1 Guile NEWS --- history of user-visible changes.
2 Copyright (C) 1996-2011 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 2.0.0 (changes since the 1.8.x series):
10 * New modules (see the manual for details)
12 ** `(srfi srfi-18)', more sophisticated multithreading support
13 ** `(srfi srfi-27)', sources of random bits
14 ** `(srfi srfi-38)', External Representation for Data With Shared Structure
15 ** `(srfi srfi-42)', eager comprehensions
16 ** `(srfi srfi-45)', primitives for expressing iterative lazy algorithms
17 ** `(srfi srfi-67)', compare procedures
18 ** `(ice-9 i18n)', internationalization support
19 ** `(ice-9 futures)', fine-grain parallelism
20 ** `(rnrs bytevectors)', the R6RS bytevector API
21 ** `(rnrs io ports)', a subset of the R6RS I/O port API
22 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
23 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
24 ** `(system foreign)', foreign function interface
25 ** `(sxml match)', a pattern matcher for SXML
26 ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library
27 ** `(system vm coverage)', a line-by-line code coverage library
28 ** `(web uri)', URI data type, parser, and unparser
29 ** `(web http)', HTTP header parsers and unparsers
30 ** `(web request)', HTTP request data type, reader, and writer
31 ** `(web response)', HTTP response data type, reader, and writer
32 ** `(web server)', Generic HTTP server
33 ** `(ice-9 poll)', a poll wrapper
34 ** `(web server http)', HTTP-over-TCP web server implementation
36 ** Replaced `(ice-9 match)' with Alex Shinn's compatible, hygienic matcher.
38 Guile's copy of Andrew K. Wright's `match' library has been replaced by
39 a compatible hygienic implementation by Alex Shinn. It is now
40 documented, see "Pattern Matching" in the manual.
42 Compared to Andrew K. Wright's `match', the new `match' lacks
43 `match-define', `match:error-control', `match:set-error-control',
44 `match:error', `match:set-error', and all structure-related procedures.
46 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
48 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
49 toolkit from Guile-Lib have been imported into Guile proper. See
50 "Standard Library" in the manual for more details.
52 ** Integration of lalr-scm, a parser generator
54 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
55 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
58 * Changes to the stand-alone interpreter
60 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
62 Compiled code loads much faster than Scheme source code, and runs around
63 3 or 4 times as fast, generating much less garbage in the process.
65 ** Evaluating Scheme code does not use the C stack.
67 Besides when compiling Guile itself, Guile no longer uses a recursive C
68 function as an evaluator. This obviates the need to check the C stack
69 pointer for overflow. Continuations still capture the C stack, however.
71 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
72 GUILE_SYSTEM_LOAD_COMPILED_PATH
74 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
75 for source files. It is a different path, however, because compiled
76 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
79 ** New read-eval-print loop (REPL) implementation
81 Running Guile with no arguments drops the user into the new REPL. See
82 "Using Guile Interactively" in the manual, for more information.
84 ** Remove old Emacs interface
86 Guile had an unused `--emacs' command line argument that was supposed to
87 help when running Guile inside Emacs. This option has been removed, and
88 the helper functions `named-module-use!' and `load-emacs-interface' have
91 ** Add `(system repl server)' module and `--listen' command-line argument
93 The `(system repl server)' module exposes procedures to listen on
94 sockets for connections, and serve REPLs to those clients. The --listen
95 command-line argument allows any Guile program to thus be remotely
98 See "Invoking Guile" for more information on `--listen'.
100 ** Command line additions
102 The guile binary now supports a new switch "-x", which can be used to
103 extend the list of filename extensions tried when loading files
106 ** New reader options: `square-brackets', `r6rs-hex-escapes',
109 The reader supports a new option (changeable via `read-options'),
110 `square-brackets', which instructs it to interpret square brackets as
111 parentheses. This option is on by default.
113 When the new `r6rs-hex-escapes' reader option is enabled, the reader
114 will recognize string escape sequences as defined in R6RS. R6RS string
115 escape sequences are incompatible with Guile's existing escapes, though,
116 so this option is off by default.
118 Additionally, Guile follows the R6RS newline escaping rules when the
119 `hungry-eol-escapes' option is enabled.
121 See "String Syntax" in the manual, for more information.
123 ** Function profiling and tracing at the REPL
125 The `,profile FORM' REPL meta-command can now be used to statistically
126 profile execution of a form, to see which functions are taking the most
127 time. See `,help profile' for more information.
129 Similarly, `,trace FORM' traces all function applications that occur
130 during the execution of `FORM'. See `,help trace' for more information.
132 ** Recursive debugging REPL on error
134 When Guile sees an error at the REPL, instead of saving the stack, Guile
135 will directly enter a recursive REPL in the dynamic context of the
136 error. See "Error Handling" in the manual, for more information.
138 A recursive REPL is the same as any other REPL, except that it
139 has been augmented with debugging information, so that one can inspect
140 the context of the error. The debugger has been integrated with the REPL
141 via a set of debugging meta-commands.
143 For example, one may access a backtrace with `,backtrace' (or
144 `,bt'). See "Interactive Debugging" in the manual, for more
147 ** New `guile-tools' commands: `compile', `disassemble'
149 Pass the `--help' command-line option to these commands for more
152 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
154 Users may now install Guile to nonstandard prefixes and just run
155 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
156 include `/path/to/lib'.
158 ** Guile's Emacs integration is now more keyboard-friendly
160 Backtraces may now be disclosed with the keyboard in addition to the
163 ** Load path change: search in version-specific paths before site paths
165 When looking for a module, Guile now searches first in Guile's
166 version-specific path (the library path), *then* in the site dir. This
167 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
168 installed. Also it should cut the number of `stat' system calls by half,
171 ** Value history in the REPL on by default
173 By default, the REPL will save computed values in variables like `$1',
174 `$2', and the like. There are programmatic and interactive interfaces to
175 control this. See "Value History" in the manual, for more information.
177 ** Readline tab completion for arguments
179 When readline is enabled, tab completion works for arguments too, not
180 just for the operator position.
182 ** Expression-oriented readline history
184 Guile's readline history now tries to operate on expressions instead of
185 input lines. Let us know what you think!
187 ** Interactive Guile follows GNU conventions
189 As recommended by the GPL, Guile now shows a brief copyright and
190 warranty disclaimer on startup, along with pointers to more information.
192 * Changes to Scheme functions and syntax
194 ** Support for R6RS libraries
196 The `library' and `import' forms from the latest Scheme report have been
197 added to Guile, in such a way that R6RS libraries share a namespace with
198 Guile modules. R6RS modules may import Guile modules, and are available
199 for Guile modules to import via use-modules and all the rest. See "R6RS
200 Libraries" in the manual for more information.
202 ** Implementations of R6RS libraries
204 Guile now has implementations for all of the libraries defined in the
205 R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS
206 Standard Libraries" in the manual for a full list of libraries.
208 ** Partial R6RS compatibility
210 Guile now has enough support for R6RS to run a reasonably large subset
213 Guile is not fully R6RS compatible. Many incompatibilities are simply
214 bugs, though some parts of Guile will remain R6RS-incompatible for the
215 foreseeable future. See "R6RS Incompatibilities" in the manual, for more
218 Please contact bug-guile@gnu.org if you have found an issue not
219 mentioned in that compatibility list.
221 ** New implementation of `primitive-eval'
223 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
224 still a C evaluator, used when building a fresh Guile to interpret the
225 compiler, so we can compile eval.scm. Thereafter all calls to
226 primitive-eval are implemented by VM-compiled code.
228 This allows all of Guile's procedures, be they interpreted or compiled,
229 to execute on the same stack, unifying multiple-value return semantics,
230 providing for proper tail recursion between interpreted and compiled
231 code, and simplifying debugging.
233 As part of this change, the evaluator no longer mutates the internal
234 representation of the code being evaluated in a thread-unsafe manner.
236 There are two negative aspects of this change, however. First, Guile
237 takes a lot longer to compile now. Also, there is less debugging
238 information available for debugging interpreted code. We hope to improve
239 both of these situations.
241 There are many changes to the internal C evalator interface, but all
242 public interfaces should be the same. See the ChangeLog for details. If
243 we have inadvertantly changed an interface that you were using, please
244 contact bug-guile@gnu.org.
246 ** Procedure removed: `the-environment'
248 This procedure was part of the interpreter's execution model, and does
249 not apply to the compiler.
251 ** No more `local-eval'
253 `local-eval' used to exist so that one could evaluate code in the
254 lexical context of a function. Since there is no way to get the lexical
255 environment any more, as that concept has no meaning for the compiler,
256 and a different meaning for the interpreter, we have removed the
259 If you think you need `local-eval', you should probably implement your
260 own metacircular evaluator. It will probably be as fast as Guile's
263 ** Scheme source files will now be compiled automatically.
265 If a compiled .go file corresponding to a .scm file is not found or is
266 not fresh, the .scm file will be compiled on the fly, and the resulting
267 .go file stored away. An advisory note will be printed on the console.
269 Note that this mechanism depends on the timestamp of the .go file being
270 newer than that of the .scm file; if the .scm or .go files are moved
271 after installation, care should be taken to preserve their original
274 Auto-compiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
275 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
276 will be created if needed.
278 To inhibit automatic compilation, set the GUILE_AUTO_COMPILE environment
279 variable to 0, or pass --no-auto-compile on the Guile command line.
281 ** New POSIX procedures: `getrlimit' and `setrlimit'
283 Note however that the interface of these functions is likely to change
284 in the next prerelease.
286 ** New POSIX procedure: `getsid'
288 Scheme binding for the `getsid' C library call.
290 ** New POSIX procedure: `getaddrinfo'
292 Scheme binding for the `getaddrinfo' C library function.
294 ** Multicast socket options
296 Support was added for the IP_MULTICAST_TTL and IP_MULTICAST_IF socket
297 options. See "Network Sockets and Communication" in the manual, for
300 ** `recv!', `recvfrom!', `send', `sendto' now deal in bytevectors
302 These socket procedures now take bytevectors as arguments, instead of
303 strings. There is some deprecated string support, however.
305 ** New GNU procedures: `setaffinity' and `getaffinity'.
307 See "Processes" in the manual, for more information.
309 ** New procedures: `compose', `negate', and `const'
311 See "Higher-Order Functions" in the manual, for more information.
313 ** New procedure in `(oops goops)': `method-formals'
315 ** New procedures in (ice-9 session): `add-value-help-handler!',
316 `remove-value-help-handler!', `add-name-help-handler!'
317 `remove-name-help-handler!', `procedure-arguments'
319 The value and name help handlers provide some minimal extensibility to
320 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
321 example, to make stexinfo help documentation available. See those
322 procedures' docstrings for more information.
324 `procedure-arguments' describes the arguments that a procedure can take,
325 combining arity and formals. For example:
327 (procedure-arguments resolve-interface)
328 => ((required . (name)) (rest . args))
330 Additionally, `module-commentary' is now publically exported from
333 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
335 These procedures created primitive fexprs for the old evaluator, and are
336 no longer supported. If you feel that you need these functions, you
337 probably need to write your own metacircular evaluator (which will
338 probably be as fast as Guile's, anyway).
340 ** New language: ECMAScript
342 Guile now ships with one other high-level language supported,
343 ECMAScript. The goal is to support all of version 3.1 of the standard,
344 but not all of the libraries are there yet. This support is not yet
345 documented; ask on the mailing list if you are interested.
347 ** New language: Brainfuck
349 Brainfuck is a toy language that closely models Turing machines. Guile's
350 brainfuck compiler is meant to be an example of implementing other
351 languages. See the manual for details, or
352 http://en.wikipedia.org/wiki/Brainfuck for more information about the
353 Brainfuck language itself.
355 ** New language: Elisp
357 Guile now has an experimental Emacs Lisp compiler and runtime. You can
358 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
359 Kraft and Brian Templeton, and all bugs to bug-guile@gnu.org.
361 ** Better documentation infrastructure for macros
363 It is now possible to introspect on the type of a macro, e.g.
364 syntax-rules, identifier-syntax, etc, and extract information about that
365 macro, such as the syntax-rules patterns or the defmacro arguments.
366 `(texinfo reflection)' takes advantage of this to give better macro
369 ** Support for arbitrary procedure metadata
371 Building on its support for docstrings, Guile now supports multiple
372 docstrings, adding them to the tail of a compiled procedure's
373 properties. For example:
379 (procedure-properties foo)
380 => ((name . foo) (documentation . "one") (documentation . "two"))
382 Also, vectors of pairs are now treated as additional metadata entries:
385 #((quz . #f) (docstring . "xyzzy"))
387 (procedure-properties bar)
388 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
390 This allows arbitrary literals to be embedded as metadata in a compiled
393 ** The psyntax expander now knows how to interpret the @ and @@ special
396 ** The psyntax expander is now hygienic with respect to modules.
398 Free variables in a macro are scoped in the module that the macro was
399 defined in, not in the module the macro is used in. For example, code
402 (define-module (foo) #:export (bar))
403 (define (helper x) ...)
405 (syntax-rules () ((_ x) (helper x))))
407 (define-module (baz) #:use-module (foo))
410 It used to be you had to export `helper' from `(foo)' as well.
411 Thankfully, this has been fixed.
413 ** Support for version information in Guile's `module' form
415 Guile modules now have a `#:version' field. See "R6RS Version
416 References", "General Information about Modules", "Using Guile Modules",
417 and "Creating Guile Modules" in the manual for more information.
419 ** Support for renaming bindings on module export
421 Wherever Guile accepts a symbol as an argument to specify a binding to
422 export, it now also accepts a pair of symbols, indicating that a binding
423 should be renamed on export. See "Creating Guile Modules" in the manual
424 for more information.
426 ** New procedure: `module-export-all!'
428 This procedure exports all current and future bindings from a module.
429 Use as `(module-export-all! (current-module))'.
431 ** New procedure `reload-module', and `,reload' REPL command
433 See "Module System Reflection" and "Module Commands" in the manual, for
436 ** `eval-case' has been deprecated, and replaced by `eval-when'.
438 The semantics of `eval-when' are easier to understand. See "Eval When"
439 in the manual, for more information.
441 ** Guile is now more strict about prohibiting definitions in expression
444 Although previous versions of Guile accepted it, the following
445 expression is not valid, in R5RS or R6RS:
447 (if test (define foo 'bar) (define foo 'baz))
449 In this specific case, it would be better to do:
451 (define foo (if test 'bar 'baz))
453 It is possible to circumvent this restriction with e.g.
454 `(module-define! (current-module) 'foo 'baz)'. Contact the list if you
457 ** Support for `letrec*'
459 Guile now supports `letrec*', a recursive lexical binding operator in
460 which the identifiers are bound in order. See "Local Bindings" in the
461 manual, for more details.
463 ** Internal definitions now expand to `letrec*'
465 Following the R6RS, internal definitions now expand to letrec* instead
466 of letrec. The following program is invalid for R5RS, but valid for
471 (define baz (+ bar 20))
474 ;; R5RS and Guile <= 1.8:
475 (foo) => Unbound variable: bar
476 ;; R6RS and Guile >= 2.0:
479 This change should not affect correct R5RS programs, or programs written
480 in earlier Guile dialects.
482 ** Macro expansion produces structures instead of s-expressions
484 In the olden days, macroexpanding an s-expression would yield another
485 s-expression. Though the lexical variables were renamed, expansions of
486 core forms like `if' and `begin' were still non-hygienic, as they relied
487 on the toplevel definitions of `if' et al being the conventional ones.
489 The solution is to expand to structures instead of s-expressions. There
490 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
491 etc. The expander already did this for compilation, producing Tree-IL
492 directly; it has been changed now to do so when expanding for the
495 ** Defmacros must now produce valid Scheme expressions.
497 It used to be that defmacros could unquote in Scheme values, as a way of
498 supporting partial evaluation, and avoiding some hygiene issues. For
501 (define (helper x) ...)
502 (define-macro (foo bar)
505 Assuming this macro is in the `(baz)' module, the direct translation of
508 (define (helper x) ...)
509 (define-macro (foo bar)
510 `((@@ (baz) helper) ,bar))
512 Of course, one could just use a hygienic macro instead:
516 ((_ bar) (helper bar))))
518 ** Guile's psyntax now supports docstrings and internal definitions.
520 The following Scheme is not strictly legal:
527 However its intent is fairly clear. Guile interprets "bar" to be the
528 docstring of `foo', and the definition of `baz' is still in definition
531 ** Support for settable identifier syntax
533 Following the R6RS, "variable transformers" are settable
534 identifier-syntax. See "Identifier macros" in the manual, for more
537 ** syntax-case treats `_' as a placeholder
539 Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches
540 anything, and binds no pattern variables. Unlike the R6RS, Guile also
541 permits `_' to be in the literals list for a pattern.
543 ** Macros need to be defined before their first use.
545 It used to be that with lazy memoization, this might work:
549 (define-macro (ref x) x)
552 But now, the body of `foo' is interpreted to mean a call to the toplevel
553 `ref' function, instead of a macro expansion. The solution is to define
554 macros before code that uses them.
556 ** Functions needed by macros at expand-time need to be present at
559 For example, this code will work at the REPL:
561 (define (double-helper x) (* x x))
562 (define-macro (double-literal x) (double-helper x))
563 (double-literal 2) => 4
565 But it will not work when a file is compiled, because the definition of
566 `double-helper' is not present at expand-time. The solution is to wrap
567 the definition of `double-helper' in `eval-when':
569 (eval-when (load compile eval)
570 (define (double-helper x) (* x x)))
571 (define-macro (double-literal x) (double-helper x))
572 (double-literal 2) => 4
574 See the documentation for eval-when for more information.
576 ** `macroexpand' produces structures, not S-expressions.
578 Given the need to maintain referential transparency, both lexically and
579 modular, the result of expanding Scheme expressions is no longer itself
580 an s-expression. If you want a human-readable approximation of the
581 result of `macroexpand', call `tree-il->scheme' from `(language
584 ** Removed function: `macroexpand-1'
586 It is unclear how to implement `macroexpand-1' with syntax-case, though
587 PLT Scheme does prove that it is possible.
589 ** New reader macros: #' #` #, #,@
591 These macros translate, respectively, to `syntax', `quasisyntax',
592 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
593 These reader macros may be overridden by `read-hash-extend'.
595 ** Incompatible change to #'
597 Guile did have a #' hash-extension, by default, which just returned the
598 subsequent datum: #'foo => foo. In the unlikely event that anyone
599 actually used this, this behavior may be reinstated via the
600 `read-hash-extend' mechanism.
602 ** `unquote' and `unquote-splicing' accept multiple expressions
604 As per the R6RS, these syntax operators can now accept any number of
605 expressions to unquote.
607 ** Scheme expresssions may be commented out with #;
609 #; comments out an entire expression. See SRFI-62 or the R6RS for more
612 ** Prompts: Delimited, composable continuations
614 Guile now has prompts as part of its primitive language. See "Prompts"
615 in the manual, for more information.
617 Expressions entered in at the REPL, or from the command line, are
618 surrounded by a prompt with the default prompt tag.
620 ** `make-stack' with a tail-called procedural narrowing argument no longer
621 works (with compiled procedures)
623 It used to be the case that a captured stack could be narrowed to select
624 calls only up to or from a certain procedure, even if that procedure
625 already tail-called another procedure. This was because the debug
626 information from the original procedure was kept on the stack.
628 Now with the new compiler, the stack only contains active frames from
629 the current continuation. A narrow to a procedure that is not in the
630 stack will result in an empty stack. To fix this, narrow to a procedure
631 that is active in the current continuation, or narrow to a specific
632 number of stack frames.
634 ** Backtraces through compiled procedures only show procedures that are
635 active in the current continuation
637 Similarly to the previous issue, backtraces in compiled code may be
638 different from backtraces in interpreted code. There are no semantic
639 differences, however. Please mail bug-guile@gnu.org if you see any
640 deficiencies with Guile's backtraces.
642 ** `positions' reader option enabled by default
644 This change allows primitive-load without --auto-compile to also
645 propagate source information through the expander, for better errors and
646 to let macros know their source locations. The compiler was already
647 turning it on anyway.
649 ** New macro: `current-source-location'
651 The macro returns the current source location (to be documented).
653 ** syntax-rules and syntax-case macros now propagate source information
654 through to the expanded code
656 This should result in better backtraces.
658 ** The currying behavior of `define' has been removed.
660 Before, `(define ((f a) b) (* a b))' would translate to
662 (define f (lambda (a) (lambda (b) (* a b))))
664 Now a syntax error is signaled, as this syntax is not supported by
665 default. Use the `(ice-9 curried-definitions)' module to get back the
668 ** New procedure, `define!'
670 `define!' is a procedure that takes two arguments, a symbol and a value,
671 and binds the value to the symbol in the current module. It's useful to
672 programmatically make definitions in the current module, and is slightly
673 less verbose than `module-define!'.
675 ** All modules have names now
677 Before, you could have anonymous modules: modules without names. Now,
678 because of hygiene and macros, all modules have names. If a module was
679 created without a name, the first time `module-name' is called on it, a
680 fresh name will be lazily generated for it.
682 ** The module namespace is now separate from the value namespace
684 It was a little-known implementation detail of Guile's module system
685 that it was built on a single hierarchical namespace of values -- that
686 if there was a module named `(foo bar)', then in the module named
687 `(foo)' there was a binding from `bar' to the `(foo bar)' module.
689 This was a neat trick, but presented a number of problems. One problem
690 was that the bindings in a module were not apparent from the module
691 itself; perhaps the `(foo)' module had a private binding for `bar', and
692 then an external contributor defined `(foo bar)'. In the end there can
693 be only one binding, so one of the two will see the wrong thing, and
694 produce an obtuse error of unclear provenance.
696 Also, the public interface of a module was also bound in the value
697 namespace, as `%module-public-interface'. This was a hack from the early
698 days of Guile's modules.
700 Both of these warts have been fixed by the addition of fields in the
701 `module' data type. Access to modules and their interfaces from the
702 value namespace has been deprecated, and all accessors use the new
703 record accessors appropriately.
705 When Guile is built with support for deprecated code, as is the default,
706 the value namespace is still searched for modules and public interfaces,
707 and a deprecation warning is raised as appropriate.
709 Finally, to support lazy loading of modules as one used to be able to do
710 with module binder procedures, Guile now has submodule binders, called
711 if a given submodule is not found. See boot-9.scm for more information.
713 ** New procedures: module-ref-submodule, module-define-submodule,
714 nested-ref-module, nested-define-module!, local-ref-module,
717 These new accessors are like their bare variants, but operate on
718 namespaces instead of values.
720 ** The (app modules) module tree is officially deprecated
722 It used to be that one could access a module named `(foo bar)' via
723 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
724 modules)' bit was a never-used and never-documented abstraction, and has
725 been deprecated. See the following mail for a full discussion:
727 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
729 The `%app' binding is also deprecated.
731 ** `module-filename' field and accessor
733 Modules now record the file in which they are defined. This field may be
734 accessed with the new `module-filename' procedure.
736 ** Modules load within a known environment
738 It takes a few procedure calls to define a module, and those procedure
739 calls need to be in scope. Now we ensure that the current module when
740 loading a module is one that has the needed bindings, instead of relying
743 ** `load' is a macro (!) that resolves paths relative to source file dir
745 The familiar Schem `load' procedure is now a macro that captures the
746 name of the source file being expanded, and dispatches to the new
747 `load-in-vicinity'. Referencing `load' by bare name returns a closure
748 that embeds the current source file name.
750 This fix allows `load' of relative paths to be resolved with respect to
751 the location of the file that calls `load'.
753 ** Many syntax errors have different texts now
755 Syntax errors still throw to the `syntax-error' key, but the arguments
756 are often different now. Perhaps in the future, Guile will switch to
757 using standard SRFI-35 conditions.
759 ** Returning multiple values to compiled code will silently truncate the
760 values to the expected number
762 For example, the interpreter would raise an error evaluating the form,
763 `(+ (values 1 2) (values 3 4))', because it would see the operands as
764 being two compound "values" objects, to which `+' does not apply.
766 The compiler, on the other hand, receives multiple values on the stack,
767 not as a compound object. Given that it must check the number of values
768 anyway, if too many values are provided for a continuation, it chooses
769 to truncate those values, effectively evaluating `(+ 1 3)' instead.
771 The idea is that the semantics that the compiler implements is more
772 intuitive, and the use of the interpreter will fade out with time.
773 This behavior is allowed both by the R5RS and the R6RS.
775 ** Multiple values in compiled code are not represented by compound
778 This change may manifest itself in the following situation:
780 (let ((val (foo))) (do-something) val)
782 In the interpreter, if `foo' returns multiple values, multiple values
783 are produced from the `let' expression. In the compiler, those values
784 are truncated to the first value, and that first value is returned. In
785 the compiler, if `foo' returns no values, an error will be raised, while
786 the interpreter would proceed.
788 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
789 behavior is more correct, however. If you wish to preserve a potentially
790 multiply-valued return, you will need to set up a multiple-value
791 continuation, using `call-with-values'.
793 ** Defmacros are now implemented in terms of syntax-case.
795 The practical ramification of this is that the `defmacro?' predicate has
796 been removed, along with `defmacro-transformer', `macro-table',
797 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
798 `defmacro:transformer'. This is because defmacros are simply macros. If
799 any of these procedures provided useful facilities to you, we encourage
800 you to contact the Guile developers.
802 ** Hygienic macros documented as the primary syntactic extension mechanism.
804 The macro documentation was finally fleshed out with some documentation
805 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
806 expansion process. See "Macros" in the manual, for details.
808 ** psyntax is now the default expander
810 Scheme code is now expanded by default by the psyntax hygienic macro
811 expander. Expansion is performed completely before compilation or
814 Notably, syntax errors will be signalled before interpretation begins.
815 In the past, many syntax errors were only detected at runtime if the
816 code in question was memoized.
818 As part of its expansion, psyntax renames all lexically-bound
819 identifiers. Original identifier names are preserved and given to the
820 compiler, but the interpreter will see the renamed variables, e.g.,
821 `x432' instead of `x'.
823 Note that the psyntax that Guile uses is a fork, as Guile already had
824 modules before incompatible modules were added to psyntax -- about 10
825 years ago! Thus there are surely a number of bugs that have been fixed
826 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
828 ** syntax-rules and syntax-case are available by default.
830 There is no longer any need to import the `(ice-9 syncase)' module
831 (which is now deprecated). The expander may be invoked directly via
832 `macroexpand', though it is normally searched for via the current module
835 Also, the helper routines for syntax-case are available in the default
836 environment as well: `syntax->datum', `datum->syntax',
837 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
838 `identifier?', and `syntax-violation'. See the R6RS for documentation.
840 ** Tail patterns in syntax-case
842 Guile has pulled in some more recent changes from the psyntax portable
843 syntax expander, to implement support for "tail patterns". Such patterns
844 are supported by syntax-rules and syntax-case. This allows a syntax-case
845 match clause to have ellipses, then a pattern at the end. For example:
849 ((_ val match-clause ... (else e e* ...))
852 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
853 tail pattern for the else clause. Thanks to Andreas Rottmann for the
854 patch, and Kent Dybvig for the code.
856 ** Lexical bindings introduced by hygienic macros may not be referenced
857 by nonhygienic macros.
859 If a lexical binding is introduced by a hygienic macro, it may not be
860 referenced by a nonhygienic macro. For example, this works:
863 (define-macro (bind-x val body)
864 `(let ((x ,val)) ,body))
865 (define-macro (ref x)
872 (define-syntax bind-x
874 ((_ val body) (let ((x val)) body))))
875 (define-macro (ref x)
879 It is not normal to run into this situation with existing code. However,
880 if you have defmacros that expand to hygienic macros, it is possible to
881 run into situations like this. For example, if you have a defmacro that
882 generates a `while' expression, the `break' bound by the `while' may not
883 be visible within other parts of your defmacro. The solution is to port
884 from defmacros to syntax-rules or syntax-case.
886 ** Macros may no longer be referenced as first-class values.
888 In the past, you could evaluate e.g. `if', and get its macro value. Now,
889 expanding this form raises a syntax error.
891 Macros still /exist/ as first-class values, but they must be
892 /referenced/ via the module system, e.g. `(module-ref (current-module)
895 ** Macros may now have docstrings.
897 `object-documentation' from `(ice-9 documentation)' may be used to
898 retrieve the docstring, once you have a macro value -- but see the above
899 note about first-class macros. Docstrings are associated with the syntax
900 transformer procedures.
902 ** `case-lambda' is now available in the default environment.
904 The binding in the default environment is equivalent to the one from the
905 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
906 to maintain compatibility with Guile 1.8 and earlier.
908 ** Procedures may now have more than one arity.
910 This can be the case, for example, in case-lambda procedures. The
911 arities of compiled procedures may be accessed via procedures from the
912 `(system vm program)' module; see "Compiled Procedures", "Optional
913 Arguments", and "Case-lambda" in the manual.
915 ** Deprecate arity access via (procedure-properties proc 'arity)
917 Instead of accessing a procedure's arity as a property, use the new
918 `procedure-minimum-arity' function, which gives the most permissive
919 arity that the the function has, in the same format as the old arity
922 ** `lambda*' and `define*' are now available in the default environment
924 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
925 compatibility purposes. No semantic change has been made (we hope).
926 Optional and keyword arguments now dispatch via special VM operations,
927 without the need to cons rest arguments, making them very fast.
929 ** New syntax: define-once
931 `define-once' is like Lisp's `defvar': it creates a toplevel binding,
932 but only if one does not exist already.
934 ** New function, `truncated-print', with `format' support
936 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
937 will ensure that the output stays within a certain width, truncating the
938 output in what is hopefully an intelligent manner. See the manual for
941 There is a new `format' specifier, `~@y', for doing a truncated
942 print (as opposed to `~y', which does a pretty-print). See the `format'
943 documentation for more details.
945 ** Better pretty-printing
947 Indentation recognizes more special forms, like `syntax-case', and read
948 macros like `quote' are printed better.
950 ** Passing a number as the destination of `format' is deprecated
952 The `format' procedure in `(ice-9 format)' now emits a deprecation
953 warning if a number is passed as its first argument.
955 Also, it used to be that you could omit passing a port to `format', in
956 some cases. This still works, but has been formally deprecated.
958 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
960 Guile now implements SRFI-4 vectors using bytevectors. Often when you
961 have a numeric vector, you end up wanting to write its bytes somewhere,
962 or have access to the underlying bytes, or read in bytes from somewhere
963 else. Bytevectors are very good at this sort of thing. But the SRFI-4
964 APIs are nicer to use when doing number-crunching, because they are
965 addressed by element and not by byte.
967 So as a compromise, Guile allows all bytevector functions to operate on
968 numeric vectors. They address the underlying bytes in the native
969 endianness, as one would expect.
971 Following the same reasoning, that it's just bytes underneath, Guile
972 also allows uniform vectors of a given type to be accessed as if they
973 were of any type. One can fill a u32vector, and access its elements with
974 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
977 In this way, uniform numeric vectors may be written to and read from
978 input/output ports using the procedures that operate on bytevectors.
980 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
981 inlined to the VM instructions for bytevector access.
983 See "SRFI-4" in the manual, for more information.
985 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
987 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
988 are part of the standard. Complex uniform vectors and the
989 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
991 Guile's default environment imports `(srfi srfi-4)', and probably should
992 import `(srfi srfi-4 gnu)' as well.
994 See "SRFI-4 Extensions" in the manual, for more information.
996 ** New syntax: include-from-path.
998 `include-from-path' is like `include', except it looks for its file in
999 the load path. It can be used to compile other files into a file.
1001 ** New syntax: quasisyntax.
1003 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
1004 documentation for more information. Thanks to Andre van Tonder for the
1007 ** `*unspecified*' is identifier syntax
1009 `*unspecified*' is no longer a variable, so it is optimized properly by
1010 the compiler, and is not `set!'-able.
1012 ** Changes and bugfixes in numerics code
1014 *** Added six new sets of fast quotient and remainder operators
1016 Added six new sets of fast quotient and remainder operator pairs with
1017 different semantics than the R5RS operators. They support not only
1018 integers, but all reals, including exact rationals and inexact
1019 floating point numbers.
1021 These procedures accept two real numbers N and D, where the divisor D
1022 must be non-zero. Each set of operators computes an integer quotient
1023 Q and a real remainder R such that N = Q*D + R and |R| < |D|. They
1024 differ only in how N/D is rounded to produce Q.
1026 `euclidean-quotient' returns the integer Q and `euclidean-remainder'
1027 returns the real R such that N = Q*D + R and 0 <= R < |D|. `euclidean/'
1028 returns both Q and R, and is more efficient than computing each
1029 separately. Note that when D > 0, `euclidean-quotient' returns
1030 floor(N/D), and when D < 0 it returns ceiling(N/D).
1032 `centered-quotient', `centered-remainder', and `centered/' are similar
1033 except that the range of remainders is -abs(D/2) <= R < abs(D/2), and
1034 `centered-quotient' rounds N/D to the nearest integer. Note that these
1035 operators are equivalent to the R6RS integer division operators `div',
1036 `mod', `div-and-mod', `div0', `mod0', and `div0-and-mod0'.
1038 `floor-quotient' and `floor-remainder' compute Q and R, respectively,
1039 where Q has been rounded toward negative infinity. `floor/' returns
1040 both Q and R, and is more efficient than computing each separately.
1041 Note that when applied to integers, `floor-remainder' is equivalent to
1042 the R5RS integer-only `modulo' operator. `ceiling-quotient',
1043 `ceiling-remainder', and `ceiling/' are similar except that Q is
1044 rounded toward positive infinity.
1046 For `truncate-quotient', `truncate-remainder', and `truncate/', Q is
1047 rounded toward zero. Note that when applied to integers,
1048 `truncate-quotient' and `truncate-remainder' are equivalent to the
1049 R5RS integer-only operators `quotient' and `remainder'.
1051 For `round-quotient', `round-remainder', and `round/', Q is rounded to
1052 the nearest integer, with ties going to the nearest even integer.
1054 *** Complex number changes
1056 Guile is now able to represent non-real complex numbers whose
1057 imaginary part is an _inexact_ zero (0.0 or -0.0), per R6RS.
1058 Previously, such numbers were immediately changed into inexact reals.
1060 (real? 0.0+0.0i) now returns #f, per R6RS, although (zero? 0.0+0.0i)
1061 still returns #t, per R6RS. (= 0 0.0+0.0i) and (= 0.0 0.0+0.0i) are
1062 #t, but the same comparisons using `eqv?' or `equal?' are #f.
1064 Like other non-real numbers, these complex numbers with inexact zero
1065 imaginary part will raise exceptions is passed to procedures requiring
1066 reals, such as `<', `>', `<=', `>=', `min', `max', `positive?',
1067 `negative?', `inf?', `nan?', `finite?', etc.
1069 **** `make-rectangular' changes
1071 scm_make_rectangular `make-rectangular' now returns a real number only
1072 if the imaginary part is an _exact_ 0. Previously, it would return a
1073 real number if the imaginary part was an inexact zero.
1075 scm_c_make_rectangular now always returns a non-real complex number,
1076 even if the imaginary part is zero. Previously, it would return a
1077 real number if the imaginary part was zero.
1079 **** `make-polar' changes
1081 scm_make_polar `make-polar' now returns a real number only if the
1082 angle or magnitude is an _exact_ 0. If the magnitude is an exact 0,
1083 it now returns an exact 0. Previously, it would return a real
1084 number if the imaginary part was an inexact zero.
1086 scm_c_make_polar now always returns a non-real complex number, even if
1087 the imaginary part is 0.0. Previously, it would return a real number
1088 if the imaginary part was 0.0.
1090 **** `imag-part' changes
1092 scm_imag_part `imag-part' now returns an exact 0 if applied to an
1093 inexact real number. Previously it returned an inexact zero in this
1096 *** `eqv?' and `equal?' now compare numbers equivalently
1098 scm_equal_p `equal?' now behaves equivalently to scm_eqv_p `eqv?' for
1099 numeric values, per R5RS. Previously, equal? worked differently,
1100 e.g. `(equal? 0.0 -0.0)' returned #t but `(eqv? 0.0 -0.0)' returned #f,
1101 and `(equal? +nan.0 +nan.0)' returned #f but `(eqv? +nan.0 +nan.0)'
1104 *** `(equal? +nan.0 +nan.0)' now returns #t
1106 Previously, `(equal? +nan.0 +nan.0)' returned #f, although
1107 `(let ((x +nan.0)) (equal? x x))' and `(eqv? +nan.0 +nan.0)'
1108 both returned #t. R5RS requires that `equal?' behave like
1109 `eqv?' when comparing numbers.
1111 *** Change in handling products `*' involving exact 0
1113 scm_product `*' now handles exact 0 differently. A product containing
1114 an exact 0 now returns an exact 0 if and only if the other arguments
1115 are all exact. An inexact zero is returned if and only if the other
1116 arguments are all finite but not all exact. If an infinite or NaN
1117 value is present, a NaN value is returned. Previously, any product
1118 containing an exact 0 yielded an exact 0, regardless of the other
1121 *** `expt' and `integer-expt' changes when the base is 0
1123 While `(expt 0 0)' is still 1, and `(expt 0 N)' for N > 0 is still
1124 zero, `(expt 0 N)' for N < 0 is now a NaN value, and likewise for
1125 integer-expt. This is more correct, and conforming to R6RS, but seems
1126 to be incompatible with R5RS, which would return 0 for all non-zero
1129 *** `expt' and `integer-expt' are more generic, less strict
1131 When raising to an exact non-negative integer exponent, `expt' and
1132 `integer-expt' are now able to exponentiate any object that can be
1133 multiplied using `*'. They can also raise an object to an exact
1134 negative integer power if its reciprocal can be taken using `/'.
1135 In order to allow this, the type of the first argument is no longer
1136 checked when raising to an exact integer power. If the exponent is 0
1137 or 1, the first parameter is not manipulated at all, and need not
1138 even support multiplication.
1140 *** Infinities are no longer integers, nor rationals
1142 scm_integer_p `integer?' and scm_rational_p `rational?' now return #f
1143 for infinities, per R6RS. Previously they returned #t for real
1144 infinities. The real infinities and NaNs are still considered real by
1145 scm_real `real?' however, per R6RS.
1147 *** NaNs are no longer rationals
1149 scm_rational_p `rational?' now returns #f for NaN values, per R6RS.
1150 Previously it returned #t for real NaN values. They are still
1151 considered real by scm_real `real?' however, per R6RS.
1153 *** `inf?' and `nan?' now throw exceptions for non-reals
1155 The domain of `inf?' and `nan?' is the real numbers. Guile now signals
1156 an error when a non-real number or non-number is passed to these
1157 procedures. (Note that NaNs _are_ considered numbers by scheme, despite
1160 *** `rationalize' bugfixes and changes
1162 Fixed bugs in scm_rationalize `rationalize'. Previously, it returned
1163 exact integers unmodified, although that was incorrect if the epsilon
1164 was at least 1 or inexact, e.g. (rationalize 4 1) should return 3 per
1165 R5RS and R6RS, but previously it returned 4. It also now handles
1166 cases involving infinities and NaNs properly, per R6RS.
1168 *** Trigonometric functions now return exact numbers in some cases
1170 scm_sin `sin', scm_cos `cos', scm_tan `tan', scm_asin `asin', scm_acos
1171 `acos', scm_atan `atan', scm_sinh `sinh', scm_cosh `cosh', scm_tanh
1172 `tanh', scm_sys_asinh `asinh', scm_sys_acosh `acosh', and
1173 scm_sys_atanh `atanh' now return exact results in some cases.
1175 *** New procedure: `finite?'
1177 Add scm_finite_p `finite?' from R6RS to guile core, which returns #t
1178 if and only if its argument is neither infinite nor a NaN. Note that
1179 this is not the same as (not (inf? x)) or (not (infinite? x)), since
1180 NaNs are neither finite nor infinite.
1182 *** Improved exactness handling for complex number parsing
1184 When parsing non-real complex numbers, exactness specifiers are now
1185 applied to each component, as is done in PLT Scheme. For complex
1186 numbers written in rectangular form, exactness specifiers are applied
1187 to the real and imaginary parts before calling scm_make_rectangular.
1188 For complex numbers written in polar form, exactness specifiers are
1189 applied to the magnitude and angle before calling scm_make_polar.
1191 Previously, exactness specifiers were applied to the number as a whole
1192 _after_ calling scm_make_rectangular or scm_make_polar.
1194 For example, (string->number "#i5.0+0i") now does the equivalent of:
1196 (make-rectangular (exact->inexact 5.0) (exact->inexact 0))
1198 which yields 5.0+0.0i. Previously it did the equivalent of:
1200 (exact->inexact (make-rectangular 5.0 0))
1204 ** Unicode characters
1206 Unicode characters may be entered in octal format via e.g. `#\454', or
1207 created via (integer->char 300). A hex external representation will
1208 probably be introduced at some point.
1212 Internally, strings are now represented either in the `latin-1'
1213 encoding, one byte per character, or in UTF-32, with four bytes per
1214 character. Strings manage their own allocation, switching if needed.
1216 Extended characters may be written in a literal string using the
1217 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
1218 or 24-bit codepoints, respectively, or entered directly in the native
1219 encoding of the port on which the string is read.
1223 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
1225 ** Support for non-ASCII source code files
1227 The default reader now handles source code files for some of the
1228 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
1229 should have an encoding declaration near the top of the file. Also,
1230 there is a new function, `file-encoding', that scans a port for a coding
1231 declaration. See the section of the manual entitled, "Character Encoding
1234 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
1235 code. This use is now discouraged. Binary input and output is
1236 currently supported by opening ports in the ISO-8859-1 locale.
1238 ** Source files default to UTF-8.
1240 If source files do not specify their encoding via a `coding:' block,
1241 the default encoding is UTF-8, instead of being taken from the current
1244 ** Interactive Guile installs the current locale.
1246 Instead of leaving the user in the "C" locale, running the Guile REPL
1247 installs the current locale. [FIXME xref?]
1249 ** Support for locale transcoding when reading from and writing to ports
1251 Ports now have an associated character encoding, and port read and write
1252 operations do conversion to and from locales automatically. Ports also
1253 have an associated strategy for how to deal with locale conversion
1256 See the documentation in the manual for the four new support functions,
1257 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
1258 and `port-conversion-strategy'.
1260 ** String and SRFI-13 functions can operate on Unicode strings
1262 ** Unicode support for SRFI-14 character sets
1264 The default character sets are no longer locale dependent and contain
1265 characters from the whole Unicode range. There is a new predefined
1266 character set, `char-set:designated', which contains all assigned
1267 Unicode characters. There is a new debugging function, `%char-set-dump'.
1269 ** Character functions operate on Unicode characters
1271 `char-upcase' and `char-downcase' use default Unicode casing rules.
1272 Character comparisons such as `char<?' and `char-ci<?' now sort based on
1273 Unicode code points.
1275 ** Global variables `scm_charnames' and `scm_charnums' are removed
1277 These variables contained the names of control characters and were
1278 used when writing characters. While these were global, they were
1279 never intended to be public API. They have been replaced with private
1282 ** EBCDIC support is removed
1284 There was an EBCDIC compile flag that altered some of the character
1285 processing. It appeared that full EBCDIC support was never completed
1286 and was unmaintained.
1288 ** Compile-time warnings
1290 Guile can warn about potentially unbound free variables. Pass the
1291 -Wunbound-variable on the `guile-tools compile' command line, or add
1292 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
1293 invocation. Warnings are also enabled by default for expressions entered
1296 Guile can also warn when you pass the wrong number of arguments to a
1297 procedure, with -Warity-mismatch, or `arity-mismatch' in the
1298 `#:warnings' as above.
1300 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
1301 warn about unused local or global (top-level) variables, and `-Wformat',
1302 to check for various errors related to the `format' procedure.
1304 ** A new `memoize-symbol' evaluator trap has been added.
1306 This trap can be used for efficiently implementing a Scheme code
1309 ** Duplicate bindings among used modules are resolved lazily.
1311 This slightly improves program startup times.
1313 ** New thread cancellation and thread cleanup API
1315 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
1317 ** New threads are in `(guile-user)' by default, not `(guile)'
1319 It used to be that a new thread entering Guile would do so in the
1320 `(guile)' module, unless this was the first time Guile was initialized,
1321 in which case it was `(guile-user)'. This has been fixed to have all
1322 new threads unknown to Guile default to `(guile-user)'.
1324 ** New helpers: `print-exception', `set-exception-printer!'
1326 These functions implement an extensible exception printer. Guile
1327 registers printers for all of the exceptions it throws. Users may add
1328 their own printers. There is also `scm_print_exception', for use by C
1329 programs. Pleasantly, this allows SRFI-35 and R6RS exceptions to be
1330 printed appropriately.
1332 ** GOOPS dispatch in scheme
1334 As an implementation detail, GOOPS dispatch is no longer implemented by
1335 special evaluator bytecodes, but rather directly via a Scheme function
1336 associated with an applicable struct. There is some VM support for the
1337 underlying primitives, like `class-of'.
1339 This change will in the future allow users to customize generic function
1340 dispatch without incurring a performance penalty, and allow us to
1341 implement method combinations.
1343 ** Applicable struct support
1345 One may now make structs from Scheme that may be applied as procedures.
1346 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
1347 That struct will be the vtable of your applicable structs; instances of
1348 that new struct are assumed to have the procedure in their first slot.
1349 `<applicable-struct-vtable>' is like Common Lisp's
1350 `funcallable-standard-class'. Likewise there is
1351 `<applicable-struct-with-setter-vtable>', which looks for the setter in
1352 the second slot. This needs to be better documented.
1356 GOOPS had a number of concepts that were relevant to the days of Tcl,
1357 but not any more: operators and entities, mainly. These objects were
1358 never documented, and it is unlikely that they were ever used. Operators
1359 were a kind of generic specific to the Tcl support. Entities were
1360 replaced by applicable structs, mentioned above.
1362 ** New struct slot allocation: "hidden"
1364 A hidden slot is readable and writable, but will not be initialized by a
1365 call to make-struct. For example in your layout you would say "ph"
1366 instead of "pw". Hidden slots are useful for adding new slots to a
1367 vtable without breaking existing invocations to make-struct.
1369 ** eqv? not a generic
1371 One used to be able to extend `eqv?' as a primitive-generic, but no
1372 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
1373 should be able to compile to static dispatch tables, it doesn't make
1374 sense to allow extensions that would subvert this optimization.
1376 ** `inet-ntop' and `inet-pton' are always available.
1378 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
1379 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
1380 functions are deprecated.
1382 ** `getopt-long' parsing errors throw to `quit', not `misc-error'
1384 This change should inhibit backtraces on argument parsing errors.
1385 `getopt-long' has been modified to print out the error that it throws
1388 ** New primitive: `tmpfile'.
1390 See "File System" in the manual.
1392 ** Random generator state may be serialized to a datum
1394 `random-state->datum' will serialize a random state to a datum, which
1395 may be written out, read back in later, and revivified using
1396 `datum->random-state'. See "Random" in the manual, for more details.
1398 ** Fix random number generator on 64-bit platforms
1400 There was a nasty bug on 64-bit platforms in which asking for a random
1401 integer with a range between 2**32 and 2**64 caused a segfault. After
1402 many embarrassing iterations, this was fixed.
1404 ** Fast bit operations.
1406 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
1407 have dedicated bytecodes. Guile is not just for symbolic computation,
1408 it's for number crunching too.
1410 ** Faster SRFI-9 record access
1412 SRFI-9 records are now implemented directly on top of Guile's structs,
1413 and their accessors are defined in such a way that normal call-sites
1414 inline to special VM opcodes, while still allowing for the general case
1415 (e.g. passing a record accessor to `apply').
1417 ** R6RS block comment support
1419 Guile now supports R6RS nested block comments. The start of a comment is
1420 marked with `#|', and the end with `|#'.
1422 ** `guile-2' cond-expand feature
1424 To test if your code is running under Guile 2.0 (or its alpha releases),
1425 test for the `guile-2' cond-expand feature. Like this:
1427 (cond-expand (guile-2 (eval-when (compile)
1428 ;; This must be evaluated at compile time.
1429 (fluid-set! current-reader my-reader)))
1431 ;; Earlier versions of Guile do not have a
1432 ;; separate compilation phase.
1433 (fluid-set! current-reader my-reader)))
1435 ** New global variables: %load-compiled-path, %load-compiled-extensions
1437 These are analogous to %load-path and %load-extensions.
1439 ** New fluid: `%file-port-name-canonicalization'
1441 This fluid parameterizes the file names that are associated with file
1442 ports. If %file-port-name-canonicalization is 'absolute, then file names
1443 are canonicalized to be absolute paths. If it is 'relative, then the
1444 name is canonicalized, but any prefix corresponding to a member of
1445 `%load-path' is stripped off. Otherwise the names are passed through
1448 In addition, the `compile-file' and `compile-and-load' procedures bind
1449 %file-port-name-canonicalization to their `#:canonicalization' keyword
1450 argument, which defaults to 'relative. In this way, one might compile
1451 "../module/ice-9/boot-9.scm", but the path that gets residualized into
1452 the .go is "ice-9/boot-9.scm".
1454 ** New procedure, `make-promise'
1456 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
1458 ** `defined?' may accept a module as its second argument
1460 Previously it only accepted internal structures from the evaluator.
1462 ** New entry into %guile-build-info: `ccachedir'
1464 ** Fix bug in `module-bound?'.
1466 `module-bound?' was returning true if a module did have a local
1467 variable, but one that was unbound, but another imported module bound
1468 the variable. This was an error, and was fixed.
1470 ** `(ice-9 syncase)' has been deprecated.
1472 As syntax-case is available by default, importing `(ice-9 syncase)' has
1473 no effect, and will trigger a deprecation warning.
1475 ** New readline history functions
1477 The (ice-9 readline) module now provides add-history, read-history,
1478 write-history and clear-history, which wrap the corresponding GNU
1479 History library functions.
1481 ** Removed deprecated uniform array procedures:
1482 dimensions->uniform-array, list->uniform-array, array-prototype
1484 Instead, use make-typed-array, list->typed-array, or array-type,
1487 ** Deprecate the old `scm-style-repl'
1489 The following bindings from boot-9 are now found in `(ice-9
1490 scm-style-repl)': `scm-style-repl', `error-catching-loop',
1491 `error-catching-repl', `bad-throw', `scm-repl-silent'
1492 `assert-repl-silence', `repl-print-unspecified',
1493 `assert-repl-print-unspecified', `scm-repl-verbose',
1494 `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl',
1495 `default-pre-unwind-handler', `handle-system-error',
1497 The following bindings have been deprecated, with no replacement:
1498 `pre-unwind-handler-dispatch'.
1500 The following bindings have been totally removed:
1501 `before-signal-stack'.
1503 Deprecated forwarding shims have been installed so that users that
1504 expect these bindings in the main namespace will still work, but receive
1505 a deprecation warning.
1507 ** `set-batch-mode?!' replaced by `ensure-batch-mode!'
1509 "Batch mode" is a flag used to tell a program that it is not running
1510 interactively. One usually turns it on after a fork. It may not be
1511 turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!',
1512 because it is a better interface, as it can only turn on batch mode, not
1515 ** Deprecate `save-stack', `the-last-stack'
1517 It used to be that the way to debug programs in Guile was to capture the
1518 stack at the time of error, drop back to the REPL, then debug that
1519 stack. But this approach didn't compose, was tricky to get right in the
1520 presence of threads, and was not very powerful.
1522 So `save-stack', `stack-saved?', and `the-last-stack' have been moved to
1523 `(ice-9 save-stack)', with deprecated bindings left in the root module.
1525 ** `top-repl' has its own module
1527 The `top-repl' binding, called with Guile is run interactively, is now
1528 is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was
1529 left in the default environment.
1531 ** `display-error' takes a frame
1533 The `display-error' / `scm_display_error' helper now takes a frame as an
1534 argument instead of a stack. Stacks are still supported in deprecated
1535 builds. Additionally, `display-error' will again source location
1536 information for the error.
1538 ** No more `(ice-9 debug)'
1540 This module had some debugging helpers that are no longer applicable to
1541 the current debugging model. Importing this module will produce a
1542 deprecation warning. Users should contact bug-guile for support.
1544 ** Remove obsolete debug-options
1546 Removed `breakpoints', `trace', `procnames', `indent', `frames',
1547 `maxdepth', and `debug' debug-options.
1549 ** `backtrace' debug option on by default
1551 Given that Guile 2.0 can always give you a backtrace, backtraces are now
1554 ** `turn-on-debugging' deprecated
1556 ** Remove obsolete print-options
1558 The `source' and `closure-hook' print options are obsolete, and have
1561 ** Remove obsolete read-options
1563 The "elisp-strings" and "elisp-vectors" read options were unused and
1564 obsolete, so they have been removed.
1566 ** Remove eval-options and trap-options
1568 Eval-options and trap-options are obsolete with the new VM and
1571 ** Remove (ice-9 debugger) and (ice-9 debugging)
1573 See "Traps" and "Interactive Debugging" in the manual, for information
1574 on their replacements.
1576 ** Remove the GDS Emacs integration
1578 See "Using Guile in Emacs" in the manual, for info on how we think you
1579 should use Guile with Emacs.
1581 ** Deprecated: `lazy-catch'
1583 `lazy-catch' was a form that captured the stack at the point of a
1584 `throw', but the dynamic state at the point of the `catch'. It was a bit
1585 crazy. Please change to use `catch', possibly with a throw-handler, or
1586 `with-throw-handler'.
1588 ** Deprecated: primitive properties
1590 The `primitive-make-property', `primitive-property-set!',
1591 `primitive-property-ref', and `primitive-property-del!' procedures were
1592 crufty and only used to implement object properties, which has a new,
1593 threadsafe implementation. Use object properties or weak hash tables
1596 ** Deprecated `@bind' syntax
1598 `@bind' was part of an older implementation of the Emacs Lisp language,
1599 and is no longer used.
1601 ** Miscellaneous other deprecations
1603 `cuserid' has been deprecated, as it only returns 8 bytes of a user's
1604 login. Use `(passwd:name (getpwuid (geteuid)))' instead.
1606 Additionally, the procedures `apply-to-args', `has-suffix?', `scheme-file-suffix'
1607 `get-option', `for-next-option', `display-usage-report',
1608 `transform-usage-lambda', `collect', and `set-batch-mode?!' have all
1611 ** Add support for unbound fluids
1613 See `make-unbound-fluid', `fluid-unset!', and `fluid-bound?' in the
1616 ** Add `variable-unset!'
1618 See "Variables" in the manual, for more details.
1620 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
1622 * Changes to the C interface
1624 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
1626 The semantics of `scm_gc_malloc ()' have been changed, in a
1627 backward-compatible way. A new allocation routine,
1628 `scm_gc_malloc_pointerless ()', was added.
1630 Libgc is a conservative GC, which we hope will make interaction with C
1631 code easier and less error-prone.
1633 ** New procedures: `scm_to_stringn', `scm_from_stringn'
1634 ** New procedures: scm_{to,from}_{utf8,latin1}_symbol{n,}
1635 ** New procedures: scm_{to,from}_{utf8,utf32,latin1}_string{n,}
1637 These new procedures convert to and from string representations in
1638 particular encodings.
1640 Users should continue to use locale encoding for user input, user
1641 output, or interacting with the C library.
1643 Use the Latin-1 functions for ASCII, and for literals in source code.
1645 Use UTF-8 functions for interaction with modern libraries which deal in
1646 UTF-8, and UTF-32 for interaction with utf32-using libraries.
1648 Otherwise, use scm_to_stringn or scm_from_stringn with a specific
1651 ** New type definitions for `scm_t_intptr' and friends.
1653 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
1654 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
1655 available to C. Have fun!
1657 ** The GH interface (deprecated in version 1.6, 2001) was removed.
1659 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
1661 This makes these internal functions technically not callable from
1664 ** Functions for handling `scm_option' now no longer require an argument
1665 indicating length of the `scm_t_option' array.
1667 ** Procedures-with-setters are now implemented using applicable structs
1669 From a user's perspective this doesn't mean very much. But if, for some
1670 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
1671 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
1674 ** Remove old evaluator closures
1676 There used to be ranges of typecodes allocated to interpreted data
1677 structures, but that it no longer the case, given that interpreted
1678 procedure are now just regular VM closures. As a result, there is a
1679 newly free tc3, and a number of removed macros. See the ChangeLog for
1682 ** Primitive procedures are now VM trampoline procedures
1684 It used to be that there were something like 12 different typecodes
1685 allocated to primitive procedures, each with its own calling convention.
1686 Now there is only one, the gsubr. This may affect user code if you were
1687 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
1688 solution is to switch to use scm_c_make_gsubr. This solution works well
1689 both with the old 1.8 and and with the current 1.9 branch.
1691 Guile's old evaluator used to have special cases for applying "gsubrs",
1692 primitive procedures with specified numbers of required, optional, and
1693 rest arguments. Now, however, Guile represents gsubrs as normal VM
1694 procedures, with appropriate bytecode to parse out the correct number of
1695 arguments, including optional and rest arguments, and then with a
1696 special bytecode to apply the gsubr.
1698 This allows primitive procedures to appear on the VM stack, allowing
1699 them to be accurately counted in profiles. Also they now have more
1700 debugging information attached to them -- their number of arguments, for
1701 example. In addition, the VM can completely inline the application
1702 mechanics, allowing for faster primitive calls.
1704 However there are some changes on the C level. There is no more
1705 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
1706 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
1707 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
1708 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
1709 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
1711 Perhaps more significantly, `scm_c_make_subr',
1712 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
1713 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
1714 and there are no more subr typecodes. Use the scm_c_make_gsubr family
1717 Normal users of gsubrs should not be affected, though, as the
1718 scm_c_make_gsubr family still is the correct way to create primitive
1721 ** Remove deprecated array C interfaces
1723 Removed the deprecated array functions `scm_i_arrayp',
1724 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
1725 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
1726 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
1727 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
1729 ** Remove unused snarf macros
1731 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
1732 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
1734 ** New functions: `scm_call_n', `scm_c_run_hookn'
1736 `scm_call_n' applies to apply a function to an array of arguments.
1737 `scm_c_run_hookn' runs a hook with an array of arguments.
1739 ** Some SMOB types changed to have static typecodes
1741 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
1742 they have statically allocated tc7 typecodes.
1744 ** Preparations for changing SMOB representation
1746 If things go right, we'll be changing the SMOB representation soon. To
1747 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
1748 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
1749 in the future. Code accessing SMOBs using SCM_CELL macros was never
1750 correct, but until now things still worked. Users should be aware of
1753 ** Changed invocation mechanics of applicable SMOBs
1755 Guile's old evaluator used to have special cases for applying SMOB
1756 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
1757 trampoline procedure for it, and use the normal mechanics to apply the
1758 trampoline. This simplifies procedure application in the normal,
1761 The upshot is that the mechanics used to apply a SMOB are different from
1762 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
1763 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
1764 deprecated. Just use the scm_call_0 family of procedures.
1766 ** Removed support shlibs for SRFIs 1, 4, 13, 14, and 60
1768 Though these SRFI support libraries did expose API, they encoded a
1769 strange version string into their library names. That version was never
1770 programmatically exported, so there was no way people could use the
1773 This was a fortunate oversight, as it allows us to remove the need for
1774 extra, needless shared libraries --- the C support code for SRFIs 4, 13,
1775 and 14 was already in core --- and allow us to incrementally return the
1776 SRFI implementation to Scheme.
1778 ** New C function: scm_module_public_interface
1780 This procedure corresponds to Scheme's `module-public-interface'.
1782 ** Undeprecate `scm_the_root_module ()'
1784 It's useful to be able to get the root module from C without doing a
1787 ** Inline vector allocation
1789 Instead of having vectors point out into the heap for their data, their
1790 data is now allocated inline to the vector object itself. The same is
1791 true for bytevectors, by default, though there is an indirection
1792 available which should allow for making a bytevector from an existing
1795 ** New struct constructors that don't involve making lists
1797 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
1798 constructors, respectively, for structs. You might find them useful.
1802 In Guile 1.8, there were debugging frames on the C stack. Now there is
1803 no more need to explicitly mark the stack in this way, because Guile has
1804 a VM stack that it knows how to walk, which simplifies the C API
1805 considerably. See the ChangeLog for details; the relevant interface is
1806 in libguile/stacks.h. The Scheme API has not been changed significantly.
1808 ** Removal of Guile's primitive object system.
1810 There were a number of pieces in `objects.[ch]' that tried to be a
1811 minimal object system, but were never documented, and were quickly
1812 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
1813 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
1814 from objects.h are no more. In the very unlikely case in which these
1815 were useful to you, we urge you to contact guile-devel.
1819 Actually the future is still in the state that it was, is, and ever
1820 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
1821 part of it. These files were experimental, never compiled, and would be
1822 better implemented in Scheme anyway. In the future, that is.
1824 ** Deprecate trampolines
1826 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
1827 so on. The point was to do some precomputation on the type of the
1828 procedure, then return a specialized "call" procedure. However this
1829 optimization wasn't actually an optimization, so it is now deprecated.
1830 Just use `scm_call_0', etc instead.
1832 ** Deprecated `scm_badargsp'
1834 This function is unused in Guile, but was part of its API.
1836 ** Better support for Lisp `nil'.
1838 The bit representation of `nil' has been tweaked so that it is now very
1839 efficient to check e.g. if a value is equal to Scheme's end-of-list or
1840 Lisp's nil. Additionally there are a heap of new, specific predicates
1841 like scm_is_null_or_nil.
1843 ** Better integration of Lisp `nil'.
1845 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
1846 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
1847 but when we start to integrate more with Emacs, it is possible that we
1848 break code that assumes that, for example, `(not x)' implies that `x' is
1849 `eq?' to `#f'. This is not a common assumption. Refactoring affected
1850 code to rely on properties instead of identities will improve code
1851 correctness. See "Nil" in the manual, for more details.
1853 ** Support for static allocation of strings, symbols, and subrs.
1855 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
1856 much of their associated data as static variables, reducing Guile's
1859 ** `scm_stat' has an additional argument, `exception_on_error'
1860 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
1862 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
1864 Previously they would use the `off_t' type, which is fragile since its
1865 definition depends on the application's value for `_FILE_OFFSET_BITS'.
1867 ** The `long_long' C type, deprecated in 1.8, has been removed
1869 ** Removed deprecated uniform array procedures: scm_make_uve,
1870 scm_array_prototype, scm_list_to_uniform_array,
1871 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
1872 scm_ra_set_contp, scm_aind, scm_raprin1
1874 These functions have been deprecated since early 2005.
1876 * Changes to the distribution
1878 ** Guile's license is now LGPLv3+
1880 In other words the GNU Lesser General Public License, version 3 or
1881 later (at the discretion of each person that chooses to redistribute
1886 Guile's build is visually quieter, due to the use of Automake 1.11's
1887 AM_SILENT_RULES. Build as `make V=1' to see all of the output.
1889 ** GOOPS documentation folded into Guile reference manual
1891 GOOPS, Guile's object system, used to be documented in separate manuals.
1892 This content is now included in Guile's manual directly.
1894 ** `guile-config' will be deprecated in favor of `pkg-config'
1896 `guile-config' has been rewritten to get its information from
1897 `pkg-config', so this should be a transparent change. Note however that
1898 guile.m4 has yet to be modified to call pkg-config instead of
1901 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
1903 Programs that use `pkg-config' to find Guile or one of its Autoconf
1904 macros should now require `guile-2.0' instead of `guile-1.8'.
1906 ** New installation directory: $(pkglibdir)/1.9/ccache
1908 If $(libdir) is /usr/lib, for example, Guile will install its .go files
1909 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
1911 ** Parallel installability fixes
1913 Guile now installs its header files to a effective-version-specific
1914 directory, and includes the effective version (e.g. 2.0) in the library
1915 name (e.g. libguile-2.0.so).
1917 This change should be transparent to users, who should detect Guile via
1918 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
1919 parallel installs for multiple versions of Guile development
1922 ** Dynamically loadable extensions may be placed in a Guile-specific path
1924 Before, Guile only searched the system library paths for extensions
1925 (e.g. /usr/lib), which meant that the names of Guile extensions had to
1926 be globally unique. Installing them to a Guile-specific extensions
1927 directory is cleaner. Use `pkg-config --variable=extensiondir
1928 guile-2.0' to get the location of the extensions directory.
1930 ** User Scheme code may be placed in a version-specific path
1932 Before, there was only one way to install user Scheme code to a
1933 version-specific Guile directory: install to Guile's own path,
1934 e.g. /usr/share/guile/2.0. The site directory,
1935 e.g. /usr/share/guile/site, was unversioned. This has been changed to
1936 add a version-specific site directory, e.g. /usr/share/guile/site/2.0,
1937 searched before the global site directory.
1939 ** New dependency: libgc
1941 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
1943 ** New dependency: GNU libunistring
1945 See http://www.gnu.org/software/libunistring/, for more information. Our
1946 Unicode support uses routines from libunistring.
1948 ** New dependency: libffi
1950 See http://sourceware.org/libffi/, for more information.
1954 Changes in 1.8.8 (since 1.8.7)
1958 ** Fix possible buffer overruns when parsing numbers
1959 ** Avoid clash with system setjmp/longjmp on IA64
1960 ** Fix `wrong type arg' exceptions with IPv6 addresses
1963 Changes in 1.8.7 (since 1.8.6)
1965 * New modules (see the manual for details)
1967 ** `(srfi srfi-98)', an interface to access environment variables
1971 ** Fix compilation with `--disable-deprecated'
1972 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
1973 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
1974 ** Fix build problem when scm_t_timespec is different from struct timespec
1975 ** Fix build when compiled with -Wundef -Werror
1976 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
1977 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
1978 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
1979 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
1980 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
1981 ** Fix reading of complex numbers where both parts are inexact decimals
1983 ** Allow @ macro to work with (ice-9 syncase)
1985 Previously, use of the @ macro in a module whose code is being
1986 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
1987 Now it works as you would expect (giving the value of the specified
1990 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
1993 Changes in 1.8.6 (since 1.8.5)
1995 * New features (see the manual for details)
1997 ** New convenience function `scm_c_symbol_length ()'
1999 ** Single stepping through code from Emacs
2001 When you use GDS to evaluate Scheme code from Emacs, you can now use
2002 `C-u' to indicate that you want to single step through that code. See
2003 `Evaluating Scheme Code' in the manual for more details.
2005 ** New "guile(1)" man page!
2007 * Changes to the distribution
2009 ** Automake's `AM_MAINTAINER_MODE' is no longer used
2011 Thus, the `--enable-maintainer-mode' configure option is no longer
2012 available: Guile is now always configured in "maintainer mode".
2014 ** `ChangeLog' files are no longer updated
2016 Instead, changes are detailed in the version control system's logs. See
2017 the top-level `ChangeLog' files for details.
2022 ** `symbol->string' now returns a read-only string, as per R5RS
2023 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
2024 ** `guile-config link' now prints `-L$libdir' before `-lguile'
2025 ** Fix memory corruption involving GOOPS' `class-redefinition'
2026 ** Fix possible deadlock in `mutex-lock'
2027 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
2028 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
2029 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
2030 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
2031 ** Fix misleading output from `(help rationalize)'
2032 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
2033 ** Fix `gcd' when called with a single, negative argument.
2034 ** Fix `Stack overflow' errors seen when building on some platforms
2035 ** Fix bug when `scm_with_guile ()' was called several times from the
2037 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
2038 dynamic environment of the call to `raise'
2039 ** Fix potential deadlock in `make-struct'
2040 ** Fix compilation problem with libltdl from Libtool 2.2.x
2041 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
2044 Changes in 1.8.5 (since 1.8.4)
2046 * Infrastructure changes
2048 ** Guile repository switched from CVS to Git
2050 The new repository can be accessed using
2051 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
2052 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
2054 ** Add support for `pkg-config'
2056 See "Autoconf Support" in the manual for details.
2058 * New modules (see the manual for details)
2062 * New features (see the manual for details)
2064 ** New `postfix' read option, for SRFI-88 keyword syntax
2065 ** Some I/O primitives have been inlined, which improves I/O performance
2066 ** New object-based traps infrastructure
2068 This is a GOOPS-based infrastructure that builds on Guile's low-level
2069 evaluator trap calls and facilitates the development of debugging
2070 features like single-stepping, breakpoints, tracing and profiling.
2071 See the `Traps' node of the manual for details.
2073 ** New support for working on Guile code from within Emacs
2075 Guile now incorporates the `GDS' library (previously distributed
2076 separately) for working on Guile code from within Emacs. See the
2077 `Using Guile In Emacs' node of the manual for details.
2081 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
2082 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
2084 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
2085 would trigger an unbound variable error for `match:andmap'.
2087 ** `(oop goops describe)' now properly provides the `describe' feature
2088 ** Fixed `args-fold' from `(srfi srfi-37)'
2090 Previously, parsing short option names of argument-less options would
2091 lead to a stack overflow.
2093 ** `(srfi srfi-35)' is now visible through `cond-expand'
2094 ** Fixed type-checking for the second argument of `eval'
2095 ** Fixed type-checking for SRFI-1 `partition'
2096 ** Fixed `struct-ref' and `struct-set!' on "light structs"
2097 ** Honor struct field access rights in GOOPS
2098 ** Changed the storage strategy of source properties, which fixes a deadlock
2099 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
2100 ** Fixed build issue for GNU/Linux on IA64
2101 ** Fixed build issues on NetBSD 1.6
2102 ** Fixed build issue on Solaris 2.10 x86_64
2103 ** Fixed build issue with DEC/Compaq/HP's compiler
2104 ** Fixed `scm_from_complex_double' build issue on FreeBSD
2105 ** Fixed `alloca' build issue on FreeBSD 6
2106 ** Removed use of non-portable makefile constructs
2107 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
2108 ** Make sure all tests honor `$TMPDIR'
2111 Changes in 1.8.4 (since 1.8.3)
2115 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
2116 ** Fixed a segmentation fault which occurred when displaying the
2117 backtrace of a stack with a promise object (made by `delay') in it.
2118 ** Make `accept' leave guile mode while blocking
2119 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
2120 ** Fixed a build problem on AIX (use of func_data identifier)
2121 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
2122 called with an associator proc that returns neither a pair nor #f.
2123 ** Secondary threads now always return a valid module for (current-module).
2124 ** Avoid MacOS build problems caused by incorrect combination of "64"
2125 system and library calls.
2126 ** `guile-snarf' now honors `$TMPDIR'
2127 ** `guile-config compile' now reports CPPFLAGS used at compile-time
2128 ** Fixed build with Sun Studio (Solaris 9)
2129 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
2130 uniform vectors on AIX.
2131 ** Fixed a deadlock that occurs upon GC with multiple threads.
2132 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
2133 ** Fixed autotool-derived build problems on AIX 6.1.
2134 ** Fixed NetBSD/alpha support
2135 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
2137 * New modules (see the manual for details)
2141 * Documentation fixes and improvements
2143 ** Removed premature breakpoint documentation
2145 The features described are not available in the series of 1.8.x
2146 releases, so the documentation was misleading and has been removed.
2148 ** More about Guile's default *random-state* variable
2150 ** GOOPS: more about how to use `next-method'
2152 * Changes to the distribution
2154 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
2156 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
2157 General Public License, and the few incorrect files have now been
2158 fixed to agree with the rest of the Guile distribution.
2160 ** Removed unnecessary extra copies of COPYING*
2162 The distribution now contains a single COPYING.LESSER at its top level.
2165 Changes in 1.8.3 (since 1.8.2)
2167 * New modules (see the manual for details)
2174 ** The `(ice-9 slib)' module now works as expected
2175 ** Expressions like "(set! 'x #t)" no longer yield a crash
2176 ** Warnings about duplicate bindings now go to stderr
2177 ** A memory leak in `make-socket-address' was fixed
2178 ** Alignment issues (e.g., on SPARC) in network routines were fixed
2179 ** A threading issue that showed up at least on NetBSD was fixed
2180 ** Build problems on Solaris and IRIX fixed
2182 * Implementation improvements
2184 ** The reader is now faster, which reduces startup time
2185 ** Procedures returned by `record-accessor' and `record-modifier' are faster
2188 Changes in 1.8.2 (since 1.8.1):
2190 * New procedures (see the manual for details)
2192 ** set-program-arguments
2195 * Incompatible changes
2197 ** The body of a top-level `define' no longer sees the binding being created
2199 In a top-level `define', the binding being created is no longer visible
2200 from the `define' body. This breaks code like
2201 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
2202 unbound in the body. However, such code was not R5RS-compliant anyway,
2207 ** Fractions were not `equal?' if stored in unreduced form.
2208 (A subtle problem, since printing a value reduced it, making it work.)
2209 ** srfi-60 `copy-bit' failed on 64-bit systems
2210 ** "guile --use-srfi" option at the REPL can replace core functions
2211 (Programs run with that option were ok, but in the interactive REPL
2212 the core bindings got priority, preventing SRFI replacements or
2214 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
2215 ** `kill' on mingw throws an error for a PID other than oneself
2216 ** Procedure names are attached to procedure-with-setters
2217 ** Array read syntax works with negative lower bound
2218 ** `array-in-bounds?' fix if an array has different lower bounds on each index
2219 ** `*' returns exact 0 for "(* inexact 0)"
2220 This follows what it always did for "(* 0 inexact)".
2221 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
2222 ** SRFI-19: `date->julian-day' did not account for timezone offset
2223 ** `ttyname' no longer crashes when passed a non-tty argument
2224 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
2225 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
2226 ** GOOPS: Fixed a bug in `method-more-specific?'
2227 ** Build problems on Solaris fixed
2228 ** Build problems on HP-UX IA64 fixed
2229 ** Build problems on MinGW fixed
2232 Changes in 1.8.1 (since 1.8.0):
2234 * LFS functions are now used to access 64-bit files on 32-bit systems.
2236 * New procedures (see the manual for details)
2238 ** primitive-_exit - [Scheme] the-root-module
2239 ** scm_primitive__exit - [C]
2240 ** make-completion-function - [Scheme] (ice-9 readline)
2241 ** scm_c_locale_stringn_to_number - [C]
2242 ** scm_srfi1_append_reverse [C]
2243 ** scm_srfi1_append_reverse_x [C]
2251 ** Build problems have been fixed on MacOS, SunOS, and QNX.
2253 ** `strftime' fix sign of %z timezone offset.
2255 ** A one-dimensional array can now be 'equal?' to a vector.
2257 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
2259 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
2261 ** `record-accessor' and `record-modifier' now have strict type checks.
2263 Record accessor and modifier procedures now throw an error if the
2264 record type of the record they're given is not the type expected.
2265 (Previously accessors returned #f and modifiers silently did nothing).
2267 ** It is now OK to use both autoload and use-modules on a given module.
2269 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
2271 Previously there was no checking on primatives like make-vector that
2272 accept "one or two" arguments. Now there is.
2274 ** The srfi-1 assoc function now calls its equality predicate properly.
2276 Previously srfi-1 assoc would call the equality predicate with the key
2277 last. According to the SRFI, the key should be first.
2279 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
2281 ** The array-set! procedure no longer segfaults when given a bit vector.
2283 ** Bugs in make-shared-array have been fixed.
2285 ** string<? and friends now follow char<? etc order on 8-bit chars.
2287 ** The format procedure now handles inf and nan values for ~f correctly.
2289 ** exact->inexact should no longer overflow when given certain large fractions.
2291 ** srfi-9 accessor and modifier procedures now have strict record type checks.
2293 This matches the srfi-9 specification.
2295 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
2297 Previously the (ice-9 ftw) procedures would ignore any file that had
2298 the same inode number as a file they had already seen, even if that
2299 file was on a different device.
2302 Changes in 1.8.0 (changes since the 1.6.x series):
2304 * Changes to the distribution
2306 ** Guile is now licensed with the GNU Lesser General Public License.
2308 ** The manual is now licensed with the GNU Free Documentation License.
2310 ** Guile now requires GNU MP (http://swox.com/gmp).
2312 Guile now uses the GNU MP library for arbitrary precision arithmetic.
2314 ** Guile now has separate private and public configuration headers.
2316 That is, things like HAVE_STRING_H no longer leak from Guile's
2319 ** Guile now provides and uses an "effective" version number.
2321 Guile now provides scm_effective_version and effective-version
2322 functions which return the "effective" version number. This is just
2323 the normal full version string without the final micro-version number,
2324 so the current effective-version is "1.8". The effective version
2325 should remain unchanged during a stable series, and should be used for
2326 items like the versioned share directory name
2327 i.e. /usr/share/guile/1.8.
2329 Providing an unchanging version number during a stable release for
2330 things like the versioned share directory can be particularly
2331 important for Guile "add-on" packages, since it provides a directory
2332 that they can install to that won't be changed out from under them
2333 with each micro release during a stable series.
2335 ** Thread implementation has changed.
2337 When you configure "--with-threads=null", you will get the usual
2338 threading API (call-with-new-thread, make-mutex, etc), but you can't
2339 actually create new threads. Also, "--with-threads=no" is now
2340 equivalent to "--with-threads=null". This means that the thread API
2341 is always present, although you might not be able to create new
2344 When you configure "--with-threads=pthreads" or "--with-threads=yes",
2345 you will get threads that are implemented with the portable POSIX
2346 threads. These threads can run concurrently (unlike the previous
2347 "coop" thread implementation), but need to cooperate for things like
2350 The default is "pthreads", unless your platform doesn't have pthreads,
2351 in which case "null" threads are used.
2353 See the manual for details, nodes "Initialization", "Multi-Threading",
2354 "Blocking", and others.
2356 ** There is the new notion of 'discouraged' features.
2358 This is a milder form of deprecation.
2360 Things that are discouraged should not be used in new code, but it is
2361 OK to leave them in old code for now. When a discouraged feature is
2362 used, no warning message is printed like there is for 'deprecated'
2363 features. Also, things that are merely discouraged are nevertheless
2364 implemented efficiently, while deprecated features can be very slow.
2366 You can omit discouraged features from libguile by configuring it with
2367 the '--disable-discouraged' option.
2369 ** Deprecation warnings can be controlled at run-time.
2371 (debug-enable 'warn-deprecated) switches them on and (debug-disable
2372 'warn-deprecated) switches them off.
2374 ** Support for SRFI 61, extended cond syntax for multiple values has
2377 This SRFI is always available.
2379 ** Support for require-extension, SRFI-55, has been added.
2381 The SRFI-55 special form `require-extension' has been added. It is
2382 available at startup, and provides a portable way to load Scheme
2383 extensions. SRFI-55 only requires support for one type of extension,
2384 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
2387 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
2389 The (srfi srfi-26) module is an implementation of SRFI-26 which
2390 provides the `cut' and `cute' syntax. These may be used to specialize
2391 parameters without currying.
2393 ** New module (srfi srfi-31)
2395 This is an implementation of SRFI-31 which provides a special form
2396 `rec' for recursive evaluation.
2398 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
2399 been merged with the core, making their functionality always
2402 The modules are still available, tho, and you could use them together
2403 with a renaming import, for example.
2405 ** Guile no longer includes its own version of libltdl.
2407 The official version is good enough now.
2409 ** The --enable-htmldoc option has been removed from 'configure'.
2411 Support for translating the documentation into HTML is now always
2412 provided. Use 'make html'.
2414 ** New module (ice-9 serialize):
2416 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
2417 don't trust the thread safety of most of your program, but where you
2418 have some section(s) of code which you consider can run in parallel to
2419 other sections. See ice-9/serialize.scm for more information.
2421 ** The configure option '--disable-arrays' has been removed.
2423 Support for arrays and uniform numeric arrays is now always included
2426 * Changes to the stand-alone interpreter
2428 ** New command line option `-L'.
2430 This option adds a directory to the front of the load path.
2432 ** New command line option `--no-debug'.
2434 Specifying `--no-debug' on the command line will keep the debugging
2435 evaluator turned off, even for interactive sessions.
2437 ** User-init file ~/.guile is now loaded with the debugging evaluator.
2439 Previously, the normal evaluator would have been used. Using the
2440 debugging evaluator gives better error messages.
2442 ** The '-e' option now 'read's its argument.
2444 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
2445 be used with '-e'. For example, you can now write a script like
2448 exec guile -e '(@ (demo) main)' -s "$0" "$@"
2451 (define-module (demo)
2455 (format #t "Demo: ~a~%" args))
2458 * Changes to Scheme functions and syntax
2460 ** Guardians have changed back to their original semantics
2462 Guardians now behave like described in the paper by Dybvig et al. In
2463 particular, they no longer make guarantees about the order in which
2464 they return objects, and they can no longer be greedy.
2466 They no longer drop cyclic data structures.
2468 The C function scm_make_guardian has been changed incompatibly and no
2469 longer takes the 'greedy_p' argument.
2471 ** New function hashx-remove!
2473 This function completes the set of 'hashx' functions.
2475 ** The concept of dynamic roots has been factored into continuation
2476 barriers and dynamic states.
2478 Each thread has a current dynamic state that carries the values of the
2479 fluids. You can create and copy dynamic states and use them as the
2480 second argument for 'eval'. See "Fluids and Dynamic States" in the
2483 To restrict the influence that captured continuations can have on the
2484 control flow, you can errect continuation barriers. See "Continuation
2485 Barriers" in the manual.
2487 The function call-with-dynamic-root now essentially temporarily
2488 installs a new dynamic state and errects a continuation barrier.
2490 ** The default load path no longer includes "." at the end.
2492 Automatically loading modules from the current directory should not
2493 happen by default. If you want to allow it in a more controlled
2494 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
2495 variable %load-path.
2497 ** The uniform vector and array support has been overhauled.
2499 It now complies with SRFI-4 and the weird prototype based uniform
2500 array creation has been deprecated. See the manual for more details.
2502 Some non-compatible changes have been made:
2503 - characters can no longer be stored into byte arrays.
2504 - strings and bit vectors are no longer considered to be uniform numeric
2506 - array-rank throws an error for non-arrays instead of returning zero.
2507 - array-ref does no longer accept non-arrays when no indices are given.
2509 There is the new notion of 'generalized vectors' and corresponding
2510 procedures like 'generalized-vector-ref'. Generalized vectors include
2511 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
2513 Arrays use generalized vectors as their storage, so that you still
2514 have arrays of characters, bits, etc. However, uniform-array-read!
2515 and uniform-array-write can no longer read/write strings and
2518 ** There is now support for copy-on-write substrings, mutation-sharing
2519 substrings and read-only strings.
2521 Three new procedures are related to this: substring/shared,
2522 substring/copy, and substring/read-only. See the manual for more
2525 ** Backtraces will now highlight the value that caused the error.
2527 By default, these values are enclosed in "{...}", such as in this
2536 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
2537 <unnamed port>:1:1: Wrong type (expecting pair): a
2538 ABORT: (wrong-type-arg)
2540 The prefix and suffix used for highlighting can be set via the two new
2541 printer options 'highlight-prefix' and 'highlight-suffix'. For
2542 example, putting this into ~/.guile will output the bad value in bold
2543 on an ANSI terminal:
2545 (print-set! highlight-prefix "\x1b[1m")
2546 (print-set! highlight-suffix "\x1b[22m")
2549 ** 'gettext' support for internationalization has been added.
2551 See the manual for details.
2553 ** New syntax '@' and '@@':
2555 You can now directly refer to variables exported from a module by
2558 (@ MODULE-NAME VARIABLE-NAME)
2560 For example (@ (ice-9 pretty-print) pretty-print) will directly access
2561 the pretty-print variable exported from the (ice-9 pretty-print)
2562 module. You don't need to 'use' that module first. You can also use
2563 '@' as a target of 'set!', as in (set! (@ mod var) val).
2565 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
2566 but it can also access variables that have not been exported. It is
2567 intended only for kluges and temporary fixes and for debugging, not
2570 ** Keyword syntax has been made more disciplined.
2572 Previously, the name of a keyword was read as a 'token' but printed as
2573 a symbol. Now, it is read as a general Scheme datum which must be a
2584 ERROR: In expression (a b c):
2588 ERROR: Unbound variable: foo
2593 ERROR: Wrong type (expecting symbol): 12
2597 ERROR: Wrong type (expecting symbol): (a b c)
2601 ** The printing of symbols that might look like keywords can be
2604 The new printer option 'quote-keywordish-symbols' controls how symbols
2605 are printed that have a colon as their first or last character. The
2606 default now is to only quote a symbol with #{...}# when the read
2607 option 'keywords' is not '#f'. Thus:
2609 guile> (define foo (string->symbol ":foo"))
2610 guile> (read-set! keywords #f)
2613 guile> (read-set! keywords 'prefix)
2616 guile> (print-set! quote-keywordish-symbols #f)
2620 ** 'while' now provides 'break' and 'continue'
2622 break and continue were previously bound in a while loop, but not
2623 documented, and continue didn't quite work properly. The undocumented
2624 parameter to break which gave a return value for the while has been
2627 ** 'call-with-current-continuation' is now also available under the name
2630 ** The module system now checks for duplicate bindings.
2632 The module system now can check for name conflicts among imported
2635 The behavior can be controlled by specifying one or more 'duplicates'
2636 handlers. For example, to make Guile return an error for every name
2639 (define-module (foo)
2644 The new default behavior of the module system when a name collision
2645 has been detected is to
2647 1. Give priority to bindings marked as a replacement.
2648 2. Issue a warning (different warning if overriding core binding).
2649 3. Give priority to the last encountered binding (this corresponds to
2652 If you want the old behavior back without replacements or warnings you
2655 (default-duplicate-binding-handler 'last)
2657 to your .guile init file.
2659 ** New define-module option: :replace
2661 :replace works as :export, but, in addition, marks the binding as a
2664 A typical example is `format' in (ice-9 format) which is a replacement
2665 for the core binding `format'.
2667 ** Adding prefixes to imported bindings in the module system
2669 There is now a new :use-module option :prefix. It can be used to add
2670 a prefix to all imported bindings.
2672 (define-module (foo)
2673 :use-module ((bar) :prefix bar:))
2675 will import all bindings exported from bar, but rename them by adding
2678 ** Conflicting generic functions can be automatically merged.
2680 When two imported bindings conflict and they are both generic
2681 functions, the two functions can now be merged automatically. This is
2682 activated with the 'duplicates' handler 'merge-generics'.
2684 ** New function: effective-version
2686 Returns the "effective" version number. This is just the normal full
2687 version string without the final micro-version number. See "Changes
2688 to the distribution" above.
2690 ** New threading functions: parallel, letpar, par-map, and friends
2692 These are convenient ways to run calculations in parallel in new
2693 threads. See "Parallel forms" in the manual for details.
2695 ** New function 'try-mutex'.
2697 This function will attempt to lock a mutex but will return immediately
2698 instead of blocking and indicate failure.
2700 ** Waiting on a condition variable can have a timeout.
2702 The function 'wait-condition-variable' now takes a third, optional
2703 argument that specifies the point in time where the waiting should be
2706 ** New function 'broadcast-condition-variable'.
2708 ** New functions 'all-threads' and 'current-thread'.
2710 ** Signals and system asyncs work better with threads.
2712 The function 'sigaction' now takes a fourth, optional, argument that
2713 specifies the thread that the handler should run in. When the
2714 argument is omitted, the handler will run in the thread that called
2717 Likewise, 'system-async-mark' takes a second, optional, argument that
2718 specifies the thread that the async should run in. When it is
2719 omitted, the async will run in the thread that called
2720 'system-async-mark'.
2722 C code can use the new functions scm_sigaction_for_thread and
2723 scm_system_async_mark_for_thread to pass the new thread argument.
2725 When a thread blocks on a mutex, a condition variable or is waiting
2726 for IO to be possible, it will still execute system asyncs. This can
2727 be used to interrupt such a thread by making it execute a 'throw', for
2730 ** The function 'system-async' is deprecated.
2732 You can now pass any zero-argument procedure to 'system-async-mark'.
2733 The function 'system-async' will just return its argument unchanged
2736 ** New functions 'call-with-blocked-asyncs' and
2737 'call-with-unblocked-asyncs'
2739 The expression (call-with-blocked-asyncs PROC) will call PROC and will
2740 block execution of system asyncs for the current thread by one level
2741 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
2742 procedure and will unblock the execution of system asyncs by one
2743 level for the current thread.
2745 Only system asyncs are affected by these functions.
2747 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
2749 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
2750 instead. Those functions are easier to use correctly and can be
2753 ** New function 'unsetenv'.
2755 ** New macro 'define-syntax-public'.
2757 It works like 'define-syntax' and also exports the defined macro (but
2760 ** There is support for Infinity and NaNs.
2762 Following PLT Scheme, Guile can now work with infinite numbers, and
2765 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
2766 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
2767 "+nan.0"). These numbers are inexact and have no exact counterpart.
2769 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
2770 sign of the dividend. The infinities are integers, and they answer #t
2771 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
2772 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
2783 ERROR: Numerical overflow
2785 Two new predicates 'inf?' and 'nan?' can be used to test for the
2788 ** Inexact zero can have a sign.
2790 Guile can now distinguish between plus and minus inexact zero, if your
2791 platform supports this, too. The two zeros are equal according to
2792 '=', but not according to 'eqv?'. For example
2803 ** Guile now has exact rationals.
2805 Guile can now represent fractions such as 1/3 exactly. Computing with
2806 them is also done exactly, of course:
2811 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
2812 for exact arguments.
2814 For example: (floor 2) now returns an exact 2 where in the past it
2815 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
2817 ** inexact->exact no longer returns only integers.
2819 Without exact rationals, the closest exact number was always an
2820 integer, but now inexact->exact returns the fraction that is exactly
2821 equal to a floating point number. For example:
2823 (inexact->exact 1.234)
2824 => 694680242521899/562949953421312
2826 When you want the old behavior, use 'round' explicitly:
2828 (inexact->exact (round 1.234))
2831 ** New function 'rationalize'.
2833 This function finds a simple fraction that is close to a given real
2834 number. For example (and compare with inexact->exact above):
2836 (rationalize (inexact->exact 1.234) 1/2000)
2839 Note that, as required by R5RS, rationalize returns only then an exact
2840 result when both its arguments are exact.
2842 ** 'odd?' and 'even?' work also for inexact integers.
2844 Previously, (odd? 1.0) would signal an error since only exact integers
2845 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
2846 returns #f and (odd? 1.5) signals an error.
2848 ** Guile now has uninterned symbols.
2850 The new function 'make-symbol' will return an uninterned symbol. This
2851 is a symbol that is unique and is guaranteed to remain unique.
2852 However, uninterned symbols can not yet be read back in.
2854 Use the new function 'symbol-interned?' to check whether a symbol is
2857 ** pretty-print has more options.
2859 The function pretty-print from the (ice-9 pretty-print) module can now
2860 also be invoked with keyword arguments that control things like
2861 maximum output width. See the manual for details.
2863 ** Variables have no longer a special behavior for `equal?'.
2865 Previously, comparing two variables with `equal?' would recursivly
2866 compare their values. This is no longer done. Variables are now only
2867 `equal?' if they are `eq?'.
2869 ** `(begin)' is now valid.
2871 You can now use an empty `begin' form. It will yield #<unspecified>
2872 when evaluated and simply be ignored in a definition context.
2874 ** Deprecated: procedure->macro
2876 Change your code to use 'define-macro' or r5rs macros. Also, be aware
2877 that macro expansion will not be done during evaluation, but prior to
2880 ** Soft ports now allow a `char-ready?' procedure
2882 The vector argument to `make-soft-port' can now have a length of
2883 either 5 or 6. (Previously the length had to be 5.) The optional 6th
2884 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
2885 that returns the number of characters that can be read immediately
2886 without the soft port blocking.
2888 ** Deprecated: undefine
2890 There is no replacement for undefine.
2892 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
2893 have been discouraged.
2895 They are relics from a time where a keyword like #:foo was used
2896 directly as a Tcl option "-foo" and thus keywords were internally
2897 stored as a symbol with a starting dash. We now store a symbol
2900 Use symbol->keyword and keyword->symbol instead.
2902 ** The `cheap' debug option is now obsolete
2904 Evaluator trap calls are now unconditionally "cheap" - in other words,
2905 they pass a debug object to the trap handler rather than a full
2906 continuation. The trap handler code can capture a full continuation
2907 by using `call-with-current-continuation' in the usual way, if it so
2910 The `cheap' option is retained for now so as not to break existing
2911 code which gets or sets it, but setting it now has no effect. It will
2912 be removed in the next major Guile release.
2914 ** Evaluator trap calls now support `tweaking'
2916 `Tweaking' means that the trap handler code can modify the Scheme
2917 expression that is about to be evaluated (in the case of an
2918 enter-frame trap) or the value that is being returned (in the case of
2919 an exit-frame trap). The trap handler code indicates that it wants to
2920 do this by returning a pair whose car is the symbol 'instead and whose
2921 cdr is the modified expression or return value.
2923 * Changes to the C interface
2925 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
2926 take a 'delete' function argument.
2928 This argument makes no sense since the delete function is used to
2929 remove a pair from an alist, and this must not be configurable.
2931 This is an incompatible change.
2933 ** The GH interface is now subject to the deprecation mechanism
2935 The GH interface has been deprecated for quite some time but now it is
2936 actually removed from Guile when it is configured with
2937 --disable-deprecated.
2939 See the manual "Transitioning away from GH" for more information.
2941 ** A new family of functions for converting between C values and
2942 Scheme values has been added.
2944 These functions follow a common naming scheme and are designed to be
2945 easier to use, thread-safe and more future-proof than the older
2948 - int scm_is_* (...)
2950 These are predicates that return a C boolean: 1 or 0. Instead of
2951 SCM_NFALSEP, you can now use scm_is_true, for example.
2953 - <type> scm_to_<type> (SCM val, ...)
2955 These are functions that convert a Scheme value into an appropriate
2956 C value. For example, you can use scm_to_int to safely convert from
2959 - SCM scm_from_<type> (<type> val, ...)
2961 These functions convert from a C type to a SCM value; for example,
2962 scm_from_int for ints.
2964 There is a huge number of these functions, for numbers, strings,
2965 symbols, vectors, etc. They are documented in the reference manual in
2966 the API section together with the types that they apply to.
2968 ** New functions for dealing with complex numbers in C have been added.
2970 The new functions are scm_c_make_rectangular, scm_c_make_polar,
2971 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
2972 They work like scm_make_rectangular etc but take or return doubles
2975 ** The function scm_make_complex has been discouraged.
2977 Use scm_c_make_rectangular instead.
2979 ** The INUM macros have been deprecated.
2981 A lot of code uses these macros to do general integer conversions,
2982 although the macros only work correctly with fixnums. Use the
2983 following alternatives.
2985 SCM_INUMP -> scm_is_integer or similar
2986 SCM_NINUMP -> !scm_is_integer or similar
2987 SCM_MAKINUM -> scm_from_int or similar
2988 SCM_INUM -> scm_to_int or similar
2990 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
2991 do the validating for you.
2993 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
2994 have been discouraged.
2996 Use the newer scm_to_<type> and scm_from_<type> functions instead for
2997 new code. The functions have been discouraged since they don't fit
3000 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
3002 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
3003 evaluates its argument twice. Use scm_is_true, etc. instead for new
3006 ** The macro SCM_EQ_P has been discouraged.
3008 Use scm_is_eq for new code, which fits better into the naming
3011 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
3014 Use the function scm_is_pair or scm_is_null instead.
3016 ** The functions scm_round and scm_truncate have been deprecated and
3017 are now available as scm_c_round and scm_c_truncate, respectively.
3019 These functions occupy the names that scm_round_number and
3020 scm_truncate_number should have.
3022 ** The functions scm_c_string2str, scm_c_substring2str, and
3023 scm_c_symbol2str have been deprecated.
3025 Use scm_to_locale_stringbuf or similar instead, maybe together with
3028 ** New functions scm_c_make_string, scm_c_string_length,
3029 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
3030 scm_c_substring_shared, scm_c_substring_copy.
3032 These are like scm_make_string, scm_length, etc. but are slightly
3033 easier to use from C.
3035 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
3036 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
3038 They export too many assumptions about the implementation of strings
3039 and symbols that are no longer true in the presence of
3040 mutation-sharing substrings and when Guile switches to some form of
3043 When working with strings, it is often best to use the normal string
3044 functions provided by Guile, such as scm_c_string_ref,
3045 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
3046 manual since many more such functions are now provided than
3049 When you want to convert a SCM string to a C string, use the
3050 scm_to_locale_string function or similar instead. For symbols, use
3051 scm_symbol_to_string and then work with that string. Because of the
3052 new string representation, scm_symbol_to_string does not need to copy
3053 and is thus quite efficient.
3055 ** Some string, symbol and keyword functions have been discouraged.
3057 They don't fit into the uniform naming scheme and are not explicit
3058 about the character encoding.
3060 Replace according to the following table:
3062 scm_allocate_string -> scm_c_make_string
3063 scm_take_str -> scm_take_locale_stringn
3064 scm_take0str -> scm_take_locale_string
3065 scm_mem2string -> scm_from_locale_stringn
3066 scm_str2string -> scm_from_locale_string
3067 scm_makfrom0str -> scm_from_locale_string
3068 scm_mem2symbol -> scm_from_locale_symboln
3069 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
3070 scm_str2symbol -> scm_from_locale_symbol
3072 SCM_SYMBOL_HASH -> scm_hashq
3073 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
3075 scm_c_make_keyword -> scm_from_locale_keyword
3077 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
3078 now also available to C code.
3080 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
3082 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
3083 the latter returns the true name of the keyword, not the 'dash name',
3084 as SCM_KEYWORDSYM used to do.
3086 ** A new way to access arrays in a thread-safe and efficient way has
3089 See the manual, node "Accessing Arrays From C".
3091 ** The old uniform vector and bitvector implementations have been
3092 unceremoniously removed.
3094 This implementation exposed the details of the tagging system of
3095 Guile. Use the new C API explained in the manual in node "Uniform
3096 Numeric Vectors" and "Bit Vectors", respectively.
3098 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
3099 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
3100 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
3101 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
3102 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
3103 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
3106 ** The macros dealing with vectors have been deprecated.
3108 Use the new functions scm_is_vector, scm_vector_elements,
3109 scm_vector_writable_elements, etc, or scm_is_simple_vector,
3110 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
3111 manual for more details.
3113 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
3114 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
3116 The following macros have been removed: SCM_VECTOR_BASE,
3117 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
3118 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
3120 ** Some C functions and macros related to arrays have been deprecated.
3122 Migrate according to the following table:
3124 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
3125 scm_make_ra -> scm_make_array
3126 scm_shap2ra -> scm_make_array
3127 scm_cvref -> scm_c_generalized_vector_ref
3128 scm_ra_set_contp -> do not use
3129 scm_aind -> scm_array_handle_pos
3130 scm_raprin1 -> scm_display or scm_write
3132 SCM_ARRAYP -> scm_is_array
3133 SCM_ARRAY_NDIM -> scm_c_array_rank
3134 SCM_ARRAY_DIMS -> scm_array_handle_dims
3135 SCM_ARRAY_CONTP -> do not use
3136 SCM_ARRAY_MEM -> do not use
3137 SCM_ARRAY_V -> scm_array_handle_elements or similar
3138 SCM_ARRAY_BASE -> do not use
3140 ** SCM_CELL_WORD_LOC has been deprecated.
3142 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
3143 to a SCM, as opposed to a pointer to a scm_t_bits.
3145 This was done to allow the correct use of pointers into the Scheme
3146 heap. Previously, the heap words were of type scm_t_bits and local
3147 variables and function arguments were of type SCM, making it
3148 non-standards-conformant to have a pointer that can point to both.
3150 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
3152 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
3153 second and third words of double smobs. Likewise for
3154 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
3156 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
3157 used to get and set the 16 exra bits in the zeroth word of a smob.
3159 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
3160 accesing the first immediate word of a smob as a SCM value, and there
3161 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
3162 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
3164 ** New way to deal with non-local exits and re-entries.
3166 There is a new set of functions that essentially do what
3167 scm_internal_dynamic_wind does, but in a way that is more convenient
3168 for C code in some situations. Here is a quick example of how to
3169 prevent a potential memory leak:
3176 scm_dynwind_begin (0);
3178 mem = scm_malloc (100);
3179 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
3181 /* MEM would leak if BAR throws an error.
3182 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
3189 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
3190 SCM_DYNWIND_END as well.
3194 For full documentation, see the node "Dynamic Wind" in the manual.
3196 ** New function scm_dynwind_free
3198 This function calls 'free' on a given pointer when a dynwind context
3199 is left. Thus the call to scm_dynwind_unwind_handler above could be
3200 replaced with simply scm_dynwind_free (mem).
3202 ** New functions scm_c_call_with_blocked_asyncs and
3203 scm_c_call_with_unblocked_asyncs
3205 Like scm_call_with_blocked_asyncs etc. but for C functions.
3207 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
3209 In addition to scm_c_call_with_blocked_asyncs you can now also use
3210 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
3211 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
3213 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
3214 SCM_REALLOW_INTS have been deprecated.
3216 They do no longer fulfill their original role of blocking signal
3217 delivery. Depending on what you want to achieve, replace a pair of
3218 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
3219 mutex, blocks asyncs, or both. See node "Critical Sections" in the
3222 ** The value 'scm_mask_ints' is no longer writable.
3224 Previously, you could set scm_mask_ints directly. This is no longer
3225 possible. Use scm_c_call_with_blocked_asyncs and
3226 scm_c_call_with_unblocked_asyncs instead.
3228 ** New way to temporarily set the current input, output or error ports
3230 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
3231 context' (see above). <foo> is one of "input", "output" or "error".
3233 ** New way to temporarily set fluids
3235 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
3236 above) to temporarily set the value of a fluid.
3238 ** New types scm_t_intmax and scm_t_uintmax.
3240 On platforms that have them, these types are identical to intmax_t and
3241 uintmax_t, respectively. On other platforms, they are identical to
3242 the largest integer types that Guile knows about.
3244 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
3246 You should not have used them.
3248 ** Many public #defines with generic names have been made private.
3250 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
3251 private or renamed with a more suitable public name.
3253 ** The macro SCM_TYP16S has been deprecated.
3255 This macro is not intended for public use.
3257 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
3259 Use scm_is_true (scm_inexact_p (...)) instead.
3261 ** The macro SCM_SLOPPY_REALP has been deprecated.
3263 Use scm_is_real instead.
3265 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
3267 Use scm_is_complex instead.
3269 ** Some preprocessor defines have been deprecated.
3271 These defines indicated whether a certain feature was present in Guile
3272 or not. Going forward, assume that the features are always present.
3274 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
3275 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
3277 The following macros have been removed completely: MEMOIZE_LOCALS,
3278 SCM_RECKLESS, SCM_CAUTIOUS.
3280 ** The preprocessor define STACK_DIRECTION has been deprecated.
3282 There should be no need to know about the stack direction for ordinary
3285 ** New function: scm_effective_version
3287 Returns the "effective" version number. This is just the normal full
3288 version string without the final micro-version number. See "Changes
3289 to the distribution" above.
3291 ** The function scm_call_with_new_thread has a new prototype.
3293 Instead of taking a list with the thunk and handler, these two
3294 arguments are now passed directly:
3296 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
3298 This is an incompatible change.
3300 ** New snarfer macro SCM_DEFINE_PUBLIC.
3302 This is like SCM_DEFINE, but also calls scm_c_export for the defined
3303 function in the init section.
3305 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
3307 ** Garbage collector rewrite.
3309 The garbage collector is cleaned up a lot, and now uses lazy
3310 sweeping. This is reflected in the output of (gc-stats); since cells
3311 are being freed when they are allocated, the cells-allocated field
3312 stays roughly constant.
3314 For malloc related triggers, the behavior is changed. It uses the same
3315 heuristic as the cell-triggered collections. It may be tuned with the
3316 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
3317 for minimum yield of malloc related triggers. The default is 40.
3318 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
3321 Debugging operations for the freelist have been deprecated, along with
3322 the C variables that control garbage collection. The environment
3323 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
3324 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
3326 For understanding the memory usage of a GUILE program, the routine
3327 gc-live-object-stats returns an alist containing the number of live
3328 objects for every type.
3331 ** The function scm_definedp has been renamed to scm_defined_p
3333 The name scm_definedp is deprecated.
3335 ** The struct scm_cell type has been renamed to scm_t_cell
3337 This is in accordance to Guile's naming scheme for types. Note that
3338 the name scm_cell is now used for a function that allocates and
3339 initializes a new cell (see below).
3341 ** New functions for memory management
3343 A new set of functions for memory management has been added since the
3344 old way (scm_must_malloc, scm_must_free, etc) was error prone and
3345 indeed, Guile itself contained some long standing bugs that could
3346 cause aborts in long running programs.
3348 The new functions are more symmetrical and do not need cooperation
3349 from smob free routines, among other improvements.
3351 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
3352 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
3353 scm_gc_free, scm_gc_register_collectable_memory, and
3354 scm_gc_unregister_collectable_memory. Refer to the manual for more
3355 details and for upgrading instructions.
3357 The old functions for memory management have been deprecated. They
3358 are: scm_must_malloc, scm_must_realloc, scm_must_free,
3359 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
3361 ** Declarations of exported features are marked with SCM_API.
3363 Every declaration of a feature that belongs to the exported Guile API
3364 has been marked by adding the macro "SCM_API" to the start of the
3365 declaration. This macro can expand into different things, the most
3366 common of which is just "extern" for Unix platforms. On Win32, it can
3367 be used to control which symbols are exported from a DLL.
3369 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
3370 will expand into "__declspec (dllimport) extern", which is needed for
3371 linking to the Guile DLL in Windows.
3373 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
3374 SCM_SRFI4_IMPORT, for the corresponding libraries.
3376 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
3378 Use the new functions scm_cell and scm_double_cell instead. The old
3379 macros had problems because with them allocation and initialization
3380 was separated and the GC could sometimes observe half initialized
3381 cells. Only careful coding by the user of SCM_NEWCELL and
3382 SCM_NEWCELL2 could make this safe and efficient.
3384 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
3386 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
3389 ** SRCBRKP has been deprecated.
3391 Use scm_c_source_property_breakpoint_p instead.
3393 ** Deprecated: scm_makmacro
3395 Change your code to use either scm_makmmacro or to define macros in
3396 Scheme, using 'define-macro'.
3398 ** New function scm_c_port_for_each.
3400 This function is like scm_port_for_each but takes a pointer to a C
3401 function as the callback instead of a SCM value.
3403 ** The names scm_internal_select, scm_thread_sleep, and
3404 scm_thread_usleep have been discouraged.
3406 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
3408 ** The GC can no longer be blocked.
3410 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
3411 The GC can now run (partially) concurrently with other code and thus
3412 blocking it is not well defined.
3414 ** Many definitions have been removed that were previously deprecated.
3416 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
3417 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
3418 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
3419 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
3420 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
3421 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
3422 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
3423 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
3424 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
3425 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
3426 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
3427 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
3428 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
3429 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
3430 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
3431 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
3432 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
3433 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
3434 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
3435 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
3436 scm_vector_set_length_x, scm_contregs, scm_debug_info,
3437 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
3438 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
3439 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
3440 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
3441 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
3442 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
3443 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
3444 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
3445 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
3446 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
3448 * Changes to bundled modules
3452 Using the (ice-9 debug) module no longer automatically switches Guile
3453 to use the debugging evaluator. If you want to switch to the
3454 debugging evaluator (which is needed for backtrace information if you
3455 hit an error), please add an explicit "(debug-enable 'debug)" to your
3456 code just after the code to use (ice-9 debug).
3459 Changes since Guile 1.4:
3461 * Changes to the distribution
3463 ** A top-level TODO file is included.
3465 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
3467 Guile now always uses three numbers to represent the version,
3468 i.e. "1.6.5". The first number, 1, is the major version number, the
3469 second number, 6, is the minor version number, and the third number,
3470 5, is the micro version number. Changes in major version number
3471 indicate major changes in Guile.
3473 Minor version numbers that are even denote stable releases, and odd
3474 minor version numbers denote development versions (which may be
3475 unstable). The micro version number indicates a minor sub-revision of
3476 a given MAJOR.MINOR release.
3478 In keeping with the new scheme, (minor-version) and scm_minor_version
3479 no longer return everything but the major version number. They now
3480 just return the minor version number. Two new functions
3481 (micro-version) and scm_micro_version have been added to report the
3482 micro version number.
3484 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
3486 ** New preprocessor definitions are available for checking versions.
3488 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
3489 SCM_MICRO_VERSION to the appropriate integer values.
3491 ** Guile now actively warns about deprecated features.
3493 The new configure option `--enable-deprecated=LEVEL' and the
3494 environment variable GUILE_WARN_DEPRECATED control this mechanism.
3495 See INSTALL and README for more information.
3497 ** Guile is much more likely to work on 64-bit architectures.
3499 Guile now compiles and passes "make check" with only two UNRESOLVED GC
3500 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
3501 for the use of a test machine, and thanks to Stefan Jahn for ia64
3504 ** New functions: setitimer and getitimer.
3506 These implement a fairly direct interface to the libc functions of the
3509 ** The #. reader extension is now disabled by default.
3511 For safety reasons, #. evaluation is disabled by default. To
3512 re-enable it, set the fluid read-eval? to #t. For example:
3514 (fluid-set! read-eval? #t)
3516 but make sure you realize the potential security risks involved. With
3517 read-eval? enabled, reading a data file from an untrusted source can
3520 ** New SRFI modules have been added:
3522 SRFI-0 `cond-expand' is now supported in Guile, without requiring
3525 (srfi srfi-1) is a library containing many useful pair- and list-processing
3528 (srfi srfi-2) exports and-let*.
3530 (srfi srfi-4) implements homogeneous numeric vector datatypes.
3532 (srfi srfi-6) is a dummy module for now, since guile already provides
3533 all of the srfi-6 procedures by default: open-input-string,
3534 open-output-string, get-output-string.
3536 (srfi srfi-8) exports receive.
3538 (srfi srfi-9) exports define-record-type.
3540 (srfi srfi-10) exports define-reader-ctor and implements the reader
3543 (srfi srfi-11) exports let-values and let*-values.
3545 (srfi srfi-13) implements the SRFI String Library.
3547 (srfi srfi-14) implements the SRFI Character-Set Library.
3549 (srfi srfi-17) implements setter and getter-with-setter and redefines
3550 some accessor procedures as procedures with getters. (such as car,
3551 cdr, vector-ref etc.)
3553 (srfi srfi-19) implements the SRFI Time/Date Library.
3555 ** New scripts / "executable modules"
3557 Subdirectory "scripts" contains Scheme modules that are packaged to
3558 also be executable as scripts. At this time, these scripts are available:
3567 See README there for more info.
3569 These scripts can be invoked from the shell with the new program
3570 "guile-tools", which keeps track of installation directory for you.
3573 $ guile-tools display-commentary srfi/*.scm
3575 guile-tools is copied to the standard $bindir on "make install".
3577 ** New module (ice-9 stack-catch):
3579 stack-catch is like catch, but saves the current state of the stack in
3580 the fluid the-last-stack. This fluid can be useful when using the
3581 debugger and when re-throwing an error.
3583 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
3585 This has been done to prevent problems on lesser operating systems
3586 that can't tolerate `*'s in file names. The exported macro continues
3587 to be named `and-let*', of course.
3589 On systems that support it, there is also a compatibility module named
3590 (ice-9 and-let*). It will go away in the next release.
3592 ** New modules (oop goops) etc.:
3595 (oop goops describe)
3597 (oop goops active-slot)
3598 (oop goops composite-slot)
3600 The Guile Object Oriented Programming System (GOOPS) has been
3601 integrated into Guile. For further information, consult the GOOPS
3602 manual and tutorial in the `doc' directory.
3604 ** New module (ice-9 rdelim).
3606 This exports the following procedures which were previously defined
3607 in the default environment:
3609 read-line read-line! read-delimited read-delimited! %read-delimited!
3610 %read-line write-line
3612 For backwards compatibility the definitions are still imported into the
3613 default environment in this version of Guile. However you should add:
3615 (use-modules (ice-9 rdelim))
3617 to any program which uses the definitions, since this may change in
3620 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
3621 can be used for similar functionality.
3623 ** New module (ice-9 rw)
3625 This is a subset of the (scsh rw) module from guile-scsh. Currently
3626 it defines two procedures:
3628 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
3630 Read characters from a port or file descriptor into a string STR.
3631 A port must have an underlying file descriptor -- a so-called
3632 fport. This procedure is scsh-compatible and can efficiently read
3635 *** New function: write-string/partial str [port_or_fdes [start [end]]]
3637 Write characters from a string STR to a port or file descriptor.
3638 A port must have an underlying file descriptor -- a so-called
3639 fport. This procedure is mostly compatible and can efficiently
3640 write large strings.
3642 ** New module (ice-9 match)
3644 This module includes Andrew K. Wright's pattern matcher. See
3645 ice-9/match.scm for brief description or
3647 http://www.star-lab.com/wright/code.html
3649 for complete documentation.
3651 ** New module (ice-9 buffered-input)
3653 This module provides procedures to construct an input port from an
3654 underlying source of input that reads and returns its input in chunks.
3655 The underlying input source is a Scheme procedure, specified by the
3656 caller, which the port invokes whenever it needs more input.
3658 This is useful when building an input port whose back end is Readline
3659 or a UI element such as the GtkEntry widget.
3663 The reference and tutorial documentation that was previously
3664 distributed separately, as `guile-doc', is now included in the core
3665 Guile distribution. The documentation consists of the following
3668 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
3671 - The Guile Reference Manual (guile.texi) contains (or is intended to
3672 contain) reference documentation on all aspects of Guile.
3674 - The GOOPS Manual (goops.texi) contains both tutorial-style and
3675 reference documentation for using GOOPS, Guile's Object Oriented
3678 - The Revised^5 Report on the Algorithmic Language Scheme
3681 See the README file in the `doc' directory for more details.
3683 ** There are a couple of examples in the examples/ directory now.
3685 * Changes to the stand-alone interpreter
3687 ** New command line option `--use-srfi'
3689 Using this option, SRFI modules can be loaded on startup and be
3690 available right from the beginning. This makes programming portable
3691 Scheme programs easier.
3693 The option `--use-srfi' expects a comma-separated list of numbers,
3694 each representing a SRFI number to be loaded into the interpreter
3695 before starting evaluating a script file or the REPL. Additionally,
3696 the feature identifier for the loaded SRFIs is recognized by
3697 `cond-expand' when using this option.
3700 $ guile --use-srfi=8,13
3701 guile> (receive (x z) (values 1 2) (+ 1 2))
3703 guile> (string-pad "bla" 20)
3706 ** Guile now always starts up in the `(guile-user)' module.
3708 Previously, scripts executed via the `-s' option would run in the
3709 `(guile)' module and the repl would run in the `(guile-user)' module.
3710 Now every user action takes place in the `(guile-user)' module by
3713 * Changes to Scheme functions and syntax
3715 ** Character classifiers work for non-ASCII characters.
3717 The predicates `char-alphabetic?', `char-numeric?',
3718 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
3719 no longer check whether their arguments are ASCII characters.
3720 Previously, a character would only be considered alphabetic when it
3721 was also ASCII, for example.
3723 ** Previously deprecated Scheme functions have been removed:
3725 tag - no replacement.
3726 fseek - replaced by seek.
3727 list* - replaced by cons*.
3729 ** It's now possible to create modules with controlled environments
3733 (use-modules (ice-9 safe))
3734 (define m (make-safe-module))
3735 ;;; m will now be a module containing only a safe subset of R5RS
3736 (eval '(+ 1 2) m) --> 3
3737 (eval 'load m) --> ERROR: Unbound variable: load
3739 ** Evaluation of "()", the empty list, is now an error.
3741 Previously, the expression "()" evaluated to the empty list. This has
3742 been changed to signal a "missing expression" error. The correct way
3743 to write the empty list as a literal constant is to use quote: "'()".
3745 ** New concept of `Guile Extensions'.
3747 A Guile Extension is just a ordinary shared library that can be linked
3748 at run-time. We found it advantageous to give this simple concept a
3749 dedicated name to distinguish the issues related to shared libraries
3750 from the issues related to the module system.
3752 *** New function: load-extension
3754 Executing (load-extension lib init) is mostly equivalent to
3756 (dynamic-call init (dynamic-link lib))
3758 except when scm_register_extension has been called previously.
3759 Whenever appropriate, you should use `load-extension' instead of
3760 dynamic-link and dynamic-call.
3762 *** New C function: scm_c_register_extension
3764 This function registers a initialization function for use by
3765 `load-extension'. Use it when you don't want specific extensions to
3766 be loaded as shared libraries (for example on platforms that don't
3767 support dynamic linking).
3769 ** Auto-loading of compiled-code modules is deprecated.
3771 Guile used to be able to automatically find and link a shared
3772 library to satisfy requests for a module. For example, the module
3773 `(foo bar)' could be implemented by placing a shared library named
3774 "foo/libbar.so" (or with a different extension) in a directory on the
3777 This has been found to be too tricky, and is no longer supported. The
3778 shared libraries are now called "extensions". You should now write a
3779 small Scheme file that calls `load-extension' to load the shared
3780 library and initialize it explicitly.
3782 The shared libraries themselves should be installed in the usual
3783 places for shared libraries, with names like "libguile-foo-bar".
3785 For example, place this into a file "foo/bar.scm"
3787 (define-module (foo bar))
3789 (load-extension "libguile-foo-bar" "foobar_init")
3791 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
3793 `eval' is now R5RS, that is it takes two arguments.
3794 The second argument is an environment specifier, i.e. either
3796 (scheme-report-environment 5)
3797 (null-environment 5)
3798 (interaction-environment)
3804 ** The module system has been made more disciplined.
3806 The function `eval' will save and restore the current module around
3807 the evaluation of the specified expression. While this expression is
3808 evaluated, `(current-module)' will now return the right module, which
3809 is the module specified as the second argument to `eval'.
3811 A consequence of this change is that `eval' is not particularly
3812 useful when you want allow the evaluated code to change what module is
3813 designated as the current module and have this change persist from one
3814 call to `eval' to the next. The read-eval-print-loop is an example
3815 where `eval' is now inadequate. To compensate, there is a new
3816 function `primitive-eval' that does not take a module specifier and
3817 that does not save/restore the current module. You should use this
3818 function together with `set-current-module', `current-module', etc
3819 when you want to have more control over the state that is carried from
3820 one eval to the next.
3822 Additionally, it has been made sure that forms that are evaluated at
3823 the top level are always evaluated with respect to the current module.
3824 Previously, subforms of top-level forms such as `begin', `case',
3825 etc. did not respect changes to the current module although these
3826 subforms are at the top-level as well.
3828 To prevent strange behavior, the forms `define-module',
3829 `use-modules', `use-syntax', and `export' have been restricted to only
3830 work on the top level. The forms `define-public' and
3831 `defmacro-public' only export the new binding on the top level. They
3832 behave just like `define' and `defmacro', respectively, when they are
3833 used in a lexical environment.
3835 Also, `export' will no longer silently re-export bindings imported
3836 from a used module. It will emit a `deprecation' warning and will
3837 cease to perform any re-export in the next version. If you actually
3838 want to re-export bindings, use the new `re-export' in place of
3839 `export'. The new `re-export' will not make copies of variables when
3840 rexporting them, as `export' did wrongly.
3842 ** Module system now allows selection and renaming of imported bindings
3844 Previously, when using `use-modules' or the `#:use-module' clause in
3845 the `define-module' form, all the bindings (association of symbols to
3846 values) for imported modules were added to the "current module" on an
3847 as-is basis. This has been changed to allow finer control through two
3848 new facilities: selection and renaming.
3850 You can now select which of the imported module's bindings are to be
3851 visible in the current module by using the `:select' clause. This
3852 clause also can be used to rename individual bindings. For example:
3854 ;; import all bindings no questions asked
3855 (use-modules (ice-9 common-list))
3857 ;; import four bindings, renaming two of them;
3858 ;; the current module sees: every some zonk-y zonk-n
3859 (use-modules ((ice-9 common-list)
3861 (remove-if . zonk-y)
3862 (remove-if-not . zonk-n))))
3864 You can also programmatically rename all selected bindings using the
3865 `:renamer' clause, which specifies a proc that takes a symbol and
3866 returns another symbol. Because it is common practice to use a prefix,
3867 we now provide the convenience procedure `symbol-prefix-proc'. For
3870 ;; import four bindings, renaming two of them specifically,
3871 ;; and all four w/ prefix "CL:";
3872 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
3873 (use-modules ((ice-9 common-list)
3875 (remove-if . zonk-y)
3876 (remove-if-not . zonk-n))
3877 :renamer (symbol-prefix-proc 'CL:)))
3879 ;; import four bindings, renaming two of them specifically,
3880 ;; and all four by upcasing.
3881 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
3882 (define (upcase-symbol sym)
3883 (string->symbol (string-upcase (symbol->string sym))))
3885 (use-modules ((ice-9 common-list)
3887 (remove-if . zonk-y)
3888 (remove-if-not . zonk-n))
3889 :renamer upcase-symbol))
3891 Note that programmatic renaming is done *after* individual renaming.
3892 Also, the above examples show `use-modules', but the same facilities are
3893 available for the `#:use-module' clause of `define-module'.
3895 See manual for more info.
3897 ** The semantics of guardians have changed.
3899 The changes are for the most part compatible. An important criterion
3900 was to keep the typical usage of guardians as simple as before, but to
3901 make the semantics safer and (as a result) more useful.
3903 *** All objects returned from guardians are now properly alive.
3905 It is now guaranteed that any object referenced by an object returned
3906 from a guardian is alive. It's now impossible for a guardian to
3907 return a "contained" object before its "containing" object.
3909 One incompatible (but probably not very important) change resulting
3910 from this is that it is no longer possible to guard objects that
3911 indirectly reference themselves (i.e. are parts of cycles). If you do
3912 so accidentally, you'll get a warning.
3914 *** There are now two types of guardians: greedy and sharing.
3916 If you call (make-guardian #t) or just (make-guardian), you'll get a
3917 greedy guardian, and for (make-guardian #f) a sharing guardian.
3919 Greedy guardians are the default because they are more "defensive".
3920 You can only greedily guard an object once. If you guard an object
3921 more than once, once in a greedy guardian and the rest of times in
3922 sharing guardians, then it is guaranteed that the object won't be
3923 returned from sharing guardians as long as it is greedily guarded
3926 Guardians returned by calls to `make-guardian' can now take one more
3927 optional parameter, which says whether to throw an error in case an
3928 attempt is made to greedily guard an object that is already greedily
3929 guarded. The default is true, i.e. throw an error. If the parameter
3930 is false, the guardian invocation returns #t if guarding was
3931 successful and #f if it wasn't.
3933 Also, since greedy guarding is, in effect, a side-effecting operation
3934 on objects, a new function is introduced: `destroy-guardian!'.
3935 Invoking this function on a guardian renders it unoperative and, if
3936 the guardian is greedy, clears the "greedily guarded" property of the
3937 objects that were guarded by it, thus undoing the side effect.
3939 Note that all this hair is hardly very important, since guardian
3940 objects are usually permanent.
3942 ** Continuations created by call-with-current-continuation now accept
3943 any number of arguments, as required by R5RS.
3945 ** New function `issue-deprecation-warning'
3947 This function is used to display the deprecation messages that are
3948 controlled by GUILE_WARN_DEPRECATION as explained in the README.
3951 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
3955 ;; `id' is deprecated. Use `identity' instead.
3960 ** New syntax `begin-deprecated'
3962 When deprecated features are included (as determined by the configure
3963 option --enable-deprecated), `begin-deprecated' is identical to
3964 `begin'. When deprecated features are excluded, it always evaluates
3965 to `#f', ignoring the body forms.
3967 ** New function `make-object-property'
3969 This function returns a new `procedure with setter' P that can be used
3970 to attach a property to objects. When calling P as
3974 where `obj' is any kind of object, it attaches `val' to `obj' in such
3975 a way that it can be retrieved by calling P as
3979 This function will replace procedure properties, symbol properties and
3980 source properties eventually.
3982 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
3984 Instead of #&optional, #&key, etc you should now use #:optional,
3985 #:key, etc. Since #:optional is a keyword, you can write it as just
3986 :optional when (read-set! keywords 'prefix) is active.
3988 The old reader syntax `#&' is still supported, but deprecated. It
3989 will be removed in the next release.
3991 ** New define-module option: pure
3993 Tells the module system not to include any bindings from the root
3998 (define-module (totally-empty-module)
4001 ** New define-module option: export NAME1 ...
4003 Export names NAME1 ...
4005 This option is required if you want to be able to export bindings from
4006 a module which doesn't import one of `define-public' or `export'.
4010 (define-module (foo)
4012 :use-module (ice-9 r5rs)
4015 ;;; Note that we're pure R5RS below this point!
4020 ** New function: object->string OBJ
4022 Return a Scheme string obtained by printing a given object.
4024 ** New function: port? X
4026 Returns a boolean indicating whether X is a port. Equivalent to
4027 `(or (input-port? X) (output-port? X))'.
4029 ** New function: file-port?
4031 Determines whether a given object is a port that is related to a file.
4033 ** New function: port-for-each proc
4035 Apply PROC to each port in the Guile port table in turn. The return
4036 value is unspecified. More specifically, PROC is applied exactly once
4037 to every port that exists in the system at the time PORT-FOR-EACH is
4038 invoked. Changes to the port table while PORT-FOR-EACH is running
4039 have no effect as far as PORT-FOR-EACH is concerned.
4041 ** New function: dup2 oldfd newfd
4043 A simple wrapper for the `dup2' system call. Copies the file
4044 descriptor OLDFD to descriptor number NEWFD, replacing the
4045 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
4046 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
4047 to move away ports which are using NEWFD. The return value is
4050 ** New function: close-fdes fd
4052 A simple wrapper for the `close' system call. Close file
4053 descriptor FD, which must be an integer. Unlike close (*note
4054 close: Ports and File Descriptors.), the file descriptor will be
4055 closed even if a port is using it. The return value is
4058 ** New function: crypt password salt
4060 Encrypts `password' using the standard unix password encryption
4063 ** New function: chroot path
4065 Change the root directory of the running process to `path'.
4067 ** New functions: getlogin, cuserid
4069 Return the login name or the user name of the current effective user
4072 ** New functions: getpriority which who, setpriority which who prio
4074 Get or set the priority of the running process.
4076 ** New function: getpass prompt
4078 Read a password from the terminal, first displaying `prompt' and
4081 ** New function: flock file operation
4083 Set/remove an advisory shared or exclusive lock on `file'.
4085 ** New functions: sethostname name, gethostname
4087 Set or get the hostname of the machine the current process is running
4090 ** New function: mkstemp! tmpl
4092 mkstemp creates a new unique file in the file system and returns a
4093 new buffered port open for reading and writing to the file. TMPL
4094 is a string specifying where the file should be created: it must
4095 end with `XXXXXX' and will be changed in place to return the name
4096 of the temporary file.
4098 ** New function: open-input-string string
4100 Return an input string port which delivers the characters from
4101 `string'. This procedure, together with `open-output-string' and
4102 `get-output-string' implements SRFI-6.
4104 ** New function: open-output-string
4106 Return an output string port which collects all data written to it.
4107 The data can then be retrieved by `get-output-string'.
4109 ** New function: get-output-string
4111 Return the contents of an output string port.
4113 ** New function: identity
4115 Return the argument.
4117 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
4118 are represented in Scheme as integers with normal host byte ordering.
4120 ** New function: inet-pton family address
4122 Convert a printable string network address into an integer. Note that
4123 unlike the C version of this function, the result is an integer with
4124 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4127 (inet-pton AF_INET "127.0.0.1") => 2130706433
4128 (inet-pton AF_INET6 "::1") => 1
4130 ** New function: inet-ntop family address
4132 Convert an integer network address into a printable string. Note that
4133 unlike the C version of this function, the input is an integer with
4134 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
4137 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
4138 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
4139 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
4143 Use `identity' instead.
4149 ** Deprecated: return-it
4153 ** Deprecated: string-character-length
4155 Use `string-length' instead.
4157 ** Deprecated: flags
4159 Use `logior' instead.
4161 ** Deprecated: close-all-ports-except.
4163 This was intended for closing ports in a child process after a fork,
4164 but it has the undesirable side effect of flushing buffers.
4165 port-for-each is more flexible.
4167 ** The (ice-9 popen) module now attempts to set up file descriptors in
4168 the child process from the current Scheme ports, instead of using the
4169 current values of file descriptors 0, 1, and 2 in the parent process.
4171 ** Removed function: builtin-weak-bindings
4173 There is no such concept as a weak binding any more.
4175 ** Removed constants: bignum-radix, scm-line-incrementors
4177 ** define-method: New syntax mandatory.
4179 The new method syntax is now mandatory:
4181 (define-method (NAME ARG-SPEC ...) BODY ...)
4182 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
4184 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
4185 REST-ARG ::= ARG-NAME
4187 If you have old code using the old syntax, import
4188 (oop goops old-define-method) before (oop goops) as in:
4190 (use-modules (oop goops old-define-method) (oop goops))
4192 ** Deprecated function: builtin-variable
4193 Removed function: builtin-bindings
4195 There is no longer a distinction between builtin or other variables.
4196 Use module system operations for all variables.
4198 ** Lazy-catch handlers are no longer allowed to return.
4200 That is, a call to `throw', `error', etc is now guaranteed to not
4203 ** Bugfixes for (ice-9 getopt-long)
4205 This module is now tested using test-suite/tests/getopt-long.test.
4206 The following bugs have been fixed:
4208 *** Parsing for options that are specified to have `optional' args now checks
4209 if the next element is an option instead of unconditionally taking it as the
4212 *** An error is now thrown for `--opt=val' when the option description
4213 does not specify `(value #t)' or `(value optional)'. This condition used to
4214 be accepted w/o error, contrary to the documentation.
4216 *** The error message for unrecognized options is now more informative.
4217 It used to be "not a record", an artifact of the implementation.
4219 *** The error message for `--opt' terminating the arg list (no value), when
4220 `(value #t)' is specified, is now more informative. It used to be "not enough
4223 *** "Clumped" single-char args now preserve trailing string, use it as arg.
4224 The expansion used to be like so:
4226 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
4228 Note that the "5d" is dropped. Now it is like so:
4230 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
4232 This enables single-char options to have adjoining arguments as long as their
4233 constituent characters are not potential single-char options.
4235 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
4237 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
4238 property `arglist', which can be retrieved by `arity'. The result is that
4239 `arity' can give more detailed information than before:
4243 guile> (use-modules (ice-9 optargs))
4244 guile> (define* (foo #:optional a b c) a)
4246 0 or more arguments in `lambda*:G0'.
4251 3 optional arguments: `a', `b' and `c'.
4252 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
4254 2 required arguments: `a' and `b', 2 keyword arguments: `c'
4255 and `d', other keywords allowed.
4256 guile> (define* (baz a b #:optional c #:rest r) a)
4258 2 required arguments: `a' and `b', 1 optional argument: `c',
4261 * Changes to the C interface
4263 ** Types have been renamed from scm_*_t to scm_t_*.
4265 This has been done for POSIX sake. It reserves identifiers ending
4266 with "_t". What a concept.
4268 The old names are still available with status `deprecated'.
4270 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
4272 ** Deprecated features have been removed.
4276 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
4277 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
4279 *** C Functions removed
4281 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
4282 scm_fseek - replaced by scm_seek.
4283 gc-thunk - replaced by after-gc-hook.
4284 gh_int2scmb - replaced by gh_bool2scm.
4285 scm_tc_dblr - replaced by scm_tc16_real.
4286 scm_tc_dblc - replaced by scm_tc16_complex.
4287 scm_list_star - replaced by scm_cons_star.
4289 ** Deprecated: scm_makfromstr
4291 Use scm_mem2string instead.
4293 ** Deprecated: scm_make_shared_substring
4295 Explicit shared substrings will disappear from Guile.
4297 Instead, "normal" strings will be implemented using sharing
4298 internally, combined with a copy-on-write strategy.
4300 ** Deprecated: scm_read_only_string_p
4302 The concept of read-only strings will disappear in next release of
4305 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
4307 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
4309 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
4311 Call a procedure with the indicated number of arguments. See "Fly
4312 Evaluation" in the manual.
4314 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
4316 Call a procedure with the indicated number of arguments and a list of
4317 further arguments. See "Fly Evaluation" in the manual.
4319 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
4321 Create a list of the given number of elements. See "List
4322 Constructors" in the manual.
4324 ** Renamed function: scm_listify has been replaced by scm_list_n.
4326 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
4327 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
4329 Use functions scm_list_N instead.
4331 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
4333 Used by an application to read arbitrary number of bytes from a port.
4334 Same semantics as libc read, except that scm_c_read only returns less
4335 than SIZE bytes if at end-of-file.
4337 Warning: Doesn't update port line and column counts!
4339 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
4341 Used by an application to write arbitrary number of bytes to an SCM
4342 port. Similar semantics as libc write. However, unlike libc
4343 write, scm_c_write writes the requested number of bytes and has no
4346 Warning: Doesn't update port line and column counts!
4348 ** New function: scm_init_guile ()
4350 In contrast to scm_boot_guile, scm_init_guile will return normally
4351 after initializing Guile. It is not available on all systems, tho.
4353 ** New functions: scm_str2symbol, scm_mem2symbol
4355 The function scm_str2symbol takes a const char* pointing to a zero-terminated
4356 field of characters and creates a scheme symbol object from that C string.
4357 The function scm_mem2symbol takes a const char* and a number of characters and
4358 creates a symbol from the characters in that memory area.
4360 ** New functions: scm_primitive_make_property
4361 scm_primitive_property_ref
4362 scm_primitive_property_set_x
4363 scm_primitive_property_del_x
4365 These functions implement a new way to deal with object properties.
4366 See libguile/properties.c for their documentation.
4368 ** New function: scm_done_free (long size)
4370 This function is the inverse of scm_done_malloc. Use it to report the
4371 amount of smob memory you free. The previous method, which involved
4372 calling scm_done_malloc with negative argument, was somewhat
4373 unintuitive (and is still available, of course).
4375 ** New function: scm_c_memq (SCM obj, SCM list)
4377 This function provides a fast C level alternative for scm_memq for the case
4378 that the list parameter is known to be a proper list. The function is a
4379 replacement for scm_sloppy_memq, but is stricter in its requirements on its
4380 list input parameter, since for anything else but a proper list the function's
4381 behaviour is undefined - it may even crash or loop endlessly. Further, for
4382 the case that the object is not found in the list, scm_c_memq returns #f which
4383 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
4385 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
4386 scm_remember_upto_here
4388 These functions replace the function scm_remember.
4390 ** Deprecated function: scm_remember
4392 Use one of the new functions scm_remember_upto_here_1,
4393 scm_remember_upto_here_2 or scm_remember_upto_here instead.
4395 ** New function: scm_allocate_string
4397 This function replaces the function scm_makstr.
4399 ** Deprecated function: scm_makstr
4401 Use the new function scm_allocate_string instead.
4403 ** New global variable scm_gc_running_p introduced.
4405 Use this variable to find out if garbage collection is being executed. Up to
4406 now applications have used scm_gc_heap_lock to test if garbage collection was
4407 running, which also works because of the fact that up to know only the garbage
4408 collector has set this variable. But, this is an implementation detail that
4409 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
4410 of this variable is (and has been) not fully safe anyway.
4412 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
4414 Use these instead of SCM_LENGTH_MAX.
4416 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
4417 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
4418 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
4420 Use these instead of SCM_LENGTH.
4422 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
4423 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
4424 SCM_SET_BITVECTOR_LENGTH
4426 Use these instead of SCM_SETLENGTH
4428 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
4429 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
4432 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
4435 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
4436 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
4439 Use these instead of SCM_SETCHARS.
4441 ** New macro: SCM_BITVECTOR_P
4443 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
4445 Use instead of SCM_COERCE_SUBSTR.
4447 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
4449 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
4451 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
4452 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
4453 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
4454 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
4455 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
4456 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
4457 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
4458 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
4459 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
4460 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
4461 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
4462 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
4463 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
4464 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
4465 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
4467 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
4468 Use scm_memory_error instead of SCM_NALLOC.
4469 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
4470 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
4471 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
4472 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
4473 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
4474 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
4475 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
4476 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
4477 Use SCM_STRINGP instead of SCM_RWSTRINGP.
4478 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
4479 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
4480 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
4481 Use a type specific setter macro instead of SCM_SETLENGTH.
4482 Use a type specific setter macro instead of SCM_SETCHARS.
4483 Use a type specific length macro instead of SCM_LENGTH_MAX.
4484 Use SCM_GCMARKP instead of SCM_GC8MARKP.
4485 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
4486 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
4487 Use SCM_TYP16 instead of SCM_GCTYP16.
4488 Use SCM_CDR instead of SCM_GCCDR.
4489 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
4490 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
4491 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
4492 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
4493 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
4494 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
4495 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
4497 ** Removed function: scm_struct_init
4499 ** Removed variable: scm_symhash_dim
4501 ** Renamed function: scm_make_cont has been replaced by
4502 scm_make_continuation, which has a different interface.
4504 ** Deprecated function: scm_call_catching_errors
4506 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
4508 ** Deprecated function: scm_strhash
4510 Use scm_string_hash instead.
4512 ** Deprecated function: scm_vector_set_length_x
4514 Instead, create a fresh vector of the desired size and copy the contents.
4516 ** scm_gensym has changed prototype
4518 scm_gensym now only takes one argument.
4520 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
4523 There is now only a single symbol type scm_tc7_symbol.
4524 The tag scm_tc7_lvector was not used anyway.
4526 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
4528 Use scm_make_smob_type and scm_set_smob_XXX instead.
4530 ** New function scm_set_smob_apply.
4532 This can be used to set an apply function to a smob type.
4534 ** Deprecated function: scm_strprint_obj
4536 Use scm_object_to_string instead.
4538 ** Deprecated function: scm_wta
4540 Use scm_wrong_type_arg, or another appropriate error signalling function
4543 ** Explicit support for obarrays has been deprecated.
4545 Use `scm_str2symbol' and the generic hashtable functions instead.
4547 ** The concept of `vcells' has been deprecated.
4549 The data type `variable' is now used exclusively. `Vcells' have been
4550 a low-level concept so you are likely not affected by this change.
4552 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
4553 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
4555 Use scm_c_define or scm_c_lookup instead, as appropriate.
4557 *** New functions: scm_c_module_lookup, scm_c_lookup,
4558 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
4559 scm_module_define, scm_define.
4561 These functions work with variables instead of with vcells.
4563 ** New functions for creating and defining `subr's and `gsubr's.
4565 The new functions more clearly distinguish between creating a subr (or
4566 gsubr) object and adding it to the current module.
4568 These new functions are available: scm_c_make_subr, scm_c_define_subr,
4569 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
4570 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
4571 scm_c_define_gsubr_with_generic.
4573 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
4574 scm_make_subr_with_generic, scm_make_gsubr,
4575 scm_make_gsubr_with_generic.
4577 Use the new ones from above instead.
4579 ** C interface to the module system has changed.
4581 While we suggest that you avoid as many explicit module system
4582 operations from C as possible for the time being, the C interface has
4583 been made more similar to the high-level Scheme module system.
4585 *** New functions: scm_c_define_module, scm_c_use_module,
4586 scm_c_export, scm_c_resolve_module.
4588 They mostly work like their Scheme namesakes. scm_c_define_module
4589 takes a function that is called a context where the new module is
4592 *** Deprecated functions: scm_the_root_module, scm_make_module,
4593 scm_ensure_user_module, scm_load_scheme_module.
4595 Use the new functions instead.
4597 ** Renamed function: scm_internal_with_fluids becomes
4600 scm_internal_with_fluids is available as a deprecated function.
4602 ** New function: scm_c_with_fluid.
4604 Just like scm_c_with_fluids, but takes one fluid and one value instead
4607 ** Deprecated typedefs: long_long, ulong_long.
4609 They are of questionable utility and they pollute the global
4612 ** Deprecated typedef: scm_sizet
4614 It is of questionable utility now that Guile requires ANSI C, and is
4617 ** Deprecated typedefs: scm_port_rw_active, scm_port,
4618 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
4619 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
4621 Made more compliant with the naming policy by adding a _t at the end.
4623 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
4624 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
4626 With the exception of the mysterious scm_2ulong2big, they are still
4627 available under new names (scm_i_mkbig etc). These functions are not
4628 intended to be used in user code. You should avoid dealing with
4629 bignums directly, and should deal with numbers in general (which can
4632 ** Change in behavior: scm_num2long, scm_num2ulong
4634 The scm_num2[u]long functions don't any longer accept an inexact
4635 argument. This change in behavior is motivated by concordance with
4636 R5RS: It is more common that a primitive doesn't want to accept an
4637 inexact for an exact.
4639 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
4640 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
4641 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
4644 These are conversion functions between the various ANSI C integral
4645 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
4646 accept an inexact argument.
4648 ** New functions: scm_float2num, scm_double2num,
4649 scm_num2float, scm_num2double.
4651 These are conversion functions between the two ANSI C float types and
4654 ** New number validation macros:
4655 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
4659 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
4661 These are just nicer-named old scm_protect_object and
4662 scm_unprotect_object.
4664 ** Deprecated functions: scm_protect_object, scm_unprotect_object
4666 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
4668 These functions can be used to register pointers to locations that
4671 ** Deprecated function: scm_create_hook.
4673 Its sins are: misleading name, non-modularity and lack of general
4677 Changes since Guile 1.3.4:
4679 * Changes to the distribution
4681 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
4683 We've changed the way we handle generated files in the Guile source
4684 repository. As a result, the procedure for building trees obtained
4685 from the nightly FTP snapshots or via CVS has changed:
4686 - You must have appropriate versions of autoconf, automake, and
4687 libtool installed on your system. See README for info on how to
4688 obtain these programs.
4689 - Before configuring the tree, you must first run the script
4690 `autogen.sh' at the top of the source tree.
4692 The Guile repository used to contain not only source files, written by
4693 humans, but also some generated files, like configure scripts and
4694 Makefile.in files. Even though the contents of these files could be
4695 derived mechanically from other files present, we thought it would
4696 make the tree easier to build if we checked them into CVS.
4698 However, this approach means that minor differences between
4699 developer's installed tools and habits affected the whole team.
4700 So we have removed the generated files from the repository, and
4701 added the autogen.sh script, which will reconstruct them
4705 ** configure now has experimental options to remove support for certain
4708 --disable-arrays omit array and uniform array support
4709 --disable-posix omit posix interfaces
4710 --disable-networking omit networking interfaces
4711 --disable-regex omit regular expression interfaces
4713 These are likely to become separate modules some day.
4715 ** New configure option --enable-debug-freelist
4717 This enables a debugging version of SCM_NEWCELL(), and also registers
4718 an extra primitive, the setter `gc-set-debug-check-freelist!'.
4720 Configure with the --enable-debug-freelist option to enable
4721 the gc-set-debug-check-freelist! primitive, and then use:
4723 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
4724 (gc-set-debug-check-freelist! #f) # turn off checking
4726 Checking of the freelist forces a traversal of the freelist and
4727 a garbage collection before each allocation of a cell. This can
4728 slow down the interpreter dramatically, so the setter should be used to
4729 turn on this extra processing only when necessary.
4731 ** New configure option --enable-debug-malloc
4733 Include code for debugging of calls to scm_must_malloc/realloc/free.
4737 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
4738 2. objects reallocated by scm_must_realloc has been allocated by
4740 3. reallocated objects are reallocated with the same what string
4742 But, most importantly, it records the number of allocated objects of
4743 each kind. This is useful when searching for memory leaks.
4745 A Guile compiled with this option provides the primitive
4746 `malloc-stats' which returns an alist with pairs of kind and the
4747 number of objects of that kind.
4749 ** All includes are now referenced relative to the root directory
4751 Since some users have had problems with mixups between Guile and
4752 system headers, we have decided to always refer to Guile headers via
4753 their parent directories. This essentially creates a "private name
4754 space" for Guile headers. This means that the compiler only is given
4755 -I options for the root build and root source directory.
4757 ** Header files kw.h and genio.h have been removed.
4759 ** The module (ice-9 getopt-gnu-style) has been removed.
4761 ** New module (ice-9 documentation)
4763 Implements the interface to documentation strings associated with
4766 ** New module (ice-9 time)
4768 Provides a macro `time', which displays execution time of a given form.
4770 ** New module (ice-9 history)
4772 Loading this module enables value history in the repl.
4774 * Changes to the stand-alone interpreter
4776 ** New command line option --debug
4778 Start Guile with debugging evaluator and backtraces enabled.
4780 This is useful when debugging your .guile init file or scripts.
4782 ** New help facility
4784 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
4785 (help REGEXP) ditto for objects with names matching REGEXP (a string)
4786 (help 'NAME) gives documentation for NAME, even if it is not an object
4787 (help ,EXPR) gives documentation for object returned by EXPR
4788 (help (my module)) gives module commentary for `(my module)'
4789 (help) gives this text
4791 `help' searches among bindings exported from loaded modules, while
4792 `apropos' searches among bindings visible from the "current" module.
4794 Examples: (help help)
4796 (help "output-string")
4798 ** `help' and `apropos' now prints full module names
4800 ** Dynamic linking now uses libltdl from the libtool package.
4802 The old system dependent code for doing dynamic linking has been
4803 replaced with calls to the libltdl functions which do all the hairy
4806 The major improvement is that you can now directly pass libtool
4807 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
4808 will be able to do the best shared library job you can get, via
4811 The way dynamic libraries are found has changed and is not really
4812 portable across platforms, probably. It is therefore recommended to
4813 use absolute filenames when possible.
4815 If you pass a filename without an extension to `dynamic-link', it will
4816 try a few appropriate ones. Thus, the most platform ignorant way is
4817 to specify a name like "libfoo", without any directories and
4820 ** Guile COOP threads are now compatible with LinuxThreads
4822 Previously, COOP threading wasn't possible in applications linked with
4823 Linux POSIX threads due to their use of the stack pointer to find the
4824 thread context. This has now been fixed with a workaround which uses
4825 the pthreads to allocate the stack.
4827 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
4829 ** Positions of erring expression in scripts
4831 With version 1.3.4, the location of the erring expression in Guile
4832 scipts is no longer automatically reported. (This should have been
4833 documented before the 1.3.4 release.)
4835 You can get this information by enabling recording of positions of
4836 source expressions and running the debugging evaluator. Put this at
4837 the top of your script (or in your "site" file):
4839 (read-enable 'positions)
4840 (debug-enable 'debug)
4842 ** Backtraces in scripts
4844 It is now possible to get backtraces in scripts.
4848 (debug-enable 'debug 'backtrace)
4850 at the top of the script.
4852 (The first options enables the debugging evaluator.
4853 The second enables backtraces.)
4855 ** Part of module system symbol lookup now implemented in C
4857 The eval closure of most modules is now implemented in C. Since this
4858 was one of the bottlenecks for loading speed, Guile now loads code
4859 substantially faster than before.
4861 ** Attempting to get the value of an unbound variable now produces
4862 an exception with a key of 'unbound-variable instead of 'misc-error.
4864 ** The initial default output port is now unbuffered if it's using a
4865 tty device. Previously in this situation it was line-buffered.
4867 ** New hook: after-gc-hook
4869 after-gc-hook takes over the role of gc-thunk. This hook is run at
4870 the first SCM_TICK after a GC. (Thus, the code is run at the same
4871 point during evaluation as signal handlers.)
4873 Note that this hook should be used only for diagnostic and debugging
4874 purposes. It is not certain that it will continue to be well-defined
4875 when this hook is run in the future.
4877 C programmers: Note the new C level hooks scm_before_gc_c_hook,
4878 scm_before_sweep_c_hook, scm_after_gc_c_hook.
4880 ** Improvements to garbage collector
4882 Guile 1.4 has a new policy for triggering heap allocation and
4883 determining the sizes of heap segments. It fixes a number of problems
4886 1. The new policy can handle two separate pools of cells
4887 (2-word/4-word) better. (The old policy would run wild, allocating
4888 more and more memory for certain programs.)
4890 2. The old code would sometimes allocate far too much heap so that the
4891 Guile process became gigantic. The new code avoids this.
4893 3. The old code would sometimes allocate too little so that few cells
4894 were freed at GC so that, in turn, too much time was spent in GC.
4896 4. The old code would often trigger heap allocation several times in a
4897 row. (The new scheme predicts how large the segments needs to be
4898 in order not to need further allocation.)
4900 All in all, the new GC policy will make larger applications more
4903 The new GC scheme also is prepared for POSIX threading. Threads can
4904 allocate private pools of cells ("clusters") with just a single
4905 function call. Allocation of single cells from such a cluster can
4906 then proceed without any need of inter-thread synchronization.
4908 ** New environment variables controlling GC parameters
4910 GUILE_MAX_SEGMENT_SIZE Maximal segment size
4913 Allocation of 2-word cell heaps:
4915 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
4918 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
4919 GC in percent of total heap size
4922 Allocation of 4-word cell heaps
4923 (used for real numbers and misc other objects):
4925 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
4927 (See entry "Way for application to customize GC parameters" under
4928 section "Changes to the scm_ interface" below.)
4930 ** Guile now implements reals using 4-word cells
4932 This speeds up computation with reals. (They were earlier allocated
4933 with `malloc'.) There is still some room for optimizations, however.
4935 ** Some further steps toward POSIX thread support have been taken
4937 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
4938 don't have much effect any longer, and many of them will be removed in
4942 are only handled at the top of the evaluator loop, immediately after
4943 I/O, and in scm_equalp.
4945 *** The GC can allocate thread private pools of pairs.
4947 * Changes to Scheme functions and syntax
4949 ** close-input-port and close-output-port are now R5RS
4951 These procedures have been turned into primitives and have R5RS behaviour.
4953 ** New procedure: simple-format PORT MESSAGE ARG1 ...
4955 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
4956 extended by the more sophisticated version in (ice-9 format)
4958 (simple-format port message . args)
4959 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
4960 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
4961 the escapes are replaced with corresponding members of ARGS:
4962 ~A formats using `display' and ~S formats using `write'.
4963 If DESTINATION is #t, then use the `current-output-port',
4964 if DESTINATION is #f, then return a string containing the formatted text.
4965 Does not add a trailing newline."
4967 ** string-ref: the second argument is no longer optional.
4969 ** string, list->string: no longer accept strings in their arguments,
4970 only characters, for compatibility with R5RS.
4972 ** New procedure: port-closed? PORT
4973 Returns #t if PORT is closed or #f if it is open.
4975 ** Deprecated: list*
4977 The list* functionality is now provided by cons* (SRFI-1 compliant)
4979 ** New procedure: cons* ARG1 ARG2 ... ARGn
4981 Like `list', but the last arg provides the tail of the constructed list,
4982 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
4984 Requires at least one argument. If given one argument, that argument
4985 is returned as result.
4987 This function is called `list*' in some other Schemes and in Common LISP.
4989 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
4991 ** New procedure: object-documentation OBJECT
4993 Returns the documentation string associated with OBJECT. The
4994 procedure uses a caching mechanism so that subsequent lookups are
4997 Exported by (ice-9 documentation).
4999 ** module-name now returns full names of modules
5001 Previously, only the last part of the name was returned (`session' for
5002 `(ice-9 session)'). Ex: `(ice-9 session)'.
5004 * Changes to the gh_ interface
5006 ** Deprecated: gh_int2scmb
5008 Use gh_bool2scm instead.
5010 * Changes to the scm_ interface
5012 ** Guile primitives now carry docstrings!
5014 Thanks to Greg Badros!
5016 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5018 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
5019 macros and must contain a docstring that is extracted into foo.doc using a new
5020 guile-doc-snarf script (that uses guile-doc-snarf.awk).
5022 However, a major overhaul of these macros is scheduled for the next release of
5025 ** Guile primitives use a new technique for validation of arguments
5027 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
5028 the readability of argument checking.
5030 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
5032 ** New macros: SCM_PACK, SCM_UNPACK
5034 Compose/decompose an SCM value.
5036 The SCM type is now treated as an abstract data type and may be defined as a
5037 long, a void* or as a struct, depending on the architecture and compile time
5038 options. This makes it easier to find several types of bugs, for example when
5039 SCM values are treated as integers without conversion. Values of the SCM type
5040 should be treated as "atomic" values. These macros are used when
5041 composing/decomposing an SCM value, either because you want to access
5042 individual bits, or because you want to treat it as an integer value.
5044 E.g., in order to set bit 7 in an SCM value x, use the expression
5046 SCM_PACK (SCM_UNPACK (x) | 0x80)
5048 ** The name property of hooks is deprecated.
5049 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
5051 You can emulate this feature by using object properties.
5053 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
5054 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
5057 These macros will be removed in a future release of Guile.
5059 ** The following types, functions and macros from numbers.h are deprecated:
5060 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
5061 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
5063 ** Port internals: the rw_random variable in the scm_port structure
5064 must be set to non-zero in any random access port. In recent Guile
5065 releases it was only set for bidirectional random-access ports.
5067 ** Port internals: the seek ptob procedure is now responsible for
5068 resetting the buffers if required. The change was made so that in the
5069 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
5070 the fport and strport ptobs can avoid resetting the buffers,
5071 in particular to avoid discarding unread chars. An existing port
5072 type can be fixed by adding something like the following to the
5073 beginning of the ptob seek procedure:
5075 if (pt->rw_active == SCM_PORT_READ)
5076 scm_end_input (object);
5077 else if (pt->rw_active == SCM_PORT_WRITE)
5078 ptob->flush (object);
5080 although to actually avoid resetting the buffers and discard unread
5081 chars requires further hacking that depends on the characteristics
5084 ** Deprecated functions: scm_fseek, scm_tag
5086 These functions are no longer used and will be removed in a future version.
5088 ** The scm_sysmissing procedure is no longer used in libguile.
5089 Unless it turns out to be unexpectedly useful to somebody, it will be
5090 removed in a future version.
5092 ** The format of error message strings has changed
5094 The two C procedures: scm_display_error and scm_error, as well as the
5095 primitive `scm-error', now use scm_simple_format to do their work.
5096 This means that the message strings of all code must be updated to use
5097 ~A where %s was used before, and ~S where %S was used before.
5099 During the period when there still are a lot of old Guiles out there,
5100 you might want to support both old and new versions of Guile.
5102 There are basically two methods to achieve this. Both methods use
5105 AC_CHECK_FUNCS(scm_simple_format)
5107 in your configure.in.
5109 Method 1: Use the string concatenation features of ANSI C's
5114 #ifdef HAVE_SCM_SIMPLE_FORMAT
5120 Then represent each of your error messages using a preprocessor macro:
5122 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
5126 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
5127 (define make-message string-append)
5129 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
5131 Method 2: Use the oldfmt function found in doc/oldfmt.c.
5135 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
5140 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
5144 ** Deprecated: coop_mutex_init, coop_condition_variable_init
5146 Don't use the functions coop_mutex_init and
5147 coop_condition_variable_init. They will change.
5149 Use scm_mutex_init and scm_cond_init instead.
5151 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
5152 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
5153 COND, as `scm_cond_wait' does, but it also bounds the duration
5154 of the wait. If COND has not been signaled before time ABSTIME,
5155 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
5156 returns the error code `ETIMEDOUT'.
5158 The ABSTIME parameter specifies an absolute time, with the same
5159 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
5160 to 00:00:00 GMT, January 1, 1970.
5162 ** New function: scm_cond_broadcast (scm_cond_t *COND)
5163 `scm_cond_broadcast' restarts all the threads that are waiting
5164 on the condition variable COND. Nothing happens if no threads are
5167 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
5168 `scm_key_create' allocates a new TSD key. The key is stored in
5169 the location pointed to by KEY. There is no limit on the number
5170 of keys allocated at a given time. The value initially associated
5171 with the returned key is `NULL' in all currently executing threads.
5173 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
5174 function associated with the key. When a thread terminates,
5175 DESTR_FUNCTION is called on the value associated with the key in
5176 that thread. The DESTR_FUNCTION is not called if a key is deleted
5177 with `scm_key_delete' or a value is changed with
5178 `scm_setspecific'. The order in which destructor functions are
5179 called at thread termination time is unspecified.
5181 Destructors are not yet implemented.
5183 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
5184 `scm_setspecific' changes the value associated with KEY in the
5185 calling thread, storing the given POINTER instead.
5187 ** New function: scm_getspecific (scm_key_t KEY)
5188 `scm_getspecific' returns the value currently associated with
5189 KEY in the calling thread.
5191 ** New function: scm_key_delete (scm_key_t KEY)
5192 `scm_key_delete' deallocates a TSD key. It does not check
5193 whether non-`NULL' values are associated with that key in the
5194 currently executing threads, nor call the destructor function
5195 associated with the key.
5197 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
5199 Initialize a C level hook HOOK with associated HOOK_DATA and type
5200 TYPE. (See scm_c_hook_run ().)
5202 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
5204 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
5205 is true, add it last, otherwise first. The same FUNC can be added
5206 multiple times if FUNC_DATA differ and vice versa.
5208 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
5210 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
5211 function is only removed if both FUNC and FUNC_DATA matches.
5213 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
5215 Run hook HOOK passing DATA to the hook functions.
5217 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
5218 returned is undefined.
5220 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
5221 returns a non-NULL value. This value is returned as the result of
5222 scm_c_hook_run. If all functions return NULL, NULL is returned.
5224 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
5225 returns a NULL value, and NULL is returned. If all functions returns
5226 a non-NULL value, the last value is returned.
5228 ** New C level GC hooks
5230 Five new C level hooks has been added to the garbage collector.
5232 scm_before_gc_c_hook
5235 are run before locking and after unlocking the heap. The system is
5236 thus in a mode where evaluation can take place. (Except that
5237 scm_before_gc_c_hook must not allocate new cells.)
5239 scm_before_mark_c_hook
5240 scm_before_sweep_c_hook
5241 scm_after_sweep_c_hook
5243 are run when the heap is locked. These are intended for extension of
5244 the GC in a modular fashion. Examples are the weaks and guardians
5247 ** Way for application to customize GC parameters
5249 The application can set up other default values for the GC heap
5250 allocation parameters
5252 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
5253 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
5254 GUILE_MAX_SEGMENT_SIZE,
5258 scm_default_init_heap_size_1, scm_default_min_yield_1,
5259 scm_default_init_heap_size_2, scm_default_min_yield_2,
5260 scm_default_max_segment_size
5262 respectively before callong scm_boot_guile.
5264 (See entry "New environment variables ..." in section
5265 "Changes to the stand-alone interpreter" above.)
5267 ** scm_protect_object/scm_unprotect_object now nest
5269 This means that you can call scm_protect_object multiple times on an
5270 object and count on the object being protected until
5271 scm_unprotect_object has been call the same number of times.
5273 The functions also have better time complexity.
5275 Still, it is usually possible to structure the application in a way
5276 that you don't need to use these functions. For example, if you use a
5277 protected standard Guile list to keep track of live objects rather
5278 than some custom data type, objects will die a natural death when they
5279 are no longer needed.
5281 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
5283 Guile does not provide the float representation for inexact real numbers any
5284 more. Now, only doubles are used to represent inexact real numbers. Further,
5285 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
5286 and scm_tc16_complex, respectively.
5288 ** Removed deprecated type scm_smobfuns
5290 ** Removed deprecated function scm_newsmob
5292 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
5294 There is an ongoing discussion among the developers whether to
5295 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
5296 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
5297 until this issue has been settled.
5299 ** Removed deprecated type tag scm_tc16_kw
5301 ** Added type tag scm_tc16_keyword
5303 (This was introduced already in release 1.3.4 but was not documented
5306 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
5308 * Changes to system call interfaces:
5310 ** The "select" procedure now tests port buffers for the ability to
5311 provide input or accept output. Previously only the underlying file
5312 descriptors were checked.
5314 ** New variable PIPE_BUF: the maximum number of bytes that can be
5315 atomically written to a pipe.
5317 ** If a facility is not available on the system when Guile is
5318 compiled, the corresponding primitive procedure will not be defined.
5319 Previously it would have been defined but would throw a system-error
5320 exception if called. Exception handlers which catch this case may
5321 need minor modification: an error will be thrown with key
5322 'unbound-variable instead of 'system-error. Alternatively it's
5323 now possible to use `defined?' to check whether the facility is
5326 ** Procedures which depend on the timezone should now give the correct
5327 result on systems which cache the TZ environment variable, even if TZ
5328 is changed without calling tzset.
5330 * Changes to the networking interfaces:
5332 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
5333 long integers between network and host format. For now, it's not
5334 particularly convenient to do this kind of thing, but consider:
5336 (define write-network-long
5337 (lambda (value port)
5338 (let ((v (make-uniform-vector 1 1 0)))
5339 (uniform-vector-set! v 0 (htonl value))
5340 (uniform-vector-write v port))))
5342 (define read-network-long
5344 (let ((v (make-uniform-vector 1 1 0)))
5345 (uniform-vector-read! v port)
5346 (ntohl (uniform-vector-ref v 0)))))
5348 ** If inet-aton fails, it now throws an error with key 'misc-error
5349 instead of 'system-error, since errno is not relevant.
5351 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
5352 specific keys instead of 'system-error. The latter is inappropriate
5353 since errno will not have been set. The keys are:
5354 'host-not-found, 'try-again, 'no-recovery and 'no-data.
5356 ** sethostent, setnetent, setprotoent, setservent: now take an
5357 optional argument STAYOPEN, which specifies whether the database
5358 remains open after a database entry is accessed randomly (e.g., using
5359 gethostbyname for the hosts database.) The default is #f. Previously
5363 Changes since Guile 1.3.2:
5365 * Changes to the stand-alone interpreter
5369 An initial version of the Guile debugger written by Chris Hanson has
5370 been added. The debugger is still under development but is included
5371 in the distribution anyway since it is already quite useful.
5377 after an error to enter the debugger. Type `help' inside the debugger
5378 for a description of available commands.
5380 If you prefer to have stack frames numbered and printed in
5381 anti-chronological order and prefer up in the stack to be down on the
5382 screen as is the case in gdb, you can put
5384 (debug-enable 'backwards)
5386 in your .guile startup file. (However, this means that Guile can't
5387 use indentation to indicate stack level.)
5389 The debugger is autoloaded into Guile at the first use.
5391 ** Further enhancements to backtraces
5393 There is a new debug option `width' which controls the maximum width
5394 on the screen of printed stack frames. Fancy printing parameters
5395 ("level" and "length" as in Common LISP) are adaptively adjusted for
5396 each stack frame to give maximum information while still fitting
5397 within the bounds. If the stack frame can't be made to fit by
5398 adjusting parameters, it is simply cut off at the end. This is marked
5401 ** Some modules are now only loaded when the repl is started
5403 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
5404 regex) are now loaded into (guile-user) only if the repl has been
5405 started. The effect is that the startup time for scripts has been
5406 reduced to 30% of what it was previously.
5408 Correctly written scripts load the modules they require at the top of
5409 the file and should not be affected by this change.
5411 ** Hooks are now represented as smobs
5413 * Changes to Scheme functions and syntax
5415 ** Readline support has changed again.
5417 The old (readline-activator) module is gone. Use (ice-9 readline)
5418 instead, which now contains all readline functionality. So the code
5419 to activate readline is now
5421 (use-modules (ice-9 readline))
5424 This should work at any time, including from the guile prompt.
5426 To avoid confusion about the terms of Guile's license, please only
5427 enable readline for your personal use; please don't make it the
5428 default for others. Here is why we make this rather odd-sounding
5431 Guile is normally licensed under a weakened form of the GNU General
5432 Public License, which allows you to link code with Guile without
5433 placing that code under the GPL. This exception is important to some
5436 However, since readline is distributed under the GNU General Public
5437 License, when you link Guile with readline, either statically or
5438 dynamically, you effectively change Guile's license to the strict GPL.
5439 Whenever you link any strictly GPL'd code into Guile, uses of Guile
5440 which are normally permitted become forbidden. This is a rather
5441 non-obvious consequence of the licensing terms.
5443 So, to make sure things remain clear, please let people choose for
5444 themselves whether to link GPL'd libraries like readline with Guile.
5446 ** regexp-substitute/global has changed slightly, but incompatibly.
5448 If you include a function in the item list, the string of the match
5449 object it receives is the same string passed to
5450 regexp-substitute/global, not some suffix of that string.
5451 Correspondingly, the match's positions are relative to the entire
5452 string, not the suffix.
5454 If the regexp can match the empty string, the way matches are chosen
5455 from the string has changed. regexp-substitute/global recognizes the
5456 same set of matches that list-matches does; see below.
5458 ** New function: list-matches REGEXP STRING [FLAGS]
5460 Return a list of match objects, one for every non-overlapping, maximal
5461 match of REGEXP in STRING. The matches appear in left-to-right order.
5462 list-matches only reports matches of the empty string if there are no
5463 other matches which begin on, end at, or include the empty match's
5466 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5468 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
5470 For each match of REGEXP in STRING, apply PROC to the match object,
5471 and the last value PROC returned, or INIT for the first call. Return
5472 the last value returned by PROC. We apply PROC to the matches as they
5473 appear from left to right.
5475 This function recognizes matches according to the same criteria as
5478 Thus, you could define list-matches like this:
5480 (define (list-matches regexp string . flags)
5481 (reverse! (apply fold-matches regexp string '() cons flags)))
5483 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5487 *** New function: hook? OBJ
5489 Return #t if OBJ is a hook, otherwise #f.
5491 *** New function: make-hook-with-name NAME [ARITY]
5493 Return a hook with name NAME and arity ARITY. The default value for
5494 ARITY is 0. The only effect of NAME is that it will appear when the
5495 hook object is printed to ease debugging.
5497 *** New function: hook-empty? HOOK
5499 Return #t if HOOK doesn't contain any procedures, otherwise #f.
5501 *** New function: hook->list HOOK
5503 Return a list of the procedures that are called when run-hook is
5506 ** `map' signals an error if its argument lists are not all the same length.
5508 This is the behavior required by R5RS, so this change is really a bug
5509 fix. But it seems to affect a lot of people's code, so we're
5510 mentioning it here anyway.
5512 ** Print-state handling has been made more transparent
5514 Under certain circumstances, ports are represented as a port with an
5515 associated print state. Earlier, this pair was represented as a pair
5516 (see "Some magic has been added to the printer" below). It is now
5517 indistinguishable (almost; see `get-print-state') from a port on the
5520 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
5522 Return a new port with the associated print state PRINT-STATE.
5524 *** New function: get-print-state OUTPUT-PORT
5526 Return the print state associated with this port if it exists,
5527 otherwise return #f.
5529 *** New function: directory-stream? OBJECT
5531 Returns true iff OBJECT is a directory stream --- the sort of object
5532 returned by `opendir'.
5534 ** New function: using-readline?
5536 Return #t if readline is in use in the current repl.
5538 ** structs will be removed in 1.4
5540 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
5541 and use GOOPS objects as the fundamental record type.
5543 * Changes to the scm_ interface
5545 ** structs will be removed in 1.4
5547 The entire current struct interface (struct.c, struct.h) will be
5548 replaced in Guile 1.4. We will merge GOOPS into libguile and use
5549 GOOPS objects as the fundamental record type.
5551 ** The internal representation of subr's has changed
5553 Instead of giving a hint to the subr name, the CAR field of the subr
5554 now contains an index to a subr entry in scm_subr_table.
5556 *** New variable: scm_subr_table
5558 An array of subr entries. A subr entry contains the name, properties
5559 and documentation associated with the subr. The properties and
5560 documentation slots are not yet used.
5562 ** A new scheme for "forwarding" calls to a builtin to a generic function
5564 It is now possible to extend the functionality of some Guile
5565 primitives by letting them defer a call to a GOOPS generic function on
5566 argument mismatch. This means that there is no loss of efficiency in
5571 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
5572 (define-method + ((x <string>) (y <string>))
5573 (string-append x y))
5575 + will still be as efficient as usual in numerical calculations, but
5576 can also be used for concatenating strings.
5578 Who will be the first one to extend Guile's numerical tower to
5579 rationals? :) [OK, there a few other things to fix before this can
5580 be made in a clean way.]
5582 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
5584 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
5586 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
5588 These do the same job as SCM_PROC and SCM_PROC1, but they also define
5589 a variable GENERIC which can be used by the dispatch macros below.
5591 [This is experimental code which may change soon.]
5593 *** New macros for forwarding control to a generic on arg type error
5595 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
5597 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
5599 These correspond to the scm_wta function call, and have the same
5600 behaviour until the user has called the GOOPS primitive
5601 `enable-primitive-generic!'. After that, these macros will apply the
5602 generic function GENERIC to the argument(s) instead of calling
5605 [This is experimental code which may change soon.]
5607 *** New macros for argument testing with generic dispatch
5609 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
5611 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
5613 These correspond to the SCM_ASSERT macro, but will defer control to
5614 GENERIC on error after `enable-primitive-generic!' has been called.
5616 [This is experimental code which may change soon.]
5618 ** New function: SCM scm_eval_body (SCM body, SCM env)
5620 Evaluates the body of a special form.
5622 ** The internal representation of struct's has changed
5624 Previously, four slots were allocated for the procedure(s) of entities
5625 and operators. The motivation for this representation had to do with
5626 the structure of the evaluator, the wish to support tail-recursive
5627 generic functions, and efficiency. Since the generic function
5628 dispatch mechanism has changed, there is no longer a need for such an
5629 expensive representation, and the representation has been simplified.
5631 This should not make any difference for most users.
5633 ** GOOPS support has been cleaned up.
5635 Some code has been moved from eval.c to objects.c and code in both of
5636 these compilation units has been cleaned up and better structured.
5638 *** New functions for applying generic functions
5640 New function: SCM scm_apply_generic (GENERIC, ARGS)
5641 New function: SCM scm_call_generic_0 (GENERIC)
5642 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
5643 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
5644 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
5646 ** Deprecated function: scm_make_named_hook
5648 It is now replaced by:
5650 ** New function: SCM scm_create_hook (const char *name, int arity)
5652 Creates a hook in the same way as make-hook above but also
5653 binds a variable named NAME to it.
5655 This is the typical way of creating a hook from C code.
5657 Currently, the variable is created in the "current" module.
5658 This might change when we get the new module system.
5660 [The behaviour is identical to scm_make_named_hook.]
5664 Changes since Guile 1.3:
5666 * Changes to mailing lists
5668 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
5670 See the README file to find current addresses for all the Guile
5673 * Changes to the distribution
5675 ** Readline support is no longer included with Guile by default.
5677 Based on the different license terms of Guile and Readline, we
5678 concluded that Guile should not *by default* cause the linking of
5679 Readline into an application program. Readline support is now offered
5680 as a separate module, which is linked into an application only when
5681 you explicitly specify it.
5683 Although Guile is GNU software, its distribution terms add a special
5684 exception to the usual GNU General Public License (GPL). Guile's
5685 license includes a clause that allows you to link Guile with non-free
5686 programs. We add this exception so as not to put Guile at a
5687 disadvantage vis-a-vis other extensibility packages that support other
5690 In contrast, the GNU Readline library is distributed under the GNU
5691 General Public License pure and simple. This means that you may not
5692 link Readline, even dynamically, into an application unless it is
5693 distributed under a free software license that is compatible the GPL.
5695 Because of this difference in distribution terms, an application that
5696 can use Guile may not be able to use Readline. Now users will be
5697 explicitly offered two independent decisions about the use of these
5700 You can activate the readline support by issuing
5702 (use-modules (readline-activator))
5705 from your ".guile" file, for example.
5707 * Changes to the stand-alone interpreter
5709 ** All builtins now print as primitives.
5710 Previously builtin procedures not belonging to the fundamental subr
5711 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
5712 Now, they print as #<primitive-procedure NAME>.
5714 ** Backtraces slightly more intelligible.
5715 gsubr-apply and macro transformer application frames no longer appear
5718 * Changes to Scheme functions and syntax
5720 ** Guile now correctly handles internal defines by rewriting them into
5721 their equivalent letrec. Previously, internal defines would
5722 incrementally add to the innermost environment, without checking
5723 whether the restrictions specified in RnRS were met. This lead to the
5724 correct behaviour when these restriction actually were met, but didn't
5725 catch all illegal uses. Such an illegal use could lead to crashes of
5726 the Guile interpreter or or other unwanted results. An example of
5727 incorrect internal defines that made Guile behave erratically:
5739 The problem with this example is that the definition of `c' uses the
5740 value of `b' directly. This confuses the meoization machine of Guile
5741 so that the second call of `b' (this time in a larger environment that
5742 also contains bindings for `c' and `d') refers to the binding of `c'
5743 instead of `a'. You could also make Guile crash with a variation on
5748 (define (b flag) (if flag a 1))
5749 (define c (1+ (b flag)))
5757 From now on, Guile will issue an `Unbound variable: b' error message
5762 A hook contains a list of functions which should be called on
5763 particular occasions in an existing program. Hooks are used for
5766 A window manager might have a hook before-window-map-hook. The window
5767 manager uses the function run-hooks to call all functions stored in
5768 before-window-map-hook each time a window is mapped. The user can
5769 store functions in the hook using add-hook!.
5771 In Guile, hooks are first class objects.
5773 *** New function: make-hook [N_ARGS]
5775 Return a hook for hook functions which can take N_ARGS arguments.
5776 The default value for N_ARGS is 0.
5778 (See also scm_make_named_hook below.)
5780 *** New function: add-hook! HOOK PROC [APPEND_P]
5782 Put PROC at the beginning of the list of functions stored in HOOK.
5783 If APPEND_P is supplied, and non-false, put PROC at the end instead.
5785 PROC must be able to take the number of arguments specified when the
5788 If PROC already exists in HOOK, then remove it first.
5790 *** New function: remove-hook! HOOK PROC
5792 Remove PROC from the list of functions in HOOK.
5794 *** New function: reset-hook! HOOK
5796 Clear the list of hook functions stored in HOOK.
5798 *** New function: run-hook HOOK ARG1 ...
5800 Run all hook functions stored in HOOK with arguments ARG1 ... .
5801 The number of arguments supplied must correspond to the number given
5802 when the hook was created.
5804 ** The function `dynamic-link' now takes optional keyword arguments.
5805 The only keyword argument that is currently defined is `:global
5806 BOOL'. With it, you can control whether the shared library will be
5807 linked in global mode or not. In global mode, the symbols from the
5808 linked library can be used to resolve references from other
5809 dynamically linked libraries. In non-global mode, the linked
5810 library is essentially invisible and can only be accessed via
5811 `dynamic-func', etc. The default is now to link in global mode.
5812 Previously, the default has been non-global mode.
5814 The `#:global' keyword is only effective on platforms that support
5815 the dlopen family of functions.
5817 ** New function `provided?'
5819 - Function: provided? FEATURE
5820 Return true iff FEATURE is supported by this installation of
5821 Guile. FEATURE must be a symbol naming a feature; the global
5822 variable `*features*' is a list of available features.
5824 ** Changes to the module (ice-9 expect):
5826 *** The expect-strings macro now matches `$' in a regular expression
5827 only at a line-break or end-of-file by default. Previously it would
5828 match the end of the string accumulated so far. The old behaviour
5829 can be obtained by setting the variable `expect-strings-exec-flags'
5832 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
5833 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
5834 in a regular expression will still match before a line-break or
5835 end-of-file. The default is `regexp/noteol'.
5837 *** The expect-strings macro now uses a variable
5838 `expect-strings-compile-flags' for the flags to be supplied to
5839 `make-regexp'. The default is `regexp/newline', which was previously
5842 *** The expect macro now supplies two arguments to a match procedure:
5843 the current accumulated string and a flag to indicate whether
5844 end-of-file has been reached. Previously only the string was supplied.
5845 If end-of-file is reached, the match procedure will be called an
5846 additional time with the same accumulated string as the previous call
5847 but with the flag set.
5849 ** New module (ice-9 format), implementing the Common Lisp `format' function.
5851 This code, and the documentation for it that appears here, was
5852 borrowed from SLIB, with minor adaptations for Guile.
5854 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
5855 An almost complete implementation of Common LISP format description
5856 according to the CL reference book `Common LISP' from Guy L.
5857 Steele, Digital Press. Backward compatible to most of the
5858 available Scheme format implementations.
5860 Returns `#t', `#f' or a string; has side effect of printing
5861 according to FORMAT-STRING. If DESTINATION is `#t', the output is
5862 to the current output port and `#t' is returned. If DESTINATION
5863 is `#f', a formatted string is returned as the result of the call.
5864 NEW: If DESTINATION is a string, DESTINATION is regarded as the
5865 format string; FORMAT-STRING is then the first argument and the
5866 output is returned as a string. If DESTINATION is a number, the
5867 output is to the current error port if available by the
5868 implementation. Otherwise DESTINATION must be an output port and
5871 FORMAT-STRING must be a string. In case of a formatting error
5872 format returns `#f' and prints a message on the current output or
5873 error port. Characters are output as if the string were output by
5874 the `display' function with the exception of those prefixed by a
5875 tilde (~). For a detailed description of the FORMAT-STRING syntax
5876 please consult a Common LISP format reference manual. For a test
5877 suite to verify this format implementation load `formatst.scm'.
5878 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
5880 Note: `format' is not reentrant, i.e. only one `format'-call may
5881 be executed at a time.
5884 *** Format Specification (Format version 3.0)
5886 Please consult a Common LISP format reference manual for a detailed
5887 description of the format string syntax. For a demonstration of the
5888 implemented directives see `formatst.scm'.
5890 This implementation supports directive parameters and modifiers (`:'
5891 and `@' characters). Multiple parameters must be separated by a comma
5892 (`,'). Parameters can be numerical parameters (positive or negative),
5893 character parameters (prefixed by a quote character (`''), variable
5894 parameters (`v'), number of rest arguments parameter (`#'), empty and
5895 default parameters. Directive characters are case independent. The
5896 general form of a directive is:
5898 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
5900 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
5902 *** Implemented CL Format Control Directives
5904 Documentation syntax: Uppercase characters represent the
5905 corresponding control directive characters. Lowercase characters
5906 represent control directive parameter descriptions.
5909 Any (print as `display' does).
5913 `~MINCOL,COLINC,MINPAD,PADCHARA'
5917 S-expression (print as `write' does).
5921 `~MINCOL,COLINC,MINPAD,PADCHARS'
5927 print number sign always.
5930 print comma separated.
5932 `~MINCOL,PADCHAR,COMMACHARD'
5938 print number sign always.
5941 print comma separated.
5943 `~MINCOL,PADCHAR,COMMACHARX'
5949 print number sign always.
5952 print comma separated.
5954 `~MINCOL,PADCHAR,COMMACHARO'
5960 print number sign always.
5963 print comma separated.
5965 `~MINCOL,PADCHAR,COMMACHARB'
5970 `~N,MINCOL,PADCHAR,COMMACHARR'
5974 print a number as a Roman numeral.
5977 print a number as an "old fashioned" Roman numeral.
5980 print a number as an ordinal English number.
5983 print a number as a cardinal English number.
5988 prints `y' and `ies'.
5991 as `~P but jumps 1 argument backward.'
5994 as `~@P but jumps 1 argument backward.'
5999 prints a character as the reader can understand it (i.e. `#\'
6003 prints a character as emacs does (eg. `^C' for ASCII 03).
6006 Fixed-format floating-point (prints a flonum like MMM.NNN).
6007 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
6009 If the number is positive a plus sign is printed.
6012 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
6013 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
6015 If the number is positive a plus sign is printed.
6018 General floating-point (prints a flonum either fixed or
6020 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
6022 If the number is positive a plus sign is printed.
6025 Dollars floating-point (prints a flonum in fixed with signs
6027 `~DIGITS,SCALE,WIDTH,PADCHAR$'
6029 If the number is positive a plus sign is printed.
6032 A sign is always printed and appears before the padding.
6035 The sign appears before the padding.
6043 print newline if not at the beginning of the output line.
6045 prints `~&' and then N-1 newlines.
6050 print N page separators.
6060 newline is ignored, white space left.
6063 newline is left, white space ignored.
6068 relative tabulation.
6074 Indirection (expects indirect arguments as a list).
6076 extracts indirect arguments from format arguments.
6079 Case conversion (converts by `string-downcase').
6081 converts by `string-capitalize'.
6084 converts by `string-capitalize-first'.
6087 converts by `string-upcase'.
6090 Argument Jumping (jumps 1 argument forward).
6092 jumps N arguments forward.
6095 jumps 1 argument backward.
6098 jumps N arguments backward.
6101 jumps to the 0th argument.
6104 jumps to the Nth argument (beginning from 0)
6106 `~[STR0~;STR1~;...~;STRN~]'
6107 Conditional Expression (numerical clause conditional).
6109 take argument from N.
6112 true test conditional.
6115 if-else-then conditional.
6121 default clause follows.
6124 Iteration (args come from the next argument (a list)).
6126 at most N iterations.
6129 args from next arg (a list of lists).
6132 args from the rest of arguments.
6135 args from the rest args (lists).
6146 aborts if N <= M <= K
6148 *** Not Implemented CL Format Control Directives
6151 print `#f' as an empty list (see below).
6154 print `#f' as an empty list (see below).
6160 (sorry I don't understand its semantics completely)
6162 *** Extended, Replaced and Additional Control Directives
6164 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
6165 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
6166 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
6167 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
6168 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
6169 COMMAWIDTH is the number of characters between two comma
6173 print a R4RS complex number as `~F~@Fi' with passed parameters for
6177 Pretty print formatting of an argument for scheme code lists.
6183 Flushes the output if format DESTINATION is a port.
6186 Print a `#\space' character
6188 print N `#\space' characters.
6191 Print a `#\tab' character
6193 print N `#\tab' characters.
6196 Takes N as an integer representation for a character. No arguments
6197 are consumed. N is converted to a character by `integer->char'. N
6198 must be a positive decimal number.
6201 Print out readproof. Prints out internal objects represented as
6202 `#<...>' as strings `"#<...>"' so that the format output can always
6203 be processed by `read'.
6206 Print out readproof. Prints out internal objects represented as
6207 `#<...>' as strings `"#<...>"' so that the format output can always
6208 be processed by `read'.
6211 Prints information and a copyright notice on the format
6214 prints format version.
6217 may also print number strings, i.e. passing a number as a string
6218 and format it accordingly.
6220 *** Configuration Variables
6222 The format module exports some configuration variables to suit the
6223 systems and users needs. There should be no modification necessary for
6224 the configuration that comes with Guile. Format detects automatically
6225 if the running scheme system implements floating point numbers and
6228 format:symbol-case-conv
6229 Symbols are converted by `symbol->string' so the case type of the
6230 printed symbols is implementation dependent.
6231 `format:symbol-case-conv' is a one arg closure which is either
6232 `#f' (no conversion), `string-upcase', `string-downcase' or
6233 `string-capitalize'. (default `#f')
6235 format:iobj-case-conv
6236 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
6237 implementation internal objects. (default `#f')
6240 The character prefixing the exponent value in `~E' printing.
6243 *** Compatibility With Other Format Implementations
6249 Downward compatible except for padding support and `~A', `~S',
6250 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
6251 `printf' padding support which is completely replaced by the CL
6252 `format' padding style.
6255 Downward compatible except for `~', which is not documented
6256 (ignores all characters inside the format string up to a newline
6257 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
6258 numerical and variable parameters and `:/@' modifiers in the CL
6262 Downward compatible except for `~A' and `~S' which print in
6263 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
6264 directive parameters or modifiers)).
6267 Downward compatible except for an optional destination parameter:
6268 S2C accepts a format call without a destination which returns a
6269 formatted string. This is equivalent to a #f destination in S2C.
6270 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
6271 parameters or modifiers)).
6274 ** Changes to string-handling functions.
6276 These functions were added to support the (ice-9 format) module, above.
6278 *** New function: string-upcase STRING
6279 *** New function: string-downcase STRING
6281 These are non-destructive versions of the existing string-upcase! and
6282 string-downcase! functions.
6284 *** New function: string-capitalize! STRING
6285 *** New function: string-capitalize STRING
6287 These functions convert the first letter of each word in the string to
6290 (string-capitalize "howdy there")
6293 As with the other functions, string-capitalize! modifies the string in
6294 place, while string-capitalize returns a modified copy of its argument.
6296 *** New function: string-ci->symbol STRING
6298 Return a symbol whose name is STRING, but having the same case as if
6299 the symbol had be read by `read'.
6301 Guile can be configured to be sensitive or insensitive to case
6302 differences in Scheme identifiers. If Guile is case-insensitive, all
6303 symbols are converted to lower case on input. The `string-ci->symbol'
6304 function returns a symbol whose name in STRING, transformed as Guile
6305 would if STRING were input.
6307 *** New function: substring-move! STRING1 START END STRING2 START
6309 Copy the substring of STRING1 from START (inclusive) to END
6310 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
6311 string, and the source and destination areas may overlap; in all
6312 cases, the function behaves as if all the characters were copied
6315 *** Extended functions: substring-move-left! substring-move-right!
6317 These functions now correctly copy arbitrarily overlapping substrings;
6318 they are both synonyms for substring-move!.
6321 ** New module (ice-9 getopt-long), with the function `getopt-long'.
6323 getopt-long is a function for parsing command-line arguments in a
6324 manner consistent with other GNU programs.
6326 (getopt-long ARGS GRAMMAR)
6327 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
6329 ARGS should be a list of strings. Its first element should be the
6330 name of the program; subsequent elements should be the arguments
6331 that were passed to the program on the command line. The
6332 `program-arguments' procedure returns a list of this form.
6334 GRAMMAR is a list of the form:
6335 ((OPTION (PROPERTY VALUE) ...) ...)
6337 Each OPTION should be a symbol. `getopt-long' will accept a
6338 command-line option named `--OPTION'.
6339 Each option can have the following (PROPERTY VALUE) pairs:
6341 (single-char CHAR) --- Accept `-CHAR' as a single-character
6342 equivalent to `--OPTION'. This is how to specify traditional
6344 (required? BOOL) --- If BOOL is true, the option is required.
6345 getopt-long will raise an error if it is not found in ARGS.
6346 (value BOOL) --- If BOOL is #t, the option accepts a value; if
6347 it is #f, it does not; and if it is the symbol
6348 `optional', the option may appear in ARGS with or
6350 (predicate FUNC) --- If the option accepts a value (i.e. you
6351 specified `(value #t)' for this option), then getopt
6352 will apply FUNC to the value, and throw an exception
6353 if it returns #f. FUNC should be a procedure which
6354 accepts a string and returns a boolean value; you may
6355 need to use quasiquotes to get it into GRAMMAR.
6357 The (PROPERTY VALUE) pairs may occur in any order, but each
6358 property may occur only once. By default, options do not have
6359 single-character equivalents, are not required, and do not take
6362 In ARGS, single-character options may be combined, in the usual
6363 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
6364 accepts values, then it must be the last option in the
6365 combination; the value is the next argument. So, for example, using
6366 the following grammar:
6367 ((apples (single-char #\a))
6368 (blimps (single-char #\b) (value #t))
6369 (catalexis (single-char #\c) (value #t)))
6370 the following argument lists would be acceptable:
6371 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
6372 for "blimps" and "catalexis")
6373 ("-ab" "bang" "-c" "couth") (same)
6374 ("-ac" "couth" "-b" "bang") (same)
6375 ("-abc" "couth" "bang") (an error, since `-b' is not the
6376 last option in its combination)
6378 If an option's value is optional, then `getopt-long' decides
6379 whether it has a value by looking at what follows it in ARGS. If
6380 the next element is a string, and it does not appear to be an
6381 option itself, then that string is the option's value.
6383 The value of a long option can appear as the next element in ARGS,
6384 or it can follow the option name, separated by an `=' character.
6385 Thus, using the same grammar as above, the following argument lists
6387 ("--apples" "Braeburn" "--blimps" "Goodyear")
6388 ("--apples=Braeburn" "--blimps" "Goodyear")
6389 ("--blimps" "Goodyear" "--apples=Braeburn")
6391 If the option "--" appears in ARGS, argument parsing stops there;
6392 subsequent arguments are returned as ordinary arguments, even if
6393 they resemble options. So, in the argument list:
6394 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
6395 `getopt-long' will recognize the `apples' option as having the
6396 value "Granny Smith", but it will not recognize the `blimp'
6397 option; it will return the strings "--blimp" and "Goodyear" as
6398 ordinary argument strings.
6400 The `getopt-long' function returns the parsed argument list as an
6401 assocation list, mapping option names --- the symbols from GRAMMAR
6402 --- onto their values, or #t if the option does not accept a value.
6403 Unused options do not appear in the alist.
6405 All arguments that are not the value of any option are returned
6406 as a list, associated with the empty list.
6408 `getopt-long' throws an exception if:
6409 - it finds an unrecognized option in ARGS
6410 - a required option is omitted
6411 - an option that requires an argument doesn't get one
6412 - an option that doesn't accept an argument does get one (this can
6413 only happen using the long option `--opt=value' syntax)
6414 - an option predicate fails
6419 `((lockfile-dir (required? #t)
6422 (predicate ,file-is-directory?))
6423 (verbose (required? #f)
6426 (x-includes (single-char #\x))
6427 (rnet-server (single-char #\y)
6428 (predicate ,string?))))
6430 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
6431 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
6433 => ((() "foo1" "-fred" "foo2" "foo3")
6434 (rnet-server . "lamprod")
6435 (x-includes . "/usr/include")
6436 (lockfile-dir . "/tmp")
6439 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
6441 It will be removed in a few releases.
6443 ** New syntax: lambda*
6444 ** New syntax: define*
6445 ** New syntax: define*-public
6446 ** New syntax: defmacro*
6447 ** New syntax: defmacro*-public
6448 Guile now supports optional arguments.
6450 `lambda*', `define*', `define*-public', `defmacro*' and
6451 `defmacro*-public' are identical to the non-* versions except that
6452 they use an extended type of parameter list that has the following BNF
6453 syntax (parentheses are literal, square brackets indicate grouping,
6454 and `*', `+' and `?' have the usual meaning):
6456 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
6457 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
6458 [[#&rest identifier]|[. identifier]]? ) | [identifier]
6460 ext-var-decl ::= identifier | ( identifier expression )
6462 The semantics are best illustrated with the following documentation
6463 and examples for `lambda*':
6466 lambda extended for optional and keyword arguments
6468 lambda* creates a procedure that takes optional arguments. These
6469 are specified by putting them inside brackets at the end of the
6470 paramater list, but before any dotted rest argument. For example,
6471 (lambda* (a b #&optional c d . e) '())
6472 creates a procedure with fixed arguments a and b, optional arguments c
6473 and d, and rest argument e. If the optional arguments are omitted
6474 in a call, the variables for them are unbound in the procedure. This
6475 can be checked with the bound? macro.
6477 lambda* can also take keyword arguments. For example, a procedure
6479 (lambda* (#&key xyzzy larch) '())
6480 can be called with any of the argument lists (#:xyzzy 11)
6481 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
6482 are given as keywords are bound to values.
6484 Optional and keyword arguments can also be given default values
6485 which they take on when they are not present in a call, by giving a
6486 two-item list in place of an optional argument, for example in:
6487 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
6488 foo is a fixed argument, bar is an optional argument with default
6489 value 42, and baz is a keyword argument with default value 73.
6490 Default value expressions are not evaluated unless they are needed
6491 and until the procedure is called.
6493 lambda* now supports two more special parameter list keywords.
6495 lambda*-defined procedures now throw an error by default if a
6496 keyword other than one of those specified is found in the actual
6497 passed arguments. However, specifying #&allow-other-keys
6498 immediately after the kyword argument declarations restores the
6499 previous behavior of ignoring unknown keywords. lambda* also now
6500 guarantees that if the same keyword is passed more than once, the
6501 last one passed is the one that takes effect. For example,
6502 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
6503 #:heads 37 #:tails 42 #:heads 99)
6504 would result in (99 47) being displayed.
6506 #&rest is also now provided as a synonym for the dotted syntax rest
6507 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
6508 all respects to lambda*. This is provided for more similarity to DSSSL,
6509 MIT-Scheme and Kawa among others, as well as for refugees from other
6512 Further documentation may be found in the optargs.scm file itself.
6514 The optional argument module also exports the macros `let-optional',
6515 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
6516 are not documented here because they may be removed in the future, but
6517 full documentation is still available in optargs.scm.
6519 ** New syntax: and-let*
6520 Guile now supports the `and-let*' form, described in the draft SRFI-2.
6522 Syntax: (land* (<clause> ...) <body> ...)
6523 Each <clause> should have one of the following forms:
6524 (<variable> <expression>)
6527 Each <variable> or <bound-variable> should be an identifier. Each
6528 <expression> should be a valid expression. The <body> should be a
6529 possibly empty sequence of expressions, like the <body> of a
6532 Semantics: A LAND* expression is evaluated by evaluating the
6533 <expression> or <bound-variable> of each of the <clause>s from
6534 left to right. The value of the first <expression> or
6535 <bound-variable> that evaluates to a false value is returned; the
6536 remaining <expression>s and <bound-variable>s are not evaluated.
6537 The <body> forms are evaluated iff all the <expression>s and
6538 <bound-variable>s evaluate to true values.
6540 The <expression>s and the <body> are evaluated in an environment
6541 binding each <variable> of the preceding (<variable> <expression>)
6542 clauses to the value of the <expression>. Later bindings
6543 shadow earlier bindings.
6545 Guile's and-let* macro was contributed by Michael Livshin.
6547 ** New sorting functions
6549 *** New function: sorted? SEQUENCE LESS?
6550 Returns `#t' when the sequence argument is in non-decreasing order
6551 according to LESS? (that is, there is no adjacent pair `... x y
6552 ...' for which `(less? y x)').
6554 Returns `#f' when the sequence contains at least one out-of-order
6555 pair. It is an error if the sequence is neither a list nor a
6558 *** New function: merge LIST1 LIST2 LESS?
6559 LIST1 and LIST2 are sorted lists.
6560 Returns the sorted list of all elements in LIST1 and LIST2.
6562 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
6563 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
6564 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
6565 (Here "<" should read "comes before".)
6567 *** New procedure: merge! LIST1 LIST2 LESS?
6568 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
6569 the result. If the code is compiled, and LESS? constructs no new
6570 pairs, no pairs at all will be allocated. The first pair of the
6571 result will be either the first pair of LIST1 or the first pair of
6574 *** New function: sort SEQUENCE LESS?
6575 Accepts either a list or a vector, and returns a new sequence
6576 which is sorted. The new sequence is the same type as the input.
6577 Always `(sorted? (sort sequence less?) less?)'. The original
6578 sequence is not altered in any way. The new sequence shares its
6579 elements with the old one; no elements are copied.
6581 *** New procedure: sort! SEQUENCE LESS
6582 Returns its sorted result in the original boxes. No new storage is
6583 allocated at all. Proper usage: (set! slist (sort! slist <))
6585 *** New function: stable-sort SEQUENCE LESS?
6586 Similar to `sort' but stable. That is, if "equal" elements are
6587 ordered a < b in the original sequence, they will have the same order
6590 *** New function: stable-sort! SEQUENCE LESS?
6591 Similar to `sort!' but stable.
6592 Uses temporary storage when sorting vectors.
6594 *** New functions: sort-list, sort-list!
6595 Added for compatibility with scsh.
6597 ** New built-in random number support
6599 *** New function: random N [STATE]
6600 Accepts a positive integer or real N and returns a number of the
6601 same type between zero (inclusive) and N (exclusive). The values
6602 returned have a uniform distribution.
6604 The optional argument STATE must be of the type produced by
6605 `copy-random-state' or `seed->random-state'. It defaults to the value
6606 of the variable `*random-state*'. This object is used to maintain the
6607 state of the pseudo-random-number generator and is altered as a side
6608 effect of the `random' operation.
6610 *** New variable: *random-state*
6611 Holds a data structure that encodes the internal state of the
6612 random-number generator that `random' uses by default. The nature
6613 of this data structure is implementation-dependent. It may be
6614 printed out and successfully read back in, but may or may not
6615 function correctly as a random-number state object in another
6618 *** New function: copy-random-state [STATE]
6619 Returns a new object of type suitable for use as the value of the
6620 variable `*random-state*' and as a second argument to `random'.
6621 If argument STATE is given, a copy of it is returned. Otherwise a
6622 copy of `*random-state*' is returned.
6624 *** New function: seed->random-state SEED
6625 Returns a new object of type suitable for use as the value of the
6626 variable `*random-state*' and as a second argument to `random'.
6627 SEED is a string or a number. A new state is generated and
6628 initialized using SEED.
6630 *** New function: random:uniform [STATE]
6631 Returns an uniformly distributed inexact real random number in the
6632 range between 0 and 1.
6634 *** New procedure: random:solid-sphere! VECT [STATE]
6635 Fills VECT with inexact real random numbers the sum of whose
6636 squares is less than 1.0. Thinking of VECT as coordinates in
6637 space of dimension N = `(vector-length VECT)', the coordinates are
6638 uniformly distributed within the unit N-shere. The sum of the
6639 squares of the numbers is returned. VECT can be either a vector
6640 or a uniform vector of doubles.
6642 *** New procedure: random:hollow-sphere! VECT [STATE]
6643 Fills VECT with inexact real random numbers the sum of whose squares
6644 is equal to 1.0. Thinking of VECT as coordinates in space of
6645 dimension n = `(vector-length VECT)', the coordinates are uniformly
6646 distributed over the surface of the unit n-shere. VECT can be either
6647 a vector or a uniform vector of doubles.
6649 *** New function: random:normal [STATE]
6650 Returns an inexact real in a normal distribution with mean 0 and
6651 standard deviation 1. For a normal distribution with mean M and
6652 standard deviation D use `(+ M (* D (random:normal)))'.
6654 *** New procedure: random:normal-vector! VECT [STATE]
6655 Fills VECT with inexact real random numbers which are independent and
6656 standard normally distributed (i.e., with mean 0 and variance 1).
6657 VECT can be either a vector or a uniform vector of doubles.
6659 *** New function: random:exp STATE
6660 Returns an inexact real in an exponential distribution with mean 1.
6661 For an exponential distribution with mean U use (* U (random:exp)).
6663 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
6665 These functions now operate on numbers in the range of a C unsigned
6668 These functions used to operate on numbers in the range of a C signed
6669 long; however, this seems inappropriate, because Guile integers don't
6672 ** New function: make-guardian
6673 This is an implementation of guardians as described in
6674 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
6675 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
6676 Programming Language Design and Implementation, June 1993
6677 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
6679 ** New functions: delq1!, delv1!, delete1!
6680 These procedures behave similar to delq! and friends but delete only
6681 one object if at all.
6683 ** New function: unread-string STRING PORT
6684 Unread STRING to PORT, that is, push it back onto the port so that
6685 next read operation will work on the pushed back characters.
6687 ** unread-char can now be called multiple times
6688 If unread-char is called multiple times, the unread characters will be
6689 read again in last-in first-out order.
6691 ** the procedures uniform-array-read! and uniform-array-write! now
6692 work on any kind of port, not just ports which are open on a file.
6694 ** Now 'l' in a port mode requests line buffering.
6696 ** The procedure truncate-file now works on string ports as well
6697 as file ports. If the size argument is omitted, the current
6698 file position is used.
6700 ** new procedure: seek PORT/FDES OFFSET WHENCE
6701 The arguments are the same as for the old fseek procedure, but it
6702 works on string ports as well as random-access file ports.
6704 ** the fseek procedure now works on string ports, since it has been
6705 redefined using seek.
6707 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
6708 size is not supplied.
6710 ** the newline procedure no longer flushes the port if it's not
6711 line-buffered: previously it did if it was the current output port.
6713 ** open-pipe and close-pipe are no longer primitive procedures, but
6714 an emulation can be obtained using `(use-modules (ice-9 popen))'.
6716 ** the freopen procedure has been removed.
6718 ** new procedure: drain-input PORT
6719 Drains PORT's read buffers (including any pushed-back characters)
6720 and returns the contents as a single string.
6722 ** New function: map-in-order PROC LIST1 LIST2 ...
6723 Version of `map' which guarantees that the procedure is applied to the
6724 lists in serial order.
6726 ** Renamed `serial-array-copy!' and `serial-array-map!' to
6727 `array-copy-in-order!' and `array-map-in-order!'. The old names are
6728 now obsolete and will go away in release 1.5.
6730 ** New syntax: collect BODY1 ...
6731 Version of `begin' which returns a list of the results of the body
6732 forms instead of the result of the last body form. In contrast to
6733 `begin', `collect' allows an empty body.
6735 ** New functions: read-history FILENAME, write-history FILENAME
6736 Read/write command line history from/to file. Returns #t on success
6737 and #f if an error occured.
6739 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
6741 These procedures return a list of definitions available in the specified
6742 argument, a relative module reference. In the case of no argument,
6743 `(current-module)' is now consulted for definitions to return, instead
6744 of simply returning #f, the former behavior.
6746 ** The #/ syntax for lists is no longer supported.
6748 Earlier versions of Scheme accepted this syntax, but printed a
6751 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
6753 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
6756 * Changes to the gh_ interface
6760 Now takes a second argument which is the result array. If this
6761 pointer is NULL, a new array is malloced (the old behaviour).
6763 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
6764 gh_scm2shorts, gh_scm2longs, gh_scm2floats
6768 * Changes to the scm_ interface
6770 ** Function: scm_make_named_hook (char* name, int n_args)
6772 Creates a hook in the same way as make-hook above but also
6773 binds a variable named NAME to it.
6775 This is the typical way of creating a hook from C code.
6777 Currently, the variable is created in the "current" module. This
6778 might change when we get the new module system.
6780 ** The smob interface
6782 The interface for creating smobs has changed. For documentation, see
6783 data-rep.info (made from guile-core/doc/data-rep.texi).
6785 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
6787 >>> This function will be removed in 1.3.4. <<<
6791 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
6792 This function adds a new smob type, named NAME, with instance size
6793 SIZE to the system. The return value is a tag that is used in
6794 creating instances of the type. If SIZE is 0, then no memory will
6795 be allocated when instances of the smob are created, and nothing
6796 will be freed by the default free function.
6798 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
6799 This function sets the smob marking procedure for the smob type
6800 specified by the tag TC. TC is the tag returned by
6801 `scm_make_smob_type'.
6803 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
6804 This function sets the smob freeing procedure for the smob type
6805 specified by the tag TC. TC is the tag returned by
6806 `scm_make_smob_type'.
6808 *** Function: void scm_set_smob_print (tc, print)
6810 - Function: void scm_set_smob_print (long tc,
6811 scm_sizet (*print) (SCM,
6815 This function sets the smob printing procedure for the smob type
6816 specified by the tag TC. TC is the tag returned by
6817 `scm_make_smob_type'.
6819 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
6820 This function sets the smob equality-testing predicate for the
6821 smob type specified by the tag TC. TC is the tag returned by
6822 `scm_make_smob_type'.
6824 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
6825 Make VALUE contain a smob instance of the type with type code TC and
6826 smob data DATA. VALUE must be previously declared as C type `SCM'.
6828 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
6829 This macro expands to a block of code that creates a smob instance
6830 of the type with type code TC and smob data DATA, and returns that
6831 `SCM' value. It should be the last piece of code in a block.
6833 ** The interfaces for using I/O ports and implementing port types
6834 (ptobs) have changed significantly. The new interface is based on
6835 shared access to buffers and a new set of ptob procedures.
6837 *** scm_newptob has been removed
6841 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
6843 - Function: SCM scm_make_port_type (char *type_name,
6844 int (*fill_buffer) (SCM port),
6845 void (*write_flush) (SCM port));
6847 Similarly to the new smob interface, there is a set of function
6848 setters by which the user can customize the behaviour of his port
6849 type. See ports.h (scm_set_port_XXX).
6851 ** scm_strport_to_string: New function: creates a new string from
6852 a string port's buffer.
6854 ** Plug in interface for random number generators
6855 The variable `scm_the_rng' in random.c contains a value and three
6856 function pointers which together define the current random number
6857 generator being used by the Scheme level interface and the random
6858 number library functions.
6860 The user is free to replace the default generator with the generator
6863 *** Variable: size_t scm_the_rng.rstate_size
6864 The size of the random state type used by the current RNG
6867 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
6868 Given the random STATE, return 32 random bits.
6870 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
6871 Seed random state STATE using string S of length N.
6873 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
6874 Given random state STATE, return a malloced copy.
6877 The default RNG is the MWC (Multiply With Carry) random number
6878 generator described by George Marsaglia at the Department of
6879 Statistics and Supercomputer Computations Research Institute, The
6880 Florida State University (http://stat.fsu.edu/~geo).
6882 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
6883 passes all tests in the DIEHARD test suite
6884 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
6885 costs one multiply and one add on platforms which either supports long
6886 longs (gcc does this on most systems) or have 64 bit longs. The cost
6887 is four multiply on other systems but this can be optimized by writing
6888 scm_i_uniform32 in assembler.
6890 These functions are provided through the scm_the_rng interface for use
6891 by libguile and the application.
6893 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
6894 Given the random STATE, return 32 random bits.
6895 Don't use this function directly. Instead go through the plugin
6896 interface (see "Plug in interface" above).
6898 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
6899 Initialize STATE using SEED of length N.
6901 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
6902 Return a malloc:ed copy of STATE. This function can easily be re-used
6903 in the interfaces to other RNGs.
6905 ** Random number library functions
6906 These functions use the current RNG through the scm_the_rng interface.
6907 It might be a good idea to use these functions from your C code so
6908 that only one random generator is used by all code in your program.
6910 The default random state is stored in:
6912 *** Variable: SCM scm_var_random_state
6913 Contains the vcell of the Scheme variable "*random-state*" which is
6914 used as default state by all random number functions in the Scheme
6919 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
6921 *** Function: scm_rstate *scm_c_default_rstate (void)
6922 This is a convenience function which returns the value of
6923 scm_var_random_state. An error message is generated if this value
6924 isn't a random state.
6926 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
6927 Make a new random state from the string SEED of length LENGTH.
6929 It is generally not a good idea to use multiple random states in a
6930 program. While subsequent random numbers generated from one random
6931 state are guaranteed to be reasonably independent, there is no such
6932 guarantee for numbers generated from different random states.
6934 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
6935 Return 32 random bits.
6937 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
6938 Return a sample from the uniform(0,1) distribution.
6940 *** Function: double scm_c_normal01 (scm_rstate *STATE)
6941 Return a sample from the normal(0,1) distribution.
6943 *** Function: double scm_c_exp1 (scm_rstate *STATE)
6944 Return a sample from the exp(1) distribution.
6946 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
6947 Return a sample from the discrete uniform(0,M) distribution.
6949 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
6950 Return a sample from the discrete uniform(0,M) distribution.
6951 M must be a bignum object. The returned value may be an INUM.
6955 Changes in Guile 1.3 (released Monday, October 19, 1998):
6957 * Changes to the distribution
6959 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
6960 To avoid conflicts, programs should name environment variables after
6961 themselves, except when there's a common practice establishing some
6964 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
6965 giving the former precedence, and printing a warning message if the
6966 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
6968 ** The header files related to multi-byte characters have been removed.
6969 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
6970 which referred to these explicitly will probably need to be rewritten,
6971 since the support for the variant string types has been removed; see
6974 ** The header files append.h and sequences.h have been removed. These
6975 files implemented non-R4RS operations which would encourage
6976 non-portable programming style and less easy-to-read code.
6978 * Changes to the stand-alone interpreter
6980 ** New procedures have been added to implement a "batch mode":
6982 *** Function: batch-mode?
6984 Returns a boolean indicating whether the interpreter is in batch
6987 *** Function: set-batch-mode?! ARG
6989 If ARG is true, switches the interpreter to batch mode. The `#f'
6990 case has not been implemented.
6992 ** Guile now provides full command-line editing, when run interactively.
6993 To use this feature, you must have the readline library installed.
6994 The Guile build process will notice it, and automatically include
6997 The readline library is available via anonymous FTP from any GNU
6998 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
7000 ** the-last-stack is now a fluid.
7002 * Changes to the procedure for linking libguile with your programs
7004 ** You can now use the `guile-config' utility to build programs that use Guile.
7006 Guile now includes a command-line utility called `guile-config', which
7007 can provide information about how to compile and link programs that
7010 *** `guile-config compile' prints any C compiler flags needed to use Guile.
7011 You should include this command's output on the command line you use
7012 to compile C or C++ code that #includes the Guile header files. It's
7013 usually just a `-I' flag to help the compiler find the Guile headers.
7016 *** `guile-config link' prints any linker flags necessary to link with Guile.
7018 This command writes to its standard output a list of flags which you
7019 must pass to the linker to link your code against the Guile library.
7020 The flags include '-lguile' itself, any other libraries the Guile
7021 library depends upon, and any `-L' flags needed to help the linker
7022 find those libraries.
7024 For example, here is a Makefile rule that builds a program named 'foo'
7025 from the object files ${FOO_OBJECTS}, and links them against Guile:
7028 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
7030 Previous Guile releases recommended that you use autoconf to detect
7031 which of a predefined set of libraries were present on your system.
7032 It is more robust to use `guile-config', since it records exactly which
7033 libraries the installed Guile library requires.
7035 This was originally called `build-guile', but was renamed to
7036 `guile-config' before Guile 1.3 was released, to be consistent with
7037 the analogous script for the GTK+ GUI toolkit, which is called
7041 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
7043 If you are using the GNU autoconf package to configure your program,
7044 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
7045 (described above) and gather the necessary values for use in your
7048 The GUILE_FLAGS macro expands to configure script code which runs the
7049 `guile-config' script, to find out where Guile's header files and
7050 libraries are installed. It sets two variables, marked for
7051 substitution, as by AC_SUBST.
7053 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
7054 code that uses Guile header files. This is almost always just a
7057 GUILE_LDFLAGS --- flags to pass to the linker to link a
7058 program against Guile. This includes `-lguile' for the Guile
7059 library itself, any libraries that Guile itself requires (like
7060 -lqthreads), and so on. It may also include a -L flag to tell the
7061 compiler where to find the libraries.
7063 GUILE_FLAGS is defined in the file guile.m4, in the top-level
7064 directory of the Guile distribution. You can copy it into your
7065 package's aclocal.m4 file, and then use it in your configure.in file.
7067 If you are using the `aclocal' program, distributed with GNU automake,
7068 to maintain your aclocal.m4 file, the Guile installation process
7069 installs guile.m4 where aclocal will find it. All you need to do is
7070 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
7071 this will copy the definition of GUILE_FLAGS into your aclocal.m4
7075 * Changes to Scheme functions and syntax
7077 ** Multi-byte strings have been removed, as have multi-byte and wide
7078 ports. We felt that these were the wrong approach to
7079 internationalization support.
7081 ** New function: readline [PROMPT]
7082 Read a line from the terminal, and allow the user to edit it,
7083 prompting with PROMPT. READLINE provides a large set of Emacs-like
7084 editing commands, lets the user recall previously typed lines, and
7085 works on almost every kind of terminal, including dumb terminals.
7087 READLINE assumes that the cursor is at the beginning of the line when
7088 it is invoked. Thus, you can't print a prompt yourself, and then call
7089 READLINE; you need to package up your prompt as a string, pass it to
7090 the function, and let READLINE print the prompt itself. This is
7091 because READLINE needs to know the prompt's screen width.
7093 For Guile to provide this function, you must have the readline
7094 library, version 2.1 or later, installed on your system. Readline is
7095 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
7096 any GNU mirror site.
7098 See also ADD-HISTORY function.
7100 ** New function: add-history STRING
7101 Add STRING as the most recent line in the history used by the READLINE
7102 command. READLINE does not add lines to the history itself; you must
7103 call ADD-HISTORY to make previous input available to the user.
7105 ** The behavior of the read-line function has changed.
7107 This function now uses standard C library functions to read the line,
7108 for speed. This means that it doesn not respect the value of
7109 scm-line-incrementors; it assumes that lines are delimited with
7112 (Note that this is read-line, the function that reads a line of text
7113 from a port, not readline, the function that reads a line from a
7114 terminal, providing full editing capabilities.)
7116 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
7118 This module provides some simple argument parsing. It exports one
7121 Function: getopt-gnu-style ARG-LS
7122 Parse a list of program arguments into an alist of option
7125 Each item in the list of program arguments is examined to see if
7126 it meets the syntax of a GNU long-named option. An argument like
7127 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
7128 returned alist, where MUMBLE is a keyword object with the same
7129 name as the argument. An argument like `--MUMBLE=FROB' produces
7130 an element of the form (MUMBLE . FROB), where FROB is a string.
7132 As a special case, the returned alist also contains a pair whose
7133 car is the symbol `rest'. The cdr of this pair is a list
7134 containing all the items in the argument list that are not options
7135 of the form mentioned above.
7137 The argument `--' is treated specially: all items in the argument
7138 list appearing after such an argument are not examined, and are
7139 returned in the special `rest' list.
7141 This function does not parse normal single-character switches.
7142 You will need to parse them out of the `rest' list yourself.
7144 ** The read syntax for byte vectors and short vectors has changed.
7146 Instead of #bytes(...), write #y(...).
7148 Instead of #short(...), write #h(...).
7150 This may seem nutty, but, like the other uniform vectors, byte vectors
7151 and short vectors want to have the same print and read syntax (and,
7152 more basic, want to have read syntax!). Changing the read syntax to
7153 use multiple characters after the hash sign breaks with the
7154 conventions used in R5RS and the conventions used for the other
7155 uniform vectors. It also introduces complexity in the current reader,
7156 both on the C and Scheme levels. (The Right solution is probably to
7157 change the syntax and prototypes for uniform vectors entirely.)
7160 ** The new module (ice-9 session) provides useful interactive functions.
7162 *** New procedure: (apropos REGEXP OPTION ...)
7164 Display a list of top-level variables whose names match REGEXP, and
7165 the modules they are imported from. Each OPTION should be one of the
7168 value --- Show the value of each matching variable.
7169 shadow --- Show bindings shadowed by subsequently imported modules.
7170 full --- Same as both `shadow' and `value'.
7174 guile> (apropos "trace" 'full)
7175 debug: trace #<procedure trace args>
7176 debug: untrace #<procedure untrace args>
7177 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
7178 the-scm-module: before-backtrace-hook ()
7179 the-scm-module: backtrace #<primitive-procedure backtrace>
7180 the-scm-module: after-backtrace-hook ()
7181 the-scm-module: has-shown-backtrace-hint? #f
7184 ** There are new functions and syntax for working with macros.
7186 Guile implements macros as a special object type. Any variable whose
7187 top-level binding is a macro object acts as a macro. The macro object
7188 specifies how the expression should be transformed before evaluation.
7190 *** Macro objects now print in a reasonable way, resembling procedures.
7192 *** New function: (macro? OBJ)
7193 True iff OBJ is a macro object.
7195 *** New function: (primitive-macro? OBJ)
7196 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
7197 macro transformers, implemented in eval.c rather than Scheme code.
7199 Why do we have this function?
7200 - For symmetry with procedure? and primitive-procedure?,
7201 - to allow custom print procedures to tell whether a macro is
7202 primitive, and display it differently, and
7203 - to allow compilers and user-written evaluators to distinguish
7204 builtin special forms from user-defined ones, which could be
7207 *** New function: (macro-type OBJ)
7208 Return a value indicating what kind of macro OBJ is. Possible return
7211 The symbol `syntax' --- a macro created by procedure->syntax.
7212 The symbol `macro' --- a macro created by procedure->macro.
7213 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
7214 The boolean #f --- if OBJ is not a macro object.
7216 *** New function: (macro-name MACRO)
7217 Return the name of the macro object MACRO's procedure, as returned by
7220 *** New function: (macro-transformer MACRO)
7221 Return the transformer procedure for MACRO.
7223 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
7225 Specify a new macro expander to use in the current module. Each
7226 MODULE is a module name, with the same meaning as in the `use-modules'
7227 form; each named module's exported bindings are added to the current
7228 top-level environment. TRANSFORMER is an expression evaluated in the
7229 resulting environment which must yield a procedure to use as the
7230 module's eval transformer: every expression evaluated in this module
7231 is passed to this function, and the result passed to the Guile
7234 *** macro-eval! is removed. Use local-eval instead.
7236 ** Some magic has been added to the printer to better handle user
7237 written printing routines (like record printers, closure printers).
7239 The problem is that these user written routines must have access to
7240 the current `print-state' to be able to handle fancy things like
7241 detection of circular references. These print-states have to be
7242 passed to the builtin printing routines (display, write, etc) to
7243 properly continue the print chain.
7245 We didn't want to change all existing print code so that it
7246 explicitly passes thru a print state in addition to a port. Instead,
7247 we extented the possible values that the builtin printing routines
7248 accept as a `port'. In addition to a normal port, they now also take
7249 a pair of a normal port and a print-state. Printing will go to the
7250 port and the print-state will be used to control the detection of
7251 circular references, etc. If the builtin function does not care for a
7252 print-state, it is simply ignored.
7254 User written callbacks are now called with such a pair as their
7255 `port', but because every function now accepts this pair as a PORT
7256 argument, you don't have to worry about that. In fact, it is probably
7257 safest to not check for these pairs.
7259 However, it is sometimes necessary to continue a print chain on a
7260 different port, for example to get a intermediate string
7261 representation of the printed value, mangle that string somehow, and
7262 then to finally print the mangled string. Use the new function
7264 inherit-print-state OLD-PORT NEW-PORT
7266 for this. It constructs a new `port' that prints to NEW-PORT but
7267 inherits the print-state of OLD-PORT.
7269 ** struct-vtable-offset renamed to vtable-offset-user
7271 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
7273 ** There is now a third optional argument to make-vtable-vtable
7274 (and fourth to make-struct) when constructing new types (vtables).
7275 This argument initializes field vtable-index-printer of the vtable.
7277 ** The detection of circular references has been extended to structs.
7278 That is, a structure that -- in the process of being printed -- prints
7279 itself does not lead to infinite recursion.
7281 ** There is now some basic support for fluids. Please read
7282 "libguile/fluid.h" to find out more. It is accessible from Scheme with
7283 the following functions and macros:
7285 Function: make-fluid
7287 Create a new fluid object. Fluids are not special variables or
7288 some other extension to the semantics of Scheme, but rather
7289 ordinary Scheme objects. You can store them into variables (that
7290 are still lexically scoped, of course) or into any other place you
7291 like. Every fluid has a initial value of `#f'.
7293 Function: fluid? OBJ
7295 Test whether OBJ is a fluid.
7297 Function: fluid-ref FLUID
7298 Function: fluid-set! FLUID VAL
7300 Access/modify the fluid FLUID. Modifications are only visible
7301 within the current dynamic root (that includes threads).
7303 Function: with-fluids* FLUIDS VALUES THUNK
7305 FLUIDS is a list of fluids and VALUES a corresponding list of
7306 values for these fluids. Before THUNK gets called the values are
7307 installed in the fluids and the old values of the fluids are
7308 saved in the VALUES list. When the flow of control leaves THUNK
7309 or reenters it, the values get swapped again. You might think of
7310 this as a `safe-fluid-excursion'. Note that the VALUES list is
7311 modified by `with-fluids*'.
7313 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
7315 The same as `with-fluids*' but with a different syntax. It looks
7316 just like `let', but both FLUID and VALUE are evaluated. Remember,
7317 fluids are not special variables but ordinary objects. FLUID
7318 should evaluate to a fluid.
7320 ** Changes to system call interfaces:
7322 *** close-port, close-input-port and close-output-port now return a
7323 boolean instead of an `unspecified' object. #t means that the port
7324 was successfully closed, while #f means it was already closed. It is
7325 also now possible for these procedures to raise an exception if an
7326 error occurs (some errors from write can be delayed until close.)
7328 *** the first argument to chmod, fcntl, ftell and fseek can now be a
7331 *** the third argument to fcntl is now optional.
7333 *** the first argument to chown can now be a file descriptor or a port.
7335 *** the argument to stat can now be a port.
7337 *** The following new procedures have been added (most use scsh
7340 *** procedure: close PORT/FD
7341 Similar to close-port (*note close-port: Closing Ports.), but also
7342 works on file descriptors. A side effect of closing a file
7343 descriptor is that any ports using that file descriptor are moved
7344 to a different file descriptor and have their revealed counts set
7347 *** procedure: port->fdes PORT
7348 Returns the integer file descriptor underlying PORT. As a side
7349 effect the revealed count of PORT is incremented.
7351 *** procedure: fdes->ports FDES
7352 Returns a list of existing ports which have FDES as an underlying
7353 file descriptor, without changing their revealed counts.
7355 *** procedure: fdes->inport FDES
7356 Returns an existing input port which has FDES as its underlying
7357 file descriptor, if one exists, and increments its revealed count.
7358 Otherwise, returns a new input port with a revealed count of 1.
7360 *** procedure: fdes->outport FDES
7361 Returns an existing output port which has FDES as its underlying
7362 file descriptor, if one exists, and increments its revealed count.
7363 Otherwise, returns a new output port with a revealed count of 1.
7365 The next group of procedures perform a `dup2' system call, if NEWFD
7366 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
7367 duplicated can be supplied as an integer or contained in a port. The
7368 type of value returned varies depending on which procedure is used.
7370 All procedures also have the side effect when performing `dup2' that
7371 any ports using NEWFD are moved to a different file descriptor and have
7372 their revealed counts set to zero.
7374 *** procedure: dup->fdes PORT/FD [NEWFD]
7375 Returns an integer file descriptor.
7377 *** procedure: dup->inport PORT/FD [NEWFD]
7378 Returns a new input port using the new file descriptor.
7380 *** procedure: dup->outport PORT/FD [NEWFD]
7381 Returns a new output port using the new file descriptor.
7383 *** procedure: dup PORT/FD [NEWFD]
7384 Returns a new port if PORT/FD is a port, with the same mode as the
7385 supplied port, otherwise returns an integer file descriptor.
7387 *** procedure: dup->port PORT/FD MODE [NEWFD]
7388 Returns a new port using the new file descriptor. MODE supplies a
7389 mode string for the port (*note open-file: File Ports.).
7391 *** procedure: setenv NAME VALUE
7392 Modifies the environment of the current process, which is also the
7393 default environment inherited by child processes.
7395 If VALUE is `#f', then NAME is removed from the environment.
7396 Otherwise, the string NAME=VALUE is added to the environment,
7397 replacing any existing string with name matching NAME.
7399 The return value is unspecified.
7401 *** procedure: truncate-file OBJ SIZE
7402 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
7403 can be a string containing a file name or an integer file
7404 descriptor or port open for output on the file. The underlying
7405 system calls are `truncate' and `ftruncate'.
7407 The return value is unspecified.
7409 *** procedure: setvbuf PORT MODE [SIZE]
7410 Set the buffering mode for PORT. MODE can be:
7418 block buffered, using a newly allocated buffer of SIZE bytes.
7419 However if SIZE is zero or unspecified, the port will be made
7422 This procedure should not be used after I/O has been performed with
7425 Ports are usually block buffered by default, with a default buffer
7426 size. Procedures e.g., *Note open-file: File Ports, which accept a
7427 mode string allow `0' to be added to request an unbuffered port.
7429 *** procedure: fsync PORT/FD
7430 Copies any unwritten data for the specified output file descriptor
7431 to disk. If PORT/FD is a port, its buffer is flushed before the
7432 underlying file descriptor is fsync'd. The return value is
7435 *** procedure: open-fdes PATH FLAGS [MODES]
7436 Similar to `open' but returns a file descriptor instead of a port.
7438 *** procedure: execle PATH ENV [ARG] ...
7439 Similar to `execl', but the environment of the new process is
7440 specified by ENV, which must be a list of strings as returned by
7441 the `environ' procedure.
7443 This procedure is currently implemented using the `execve' system
7444 call, but we call it `execle' because of its Scheme calling
7447 *** procedure: strerror ERRNO
7448 Returns the Unix error message corresponding to ERRNO, an integer.
7450 *** procedure: primitive-exit [STATUS]
7451 Terminate the current process without unwinding the Scheme stack.
7452 This is would typically be useful after a fork. The exit status
7453 is STATUS if supplied, otherwise zero.
7455 *** procedure: times
7456 Returns an object with information about real and processor time.
7457 The following procedures accept such an object as an argument and
7458 return a selected component:
7461 The current real time, expressed as time units relative to an
7465 The CPU time units used by the calling process.
7468 The CPU time units used by the system on behalf of the
7472 The CPU time units used by terminated child processes of the
7473 calling process, whose status has been collected (e.g., using
7477 Similarly, the CPU times units used by the system on behalf of
7478 terminated child processes.
7480 ** Removed: list-length
7481 ** Removed: list-append, list-append!
7482 ** Removed: list-reverse, list-reverse!
7484 ** array-map renamed to array-map!
7486 ** serial-array-map renamed to serial-array-map!
7488 ** catch doesn't take #f as first argument any longer
7490 Previously, it was possible to pass #f instead of a key to `catch'.
7491 That would cause `catch' to pass a jump buffer object to the procedure
7492 passed as second argument. The procedure could then use this jump
7493 buffer objekt as an argument to throw.
7495 This mechanism has been removed since its utility doesn't motivate the
7496 extra complexity it introduces.
7498 ** The `#/' notation for lists now provokes a warning message from Guile.
7499 This syntax will be removed from Guile in the near future.
7501 To disable the warning message, set the GUILE_HUSH environment
7502 variable to any non-empty value.
7504 ** The newline character now prints as `#\newline', following the
7505 normal Scheme notation, not `#\nl'.
7507 * Changes to the gh_ interface
7509 ** The gh_enter function now takes care of loading the Guile startup files.
7510 gh_enter works by calling scm_boot_guile; see the remarks below.
7512 ** Function: void gh_write (SCM x)
7514 Write the printed representation of the scheme object x to the current
7515 output port. Corresponds to the scheme level `write'.
7517 ** gh_list_length renamed to gh_length.
7519 ** vector handling routines
7521 Several major changes. In particular, gh_vector() now resembles
7522 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
7523 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
7524 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
7525 vector-related gh_ functions have been implemented.
7527 ** pair and list routines
7529 Implemented several of the R4RS pair and list functions that were
7532 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
7534 New function. Converts double arrays back and forth between Scheme
7537 * Changes to the scm_ interface
7539 ** The function scm_boot_guile now takes care of loading the startup files.
7541 Guile's primary initialization function, scm_boot_guile, now takes
7542 care of loading `boot-9.scm', in the `ice-9' module, to initialize
7543 Guile, define the module system, and put together some standard
7544 bindings. It also loads `init.scm', which is intended to hold
7545 site-specific initialization code.
7547 Since Guile cannot operate properly until boot-9.scm is loaded, there
7548 is no reason to separate loading boot-9.scm from Guile's other
7549 initialization processes.
7551 This job used to be done by scm_compile_shell_switches, which didn't
7552 make much sense; in particular, it meant that people using Guile for
7553 non-shell-like applications had to jump through hoops to get Guile
7554 initialized properly.
7556 ** The function scm_compile_shell_switches no longer loads the startup files.
7557 Now, Guile always loads the startup files, whenever it is initialized;
7558 see the notes above for scm_boot_guile and scm_load_startup_files.
7560 ** Function: scm_load_startup_files
7561 This new function takes care of loading Guile's initialization file
7562 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
7563 this is always called by the Guile initialization process, it's
7564 probably not too useful to call this yourself, but it's there anyway.
7566 ** The semantics of smob marking have changed slightly.
7568 The smob marking function (the `mark' member of the scm_smobfuns
7569 structure) is no longer responsible for setting the mark bit on the
7570 smob. The generic smob handling code in the garbage collector will
7571 set this bit. The mark function need only ensure that any other
7572 objects the smob refers to get marked.
7574 Note that this change means that the smob's GC8MARK bit is typically
7575 already set upon entry to the mark function. Thus, marking functions
7576 which look like this:
7579 if (SCM_GC8MARKP (ptr))
7581 SCM_SETGC8MARK (ptr);
7582 ... mark objects to which the smob refers ...
7585 are now incorrect, since they will return early, and fail to mark any
7586 other objects the smob refers to. Some code in the Guile library used
7589 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
7591 If you have implemented your own I/O port type, by writing the
7592 functions required by the scm_ptobfuns and then calling scm_newptob,
7593 you will need to change your functions slightly.
7595 The functions in a scm_ptobfuns structure now expect the port itself
7596 as their argument; they used to expect the `stream' member of the
7597 port's scm_port_table structure. This allows functions in an
7598 scm_ptobfuns structure to easily access the port's cell (and any flags
7599 it its CAR), and the port's scm_port_table structure.
7601 Guile now passes the I/O port itself as the `port' argument in the
7602 following scm_ptobfuns functions:
7604 int (*free) (SCM port);
7605 int (*fputc) (int, SCM port);
7606 int (*fputs) (char *, SCM port);
7607 scm_sizet (*fwrite) SCM_P ((char *ptr,
7611 int (*fflush) (SCM port);
7612 int (*fgetc) (SCM port);
7613 int (*fclose) (SCM port);
7615 The interfaces to the `mark', `print', `equalp', and `fgets' methods
7618 If you have existing code which defines its own port types, it is easy
7619 to convert your code to the new interface; simply apply SCM_STREAM to
7620 the port argument to yield the value you code used to expect.
7622 Note that since both the port and the stream have the same type in the
7623 C code --- they are both SCM values --- the C compiler will not remind
7624 you if you forget to update your scm_ptobfuns functions.
7627 ** Function: int scm_internal_select (int fds,
7631 struct timeval *timeout);
7633 This is a replacement for the `select' function provided by the OS.
7634 It enables I/O blocking and sleeping to happen for one cooperative
7635 thread without blocking other threads. It also avoids busy-loops in
7636 these situations. It is intended that all I/O blocking and sleeping
7637 will finally go through this function. Currently, this function is
7638 only available on systems providing `gettimeofday' and `select'.
7640 ** Function: SCM scm_internal_stack_catch (SCM tag,
7641 scm_catch_body_t body,
7643 scm_catch_handler_t handler,
7646 A new sibling to the other two C level `catch' functions
7647 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
7648 the stack to be saved automatically into the variable `the-last-stack'
7649 (scm_the_last_stack_var) on error. This is necessary if you want to
7650 use advanced error reporting, such as calling scm_display_error and
7651 scm_display_backtrace. (They both take a stack object as argument.)
7653 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
7655 scm_catch_handler_t handler,
7658 Spawns a new thread. It does a job similar to
7659 scm_call_with_new_thread but takes arguments more suitable when
7660 spawning threads from application C code.
7662 ** The hook scm_error_callback has been removed. It was originally
7663 intended as a way for the user to install his own error handler. But
7664 that method works badly since it intervenes between throw and catch,
7665 thereby changing the semantics of expressions like (catch #t ...).
7666 The correct way to do it is to use one of the C level catch functions
7667 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
7669 ** Removed functions:
7671 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
7672 scm_list_reverse, scm_list_reverse_x
7674 ** New macros: SCM_LISTn where n is one of the integers 0-9.
7676 These can be used for pretty list creation from C. The idea is taken
7677 from Erick Gallesio's STk.
7679 ** scm_array_map renamed to scm_array_map_x
7681 ** mbstrings are now removed
7683 This means that the type codes scm_tc7_mb_string and
7684 scm_tc7_mb_substring has been removed.
7686 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
7688 Since we no longer support multi-byte strings, these I/O functions
7689 have been simplified, and renamed. Here are their old names, and
7690 their new names and arguments:
7692 scm_gen_putc -> void scm_putc (int c, SCM port);
7693 scm_gen_puts -> void scm_puts (char *s, SCM port);
7694 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
7695 scm_gen_getc -> void scm_getc (SCM port);
7698 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
7700 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
7702 SCM_TYP7S now masks away the bit which distinguishes substrings from
7705 ** scm_catch_body_t: Backward incompatible change!
7707 Body functions to scm_internal_catch and friends do not any longer
7708 take a second argument. This is because it is no longer possible to
7709 pass a #f arg to catch.
7711 ** Calls to scm_protect_object and scm_unprotect now nest properly.
7713 The function scm_protect_object protects its argument from being freed
7714 by the garbage collector. scm_unprotect_object removes that
7717 These functions now nest properly. That is, for every object O, there
7718 is a counter which scm_protect_object(O) increments and
7719 scm_unprotect_object(O) decrements, if the counter is greater than
7720 zero. Every object's counter is zero when it is first created. If an
7721 object's counter is greater than zero, the garbage collector will not
7722 reclaim its storage.
7724 This allows you to use scm_protect_object in your code without
7725 worrying that some other function you call will call
7726 scm_unprotect_object, and allow it to be freed. Assuming that the
7727 functions you call are well-behaved, and unprotect only those objects
7728 they protect, you can follow the same rule and have confidence that
7729 objects will be freed only at appropriate times.
7732 Changes in Guile 1.2 (released Tuesday, June 24 1997):
7734 * Changes to the distribution
7736 ** Nightly snapshots are now available from ftp.red-bean.com.
7737 The old server, ftp.cyclic.com, has been relinquished to its rightful
7740 Nightly snapshots of the Guile development sources are now available via
7741 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
7743 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
7744 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
7746 ** To run Guile without installing it, the procedure has changed a bit.
7748 If you used a separate build directory to compile Guile, you'll need
7749 to include the build directory in SCHEME_LOAD_PATH, as well as the
7750 source directory. See the `INSTALL' file for examples.
7752 * Changes to the procedure for linking libguile with your programs
7754 ** The standard Guile load path for Scheme code now includes
7755 $(datadir)/guile (usually /usr/local/share/guile). This means that
7756 you can install your own Scheme files there, and Guile will find them.
7757 (Previous versions of Guile only checked a directory whose name
7758 contained the Guile version number, so you had to re-install or move
7759 your Scheme sources each time you installed a fresh version of Guile.)
7761 The load path also includes $(datadir)/guile/site; we recommend
7762 putting individual Scheme files there. If you want to install a
7763 package with multiple source files, create a directory for them under
7766 ** Guile 1.2 will now use the Rx regular expression library, if it is
7767 installed on your system. When you are linking libguile into your own
7768 programs, this means you will have to link against -lguile, -lqt (if
7769 you configured Guile with thread support), and -lrx.
7771 If you are using autoconf to generate configuration scripts for your
7772 application, the following lines should suffice to add the appropriate
7773 libraries to your link command:
7775 ### Find Rx, quickthreads and libguile.
7776 AC_CHECK_LIB(rx, main)
7777 AC_CHECK_LIB(qt, main)
7778 AC_CHECK_LIB(guile, scm_shell)
7780 The Guile 1.2 distribution does not contain sources for the Rx
7781 library, as Guile 1.0 did. If you want to use Rx, you'll need to
7782 retrieve it from a GNU FTP site and install it separately.
7784 * Changes to Scheme functions and syntax
7786 ** The dynamic linking features of Guile are now enabled by default.
7787 You can disable them by giving the `--disable-dynamic-linking' option
7790 (dynamic-link FILENAME)
7792 Find the object file denoted by FILENAME (a string) and link it
7793 into the running Guile application. When everything works out,
7794 return a Scheme object suitable for representing the linked object
7795 file. Otherwise an error is thrown. How object files are
7796 searched is system dependent.
7798 (dynamic-object? VAL)
7800 Determine whether VAL represents a dynamically linked object file.
7802 (dynamic-unlink DYNOBJ)
7804 Unlink the indicated object file from the application. DYNOBJ
7805 should be one of the values returned by `dynamic-link'.
7807 (dynamic-func FUNCTION DYNOBJ)
7809 Search the C function indicated by FUNCTION (a string or symbol)
7810 in DYNOBJ and return some Scheme object that can later be used
7811 with `dynamic-call' to actually call this function. Right now,
7812 these Scheme objects are formed by casting the address of the
7813 function to `long' and converting this number to its Scheme
7816 (dynamic-call FUNCTION DYNOBJ)
7818 Call the C function indicated by FUNCTION and DYNOBJ. The
7819 function is passed no arguments and its return value is ignored.
7820 When FUNCTION is something returned by `dynamic-func', call that
7821 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
7822 etc.), look it up in DYNOBJ; this is equivalent to
7824 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
7826 Interrupts are deferred while the C function is executing (with
7827 SCM_DEFER_INTS/SCM_ALLOW_INTS).
7829 (dynamic-args-call FUNCTION DYNOBJ ARGS)
7831 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
7832 some arguments and return its return value. The C function is
7833 expected to take two arguments and return an `int', just like
7836 int c_func (int argc, char **argv);
7838 ARGS must be a list of strings and is converted into an array of
7839 `char *'. The array is passed in ARGV and its size in ARGC. The
7840 return value is converted to a Scheme number and returned from the
7841 call to `dynamic-args-call'.
7843 When dynamic linking is disabled or not supported on your system,
7844 the above functions throw errors, but they are still available.
7846 Here is a small example that works on GNU/Linux:
7848 (define libc-obj (dynamic-link "libc.so"))
7849 (dynamic-args-call 'rand libc-obj '())
7851 See the file `libguile/DYNAMIC-LINKING' for additional comments.
7853 ** The #/ syntax for module names is depreciated, and will be removed
7854 in a future version of Guile. Instead of
7862 The latter syntax is more consistent with existing Lisp practice.
7864 ** Guile now does fancier printing of structures. Structures are the
7865 underlying implementation for records, which in turn are used to
7866 implement modules, so all of these object now print differently and in
7867 a more informative way.
7869 The Scheme printer will examine the builtin variable *struct-printer*
7870 whenever it needs to print a structure object. When this variable is
7871 not `#f' it is deemed to be a procedure and will be applied to the
7872 structure object and the output port. When *struct-printer* is `#f'
7873 or the procedure return `#f' the structure object will be printed in
7874 the boring #<struct 80458270> form.
7876 This hook is used by some routines in ice-9/boot-9.scm to implement
7877 type specific printing routines. Please read the comments there about
7880 One of the more specific uses of structs are records. The printing
7881 procedure that could be passed to MAKE-RECORD-TYPE is now actually
7882 called. It should behave like a *struct-printer* procedure (described
7885 ** Guile now supports a new R4RS-compliant syntax for keywords. A
7886 token of the form #:NAME, where NAME has the same syntax as a Scheme
7887 symbol, is the external representation of the keyword named NAME.
7888 Keyword objects print using this syntax as well, so values containing
7889 keyword objects can be read back into Guile. When used in an
7890 expression, keywords are self-quoting objects.
7892 Guile suports this read syntax, and uses this print syntax, regardless
7893 of the current setting of the `keyword' read option. The `keyword'
7894 read option only controls whether Guile recognizes the `:NAME' syntax,
7895 which is incompatible with R4RS. (R4RS says such token represent
7898 ** Guile has regular expression support again. Guile 1.0 included
7899 functions for matching regular expressions, based on the Rx library.
7900 In Guile 1.1, the Guile/Rx interface was removed to simplify the
7901 distribution, and thus Guile had no regular expression support. Guile
7902 1.2 again supports the most commonly used functions, and supports all
7903 of SCSH's regular expression functions.
7905 If your system does not include a POSIX regular expression library,
7906 and you have not linked Guile with a third-party regexp library such as
7907 Rx, these functions will not be available. You can tell whether your
7908 Guile installation includes regular expression support by checking
7909 whether the `*features*' list includes the `regex' symbol.
7911 *** regexp functions
7913 By default, Guile supports POSIX extended regular expressions. That
7914 means that the characters `(', `)', `+' and `?' are special, and must
7915 be escaped if you wish to match the literal characters.
7917 This regular expression interface was modeled after that implemented
7918 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
7919 with SCSH regular expressions.
7921 **** Function: string-match PATTERN STR [START]
7922 Compile the string PATTERN into a regular expression and compare
7923 it with STR. The optional numeric argument START specifies the
7924 position of STR at which to begin matching.
7926 `string-match' returns a "match structure" which describes what,
7927 if anything, was matched by the regular expression. *Note Match
7928 Structures::. If STR does not match PATTERN at all,
7929 `string-match' returns `#f'.
7931 Each time `string-match' is called, it must compile its PATTERN
7932 argument into a regular expression structure. This operation is
7933 expensive, which makes `string-match' inefficient if the same regular
7934 expression is used several times (for example, in a loop). For better
7935 performance, you can compile a regular expression in advance and then
7936 match strings against the compiled regexp.
7938 **** Function: make-regexp STR [FLAGS]
7939 Compile the regular expression described by STR, and return the
7940 compiled regexp structure. If STR does not describe a legal
7941 regular expression, `make-regexp' throws a
7942 `regular-expression-syntax' error.
7944 FLAGS may be the bitwise-or of one or more of the following:
7946 **** Constant: regexp/extended
7947 Use POSIX Extended Regular Expression syntax when interpreting
7948 STR. If not set, POSIX Basic Regular Expression syntax is used.
7949 If the FLAGS argument is omitted, we assume regexp/extended.
7951 **** Constant: regexp/icase
7952 Do not differentiate case. Subsequent searches using the
7953 returned regular expression will be case insensitive.
7955 **** Constant: regexp/newline
7956 Match-any-character operators don't match a newline.
7958 A non-matching list ([^...]) not containing a newline matches a
7961 Match-beginning-of-line operator (^) matches the empty string
7962 immediately after a newline, regardless of whether the FLAGS
7963 passed to regexp-exec contain regexp/notbol.
7965 Match-end-of-line operator ($) matches the empty string
7966 immediately before a newline, regardless of whether the FLAGS
7967 passed to regexp-exec contain regexp/noteol.
7969 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
7970 Match the compiled regular expression REGEXP against `str'. If
7971 the optional integer START argument is provided, begin matching
7972 from that position in the string. Return a match structure
7973 describing the results of the match, or `#f' if no match could be
7976 FLAGS may be the bitwise-or of one or more of the following:
7978 **** Constant: regexp/notbol
7979 The match-beginning-of-line operator always fails to match (but
7980 see the compilation flag regexp/newline above) This flag may be
7981 used when different portions of a string are passed to
7982 regexp-exec and the beginning of the string should not be
7983 interpreted as the beginning of the line.
7985 **** Constant: regexp/noteol
7986 The match-end-of-line operator always fails to match (but see the
7987 compilation flag regexp/newline above)
7989 **** Function: regexp? OBJ
7990 Return `#t' if OBJ is a compiled regular expression, or `#f'
7993 Regular expressions are commonly used to find patterns in one string
7994 and replace them with the contents of another string.
7996 **** Function: regexp-substitute PORT MATCH [ITEM...]
7997 Write to the output port PORT selected contents of the match
7998 structure MATCH. Each ITEM specifies what should be written, and
7999 may be one of the following arguments:
8001 * A string. String arguments are written out verbatim.
8003 * An integer. The submatch with that number is written.
8005 * The symbol `pre'. The portion of the matched string preceding
8006 the regexp match is written.
8008 * The symbol `post'. The portion of the matched string
8009 following the regexp match is written.
8011 PORT may be `#f', in which case nothing is written; instead,
8012 `regexp-substitute' constructs a string from the specified ITEMs
8015 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
8016 Similar to `regexp-substitute', but can be used to perform global
8017 substitutions on STR. Instead of taking a match structure as an
8018 argument, `regexp-substitute/global' takes two string arguments: a
8019 REGEXP string describing a regular expression, and a TARGET string
8020 which should be matched against this regular expression.
8022 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
8025 * A function may be supplied. When this function is called, it
8026 will be passed one argument: a match structure for a given
8027 regular expression match. It should return a string to be
8028 written out to PORT.
8030 * The `post' symbol causes `regexp-substitute/global' to recurse
8031 on the unmatched portion of STR. This *must* be supplied in
8032 order to perform global search-and-replace on STR; if it is
8033 not present among the ITEMs, then `regexp-substitute/global'
8034 will return after processing a single match.
8036 *** Match Structures
8038 A "match structure" is the object returned by `string-match' and
8039 `regexp-exec'. It describes which portion of a string, if any, matched
8040 the given regular expression. Match structures include: a reference to
8041 the string that was checked for matches; the starting and ending
8042 positions of the regexp match; and, if the regexp included any
8043 parenthesized subexpressions, the starting and ending positions of each
8046 In each of the regexp match functions described below, the `match'
8047 argument must be a match structure returned by a previous call to
8048 `string-match' or `regexp-exec'. Most of these functions return some
8049 information about the original target string that was matched against a
8050 regular expression; we will call that string TARGET for easy reference.
8052 **** Function: regexp-match? OBJ
8053 Return `#t' if OBJ is a match structure returned by a previous
8054 call to `regexp-exec', or `#f' otherwise.
8056 **** Function: match:substring MATCH [N]
8057 Return the portion of TARGET matched by subexpression number N.
8058 Submatch 0 (the default) represents the entire regexp match. If
8059 the regular expression as a whole matched, but the subexpression
8060 number N did not match, return `#f'.
8062 **** Function: match:start MATCH [N]
8063 Return the starting position of submatch number N.
8065 **** Function: match:end MATCH [N]
8066 Return the ending position of submatch number N.
8068 **** Function: match:prefix MATCH
8069 Return the unmatched portion of TARGET preceding the regexp match.
8071 **** Function: match:suffix MATCH
8072 Return the unmatched portion of TARGET following the regexp match.
8074 **** Function: match:count MATCH
8075 Return the number of parenthesized subexpressions from MATCH.
8076 Note that the entire regular expression match itself counts as a
8077 subexpression, and failed submatches are included in the count.
8079 **** Function: match:string MATCH
8080 Return the original TARGET string.
8082 *** Backslash Escapes
8084 Sometimes you will want a regexp to match characters like `*' or `$'
8085 exactly. For example, to check whether a particular string represents
8086 a menu entry from an Info node, it would be useful to match it against
8087 a regexp like `^* [^:]*::'. However, this won't work; because the
8088 asterisk is a metacharacter, it won't match the `*' at the beginning of
8089 the string. In this case, we want to make the first asterisk un-magic.
8091 You can do this by preceding the metacharacter with a backslash
8092 character `\'. (This is also called "quoting" the metacharacter, and
8093 is known as a "backslash escape".) When Guile sees a backslash in a
8094 regular expression, it considers the following glyph to be an ordinary
8095 character, no matter what special meaning it would ordinarily have.
8096 Therefore, we can make the above example work by changing the regexp to
8097 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
8098 to match only a single asterisk in the target string.
8100 Since the backslash is itself a metacharacter, you may force a
8101 regexp to match a backslash in the target string by preceding the
8102 backslash with itself. For example, to find variable references in a
8103 TeX program, you might want to find occurrences of the string `\let\'
8104 followed by any number of alphabetic characters. The regular expression
8105 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
8106 each match a single backslash in the target string.
8108 **** Function: regexp-quote STR
8109 Quote each special character found in STR with a backslash, and
8110 return the resulting string.
8112 *Very important:* Using backslash escapes in Guile source code (as
8113 in Emacs Lisp or C) can be tricky, because the backslash character has
8114 special meaning for the Guile reader. For example, if Guile encounters
8115 the character sequence `\n' in the middle of a string while processing
8116 Scheme code, it replaces those characters with a newline character.
8117 Similarly, the character sequence `\t' is replaced by a horizontal tab.
8118 Several of these "escape sequences" are processed by the Guile reader
8119 before your code is executed. Unrecognized escape sequences are
8120 ignored: if the characters `\*' appear in a string, they will be
8121 translated to the single character `*'.
8123 This translation is obviously undesirable for regular expressions,
8124 since we want to be able to include backslashes in a string in order to
8125 escape regexp metacharacters. Therefore, to make sure that a backslash
8126 is preserved in a string in your Guile program, you must use *two*
8127 consecutive backslashes:
8129 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
8131 The string in this example is preprocessed by the Guile reader before
8132 any code is executed. The resulting argument to `make-regexp' is the
8133 string `^\* [^:]*', which is what we really want.
8135 This also means that in order to write a regular expression that
8136 matches a single backslash character, the regular expression string in
8137 the source code must include *four* backslashes. Each consecutive pair
8138 of backslashes gets translated by the Guile reader to a single
8139 backslash, and the resulting double-backslash is interpreted by the
8140 regexp engine as matching a single backslash character. Hence:
8142 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
8144 The reason for the unwieldiness of this syntax is historical. Both
8145 regular expression pattern matchers and Unix string processing systems
8146 have traditionally used backslashes with the special meanings described
8147 above. The POSIX regular expression specification and ANSI C standard
8148 both require these semantics. Attempting to abandon either convention
8149 would cause other kinds of compatibility problems, possibly more severe
8150 ones. Therefore, without extending the Scheme reader to support
8151 strings with different quoting conventions (an ungainly and confusing
8152 extension when implemented in other languages), we must adhere to this
8153 cumbersome escape syntax.
8155 * Changes to the gh_ interface
8157 * Changes to the scm_ interface
8159 * Changes to system call interfaces:
8161 ** The value returned by `raise' is now unspecified. It throws an exception
8164 *** A new procedure `sigaction' can be used to install signal handlers
8166 (sigaction signum [action] [flags])
8168 signum is the signal number, which can be specified using the value
8171 If action is omitted, sigaction returns a pair: the CAR is the current
8172 signal hander, which will be either an integer with the value SIG_DFL
8173 (default action) or SIG_IGN (ignore), or the Scheme procedure which
8174 handles the signal, or #f if a non-Scheme procedure handles the
8175 signal. The CDR contains the current sigaction flags for the handler.
8177 If action is provided, it is installed as the new handler for signum.
8178 action can be a Scheme procedure taking one argument, or the value of
8179 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
8180 whatever signal handler was installed before sigaction was first used.
8181 Flags can optionally be specified for the new handler (SA_RESTART is
8182 always used if the system provides it, so need not be specified.) The
8183 return value is a pair with information about the old handler as
8186 This interface does not provide access to the "signal blocking"
8187 facility. Maybe this is not needed, since the thread support may
8188 provide solutions to the problem of consistent access to data
8191 *** A new procedure `flush-all-ports' is equivalent to running
8192 `force-output' on every port open for output.
8194 ** Guile now provides information on how it was built, via the new
8195 global variable, %guile-build-info. This variable records the values
8196 of the standard GNU makefile directory variables as an assocation
8197 list, mapping variable names (symbols) onto directory paths (strings).
8198 For example, to find out where the Guile link libraries were
8199 installed, you can say:
8201 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
8204 * Changes to the scm_ interface
8206 ** The new function scm_handle_by_message_noexit is just like the
8207 existing scm_handle_by_message function, except that it doesn't call
8208 exit to terminate the process. Instead, it prints a message and just
8209 returns #f. This might be a more appropriate catch-all handler for
8210 new dynamic roots and threads.
8213 Changes in Guile 1.1 (released Friday, May 16 1997):
8215 * Changes to the distribution.
8217 The Guile 1.0 distribution has been split up into several smaller
8219 guile-core --- the Guile interpreter itself.
8220 guile-tcltk --- the interface between the Guile interpreter and
8221 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
8222 is a toolkit for building graphical user interfaces.
8223 guile-rgx-ctax --- the interface between Guile and the Rx regular
8224 expression matcher, and the translator for the Ctax
8225 programming language. These are packaged together because the
8226 Ctax translator uses Rx to parse Ctax source code.
8228 This NEWS file describes the changes made to guile-core since the 1.0
8231 We no longer distribute the documentation, since it was either out of
8232 date, or incomplete. As soon as we have current documentation, we
8237 * Changes to the stand-alone interpreter
8239 ** guile now accepts command-line arguments compatible with SCSH, Olin
8240 Shivers' Scheme Shell.
8242 In general, arguments are evaluated from left to right, but there are
8243 exceptions. The following switches stop argument processing, and
8244 stash all remaining command-line arguments as the value returned by
8245 the (command-line) function.
8246 -s SCRIPT load Scheme source code from FILE, and exit
8247 -c EXPR evalute Scheme expression EXPR, and exit
8248 -- stop scanning arguments; run interactively
8250 The switches below are processed as they are encountered.
8251 -l FILE load Scheme source code from FILE
8252 -e FUNCTION after reading script, apply FUNCTION to
8253 command line arguments
8254 -ds do -s script at this point
8255 --emacs enable Emacs protocol (experimental)
8256 -h, --help display this help and exit
8257 -v, --version display version information and exit
8258 \ read arguments from following script lines
8260 So, for example, here is a Guile script named `ekko' (thanks, Olin)
8261 which re-implements the traditional "echo" command:
8263 #!/usr/local/bin/guile -s
8266 (map (lambda (arg) (display arg) (display " "))
8270 (main (command-line))
8272 Suppose we invoke this script as follows:
8274 ekko a speckled gecko
8276 Through the magic of Unix script processing (triggered by the `#!'
8277 token at the top of the file), /usr/local/bin/guile receives the
8278 following list of command-line arguments:
8280 ("-s" "./ekko" "a" "speckled" "gecko")
8282 Unix inserts the name of the script after the argument specified on
8283 the first line of the file (in this case, "-s"), and then follows that
8284 with the arguments given to the script. Guile loads the script, which
8285 defines the `main' function, and then applies it to the list of
8286 remaining command-line arguments, ("a" "speckled" "gecko").
8288 In Unix, the first line of a script file must take the following form:
8290 #!INTERPRETER ARGUMENT
8292 where INTERPRETER is the absolute filename of the interpreter
8293 executable, and ARGUMENT is a single command-line argument to pass to
8296 You may only pass one argument to the interpreter, and its length is
8297 limited. These restrictions can be annoying to work around, so Guile
8298 provides a general mechanism (borrowed from, and compatible with,
8299 SCSH) for circumventing them.
8301 If the ARGUMENT in a Guile script is a single backslash character,
8302 `\', Guile will open the script file, parse arguments from its second
8303 and subsequent lines, and replace the `\' with them. So, for example,
8304 here is another implementation of the `ekko' script:
8306 #!/usr/local/bin/guile \
8310 (for-each (lambda (arg) (display arg) (display " "))
8314 If the user invokes this script as follows:
8316 ekko a speckled gecko
8318 Unix expands this into
8320 /usr/local/bin/guile \ ekko a speckled gecko
8322 When Guile sees the `\' argument, it replaces it with the arguments
8323 read from the second line of the script, producing:
8325 /usr/local/bin/guile -e main -s ekko a speckled gecko
8327 This tells Guile to load the `ekko' script, and apply the function
8328 `main' to the argument list ("a" "speckled" "gecko").
8330 Here is how Guile parses the command-line arguments:
8331 - Each space character terminates an argument. This means that two
8332 spaces in a row introduce an empty-string argument.
8333 - The tab character is not permitted (unless you quote it with the
8334 backslash character, as described below), to avoid confusion.
8335 - The newline character terminates the sequence of arguments, and will
8336 also terminate a final non-empty argument. (However, a newline
8337 following a space will not introduce a final empty-string argument;
8338 it only terminates the argument list.)
8339 - The backslash character is the escape character. It escapes
8340 backslash, space, tab, and newline. The ANSI C escape sequences
8341 like \n and \t are also supported. These produce argument
8342 constituents; the two-character combination \n doesn't act like a
8343 terminating newline. The escape sequence \NNN for exactly three
8344 octal digits reads as the character whose ASCII code is NNN. As
8345 above, characters produced this way are argument constituents.
8346 Backslash followed by other characters is not allowed.
8348 * Changes to the procedure for linking libguile with your programs
8350 ** Guile now builds and installs a shared guile library, if your
8351 system support shared libraries. (It still builds a static library on
8352 all systems.) Guile automatically detects whether your system
8353 supports shared libraries. To prevent Guile from buildisg shared
8354 libraries, pass the `--disable-shared' flag to the configure script.
8356 Guile takes longer to compile when it builds shared libraries, because
8357 it must compile every file twice --- once to produce position-
8358 independent object code, and once to produce normal object code.
8360 ** The libthreads library has been merged into libguile.
8362 To link a program against Guile, you now need only link against
8363 -lguile and -lqt; -lthreads is no longer needed. If you are using
8364 autoconf to generate configuration scripts for your application, the
8365 following lines should suffice to add the appropriate libraries to
8368 ### Find quickthreads and libguile.
8369 AC_CHECK_LIB(qt, main)
8370 AC_CHECK_LIB(guile, scm_shell)
8372 * Changes to Scheme functions
8374 ** Guile Scheme's special syntax for keyword objects is now optional,
8375 and disabled by default.
8377 The syntax variation from R4RS made it difficult to port some
8378 interesting packages to Guile. The routines which accepted keyword
8379 arguments (mostly in the module system) have been modified to also
8380 accept symbols whose names begin with `:'.
8382 To change the keyword syntax, you must first import the (ice-9 debug)
8384 (use-modules (ice-9 debug))
8386 Then you can enable the keyword syntax as follows:
8387 (read-set! keywords 'prefix)
8389 To disable keyword syntax, do this:
8390 (read-set! keywords #f)
8392 ** Many more primitive functions accept shared substrings as
8393 arguments. In the past, these functions required normal, mutable
8394 strings as arguments, although they never made use of this
8397 ** The uniform array functions now operate on byte vectors. These
8398 functions are `array-fill!', `serial-array-copy!', `array-copy!',
8399 `serial-array-map', `array-map', `array-for-each', and
8402 ** The new functions `trace' and `untrace' implement simple debugging
8403 support for Scheme functions.
8405 The `trace' function accepts any number of procedures as arguments,
8406 and tells the Guile interpreter to display each procedure's name and
8407 arguments each time the procedure is invoked. When invoked with no
8408 arguments, `trace' returns the list of procedures currently being
8411 The `untrace' function accepts any number of procedures as arguments,
8412 and tells the Guile interpreter not to trace them any more. When
8413 invoked with no arguments, `untrace' untraces all curretly traced
8416 The tracing in Guile has an advantage over most other systems: we
8417 don't create new procedure objects, but mark the procedure objects
8418 themselves. This means that anonymous and internal procedures can be
8421 ** The function `assert-repl-prompt' has been renamed to
8422 `set-repl-prompt!'. It takes one argument, PROMPT.
8423 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
8424 - If PROMPT is a string, we use it as a prompt.
8425 - If PROMPT is a procedure accepting no arguments, we call it, and
8426 display the result as a prompt.
8427 - Otherwise, we display "> ".
8429 ** The new function `eval-string' reads Scheme expressions from a
8430 string and evaluates them, returning the value of the last expression
8431 in the string. If the string contains no expressions, it returns an
8434 ** The new function `thunk?' returns true iff its argument is a
8435 procedure of zero arguments.
8437 ** `defined?' is now a builtin function, instead of syntax. This
8438 means that its argument should be quoted. It returns #t iff its
8439 argument is bound in the current module.
8441 ** The new syntax `use-modules' allows you to add new modules to your
8442 environment without re-typing a complete `define-module' form. It
8443 accepts any number of module names as arguments, and imports their
8444 public bindings into the current module.
8446 ** The new function (module-defined? NAME MODULE) returns true iff
8447 NAME, a symbol, is defined in MODULE, a module object.
8449 ** The new function `builtin-bindings' creates and returns a hash
8450 table containing copies of all the root module's bindings.
8452 ** The new function `builtin-weak-bindings' does the same as
8453 `builtin-bindings', but creates a doubly-weak hash table.
8455 ** The `equal?' function now considers variable objects to be
8456 equivalent if they have the same name and the same value.
8458 ** The new function `command-line' returns the command-line arguments
8459 given to Guile, as a list of strings.
8461 When using guile as a script interpreter, `command-line' returns the
8462 script's arguments; those processed by the interpreter (like `-s' or
8463 `-c') are omitted. (In other words, you get the normal, expected
8464 behavior.) Any application that uses scm_shell to process its
8465 command-line arguments gets this behavior as well.
8467 ** The new function `load-user-init' looks for a file called `.guile'
8468 in the user's home directory, and loads it if it exists. This is
8469 mostly for use by the code generated by scm_compile_shell_switches,
8470 but we thought it might also be useful in other circumstances.
8472 ** The new function `log10' returns the base-10 logarithm of its
8475 ** Changes to I/O functions
8477 *** The functions `read', `primitive-load', `read-and-eval!', and
8478 `primitive-load-path' no longer take optional arguments controlling
8479 case insensitivity and a `#' parser.
8481 Case sensitivity is now controlled by a read option called
8482 `case-insensitive'. The user can add new `#' syntaxes with the
8483 `read-hash-extend' function (see below).
8485 *** The new function `read-hash-extend' allows the user to change the
8486 syntax of Guile Scheme in a somewhat controlled way.
8488 (read-hash-extend CHAR PROC)
8489 When parsing S-expressions, if we read a `#' character followed by
8490 the character CHAR, use PROC to parse an object from the stream.
8491 If PROC is #f, remove any parsing procedure registered for CHAR.
8493 The reader applies PROC to two arguments: CHAR and an input port.
8495 *** The new functions read-delimited and read-delimited! provide a
8496 general mechanism for doing delimited input on streams.
8498 (read-delimited DELIMS [PORT HANDLE-DELIM])
8499 Read until we encounter one of the characters in DELIMS (a string),
8500 or end-of-file. PORT is the input port to read from; it defaults to
8501 the current input port. The HANDLE-DELIM parameter determines how
8502 the terminating character is handled; it should be one of the
8505 'trim omit delimiter from result
8506 'peek leave delimiter character in input stream
8507 'concat append delimiter character to returned value
8508 'split return a pair: (RESULT . TERMINATOR)
8510 HANDLE-DELIM defaults to 'peek.
8512 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
8513 A side-effecting variant of `read-delimited'.
8515 The data is written into the string BUF at the indices in the
8516 half-open interval [START, END); the default interval is the whole
8517 string: START = 0 and END = (string-length BUF). The values of
8518 START and END must specify a well-defined interval in BUF, i.e.
8519 0 <= START <= END <= (string-length BUF).
8521 It returns NBYTES, the number of bytes read. If the buffer filled
8522 up without a delimiter character being found, it returns #f. If the
8523 port is at EOF when the read starts, it returns the EOF object.
8525 If an integer is returned (i.e., the read is successfully terminated
8526 by reading a delimiter character), then the HANDLE-DELIM parameter
8527 determines how to handle the terminating character. It is described
8528 above, and defaults to 'peek.
8530 (The descriptions of these functions were borrowed from the SCSH
8531 manual, by Olin Shivers and Brian Carlstrom.)
8533 *** The `%read-delimited!' function is the primitive used to implement
8534 `read-delimited' and `read-delimited!'.
8536 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
8538 This returns a pair of values: (TERMINATOR . NUM-READ).
8539 - TERMINATOR describes why the read was terminated. If it is a
8540 character or the eof object, then that is the value that terminated
8541 the read. If it is #f, the function filled the buffer without finding
8542 a delimiting character.
8543 - NUM-READ is the number of characters read into BUF.
8545 If the read is successfully terminated by reading a delimiter
8546 character, then the gobble? parameter determines what to do with the
8547 terminating character. If true, the character is removed from the
8548 input stream; if false, the character is left in the input stream
8549 where a subsequent read operation will retrieve it. In either case,
8550 the character is also the first value returned by the procedure call.
8552 (The descriptions of this function was borrowed from the SCSH manual,
8553 by Olin Shivers and Brian Carlstrom.)
8555 *** The `read-line' and `read-line!' functions have changed; they now
8556 trim the terminator by default; previously they appended it to the
8557 returned string. For the old behavior, use (read-line PORT 'concat).
8559 *** The functions `uniform-array-read!' and `uniform-array-write!' now
8560 take new optional START and END arguments, specifying the region of
8561 the array to read and write.
8563 *** The `ungetc-char-ready?' function has been removed. We feel it's
8564 inappropriate for an interface to expose implementation details this
8567 ** Changes to the Unix library and system call interface
8569 *** The new fcntl function provides access to the Unix `fcntl' system
8572 (fcntl PORT COMMAND VALUE)
8573 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
8574 Values for COMMAND are:
8576 F_DUPFD duplicate a file descriptor
8577 F_GETFD read the descriptor's close-on-exec flag
8578 F_SETFD set the descriptor's close-on-exec flag to VALUE
8579 F_GETFL read the descriptor's flags, as set on open
8580 F_SETFL set the descriptor's flags, as set on open to VALUE
8581 F_GETOWN return the process ID of a socket's owner, for SIGIO
8582 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
8583 FD_CLOEXEC not sure what this is
8585 For details, see the documentation for the fcntl system call.
8587 *** The arguments to `select' have changed, for compatibility with
8588 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
8589 expected behavior. The MILLISECONDS parameter has been changed to
8590 MICROSECONDS, to more closely resemble the underlying system call.
8591 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
8592 corresponding return set will be the same.
8594 *** The arguments to the `mknod' system call have changed. They are
8597 (mknod PATH TYPE PERMS DEV)
8598 Create a new file (`node') in the file system. PATH is the name of
8599 the file to create. TYPE is the kind of file to create; it should
8600 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
8601 permission bits to give the newly created file. If TYPE is
8602 'block-special or 'char-special, DEV specifies which device the
8603 special file refers to; its interpretation depends on the kind of
8604 special file being created.
8606 *** The `fork' function has been renamed to `primitive-fork', to avoid
8607 clashing with various SCSH forks.
8609 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
8610 and `recvfrom!'. They no longer accept a size for a second argument;
8611 you must pass a string to hold the received value. They no longer
8612 return the buffer. Instead, `recv' returns the length of the message
8613 received, and `recvfrom' returns a pair containing the packet's length
8614 and originating address.
8616 *** The file descriptor datatype has been removed, as have the
8617 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
8618 We plan to replace these functions with a SCSH-compatible interface.
8620 *** The `create' function has been removed; it's just a special case
8623 *** There are new functions to break down process termination status
8624 values. In the descriptions below, STATUS is a value returned by
8627 (status:exit-val STATUS)
8628 If the child process exited normally, this function returns the exit
8629 code for the child process (i.e., the value passed to exit, or
8630 returned from main). If the child process did not exit normally,
8631 this function returns #f.
8633 (status:stop-sig STATUS)
8634 If the child process was suspended by a signal, this function
8635 returns the signal that suspended the child. Otherwise, it returns
8638 (status:term-sig STATUS)
8639 If the child process terminated abnormally, this function returns
8640 the signal that terminated the child. Otherwise, this function
8643 POSIX promises that exactly one of these functions will return true on
8644 a valid STATUS value.
8646 These functions are compatible with SCSH.
8648 *** There are new accessors and setters for the broken-out time vectors
8649 returned by `localtime', `gmtime', and that ilk. They are:
8651 Component Accessor Setter
8652 ========================= ============ ============
8653 seconds tm:sec set-tm:sec
8654 minutes tm:min set-tm:min
8655 hours tm:hour set-tm:hour
8656 day of the month tm:mday set-tm:mday
8657 month tm:mon set-tm:mon
8658 year tm:year set-tm:year
8659 day of the week tm:wday set-tm:wday
8660 day in the year tm:yday set-tm:yday
8661 daylight saving time tm:isdst set-tm:isdst
8662 GMT offset, seconds tm:gmtoff set-tm:gmtoff
8663 name of time zone tm:zone set-tm:zone
8665 *** There are new accessors for the vectors returned by `uname',
8666 describing the host system:
8669 ============================================== ================
8670 name of the operating system implementation utsname:sysname
8671 network name of this machine utsname:nodename
8672 release level of the operating system utsname:release
8673 version level of the operating system utsname:version
8674 machine hardware platform utsname:machine
8676 *** There are new accessors for the vectors returned by `getpw',
8677 `getpwnam', `getpwuid', and `getpwent', describing entries from the
8678 system's user database:
8681 ====================== =================
8682 user name passwd:name
8683 user password passwd:passwd
8686 real name passwd:gecos
8687 home directory passwd:dir
8688 shell program passwd:shell
8690 *** There are new accessors for the vectors returned by `getgr',
8691 `getgrnam', `getgrgid', and `getgrent', describing entries from the
8692 system's group database:
8695 ======================= ============
8696 group name group:name
8697 group password group:passwd
8699 group members group:mem
8701 *** There are new accessors for the vectors returned by `gethost',
8702 `gethostbyaddr', `gethostbyname', and `gethostent', describing
8706 ========================= ===============
8707 official name of host hostent:name
8708 alias list hostent:aliases
8709 host address type hostent:addrtype
8710 length of address hostent:length
8711 list of addresses hostent:addr-list
8713 *** There are new accessors for the vectors returned by `getnet',
8714 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
8718 ========================= ===============
8719 official name of net netent:name
8720 alias list netent:aliases
8721 net number type netent:addrtype
8722 net number netent:net
8724 *** There are new accessors for the vectors returned by `getproto',
8725 `getprotobyname', `getprotobynumber', and `getprotoent', describing
8729 ========================= ===============
8730 official protocol name protoent:name
8731 alias list protoent:aliases
8732 protocol number protoent:proto
8734 *** There are new accessors for the vectors returned by `getserv',
8735 `getservbyname', `getservbyport', and `getservent', describing
8739 ========================= ===============
8740 official service name servent:name
8741 alias list servent:aliases
8742 port number servent:port
8743 protocol to use servent:proto
8745 *** There are new accessors for the sockaddr structures returned by
8746 `accept', `getsockname', `getpeername', `recvfrom!':
8749 ======================================== ===============
8750 address format (`family') sockaddr:fam
8751 path, for file domain addresses sockaddr:path
8752 address, for internet domain addresses sockaddr:addr
8753 TCP or UDP port, for internet sockaddr:port
8755 *** The `getpwent', `getgrent', `gethostent', `getnetent',
8756 `getprotoent', and `getservent' functions now return #f at the end of
8757 the user database. (They used to throw an exception.)
8759 Note that calling MUMBLEent function is equivalent to calling the
8760 corresponding MUMBLE function with no arguments.
8762 *** The `setpwent', `setgrent', `sethostent', `setnetent',
8763 `setprotoent', and `setservent' routines now take no arguments.
8765 *** The `gethost', `getproto', `getnet', and `getserv' functions now
8766 provide more useful information when they throw an exception.
8768 *** The `lnaof' function has been renamed to `inet-lnaof'.
8770 *** Guile now claims to have the `current-time' feature.
8772 *** The `mktime' function now takes an optional second argument ZONE,
8773 giving the time zone to use for the conversion. ZONE should be a
8774 string, in the same format as expected for the "TZ" environment variable.
8776 *** The `strptime' function now returns a pair (TIME . COUNT), where
8777 TIME is the parsed time as a vector, and COUNT is the number of
8778 characters from the string left unparsed. This function used to
8779 return the remaining characters as a string.
8781 *** The `gettimeofday' function has replaced the old `time+ticks' function.
8782 The return value is now (SECONDS . MICROSECONDS); the fractional
8783 component is no longer expressed in "ticks".
8785 *** The `ticks/sec' constant has been removed, in light of the above change.
8787 * Changes to the gh_ interface
8789 ** gh_eval_str() now returns an SCM object which is the result of the
8792 ** gh_scm2str() now copies the Scheme data to a caller-provided C
8795 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
8796 and returns the array
8798 ** gh_scm2str0() is gone: there is no need to distinguish
8799 null-terminated from non-null-terminated, since gh_scm2newstr() allows
8800 the user to interpret the data both ways.
8802 * Changes to the scm_ interface
8804 ** The new function scm_symbol_value0 provides an easy way to get a
8805 symbol's value from C code:
8807 SCM scm_symbol_value0 (char *NAME)
8808 Return the value of the symbol named by the null-terminated string
8809 NAME in the current module. If the symbol named NAME is unbound in
8810 the current module, return SCM_UNDEFINED.
8812 ** The new function scm_sysintern0 creates new top-level variables,
8813 without assigning them a value.
8815 SCM scm_sysintern0 (char *NAME)
8816 Create a new Scheme top-level variable named NAME. NAME is a
8817 null-terminated string. Return the variable's value cell.
8819 ** The function scm_internal_catch is the guts of catch. It handles
8820 all the mechanics of setting up a catch target, invoking the catch
8821 body, and perhaps invoking the handler if the body does a throw.
8823 The function is designed to be usable from C code, but is general
8824 enough to implement all the semantics Guile Scheme expects from throw.
8826 TAG is the catch tag. Typically, this is a symbol, but this function
8827 doesn't actually care about that.
8829 BODY is a pointer to a C function which runs the body of the catch;
8830 this is the code you can throw from. We call it like this:
8831 BODY (BODY_DATA, JMPBUF)
8833 BODY_DATA is just the BODY_DATA argument we received; we pass it
8834 through to BODY as its first argument. The caller can make
8835 BODY_DATA point to anything useful that BODY might need.
8836 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
8837 which we have just created and initialized.
8839 HANDLER is a pointer to a C function to deal with a throw to TAG,
8840 should one occur. We call it like this:
8841 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
8843 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
8844 same idea as BODY_DATA above.
8845 THROWN_TAG is the tag that the user threw to; usually this is
8846 TAG, but it could be something else if TAG was #t (i.e., a
8847 catch-all), or the user threw to a jmpbuf.
8848 THROW_ARGS is the list of arguments the user passed to the THROW
8851 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
8852 is just a pointer we pass through to HANDLER. We don't actually
8853 use either of those pointers otherwise ourselves. The idea is
8854 that, if our caller wants to communicate something to BODY or
8855 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
8856 HANDLER can then use. Think of it as a way to make BODY and
8857 HANDLER closures, not just functions; MUMBLE_DATA points to the
8860 Of course, it's up to the caller to make sure that any data a
8861 MUMBLE_DATA needs is protected from GC. A common way to do this is
8862 to make MUMBLE_DATA a pointer to data stored in an automatic
8863 structure variable; since the collector must scan the stack for
8864 references anyway, this assures that any references in MUMBLE_DATA
8867 ** The new function scm_internal_lazy_catch is exactly like
8868 scm_internal_catch, except:
8870 - It does not unwind the stack (this is the major difference).
8871 - If handler returns, its value is returned from the throw.
8872 - BODY always receives #f as its JMPBUF argument (since there's no
8873 jmpbuf associated with a lazy catch, because we don't unwind the
8876 ** scm_body_thunk is a new body function you can pass to
8877 scm_internal_catch if you want the body to be like Scheme's `catch'
8878 --- a thunk, or a function of one argument if the tag is #f.
8880 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
8881 contains the Scheme procedure to invoke as the body, and the tag
8882 we're catching. If the tag is #f, then we pass JMPBUF (created by
8883 scm_internal_catch) to the body procedure; otherwise, the body gets
8886 ** scm_handle_by_proc is a new handler function you can pass to
8887 scm_internal_catch if you want the handler to act like Scheme's catch
8888 --- call a procedure with the tag and the throw arguments.
8890 If the user does a throw to this catch, this function runs a handler
8891 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
8892 variable holding the Scheme procedure object to invoke. It ought to
8893 be a pointer to an automatic variable (i.e., one living on the stack),
8894 or the procedure object should be otherwise protected from GC.
8896 ** scm_handle_by_message is a new handler function to use with
8897 `scm_internal_catch' if you want Guile to print a message and die.
8898 It's useful for dealing with throws to uncaught keys at the top level.
8900 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
8901 message header to print; if zero, we use "guile" instead. That
8902 text is followed by a colon, then the message described by ARGS.
8904 ** The return type of scm_boot_guile is now void; the function does
8905 not return a value, and indeed, never returns at all.
8907 ** The new function scm_shell makes it easy for user applications to
8908 process command-line arguments in a way that is compatible with the
8909 stand-alone guile interpreter (which is in turn compatible with SCSH,
8912 To use the scm_shell function, first initialize any guile modules
8913 linked into your application, and then call scm_shell with the values
8914 of ARGC and ARGV your `main' function received. scm_shell will add
8915 any SCSH-style meta-arguments from the top of the script file to the
8916 argument vector, and then process the command-line arguments. This
8917 generally means loading a script file or starting up an interactive
8918 command interpreter. For details, see "Changes to the stand-alone
8921 ** The new functions scm_get_meta_args and scm_count_argv help you
8922 implement the SCSH-style meta-argument, `\'.
8924 char **scm_get_meta_args (int ARGC, char **ARGV)
8925 If the second element of ARGV is a string consisting of a single
8926 backslash character (i.e. "\\" in Scheme notation), open the file
8927 named by the following argument, parse arguments from it, and return
8928 the spliced command line. The returned array is terminated by a
8931 For details of argument parsing, see above, under "guile now accepts
8932 command-line arguments compatible with SCSH..."
8934 int scm_count_argv (char **ARGV)
8935 Count the arguments in ARGV, assuming it is terminated by a null
8938 For an example of how these functions might be used, see the source
8939 code for the function scm_shell in libguile/script.c.
8941 You will usually want to use scm_shell instead of calling this
8944 ** The new function scm_compile_shell_switches turns an array of
8945 command-line arguments into Scheme code to carry out the actions they
8946 describe. Given ARGC and ARGV, it returns a Scheme expression to
8947 evaluate, and calls scm_set_program_arguments to make any remaining
8948 command-line arguments available to the Scheme code. For example,
8949 given the following arguments:
8951 -e main -s ekko a speckled gecko
8953 scm_set_program_arguments will return the following expression:
8955 (begin (load "ekko") (main (command-line)) (quit))
8957 You will usually want to use scm_shell instead of calling this
8960 ** The function scm_shell_usage prints a usage message appropriate for
8961 an interpreter that uses scm_compile_shell_switches to handle its
8962 command-line arguments.
8964 void scm_shell_usage (int FATAL, char *MESSAGE)
8965 Print a usage message to the standard error output. If MESSAGE is
8966 non-zero, write it before the usage message, followed by a newline.
8967 If FATAL is non-zero, exit the process, using FATAL as the
8968 termination status. (If you want to be compatible with Guile,
8969 always use 1 as the exit status when terminating due to command-line
8972 You will usually want to use scm_shell instead of calling this
8975 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
8976 expressions. It used to return SCM_EOL. Earth-shattering.
8978 ** The macros for declaring scheme objects in C code have been
8979 rearranged slightly. They are now:
8981 SCM_SYMBOL (C_NAME, SCHEME_NAME)
8982 Declare a static SCM variable named C_NAME, and initialize it to
8983 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
8984 be a C identifier, and SCHEME_NAME should be a C string.
8986 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
8987 Just like SCM_SYMBOL, but make C_NAME globally visible.
8989 SCM_VCELL (C_NAME, SCHEME_NAME)
8990 Create a global variable at the Scheme level named SCHEME_NAME.
8991 Declare a static SCM variable named C_NAME, and initialize it to
8992 point to the Scheme variable's value cell.
8994 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
8995 Just like SCM_VCELL, but make C_NAME globally visible.
8997 The `guile-snarf' script writes initialization code for these macros
8998 to its standard output, given C source code as input.
9000 The SCM_GLOBAL macro is gone.
9002 ** The scm_read_line and scm_read_line_x functions have been replaced
9003 by Scheme code based on the %read-delimited! procedure (known to C
9004 code as scm_read_delimited_x). See its description above for more
9007 ** The function scm_sys_open has been renamed to scm_open. It now
9008 returns a port instead of an FD object.
9010 * The dynamic linking support has changed. For more information, see
9011 libguile/DYNAMIC-LINKING.
9016 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
9019 * Changes to the 'guile' program:
9021 ** Guile now loads some new files when it starts up. Guile first
9022 searches the load path for init.scm, and loads it if found. Then, if
9023 Guile is not being used to execute a script, and the user's home
9024 directory contains a file named `.guile', Guile loads that.
9026 ** You can now use Guile as a shell script interpreter.
9028 To paraphrase the SCSH manual:
9030 When Unix tries to execute an executable file whose first two
9031 characters are the `#!', it treats the file not as machine code to
9032 be directly executed by the native processor, but as source code
9033 to be executed by some interpreter. The interpreter to use is
9034 specified immediately after the #! sequence on the first line of
9035 the source file. The kernel reads in the name of the interpreter,
9036 and executes that instead. It passes the interpreter the source
9037 filename as its first argument, with the original arguments
9038 following. Consult the Unix man page for the `exec' system call
9039 for more information.
9041 Now you can use Guile as an interpreter, using a mechanism which is a
9042 compatible subset of that provided by SCSH.
9044 Guile now recognizes a '-s' command line switch, whose argument is the
9045 name of a file of Scheme code to load. It also treats the two
9046 characters `#!' as the start of a comment, terminated by `!#'. Thus,
9047 to make a file of Scheme code directly executable by Unix, insert the
9048 following two lines at the top of the file:
9050 #!/usr/local/bin/guile -s
9053 Guile treats the argument of the `-s' command-line switch as the name
9054 of a file of Scheme code to load, and treats the sequence `#!' as the
9055 start of a block comment, terminated by `!#'.
9057 For example, here's a version of 'echo' written in Scheme:
9059 #!/usr/local/bin/guile -s
9061 (let loop ((args (cdr (program-arguments))))
9064 (display (car args))
9065 (if (pair? (cdr args))
9067 (loop (cdr args)))))
9070 Why does `#!' start a block comment terminated by `!#', instead of the
9071 end of the line? That is the notation SCSH uses, and although we
9072 don't yet support the other SCSH features that motivate that choice,
9073 we would like to be backward-compatible with any existing Guile
9074 scripts once we do. Furthermore, if the path to Guile on your system
9075 is too long for your kernel, you can start the script with this
9079 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
9082 Note that some very old Unix systems don't support the `#!' syntax.
9085 ** You can now run Guile without installing it.
9087 Previous versions of the interactive Guile interpreter (`guile')
9088 couldn't start up unless Guile's Scheme library had been installed;
9089 they used the value of the environment variable `SCHEME_LOAD_PATH'
9090 later on in the startup process, but not to find the startup code
9091 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
9094 To run Guile without installing it, build it in the normal way, and
9095 then set the environment variable `SCHEME_LOAD_PATH' to a
9096 colon-separated list of directories, including the top-level directory
9097 of the Guile sources. For example, if you unpacked Guile so that the
9098 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
9101 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
9104 ** Guile's read-eval-print loop no longer prints #<unspecified>
9105 results. If the user wants to see this, she can evaluate the
9106 expression (assert-repl-print-unspecified #t), perhaps in her startup
9109 ** Guile no longer shows backtraces by default when an error occurs;
9110 however, it does display a message saying how to get one, and how to
9111 request that they be displayed by default. After an error, evaluate
9113 to see a backtrace, and
9114 (debug-enable 'backtrace)
9115 to see them by default.
9119 * Changes to Guile Scheme:
9121 ** Guile now distinguishes between #f and the empty list.
9123 This is for compatibility with the IEEE standard, the (possibly)
9124 upcoming Revised^5 Report on Scheme, and many extant Scheme
9127 Guile used to have #f and '() denote the same object, to make Scheme's
9128 type system more compatible with Emacs Lisp's. However, the change
9129 caused too much trouble for Scheme programmers, and we found another
9130 way to reconcile Emacs Lisp with Scheme that didn't require this.
9133 ** Guile's delq, delv, delete functions, and their destructive
9134 counterparts, delq!, delv!, and delete!, now remove all matching
9135 elements from the list, not just the first. This matches the behavior
9136 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
9137 functions which inspired them.
9139 I recognize that this change may break code in subtle ways, but it
9140 seems best to make the change before the FSF's first Guile release,
9144 ** The compiled-library-path function has been deleted from libguile.
9146 ** The facilities for loading Scheme source files have changed.
9148 *** The variable %load-path now tells Guile which directories to search
9149 for Scheme code. Its value is a list of strings, each of which names
9152 *** The variable %load-extensions now tells Guile which extensions to
9153 try appending to a filename when searching the load path. Its value
9154 is a list of strings. Its default value is ("" ".scm").
9156 *** (%search-load-path FILENAME) searches the directories listed in the
9157 value of the %load-path variable for a Scheme file named FILENAME,
9158 with all the extensions listed in %load-extensions. If it finds a
9159 match, then it returns its full filename. If FILENAME is absolute, it
9160 returns it unchanged. Otherwise, it returns #f.
9162 %search-load-path will not return matches that refer to directories.
9164 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
9165 uses %seach-load-path to find a file named FILENAME, and loads it if
9166 it finds it. If it can't read FILENAME for any reason, it throws an
9169 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
9172 *** load uses the same searching semantics as primitive-load.
9174 *** The functions %try-load, try-load-with-path, %load, load-with-path,
9175 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
9176 path, and load-module-with-path have been deleted. The functions
9177 above should serve their purposes.
9179 *** If the value of the variable %load-hook is a procedure,
9180 `primitive-load' applies its value to the name of the file being
9181 loaded (without the load path directory name prepended). If its value
9182 is #f, it is ignored. Otherwise, an error occurs.
9184 This is mostly useful for printing load notification messages.
9187 ** The function `eval!' is no longer accessible from the scheme level.
9188 We can't allow operations which introduce glocs into the scheme level,
9189 because Guile's type system can't handle these as data. Use `eval' or
9190 `read-and-eval!' (see below) as replacement.
9192 ** The new function read-and-eval! reads an expression from PORT,
9193 evaluates it, and returns the result. This is more efficient than
9194 simply calling `read' and `eval', since it is not necessary to make a
9195 copy of the expression for the evaluator to munge.
9197 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
9198 for the `read' function.
9201 ** The function `int?' has been removed; its definition was identical
9202 to that of `integer?'.
9204 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
9205 use the R4RS names for these functions.
9207 ** The function object-properties no longer returns the hash handle;
9208 it simply returns the object's property list.
9210 ** Many functions have been changed to throw errors, instead of
9211 returning #f on failure. The point of providing exception handling in
9212 the language is to simplify the logic of user code, but this is less
9213 useful if Guile's primitives don't throw exceptions.
9215 ** The function `fileno' has been renamed from `%fileno'.
9217 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
9220 * Changes to Guile's C interface:
9222 ** The library's initialization procedure has been simplified.
9223 scm_boot_guile now has the prototype:
9225 void scm_boot_guile (int ARGC,
9227 void (*main_func) (),
9230 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
9231 MAIN_FUNC should do all the work of the program (initializing other
9232 packages, reading user input, etc.) before returning. When MAIN_FUNC
9233 returns, call exit (0); this function never returns. If you want some
9234 other exit value, MAIN_FUNC may call exit itself.
9236 scm_boot_guile arranges for program-arguments to return the strings
9237 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
9238 scm_set_program_arguments with the final list, so Scheme code will
9239 know which arguments have been processed.
9241 scm_boot_guile establishes a catch-all catch handler which prints an
9242 error message and exits the process. This means that Guile exits in a
9243 coherent way when system errors occur and the user isn't prepared to
9244 handle it. If the user doesn't like this behavior, they can establish
9245 their own universal catcher in MAIN_FUNC to shadow this one.
9247 Why must the caller do all the real work from MAIN_FUNC? The garbage
9248 collector assumes that all local variables of type SCM will be above
9249 scm_boot_guile's stack frame on the stack. If you try to manipulate
9250 SCM values after this function returns, it's the luck of the draw
9251 whether the GC will be able to find the objects you allocate. So,
9252 scm_boot_guile function exits, rather than returning, to discourage
9253 people from making that mistake.
9255 The IN, OUT, and ERR arguments were removed; there are other
9256 convenient ways to override these when desired.
9258 The RESULT argument was deleted; this function should never return.
9260 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
9264 ** Guile's header files should no longer conflict with your system's
9267 In order to compile code which #included <libguile.h>, previous
9268 versions of Guile required you to add a directory containing all the
9269 Guile header files to your #include path. This was a problem, since
9270 Guile's header files have names which conflict with many systems'
9273 Now only <libguile.h> need appear in your #include path; you must
9274 refer to all Guile's other header files as <libguile/mumble.h>.
9275 Guile's installation procedure puts libguile.h in $(includedir), and
9276 the rest in $(includedir)/libguile.
9279 ** Two new C functions, scm_protect_object and scm_unprotect_object,
9280 have been added to the Guile library.
9282 scm_protect_object (OBJ) protects OBJ from the garbage collector.
9283 OBJ will not be freed, even if all other references are dropped,
9284 until someone does scm_unprotect_object (OBJ). Both functions
9287 Note that calls to scm_protect_object do not nest. You can call
9288 scm_protect_object any number of times on a given object, and the
9289 next call to scm_unprotect_object will unprotect it completely.
9291 Basically, scm_protect_object and scm_unprotect_object just
9292 maintain a list of references to things. Since the GC knows about
9293 this list, all objects it mentions stay alive. scm_protect_object
9294 adds its argument to the list; scm_unprotect_object remove its
9295 argument from the list.
9298 ** scm_eval_0str now returns the value of the last expression
9301 ** The new function scm_read_0str reads an s-expression from a
9302 null-terminated string, and returns it.
9304 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
9305 to a Scheme port object.
9307 ** The new function `scm_set_program_arguments' allows C code to set
9308 the value returned by the Scheme `program-arguments' function.
9313 * Guile no longer includes sophisticated Tcl/Tk support.
9315 The old Tcl/Tk support was unsatisfying to us, because it required the
9316 user to link against the Tcl library, as well as Tk and Guile. The
9317 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
9318 referring to widgets by names, rather than exporting widgets to Scheme
9319 code as a special datatype.
9321 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
9322 maintainers described some very interesting changes in progress to the
9323 Tcl/Tk internals, which would facilitate clean interfaces between lone
9324 Tk and other interpreters --- even for garbage-collected languages
9325 like Scheme. They expected the new Tk to be publicly available in the
9328 Since it seems that Guile might soon have a new, cleaner interface to
9329 lone Tk, and that the old Guile/Tk glue code would probably need to be
9330 completely rewritten, we (Jim Blandy and Richard Stallman) have
9331 decided not to support the old code. We'll spend the time instead on
9332 a good interface to the newer Tk, as soon as it is available.
9334 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
9337 Copyright information:
9339 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
9341 Permission is granted to anyone to make or distribute verbatim copies
9342 of this document as received, in any medium, provided that the
9343 copyright notice and this permission notice are preserved,
9344 thus giving the recipient permission to redistribute in turn.
9346 Permission is granted to distribute modified versions
9347 of this document, or of portions of it,
9348 under the above conditions, provided also that they
9349 carry prominent notices stating who last changed them.
9354 paragraph-separate: "[
\f]*$"