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 ** Fluids can now have default values.
30 Fluids are used for dynamic and thread-local binding. They have always
31 inherited their values from the context or thread that created them.
32 However, there was a case in which a new thread would enter Guile, and
33 the default values of all the fluids would be `#f' for that thread.
35 This has now been fixed so that `make-fluid' has an optional default
36 value for fluids in unrelated dynamic roots, which defaults to `#f'.
38 ** Garbage collector tuning.
40 The garbage collector has now been tuned to run more often under some
43 *** Unmanaged allocation
45 The new `scm_gc_register_allocation' function will notify the collector
46 of unmanaged allocation. This will cause the collector to run sooner.
47 Guile's `scm_malloc', `scm_calloc', and `scm_realloc' unmanaged
48 allocators eventually call this function. This leads to better
49 performance under steady-state unmanaged allocation.
51 *** Transient allocation
53 When the collector runs, it will try to record the total memory
54 footprint of a process, if the platform supports this information. If
55 the memory footprint is growing, the collector will run more frequently.
56 This reduces the increase of the resident size of a process in response
57 to a transient increase in allocation.
59 *** Management of threads, bignums
61 Creating a thread will allocate a fair amount of memory. Guile now does
62 some GC work (using `GC_collect_a_little') when allocating a thread.
63 This leads to a better memory footprint when creating many short-lived
66 Similarly, bignums can occupy a lot of memory. Guile now offers hooks
67 to enable custom GMP allocators that end up calling
68 `scm_gc_register_allocation'. These allocators are enabled by default
69 when running Guile from the command-line. To enable them in libraries,
70 set the `scm_install_gmp_memory_functions' variable to a nonzero value
73 ** SRFI-39 parameters are available by default.
75 Guile now includes support for parameters, as defined by SRFI-39, in the
76 default environment. See "Parameters" in the manual, for more
77 information. `current-input-port', `current-output-port', and
78 `current-error-port' are now parameters.
80 ** Add `current-warning-port'
82 Guile now outputs warnings on a separate port, `current-warning-port',
83 initialized to the value that `current-error-port' has on startup.
87 Following Racket's lead, Guile now supports syntax parameters. See
88 "Syntax parameters" in the manual, for more.
90 Also see Barzilay, Culpepper, and Flatt's 2011 SFP workshop paper,
91 "Keeping it Clean with syntax-parameterize".
93 ** Parse command-line arguments from the locale encoding.
95 Guile now attempts to parse command-line arguments using the user's
96 locale. However for backwards compatibility with other 2.0.x releases,
97 it does so without actually calling `setlocale'. Please report any bugs
98 in this facility to bug-guile@gnu.org.
102 ** (ice-9 session): `apropos-hook'
103 ** New print option: `escape-newlines', defaults to #t.
104 ** (ice-9 ftw): `file-system-fold', `file-system-tree', `scandir'
108 ** Fix R6RS `fold-left' so the accumulator is the first argument.
109 ** fix <dynwind> serialization.
110 ** Fix bugs in the new `peval' optimizer.
111 ** Allow values bound in non-tail let expressions to be collected.
112 ** Fix bit-set*! bug from 2005.
113 ** Fix bug in `make-repl' when `lang' is actually a language.
114 ** Hack the port-column of current-output-port after printing a prompt.
115 ** FFI: Hold a weak reference to the CIF made by `procedure->pointer'.
116 ** FFI: Hold a weak reference to the procedure passed to `procedure->pointer'.
117 ** FFI: Properly unpack small integer return values in closure call.
118 ** Allow overlapping regions to be passed to `bytevector-copy!'.
119 ** Fix `validate-target' in (system base target).
120 ** `,language' at REPL sets the current-language fluid.
121 ** `primitive-load' returns the value(s) of the last expression.
122 ** Add an exception printer for `getaddrinfo-error'.
123 ** Add a deprecated alias for $expt.
124 ** Document invalidity of (begin) as expression; add back-compat shim.
125 ** Web: Allow URIs with empty authorities, like "file:///etc/hosts".
126 ** HTTP: Fix validators for various list-style headers.
127 ** HTTP: Extend handling of "Cache-Control" header.
128 ** HTTP: Fix qstring writing of cache-extension values
129 ** HTTP: `write-request-line' writes absolute paths, not absolute URIs.
130 ** HTTP: Permit non-date values for Expires header.
131 ** FreeBSD build fixes.
132 ** Fix generalized-vector-{ref,set!} for slices.
133 ** Fix erroneous check in `set-procedure-properties!'.
134 ** Don't leak file descriptors when mmaping objcode.
135 ** Fix bugs related to mutation, the null string, and shared substrings.
136 ** Deprecate SCM_ASRTGO.
137 ** Add deprecated shim for `scm_display_error' with stack as first argument.
138 ** i18n: Fix gc_malloc/free mismatch on non-GNU systems.
139 ** Make sure `regexp-quote' tests use Unicode-capable string ports.
140 ** Have `cpu-word-size' error out on unknown CPUs; add support for MIPSEL.
141 ** `scm_from_stringn' always returns unique strings.
142 ** Empty substrings no longer reference the original stringbuf.
143 ** `scm_i_substring_copy' tries to narrow the substring.
144 ** Avoid calling `u32_conv_from_encoding' on the null string.
146 Changes in 2.0.3 (since 2.0.2):
150 ** Guile has a new optimizer, `peval'.
152 `Peval' is a partial evaluator that performs constant folding, dead code
153 elimination, copy propagation, and inlining. By default it runs on
154 every piece of code that Guile compiles, to fold computations that can
155 happen at compile-time, so they don't have to happen at runtime.
157 If we did our job right, the only impact you would see would be your
158 programs getting faster. But if you notice slowdowns or bloated code,
159 please send a mail to bug-guile@gnu.org with details.
161 Thanks to William R. Cook, Oscar Waddell, and Kent Dybvig for inspiring
162 peval and its implementation.
164 You can see what peval does on a given piece of code by running the new
165 `,optimize' REPL meta-command, and comparing it to the output of
166 `,expand'. See "Compile Commands" in the manual, for more.
168 ** Fewer calls to `stat'.
170 Guile now stats only the .go file and the .scm file when loading a fresh
175 ** New module: `(web client)', a simple synchronous web client.
177 See "Web Client" in the manual, for more.
179 ** Users can now install compiled `.go' files.
181 See "Installing Site Packages" in the manual.
183 ** Remove Front-Cover and Back-Cover text from the manual.
185 The manual is still under the GNU Free Documentation License, but no
186 longer has any invariant sections.
188 ** More helpful `guild help'.
190 `guild' is Guile's multi-tool, for use in shell scripting. Now it has a
191 nicer interface for querying the set of existing commands, and getting
192 help on those commands. Try it out and see!
194 ** New macro: `define-syntax-rule'
196 `define-syntax-rule' is a shorthand to make a `syntax-rules' macro with
197 one clause. See "Syntax Rules" in the manual, for more.
199 ** The `,time' REPL meta-command now has more precision.
201 The output of this command now has microsecond precision, instead of
202 10-millisecond precision.
204 ** `(ice-9 match)' can now match records.
206 See "Pattern Matching" in the manual, for more on matching records.
208 ** New module: `(language tree-il debug)'.
210 This module provides a tree-il verifier. This is useful for people that
211 generate tree-il, usually as part of a language compiler.
213 ** New functions: `scm_is_exact', `scm_is_inexact'.
215 These provide a nice C interface for Scheme's `exact?' and `inexact?',
220 See the git log (or the ChangeLog) for more details on these bugs.
222 ** Fix order of importing modules and resolving duplicates handlers.
223 ** Fix a number of bugs involving extended (merged) generics.
224 ** Fix invocation of merge-generics duplicate handler.
225 ** Fix write beyond array end in arrays.c.
226 ** Fix read beyond end of hashtable size array in hashtab.c.
227 ** (web http): Locale-independent parsing and serialization of dates.
228 ** Ensure presence of Host header in HTTP/1.1 requests.
229 ** Fix take-right and drop-right for improper lists.
230 ** Fix leak in get_current_locale().
231 ** Fix recursive define-inlinable expansions.
232 ** Check that srfi-1 procedure arguments are procedures.
233 ** Fix r6rs `map' for multiple returns.
234 ** Fix scm_tmpfile leak on POSIX platforms.
235 ** Fix a couple of leaks (objcode->bytecode, make-boot-program).
236 ** Fix guile-lib back-compatibility for module-stexi-documentation.
237 ** Fix --listen option to allow other ports.
238 ** Fix scm_to_latin1_stringn for substrings.
239 ** Fix compilation of untyped arrays of rank not 1.
240 ** Fix unparse-tree-il of <dynset>.
241 ** Fix reading of #||||#.
242 ** Fix segfault in GOOPS when class fields are redefined.
243 ** Prefer poll(2) over select(2) to allow file descriptors above FD_SETSIZE.
246 Changes in 2.0.2 (since 2.0.1):
250 ** `guile-tools' renamed to `guild'
252 The new name is shorter. Its intended future use is for a CPAN-like
253 system for Guile wizards and journeyfolk to band together to share code;
254 hence the name. `guile-tools' is provided as a backward-compatible
255 symbolic link. See "Using Guile Tools" in the manual, for more.
257 ** New control operators: `shift' and `reset'
259 See "Shift and Reset" in the manual, for more information.
261 ** `while' as an expression
263 Previously the return value of `while' was unspecified. Now its
264 values are specified both in the case of normal termination, and via
265 termination by invoking `break', possibly with arguments. See "while
266 do" in the manual for more.
268 ** Disallow access to handles of weak hash tables
270 `hash-get-handle' and `hash-create-handle!' are no longer permitted to
271 be called on weak hash tables, because the fields in a weak handle could
272 be nulled out by the garbage collector at any time, but yet they are
273 otherwise indistinguishable from pairs. Use `hash-ref' and `hash-set!'
276 ** More precision for `get-internal-run-time', `get-internal-real-time'
278 On 64-bit systems which support POSIX clocks, Guile's internal timing
279 procedures offer nanosecond resolution instead of the 10-millisecond
280 resolution previously available. 32-bit systems now use 1-millisecond
283 ** Guile now measures time spent in GC
285 `gc-stats' now returns a meaningful value for `gc-time-taken'.
289 The statprof profiler now exports a `gcprof' procedure, driven by the
290 `after-gc-hook', to see which parts of your program are causing GC. Let
291 us know if you find it useful.
293 ** `map', `for-each' and some others now implemented in Scheme
295 We would not mention this in NEWS, as it is not a user-visible change,
296 if it were not for one thing: `map' and `for-each' are no longer
297 primitive generics. Instead they are normal bindings, which can be
298 wrapped by normal generics. This fixes some modularity issues between
299 core `map', SRFI-1 `map', and GOOPS.
301 Also it's pretty cool that we can do this without a performance impact.
303 ** Add `scm_peek_byte_or_eof'.
305 This helper is like `scm_peek_char_or_eof', but for bytes instead of
308 ** Implement #:stop-at-first-non-option option for getopt-long
310 See "getopt-long Reference" in the manual, for more information.
312 ** Improve R6RS conformance for conditions in the I/O libraries
314 The `(rnrs io simple)' module now raises the correct R6RS conditions in
315 error cases. `(rnrs io ports)' is also more correct now, though it is
316 still a work in progress.
318 ** All deprecated routines emit warnings
320 A few deprecated routines were lacking deprecation warnings. This has
325 ** Constants in compiled code now share state better
327 Constants with shared state, like `("foo")' and `"foo"', now share state
328 as much as possible, in the entire compilation unit. This cuts compiled
329 `.go' file sizes in half, generally, and speeds startup.
331 ** VLists: optimize `vlist-fold-right', and add `vhash-fold-right'
333 These procedures are now twice as fast as they were.
335 ** UTF-8 ports to bypass `iconv' entirely
337 This reduces memory usage in a very common case.
341 The compiler is now about 40% faster. (Note that this is only the case
342 once the compiler is itself compiled, so the build still takes as long
347 Some assertions that were mostly useful for sanity-checks on the
348 bytecode compiler are now off for both "regular" and "debug" engines.
349 This together with a fix to cache a TLS access and some other tweaks
350 improve the VM's performance by about 20%.
352 ** SRFI-1 list-set optimizations
354 lset-adjoin and lset-union now have fast paths for eq? sets.
356 ** `memq', `memv' optimizations
358 These procedures are now at least twice as fast than in 2.0.1.
362 ** Deprecate scm_whash API
364 `scm_whash_get_handle', `SCM_WHASHFOUNDP', `SCM_WHASHREF',
365 `SCM_WHASHSET', `scm_whash_create_handle', `scm_whash_lookup', and
366 `scm_whash_insert' are now deprecated. Use the normal hash table API
369 ** Deprecate scm_struct_table
371 `SCM_STRUCT_TABLE_NAME', `SCM_SET_STRUCT_TABLE_NAME',
372 `SCM_STRUCT_TABLE_CLASS', `SCM_SET_STRUCT_TABLE_CLASS',
373 `scm_struct_table', and `scm_struct_create_handle' are now deprecated.
374 These routines formed part of the internals of the map between structs
377 ** Deprecate scm_internal_dynamic_wind
379 The `scm_t_inner' type and `scm_internal_dynamic_wind' are deprecated,
380 as the `scm_dynwind' API is better, and this API encourages users to
381 stuff SCM values into pointers.
383 ** Deprecate scm_immutable_cell, scm_immutable_double_cell
385 These routines are deprecated, as the GC_STUBBORN API doesn't do
390 Andreas Rottman kindly transcribed the missing parts of the `(rnrs io
391 ports)' documentation from the R6RS documentation. Thanks Andreas!
395 ** Fix double-loading of script in -ds case
396 ** -x error message fix
397 ** iconveh-related cross-compilation fixes
398 ** Fix small integer return value packing on big endian machines.
399 ** Fix hash-set! in weak-value table from non-immediate to immediate
400 ** Fix call-with-input-file & relatives for multiple values
401 ** Fix `hash' for inf and nan
402 ** Fix libguile internal type errors caught by typing-strictness==2
403 ** Fix compile error in MinGW fstat socket detection
404 ** Fix generation of auto-compiled file names on MinGW
405 ** Fix multithreaded access to internal hash tables
406 ** Emit a 1-based line number in error messages
407 ** Fix define-module ordering
408 ** Fix several POSIX functions to use the locale encoding
409 ** Add type and range checks to the complex generalized vector accessors
410 ** Fix unaligned accesses for bytevectors of complex numbers
412 ** Fix erroneous VM stack overflow for canceled threads
415 Changes in 2.0.1 (since 2.0.0):
419 ** guile.m4 supports linking with rpath
421 The GUILE_FLAGS macro now sets GUILE_LIBS and GUILE_LTLIBS, which
422 include appropriate directives to the linker to include libguile-2.0.so
423 in the runtime library lookup path.
425 ** `begin' expands macros in its body before other expressions
427 This enables support for programs like the following:
432 (or (= x 0) (odd? (- x 1)))))
435 ((odd? x) (not (even? x)))))
438 ** REPL reader usability enhancements
440 The REPL now flushes input after a read error, which should prevent one
441 error from causing other errors. The REPL also now interprets comments
444 ** REPL output has configurable width
446 The REPL now defaults to output with the current terminal's width, in
447 columns. See "Debug Commands" in the manual for more information on
450 ** Better C access to the module system
452 Guile now has convenient C accessors to look up variables or values in
453 modules and their public interfaces. See `scm_c_public_ref' and friends
454 in "Accessing Modules from C" in the manual.
456 ** Added `scm_call_5', `scm_call_6'
458 See "Fly Evaluation" in the manual.
460 ** Added `scm_from_latin1_keyword', `scm_from_utf8_keyword'
462 See "Keyword Procedures" in the manual, for more. Note that
463 `scm_from_locale_keyword' should not be used when the name is a C string
466 ** R6RS unicode and string I/O work
468 Added efficient implementations of `get-string-n' and `get-string-n!'
469 for binary ports. Exported `current-input-port', `current-output-port'
470 and `current-error-port' from `(rnrs io ports)', and enhanced support
473 ** Added `pointer->scm', `scm->pointer' to `(system foreign)'
475 These procedure are useful if one needs to pass and receive SCM values
476 to and from foreign functions. See "Foreign Variables" in the manual,
479 ** Added `heap-allocated-since-gc' to `(gc-stats)'
481 Also fixed the long-standing bug in the REPL `,stat' command.
483 ** Add `on-error' REPL option
485 This option controls what happens when an error occurs at the REPL, and
486 defaults to `debug', indicating that Guile should enter the debugger.
487 Other values include `report', which will simply print a backtrace
488 without entering the debugger. See "System Commands" in the manual.
490 ** Enforce immutability of string literals
492 Attempting to mutate a string literal now causes a runtime error.
494 ** Fix pthread redirection
496 Guile 2.0.0 shipped with headers that, if configured with pthread
497 support, would re-define `pthread_create', `pthread_join', and other API
498 to redirect to the BDW-GC wrappers, `GC_pthread_create', etc. This was
499 unintended, and not necessary: because threads must enter Guile with
500 `scm_with_guile', Guile can handle thread registration itself, without
501 needing to make the GC aware of all threads. This oversight has been
504 ** `with-continuation-barrier' now unwinds on `quit'
506 A throw to `quit' in a continuation barrier will cause Guile to exit.
507 Before, it would do so before unwinding to the barrier, which would
508 prevent cleanup handlers from running. This has been fixed so that it
509 exits only after unwinding.
511 ** `string->pointer' and `pointer->string' have optional encoding arg
513 This allows users of the FFI to more easily deal in strings with
514 particular (non-locale) encodings, like "utf-8". See "Void Pointers and
515 Byte Access" in the manual, for more.
517 ** R6RS fixnum arithmetic optimizations
519 R6RS fixnum operations are are still slower than generic arithmetic,
522 ** New procedure: `define-inlinable'
524 See "Inlinable Procedures" in the manual, for more.
526 ** New procedure: `exact-integer-sqrt'
528 See "Integer Operations" in the manual, for more.
530 ** "Extended read syntax" for symbols parses better
532 In #{foo}# symbols, backslashes are now treated as escapes, as the
533 symbol-printing code intended. Additionally, "\x" within #{foo}# is now
534 interpreted as starting an R6RS hex escape. This is backward compatible
535 because the symbol printer would never produce a "\x" before. The
536 printer also works better too.
538 ** Added `--fresh-auto-compile' option
540 This allows a user to invalidate the auto-compilation cache. It's
541 usually not needed. See "Compilation" in the manual, for a discussion.
545 ** GOOPS documentation updates
549 Thanks to Mark Harig for improvements to guile.1.
551 ** SRFI-23 documented
553 The humble `error' SRFI now has an entry in the manual.
557 ** `(ice-9 binary-ports)': "R6RS I/O Ports", in the manual
558 ** `(ice-9 eval-string)': "Fly Evaluation", in the manual
559 ** `(ice-9 command-line)', not documented yet
563 ** Fixed `iconv_t' memory leak on close-port
564 ** Fixed some leaks with weak hash tables
565 ** Export `vhash-delq' and `vhash-delv' from `(ice-9 vlist)'
566 ** `after-gc-hook' works again
567 ** `define-record-type' now allowed in nested contexts
568 ** `exact-integer-sqrt' now handles large integers correctly
569 ** Fixed C extension examples in manual
570 ** `vhash-delete' honors HASH argument
571 ** Make `locale-digit-grouping' more robust
572 ** Default exception printer robustness fixes
573 ** Fix presence of non-I CPPFLAGS in `guile-2.0.pc'
574 ** `read' updates line/column numbers when reading SCSH block comments
575 ** Fix imports of multiple custom interfaces of same module
576 ** Fix encoding scanning for non-seekable ports
577 ** Fix `setter' when called with a non-setter generic
578 ** Fix f32 and f64 bytevectors to not accept rationals
579 ** Fix description of the R6RS `finite?' in manual
580 ** Quotient, remainder and modulo accept inexact integers again
581 ** Fix `continue' within `while' to take zero arguments
582 ** Fix alignment for structures in FFI
583 ** Fix port-filename of stdin, stdout, stderr to match the docs
584 ** Fix weak hash table-related bug in `define-wrapped-pointer-type'
585 ** Fix partial continuation application with pending procedure calls
586 ** scm_{to,from}_locale_string use current locale, not current ports
587 ** Fix thread cleanup, by using a pthread_key destructor
588 ** Fix `quit' at the REPL
589 ** Fix a failure to sync regs in vm bytevector ops
590 ** Fix (texinfo reflection) to handle nested structures like syntax patterns
591 ** Fix stexi->html double translation
592 ** Fix tree-il->scheme fix for <prompt>
593 ** Fix compilation of <prompt> in <fix> in single-value context
594 ** Fix race condition in ensure-writable-dir
595 ** Fix error message on ,disassemble "non-procedure"
596 ** Fix prompt and abort with the boot evaluator
597 ** Fix `procedure->pointer' for functions returning `void'
598 ** Fix error reporting in dynamic-pointer
599 ** Fix problems detecting coding: in block comments
600 ** Fix duplicate load-path and load-compiled-path in compilation environment
601 ** Add fallback read(2) suppport for .go files if mmap(2) unavailable
602 ** Fix c32vector-set!, c64vector-set!
603 ** Fix mistakenly deprecated read syntax for uniform complex vectors
604 ** Fix parsing of exact numbers with negative exponents
605 ** Ignore SIGPIPE in (system repl server)
606 ** Fix optional second arg to R6RS log function
607 ** Fix R6RS `assert' to return true value.
608 ** Fix fencepost error when seeking in bytevector input ports
609 ** Gracefully handle `setlocale' errors when starting the REPL
610 ** Improve support of the `--disable-posix' configure option
611 ** Make sure R6RS binary ports pass `binary-port?' regardless of the locale
612 ** Gracefully handle unterminated UTF-8 sequences instead of hitting an `assert'
616 Changes in 2.0.0 (changes since the 1.8.x series):
618 * New modules (see the manual for details)
620 ** `(srfi srfi-18)', more sophisticated multithreading support
621 ** `(srfi srfi-27)', sources of random bits
622 ** `(srfi srfi-38)', External Representation for Data With Shared Structure
623 ** `(srfi srfi-42)', eager comprehensions
624 ** `(srfi srfi-45)', primitives for expressing iterative lazy algorithms
625 ** `(srfi srfi-67)', compare procedures
626 ** `(ice-9 i18n)', internationalization support
627 ** `(ice-9 futures)', fine-grain parallelism
628 ** `(rnrs bytevectors)', the R6RS bytevector API
629 ** `(rnrs io ports)', a subset of the R6RS I/O port API
630 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
631 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
632 ** `(system foreign)', foreign function interface
633 ** `(sxml match)', a pattern matcher for SXML
634 ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library
635 ** `(system vm coverage)', a line-by-line code coverage library
636 ** `(web uri)', URI data type, parser, and unparser
637 ** `(web http)', HTTP header parsers and unparsers
638 ** `(web request)', HTTP request data type, reader, and writer
639 ** `(web response)', HTTP response data type, reader, and writer
640 ** `(web server)', Generic HTTP server
641 ** `(ice-9 poll)', a poll wrapper
642 ** `(web server http)', HTTP-over-TCP web server implementation
644 ** Replaced `(ice-9 match)' with Alex Shinn's compatible, hygienic matcher.
646 Guile's copy of Andrew K. Wright's `match' library has been replaced by
647 a compatible hygienic implementation by Alex Shinn. It is now
648 documented, see "Pattern Matching" in the manual.
650 Compared to Andrew K. Wright's `match', the new `match' lacks
651 `match-define', `match:error-control', `match:set-error-control',
652 `match:error', `match:set-error', and all structure-related procedures.
654 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
656 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
657 toolkit from Guile-Lib have been imported into Guile proper. See
658 "Standard Library" in the manual for more details.
660 ** Integration of lalr-scm, a parser generator
662 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
663 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
666 * Changes to the stand-alone interpreter
668 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
670 Compiled code loads much faster than Scheme source code, and runs around
671 3 or 4 times as fast, generating much less garbage in the process.
673 ** Evaluating Scheme code does not use the C stack.
675 Besides when compiling Guile itself, Guile no longer uses a recursive C
676 function as an evaluator. This obviates the need to check the C stack
677 pointer for overflow. Continuations still capture the C stack, however.
679 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
680 GUILE_SYSTEM_LOAD_COMPILED_PATH
682 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
683 for source files. It is a different path, however, because compiled
684 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
687 ** New read-eval-print loop (REPL) implementation
689 Running Guile with no arguments drops the user into the new REPL. See
690 "Using Guile Interactively" in the manual, for more information.
692 ** Remove old Emacs interface
694 Guile had an unused `--emacs' command line argument that was supposed to
695 help when running Guile inside Emacs. This option has been removed, and
696 the helper functions `named-module-use!' and `load-emacs-interface' have
699 ** Add `(system repl server)' module and `--listen' command-line argument
701 The `(system repl server)' module exposes procedures to listen on
702 sockets for connections, and serve REPLs to those clients. The --listen
703 command-line argument allows any Guile program to thus be remotely
706 See "Invoking Guile" for more information on `--listen'.
708 ** Command line additions
710 The guile binary now supports a new switch "-x", which can be used to
711 extend the list of filename extensions tried when loading files
714 ** New reader options: `square-brackets', `r6rs-hex-escapes',
717 The reader supports a new option (changeable via `read-options'),
718 `square-brackets', which instructs it to interpret square brackets as
719 parentheses. This option is on by default.
721 When the new `r6rs-hex-escapes' reader option is enabled, the reader
722 will recognize string escape sequences as defined in R6RS. R6RS string
723 escape sequences are incompatible with Guile's existing escapes, though,
724 so this option is off by default.
726 Additionally, Guile follows the R6RS newline escaping rules when the
727 `hungry-eol-escapes' option is enabled.
729 See "String Syntax" in the manual, for more information.
731 ** Function profiling and tracing at the REPL
733 The `,profile FORM' REPL meta-command can now be used to statistically
734 profile execution of a form, to see which functions are taking the most
735 time. See `,help profile' for more information.
737 Similarly, `,trace FORM' traces all function applications that occur
738 during the execution of `FORM'. See `,help trace' for more information.
740 ** Recursive debugging REPL on error
742 When Guile sees an error at the REPL, instead of saving the stack, Guile
743 will directly enter a recursive REPL in the dynamic context of the
744 error. See "Error Handling" in the manual, for more information.
746 A recursive REPL is the same as any other REPL, except that it
747 has been augmented with debugging information, so that one can inspect
748 the context of the error. The debugger has been integrated with the REPL
749 via a set of debugging meta-commands.
751 For example, one may access a backtrace with `,backtrace' (or
752 `,bt'). See "Interactive Debugging" in the manual, for more
755 ** New `guile-tools' commands: `compile', `disassemble'
757 Pass the `--help' command-line option to these commands for more
760 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
762 Users may now install Guile to nonstandard prefixes and just run
763 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
764 include `/path/to/lib'.
766 ** Guile's Emacs integration is now more keyboard-friendly
768 Backtraces may now be disclosed with the keyboard in addition to the
771 ** Load path change: search in version-specific paths before site paths
773 When looking for a module, Guile now searches first in Guile's
774 version-specific path (the library path), *then* in the site dir. This
775 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
776 installed. Also it should cut the number of `stat' system calls by half,
779 ** Value history in the REPL on by default
781 By default, the REPL will save computed values in variables like `$1',
782 `$2', and the like. There are programmatic and interactive interfaces to
783 control this. See "Value History" in the manual, for more information.
785 ** Readline tab completion for arguments
787 When readline is enabled, tab completion works for arguments too, not
788 just for the operator position.
790 ** Expression-oriented readline history
792 Guile's readline history now tries to operate on expressions instead of
793 input lines. Let us know what you think!
795 ** Interactive Guile follows GNU conventions
797 As recommended by the GPL, Guile now shows a brief copyright and
798 warranty disclaimer on startup, along with pointers to more information.
800 * Changes to Scheme functions and syntax
802 ** Support for R6RS libraries
804 The `library' and `import' forms from the latest Scheme report have been
805 added to Guile, in such a way that R6RS libraries share a namespace with
806 Guile modules. R6RS modules may import Guile modules, and are available
807 for Guile modules to import via use-modules and all the rest. See "R6RS
808 Libraries" in the manual for more information.
810 ** Implementations of R6RS libraries
812 Guile now has implementations for all of the libraries defined in the
813 R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS
814 Standard Libraries" in the manual for a full list of libraries.
816 ** Partial R6RS compatibility
818 Guile now has enough support for R6RS to run a reasonably large subset
821 Guile is not fully R6RS compatible. Many incompatibilities are simply
822 bugs, though some parts of Guile will remain R6RS-incompatible for the
823 foreseeable future. See "R6RS Incompatibilities" in the manual, for more
826 Please contact bug-guile@gnu.org if you have found an issue not
827 mentioned in that compatibility list.
829 ** New implementation of `primitive-eval'
831 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
832 still a C evaluator, used when building a fresh Guile to interpret the
833 compiler, so we can compile eval.scm. Thereafter all calls to
834 primitive-eval are implemented by VM-compiled code.
836 This allows all of Guile's procedures, be they interpreted or compiled,
837 to execute on the same stack, unifying multiple-value return semantics,
838 providing for proper tail recursion between interpreted and compiled
839 code, and simplifying debugging.
841 As part of this change, the evaluator no longer mutates the internal
842 representation of the code being evaluated in a thread-unsafe manner.
844 There are two negative aspects of this change, however. First, Guile
845 takes a lot longer to compile now. Also, there is less debugging
846 information available for debugging interpreted code. We hope to improve
847 both of these situations.
849 There are many changes to the internal C evalator interface, but all
850 public interfaces should be the same. See the ChangeLog for details. If
851 we have inadvertantly changed an interface that you were using, please
852 contact bug-guile@gnu.org.
854 ** Procedure removed: `the-environment'
856 This procedure was part of the interpreter's execution model, and does
857 not apply to the compiler.
859 ** No more `local-eval'
861 `local-eval' used to exist so that one could evaluate code in the
862 lexical context of a function. Since there is no way to get the lexical
863 environment any more, as that concept has no meaning for the compiler,
864 and a different meaning for the interpreter, we have removed the
867 If you think you need `local-eval', you should probably implement your
868 own metacircular evaluator. It will probably be as fast as Guile's
871 ** Scheme source files will now be compiled automatically.
873 If a compiled .go file corresponding to a .scm file is not found or is
874 not fresh, the .scm file will be compiled on the fly, and the resulting
875 .go file stored away. An advisory note will be printed on the console.
877 Note that this mechanism depends on the timestamp of the .go file being
878 newer than that of the .scm file; if the .scm or .go files are moved
879 after installation, care should be taken to preserve their original
882 Auto-compiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
883 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
884 will be created if needed.
886 To inhibit automatic compilation, set the GUILE_AUTO_COMPILE environment
887 variable to 0, or pass --no-auto-compile on the Guile command line.
889 ** New POSIX procedures: `getrlimit' and `setrlimit'
891 Note however that the interface of these functions is likely to change
892 in the next prerelease.
894 ** New POSIX procedure: `getsid'
896 Scheme binding for the `getsid' C library call.
898 ** New POSIX procedure: `getaddrinfo'
900 Scheme binding for the `getaddrinfo' C library function.
902 ** Multicast socket options
904 Support was added for the IP_MULTICAST_TTL and IP_MULTICAST_IF socket
905 options. See "Network Sockets and Communication" in the manual, for
908 ** `recv!', `recvfrom!', `send', `sendto' now deal in bytevectors
910 These socket procedures now take bytevectors as arguments, instead of
911 strings. There is some deprecated string support, however.
913 ** New GNU procedures: `setaffinity' and `getaffinity'.
915 See "Processes" in the manual, for more information.
917 ** New procedures: `compose', `negate', and `const'
919 See "Higher-Order Functions" in the manual, for more information.
921 ** New procedure in `(oops goops)': `method-formals'
923 ** New procedures in (ice-9 session): `add-value-help-handler!',
924 `remove-value-help-handler!', `add-name-help-handler!'
925 `remove-name-help-handler!', `procedure-arguments'
927 The value and name help handlers provide some minimal extensibility to
928 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
929 example, to make stexinfo help documentation available. See those
930 procedures' docstrings for more information.
932 `procedure-arguments' describes the arguments that a procedure can take,
933 combining arity and formals. For example:
935 (procedure-arguments resolve-interface)
936 => ((required . (name)) (rest . args))
938 Additionally, `module-commentary' is now publically exported from
941 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
943 These procedures created primitive fexprs for the old evaluator, and are
944 no longer supported. If you feel that you need these functions, you
945 probably need to write your own metacircular evaluator (which will
946 probably be as fast as Guile's, anyway).
948 ** New language: ECMAScript
950 Guile now ships with one other high-level language supported,
951 ECMAScript. The goal is to support all of version 3.1 of the standard,
952 but not all of the libraries are there yet. This support is not yet
953 documented; ask on the mailing list if you are interested.
955 ** New language: Brainfuck
957 Brainfuck is a toy language that closely models Turing machines. Guile's
958 brainfuck compiler is meant to be an example of implementing other
959 languages. See the manual for details, or
960 http://en.wikipedia.org/wiki/Brainfuck for more information about the
961 Brainfuck language itself.
963 ** New language: Elisp
965 Guile now has an experimental Emacs Lisp compiler and runtime. You can
966 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
967 Kraft and Brian Templeton, and all bugs to bug-guile@gnu.org.
969 ** Better documentation infrastructure for macros
971 It is now possible to introspect on the type of a macro, e.g.
972 syntax-rules, identifier-syntax, etc, and extract information about that
973 macro, such as the syntax-rules patterns or the defmacro arguments.
974 `(texinfo reflection)' takes advantage of this to give better macro
977 ** Support for arbitrary procedure metadata
979 Building on its support for docstrings, Guile now supports multiple
980 docstrings, adding them to the tail of a compiled procedure's
981 properties. For example:
987 (procedure-properties foo)
988 => ((name . foo) (documentation . "one") (documentation . "two"))
990 Also, vectors of pairs are now treated as additional metadata entries:
993 #((quz . #f) (docstring . "xyzzy"))
995 (procedure-properties bar)
996 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
998 This allows arbitrary literals to be embedded as metadata in a compiled
1001 ** The psyntax expander now knows how to interpret the @ and @@ special
1004 ** The psyntax expander is now hygienic with respect to modules.
1006 Free variables in a macro are scoped in the module that the macro was
1007 defined in, not in the module the macro is used in. For example, code
1008 like this works now:
1010 (define-module (foo) #:export (bar))
1011 (define (helper x) ...)
1013 (syntax-rules () ((_ x) (helper x))))
1015 (define-module (baz) #:use-module (foo))
1018 It used to be you had to export `helper' from `(foo)' as well.
1019 Thankfully, this has been fixed.
1021 ** Support for version information in Guile's `module' form
1023 Guile modules now have a `#:version' field. See "R6RS Version
1024 References", "General Information about Modules", "Using Guile Modules",
1025 and "Creating Guile Modules" in the manual for more information.
1027 ** Support for renaming bindings on module export
1029 Wherever Guile accepts a symbol as an argument to specify a binding to
1030 export, it now also accepts a pair of symbols, indicating that a binding
1031 should be renamed on export. See "Creating Guile Modules" in the manual
1032 for more information.
1034 ** New procedure: `module-export-all!'
1036 This procedure exports all current and future bindings from a module.
1037 Use as `(module-export-all! (current-module))'.
1039 ** New procedure `reload-module', and `,reload' REPL command
1041 See "Module System Reflection" and "Module Commands" in the manual, for
1044 ** `eval-case' has been deprecated, and replaced by `eval-when'.
1046 The semantics of `eval-when' are easier to understand. See "Eval When"
1047 in the manual, for more information.
1049 ** Guile is now more strict about prohibiting definitions in expression
1052 Although previous versions of Guile accepted it, the following
1053 expression is not valid, in R5RS or R6RS:
1055 (if test (define foo 'bar) (define foo 'baz))
1057 In this specific case, it would be better to do:
1059 (define foo (if test 'bar 'baz))
1061 It is possible to circumvent this restriction with e.g.
1062 `(module-define! (current-module) 'foo 'baz)'. Contact the list if you
1065 ** Support for `letrec*'
1067 Guile now supports `letrec*', a recursive lexical binding operator in
1068 which the identifiers are bound in order. See "Local Bindings" in the
1069 manual, for more details.
1071 ** Internal definitions now expand to `letrec*'
1073 Following the R6RS, internal definitions now expand to letrec* instead
1074 of letrec. The following program is invalid for R5RS, but valid for
1079 (define baz (+ bar 20))
1082 ;; R5RS and Guile <= 1.8:
1083 (foo) => Unbound variable: bar
1084 ;; R6RS and Guile >= 2.0:
1087 This change should not affect correct R5RS programs, or programs written
1088 in earlier Guile dialects.
1090 ** Macro expansion produces structures instead of s-expressions
1092 In the olden days, macroexpanding an s-expression would yield another
1093 s-expression. Though the lexical variables were renamed, expansions of
1094 core forms like `if' and `begin' were still non-hygienic, as they relied
1095 on the toplevel definitions of `if' et al being the conventional ones.
1097 The solution is to expand to structures instead of s-expressions. There
1098 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
1099 etc. The expander already did this for compilation, producing Tree-IL
1100 directly; it has been changed now to do so when expanding for the
1103 ** Defmacros must now produce valid Scheme expressions.
1105 It used to be that defmacros could unquote in Scheme values, as a way of
1106 supporting partial evaluation, and avoiding some hygiene issues. For
1109 (define (helper x) ...)
1110 (define-macro (foo bar)
1113 Assuming this macro is in the `(baz)' module, the direct translation of
1116 (define (helper x) ...)
1117 (define-macro (foo bar)
1118 `((@@ (baz) helper) ,bar))
1120 Of course, one could just use a hygienic macro instead:
1124 ((_ bar) (helper bar))))
1126 ** Guile's psyntax now supports docstrings and internal definitions.
1128 The following Scheme is not strictly legal:
1135 However its intent is fairly clear. Guile interprets "bar" to be the
1136 docstring of `foo', and the definition of `baz' is still in definition
1139 ** Support for settable identifier syntax
1141 Following the R6RS, "variable transformers" are settable
1142 identifier-syntax. See "Identifier macros" in the manual, for more
1145 ** syntax-case treats `_' as a placeholder
1147 Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches
1148 anything, and binds no pattern variables. Unlike the R6RS, Guile also
1149 permits `_' to be in the literals list for a pattern.
1151 ** Macros need to be defined before their first use.
1153 It used to be that with lazy memoization, this might work:
1157 (define-macro (ref x) x)
1160 But now, the body of `foo' is interpreted to mean a call to the toplevel
1161 `ref' function, instead of a macro expansion. The solution is to define
1162 macros before code that uses them.
1164 ** Functions needed by macros at expand-time need to be present at
1167 For example, this code will work at the REPL:
1169 (define (double-helper x) (* x x))
1170 (define-macro (double-literal x) (double-helper x))
1171 (double-literal 2) => 4
1173 But it will not work when a file is compiled, because the definition of
1174 `double-helper' is not present at expand-time. The solution is to wrap
1175 the definition of `double-helper' in `eval-when':
1177 (eval-when (load compile eval)
1178 (define (double-helper x) (* x x)))
1179 (define-macro (double-literal x) (double-helper x))
1180 (double-literal 2) => 4
1182 See the documentation for eval-when for more information.
1184 ** `macroexpand' produces structures, not S-expressions.
1186 Given the need to maintain referential transparency, both lexically and
1187 modular, the result of expanding Scheme expressions is no longer itself
1188 an s-expression. If you want a human-readable approximation of the
1189 result of `macroexpand', call `tree-il->scheme' from `(language
1192 ** Removed function: `macroexpand-1'
1194 It is unclear how to implement `macroexpand-1' with syntax-case, though
1195 PLT Scheme does prove that it is possible.
1197 ** New reader macros: #' #` #, #,@
1199 These macros translate, respectively, to `syntax', `quasisyntax',
1200 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
1201 These reader macros may be overridden by `read-hash-extend'.
1203 ** Incompatible change to #'
1205 Guile did have a #' hash-extension, by default, which just returned the
1206 subsequent datum: #'foo => foo. In the unlikely event that anyone
1207 actually used this, this behavior may be reinstated via the
1208 `read-hash-extend' mechanism.
1210 ** `unquote' and `unquote-splicing' accept multiple expressions
1212 As per the R6RS, these syntax operators can now accept any number of
1213 expressions to unquote.
1215 ** Scheme expresssions may be commented out with #;
1217 #; comments out an entire expression. See SRFI-62 or the R6RS for more
1220 ** Prompts: Delimited, composable continuations
1222 Guile now has prompts as part of its primitive language. See "Prompts"
1223 in the manual, for more information.
1225 Expressions entered in at the REPL, or from the command line, are
1226 surrounded by a prompt with the default prompt tag.
1228 ** `make-stack' with a tail-called procedural narrowing argument no longer
1229 works (with compiled procedures)
1231 It used to be the case that a captured stack could be narrowed to select
1232 calls only up to or from a certain procedure, even if that procedure
1233 already tail-called another procedure. This was because the debug
1234 information from the original procedure was kept on the stack.
1236 Now with the new compiler, the stack only contains active frames from
1237 the current continuation. A narrow to a procedure that is not in the
1238 stack will result in an empty stack. To fix this, narrow to a procedure
1239 that is active in the current continuation, or narrow to a specific
1240 number of stack frames.
1242 ** Backtraces through compiled procedures only show procedures that are
1243 active in the current continuation
1245 Similarly to the previous issue, backtraces in compiled code may be
1246 different from backtraces in interpreted code. There are no semantic
1247 differences, however. Please mail bug-guile@gnu.org if you see any
1248 deficiencies with Guile's backtraces.
1250 ** `positions' reader option enabled by default
1252 This change allows primitive-load without --auto-compile to also
1253 propagate source information through the expander, for better errors and
1254 to let macros know their source locations. The compiler was already
1255 turning it on anyway.
1257 ** New macro: `current-source-location'
1259 The macro returns the current source location (to be documented).
1261 ** syntax-rules and syntax-case macros now propagate source information
1262 through to the expanded code
1264 This should result in better backtraces.
1266 ** The currying behavior of `define' has been removed.
1268 Before, `(define ((f a) b) (* a b))' would translate to
1270 (define f (lambda (a) (lambda (b) (* a b))))
1272 Now a syntax error is signaled, as this syntax is not supported by
1273 default. Use the `(ice-9 curried-definitions)' module to get back the
1276 ** New procedure, `define!'
1278 `define!' is a procedure that takes two arguments, a symbol and a value,
1279 and binds the value to the symbol in the current module. It's useful to
1280 programmatically make definitions in the current module, and is slightly
1281 less verbose than `module-define!'.
1283 ** All modules have names now
1285 Before, you could have anonymous modules: modules without names. Now,
1286 because of hygiene and macros, all modules have names. If a module was
1287 created without a name, the first time `module-name' is called on it, a
1288 fresh name will be lazily generated for it.
1290 ** The module namespace is now separate from the value namespace
1292 It was a little-known implementation detail of Guile's module system
1293 that it was built on a single hierarchical namespace of values -- that
1294 if there was a module named `(foo bar)', then in the module named
1295 `(foo)' there was a binding from `bar' to the `(foo bar)' module.
1297 This was a neat trick, but presented a number of problems. One problem
1298 was that the bindings in a module were not apparent from the module
1299 itself; perhaps the `(foo)' module had a private binding for `bar', and
1300 then an external contributor defined `(foo bar)'. In the end there can
1301 be only one binding, so one of the two will see the wrong thing, and
1302 produce an obtuse error of unclear provenance.
1304 Also, the public interface of a module was also bound in the value
1305 namespace, as `%module-public-interface'. This was a hack from the early
1306 days of Guile's modules.
1308 Both of these warts have been fixed by the addition of fields in the
1309 `module' data type. Access to modules and their interfaces from the
1310 value namespace has been deprecated, and all accessors use the new
1311 record accessors appropriately.
1313 When Guile is built with support for deprecated code, as is the default,
1314 the value namespace is still searched for modules and public interfaces,
1315 and a deprecation warning is raised as appropriate.
1317 Finally, to support lazy loading of modules as one used to be able to do
1318 with module binder procedures, Guile now has submodule binders, called
1319 if a given submodule is not found. See boot-9.scm for more information.
1321 ** New procedures: module-ref-submodule, module-define-submodule,
1322 nested-ref-module, nested-define-module!, local-ref-module,
1325 These new accessors are like their bare variants, but operate on
1326 namespaces instead of values.
1328 ** The (app modules) module tree is officially deprecated
1330 It used to be that one could access a module named `(foo bar)' via
1331 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
1332 modules)' bit was a never-used and never-documented abstraction, and has
1333 been deprecated. See the following mail for a full discussion:
1335 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
1337 The `%app' binding is also deprecated.
1339 ** `module-filename' field and accessor
1341 Modules now record the file in which they are defined. This field may be
1342 accessed with the new `module-filename' procedure.
1344 ** Modules load within a known environment
1346 It takes a few procedure calls to define a module, and those procedure
1347 calls need to be in scope. Now we ensure that the current module when
1348 loading a module is one that has the needed bindings, instead of relying
1351 ** `load' is a macro (!) that resolves paths relative to source file dir
1353 The familiar Schem `load' procedure is now a macro that captures the
1354 name of the source file being expanded, and dispatches to the new
1355 `load-in-vicinity'. Referencing `load' by bare name returns a closure
1356 that embeds the current source file name.
1358 This fix allows `load' of relative paths to be resolved with respect to
1359 the location of the file that calls `load'.
1361 ** Many syntax errors have different texts now
1363 Syntax errors still throw to the `syntax-error' key, but the arguments
1364 are often different now. Perhaps in the future, Guile will switch to
1365 using standard SRFI-35 conditions.
1367 ** Returning multiple values to compiled code will silently truncate the
1368 values to the expected number
1370 For example, the interpreter would raise an error evaluating the form,
1371 `(+ (values 1 2) (values 3 4))', because it would see the operands as
1372 being two compound "values" objects, to which `+' does not apply.
1374 The compiler, on the other hand, receives multiple values on the stack,
1375 not as a compound object. Given that it must check the number of values
1376 anyway, if too many values are provided for a continuation, it chooses
1377 to truncate those values, effectively evaluating `(+ 1 3)' instead.
1379 The idea is that the semantics that the compiler implements is more
1380 intuitive, and the use of the interpreter will fade out with time.
1381 This behavior is allowed both by the R5RS and the R6RS.
1383 ** Multiple values in compiled code are not represented by compound
1386 This change may manifest itself in the following situation:
1388 (let ((val (foo))) (do-something) val)
1390 In the interpreter, if `foo' returns multiple values, multiple values
1391 are produced from the `let' expression. In the compiler, those values
1392 are truncated to the first value, and that first value is returned. In
1393 the compiler, if `foo' returns no values, an error will be raised, while
1394 the interpreter would proceed.
1396 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
1397 behavior is more correct, however. If you wish to preserve a potentially
1398 multiply-valued return, you will need to set up a multiple-value
1399 continuation, using `call-with-values'.
1401 ** Defmacros are now implemented in terms of syntax-case.
1403 The practical ramification of this is that the `defmacro?' predicate has
1404 been removed, along with `defmacro-transformer', `macro-table',
1405 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
1406 `defmacro:transformer'. This is because defmacros are simply macros. If
1407 any of these procedures provided useful facilities to you, we encourage
1408 you to contact the Guile developers.
1410 ** Hygienic macros documented as the primary syntactic extension mechanism.
1412 The macro documentation was finally fleshed out with some documentation
1413 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
1414 expansion process. See "Macros" in the manual, for details.
1416 ** psyntax is now the default expander
1418 Scheme code is now expanded by default by the psyntax hygienic macro
1419 expander. Expansion is performed completely before compilation or
1422 Notably, syntax errors will be signalled before interpretation begins.
1423 In the past, many syntax errors were only detected at runtime if the
1424 code in question was memoized.
1426 As part of its expansion, psyntax renames all lexically-bound
1427 identifiers. Original identifier names are preserved and given to the
1428 compiler, but the interpreter will see the renamed variables, e.g.,
1429 `x432' instead of `x'.
1431 Note that the psyntax that Guile uses is a fork, as Guile already had
1432 modules before incompatible modules were added to psyntax -- about 10
1433 years ago! Thus there are surely a number of bugs that have been fixed
1434 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
1436 ** syntax-rules and syntax-case are available by default.
1438 There is no longer any need to import the `(ice-9 syncase)' module
1439 (which is now deprecated). The expander may be invoked directly via
1440 `macroexpand', though it is normally searched for via the current module
1443 Also, the helper routines for syntax-case are available in the default
1444 environment as well: `syntax->datum', `datum->syntax',
1445 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
1446 `identifier?', and `syntax-violation'. See the R6RS for documentation.
1448 ** Tail patterns in syntax-case
1450 Guile has pulled in some more recent changes from the psyntax portable
1451 syntax expander, to implement support for "tail patterns". Such patterns
1452 are supported by syntax-rules and syntax-case. This allows a syntax-case
1453 match clause to have ellipses, then a pattern at the end. For example:
1456 (syntax-rules (else)
1457 ((_ val match-clause ... (else e e* ...))
1460 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
1461 tail pattern for the else clause. Thanks to Andreas Rottmann for the
1462 patch, and Kent Dybvig for the code.
1464 ** Lexical bindings introduced by hygienic macros may not be referenced
1465 by nonhygienic macros.
1467 If a lexical binding is introduced by a hygienic macro, it may not be
1468 referenced by a nonhygienic macro. For example, this works:
1471 (define-macro (bind-x val body)
1472 `(let ((x ,val)) ,body))
1473 (define-macro (ref x)
1475 (bind-x 10 (ref x)))
1480 (define-syntax bind-x
1482 ((_ val body) (let ((x val)) body))))
1483 (define-macro (ref x)
1485 (bind-x 10 (ref x)))
1487 It is not normal to run into this situation with existing code. However,
1488 if you have defmacros that expand to hygienic macros, it is possible to
1489 run into situations like this. For example, if you have a defmacro that
1490 generates a `while' expression, the `break' bound by the `while' may not
1491 be visible within other parts of your defmacro. The solution is to port
1492 from defmacros to syntax-rules or syntax-case.
1494 ** Macros may no longer be referenced as first-class values.
1496 In the past, you could evaluate e.g. `if', and get its macro value. Now,
1497 expanding this form raises a syntax error.
1499 Macros still /exist/ as first-class values, but they must be
1500 /referenced/ via the module system, e.g. `(module-ref (current-module)
1503 ** Macros may now have docstrings.
1505 `object-documentation' from `(ice-9 documentation)' may be used to
1506 retrieve the docstring, once you have a macro value -- but see the above
1507 note about first-class macros. Docstrings are associated with the syntax
1508 transformer procedures.
1510 ** `case-lambda' is now available in the default environment.
1512 The binding in the default environment is equivalent to the one from the
1513 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
1514 to maintain compatibility with Guile 1.8 and earlier.
1516 ** Procedures may now have more than one arity.
1518 This can be the case, for example, in case-lambda procedures. The
1519 arities of compiled procedures may be accessed via procedures from the
1520 `(system vm program)' module; see "Compiled Procedures", "Optional
1521 Arguments", and "Case-lambda" in the manual.
1523 ** Deprecate arity access via (procedure-properties proc 'arity)
1525 Instead of accessing a procedure's arity as a property, use the new
1526 `procedure-minimum-arity' function, which gives the most permissive
1527 arity that the function has, in the same format as the old arity
1530 ** `lambda*' and `define*' are now available in the default environment
1532 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
1533 compatibility purposes. No semantic change has been made (we hope).
1534 Optional and keyword arguments now dispatch via special VM operations,
1535 without the need to cons rest arguments, making them very fast.
1537 ** New syntax: define-once
1539 `define-once' is like Lisp's `defvar': it creates a toplevel binding,
1540 but only if one does not exist already.
1542 ** New function, `truncated-print', with `format' support
1544 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
1545 will ensure that the output stays within a certain width, truncating the
1546 output in what is hopefully an intelligent manner. See the manual for
1549 There is a new `format' specifier, `~@y', for doing a truncated
1550 print (as opposed to `~y', which does a pretty-print). See the `format'
1551 documentation for more details.
1553 ** Better pretty-printing
1555 Indentation recognizes more special forms, like `syntax-case', and read
1556 macros like `quote' are printed better.
1558 ** Passing a number as the destination of `format' is deprecated
1560 The `format' procedure in `(ice-9 format)' now emits a deprecation
1561 warning if a number is passed as its first argument.
1563 Also, it used to be that you could omit passing a port to `format', in
1564 some cases. This still works, but has been formally deprecated.
1566 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
1568 Guile now implements SRFI-4 vectors using bytevectors. Often when you
1569 have a numeric vector, you end up wanting to write its bytes somewhere,
1570 or have access to the underlying bytes, or read in bytes from somewhere
1571 else. Bytevectors are very good at this sort of thing. But the SRFI-4
1572 APIs are nicer to use when doing number-crunching, because they are
1573 addressed by element and not by byte.
1575 So as a compromise, Guile allows all bytevector functions to operate on
1576 numeric vectors. They address the underlying bytes in the native
1577 endianness, as one would expect.
1579 Following the same reasoning, that it's just bytes underneath, Guile
1580 also allows uniform vectors of a given type to be accessed as if they
1581 were of any type. One can fill a u32vector, and access its elements with
1582 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
1585 In this way, uniform numeric vectors may be written to and read from
1586 input/output ports using the procedures that operate on bytevectors.
1588 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
1589 inlined to the VM instructions for bytevector access.
1591 See "SRFI-4" in the manual, for more information.
1593 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
1595 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
1596 are part of the standard. Complex uniform vectors and the
1597 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
1599 Guile's default environment imports `(srfi srfi-4)', and probably should
1600 import `(srfi srfi-4 gnu)' as well.
1602 See "SRFI-4 Extensions" in the manual, for more information.
1604 ** New syntax: include-from-path.
1606 `include-from-path' is like `include', except it looks for its file in
1607 the load path. It can be used to compile other files into a file.
1609 ** New syntax: quasisyntax.
1611 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
1612 documentation for more information. Thanks to Andre van Tonder for the
1615 ** `*unspecified*' is identifier syntax
1617 `*unspecified*' is no longer a variable, so it is optimized properly by
1618 the compiler, and is not `set!'-able.
1620 ** Changes and bugfixes in numerics code
1622 *** Added six new sets of fast quotient and remainder operators
1624 Added six new sets of fast quotient and remainder operator pairs with
1625 different semantics than the R5RS operators. They support not only
1626 integers, but all reals, including exact rationals and inexact
1627 floating point numbers.
1629 These procedures accept two real numbers N and D, where the divisor D
1630 must be non-zero. Each set of operators computes an integer quotient
1631 Q and a real remainder R such that N = Q*D + R and |R| < |D|. They
1632 differ only in how N/D is rounded to produce Q.
1634 `euclidean-quotient' returns the integer Q and `euclidean-remainder'
1635 returns the real R such that N = Q*D + R and 0 <= R < |D|. `euclidean/'
1636 returns both Q and R, and is more efficient than computing each
1637 separately. Note that when D > 0, `euclidean-quotient' returns
1638 floor(N/D), and when D < 0 it returns ceiling(N/D).
1640 `centered-quotient', `centered-remainder', and `centered/' are similar
1641 except that the range of remainders is -abs(D/2) <= R < abs(D/2), and
1642 `centered-quotient' rounds N/D to the nearest integer. Note that these
1643 operators are equivalent to the R6RS integer division operators `div',
1644 `mod', `div-and-mod', `div0', `mod0', and `div0-and-mod0'.
1646 `floor-quotient' and `floor-remainder' compute Q and R, respectively,
1647 where Q has been rounded toward negative infinity. `floor/' returns
1648 both Q and R, and is more efficient than computing each separately.
1649 Note that when applied to integers, `floor-remainder' is equivalent to
1650 the R5RS integer-only `modulo' operator. `ceiling-quotient',
1651 `ceiling-remainder', and `ceiling/' are similar except that Q is
1652 rounded toward positive infinity.
1654 For `truncate-quotient', `truncate-remainder', and `truncate/', Q is
1655 rounded toward zero. Note that when applied to integers,
1656 `truncate-quotient' and `truncate-remainder' are equivalent to the
1657 R5RS integer-only operators `quotient' and `remainder'.
1659 For `round-quotient', `round-remainder', and `round/', Q is rounded to
1660 the nearest integer, with ties going to the nearest even integer.
1662 *** Complex number changes
1664 Guile is now able to represent non-real complex numbers whose
1665 imaginary part is an _inexact_ zero (0.0 or -0.0), per R6RS.
1666 Previously, such numbers were immediately changed into inexact reals.
1668 (real? 0.0+0.0i) now returns #f, per R6RS, although (zero? 0.0+0.0i)
1669 still returns #t, per R6RS. (= 0 0.0+0.0i) and (= 0.0 0.0+0.0i) are
1670 #t, but the same comparisons using `eqv?' or `equal?' are #f.
1672 Like other non-real numbers, these complex numbers with inexact zero
1673 imaginary part will raise exceptions is passed to procedures requiring
1674 reals, such as `<', `>', `<=', `>=', `min', `max', `positive?',
1675 `negative?', `inf?', `nan?', `finite?', etc.
1677 **** `make-rectangular' changes
1679 scm_make_rectangular `make-rectangular' now returns a real number only
1680 if the imaginary part is an _exact_ 0. Previously, it would return a
1681 real number if the imaginary part was an inexact zero.
1683 scm_c_make_rectangular now always returns a non-real complex number,
1684 even if the imaginary part is zero. Previously, it would return a
1685 real number if the imaginary part was zero.
1687 **** `make-polar' changes
1689 scm_make_polar `make-polar' now returns a real number only if the
1690 angle or magnitude is an _exact_ 0. If the magnitude is an exact 0,
1691 it now returns an exact 0. Previously, it would return a real
1692 number if the imaginary part was an inexact zero.
1694 scm_c_make_polar now always returns a non-real complex number, even if
1695 the imaginary part is 0.0. Previously, it would return a real number
1696 if the imaginary part was 0.0.
1698 **** `imag-part' changes
1700 scm_imag_part `imag-part' now returns an exact 0 if applied to an
1701 inexact real number. Previously it returned an inexact zero in this
1704 *** `eqv?' and `equal?' now compare numbers equivalently
1706 scm_equal_p `equal?' now behaves equivalently to scm_eqv_p `eqv?' for
1707 numeric values, per R5RS. Previously, equal? worked differently,
1708 e.g. `(equal? 0.0 -0.0)' returned #t but `(eqv? 0.0 -0.0)' returned #f,
1709 and `(equal? +nan.0 +nan.0)' returned #f but `(eqv? +nan.0 +nan.0)'
1712 *** `(equal? +nan.0 +nan.0)' now returns #t
1714 Previously, `(equal? +nan.0 +nan.0)' returned #f, although
1715 `(let ((x +nan.0)) (equal? x x))' and `(eqv? +nan.0 +nan.0)'
1716 both returned #t. R5RS requires that `equal?' behave like
1717 `eqv?' when comparing numbers.
1719 *** Change in handling products `*' involving exact 0
1721 scm_product `*' now handles exact 0 differently. A product containing
1722 an exact 0 now returns an exact 0 if and only if the other arguments
1723 are all exact. An inexact zero is returned if and only if the other
1724 arguments are all finite but not all exact. If an infinite or NaN
1725 value is present, a NaN value is returned. Previously, any product
1726 containing an exact 0 yielded an exact 0, regardless of the other
1729 *** `expt' and `integer-expt' changes when the base is 0
1731 While `(expt 0 0)' is still 1, and `(expt 0 N)' for N > 0 is still
1732 zero, `(expt 0 N)' for N < 0 is now a NaN value, and likewise for
1733 integer-expt. This is more correct, and conforming to R6RS, but seems
1734 to be incompatible with R5RS, which would return 0 for all non-zero
1737 *** `expt' and `integer-expt' are more generic, less strict
1739 When raising to an exact non-negative integer exponent, `expt' and
1740 `integer-expt' are now able to exponentiate any object that can be
1741 multiplied using `*'. They can also raise an object to an exact
1742 negative integer power if its reciprocal can be taken using `/'.
1743 In order to allow this, the type of the first argument is no longer
1744 checked when raising to an exact integer power. If the exponent is 0
1745 or 1, the first parameter is not manipulated at all, and need not
1746 even support multiplication.
1748 *** Infinities are no longer integers, nor rationals
1750 scm_integer_p `integer?' and scm_rational_p `rational?' now return #f
1751 for infinities, per R6RS. Previously they returned #t for real
1752 infinities. The real infinities and NaNs are still considered real by
1753 scm_real `real?' however, per R6RS.
1755 *** NaNs are no longer rationals
1757 scm_rational_p `rational?' now returns #f for NaN values, per R6RS.
1758 Previously it returned #t for real NaN values. They are still
1759 considered real by scm_real `real?' however, per R6RS.
1761 *** `inf?' and `nan?' now throw exceptions for non-reals
1763 The domain of `inf?' and `nan?' is the real numbers. Guile now signals
1764 an error when a non-real number or non-number is passed to these
1765 procedures. (Note that NaNs _are_ considered numbers by scheme, despite
1768 *** `rationalize' bugfixes and changes
1770 Fixed bugs in scm_rationalize `rationalize'. Previously, it returned
1771 exact integers unmodified, although that was incorrect if the epsilon
1772 was at least 1 or inexact, e.g. (rationalize 4 1) should return 3 per
1773 R5RS and R6RS, but previously it returned 4. It also now handles
1774 cases involving infinities and NaNs properly, per R6RS.
1776 *** Trigonometric functions now return exact numbers in some cases
1778 scm_sin `sin', scm_cos `cos', scm_tan `tan', scm_asin `asin', scm_acos
1779 `acos', scm_atan `atan', scm_sinh `sinh', scm_cosh `cosh', scm_tanh
1780 `tanh', scm_sys_asinh `asinh', scm_sys_acosh `acosh', and
1781 scm_sys_atanh `atanh' now return exact results in some cases.
1783 *** New procedure: `finite?'
1785 Add scm_finite_p `finite?' from R6RS to guile core, which returns #t
1786 if and only if its argument is neither infinite nor a NaN. Note that
1787 this is not the same as (not (inf? x)) or (not (infinite? x)), since
1788 NaNs are neither finite nor infinite.
1790 *** Improved exactness handling for complex number parsing
1792 When parsing non-real complex numbers, exactness specifiers are now
1793 applied to each component, as is done in PLT Scheme. For complex
1794 numbers written in rectangular form, exactness specifiers are applied
1795 to the real and imaginary parts before calling scm_make_rectangular.
1796 For complex numbers written in polar form, exactness specifiers are
1797 applied to the magnitude and angle before calling scm_make_polar.
1799 Previously, exactness specifiers were applied to the number as a whole
1800 _after_ calling scm_make_rectangular or scm_make_polar.
1802 For example, (string->number "#i5.0+0i") now does the equivalent of:
1804 (make-rectangular (exact->inexact 5.0) (exact->inexact 0))
1806 which yields 5.0+0.0i. Previously it did the equivalent of:
1808 (exact->inexact (make-rectangular 5.0 0))
1812 ** Unicode characters
1814 Unicode characters may be entered in octal format via e.g. `#\454', or
1815 created via (integer->char 300). A hex external representation will
1816 probably be introduced at some point.
1820 Internally, strings are now represented either in the `latin-1'
1821 encoding, one byte per character, or in UTF-32, with four bytes per
1822 character. Strings manage their own allocation, switching if needed.
1824 Extended characters may be written in a literal string using the
1825 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
1826 or 24-bit codepoints, respectively, or entered directly in the native
1827 encoding of the port on which the string is read.
1831 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
1833 ** Support for non-ASCII source code files
1835 The default reader now handles source code files for some of the
1836 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
1837 should have an encoding declaration near the top of the file. Also,
1838 there is a new function, `file-encoding', that scans a port for a coding
1839 declaration. See the section of the manual entitled, "Character Encoding
1842 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
1843 code. This use is now discouraged. Binary input and output is
1844 currently supported by opening ports in the ISO-8859-1 locale.
1846 ** Source files default to UTF-8.
1848 If source files do not specify their encoding via a `coding:' block,
1849 the default encoding is UTF-8, instead of being taken from the current
1852 ** Interactive Guile installs the current locale.
1854 Instead of leaving the user in the "C" locale, running the Guile REPL
1855 installs the current locale. [FIXME xref?]
1857 ** Support for locale transcoding when reading from and writing to ports
1859 Ports now have an associated character encoding, and port read and write
1860 operations do conversion to and from locales automatically. Ports also
1861 have an associated strategy for how to deal with locale conversion
1864 See the documentation in the manual for the four new support functions,
1865 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
1866 and `port-conversion-strategy'.
1868 ** String and SRFI-13 functions can operate on Unicode strings
1870 ** Unicode support for SRFI-14 character sets
1872 The default character sets are no longer locale dependent and contain
1873 characters from the whole Unicode range. There is a new predefined
1874 character set, `char-set:designated', which contains all assigned
1875 Unicode characters. There is a new debugging function, `%char-set-dump'.
1877 ** Character functions operate on Unicode characters
1879 `char-upcase' and `char-downcase' use default Unicode casing rules.
1880 Character comparisons such as `char<?' and `char-ci<?' now sort based on
1881 Unicode code points.
1883 ** Global variables `scm_charnames' and `scm_charnums' are removed
1885 These variables contained the names of control characters and were
1886 used when writing characters. While these were global, they were
1887 never intended to be public API. They have been replaced with private
1890 ** EBCDIC support is removed
1892 There was an EBCDIC compile flag that altered some of the character
1893 processing. It appeared that full EBCDIC support was never completed
1894 and was unmaintained.
1896 ** Compile-time warnings
1898 Guile can warn about potentially unbound free variables. Pass the
1899 -Wunbound-variable on the `guile-tools compile' command line, or add
1900 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
1901 invocation. Warnings are also enabled by default for expressions entered
1904 Guile can also warn when you pass the wrong number of arguments to a
1905 procedure, with -Warity-mismatch, or `arity-mismatch' in the
1906 `#:warnings' as above.
1908 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
1909 warn about unused local or global (top-level) variables, and `-Wformat',
1910 to check for various errors related to the `format' procedure.
1912 ** A new `memoize-symbol' evaluator trap has been added.
1914 This trap can be used for efficiently implementing a Scheme code
1917 ** Duplicate bindings among used modules are resolved lazily.
1919 This slightly improves program startup times.
1921 ** New thread cancellation and thread cleanup API
1923 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
1925 ** New threads are in `(guile-user)' by default, not `(guile)'
1927 It used to be that a new thread entering Guile would do so in the
1928 `(guile)' module, unless this was the first time Guile was initialized,
1929 in which case it was `(guile-user)'. This has been fixed to have all
1930 new threads unknown to Guile default to `(guile-user)'.
1932 ** New helpers: `print-exception', `set-exception-printer!'
1934 These functions implement an extensible exception printer. Guile
1935 registers printers for all of the exceptions it throws. Users may add
1936 their own printers. There is also `scm_print_exception', for use by C
1937 programs. Pleasantly, this allows SRFI-35 and R6RS exceptions to be
1938 printed appropriately.
1940 ** GOOPS dispatch in scheme
1942 As an implementation detail, GOOPS dispatch is no longer implemented by
1943 special evaluator bytecodes, but rather directly via a Scheme function
1944 associated with an applicable struct. There is some VM support for the
1945 underlying primitives, like `class-of'.
1947 This change will in the future allow users to customize generic function
1948 dispatch without incurring a performance penalty, and allow us to
1949 implement method combinations.
1951 ** Applicable struct support
1953 One may now make structs from Scheme that may be applied as procedures.
1954 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
1955 That struct will be the vtable of your applicable structs; instances of
1956 that new struct are assumed to have the procedure in their first slot.
1957 `<applicable-struct-vtable>' is like Common Lisp's
1958 `funcallable-standard-class'. Likewise there is
1959 `<applicable-struct-with-setter-vtable>', which looks for the setter in
1960 the second slot. This needs to be better documented.
1964 GOOPS had a number of concepts that were relevant to the days of Tcl,
1965 but not any more: operators and entities, mainly. These objects were
1966 never documented, and it is unlikely that they were ever used. Operators
1967 were a kind of generic specific to the Tcl support. Entities were
1968 replaced by applicable structs, mentioned above.
1970 ** New struct slot allocation: "hidden"
1972 A hidden slot is readable and writable, but will not be initialized by a
1973 call to make-struct. For example in your layout you would say "ph"
1974 instead of "pw". Hidden slots are useful for adding new slots to a
1975 vtable without breaking existing invocations to make-struct.
1977 ** eqv? not a generic
1979 One used to be able to extend `eqv?' as a primitive-generic, but no
1980 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
1981 should be able to compile to static dispatch tables, it doesn't make
1982 sense to allow extensions that would subvert this optimization.
1984 ** `inet-ntop' and `inet-pton' are always available.
1986 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
1987 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
1988 functions are deprecated.
1990 ** `getopt-long' parsing errors throw to `quit', not `misc-error'
1992 This change should inhibit backtraces on argument parsing errors.
1993 `getopt-long' has been modified to print out the error that it throws
1996 ** New primitive: `tmpfile'.
1998 See "File System" in the manual.
2000 ** Random generator state may be serialized to a datum
2002 `random-state->datum' will serialize a random state to a datum, which
2003 may be written out, read back in later, and revivified using
2004 `datum->random-state'. See "Random" in the manual, for more details.
2006 ** Fix random number generator on 64-bit platforms
2008 There was a nasty bug on 64-bit platforms in which asking for a random
2009 integer with a range between 2**32 and 2**64 caused a segfault. After
2010 many embarrassing iterations, this was fixed.
2012 ** Fast bit operations.
2014 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
2015 have dedicated bytecodes. Guile is not just for symbolic computation,
2016 it's for number crunching too.
2018 ** Faster SRFI-9 record access
2020 SRFI-9 records are now implemented directly on top of Guile's structs,
2021 and their accessors are defined in such a way that normal call-sites
2022 inline to special VM opcodes, while still allowing for the general case
2023 (e.g. passing a record accessor to `apply').
2025 ** R6RS block comment support
2027 Guile now supports R6RS nested block comments. The start of a comment is
2028 marked with `#|', and the end with `|#'.
2030 ** `guile-2' cond-expand feature
2032 To test if your code is running under Guile 2.0 (or its alpha releases),
2033 test for the `guile-2' cond-expand feature. Like this:
2035 (cond-expand (guile-2 (eval-when (compile)
2036 ;; This must be evaluated at compile time.
2037 (fluid-set! current-reader my-reader)))
2039 ;; Earlier versions of Guile do not have a
2040 ;; separate compilation phase.
2041 (fluid-set! current-reader my-reader)))
2043 ** New global variables: %load-compiled-path, %load-compiled-extensions
2045 These are analogous to %load-path and %load-extensions.
2047 ** New fluid: `%file-port-name-canonicalization'
2049 This fluid parameterizes the file names that are associated with file
2050 ports. If %file-port-name-canonicalization is 'absolute, then file names
2051 are canonicalized to be absolute paths. If it is 'relative, then the
2052 name is canonicalized, but any prefix corresponding to a member of
2053 `%load-path' is stripped off. Otherwise the names are passed through
2056 In addition, the `compile-file' and `compile-and-load' procedures bind
2057 %file-port-name-canonicalization to their `#:canonicalization' keyword
2058 argument, which defaults to 'relative. In this way, one might compile
2059 "../module/ice-9/boot-9.scm", but the path that gets residualized into
2060 the .go is "ice-9/boot-9.scm".
2062 ** New procedure, `make-promise'
2064 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
2066 ** `defined?' may accept a module as its second argument
2068 Previously it only accepted internal structures from the evaluator.
2070 ** New entry into %guile-build-info: `ccachedir'
2072 ** Fix bug in `module-bound?'.
2074 `module-bound?' was returning true if a module did have a local
2075 variable, but one that was unbound, but another imported module bound
2076 the variable. This was an error, and was fixed.
2078 ** `(ice-9 syncase)' has been deprecated.
2080 As syntax-case is available by default, importing `(ice-9 syncase)' has
2081 no effect, and will trigger a deprecation warning.
2083 ** New readline history functions
2085 The (ice-9 readline) module now provides add-history, read-history,
2086 write-history and clear-history, which wrap the corresponding GNU
2087 History library functions.
2089 ** Removed deprecated uniform array procedures:
2090 dimensions->uniform-array, list->uniform-array, array-prototype
2092 Instead, use make-typed-array, list->typed-array, or array-type,
2095 ** Deprecate the old `scm-style-repl'
2097 The following bindings from boot-9 are now found in `(ice-9
2098 scm-style-repl)': `scm-style-repl', `error-catching-loop',
2099 `error-catching-repl', `bad-throw', `scm-repl-silent'
2100 `assert-repl-silence', `repl-print-unspecified',
2101 `assert-repl-print-unspecified', `scm-repl-verbose',
2102 `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl',
2103 `default-pre-unwind-handler', `handle-system-error',
2105 The following bindings have been deprecated, with no replacement:
2106 `pre-unwind-handler-dispatch'.
2108 The following bindings have been totally removed:
2109 `before-signal-stack'.
2111 Deprecated forwarding shims have been installed so that users that
2112 expect these bindings in the main namespace will still work, but receive
2113 a deprecation warning.
2115 ** `set-batch-mode?!' replaced by `ensure-batch-mode!'
2117 "Batch mode" is a flag used to tell a program that it is not running
2118 interactively. One usually turns it on after a fork. It may not be
2119 turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!',
2120 because it is a better interface, as it can only turn on batch mode, not
2123 ** Deprecate `save-stack', `the-last-stack'
2125 It used to be that the way to debug programs in Guile was to capture the
2126 stack at the time of error, drop back to the REPL, then debug that
2127 stack. But this approach didn't compose, was tricky to get right in the
2128 presence of threads, and was not very powerful.
2130 So `save-stack', `stack-saved?', and `the-last-stack' have been moved to
2131 `(ice-9 save-stack)', with deprecated bindings left in the root module.
2133 ** `top-repl' has its own module
2135 The `top-repl' binding, called with Guile is run interactively, is now
2136 is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was
2137 left in the default environment.
2139 ** `display-error' takes a frame
2141 The `display-error' / `scm_display_error' helper now takes a frame as an
2142 argument instead of a stack. Stacks are still supported in deprecated
2143 builds. Additionally, `display-error' will again source location
2144 information for the error.
2146 ** No more `(ice-9 debug)'
2148 This module had some debugging helpers that are no longer applicable to
2149 the current debugging model. Importing this module will produce a
2150 deprecation warning. Users should contact bug-guile for support.
2152 ** Remove obsolete debug-options
2154 Removed `breakpoints', `trace', `procnames', `indent', `frames',
2155 `maxdepth', and `debug' debug-options.
2157 ** `backtrace' debug option on by default
2159 Given that Guile 2.0 can always give you a backtrace, backtraces are now
2162 ** `turn-on-debugging' deprecated
2164 ** Remove obsolete print-options
2166 The `source' and `closure-hook' print options are obsolete, and have
2169 ** Remove obsolete read-options
2171 The "elisp-strings" and "elisp-vectors" read options were unused and
2172 obsolete, so they have been removed.
2174 ** Remove eval-options and trap-options
2176 Eval-options and trap-options are obsolete with the new VM and
2179 ** Remove (ice-9 debugger) and (ice-9 debugging)
2181 See "Traps" and "Interactive Debugging" in the manual, for information
2182 on their replacements.
2184 ** Remove the GDS Emacs integration
2186 See "Using Guile in Emacs" in the manual, for info on how we think you
2187 should use Guile with Emacs.
2189 ** Deprecated: `lazy-catch'
2191 `lazy-catch' was a form that captured the stack at the point of a
2192 `throw', but the dynamic state at the point of the `catch'. It was a bit
2193 crazy. Please change to use `catch', possibly with a throw-handler, or
2194 `with-throw-handler'.
2196 ** Deprecated: primitive properties
2198 The `primitive-make-property', `primitive-property-set!',
2199 `primitive-property-ref', and `primitive-property-del!' procedures were
2200 crufty and only used to implement object properties, which has a new,
2201 threadsafe implementation. Use object properties or weak hash tables
2204 ** Deprecated `@bind' syntax
2206 `@bind' was part of an older implementation of the Emacs Lisp language,
2207 and is no longer used.
2209 ** Miscellaneous other deprecations
2211 `cuserid' has been deprecated, as it only returns 8 bytes of a user's
2212 login. Use `(passwd:name (getpwuid (geteuid)))' instead.
2214 Additionally, the procedures `apply-to-args', `has-suffix?', `scheme-file-suffix'
2215 `get-option', `for-next-option', `display-usage-report',
2216 `transform-usage-lambda', `collect', and `set-batch-mode?!' have all
2219 ** Add support for unbound fluids
2221 See `make-unbound-fluid', `fluid-unset!', and `fluid-bound?' in the
2224 ** Add `variable-unset!'
2226 See "Variables" in the manual, for more details.
2228 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
2230 * Changes to the C interface
2232 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
2234 The semantics of `scm_gc_malloc ()' have been changed, in a
2235 backward-compatible way. A new allocation routine,
2236 `scm_gc_malloc_pointerless ()', was added.
2238 Libgc is a conservative GC, which we hope will make interaction with C
2239 code easier and less error-prone.
2241 ** New procedures: `scm_to_stringn', `scm_from_stringn'
2242 ** New procedures: scm_{to,from}_{utf8,latin1}_symbol{n,}
2243 ** New procedures: scm_{to,from}_{utf8,utf32,latin1}_string{n,}
2245 These new procedures convert to and from string representations in
2246 particular encodings.
2248 Users should continue to use locale encoding for user input, user
2249 output, or interacting with the C library.
2251 Use the Latin-1 functions for ASCII, and for literals in source code.
2253 Use UTF-8 functions for interaction with modern libraries which deal in
2254 UTF-8, and UTF-32 for interaction with utf32-using libraries.
2256 Otherwise, use scm_to_stringn or scm_from_stringn with a specific
2259 ** New type definitions for `scm_t_intptr' and friends.
2261 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
2262 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
2263 available to C. Have fun!
2265 ** The GH interface (deprecated in version 1.6, 2001) was removed.
2267 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
2269 This makes these internal functions technically not callable from
2272 ** Functions for handling `scm_option' now no longer require an argument
2273 indicating length of the `scm_t_option' array.
2275 ** Procedures-with-setters are now implemented using applicable structs
2277 From a user's perspective this doesn't mean very much. But if, for some
2278 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
2279 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
2282 ** Remove old evaluator closures
2284 There used to be ranges of typecodes allocated to interpreted data
2285 structures, but that it no longer the case, given that interpreted
2286 procedure are now just regular VM closures. As a result, there is a
2287 newly free tc3, and a number of removed macros. See the ChangeLog for
2290 ** Primitive procedures are now VM trampoline procedures
2292 It used to be that there were something like 12 different typecodes
2293 allocated to primitive procedures, each with its own calling convention.
2294 Now there is only one, the gsubr. This may affect user code if you were
2295 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
2296 solution is to switch to use scm_c_make_gsubr. This solution works well
2297 both with the old 1.8 and with the current 1.9 branch.
2299 Guile's old evaluator used to have special cases for applying "gsubrs",
2300 primitive procedures with specified numbers of required, optional, and
2301 rest arguments. Now, however, Guile represents gsubrs as normal VM
2302 procedures, with appropriate bytecode to parse out the correct number of
2303 arguments, including optional and rest arguments, and then with a
2304 special bytecode to apply the gsubr.
2306 This allows primitive procedures to appear on the VM stack, allowing
2307 them to be accurately counted in profiles. Also they now have more
2308 debugging information attached to them -- their number of arguments, for
2309 example. In addition, the VM can completely inline the application
2310 mechanics, allowing for faster primitive calls.
2312 However there are some changes on the C level. There is no more
2313 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
2314 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
2315 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
2316 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
2317 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
2319 Perhaps more significantly, `scm_c_make_subr',
2320 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
2321 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
2322 and there are no more subr typecodes. Use the scm_c_make_gsubr family
2325 Normal users of gsubrs should not be affected, though, as the
2326 scm_c_make_gsubr family still is the correct way to create primitive
2329 ** Remove deprecated array C interfaces
2331 Removed the deprecated array functions `scm_i_arrayp',
2332 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
2333 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
2334 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
2335 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
2337 ** Remove unused snarf macros
2339 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
2340 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
2342 ** New functions: `scm_call_n', `scm_c_run_hookn'
2344 `scm_call_n' applies to apply a function to an array of arguments.
2345 `scm_c_run_hookn' runs a hook with an array of arguments.
2347 ** Some SMOB types changed to have static typecodes
2349 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
2350 they have statically allocated tc7 typecodes.
2352 ** Preparations for changing SMOB representation
2354 If things go right, we'll be changing the SMOB representation soon. To
2355 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
2356 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
2357 in the future. Code accessing SMOBs using SCM_CELL macros was never
2358 correct, but until now things still worked. Users should be aware of
2361 ** Changed invocation mechanics of applicable SMOBs
2363 Guile's old evaluator used to have special cases for applying SMOB
2364 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
2365 trampoline procedure for it, and use the normal mechanics to apply the
2366 trampoline. This simplifies procedure application in the normal,
2369 The upshot is that the mechanics used to apply a SMOB are different from
2370 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
2371 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
2372 deprecated. Just use the scm_call_0 family of procedures.
2374 ** Removed support shlibs for SRFIs 1, 4, 13, 14, and 60
2376 Though these SRFI support libraries did expose API, they encoded a
2377 strange version string into their library names. That version was never
2378 programmatically exported, so there was no way people could use the
2381 This was a fortunate oversight, as it allows us to remove the need for
2382 extra, needless shared libraries --- the C support code for SRFIs 4, 13,
2383 and 14 was already in core --- and allow us to incrementally return the
2384 SRFI implementation to Scheme.
2386 ** New C function: scm_module_public_interface
2388 This procedure corresponds to Scheme's `module-public-interface'.
2390 ** Undeprecate `scm_the_root_module ()'
2392 It's useful to be able to get the root module from C without doing a
2395 ** Inline vector allocation
2397 Instead of having vectors point out into the heap for their data, their
2398 data is now allocated inline to the vector object itself. The same is
2399 true for bytevectors, by default, though there is an indirection
2400 available which should allow for making a bytevector from an existing
2403 ** New struct constructors that don't involve making lists
2405 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
2406 constructors, respectively, for structs. You might find them useful.
2410 In Guile 1.8, there were debugging frames on the C stack. Now there is
2411 no more need to explicitly mark the stack in this way, because Guile has
2412 a VM stack that it knows how to walk, which simplifies the C API
2413 considerably. See the ChangeLog for details; the relevant interface is
2414 in libguile/stacks.h. The Scheme API has not been changed significantly.
2416 ** Removal of Guile's primitive object system.
2418 There were a number of pieces in `objects.[ch]' that tried to be a
2419 minimal object system, but were never documented, and were quickly
2420 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
2421 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
2422 from objects.h are no more. In the very unlikely case in which these
2423 were useful to you, we urge you to contact guile-devel.
2427 Actually the future is still in the state that it was, is, and ever
2428 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
2429 part of it. These files were experimental, never compiled, and would be
2430 better implemented in Scheme anyway. In the future, that is.
2432 ** Deprecate trampolines
2434 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
2435 so on. The point was to do some precomputation on the type of the
2436 procedure, then return a specialized "call" procedure. However this
2437 optimization wasn't actually an optimization, so it is now deprecated.
2438 Just use `scm_call_0', etc instead.
2440 ** Deprecated `scm_badargsp'
2442 This function is unused in Guile, but was part of its API.
2444 ** Better support for Lisp `nil'.
2446 The bit representation of `nil' has been tweaked so that it is now very
2447 efficient to check e.g. if a value is equal to Scheme's end-of-list or
2448 Lisp's nil. Additionally there are a heap of new, specific predicates
2449 like scm_is_null_or_nil.
2451 ** Better integration of Lisp `nil'.
2453 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
2454 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
2455 but when we start to integrate more with Emacs, it is possible that we
2456 break code that assumes that, for example, `(not x)' implies that `x' is
2457 `eq?' to `#f'. This is not a common assumption. Refactoring affected
2458 code to rely on properties instead of identities will improve code
2459 correctness. See "Nil" in the manual, for more details.
2461 ** Support for static allocation of strings, symbols, and subrs.
2463 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
2464 much of their associated data as static variables, reducing Guile's
2467 ** `scm_stat' has an additional argument, `exception_on_error'
2468 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
2470 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
2472 Previously they would use the `off_t' type, which is fragile since its
2473 definition depends on the application's value for `_FILE_OFFSET_BITS'.
2475 ** The `long_long' C type, deprecated in 1.8, has been removed
2477 ** Removed deprecated uniform array procedures: scm_make_uve,
2478 scm_array_prototype, scm_list_to_uniform_array,
2479 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
2480 scm_ra_set_contp, scm_aind, scm_raprin1
2482 These functions have been deprecated since early 2005.
2484 * Changes to the distribution
2486 ** Guile's license is now LGPLv3+
2488 In other words the GNU Lesser General Public License, version 3 or
2489 later (at the discretion of each person that chooses to redistribute
2494 Guile's build is visually quieter, due to the use of Automake 1.11's
2495 AM_SILENT_RULES. Build as `make V=1' to see all of the output.
2497 ** GOOPS documentation folded into Guile reference manual
2499 GOOPS, Guile's object system, used to be documented in separate manuals.
2500 This content is now included in Guile's manual directly.
2502 ** `guile-config' will be deprecated in favor of `pkg-config'
2504 `guile-config' has been rewritten to get its information from
2505 `pkg-config', so this should be a transparent change. Note however that
2506 guile.m4 has yet to be modified to call pkg-config instead of
2509 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
2511 Programs that use `pkg-config' to find Guile or one of its Autoconf
2512 macros should now require `guile-2.0' instead of `guile-1.8'.
2514 ** New installation directory: $(pkglibdir)/1.9/ccache
2516 If $(libdir) is /usr/lib, for example, Guile will install its .go files
2517 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
2519 ** Parallel installability fixes
2521 Guile now installs its header files to a effective-version-specific
2522 directory, and includes the effective version (e.g. 2.0) in the library
2523 name (e.g. libguile-2.0.so).
2525 This change should be transparent to users, who should detect Guile via
2526 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
2527 parallel installs for multiple versions of Guile development
2530 ** Dynamically loadable extensions may be placed in a Guile-specific path
2532 Before, Guile only searched the system library paths for extensions
2533 (e.g. /usr/lib), which meant that the names of Guile extensions had to
2534 be globally unique. Installing them to a Guile-specific extensions
2535 directory is cleaner. Use `pkg-config --variable=extensiondir
2536 guile-2.0' to get the location of the extensions directory.
2538 ** User Scheme code may be placed in a version-specific path
2540 Before, there was only one way to install user Scheme code to a
2541 version-specific Guile directory: install to Guile's own path,
2542 e.g. /usr/share/guile/2.0. The site directory,
2543 e.g. /usr/share/guile/site, was unversioned. This has been changed to
2544 add a version-specific site directory, e.g. /usr/share/guile/site/2.0,
2545 searched before the global site directory.
2547 ** New dependency: libgc
2549 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
2551 ** New dependency: GNU libunistring
2553 See http://www.gnu.org/software/libunistring/, for more information. Our
2554 Unicode support uses routines from libunistring.
2556 ** New dependency: libffi
2558 See http://sourceware.org/libffi/, for more information.
2562 Changes in 1.8.8 (since 1.8.7)
2566 ** Fix possible buffer overruns when parsing numbers
2567 ** Avoid clash with system setjmp/longjmp on IA64
2568 ** Fix `wrong type arg' exceptions with IPv6 addresses
2571 Changes in 1.8.7 (since 1.8.6)
2573 * New modules (see the manual for details)
2575 ** `(srfi srfi-98)', an interface to access environment variables
2579 ** Fix compilation with `--disable-deprecated'
2580 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
2581 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
2582 ** Fix build problem when scm_t_timespec is different from struct timespec
2583 ** Fix build when compiled with -Wundef -Werror
2584 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
2585 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
2586 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
2587 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
2588 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
2589 ** Fix reading of complex numbers where both parts are inexact decimals
2591 ** Allow @ macro to work with (ice-9 syncase)
2593 Previously, use of the @ macro in a module whose code is being
2594 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
2595 Now it works as you would expect (giving the value of the specified
2598 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
2601 Changes in 1.8.6 (since 1.8.5)
2603 * New features (see the manual for details)
2605 ** New convenience function `scm_c_symbol_length ()'
2607 ** Single stepping through code from Emacs
2609 When you use GDS to evaluate Scheme code from Emacs, you can now use
2610 `C-u' to indicate that you want to single step through that code. See
2611 `Evaluating Scheme Code' in the manual for more details.
2613 ** New "guile(1)" man page!
2615 * Changes to the distribution
2617 ** Automake's `AM_MAINTAINER_MODE' is no longer used
2619 Thus, the `--enable-maintainer-mode' configure option is no longer
2620 available: Guile is now always configured in "maintainer mode".
2622 ** `ChangeLog' files are no longer updated
2624 Instead, changes are detailed in the version control system's logs. See
2625 the top-level `ChangeLog' files for details.
2630 ** `symbol->string' now returns a read-only string, as per R5RS
2631 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
2632 ** `guile-config link' now prints `-L$libdir' before `-lguile'
2633 ** Fix memory corruption involving GOOPS' `class-redefinition'
2634 ** Fix possible deadlock in `mutex-lock'
2635 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
2636 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
2637 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
2638 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
2639 ** Fix misleading output from `(help rationalize)'
2640 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
2641 ** Fix `gcd' when called with a single, negative argument.
2642 ** Fix `Stack overflow' errors seen when building on some platforms
2643 ** Fix bug when `scm_with_guile ()' was called several times from the
2645 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
2646 dynamic environment of the call to `raise'
2647 ** Fix potential deadlock in `make-struct'
2648 ** Fix compilation problem with libltdl from Libtool 2.2.x
2649 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
2652 Changes in 1.8.5 (since 1.8.4)
2654 * Infrastructure changes
2656 ** Guile repository switched from CVS to Git
2658 The new repository can be accessed using
2659 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
2660 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
2662 ** Add support for `pkg-config'
2664 See "Autoconf Support" in the manual for details.
2666 * New modules (see the manual for details)
2670 * New features (see the manual for details)
2672 ** New `postfix' read option, for SRFI-88 keyword syntax
2673 ** Some I/O primitives have been inlined, which improves I/O performance
2674 ** New object-based traps infrastructure
2676 This is a GOOPS-based infrastructure that builds on Guile's low-level
2677 evaluator trap calls and facilitates the development of debugging
2678 features like single-stepping, breakpoints, tracing and profiling.
2679 See the `Traps' node of the manual for details.
2681 ** New support for working on Guile code from within Emacs
2683 Guile now incorporates the `GDS' library (previously distributed
2684 separately) for working on Guile code from within Emacs. See the
2685 `Using Guile In Emacs' node of the manual for details.
2689 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
2690 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
2692 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
2693 would trigger an unbound variable error for `match:andmap'.
2695 ** `(oop goops describe)' now properly provides the `describe' feature
2696 ** Fixed `args-fold' from `(srfi srfi-37)'
2698 Previously, parsing short option names of argument-less options would
2699 lead to a stack overflow.
2701 ** `(srfi srfi-35)' is now visible through `cond-expand'
2702 ** Fixed type-checking for the second argument of `eval'
2703 ** Fixed type-checking for SRFI-1 `partition'
2704 ** Fixed `struct-ref' and `struct-set!' on "light structs"
2705 ** Honor struct field access rights in GOOPS
2706 ** Changed the storage strategy of source properties, which fixes a deadlock
2707 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
2708 ** Fixed build issue for GNU/Linux on IA64
2709 ** Fixed build issues on NetBSD 1.6
2710 ** Fixed build issue on Solaris 2.10 x86_64
2711 ** Fixed build issue with DEC/Compaq/HP's compiler
2712 ** Fixed `scm_from_complex_double' build issue on FreeBSD
2713 ** Fixed `alloca' build issue on FreeBSD 6
2714 ** Removed use of non-portable makefile constructs
2715 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
2716 ** Make sure all tests honor `$TMPDIR'
2719 Changes in 1.8.4 (since 1.8.3)
2723 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
2724 ** Fixed a segmentation fault which occurred when displaying the
2725 backtrace of a stack with a promise object (made by `delay') in it.
2726 ** Make `accept' leave guile mode while blocking
2727 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
2728 ** Fixed a build problem on AIX (use of func_data identifier)
2729 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
2730 called with an associator proc that returns neither a pair nor #f.
2731 ** Secondary threads now always return a valid module for (current-module).
2732 ** Avoid MacOS build problems caused by incorrect combination of "64"
2733 system and library calls.
2734 ** `guile-snarf' now honors `$TMPDIR'
2735 ** `guile-config compile' now reports CPPFLAGS used at compile-time
2736 ** Fixed build with Sun Studio (Solaris 9)
2737 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
2738 uniform vectors on AIX.
2739 ** Fixed a deadlock that occurs upon GC with multiple threads.
2740 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
2741 ** Fixed autotool-derived build problems on AIX 6.1.
2742 ** Fixed NetBSD/alpha support
2743 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
2745 * New modules (see the manual for details)
2749 * Documentation fixes and improvements
2751 ** Removed premature breakpoint documentation
2753 The features described are not available in the series of 1.8.x
2754 releases, so the documentation was misleading and has been removed.
2756 ** More about Guile's default *random-state* variable
2758 ** GOOPS: more about how to use `next-method'
2760 * Changes to the distribution
2762 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
2764 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
2765 General Public License, and the few incorrect files have now been
2766 fixed to agree with the rest of the Guile distribution.
2768 ** Removed unnecessary extra copies of COPYING*
2770 The distribution now contains a single COPYING.LESSER at its top level.
2773 Changes in 1.8.3 (since 1.8.2)
2775 * New modules (see the manual for details)
2782 ** The `(ice-9 slib)' module now works as expected
2783 ** Expressions like "(set! 'x #t)" no longer yield a crash
2784 ** Warnings about duplicate bindings now go to stderr
2785 ** A memory leak in `make-socket-address' was fixed
2786 ** Alignment issues (e.g., on SPARC) in network routines were fixed
2787 ** A threading issue that showed up at least on NetBSD was fixed
2788 ** Build problems on Solaris and IRIX fixed
2790 * Implementation improvements
2792 ** The reader is now faster, which reduces startup time
2793 ** Procedures returned by `record-accessor' and `record-modifier' are faster
2796 Changes in 1.8.2 (since 1.8.1):
2798 * New procedures (see the manual for details)
2800 ** set-program-arguments
2803 * Incompatible changes
2805 ** The body of a top-level `define' no longer sees the binding being created
2807 In a top-level `define', the binding being created is no longer visible
2808 from the `define' body. This breaks code like
2809 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
2810 unbound in the body. However, such code was not R5RS-compliant anyway,
2815 ** Fractions were not `equal?' if stored in unreduced form.
2816 (A subtle problem, since printing a value reduced it, making it work.)
2817 ** srfi-60 `copy-bit' failed on 64-bit systems
2818 ** "guile --use-srfi" option at the REPL can replace core functions
2819 (Programs run with that option were ok, but in the interactive REPL
2820 the core bindings got priority, preventing SRFI replacements or
2822 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
2823 ** `kill' on mingw throws an error for a PID other than oneself
2824 ** Procedure names are attached to procedure-with-setters
2825 ** Array read syntax works with negative lower bound
2826 ** `array-in-bounds?' fix if an array has different lower bounds on each index
2827 ** `*' returns exact 0 for "(* inexact 0)"
2828 This follows what it always did for "(* 0 inexact)".
2829 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
2830 ** SRFI-19: `date->julian-day' did not account for timezone offset
2831 ** `ttyname' no longer crashes when passed a non-tty argument
2832 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
2833 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
2834 ** GOOPS: Fixed a bug in `method-more-specific?'
2835 ** Build problems on Solaris fixed
2836 ** Build problems on HP-UX IA64 fixed
2837 ** Build problems on MinGW fixed
2840 Changes in 1.8.1 (since 1.8.0):
2842 * LFS functions are now used to access 64-bit files on 32-bit systems.
2844 * New procedures (see the manual for details)
2846 ** primitive-_exit - [Scheme] the-root-module
2847 ** scm_primitive__exit - [C]
2848 ** make-completion-function - [Scheme] (ice-9 readline)
2849 ** scm_c_locale_stringn_to_number - [C]
2850 ** scm_srfi1_append_reverse [C]
2851 ** scm_srfi1_append_reverse_x [C]
2859 ** Build problems have been fixed on MacOS, SunOS, and QNX.
2861 ** `strftime' fix sign of %z timezone offset.
2863 ** A one-dimensional array can now be 'equal?' to a vector.
2865 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
2867 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
2869 ** `record-accessor' and `record-modifier' now have strict type checks.
2871 Record accessor and modifier procedures now throw an error if the
2872 record type of the record they're given is not the type expected.
2873 (Previously accessors returned #f and modifiers silently did nothing).
2875 ** It is now OK to use both autoload and use-modules on a given module.
2877 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
2879 Previously there was no checking on primatives like make-vector that
2880 accept "one or two" arguments. Now there is.
2882 ** The srfi-1 assoc function now calls its equality predicate properly.
2884 Previously srfi-1 assoc would call the equality predicate with the key
2885 last. According to the SRFI, the key should be first.
2887 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
2889 ** The array-set! procedure no longer segfaults when given a bit vector.
2891 ** Bugs in make-shared-array have been fixed.
2893 ** string<? and friends now follow char<? etc order on 8-bit chars.
2895 ** The format procedure now handles inf and nan values for ~f correctly.
2897 ** exact->inexact should no longer overflow when given certain large fractions.
2899 ** srfi-9 accessor and modifier procedures now have strict record type checks.
2901 This matches the srfi-9 specification.
2903 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
2905 Previously the (ice-9 ftw) procedures would ignore any file that had
2906 the same inode number as a file they had already seen, even if that
2907 file was on a different device.
2910 Changes in 1.8.0 (changes since the 1.6.x series):
2912 * Changes to the distribution
2914 ** Guile is now licensed with the GNU Lesser General Public License.
2916 ** The manual is now licensed with the GNU Free Documentation License.
2918 ** Guile now requires GNU MP (http://swox.com/gmp).
2920 Guile now uses the GNU MP library for arbitrary precision arithmetic.
2922 ** Guile now has separate private and public configuration headers.
2924 That is, things like HAVE_STRING_H no longer leak from Guile's
2927 ** Guile now provides and uses an "effective" version number.
2929 Guile now provides scm_effective_version and effective-version
2930 functions which return the "effective" version number. This is just
2931 the normal full version string without the final micro-version number,
2932 so the current effective-version is "1.8". The effective version
2933 should remain unchanged during a stable series, and should be used for
2934 items like the versioned share directory name
2935 i.e. /usr/share/guile/1.8.
2937 Providing an unchanging version number during a stable release for
2938 things like the versioned share directory can be particularly
2939 important for Guile "add-on" packages, since it provides a directory
2940 that they can install to that won't be changed out from under them
2941 with each micro release during a stable series.
2943 ** Thread implementation has changed.
2945 When you configure "--with-threads=null", you will get the usual
2946 threading API (call-with-new-thread, make-mutex, etc), but you can't
2947 actually create new threads. Also, "--with-threads=no" is now
2948 equivalent to "--with-threads=null". This means that the thread API
2949 is always present, although you might not be able to create new
2952 When you configure "--with-threads=pthreads" or "--with-threads=yes",
2953 you will get threads that are implemented with the portable POSIX
2954 threads. These threads can run concurrently (unlike the previous
2955 "coop" thread implementation), but need to cooperate for things like
2958 The default is "pthreads", unless your platform doesn't have pthreads,
2959 in which case "null" threads are used.
2961 See the manual for details, nodes "Initialization", "Multi-Threading",
2962 "Blocking", and others.
2964 ** There is the new notion of 'discouraged' features.
2966 This is a milder form of deprecation.
2968 Things that are discouraged should not be used in new code, but it is
2969 OK to leave them in old code for now. When a discouraged feature is
2970 used, no warning message is printed like there is for 'deprecated'
2971 features. Also, things that are merely discouraged are nevertheless
2972 implemented efficiently, while deprecated features can be very slow.
2974 You can omit discouraged features from libguile by configuring it with
2975 the '--disable-discouraged' option.
2977 ** Deprecation warnings can be controlled at run-time.
2979 (debug-enable 'warn-deprecated) switches them on and (debug-disable
2980 'warn-deprecated) switches them off.
2982 ** Support for SRFI 61, extended cond syntax for multiple values has
2985 This SRFI is always available.
2987 ** Support for require-extension, SRFI-55, has been added.
2989 The SRFI-55 special form `require-extension' has been added. It is
2990 available at startup, and provides a portable way to load Scheme
2991 extensions. SRFI-55 only requires support for one type of extension,
2992 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
2995 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
2997 The (srfi srfi-26) module is an implementation of SRFI-26 which
2998 provides the `cut' and `cute' syntax. These may be used to specialize
2999 parameters without currying.
3001 ** New module (srfi srfi-31)
3003 This is an implementation of SRFI-31 which provides a special form
3004 `rec' for recursive evaluation.
3006 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
3007 been merged with the core, making their functionality always
3010 The modules are still available, tho, and you could use them together
3011 with a renaming import, for example.
3013 ** Guile no longer includes its own version of libltdl.
3015 The official version is good enough now.
3017 ** The --enable-htmldoc option has been removed from 'configure'.
3019 Support for translating the documentation into HTML is now always
3020 provided. Use 'make html'.
3022 ** New module (ice-9 serialize):
3024 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
3025 don't trust the thread safety of most of your program, but where you
3026 have some section(s) of code which you consider can run in parallel to
3027 other sections. See ice-9/serialize.scm for more information.
3029 ** The configure option '--disable-arrays' has been removed.
3031 Support for arrays and uniform numeric arrays is now always included
3034 * Changes to the stand-alone interpreter
3036 ** New command line option `-L'.
3038 This option adds a directory to the front of the load path.
3040 ** New command line option `--no-debug'.
3042 Specifying `--no-debug' on the command line will keep the debugging
3043 evaluator turned off, even for interactive sessions.
3045 ** User-init file ~/.guile is now loaded with the debugging evaluator.
3047 Previously, the normal evaluator would have been used. Using the
3048 debugging evaluator gives better error messages.
3050 ** The '-e' option now 'read's its argument.
3052 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
3053 be used with '-e'. For example, you can now write a script like
3056 exec guile -e '(@ (demo) main)' -s "$0" "$@"
3059 (define-module (demo)
3063 (format #t "Demo: ~a~%" args))
3066 * Changes to Scheme functions and syntax
3068 ** Guardians have changed back to their original semantics
3070 Guardians now behave like described in the paper by Dybvig et al. In
3071 particular, they no longer make guarantees about the order in which
3072 they return objects, and they can no longer be greedy.
3074 They no longer drop cyclic data structures.
3076 The C function scm_make_guardian has been changed incompatibly and no
3077 longer takes the 'greedy_p' argument.
3079 ** New function hashx-remove!
3081 This function completes the set of 'hashx' functions.
3083 ** The concept of dynamic roots has been factored into continuation
3084 barriers and dynamic states.
3086 Each thread has a current dynamic state that carries the values of the
3087 fluids. You can create and copy dynamic states and use them as the
3088 second argument for 'eval'. See "Fluids and Dynamic States" in the
3091 To restrict the influence that captured continuations can have on the
3092 control flow, you can errect continuation barriers. See "Continuation
3093 Barriers" in the manual.
3095 The function call-with-dynamic-root now essentially temporarily
3096 installs a new dynamic state and errects a continuation barrier.
3098 ** The default load path no longer includes "." at the end.
3100 Automatically loading modules from the current directory should not
3101 happen by default. If you want to allow it in a more controlled
3102 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
3103 variable %load-path.
3105 ** The uniform vector and array support has been overhauled.
3107 It now complies with SRFI-4 and the weird prototype based uniform
3108 array creation has been deprecated. See the manual for more details.
3110 Some non-compatible changes have been made:
3111 - characters can no longer be stored into byte arrays.
3112 - strings and bit vectors are no longer considered to be uniform numeric
3114 - array-rank throws an error for non-arrays instead of returning zero.
3115 - array-ref does no longer accept non-arrays when no indices are given.
3117 There is the new notion of 'generalized vectors' and corresponding
3118 procedures like 'generalized-vector-ref'. Generalized vectors include
3119 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
3121 Arrays use generalized vectors as their storage, so that you still
3122 have arrays of characters, bits, etc. However, uniform-array-read!
3123 and uniform-array-write can no longer read/write strings and
3126 ** There is now support for copy-on-write substrings, mutation-sharing
3127 substrings and read-only strings.
3129 Three new procedures are related to this: substring/shared,
3130 substring/copy, and substring/read-only. See the manual for more
3133 ** Backtraces will now highlight the value that caused the error.
3135 By default, these values are enclosed in "{...}", such as in this
3144 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
3145 <unnamed port>:1:1: Wrong type (expecting pair): a
3146 ABORT: (wrong-type-arg)
3148 The prefix and suffix used for highlighting can be set via the two new
3149 printer options 'highlight-prefix' and 'highlight-suffix'. For
3150 example, putting this into ~/.guile will output the bad value in bold
3151 on an ANSI terminal:
3153 (print-set! highlight-prefix "\x1b[1m")
3154 (print-set! highlight-suffix "\x1b[22m")
3157 ** 'gettext' support for internationalization has been added.
3159 See the manual for details.
3161 ** New syntax '@' and '@@':
3163 You can now directly refer to variables exported from a module by
3166 (@ MODULE-NAME VARIABLE-NAME)
3168 For example (@ (ice-9 pretty-print) pretty-print) will directly access
3169 the pretty-print variable exported from the (ice-9 pretty-print)
3170 module. You don't need to 'use' that module first. You can also use
3171 '@' as a target of 'set!', as in (set! (@ mod var) val).
3173 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
3174 but it can also access variables that have not been exported. It is
3175 intended only for kluges and temporary fixes and for debugging, not
3178 ** Keyword syntax has been made more disciplined.
3180 Previously, the name of a keyword was read as a 'token' but printed as
3181 a symbol. Now, it is read as a general Scheme datum which must be a
3192 ERROR: In expression (a b c):
3196 ERROR: Unbound variable: foo
3201 ERROR: Wrong type (expecting symbol): 12
3205 ERROR: Wrong type (expecting symbol): (a b c)
3209 ** The printing of symbols that might look like keywords can be
3212 The new printer option 'quote-keywordish-symbols' controls how symbols
3213 are printed that have a colon as their first or last character. The
3214 default now is to only quote a symbol with #{...}# when the read
3215 option 'keywords' is not '#f'. Thus:
3217 guile> (define foo (string->symbol ":foo"))
3218 guile> (read-set! keywords #f)
3221 guile> (read-set! keywords 'prefix)
3224 guile> (print-set! quote-keywordish-symbols #f)
3228 ** 'while' now provides 'break' and 'continue'
3230 break and continue were previously bound in a while loop, but not
3231 documented, and continue didn't quite work properly. The undocumented
3232 parameter to break which gave a return value for the while has been
3235 ** 'call-with-current-continuation' is now also available under the name
3238 ** The module system now checks for duplicate bindings.
3240 The module system now can check for name conflicts among imported
3243 The behavior can be controlled by specifying one or more 'duplicates'
3244 handlers. For example, to make Guile return an error for every name
3247 (define-module (foo)
3252 The new default behavior of the module system when a name collision
3253 has been detected is to
3255 1. Give priority to bindings marked as a replacement.
3256 2. Issue a warning (different warning if overriding core binding).
3257 3. Give priority to the last encountered binding (this corresponds to
3260 If you want the old behavior back without replacements or warnings you
3263 (default-duplicate-binding-handler 'last)
3265 to your .guile init file.
3267 ** New define-module option: :replace
3269 :replace works as :export, but, in addition, marks the binding as a
3272 A typical example is `format' in (ice-9 format) which is a replacement
3273 for the core binding `format'.
3275 ** Adding prefixes to imported bindings in the module system
3277 There is now a new :use-module option :prefix. It can be used to add
3278 a prefix to all imported bindings.
3280 (define-module (foo)
3281 :use-module ((bar) :prefix bar:))
3283 will import all bindings exported from bar, but rename them by adding
3286 ** Conflicting generic functions can be automatically merged.
3288 When two imported bindings conflict and they are both generic
3289 functions, the two functions can now be merged automatically. This is
3290 activated with the 'duplicates' handler 'merge-generics'.
3292 ** New function: effective-version
3294 Returns the "effective" version number. This is just the normal full
3295 version string without the final micro-version number. See "Changes
3296 to the distribution" above.
3298 ** New threading functions: parallel, letpar, par-map, and friends
3300 These are convenient ways to run calculations in parallel in new
3301 threads. See "Parallel forms" in the manual for details.
3303 ** New function 'try-mutex'.
3305 This function will attempt to lock a mutex but will return immediately
3306 instead of blocking and indicate failure.
3308 ** Waiting on a condition variable can have a timeout.
3310 The function 'wait-condition-variable' now takes a third, optional
3311 argument that specifies the point in time where the waiting should be
3314 ** New function 'broadcast-condition-variable'.
3316 ** New functions 'all-threads' and 'current-thread'.
3318 ** Signals and system asyncs work better with threads.
3320 The function 'sigaction' now takes a fourth, optional, argument that
3321 specifies the thread that the handler should run in. When the
3322 argument is omitted, the handler will run in the thread that called
3325 Likewise, 'system-async-mark' takes a second, optional, argument that
3326 specifies the thread that the async should run in. When it is
3327 omitted, the async will run in the thread that called
3328 'system-async-mark'.
3330 C code can use the new functions scm_sigaction_for_thread and
3331 scm_system_async_mark_for_thread to pass the new thread argument.
3333 When a thread blocks on a mutex, a condition variable or is waiting
3334 for IO to be possible, it will still execute system asyncs. This can
3335 be used to interrupt such a thread by making it execute a 'throw', for
3338 ** The function 'system-async' is deprecated.
3340 You can now pass any zero-argument procedure to 'system-async-mark'.
3341 The function 'system-async' will just return its argument unchanged
3344 ** New functions 'call-with-blocked-asyncs' and
3345 'call-with-unblocked-asyncs'
3347 The expression (call-with-blocked-asyncs PROC) will call PROC and will
3348 block execution of system asyncs for the current thread by one level
3349 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
3350 procedure and will unblock the execution of system asyncs by one
3351 level for the current thread.
3353 Only system asyncs are affected by these functions.
3355 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
3357 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
3358 instead. Those functions are easier to use correctly and can be
3361 ** New function 'unsetenv'.
3363 ** New macro 'define-syntax-public'.
3365 It works like 'define-syntax' and also exports the defined macro (but
3368 ** There is support for Infinity and NaNs.
3370 Following PLT Scheme, Guile can now work with infinite numbers, and
3373 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
3374 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
3375 "+nan.0"). These numbers are inexact and have no exact counterpart.
3377 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
3378 sign of the dividend. The infinities are integers, and they answer #t
3379 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
3380 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
3391 ERROR: Numerical overflow
3393 Two new predicates 'inf?' and 'nan?' can be used to test for the
3396 ** Inexact zero can have a sign.
3398 Guile can now distinguish between plus and minus inexact zero, if your
3399 platform supports this, too. The two zeros are equal according to
3400 '=', but not according to 'eqv?'. For example
3411 ** Guile now has exact rationals.
3413 Guile can now represent fractions such as 1/3 exactly. Computing with
3414 them is also done exactly, of course:
3419 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
3420 for exact arguments.
3422 For example: (floor 2) now returns an exact 2 where in the past it
3423 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
3425 ** inexact->exact no longer returns only integers.
3427 Without exact rationals, the closest exact number was always an
3428 integer, but now inexact->exact returns the fraction that is exactly
3429 equal to a floating point number. For example:
3431 (inexact->exact 1.234)
3432 => 694680242521899/562949953421312
3434 When you want the old behavior, use 'round' explicitly:
3436 (inexact->exact (round 1.234))
3439 ** New function 'rationalize'.
3441 This function finds a simple fraction that is close to a given real
3442 number. For example (and compare with inexact->exact above):
3444 (rationalize (inexact->exact 1.234) 1/2000)
3447 Note that, as required by R5RS, rationalize returns only then an exact
3448 result when both its arguments are exact.
3450 ** 'odd?' and 'even?' work also for inexact integers.
3452 Previously, (odd? 1.0) would signal an error since only exact integers
3453 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
3454 returns #f and (odd? 1.5) signals an error.
3456 ** Guile now has uninterned symbols.
3458 The new function 'make-symbol' will return an uninterned symbol. This
3459 is a symbol that is unique and is guaranteed to remain unique.
3460 However, uninterned symbols can not yet be read back in.
3462 Use the new function 'symbol-interned?' to check whether a symbol is
3465 ** pretty-print has more options.
3467 The function pretty-print from the (ice-9 pretty-print) module can now
3468 also be invoked with keyword arguments that control things like
3469 maximum output width. See the manual for details.
3471 ** Variables have no longer a special behavior for `equal?'.
3473 Previously, comparing two variables with `equal?' would recursivly
3474 compare their values. This is no longer done. Variables are now only
3475 `equal?' if they are `eq?'.
3477 ** `(begin)' is now valid.
3479 You can now use an empty `begin' form. It will yield #<unspecified>
3480 when evaluated and simply be ignored in a definition context.
3482 ** Deprecated: procedure->macro
3484 Change your code to use 'define-macro' or r5rs macros. Also, be aware
3485 that macro expansion will not be done during evaluation, but prior to
3488 ** Soft ports now allow a `char-ready?' procedure
3490 The vector argument to `make-soft-port' can now have a length of
3491 either 5 or 6. (Previously the length had to be 5.) The optional 6th
3492 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
3493 that returns the number of characters that can be read immediately
3494 without the soft port blocking.
3496 ** Deprecated: undefine
3498 There is no replacement for undefine.
3500 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
3501 have been discouraged.
3503 They are relics from a time where a keyword like #:foo was used
3504 directly as a Tcl option "-foo" and thus keywords were internally
3505 stored as a symbol with a starting dash. We now store a symbol
3508 Use symbol->keyword and keyword->symbol instead.
3510 ** The `cheap' debug option is now obsolete
3512 Evaluator trap calls are now unconditionally "cheap" - in other words,
3513 they pass a debug object to the trap handler rather than a full
3514 continuation. The trap handler code can capture a full continuation
3515 by using `call-with-current-continuation' in the usual way, if it so
3518 The `cheap' option is retained for now so as not to break existing
3519 code which gets or sets it, but setting it now has no effect. It will
3520 be removed in the next major Guile release.
3522 ** Evaluator trap calls now support `tweaking'
3524 `Tweaking' means that the trap handler code can modify the Scheme
3525 expression that is about to be evaluated (in the case of an
3526 enter-frame trap) or the value that is being returned (in the case of
3527 an exit-frame trap). The trap handler code indicates that it wants to
3528 do this by returning a pair whose car is the symbol 'instead and whose
3529 cdr is the modified expression or return value.
3531 * Changes to the C interface
3533 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
3534 take a 'delete' function argument.
3536 This argument makes no sense since the delete function is used to
3537 remove a pair from an alist, and this must not be configurable.
3539 This is an incompatible change.
3541 ** The GH interface is now subject to the deprecation mechanism
3543 The GH interface has been deprecated for quite some time but now it is
3544 actually removed from Guile when it is configured with
3545 --disable-deprecated.
3547 See the manual "Transitioning away from GH" for more information.
3549 ** A new family of functions for converting between C values and
3550 Scheme values has been added.
3552 These functions follow a common naming scheme and are designed to be
3553 easier to use, thread-safe and more future-proof than the older
3556 - int scm_is_* (...)
3558 These are predicates that return a C boolean: 1 or 0. Instead of
3559 SCM_NFALSEP, you can now use scm_is_true, for example.
3561 - <type> scm_to_<type> (SCM val, ...)
3563 These are functions that convert a Scheme value into an appropriate
3564 C value. For example, you can use scm_to_int to safely convert from
3567 - SCM scm_from_<type> (<type> val, ...)
3569 These functions convert from a C type to a SCM value; for example,
3570 scm_from_int for ints.
3572 There is a huge number of these functions, for numbers, strings,
3573 symbols, vectors, etc. They are documented in the reference manual in
3574 the API section together with the types that they apply to.
3576 ** New functions for dealing with complex numbers in C have been added.
3578 The new functions are scm_c_make_rectangular, scm_c_make_polar,
3579 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
3580 They work like scm_make_rectangular etc but take or return doubles
3583 ** The function scm_make_complex has been discouraged.
3585 Use scm_c_make_rectangular instead.
3587 ** The INUM macros have been deprecated.
3589 A lot of code uses these macros to do general integer conversions,
3590 although the macros only work correctly with fixnums. Use the
3591 following alternatives.
3593 SCM_INUMP -> scm_is_integer or similar
3594 SCM_NINUMP -> !scm_is_integer or similar
3595 SCM_MAKINUM -> scm_from_int or similar
3596 SCM_INUM -> scm_to_int or similar
3598 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
3599 do the validating for you.
3601 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
3602 have been discouraged.
3604 Use the newer scm_to_<type> and scm_from_<type> functions instead for
3605 new code. The functions have been discouraged since they don't fit
3608 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
3610 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
3611 evaluates its argument twice. Use scm_is_true, etc. instead for new
3614 ** The macro SCM_EQ_P has been discouraged.
3616 Use scm_is_eq for new code, which fits better into the naming
3619 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
3622 Use the function scm_is_pair or scm_is_null instead.
3624 ** The functions scm_round and scm_truncate have been deprecated and
3625 are now available as scm_c_round and scm_c_truncate, respectively.
3627 These functions occupy the names that scm_round_number and
3628 scm_truncate_number should have.
3630 ** The functions scm_c_string2str, scm_c_substring2str, and
3631 scm_c_symbol2str have been deprecated.
3633 Use scm_to_locale_stringbuf or similar instead, maybe together with
3636 ** New functions scm_c_make_string, scm_c_string_length,
3637 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
3638 scm_c_substring_shared, scm_c_substring_copy.
3640 These are like scm_make_string, scm_length, etc. but are slightly
3641 easier to use from C.
3643 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
3644 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
3646 They export too many assumptions about the implementation of strings
3647 and symbols that are no longer true in the presence of
3648 mutation-sharing substrings and when Guile switches to some form of
3651 When working with strings, it is often best to use the normal string
3652 functions provided by Guile, such as scm_c_string_ref,
3653 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
3654 manual since many more such functions are now provided than
3657 When you want to convert a SCM string to a C string, use the
3658 scm_to_locale_string function or similar instead. For symbols, use
3659 scm_symbol_to_string and then work with that string. Because of the
3660 new string representation, scm_symbol_to_string does not need to copy
3661 and is thus quite efficient.
3663 ** Some string, symbol and keyword functions have been discouraged.
3665 They don't fit into the uniform naming scheme and are not explicit
3666 about the character encoding.
3668 Replace according to the following table:
3670 scm_allocate_string -> scm_c_make_string
3671 scm_take_str -> scm_take_locale_stringn
3672 scm_take0str -> scm_take_locale_string
3673 scm_mem2string -> scm_from_locale_stringn
3674 scm_str2string -> scm_from_locale_string
3675 scm_makfrom0str -> scm_from_locale_string
3676 scm_mem2symbol -> scm_from_locale_symboln
3677 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
3678 scm_str2symbol -> scm_from_locale_symbol
3680 SCM_SYMBOL_HASH -> scm_hashq
3681 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
3683 scm_c_make_keyword -> scm_from_locale_keyword
3685 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
3686 now also available to C code.
3688 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
3690 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
3691 the latter returns the true name of the keyword, not the 'dash name',
3692 as SCM_KEYWORDSYM used to do.
3694 ** A new way to access arrays in a thread-safe and efficient way has
3697 See the manual, node "Accessing Arrays From C".
3699 ** The old uniform vector and bitvector implementations have been
3700 unceremoniously removed.
3702 This implementation exposed the details of the tagging system of
3703 Guile. Use the new C API explained in the manual in node "Uniform
3704 Numeric Vectors" and "Bit Vectors", respectively.
3706 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
3707 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
3708 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
3709 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
3710 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
3711 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
3714 ** The macros dealing with vectors have been deprecated.
3716 Use the new functions scm_is_vector, scm_vector_elements,
3717 scm_vector_writable_elements, etc, or scm_is_simple_vector,
3718 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
3719 manual for more details.
3721 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
3722 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
3724 The following macros have been removed: SCM_VECTOR_BASE,
3725 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
3726 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
3728 ** Some C functions and macros related to arrays have been deprecated.
3730 Migrate according to the following table:
3732 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
3733 scm_make_ra -> scm_make_array
3734 scm_shap2ra -> scm_make_array
3735 scm_cvref -> scm_c_generalized_vector_ref
3736 scm_ra_set_contp -> do not use
3737 scm_aind -> scm_array_handle_pos
3738 scm_raprin1 -> scm_display or scm_write
3740 SCM_ARRAYP -> scm_is_array
3741 SCM_ARRAY_NDIM -> scm_c_array_rank
3742 SCM_ARRAY_DIMS -> scm_array_handle_dims
3743 SCM_ARRAY_CONTP -> do not use
3744 SCM_ARRAY_MEM -> do not use
3745 SCM_ARRAY_V -> scm_array_handle_elements or similar
3746 SCM_ARRAY_BASE -> do not use
3748 ** SCM_CELL_WORD_LOC has been deprecated.
3750 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
3751 to a SCM, as opposed to a pointer to a scm_t_bits.
3753 This was done to allow the correct use of pointers into the Scheme
3754 heap. Previously, the heap words were of type scm_t_bits and local
3755 variables and function arguments were of type SCM, making it
3756 non-standards-conformant to have a pointer that can point to both.
3758 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
3760 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
3761 second and third words of double smobs. Likewise for
3762 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
3764 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
3765 used to get and set the 16 exra bits in the zeroth word of a smob.
3767 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
3768 accesing the first immediate word of a smob as a SCM value, and there
3769 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
3770 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
3772 ** New way to deal with non-local exits and re-entries.
3774 There is a new set of functions that essentially do what
3775 scm_internal_dynamic_wind does, but in a way that is more convenient
3776 for C code in some situations. Here is a quick example of how to
3777 prevent a potential memory leak:
3784 scm_dynwind_begin (0);
3786 mem = scm_malloc (100);
3787 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
3789 /* MEM would leak if BAR throws an error.
3790 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
3797 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
3798 SCM_DYNWIND_END as well.
3802 For full documentation, see the node "Dynamic Wind" in the manual.
3804 ** New function scm_dynwind_free
3806 This function calls 'free' on a given pointer when a dynwind context
3807 is left. Thus the call to scm_dynwind_unwind_handler above could be
3808 replaced with simply scm_dynwind_free (mem).
3810 ** New functions scm_c_call_with_blocked_asyncs and
3811 scm_c_call_with_unblocked_asyncs
3813 Like scm_call_with_blocked_asyncs etc. but for C functions.
3815 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
3817 In addition to scm_c_call_with_blocked_asyncs you can now also use
3818 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
3819 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
3821 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
3822 SCM_REALLOW_INTS have been deprecated.
3824 They do no longer fulfill their original role of blocking signal
3825 delivery. Depending on what you want to achieve, replace a pair of
3826 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
3827 mutex, blocks asyncs, or both. See node "Critical Sections" in the
3830 ** The value 'scm_mask_ints' is no longer writable.
3832 Previously, you could set scm_mask_ints directly. This is no longer
3833 possible. Use scm_c_call_with_blocked_asyncs and
3834 scm_c_call_with_unblocked_asyncs instead.
3836 ** New way to temporarily set the current input, output or error ports
3838 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
3839 context' (see above). <foo> is one of "input", "output" or "error".
3841 ** New way to temporarily set fluids
3843 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
3844 above) to temporarily set the value of a fluid.
3846 ** New types scm_t_intmax and scm_t_uintmax.
3848 On platforms that have them, these types are identical to intmax_t and
3849 uintmax_t, respectively. On other platforms, they are identical to
3850 the largest integer types that Guile knows about.
3852 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
3854 You should not have used them.
3856 ** Many public #defines with generic names have been made private.
3858 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
3859 private or renamed with a more suitable public name.
3861 ** The macro SCM_TYP16S has been deprecated.
3863 This macro is not intended for public use.
3865 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
3867 Use scm_is_true (scm_inexact_p (...)) instead.
3869 ** The macro SCM_SLOPPY_REALP has been deprecated.
3871 Use scm_is_real instead.
3873 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
3875 Use scm_is_complex instead.
3877 ** Some preprocessor defines have been deprecated.
3879 These defines indicated whether a certain feature was present in Guile
3880 or not. Going forward, assume that the features are always present.
3882 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
3883 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
3885 The following macros have been removed completely: MEMOIZE_LOCALS,
3886 SCM_RECKLESS, SCM_CAUTIOUS.
3888 ** The preprocessor define STACK_DIRECTION has been deprecated.
3890 There should be no need to know about the stack direction for ordinary
3893 ** New function: scm_effective_version
3895 Returns the "effective" version number. This is just the normal full
3896 version string without the final micro-version number. See "Changes
3897 to the distribution" above.
3899 ** The function scm_call_with_new_thread has a new prototype.
3901 Instead of taking a list with the thunk and handler, these two
3902 arguments are now passed directly:
3904 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
3906 This is an incompatible change.
3908 ** New snarfer macro SCM_DEFINE_PUBLIC.
3910 This is like SCM_DEFINE, but also calls scm_c_export for the defined
3911 function in the init section.
3913 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
3915 ** Garbage collector rewrite.
3917 The garbage collector is cleaned up a lot, and now uses lazy
3918 sweeping. This is reflected in the output of (gc-stats); since cells
3919 are being freed when they are allocated, the cells-allocated field
3920 stays roughly constant.
3922 For malloc related triggers, the behavior is changed. It uses the same
3923 heuristic as the cell-triggered collections. It may be tuned with the
3924 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
3925 for minimum yield of malloc related triggers. The default is 40.
3926 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
3929 Debugging operations for the freelist have been deprecated, along with
3930 the C variables that control garbage collection. The environment
3931 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
3932 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
3934 For understanding the memory usage of a GUILE program, the routine
3935 gc-live-object-stats returns an alist containing the number of live
3936 objects for every type.
3939 ** The function scm_definedp has been renamed to scm_defined_p
3941 The name scm_definedp is deprecated.
3943 ** The struct scm_cell type has been renamed to scm_t_cell
3945 This is in accordance to Guile's naming scheme for types. Note that
3946 the name scm_cell is now used for a function that allocates and
3947 initializes a new cell (see below).
3949 ** New functions for memory management
3951 A new set of functions for memory management has been added since the
3952 old way (scm_must_malloc, scm_must_free, etc) was error prone and
3953 indeed, Guile itself contained some long standing bugs that could
3954 cause aborts in long running programs.
3956 The new functions are more symmetrical and do not need cooperation
3957 from smob free routines, among other improvements.
3959 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
3960 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
3961 scm_gc_free, scm_gc_register_collectable_memory, and
3962 scm_gc_unregister_collectable_memory. Refer to the manual for more
3963 details and for upgrading instructions.
3965 The old functions for memory management have been deprecated. They
3966 are: scm_must_malloc, scm_must_realloc, scm_must_free,
3967 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
3969 ** Declarations of exported features are marked with SCM_API.
3971 Every declaration of a feature that belongs to the exported Guile API
3972 has been marked by adding the macro "SCM_API" to the start of the
3973 declaration. This macro can expand into different things, the most
3974 common of which is just "extern" for Unix platforms. On Win32, it can
3975 be used to control which symbols are exported from a DLL.
3977 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
3978 will expand into "__declspec (dllimport) extern", which is needed for
3979 linking to the Guile DLL in Windows.
3981 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
3982 SCM_SRFI4_IMPORT, for the corresponding libraries.
3984 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
3986 Use the new functions scm_cell and scm_double_cell instead. The old
3987 macros had problems because with them allocation and initialization
3988 was separated and the GC could sometimes observe half initialized
3989 cells. Only careful coding by the user of SCM_NEWCELL and
3990 SCM_NEWCELL2 could make this safe and efficient.
3992 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
3994 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
3997 ** SRCBRKP has been deprecated.
3999 Use scm_c_source_property_breakpoint_p instead.
4001 ** Deprecated: scm_makmacro
4003 Change your code to use either scm_makmmacro or to define macros in
4004 Scheme, using 'define-macro'.
4006 ** New function scm_c_port_for_each.
4008 This function is like scm_port_for_each but takes a pointer to a C
4009 function as the callback instead of a SCM value.
4011 ** The names scm_internal_select, scm_thread_sleep, and
4012 scm_thread_usleep have been discouraged.
4014 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
4016 ** The GC can no longer be blocked.
4018 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
4019 The GC can now run (partially) concurrently with other code and thus
4020 blocking it is not well defined.
4022 ** Many definitions have been removed that were previously deprecated.
4024 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
4025 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
4026 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
4027 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
4028 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
4029 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
4030 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
4031 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
4032 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
4033 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
4034 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
4035 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
4036 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
4037 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
4038 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
4039 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
4040 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
4041 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
4042 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
4043 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
4044 scm_vector_set_length_x, scm_contregs, scm_debug_info,
4045 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
4046 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
4047 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
4048 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
4049 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
4050 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
4051 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
4052 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
4053 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
4054 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
4056 * Changes to bundled modules
4060 Using the (ice-9 debug) module no longer automatically switches Guile
4061 to use the debugging evaluator. If you want to switch to the
4062 debugging evaluator (which is needed for backtrace information if you
4063 hit an error), please add an explicit "(debug-enable 'debug)" to your
4064 code just after the code to use (ice-9 debug).
4067 Changes since Guile 1.4:
4069 * Changes to the distribution
4071 ** A top-level TODO file is included.
4073 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
4075 Guile now always uses three numbers to represent the version,
4076 i.e. "1.6.5". The first number, 1, is the major version number, the
4077 second number, 6, is the minor version number, and the third number,
4078 5, is the micro version number. Changes in major version number
4079 indicate major changes in Guile.
4081 Minor version numbers that are even denote stable releases, and odd
4082 minor version numbers denote development versions (which may be
4083 unstable). The micro version number indicates a minor sub-revision of
4084 a given MAJOR.MINOR release.
4086 In keeping with the new scheme, (minor-version) and scm_minor_version
4087 no longer return everything but the major version number. They now
4088 just return the minor version number. Two new functions
4089 (micro-version) and scm_micro_version have been added to report the
4090 micro version number.
4092 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
4094 ** New preprocessor definitions are available for checking versions.
4096 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
4097 SCM_MICRO_VERSION to the appropriate integer values.
4099 ** Guile now actively warns about deprecated features.
4101 The new configure option `--enable-deprecated=LEVEL' and the
4102 environment variable GUILE_WARN_DEPRECATED control this mechanism.
4103 See INSTALL and README for more information.
4105 ** Guile is much more likely to work on 64-bit architectures.
4107 Guile now compiles and passes "make check" with only two UNRESOLVED GC
4108 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
4109 for the use of a test machine, and thanks to Stefan Jahn for ia64
4112 ** New functions: setitimer and getitimer.
4114 These implement a fairly direct interface to the libc functions of the
4117 ** The #. reader extension is now disabled by default.
4119 For safety reasons, #. evaluation is disabled by default. To
4120 re-enable it, set the fluid read-eval? to #t. For example:
4122 (fluid-set! read-eval? #t)
4124 but make sure you realize the potential security risks involved. With
4125 read-eval? enabled, reading a data file from an untrusted source can
4128 ** New SRFI modules have been added:
4130 SRFI-0 `cond-expand' is now supported in Guile, without requiring
4133 (srfi srfi-1) is a library containing many useful pair- and list-processing
4136 (srfi srfi-2) exports and-let*.
4138 (srfi srfi-4) implements homogeneous numeric vector datatypes.
4140 (srfi srfi-6) is a dummy module for now, since guile already provides
4141 all of the srfi-6 procedures by default: open-input-string,
4142 open-output-string, get-output-string.
4144 (srfi srfi-8) exports receive.
4146 (srfi srfi-9) exports define-record-type.
4148 (srfi srfi-10) exports define-reader-ctor and implements the reader
4151 (srfi srfi-11) exports let-values and let*-values.
4153 (srfi srfi-13) implements the SRFI String Library.
4155 (srfi srfi-14) implements the SRFI Character-Set Library.
4157 (srfi srfi-17) implements setter and getter-with-setter and redefines
4158 some accessor procedures as procedures with getters. (such as car,
4159 cdr, vector-ref etc.)
4161 (srfi srfi-19) implements the SRFI Time/Date Library.
4163 ** New scripts / "executable modules"
4165 Subdirectory "scripts" contains Scheme modules that are packaged to
4166 also be executable as scripts. At this time, these scripts are available:
4175 See README there for more info.
4177 These scripts can be invoked from the shell with the new program
4178 "guile-tools", which keeps track of installation directory for you.
4181 $ guile-tools display-commentary srfi/*.scm
4183 guile-tools is copied to the standard $bindir on "make install".
4185 ** New module (ice-9 stack-catch):
4187 stack-catch is like catch, but saves the current state of the stack in
4188 the fluid the-last-stack. This fluid can be useful when using the
4189 debugger and when re-throwing an error.
4191 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
4193 This has been done to prevent problems on lesser operating systems
4194 that can't tolerate `*'s in file names. The exported macro continues
4195 to be named `and-let*', of course.
4197 On systems that support it, there is also a compatibility module named
4198 (ice-9 and-let*). It will go away in the next release.
4200 ** New modules (oop goops) etc.:
4203 (oop goops describe)
4205 (oop goops active-slot)
4206 (oop goops composite-slot)
4208 The Guile Object Oriented Programming System (GOOPS) has been
4209 integrated into Guile. For further information, consult the GOOPS
4210 manual and tutorial in the `doc' directory.
4212 ** New module (ice-9 rdelim).
4214 This exports the following procedures which were previously defined
4215 in the default environment:
4217 read-line read-line! read-delimited read-delimited! %read-delimited!
4218 %read-line write-line
4220 For backwards compatibility the definitions are still imported into the
4221 default environment in this version of Guile. However you should add:
4223 (use-modules (ice-9 rdelim))
4225 to any program which uses the definitions, since this may change in
4228 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
4229 can be used for similar functionality.
4231 ** New module (ice-9 rw)
4233 This is a subset of the (scsh rw) module from guile-scsh. Currently
4234 it defines two procedures:
4236 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
4238 Read characters from a port or file descriptor into a string STR.
4239 A port must have an underlying file descriptor -- a so-called
4240 fport. This procedure is scsh-compatible and can efficiently read
4243 *** New function: write-string/partial str [port_or_fdes [start [end]]]
4245 Write characters from a string STR to a port or file descriptor.
4246 A port must have an underlying file descriptor -- a so-called
4247 fport. This procedure is mostly compatible and can efficiently
4248 write large strings.
4250 ** New module (ice-9 match)
4252 This module includes Andrew K. Wright's pattern matcher. See
4253 ice-9/match.scm for brief description or
4255 http://www.star-lab.com/wright/code.html
4257 for complete documentation.
4259 ** New module (ice-9 buffered-input)
4261 This module provides procedures to construct an input port from an
4262 underlying source of input that reads and returns its input in chunks.
4263 The underlying input source is a Scheme procedure, specified by the
4264 caller, which the port invokes whenever it needs more input.
4266 This is useful when building an input port whose back end is Readline
4267 or a UI element such as the GtkEntry widget.
4271 The reference and tutorial documentation that was previously
4272 distributed separately, as `guile-doc', is now included in the core
4273 Guile distribution. The documentation consists of the following
4276 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
4279 - The Guile Reference Manual (guile.texi) contains (or is intended to
4280 contain) reference documentation on all aspects of Guile.
4282 - The GOOPS Manual (goops.texi) contains both tutorial-style and
4283 reference documentation for using GOOPS, Guile's Object Oriented
4286 - The Revised^5 Report on the Algorithmic Language Scheme
4289 See the README file in the `doc' directory for more details.
4291 ** There are a couple of examples in the examples/ directory now.
4293 * Changes to the stand-alone interpreter
4295 ** New command line option `--use-srfi'
4297 Using this option, SRFI modules can be loaded on startup and be
4298 available right from the beginning. This makes programming portable
4299 Scheme programs easier.
4301 The option `--use-srfi' expects a comma-separated list of numbers,
4302 each representing a SRFI number to be loaded into the interpreter
4303 before starting evaluating a script file or the REPL. Additionally,
4304 the feature identifier for the loaded SRFIs is recognized by
4305 `cond-expand' when using this option.
4308 $ guile --use-srfi=8,13
4309 guile> (receive (x z) (values 1 2) (+ 1 2))
4311 guile> (string-pad "bla" 20)
4314 ** Guile now always starts up in the `(guile-user)' module.
4316 Previously, scripts executed via the `-s' option would run in the
4317 `(guile)' module and the repl would run in the `(guile-user)' module.
4318 Now every user action takes place in the `(guile-user)' module by
4321 * Changes to Scheme functions and syntax
4323 ** Character classifiers work for non-ASCII characters.
4325 The predicates `char-alphabetic?', `char-numeric?',
4326 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
4327 no longer check whether their arguments are ASCII characters.
4328 Previously, a character would only be considered alphabetic when it
4329 was also ASCII, for example.
4331 ** Previously deprecated Scheme functions have been removed:
4333 tag - no replacement.
4334 fseek - replaced by seek.
4335 list* - replaced by cons*.
4337 ** It's now possible to create modules with controlled environments
4341 (use-modules (ice-9 safe))
4342 (define m (make-safe-module))
4343 ;;; m will now be a module containing only a safe subset of R5RS
4344 (eval '(+ 1 2) m) --> 3
4345 (eval 'load m) --> ERROR: Unbound variable: load
4347 ** Evaluation of "()", the empty list, is now an error.
4349 Previously, the expression "()" evaluated to the empty list. This has
4350 been changed to signal a "missing expression" error. The correct way
4351 to write the empty list as a literal constant is to use quote: "'()".
4353 ** New concept of `Guile Extensions'.
4355 A Guile Extension is just a ordinary shared library that can be linked
4356 at run-time. We found it advantageous to give this simple concept a
4357 dedicated name to distinguish the issues related to shared libraries
4358 from the issues related to the module system.
4360 *** New function: load-extension
4362 Executing (load-extension lib init) is mostly equivalent to
4364 (dynamic-call init (dynamic-link lib))
4366 except when scm_register_extension has been called previously.
4367 Whenever appropriate, you should use `load-extension' instead of
4368 dynamic-link and dynamic-call.
4370 *** New C function: scm_c_register_extension
4372 This function registers a initialization function for use by
4373 `load-extension'. Use it when you don't want specific extensions to
4374 be loaded as shared libraries (for example on platforms that don't
4375 support dynamic linking).
4377 ** Auto-loading of compiled-code modules is deprecated.
4379 Guile used to be able to automatically find and link a shared
4380 library to satisfy requests for a module. For example, the module
4381 `(foo bar)' could be implemented by placing a shared library named
4382 "foo/libbar.so" (or with a different extension) in a directory on the
4385 This has been found to be too tricky, and is no longer supported. The
4386 shared libraries are now called "extensions". You should now write a
4387 small Scheme file that calls `load-extension' to load the shared
4388 library and initialize it explicitly.
4390 The shared libraries themselves should be installed in the usual
4391 places for shared libraries, with names like "libguile-foo-bar".
4393 For example, place this into a file "foo/bar.scm"
4395 (define-module (foo bar))
4397 (load-extension "libguile-foo-bar" "foobar_init")
4399 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
4401 `eval' is now R5RS, that is it takes two arguments.
4402 The second argument is an environment specifier, i.e. either
4404 (scheme-report-environment 5)
4405 (null-environment 5)
4406 (interaction-environment)
4412 ** The module system has been made more disciplined.
4414 The function `eval' will save and restore the current module around
4415 the evaluation of the specified expression. While this expression is
4416 evaluated, `(current-module)' will now return the right module, which
4417 is the module specified as the second argument to `eval'.
4419 A consequence of this change is that `eval' is not particularly
4420 useful when you want allow the evaluated code to change what module is
4421 designated as the current module and have this change persist from one
4422 call to `eval' to the next. The read-eval-print-loop is an example
4423 where `eval' is now inadequate. To compensate, there is a new
4424 function `primitive-eval' that does not take a module specifier and
4425 that does not save/restore the current module. You should use this
4426 function together with `set-current-module', `current-module', etc
4427 when you want to have more control over the state that is carried from
4428 one eval to the next.
4430 Additionally, it has been made sure that forms that are evaluated at
4431 the top level are always evaluated with respect to the current module.
4432 Previously, subforms of top-level forms such as `begin', `case',
4433 etc. did not respect changes to the current module although these
4434 subforms are at the top-level as well.
4436 To prevent strange behavior, the forms `define-module',
4437 `use-modules', `use-syntax', and `export' have been restricted to only
4438 work on the top level. The forms `define-public' and
4439 `defmacro-public' only export the new binding on the top level. They
4440 behave just like `define' and `defmacro', respectively, when they are
4441 used in a lexical environment.
4443 Also, `export' will no longer silently re-export bindings imported
4444 from a used module. It will emit a `deprecation' warning and will
4445 cease to perform any re-export in the next version. If you actually
4446 want to re-export bindings, use the new `re-export' in place of
4447 `export'. The new `re-export' will not make copies of variables when
4448 rexporting them, as `export' did wrongly.
4450 ** Module system now allows selection and renaming of imported bindings
4452 Previously, when using `use-modules' or the `#:use-module' clause in
4453 the `define-module' form, all the bindings (association of symbols to
4454 values) for imported modules were added to the "current module" on an
4455 as-is basis. This has been changed to allow finer control through two
4456 new facilities: selection and renaming.
4458 You can now select which of the imported module's bindings are to be
4459 visible in the current module by using the `:select' clause. This
4460 clause also can be used to rename individual bindings. For example:
4462 ;; import all bindings no questions asked
4463 (use-modules (ice-9 common-list))
4465 ;; import four bindings, renaming two of them;
4466 ;; the current module sees: every some zonk-y zonk-n
4467 (use-modules ((ice-9 common-list)
4469 (remove-if . zonk-y)
4470 (remove-if-not . zonk-n))))
4472 You can also programmatically rename all selected bindings using the
4473 `:renamer' clause, which specifies a proc that takes a symbol and
4474 returns another symbol. Because it is common practice to use a prefix,
4475 we now provide the convenience procedure `symbol-prefix-proc'. For
4478 ;; import four bindings, renaming two of them specifically,
4479 ;; and all four w/ prefix "CL:";
4480 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
4481 (use-modules ((ice-9 common-list)
4483 (remove-if . zonk-y)
4484 (remove-if-not . zonk-n))
4485 :renamer (symbol-prefix-proc 'CL:)))
4487 ;; import four bindings, renaming two of them specifically,
4488 ;; and all four by upcasing.
4489 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
4490 (define (upcase-symbol sym)
4491 (string->symbol (string-upcase (symbol->string sym))))
4493 (use-modules ((ice-9 common-list)
4495 (remove-if . zonk-y)
4496 (remove-if-not . zonk-n))
4497 :renamer upcase-symbol))
4499 Note that programmatic renaming is done *after* individual renaming.
4500 Also, the above examples show `use-modules', but the same facilities are
4501 available for the `#:use-module' clause of `define-module'.
4503 See manual for more info.
4505 ** The semantics of guardians have changed.
4507 The changes are for the most part compatible. An important criterion
4508 was to keep the typical usage of guardians as simple as before, but to
4509 make the semantics safer and (as a result) more useful.
4511 *** All objects returned from guardians are now properly alive.
4513 It is now guaranteed that any object referenced by an object returned
4514 from a guardian is alive. It's now impossible for a guardian to
4515 return a "contained" object before its "containing" object.
4517 One incompatible (but probably not very important) change resulting
4518 from this is that it is no longer possible to guard objects that
4519 indirectly reference themselves (i.e. are parts of cycles). If you do
4520 so accidentally, you'll get a warning.
4522 *** There are now two types of guardians: greedy and sharing.
4524 If you call (make-guardian #t) or just (make-guardian), you'll get a
4525 greedy guardian, and for (make-guardian #f) a sharing guardian.
4527 Greedy guardians are the default because they are more "defensive".
4528 You can only greedily guard an object once. If you guard an object
4529 more than once, once in a greedy guardian and the rest of times in
4530 sharing guardians, then it is guaranteed that the object won't be
4531 returned from sharing guardians as long as it is greedily guarded
4534 Guardians returned by calls to `make-guardian' can now take one more
4535 optional parameter, which says whether to throw an error in case an
4536 attempt is made to greedily guard an object that is already greedily
4537 guarded. The default is true, i.e. throw an error. If the parameter
4538 is false, the guardian invocation returns #t if guarding was
4539 successful and #f if it wasn't.
4541 Also, since greedy guarding is, in effect, a side-effecting operation
4542 on objects, a new function is introduced: `destroy-guardian!'.
4543 Invoking this function on a guardian renders it unoperative and, if
4544 the guardian is greedy, clears the "greedily guarded" property of the
4545 objects that were guarded by it, thus undoing the side effect.
4547 Note that all this hair is hardly very important, since guardian
4548 objects are usually permanent.
4550 ** Continuations created by call-with-current-continuation now accept
4551 any number of arguments, as required by R5RS.
4553 ** New function `issue-deprecation-warning'
4555 This function is used to display the deprecation messages that are
4556 controlled by GUILE_WARN_DEPRECATION as explained in the README.
4559 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
4563 ;; `id' is deprecated. Use `identity' instead.
4568 ** New syntax `begin-deprecated'
4570 When deprecated features are included (as determined by the configure
4571 option --enable-deprecated), `begin-deprecated' is identical to
4572 `begin'. When deprecated features are excluded, it always evaluates
4573 to `#f', ignoring the body forms.
4575 ** New function `make-object-property'
4577 This function returns a new `procedure with setter' P that can be used
4578 to attach a property to objects. When calling P as
4582 where `obj' is any kind of object, it attaches `val' to `obj' in such
4583 a way that it can be retrieved by calling P as
4587 This function will replace procedure properties, symbol properties and
4588 source properties eventually.
4590 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
4592 Instead of #&optional, #&key, etc you should now use #:optional,
4593 #:key, etc. Since #:optional is a keyword, you can write it as just
4594 :optional when (read-set! keywords 'prefix) is active.
4596 The old reader syntax `#&' is still supported, but deprecated. It
4597 will be removed in the next release.
4599 ** New define-module option: pure
4601 Tells the module system not to include any bindings from the root
4606 (define-module (totally-empty-module)
4609 ** New define-module option: export NAME1 ...
4611 Export names NAME1 ...
4613 This option is required if you want to be able to export bindings from
4614 a module which doesn't import one of `define-public' or `export'.
4618 (define-module (foo)
4620 :use-module (ice-9 r5rs)
4623 ;;; Note that we're pure R5RS below this point!
4628 ** New function: object->string OBJ
4630 Return a Scheme string obtained by printing a given object.
4632 ** New function: port? X
4634 Returns a boolean indicating whether X is a port. Equivalent to
4635 `(or (input-port? X) (output-port? X))'.
4637 ** New function: file-port?
4639 Determines whether a given object is a port that is related to a file.
4641 ** New function: port-for-each proc
4643 Apply PROC to each port in the Guile port table in turn. The return
4644 value is unspecified. More specifically, PROC is applied exactly once
4645 to every port that exists in the system at the time PORT-FOR-EACH is
4646 invoked. Changes to the port table while PORT-FOR-EACH is running
4647 have no effect as far as PORT-FOR-EACH is concerned.
4649 ** New function: dup2 oldfd newfd
4651 A simple wrapper for the `dup2' system call. Copies the file
4652 descriptor OLDFD to descriptor number NEWFD, replacing the
4653 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
4654 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
4655 to move away ports which are using NEWFD. The return value is
4658 ** New function: close-fdes fd
4660 A simple wrapper for the `close' system call. Close file
4661 descriptor FD, which must be an integer. Unlike close (*note
4662 close: Ports and File Descriptors.), the file descriptor will be
4663 closed even if a port is using it. The return value is
4666 ** New function: crypt password salt
4668 Encrypts `password' using the standard unix password encryption
4671 ** New function: chroot path
4673 Change the root directory of the running process to `path'.
4675 ** New functions: getlogin, cuserid
4677 Return the login name or the user name of the current effective user
4680 ** New functions: getpriority which who, setpriority which who prio
4682 Get or set the priority of the running process.
4684 ** New function: getpass prompt
4686 Read a password from the terminal, first displaying `prompt' and
4689 ** New function: flock file operation
4691 Set/remove an advisory shared or exclusive lock on `file'.
4693 ** New functions: sethostname name, gethostname
4695 Set or get the hostname of the machine the current process is running
4698 ** New function: mkstemp! tmpl
4700 mkstemp creates a new unique file in the file system and returns a
4701 new buffered port open for reading and writing to the file. TMPL
4702 is a string specifying where the file should be created: it must
4703 end with `XXXXXX' and will be changed in place to return the name
4704 of the temporary file.
4706 ** New function: open-input-string string
4708 Return an input string port which delivers the characters from
4709 `string'. This procedure, together with `open-output-string' and
4710 `get-output-string' implements SRFI-6.
4712 ** New function: open-output-string
4714 Return an output string port which collects all data written to it.
4715 The data can then be retrieved by `get-output-string'.
4717 ** New function: get-output-string
4719 Return the contents of an output string port.
4721 ** New function: identity
4723 Return the argument.
4725 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
4726 are represented in Scheme as integers with normal host byte ordering.
4728 ** New function: inet-pton family address
4730 Convert a printable string network address into an integer. Note that
4731 unlike the C version of this function, the result is an integer with
4732 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4735 (inet-pton AF_INET "127.0.0.1") => 2130706433
4736 (inet-pton AF_INET6 "::1") => 1
4738 ** New function: inet-ntop family address
4740 Convert an integer network address into a printable string. Note that
4741 unlike the C version of this function, the input is an integer with
4742 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4745 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
4746 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
4747 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
4751 Use `identity' instead.
4757 ** Deprecated: return-it
4761 ** Deprecated: string-character-length
4763 Use `string-length' instead.
4765 ** Deprecated: flags
4767 Use `logior' instead.
4769 ** Deprecated: close-all-ports-except.
4771 This was intended for closing ports in a child process after a fork,
4772 but it has the undesirable side effect of flushing buffers.
4773 port-for-each is more flexible.
4775 ** The (ice-9 popen) module now attempts to set up file descriptors in
4776 the child process from the current Scheme ports, instead of using the
4777 current values of file descriptors 0, 1, and 2 in the parent process.
4779 ** Removed function: builtin-weak-bindings
4781 There is no such concept as a weak binding any more.
4783 ** Removed constants: bignum-radix, scm-line-incrementors
4785 ** define-method: New syntax mandatory.
4787 The new method syntax is now mandatory:
4789 (define-method (NAME ARG-SPEC ...) BODY ...)
4790 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
4792 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
4793 REST-ARG ::= ARG-NAME
4795 If you have old code using the old syntax, import
4796 (oop goops old-define-method) before (oop goops) as in:
4798 (use-modules (oop goops old-define-method) (oop goops))
4800 ** Deprecated function: builtin-variable
4801 Removed function: builtin-bindings
4803 There is no longer a distinction between builtin or other variables.
4804 Use module system operations for all variables.
4806 ** Lazy-catch handlers are no longer allowed to return.
4808 That is, a call to `throw', `error', etc is now guaranteed to not
4811 ** Bugfixes for (ice-9 getopt-long)
4813 This module is now tested using test-suite/tests/getopt-long.test.
4814 The following bugs have been fixed:
4816 *** Parsing for options that are specified to have `optional' args now checks
4817 if the next element is an option instead of unconditionally taking it as the
4820 *** An error is now thrown for `--opt=val' when the option description
4821 does not specify `(value #t)' or `(value optional)'. This condition used to
4822 be accepted w/o error, contrary to the documentation.
4824 *** The error message for unrecognized options is now more informative.
4825 It used to be "not a record", an artifact of the implementation.
4827 *** The error message for `--opt' terminating the arg list (no value), when
4828 `(value #t)' is specified, is now more informative. It used to be "not enough
4831 *** "Clumped" single-char args now preserve trailing string, use it as arg.
4832 The expansion used to be like so:
4834 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
4836 Note that the "5d" is dropped. Now it is like so:
4838 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
4840 This enables single-char options to have adjoining arguments as long as their
4841 constituent characters are not potential single-char options.
4843 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
4845 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
4846 property `arglist', which can be retrieved by `arity'. The result is that
4847 `arity' can give more detailed information than before:
4851 guile> (use-modules (ice-9 optargs))
4852 guile> (define* (foo #:optional a b c) a)
4854 0 or more arguments in `lambda*:G0'.
4859 3 optional arguments: `a', `b' and `c'.
4860 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
4862 2 required arguments: `a' and `b', 2 keyword arguments: `c'
4863 and `d', other keywords allowed.
4864 guile> (define* (baz a b #:optional c #:rest r) a)
4866 2 required arguments: `a' and `b', 1 optional argument: `c',
4869 * Changes to the C interface
4871 ** Types have been renamed from scm_*_t to scm_t_*.
4873 This has been done for POSIX sake. It reserves identifiers ending
4874 with "_t". What a concept.
4876 The old names are still available with status `deprecated'.
4878 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
4880 ** Deprecated features have been removed.
4884 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
4885 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
4887 *** C Functions removed
4889 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
4890 scm_fseek - replaced by scm_seek.
4891 gc-thunk - replaced by after-gc-hook.
4892 gh_int2scmb - replaced by gh_bool2scm.
4893 scm_tc_dblr - replaced by scm_tc16_real.
4894 scm_tc_dblc - replaced by scm_tc16_complex.
4895 scm_list_star - replaced by scm_cons_star.
4897 ** Deprecated: scm_makfromstr
4899 Use scm_mem2string instead.
4901 ** Deprecated: scm_make_shared_substring
4903 Explicit shared substrings will disappear from Guile.
4905 Instead, "normal" strings will be implemented using sharing
4906 internally, combined with a copy-on-write strategy.
4908 ** Deprecated: scm_read_only_string_p
4910 The concept of read-only strings will disappear in next release of
4913 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
4915 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
4917 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
4919 Call a procedure with the indicated number of arguments. See "Fly
4920 Evaluation" in the manual.
4922 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
4924 Call a procedure with the indicated number of arguments and a list of
4925 further arguments. See "Fly Evaluation" in the manual.
4927 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
4929 Create a list of the given number of elements. See "List
4930 Constructors" in the manual.
4932 ** Renamed function: scm_listify has been replaced by scm_list_n.
4934 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
4935 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
4937 Use functions scm_list_N instead.
4939 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
4941 Used by an application to read arbitrary number of bytes from a port.
4942 Same semantics as libc read, except that scm_c_read only returns less
4943 than SIZE bytes if at end-of-file.
4945 Warning: Doesn't update port line and column counts!
4947 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
4949 Used by an application to write arbitrary number of bytes to an SCM
4950 port. Similar semantics as libc write. However, unlike libc
4951 write, scm_c_write writes the requested number of bytes and has no
4954 Warning: Doesn't update port line and column counts!
4956 ** New function: scm_init_guile ()
4958 In contrast to scm_boot_guile, scm_init_guile will return normally
4959 after initializing Guile. It is not available on all systems, tho.
4961 ** New functions: scm_str2symbol, scm_mem2symbol
4963 The function scm_str2symbol takes a const char* pointing to a zero-terminated
4964 field of characters and creates a scheme symbol object from that C string.
4965 The function scm_mem2symbol takes a const char* and a number of characters and
4966 creates a symbol from the characters in that memory area.
4968 ** New functions: scm_primitive_make_property
4969 scm_primitive_property_ref
4970 scm_primitive_property_set_x
4971 scm_primitive_property_del_x
4973 These functions implement a new way to deal with object properties.
4974 See libguile/properties.c for their documentation.
4976 ** New function: scm_done_free (long size)
4978 This function is the inverse of scm_done_malloc. Use it to report the
4979 amount of smob memory you free. The previous method, which involved
4980 calling scm_done_malloc with negative argument, was somewhat
4981 unintuitive (and is still available, of course).
4983 ** New function: scm_c_memq (SCM obj, SCM list)
4985 This function provides a fast C level alternative for scm_memq for the case
4986 that the list parameter is known to be a proper list. The function is a
4987 replacement for scm_sloppy_memq, but is stricter in its requirements on its
4988 list input parameter, since for anything else but a proper list the function's
4989 behaviour is undefined - it may even crash or loop endlessly. Further, for
4990 the case that the object is not found in the list, scm_c_memq returns #f which
4991 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
4993 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
4994 scm_remember_upto_here
4996 These functions replace the function scm_remember.
4998 ** Deprecated function: scm_remember
5000 Use one of the new functions scm_remember_upto_here_1,
5001 scm_remember_upto_here_2 or scm_remember_upto_here instead.
5003 ** New function: scm_allocate_string
5005 This function replaces the function scm_makstr.
5007 ** Deprecated function: scm_makstr
5009 Use the new function scm_allocate_string instead.
5011 ** New global variable scm_gc_running_p introduced.
5013 Use this variable to find out if garbage collection is being executed. Up to
5014 now applications have used scm_gc_heap_lock to test if garbage collection was
5015 running, which also works because of the fact that up to know only the garbage
5016 collector has set this variable. But, this is an implementation detail that
5017 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
5018 of this variable is (and has been) not fully safe anyway.
5020 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
5022 Use these instead of SCM_LENGTH_MAX.
5024 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
5025 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
5026 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
5028 Use these instead of SCM_LENGTH.
5030 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
5031 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
5032 SCM_SET_BITVECTOR_LENGTH
5034 Use these instead of SCM_SETLENGTH
5036 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
5037 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
5040 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
5043 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
5044 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
5047 Use these instead of SCM_SETCHARS.
5049 ** New macro: SCM_BITVECTOR_P
5051 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
5053 Use instead of SCM_COERCE_SUBSTR.
5055 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
5057 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
5059 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
5060 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
5061 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
5062 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
5063 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
5064 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
5065 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
5066 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
5067 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
5068 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
5069 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
5070 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
5071 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
5072 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
5073 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
5075 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
5076 Use scm_memory_error instead of SCM_NALLOC.
5077 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
5078 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
5079 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
5080 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
5081 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
5082 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
5083 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
5084 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
5085 Use SCM_STRINGP instead of SCM_RWSTRINGP.
5086 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
5087 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
5088 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
5089 Use a type specific setter macro instead of SCM_SETLENGTH.
5090 Use a type specific setter macro instead of SCM_SETCHARS.
5091 Use a type specific length macro instead of SCM_LENGTH_MAX.
5092 Use SCM_GCMARKP instead of SCM_GC8MARKP.
5093 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
5094 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
5095 Use SCM_TYP16 instead of SCM_GCTYP16.
5096 Use SCM_CDR instead of SCM_GCCDR.
5097 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
5098 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
5099 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
5100 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
5101 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
5102 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
5103 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
5105 ** Removed function: scm_struct_init
5107 ** Removed variable: scm_symhash_dim
5109 ** Renamed function: scm_make_cont has been replaced by
5110 scm_make_continuation, which has a different interface.
5112 ** Deprecated function: scm_call_catching_errors
5114 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
5116 ** Deprecated function: scm_strhash
5118 Use scm_string_hash instead.
5120 ** Deprecated function: scm_vector_set_length_x
5122 Instead, create a fresh vector of the desired size and copy the contents.
5124 ** scm_gensym has changed prototype
5126 scm_gensym now only takes one argument.
5128 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
5131 There is now only a single symbol type scm_tc7_symbol.
5132 The tag scm_tc7_lvector was not used anyway.
5134 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
5136 Use scm_make_smob_type and scm_set_smob_XXX instead.
5138 ** New function scm_set_smob_apply.
5140 This can be used to set an apply function to a smob type.
5142 ** Deprecated function: scm_strprint_obj
5144 Use scm_object_to_string instead.
5146 ** Deprecated function: scm_wta
5148 Use scm_wrong_type_arg, or another appropriate error signalling function
5151 ** Explicit support for obarrays has been deprecated.
5153 Use `scm_str2symbol' and the generic hashtable functions instead.
5155 ** The concept of `vcells' has been deprecated.
5157 The data type `variable' is now used exclusively. `Vcells' have been
5158 a low-level concept so you are likely not affected by this change.
5160 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
5161 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
5163 Use scm_c_define or scm_c_lookup instead, as appropriate.
5165 *** New functions: scm_c_module_lookup, scm_c_lookup,
5166 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
5167 scm_module_define, scm_define.
5169 These functions work with variables instead of with vcells.
5171 ** New functions for creating and defining `subr's and `gsubr's.
5173 The new functions more clearly distinguish between creating a subr (or
5174 gsubr) object and adding it to the current module.
5176 These new functions are available: scm_c_make_subr, scm_c_define_subr,
5177 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
5178 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
5179 scm_c_define_gsubr_with_generic.
5181 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
5182 scm_make_subr_with_generic, scm_make_gsubr,
5183 scm_make_gsubr_with_generic.
5185 Use the new ones from above instead.
5187 ** C interface to the module system has changed.
5189 While we suggest that you avoid as many explicit module system
5190 operations from C as possible for the time being, the C interface has
5191 been made more similar to the high-level Scheme module system.
5193 *** New functions: scm_c_define_module, scm_c_use_module,
5194 scm_c_export, scm_c_resolve_module.
5196 They mostly work like their Scheme namesakes. scm_c_define_module
5197 takes a function that is called a context where the new module is
5200 *** Deprecated functions: scm_the_root_module, scm_make_module,
5201 scm_ensure_user_module, scm_load_scheme_module.
5203 Use the new functions instead.
5205 ** Renamed function: scm_internal_with_fluids becomes
5208 scm_internal_with_fluids is available as a deprecated function.
5210 ** New function: scm_c_with_fluid.
5212 Just like scm_c_with_fluids, but takes one fluid and one value instead
5215 ** Deprecated typedefs: long_long, ulong_long.
5217 They are of questionable utility and they pollute the global
5220 ** Deprecated typedef: scm_sizet
5222 It is of questionable utility now that Guile requires ANSI C, and is
5225 ** Deprecated typedefs: scm_port_rw_active, scm_port,
5226 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
5227 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
5229 Made more compliant with the naming policy by adding a _t at the end.
5231 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
5232 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
5234 With the exception of the mysterious scm_2ulong2big, they are still
5235 available under new names (scm_i_mkbig etc). These functions are not
5236 intended to be used in user code. You should avoid dealing with
5237 bignums directly, and should deal with numbers in general (which can
5240 ** Change in behavior: scm_num2long, scm_num2ulong
5242 The scm_num2[u]long functions don't any longer accept an inexact
5243 argument. This change in behavior is motivated by concordance with
5244 R5RS: It is more common that a primitive doesn't want to accept an
5245 inexact for an exact.
5247 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
5248 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
5249 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
5252 These are conversion functions between the various ANSI C integral
5253 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
5254 accept an inexact argument.
5256 ** New functions: scm_float2num, scm_double2num,
5257 scm_num2float, scm_num2double.
5259 These are conversion functions between the two ANSI C float types and
5262 ** New number validation macros:
5263 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
5267 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
5269 These are just nicer-named old scm_protect_object and
5270 scm_unprotect_object.
5272 ** Deprecated functions: scm_protect_object, scm_unprotect_object
5274 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
5276 These functions can be used to register pointers to locations that
5279 ** Deprecated function: scm_create_hook.
5281 Its sins are: misleading name, non-modularity and lack of general
5285 Changes since Guile 1.3.4:
5287 * Changes to the distribution
5289 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
5291 We've changed the way we handle generated files in the Guile source
5292 repository. As a result, the procedure for building trees obtained
5293 from the nightly FTP snapshots or via CVS has changed:
5294 - You must have appropriate versions of autoconf, automake, and
5295 libtool installed on your system. See README for info on how to
5296 obtain these programs.
5297 - Before configuring the tree, you must first run the script
5298 `autogen.sh' at the top of the source tree.
5300 The Guile repository used to contain not only source files, written by
5301 humans, but also some generated files, like configure scripts and
5302 Makefile.in files. Even though the contents of these files could be
5303 derived mechanically from other files present, we thought it would
5304 make the tree easier to build if we checked them into CVS.
5306 However, this approach means that minor differences between
5307 developer's installed tools and habits affected the whole team.
5308 So we have removed the generated files from the repository, and
5309 added the autogen.sh script, which will reconstruct them
5313 ** configure now has experimental options to remove support for certain
5316 --disable-arrays omit array and uniform array support
5317 --disable-posix omit posix interfaces
5318 --disable-networking omit networking interfaces
5319 --disable-regex omit regular expression interfaces
5321 These are likely to become separate modules some day.
5323 ** New configure option --enable-debug-freelist
5325 This enables a debugging version of SCM_NEWCELL(), and also registers
5326 an extra primitive, the setter `gc-set-debug-check-freelist!'.
5328 Configure with the --enable-debug-freelist option to enable
5329 the gc-set-debug-check-freelist! primitive, and then use:
5331 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
5332 (gc-set-debug-check-freelist! #f) # turn off checking
5334 Checking of the freelist forces a traversal of the freelist and
5335 a garbage collection before each allocation of a cell. This can
5336 slow down the interpreter dramatically, so the setter should be used to
5337 turn on this extra processing only when necessary.
5339 ** New configure option --enable-debug-malloc
5341 Include code for debugging of calls to scm_must_malloc/realloc/free.
5345 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
5346 2. objects reallocated by scm_must_realloc has been allocated by
5348 3. reallocated objects are reallocated with the same what string
5350 But, most importantly, it records the number of allocated objects of
5351 each kind. This is useful when searching for memory leaks.
5353 A Guile compiled with this option provides the primitive
5354 `malloc-stats' which returns an alist with pairs of kind and the
5355 number of objects of that kind.
5357 ** All includes are now referenced relative to the root directory
5359 Since some users have had problems with mixups between Guile and
5360 system headers, we have decided to always refer to Guile headers via
5361 their parent directories. This essentially creates a "private name
5362 space" for Guile headers. This means that the compiler only is given
5363 -I options for the root build and root source directory.
5365 ** Header files kw.h and genio.h have been removed.
5367 ** The module (ice-9 getopt-gnu-style) has been removed.
5369 ** New module (ice-9 documentation)
5371 Implements the interface to documentation strings associated with
5374 ** New module (ice-9 time)
5376 Provides a macro `time', which displays execution time of a given form.
5378 ** New module (ice-9 history)
5380 Loading this module enables value history in the repl.
5382 * Changes to the stand-alone interpreter
5384 ** New command line option --debug
5386 Start Guile with debugging evaluator and backtraces enabled.
5388 This is useful when debugging your .guile init file or scripts.
5390 ** New help facility
5392 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
5393 (help REGEXP) ditto for objects with names matching REGEXP (a string)
5394 (help 'NAME) gives documentation for NAME, even if it is not an object
5395 (help ,EXPR) gives documentation for object returned by EXPR
5396 (help (my module)) gives module commentary for `(my module)'
5397 (help) gives this text
5399 `help' searches among bindings exported from loaded modules, while
5400 `apropos' searches among bindings visible from the "current" module.
5402 Examples: (help help)
5404 (help "output-string")
5406 ** `help' and `apropos' now prints full module names
5408 ** Dynamic linking now uses libltdl from the libtool package.
5410 The old system dependent code for doing dynamic linking has been
5411 replaced with calls to the libltdl functions which do all the hairy
5414 The major improvement is that you can now directly pass libtool
5415 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
5416 will be able to do the best shared library job you can get, via
5419 The way dynamic libraries are found has changed and is not really
5420 portable across platforms, probably. It is therefore recommended to
5421 use absolute filenames when possible.
5423 If you pass a filename without an extension to `dynamic-link', it will
5424 try a few appropriate ones. Thus, the most platform ignorant way is
5425 to specify a name like "libfoo", without any directories and
5428 ** Guile COOP threads are now compatible with LinuxThreads
5430 Previously, COOP threading wasn't possible in applications linked with
5431 Linux POSIX threads due to their use of the stack pointer to find the
5432 thread context. This has now been fixed with a workaround which uses
5433 the pthreads to allocate the stack.
5435 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
5437 ** Positions of erring expression in scripts
5439 With version 1.3.4, the location of the erring expression in Guile
5440 scipts is no longer automatically reported. (This should have been
5441 documented before the 1.3.4 release.)
5443 You can get this information by enabling recording of positions of
5444 source expressions and running the debugging evaluator. Put this at
5445 the top of your script (or in your "site" file):
5447 (read-enable 'positions)
5448 (debug-enable 'debug)
5450 ** Backtraces in scripts
5452 It is now possible to get backtraces in scripts.
5456 (debug-enable 'debug 'backtrace)
5458 at the top of the script.
5460 (The first options enables the debugging evaluator.
5461 The second enables backtraces.)
5463 ** Part of module system symbol lookup now implemented in C
5465 The eval closure of most modules is now implemented in C. Since this
5466 was one of the bottlenecks for loading speed, Guile now loads code
5467 substantially faster than before.
5469 ** Attempting to get the value of an unbound variable now produces
5470 an exception with a key of 'unbound-variable instead of 'misc-error.
5472 ** The initial default output port is now unbuffered if it's using a
5473 tty device. Previously in this situation it was line-buffered.
5475 ** New hook: after-gc-hook
5477 after-gc-hook takes over the role of gc-thunk. This hook is run at
5478 the first SCM_TICK after a GC. (Thus, the code is run at the same
5479 point during evaluation as signal handlers.)
5481 Note that this hook should be used only for diagnostic and debugging
5482 purposes. It is not certain that it will continue to be well-defined
5483 when this hook is run in the future.
5485 C programmers: Note the new C level hooks scm_before_gc_c_hook,
5486 scm_before_sweep_c_hook, scm_after_gc_c_hook.
5488 ** Improvements to garbage collector
5490 Guile 1.4 has a new policy for triggering heap allocation and
5491 determining the sizes of heap segments. It fixes a number of problems
5494 1. The new policy can handle two separate pools of cells
5495 (2-word/4-word) better. (The old policy would run wild, allocating
5496 more and more memory for certain programs.)
5498 2. The old code would sometimes allocate far too much heap so that the
5499 Guile process became gigantic. The new code avoids this.
5501 3. The old code would sometimes allocate too little so that few cells
5502 were freed at GC so that, in turn, too much time was spent in GC.
5504 4. The old code would often trigger heap allocation several times in a
5505 row. (The new scheme predicts how large the segments needs to be
5506 in order not to need further allocation.)
5508 All in all, the new GC policy will make larger applications more
5511 The new GC scheme also is prepared for POSIX threading. Threads can
5512 allocate private pools of cells ("clusters") with just a single
5513 function call. Allocation of single cells from such a cluster can
5514 then proceed without any need of inter-thread synchronization.
5516 ** New environment variables controlling GC parameters
5518 GUILE_MAX_SEGMENT_SIZE Maximal segment size
5521 Allocation of 2-word cell heaps:
5523 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
5526 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
5527 GC in percent of total heap size
5530 Allocation of 4-word cell heaps
5531 (used for real numbers and misc other objects):
5533 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
5535 (See entry "Way for application to customize GC parameters" under
5536 section "Changes to the scm_ interface" below.)
5538 ** Guile now implements reals using 4-word cells
5540 This speeds up computation with reals. (They were earlier allocated
5541 with `malloc'.) There is still some room for optimizations, however.
5543 ** Some further steps toward POSIX thread support have been taken
5545 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
5546 don't have much effect any longer, and many of them will be removed in
5550 are only handled at the top of the evaluator loop, immediately after
5551 I/O, and in scm_equalp.
5553 *** The GC can allocate thread private pools of pairs.
5555 * Changes to Scheme functions and syntax
5557 ** close-input-port and close-output-port are now R5RS
5559 These procedures have been turned into primitives and have R5RS behaviour.
5561 ** New procedure: simple-format PORT MESSAGE ARG1 ...
5563 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
5564 extended by the more sophisticated version in (ice-9 format)
5566 (simple-format port message . args)
5567 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
5568 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
5569 the escapes are replaced with corresponding members of ARGS:
5570 ~A formats using `display' and ~S formats using `write'.
5571 If DESTINATION is #t, then use the `current-output-port',
5572 if DESTINATION is #f, then return a string containing the formatted text.
5573 Does not add a trailing newline."
5575 ** string-ref: the second argument is no longer optional.
5577 ** string, list->string: no longer accept strings in their arguments,
5578 only characters, for compatibility with R5RS.
5580 ** New procedure: port-closed? PORT
5581 Returns #t if PORT is closed or #f if it is open.
5583 ** Deprecated: list*
5585 The list* functionality is now provided by cons* (SRFI-1 compliant)
5587 ** New procedure: cons* ARG1 ARG2 ... ARGn
5589 Like `list', but the last arg provides the tail of the constructed list,
5590 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
5592 Requires at least one argument. If given one argument, that argument
5593 is returned as result.
5595 This function is called `list*' in some other Schemes and in Common LISP.
5597 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
5599 ** New procedure: object-documentation OBJECT
5601 Returns the documentation string associated with OBJECT. The
5602 procedure uses a caching mechanism so that subsequent lookups are
5605 Exported by (ice-9 documentation).
5607 ** module-name now returns full names of modules
5609 Previously, only the last part of the name was returned (`session' for
5610 `(ice-9 session)'). Ex: `(ice-9 session)'.
5612 * Changes to the gh_ interface
5614 ** Deprecated: gh_int2scmb
5616 Use gh_bool2scm instead.
5618 * Changes to the scm_ interface
5620 ** Guile primitives now carry docstrings!
5622 Thanks to Greg Badros!
5624 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5626 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5627 macros and must contain a docstring that is extracted into foo.doc using a new
5628 guile-doc-snarf script (that uses guile-doc-snarf.awk).
5630 However, a major overhaul of these macros is scheduled for the next release of
5633 ** Guile primitives use a new technique for validation of arguments
5635 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
5636 the readability of argument checking.
5638 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
5640 ** New macros: SCM_PACK, SCM_UNPACK
5642 Compose/decompose an SCM value.
5644 The SCM type is now treated as an abstract data type and may be defined as a
5645 long, a void* or as a struct, depending on the architecture and compile time
5646 options. This makes it easier to find several types of bugs, for example when
5647 SCM values are treated as integers without conversion. Values of the SCM type
5648 should be treated as "atomic" values. These macros are used when
5649 composing/decomposing an SCM value, either because you want to access
5650 individual bits, or because you want to treat it as an integer value.
5652 E.g., in order to set bit 7 in an SCM value x, use the expression
5654 SCM_PACK (SCM_UNPACK (x) | 0x80)
5656 ** The name property of hooks is deprecated.
5657 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
5659 You can emulate this feature by using object properties.
5661 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
5662 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
5665 These macros will be removed in a future release of Guile.
5667 ** The following types, functions and macros from numbers.h are deprecated:
5668 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
5669 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
5671 ** Port internals: the rw_random variable in the scm_port structure
5672 must be set to non-zero in any random access port. In recent Guile
5673 releases it was only set for bidirectional random-access ports.
5675 ** Port internals: the seek ptob procedure is now responsible for
5676 resetting the buffers if required. The change was made so that in the
5677 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
5678 the fport and strport ptobs can avoid resetting the buffers,
5679 in particular to avoid discarding unread chars. An existing port
5680 type can be fixed by adding something like the following to the
5681 beginning of the ptob seek procedure:
5683 if (pt->rw_active == SCM_PORT_READ)
5684 scm_end_input (object);
5685 else if (pt->rw_active == SCM_PORT_WRITE)
5686 ptob->flush (object);
5688 although to actually avoid resetting the buffers and discard unread
5689 chars requires further hacking that depends on the characteristics
5692 ** Deprecated functions: scm_fseek, scm_tag
5694 These functions are no longer used and will be removed in a future version.
5696 ** The scm_sysmissing procedure is no longer used in libguile.
5697 Unless it turns out to be unexpectedly useful to somebody, it will be
5698 removed in a future version.
5700 ** The format of error message strings has changed
5702 The two C procedures: scm_display_error and scm_error, as well as the
5703 primitive `scm-error', now use scm_simple_format to do their work.
5704 This means that the message strings of all code must be updated to use
5705 ~A where %s was used before, and ~S where %S was used before.
5707 During the period when there still are a lot of old Guiles out there,
5708 you might want to support both old and new versions of Guile.
5710 There are basically two methods to achieve this. Both methods use
5713 AC_CHECK_FUNCS(scm_simple_format)
5715 in your configure.in.
5717 Method 1: Use the string concatenation features of ANSI C's
5722 #ifdef HAVE_SCM_SIMPLE_FORMAT
5728 Then represent each of your error messages using a preprocessor macro:
5730 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
5734 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
5735 (define make-message string-append)
5737 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
5739 Method 2: Use the oldfmt function found in doc/oldfmt.c.
5743 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
5748 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
5752 ** Deprecated: coop_mutex_init, coop_condition_variable_init
5754 Don't use the functions coop_mutex_init and
5755 coop_condition_variable_init. They will change.
5757 Use scm_mutex_init and scm_cond_init instead.
5759 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
5760 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
5761 COND, as `scm_cond_wait' does, but it also bounds the duration
5762 of the wait. If COND has not been signaled before time ABSTIME,
5763 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
5764 returns the error code `ETIMEDOUT'.
5766 The ABSTIME parameter specifies an absolute time, with the same
5767 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
5768 to 00:00:00 GMT, January 1, 1970.
5770 ** New function: scm_cond_broadcast (scm_cond_t *COND)
5771 `scm_cond_broadcast' restarts all the threads that are waiting
5772 on the condition variable COND. Nothing happens if no threads are
5775 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
5776 `scm_key_create' allocates a new TSD key. The key is stored in
5777 the location pointed to by KEY. There is no limit on the number
5778 of keys allocated at a given time. The value initially associated
5779 with the returned key is `NULL' in all currently executing threads.
5781 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
5782 function associated with the key. When a thread terminates,
5783 DESTR_FUNCTION is called on the value associated with the key in
5784 that thread. The DESTR_FUNCTION is not called if a key is deleted
5785 with `scm_key_delete' or a value is changed with
5786 `scm_setspecific'. The order in which destructor functions are
5787 called at thread termination time is unspecified.
5789 Destructors are not yet implemented.
5791 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
5792 `scm_setspecific' changes the value associated with KEY in the
5793 calling thread, storing the given POINTER instead.
5795 ** New function: scm_getspecific (scm_key_t KEY)
5796 `scm_getspecific' returns the value currently associated with
5797 KEY in the calling thread.
5799 ** New function: scm_key_delete (scm_key_t KEY)
5800 `scm_key_delete' deallocates a TSD key. It does not check
5801 whether non-`NULL' values are associated with that key in the
5802 currently executing threads, nor call the destructor function
5803 associated with the key.
5805 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
5807 Initialize a C level hook HOOK with associated HOOK_DATA and type
5808 TYPE. (See scm_c_hook_run ().)
5810 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
5812 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
5813 is true, add it last, otherwise first. The same FUNC can be added
5814 multiple times if FUNC_DATA differ and vice versa.
5816 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
5818 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
5819 function is only removed if both FUNC and FUNC_DATA matches.
5821 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
5823 Run hook HOOK passing DATA to the hook functions.
5825 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
5826 returned is undefined.
5828 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
5829 returns a non-NULL value. This value is returned as the result of
5830 scm_c_hook_run. If all functions return NULL, NULL is returned.
5832 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
5833 returns a NULL value, and NULL is returned. If all functions returns
5834 a non-NULL value, the last value is returned.
5836 ** New C level GC hooks
5838 Five new C level hooks has been added to the garbage collector.
5840 scm_before_gc_c_hook
5843 are run before locking and after unlocking the heap. The system is
5844 thus in a mode where evaluation can take place. (Except that
5845 scm_before_gc_c_hook must not allocate new cells.)
5847 scm_before_mark_c_hook
5848 scm_before_sweep_c_hook
5849 scm_after_sweep_c_hook
5851 are run when the heap is locked. These are intended for extension of
5852 the GC in a modular fashion. Examples are the weaks and guardians
5855 ** Way for application to customize GC parameters
5857 The application can set up other default values for the GC heap
5858 allocation parameters
5860 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
5861 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
5862 GUILE_MAX_SEGMENT_SIZE,
5866 scm_default_init_heap_size_1, scm_default_min_yield_1,
5867 scm_default_init_heap_size_2, scm_default_min_yield_2,
5868 scm_default_max_segment_size
5870 respectively before callong scm_boot_guile.
5872 (See entry "New environment variables ..." in section
5873 "Changes to the stand-alone interpreter" above.)
5875 ** scm_protect_object/scm_unprotect_object now nest
5877 This means that you can call scm_protect_object multiple times on an
5878 object and count on the object being protected until
5879 scm_unprotect_object has been call the same number of times.
5881 The functions also have better time complexity.
5883 Still, it is usually possible to structure the application in a way
5884 that you don't need to use these functions. For example, if you use a
5885 protected standard Guile list to keep track of live objects rather
5886 than some custom data type, objects will die a natural death when they
5887 are no longer needed.
5889 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
5891 Guile does not provide the float representation for inexact real numbers any
5892 more. Now, only doubles are used to represent inexact real numbers. Further,
5893 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
5894 and scm_tc16_complex, respectively.
5896 ** Removed deprecated type scm_smobfuns
5898 ** Removed deprecated function scm_newsmob
5900 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
5902 There is an ongoing discussion among the developers whether to
5903 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
5904 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
5905 until this issue has been settled.
5907 ** Removed deprecated type tag scm_tc16_kw
5909 ** Added type tag scm_tc16_keyword
5911 (This was introduced already in release 1.3.4 but was not documented
5914 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
5916 * Changes to system call interfaces:
5918 ** The "select" procedure now tests port buffers for the ability to
5919 provide input or accept output. Previously only the underlying file
5920 descriptors were checked.
5922 ** New variable PIPE_BUF: the maximum number of bytes that can be
5923 atomically written to a pipe.
5925 ** If a facility is not available on the system when Guile is
5926 compiled, the corresponding primitive procedure will not be defined.
5927 Previously it would have been defined but would throw a system-error
5928 exception if called. Exception handlers which catch this case may
5929 need minor modification: an error will be thrown with key
5930 'unbound-variable instead of 'system-error. Alternatively it's
5931 now possible to use `defined?' to check whether the facility is
5934 ** Procedures which depend on the timezone should now give the correct
5935 result on systems which cache the TZ environment variable, even if TZ
5936 is changed without calling tzset.
5938 * Changes to the networking interfaces:
5940 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
5941 long integers between network and host format. For now, it's not
5942 particularly convenient to do this kind of thing, but consider:
5944 (define write-network-long
5945 (lambda (value port)
5946 (let ((v (make-uniform-vector 1 1 0)))
5947 (uniform-vector-set! v 0 (htonl value))
5948 (uniform-vector-write v port))))
5950 (define read-network-long
5952 (let ((v (make-uniform-vector 1 1 0)))
5953 (uniform-vector-read! v port)
5954 (ntohl (uniform-vector-ref v 0)))))
5956 ** If inet-aton fails, it now throws an error with key 'misc-error
5957 instead of 'system-error, since errno is not relevant.
5959 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
5960 specific keys instead of 'system-error. The latter is inappropriate
5961 since errno will not have been set. The keys are:
5962 'host-not-found, 'try-again, 'no-recovery and 'no-data.
5964 ** sethostent, setnetent, setprotoent, setservent: now take an
5965 optional argument STAYOPEN, which specifies whether the database
5966 remains open after a database entry is accessed randomly (e.g., using
5967 gethostbyname for the hosts database.) The default is #f. Previously
5971 Changes since Guile 1.3.2:
5973 * Changes to the stand-alone interpreter
5977 An initial version of the Guile debugger written by Chris Hanson has
5978 been added. The debugger is still under development but is included
5979 in the distribution anyway since it is already quite useful.
5985 after an error to enter the debugger. Type `help' inside the debugger
5986 for a description of available commands.
5988 If you prefer to have stack frames numbered and printed in
5989 anti-chronological order and prefer up in the stack to be down on the
5990 screen as is the case in gdb, you can put
5992 (debug-enable 'backwards)
5994 in your .guile startup file. (However, this means that Guile can't
5995 use indentation to indicate stack level.)
5997 The debugger is autoloaded into Guile at the first use.
5999 ** Further enhancements to backtraces
6001 There is a new debug option `width' which controls the maximum width
6002 on the screen of printed stack frames. Fancy printing parameters
6003 ("level" and "length" as in Common LISP) are adaptively adjusted for
6004 each stack frame to give maximum information while still fitting
6005 within the bounds. If the stack frame can't be made to fit by
6006 adjusting parameters, it is simply cut off at the end. This is marked
6009 ** Some modules are now only loaded when the repl is started
6011 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
6012 regex) are now loaded into (guile-user) only if the repl has been
6013 started. The effect is that the startup time for scripts has been
6014 reduced to 30% of what it was previously.
6016 Correctly written scripts load the modules they require at the top of
6017 the file and should not be affected by this change.
6019 ** Hooks are now represented as smobs
6021 * Changes to Scheme functions and syntax
6023 ** Readline support has changed again.
6025 The old (readline-activator) module is gone. Use (ice-9 readline)
6026 instead, which now contains all readline functionality. So the code
6027 to activate readline is now
6029 (use-modules (ice-9 readline))
6032 This should work at any time, including from the guile prompt.
6034 To avoid confusion about the terms of Guile's license, please only
6035 enable readline for your personal use; please don't make it the
6036 default for others. Here is why we make this rather odd-sounding
6039 Guile is normally licensed under a weakened form of the GNU General
6040 Public License, which allows you to link code with Guile without
6041 placing that code under the GPL. This exception is important to some
6044 However, since readline is distributed under the GNU General Public
6045 License, when you link Guile with readline, either statically or
6046 dynamically, you effectively change Guile's license to the strict GPL.
6047 Whenever you link any strictly GPL'd code into Guile, uses of Guile
6048 which are normally permitted become forbidden. This is a rather
6049 non-obvious consequence of the licensing terms.
6051 So, to make sure things remain clear, please let people choose for
6052 themselves whether to link GPL'd libraries like readline with Guile.
6054 ** regexp-substitute/global has changed slightly, but incompatibly.
6056 If you include a function in the item list, the string of the match
6057 object it receives is the same string passed to
6058 regexp-substitute/global, not some suffix of that string.
6059 Correspondingly, the match's positions are relative to the entire
6060 string, not the suffix.
6062 If the regexp can match the empty string, the way matches are chosen
6063 from the string has changed. regexp-substitute/global recognizes the
6064 same set of matches that list-matches does; see below.
6066 ** New function: list-matches REGEXP STRING [FLAGS]
6068 Return a list of match objects, one for every non-overlapping, maximal
6069 match of REGEXP in STRING. The matches appear in left-to-right order.
6070 list-matches only reports matches of the empty string if there are no
6071 other matches which begin on, end at, or include the empty match's
6074 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6076 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
6078 For each match of REGEXP in STRING, apply PROC to the match object,
6079 and the last value PROC returned, or INIT for the first call. Return
6080 the last value returned by PROC. We apply PROC to the matches as they
6081 appear from left to right.
6083 This function recognizes matches according to the same criteria as
6086 Thus, you could define list-matches like this:
6088 (define (list-matches regexp string . flags)
6089 (reverse! (apply fold-matches regexp string '() cons flags)))
6091 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
6095 *** New function: hook? OBJ
6097 Return #t if OBJ is a hook, otherwise #f.
6099 *** New function: make-hook-with-name NAME [ARITY]
6101 Return a hook with name NAME and arity ARITY. The default value for
6102 ARITY is 0. The only effect of NAME is that it will appear when the
6103 hook object is printed to ease debugging.
6105 *** New function: hook-empty? HOOK
6107 Return #t if HOOK doesn't contain any procedures, otherwise #f.
6109 *** New function: hook->list HOOK
6111 Return a list of the procedures that are called when run-hook is
6114 ** `map' signals an error if its argument lists are not all the same length.
6116 This is the behavior required by R5RS, so this change is really a bug
6117 fix. But it seems to affect a lot of people's code, so we're
6118 mentioning it here anyway.
6120 ** Print-state handling has been made more transparent
6122 Under certain circumstances, ports are represented as a port with an
6123 associated print state. Earlier, this pair was represented as a pair
6124 (see "Some magic has been added to the printer" below). It is now
6125 indistinguishable (almost; see `get-print-state') from a port on the
6128 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
6130 Return a new port with the associated print state PRINT-STATE.
6132 *** New function: get-print-state OUTPUT-PORT
6134 Return the print state associated with this port if it exists,
6135 otherwise return #f.
6137 *** New function: directory-stream? OBJECT
6139 Returns true iff OBJECT is a directory stream --- the sort of object
6140 returned by `opendir'.
6142 ** New function: using-readline?
6144 Return #t if readline is in use in the current repl.
6146 ** structs will be removed in 1.4
6148 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
6149 and use GOOPS objects as the fundamental record type.
6151 * Changes to the scm_ interface
6153 ** structs will be removed in 1.4
6155 The entire current struct interface (struct.c, struct.h) will be
6156 replaced in Guile 1.4. We will merge GOOPS into libguile and use
6157 GOOPS objects as the fundamental record type.
6159 ** The internal representation of subr's has changed
6161 Instead of giving a hint to the subr name, the CAR field of the subr
6162 now contains an index to a subr entry in scm_subr_table.
6164 *** New variable: scm_subr_table
6166 An array of subr entries. A subr entry contains the name, properties
6167 and documentation associated with the subr. The properties and
6168 documentation slots are not yet used.
6170 ** A new scheme for "forwarding" calls to a builtin to a generic function
6172 It is now possible to extend the functionality of some Guile
6173 primitives by letting them defer a call to a GOOPS generic function on
6174 argument mismatch. This means that there is no loss of efficiency in
6179 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
6180 (define-method + ((x <string>) (y <string>))
6181 (string-append x y))
6183 + will still be as efficient as usual in numerical calculations, but
6184 can also be used for concatenating strings.
6186 Who will be the first one to extend Guile's numerical tower to
6187 rationals? :) [OK, there a few other things to fix before this can
6188 be made in a clean way.]
6190 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
6192 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
6194 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
6196 These do the same job as SCM_PROC and SCM_PROC1, but they also define
6197 a variable GENERIC which can be used by the dispatch macros below.
6199 [This is experimental code which may change soon.]
6201 *** New macros for forwarding control to a generic on arg type error
6203 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
6205 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
6207 These correspond to the scm_wta function call, and have the same
6208 behaviour until the user has called the GOOPS primitive
6209 `enable-primitive-generic!'. After that, these macros will apply the
6210 generic function GENERIC to the argument(s) instead of calling
6213 [This is experimental code which may change soon.]
6215 *** New macros for argument testing with generic dispatch
6217 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
6219 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
6221 These correspond to the SCM_ASSERT macro, but will defer control to
6222 GENERIC on error after `enable-primitive-generic!' has been called.
6224 [This is experimental code which may change soon.]
6226 ** New function: SCM scm_eval_body (SCM body, SCM env)
6228 Evaluates the body of a special form.
6230 ** The internal representation of struct's has changed
6232 Previously, four slots were allocated for the procedure(s) of entities
6233 and operators. The motivation for this representation had to do with
6234 the structure of the evaluator, the wish to support tail-recursive
6235 generic functions, and efficiency. Since the generic function
6236 dispatch mechanism has changed, there is no longer a need for such an
6237 expensive representation, and the representation has been simplified.
6239 This should not make any difference for most users.
6241 ** GOOPS support has been cleaned up.
6243 Some code has been moved from eval.c to objects.c and code in both of
6244 these compilation units has been cleaned up and better structured.
6246 *** New functions for applying generic functions
6248 New function: SCM scm_apply_generic (GENERIC, ARGS)
6249 New function: SCM scm_call_generic_0 (GENERIC)
6250 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
6251 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
6252 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
6254 ** Deprecated function: scm_make_named_hook
6256 It is now replaced by:
6258 ** New function: SCM scm_create_hook (const char *name, int arity)
6260 Creates a hook in the same way as make-hook above but also
6261 binds a variable named NAME to it.
6263 This is the typical way of creating a hook from C code.
6265 Currently, the variable is created in the "current" module.
6266 This might change when we get the new module system.
6268 [The behaviour is identical to scm_make_named_hook.]
6272 Changes since Guile 1.3:
6274 * Changes to mailing lists
6276 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
6278 See the README file to find current addresses for all the Guile
6281 * Changes to the distribution
6283 ** Readline support is no longer included with Guile by default.
6285 Based on the different license terms of Guile and Readline, we
6286 concluded that Guile should not *by default* cause the linking of
6287 Readline into an application program. Readline support is now offered
6288 as a separate module, which is linked into an application only when
6289 you explicitly specify it.
6291 Although Guile is GNU software, its distribution terms add a special
6292 exception to the usual GNU General Public License (GPL). Guile's
6293 license includes a clause that allows you to link Guile with non-free
6294 programs. We add this exception so as not to put Guile at a
6295 disadvantage vis-a-vis other extensibility packages that support other
6298 In contrast, the GNU Readline library is distributed under the GNU
6299 General Public License pure and simple. This means that you may not
6300 link Readline, even dynamically, into an application unless it is
6301 distributed under a free software license that is compatible the GPL.
6303 Because of this difference in distribution terms, an application that
6304 can use Guile may not be able to use Readline. Now users will be
6305 explicitly offered two independent decisions about the use of these
6308 You can activate the readline support by issuing
6310 (use-modules (readline-activator))
6313 from your ".guile" file, for example.
6315 * Changes to the stand-alone interpreter
6317 ** All builtins now print as primitives.
6318 Previously builtin procedures not belonging to the fundamental subr
6319 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
6320 Now, they print as #<primitive-procedure NAME>.
6322 ** Backtraces slightly more intelligible.
6323 gsubr-apply and macro transformer application frames no longer appear
6326 * Changes to Scheme functions and syntax
6328 ** Guile now correctly handles internal defines by rewriting them into
6329 their equivalent letrec. Previously, internal defines would
6330 incrementally add to the innermost environment, without checking
6331 whether the restrictions specified in RnRS were met. This lead to the
6332 correct behaviour when these restriction actually were met, but didn't
6333 catch all illegal uses. Such an illegal use could lead to crashes of
6334 the Guile interpreter or other unwanted results. An example of
6335 incorrect internal defines that made Guile behave erratically:
6347 The problem with this example is that the definition of `c' uses the
6348 value of `b' directly. This confuses the meoization machine of Guile
6349 so that the second call of `b' (this time in a larger environment that
6350 also contains bindings for `c' and `d') refers to the binding of `c'
6351 instead of `a'. You could also make Guile crash with a variation on
6356 (define (b flag) (if flag a 1))
6357 (define c (1+ (b flag)))
6365 From now on, Guile will issue an `Unbound variable: b' error message
6370 A hook contains a list of functions which should be called on
6371 particular occasions in an existing program. Hooks are used for
6374 A window manager might have a hook before-window-map-hook. The window
6375 manager uses the function run-hooks to call all functions stored in
6376 before-window-map-hook each time a window is mapped. The user can
6377 store functions in the hook using add-hook!.
6379 In Guile, hooks are first class objects.
6381 *** New function: make-hook [N_ARGS]
6383 Return a hook for hook functions which can take N_ARGS arguments.
6384 The default value for N_ARGS is 0.
6386 (See also scm_make_named_hook below.)
6388 *** New function: add-hook! HOOK PROC [APPEND_P]
6390 Put PROC at the beginning of the list of functions stored in HOOK.
6391 If APPEND_P is supplied, and non-false, put PROC at the end instead.
6393 PROC must be able to take the number of arguments specified when the
6396 If PROC already exists in HOOK, then remove it first.
6398 *** New function: remove-hook! HOOK PROC
6400 Remove PROC from the list of functions in HOOK.
6402 *** New function: reset-hook! HOOK
6404 Clear the list of hook functions stored in HOOK.
6406 *** New function: run-hook HOOK ARG1 ...
6408 Run all hook functions stored in HOOK with arguments ARG1 ... .
6409 The number of arguments supplied must correspond to the number given
6410 when the hook was created.
6412 ** The function `dynamic-link' now takes optional keyword arguments.
6413 The only keyword argument that is currently defined is `:global
6414 BOOL'. With it, you can control whether the shared library will be
6415 linked in global mode or not. In global mode, the symbols from the
6416 linked library can be used to resolve references from other
6417 dynamically linked libraries. In non-global mode, the linked
6418 library is essentially invisible and can only be accessed via
6419 `dynamic-func', etc. The default is now to link in global mode.
6420 Previously, the default has been non-global mode.
6422 The `#:global' keyword is only effective on platforms that support
6423 the dlopen family of functions.
6425 ** New function `provided?'
6427 - Function: provided? FEATURE
6428 Return true iff FEATURE is supported by this installation of
6429 Guile. FEATURE must be a symbol naming a feature; the global
6430 variable `*features*' is a list of available features.
6432 ** Changes to the module (ice-9 expect):
6434 *** The expect-strings macro now matches `$' in a regular expression
6435 only at a line-break or end-of-file by default. Previously it would
6436 match the end of the string accumulated so far. The old behaviour
6437 can be obtained by setting the variable `expect-strings-exec-flags'
6440 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
6441 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
6442 in a regular expression will still match before a line-break or
6443 end-of-file. The default is `regexp/noteol'.
6445 *** The expect-strings macro now uses a variable
6446 `expect-strings-compile-flags' for the flags to be supplied to
6447 `make-regexp'. The default is `regexp/newline', which was previously
6450 *** The expect macro now supplies two arguments to a match procedure:
6451 the current accumulated string and a flag to indicate whether
6452 end-of-file has been reached. Previously only the string was supplied.
6453 If end-of-file is reached, the match procedure will be called an
6454 additional time with the same accumulated string as the previous call
6455 but with the flag set.
6457 ** New module (ice-9 format), implementing the Common Lisp `format' function.
6459 This code, and the documentation for it that appears here, was
6460 borrowed from SLIB, with minor adaptations for Guile.
6462 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
6463 An almost complete implementation of Common LISP format description
6464 according to the CL reference book `Common LISP' from Guy L.
6465 Steele, Digital Press. Backward compatible to most of the
6466 available Scheme format implementations.
6468 Returns `#t', `#f' or a string; has side effect of printing
6469 according to FORMAT-STRING. If DESTINATION is `#t', the output is
6470 to the current output port and `#t' is returned. If DESTINATION
6471 is `#f', a formatted string is returned as the result of the call.
6472 NEW: If DESTINATION is a string, DESTINATION is regarded as the
6473 format string; FORMAT-STRING is then the first argument and the
6474 output is returned as a string. If DESTINATION is a number, the
6475 output is to the current error port if available by the
6476 implementation. Otherwise DESTINATION must be an output port and
6479 FORMAT-STRING must be a string. In case of a formatting error
6480 format returns `#f' and prints a message on the current output or
6481 error port. Characters are output as if the string were output by
6482 the `display' function with the exception of those prefixed by a
6483 tilde (~). For a detailed description of the FORMAT-STRING syntax
6484 please consult a Common LISP format reference manual. For a test
6485 suite to verify this format implementation load `formatst.scm'.
6486 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
6488 Note: `format' is not reentrant, i.e. only one `format'-call may
6489 be executed at a time.
6492 *** Format Specification (Format version 3.0)
6494 Please consult a Common LISP format reference manual for a detailed
6495 description of the format string syntax. For a demonstration of the
6496 implemented directives see `formatst.scm'.
6498 This implementation supports directive parameters and modifiers (`:'
6499 and `@' characters). Multiple parameters must be separated by a comma
6500 (`,'). Parameters can be numerical parameters (positive or negative),
6501 character parameters (prefixed by a quote character (`''), variable
6502 parameters (`v'), number of rest arguments parameter (`#'), empty and
6503 default parameters. Directive characters are case independent. The
6504 general form of a directive is:
6506 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
6508 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
6510 *** Implemented CL Format Control Directives
6512 Documentation syntax: Uppercase characters represent the
6513 corresponding control directive characters. Lowercase characters
6514 represent control directive parameter descriptions.
6517 Any (print as `display' does).
6521 `~MINCOL,COLINC,MINPAD,PADCHARA'
6525 S-expression (print as `write' does).
6529 `~MINCOL,COLINC,MINPAD,PADCHARS'
6535 print number sign always.
6538 print comma separated.
6540 `~MINCOL,PADCHAR,COMMACHARD'
6546 print number sign always.
6549 print comma separated.
6551 `~MINCOL,PADCHAR,COMMACHARX'
6557 print number sign always.
6560 print comma separated.
6562 `~MINCOL,PADCHAR,COMMACHARO'
6568 print number sign always.
6571 print comma separated.
6573 `~MINCOL,PADCHAR,COMMACHARB'
6578 `~N,MINCOL,PADCHAR,COMMACHARR'
6582 print a number as a Roman numeral.
6585 print a number as an "old fashioned" Roman numeral.
6588 print a number as an ordinal English number.
6591 print a number as a cardinal English number.
6596 prints `y' and `ies'.
6599 as `~P but jumps 1 argument backward.'
6602 as `~@P but jumps 1 argument backward.'
6607 prints a character as the reader can understand it (i.e. `#\'
6611 prints a character as emacs does (eg. `^C' for ASCII 03).
6614 Fixed-format floating-point (prints a flonum like MMM.NNN).
6615 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
6617 If the number is positive a plus sign is printed.
6620 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
6621 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
6623 If the number is positive a plus sign is printed.
6626 General floating-point (prints a flonum either fixed or
6628 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
6630 If the number is positive a plus sign is printed.
6633 Dollars floating-point (prints a flonum in fixed with signs
6635 `~DIGITS,SCALE,WIDTH,PADCHAR$'
6637 If the number is positive a plus sign is printed.
6640 A sign is always printed and appears before the padding.
6643 The sign appears before the padding.
6651 print newline if not at the beginning of the output line.
6653 prints `~&' and then N-1 newlines.
6658 print N page separators.
6668 newline is ignored, white space left.
6671 newline is left, white space ignored.
6676 relative tabulation.
6682 Indirection (expects indirect arguments as a list).
6684 extracts indirect arguments from format arguments.
6687 Case conversion (converts by `string-downcase').
6689 converts by `string-capitalize'.
6692 converts by `string-capitalize-first'.
6695 converts by `string-upcase'.
6698 Argument Jumping (jumps 1 argument forward).
6700 jumps N arguments forward.
6703 jumps 1 argument backward.
6706 jumps N arguments backward.
6709 jumps to the 0th argument.
6712 jumps to the Nth argument (beginning from 0)
6714 `~[STR0~;STR1~;...~;STRN~]'
6715 Conditional Expression (numerical clause conditional).
6717 take argument from N.
6720 true test conditional.
6723 if-else-then conditional.
6729 default clause follows.
6732 Iteration (args come from the next argument (a list)).
6734 at most N iterations.
6737 args from next arg (a list of lists).
6740 args from the rest of arguments.
6743 args from the rest args (lists).
6754 aborts if N <= M <= K
6756 *** Not Implemented CL Format Control Directives
6759 print `#f' as an empty list (see below).
6762 print `#f' as an empty list (see below).
6768 (sorry I don't understand its semantics completely)
6770 *** Extended, Replaced and Additional Control Directives
6772 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
6773 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
6774 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
6775 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
6776 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
6777 COMMAWIDTH is the number of characters between two comma
6781 print a R4RS complex number as `~F~@Fi' with passed parameters for
6785 Pretty print formatting of an argument for scheme code lists.
6791 Flushes the output if format DESTINATION is a port.
6794 Print a `#\space' character
6796 print N `#\space' characters.
6799 Print a `#\tab' character
6801 print N `#\tab' characters.
6804 Takes N as an integer representation for a character. No arguments
6805 are consumed. N is converted to a character by `integer->char'. N
6806 must be a positive decimal number.
6809 Print out readproof. Prints out internal objects represented as
6810 `#<...>' as strings `"#<...>"' so that the format output can always
6811 be processed by `read'.
6814 Print out readproof. Prints out internal objects represented as
6815 `#<...>' as strings `"#<...>"' so that the format output can always
6816 be processed by `read'.
6819 Prints information and a copyright notice on the format
6822 prints format version.
6825 may also print number strings, i.e. passing a number as a string
6826 and format it accordingly.
6828 *** Configuration Variables
6830 The format module exports some configuration variables to suit the
6831 systems and users needs. There should be no modification necessary for
6832 the configuration that comes with Guile. Format detects automatically
6833 if the running scheme system implements floating point numbers and
6836 format:symbol-case-conv
6837 Symbols are converted by `symbol->string' so the case type of the
6838 printed symbols is implementation dependent.
6839 `format:symbol-case-conv' is a one arg closure which is either
6840 `#f' (no conversion), `string-upcase', `string-downcase' or
6841 `string-capitalize'. (default `#f')
6843 format:iobj-case-conv
6844 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
6845 implementation internal objects. (default `#f')
6848 The character prefixing the exponent value in `~E' printing.
6851 *** Compatibility With Other Format Implementations
6857 Downward compatible except for padding support and `~A', `~S',
6858 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
6859 `printf' padding support which is completely replaced by the CL
6860 `format' padding style.
6863 Downward compatible except for `~', which is not documented
6864 (ignores all characters inside the format string up to a newline
6865 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
6866 numerical and variable parameters and `:/@' modifiers in the CL
6870 Downward compatible except for `~A' and `~S' which print in
6871 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
6872 directive parameters or modifiers)).
6875 Downward compatible except for an optional destination parameter:
6876 S2C accepts a format call without a destination which returns a
6877 formatted string. This is equivalent to a #f destination in S2C.
6878 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
6879 parameters or modifiers)).
6882 ** Changes to string-handling functions.
6884 These functions were added to support the (ice-9 format) module, above.
6886 *** New function: string-upcase STRING
6887 *** New function: string-downcase STRING
6889 These are non-destructive versions of the existing string-upcase! and
6890 string-downcase! functions.
6892 *** New function: string-capitalize! STRING
6893 *** New function: string-capitalize STRING
6895 These functions convert the first letter of each word in the string to
6898 (string-capitalize "howdy there")
6901 As with the other functions, string-capitalize! modifies the string in
6902 place, while string-capitalize returns a modified copy of its argument.
6904 *** New function: string-ci->symbol STRING
6906 Return a symbol whose name is STRING, but having the same case as if
6907 the symbol had be read by `read'.
6909 Guile can be configured to be sensitive or insensitive to case
6910 differences in Scheme identifiers. If Guile is case-insensitive, all
6911 symbols are converted to lower case on input. The `string-ci->symbol'
6912 function returns a symbol whose name in STRING, transformed as Guile
6913 would if STRING were input.
6915 *** New function: substring-move! STRING1 START END STRING2 START
6917 Copy the substring of STRING1 from START (inclusive) to END
6918 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
6919 string, and the source and destination areas may overlap; in all
6920 cases, the function behaves as if all the characters were copied
6923 *** Extended functions: substring-move-left! substring-move-right!
6925 These functions now correctly copy arbitrarily overlapping substrings;
6926 they are both synonyms for substring-move!.
6929 ** New module (ice-9 getopt-long), with the function `getopt-long'.
6931 getopt-long is a function for parsing command-line arguments in a
6932 manner consistent with other GNU programs.
6934 (getopt-long ARGS GRAMMAR)
6935 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
6937 ARGS should be a list of strings. Its first element should be the
6938 name of the program; subsequent elements should be the arguments
6939 that were passed to the program on the command line. The
6940 `program-arguments' procedure returns a list of this form.
6942 GRAMMAR is a list of the form:
6943 ((OPTION (PROPERTY VALUE) ...) ...)
6945 Each OPTION should be a symbol. `getopt-long' will accept a
6946 command-line option named `--OPTION'.
6947 Each option can have the following (PROPERTY VALUE) pairs:
6949 (single-char CHAR) --- Accept `-CHAR' as a single-character
6950 equivalent to `--OPTION'. This is how to specify traditional
6952 (required? BOOL) --- If BOOL is true, the option is required.
6953 getopt-long will raise an error if it is not found in ARGS.
6954 (value BOOL) --- If BOOL is #t, the option accepts a value; if
6955 it is #f, it does not; and if it is the symbol
6956 `optional', the option may appear in ARGS with or
6958 (predicate FUNC) --- If the option accepts a value (i.e. you
6959 specified `(value #t)' for this option), then getopt
6960 will apply FUNC to the value, and throw an exception
6961 if it returns #f. FUNC should be a procedure which
6962 accepts a string and returns a boolean value; you may
6963 need to use quasiquotes to get it into GRAMMAR.
6965 The (PROPERTY VALUE) pairs may occur in any order, but each
6966 property may occur only once. By default, options do not have
6967 single-character equivalents, are not required, and do not take
6970 In ARGS, single-character options may be combined, in the usual
6971 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
6972 accepts values, then it must be the last option in the
6973 combination; the value is the next argument. So, for example, using
6974 the following grammar:
6975 ((apples (single-char #\a))
6976 (blimps (single-char #\b) (value #t))
6977 (catalexis (single-char #\c) (value #t)))
6978 the following argument lists would be acceptable:
6979 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
6980 for "blimps" and "catalexis")
6981 ("-ab" "bang" "-c" "couth") (same)
6982 ("-ac" "couth" "-b" "bang") (same)
6983 ("-abc" "couth" "bang") (an error, since `-b' is not the
6984 last option in its combination)
6986 If an option's value is optional, then `getopt-long' decides
6987 whether it has a value by looking at what follows it in ARGS. If
6988 the next element is a string, and it does not appear to be an
6989 option itself, then that string is the option's value.
6991 The value of a long option can appear as the next element in ARGS,
6992 or it can follow the option name, separated by an `=' character.
6993 Thus, using the same grammar as above, the following argument lists
6995 ("--apples" "Braeburn" "--blimps" "Goodyear")
6996 ("--apples=Braeburn" "--blimps" "Goodyear")
6997 ("--blimps" "Goodyear" "--apples=Braeburn")
6999 If the option "--" appears in ARGS, argument parsing stops there;
7000 subsequent arguments are returned as ordinary arguments, even if
7001 they resemble options. So, in the argument list:
7002 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
7003 `getopt-long' will recognize the `apples' option as having the
7004 value "Granny Smith", but it will not recognize the `blimp'
7005 option; it will return the strings "--blimp" and "Goodyear" as
7006 ordinary argument strings.
7008 The `getopt-long' function returns the parsed argument list as an
7009 assocation list, mapping option names --- the symbols from GRAMMAR
7010 --- onto their values, or #t if the option does not accept a value.
7011 Unused options do not appear in the alist.
7013 All arguments that are not the value of any option are returned
7014 as a list, associated with the empty list.
7016 `getopt-long' throws an exception if:
7017 - it finds an unrecognized option in ARGS
7018 - a required option is omitted
7019 - an option that requires an argument doesn't get one
7020 - an option that doesn't accept an argument does get one (this can
7021 only happen using the long option `--opt=value' syntax)
7022 - an option predicate fails
7027 `((lockfile-dir (required? #t)
7030 (predicate ,file-is-directory?))
7031 (verbose (required? #f)
7034 (x-includes (single-char #\x))
7035 (rnet-server (single-char #\y)
7036 (predicate ,string?))))
7038 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
7039 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
7041 => ((() "foo1" "-fred" "foo2" "foo3")
7042 (rnet-server . "lamprod")
7043 (x-includes . "/usr/include")
7044 (lockfile-dir . "/tmp")
7047 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
7049 It will be removed in a few releases.
7051 ** New syntax: lambda*
7052 ** New syntax: define*
7053 ** New syntax: define*-public
7054 ** New syntax: defmacro*
7055 ** New syntax: defmacro*-public
7056 Guile now supports optional arguments.
7058 `lambda*', `define*', `define*-public', `defmacro*' and
7059 `defmacro*-public' are identical to the non-* versions except that
7060 they use an extended type of parameter list that has the following BNF
7061 syntax (parentheses are literal, square brackets indicate grouping,
7062 and `*', `+' and `?' have the usual meaning):
7064 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
7065 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
7066 [[#&rest identifier]|[. identifier]]? ) | [identifier]
7068 ext-var-decl ::= identifier | ( identifier expression )
7070 The semantics are best illustrated with the following documentation
7071 and examples for `lambda*':
7074 lambda extended for optional and keyword arguments
7076 lambda* creates a procedure that takes optional arguments. These
7077 are specified by putting them inside brackets at the end of the
7078 paramater list, but before any dotted rest argument. For example,
7079 (lambda* (a b #&optional c d . e) '())
7080 creates a procedure with fixed arguments a and b, optional arguments c
7081 and d, and rest argument e. If the optional arguments are omitted
7082 in a call, the variables for them are unbound in the procedure. This
7083 can be checked with the bound? macro.
7085 lambda* can also take keyword arguments. For example, a procedure
7087 (lambda* (#&key xyzzy larch) '())
7088 can be called with any of the argument lists (#:xyzzy 11)
7089 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
7090 are given as keywords are bound to values.
7092 Optional and keyword arguments can also be given default values
7093 which they take on when they are not present in a call, by giving a
7094 two-item list in place of an optional argument, for example in:
7095 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
7096 foo is a fixed argument, bar is an optional argument with default
7097 value 42, and baz is a keyword argument with default value 73.
7098 Default value expressions are not evaluated unless they are needed
7099 and until the procedure is called.
7101 lambda* now supports two more special parameter list keywords.
7103 lambda*-defined procedures now throw an error by default if a
7104 keyword other than one of those specified is found in the actual
7105 passed arguments. However, specifying #&allow-other-keys
7106 immediately after the kyword argument declarations restores the
7107 previous behavior of ignoring unknown keywords. lambda* also now
7108 guarantees that if the same keyword is passed more than once, the
7109 last one passed is the one that takes effect. For example,
7110 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
7111 #:heads 37 #:tails 42 #:heads 99)
7112 would result in (99 47) being displayed.
7114 #&rest is also now provided as a synonym for the dotted syntax rest
7115 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
7116 all respects to lambda*. This is provided for more similarity to DSSSL,
7117 MIT-Scheme and Kawa among others, as well as for refugees from other
7120 Further documentation may be found in the optargs.scm file itself.
7122 The optional argument module also exports the macros `let-optional',
7123 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
7124 are not documented here because they may be removed in the future, but
7125 full documentation is still available in optargs.scm.
7127 ** New syntax: and-let*
7128 Guile now supports the `and-let*' form, described in the draft SRFI-2.
7130 Syntax: (land* (<clause> ...) <body> ...)
7131 Each <clause> should have one of the following forms:
7132 (<variable> <expression>)
7135 Each <variable> or <bound-variable> should be an identifier. Each
7136 <expression> should be a valid expression. The <body> should be a
7137 possibly empty sequence of expressions, like the <body> of a
7140 Semantics: A LAND* expression is evaluated by evaluating the
7141 <expression> or <bound-variable> of each of the <clause>s from
7142 left to right. The value of the first <expression> or
7143 <bound-variable> that evaluates to a false value is returned; the
7144 remaining <expression>s and <bound-variable>s are not evaluated.
7145 The <body> forms are evaluated iff all the <expression>s and
7146 <bound-variable>s evaluate to true values.
7148 The <expression>s and the <body> are evaluated in an environment
7149 binding each <variable> of the preceding (<variable> <expression>)
7150 clauses to the value of the <expression>. Later bindings
7151 shadow earlier bindings.
7153 Guile's and-let* macro was contributed by Michael Livshin.
7155 ** New sorting functions
7157 *** New function: sorted? SEQUENCE LESS?
7158 Returns `#t' when the sequence argument is in non-decreasing order
7159 according to LESS? (that is, there is no adjacent pair `... x y
7160 ...' for which `(less? y x)').
7162 Returns `#f' when the sequence contains at least one out-of-order
7163 pair. It is an error if the sequence is neither a list nor a
7166 *** New function: merge LIST1 LIST2 LESS?
7167 LIST1 and LIST2 are sorted lists.
7168 Returns the sorted list of all elements in LIST1 and LIST2.
7170 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
7171 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
7172 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
7173 (Here "<" should read "comes before".)
7175 *** New procedure: merge! LIST1 LIST2 LESS?
7176 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
7177 the result. If the code is compiled, and LESS? constructs no new
7178 pairs, no pairs at all will be allocated. The first pair of the
7179 result will be either the first pair of LIST1 or the first pair of
7182 *** New function: sort SEQUENCE LESS?
7183 Accepts either a list or a vector, and returns a new sequence
7184 which is sorted. The new sequence is the same type as the input.
7185 Always `(sorted? (sort sequence less?) less?)'. The original
7186 sequence is not altered in any way. The new sequence shares its
7187 elements with the old one; no elements are copied.
7189 *** New procedure: sort! SEQUENCE LESS
7190 Returns its sorted result in the original boxes. No new storage is
7191 allocated at all. Proper usage: (set! slist (sort! slist <))
7193 *** New function: stable-sort SEQUENCE LESS?
7194 Similar to `sort' but stable. That is, if "equal" elements are
7195 ordered a < b in the original sequence, they will have the same order
7198 *** New function: stable-sort! SEQUENCE LESS?
7199 Similar to `sort!' but stable.
7200 Uses temporary storage when sorting vectors.
7202 *** New functions: sort-list, sort-list!
7203 Added for compatibility with scsh.
7205 ** New built-in random number support
7207 *** New function: random N [STATE]
7208 Accepts a positive integer or real N and returns a number of the
7209 same type between zero (inclusive) and N (exclusive). The values
7210 returned have a uniform distribution.
7212 The optional argument STATE must be of the type produced by
7213 `copy-random-state' or `seed->random-state'. It defaults to the value
7214 of the variable `*random-state*'. This object is used to maintain the
7215 state of the pseudo-random-number generator and is altered as a side
7216 effect of the `random' operation.
7218 *** New variable: *random-state*
7219 Holds a data structure that encodes the internal state of the
7220 random-number generator that `random' uses by default. The nature
7221 of this data structure is implementation-dependent. It may be
7222 printed out and successfully read back in, but may or may not
7223 function correctly as a random-number state object in another
7226 *** New function: copy-random-state [STATE]
7227 Returns a new object of type suitable for use as the value of the
7228 variable `*random-state*' and as a second argument to `random'.
7229 If argument STATE is given, a copy of it is returned. Otherwise a
7230 copy of `*random-state*' is returned.
7232 *** New function: seed->random-state SEED
7233 Returns a new object of type suitable for use as the value of the
7234 variable `*random-state*' and as a second argument to `random'.
7235 SEED is a string or a number. A new state is generated and
7236 initialized using SEED.
7238 *** New function: random:uniform [STATE]
7239 Returns an uniformly distributed inexact real random number in the
7240 range between 0 and 1.
7242 *** New procedure: random:solid-sphere! VECT [STATE]
7243 Fills VECT with inexact real random numbers the sum of whose
7244 squares is less than 1.0. Thinking of VECT as coordinates in
7245 space of dimension N = `(vector-length VECT)', the coordinates are
7246 uniformly distributed within the unit N-shere. The sum of the
7247 squares of the numbers is returned. VECT can be either a vector
7248 or a uniform vector of doubles.
7250 *** New procedure: random:hollow-sphere! VECT [STATE]
7251 Fills VECT with inexact real random numbers the sum of whose squares
7252 is equal to 1.0. Thinking of VECT as coordinates in space of
7253 dimension n = `(vector-length VECT)', the coordinates are uniformly
7254 distributed over the surface of the unit n-shere. VECT can be either
7255 a vector or a uniform vector of doubles.
7257 *** New function: random:normal [STATE]
7258 Returns an inexact real in a normal distribution with mean 0 and
7259 standard deviation 1. For a normal distribution with mean M and
7260 standard deviation D use `(+ M (* D (random:normal)))'.
7262 *** New procedure: random:normal-vector! VECT [STATE]
7263 Fills VECT with inexact real random numbers which are independent and
7264 standard normally distributed (i.e., with mean 0 and variance 1).
7265 VECT can be either a vector or a uniform vector of doubles.
7267 *** New function: random:exp STATE
7268 Returns an inexact real in an exponential distribution with mean 1.
7269 For an exponential distribution with mean U use (* U (random:exp)).
7271 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
7273 These functions now operate on numbers in the range of a C unsigned
7276 These functions used to operate on numbers in the range of a C signed
7277 long; however, this seems inappropriate, because Guile integers don't
7280 ** New function: make-guardian
7281 This is an implementation of guardians as described in
7282 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
7283 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
7284 Programming Language Design and Implementation, June 1993
7285 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
7287 ** New functions: delq1!, delv1!, delete1!
7288 These procedures behave similar to delq! and friends but delete only
7289 one object if at all.
7291 ** New function: unread-string STRING PORT
7292 Unread STRING to PORT, that is, push it back onto the port so that
7293 next read operation will work on the pushed back characters.
7295 ** unread-char can now be called multiple times
7296 If unread-char is called multiple times, the unread characters will be
7297 read again in last-in first-out order.
7299 ** the procedures uniform-array-read! and uniform-array-write! now
7300 work on any kind of port, not just ports which are open on a file.
7302 ** Now 'l' in a port mode requests line buffering.
7304 ** The procedure truncate-file now works on string ports as well
7305 as file ports. If the size argument is omitted, the current
7306 file position is used.
7308 ** new procedure: seek PORT/FDES OFFSET WHENCE
7309 The arguments are the same as for the old fseek procedure, but it
7310 works on string ports as well as random-access file ports.
7312 ** the fseek procedure now works on string ports, since it has been
7313 redefined using seek.
7315 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
7316 size is not supplied.
7318 ** the newline procedure no longer flushes the port if it's not
7319 line-buffered: previously it did if it was the current output port.
7321 ** open-pipe and close-pipe are no longer primitive procedures, but
7322 an emulation can be obtained using `(use-modules (ice-9 popen))'.
7324 ** the freopen procedure has been removed.
7326 ** new procedure: drain-input PORT
7327 Drains PORT's read buffers (including any pushed-back characters)
7328 and returns the contents as a single string.
7330 ** New function: map-in-order PROC LIST1 LIST2 ...
7331 Version of `map' which guarantees that the procedure is applied to the
7332 lists in serial order.
7334 ** Renamed `serial-array-copy!' and `serial-array-map!' to
7335 `array-copy-in-order!' and `array-map-in-order!'. The old names are
7336 now obsolete and will go away in release 1.5.
7338 ** New syntax: collect BODY1 ...
7339 Version of `begin' which returns a list of the results of the body
7340 forms instead of the result of the last body form. In contrast to
7341 `begin', `collect' allows an empty body.
7343 ** New functions: read-history FILENAME, write-history FILENAME
7344 Read/write command line history from/to file. Returns #t on success
7345 and #f if an error occured.
7347 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
7349 These procedures return a list of definitions available in the specified
7350 argument, a relative module reference. In the case of no argument,
7351 `(current-module)' is now consulted for definitions to return, instead
7352 of simply returning #f, the former behavior.
7354 ** The #/ syntax for lists is no longer supported.
7356 Earlier versions of Scheme accepted this syntax, but printed a
7359 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
7361 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
7364 * Changes to the gh_ interface
7368 Now takes a second argument which is the result array. If this
7369 pointer is NULL, a new array is malloced (the old behaviour).
7371 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
7372 gh_scm2shorts, gh_scm2longs, gh_scm2floats
7376 * Changes to the scm_ interface
7378 ** Function: scm_make_named_hook (char* name, int n_args)
7380 Creates a hook in the same way as make-hook above but also
7381 binds a variable named NAME to it.
7383 This is the typical way of creating a hook from C code.
7385 Currently, the variable is created in the "current" module. This
7386 might change when we get the new module system.
7388 ** The smob interface
7390 The interface for creating smobs has changed. For documentation, see
7391 data-rep.info (made from guile-core/doc/data-rep.texi).
7393 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
7395 >>> This function will be removed in 1.3.4. <<<
7399 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
7400 This function adds a new smob type, named NAME, with instance size
7401 SIZE to the system. The return value is a tag that is used in
7402 creating instances of the type. If SIZE is 0, then no memory will
7403 be allocated when instances of the smob are created, and nothing
7404 will be freed by the default free function.
7406 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
7407 This function sets the smob marking procedure for the smob type
7408 specified by the tag TC. TC is the tag returned by
7409 `scm_make_smob_type'.
7411 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
7412 This function sets the smob freeing procedure for the smob type
7413 specified by the tag TC. TC is the tag returned by
7414 `scm_make_smob_type'.
7416 *** Function: void scm_set_smob_print (tc, print)
7418 - Function: void scm_set_smob_print (long tc,
7419 scm_sizet (*print) (SCM,
7423 This function sets the smob printing procedure for the smob type
7424 specified by the tag TC. TC is the tag returned by
7425 `scm_make_smob_type'.
7427 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
7428 This function sets the smob equality-testing predicate for the
7429 smob type specified by the tag TC. TC is the tag returned by
7430 `scm_make_smob_type'.
7432 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
7433 Make VALUE contain a smob instance of the type with type code TC and
7434 smob data DATA. VALUE must be previously declared as C type `SCM'.
7436 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
7437 This macro expands to a block of code that creates a smob instance
7438 of the type with type code TC and smob data DATA, and returns that
7439 `SCM' value. It should be the last piece of code in a block.
7441 ** The interfaces for using I/O ports and implementing port types
7442 (ptobs) have changed significantly. The new interface is based on
7443 shared access to buffers and a new set of ptob procedures.
7445 *** scm_newptob has been removed
7449 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
7451 - Function: SCM scm_make_port_type (char *type_name,
7452 int (*fill_buffer) (SCM port),
7453 void (*write_flush) (SCM port));
7455 Similarly to the new smob interface, there is a set of function
7456 setters by which the user can customize the behaviour of his port
7457 type. See ports.h (scm_set_port_XXX).
7459 ** scm_strport_to_string: New function: creates a new string from
7460 a string port's buffer.
7462 ** Plug in interface for random number generators
7463 The variable `scm_the_rng' in random.c contains a value and three
7464 function pointers which together define the current random number
7465 generator being used by the Scheme level interface and the random
7466 number library functions.
7468 The user is free to replace the default generator with the generator
7471 *** Variable: size_t scm_the_rng.rstate_size
7472 The size of the random state type used by the current RNG
7475 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
7476 Given the random STATE, return 32 random bits.
7478 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
7479 Seed random state STATE using string S of length N.
7481 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
7482 Given random state STATE, return a malloced copy.
7485 The default RNG is the MWC (Multiply With Carry) random number
7486 generator described by George Marsaglia at the Department of
7487 Statistics and Supercomputer Computations Research Institute, The
7488 Florida State University (http://stat.fsu.edu/~geo).
7490 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
7491 passes all tests in the DIEHARD test suite
7492 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
7493 costs one multiply and one add on platforms which either supports long
7494 longs (gcc does this on most systems) or have 64 bit longs. The cost
7495 is four multiply on other systems but this can be optimized by writing
7496 scm_i_uniform32 in assembler.
7498 These functions are provided through the scm_the_rng interface for use
7499 by libguile and the application.
7501 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
7502 Given the random STATE, return 32 random bits.
7503 Don't use this function directly. Instead go through the plugin
7504 interface (see "Plug in interface" above).
7506 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
7507 Initialize STATE using SEED of length N.
7509 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
7510 Return a malloc:ed copy of STATE. This function can easily be re-used
7511 in the interfaces to other RNGs.
7513 ** Random number library functions
7514 These functions use the current RNG through the scm_the_rng interface.
7515 It might be a good idea to use these functions from your C code so
7516 that only one random generator is used by all code in your program.
7518 The default random state is stored in:
7520 *** Variable: SCM scm_var_random_state
7521 Contains the vcell of the Scheme variable "*random-state*" which is
7522 used as default state by all random number functions in the Scheme
7527 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
7529 *** Function: scm_rstate *scm_c_default_rstate (void)
7530 This is a convenience function which returns the value of
7531 scm_var_random_state. An error message is generated if this value
7532 isn't a random state.
7534 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
7535 Make a new random state from the string SEED of length LENGTH.
7537 It is generally not a good idea to use multiple random states in a
7538 program. While subsequent random numbers generated from one random
7539 state are guaranteed to be reasonably independent, there is no such
7540 guarantee for numbers generated from different random states.
7542 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
7543 Return 32 random bits.
7545 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
7546 Return a sample from the uniform(0,1) distribution.
7548 *** Function: double scm_c_normal01 (scm_rstate *STATE)
7549 Return a sample from the normal(0,1) distribution.
7551 *** Function: double scm_c_exp1 (scm_rstate *STATE)
7552 Return a sample from the exp(1) distribution.
7554 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
7555 Return a sample from the discrete uniform(0,M) distribution.
7557 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
7558 Return a sample from the discrete uniform(0,M) distribution.
7559 M must be a bignum object. The returned value may be an INUM.
7563 Changes in Guile 1.3 (released Monday, October 19, 1998):
7565 * Changes to the distribution
7567 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
7568 To avoid conflicts, programs should name environment variables after
7569 themselves, except when there's a common practice establishing some
7572 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
7573 giving the former precedence, and printing a warning message if the
7574 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
7576 ** The header files related to multi-byte characters have been removed.
7577 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
7578 which referred to these explicitly will probably need to be rewritten,
7579 since the support for the variant string types has been removed; see
7582 ** The header files append.h and sequences.h have been removed. These
7583 files implemented non-R4RS operations which would encourage
7584 non-portable programming style and less easy-to-read code.
7586 * Changes to the stand-alone interpreter
7588 ** New procedures have been added to implement a "batch mode":
7590 *** Function: batch-mode?
7592 Returns a boolean indicating whether the interpreter is in batch
7595 *** Function: set-batch-mode?! ARG
7597 If ARG is true, switches the interpreter to batch mode. The `#f'
7598 case has not been implemented.
7600 ** Guile now provides full command-line editing, when run interactively.
7601 To use this feature, you must have the readline library installed.
7602 The Guile build process will notice it, and automatically include
7605 The readline library is available via anonymous FTP from any GNU
7606 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
7608 ** the-last-stack is now a fluid.
7610 * Changes to the procedure for linking libguile with your programs
7612 ** You can now use the `guile-config' utility to build programs that use Guile.
7614 Guile now includes a command-line utility called `guile-config', which
7615 can provide information about how to compile and link programs that
7618 *** `guile-config compile' prints any C compiler flags needed to use Guile.
7619 You should include this command's output on the command line you use
7620 to compile C or C++ code that #includes the Guile header files. It's
7621 usually just a `-I' flag to help the compiler find the Guile headers.
7624 *** `guile-config link' prints any linker flags necessary to link with Guile.
7626 This command writes to its standard output a list of flags which you
7627 must pass to the linker to link your code against the Guile library.
7628 The flags include '-lguile' itself, any other libraries the Guile
7629 library depends upon, and any `-L' flags needed to help the linker
7630 find those libraries.
7632 For example, here is a Makefile rule that builds a program named 'foo'
7633 from the object files ${FOO_OBJECTS}, and links them against Guile:
7636 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
7638 Previous Guile releases recommended that you use autoconf to detect
7639 which of a predefined set of libraries were present on your system.
7640 It is more robust to use `guile-config', since it records exactly which
7641 libraries the installed Guile library requires.
7643 This was originally called `build-guile', but was renamed to
7644 `guile-config' before Guile 1.3 was released, to be consistent with
7645 the analogous script for the GTK+ GUI toolkit, which is called
7649 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
7651 If you are using the GNU autoconf package to configure your program,
7652 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
7653 (described above) and gather the necessary values for use in your
7656 The GUILE_FLAGS macro expands to configure script code which runs the
7657 `guile-config' script, to find out where Guile's header files and
7658 libraries are installed. It sets two variables, marked for
7659 substitution, as by AC_SUBST.
7661 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
7662 code that uses Guile header files. This is almost always just a
7665 GUILE_LDFLAGS --- flags to pass to the linker to link a
7666 program against Guile. This includes `-lguile' for the Guile
7667 library itself, any libraries that Guile itself requires (like
7668 -lqthreads), and so on. It may also include a -L flag to tell the
7669 compiler where to find the libraries.
7671 GUILE_FLAGS is defined in the file guile.m4, in the top-level
7672 directory of the Guile distribution. You can copy it into your
7673 package's aclocal.m4 file, and then use it in your configure.in file.
7675 If you are using the `aclocal' program, distributed with GNU automake,
7676 to maintain your aclocal.m4 file, the Guile installation process
7677 installs guile.m4 where aclocal will find it. All you need to do is
7678 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
7679 this will copy the definition of GUILE_FLAGS into your aclocal.m4
7683 * Changes to Scheme functions and syntax
7685 ** Multi-byte strings have been removed, as have multi-byte and wide
7686 ports. We felt that these were the wrong approach to
7687 internationalization support.
7689 ** New function: readline [PROMPT]
7690 Read a line from the terminal, and allow the user to edit it,
7691 prompting with PROMPT. READLINE provides a large set of Emacs-like
7692 editing commands, lets the user recall previously typed lines, and
7693 works on almost every kind of terminal, including dumb terminals.
7695 READLINE assumes that the cursor is at the beginning of the line when
7696 it is invoked. Thus, you can't print a prompt yourself, and then call
7697 READLINE; you need to package up your prompt as a string, pass it to
7698 the function, and let READLINE print the prompt itself. This is
7699 because READLINE needs to know the prompt's screen width.
7701 For Guile to provide this function, you must have the readline
7702 library, version 2.1 or later, installed on your system. Readline is
7703 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
7704 any GNU mirror site.
7706 See also ADD-HISTORY function.
7708 ** New function: add-history STRING
7709 Add STRING as the most recent line in the history used by the READLINE
7710 command. READLINE does not add lines to the history itself; you must
7711 call ADD-HISTORY to make previous input available to the user.
7713 ** The behavior of the read-line function has changed.
7715 This function now uses standard C library functions to read the line,
7716 for speed. This means that it doesn not respect the value of
7717 scm-line-incrementors; it assumes that lines are delimited with
7720 (Note that this is read-line, the function that reads a line of text
7721 from a port, not readline, the function that reads a line from a
7722 terminal, providing full editing capabilities.)
7724 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
7726 This module provides some simple argument parsing. It exports one
7729 Function: getopt-gnu-style ARG-LS
7730 Parse a list of program arguments into an alist of option
7733 Each item in the list of program arguments is examined to see if
7734 it meets the syntax of a GNU long-named option. An argument like
7735 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
7736 returned alist, where MUMBLE is a keyword object with the same
7737 name as the argument. An argument like `--MUMBLE=FROB' produces
7738 an element of the form (MUMBLE . FROB), where FROB is a string.
7740 As a special case, the returned alist also contains a pair whose
7741 car is the symbol `rest'. The cdr of this pair is a list
7742 containing all the items in the argument list that are not options
7743 of the form mentioned above.
7745 The argument `--' is treated specially: all items in the argument
7746 list appearing after such an argument are not examined, and are
7747 returned in the special `rest' list.
7749 This function does not parse normal single-character switches.
7750 You will need to parse them out of the `rest' list yourself.
7752 ** The read syntax for byte vectors and short vectors has changed.
7754 Instead of #bytes(...), write #y(...).
7756 Instead of #short(...), write #h(...).
7758 This may seem nutty, but, like the other uniform vectors, byte vectors
7759 and short vectors want to have the same print and read syntax (and,
7760 more basic, want to have read syntax!). Changing the read syntax to
7761 use multiple characters after the hash sign breaks with the
7762 conventions used in R5RS and the conventions used for the other
7763 uniform vectors. It also introduces complexity in the current reader,
7764 both on the C and Scheme levels. (The Right solution is probably to
7765 change the syntax and prototypes for uniform vectors entirely.)
7768 ** The new module (ice-9 session) provides useful interactive functions.
7770 *** New procedure: (apropos REGEXP OPTION ...)
7772 Display a list of top-level variables whose names match REGEXP, and
7773 the modules they are imported from. Each OPTION should be one of the
7776 value --- Show the value of each matching variable.
7777 shadow --- Show bindings shadowed by subsequently imported modules.
7778 full --- Same as both `shadow' and `value'.
7782 guile> (apropos "trace" 'full)
7783 debug: trace #<procedure trace args>
7784 debug: untrace #<procedure untrace args>
7785 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
7786 the-scm-module: before-backtrace-hook ()
7787 the-scm-module: backtrace #<primitive-procedure backtrace>
7788 the-scm-module: after-backtrace-hook ()
7789 the-scm-module: has-shown-backtrace-hint? #f
7792 ** There are new functions and syntax for working with macros.
7794 Guile implements macros as a special object type. Any variable whose
7795 top-level binding is a macro object acts as a macro. The macro object
7796 specifies how the expression should be transformed before evaluation.
7798 *** Macro objects now print in a reasonable way, resembling procedures.
7800 *** New function: (macro? OBJ)
7801 True iff OBJ is a macro object.
7803 *** New function: (primitive-macro? OBJ)
7804 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
7805 macro transformers, implemented in eval.c rather than Scheme code.
7807 Why do we have this function?
7808 - For symmetry with procedure? and primitive-procedure?,
7809 - to allow custom print procedures to tell whether a macro is
7810 primitive, and display it differently, and
7811 - to allow compilers and user-written evaluators to distinguish
7812 builtin special forms from user-defined ones, which could be
7815 *** New function: (macro-type OBJ)
7816 Return a value indicating what kind of macro OBJ is. Possible return
7819 The symbol `syntax' --- a macro created by procedure->syntax.
7820 The symbol `macro' --- a macro created by procedure->macro.
7821 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
7822 The boolean #f --- if OBJ is not a macro object.
7824 *** New function: (macro-name MACRO)
7825 Return the name of the macro object MACRO's procedure, as returned by
7828 *** New function: (macro-transformer MACRO)
7829 Return the transformer procedure for MACRO.
7831 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
7833 Specify a new macro expander to use in the current module. Each
7834 MODULE is a module name, with the same meaning as in the `use-modules'
7835 form; each named module's exported bindings are added to the current
7836 top-level environment. TRANSFORMER is an expression evaluated in the
7837 resulting environment which must yield a procedure to use as the
7838 module's eval transformer: every expression evaluated in this module
7839 is passed to this function, and the result passed to the Guile
7842 *** macro-eval! is removed. Use local-eval instead.
7844 ** Some magic has been added to the printer to better handle user
7845 written printing routines (like record printers, closure printers).
7847 The problem is that these user written routines must have access to
7848 the current `print-state' to be able to handle fancy things like
7849 detection of circular references. These print-states have to be
7850 passed to the builtin printing routines (display, write, etc) to
7851 properly continue the print chain.
7853 We didn't want to change all existing print code so that it
7854 explicitly passes thru a print state in addition to a port. Instead,
7855 we extented the possible values that the builtin printing routines
7856 accept as a `port'. In addition to a normal port, they now also take
7857 a pair of a normal port and a print-state. Printing will go to the
7858 port and the print-state will be used to control the detection of
7859 circular references, etc. If the builtin function does not care for a
7860 print-state, it is simply ignored.
7862 User written callbacks are now called with such a pair as their
7863 `port', but because every function now accepts this pair as a PORT
7864 argument, you don't have to worry about that. In fact, it is probably
7865 safest to not check for these pairs.
7867 However, it is sometimes necessary to continue a print chain on a
7868 different port, for example to get a intermediate string
7869 representation of the printed value, mangle that string somehow, and
7870 then to finally print the mangled string. Use the new function
7872 inherit-print-state OLD-PORT NEW-PORT
7874 for this. It constructs a new `port' that prints to NEW-PORT but
7875 inherits the print-state of OLD-PORT.
7877 ** struct-vtable-offset renamed to vtable-offset-user
7879 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
7881 ** There is now a third optional argument to make-vtable-vtable
7882 (and fourth to make-struct) when constructing new types (vtables).
7883 This argument initializes field vtable-index-printer of the vtable.
7885 ** The detection of circular references has been extended to structs.
7886 That is, a structure that -- in the process of being printed -- prints
7887 itself does not lead to infinite recursion.
7889 ** There is now some basic support for fluids. Please read
7890 "libguile/fluid.h" to find out more. It is accessible from Scheme with
7891 the following functions and macros:
7893 Function: make-fluid
7895 Create a new fluid object. Fluids are not special variables or
7896 some other extension to the semantics of Scheme, but rather
7897 ordinary Scheme objects. You can store them into variables (that
7898 are still lexically scoped, of course) or into any other place you
7899 like. Every fluid has a initial value of `#f'.
7901 Function: fluid? OBJ
7903 Test whether OBJ is a fluid.
7905 Function: fluid-ref FLUID
7906 Function: fluid-set! FLUID VAL
7908 Access/modify the fluid FLUID. Modifications are only visible
7909 within the current dynamic root (that includes threads).
7911 Function: with-fluids* FLUIDS VALUES THUNK
7913 FLUIDS is a list of fluids and VALUES a corresponding list of
7914 values for these fluids. Before THUNK gets called the values are
7915 installed in the fluids and the old values of the fluids are
7916 saved in the VALUES list. When the flow of control leaves THUNK
7917 or reenters it, the values get swapped again. You might think of
7918 this as a `safe-fluid-excursion'. Note that the VALUES list is
7919 modified by `with-fluids*'.
7921 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
7923 The same as `with-fluids*' but with a different syntax. It looks
7924 just like `let', but both FLUID and VALUE are evaluated. Remember,
7925 fluids are not special variables but ordinary objects. FLUID
7926 should evaluate to a fluid.
7928 ** Changes to system call interfaces:
7930 *** close-port, close-input-port and close-output-port now return a
7931 boolean instead of an `unspecified' object. #t means that the port
7932 was successfully closed, while #f means it was already closed. It is
7933 also now possible for these procedures to raise an exception if an
7934 error occurs (some errors from write can be delayed until close.)
7936 *** the first argument to chmod, fcntl, ftell and fseek can now be a
7939 *** the third argument to fcntl is now optional.
7941 *** the first argument to chown can now be a file descriptor or a port.
7943 *** the argument to stat can now be a port.
7945 *** The following new procedures have been added (most use scsh
7948 *** procedure: close PORT/FD
7949 Similar to close-port (*note close-port: Closing Ports.), but also
7950 works on file descriptors. A side effect of closing a file
7951 descriptor is that any ports using that file descriptor are moved
7952 to a different file descriptor and have their revealed counts set
7955 *** procedure: port->fdes PORT
7956 Returns the integer file descriptor underlying PORT. As a side
7957 effect the revealed count of PORT is incremented.
7959 *** procedure: fdes->ports FDES
7960 Returns a list of existing ports which have FDES as an underlying
7961 file descriptor, without changing their revealed counts.
7963 *** procedure: fdes->inport FDES
7964 Returns an existing input port which has FDES as its underlying
7965 file descriptor, if one exists, and increments its revealed count.
7966 Otherwise, returns a new input port with a revealed count of 1.
7968 *** procedure: fdes->outport FDES
7969 Returns an existing output port which has FDES as its underlying
7970 file descriptor, if one exists, and increments its revealed count.
7971 Otherwise, returns a new output port with a revealed count of 1.
7973 The next group of procedures perform a `dup2' system call, if NEWFD
7974 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
7975 duplicated can be supplied as an integer or contained in a port. The
7976 type of value returned varies depending on which procedure is used.
7978 All procedures also have the side effect when performing `dup2' that
7979 any ports using NEWFD are moved to a different file descriptor and have
7980 their revealed counts set to zero.
7982 *** procedure: dup->fdes PORT/FD [NEWFD]
7983 Returns an integer file descriptor.
7985 *** procedure: dup->inport PORT/FD [NEWFD]
7986 Returns a new input port using the new file descriptor.
7988 *** procedure: dup->outport PORT/FD [NEWFD]
7989 Returns a new output port using the new file descriptor.
7991 *** procedure: dup PORT/FD [NEWFD]
7992 Returns a new port if PORT/FD is a port, with the same mode as the
7993 supplied port, otherwise returns an integer file descriptor.
7995 *** procedure: dup->port PORT/FD MODE [NEWFD]
7996 Returns a new port using the new file descriptor. MODE supplies a
7997 mode string for the port (*note open-file: File Ports.).
7999 *** procedure: setenv NAME VALUE
8000 Modifies the environment of the current process, which is also the
8001 default environment inherited by child processes.
8003 If VALUE is `#f', then NAME is removed from the environment.
8004 Otherwise, the string NAME=VALUE is added to the environment,
8005 replacing any existing string with name matching NAME.
8007 The return value is unspecified.
8009 *** procedure: truncate-file OBJ SIZE
8010 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
8011 can be a string containing a file name or an integer file
8012 descriptor or port open for output on the file. The underlying
8013 system calls are `truncate' and `ftruncate'.
8015 The return value is unspecified.
8017 *** procedure: setvbuf PORT MODE [SIZE]
8018 Set the buffering mode for PORT. MODE can be:
8026 block buffered, using a newly allocated buffer of SIZE bytes.
8027 However if SIZE is zero or unspecified, the port will be made
8030 This procedure should not be used after I/O has been performed with
8033 Ports are usually block buffered by default, with a default buffer
8034 size. Procedures e.g., *Note open-file: File Ports, which accept a
8035 mode string allow `0' to be added to request an unbuffered port.
8037 *** procedure: fsync PORT/FD
8038 Copies any unwritten data for the specified output file descriptor
8039 to disk. If PORT/FD is a port, its buffer is flushed before the
8040 underlying file descriptor is fsync'd. The return value is
8043 *** procedure: open-fdes PATH FLAGS [MODES]
8044 Similar to `open' but returns a file descriptor instead of a port.
8046 *** procedure: execle PATH ENV [ARG] ...
8047 Similar to `execl', but the environment of the new process is
8048 specified by ENV, which must be a list of strings as returned by
8049 the `environ' procedure.
8051 This procedure is currently implemented using the `execve' system
8052 call, but we call it `execle' because of its Scheme calling
8055 *** procedure: strerror ERRNO
8056 Returns the Unix error message corresponding to ERRNO, an integer.
8058 *** procedure: primitive-exit [STATUS]
8059 Terminate the current process without unwinding the Scheme stack.
8060 This is would typically be useful after a fork. The exit status
8061 is STATUS if supplied, otherwise zero.
8063 *** procedure: times
8064 Returns an object with information about real and processor time.
8065 The following procedures accept such an object as an argument and
8066 return a selected component:
8069 The current real time, expressed as time units relative to an
8073 The CPU time units used by the calling process.
8076 The CPU time units used by the system on behalf of the
8080 The CPU time units used by terminated child processes of the
8081 calling process, whose status has been collected (e.g., using
8085 Similarly, the CPU times units used by the system on behalf of
8086 terminated child processes.
8088 ** Removed: list-length
8089 ** Removed: list-append, list-append!
8090 ** Removed: list-reverse, list-reverse!
8092 ** array-map renamed to array-map!
8094 ** serial-array-map renamed to serial-array-map!
8096 ** catch doesn't take #f as first argument any longer
8098 Previously, it was possible to pass #f instead of a key to `catch'.
8099 That would cause `catch' to pass a jump buffer object to the procedure
8100 passed as second argument. The procedure could then use this jump
8101 buffer objekt as an argument to throw.
8103 This mechanism has been removed since its utility doesn't motivate the
8104 extra complexity it introduces.
8106 ** The `#/' notation for lists now provokes a warning message from Guile.
8107 This syntax will be removed from Guile in the near future.
8109 To disable the warning message, set the GUILE_HUSH environment
8110 variable to any non-empty value.
8112 ** The newline character now prints as `#\newline', following the
8113 normal Scheme notation, not `#\nl'.
8115 * Changes to the gh_ interface
8117 ** The gh_enter function now takes care of loading the Guile startup files.
8118 gh_enter works by calling scm_boot_guile; see the remarks below.
8120 ** Function: void gh_write (SCM x)
8122 Write the printed representation of the scheme object x to the current
8123 output port. Corresponds to the scheme level `write'.
8125 ** gh_list_length renamed to gh_length.
8127 ** vector handling routines
8129 Several major changes. In particular, gh_vector() now resembles
8130 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
8131 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
8132 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
8133 vector-related gh_ functions have been implemented.
8135 ** pair and list routines
8137 Implemented several of the R4RS pair and list functions that were
8140 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
8142 New function. Converts double arrays back and forth between Scheme
8145 * Changes to the scm_ interface
8147 ** The function scm_boot_guile now takes care of loading the startup files.
8149 Guile's primary initialization function, scm_boot_guile, now takes
8150 care of loading `boot-9.scm', in the `ice-9' module, to initialize
8151 Guile, define the module system, and put together some standard
8152 bindings. It also loads `init.scm', which is intended to hold
8153 site-specific initialization code.
8155 Since Guile cannot operate properly until boot-9.scm is loaded, there
8156 is no reason to separate loading boot-9.scm from Guile's other
8157 initialization processes.
8159 This job used to be done by scm_compile_shell_switches, which didn't
8160 make much sense; in particular, it meant that people using Guile for
8161 non-shell-like applications had to jump through hoops to get Guile
8162 initialized properly.
8164 ** The function scm_compile_shell_switches no longer loads the startup files.
8165 Now, Guile always loads the startup files, whenever it is initialized;
8166 see the notes above for scm_boot_guile and scm_load_startup_files.
8168 ** Function: scm_load_startup_files
8169 This new function takes care of loading Guile's initialization file
8170 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
8171 this is always called by the Guile initialization process, it's
8172 probably not too useful to call this yourself, but it's there anyway.
8174 ** The semantics of smob marking have changed slightly.
8176 The smob marking function (the `mark' member of the scm_smobfuns
8177 structure) is no longer responsible for setting the mark bit on the
8178 smob. The generic smob handling code in the garbage collector will
8179 set this bit. The mark function need only ensure that any other
8180 objects the smob refers to get marked.
8182 Note that this change means that the smob's GC8MARK bit is typically
8183 already set upon entry to the mark function. Thus, marking functions
8184 which look like this:
8187 if (SCM_GC8MARKP (ptr))
8189 SCM_SETGC8MARK (ptr);
8190 ... mark objects to which the smob refers ...
8193 are now incorrect, since they will return early, and fail to mark any
8194 other objects the smob refers to. Some code in the Guile library used
8197 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
8199 If you have implemented your own I/O port type, by writing the
8200 functions required by the scm_ptobfuns and then calling scm_newptob,
8201 you will need to change your functions slightly.
8203 The functions in a scm_ptobfuns structure now expect the port itself
8204 as their argument; they used to expect the `stream' member of the
8205 port's scm_port_table structure. This allows functions in an
8206 scm_ptobfuns structure to easily access the port's cell (and any flags
8207 it its CAR), and the port's scm_port_table structure.
8209 Guile now passes the I/O port itself as the `port' argument in the
8210 following scm_ptobfuns functions:
8212 int (*free) (SCM port);
8213 int (*fputc) (int, SCM port);
8214 int (*fputs) (char *, SCM port);
8215 scm_sizet (*fwrite) SCM_P ((char *ptr,
8219 int (*fflush) (SCM port);
8220 int (*fgetc) (SCM port);
8221 int (*fclose) (SCM port);
8223 The interfaces to the `mark', `print', `equalp', and `fgets' methods
8226 If you have existing code which defines its own port types, it is easy
8227 to convert your code to the new interface; simply apply SCM_STREAM to
8228 the port argument to yield the value you code used to expect.
8230 Note that since both the port and the stream have the same type in the
8231 C code --- they are both SCM values --- the C compiler will not remind
8232 you if you forget to update your scm_ptobfuns functions.
8235 ** Function: int scm_internal_select (int fds,
8239 struct timeval *timeout);
8241 This is a replacement for the `select' function provided by the OS.
8242 It enables I/O blocking and sleeping to happen for one cooperative
8243 thread without blocking other threads. It also avoids busy-loops in
8244 these situations. It is intended that all I/O blocking and sleeping
8245 will finally go through this function. Currently, this function is
8246 only available on systems providing `gettimeofday' and `select'.
8248 ** Function: SCM scm_internal_stack_catch (SCM tag,
8249 scm_catch_body_t body,
8251 scm_catch_handler_t handler,
8254 A new sibling to the other two C level `catch' functions
8255 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
8256 the stack to be saved automatically into the variable `the-last-stack'
8257 (scm_the_last_stack_var) on error. This is necessary if you want to
8258 use advanced error reporting, such as calling scm_display_error and
8259 scm_display_backtrace. (They both take a stack object as argument.)
8261 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
8263 scm_catch_handler_t handler,
8266 Spawns a new thread. It does a job similar to
8267 scm_call_with_new_thread but takes arguments more suitable when
8268 spawning threads from application C code.
8270 ** The hook scm_error_callback has been removed. It was originally
8271 intended as a way for the user to install his own error handler. But
8272 that method works badly since it intervenes between throw and catch,
8273 thereby changing the semantics of expressions like (catch #t ...).
8274 The correct way to do it is to use one of the C level catch functions
8275 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
8277 ** Removed functions:
8279 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
8280 scm_list_reverse, scm_list_reverse_x
8282 ** New macros: SCM_LISTn where n is one of the integers 0-9.
8284 These can be used for pretty list creation from C. The idea is taken
8285 from Erick Gallesio's STk.
8287 ** scm_array_map renamed to scm_array_map_x
8289 ** mbstrings are now removed
8291 This means that the type codes scm_tc7_mb_string and
8292 scm_tc7_mb_substring has been removed.
8294 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
8296 Since we no longer support multi-byte strings, these I/O functions
8297 have been simplified, and renamed. Here are their old names, and
8298 their new names and arguments:
8300 scm_gen_putc -> void scm_putc (int c, SCM port);
8301 scm_gen_puts -> void scm_puts (char *s, SCM port);
8302 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
8303 scm_gen_getc -> void scm_getc (SCM port);
8306 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
8308 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
8310 SCM_TYP7S now masks away the bit which distinguishes substrings from
8313 ** scm_catch_body_t: Backward incompatible change!
8315 Body functions to scm_internal_catch and friends do not any longer
8316 take a second argument. This is because it is no longer possible to
8317 pass a #f arg to catch.
8319 ** Calls to scm_protect_object and scm_unprotect now nest properly.
8321 The function scm_protect_object protects its argument from being freed
8322 by the garbage collector. scm_unprotect_object removes that
8325 These functions now nest properly. That is, for every object O, there
8326 is a counter which scm_protect_object(O) increments and
8327 scm_unprotect_object(O) decrements, if the counter is greater than
8328 zero. Every object's counter is zero when it is first created. If an
8329 object's counter is greater than zero, the garbage collector will not
8330 reclaim its storage.
8332 This allows you to use scm_protect_object in your code without
8333 worrying that some other function you call will call
8334 scm_unprotect_object, and allow it to be freed. Assuming that the
8335 functions you call are well-behaved, and unprotect only those objects
8336 they protect, you can follow the same rule and have confidence that
8337 objects will be freed only at appropriate times.
8340 Changes in Guile 1.2 (released Tuesday, June 24 1997):
8342 * Changes to the distribution
8344 ** Nightly snapshots are now available from ftp.red-bean.com.
8345 The old server, ftp.cyclic.com, has been relinquished to its rightful
8348 Nightly snapshots of the Guile development sources are now available via
8349 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
8351 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
8352 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
8354 ** To run Guile without installing it, the procedure has changed a bit.
8356 If you used a separate build directory to compile Guile, you'll need
8357 to include the build directory in SCHEME_LOAD_PATH, as well as the
8358 source directory. See the `INSTALL' file for examples.
8360 * Changes to the procedure for linking libguile with your programs
8362 ** The standard Guile load path for Scheme code now includes
8363 $(datadir)/guile (usually /usr/local/share/guile). This means that
8364 you can install your own Scheme files there, and Guile will find them.
8365 (Previous versions of Guile only checked a directory whose name
8366 contained the Guile version number, so you had to re-install or move
8367 your Scheme sources each time you installed a fresh version of Guile.)
8369 The load path also includes $(datadir)/guile/site; we recommend
8370 putting individual Scheme files there. If you want to install a
8371 package with multiple source files, create a directory for them under
8374 ** Guile 1.2 will now use the Rx regular expression library, if it is
8375 installed on your system. When you are linking libguile into your own
8376 programs, this means you will have to link against -lguile, -lqt (if
8377 you configured Guile with thread support), and -lrx.
8379 If you are using autoconf to generate configuration scripts for your
8380 application, the following lines should suffice to add the appropriate
8381 libraries to your link command:
8383 ### Find Rx, quickthreads and libguile.
8384 AC_CHECK_LIB(rx, main)
8385 AC_CHECK_LIB(qt, main)
8386 AC_CHECK_LIB(guile, scm_shell)
8388 The Guile 1.2 distribution does not contain sources for the Rx
8389 library, as Guile 1.0 did. If you want to use Rx, you'll need to
8390 retrieve it from a GNU FTP site and install it separately.
8392 * Changes to Scheme functions and syntax
8394 ** The dynamic linking features of Guile are now enabled by default.
8395 You can disable them by giving the `--disable-dynamic-linking' option
8398 (dynamic-link FILENAME)
8400 Find the object file denoted by FILENAME (a string) and link it
8401 into the running Guile application. When everything works out,
8402 return a Scheme object suitable for representing the linked object
8403 file. Otherwise an error is thrown. How object files are
8404 searched is system dependent.
8406 (dynamic-object? VAL)
8408 Determine whether VAL represents a dynamically linked object file.
8410 (dynamic-unlink DYNOBJ)
8412 Unlink the indicated object file from the application. DYNOBJ
8413 should be one of the values returned by `dynamic-link'.
8415 (dynamic-func FUNCTION DYNOBJ)
8417 Search the C function indicated by FUNCTION (a string or symbol)
8418 in DYNOBJ and return some Scheme object that can later be used
8419 with `dynamic-call' to actually call this function. Right now,
8420 these Scheme objects are formed by casting the address of the
8421 function to `long' and converting this number to its Scheme
8424 (dynamic-call FUNCTION DYNOBJ)
8426 Call the C function indicated by FUNCTION and DYNOBJ. The
8427 function is passed no arguments and its return value is ignored.
8428 When FUNCTION is something returned by `dynamic-func', call that
8429 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
8430 etc.), look it up in DYNOBJ; this is equivalent to
8432 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
8434 Interrupts are deferred while the C function is executing (with
8435 SCM_DEFER_INTS/SCM_ALLOW_INTS).
8437 (dynamic-args-call FUNCTION DYNOBJ ARGS)
8439 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
8440 some arguments and return its return value. The C function is
8441 expected to take two arguments and return an `int', just like
8444 int c_func (int argc, char **argv);
8446 ARGS must be a list of strings and is converted into an array of
8447 `char *'. The array is passed in ARGV and its size in ARGC. The
8448 return value is converted to a Scheme number and returned from the
8449 call to `dynamic-args-call'.
8451 When dynamic linking is disabled or not supported on your system,
8452 the above functions throw errors, but they are still available.
8454 Here is a small example that works on GNU/Linux:
8456 (define libc-obj (dynamic-link "libc.so"))
8457 (dynamic-args-call 'rand libc-obj '())
8459 See the file `libguile/DYNAMIC-LINKING' for additional comments.
8461 ** The #/ syntax for module names is depreciated, and will be removed
8462 in a future version of Guile. Instead of
8470 The latter syntax is more consistent with existing Lisp practice.
8472 ** Guile now does fancier printing of structures. Structures are the
8473 underlying implementation for records, which in turn are used to
8474 implement modules, so all of these object now print differently and in
8475 a more informative way.
8477 The Scheme printer will examine the builtin variable *struct-printer*
8478 whenever it needs to print a structure object. When this variable is
8479 not `#f' it is deemed to be a procedure and will be applied to the
8480 structure object and the output port. When *struct-printer* is `#f'
8481 or the procedure return `#f' the structure object will be printed in
8482 the boring #<struct 80458270> form.
8484 This hook is used by some routines in ice-9/boot-9.scm to implement
8485 type specific printing routines. Please read the comments there about
8488 One of the more specific uses of structs are records. The printing
8489 procedure that could be passed to MAKE-RECORD-TYPE is now actually
8490 called. It should behave like a *struct-printer* procedure (described
8493 ** Guile now supports a new R4RS-compliant syntax for keywords. A
8494 token of the form #:NAME, where NAME has the same syntax as a Scheme
8495 symbol, is the external representation of the keyword named NAME.
8496 Keyword objects print using this syntax as well, so values containing
8497 keyword objects can be read back into Guile. When used in an
8498 expression, keywords are self-quoting objects.
8500 Guile suports this read syntax, and uses this print syntax, regardless
8501 of the current setting of the `keyword' read option. The `keyword'
8502 read option only controls whether Guile recognizes the `:NAME' syntax,
8503 which is incompatible with R4RS. (R4RS says such token represent
8506 ** Guile has regular expression support again. Guile 1.0 included
8507 functions for matching regular expressions, based on the Rx library.
8508 In Guile 1.1, the Guile/Rx interface was removed to simplify the
8509 distribution, and thus Guile had no regular expression support. Guile
8510 1.2 again supports the most commonly used functions, and supports all
8511 of SCSH's regular expression functions.
8513 If your system does not include a POSIX regular expression library,
8514 and you have not linked Guile with a third-party regexp library such as
8515 Rx, these functions will not be available. You can tell whether your
8516 Guile installation includes regular expression support by checking
8517 whether the `*features*' list includes the `regex' symbol.
8519 *** regexp functions
8521 By default, Guile supports POSIX extended regular expressions. That
8522 means that the characters `(', `)', `+' and `?' are special, and must
8523 be escaped if you wish to match the literal characters.
8525 This regular expression interface was modeled after that implemented
8526 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
8527 with SCSH regular expressions.
8529 **** Function: string-match PATTERN STR [START]
8530 Compile the string PATTERN into a regular expression and compare
8531 it with STR. The optional numeric argument START specifies the
8532 position of STR at which to begin matching.
8534 `string-match' returns a "match structure" which describes what,
8535 if anything, was matched by the regular expression. *Note Match
8536 Structures::. If STR does not match PATTERN at all,
8537 `string-match' returns `#f'.
8539 Each time `string-match' is called, it must compile its PATTERN
8540 argument into a regular expression structure. This operation is
8541 expensive, which makes `string-match' inefficient if the same regular
8542 expression is used several times (for example, in a loop). For better
8543 performance, you can compile a regular expression in advance and then
8544 match strings against the compiled regexp.
8546 **** Function: make-regexp STR [FLAGS]
8547 Compile the regular expression described by STR, and return the
8548 compiled regexp structure. If STR does not describe a legal
8549 regular expression, `make-regexp' throws a
8550 `regular-expression-syntax' error.
8552 FLAGS may be the bitwise-or of one or more of the following:
8554 **** Constant: regexp/extended
8555 Use POSIX Extended Regular Expression syntax when interpreting
8556 STR. If not set, POSIX Basic Regular Expression syntax is used.
8557 If the FLAGS argument is omitted, we assume regexp/extended.
8559 **** Constant: regexp/icase
8560 Do not differentiate case. Subsequent searches using the
8561 returned regular expression will be case insensitive.
8563 **** Constant: regexp/newline
8564 Match-any-character operators don't match a newline.
8566 A non-matching list ([^...]) not containing a newline matches a
8569 Match-beginning-of-line operator (^) matches the empty string
8570 immediately after a newline, regardless of whether the FLAGS
8571 passed to regexp-exec contain regexp/notbol.
8573 Match-end-of-line operator ($) matches the empty string
8574 immediately before a newline, regardless of whether the FLAGS
8575 passed to regexp-exec contain regexp/noteol.
8577 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
8578 Match the compiled regular expression REGEXP against `str'. If
8579 the optional integer START argument is provided, begin matching
8580 from that position in the string. Return a match structure
8581 describing the results of the match, or `#f' if no match could be
8584 FLAGS may be the bitwise-or of one or more of the following:
8586 **** Constant: regexp/notbol
8587 The match-beginning-of-line operator always fails to match (but
8588 see the compilation flag regexp/newline above) This flag may be
8589 used when different portions of a string are passed to
8590 regexp-exec and the beginning of the string should not be
8591 interpreted as the beginning of the line.
8593 **** Constant: regexp/noteol
8594 The match-end-of-line operator always fails to match (but see the
8595 compilation flag regexp/newline above)
8597 **** Function: regexp? OBJ
8598 Return `#t' if OBJ is a compiled regular expression, or `#f'
8601 Regular expressions are commonly used to find patterns in one string
8602 and replace them with the contents of another string.
8604 **** Function: regexp-substitute PORT MATCH [ITEM...]
8605 Write to the output port PORT selected contents of the match
8606 structure MATCH. Each ITEM specifies what should be written, and
8607 may be one of the following arguments:
8609 * A string. String arguments are written out verbatim.
8611 * An integer. The submatch with that number is written.
8613 * The symbol `pre'. The portion of the matched string preceding
8614 the regexp match is written.
8616 * The symbol `post'. The portion of the matched string
8617 following the regexp match is written.
8619 PORT may be `#f', in which case nothing is written; instead,
8620 `regexp-substitute' constructs a string from the specified ITEMs
8623 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
8624 Similar to `regexp-substitute', but can be used to perform global
8625 substitutions on STR. Instead of taking a match structure as an
8626 argument, `regexp-substitute/global' takes two string arguments: a
8627 REGEXP string describing a regular expression, and a TARGET string
8628 which should be matched against this regular expression.
8630 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
8633 * A function may be supplied. When this function is called, it
8634 will be passed one argument: a match structure for a given
8635 regular expression match. It should return a string to be
8636 written out to PORT.
8638 * The `post' symbol causes `regexp-substitute/global' to recurse
8639 on the unmatched portion of STR. This *must* be supplied in
8640 order to perform global search-and-replace on STR; if it is
8641 not present among the ITEMs, then `regexp-substitute/global'
8642 will return after processing a single match.
8644 *** Match Structures
8646 A "match structure" is the object returned by `string-match' and
8647 `regexp-exec'. It describes which portion of a string, if any, matched
8648 the given regular expression. Match structures include: a reference to
8649 the string that was checked for matches; the starting and ending
8650 positions of the regexp match; and, if the regexp included any
8651 parenthesized subexpressions, the starting and ending positions of each
8654 In each of the regexp match functions described below, the `match'
8655 argument must be a match structure returned by a previous call to
8656 `string-match' or `regexp-exec'. Most of these functions return some
8657 information about the original target string that was matched against a
8658 regular expression; we will call that string TARGET for easy reference.
8660 **** Function: regexp-match? OBJ
8661 Return `#t' if OBJ is a match structure returned by a previous
8662 call to `regexp-exec', or `#f' otherwise.
8664 **** Function: match:substring MATCH [N]
8665 Return the portion of TARGET matched by subexpression number N.
8666 Submatch 0 (the default) represents the entire regexp match. If
8667 the regular expression as a whole matched, but the subexpression
8668 number N did not match, return `#f'.
8670 **** Function: match:start MATCH [N]
8671 Return the starting position of submatch number N.
8673 **** Function: match:end MATCH [N]
8674 Return the ending position of submatch number N.
8676 **** Function: match:prefix MATCH
8677 Return the unmatched portion of TARGET preceding the regexp match.
8679 **** Function: match:suffix MATCH
8680 Return the unmatched portion of TARGET following the regexp match.
8682 **** Function: match:count MATCH
8683 Return the number of parenthesized subexpressions from MATCH.
8684 Note that the entire regular expression match itself counts as a
8685 subexpression, and failed submatches are included in the count.
8687 **** Function: match:string MATCH
8688 Return the original TARGET string.
8690 *** Backslash Escapes
8692 Sometimes you will want a regexp to match characters like `*' or `$'
8693 exactly. For example, to check whether a particular string represents
8694 a menu entry from an Info node, it would be useful to match it against
8695 a regexp like `^* [^:]*::'. However, this won't work; because the
8696 asterisk is a metacharacter, it won't match the `*' at the beginning of
8697 the string. In this case, we want to make the first asterisk un-magic.
8699 You can do this by preceding the metacharacter with a backslash
8700 character `\'. (This is also called "quoting" the metacharacter, and
8701 is known as a "backslash escape".) When Guile sees a backslash in a
8702 regular expression, it considers the following glyph to be an ordinary
8703 character, no matter what special meaning it would ordinarily have.
8704 Therefore, we can make the above example work by changing the regexp to
8705 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
8706 to match only a single asterisk in the target string.
8708 Since the backslash is itself a metacharacter, you may force a
8709 regexp to match a backslash in the target string by preceding the
8710 backslash with itself. For example, to find variable references in a
8711 TeX program, you might want to find occurrences of the string `\let\'
8712 followed by any number of alphabetic characters. The regular expression
8713 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
8714 each match a single backslash in the target string.
8716 **** Function: regexp-quote STR
8717 Quote each special character found in STR with a backslash, and
8718 return the resulting string.
8720 *Very important:* Using backslash escapes in Guile source code (as
8721 in Emacs Lisp or C) can be tricky, because the backslash character has
8722 special meaning for the Guile reader. For example, if Guile encounters
8723 the character sequence `\n' in the middle of a string while processing
8724 Scheme code, it replaces those characters with a newline character.
8725 Similarly, the character sequence `\t' is replaced by a horizontal tab.
8726 Several of these "escape sequences" are processed by the Guile reader
8727 before your code is executed. Unrecognized escape sequences are
8728 ignored: if the characters `\*' appear in a string, they will be
8729 translated to the single character `*'.
8731 This translation is obviously undesirable for regular expressions,
8732 since we want to be able to include backslashes in a string in order to
8733 escape regexp metacharacters. Therefore, to make sure that a backslash
8734 is preserved in a string in your Guile program, you must use *two*
8735 consecutive backslashes:
8737 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
8739 The string in this example is preprocessed by the Guile reader before
8740 any code is executed. The resulting argument to `make-regexp' is the
8741 string `^\* [^:]*', which is what we really want.
8743 This also means that in order to write a regular expression that
8744 matches a single backslash character, the regular expression string in
8745 the source code must include *four* backslashes. Each consecutive pair
8746 of backslashes gets translated by the Guile reader to a single
8747 backslash, and the resulting double-backslash is interpreted by the
8748 regexp engine as matching a single backslash character. Hence:
8750 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
8752 The reason for the unwieldiness of this syntax is historical. Both
8753 regular expression pattern matchers and Unix string processing systems
8754 have traditionally used backslashes with the special meanings described
8755 above. The POSIX regular expression specification and ANSI C standard
8756 both require these semantics. Attempting to abandon either convention
8757 would cause other kinds of compatibility problems, possibly more severe
8758 ones. Therefore, without extending the Scheme reader to support
8759 strings with different quoting conventions (an ungainly and confusing
8760 extension when implemented in other languages), we must adhere to this
8761 cumbersome escape syntax.
8763 * Changes to the gh_ interface
8765 * Changes to the scm_ interface
8767 * Changes to system call interfaces:
8769 ** The value returned by `raise' is now unspecified. It throws an exception
8772 *** A new procedure `sigaction' can be used to install signal handlers
8774 (sigaction signum [action] [flags])
8776 signum is the signal number, which can be specified using the value
8779 If action is omitted, sigaction returns a pair: the CAR is the current
8780 signal hander, which will be either an integer with the value SIG_DFL
8781 (default action) or SIG_IGN (ignore), or the Scheme procedure which
8782 handles the signal, or #f if a non-Scheme procedure handles the
8783 signal. The CDR contains the current sigaction flags for the handler.
8785 If action is provided, it is installed as the new handler for signum.
8786 action can be a Scheme procedure taking one argument, or the value of
8787 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
8788 whatever signal handler was installed before sigaction was first used.
8789 Flags can optionally be specified for the new handler (SA_RESTART is
8790 always used if the system provides it, so need not be specified.) The
8791 return value is a pair with information about the old handler as
8794 This interface does not provide access to the "signal blocking"
8795 facility. Maybe this is not needed, since the thread support may
8796 provide solutions to the problem of consistent access to data
8799 *** A new procedure `flush-all-ports' is equivalent to running
8800 `force-output' on every port open for output.
8802 ** Guile now provides information on how it was built, via the new
8803 global variable, %guile-build-info. This variable records the values
8804 of the standard GNU makefile directory variables as an assocation
8805 list, mapping variable names (symbols) onto directory paths (strings).
8806 For example, to find out where the Guile link libraries were
8807 installed, you can say:
8809 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
8812 * Changes to the scm_ interface
8814 ** The new function scm_handle_by_message_noexit is just like the
8815 existing scm_handle_by_message function, except that it doesn't call
8816 exit to terminate the process. Instead, it prints a message and just
8817 returns #f. This might be a more appropriate catch-all handler for
8818 new dynamic roots and threads.
8821 Changes in Guile 1.1 (released Friday, May 16 1997):
8823 * Changes to the distribution.
8825 The Guile 1.0 distribution has been split up into several smaller
8827 guile-core --- the Guile interpreter itself.
8828 guile-tcltk --- the interface between the Guile interpreter and
8829 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
8830 is a toolkit for building graphical user interfaces.
8831 guile-rgx-ctax --- the interface between Guile and the Rx regular
8832 expression matcher, and the translator for the Ctax
8833 programming language. These are packaged together because the
8834 Ctax translator uses Rx to parse Ctax source code.
8836 This NEWS file describes the changes made to guile-core since the 1.0
8839 We no longer distribute the documentation, since it was either out of
8840 date, or incomplete. As soon as we have current documentation, we
8845 * Changes to the stand-alone interpreter
8847 ** guile now accepts command-line arguments compatible with SCSH, Olin
8848 Shivers' Scheme Shell.
8850 In general, arguments are evaluated from left to right, but there are
8851 exceptions. The following switches stop argument processing, and
8852 stash all remaining command-line arguments as the value returned by
8853 the (command-line) function.
8854 -s SCRIPT load Scheme source code from FILE, and exit
8855 -c EXPR evalute Scheme expression EXPR, and exit
8856 -- stop scanning arguments; run interactively
8858 The switches below are processed as they are encountered.
8859 -l FILE load Scheme source code from FILE
8860 -e FUNCTION after reading script, apply FUNCTION to
8861 command line arguments
8862 -ds do -s script at this point
8863 --emacs enable Emacs protocol (experimental)
8864 -h, --help display this help and exit
8865 -v, --version display version information and exit
8866 \ read arguments from following script lines
8868 So, for example, here is a Guile script named `ekko' (thanks, Olin)
8869 which re-implements the traditional "echo" command:
8871 #!/usr/local/bin/guile -s
8874 (map (lambda (arg) (display arg) (display " "))
8878 (main (command-line))
8880 Suppose we invoke this script as follows:
8882 ekko a speckled gecko
8884 Through the magic of Unix script processing (triggered by the `#!'
8885 token at the top of the file), /usr/local/bin/guile receives the
8886 following list of command-line arguments:
8888 ("-s" "./ekko" "a" "speckled" "gecko")
8890 Unix inserts the name of the script after the argument specified on
8891 the first line of the file (in this case, "-s"), and then follows that
8892 with the arguments given to the script. Guile loads the script, which
8893 defines the `main' function, and then applies it to the list of
8894 remaining command-line arguments, ("a" "speckled" "gecko").
8896 In Unix, the first line of a script file must take the following form:
8898 #!INTERPRETER ARGUMENT
8900 where INTERPRETER is the absolute filename of the interpreter
8901 executable, and ARGUMENT is a single command-line argument to pass to
8904 You may only pass one argument to the interpreter, and its length is
8905 limited. These restrictions can be annoying to work around, so Guile
8906 provides a general mechanism (borrowed from, and compatible with,
8907 SCSH) for circumventing them.
8909 If the ARGUMENT in a Guile script is a single backslash character,
8910 `\', Guile will open the script file, parse arguments from its second
8911 and subsequent lines, and replace the `\' with them. So, for example,
8912 here is another implementation of the `ekko' script:
8914 #!/usr/local/bin/guile \
8918 (for-each (lambda (arg) (display arg) (display " "))
8922 If the user invokes this script as follows:
8924 ekko a speckled gecko
8926 Unix expands this into
8928 /usr/local/bin/guile \ ekko a speckled gecko
8930 When Guile sees the `\' argument, it replaces it with the arguments
8931 read from the second line of the script, producing:
8933 /usr/local/bin/guile -e main -s ekko a speckled gecko
8935 This tells Guile to load the `ekko' script, and apply the function
8936 `main' to the argument list ("a" "speckled" "gecko").
8938 Here is how Guile parses the command-line arguments:
8939 - Each space character terminates an argument. This means that two
8940 spaces in a row introduce an empty-string argument.
8941 - The tab character is not permitted (unless you quote it with the
8942 backslash character, as described below), to avoid confusion.
8943 - The newline character terminates the sequence of arguments, and will
8944 also terminate a final non-empty argument. (However, a newline
8945 following a space will not introduce a final empty-string argument;
8946 it only terminates the argument list.)
8947 - The backslash character is the escape character. It escapes
8948 backslash, space, tab, and newline. The ANSI C escape sequences
8949 like \n and \t are also supported. These produce argument
8950 constituents; the two-character combination \n doesn't act like a
8951 terminating newline. The escape sequence \NNN for exactly three
8952 octal digits reads as the character whose ASCII code is NNN. As
8953 above, characters produced this way are argument constituents.
8954 Backslash followed by other characters is not allowed.
8956 * Changes to the procedure for linking libguile with your programs
8958 ** Guile now builds and installs a shared guile library, if your
8959 system support shared libraries. (It still builds a static library on
8960 all systems.) Guile automatically detects whether your system
8961 supports shared libraries. To prevent Guile from buildisg shared
8962 libraries, pass the `--disable-shared' flag to the configure script.
8964 Guile takes longer to compile when it builds shared libraries, because
8965 it must compile every file twice --- once to produce position-
8966 independent object code, and once to produce normal object code.
8968 ** The libthreads library has been merged into libguile.
8970 To link a program against Guile, you now need only link against
8971 -lguile and -lqt; -lthreads is no longer needed. If you are using
8972 autoconf to generate configuration scripts for your application, the
8973 following lines should suffice to add the appropriate libraries to
8976 ### Find quickthreads and libguile.
8977 AC_CHECK_LIB(qt, main)
8978 AC_CHECK_LIB(guile, scm_shell)
8980 * Changes to Scheme functions
8982 ** Guile Scheme's special syntax for keyword objects is now optional,
8983 and disabled by default.
8985 The syntax variation from R4RS made it difficult to port some
8986 interesting packages to Guile. The routines which accepted keyword
8987 arguments (mostly in the module system) have been modified to also
8988 accept symbols whose names begin with `:'.
8990 To change the keyword syntax, you must first import the (ice-9 debug)
8992 (use-modules (ice-9 debug))
8994 Then you can enable the keyword syntax as follows:
8995 (read-set! keywords 'prefix)
8997 To disable keyword syntax, do this:
8998 (read-set! keywords #f)
9000 ** Many more primitive functions accept shared substrings as
9001 arguments. In the past, these functions required normal, mutable
9002 strings as arguments, although they never made use of this
9005 ** The uniform array functions now operate on byte vectors. These
9006 functions are `array-fill!', `serial-array-copy!', `array-copy!',
9007 `serial-array-map', `array-map', `array-for-each', and
9010 ** The new functions `trace' and `untrace' implement simple debugging
9011 support for Scheme functions.
9013 The `trace' function accepts any number of procedures as arguments,
9014 and tells the Guile interpreter to display each procedure's name and
9015 arguments each time the procedure is invoked. When invoked with no
9016 arguments, `trace' returns the list of procedures currently being
9019 The `untrace' function accepts any number of procedures as arguments,
9020 and tells the Guile interpreter not to trace them any more. When
9021 invoked with no arguments, `untrace' untraces all curretly traced
9024 The tracing in Guile has an advantage over most other systems: we
9025 don't create new procedure objects, but mark the procedure objects
9026 themselves. This means that anonymous and internal procedures can be
9029 ** The function `assert-repl-prompt' has been renamed to
9030 `set-repl-prompt!'. It takes one argument, PROMPT.
9031 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
9032 - If PROMPT is a string, we use it as a prompt.
9033 - If PROMPT is a procedure accepting no arguments, we call it, and
9034 display the result as a prompt.
9035 - Otherwise, we display "> ".
9037 ** The new function `eval-string' reads Scheme expressions from a
9038 string and evaluates them, returning the value of the last expression
9039 in the string. If the string contains no expressions, it returns an
9042 ** The new function `thunk?' returns true iff its argument is a
9043 procedure of zero arguments.
9045 ** `defined?' is now a builtin function, instead of syntax. This
9046 means that its argument should be quoted. It returns #t iff its
9047 argument is bound in the current module.
9049 ** The new syntax `use-modules' allows you to add new modules to your
9050 environment without re-typing a complete `define-module' form. It
9051 accepts any number of module names as arguments, and imports their
9052 public bindings into the current module.
9054 ** The new function (module-defined? NAME MODULE) returns true iff
9055 NAME, a symbol, is defined in MODULE, a module object.
9057 ** The new function `builtin-bindings' creates and returns a hash
9058 table containing copies of all the root module's bindings.
9060 ** The new function `builtin-weak-bindings' does the same as
9061 `builtin-bindings', but creates a doubly-weak hash table.
9063 ** The `equal?' function now considers variable objects to be
9064 equivalent if they have the same name and the same value.
9066 ** The new function `command-line' returns the command-line arguments
9067 given to Guile, as a list of strings.
9069 When using guile as a script interpreter, `command-line' returns the
9070 script's arguments; those processed by the interpreter (like `-s' or
9071 `-c') are omitted. (In other words, you get the normal, expected
9072 behavior.) Any application that uses scm_shell to process its
9073 command-line arguments gets this behavior as well.
9075 ** The new function `load-user-init' looks for a file called `.guile'
9076 in the user's home directory, and loads it if it exists. This is
9077 mostly for use by the code generated by scm_compile_shell_switches,
9078 but we thought it might also be useful in other circumstances.
9080 ** The new function `log10' returns the base-10 logarithm of its
9083 ** Changes to I/O functions
9085 *** The functions `read', `primitive-load', `read-and-eval!', and
9086 `primitive-load-path' no longer take optional arguments controlling
9087 case insensitivity and a `#' parser.
9089 Case sensitivity is now controlled by a read option called
9090 `case-insensitive'. The user can add new `#' syntaxes with the
9091 `read-hash-extend' function (see below).
9093 *** The new function `read-hash-extend' allows the user to change the
9094 syntax of Guile Scheme in a somewhat controlled way.
9096 (read-hash-extend CHAR PROC)
9097 When parsing S-expressions, if we read a `#' character followed by
9098 the character CHAR, use PROC to parse an object from the stream.
9099 If PROC is #f, remove any parsing procedure registered for CHAR.
9101 The reader applies PROC to two arguments: CHAR and an input port.
9103 *** The new functions read-delimited and read-delimited! provide a
9104 general mechanism for doing delimited input on streams.
9106 (read-delimited DELIMS [PORT HANDLE-DELIM])
9107 Read until we encounter one of the characters in DELIMS (a string),
9108 or end-of-file. PORT is the input port to read from; it defaults to
9109 the current input port. The HANDLE-DELIM parameter determines how
9110 the terminating character is handled; it should be one of the
9113 'trim omit delimiter from result
9114 'peek leave delimiter character in input stream
9115 'concat append delimiter character to returned value
9116 'split return a pair: (RESULT . TERMINATOR)
9118 HANDLE-DELIM defaults to 'peek.
9120 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
9121 A side-effecting variant of `read-delimited'.
9123 The data is written into the string BUF at the indices in the
9124 half-open interval [START, END); the default interval is the whole
9125 string: START = 0 and END = (string-length BUF). The values of
9126 START and END must specify a well-defined interval in BUF, i.e.
9127 0 <= START <= END <= (string-length BUF).
9129 It returns NBYTES, the number of bytes read. If the buffer filled
9130 up without a delimiter character being found, it returns #f. If the
9131 port is at EOF when the read starts, it returns the EOF object.
9133 If an integer is returned (i.e., the read is successfully terminated
9134 by reading a delimiter character), then the HANDLE-DELIM parameter
9135 determines how to handle the terminating character. It is described
9136 above, and defaults to 'peek.
9138 (The descriptions of these functions were borrowed from the SCSH
9139 manual, by Olin Shivers and Brian Carlstrom.)
9141 *** The `%read-delimited!' function is the primitive used to implement
9142 `read-delimited' and `read-delimited!'.
9144 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
9146 This returns a pair of values: (TERMINATOR . NUM-READ).
9147 - TERMINATOR describes why the read was terminated. If it is a
9148 character or the eof object, then that is the value that terminated
9149 the read. If it is #f, the function filled the buffer without finding
9150 a delimiting character.
9151 - NUM-READ is the number of characters read into BUF.
9153 If the read is successfully terminated by reading a delimiter
9154 character, then the gobble? parameter determines what to do with the
9155 terminating character. If true, the character is removed from the
9156 input stream; if false, the character is left in the input stream
9157 where a subsequent read operation will retrieve it. In either case,
9158 the character is also the first value returned by the procedure call.
9160 (The descriptions of this function was borrowed from the SCSH manual,
9161 by Olin Shivers and Brian Carlstrom.)
9163 *** The `read-line' and `read-line!' functions have changed; they now
9164 trim the terminator by default; previously they appended it to the
9165 returned string. For the old behavior, use (read-line PORT 'concat).
9167 *** The functions `uniform-array-read!' and `uniform-array-write!' now
9168 take new optional START and END arguments, specifying the region of
9169 the array to read and write.
9171 *** The `ungetc-char-ready?' function has been removed. We feel it's
9172 inappropriate for an interface to expose implementation details this
9175 ** Changes to the Unix library and system call interface
9177 *** The new fcntl function provides access to the Unix `fcntl' system
9180 (fcntl PORT COMMAND VALUE)
9181 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
9182 Values for COMMAND are:
9184 F_DUPFD duplicate a file descriptor
9185 F_GETFD read the descriptor's close-on-exec flag
9186 F_SETFD set the descriptor's close-on-exec flag to VALUE
9187 F_GETFL read the descriptor's flags, as set on open
9188 F_SETFL set the descriptor's flags, as set on open to VALUE
9189 F_GETOWN return the process ID of a socket's owner, for SIGIO
9190 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
9191 FD_CLOEXEC not sure what this is
9193 For details, see the documentation for the fcntl system call.
9195 *** The arguments to `select' have changed, for compatibility with
9196 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
9197 expected behavior. The MILLISECONDS parameter has been changed to
9198 MICROSECONDS, to more closely resemble the underlying system call.
9199 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
9200 corresponding return set will be the same.
9202 *** The arguments to the `mknod' system call have changed. They are
9205 (mknod PATH TYPE PERMS DEV)
9206 Create a new file (`node') in the file system. PATH is the name of
9207 the file to create. TYPE is the kind of file to create; it should
9208 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
9209 permission bits to give the newly created file. If TYPE is
9210 'block-special or 'char-special, DEV specifies which device the
9211 special file refers to; its interpretation depends on the kind of
9212 special file being created.
9214 *** The `fork' function has been renamed to `primitive-fork', to avoid
9215 clashing with various SCSH forks.
9217 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
9218 and `recvfrom!'. They no longer accept a size for a second argument;
9219 you must pass a string to hold the received value. They no longer
9220 return the buffer. Instead, `recv' returns the length of the message
9221 received, and `recvfrom' returns a pair containing the packet's length
9222 and originating address.
9224 *** The file descriptor datatype has been removed, as have the
9225 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
9226 We plan to replace these functions with a SCSH-compatible interface.
9228 *** The `create' function has been removed; it's just a special case
9231 *** There are new functions to break down process termination status
9232 values. In the descriptions below, STATUS is a value returned by
9235 (status:exit-val STATUS)
9236 If the child process exited normally, this function returns the exit
9237 code for the child process (i.e., the value passed to exit, or
9238 returned from main). If the child process did not exit normally,
9239 this function returns #f.
9241 (status:stop-sig STATUS)
9242 If the child process was suspended by a signal, this function
9243 returns the signal that suspended the child. Otherwise, it returns
9246 (status:term-sig STATUS)
9247 If the child process terminated abnormally, this function returns
9248 the signal that terminated the child. Otherwise, this function
9251 POSIX promises that exactly one of these functions will return true on
9252 a valid STATUS value.
9254 These functions are compatible with SCSH.
9256 *** There are new accessors and setters for the broken-out time vectors
9257 returned by `localtime', `gmtime', and that ilk. They are:
9259 Component Accessor Setter
9260 ========================= ============ ============
9261 seconds tm:sec set-tm:sec
9262 minutes tm:min set-tm:min
9263 hours tm:hour set-tm:hour
9264 day of the month tm:mday set-tm:mday
9265 month tm:mon set-tm:mon
9266 year tm:year set-tm:year
9267 day of the week tm:wday set-tm:wday
9268 day in the year tm:yday set-tm:yday
9269 daylight saving time tm:isdst set-tm:isdst
9270 GMT offset, seconds tm:gmtoff set-tm:gmtoff
9271 name of time zone tm:zone set-tm:zone
9273 *** There are new accessors for the vectors returned by `uname',
9274 describing the host system:
9277 ============================================== ================
9278 name of the operating system implementation utsname:sysname
9279 network name of this machine utsname:nodename
9280 release level of the operating system utsname:release
9281 version level of the operating system utsname:version
9282 machine hardware platform utsname:machine
9284 *** There are new accessors for the vectors returned by `getpw',
9285 `getpwnam', `getpwuid', and `getpwent', describing entries from the
9286 system's user database:
9289 ====================== =================
9290 user name passwd:name
9291 user password passwd:passwd
9294 real name passwd:gecos
9295 home directory passwd:dir
9296 shell program passwd:shell
9298 *** There are new accessors for the vectors returned by `getgr',
9299 `getgrnam', `getgrgid', and `getgrent', describing entries from the
9300 system's group database:
9303 ======================= ============
9304 group name group:name
9305 group password group:passwd
9307 group members group:mem
9309 *** There are new accessors for the vectors returned by `gethost',
9310 `gethostbyaddr', `gethostbyname', and `gethostent', describing
9314 ========================= ===============
9315 official name of host hostent:name
9316 alias list hostent:aliases
9317 host address type hostent:addrtype
9318 length of address hostent:length
9319 list of addresses hostent:addr-list
9321 *** There are new accessors for the vectors returned by `getnet',
9322 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
9326 ========================= ===============
9327 official name of net netent:name
9328 alias list netent:aliases
9329 net number type netent:addrtype
9330 net number netent:net
9332 *** There are new accessors for the vectors returned by `getproto',
9333 `getprotobyname', `getprotobynumber', and `getprotoent', describing
9337 ========================= ===============
9338 official protocol name protoent:name
9339 alias list protoent:aliases
9340 protocol number protoent:proto
9342 *** There are new accessors for the vectors returned by `getserv',
9343 `getservbyname', `getservbyport', and `getservent', describing
9347 ========================= ===============
9348 official service name servent:name
9349 alias list servent:aliases
9350 port number servent:port
9351 protocol to use servent:proto
9353 *** There are new accessors for the sockaddr structures returned by
9354 `accept', `getsockname', `getpeername', `recvfrom!':
9357 ======================================== ===============
9358 address format (`family') sockaddr:fam
9359 path, for file domain addresses sockaddr:path
9360 address, for internet domain addresses sockaddr:addr
9361 TCP or UDP port, for internet sockaddr:port
9363 *** The `getpwent', `getgrent', `gethostent', `getnetent',
9364 `getprotoent', and `getservent' functions now return #f at the end of
9365 the user database. (They used to throw an exception.)
9367 Note that calling MUMBLEent function is equivalent to calling the
9368 corresponding MUMBLE function with no arguments.
9370 *** The `setpwent', `setgrent', `sethostent', `setnetent',
9371 `setprotoent', and `setservent' routines now take no arguments.
9373 *** The `gethost', `getproto', `getnet', and `getserv' functions now
9374 provide more useful information when they throw an exception.
9376 *** The `lnaof' function has been renamed to `inet-lnaof'.
9378 *** Guile now claims to have the `current-time' feature.
9380 *** The `mktime' function now takes an optional second argument ZONE,
9381 giving the time zone to use for the conversion. ZONE should be a
9382 string, in the same format as expected for the "TZ" environment variable.
9384 *** The `strptime' function now returns a pair (TIME . COUNT), where
9385 TIME is the parsed time as a vector, and COUNT is the number of
9386 characters from the string left unparsed. This function used to
9387 return the remaining characters as a string.
9389 *** The `gettimeofday' function has replaced the old `time+ticks' function.
9390 The return value is now (SECONDS . MICROSECONDS); the fractional
9391 component is no longer expressed in "ticks".
9393 *** The `ticks/sec' constant has been removed, in light of the above change.
9395 * Changes to the gh_ interface
9397 ** gh_eval_str() now returns an SCM object which is the result of the
9400 ** gh_scm2str() now copies the Scheme data to a caller-provided C
9403 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
9404 and returns the array
9406 ** gh_scm2str0() is gone: there is no need to distinguish
9407 null-terminated from non-null-terminated, since gh_scm2newstr() allows
9408 the user to interpret the data both ways.
9410 * Changes to the scm_ interface
9412 ** The new function scm_symbol_value0 provides an easy way to get a
9413 symbol's value from C code:
9415 SCM scm_symbol_value0 (char *NAME)
9416 Return the value of the symbol named by the null-terminated string
9417 NAME in the current module. If the symbol named NAME is unbound in
9418 the current module, return SCM_UNDEFINED.
9420 ** The new function scm_sysintern0 creates new top-level variables,
9421 without assigning them a value.
9423 SCM scm_sysintern0 (char *NAME)
9424 Create a new Scheme top-level variable named NAME. NAME is a
9425 null-terminated string. Return the variable's value cell.
9427 ** The function scm_internal_catch is the guts of catch. It handles
9428 all the mechanics of setting up a catch target, invoking the catch
9429 body, and perhaps invoking the handler if the body does a throw.
9431 The function is designed to be usable from C code, but is general
9432 enough to implement all the semantics Guile Scheme expects from throw.
9434 TAG is the catch tag. Typically, this is a symbol, but this function
9435 doesn't actually care about that.
9437 BODY is a pointer to a C function which runs the body of the catch;
9438 this is the code you can throw from. We call it like this:
9439 BODY (BODY_DATA, JMPBUF)
9441 BODY_DATA is just the BODY_DATA argument we received; we pass it
9442 through to BODY as its first argument. The caller can make
9443 BODY_DATA point to anything useful that BODY might need.
9444 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
9445 which we have just created and initialized.
9447 HANDLER is a pointer to a C function to deal with a throw to TAG,
9448 should one occur. We call it like this:
9449 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
9451 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
9452 same idea as BODY_DATA above.
9453 THROWN_TAG is the tag that the user threw to; usually this is
9454 TAG, but it could be something else if TAG was #t (i.e., a
9455 catch-all), or the user threw to a jmpbuf.
9456 THROW_ARGS is the list of arguments the user passed to the THROW
9459 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
9460 is just a pointer we pass through to HANDLER. We don't actually
9461 use either of those pointers otherwise ourselves. The idea is
9462 that, if our caller wants to communicate something to BODY or
9463 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
9464 HANDLER can then use. Think of it as a way to make BODY and
9465 HANDLER closures, not just functions; MUMBLE_DATA points to the
9468 Of course, it's up to the caller to make sure that any data a
9469 MUMBLE_DATA needs is protected from GC. A common way to do this is
9470 to make MUMBLE_DATA a pointer to data stored in an automatic
9471 structure variable; since the collector must scan the stack for
9472 references anyway, this assures that any references in MUMBLE_DATA
9475 ** The new function scm_internal_lazy_catch is exactly like
9476 scm_internal_catch, except:
9478 - It does not unwind the stack (this is the major difference).
9479 - If handler returns, its value is returned from the throw.
9480 - BODY always receives #f as its JMPBUF argument (since there's no
9481 jmpbuf associated with a lazy catch, because we don't unwind the
9484 ** scm_body_thunk is a new body function you can pass to
9485 scm_internal_catch if you want the body to be like Scheme's `catch'
9486 --- a thunk, or a function of one argument if the tag is #f.
9488 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
9489 contains the Scheme procedure to invoke as the body, and the tag
9490 we're catching. If the tag is #f, then we pass JMPBUF (created by
9491 scm_internal_catch) to the body procedure; otherwise, the body gets
9494 ** scm_handle_by_proc is a new handler function you can pass to
9495 scm_internal_catch if you want the handler to act like Scheme's catch
9496 --- call a procedure with the tag and the throw arguments.
9498 If the user does a throw to this catch, this function runs a handler
9499 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
9500 variable holding the Scheme procedure object to invoke. It ought to
9501 be a pointer to an automatic variable (i.e., one living on the stack),
9502 or the procedure object should be otherwise protected from GC.
9504 ** scm_handle_by_message is a new handler function to use with
9505 `scm_internal_catch' if you want Guile to print a message and die.
9506 It's useful for dealing with throws to uncaught keys at the top level.
9508 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
9509 message header to print; if zero, we use "guile" instead. That
9510 text is followed by a colon, then the message described by ARGS.
9512 ** The return type of scm_boot_guile is now void; the function does
9513 not return a value, and indeed, never returns at all.
9515 ** The new function scm_shell makes it easy for user applications to
9516 process command-line arguments in a way that is compatible with the
9517 stand-alone guile interpreter (which is in turn compatible with SCSH,
9520 To use the scm_shell function, first initialize any guile modules
9521 linked into your application, and then call scm_shell with the values
9522 of ARGC and ARGV your `main' function received. scm_shell will add
9523 any SCSH-style meta-arguments from the top of the script file to the
9524 argument vector, and then process the command-line arguments. This
9525 generally means loading a script file or starting up an interactive
9526 command interpreter. For details, see "Changes to the stand-alone
9529 ** The new functions scm_get_meta_args and scm_count_argv help you
9530 implement the SCSH-style meta-argument, `\'.
9532 char **scm_get_meta_args (int ARGC, char **ARGV)
9533 If the second element of ARGV is a string consisting of a single
9534 backslash character (i.e. "\\" in Scheme notation), open the file
9535 named by the following argument, parse arguments from it, and return
9536 the spliced command line. The returned array is terminated by a
9539 For details of argument parsing, see above, under "guile now accepts
9540 command-line arguments compatible with SCSH..."
9542 int scm_count_argv (char **ARGV)
9543 Count the arguments in ARGV, assuming it is terminated by a null
9546 For an example of how these functions might be used, see the source
9547 code for the function scm_shell in libguile/script.c.
9549 You will usually want to use scm_shell instead of calling this
9552 ** The new function scm_compile_shell_switches turns an array of
9553 command-line arguments into Scheme code to carry out the actions they
9554 describe. Given ARGC and ARGV, it returns a Scheme expression to
9555 evaluate, and calls scm_set_program_arguments to make any remaining
9556 command-line arguments available to the Scheme code. For example,
9557 given the following arguments:
9559 -e main -s ekko a speckled gecko
9561 scm_set_program_arguments will return the following expression:
9563 (begin (load "ekko") (main (command-line)) (quit))
9565 You will usually want to use scm_shell instead of calling this
9568 ** The function scm_shell_usage prints a usage message appropriate for
9569 an interpreter that uses scm_compile_shell_switches to handle its
9570 command-line arguments.
9572 void scm_shell_usage (int FATAL, char *MESSAGE)
9573 Print a usage message to the standard error output. If MESSAGE is
9574 non-zero, write it before the usage message, followed by a newline.
9575 If FATAL is non-zero, exit the process, using FATAL as the
9576 termination status. (If you want to be compatible with Guile,
9577 always use 1 as the exit status when terminating due to command-line
9580 You will usually want to use scm_shell instead of calling this
9583 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
9584 expressions. It used to return SCM_EOL. Earth-shattering.
9586 ** The macros for declaring scheme objects in C code have been
9587 rearranged slightly. They are now:
9589 SCM_SYMBOL (C_NAME, SCHEME_NAME)
9590 Declare a static SCM variable named C_NAME, and initialize it to
9591 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
9592 be a C identifier, and SCHEME_NAME should be a C string.
9594 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
9595 Just like SCM_SYMBOL, but make C_NAME globally visible.
9597 SCM_VCELL (C_NAME, SCHEME_NAME)
9598 Create a global variable at the Scheme level named SCHEME_NAME.
9599 Declare a static SCM variable named C_NAME, and initialize it to
9600 point to the Scheme variable's value cell.
9602 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
9603 Just like SCM_VCELL, but make C_NAME globally visible.
9605 The `guile-snarf' script writes initialization code for these macros
9606 to its standard output, given C source code as input.
9608 The SCM_GLOBAL macro is gone.
9610 ** The scm_read_line and scm_read_line_x functions have been replaced
9611 by Scheme code based on the %read-delimited! procedure (known to C
9612 code as scm_read_delimited_x). See its description above for more
9615 ** The function scm_sys_open has been renamed to scm_open. It now
9616 returns a port instead of an FD object.
9618 * The dynamic linking support has changed. For more information, see
9619 libguile/DYNAMIC-LINKING.
9624 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
9627 * Changes to the 'guile' program:
9629 ** Guile now loads some new files when it starts up. Guile first
9630 searches the load path for init.scm, and loads it if found. Then, if
9631 Guile is not being used to execute a script, and the user's home
9632 directory contains a file named `.guile', Guile loads that.
9634 ** You can now use Guile as a shell script interpreter.
9636 To paraphrase the SCSH manual:
9638 When Unix tries to execute an executable file whose first two
9639 characters are the `#!', it treats the file not as machine code to
9640 be directly executed by the native processor, but as source code
9641 to be executed by some interpreter. The interpreter to use is
9642 specified immediately after the #! sequence on the first line of
9643 the source file. The kernel reads in the name of the interpreter,
9644 and executes that instead. It passes the interpreter the source
9645 filename as its first argument, with the original arguments
9646 following. Consult the Unix man page for the `exec' system call
9647 for more information.
9649 Now you can use Guile as an interpreter, using a mechanism which is a
9650 compatible subset of that provided by SCSH.
9652 Guile now recognizes a '-s' command line switch, whose argument is the
9653 name of a file of Scheme code to load. It also treats the two
9654 characters `#!' as the start of a comment, terminated by `!#'. Thus,
9655 to make a file of Scheme code directly executable by Unix, insert the
9656 following two lines at the top of the file:
9658 #!/usr/local/bin/guile -s
9661 Guile treats the argument of the `-s' command-line switch as the name
9662 of a file of Scheme code to load, and treats the sequence `#!' as the
9663 start of a block comment, terminated by `!#'.
9665 For example, here's a version of 'echo' written in Scheme:
9667 #!/usr/local/bin/guile -s
9669 (let loop ((args (cdr (program-arguments))))
9672 (display (car args))
9673 (if (pair? (cdr args))
9675 (loop (cdr args)))))
9678 Why does `#!' start a block comment terminated by `!#', instead of the
9679 end of the line? That is the notation SCSH uses, and although we
9680 don't yet support the other SCSH features that motivate that choice,
9681 we would like to be backward-compatible with any existing Guile
9682 scripts once we do. Furthermore, if the path to Guile on your system
9683 is too long for your kernel, you can start the script with this
9687 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
9690 Note that some very old Unix systems don't support the `#!' syntax.
9693 ** You can now run Guile without installing it.
9695 Previous versions of the interactive Guile interpreter (`guile')
9696 couldn't start up unless Guile's Scheme library had been installed;
9697 they used the value of the environment variable `SCHEME_LOAD_PATH'
9698 later on in the startup process, but not to find the startup code
9699 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
9702 To run Guile without installing it, build it in the normal way, and
9703 then set the environment variable `SCHEME_LOAD_PATH' to a
9704 colon-separated list of directories, including the top-level directory
9705 of the Guile sources. For example, if you unpacked Guile so that the
9706 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
9709 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
9712 ** Guile's read-eval-print loop no longer prints #<unspecified>
9713 results. If the user wants to see this, she can evaluate the
9714 expression (assert-repl-print-unspecified #t), perhaps in her startup
9717 ** Guile no longer shows backtraces by default when an error occurs;
9718 however, it does display a message saying how to get one, and how to
9719 request that they be displayed by default. After an error, evaluate
9721 to see a backtrace, and
9722 (debug-enable 'backtrace)
9723 to see them by default.
9727 * Changes to Guile Scheme:
9729 ** Guile now distinguishes between #f and the empty list.
9731 This is for compatibility with the IEEE standard, the (possibly)
9732 upcoming Revised^5 Report on Scheme, and many extant Scheme
9735 Guile used to have #f and '() denote the same object, to make Scheme's
9736 type system more compatible with Emacs Lisp's. However, the change
9737 caused too much trouble for Scheme programmers, and we found another
9738 way to reconcile Emacs Lisp with Scheme that didn't require this.
9741 ** Guile's delq, delv, delete functions, and their destructive
9742 counterparts, delq!, delv!, and delete!, now remove all matching
9743 elements from the list, not just the first. This matches the behavior
9744 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
9745 functions which inspired them.
9747 I recognize that this change may break code in subtle ways, but it
9748 seems best to make the change before the FSF's first Guile release,
9752 ** The compiled-library-path function has been deleted from libguile.
9754 ** The facilities for loading Scheme source files have changed.
9756 *** The variable %load-path now tells Guile which directories to search
9757 for Scheme code. Its value is a list of strings, each of which names
9760 *** The variable %load-extensions now tells Guile which extensions to
9761 try appending to a filename when searching the load path. Its value
9762 is a list of strings. Its default value is ("" ".scm").
9764 *** (%search-load-path FILENAME) searches the directories listed in the
9765 value of the %load-path variable for a Scheme file named FILENAME,
9766 with all the extensions listed in %load-extensions. If it finds a
9767 match, then it returns its full filename. If FILENAME is absolute, it
9768 returns it unchanged. Otherwise, it returns #f.
9770 %search-load-path will not return matches that refer to directories.
9772 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
9773 uses %seach-load-path to find a file named FILENAME, and loads it if
9774 it finds it. If it can't read FILENAME for any reason, it throws an
9777 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
9780 *** load uses the same searching semantics as primitive-load.
9782 *** The functions %try-load, try-load-with-path, %load, load-with-path,
9783 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
9784 path, and load-module-with-path have been deleted. The functions
9785 above should serve their purposes.
9787 *** If the value of the variable %load-hook is a procedure,
9788 `primitive-load' applies its value to the name of the file being
9789 loaded (without the load path directory name prepended). If its value
9790 is #f, it is ignored. Otherwise, an error occurs.
9792 This is mostly useful for printing load notification messages.
9795 ** The function `eval!' is no longer accessible from the scheme level.
9796 We can't allow operations which introduce glocs into the scheme level,
9797 because Guile's type system can't handle these as data. Use `eval' or
9798 `read-and-eval!' (see below) as replacement.
9800 ** The new function read-and-eval! reads an expression from PORT,
9801 evaluates it, and returns the result. This is more efficient than
9802 simply calling `read' and `eval', since it is not necessary to make a
9803 copy of the expression for the evaluator to munge.
9805 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
9806 for the `read' function.
9809 ** The function `int?' has been removed; its definition was identical
9810 to that of `integer?'.
9812 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
9813 use the R4RS names for these functions.
9815 ** The function object-properties no longer returns the hash handle;
9816 it simply returns the object's property list.
9818 ** Many functions have been changed to throw errors, instead of
9819 returning #f on failure. The point of providing exception handling in
9820 the language is to simplify the logic of user code, but this is less
9821 useful if Guile's primitives don't throw exceptions.
9823 ** The function `fileno' has been renamed from `%fileno'.
9825 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
9828 * Changes to Guile's C interface:
9830 ** The library's initialization procedure has been simplified.
9831 scm_boot_guile now has the prototype:
9833 void scm_boot_guile (int ARGC,
9835 void (*main_func) (),
9838 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
9839 MAIN_FUNC should do all the work of the program (initializing other
9840 packages, reading user input, etc.) before returning. When MAIN_FUNC
9841 returns, call exit (0); this function never returns. If you want some
9842 other exit value, MAIN_FUNC may call exit itself.
9844 scm_boot_guile arranges for program-arguments to return the strings
9845 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
9846 scm_set_program_arguments with the final list, so Scheme code will
9847 know which arguments have been processed.
9849 scm_boot_guile establishes a catch-all catch handler which prints an
9850 error message and exits the process. This means that Guile exits in a
9851 coherent way when system errors occur and the user isn't prepared to
9852 handle it. If the user doesn't like this behavior, they can establish
9853 their own universal catcher in MAIN_FUNC to shadow this one.
9855 Why must the caller do all the real work from MAIN_FUNC? The garbage
9856 collector assumes that all local variables of type SCM will be above
9857 scm_boot_guile's stack frame on the stack. If you try to manipulate
9858 SCM values after this function returns, it's the luck of the draw
9859 whether the GC will be able to find the objects you allocate. So,
9860 scm_boot_guile function exits, rather than returning, to discourage
9861 people from making that mistake.
9863 The IN, OUT, and ERR arguments were removed; there are other
9864 convenient ways to override these when desired.
9866 The RESULT argument was deleted; this function should never return.
9868 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
9872 ** Guile's header files should no longer conflict with your system's
9875 In order to compile code which #included <libguile.h>, previous
9876 versions of Guile required you to add a directory containing all the
9877 Guile header files to your #include path. This was a problem, since
9878 Guile's header files have names which conflict with many systems'
9881 Now only <libguile.h> need appear in your #include path; you must
9882 refer to all Guile's other header files as <libguile/mumble.h>.
9883 Guile's installation procedure puts libguile.h in $(includedir), and
9884 the rest in $(includedir)/libguile.
9887 ** Two new C functions, scm_protect_object and scm_unprotect_object,
9888 have been added to the Guile library.
9890 scm_protect_object (OBJ) protects OBJ from the garbage collector.
9891 OBJ will not be freed, even if all other references are dropped,
9892 until someone does scm_unprotect_object (OBJ). Both functions
9895 Note that calls to scm_protect_object do not nest. You can call
9896 scm_protect_object any number of times on a given object, and the
9897 next call to scm_unprotect_object will unprotect it completely.
9899 Basically, scm_protect_object and scm_unprotect_object just
9900 maintain a list of references to things. Since the GC knows about
9901 this list, all objects it mentions stay alive. scm_protect_object
9902 adds its argument to the list; scm_unprotect_object remove its
9903 argument from the list.
9906 ** scm_eval_0str now returns the value of the last expression
9909 ** The new function scm_read_0str reads an s-expression from a
9910 null-terminated string, and returns it.
9912 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
9913 to a Scheme port object.
9915 ** The new function `scm_set_program_arguments' allows C code to set
9916 the value returned by the Scheme `program-arguments' function.
9921 * Guile no longer includes sophisticated Tcl/Tk support.
9923 The old Tcl/Tk support was unsatisfying to us, because it required the
9924 user to link against the Tcl library, as well as Tk and Guile. The
9925 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
9926 referring to widgets by names, rather than exporting widgets to Scheme
9927 code as a special datatype.
9929 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
9930 maintainers described some very interesting changes in progress to the
9931 Tcl/Tk internals, which would facilitate clean interfaces between lone
9932 Tk and other interpreters --- even for garbage-collected languages
9933 like Scheme. They expected the new Tk to be publicly available in the
9936 Since it seems that Guile might soon have a new, cleaner interface to
9937 lone Tk, and that the old Guile/Tk glue code would probably need to be
9938 completely rewritten, we (Jim Blandy and Richard Stallman) have
9939 decided not to support the old code. We'll spend the time instead on
9940 a good interface to the newer Tk, as soon as it is available.
9942 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
9945 Copyright information:
9947 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
9949 Permission is granted to anyone to make or distribute verbatim copies
9950 of this document as received, in any medium, provided that the
9951 copyright notice and this permission notice are preserved,
9952 thus giving the recipient permission to redistribute in turn.
9954 Permission is granted to distribute modified versions
9955 of this document, or of portions of it,
9956 under the above conditions, provided also that they
9957 carry prominent notices stating who last changed them.
9962 paragraph-separate: "[
\f]*$"