751 ** New functions: sethostname name, gethostname
4753 Set or get the hostname of the machine the current process is running
4756 ** New function: mkstemp! tmpl
4758 mkstemp creates a new unique file in the file system and returns a
4759 new buffered port open for reading and writing to the file. TMPL
4760 is a string specifying where the file should be created: it must
4761 end with `XXXXXX' and will be changed in place to return the name
4762 of the temporary file.
4764 ** New function: open-input-string string
4766 Return an input string port which delivers the characters from
4767 `string'. This procedure, together with `open-output-string' and
4768 `get-output-string' implements SRFI-6.
4770 ** New function: open-output-string
4772 Return an output string port which collects all data written to it.
4773 The data can then be retrieved by `get-output-string'.
4775 ** New function: get-output-string
4777 Return the contents of an output string port.
4779 ** New function: identity
4781 Return the argument.
4783 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
4784 are represented in Scheme as integers with normal host byte ordering.
4786 ** New function: inet-pton family address
4788 Convert a printable string network address into an integer. Note that
4789 unlike the C version of this function, the result is an integer with
4790 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4793 (inet-pton AF_INET "127.0.0.1") => 2130706433
4794 (inet-pton AF_INET6 "::1") => 1
4796 ** New function: inet-ntop family address
4798 Convert an integer network address into a printable string. Note that
4799 unlike the C version of this function, the input is an integer with
4800 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4803 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
4804 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
4805 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
4809 Use `identity' instead.
4815 ** Deprecated: return-it
4819 ** Deprecated: string-character-length
4821 Use `string-length' instead.
4823 ** Deprecated: flags
4825 Use `logior' instead.
4827 ** Deprecated: close-all-ports-except.
4829 This was intended for closing ports in a child process after a fork,
4830 but it has the undesirable side effect of flushing buffers.
4831 port-for-each is more flexible.
4833 ** The (ice-9 popen) module now attempts to set up file descriptors in
4834 the child process from the current Scheme ports, instead of using the
4835 current values of file descriptors 0, 1, and 2 in the parent process.
4837 ** Removed function: builtin-weak-bindings
4839 There is no such concept as a weak binding any more.
4841 ** Removed constants: bignum-radix, scm-line-incrementors
4843 ** define-method: New syntax mandatory.
4845 The new method syntax is now mandatory:
4847 (define-method (NAME ARG-SPEC ...) BODY ...)
4848 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
4850 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
4851 REST-ARG ::= ARG-NAME
4853 If you have old code using the old syntax, import
4854 (oop goops old-define-method) before (oop goops) as in:
4856 (use-modules (oop goops old-define-method) (oop goops))
4858 ** Deprecated function: builtin-variable
4859 Removed function: builtin-bindings
4861 There is no longer a distinction between builtin or other variables.
4862 Use module system operations for all variables.
4864 ** Lazy-catch handlers are no longer allowed to return.
4866 That is, a call to `throw', `error', etc is now guaranteed to not
4869 ** Bugfixes for (ice-9 getopt-long)
4871 This module is now tested using test-suite/tests/getopt-long.test.
4872 The following bugs have been fixed:
4874 *** Parsing for options that are specified to have `optional' args now checks
4875 if the next element is an option instead of unconditionally taking it as the
4878 *** An error is now thrown for `--opt=val' when the option description
4879 does not specify `(value #t)' or `(value optional)'. This condition used to
4880 be accepted w/o error, contrary to the documentation.
4882 *** The error message for unrecognized options is now more informative.
4883 It used to be "not a record", an artifact of the implementation.
4885 *** The error message for `--opt' terminating the arg list (no value), when
4886 `(value #t)' is specified, is now more informative. It used to be "not enough
4889 *** "Clumped" single-char args now preserve trailing string, use it as arg.
4890 The expansion used to be like so:
4892 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
4894 Note that the "5d" is dropped. Now it is like so:
4896 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
4898 This enables single-char options to have adjoining arguments as long as their
4899 constituent characters are not potential single-char options.
4901 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
4903 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
4904 property `arglist', which can be retrieved by `arity'. The result is that
4905 `arity' can give more detailed information than before:
4909 guile> (use-modules (ice-9 optargs))
4910 guile> (define* (foo #:optional a b c) a)
4912 0 or more arguments in `lambda*:G0'.
4917 3 optional arguments: `a', `b' and `c'.
4918 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
4920 2 required arguments: `a' and `b', 2 keyword arguments: `c'
4921 and `d', other keywords allowed.
4922 guile> (define* (baz a b #:optional c #:rest r) a)
4924 2 required arguments: `a' and `b', 1 optional argument: `c',
4927 * Changes to the C interface
4929 ** Types have been renamed from scm_*_t to scm_t_*.
4931 This has been done for POSIX sake. It reserves identifiers ending
4932 with "_t". What a concept.
4934 The old names are still available with status `deprecated'.
4936 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
4938 ** Deprecated features have been removed.
4942 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
4943 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
4945 *** C Functions removed
4947 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
4948 scm_fseek - replaced by scm_seek.
4949 gc-thunk - replaced by after-gc-hook.
4950 gh_int2scmb - replaced by gh_bool2scm.
4951 scm_tc_dblr - replaced by scm_tc16_real.
4952 scm_tc_dblc - replaced by scm_tc16_complex.
4953 scm_list_star - replaced by scm_cons_star.
4955 ** Deprecated: scm_makfromstr
4957 Use scm_mem2string instead.
4959 ** Deprecated: scm_make_shared_substring
4961 Explicit shared substrings will disappear from Guile.
4963 Instead, "normal" strings will be implemented using sharing
4964 internally, combined with a copy-on-write strategy.
4966 ** Deprecated: scm_read_only_string_p
4968 The concept of read-only strings will disappear in next release of
4971 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
4973 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
4975 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
4977 Call a procedure with the indicated number of arguments. See "Fly
4978 Evaluation" in the manual.
4980 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
4982 Call a procedure with the indicated number of arguments and a list of
4983 further arguments. See "Fly Evaluation" in the manual.
4985 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
4987 Create a list of the given number of elements. See "List
4988 Constructors" in the manual.
4990 ** Renamed function: scm_listify has been replaced by scm_list_n.
4992 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
4993 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
4995 Use functions scm_list_N instead.
4997 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
4999 Used by an application to read arbitrary number of bytes from a port.
5000 Same semantics as libc read, except that scm_c_read only returns less
5001 than SIZE bytes if at end-of-file.
5003 Warning: Doesn't update port line and column counts!
5005 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
5007 Used by an application to write arbitrary number of bytes to an SCM
5008 port. Similar semantics as libc write. However, unlike libc
5009 write, scm_c_write writes the requested number of bytes and has no
5012 Warning: Doesn't update port line and column counts!
5014 ** New function: scm_init_guile ()
5016 In contrast to scm_boot_guile, scm_init_guile will return normally
5017 after initializing Guile. It is not available on all systems, tho.
5019 ** New functions: scm_str2symbol, scm_mem2symbol
5021 The function scm_str2symbol takes a const char* pointing to a zero-terminated
5022 field of characters and creates a scheme symbol object from that C string.
5023 The function scm_mem2symbol takes a const char* and a number of characters and
5024 creates a symbol from the characters in that memory area.
5026 ** New functions: scm_primitive_make_property
5027 scm_primitive_property_ref
5028 scm_primitive_property_set_x
5029 scm_primitive_property_del_x
5031 These functions implement a new way to deal with object properties.
5032 See libguile/properties.c for their documentation.
5034 ** New function: scm_done_free (long size)
5036 This function is the inverse of scm_done_malloc. Use it to report the
5037 amount of smob memory you free. The previous method, which involved
5038 calling scm_done_malloc with negative argument, was somewhat
5039 unintuitive (and is still available, of course).
5041 ** New function: scm_c_memq (SCM obj, SCM list)
5043 This function provides a fast C level alternative for scm_memq for the case
5044 that the list parameter is known to be a proper list. The function is a
5045 replacement for scm_sloppy_memq, but is stricter in its requirements on its
5046 list input parameter, since for anything else but a proper list the function's
5047 behaviour is undefined - it may even crash or loop endlessly. Further, for
5048 the case that the object is not found in the list, scm_c_memq returns #f which
5049 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
5051 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
5052 scm_remember_upto_here
5054 These functions replace the function scm_remember.
5056 ** Deprecated function: scm_remember
5058 Use one of the new functions scm_remember_upto_here_1,
5059 scm_remember_upto_here_2 or scm_remember_upto_here instead.
5061 ** New function: scm_allocate_string
5063 This function replaces the function scm_makstr.
5065 ** Deprecated function: scm_makstr
5067 Use the new function scm_allocate_string instead.
5069 ** New global variable scm_gc_running_p introduced.
5071 Use this variable to find out if garbage collection is being executed. Up to
5072 now applications have used scm_gc_heap_lock to test if garbage collection was
5073 running, which also works because of the fact that up to know only the garbage
5074 collector has set this variable. But, this is an implementation detail that
5075 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
5076 of this variable is (and has been) not fully safe anyway.
5078 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
5080 Use these instead of SCM_LENGTH_MAX.
5082 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
5083 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
5084 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
5086 Use these instead of SCM_LENGTH.
5088 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
5089 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
5090 SCM_SET_BITVECTOR_LENGTH
5092 Use these instead of SCM_SETLENGTH
5094 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
5095 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
5098 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
5101 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
5102 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
5105 Use these instead of SCM_SETCHARS.
5107 ** New macro: SCM_BITVECTOR_P
5109 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
5111 Use instead of SCM_COERCE_SUBSTR.
5113 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
5115 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
5117 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
5118 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
5119 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
5120 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
5121 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
5122 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
5123 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
5124 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
5125 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
5126 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
5127 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
5128 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
5129 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
5130 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
5131 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
5133 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
5134 Use scm_memory_error instead of SCM_NALLOC.
5135 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
5136 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
5137 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
5138 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
5139 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
5140 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
5141 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
5142 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
5143 Use SCM_STRINGP instead of SCM_RWSTRINGP.
5144 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
5145 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
5146 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
5147 Use a type specific setter macro instead of SCM_SETLENGTH.
5148 Use a type specific setter macro instead of SCM_SETCHARS.
5149 Use a type specific length macro instead of SCM_LENGTH_MAX.
5150 Use SCM_GCMARKP instead of SCM_GC8MARKP.
5151 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
5152 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
5153 Use SCM_TYP16 instead of SCM_GCTYP16.
5154 Use SCM_CDR instead of SCM_GCCDR.
5155 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
5156 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
5157 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
5158 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
5159 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
5160 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
5161 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
5163 ** Removed function: scm_struct_init
5165 ** Removed variable: scm_symhash_dim
5167 ** Renamed function: scm_make_cont has been replaced by
5168 scm_make_continuation, which has a different interface.
5170 ** Deprecated function: scm_call_catching_errors
5172 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
5174 ** Deprecated function: scm_strhash
5176 Use scm_string_hash instead.
5178 ** Deprecated function: scm_vector_set_length_x
5180 Instead, create a fresh vector of the desired size and copy the contents.
5182 ** scm_gensym has changed prototype
5184 scm_gensym now only takes one argument.
5186 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
5189 There is now only a single symbol type scm_tc7_symbol.
5190 The tag scm_tc7_lvector was not used anyway.
5192 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
5194 Use scm_make_smob_type and scm_set_smob_XXX instead.
5196 ** New function scm_set_smob_apply.
5198 This can be used to set an apply function to a smob type.
5200 ** Deprecated function: scm_strprint_obj
5202 Use scm_object_to_string instead.
5204 ** Deprecated function: scm_wta
5206 Use scm_wrong_type_arg, or another appropriate error signalling function
5209 ** Explicit support for obarrays has been deprecated.
5211 Use `scm_str2symbol' and the generic hashtable functions instead.
5213 ** The concept of `vcells' has been deprecated.
5215 The data type `variable' is now used exclusively. `Vcells' have been
5216 a low-level concept so you are likely not affected by this change.
5218 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
5219 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
5221 Use scm_c_define or scm_c_lookup instead, as appropriate.
5223 *** New functions: scm_c_module_lookup, scm_c_lookup,
5224 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
5225 scm_module_define, scm_define.
5227 These functions work with variables instead of with vcells.
5229 ** New functions for creating and defining `subr's and `gsubr's.
5231 The new functions more clearly distinguish between creating a subr (or
5232 gsubr) object and adding it to the current module.
5234 These new functions are available: scm_c_make_subr, scm_c_define_subr,
5235 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
5236 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
5237 scm_c_define_gsubr_with_generic.
5239 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
5240 scm_make_subr_with_generic, scm_make_gsubr,
5241 scm_make_gsubr_with_generic.
5243 Use the new ones from above instead.
5245 ** C interface to the module system has changed.
5247 While we suggest that you avoid as many explicit module system
5248 operations from C as possible for the time being, the C interface has
5249 been made more similar to the high-level Scheme module system.
5251 *** New functions: scm_c_define_module, scm_c_use_module,
5252 scm_c_export, scm_c_resolve_module.
5254 They mostly work like their Scheme namesakes. scm_c_define_module
5255 takes a function that is called a context where the new module is
5258 *** Deprecated functions: scm_the_root_module, scm_make_module,
5259 scm_ensure_user_module, scm_load_scheme_module.
5261 Use the new functions instead.
5263 ** Renamed function: scm_internal_with_fluids becomes
5266 scm_internal_with_fluids is available as a deprecated function.
5268 ** New function: scm_c_with_fluid.
5270 Just like scm_c_with_fluids, but takes one fluid and one value instead
5273 ** Deprecated typedefs: long_long, ulong_long.
5275 They are of questionable utility and they pollute the global
5278 ** Deprecated typedef: scm_sizet
5280 It is of questionable utility now that Guile requires ANSI C, and is
5283 ** Deprecated typedefs: scm_port_rw_active, scm_port,
5284 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
5285 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
5287 Made more compliant with the naming policy by adding a _t at the end.
5289 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
5290 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
5292 With the exception of the mysterious scm_2ulong2big, they are still
5293 available under new names (scm_i_mkbig etc). These functions are not
5294 intended to be used in user code. You should avoid dealing with
5295 bignums directly, and should deal with numbers in general (which can
5298 ** Change in behavior: scm_num2long, scm_num2ulong
5300 The scm_num2[u]long functions don't any longer accept an inexact
5301 argument. This change in behavior is motivated by concordance with
5302 R5RS: It is more common that a primitive doesn't want to accept an
5303 inexact for an exact.
5305 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
5306 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
5307 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
5310 These are conversion functions between the various ANSI C integral
5311 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
5312 accept an inexact argument.
5314 ** New functions: scm_float2num, scm_double2num,
5315 scm_num2float, scm_num2double.
5317 These are conversion functions between the two ANSI C float types and
5320 ** New number validation macros:
5321 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
5325 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
5327 These are just nicer-named old scm_protect_object and
5328 scm_unprotect_object.
5330 ** Deprecated functions: scm_protect_object, scm_unprotect_object
5332 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
5334 These functions can be used to register pointers to locations that
5337 ** Deprecated function: scm_create_hook.
5339 Its sins are: misleading name, non-modularity and lack of general
5343 Changes since Guile 1.3.4:
5345 * Changes to the distribution
5347 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
5349 We've changed the way we handle generated files in the Guile source
5350 repository. As a result, the procedure for building trees obtained
5351 from the nightly FTP snapshots or via CVS has changed:
5352 - You must have appropriate versions of autoconf, automake, and
5353 libtool installed on your system. See README for info on how to
5354 obtain these programs.
5355 - Before configuring the tree, you must first run the script
5356 `autogen.sh' at the top of the source tree.
5358 The Guile repository used to contain not only source files, written by
5359 humans, but also some generated files, like configure scripts and
5360 Makefile.in files. Even though the contents of these files could be
5361 derived mechanically from other files present, we thought it would
5362 make the tree easier to build if we checked them into CVS.
5364 However, this approach means that minor differences between
5365 developer's installed tools and habits affected the whole team.
5366 So we have removed the generated files from the repository, and
5367 added the autogen.sh script, which will reconstruct them
5371 ** configure now has experimental options to remove support for certain
5374 --disable-arrays omit array and uniform array support
5375 --disable-posix omit posix interfaces
5376 --disable-networking omit networking interfaces
5377 --disable-regex omit regular expression interfaces
5379 These are likely to become separate modules some day.
5381 ** New configure option --enable-debug-freelist
5383 This enables a debugging version of SCM_NEWCELL(), and also registers
5384 an extra primitive, the setter `gc-set-debug-check-freelist!'.
5386 Configure with the --enable-debug-freelist option to enable
5387 the gc-set-debug-check-freelist! primitive, and then use:
5389 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
5390 (gc-set-debug-check-freelist! #f) # turn off checking
5392 Checking of the freelist forces a traversal of the freelist and
5393 a garbage collection before each allocation of a cell. This can
5394 slow down the interpreter dramatically, so the setter should be used to
5395 turn on this extra processing only when necessary.
5397 ** New configure option --enable-debug-malloc
5399 Include code for debugging of calls to scm_must_malloc/realloc/free.
5403 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
5404 2. objects reallocated by scm_must_realloc has been allocated by
5406 3. reallocated objects are reallocated with the same what string
5408 But, most importantly, it records the number of allocated objects of
5409 each kind. This is useful when searching for memory leaks.
5411 A Guile compiled with this option provides the primitive
5412 `malloc-stats' which returns an alist with pairs of kind and the
5413 number of objects of that kind.
5415 ** All includes are now referenced relative to the root directory
5417 Since some users have had problems with mixups between Guile and
5418 system headers, we have decided to always refer to Guile headers via
5419 their parent directories. This essentially creates a "private name
5420 space" for Guile headers. This means that the compiler only is given
5421 -I options for the root build and root source directory.
5423 ** Header files kw.h and genio.h have been removed.
5425 ** The module (ice-9 getopt-gnu-style) has been removed.
5427 ** New module (ice-9 documentation)
5429 Implements the interface to documentation strings associated with
5432 ** New module (ice-9 time)
5434 Provides a macro `time', which displays execution time of a given form.
5436 ** New module (ice-9 history)
5438 Loading this module enables value history in the repl.
5440 * Changes to the stand-alone interpreter
5442 ** New command line option --debug
5444 Start Guile with debugging evaluator and backtraces enabled.
5446 This is useful when debugging your .guile init file or scripts.
5448 ** New help facility
5450 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
5451 (help REGEXP) ditto for objects with names matching REGEXP (a string)
5452 (help 'NAME) gives documentation for NAME, even if it is not an object
5453 (help ,EXPR) gives documentation for object returned by EXPR
5454 (help (my module)) gives module commentary for `(my module)'
5455 (help) gives this text
5457 `help' searches among bindings exported from loaded modules, while
5458 `apropos' searches among bindings visible from the "current" module.
5460 Examples: (help help)
5462 (help "output-string")
5464 ** `help' and `apropos' now prints full module names
5466 ** Dynamic linking now uses libltdl from the libtool package.
5468 The old system dependent code for doing dynamic linking has been
5469 replaced with calls to the libltdl functions which do all the hairy
5472 The major improvement is that you can now directly pass libtool
5473 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
5474 will be able to do the best shared library job you can get, via
5477 The way dynamic libraries are found has changed and is not really
5478 portable across platforms, probably. It is therefore recommended to
5479 use absolute filenames when possible.
5481 If you pass a filename without an extension to `dynamic-link', it will
5482 try a few appropriate ones. Thus, the most platform ignorant way is
5483 to specify a name like "libfoo", without any directories and
5486 ** Guile COOP threads are now compatible with LinuxThreads
5488 Previously, COOP threading wasn't possible in applications linked with
5489 Linux POSIX threads due to their use of the stack pointer to find the
5490 thread context. This has now been fixed with a workaround which uses
5491 the pthreads to allocate the stack.
5493 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
5495 ** Positions of erring expression in scripts
5497 With version 1.3.4, the location of the erring expression in Guile
5498 scipts is no longer automatically reported. (This should have been
5499 documented before the 1.3.4 release.)
5501 You can get this information by enabling recording of positions of
5502 source expressions and running the debugging evaluator. Put this at
5503 the top of your script (or in your "site" file):
5505 (read-enable 'positions)
5506 (debug-enable 'debug)
5508 ** Backtraces in scripts
5510 It is now possible to get backtraces in scripts.
5514 (debug-enable 'debug 'backtrace)
5516 at the top of the script.
5518 (The first options enables the debugging evaluator.
5519 The second enables backtraces.)
5521 ** Part of module system symbol lookup now implemented in C
5523 The eval closure of most modules is now implemented in C. Since this
5524 was one of the bottlenecks for loading speed, Guile now loads code
5525 substantially faster than before.
5527 ** Attempting to get the value of an unbound variable now produces
5528 an exception with a key of 'unbound-variable instead of 'misc-error.
5530 ** The initial default output port is now unbuffered if it's using a
5531 tty device. Previously in this situation it was line-buffered.
5533 ** New hook: after-gc-hook
5535 after-gc-hook takes over the role of gc-thunk. This hook is run at
5536 the first SCM_TICK after a GC. (Thus, the code is run at the same
5537 point during evaluation as signal handlers.)
5539 Note that this hook should be used only for diagnostic and debugging
5540 purposes. It is not certain that it will continue to be well-defined
5541 when this hook is run in the future.
5543 C programmers: Note the new C level hooks scm_before_gc_c_hook,
5544 scm_before_sweep_c_hook, scm_after_gc_c_hook.
5546 ** Improvements to garbage collector
5548 Guile 1.4 has a new policy for triggering heap allocation and
5549 determining the sizes of heap segments. It fixes a number of problems
5552 1. The new policy can handle two separate pools of cells
5553 (2-word/4-word) better. (The old policy would run wild, allocating
5554 more and more memory for certain programs.)
5556 2. The old code would sometimes allocate far too much heap so that the
5557 Guile process became gigantic. The new code avoids this.
5559 3. The old code would sometimes allocate too little so that few cells
5560 were freed at GC so that, in turn, too much time was spent in GC.
5562 4. The old code would often trigger heap allocation several times in a
5563 row. (The new scheme predicts how large the segments needs to be
5564 in order not to need further allocation.)
5566 All in all, the new GC policy will make larger applications more
5569 The new GC scheme also is prepared for POSIX threading. Threads can
5570 allocate private pools of cells ("clusters") with just a single
5571 function call. Allocation of single cells from such a cluster can
5572 then proceed without any need of inter-thread synchronization.
5574 ** New environment variables controlling GC parameters
5576 GUILE_MAX_SEGMENT_SIZE Maximal segment size
5579 Allocation of 2-word cell heaps:
5581 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
5584 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
5585 GC in percent of total heap size
5588 Allocation of 4-word cell heaps
5589 (used for real numbers and misc other objects):
5591 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
5593 (See entry "Way for application to customize GC parameters" under
5594 section "Changes to the scm_ interface" below.)
5596 ** Guile now implements reals using 4-word cells
5598 This speeds up computation with reals. (They were earlier allocated
5599 with `malloc'.) There is still some room for optimizations, however.
5601 ** Some further steps toward POSIX thread support have been taken
5603 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
5604 don't have much effect any longer, and many of them will be removed in
5608 are only handled at the top of the evaluator loop, immediately after
5609 I/O, and in scm_equalp.
5611 *** The GC can allocate thread private pools of pairs.
5613 * Changes to Scheme functions and syntax
5615 ** close-input-port and close-output-port are now R5RS
5617 These procedures have been turned into primitives and have R5RS behaviour.
5619 ** New procedure: simple-format PORT MESSAGE ARG1 ...
5621 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
5622 extended by the more sophisticated version in (ice-9 format)
5624 (simple-format port message . args)
5625 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
5626 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
5627 the escapes are replaced with corresponding members of ARGS:
5628 ~A formats using `display' and ~S formats using `write'.
5629 If DESTINATION is #t, then use the `current-output-port',
5630 if DESTINATION is #f, then return a string containing the formatted text.
5631 Does not add a trailing newline."
5633 ** string-ref: the second argument is no longer optional.
5635 ** string, list->string: no longer accept strings in their arguments,
5636 only characters, for compatibility with R5RS.
5638 ** New procedure: port-closed? PORT
5639 Returns #t if PORT is closed or #f if it is open.
5641 ** Deprecated: list*
5643 The list* functionality is now provided by cons* (SRFI-1 compliant)
5645 ** New procedure: cons* ARG1 ARG2 ... ARGn
5647 Like `list', but the last arg provides the tail of the constructed list,
5648 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
5650 Requires at least one argument. If given one argument, that argument
5651 is returned as result.
5653 This function is called `list*' in some other Schemes and in Common LISP.
5655 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
5657 ** New procedure: object-documentation OBJECT
5659 Returns the documentation string associated with OBJECT. The
5660 procedure uses a caching mechanism so that subsequent lookups are
5663 Exported by (ice-9 documentation).
5665 ** module-name now returns full names of modules
5667 Previously, only the last part of the name was returned (`session' for
5668 `(ice-9 session)'). Ex: `(ice-9 session)'.
5670 * Changes to the gh_ interface
5672 ** Deprecated: gh_int2scmb
5674 Use gh_bool2scm instead.
5676 * Changes to the scm_ interface
5678 ** Guile primitives now carry docstrings!
5680 Thanks to Greg Badros!
5682 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5684 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5685 macros and must contain a docstring that is extracted into foo.doc using a new
5686 guile-doc-snarf script (that uses guile-doc-snarf.awk).
5688 However, a major overhaul of these macros is scheduled for the next release of
5691 ** Guile primitives use a new technique for validation of arguments
5693 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
5694 the readability of argument checking.
5696 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
5698 ** New macros: SCM_PACK, SCM_UNPACK
5700 Compose/decompose an SCM value.
5702 The SCM type is now treated as an abstract data type and may be defined as a
5703 long, a void* or as a struct, depending on the architecture and compile time
5704 options. This makes it easier to find several types of bugs, for example when
5705 SCM values are treated as integers without conversion. Values of the SCM type
5706 should be treated as "atomic" values. These macros are used when
5707 composing/decomposing an SCM value, either because you want to access
5708 individual bits, or because you want to treat it as an integer value.
5710 E.g., in order to set bit 7 in an SCM value x, use the expression
5712 SCM_PACK (SCM_UNPACK (x) | 0x80)
5714 ** The name property of hooks is deprecated.
5715 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
5717 You can emulate this feature by using object properties.
5719 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
5720 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
5723 These macros will be removed in a future release of Guile.
5725 ** The following types, functions and macros from numbers.h are deprecated:
5726 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
5727 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
5729 ** Port internals: the rw_random variable in the scm_port structure
5730 must be set to non-zero in any random access port. In recent Guile
5731 releases it was only set for bidirectional random-access ports.
5733 ** Port internals: the seek ptob procedure is now responsible for
5734 resetting the buffers if required. The change was made so that in the
5735 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
5736 the fport and strport ptobs can avoid resetting the buffers,
5737 in particular to avoid discarding unread chars. An existing port
5738 type can be fixed by adding something like the following to the
5739 beginning of the ptob seek procedure:
5741 if (pt->rw_active == SCM_PORT_READ)
5742 scm_end_input (object);
5743 else if (pt->rw_active == SCM_PORT_WRITE)
5744 ptob->flush (object);
5746 although to actually avoid resetting the buffers and discard unread
5747 chars requires further hacking that depends on the characteristics
5750 ** Deprecated functions: scm_fseek, scm_tag
5752 These functions are no longer used and will be removed in a future version.
5754 ** The scm_sysmissing procedure is no longer used in libguile.
5755 Unless it turns out to be unexpectedly useful to somebody, it will be
5756 removed in a future version.
5758 ** The format of error message strings has changed
5760 The two C procedures: scm_display_error and scm_error, as well as the
5761 primitive `scm-error', now use scm_simple_format to do their work.
5762 This means that the message strings of all code must be updated to use
5763 ~A where %s was used before, and ~S where %S was used before.
5765 During the period when there still are a lot of old Guiles out there,
5766 you might want to support both old and new versions of Guile.
5768 There are basically two methods to achieve this. Both methods use
5771 AC_CHECK_FUNCS(scm_simple_format)
5773 in your configure.in.
5775 Method 1: Use the string concatenation features of ANSI C's
5780 #ifdef HAVE_SCM_SIMPLE_FORMAT
5786 Then represent each of your error messages using a preprocessor macro:
5788 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
5792 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
5793 (define make-message string-append)
5795 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
5797 Method 2: Use the oldfmt function found in doc/oldfmt.c.
5801 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
5806 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
5810 ** Deprecated: coop_mutex_init, coop_condition_variable_init
5812 Don't use the functions coop_mutex_init and
5813 coop_condition_variable_init. They will change.
5815 Use scm_mutex_init and scm_cond_init instead.
5817 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
5818 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
5819 COND, as `scm_cond_wait' does, but it also bounds the duration
5820 of the wait. If COND has not been signaled before time ABSTIME,
5821 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
5822 returns the error code `ETIMEDOUT'.
5824 The ABSTIME parameter specifies an absolute time, with the same
5825 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
5826 to 00:00:00 GMT, January 1, 1970.
5828 ** New function: scm_cond_broadcast (scm_cond_t *COND)
5829 `scm_cond_broadcast' restarts all the threads that are waiting
5830 on the condition variable COND. Nothing happens if no threads are
5833 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
5834 `scm_key_create' allocates a new TSD key. The key is stored in
5835 the location pointed to by KEY. There is no limit on the number
5836 of keys allocated at a given time. The value initially associated
5837 with the returned key is `NULL' in all currently executing threads.
5839 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
5840 function associated with the key. When a thread terminates,
5841 DESTR_FUNCTION is called on the value associated with the key in
5842 that thread. The DESTR_FUNCTION is not called if a key is deleted
5843 with `scm_key_delete' or a value is changed with
5844 `scm_setspecific'. The order in which destructor functions are
5845 called at thread termination time is unspecified.
5847 Destructors are not yet implemented.
5849 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
5850 `scm_setspecific' changes the value associated with KEY in the
5851 calling thread, storing the given POINTER instead.
5853 ** New function: scm_getspecific (scm_key_t KEY)
5854 `scm_getspecific' returns the value currently associated with
5855 KEY in the calling thread.
5857 ** New function: scm_key_delete (scm_key_t KEY)
5858 `scm_key_delete' deallocates a TSD key. It does not check
5859 whether non-`NULL' values are associated with that key in the
5860 currently executing threads, nor call the destructor function
5861 associated with the key.
5863 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
5865 Initialize a C level hook HOOK with associated HOOK_DATA and type
5866 TYPE. (See scm_c_hook_run ().)
5868 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
5870 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
5871 is true, add it last, otherwise first. The same FUNC can be added
5872 multiple times if FUNC_DATA differ and vice versa.
5874 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
5876 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
5877 function is only removed if both FUNC and FUNC_DATA matches.
5879 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
5881 Run hook HOOK passing DATA to the hook functions.
5883 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
5884 returned is undefined.
5886 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
5887 returns a non-NULL value. This value is returned as the result of
5888 scm_c_hook_run. If all functions return NULL, NULL is returned.
5890 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
5891 returns a NULL value, and NULL is returned. If all functions returns
5892 a non-NULL value, the last value is returned.
5894 ** New C level GC hooks
5896 Five new C level hooks has been added to the garbage collector.
5898 scm_before_gc_c_hook
5901 are run before locking and after unlocking the heap. The system is
5902 thus in a mode where evaluation can take place. (Except that
5903 scm_before_gc_c_hook must not allocate new cells.)
5905 scm_before_mark_c_hook
5906 scm_before_sweep_c_hook
5907 scm_after_sweep_c_hook
5909 are run when the heap is locked. These are intended for extension of
5910 the GC in a modular fashion. Examples are the weaks and guardians
5913 ** Way for application to customize GC parameters
5915 The application can set up other default values for the GC heap
5916 allocation parameters
5918 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
5919 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
5920 GUILE_MAX_SEGMENT_SIZE,
5924 scm_default_init_heap_size_1, scm_default_min_yield_1,
5925 scm_default_init_heap_size_2, scm_default_min_yield_2,
5926 scm_default_max_segment_size
5928 respectively before callong scm_boot_guile.
5930 (See entry "New environment variables ..." in section
5931 "Changes to the stand-alone interpreter" above.)
5933 ** scm_protect_object/scm_unprotect_object now nest
5935 This means that you can call scm_protect_object multiple times on an
5936 object and count on the object being protected until
5937 scm_unprotect_object has been call the same number of times.
5939 The functions also have better time complexity.
5941 Still, it is usually possible to structure the application in a way
5942 that you don't need to use these functions. For example, if you use a
5943 protected standard Guile list to keep track of live objects rather
5944 than some custom data type, objects will die a natural death when they
5945 are no longer needed.
5947 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
5949 Guile does not provide the float representation for inexact real numbers any
5950 more. Now, only doubles are used to represent inexact real numbers. Further,
5951 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
5952 and scm_tc16_complex, respectively.
5954 ** Removed deprecated type scm_smobfuns
5956 ** Removed deprecated function scm_newsmob
5958 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
5960 There is an ongoing discussion among the developers whether to
5961 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
5962 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
5963 until this issue has been settled.
5965 ** Removed deprecated type tag scm_tc16_kw
5967 ** Added type tag scm_tc16_keyword
5969 (This was introduced already in release 1.3.4 but was not documented
5972 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
5974 * Changes to system call interfaces:
5976 ** The "select" procedure now tests port buffers for the ability to
5977 provide input or accept output. Previously only the underlying file
5978 descriptors were checked.
5980 ** New variable PIPE_BUF: the maximum number of bytes that can be
5981 atomically written to a pipe.
5983 ** If a facility is not available on the system when Guile is
5984 compiled, the corresponding primitive procedure will not be defined.
5985 Previously it would have been defined but would throw a system-error
5986 exception if called. Exception handlers which catch this case may
5987 need minor modification: an error will be thrown with key
5988 'unbound-variable instead of 'system-error. Alternatively it's
5989 now possible to use `defined?' to check whether the facility is
5992 ** Procedures which depend on the timezone should now give the correct
5993 result on systems which cache the TZ environment variable, even if TZ
5994 is changed without calling tzset.
5996 * Changes to the networking interfaces:
5998 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
5999 long integers between network and host format. For now, it's not
6000 particularly convenient to do this kind of thing, but consider:
6002 (define write-network-long
6003 (lambda (value port)
6004 (let ((v (make-uniform-vector 1 1 0)))
6005 (uniform-vector-set! v 0 (htonl value))
6006 (uniform-vector-write v port))))
6008 (define read-network-long
6010 (let ((v (make-uniform-vector 1 1 0)))
6011 (uniform-vector-read! v port)
6012 (ntohl (uniform-vector-ref v 0)))))
6014 ** If inet-aton fails, it now throws an error with key 'misc-error
6015 instead of 'system-error, since errno is not relevant.
6017 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
6018 specific keys instead of 'system-error. The latter is inappropriate
6019 since errno will not have been set. The keys are:
6020 'host-not-found, 'try-again, 'no-recovery and 'no-data.
6022 ** sethostent, setnetent, setprotoent, setservent: now take an
6023 optional argument STAYOPEN, which specifies whether the database
6024 remains open after a database entry is accessed randomly (e.g., using
6025 gethostbyname for the hosts database.) The default is #f. Previously
6029 Changes since Guile 1.3.2:
6031 * Changes to the stand-alone interpreter
6035 An initial version of the Guile debugger written by Chris Hanson has
6036 been added. The debugger is still under development but is included
6037 in the distribution anyway since it is already quite useful.
6043 after an error to enter the debugger. Type `help' inside the debugger
6044 for a description of available commands.
6046 If you prefer to have stack frames numbered and printed in
6047 anti-chronological order and prefer up in the stack to be down on the
6048 screen as is the case in gdb, you can put
6050 (debug-enable 'backwards)
6052 in your .guile startup file. (However, this means that Guile can't
6053 use indentation to indicate stack level.)
6055 The debugger is autoloaded into Guile at the first use.
6057 ** Further enhancements to backtraces
6059 There is a new debug option `width' which controls the maximum width
6060 on the screen of printed stack frames. Fancy printing parameters
6061 ("level" and "length" as in Common LISP) are adaptively adjusted for
6062 each stack frame to give maximum information while still fitting
6063 within the bounds. If the stack frame can't be made to fit by
6064 adjusting parameters, it is simply cut off at the end. This is marked
6067 ** Some modules are now only loaded when the repl is started
6069 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
6070 regex) are now loaded into (guile-user) only if the repl has been
6071 started. The effect is that the startup time for scripts has been
6072 reduced to 30% of what it was previously.
6074 Correctly written scripts load the modules they require at the top of
6075 the file and should not be affected by this change.
6077 ** Hooks are now represented as smobs
6079 * Changes to Scheme functions and syntax
6081 ** Readline support has changed again.
6083 The old (readline-activator) module is gone. Use (ice-9 readline)
6084 instead, which now contains all readline functionality. So the code
6085 to activate readline is now
6087 (use-modules (ice-9 readline))
6090 This should work at any time, including from the guile prompt.
6092 To avoid confusion about the terms of Guile's license, please only
6093 enable readline for your personal use; please don't make it the
6094 default for others. Here is why we make this rather odd-sounding
6097 Guile is normally licensed under a weakened form of the GNU General
6098 Public License, which allows you to link code with Guile without
6099 placing that code under the GPL. This exception is important to some
6102 However, since readline is distributed under the GNU General Public
6103 License, when you link Guile with readline, either statically or
6104 dynamically, you effectively change Guile's license to the strict GPL.
6105 Whenever you link any strictly GPL'd code into Guile, uses of Guile
6106 which are normally permitted become forbidden. This is a rather
6107 non-obvious consequence of the licensing terms.
6109 So, to make sure things remain clear, please let people choose for
6110 themselves whether to link GPL'd libraries like readline with Guile.
6112 ** regexp-substitute/global has changed slightly, but incompatibly.
6114 If you include a function in the item list, the string of the match
6115 object it receives is the same string passed to
6116 regexp-substitute/global, not some suffix of that string.
6117 Correspondingly, the match's positions are relative to the entire
6118 string, not the suffix.
6120 If the regexp can match the empty string, the way matches are chosen
6121 from the string has changed. regexp-substitute/global recognizes the
6122 same set of matches that list-matches does; see below.
6124 ** New function: list-matches REGEXP STRING [FLAGS]
6126 Return a list of match objects, one for every non-overlapping, maximal
6127 match of REGEXP in STRING. The matches appear in left-to-right order.
6128 list-matches only reports matches of the empty string if there are no
6129 other matches which begin on, end at, or include the empty match's
6132 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6134 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
6136 For each match of REGEXP in STRING, apply PROC to the match object,
6137 and the last value PROC returned, or INIT for the first call. Return
6138 the last value returned by PROC. We apply PROC to the matches as they
6139 appear from left to right.
6141 This function recognizes matches according to the same criteria as
6144 Thus, you could define list-matches like this:
6146 (define (list-matches regexp string . flags)
6147 (reverse! (apply fold-matches regexp string '() cons flags)))
6149 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6153 *** New function: hook? OBJ
6155 Return #t if OBJ is a hook, otherwise #f.
6157 *** New function: make-hook-with-name NAME [ARITY]
6159 Return a hook with name NAME and arity ARITY. The default value for
6160 ARITY is 0. The only effect of NAME is that it will appear when the
6161 hook object is printed to ease debugging.
6163 *** New function: hook-empty? HOOK
6165 Return #t if HOOK doesn't contain any procedures, otherwise #f.
6167 *** New function: hook->list HOOK
6169 Return a list of the procedures that are called when run-hook is
6172 ** `map' signals an error if its argument lists are not all the same length.
6174 This is the behavior required by R5RS, so this change is really a bug
6175 fix. But it seems to affect a lot of people's code, so we're
6176 mentioning it here anyway.
6178 ** Print-state handling has been made more transparent
6180 Under certain circumstances, ports are represented as a port with an
6181 associated print state. Earlier, this pair was represented as a pair
6182 (see "Some magic has been added to the printer" below). It is now
6183 indistinguishable (almost; see `get-print-state') from a port on the
6186 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
6188 Return a new port with the associated print state PRINT-STATE.
6190 *** New function: get-print-state OUTPUT-PORT
6192 Return the print state associated with this port if it exists,
6193 otherwise return #f.
6195 *** New function: directory-stream? OBJECT
6197 Returns true iff OBJECT is a directory stream --- the sort of object
6198 returned by `opendir'.
6200 ** New function: using-readline?
6202 Return #t if readline is in use in the current repl.
6204 ** structs will be removed in 1.4
6206 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
6207 and use GOOPS objects as the fundamental record type.
6209 * Changes to the scm_ interface
6211 ** structs will be removed in 1.4
6213 The entire current struct interface (struct.c, struct.h) will be
6214 replaced in Guile 1.4. We will merge GOOPS into libguile and use
6215 GOOPS objects as the fundamental record type.
6217 ** The internal representation of subr's has changed
6219 Instead of giving a hint to the subr name, the CAR field of the subr
6220 now contains an index to a subr entry in scm_subr_table.
6222 *** New variable: scm_subr_table
6224 An array of subr entries. A subr entry contains the name, properties
6225 and documentation associated with the subr. The properties and
6226 documentation slots are not yet used.
6228 ** A new scheme for "forwarding" calls to a builtin to a generic function
6230 It is now possible to extend the functionality of some Guile
6231 primitives by letting them defer a call to a GOOPS generic function on
6232 argument mismatch. This means that there is no loss of efficiency in
6237 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
6238 (define-method + ((x <string>) (y <string>))
6239 (string-append x y))
6241 + will still be as efficient as usual in numerical calculations, but
6242 can also be used for concatenating strings.
6244 Who will be the first one to extend Guile's numerical tower to
6245 rationals? :) [OK, there a few other things to fix before this can
6246 be made in a clean way.]
6248 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
6250 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
6252 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
6254 These do the same job as SCM_PROC and SCM_PROC1, but they also define
6255 a variable GENERIC which can be used by the dispatch macros below.
6257 [This is experimental code which may change soon.]
6259 *** New macros for forwarding control to a generic on arg type error
6261 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
6263 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
6265 These correspond to the scm_wta function call, and have the same
6266 behaviour until the user has called the GOOPS primitive
6267 `enable-primitive-generic!'. After that, these macros will apply the
6268 generic function GENERIC to the argument(s) instead of calling
6271 [This is experimental code which may change soon.]
6273 *** New macros for argument testing with generic dispatch
6275 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
6277 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
6279 These correspond to the SCM_ASSERT macro, but will defer control to
6280 GENERIC on error after `enable-primitive-generic!' has been called.
6282 [This is experimental code which may change soon.]
6284 ** New function: SCM scm_eval_body (SCM body, SCM env)
6286 Evaluates the body of a special form.
6288 ** The internal representation of struct's has changed
6290 Previously, four slots were allocated for the procedure(s) of entities
6291 and operators. The motivation for this representation had to do with
6292 the structure of the evaluator, the wish to support tail-recursive
6293 generic functions, and efficiency. Since the generic function
6294 dispatch mechanism has changed, there is no longer a need for such an
6295 expensive representation, and the representation has been simplified.
6297 This should not make any difference for most users.
6299 ** GOOPS support has been cleaned up.
6301 Some code has been moved from eval.c to objects.c and code in both of
6302 these compilation units has been cleaned up and better structured.
6304 *** New functions for applying generic functions
6306 New function: SCM scm_apply_generic (GENERIC, ARGS)
6307 New function: SCM scm_call_generic_0 (GENERIC)
6308 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
6309 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
6310 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
6312 ** Deprecated function: scm_make_named_hook
6314 It is now replaced by:
6316 ** New function: SCM scm_create_hook (const char *name, int arity)
6318 Creates a hook in the same way as make-hook above but also
6319 binds a variable named NAME to it.
6321 This is the typical way of creating a hook from C code.
6323 Currently, the variable is created in the "current" module.
6324 This might change when we get the new module system.
6326 [The behaviour is identical to scm_make_named_hook.]
6330 Changes since Guile 1.3:
6332 * Changes to mailing lists
6334 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
6336 See the README file to find current addresses for all the Guile
6339 * Changes to the distribution
6341 ** Readline support is no longer included with Guile by default.
6343 Based on the different license terms of Guile and Readline, we
6344 concluded that Guile should not *by default* cause the linking of
6345 Readline into an application program. Readline support is now offered
6346 as a separate module, which is linked into an application only when
6347 you explicitly specify it.
6349 Although Guile is GNU software, its distribution terms add a special
6350 exception to the usual GNU General Public License (GPL). Guile's
6351 license includes a clause that allows you to link Guile with non-free
6352 programs. We add this exception so as not to put Guile at a
6353 disadvantage vis-a-vis other extensibility packages that support other
6356 In contrast, the GNU Readline library is distributed under the GNU
6357 General Public License pure and simple. This means that you may not
6358 link Readline, even dynamically, into an application unless it is
6359 distributed under a free software license that is compatible the GPL.
6361 Because of this difference in distribution terms, an application that
6362 can use Guile may not be able to use Readline. Now users will be
6363 explicitly offered two independent decisions about the use of these
6366 You can activate the readline support by issuing
6368 (use-modules (readline-activator))
6371 from your ".guile" file, for example.
6373 * Changes to the stand-alone interpreter
6375 ** All builtins now print as primitives.
6376 Previously builtin procedures not belonging to the fundamental subr
6377 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
6378 Now, they print as #<primitive-procedure NAME>.
6380 ** Backtraces slightly more intelligible.
6381 gsubr-apply and macro transformer application frames no longer appear
6384 * Changes to Scheme functions and syntax
6386 ** Guile now correctly handles internal defines by rewriting them into
6387 their equivalent letrec. Previously, internal defines would
6388 incrementally add to the innermost environment, without checking
6389 whether the restrictions specified in RnRS were met. This lead to the
6390 correct behaviour when these restriction actually were met, but didn't
6391 catch all illegal uses. Such an illegal use could lead to crashes of
6392 the Guile interpreter or other unwanted results. An example of
6393 incorrect internal defines that made Guile behave erratically:
6405 The problem with this example is that the definition of `c' uses the
6406 value of `b' directly. This confuses the meoization machine of Guile
6407 so that the second call of `b' (this time in a larger environment that
6408 also contains bindings for `c' and `d') refers to the binding of `c'
6409 instead of `a'. You could also make Guile crash with a variation on
6414 (define (b flag) (if flag a 1))
6415 (define c (1+ (b flag)))
6423 From now on, Guile will issue an `Unbound variable: b' error message
6428 A hook contains a list of functions which should be called on
6429 particular occasions in an existing program. Hooks are used for
6432 A window manager might have a hook before-window-map-hook. The window
6433 manager uses the function run-hooks to call all functions stored in
6434 before-window-map-hook each time a window is mapped. The user can
6435 store functions in the hook using add-hook!.
6437 In Guile, hooks are first class objects.
6439 *** New function: make-hook [N_ARGS]
6441 Return a hook for hook functions which can take N_ARGS arguments.
6442 The default value for N_ARGS is 0.
6444 (See also scm_make_named_hook below.)
6446 *** New function: add-hook! HOOK PROC [APPEND_P]
6448 Put PROC at the beginning of the list of functions stored in HOOK.
6449 If APPEND_P is supplied, and non-false, put PROC at the end instead.
6451 PROC must be able to take the number of arguments specified when the
6454 If PROC already exists in HOOK, then remove it first.
6456 *** New function: remove-hook! HOOK PROC
6458 Remove PROC from the list of functions in HOOK.
6460 *** New function: reset-hook! HOOK
6462 Clear the list of hook functions stored in HOOK.
6464 *** New function: run-hook HOOK ARG1 ...
6466 Run all hook functions stored in HOOK with arguments ARG1 ... .
6467 The number of arguments supplied must correspond to the number given
6468 when the hook was created.
6470 ** The function `dynamic-link' now takes optional keyword arguments.
6471 The only keyword argument that is currently defined is `:global
6472 BOOL'. With it, you can control whether the shared library will be
6473 linked in global mode or not. In global mode, the symbols from the
6474 linked library can be used to resolve references from other
6475 dynamically linked libraries. In non-global mode, the linked
6476 library is essentially invisible and can only be accessed via
6477 `dynamic-func', etc. The default is now to link in global mode.
6478 Previously, the default has been non-global mode.
6480 The `#:global' keyword is only effective on platforms that support
6481 the dlopen family of functions.
6483 ** New function `provided?'
6485 - Function: provided? FEATURE
6486 Return true iff FEATURE is supported by this installation of
6487 Guile. FEATURE must be a symbol naming a feature; the global
6488 variable `*features*' is a list of available features.
6490 ** Changes to the module (ice-9 expect):
6492 *** The expect-strings macro now matches `$' in a regular expression
6493 only at a line-break or end-of-file by default. Previously it would
6494 match the end of the string accumulated so far. The old behaviour
6495 can be obtained by setting the variable `expect-strings-exec-flags'
6498 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
6499 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
6500 in a regular expression will still match before a line-break or
6501 end-of-file. The default is `regexp/noteol'.
6503 *** The expect-strings macro now uses a variable
6504 `expect-strings-compile-flags' for the flags to be supplied to
6505 `make-regexp'. The default is `regexp/newline', which was previously
6508 *** The expect macro now supplies two arguments to a match procedure:
6509 the current accumulated string and a flag to indicate whether
6510 end-of-file has been reached. Previously only the string was supplied.
6511 If end-of-file is reached, the match procedure will be called an
6512 additional time with the same accumulated string as the previous call
6513 but with the flag set.
6515 ** New module (ice-9 format), implementing the Common Lisp `format' function.
6517 This code, and the documentation for it that appears here, was
6518 borrowed from SLIB, with minor adaptations for Guile.
6520 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
6521 An almost complete implementation of Common LISP format description
6522 according to the CL reference book `Common LISP' from Guy L.
6523 Steele, Digital Press. Backward compatible to most of the
6524 available Scheme format implementations.
6526 Returns `#t', `#f' or a string; has side effect of printing
6527 according to FORMAT-STRING. If DESTINATION is `#t', the output is
6528 to the current output port and `#t' is returned. If DESTINATION
6529 is `#f', a formatted string is returned as the result of the call.
6530 NEW: If DESTINATION is a string, DESTINATION is regarded as the
6531 format string; FORMAT-STRING is then the first argument and the
6532 output is returned as a string. If DESTINATION is a number, the
6533 output is to the current error port if available by the
6534 implementation. Otherwise DESTINATION must be an output port and
6537 FORMAT-STRING must be a string. In case of a formatting error
6538 format returns `#f' and prints a message on the current output or
6539 error port. Characters are output as if the string were output by
6540 the `display' function with the exception of those prefixed by a
6541 tilde (~). For a detailed description of the FORMAT-STRING syntax
6542 please consult a Common LISP format reference manual. For a test
6543 suite to verify this format implementation load `formatst.scm'.
6544 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
6546 Note: `format' is not reentrant, i.e. only one `format'-call may
6547 be executed at a time.
6550 *** Format Specification (Format version 3.0)
6552 Please consult a Common LISP format reference manual for a detailed
6553 description of the format string syntax. For a demonstration of the
6554 implemented directives see `formatst.scm'.
6556 This implementation supports directive parameters and modifiers (`:'
6557 and `@' characters). Multiple parameters must be separated by a comma
6558 (`,'). Parameters can be numerical parameters (positive or negative),
6559 character parameters (prefixed by a quote character (`''), variable
6560 parameters (`v'), number of rest arguments parameter (`#'), empty and
6561 default parameters. Directive characters are case independent. The
6562 general form of a directive is:
6564 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
6566 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
6568 *** Implemented CL Format Control Directives
6570 Documentation syntax: Uppercase characters represent the
6571 corresponding control directive characters. Lowercase characters
6572 represent control directive parameter descriptions.
6575 Any (print as `display' does).
6579 `~MINCOL,COLINC,MINPAD,PADCHARA'
6583 S-expression (print as `write' does).
6587 `~MINCOL,COLINC,MINPAD,PADCHARS'
6593 print number sign always.
6596 print comma separated.
6598 `~MINCOL,PADCHAR,COMMACHARD'
6604 print number sign always.
6607 print comma separated.
6609 `~MINCOL,PADCHAR,COMMACHARX'
6615 print number sign always.
6618 print comma separated.
6620 `~MINCOL,PADCHAR,COMMACHARO'
6626 print number sign always.
6629 print comma separated.
6631 `~MINCOL,PADCHAR,COMMACHARB'
6636 `~N,MINCOL,PADCHAR,COMMACHARR'
6640 print a number as a Roman numeral.
6643 print a number as an "old fashioned" Roman numeral.
6646 print a number as an ordinal English number.
6649 print a number as a cardinal English number.
6654 prints `y' and `ies'.
6657 as `~P but jumps 1 argument backward.'
6660 as `~@P but jumps 1 argument backward.'
6665 prints a character as the reader can understand it (i.e. `#\'
6669 prints a character as emacs does (eg. `^C' for ASCII 03).
6672 Fixed-format floating-point (prints a flonum like MMM.NNN).
6673 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
6675 If the number is positive a plus sign is printed.
6678 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
6679 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
6681 If the number is positive a plus sign is printed.
6684 General floating-point (prints a flonum either fixed or
6686 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
6688 If the number is positive a plus sign is printed.
6691 Dollars floating-point (prints a flonum in fixed with signs
6693 `~DIGITS,SCALE,WIDTH,PADCHAR$'
6695 If the number is positive a plus sign is printed.
6698 A sign is always printed and appears before the padding.
6701 The sign appears before the padding.
6709 print newline if not at the beginning of the output line.
6711 prints `~&' and then N-1 newlines.
6716 print N page separators.
6726 newline is ignored, white space left.
6729 newline is left, white space ignored.
6734 relative tabulation.
6740 Indirection (expects indirect arguments as a list).
6742 extracts indirect arguments from format arguments.
6745 Case conversion (converts by `string-downcase').
6747 converts by `string-capitalize'.
6750 converts by `string-capitalize-first'.
6753 converts by `string-upcase'.
6756 Argument Jumping (jumps 1 argument forward).
6758 jumps N arguments forward.
6761 jumps 1 argument backward.
6764 jumps N arguments backward.
6767 jumps to the 0th argument.
6770 jumps to the Nth argument (beginning from 0)
6772 `~[STR0~;STR1~;...~;STRN~]'
6773 Conditional Expression (numerical clause conditional).
6775 take argument from N.
6778 true test conditional.
6781 if-else-then conditional.
6787 default clause follows.
6790 Iteration (args come from the next argument (a list)).
6792 at most N iterations.
6795 args from next arg (a list of lists).
6798 args from the rest of arguments.
6801 args from the rest args (lists).
6812 aborts if N <= M <= K
6814 *** Not Implemented CL Format Control Directives
6817 print `#f' as an empty list (see below).
6820 print `#f' as an empty list (see below).
6826 (sorry I don't understand its semantics completely)
6828 *** Extended, Replaced and Additional Control Directives
6830 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
6831 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
6832 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
6833 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
6834 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
6835 COMMAWIDTH is the number of characters between two comma
6839 print a R4RS complex number as `~F~@Fi' with passed parameters for
6843 Pretty print formatting of an argument for scheme code lists.
6849 Flushes the output if format DESTINATION is a port.
6852 Print a `#\space' character
6854 print N `#\space' characters.
6857 Print a `#\tab' character
6859 print N `#\tab' characters.
6862 Takes N as an integer representation for a character. No arguments
6863 are consumed. N is converted to a character by `integer->char'. N
6864 must be a positive decimal number.
6867 Print out readproof. Prints out internal objects represented as
6868 `#<...>' as strings `"#<...>"' so that the format output can always
6869 be processed by `read'.
6872 Print out readproof. Prints out internal objects represented as
6873 `#<...>' as strings `"#<...>"' so that the format output can always
6874 be processed by `read'.
6877 Prints information and a copyright notice on the format
6880 prints format version.
6883 may also print number strings, i.e. passing a number as a string
6884 and format it accordingly.
6886 *** Configuration Variables
6888 The format module exports some configuration variables to suit the
6889 systems and users needs. There should be no modification necessary for
6890 the configuration that comes with Guile. Format detects automatically
6891 if the running scheme system implements floating point numbers and
6894 format:symbol-case-conv
6895 Symbols are converted by `symbol->string' so the case type of the
6896 printed symbols is implementation dependent.
6897 `format:symbol-case-conv' is a one arg closure which is either
6898 `#f' (no conversion), `string-upcase', `string-downcase' or
6899 `string-capitalize'. (default `#f')
6901 format:iobj-case-conv
6902 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
6903 implementation internal objects. (default `#f')
6906 The character prefixing the exponent value in `~E' printing.
6909 *** Compatibility With Other Format Implementations
6915 Downward compatible except for padding support and `~A', `~S',
6916 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
6917 `printf' padding support which is completely replaced by the CL
6918 `format' padding style.
6921 Downward compatible except for `~', which is not documented
6922 (ignores all characters inside the format string up to a newline
6923 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
6924 numerical and variable parameters and `:/@' modifiers in the CL
6928 Downward compatible except for `~A' and `~S' which print in
6929 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
6930 directive parameters or modifiers)).
6933 Downward compatible except for an optional destination parameter:
6934 S2C accepts a format call without a destination which returns a
6935 formatted string. This is equivalent to a #f destination in S2C.
6936 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
6937 parameters or modifiers)).
6940 ** Changes to string-handling functions.
6942 These functions were added to support the (ice-9 format) module, above.
6944 *** New function: string-upcase STRING
6945 *** New function: string-downcase STRING
6947 These are non-destructive versions of the existing string-upcase! and
6948 string-downcase! functions.
6950 *** New function: string-capitalize! STRING
6951 *** New function: string-capitalize STRING
6953 These functions convert the first letter of each word in the string to
6956 (string-capitalize "howdy there")
6959 As with the other functions, string-capitalize! modifies the string in
6960 place, while string-capitalize returns a modified copy of its argument.
6962 *** New function: string-ci->symbol STRING
6964 Return a symbol whose name is STRING, but having the same case as if
6965 the symbol had be read by `read'.
6967 Guile can be configured to be sensitive or insensitive to case
6968 differences in Scheme identifiers. If Guile is case-insensitive, all
6969 symbols are converted to lower case on input. The `string-ci->symbol'
6970 function returns a symbol whose name in STRING, transformed as Guile
6971 would if STRING were input.
6973 *** New function: substring-move! STRING1 START END STRING2 START
6975 Copy the substring of STRING1 from START (inclusive) to END
6976 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
6977 string, and the source and destination areas may overlap; in all
6978 cases, the function behaves as if all the characters were copied
6981 *** Extended functions: substring-move-left! substring-move-right!
6983 These functions now correctly copy arbitrarily overlapping substrings;
6984 they are both synonyms for substring-move!.
6987 ** New module (ice-9 getopt-long), with the function `getopt-long'.
6989 getopt-long is a function for parsing command-line arguments in a
6990 manner consistent with other GNU programs.
6992 (getopt-long ARGS GRAMMAR)
6993 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
6995 ARGS should be a list of strings. Its first element should be the
6996 name of the program; subsequent elements should be the arguments
6997 that were passed to the program on the command line. The
6998 `program-arguments' procedure returns a list of this form.
7000 GRAMMAR is a list of the form:
7001 ((OPTION (PROPERTY VALUE) ...) ...)
7003 Each OPTION should be a symbol. `getopt-long' will accept a
7004 command-line option named `--OPTION'.
7005 Each option can have the following (PROPERTY VALUE) pairs:
7007 (single-char CHAR) --- Accept `-CHAR' as a single-character
7008 equivalent to `--OPTION'. This is how to specify traditional
7010 (required? BOOL) --- If BOOL is true, the option is required.
7011 getopt-long will raise an error if it is not found in ARGS.
7012 (value BOOL) --- If BOOL is #t, the option accepts a value; if
7013 it is #f, it does not; and if it is the symbol
7014 `optional', the option may appear in ARGS with or
7016 (predicate FUNC) --- If the option accepts a value (i.e. you
7017 specified `(value #t)' for this option), then getopt
7018 will apply FUNC to the value, and throw an exception
7019 if it returns #f. FUNC should be a procedure which
7020 accepts a string and returns a boolean value; you may
7021 need to use quasiquotes to get it into GRAMMAR.
7023 The (PROPERTY VALUE) pairs may occur in any order, but each
7024 property may occur only once. By default, options do not have
7025 single-character equivalents, are not required, and do not take
7028 In ARGS, single-character options may be combined, in the usual
7029 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
7030 accepts values, then it must be the last option in the
7031 combination; the value is the next argument. So, for example, using
7032 the following grammar:
7033 ((apples (single-char #\a))
7034 (blimps (single-char #\b) (value #t))
7035 (catalexis (single-char #\c) (value #t)))
7036 the following argument lists would be acceptable:
7037 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
7038 for "blimps" and "catalexis")
7039 ("-ab" "bang" "-c" "couth") (same)
7040 ("-ac" "couth" "-b" "bang") (same)
7041 ("-abc" "couth" "bang") (an error, since `-b' is not the
7042 last option in its combination)
7044 If an option's value is optional, then `getopt-long' decides
7045 whether it has a value by looking at what follows it in ARGS. If
7046 the next element is a string, and it does not appear to be an
7047 option itself, then that string is the option's value.
7049 The value of a long option can appear as the next element in ARGS,
7050 or it can follow the option name, separated by an `=' character.
7051 Thus, using the same grammar as above, the following argument lists
7053 ("--apples" "Braeburn" "--blimps" "Goodyear")
7054 ("--apples=Braeburn" "--blimps" "Goodyear")
7055 ("--blimps" "Goodyear" "--apples=Braeburn")
7057 If the option "--" appears in ARGS, argument parsing stops there;
7058 subsequent arguments are returned as ordinary arguments, even if
7059 they resemble options. So, in the argument list:
7060 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
7061 `getopt-long' will recognize the `apples' option as having the
7062 value "Granny Smith", but it will not recognize the `blimp'
7063 option; it will return the strings "--blimp" and "Goodyear" as
7064 ordinary argument strings.
7066 The `getopt-long' function returns the parsed argument list as an
7067 assocation list, mapping option names --- the symbols from GRAMMAR
7068 --- onto their values, or #t if the option does not accept a value.
7069 Unused options do not appear in the alist.
7071 All arguments that are not the value of any option are returned
7072 as a list, associated with the empty list.
7074 `getopt-long' throws an exception if:
7075 - it finds an unrecognized option in ARGS
7076 - a required option is omitted
7077 - an option that requires an argument doesn't get one
7078 - an option that doesn't accept an argument does get one (this can
7079 only happen using the long option `--opt=value' syntax)
7080 - an option predicate fails
7085 `((lockfile-dir (required? #t)
7088 (predicate ,file-is-directory?))
7089 (verbose (required? #f)
7092 (x-includes (single-char #\x))
7093 (rnet-server (single-char #\y)
7094 (predicate ,string?))))
7096 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
7097 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
7099 => ((() "foo1" "-fred" "foo2" "foo3")
7100 (rnet-server . "lamprod")
7101 (x-includes . "/usr/include")
7102 (lockfile-dir . "/tmp")
7105 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
7107 It will be removed in a few releases.
7109 ** New syntax: lambda*
7110 ** New syntax: define*
7111 ** New syntax: define*-public
7112 ** New syntax: defmacro*
7113 ** New syntax: defmacro*-public
7114 Guile now supports optional arguments.
7116 `lambda*', `define*', `define*-public', `defmacro*' and
7117 `defmacro*-public' are identical to the non-* versions except that
7118 they use an extended type of parameter list that has the following BNF
7119 syntax (parentheses are literal, square brackets indicate grouping,
7120 and `*', `+' and `?' have the usual meaning):
7122 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
7123 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
7124 [[#&rest identifier]|[. identifier]]? ) | [identifier]
7126 ext-var-decl ::= identifier | ( identifier expression )
7128 The semantics are best illustrated with the following documentation
7129 and examples for `lambda*':
7132 lambda extended for optional and keyword arguments
7134 lambda* creates a procedure that takes optional arguments. These
7135 are specified by putting them inside brackets at the end of the
7136 paramater list, but before any dotted rest argument. For example,
7137 (lambda* (a b #&optional c d . e) '())
7138 creates a procedure with fixed arguments a and b, optional arguments c
7139 and d, and rest argument e. If the optional arguments are omitted
7140 in a call, the variables for them are unbound in the procedure. This
7141 can be checked with the bound? macro.
7143 lambda* can also take keyword arguments. For example, a procedure
7145 (lambda* (#&key xyzzy larch) '())
7146 can be called with any of the argument lists (#:xyzzy 11)
7147 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
7148 are given as keywords are bound to values.
7150 Optional and keyword arguments can also be given default values
7151 which they take on when they are not present in a call, by giving a
7152 two-item list in place of an optional argument, for example in:
7153 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
7154 foo is a fixed argument, bar is an optional argument with default
7155 value 42, and baz is a keyword argument with default value 73.
7156 Default value expressions are not evaluated unless they are needed
7157 and until the procedure is called.
7159 lambda* now supports two more special parameter list keywords.
7161 lambda*-defined procedures now throw an error by default if a
7162 keyword other than one of those specified is found in the actual
7163 passed arguments. However, specifying #&allow-other-keys
7164 immediately after the kyword argument declarations restores the
7165 previous behavior of ignoring unknown keywords. lambda* also now
7166 guarantees that if the same keyword is passed more than once, the
7167 last one passed is the one that takes effect. For example,
7168 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
7169 #:heads 37 #:tails 42 #:heads 99)
7170 would result in (99 47) being displayed.
7172 #&rest is also now provided as a synonym for the dotted syntax rest
7173 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
7174 all respects to lambda*. This is provided for more similarity to DSSSL,
7175 MIT-Scheme and Kawa among others, as well as for refugees from other
7178 Further documentation may be found in the optargs.scm file itself.
7180 The optional argument module also exports the macros `let-optional',
7181 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
7182 are not documented here because they may be removed in the future, but
7183 full documentation is still available in optargs.scm.
7185 ** New syntax: and-let*
7186 Guile now supports the `and-let*' form, described in the draft SRFI-2.
7188 Syntax: (land* (<clause> ...) <body> ...)
7189 Each <clause> should have one of the following forms:
7190 (<variable> <expression>)
7193 Each <variable> or <bound-variable> should be an identifier. Each
7194 <expression> should be a valid expression. The <body> should be a
7195 possibly empty sequence of expressions, like the <body> of a
7198 Semantics: A LAND* expression is evaluated by evaluating the
7199 <expression> or <bound-variable> of each of the <clause>s from
7200 left to right. The value of the first <expression> or
7201 <bound-variable> that evaluates to a false value is returned; the
7202 remaining <expression>s and <bound-variable>s are not evaluated.
7203 The <body> forms are evaluated iff all the <expression>s and
7204 <bound-variable>s evaluate to true values.
7206 The <expression>s and the <body> are evaluated in an environment
7207 binding each <variable> of the preceding (<variable> <expression>)
7208 clauses to the value of the <expression>. Later bindings
7209 shadow earlier bindings.
7211 Guile's and-let* macro was contributed by Michael Livshin.
7213 ** New sorting functions
7215 *** New function: sorted? SEQUENCE LESS?
7216 Returns `#t' when the sequence argument is in non-decreasing order
7217 according to LESS? (that is, there is no adjacent pair `... x y
7218 ...' for which `(less? y x)').
7220 Returns `#f' when the sequence contains at least one out-of-order
7221 pair. It is an error if the sequence is neither a list nor a
7224 *** New function: merge LIST1 LIST2 LESS?
7225 LIST1 and LIST2 are sorted lists.
7226 Returns the sorted list of all elements in LIST1 and LIST2.
7228 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
7229 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
7230 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
7231 (Here "<" should read "comes before".)
7233 *** New procedure: merge! LIST1 LIST2 LESS?
7234 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
7235 the result. If the code is compiled, and LESS? constructs no new
7236 pairs, no pairs at all will be allocated. The first pair of the
7237 result will be either the first pair of LIST1 or the first pair of
7240 *** New function: sort SEQUENCE LESS?
7241 Accepts either a list or a vector, and returns a new sequence
7242 which is sorted. The new sequence is the same type as the input.
7243 Always `(sorted? (sort sequence less?) less?)'. The original
7244 sequence is not altered in any way. The new sequence shares its
7245 elements with the old one; no elements are copied.
7247 *** New procedure: sort! SEQUENCE LESS
7248 Returns its sorted result in the original boxes. No new storage is
7249 allocated at all. Proper usage: (set! slist (sort! slist <))
7251 *** New function: stable-sort SEQUENCE LESS?
7252 Similar to `sort' but stable. That is, if "equal" elements are
7253 ordered a < b in the original sequence, they will have the same order
7256 *** New function: stable-sort! SEQUENCE LESS?
7257 Similar to `sort!' but stable.
7258 Uses temporary storage when sorting vectors.
7260 *** New functions: sort-list, sort-list!
7261 Added for compatibility with scsh.
7263 ** New built-in random number support
7265 *** New function: random N [STATE]
7266 Accepts a positive integer or real N and returns a number of the
7267 same type between zero (inclusive) and N (exclusive). The values
7268 returned have a uniform distribution.
7270 The optional argument STATE must be of the type produced by
7271 `copy-random-state' or `seed->random-state'. It defaults to the value
7272 of the variable `*random-state*'. This object is used to maintain the
7273 state of the pseudo-random-number generator and is altered as a side
7274 effect of the `random' operation.
7276 *** New variable: *random-state*
7277 Holds a data structure that encodes the internal state of the
7278 random-number generator that `random' uses by default. The nature
7279 of this data structure is implementation-dependent. It may be
7280 printed out and successfully read back in, but may or may not
7281 function correctly as a random-number state object in another
7284 *** New function: copy-random-state [STATE]
7285 Returns a new object of type suitable for use as the value of the
7286 variable `*random-state*' and as a second argument to `random'.
7287 If argument STATE is given, a copy of it is returned. Otherwise a
7288 copy of `*random-state*' is returned.
7290 *** New function: seed->random-state SEED
7291 Returns a new object of type suitable for use as the value of the
7292 variable `*random-state*' and as a second argument to `random'.
7293 SEED is a string or a number. A new state is generated and
7294 initialized using SEED.
7296 *** New function: random:uniform [STATE]
7297 Returns an uniformly distributed inexact real random number in the
7298 range between 0 and 1.
7300 *** New procedure: random:solid-sphere! VECT [STATE]
7301 Fills VECT with inexact real random numbers the sum of whose
7302 squares is less than 1.0. Thinking of VECT as coordinates in
7303 space of dimension N = `(vector-length VECT)', the coordinates are
7304 uniformly distributed within the unit N-shere. The sum of the
7305 squares of the numbers is returned. VECT can be either a vector
7306 or a uniform vector of doubles.
7308 *** New procedure: random:hollow-sphere! VECT [STATE]
7309 Fills VECT with inexact real random numbers the sum of whose squares
7310 is equal to 1.0. Thinking of VECT as coordinates in space of
7311 dimension n = `(vector-length VECT)', the coordinates are uniformly
7312 distributed over the surface of the unit n-shere. VECT can be either
7313 a vector or a uniform vector of doubles.
7315 *** New function: random:normal [STATE]
7316 Returns an inexact real in a normal distribution with mean 0 and
7317 standard deviation 1. For a normal distribution with mean M and
7318 standard deviation D use `(+ M (* D (random:normal)))'.
7320 *** New procedure: random:normal-vector! VECT [STATE]
7321 Fills VECT with inexact real random numbers which are independent and
7322 standard normally distributed (i.e., with mean 0 and variance 1).
7323 VECT can be either a vector or a uniform vector of doubles.
7325 *** New function: random:exp STATE
7326 Returns an inexact real in an exponential distribution with mean 1.
7327 For an exponential distribution with mean U use (* U (random:exp)).
7329 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
7331 These functions now operate on numbers in the range of a C unsigned
7334 These functions used to operate on numbers in the range of a C signed
7335 long; however, this seems inappropriate, because Guile integers don't
7338 ** New function: make-guardian
7339 This is an implementation of guardians as described in
7340 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
7341 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
7342 Programming Language Design and Implementation, June 1993
7343 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
7345 ** New functions: delq1!, delv1!, delete1!
7346 These procedures behave similar to delq! and friends but delete only
7347 one object if at all.
7349 ** New function: unread-string STRING PORT
7350 Unread STRING to PORT, that is, push it back onto the port so that
7351 next read operation will work on the pushed back characters.
7353 ** unread-char can now be called multiple times
7354 If unread-char is called multiple times, the unread characters will be
7355 read again in last-in first-out order.
7357 ** the procedures uniform-array-read! and uniform-array-write! now
7358 work on any kind of port, not just ports which are open on a file.
7360 ** Now 'l' in a port mode requests line buffering.
7362 ** The procedure truncate-file now works on string ports as well
7363 as file ports. If the size argument is omitted, the current
7364 file position is used.
7366 ** new procedure: seek PORT/FDES OFFSET WHENCE
7367 The arguments are the same as for the old fseek procedure, but it
7368 works on string ports as well as random-access file ports.
7370 ** the fseek procedure now works on string ports, since it has been
7371 redefined using seek.
7373 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
7374 size is not supplied.
7376 ** the newline procedure no longer flushes the port if it's not
7377 line-buffered: previously it did if it was the current output port.
7379 ** open-pipe and close-pipe are no longer primitive procedures, but
7380 an emulation can be obtained using `(use-modules (ice-9 popen))'.
7382 ** the freopen procedure has been removed.
7384 ** new procedure: drain-input PORT
7385 Drains PORT's read buffers (including any pushed-back characters)
7386 and returns the contents as a single string.
7388 ** New function: map-in-order PROC LIST1 LIST2 ...
7389 Version of `map' which guarantees that the procedure is applied to the
7390 lists in serial order.
7392 ** Renamed `serial-array-copy!' and `serial-array-map!' to
7393 `array-copy-in-order!' and `array-map-in-order!'. The old names are
7394 now obsolete and will go away in release 1.5.
7396 ** New syntax: collect BODY1 ...
7397 Version of `begin' which returns a list of the results of the body
7398 forms instead of the result of the last body form. In contrast to
7399 `begin', `collect' allows an empty body.
7401 ** New functions: read-history FILENAME, write-history FILENAME
7402 Read/write command line history from/to file. Returns #t on success
7403 and #f if an error occured.
7405 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
7407 These procedures return a list of definitions available in the specified
7408 argument, a relative module reference. In the case of no argument,
7409 `(current-module)' is now consulted for definitions to return, instead
7410 of simply returning #f, the former behavior.
7412 ** The #/ syntax for lists is no longer supported.
7414 Earlier versions of Scheme accepted this syntax, but printed a
7417 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
7419 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
7422 * Changes to the gh_ interface
7426 Now takes a second argument which is the result array. If this
7427 pointer is NULL, a new array is malloced (the old behaviour).
7429 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
7430 gh_scm2shorts, gh_scm2longs, gh_scm2floats
7434 * Changes to the scm_ interface
7436 ** Function: scm_make_named_hook (char* name, int n_args)
7438 Creates a hook in the same way as make-hook above but also
7439 binds a variable named NAME to it.
7441 This is the typical way of creating a hook from C code.
7443 Currently, the variable is created in the "current" module. This
7444 might change when we get the new module system.
7446 ** The smob interface
7448 The interface for creating smobs has changed. For documentation, see
7449 data-rep.info (made from guile-core/doc/data-rep.texi).
7451 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
7453 >>> This function will be removed in 1.3.4. <<<
7457 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
7458 This function adds a new smob type, named NAME, with instance size
7459 SIZE to the system. The return value is a tag that is used in
7460 creating instances of the type. If SIZE is 0, then no memory will
7461 be allocated when instances of the smob are created, and nothing
7462 will be freed by the default free function.
7464 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
7465 This function sets the smob marking procedure for the smob type
7466 specified by the tag TC. TC is the tag returned by
7467 `scm_make_smob_type'.
7469 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
7470 This function sets the smob freeing procedure for the smob type
7471 specified by the tag TC. TC is the tag returned by
7472 `scm_make_smob_type'.
7474 *** Function: void scm_set_smob_print (tc, print)
7476 - Function: void scm_set_smob_print (long tc,
7477 scm_sizet (*print) (SCM,
7481 This function sets the smob printing procedure for the smob type
7482 specified by the tag TC. TC is the tag returned by
7483 `scm_make_smob_type'.
7485 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
7486 This function sets the smob equality-testing predicate for the
7487 smob type specified by the tag TC. TC is the tag returned by
7488 `scm_make_smob_type'.
7490 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
7491 Make VALUE contain a smob instance of the type with type code TC and
7492 smob data DATA. VALUE must be previously declared as C type `SCM'.
7494 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
7495 This macro expands to a block of code that creates a smob instance
7496 of the type with type code TC and smob data DATA, and returns that
7497 `SCM' value. It should be the last piece of code in a block.
7499 ** The interfaces for using I/O ports and implementing port types
7500 (ptobs) have changed significantly. The new interface is based on
7501 shared access to buffers and a new set of ptob procedures.
7503 *** scm_newptob has been removed
7507 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
7509 - Function: SCM scm_make_port_type (char *type_name,
7510 int (*fill_buffer) (SCM port),
7511 void (*write_flush) (SCM port));
7513 Similarly to the new smob interface, there is a set of function
7514 setters by which the user can customize the behaviour of his port
7515 type. See ports.h (scm_set_port_XXX).
7517 ** scm_strport_to_string: New function: creates a new string from
7518 a string port's buffer.
7520 ** Plug in interface for random number generators
7521 The variable `scm_the_rng' in random.c contains a value and three
7522 function pointers which together define the current random number
7523 generator being used by the Scheme level interface and the random
7524 number library functions.
7526 The user is free to replace the default generator with the generator
7529 *** Variable: size_t scm_the_rng.rstate_size
7530 The size of the random state type used by the current RNG
7533 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
7534 Given the random STATE, return 32 random bits.
7536 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
7537 Seed random state STATE using string S of length N.
7539 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
7540 Given random state STATE, return a malloced copy.
7543 The default RNG is the MWC (Multiply With Carry) random number
7544 generator described by George Marsaglia at the Department of
7545 Statistics and Supercomputer Computations Research Institute, The
7546 Florida State University (http://stat.fsu.edu/~geo).
7548 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
7549 passes all tests in the DIEHARD test suite
7550 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
7551 costs one multiply and one add on platforms which either supports long
7552 longs (gcc does this on most systems) or have 64 bit longs. The cost
7553 is four multiply on other systems but this can be optimized by writing
7554 scm_i_uniform32 in assembler.
7556 These functions are provided through the scm_the_rng interface for use
7557 by libguile and the application.
7559 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
7560 Given the random STATE, return 32 random bits.
7561 Don't use this function directly. Instead go through the plugin
7562 interface (see "Plug in interface" above).
7564 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
7565 Initialize STATE using SEED of length N.
7567 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
7568 Return a malloc:ed copy of STATE. This function can easily be re-used
7569 in the interfaces to other RNGs.
7571 ** Random number library functions
7572 These functions use the current RNG through the scm_the_rng interface.
7573 It might be a good idea to use these functions from your C code so
7574 that only one random generator is used by all code in your program.
7576 The default random state is stored in:
7578 *** Variable: SCM scm_var_random_state
7579 Contains the vcell of the Scheme variable "*random-state*" which is
7580 used as default state by all random number functions in the Scheme
7585 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
7587 *** Function: scm_rstate *scm_c_default_rstate (void)
7588 This is a convenience function which returns the value of
7589 scm_var_random_state. An error message is generated if this value
7590 isn't a random state.
7592 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
7593 Make a new random state from the string SEED of length LENGTH.
7595 It is generally not a good idea to use multiple random states in a
7596 program. While subsequent random numbers generated from one random
7597 state are guaranteed to be reasonably independent, there is no such
7598 guarantee for numbers generated from different random states.
7600 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
7601 Return 32 random bits.
7603 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
7604 Return a sample from the uniform(0,1) distribution.
7606 *** Function: double scm_c_normal01 (scm_rstate *STATE)
7607 Return a sample from the normal(0,1) distribution.
7609 *** Function: double scm_c_exp1 (scm_rstate *STATE)
7610 Return a sample from the exp(1) distribution.
7612 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
7613 Return a sample from the discrete uniform(0,M) distribution.
7615 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
7616 Return a sample from the discrete uniform(0,M) distribution.
7617 M must be a bignum object. The returned value may be an INUM.
7621 Changes in Guile 1.3 (released Monday, October 19, 1998):
7623 * Changes to the distribution
7625 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
7626 To avoid conflicts, programs should name environment variables after
7627 themselves, except when there's a common practice establishing some
7630 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
7631 giving the former precedence, and printing a warning message if the
7632 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
7634 ** The header files related to multi-byte characters have been removed.
7635 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
7636 which referred to these explicitly will probably need to be rewritten,
7637 since the support for the variant string types has been removed; see
7640 ** The header files append.h and sequences.h have been removed. These
7641 files implemented non-R4RS operations which would encourage
7642 non-portable programming style and less easy-to-read code.
7644 * Changes to the stand-alone interpreter
7646 ** New procedures have been added to implement a "batch mode":
7648 *** Function: batch-mode?
7650 Returns a boolean indicating whether the interpreter is in batch
7653 *** Function: set-batch-mode?! ARG
7655 If ARG is true, switches the interpreter to batch mode. The `#f'
7656 case has not been implemented.
7658 ** Guile now provides full command-line editing, when run interactively.
7659 To use this feature, you must have the readline library installed.
7660 The Guile build process will notice it, and automatically include
7663 The readline library is available via anonymous FTP from any GNU
7664 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
7666 ** the-last-stack is now a fluid.
7668 * Changes to the procedure for linking libguile with your programs
7670 ** You can now use the `guile-config' utility to build programs that use Guile.
7672 Guile now includes a command-line utility called `guile-config', which
7673 can provide information about how to compile and link programs that
7676 *** `guile-config compile' prints any C compiler flags needed to use Guile.
7677 You should include this command's output on the command line you use
7678 to compile C or C++ code that #includes the Guile header files. It's
7679 usually just a `-I' flag to help the compiler find the Guile headers.
7682 *** `guile-config link' prints any linker flags necessary to link with Guile.
7684 This command writes to its standard output a list of flags which you
7685 must pass to the linker to link your code against the Guile library.
7686 The flags include '-lguile' itself, any other libraries the Guile
7687 library depends upon, and any `-L' flags needed to help the linker
7688 find those libraries.
7690 For example, here is a Makefile rule that builds a program named 'foo'
7691 from the object files ${FOO_OBJECTS}, and links them against Guile:
7694 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
7696 Previous Guile releases recommended that you use autoconf to detect
7697 which of a predefined set of libraries were present on your system.
7698 It is more robust to use `guile-config', since it records exactly which
7699 libraries the installed Guile library requires.
7701 This was originally called `build-guile', but was renamed to
7702 `guile-config' before Guile 1.3 was released, to be consistent with
7703 the analogous script for the GTK+ GUI toolkit, which is called
7707 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
7709 If you are using the GNU autoconf package to configure your program,
7710 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
7711 (described above) and gather the necessary values for use in your
7714 The GUILE_FLAGS macro expands to configure script code which runs the
7715 `guile-config' script, to find out where Guile's header files and
7716 libraries are installed. It sets two variables, marked for
7717 substitution, as by AC_SUBST.
7719 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
7720 code that uses Guile header files. This is almost always just a
7723 GUILE_LDFLAGS --- flags to pass to the linker to link a
7724 program against Guile. This includes `-lguile' for the Guile
7725 library itself, any libraries that Guile itself requires (like
7726 -lqthreads), and so on. It may also include a -L flag to tell the
7727 compiler where to find the libraries.
7729 GUILE_FLAGS is defined in the file guile.m4, in the top-level
7730 directory of the Guile distribution. You can copy it into your
7731 package's aclocal.m4 file, and then use it in your configure.in file.
7733 If you are using the `aclocal' program, distributed with GNU automake,
7734 to maintain your aclocal.m4 file, the Guile installation process
7735 installs guile.m4 where aclocal will find it. All you need to do is
7736 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
7737 this will copy the definition of GUILE_FLAGS into your aclocal.m4
7741 * Changes to Scheme functions and syntax
7743 ** Multi-byte strings have been removed, as have multi-byte and wide
7744 ports. We felt that these were the wrong approach to
7745 internationalization support.
7747 ** New function: readline [PROMPT]
7748 Read a line from the terminal, and allow the user to edit it,
7749 prompting with PROMPT. READLINE provides a large set of Emacs-like
7750 editing commands, lets the user recall previously typed lines, and
7751 works on almost every kind of terminal, including dumb terminals.
7753 READLINE assumes that the cursor is at the beginning of the line when
7754 it is invoked. Thus, you can't print a prompt yourself, and then call
7755 READLINE; you need to package up your prompt as a string, pass it to
7756 the function, and let READLINE print the prompt itself. This is
7757 because READLINE needs to know the prompt's screen width.
7759 For Guile to provide this function, you must have the readline
7760 library, version 2.1 or later, installed on your system. Readline is
7761 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
7762 any GNU mirror site.
7764 See also ADD-HISTORY function.
7766 ** New function: add-history STRING
7767 Add STRING as the most recent line in the history used by the READLINE
7768 command. READLINE does not add lines to the history itself; you must
7769 call ADD-HISTORY to make previous input available to the user.
7771 ** The behavior of the read-line function has changed.
7773 This function now uses standard C library functions to read the line,
7774 for speed. This means that it doesn not respect the value of
7775 scm-line-incrementors; it assumes that lines are delimited with
7778 (Note that this is read-line, the function that reads a line of text
7779 from a port, not readline, the function that reads a line from a
7780 terminal, providing full editing capabilities.)
7782 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
7784 This module provides some simple argument parsing. It exports one
7787 Function: getopt-gnu-style ARG-LS
7788 Parse a list of program arguments into an alist of option
7791 Each item in the list of program arguments is examined to see if
7792 it meets the syntax of a GNU long-named option. An argument like
7793 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
7794 returned alist, where MUMBLE is a keyword object with the same
7795 name as the argument. An argument like `--MUMBLE=FROB' produces
7796 an element of the form (MUMBLE . FROB), where FROB is a string.
7798 As a special case, the returned alist also contains a pair whose
7799 car is the symbol `rest'. The cdr of this pair is a list
7800 containing all the items in the argument list that are not options
7801 of the form mentioned above.
7803 The argument `--' is treated specially: all items in the argument
7804 list appearing after such an argument are not examined, and are
7805 returned in the special `rest' list.
7807 This function does not parse normal single-character switches.
7808 You will need to parse them out of the `rest' list yourself.
7810 ** The read syntax for byte vectors and short vectors has changed.
7812 Instead of #bytes(...), write #y(...).
7814 Instead of #short(...), write #h(...).
7816 This may seem nutty, but, like the other uniform vectors, byte vectors
7817 and short vectors want to have the same print and read syntax (and,
7818 more basic, want to have read syntax!). Changing the read syntax to
7819 use multiple characters after the hash sign breaks with the
7820 conventions used in R5RS and the conventions used for the other
7821 uniform vectors. It also introduces complexity in the current reader,
7822 both on the C and Scheme levels. (The Right solution is probably to
7823 change the syntax and prototypes for uniform vectors entirely.)
7826 ** The new module (ice-9 session) provides useful interactive functions.
7828 *** New procedure: (apropos REGEXP OPTION ...)
7830 Display a list of top-level variables whose names match REGEXP, and
7831 the modules they are imported from. Each OPTION should be one of the
7834 value --- Show the value of each matching variable.
7835 shadow --- Show bindings shadowed by subsequently imported modules.
7836 full --- Same as both `shadow' and `value'.
7840 guile> (apropos "trace" 'full)
7841 debug: trace #<procedure trace args>
7842 debug: untrace #<procedure untrace args>
7843 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
7844 the-scm-module: before-backtrace-hook ()
7845 the-scm-module: backtrace #<primitive-procedure backtrace>
7846 the-scm-module: after-backtrace-hook ()
7847 the-scm-module: has-shown-backtrace-hint? #f
7850 ** There are new functions and syntax for working with macros.
7852 Guile implements macros as a special object type. Any variable whose
7853 top-level binding is a macro object acts as a macro. The macro object
7854 specifies how the expression should be transformed before evaluation.
7856 *** Macro objects now print in a reasonable way, resembling procedures.
7858 *** New function: (macro? OBJ)
7859 True iff OBJ is a macro object.
7861 *** New function: (primitive-macro? OBJ)
7862 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
7863 macro transformers, implemented in eval.c rather than Scheme code.
7865 Why do we have this function?
7866 - For symmetry with procedure? and primitive-procedure?,
7867 - to allow custom print procedures to tell whether a macro is
7868 primitive, and display it differently, and
7869 - to allow compilers and user-written evaluators to distinguish
7870 builtin special forms from user-defined ones, which could be
7873 *** New function: (macro-type OBJ)
7874 Return a value indicating what kind of macro OBJ is. Possible return
7877 The symbol `syntax' --- a macro created by procedure->syntax.
7878 The symbol `macro' --- a macro created by procedure->macro.
7879 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
7880 The boolean #f --- if OBJ is not a macro object.
7882 *** New function: (macro-name MACRO)
7883 Return the name of the macro object MACRO's procedure, as returned by
7886 *** New function: (macro-transformer MACRO)
7887 Return the transformer procedure for MACRO.
7889 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
7891 Specify a new macro expander to use in the current module. Each
7892 MODULE is a module name, with the same meaning as in the `use-modules'
7893 form; each named module's exported bindings are added to the current
7894 top-level environment. TRANSFORMER is an expression evaluated in the
7895 resulting environment which must yield a procedure to use as the
7896 module's eval transformer: every expression evaluated in this module
7897 is passed to this function, and the result passed to the Guile
7900 *** macro-eval! is removed. Use local-eval instead.
7902 ** Some magic has been added to the printer to better handle user
7903 written printing routines (like record printers, closure printers).
7905 The problem is that these user written routines must have access to
7906 the current `print-state' to be able to handle fancy things like
7907 detection of circular references. These print-states have to be
7908 passed to the builtin printing routines (display, write, etc) to
7909 properly continue the print chain.
7911 We didn't want to change all existing print code so that it
7912 explicitly passes thru a print state in addition to a port. Instead,
7913 we extented the possible values that the builtin printing routines
7914 accept as a `port'. In addition to a normal port, they now also take
7915 a pair of a normal port and a print-state. Printing will go to the
7916 port and the print-state will be used to control the detection of
7917 circular references, etc. If the builtin function does not care for a
7918 print-state, it is simply ignored.
7920 User written callbacks are now called with such a pair as their
7921 `port', but because every function now accepts this pair as a PORT
7922 argument, you don't have to worry about that. In fact, it is probably
7923 safest to not check for these pairs.
7925 However, it is sometimes necessary to continue a print chain on a
7926 different port, for example to get a intermediate string
7927 representation of the printed value, mangle that string somehow, and
7928 then to finally print the mangled string. Use the new function
7930 inherit-print-state OLD-PORT NEW-PORT
7932 for this. It constructs a new `port' that prints to NEW-PORT but
7933 inherits the print-state of OLD-PORT.
7935 ** struct-vtable-offset renamed to vtable-offset-user
7937 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
7939 ** There is now a third optional argument to make-vtable-vtable
7940 (and fourth to make-struct) when constructing new types (vtables).
7941 This argument initializes field vtable-index-printer of the vtable.
7943 ** The detection of circular references has been extended to structs.
7944 That is, a structure that -- in the process of being printed -- prints
7945 itself does not lead to infinite recursion.
7947 ** There is now some basic support for fluids. Please read
7948 "libguile/fluid.h" to find out more. It is accessible from Scheme with
7949 the following functions and macros:
7951 Function: make-fluid
7953 Create a new fluid object. Fluids are not special variables or
7954 some other extension to the semantics of Scheme, but rather
7955 ordinary Scheme objects. You can store them into variables (that
7956 are still lexically scoped, of course) or into any other place you
7957 like. Every fluid has a initial value of `#f'.
7959 Function: fluid? OBJ
7961 Test whether OBJ is a fluid.
7963 Function: fluid-ref FLUID
7964 Function: fluid-set! FLUID VAL
7966 Access/modify the fluid FLUID. Modifications are only visible
7967 within the current dynamic root (that includes threads).
7969 Function: with-fluids* FLUIDS VALUES THUNK
7971 FLUIDS is a list of fluids and VALUES a corresponding list of
7972 values for these fluids. Before THUNK gets called the values are
7973 installed in the fluids and the old values of the fluids are
7974 saved in the VALUES list. When the flow of control leaves THUNK
7975 or reenters it, the values get swapped again. You might think of
7976 this as a `safe-fluid-excursion'. Note that the VALUES list is
7977 modified by `with-fluids*'.
7979 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
7981 The same as `with-fluids*' but with a different syntax. It looks
7982 just like `let', but both FLUID and VALUE are evaluated. Remember,
7983 fluids are not special variables but ordinary objects. FLUID
7984 should evaluate to a fluid.
7986 ** Changes to system call interfaces:
7988 *** close-port, close-input-port and close-output-port now return a
7989 boolean instead of an `unspecified' object. #t means that the port
7990 was successfully closed, while #f means it was already closed. It is
7991 also now possible for these procedures to raise an exception if an
7992 error occurs (some errors from write can be delayed until close.)
7994 *** the first argument to chmod, fcntl, ftell and fseek can now be a
7997 *** the third argument to fcntl is now optional.
7999 *** the first argument to chown can now be a file descriptor or a port.
8001 *** the argument to stat can now be a port.
8003 *** The following new procedures have been added (most use scsh
8006 *** procedure: close PORT/FD
8007 Similar to close-port (*note close-port: Closing Ports.), but also
8008 works on file descriptors. A side effect of closing a file
8009 descriptor is that any ports using that file descriptor are moved
8010 to a different file descriptor and have their revealed counts set
8013 *** procedure: port->fdes PORT
8014 Returns the integer file descriptor underlying PORT. As a side
8015 effect the revealed count of PORT is incremented.
8017 *** procedure: fdes->ports FDES
8018 Returns a list of existing ports which have FDES as an underlying
8019 file descriptor, without changing their revealed counts.
8021 *** procedure: fdes->inport FDES
8022 Returns an existing input port which has FDES as its underlying
8023 file descriptor, if one exists, and increments its revealed count.
8024 Otherwise, returns a new input port with a revealed count of 1.
8026 *** procedure: fdes->outport FDES
8027 Returns an existing output port which has FDES as its underlying
8028 file descriptor, if one exists, and increments its revealed count.
8029 Otherwise, returns a new output port with a revealed count of 1.
8031 The next group of procedures perform a `dup2' system call, if NEWFD
8032 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
8033 duplicated can be supplied as an integer or contained in a port. The
8034 type of value returned varies depending on which procedure is used.
8036 All procedures also have the side effect when performing `dup2' that
8037 any ports using NEWFD are moved to a different file descriptor and have
8038 their revealed counts set to zero.
8040 *** procedure: dup->fdes PORT/FD [NEWFD]
8041 Returns an integer file descriptor.
8043 *** procedure: dup->inport PORT/FD [NEWFD]
8044 Returns a new input port using the new file descriptor.
8046 *** procedure: dup->outport PORT/FD [NEWFD]
8047 Returns a new output port using the new file descriptor.
8049 *** procedure: dup PORT/FD [NEWFD]
8050 Returns a new port if PORT/FD is a port, with the same mode as the
8051 supplied port, otherwise returns an integer file descriptor.
8053 *** procedure: dup->port PORT/FD MODE [NEWFD]
8054 Returns a new port using the new file descriptor. MODE supplies a
8055 mode string for the port (*note open-file: File Ports.).
8057 *** procedure: setenv NAME VALUE
8058 Modifies the environment of the current process, which is also the
8059 default environment inherited by child processes.
8061 If VALUE is `#f', then NAME is removed from the environment.
8062 Otherwise, the string NAME=VALUE is added to the environment,
8063 replacing any existing string with name matching NAME.
8065 The return value is unspecified.
8067 *** procedure: truncate-file OBJ SIZE
8068 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
8069 can be a string containing a file name or an integer file
8070 descriptor or port open for output on the file. The underlying
8071 system calls are `truncate' and `ftruncate'.
8073 The return value is unspecified.
8075 *** procedure: setvbuf PORT MODE [SIZE]
8076 Set the buffering mode for PORT. MODE can be:
8084 block buffered, using a newly allocated buffer of SIZE bytes.
8085 However if SIZE is zero or unspecified, the port will be made
8088 This procedure should not be used after I/O has been performed with
8091 Ports are usually block buffered by default, with a default buffer
8092 size. Procedures e.g., *Note open-file: File Ports, which accept a
8093 mode string allow `0' to be added to request an unbuffered port.
8095 *** procedure: fsync PORT/FD
8096 Copies any unwritten data for the specified output file descriptor
8097 to disk. If PORT/FD is a port, its buffer is flushed before the
8098 underlying file descriptor is fsync'd. The return value is
8101 *** procedure: open-fdes PATH FLAGS [MODES]
8102 Similar to `open' but returns a file descriptor instead of a port.
8104 *** procedure: execle PATH ENV [ARG] ...
8105 Similar to `execl', but the environment of the new process is
8106 specified by ENV, which must be a list of strings as returned by
8107 the `environ' procedure.
8109 This procedure is currently implemented using the `execve' system
8110 call, but we call it `execle' because of its Scheme calling
8113 *** procedure: strerror ERRNO
8114 Returns the Unix error message corresponding to ERRNO, an integer.
8116 *** procedure: primitive-exit [STATUS]
8117 Terminate the current process without unwinding the Scheme stack.
8118 This is would typically be useful after a fork. The exit status
8119 is STATUS if supplied, otherwise zero.
8121 *** procedure: times
8122 Returns an object with information about real and processor time.
8123 The following procedures accept such an object as an argument and
8124 return a selected component:
8127 The current real time, expressed as time units relative to an
8131 The CPU time units used by the calling process.
8134 The CPU time units used by the system on behalf of the
8138 The CPU time units used by terminated child processes of the
8139 calling process, whose status has been collected (e.g., using
8143 Similarly, the CPU times units used by the system on behalf of
8144 terminated child processes.
8146 ** Removed: list-length
8147 ** Removed: list-append, list-append!
8148 ** Removed: list-reverse, list-reverse!
8150 ** array-map renamed to array-map!
8152 ** serial-array-map renamed to serial-array-map!
8154 ** catch doesn't take #f as first argument any longer
8156 Previously, it was possible to pass #f instead of a key to `catch'.
8157 That would cause `catch' to pass a jump buffer object to the procedure
8158 passed as second argument. The procedure could then use this jump
8159 buffer objekt as an argument to throw.
8161 This mechanism has been removed since its utility doesn't motivate the
8162 extra complexity it introduces.
8164 ** The `#/' notation for lists now provokes a warning message from Guile.
8165 This syntax will be removed from Guile in the near future.
8167 To disable the warning message, set the GUILE_HUSH environment
8168 variable to any non-empty value.
8170 ** The newline character now prints as `#\newline', following the
8171 normal Scheme notation, not `#\nl'.
8173 * Changes to the gh_ interface
8175 ** The gh_enter function now takes care of loading the Guile startup files.
8176 gh_enter works by calling scm_boot_guile; see the remarks below.
8178 ** Function: void gh_write (SCM x)
8180 Write the printed representation of the scheme object x to the current
8181 output port. Corresponds to the scheme level `write'.
8183 ** gh_list_length renamed to gh_length.
8185 ** vector handling routines
8187 Several major changes. In particular, gh_vector() now resembles
8188 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
8189 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
8190 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
8191 vector-related gh_ functions have been implemented.
8193 ** pair and list routines
8195 Implemented several of the R4RS pair and list functions that were
8198 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
8200 New function. Converts double arrays back and forth between Scheme
8203 * Changes to the scm_ interface
8205 ** The function scm_boot_guile now takes care of loading the startup files.
8207 Guile's primary initialization function, scm_boot_guile, now takes
8208 care of loading `boot-9.scm', in the `ice-9' module, to initialize
8209 Guile, define the module system, and put together some standard
8210 bindings. It also loads `init.scm', which is intended to hold
8211 site-specific initialization code.
8213 Since Guile cannot operate properly until boot-9.scm is loaded, there
8214 is no reason to separate loading boot-9.scm from Guile's other
8215 initialization processes.
8217 This job used to be done by scm_compile_shell_switches, which didn't
8218 make much sense; in particular, it meant that people using Guile for
8219 non-shell-like applications had to jump through hoops to get Guile
8220 initialized properly.
8222 ** The function scm_compile_shell_switches no longer loads the startup files.
8223 Now, Guile always loads the startup files, whenever it is initialized;
8224 see the notes above for scm_boot_guile and scm_load_startup_files.
8226 ** Function: scm_load_startup_files
8227 This new function takes care of loading Guile's initialization file
8228 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
8229 this is always called by the Guile initialization process, it's
8230 probably not too useful to call this yourself, but it's there anyway.
8232 ** The semantics of smob marking have changed slightly.
8234 The smob marking function (the `mark' member of the scm_smobfuns
8235 structure) is no longer responsible for setting the mark bit on the
8236 smob. The generic smob handling code in the garbage collector will
8237 set this bit. The mark function need only ensure that any other
8238 objects the smob refers to get marked.
8240 Note that this change means that the smob's GC8MARK bit is typically
8241 already set upon entry to the mark function. Thus, marking functions
8242 which look like this:
8245 if (SCM_GC8MARKP (ptr))
8247 SCM_SETGC8MARK (ptr);
8248 ... mark objects to which the smob refers ...
8251 are now incorrect, since they will return early, and fail to mark any
8252 other objects the smob refers to. Some code in the Guile library used
8255 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
8257 If you have implemented your own I/O port type, by writing the
8258 functions required by the scm_ptobfuns and then calling scm_newptob,
8259 you will need to change your functions slightly.
8261 The functions in a scm_ptobfuns structure now expect the port itself
8262 as their argument; they used to expect the `stream' member of the
8263 port's scm_port_table structure. This allows functions in an
8264 scm_ptobfuns structure to easily access the port's cell (and any flags
8265 it its CAR), and the port's scm_port_table structure.
8267 Guile now passes the I/O port itself as the `port' argument in the
8268 following scm_ptobfuns functions:
8270 int (*free) (SCM port);
8271 int (*fputc) (int, SCM port);
8272 int (*fputs) (char *, SCM port);
8273 scm_sizet (*fwrite) SCM_P ((char *ptr,
8277 int (*fflush) (SCM port);
8278 int (*fgetc) (SCM port);
8279 int (*fclose) (SCM port);
8281 The interfaces to the `mark', `print', `equalp', and `fgets' methods
8284 If you have existing code which defines its own port types, it is easy
8285 to convert your code to the new interface; simply apply SCM_STREAM to
8286 the port argument to yield the value you code used to expect.
8288 Note that since both the port and the stream have the same type in the
8289 C code --- they are both SCM values --- the C compiler will not remind
8290 you if you forget to update your scm_ptobfuns functions.
8293 ** Function: int scm_internal_select (int fds,
8297 struct timeval *timeout);
8299 This is a replacement for the `select' function provided by the OS.
8300 It enables I/O blocking and sleeping to happen for one cooperative
8301 thread without blocking other threads. It also avoids busy-loops in
8302 these situations. It is intended that all I/O blocking and sleeping
8303 will finally go through this function. Currently, this function is
8304 only available on systems providing `gettimeofday' and `select'.
8306 ** Function: SCM scm_internal_stack_catch (SCM tag,
8307 scm_catch_body_t body,
8309 scm_catch_handler_t handler,
8312 A new sibling to the other two C level `catch' functions
8313 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
8314 the stack to be saved automatically into the variable `the-last-stack'
8315 (scm_the_last_stack_var) on error. This is necessary if you want to
8316 use advanced error reporting, such as calling scm_display_error and
8317 scm_display_backtrace. (They both take a stack object as argument.)
8319 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
8321 scm_catch_handler_t handler,
8324 Spawns a new thread. It does a job similar to
8325 scm_call_with_new_thread but takes arguments more suitable when
8326 spawning threads from application C code.
8328 ** The hook scm_error_callback has been removed. It was originally
8329 intended as a way for the user to install his own error handler. But
8330 that method works badly since it intervenes between throw and catch,
8331 thereby changing the semantics of expressions like (catch #t ...).
8332 The correct way to do it is to use one of the C level catch functions
8333 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
8335 ** Removed functions:
8337 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
8338 scm_list_reverse, scm_list_reverse_x
8340 ** New macros: SCM_LISTn where n is one of the integers 0-9.
8342 These can be used for pretty list creation from C. The idea is taken
8343 from Erick Gallesio's STk.
8345 ** scm_array_map renamed to scm_array_map_x
8347 ** mbstrings are now removed
8349 This means that the type codes scm_tc7_mb_string and
8350 scm_tc7_mb_substring has been removed.
8352 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
8354 Since we no longer support multi-byte strings, these I/O functions
8355 have been simplified, and renamed. Here are their old names, and
8356 their new names and arguments:
8358 scm_gen_putc -> void scm_putc (int c, SCM port);
8359 scm_gen_puts -> void scm_puts (char *s, SCM port);
8360 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
8361 scm_gen_getc -> void scm_getc (SCM port);
8364 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
8366 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
8368 SCM_TYP7S now masks away the bit which distinguishes substrings from
8371 ** scm_catch_body_t: Backward incompatible change!
8373 Body functions to scm_internal_catch and friends do not any longer
8374 take a second argument. This is because it is no longer possible to
8375 pass a #f arg to catch.
8377 ** Calls to scm_protect_object and scm_unprotect now nest properly.
8379 The function scm_protect_object protects its argument from being freed
8380 by the garbage collector. scm_unprotect_object removes that
8383 These functions now nest properly. That is, for every object O, there
8384 is a counter which scm_protect_object(O) increments and
8385 scm_unprotect_object(O) decrements, if the counter is greater than
8386 zero. Every object's counter is zero when it is first created. If an
8387 object's counter is greater than zero, the garbage collector will not
8388 reclaim its storage.
8390 This allows you to use scm_protect_object in your code without
8391 worrying that some other function you call will call
8392 scm_unprotect_object, and allow it to be freed. Assuming that the
8393 functions you call are well-behaved, and unprotect only those objects
8394 they protect, you can follow the same rule and have confidence that
8395 objects will be freed only at appropriate times.
8398 Changes in Guile 1.2 (released Tuesday, June 24 1997):
8400 * Changes to the distribution
8402 ** Nightly snapshots are now available from ftp.red-bean.com.
8403 The old server, ftp.cyclic.com, has been relinquished to its rightful
8406 Nightly snapshots of the Guile development sources are now available via
8407 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
8409 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
8410 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
8412 ** To run Guile without installing it, the procedure has changed a bit.
8414 If you used a separate build directory to compile Guile, you'll need
8415 to include the build directory in SCHEME_LOAD_PATH, as well as the
8416 source directory. See the `INSTALL' file for examples.
8418 * Changes to the procedure for linking libguile with your programs
8420 ** The standard Guile load path for Scheme code now includes
8421 $(datadir)/guile (usually /usr/local/share/guile). This means that
8422 you can install your own Scheme files there, and Guile will find them.
8423 (Previous versions of Guile only checked a directory whose name
8424 contained the Guile version number, so you had to re-install or move
8425 your Scheme sources each time you installed a fresh version of Guile.)
8427 The load path also includes $(datadir)/guile/site; we recommend
8428 putting individual Scheme files there. If you want to install a
8429 package with multiple source files, create a directory for them under
8432 ** Guile 1.2 will now use the Rx regular expression library, if it is
8433 installed on your system. When you are linking libguile into your own
8434 programs, this means you will have to link against -lguile, -lqt (if
8435 you configured Guile with thread support), and -lrx.
8437 If you are using autoconf to generate configuration scripts for your
8438 application, the following lines should suffice to add the appropriate
8439 libraries to your link command:
8441 ### Find Rx, quickthreads and libguile.
8442 AC_CHECK_LIB(rx, main)
8443 AC_CHECK_LIB(qt, main)
8444 AC_CHECK_LIB(guile, scm_shell)
8446 The Guile 1.2 distribution does not contain sources for the Rx
8447 library, as Guile 1.0 did. If you want to use Rx, you'll need to
8448 retrieve it from a GNU FTP site and install it separately.
8450 * Changes to Scheme functions and syntax
8452 ** The dynamic linking features of Guile are now enabled by default.
8453 You can disable them by giving the `--disable-dynamic-linking' option
8456 (dynamic-link FILENAME)
8458 Find the object file denoted by FILENAME (a string) and link it
8459 into the running Guile application. When everything works out,
8460 return a Scheme object suitable for representing the linked object
8461 file. Otherwise an error is thrown. How object files are
8462 searched is system dependent.
8464 (dynamic-object? VAL)
8466 Determine whether VAL represents a dynamically linked object file.
8468 (dynamic-unlink DYNOBJ)
8470 Unlink the indicated object file from the application. DYNOBJ
8471 should be one of the values returned by `dynamic-link'.
8473 (dynamic-func FUNCTION DYNOBJ)
8475 Search the C function indicated by FUNCTION (a string or symbol)
8476 in DYNOBJ and return some Scheme object that can later be used
8477 with `dynamic-call' to actually call this function. Right now,
8478 these Scheme objects are formed by casting the address of the
8479 function to `long' and converting this number to its Scheme
8482 (dynamic-call FUNCTION DYNOBJ)
8484 Call the C function indicated by FUNCTION and DYNOBJ. The
8485 function is passed no arguments and its return value is ignored.
8486 When FUNCTION is something returned by `dynamic-func', call that
8487 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
8488 etc.), look it up in DYNOBJ; this is equivalent to
8490 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
8492 Interrupts are deferred while the C function is executing (with
8493 SCM_DEFER_INTS/SCM_ALLOW_INTS).
8495 (dynamic-args-call FUNCTION DYNOBJ ARGS)
8497 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
8498 some arguments and return its return value. The C function is
8499 expected to take two arguments and return an `int', just like
8502 int c_func (int argc, char **argv);
8504 ARGS must be a list of strings and is converted into an array of
8505 `char *'. The array is passed in ARGV and its size in ARGC. The
8506 return value is converted to a Scheme number and returned from the
8507 call to `dynamic-args-call'.
8509 When dynamic linking is disabled or not supported on your system,
8510 the above functions throw errors, but they are still available.
8512 Here is a small example that works on GNU/Linux:
8514 (define libc-obj (dynamic-link "libc.so"))
8515 (dynamic-args-call 'rand libc-obj '())
8517 See the file `libguile/DYNAMIC-LINKING' for additional comments.
8519 ** The #/ syntax for module names is depreciated, and will be removed
8520 in a future version of Guile. Instead of
8528 The latter syntax is more consistent with existing Lisp practice.
8530 ** Guile now does fancier printing of structures. Structures are the
8531 underlying implementation for records, which in turn are used to
8532 implement modules, so all of these object now print differently and in
8533 a more informative way.
8535 The Scheme printer will examine the builtin variable *struct-printer*
8536 whenever it needs to print a structure object. When this variable is
8537 not `#f' it is deemed to be a procedure and will be applied to the
8538 structure object and the output port. When *struct-printer* is `#f'
8539 or the procedure return `#f' the structure object will be printed in
8540 the boring #<struct 80458270> form.
8542 This hook is used by some routines in ice-9/boot-9.scm to implement
8543 type specific printing routines. Please read the comments there about
8546 One of the more specific uses of structs are records. The printing
8547 procedure that could be passed to MAKE-RECORD-TYPE is now actually
8548 called. It should behave like a *struct-printer* procedure (described
8551 ** Guile now supports a new R4RS-compliant syntax for keywords. A
8552 token of the form #:NAME, where NAME has the same syntax as a Scheme
8553 symbol, is the external representation of the keyword named NAME.
8554 Keyword objects print using this syntax as well, so values containing
8555 keyword objects can be read back into Guile. When used in an
8556 expression, keywords are self-quoting objects.
8558 Guile suports this read syntax, and uses this print syntax, regardless
8559 of the current setting of the `keyword' read option. The `keyword'
8560 read option only controls whether Guile recognizes the `:NAME' syntax,
8561 which is incompatible with R4RS. (R4RS says such token represent
8564 ** Guile has regular expression support again. Guile 1.0 included
8565 functions for matching regular expressions, based on the Rx library.
8566 In Guile 1.1, the Guile/Rx interface was removed to simplify the
8567 distribution, and thus Guile had no regular expression support. Guile
8568 1.2 again supports the most commonly used functions, and supports all
8569 of SCSH's regular expression functions.
8571 If your system does not include a POSIX regular expression library,
8572 and you have not linked Guile with a third-party regexp library such as
8573 Rx, these functions will not be available. You can tell whether your
8574 Guile installation includes regular expression support by checking
8575 whether the `*features*' list includes the `regex' symbol.
8577 *** regexp functions
8579 By default, Guile supports POSIX extended regular expressions. That
8580 means that the characters `(', `)', `+' and `?' are special, and must
8581 be escaped if you wish to match the literal characters.
8583 This regular expression interface was modeled after that implemented
8584 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
8585 with SCSH regular expressions.
8587 **** Function: string-match PATTERN STR [START]
8588 Compile the string PATTERN into a regular expression and compare
8589 it with STR. The optional numeric argument START specifies the
8590 position of STR at which to begin matching.
8592 `string-match' returns a "match structure" which describes what,
8593 if anything, was matched by the regular expression. *Note Match
8594 Structures::. If STR does not match PATTERN at all,
8595 `string-match' returns `#f'.
8597 Each time `string-match' is called, it must compile its PATTERN
8598 argument into a regular expression structure. This operation is
8599 expensive, which makes `string-match' inefficient if the same regular
8600 expression is used several times (for example, in a loop). For better
8601 performance, you can compile a regular expression in advance and then
8602 match strings against the compiled regexp.
8604 **** Function: make-regexp STR [FLAGS]
8605 Compile the regular expression described by STR, and return the
8606 compiled regexp structure. If STR does not describe a legal
8607 regular expression, `make-regexp' throws a
8608 `regular-expression-syntax' error.
8610 FLAGS may be the bitwise-or of one or more of the following:
8612 **** Constant: regexp/extended
8613 Use POSIX Extended Regular Expression syntax when interpreting
8614 STR. If not set, POSIX Basic Regular Expression syntax is used.
8615 If the FLAGS argument is omitted, we assume regexp/extended.
8617 **** Constant: regexp/icase
8618 Do not differentiate case. Subsequent searches using the
8619 returned regular expression will be case insensitive.
8621 **** Constant: regexp/newline
8622 Match-any-character operators don't match a newline.
8624 A non-matching list ([^...]) not containing a newline matches a
8627 Match-beginning-of-line operator (^) matches the empty string
8628 immediately after a newline, regardless of whether the FLAGS
8629 passed to regexp-exec contain regexp/notbol.
8631 Match-end-of-line operator ($) matches the empty string
8632 immediately before a newline, regardless of whether the FLAGS
8633 passed to regexp-exec contain regexp/noteol.
8635 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
8636 Match the compiled regular expression REGEXP against `str'. If
8637 the optional integer START argument is provided, begin matching
8638 from that position in the string. Return a match structure
8639 describing the results of the match, or `#f' if no match could be
8642 FLAGS may be the bitwise-or of one or more of the following:
8644 **** Constant: regexp/notbol
8645 The match-beginning-of-line operator always fails to match (but
8646 see the compilation flag regexp/newline above) This flag may be
8647 used when different portions of a string are passed to
8648 regexp-exec and the beginning of the string should not be
8649 interpreted as the beginning of the line.
8651 **** Constant: regexp/noteol
8652 The match-end-of-line operator always fails to match (but see the
8653 compilation flag regexp/newline above)
8655 **** Function: regexp? OBJ
8656 Return `#t' if OBJ is a compiled regular expression, or `#f'
8659 Regular expressions are commonly used to find patterns in one string
8660 and replace them with the contents of another string.
8662 **** Function: regexp-substitute PORT MATCH [ITEM...]
8663 Write to the output port PORT selected contents of the match
8664 structure MATCH. Each ITEM specifies what should be written, and
8665 may be one of the following arguments:
8667 * A string. String arguments are written out verbatim.
8669 * An integer. The submatch with that number is written.
8671 * The symbol `pre'. The portion of the matched string preceding
8672 the regexp match is written.
8674 * The symbol `post'. The portion of the matched string
8675 following the regexp match is written.
8677 PORT may be `#f', in which case nothing is written; instead,
8678 `regexp-substitute' constructs a string from the specified ITEMs
8681 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
8682 Similar to `regexp-substitute', but can be used to perform global
8683 substitutions on STR. Instead of taking a match structure as an
8684 argument, `regexp-substitute/global' takes two string arguments: a
8685 REGEXP string describing a regular expression, and a TARGET string
8686 which should be matched against this regular expression.
8688 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
8691 * A function may be supplied. When this function is called, it
8692 will be passed one argument: a match structure for a given
8693 regular expression match. It should return a string to be
8694 written out to PORT.
8696 * The `post' symbol causes `regexp-substitute/global' to recurse
8697 on the unmatched portion of STR. This *must* be supplied in
8698 order to perform global search-and-replace on STR; if it is
8699 not present among the ITEMs, then `regexp-substitute/global'
8700 will return after processing a single match.
8702 *** Match Structures
8704 A "match structure" is the object returned by `string-match' and
8705 `regexp-exec'. It describes which portion of a string, if any, matched
8706 the given regular expression. Match structures include: a reference to
8707 the string that was checked for matches; the starting and ending
8708 positions of the regexp match; and, if the regexp included any
8709 parenthesized subexpressions, the starting and ending positions of each
8712 In each of the regexp match functions described below, the `match'
8713 argument must be a match structure returned by a previous call to
8714 `string-match' or `regexp-exec'. Most of these functions return some
8715 information about the original target string that was matched against a
8716 regular expression; we will call that string TARGET for easy reference.
8718 **** Function: regexp-match? OBJ
8719 Return `#t' if OBJ is a match structure returned by a previous
8720 call to `regexp-exec', or `#f' otherwise.
8722 **** Function: match:substring MATCH [N]
8723 Return the portion of TARGET matched by subexpression number N.
8724 Submatch 0 (the default) represents the entire regexp match. If
8725 the regular expression as a whole matched, but the subexpression
8726 number N did not match, return `#f'.
8728 **** Function: match:start MATCH [N]
8729 Return the starting position of submatch number N.
8731 **** Function: match:end MATCH [N]
8732 Return the ending position of submatch number N.
8734 **** Function: match:prefix MATCH
8735 Return the unmatched portion of TARGET preceding the regexp match.
8737 **** Function: match:suffix MATCH
8738 Return the unmatched portion of TARGET following the regexp match.
8740 **** Function: match:count MATCH
8741 Return the number of parenthesized subexpressions from MATCH.
8742 Note that the entire regular expression match itself counts as a
8743 subexpression, and failed submatches are included in the count.
8745 **** Function: match:string MATCH
8746 Return the original TARGET string.
8748 *** Backslash Escapes
8750 Sometimes you will want a regexp to match characters like `*' or `$'
8751 exactly. For example, to check whether a particular string represents
8752 a menu entry from an Info node, it would be useful to match it against
8753 a regexp like `^* [^:]*::'. However, this won't work; because the
8754 asterisk is a metacharacter, it won't match the `*' at the beginning of
8755 the string. In this case, we want to make the first asterisk un-magic.
8757 You can do this by preceding the metacharacter with a backslash
8758 character `\'. (This is also called "quoting" the metacharacter, and
8759 is known as a "backslash escape".) When Guile sees a backslash in a
8760 regular expression, it considers the following glyph to be an ordinary
8761 character, no matter what special meaning it would ordinarily have.
8762 Therefore, we can make the above example work by changing the regexp to
8763 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
8764 to match only a single asterisk in the target string.
8766 Since the backslash is itself a metacharacter, you may force a
8767 regexp to match a backslash in the target string by preceding the
8768 backslash with itself. For example, to find variable references in a
8769 TeX program, you might want to find occurrences of the string `\let\'
8770 followed by any number of alphabetic characters. The regular expression
8771 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
8772 each match a single backslash in the target string.
8774 **** Function: regexp-quote STR
8775 Quote each special character found in STR with a backslash, and
8776 return the resulting string.
8778 *Very important:* Using backslash escapes in Guile source code (as
8779 in Emacs Lisp or C) can be tricky, because the backslash character has
8780 special meaning for the Guile reader. For example, if Guile encounters
8781 the character sequence `\n' in the middle of a string while processing
8782 Scheme code, it replaces those characters with a newline character.
8783 Similarly, the character sequence `\t' is replaced by a horizontal tab.
8784 Several of these "escape sequences" are processed by the Guile reader
8785 before your code is executed. Unrecognized escape sequences are
8786 ignored: if the characters `\*' appear in a string, they will be
8787 translated to the single character `*'.
8789 This translation is obviously undesirable for regular expressions,
8790 since we want to be able to include backslashes in a string in order to
8791 escape regexp metacharacters. Therefore, to make sure that a backslash
8792 is preserved in a string in your Guile program, you must use *two*
8793 consecutive backslashes:
8795 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
8797 The string in this example is preprocessed by the Guile reader before
8798 any code is executed. The resulting argument to `make-regexp' is the
8799 string `^\* [^:]*', which is what we really want.
8801 This also means that in order to write a regular expression that
8802 matches a single backslash character, the regular expression string in
8803 the source code must include *four* backslashes. Each consecutive pair
8804 of backslashes gets translated by the Guile reader to a single
8805 backslash, and the resulting double-backslash is interpreted by the
8806 regexp engine as matching a single backslash character. Hence:
8808 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
8810 The reason for the unwieldiness of this syntax is historical. Both
8811 regular expression pattern matchers and Unix string processing systems
8812 have traditionally used backslashes with the special meanings described
8813 above. The POSIX regular expression specification and ANSI C standard
8814 both require these semantics. Attempting to abandon either convention
8815 would cause other kinds of compatibility problems, possibly more severe
8816 ones. Therefore, without extending the Scheme reader to support
8817 strings with different quoting conventions (an ungainly and confusing
8818 extension when implemented in other languages), we must adhere to this
8819 cumbersome escape syntax.
8821 * Changes to the gh_ interface
8823 * Changes to the scm_ interface
8825 * Changes to system call interfaces:
8827 ** The value returned by `raise' is now unspecified. It throws an exception
8830 *** A new procedure `sigaction' can be used to install signal handlers
8832 (sigaction signum [action] [flags])
8834 signum is the signal number, which can be specified using the value
8837 If action is omitted, sigaction returns a pair: the CAR is the current
8838 signal hander, which will be either an integer with the value SIG_DFL
8839 (default action) or SIG_IGN (ignore), or the Scheme procedure which
8840 handles the signal, or #f if a non-Scheme procedure handles the
8841 signal. The CDR contains the current sigaction flags for the handler.
8843 If action is provided, it is installed as the new handler for signum.
8844 action can be a Scheme procedure taking one argument, or the value of
8845 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
8846 whatever signal handler was installed before sigaction was first used.
8847 Flags can optionally be specified for the new handler (SA_RESTART is
8848 always used if the system provides it, so need not be specified.) The
8849 return value is a pair with information about the old handler as
8852 This interface does not provide access to the "signal blocking"
8853 facility. Maybe this is not needed, since the thread support may
8854 provide solutions to the problem of consistent access to data
8857 *** A new procedure `flush-all-ports' is equivalent to running
8858 `force-output' on every port open for output.
8860 ** Guile now provides information on how it was built, via the new
8861 global variable, %guile-build-info. This variable records the values
8862 of the standard GNU makefile directory variables as an assocation
8863 list, mapping variable names (symbols) onto directory paths (strings).
8864 For example, to find out where the Guile link libraries were
8865 installed, you can say:
8867 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
8870 * Changes to the scm_ interface
8872 ** The new function scm_handle_by_message_noexit is just like the
8873 existing scm_handle_by_message function, except that it doesn't call
8874 exit to terminate the process. Instead, it prints a message and just
8875 returns #f. This might be a more appropriate catch-all handler for
8876 new dynamic roots and threads.
8879 Changes in Guile 1.1 (released Friday, May 16 1997):
8881 * Changes to the distribution.
8883 The Guile 1.0 distribution has been split up into several smaller
8885 guile-core --- the Guile interpreter itself.
8886 guile-tcltk --- the interface between the Guile interpreter and
8887 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
8888 is a toolkit for building graphical user interfaces.
8889 guile-rgx-ctax --- the interface between Guile and the Rx regular
8890 expression matcher, and the translator for the Ctax
8891 programming language. These are packaged together because the
8892 Ctax translator uses Rx to parse Ctax source code.
8894 This NEWS file describes the changes made to guile-core since the 1.0
8897 We no longer distribute the documentation, since it was either out of
8898 date, or incomplete. As soon as we have current documentation, we
8903 * Changes to the stand-alone interpreter
8905 ** guile now accepts command-line arguments compatible with SCSH, Olin
8906 Shivers' Scheme Shell.
8908 In general, arguments are evaluated from left to right, but there are
8909 exceptions. The following switches stop argument processing, and
8910 stash all remaining command-line arguments as the value returned by
8911 the (command-line) function.
8912 -s SCRIPT load Scheme source code from FILE, and exit
8913 -c EXPR evalute Scheme expression EXPR, and exit
8914 -- stop scanning arguments; run interactively
8916 The switches below are processed as they are encountered.
8917 -l FILE load Scheme source code from FILE
8918 -e FUNCTION after reading script, apply FUNCTION to
8919 command line arguments
8920 -ds do -s script at this point
8921 --emacs enable Emacs protocol (experimental)
8922 -h, --help display this help and exit
8923 -v, --version display version information and exit
8924 \ read arguments from following script lines
8926 So, for example, here is a Guile script named `ekko' (thanks, Olin)
8927 which re-implements the traditional "echo" command:
8929 #!/usr/local/bin/guile -s
8932 (map (lambda (arg) (display arg) (display " "))
8936 (main (command-line))
8938 Suppose we invoke this script as follows:
8940 ekko a speckled gecko
8942 Through the magic of Unix script processing (triggered by the `#!'
8943 token at the top of the file), /usr/local/bin/guile receives the
8944 following list of command-line arguments:
8946 ("-s" "./ekko" "a" "speckled" "gecko")
8948 Unix inserts the name of the script after the argument specified on
8949 the first line of the file (in this case, "-s"), and then follows that
8950 with the arguments given to the script. Guile loads the script, which
8951 defines the `main' function, and then applies it to the list of
8952 remaining command-line arguments, ("a" "speckled" "gecko").
8954 In Unix, the first line of a script file must take the following form:
8956 #!INTERPRETER ARGUMENT
8958 where INTERPRETER is the absolute filename of the interpreter
8959 executable, and ARGUMENT is a single command-line argument to pass to
8962 You may only pass one argument to the interpreter, and its length is
8963 limited. These restrictions can be annoying to work around, so Guile
8964 provides a general mechanism (borrowed from, and compatible with,
8965 SCSH) for circumventing them.
8967 If the ARGUMENT in a Guile script is a single backslash character,
8968 `\', Guile will open the script file, parse arguments from its second
8969 and subsequent lines, and replace the `\' with them. So, for example,
8970 here is another implementation of the `ekko' script:
8972 #!/usr/local/bin/guile \
8976 (for-each (lambda (arg) (display arg) (display " "))
8980 If the user invokes this script as follows:
8982 ekko a speckled gecko
8984 Unix expands this into
8986 /usr/local/bin/guile \ ekko a speckled gecko
8988 When Guile sees the `\' argument, it replaces it with the arguments
8989 read from the second line of the script, producing:
8991 /usr/local/bin/guile -e main -s ekko a speckled gecko
8993 This tells Guile to load the `ekko' script, and apply the function
8994 `main' to the argument list ("a" "speckled" "gecko").
8996 Here is how Guile parses the command-line arguments:
8997 - Each space character terminates an argument. This means that two
8998 spaces in a row introduce an empty-string argument.
8999 - The tab character is not permitted (unless you quote it with the
9000 backslash character, as described below), to avoid confusion.
9001 - The newline character terminates the sequence of arguments, and will
9002 also terminate a final non-empty argument. (However, a newline
9003 following a space will not introduce a final empty-string argument;
9004 it only terminates the argument list.)
9005 - The backslash character is the escape character. It escapes
9006 backslash, space, tab, and newline. The ANSI C escape sequences
9007 like \n and \t are also supported. These produce argument
9008 constituents; the two-character combination \n doesn't act like a
9009 terminating newline. The escape sequence \NNN for exactly three
9010 octal digits reads as the character whose ASCII code is NNN. As
9011 above, characters produced this way are argument constituents.
9012 Backslash followed by other characters is not allowed.
9014 * Changes to the procedure for linking libguile with your programs
9016 ** Guile now builds and installs a shared guile library, if your
9017 system support shared libraries. (It still builds a static library on
9018 all systems.) Guile automatically detects whether your system
9019 supports shared libraries. To prevent Guile from buildisg shared
9020 libraries, pass the `--disable-shared' flag to the configure script.
9022 Guile takes longer to compile when it builds shared libraries, because
9023 it must compile every file twice --- once to produce position-
9024 independent object code, and once to produce normal object code.
9026 ** The libthreads library has been merged into libguile.
9028 To link a program against Guile, you now need only link against
9029 -lguile and -lqt; -lthreads is no longer needed. If you are using
9030 autoconf to generate configuration scripts for your application, the
9031 following lines should suffice to add the appropriate libraries to
9034 ### Find quickthreads and libguile.
9035 AC_CHECK_LIB(qt, main)
9036 AC_CHECK_LIB(guile, scm_shell)
9038 * Changes to Scheme functions
9040 ** Guile Scheme's special syntax for keyword objects is now optional,
9041 and disabled by default.
9043 The syntax variation from R4RS made it difficult to port some
9044 interesting packages to Guile. The routines which accepted keyword
9045 arguments (mostly in the module system) have been modified to also
9046 accept symbols whose names begin with `:'.
9048 To change the keyword syntax, you must first import the (ice-9 debug)
9050 (use-modules (ice-9 debug))
9052 Then you can enable the keyword syntax as follows:
9053 (read-set! keywords 'prefix)
9055 To disable keyword syntax, do this:
9056 (read-set! keywords #f)
9058 ** Many more primitive functions accept shared substrings as
9059 arguments. In the past, these functions required normal, mutable
9060 strings as arguments, although they never made use of this
9063 ** The uniform array functions now operate on byte vectors. These
9064 functions are `array-fill!', `serial-array-copy!', `array-copy!',
9065 `serial-array-map', `array-map', `array-for-each', and
9068 ** The new functions `trace' and `untrace' implement simple debugging
9069 support for Scheme functions.
9071 The `trace' function accepts any number of procedures as arguments,
9072 and tells the Guile interpreter to display each procedure's name and
9073 arguments each time the procedure is invoked. When invoked with no
9074 arguments, `trace' returns the list of procedures currently being
9077 The `untrace' function accepts any number of procedures as arguments,
9078 and tells the Guile interpreter not to trace them any more. When
9079 invoked with no arguments, `untrace' untraces all curretly traced
9082 The tracing in Guile has an advantage over most other systems: we
9083 don't create new procedure objects, but mark the procedure objects
9084 themselves. This means that anonymous and internal procedures can be
9087 ** The function `assert-repl-prompt' has been renamed to
9088 `set-repl-prompt!'. It takes one argument, PROMPT.
9089 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
9090 - If PROMPT is a string, we use it as a prompt.
9091 - If PROMPT is a procedure accepting no arguments, we call it, and
9092 display the result as a prompt.
9093 - Otherwise, we display "> ".
9095 ** The new function `eval-string' reads Scheme expressions from a
9096 string and evaluates them, returning the value of the last expression
9097 in the string. If the string contains no expressions, it returns an
9100 ** The new function `thunk?' returns true iff its argument is a
9101 procedure of zero arguments.
9103 ** `defined?' is now a builtin function, instead of syntax. This
9104 means that its argument should be quoted. It returns #t iff its
9105 argument is bound in the current module.
9107 ** The new syntax `use-modules' allows you to add new modules to your
9108 environment without re-typing a complete `define-module' form. It
9109 accepts any number of module names as arguments, and imports their
9110 public bindings into the current module.
9112 ** The new function (module-defined? NAME MODULE) returns true iff
9113 NAME, a symbol, is defined in MODULE, a module object.
9115 ** The new function `builtin-bindings' creates and returns a hash
9116 table containing copies of all the root module's bindings.
9118 ** The new function `builtin-weak-bindings' does the same as
9119 `builtin-bindings', but creates a doubly-weak hash table.
9121 ** The `equal?' function now considers variable objects to be
9122 equivalent if they have the same name and the same value.
9124 ** The new function `command-line' returns the command-line arguments
9125 given to Guile, as a list of strings.
9127 When using guile as a script interpreter, `command-line' returns the
9128 script's arguments; those processed by the interpreter (like `-s' or
9129 `-c') are omitted. (In other words, you get the normal, expected
9130 behavior.) Any application that uses scm_shell to process its
9131 command-line arguments gets this behavior as well.
9133 ** The new function `load-user-init' looks for a file called `.guile'
9134 in the user's home directory, and loads it if it exists. This is
9135 mostly for use by the code generated by scm_compile_shell_switches,
9136 but we thought it might also be useful in other circumstances.
9138 ** The new function `log10' returns the base-10 logarithm of its
9141 ** Changes to I/O functions
9143 *** The functions `read', `primitive-load', `read-and-eval!', and
9144 `primitive-load-path' no longer take optional arguments controlling
9145 case insensitivity and a `#' parser.
9147 Case sensitivity is now controlled by a read option called
9148 `case-insensitive'. The user can add new `#' syntaxes with the
9149 `read-hash-extend' function (see below).
9151 *** The new function `read-hash-extend' allows the user to change the
9152 syntax of Guile Scheme in a somewhat controlled way.
9154 (read-hash-extend CHAR PROC)
9155 When parsing S-expressions, if we read a `#' character followed by
9156 the character CHAR, use PROC to parse an object from the stream.
9157 If PROC is #f, remove any parsing procedure registered for CHAR.
9159 The reader applies PROC to two arguments: CHAR and an input port.
9161 *** The new functions read-delimited and read-delimited! provide a
9162 general mechanism for doing delimited input on streams.
9164 (read-delimited DELIMS [PORT HANDLE-DELIM])
9165 Read until we encounter one of the characters in DELIMS (a string),
9166 or end-of-file. PORT is the input port to read from; it defaults to
9167 the current input port. The HANDLE-DELIM parameter determines how
9168 the terminating character is handled; it should be one of the
9171 'trim omit delimiter from result
9172 'peek leave delimiter character in input stream
9173 'concat append delimiter character to returned value
9174 'split return a pair: (RESULT . TERMINATOR)
9176 HANDLE-DELIM defaults to 'peek.
9178 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
9179 A side-effecting variant of `read-delimited'.
9181 The data is written into the string BUF at the indices in the
9182 half-open interval [START, END); the default interval is the whole
9183 string: START = 0 and END = (string-length BUF). The values of
9184 START and END must specify a well-defined interval in BUF, i.e.
9185 0 <= START <= END <= (string-length BUF).
9187 It returns NBYTES, the number of bytes read. If the buffer filled
9188 up without a delimiter character being found, it returns #f. If the
9189 port is at EOF when the read starts, it returns the EOF object.
9191 If an integer is returned (i.e., the read is successfully terminated
9192 by reading a delimiter character), then the HANDLE-DELIM parameter
9193 determines how to handle the terminating character. It is described
9194 above, and defaults to 'peek.
9196 (The descriptions of these functions were borrowed from the SCSH
9197 manual, by Olin Shivers and Brian Carlstrom.)
9199 *** The `%read-delimited!' function is the primitive used to implement
9200 `read-delimited' and `read-delimited!'.
9202 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
9204 This returns a pair of values: (TERMINATOR . NUM-READ).
9205 - TERMINATOR describes why the read was terminated. If it is a
9206 character or the eof object, then that is the value that terminated
9207 the read. If it is #f, the function filled the buffer without finding
9208 a delimiting character.
9209 - NUM-READ is the number of characters read into BUF.
9211 If the read is successfully terminated by reading a delimiter
9212 character, then the gobble? parameter determines what to do with the
9213 terminating character. If true, the character is removed from the
9214 input stream; if false, the character is left in the input stream
9215 where a subsequent read operation will retrieve it. In either case,
9216 the character is also the first value returned by the procedure call.
9218 (The descriptions of this function was borrowed from the SCSH manual,
9219 by Olin Shivers and Brian Carlstrom.)
9221 *** The `read-line' and `read-line!' functions have changed; they now
9222 trim the terminator by default; previously they appended it to the
9223 returned string. For the old behavior, use (read-line PORT 'concat).
9225 *** The functions `uniform-array-read!' and `uniform-array-write!' now
9226 take new optional START and END arguments, specifying the region of
9227 the array to read and write.
9229 *** The `ungetc-char-ready?' function has been removed. We feel it's
9230 inappropriate for an interface to expose implementation details this
9233 ** Changes to the Unix library and system call interface
9235 *** The new fcntl function provides access to the Unix `fcntl' system
9238 (fcntl PORT COMMAND VALUE)
9239 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
9240 Values for COMMAND are:
9242 F_DUPFD duplicate a file descriptor
9243 F_GETFD read the descriptor's close-on-exec flag
9244 F_SETFD set the descriptor's close-on-exec flag to VALUE
9245 F_GETFL read the descriptor's flags, as set on open
9246 F_SETFL set the descriptor's flags, as set on open to VALUE
9247 F_GETOWN return the process ID of a socket's owner, for SIGIO
9248 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
9249 FD_CLOEXEC not sure what this is
9251 For details, see the documentation for the fcntl system call.
9253 *** The arguments to `select' have changed, for compatibility with
9254 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
9255 expected behavior. The MILLISECONDS parameter has been changed to
9256 MICROSECONDS, to more closely resemble the underlying system call.
9257 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
9258 corresponding return set will be the same.
9260 *** The arguments to the `mknod' system call have changed. They are
9263 (mknod PATH TYPE PERMS DEV)
9264 Create a new file (`node') in the file system. PATH is the name of
9265 the file to create. TYPE is the kind of file to create; it should
9266 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
9267 permission bits to give the newly created file. If TYPE is
9268 'block-special or 'char-special, DEV specifies which device the
9269 special file refers to; its interpretation depends on the kind of
9270 special file being created.
9272 *** The `fork' function has been renamed to `primitive-fork', to avoid
9273 clashing with various SCSH forks.
9275 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
9276 and `recvfrom!'. They no longer accept a size for a second argument;
9277 you must pass a string to hold the received value. They no longer
9278 return the buffer. Instead, `recv' returns the length of the message
9279 received, and `recvfrom' returns a pair containing the packet's length
9280 and originating address.
9282 *** The file descriptor datatype has been removed, as have the
9283 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
9284 We plan to replace these functions with a SCSH-compatible interface.
9286 *** The `create' function has been removed; it's just a special case
9289 *** There are new functions to break down process termination status
9290 values. In the descriptions below, STATUS is a value returned by
9293 (status:exit-val STATUS)
9294 If the child process exited normally, this function returns the exit
9295 code for the child process (i.e., the value passed to exit, or
9296 returned from main). If the child process did not exit normally,
9297 this function returns #f.
9299 (status:stop-sig STATUS)
9300 If the child process was suspended by a signal, this function
9301 returns the signal that suspended the child. Otherwise, it returns
9304 (status:term-sig STATUS)
9305 If the child process terminated abnormally, this function returns
9306 the signal that terminated the child. Otherwise, this function
9309 POSIX promises that exactly one of these functions will return true on
9310 a valid STATUS value.
9312 These functions are compatible with SCSH.
9314 *** There are new accessors and setters for the broken-out time vectors
9315 returned by `localtime', `gmtime', and that ilk. They are:
9317 Component Accessor Setter
9318 ========================= ============ ============
9319 seconds tm:sec set-tm:sec
9320 minutes tm:min set-tm:min
9321 hours tm:hour set-tm:hour
9322 day of the month tm:mday set-tm:mday
9323 month tm:mon set-tm:mon
9324 year tm:year set-tm:year
9325 day of the week tm:wday set-tm:wday
9326 day in the year tm:yday set-tm:yday
9327 daylight saving time tm:isdst set-tm:isdst
9328 GMT offset, seconds tm:gmtoff set-tm:gmtoff
9329 name of time zone tm:zone set-tm:zone
9331 *** There are new accessors for the vectors returned by `uname',
9332 describing the host system:
9335 ============================================== ================
9336 name of the operating system implementation utsname:sysname
9337 network name of this machine utsname:nodename
9338 release level of the operating system utsname:release
9339 version level of the operating system utsname:version
9340 machine hardware platform utsname:machine
9342 *** There are new accessors for the vectors returned by `getpw',
9343 `getpwnam', `getpwuid', and `getpwent', describing entries from the
9344 system's user database:
9347 ====================== =================
9348 user name passwd:name
9349 user password passwd:passwd
9352 real name passwd:gecos
9353 home directory passwd:dir
9354 shell program passwd:shell
9356 *** There are new accessors for the vectors returned by `getgr',
9357 `getgrnam', `getgrgid', and `getgrent', describing entries from the
9358 system's group database:
9361 ======================= ============
9362 group name group:name
9363 group password group:passwd
9365 group members group:mem
9367 *** There are new accessors for the vectors returned by `gethost',
9368 `gethostbyaddr', `gethostbyname', and `gethostent', describing
9372 ========================= ===============
9373 official name of host hostent:name
9374 alias list hostent:aliases
9375 host address type hostent:addrtype
9376 length of address hostent:length
9377 list of addresses hostent:addr-list
9379 *** There are new accessors for the vectors returned by `getnet',
9380 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
9384 ========================= ===============
9385 official name of net netent:name
9386 alias list netent:aliases
9387 net number type netent:addrtype
9388 net number netent:net
9390 *** There are new accessors for the vectors returned by `getproto',
9391 `getprotobyname', `getprotobynumber', and `getprotoent', describing
9395 ========================= ===============
9396 official protocol name protoent:name
9397 alias list protoent:aliases
9398 protocol number protoent:proto
9400 *** There are new accessors for the vectors returned by `getserv',
9401 `getservbyname', `getservbyport', and `getservent', describing
9405 ========================= ===============
9406 official service name servent:name
9407 alias list servent:aliases
9408 port number servent:port
9409 protocol to use servent:proto
9411 *** There are new accessors for the sockaddr structures returned by
9412 `accept', `getsockname', `getpeername', `recvfrom!':
9415 ======================================== ===============
9416 address format (`family') sockaddr:fam
9417 path, for file domain addresses sockaddr:path
9418 address, for internet domain addresses sockaddr:addr
9419 TCP or UDP port, for internet sockaddr:port
9421 *** The `getpwent', `getgrent', `gethostent', `getnetent',
9422 `getprotoent', and `getservent' functions now return #f at the end of
9423 the user database. (They used to throw an exception.)
9425 Note that calling MUMBLEent function is equivalent to calling the
9426 corresponding MUMBLE function with no arguments.
9428 *** The `setpwent', `setgrent', `sethostent', `setnetent',
9429 `setprotoent', and `setservent' routines now take no arguments.
9431 *** The `gethost', `getproto', `getnet', and `getserv' functions now
9432 provide more useful information when they throw an exception.
9434 *** The `lnaof' function has been renamed to `inet-lnaof'.
9436 *** Guile now claims to have the `current-time' feature.
9438 *** The `mktime' function now takes an optional second argument ZONE,
9439 giving the time zone to use for the conversion. ZONE should be a
9440 string, in the same format as expected for the "TZ" environment variable.
9442 *** The `strptime' function now returns a pair (TIME . COUNT), where
9443 TIME is the parsed time as a vector, and COUNT is the number of
9444 characters from the string left unparsed. This function used to
9445 return the remaining characters as a string.
9447 *** The `gettimeofday' function has replaced the old `time+ticks' function.
9448 The return value is now (SECONDS . MICROSECONDS); the fractional
9449 component is no longer expressed in "ticks".
9451 *** The `ticks/sec' constant has been removed, in light of the above change.
9453 * Changes to the gh_ interface
9455 ** gh_eval_str() now returns an SCM object which is the result of the
9458 ** gh_scm2str() now copies the Scheme data to a caller-provided C
9461 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
9462 and returns the array
9464 ** gh_scm2str0() is gone: there is no need to distinguish
9465 null-terminated from non-null-terminated, since gh_scm2newstr() allows
9466 the user to interpret the data both ways.
9468 * Changes to the scm_ interface
9470 ** The new function scm_symbol_value0 provides an easy way to get a
9471 symbol's value from C code:
9473 SCM scm_symbol_value0 (char *NAME)
9474 Return the value of the symbol named by the null-terminated string
9475 NAME in the current module. If the symbol named NAME is unbound in
9476 the current module, return SCM_UNDEFINED.
9478 ** The new function scm_sysintern0 creates new top-level variables,
9479 without assigning them a value.
9481 SCM scm_sysintern0 (char *NAME)
9482 Create a new Scheme top-level variable named NAME. NAME is a
9483 null-terminated string. Return the variable's value cell.
9485 ** The function scm_internal_catch is the guts of catch. It handles
9486 all the mechanics of setting up a catch target, invoking the catch
9487 body, and perhaps invoking the handler if the body does a throw.
9489 The function is designed to be usable from C code, but is general
9490 enough to implement all the semantics Guile Scheme expects from throw.
9492 TAG is the catch tag. Typically, this is a symbol, but this function
9493 doesn't actually care about that.
9495 BODY is a pointer to a C function which runs the body of the catch;
9496 this is the code you can throw from. We call it like this:
9497 BODY (BODY_DATA, JMPBUF)
9499 BODY_DATA is just the BODY_DATA argument we received; we pass it
9500 through to BODY as its first argument. The caller can make
9501 BODY_DATA point to anything useful that BODY might need.
9502 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
9503 which we have just created and initialized.
9505 HANDLER is a pointer to a C function to deal with a throw to TAG,
9506 should one occur. We call it like this:
9507 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
9509 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
9510 same idea as BODY_DATA above.
9511 THROWN_TAG is the tag that the user threw to; usually this is
9512 TAG, but it could be something else if TAG was #t (i.e., a
9513 catch-all), or the user threw to a jmpbuf.
9514 THROW_ARGS is the list of arguments the user passed to the THROW
9517 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
9518 is just a pointer we pass through to HANDLER. We don't actually
9519 use either of those pointers otherwise ourselves. The idea is
9520 that, if our caller wants to communicate something to BODY or
9521 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
9522 HANDLER can then use. Think of it as a way to make BODY and
9523 HANDLER closures, not just functions; MUMBLE_DATA points to the
9526 Of course, it's up to the caller to make sure that any data a
9527 MUMBLE_DATA needs is protected from GC. A common way to do this is
9528 to make MUMBLE_DATA a pointer to data stored in an automatic
9529 structure variable; since the collector must scan the stack for
9530 references anyway, this assures that any references in MUMBLE_DATA
9533 ** The new function scm_internal_lazy_catch is exactly like
9534 scm_internal_catch, except:
9536 - It does not unwind the stack (this is the major difference).
9537 - If handler returns, its value is returned from the throw.
9538 - BODY always receives #f as its JMPBUF argument (since there's no
9539 jmpbuf associated with a lazy catch, because we don't unwind the
9542 ** scm_body_thunk is a new body function you can pass to
9543 scm_internal_catch if you want the body to be like Scheme's `catch'
9544 --- a thunk, or a function of one argument if the tag is #f.
9546 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
9547 contains the Scheme procedure to invoke as the body, and the tag
9548 we're catching. If the tag is #f, then we pass JMPBUF (created by
9549 scm_internal_catch) to the body procedure; otherwise, the body gets
9552 ** scm_handle_by_proc is a new handler function you can pass to
9553 scm_internal_catch if you want the handler to act like Scheme's catch
9554 --- call a procedure with the tag and the throw arguments.
9556 If the user does a throw to this catch, this function runs a handler
9557 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
9558 variable holding the Scheme procedure object to invoke. It ought to
9559 be a pointer to an automatic variable (i.e., one living on the stack),
9560 or the procedure object should be otherwise protected from GC.
9562 ** scm_handle_by_message is a new handler function to use with
9563 `scm_internal_catch' if you want Guile to print a message and die.
9564 It's useful for dealing with throws to uncaught keys at the top level.
9566 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
9567 message header to print; if zero, we use "guile" instead. That
9568 text is followed by a colon, then the message described by ARGS.
9570 ** The return type of scm_boot_guile is now void; the function does
9571 not return a value, and indeed, never returns at all.
9573 ** The new function scm_shell makes it easy for user applications to
9574 process command-line arguments in a way that is compatible with the
9575 stand-alone guile interpreter (which is in turn compatible with SCSH,
9578 To use the scm_shell function, first initialize any guile modules
9579 linked into your application, and then call scm_shell with the values
9580 of ARGC and ARGV your `main' function received. scm_shell will add
9581 any SCSH-style meta-arguments from the top of the script file to the
9582 argument vector, and then process the command-line arguments. This
9583 generally means loading a script file or starting up an interactive
9584 command interpreter. For details, see "Changes to the stand-alone
9587 ** The new functions scm_get_meta_args and scm_count_argv help you
9588 implement the SCSH-style meta-argument, `\'.
9590 char **scm_get_meta_args (int ARGC, char **ARGV)
9591 If the second element of ARGV is a string consisting of a single
9592 backslash character (i.e. "\\" in Scheme notation), open the file
9593 named by the following argument, parse arguments from it, and return
9594 the spliced command line. The returned array is terminated by a
9597 For details of argument parsing, see above, under "guile now accepts
9598 command-line arguments compatible with SCSH..."
9600 int scm_count_argv (char **ARGV)
9601 Count the arguments in ARGV, assuming it is terminated by a null
9604 For an example of how these functions might be used, see the source
9605 code for the function scm_shell in libguile/script.c.
9607 You will usually want to use scm_shell instead of calling this
9610 ** The new function scm_compile_shell_switches turns an array of
9611 command-line arguments into Scheme code to carry out the actions they
9612 describe. Given ARGC and ARGV, it returns a Scheme expression to
9613 evaluate, and calls scm_set_program_arguments to make any remaining
9614 command-line arguments available to the Scheme code. For example,
9615 given the following arguments:
9617 -e main -s ekko a speckled gecko
9619 scm_set_program_arguments will return the following expression:
9621 (begin (load "ekko") (main (command-line)) (quit))
9623 You will usually want to use scm_shell instead of calling this
9626 ** The function scm_shell_usage prints a usage message appropriate for
9627 an interpreter that uses scm_compile_shell_switches to handle its
9628 command-line arguments.
9630 void scm_shell_usage (int FATAL, char *MESSAGE)
9631 Print a usage message to the standard error output. If MESSAGE is
9632 non-zero, write it before the usage message, followed by a newline.
9633 If FATAL is non-zero, exit the process, using FATAL as the
9634 termination status. (If you want to be compatible with Guile,
9635 always use 1 as the exit status when terminating due to command-line
9638 You will usually want to use scm_shell instead of calling this
9641 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
9642 expressions. It used to return SCM_EOL. Earth-shattering.
9644 ** The macros for declaring scheme objects in C code have been
9645 rearranged slightly. They are now:
9647 SCM_SYMBOL (C_NAME, SCHEME_NAME)
9648 Declare a static SCM variable named C_NAME, and initialize it to
9649 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
9650 be a C identifier, and SCHEME_NAME should be a C string.
9652 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
9653 Just like SCM_SYMBOL, but make C_NAME globally visible.
9655 SCM_VCELL (C_NAME, SCHEME_NAME)
9656 Create a global variable at the Scheme level named SCHEME_NAME.
9657 Declare a static SCM variable named C_NAME, and initialize it to
9658 point to the Scheme variable's value cell.
9660 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
9661 Just like SCM_VCELL, but make C_NAME globally visible.
9663 The `guile-snarf' script writes initialization code for these macros
9664 to its standard output, given C source code as input.
9666 The SCM_GLOBAL macro is gone.
9668 ** The scm_read_line and scm_read_line_x functions have been replaced
9669 by Scheme code based on the %read-delimited! procedure (known to C
9670 code as scm_read_delimited_x). See its description above for more
9673 ** The function scm_sys_open has been renamed to scm_open. It now
9674 returns a port instead of an FD object.
9676 * The dynamic linking support has changed. For more information, see
9677 libguile/DYNAMIC-LINKING.
9682 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
9685 * Changes to the 'guile' program:
9687 ** Guile now loads some new files when it starts up. Guile first
9688 searches the load path for init.scm, and loads it if found. Then, if
9689 Guile is not being used to execute a script, and the user's home
9690 directory contains a file named `.guile', Guile loads that.
9692 ** You can now use Guile as a shell script interpreter.
9694 To paraphrase the SCSH manual:
9696 When Unix tries to execute an executable file whose first two
9697 characters are the `#!', it treats the file not as machine code to
9698 be directly executed by the native processor, but as source code
9699 to be executed by some interpreter. The interpreter to use is
9700 specified immediately after the #! sequence on the first line of
9701 the source file. The kernel reads in the name of the interpreter,
9702 and executes that instead. It passes the interpreter the source
9703 filename as its first argument, with the original arguments
9704 following. Consult the Unix man page for the `exec' system call
9705 for more information.
9707 Now you can use Guile as an interpreter, using a mechanism which is a
9708 compatible subset of that provided by SCSH.
9710 Guile now recognizes a '-s' command line switch, whose argument is the
9711 name of a file of Scheme code to load. It also treats the two
9712 characters `#!' as the start of a comment, terminated by `!#'. Thus,
9713 to make a file of Scheme code directly executable by Unix, insert the
9714 following two lines at the top of the file:
9716 #!/usr/local/bin/guile -s
9719 Guile treats the argument of the `-s' command-line switch as the name
9720 of a file of Scheme code to load, and treats the sequence `#!' as the
9721 start of a block comment, terminated by `!#'.
9723 For example, here's a version of 'echo' written in Scheme:
9725 #!/usr/local/bin/guile -s
9727 (let loop ((args (cdr (program-arguments))))
9730 (display (car args))
9731 (if (pair? (cdr args))
9733 (loop (cdr args)))))
9736 Why does `#!' start a block comment terminated by `!#', instead of the
9737 end of the line? That is the notation SCSH uses, and although we
9738 don't yet support the other SCSH features that motivate that choice,
9739 we would like to be backward-compatible with any existing Guile
9740 scripts once we do. Furthermore, if the path to Guile on your system
9741 is too long for your kernel, you can start the script with this
9745 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
9748 Note that some very old Unix systems don't support the `#!' syntax.
9751 ** You can now run Guile without installing it.
9753 Previous versions of the interactive Guile interpreter (`guile')
9754 couldn't start up unless Guile's Scheme library had been installed;
9755 they used the value of the environment variable `SCHEME_LOAD_PATH'
9756 later on in the startup process, but not to find the startup code
9757 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
9760 To run Guile without installing it, build it in the normal way, and
9761 then set the environment variable `SCHEME_LOAD_PATH' to a
9762 colon-separated list of directories, including the top-level directory
9763 of the Guile sources. For example, if you unpacked Guile so that the
9764 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
9767 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
9770 ** Guile's read-eval-print loop no longer prints #<unspecified>
9771 results. If the user wants to see this, she can evaluate the
9772 expression (assert-repl-print-unspecified #t), perhaps in her startup
9775 ** Guile no longer shows backtraces by default when an error occurs;
9776 however, it does display a message saying how to get one, and how to
9777 request that they be displayed by default. After an error, evaluate
9779 to see a backtrace, and
9780 (debug-enable 'backtrace)
9781 to see them by default.
9785 * Changes to Guile Scheme:
9787 ** Guile now distinguishes between #f and the empty list.
9789 This is for compatibility with the IEEE standard, the (possibly)
9790 upcoming Revised^5 Report on Scheme, and many extant Scheme
9793 Guile used to have #f and '() denote the same object, to make Scheme's
9794 type system more compatible with Emacs Lisp's. However, the change
9795 caused too much trouble for Scheme programmers, and we found another
9796 way to reconcile Emacs Lisp with Scheme that didn't require this.
9799 ** Guile's delq, delv, delete functions, and their destructive
9800 counterparts, delq!, delv!, and delete!, now remove all matching
9801 elements from the list, not just the first. This matches the behavior
9802 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
9803 functions which inspired them.
9805 I recognize that this change may break code in subtle ways, but it
9806 seems best to make the change before the FSF's first Guile release,
9810 ** The compiled-library-path function has been deleted from libguile.
9812 ** The facilities for loading Scheme source files have changed.
9814 *** The variable %load-path now tells Guile which directories to search
9815 for Scheme code. Its value is a list of strings, each of which names
9818 *** The variable %load-extensions now tells Guile which extensions to
9819 try appending to a filename when searching the load path. Its value
9820 is a list of strings. Its default value is ("" ".scm").
9822 *** (%search-load-path FILENAME) searches the directories listed in the
9823 value of the %load-path variable for a Scheme file named FILENAME,
9824 with all the extensions listed in %load-extensions. If it finds a
9825 match, then it returns its full filename. If FILENAME is absolute, it
9826 returns it unchanged. Otherwise, it returns #f.
9828 %search-load-path will not return matches that refer to directories.
9830 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
9831 uses %seach-load-path to find a file named FILENAME, and loads it if
9832 it finds it. If it can't read FILENAME for any reason, it throws an
9835 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
9838 *** load uses the same searching semantics as primitive-load.
9840 *** The functions %try-load, try-load-with-path, %load, load-with-path,
9841 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
9842 path, and load-module-with-path have been deleted. The functions
9843 above should serve their purposes.
9845 *** If the value of the variable %load-hook is a procedure,
9846 `primitive-load' applies its value to the name of the file being
9847 loaded (without the load path directory name prepended). If its value
9848 is #f, it is ignored. Otherwise, an error occurs.
9850 This is mostly useful for printing load notification messages.
9853 ** The function `eval!' is no longer accessible from the scheme level.
9854 We can't allow operations which introduce glocs into the scheme level,
9855 because Guile's type system can't handle these as data. Use `eval' or
9856 `read-and-eval!' (see below) as replacement.
9858 ** The new function read-and-eval! reads an expression from PORT,
9859 evaluates it, and returns the result. This is more efficient than
9860 simply calling `read' and `eval', since it is not necessary to make a
9861 copy of the expression for the evaluator to munge.
9863 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
9864 for the `read' function.
9867 ** The function `int?' has been removed; its definition was identical
9868 to that of `integer?'.
9870 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
9871 use the R4RS names for these functions.
9873 ** The function object-properties no longer returns the hash handle;
9874 it simply returns the object's property list.
9876 ** Many functions have been changed to throw errors, instead of
9877 returning #f on failure. The point of providing exception handling in
9878 the language is to simplify the logic of user code, but this is less
9879 useful if Guile's primitives don't throw exceptions.
9881 ** The function `fileno' has been renamed from `%fileno'.
9883 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
9886 * Changes to Guile's C interface:
9888 ** The library's initialization procedure has been simplified.
9889 scm_boot_guile now has the prototype:
9891 void scm_boot_guile (int ARGC,
9893 void (*main_func) (),
9896 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
9897 MAIN_FUNC should do all the work of the program (initializing other
9898 packages, reading user input, etc.) before returning. When MAIN_FUNC
9899 returns, call exit (0); this function never returns. If you want some
9900 other exit value, MAIN_FUNC may call exit itself.
9902 scm_boot_guile arranges for program-arguments to return the strings
9903 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
9904 scm_set_program_arguments with the final list, so Scheme code will
9905 know which arguments have been processed.
9907 scm_boot_guile establishes a catch-all catch handler which prints an
9908 error message and exits the process. This means that Guile exits in a
9909 coherent way when system errors occur and the user isn't prepared to
9910 handle it. If the user doesn't like this behavior, they can establish
9911 their own universal catcher in MAIN_FUNC to shadow this one.
9913 Why must the caller do all the real work from MAIN_FUNC? The garbage
9914 collector assumes that all local variables of type SCM will be above
9915 scm_boot_guile's stack frame on the stack. If you try to manipulate
9916 SCM values after this function returns, it's the luck of the draw
9917 whether the GC will be able to find the objects you allocate. So,
9918 scm_boot_guile function exits, rather than returning, to discourage
9919 people from making that mistake.
9921 The IN, OUT, and ERR arguments were removed; there are other
9922 convenient ways to override these when desired.
9924 The RESULT argument was deleted; this function should never return.
9926 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
9930 ** Guile's header files should no longer conflict with your system's
9933 In order to compile code which #included <libguile.h>, previous
9934 versions of Guile required you to add a directory containing all the
9935 Guile header files to your #include path. This was a problem, since
9936 Guile's header files have names which conflict with many systems'
9939 Now only <libguile.h> need appear in your #include path; you must
9940 refer to all Guile's other header files as <libguile/mumble.h>.
9941 Guile's installation procedure puts libguile.h in $(includedir), and
9942 the rest in $(includedir)/libguile.
9945 ** Two new C functions, scm_protect_object and scm_unprotect_object,
9946 have been added to the Guile library.
9948 scm_protect_object (OBJ) protects OBJ from the garbage collector.
9949 OBJ will not be freed, even if all other references are dropped,
9950 until someone does scm_unprotect_object (OBJ). Both functions
9953 Note that calls to scm_protect_object do not nest. You can call
9954 scm_protect_object any number of times on a given object, and the
9955 next call to scm_unprotect_object will unprotect it completely.
9957 Basically, scm_protect_object and scm_unprotect_object just
9958 maintain a list of references to things. Since the GC knows about
9959 this list, all objects it mentions stay alive. scm_protect_object
9960 adds its argument to the list; scm_unprotect_object remove its
9961 argument from the list.
9964 ** scm_eval_0str now returns the value of the last expression
9967 ** The new function scm_read_0str reads an s-expression from a
9968 null-terminated string, and returns it.
9970 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
9971 to a Scheme port object.
9973 ** The new function `scm_set_program_arguments' allows C code to set
9974 the value returned by the Scheme `program-arguments' function.
9979 * Guile no longer includes sophisticated Tcl/Tk support.
9981 The old Tcl/Tk support was unsatisfying to us, because it required the
9982 user to link against the Tcl library, as well as Tk and Guile. The
9983 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
9984 referring to widgets by names, rather than exporting widgets to Scheme
9985 code as a special datatype.
9987 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
9988 maintainers described some very interesting changes in progress to the
9989 Tcl/Tk internals, which would facilitate clean interfaces between lone
9990 Tk and other interpreters --- even for garbage-collected languages
9991 like Scheme. They expected the new Tk to be publicly available in the
9994 Since it seems that Guile might soon have a new, cleaner interface to
9995 lone Tk, and that the old Guile/Tk glue code would probably need to be
9996 completely rewritten, we (Jim Blandy and Richard Stallman) have
9997 decided not to support the old code. We'll spend the time instead on
9998 a good interface to the newer Tk, as soon as it is available.
10000 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
10003 Copyright information:
10005 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
10007 Permission is granted to anyone to make or distribute verbatim copies
10008 of this document as received, in any medium, provided that the
10009 copyright notice and this permission notice are preserved,
10010 thus giving the recipient permission to redistribute in turn.
10012 Permission is granted to distribute modified versions
10013 of this document, or of portions of it,
10014 under the above conditions, provided also that they
10015 carry prominent notices stating who last changed them.
10020 paragraph-separate: "[
\f]*$"