1 Guile NEWS --- history of user-visible changes.
2 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send Guile bug reports to bug-guile@gnu.org.
8 (During the 1.9 series, we will keep an incremental NEWS for the latest
9 prerelease, and a full NEWS corresponding to 1.8 -> 2.0.)
11 Changes in 1.9.11 (since the 1.9.10 prerelease):
13 ** New module: (sxml match)
15 Guile has incorporated Jim Bender's `sxml-match' library. See
16 "sxml-match' in the manual for more information. Thanks, Jim!
18 ** New module: (srfi srfi-9 gnu)
20 This module adds an extension to srfi-9, `set-record-type-printer!'. See
21 ... in the manual for more information.
23 ** Support for R6RS libraries
25 The `library' and `import' forms from the latest Scheme report have been
26 added to Guile, in such a way that R6RS libraries share a namespace with
27 Guile modules. R6RS modules may import Guile modules, and are available
28 for Guile modules to import via use-modules and all the rest. See "R6RS
29 Libraries" in the manual for more information.
31 ** Implementations of R6RS libraries
33 Guile now has implementations for all of the libraries defined in the
34 R6RS. Thanks to Julian Graham for this excellent hack. See ... in the
35 manual for a full list of libraries.
37 ** Partial R6RS compatibility
39 Guile now has enough support for R6RS to run a reasonably large subset
40 of R6RS programs. Guile is not fully R6RS compatible. Many
41 incompatibilities are simply bugs, though some parts of Guile will
42 remain R6RS-incompatible for the foreseeable future. See ... in the
43 manual Please contact bug-guile@gnu.org if you have found an issue not
44 mentioned in that compatibility list.
46 FIXME: put this list in the manual:
48 (rnrs arithmetic bitwise)
49 (rnrs arithmetic flonums)
57 (rnrs mutable-strings)
60 (rnrs record syntactic)
61 (rnrs records inspection)
62 (rnrs records procedural)
66 ** Macro expansion produces structures instead of s-expressions
68 In the olden days, macroexpanding an s-expression would yield another
69 s-expression. Though the lexical variables were renamed, expansions of
70 core forms like `if' and `begin' were still non-hygienic, as they relied
71 on the toplevel definitions of `if' et al being the conventional ones.
73 The solution is to expand to structures instead of s-expressions. There
74 is an `if' structure, a `begin' structure, a `toplevel-ref' structure,
75 etc. The expander already did this for compilation, producing Tree-IL
76 directly; it has been changed now to do so when expanding for the
79 The real truth is somewhat more involved: Tree-IL doesn't exist until
80 modules have been booted, but we need the expander to boot modules, and
81 additionally we need a boot expander before psyntax is loaded. So a
82 subset of Tree-IL is defined in C, and the boot expander produces these
83 "macroexpanded" structures. Psyntax has been modified to produce those
84 structures as well. When Tree-IL loads, it incorporates those structures
85 directly as part of its language.
87 Finally, the evaluator has been adapted to accept these "expanded"
88 structures, and enhanced to better support the gamut of this subset of
89 Tree-IL, including `lambda*' and `case-lambda'. This was a much-needed
90 harmonization between the compiler, expander, and evaluator.
92 ** Deprecated `scm_badargsp'
94 This function is unused in Guile, but was part of its API.
96 ** `sxml->xml' enhancement
98 `sxml->xml' from `(sxml simple)' can now handle the result of
99 `xml->sxml'. See bug #29260 for more information.
101 ** New module: (system vm coverage)
103 This new module can produce code coverage reports for compiled Scheme
104 code on a line-by-line level. See "Code Coverage" in the manual for more
109 The frame objects passed to VM hook procedures are now allocated on the
110 stack instead of the heap, making the next-instruction hook practical to
113 ** Add `program-free-variables' to `(system vm program)'.
117 ** New `eval-when' situation: `expand'
119 Sometimes it's important to cause side-effects while expanding an
120 expression, even in eval mode. This situation is used in
121 `define-module', `use-modules', et al, in order to affect the current
122 module and its set of syntax expanders.
124 ** Better module-level hygiene
126 Instead of attempting to track changes to the current module when
127 expanding toplevel sequences, we instead preserve referential
128 transparency relative to where the macro itself was defined. If the
129 macro should expand to expressions in the context of the new module, it
130 should wrap those expressions in `@@', which has been enhanced to accept
131 generic expressions, not just identifier references. For example, part
132 of the definition of the R6RS `library' form:
135 (define-module (name name* ...) #:pure #:version (version ...))
137 (re-export r ...) (export e ...)
138 (@@ (name name* ...) body)
141 In this example the `import' refers to the `import' definition in the
142 module where the `library' macro is defined, not in the new module.
144 ** Module system macros rewritten as hygienic macros
146 `define-module', `use-modules', `export', and other such macros have
147 been rewritten as hygienic macros. This allows the necessary referential
148 transparency for the R6RS `library' for to do the right thing.
150 ** Compiler and VM documentation updated
152 The documentation for the compiler and VM had slipped out of date; it
153 has been brought back... to the future!
155 ** Tree-IL field renaming: `vars' -> `gensyms'
157 The `vars' fields of <let>, <letrec>, <fix>, and <lambda-case> has been
158 renamed to `gensyms', for clarity, and to match <lexical-ref>.
160 ** Removed `version' field from <language>
162 Language versions weren't being updated or used in any worthwhile way;
163 they have been removed, for now at least.
165 ** New procedure: `module-export-all!'
167 This procedure exports all current and future bindings from a module.
168 Use as `(module-export-all! (current-module))'.
172 The introductory sections of the manual have been reorganized
173 significantly, making it more accessible to new users of Guile. Check it
176 ** The module namespace is now separate from the value namespace
178 It was a little-known implementation detail of Guile's module system
179 that it was built on a single hierarchical namespace of values -- that
180 if there was a module named `(foo bar)', then there was a also module
181 named `(foo)' with a binding from `bar' to the `(foo bar)' module.
183 This was a neat trick, but presented a number of problems. One problem
184 was that the bindings in a module were not apparent from the module
185 itself; perhaps the `(foo)' module had a private binding for `bar', and
186 then an external contributor defined `(foo bar)'. In the end there can
187 be only one binding, so one of the two will see the wrong thing, and
188 produce an obtuse error of unclear provenance.
190 Also, the public interface of a module was also bound in the value
191 namespace, as `%module-public-interface'. This was a hack from the early
192 days of Guile's modules.
194 Both of these warts have been fixed by the addition of fields in the
195 `module' data type. Access to modules and their interfaces from the
196 value namespace has been deprecated, and all accessors use the new
197 record accessors appropriately.
199 When Guile is built with support for deprecated code, as is the default,
200 the value namespace is still searched for modules and public interfaces,
201 and a deprecation warning is raised as appropriate.
203 Finally, to support lazy loading of modules as one used to be able to do
204 with module binder procedures, Guile now has submodule binders, called
205 if a given submodule is not found. See boot-9.scm for more information.
207 ** New procedures: module-ref-submodule, module-define-submodule,
208 nested-ref-module, nested-define-module!, local-ref-module,
211 These new accessors are like their bare variants, but operate on
212 namespaces instead of values.
214 ** The (app modules) module tree is officially deprecated
216 It used to be that one could access a module named `(foo bar)' via
217 `(nested-ref the-root-module '(app modules foo bar))'. The `(app
218 modules)' bit was a never-used and never-documented abstraction, and has
219 been deprecated. See the following mail for a full discussion:
221 http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html
223 The `%app' binding is also deprecated.
225 ** Deprecated `@bind' syntax
227 `@bind' was part of an older implementation of the Emacs Lisp language,
228 and is no longer used.
230 ** New fluid: `%file-port-name-canonicalization'
232 This fluid parameterizes the file names that are associated with file
233 ports. If %file-port-name-canonicalization is 'absolute, then file names
234 are canonicalized to be absolute paths. If it is 'relative, then the
235 name is canonicalized, but any prefix corresponding to a member of
236 `%load-path' is stripped off. Otherwise the names are passed through
239 ** Source file name canonicalization in `compile-file', `compile-and-load'
241 These file-compiling procedures now bind
242 %file-port-name-canonicalization to their `#:canonicalization' keyword
243 argument, which defaults to 'relative. In this way, one might compile
244 "../module/ice-9/boot-9.scm", but the path that gets residualized into
245 the .go is "ice-9/boot-9.scm".
247 ** Deprecate arity access via (procedure-properties proc 'arity)
249 Instead of accessing a procedure's arity as a property, use the new
250 `procedure-minimum-arity' function, which gives the most permissive
251 arity that the the function has, in the same format as the old arity
254 ** Remove redundant accessors: program-name, program-documentation,
255 program-properties, program-property
257 Instead, just use procedure-name, procedure-documentation,
258 procedure-properties, and procedure-property.
260 ** Enhance documentation for support of Emacs Lisp's `nil'
262 See "Nil" in the manual, for more details.
264 ** Enhance documentation for support of other languages
266 See "Other Languages" in the manual, for more details.
268 ** And of course, the usual collection of bugfixes
270 Interested users should see the ChangeLog for more information.
274 Changes in 1.9.x (since the 1.8.x series):
276 * New modules (see the manual for details)
278 ** `(srfi srfi-18)', more sophisticated multithreading support
279 ** `(ice-9 i18n)', internationalization support
280 ** `(rnrs bytevector)', the R6RS bytevector API
281 ** `(rnrs io ports)', a subset of the R6RS I/O port API
282 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
283 ** `(ice-9 vlist)', lists with constant-time random access; hash lists
284 ** `(system foreign)', foreign function interface
286 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
288 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
289 toolkit from Guile-Lib have been imported into Guile proper. See
290 "Standard Library" in the manual for more details.
292 ** Integration of lalr-scm, a parser generator
294 Guile has included Dominique Boucher's fine `lalr-scm' parser generator
295 as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more
298 * Changes to the stand-alone interpreter
300 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
302 Compiled code loads much faster than Scheme source code, and runs around
303 3 or 4 times as fast, generating much less garbage in the process.
305 ** Evaluating Scheme code does not use the C stack.
307 Besides when compiling Guile itself, Guile no longer uses a recursive C
308 function as an evaluator. This obviates the need to check the C stack
309 pointer for overflow. Continuations still capture the C stack, however.
311 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
312 GUILE_SYSTEM_LOAD_COMPILED_PATH
314 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
315 for source files. It is a different path, however, because compiled
316 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
319 ** New read-eval-print loop (REPL) implementation
321 Running Guile with no arguments drops the user into the new REPL. While
322 it is self-documenting to an extent, the new REPL has not yet been
323 documented in the manual. This will be fixed before 2.0.
325 ** New reader options: `square-brackets' and `r6rs-hex-escapes'
327 The reader supports a new option (changeable via `read-options'),
328 `square-brackets', which instructs it to interpret square brackets as
329 parentheses. This option is on by default.
331 When the new `r6rs-hex-escapes' reader option is enabled, the reader
332 will recognize string escape sequences as defined in R6RS.
334 ** Function profiling and tracing at the REPL
336 The `,profile FORM' REPL meta-command can now be used to statistically
337 profile execution of a form, to see which functions are taking the most
338 time. See `,help profile' for more information.
340 Similarly, `,trace FORM' traces all function applications that occur
341 during the execution of `FORM'. See `,help trace' for more information.
345 By default, if an exception is raised at the REPL and not caught by user
346 code, Guile will drop the user into a debugger. The user may request a
347 backtrace, inspect frames, or continue raising the exception. Full
348 documentation is available from within the debugger.
350 ** New `guile-tools' commands: `compile', `disassemble'
352 Pass the `--help' command-line option to these commands for more
355 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
357 Users may now install Guile to nonstandard prefixes and just run
358 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
359 include `/path/to/lib'.
361 ** Guile's Emacs integration is now more keyboard-friendly
363 Backtraces may now be disclosed with the keyboard in addition to the
366 ** Load path change: search in version-specific paths before site paths
368 When looking for a module, Guile now searches first in Guile's
369 version-specific path (the library path), *then* in the site dir. This
370 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
371 installed. Also it should cut the number of `stat' system calls by half,
374 ** Interactive Guile follows GNU conventions
376 As recommended by the GPL, Guile now shows a brief copyright and
377 warranty disclaimer on startup, along with pointers to more information.
379 * Changes to Scheme functions and syntax
381 ** New implementation of `primitive-eval'
383 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
384 still a C evaluator, used when building a fresh Guile to interpret the
385 compiler, so we can compile eval.scm. Thereafter all calls to
386 primitive-eval are implemented by VM-compiled code.
388 This allows all of Guile's procedures, be they interpreted or compiled,
389 to execute on the same stack, unifying multiple-value return semantics,
390 providing for proper tail recursion between interpreted and compiled
391 code, and simplifying debugging.
393 As part of this change, the evaluator no longer mutates the internal
394 representation of the code being evaluated in a thread-unsafe manner.
396 There are two negative aspects of this change, however. First, Guile
397 takes a lot longer to compile now. Also, there is less debugging
398 information available for debugging interpreted code. We hope to improve
399 both of these situations.
401 There are many changes to the internal C evalator interface, but all
402 public interfaces should be the same. See the ChangeLog for details. If
403 we have inadvertantly changed an interface that you were using, please
404 contact bug-guile@gnu.org.
406 ** Procedure removed: `the-environment'
408 This procedure was part of the interpreter's execution model, and does
409 not apply to the compiler.
411 ** No more `local-eval'
413 `local-eval' used to exist so that one could evaluate code in the
414 lexical context of a function. Since there is no way to get the lexical
415 environment any more, as that concept has no meaning for the compiler,
416 and a different meaning for the interpreter, we have removed the
419 If you think you need `local-eval', you should probably implement your
420 own metacircular evaluator. It will probably be as fast as Guile's
423 ** Scheme source files will now be compiled automatically.
425 If a compiled .go file corresponding to a .scm file is not found or is
426 not fresh, the .scm file will be compiled on the fly, and the resulting
427 .go file stored away. An advisory note will be printed on the console.
429 Note that this mechanism depends on preservation of the .scm and .go
430 modification times; if the .scm or .go files are moved after
431 installation, care should be taken to preserve their original
434 Autocompiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
435 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
436 will be created if needed.
438 To inhibit autocompilation, set the GUILE_AUTO_COMPILE environment
439 variable to 0, or pass --no-autocompile on the Guile command line.
441 ** New POSIX procedures: `getrlimit' and `setrlimit'
443 Note however that the interface of these functions is likely to change
444 in the next prerelease.
446 ** New POSIX procedure: `getsid'
448 Scheme binding for the `getsid' C library call.
450 ** New POSIX procedure: `getaddrinfo'
452 Scheme binding for the `getaddrinfo' C library function.
454 ** New procedure in `(oops goops)': `method-formals'
456 ** New procedures in (ice-9 session): `add-value-help-handler!',
457 `remove-value-help-handler!', `add-name-help-handler!'
458 `remove-name-help-handler!', `procedure-arguments'
460 The value and name help handlers provide some minimal extensibility to
461 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
462 example, to make stexinfo help documentation available. See those
463 procedures' docstrings for more information.
465 `procedure-arguments' describes the arguments that a procedure can take,
466 combining arity and formals. For example:
468 (procedure-arguments resolve-interface)
469 => ((required . (name)) (rest . args))
471 Additionally, `module-commentary' is now publically exported from
474 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
476 These procedures created primitive fexprs for the old evaluator, and are
477 no longer supported. If you feel that you need these functions, you
478 probably need to write your own metacircular evaluator (which will
479 probably be as fast as Guile's, anyway).
481 ** New language: ECMAScript
483 Guile now ships with one other high-level language supported,
484 ECMAScript. The goal is to support all of version 3.1 of the standard,
485 but not all of the libraries are there yet. This support is not yet
486 documented; ask on the mailing list if you are interested.
488 ** New language: Brainfuck
490 Brainfuck is a toy language that closely models Turing machines. Guile's
491 brainfuck compiler is meant to be an example of implementing other
492 languages. See the manual for details, or
493 http://en.wikipedia.org/wiki/Brainfuck for more information about the
494 Brainfuck language itself.
496 ** New language: Elisp
498 Guile now has an experimental Emacs Lisp compiler and runtime. You can
499 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
500 Kraft, and all bugs to bug-guile@gnu.org.
502 ** Better documentation infrastructure for macros
504 It is now possible to introspect on the type of a macro, e.g.
505 syntax-rules, identifier-syntax, etc, and extract information about that
506 macro, such as the syntax-rules patterns or the defmacro arguments.
507 `(texinfo reflection)' takes advantage of this to give better macro
510 ** Support for arbitrary procedure metadata
512 Building on its support for docstrings, Guile now supports multiple
513 docstrings, adding them to the tail of a compiled procedure's
514 properties. For example:
520 (procedure-properties foo)
521 => ((name . foo) (documentation . "one") (documentation . "two"))
523 Also, vectors of pairs are now treated as additional metadata entries:
526 #((quz . #f) (docstring . "xyzzy"))
528 (procedure-properties bar)
529 => ((name . bar) (quz . #f) (docstring . "xyzzy"))
531 This allows arbitrary literals to be embedded as metadata in a compiled
534 ** The psyntax expander now knows how to interpret the @ and @@ special
537 ** The psyntax expander is now hygienic with respect to modules.
539 Free variables in a macro are scoped in the module that the macro was
540 defined in, not in the module the macro is used in. For example, code
543 (define-module (foo) #:export (bar))
544 (define (helper x) ...)
546 (syntax-rules () ((_ x) (helper x))))
548 (define-module (baz) #:use-module (foo))
551 It used to be you had to export `helper' from `(foo)' as well.
552 Thankfully, this has been fixed.
554 ** Complete support for version information in Guile's `module' form
556 Guile modules now have a `#:version' field. They may be loaded by
557 version as well. See "R6RS Version References", "General Information
558 about Modules", "Using Guile Modules", and "Creating Guile Modules" in
559 the manual for more information.
561 ** Support for renaming bindings on module export
563 Wherever Guile accepts a symbol as an argument to specify a binding to
564 export, it now also accepts a pair of symbols, indicating that a binding
565 should be renamed on export. See "Creating Guile Modules" in the manual
566 for more information.
568 ** `eval-case' has been deprecated, and replaced by `eval-when'.
570 The semantics of `eval-when' are easier to understand. See "Eval When"
571 in the manual, for more information.
573 ** Guile is now more strict about prohibiting definitions in expression
576 Although previous versions of Guile accepted it, the following
577 expression is not valid, in R5RS or R6RS:
579 (if test (define foo 'bar) (define foo 'baz))
581 In this specific case, it would be better to do:
583 (define foo (if test 'bar 'baz))
585 It is certainly possible to circumvent this resriction with e.g.
586 `(module-define! (current-module) 'foo 'baz)'. We would appreciate
587 feedback about this change (a consequence of using psyntax as the
588 default expander), and may choose to revisit this situation before 2.0
589 in response to user feedback.
591 ** Defmacros must now produce valid Scheme expressions.
593 It used to be that defmacros could unquote in Scheme values, as a way of
594 supporting partial evaluation, and avoiding some hygiene issues. For
597 (define (helper x) ...)
598 (define-macro (foo bar)
601 Assuming this macro is in the `(baz)' module, the direct translation of
604 (define (helper x) ...)
605 (define-macro (foo bar)
606 `((@@ (baz) helper) ,bar))
608 Of course, one could just use a hygienic macro instead:
612 ((_ bar) (helper bar))))
614 ** Guile's psyntax now supports docstrings and internal definitions.
616 The following Scheme is not strictly legal:
623 However its intent is fairly clear. Guile interprets "bar" to be the
624 docstring of `foo', and the definition of `baz' is still in definition
627 ** Macros need to be defined before their first use.
629 It used to be that with lazy memoization, this might work:
633 (define-macro (ref x) x)
636 But now, the body of `foo' is interpreted to mean a call to the toplevel
637 `ref' function, instead of a macro expansion. The solution is to define
638 macros before code that uses them.
640 ** Functions needed by macros at expand-time need to be present at
643 For example, this code will work at the REPL:
645 (define (double-helper x) (* x x))
646 (define-macro (double-literal x) (double-helper x))
647 (double-literal 2) => 4
649 But it will not work when a file is compiled, because the definition of
650 `double-helper' is not present at expand-time. The solution is to wrap
651 the definition of `double-helper' in `eval-when':
653 (eval-when (load compile eval)
654 (define (double-helper x) (* x x)))
655 (define-macro (double-literal x) (double-helper x))
656 (double-literal 2) => 4
658 See the documentation for eval-when for more information.
660 ** `macroexpand' produces structures, not S-expressions.
662 Given the need to maintain referential transparency, both lexically and
663 modular, the result of expanding Scheme expressions is no longer itself
664 an s-expression. If you want a human-readable approximation of the
665 result of `macroexpand', call `tree-il->scheme' from `(language
668 ** Removed function: `macroexpand-1'
670 It is unclear how to implement `macroexpand-1' with syntax-case, though
671 PLT Scheme does prove that it is possible.
673 ** New reader macros: #' #` #, #,@
675 These macros translate, respectively, to `syntax', `quasisyntax',
676 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
677 These reader macros may be overridden by `read-hash-extend'.
679 ** Incompatible change to #'
681 Guile did have a #' hash-extension, by default, which just returned the
682 subsequent datum: #'foo => foo. In the unlikely event that anyone
683 actually used this, this behavior may be reinstated via the
684 `read-hash-extend' mechanism.
686 ** Scheme expresssions may be commented out with #;
688 #; comments out an entire expression. See SRFI-62 or the R6RS for more
691 ** Prompts: Delimited, composable continuations
693 Guile now has prompts as part of its primitive language. See "Prompts"
694 in the manual, for more information.
696 Expressions entered in at the REPL, or from the command line, are
697 surrounded by a prompt with the default prompt tag.
699 ** `make-stack' with a tail-called procedural narrowing argument no longer
700 works (with compiled procedures)
702 It used to be the case that a captured stack could be narrowed to select
703 calls only up to or from a certain procedure, even if that procedure
704 already tail-called another procedure. This was because the debug
705 information from the original procedure was kept on the stack.
707 Now with the new compiler, the stack only contains active frames from
708 the current continuation. A narrow to a procedure that is not in the
709 stack will result in an empty stack. To fix this, narrow to a procedure
710 that is active in the current continuation, or narrow to a specific
711 number of stack frames.
713 ** Backtraces through compiled procedures only show procedures that are
714 active in the current continuation
716 Similarly to the previous issue, backtraces in compiled code may be
717 different from backtraces in interpreted code. There are no semantic
718 differences, however. Please mail bug-guile@gnu.org if you see any
719 deficiencies with Guile's backtraces.
721 ** syntax-rules and syntax-case macros now propagate source information
722 through to the expanded code
724 This should result in better backtraces.
726 ** The currying behavior of `define' has been removed.
728 Before, `(define ((f a) b) (* a b))' would translate to
730 (define f (lambda (a) (lambda (b) (* a b))))
732 Now a syntax error is signaled, as this syntax is not supported by
733 default. Use the `(ice-9 curried-definitions)' module to get back the
736 ** New procedure, `define!'
738 `define!' is a procedure that takes two arguments, a symbol and a value,
739 and binds the value to the symbol in the current module. It's useful to
740 programmatically make definitions in the current module, and is slightly
741 less verbose than `module-define!'.
743 ** All modules have names now
745 Before, you could have anonymous modules: modules without names. Now,
746 because of hygiene and macros, all modules have names. If a module was
747 created without a name, the first time `module-name' is called on it, a
748 fresh name will be lazily generated for it.
750 ** Many syntax errors have different texts now
752 Syntax errors still throw to the `syntax-error' key, but the arguments
753 are often different now. Perhaps in the future, Guile will switch to
754 using standard SRFI-35 conditions.
756 ** Returning multiple values to compiled code will silently truncate the
757 values to the expected number
759 For example, the interpreter would raise an error evaluating the form,
760 `(+ (values 1 2) (values 3 4))', because it would see the operands as
761 being two compound "values" objects, to which `+' does not apply.
763 The compiler, on the other hand, receives multiple values on the stack,
764 not as a compound object. Given that it must check the number of values
765 anyway, if too many values are provided for a continuation, it chooses
766 to truncate those values, effectively evaluating `(+ 1 3)' instead.
768 The idea is that the semantics that the compiler implements is more
769 intuitive, and the use of the interpreter will fade out with time.
770 This behavior is allowed both by the R5RS and the R6RS.
772 ** Multiple values in compiled code are not represented by compound
775 This change may manifest itself in the following situation:
777 (let ((val (foo))) (do-something) val)
779 In the interpreter, if `foo' returns multiple values, multiple values
780 are produced from the `let' expression. In the compiler, those values
781 are truncated to the first value, and that first value is returned. In
782 the compiler, if `foo' returns no values, an error will be raised, while
783 the interpreter would proceed.
785 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
786 behavior is more correct, however. If you wish to preserve a potentially
787 multiply-valued return, you will need to set up a multiple-value
788 continuation, using `call-with-values'.
790 ** Defmacros are now implemented in terms of syntax-case.
792 The practical ramification of this is that the `defmacro?' predicate has
793 been removed, along with `defmacro-transformer', `macro-table',
794 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
795 `defmacro:transformer'. This is because defmacros are simply macros. If
796 any of these procedures provided useful facilities to you, we encourage
797 you to contact the Guile developers.
799 ** Hygienic macros documented as the primary syntactic extension mechanism.
801 The macro documentation was finally fleshed out with some documentation
802 on `syntax-rules' and `syntax-case' macros, and other parts of the macro
803 expansion process. See "Macros" in the manual, for details.
805 ** psyntax is now the default expander
807 Scheme code is now expanded by default by the psyntax hygienic macro
808 expander. Expansion is performed completely before compilation or
811 Notably, syntax errors will be signalled before interpretation begins.
812 In the past, many syntax errors were only detected at runtime if the
813 code in question was memoized.
815 As part of its expansion, psyntax renames all lexically-bound
816 identifiers. Original identifier names are preserved and given to the
817 compiler, but the interpreter will see the renamed variables, e.g.,
818 `x432' instead of `x'.
820 Note that the psyntax that Guile uses is a fork, as Guile already had
821 modules before incompatible modules were added to psyntax -- about 10
822 years ago! Thus there are surely a number of bugs that have been fixed
823 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
825 ** syntax-rules and syntax-case are available by default.
827 There is no longer any need to import the `(ice-9 syncase)' module
828 (which is now deprecated). The expander may be invoked directly via
829 `macroexpand', though it is normally searched for via the current module
832 Also, the helper routines for syntax-case are available in the default
833 environment as well: `syntax->datum', `datum->syntax',
834 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
835 `identifier?', and `syntax-violation'. See the R6RS for documentation.
837 ** Tail patterns in syntax-case
839 Guile has pulled in some more recent changes from the psyntax portable
840 syntax expander, to implement support for "tail patterns". Such patterns
841 are supported by syntax-rules and syntax-case. This allows a syntax-case
842 match clause to have ellipses, then a pattern at the end. For example:
846 ((_ val match-clause ... (else e e* ...))
849 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
850 tail pattern for the else clause. Thanks to Andreas Rottmann for the
851 patch, and Kent Dybvig for the code.
853 ** Lexical bindings introduced by hygienic macros may not be referenced
854 by nonhygienic macros.
856 If a lexical binding is introduced by a hygienic macro, it may not be
857 referenced by a nonhygienic macro. For example, this works:
860 (define-macro (bind-x val body)
861 `(let ((x ,val)) ,body))
862 (define-macro (ref x)
869 (define-syntax bind-x
871 ((_ val body) (let ((x val)) body))))
872 (define-macro (ref x)
876 It is not normal to run into this situation with existing code. However,
877 as code is ported over from defmacros to syntax-case, it is possible to
878 run into situations like this. In the future, Guile will probably port
879 its `while' macro to syntax-case, which makes this issue one to know
882 ** Macros may no longer be referenced as first-class values.
884 In the past, you could evaluate e.g. `if', and get its macro value. Now,
885 expanding this form raises a syntax error.
887 Macros still /exist/ as first-class values, but they must be
888 /referenced/ via the module system, e.g. `(module-ref (current-module)
891 ** Macros may now have docstrings.
893 `object-documentation' from `(ice-9 documentation)' may be used to
894 retrieve the docstring, once you have a macro value -- but see the above
895 note about first-class macros. Docstrings are associated with the syntax
896 transformer procedures.
898 ** `case-lambda' is now available in the default environment.
900 The binding in the default environment is equivalent to the one from the
901 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
902 to maintain compatibility with Guile 1.8 and earlier.
904 ** Procedures may now have more than one arity.
906 This can be the case, for example, in case-lambda procedures. The
907 arities of compiled procedures may be accessed via procedures from the
908 `(system vm program)' module; see "Compiled Procedures", "Optional
909 Arguments", and "Case-lambda" in the manual.
911 ** `lambda*' and `define*' are now available in the default environment
913 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
914 compatibility purposes. No semantic change has been made (we hope).
915 Optional and keyword arguments now dispatch via special VM operations,
916 without the need to cons rest arguments, making them very fast.
918 ** New function, `truncated-print', with `format' support
920 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
921 will ensure that the output stays within a certain width, truncating the
922 output in what is hopefully an intelligent manner. See the manual for
925 There is a new `format' specifier, `~@y', for doing a truncated
926 print (as opposed to `~y', which does a pretty-print). See the `format'
927 documentation for more details.
929 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
931 Guile now implements SRFI-4 vectors using bytevectors. Often when you
932 have a numeric vector, you end up wanting to write its bytes somewhere,
933 or have access to the underlying bytes, or read in bytes from somewhere
934 else. Bytevectors are very good at this sort of thing. But the SRFI-4
935 APIs are nicer to use when doing number-crunching, because they are
936 addressed by element and not by byte.
938 So as a compromise, Guile allows all bytevector functions to operate on
939 numeric vectors. They address the underlying bytes in the native
940 endianness, as one would expect.
942 Following the same reasoning, that it's just bytes underneath, Guile
943 also allows uniform vectors of a given type to be accessed as if they
944 were of any type. One can fill a u32vector, and access its elements with
945 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
948 In this way, uniform numeric vectors may be written to and read from
949 input/output ports using the procedures that operate on bytevectors.
951 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
952 inlined to the VM instructions for bytevector access.
954 See "SRFI-4" in the manual, for more information.
956 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
958 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
959 are part of the standard. Complex uniform vectors and the
960 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
962 Guile's default environment imports `(srfi srfi-4)', and probably should
963 import `(srfi srfi-4 gnu)' as well.
965 See "SRFI-4 Extensions" in the manual, for more information.
967 ** New syntax: include-from-path.
969 `include-from-path' is like `include', except it looks for its file in
970 the load path. It can be used to compile other files into a file.
972 ** New syntax: quasisyntax.
974 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
975 documentation for more information. Thanks to Andre van Tonder for the
978 ** Unicode characters
980 Unicode characters may be entered in octal format via e.g. `#\454', or
981 created via (integer->char 300). A hex external representation will
982 probably be introduced at some point.
986 Internally, strings are now represented either in the `latin-1'
987 encoding, one byte per character, or in UTF-32, with four bytes per
988 character. Strings manage their own allocation, switching if needed.
990 Extended characters may be written in a literal string using the
991 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
992 or 24-bit codepoints, respectively, or entered directly in the native
993 encoding of the port on which the string is read.
997 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
999 ** Support for non-ASCII source code files
1001 The default reader now handles source code files for some of the
1002 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
1003 should have an encoding declaration near the top of the file. Also,
1004 there is a new function, `file-encoding', that scans a port for a coding
1005 declaration. See the section of the manual entitled, "Character Encoding
1008 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
1009 code. This use is now discouraged.
1011 ** Support for locale transcoding when reading from and writing to ports
1013 Ports now have an associated character encoding, and port read and write
1014 operations do conversion to and from locales automatically. Ports also
1015 have an associated strategy for how to deal with locale conversion
1018 See the documentation in the manual for the four new support functions,
1019 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
1020 and `port-conversion-strategy'.
1022 ** String and SRFI-13 functions can operate on Unicode strings
1024 ** Unicode support for SRFI-14 character sets
1026 The default character sets are no longer locale dependent and contain
1027 characters from the whole Unicode range. There is a new predefined
1028 character set, `char-set:designated', which contains all assigned
1029 Unicode characters. There is a new debugging function, `%char-set-dump'.
1031 ** Character functions operate on Unicode characters
1033 `char-upcase' and `char-downcase' use default Unicode casing rules.
1034 Character comparisons such as `char<?' and `char-ci<?' now sort based on
1035 Unicode code points.
1037 ** Global variables `scm_charnames' and `scm_charnums' are removed
1039 These variables contained the names of control characters and were
1040 used when writing characters. While these were global, they were
1041 never intended to be public API. They have been replaced with private
1044 ** EBCDIC support is removed
1046 There was an EBCDIC compile flag that altered some of the character
1047 processing. It appeared that full EBCDIC support was never completed
1048 and was unmaintained.
1050 ** Compile-time warnings
1052 Guile can warn about potentially unbound free variables. Pass the
1053 -Wunbound-variable on the `guile-tools compile' command line, or add
1054 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
1057 Guile can also warn when you pass the wrong number of arguments to a
1058 procedure, with -Warity-mismatch, or `arity-mismatch' in the
1059 `#:warnings' as above.
1061 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
1062 warn about unused local or global (top-level) variables.
1064 ** A new `memoize-symbol' evaluator trap has been added.
1066 This trap can be used for efficiently implementing a Scheme code
1069 ** Duplicate bindings among used modules are resolved lazily.
1071 This slightly improves program startup times.
1073 ** New thread cancellation and thread cleanup API
1075 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
1077 ** GOOPS dispatch in scheme
1079 As an implementation detail, GOOPS dispatch is no longer implemented by
1080 special evaluator bytecodes, but rather directly via a Scheme function
1081 associated with an applicable struct. There is some VM support for the
1082 underlying primitives, like `class-of'.
1084 This change will in the future allow users to customize generic function
1085 dispatch without incurring a performance penalty, and allow us to
1086 implement method combinations.
1088 ** Applicable struct support
1090 One may now make structs from Scheme that may be applied as procedures.
1091 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
1092 That struct will be the vtable of your applicable structs; instances of
1093 that new struct are assumed to have the procedure in their first slot.
1094 `<applicable-struct-vtable>' is like Common Lisp's
1095 `funcallable-standard-class'. Likewise there is
1096 `<applicable-struct-with-setter-vtable>', which looks for the setter in
1097 the second slot. This needs to be better documented.
1101 GOOPS had a number of concepts that were relevant to the days of Tcl,
1102 but not any more: operators and entities, mainly. These objects were
1103 never documented, and it is unlikely that they were ever used. Operators
1104 were a kind of generic specific to the Tcl support. Entities were
1105 replaced by applicable structs, mentioned above.
1107 ** New struct slot allocation: "hidden"
1109 A hidden slot is readable and writable, but will not be initialized by a
1110 call to make-struct. For example in your layout you would say "ph"
1111 instead of "pw". Hidden slots are useful for adding new slots to a
1112 vtable without breaking existing invocations to make-struct.
1114 ** eqv? not a generic
1116 One used to be able to extend `eqv?' as a primitive-generic, but no
1117 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
1118 should be able to compile to static dispatch tables, it doesn't make
1119 sense to allow extensions that would subvert this optimization.
1121 ** `inet-ntop' and `inet-pton' are always available.
1123 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
1124 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
1125 functions are deprecated.
1127 ** Fast bit operations.
1129 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
1130 have dedicated bytecodes. Guile is not just for symbolic computation,
1131 it's for number crunching too.
1133 ** Faster SRFI-9 record access
1135 SRFI-9 records are now implemented directly on top of Guile's structs,
1136 and their accessors are defined in such a way that normal call-sites
1137 inline to special VM opcodes, while still allowing for the general case
1138 (e.g. passing a record accessor to `apply').
1140 ** R6RS block comment support
1142 Guile now supports R6RS nested block comments. The start of a comment is
1143 marked with `#|', and the end with `|#'.
1145 ** `guile-2' cond-expand feature
1147 To test if your code is running under Guile 2.0 (or its alpha releases),
1148 test for the `guile-2' cond-expand feature. Like this:
1150 (cond-expand (guile-2 (eval-when (compile)
1151 ;; This must be evaluated at compile time.
1152 (fluid-set! current-reader my-reader)))
1154 ;; Earlier versions of Guile do not have a
1155 ;; separate compilation phase.
1156 (fluid-set! current-reader my-reader)))
1158 ** Fix bad interaction between `false-if-exception' and stack-call.
1160 Exceptions thrown by `false-if-exception' were erronously causing the
1161 stack to be saved, causing later errors to show the incorrectly-saved
1162 backtrace. This has been fixed.
1164 ** New global variables: %load-compiled-path, %load-compiled-extensions
1166 These are analogous to %load-path and %load-extensions.
1168 ** New procedure, `make-promise'
1170 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
1172 ** `defined?' may accept a module as its second argument
1174 Previously it only accepted internal structures from the evaluator.
1176 ** New entry into %guile-build-info: `ccachedir'
1178 ** Fix bug in `module-bound?'.
1180 `module-bound?' was returning true if a module did have a local
1181 variable, but one that was unbound, but another imported module bound
1182 the variable. This was an error, and was fixed.
1184 ** `(ice-9 syncase)' has been deprecated.
1186 As syntax-case is available by default, importing `(ice-9 syncase)' has
1187 no effect, and will trigger a deprecation warning.
1189 ** New readline history functions
1191 The (ice-9 readline) module now provides add-history, read-history,
1192 write-history and clear-history, which wrap the corresponding GNU
1193 History library functions.
1195 ** Removed deprecated uniform array procedures:
1196 dimensions->uniform-array, list->uniform-array, array-prototype
1198 Instead, use make-typed-array, list->typed-array, or array-type,
1201 ** Deprecated: `lazy-catch'
1203 `lazy-catch' was a form that captured the stack at the point of a
1204 `throw', but the dynamic state at the point of the `catch'. It was a bit
1205 crazy. Please change to use `catch', possibly with a throw-handler, or
1206 `with-throw-handler'.
1208 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
1210 * Changes to the C interface
1212 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
1214 The semantics of `scm_gc_malloc ()' have been changed, in a
1215 backward-compatible way. A new allocation routine,
1216 `scm_gc_malloc_pointerless ()', was added.
1218 Libgc is a conservative GC, which we hope will make interaction with C
1219 code easier and less error-prone.
1221 ** New type definitions for `scm_t_intptr' and friends.
1223 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
1224 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
1225 available to C. Have fun!
1227 ** The GH interface (deprecated in version 1.6, 2001) was removed.
1229 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
1231 This makes these internal functions technically not callable from
1234 ** Functions for handling `scm_option' now no longer require an argument
1235 indicating length of the `scm_t_option' array.
1237 ** Procedures-with-setters are now implemented using applicable structs
1239 From a user's perspective this doesn't mean very much. But if, for some
1240 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
1241 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
1244 ** Remove old evaluator closures
1246 There used to be ranges of typecodes allocated to interpreted data
1247 structures, but that it no longer the case, given that interpreted
1248 procedure are now just regular VM closures. As a result, there is a
1249 newly free tc3, and a number of removed macros. See the ChangeLog for
1252 ** Primitive procedures are now VM trampoline procedures
1254 It used to be that there were something like 12 different typecodes
1255 allocated to primitive procedures, each with its own calling convention.
1256 Now there is only one, the gsubr. This may affect user code if you were
1257 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
1258 solution is to switch to use scm_c_make_gsubr. This solution works well
1259 both with the old 1.8 and and with the current 1.9 branch.
1261 Guile's old evaluator used to have special cases for applying "gsubrs",
1262 primitive procedures with specified numbers of required, optional, and
1263 rest arguments. Now, however, Guile represents gsubrs as normal VM
1264 procedures, with appropriate bytecode to parse out the correct number of
1265 arguments, including optional and rest arguments, and then with a
1266 special bytecode to apply the gsubr.
1268 This allows primitive procedures to appear on the VM stack, allowing
1269 them to be accurately counted in profiles. Also they now have more
1270 debugging information attached to them -- their number of arguments, for
1271 example. In addition, the VM can completely inline the application
1272 mechanics, allowing for faster primitive calls.
1274 However there are some changes on the C level. There is no more
1275 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
1276 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
1277 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
1278 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
1279 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
1281 Perhaps more significantly, `scm_c_make_subr',
1282 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
1283 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
1284 and there are no more subr typecodes. Use the scm_c_make_gsubr family
1287 Normal users of gsubrs should not be affected, though, as the
1288 scm_c_make_gsubr family still is the correct way to create primitive
1291 ** Remove deprecated array C interfaces
1293 Removed the deprecated array functions `scm_i_arrayp',
1294 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
1295 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
1296 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
1297 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
1299 ** Remove unused snarf macros
1301 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
1302 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
1304 ** New functions: `scm_call_n', `scm_c_run_hookn'
1306 `scm_call_n' applies to apply a function to an array of arguments.
1307 `scm_c_run_hookn' runs a hook with an array of arguments.
1309 ** Some SMOB types changed to have static typecodes
1311 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
1312 they have statically allocated tc7 typecodes.
1314 ** Preparations for changing SMOB representation
1316 If things go right, we'll be changing the SMOB representation soon. To
1317 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
1318 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
1319 in the future. Code accessing SMOBs using SCM_CELL macros was never
1320 correct, but until now things still worked. Users should be aware of
1323 ** Changed invocation mechanics of applicable SMOBs
1325 Guile's old evaluator used to have special cases for applying SMOB
1326 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
1327 trampoline procedure for it, and use the normal mechanics to apply the
1328 trampoline. This simplifies procedure application in the normal,
1331 The upshot is that the mechanics used to apply a SMOB are different from
1332 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
1333 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
1334 deprecated. Just use the scm_call_0 family of procedures.
1336 ** New C function: scm_module_public_interface
1338 This procedure corresponds to Scheme's `module-public-interface'.
1340 ** Undeprecate `scm_the_root_module ()'
1342 It's useful to be able to get the root module from C without doing a
1345 ** Inline vector allocation
1347 Instead of having vectors point out into the heap for their data, their
1348 data is now allocated inline to the vector object itself. The same is
1349 true for bytevectors, by default, though there is an indirection
1350 available which should allow for making a bytevector from an existing
1353 ** New struct constructors that don't involve making lists
1355 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
1356 constructors, respectively, for structs. You might find them useful.
1360 In Guile 1.8, there were debugging frames on the C stack. Now there is
1361 no more need to explicitly mark the stack in this way, because Guile has
1362 a VM stack that it knows how to walk, which simplifies the C API
1363 considerably. See the ChangeLog for details; the relevant interface is
1364 in libguile/stacks.h. The Scheme API has not been changed significantly.
1366 ** Removal of Guile's primitive object system.
1368 There were a number of pieces in `objects.[ch]' that tried to be a
1369 minimal object system, but were never documented, and were quickly
1370 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
1371 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
1372 from objects.h are no more. In the very unlikely case in which these
1373 were useful to you, we urge you to contact guile-devel.
1377 Actually the future is still in the state that it was, is, and ever
1378 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
1379 part of it. These files were experimental, never compiled, and would be
1380 better implemented in Scheme anyway. In the future, that is.
1382 ** Deprecate trampolines
1384 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
1385 so on. The point was to do some precomputation on the type of the
1386 procedure, then return a specialized "call" procedure. However this
1387 optimization wasn't actually an optimization, so it is now deprecated.
1388 Just use `scm_call_0', etc instead.
1390 ** Better support for Lisp `nil'.
1392 The bit representation of `nil' has been tweaked so that it is now very
1393 efficient to check e.g. if a value is equal to Scheme's end-of-list or
1394 Lisp's nil. Additionally there are a heap of new, specific predicates
1395 like scm_is_null_or_nil.
1397 ** Better integration of Lisp `nil'.
1399 `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now
1400 for Lisp's `nil'. This shouldn't affect any Scheme code at this point,
1401 but when we start to integrate more with Emacs, it is possible that we
1402 break code that assumes that, for example, `(not x)' implies that `x' is
1403 `eq?' to `#f'. This is not a common assumption. Refactoring affected
1404 code to rely on properties instead of identities will improve code
1405 correctness. See "Nil" in the manual, for more details.
1407 ** Support for static allocation of strings, symbols, and subrs.
1409 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
1410 much of their associated data as static variables, reducing Guile's
1413 ** `scm_stat' has an additional argument, `exception_on_error'
1414 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
1416 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
1418 Previously they would use the `off_t' type, which is fragile since its
1419 definition depends on the application's value for `_FILE_OFFSET_BITS'.
1421 ** The `long_long' C type, deprecated in 1.8, has been removed
1423 ** Removed deprecated uniform array procedures: scm_make_uve,
1424 scm_array_prototype, scm_list_to_uniform_array,
1425 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
1426 scm_ra_set_contp, scm_aind, scm_raprin1
1428 These functions have been deprecated since early 2005.
1430 * Changes to the distribution
1432 ** Guile's license is now LGPLv3+
1434 In other words the GNU Lesser General Public License, version 3 or
1435 later (at the discretion of each person that chooses to redistribute
1438 ** GOOPS documentation folded into Guile reference manual
1440 GOOPS, Guile's object system, used to be documented in separate manuals.
1441 This content is now included in Guile's manual directly.
1443 ** `guile-config' will be deprecated in favor of `pkg-config'
1445 `guile-config' has been rewritten to get its information from
1446 `pkg-config', so this should be a transparent change. Note however that
1447 guile.m4 has yet to be modified to call pkg-config instead of
1450 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
1452 Programs that use `pkg-config' to find Guile or one of its Autoconf
1453 macros should now require `guile-2.0' instead of `guile-1.8'.
1455 ** New installation directory: $(pkglibdir)/1.9/ccache
1457 If $(libdir) is /usr/lib, for example, Guile will install its .go files
1458 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
1460 ** Parallel installability fixes
1462 Guile now installs its header files to a effective-version-specific
1463 directory, and includes the effective version (e.g. 2.0) in the library
1464 name (e.g. libguile-2.0.so).
1466 This change should be transparent to users, who should detect Guile via
1467 the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow
1468 parallel installs for multiple versions of Guile development
1471 ** Dynamically loadable extensions may be placed in a Guile-specific path
1473 Before, Guile only searched the system library paths for extensions
1474 (e.g. /usr/lib), which meant that the names of Guile extensions had to
1475 be globally unique. Installing them to a Guile-specific extensions
1476 directory is cleaner. Use `pkg-config --variable=extensionsdir
1477 guile-2.0' to get the location of the extensions directory.
1479 ** New dependency: libgc
1481 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
1483 ** New dependency: GNU libunistring
1485 See http://www.gnu.org/software/libunistring/, for more information. Our
1486 Unicode support uses routines from libunistring.
1488 ** New dependency: libffi
1490 See http://sourceware.org/libffi/, for more information.
1494 Changes in 1.8.8 (since 1.8.7)
1498 ** Fix possible buffer overruns when parsing numbers
1499 ** Avoid clash with system setjmp/longjmp on IA64
1500 ** Fix `wrong type arg' exceptions with IPv6 addresses
1503 Changes in 1.8.7 (since 1.8.6)
1505 * New modules (see the manual for details)
1507 ** `(srfi srfi-98)', an interface to access environment variables
1511 ** Fix compilation with `--disable-deprecated'
1512 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
1513 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
1514 ** Fix build problem when scm_t_timespec is different from struct timespec
1515 ** Fix build when compiled with -Wundef -Werror
1516 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
1517 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
1518 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
1519 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
1520 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
1521 ** Fix reading of complex numbers where both parts are inexact decimals
1523 ** Allow @ macro to work with (ice-9 syncase)
1525 Previously, use of the @ macro in a module whose code is being
1526 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
1527 Now it works as you would expect (giving the value of the specified
1530 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
1533 Changes in 1.8.6 (since 1.8.5)
1535 * New features (see the manual for details)
1537 ** New convenience function `scm_c_symbol_length ()'
1539 ** Single stepping through code from Emacs
1541 When you use GDS to evaluate Scheme code from Emacs, you can now use
1542 `C-u' to indicate that you want to single step through that code. See
1543 `Evaluating Scheme Code' in the manual for more details.
1545 ** New "guile(1)" man page!
1547 * Changes to the distribution
1549 ** Automake's `AM_MAINTAINER_MODE' is no longer used
1551 Thus, the `--enable-maintainer-mode' configure option is no longer
1552 available: Guile is now always configured in "maintainer mode".
1554 ** `ChangeLog' files are no longer updated
1556 Instead, changes are detailed in the version control system's logs. See
1557 the top-level `ChangeLog' files for details.
1562 ** `symbol->string' now returns a read-only string, as per R5RS
1563 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
1564 ** `guile-config link' now prints `-L$libdir' before `-lguile'
1565 ** Fix memory corruption involving GOOPS' `class-redefinition'
1566 ** Fix possible deadlock in `mutex-lock'
1567 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
1568 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
1569 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
1570 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
1571 ** Fix misleading output from `(help rationalize)'
1572 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
1573 ** Fix `gcd' when called with a single, negative argument.
1574 ** Fix `Stack overflow' errors seen when building on some platforms
1575 ** Fix bug when `scm_with_guile ()' was called several times from the
1577 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
1578 dynamic environment of the call to `raise'
1579 ** Fix potential deadlock in `make-struct'
1580 ** Fix compilation problem with libltdl from Libtool 2.2.x
1581 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
1584 Changes in 1.8.5 (since 1.8.4)
1586 * Infrastructure changes
1588 ** Guile repository switched from CVS to Git
1590 The new repository can be accessed using
1591 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
1592 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
1594 ** Add support for `pkg-config'
1596 See "Autoconf Support" in the manual for details.
1598 * New modules (see the manual for details)
1602 * New features (see the manual for details)
1604 ** New `postfix' read option, for SRFI-88 keyword syntax
1605 ** Some I/O primitives have been inlined, which improves I/O performance
1606 ** New object-based traps infrastructure
1608 This is a GOOPS-based infrastructure that builds on Guile's low-level
1609 evaluator trap calls and facilitates the development of debugging
1610 features like single-stepping, breakpoints, tracing and profiling.
1611 See the `Traps' node of the manual for details.
1613 ** New support for working on Guile code from within Emacs
1615 Guile now incorporates the `GDS' library (previously distributed
1616 separately) for working on Guile code from within Emacs. See the
1617 `Using Guile In Emacs' node of the manual for details.
1621 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
1622 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
1624 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
1625 would trigger an unbound variable error for `match:andmap'.
1627 ** `(oop goops describe)' now properly provides the `describe' feature
1628 ** Fixed `args-fold' from `(srfi srfi-37)'
1630 Previously, parsing short option names of argument-less options would
1631 lead to a stack overflow.
1633 ** `(srfi srfi-35)' is now visible through `cond-expand'
1634 ** Fixed type-checking for the second argument of `eval'
1635 ** Fixed type-checking for SRFI-1 `partition'
1636 ** Fixed `struct-ref' and `struct-set!' on "light structs"
1637 ** Honor struct field access rights in GOOPS
1638 ** Changed the storage strategy of source properties, which fixes a deadlock
1639 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
1640 ** Fixed build issue for GNU/Linux on IA64
1641 ** Fixed build issues on NetBSD 1.6
1642 ** Fixed build issue on Solaris 2.10 x86_64
1643 ** Fixed build issue with DEC/Compaq/HP's compiler
1644 ** Fixed `scm_from_complex_double' build issue on FreeBSD
1645 ** Fixed `alloca' build issue on FreeBSD 6
1646 ** Removed use of non-portable makefile constructs
1647 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
1648 ** Make sure all tests honor `$TMPDIR'
1651 Changes in 1.8.4 (since 1.8.3)
1655 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
1656 ** Fixed a segmentation fault which occurred when displaying the
1657 backtrace of a stack with a promise object (made by `delay') in it.
1658 ** Make `accept' leave guile mode while blocking
1659 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
1660 ** Fixed a build problem on AIX (use of func_data identifier)
1661 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
1662 called with an associator proc that returns neither a pair nor #f.
1663 ** Secondary threads now always return a valid module for (current-module).
1664 ** Avoid MacOS build problems caused by incorrect combination of "64"
1665 system and library calls.
1666 ** `guile-snarf' now honors `$TMPDIR'
1667 ** `guile-config compile' now reports CPPFLAGS used at compile-time
1668 ** Fixed build with Sun Studio (Solaris 9)
1669 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
1670 uniform vectors on AIX.
1671 ** Fixed a deadlock that occurs upon GC with multiple threads.
1672 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
1673 ** Fixed autotool-derived build problems on AIX 6.1.
1674 ** Fixed NetBSD/alpha support
1675 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
1677 * New modules (see the manual for details)
1681 * Documentation fixes and improvements
1683 ** Removed premature breakpoint documentation
1685 The features described are not available in the series of 1.8.x
1686 releases, so the documentation was misleading and has been removed.
1688 ** More about Guile's default *random-state* variable
1690 ** GOOPS: more about how to use `next-method'
1692 * Changes to the distribution
1694 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
1696 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
1697 General Public License, and the few incorrect files have now been
1698 fixed to agree with the rest of the Guile distribution.
1700 ** Removed unnecessary extra copies of COPYING*
1702 The distribution now contains a single COPYING.LESSER at its top level.
1705 Changes in 1.8.3 (since 1.8.2)
1707 * New modules (see the manual for details)
1714 ** The `(ice-9 slib)' module now works as expected
1715 ** Expressions like "(set! 'x #t)" no longer yield a crash
1716 ** Warnings about duplicate bindings now go to stderr
1717 ** A memory leak in `make-socket-address' was fixed
1718 ** Alignment issues (e.g., on SPARC) in network routines were fixed
1719 ** A threading issue that showed up at least on NetBSD was fixed
1720 ** Build problems on Solaris and IRIX fixed
1722 * Implementation improvements
1724 ** The reader is now faster, which reduces startup time
1725 ** Procedures returned by `record-accessor' and `record-modifier' are faster
1728 Changes in 1.8.2 (since 1.8.1):
1730 * New procedures (see the manual for details)
1732 ** set-program-arguments
1735 * Incompatible changes
1737 ** The body of a top-level `define' no longer sees the binding being created
1739 In a top-level `define', the binding being created is no longer visible
1740 from the `define' body. This breaks code like
1741 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
1742 unbound in the body. However, such code was not R5RS-compliant anyway,
1747 ** Fractions were not `equal?' if stored in unreduced form.
1748 (A subtle problem, since printing a value reduced it, making it work.)
1749 ** srfi-60 `copy-bit' failed on 64-bit systems
1750 ** "guile --use-srfi" option at the REPL can replace core functions
1751 (Programs run with that option were ok, but in the interactive REPL
1752 the core bindings got priority, preventing SRFI replacements or
1754 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
1755 ** `kill' on mingw throws an error for a PID other than oneself
1756 ** Procedure names are attached to procedure-with-setters
1757 ** Array read syntax works with negative lower bound
1758 ** `array-in-bounds?' fix if an array has different lower bounds on each index
1759 ** `*' returns exact 0 for "(* inexact 0)"
1760 This follows what it always did for "(* 0 inexact)".
1761 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
1762 ** SRFI-19: `date->julian-day' did not account for timezone offset
1763 ** `ttyname' no longer crashes when passed a non-tty argument
1764 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
1765 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
1766 ** GOOPS: Fixed a bug in `method-more-specific?'
1767 ** Build problems on Solaris fixed
1768 ** Build problems on HP-UX IA64 fixed
1769 ** Build problems on MinGW fixed
1772 Changes in 1.8.1 (since 1.8.0):
1774 * LFS functions are now used to access 64-bit files on 32-bit systems.
1776 * New procedures (see the manual for details)
1778 ** primitive-_exit - [Scheme] the-root-module
1779 ** scm_primitive__exit - [C]
1780 ** make-completion-function - [Scheme] (ice-9 readline)
1781 ** scm_c_locale_stringn_to_number - [C]
1782 ** scm_srfi1_append_reverse [C]
1783 ** scm_srfi1_append_reverse_x [C]
1791 ** Build problems have been fixed on MacOS, SunOS, and QNX.
1793 ** `strftime' fix sign of %z timezone offset.
1795 ** A one-dimensional array can now be 'equal?' to a vector.
1797 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
1799 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
1801 ** `record-accessor' and `record-modifier' now have strict type checks.
1803 Record accessor and modifier procedures now throw an error if the
1804 record type of the record they're given is not the type expected.
1805 (Previously accessors returned #f and modifiers silently did nothing).
1807 ** It is now OK to use both autoload and use-modules on a given module.
1809 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
1811 Previously there was no checking on primatives like make-vector that
1812 accept "one or two" arguments. Now there is.
1814 ** The srfi-1 assoc function now calls its equality predicate properly.
1816 Previously srfi-1 assoc would call the equality predicate with the key
1817 last. According to the SRFI, the key should be first.
1819 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
1821 ** The array-set! procedure no longer segfaults when given a bit vector.
1823 ** Bugs in make-shared-array have been fixed.
1825 ** string<? and friends now follow char<? etc order on 8-bit chars.
1827 ** The format procedure now handles inf and nan values for ~f correctly.
1829 ** exact->inexact should no longer overflow when given certain large fractions.
1831 ** srfi-9 accessor and modifier procedures now have strict record type checks.
1833 This matches the srfi-9 specification.
1835 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
1837 Previously the (ice-9 ftw) procedures would ignore any file that had
1838 the same inode number as a file they had already seen, even if that
1839 file was on a different device.
1842 Changes in 1.8.0 (changes since the 1.6.x series):
1844 * Changes to the distribution
1846 ** Guile is now licensed with the GNU Lesser General Public License.
1848 ** The manual is now licensed with the GNU Free Documentation License.
1850 ** Guile now requires GNU MP (http://swox.com/gmp).
1852 Guile now uses the GNU MP library for arbitrary precision arithmetic.
1854 ** Guile now has separate private and public configuration headers.
1856 That is, things like HAVE_STRING_H no longer leak from Guile's
1859 ** Guile now provides and uses an "effective" version number.
1861 Guile now provides scm_effective_version and effective-version
1862 functions which return the "effective" version number. This is just
1863 the normal full version string without the final micro-version number,
1864 so the current effective-version is "1.8". The effective version
1865 should remain unchanged during a stable series, and should be used for
1866 items like the versioned share directory name
1867 i.e. /usr/share/guile/1.8.
1869 Providing an unchanging version number during a stable release for
1870 things like the versioned share directory can be particularly
1871 important for Guile "add-on" packages, since it provides a directory
1872 that they can install to that won't be changed out from under them
1873 with each micro release during a stable series.
1875 ** Thread implementation has changed.
1877 When you configure "--with-threads=null", you will get the usual
1878 threading API (call-with-new-thread, make-mutex, etc), but you can't
1879 actually create new threads. Also, "--with-threads=no" is now
1880 equivalent to "--with-threads=null". This means that the thread API
1881 is always present, although you might not be able to create new
1884 When you configure "--with-threads=pthreads" or "--with-threads=yes",
1885 you will get threads that are implemented with the portable POSIX
1886 threads. These threads can run concurrently (unlike the previous
1887 "coop" thread implementation), but need to cooperate for things like
1890 The default is "pthreads", unless your platform doesn't have pthreads,
1891 in which case "null" threads are used.
1893 See the manual for details, nodes "Initialization", "Multi-Threading",
1894 "Blocking", and others.
1896 ** There is the new notion of 'discouraged' features.
1898 This is a milder form of deprecation.
1900 Things that are discouraged should not be used in new code, but it is
1901 OK to leave them in old code for now. When a discouraged feature is
1902 used, no warning message is printed like there is for 'deprecated'
1903 features. Also, things that are merely discouraged are nevertheless
1904 implemented efficiently, while deprecated features can be very slow.
1906 You can omit discouraged features from libguile by configuring it with
1907 the '--disable-discouraged' option.
1909 ** Deprecation warnings can be controlled at run-time.
1911 (debug-enable 'warn-deprecated) switches them on and (debug-disable
1912 'warn-deprecated) switches them off.
1914 ** Support for SRFI 61, extended cond syntax for multiple values has
1917 This SRFI is always available.
1919 ** Support for require-extension, SRFI-55, has been added.
1921 The SRFI-55 special form `require-extension' has been added. It is
1922 available at startup, and provides a portable way to load Scheme
1923 extensions. SRFI-55 only requires support for one type of extension,
1924 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
1927 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
1929 The (srfi srfi-26) module is an implementation of SRFI-26 which
1930 provides the `cut' and `cute' syntax. These may be used to specialize
1931 parameters without currying.
1933 ** New module (srfi srfi-31)
1935 This is an implementation of SRFI-31 which provides a special form
1936 `rec' for recursive evaluation.
1938 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
1939 been merged with the core, making their functionality always
1942 The modules are still available, tho, and you could use them together
1943 with a renaming import, for example.
1945 ** Guile no longer includes its own version of libltdl.
1947 The official version is good enough now.
1949 ** The --enable-htmldoc option has been removed from 'configure'.
1951 Support for translating the documentation into HTML is now always
1952 provided. Use 'make html'.
1954 ** New module (ice-9 serialize):
1956 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
1957 don't trust the thread safety of most of your program, but where you
1958 have some section(s) of code which you consider can run in parallel to
1959 other sections. See ice-9/serialize.scm for more information.
1961 ** The configure option '--disable-arrays' has been removed.
1963 Support for arrays and uniform numeric arrays is now always included
1966 * Changes to the stand-alone interpreter
1968 ** New command line option `-L'.
1970 This option adds a directory to the front of the load path.
1972 ** New command line option `--no-debug'.
1974 Specifying `--no-debug' on the command line will keep the debugging
1975 evaluator turned off, even for interactive sessions.
1977 ** User-init file ~/.guile is now loaded with the debugging evaluator.
1979 Previously, the normal evaluator would have been used. Using the
1980 debugging evaluator gives better error messages.
1982 ** The '-e' option now 'read's its argument.
1984 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
1985 be used with '-e'. For example, you can now write a script like
1988 exec guile -e '(@ (demo) main)' -s "$0" "$@"
1991 (define-module (demo)
1995 (format #t "Demo: ~a~%" args))
1998 * Changes to Scheme functions and syntax
2000 ** Guardians have changed back to their original semantics
2002 Guardians now behave like described in the paper by Dybvig et al. In
2003 particular, they no longer make guarantees about the order in which
2004 they return objects, and they can no longer be greedy.
2006 They no longer drop cyclic data structures.
2008 The C function scm_make_guardian has been changed incompatibly and no
2009 longer takes the 'greedy_p' argument.
2011 ** New function hashx-remove!
2013 This function completes the set of 'hashx' functions.
2015 ** The concept of dynamic roots has been factored into continuation
2016 barriers and dynamic states.
2018 Each thread has a current dynamic state that carries the values of the
2019 fluids. You can create and copy dynamic states and use them as the
2020 second argument for 'eval'. See "Fluids and Dynamic States" in the
2023 To restrict the influence that captured continuations can have on the
2024 control flow, you can errect continuation barriers. See "Continuation
2025 Barriers" in the manual.
2027 The function call-with-dynamic-root now essentially temporarily
2028 installs a new dynamic state and errects a continuation barrier.
2030 ** The default load path no longer includes "." at the end.
2032 Automatically loading modules from the current directory should not
2033 happen by default. If you want to allow it in a more controlled
2034 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
2035 variable %load-path.
2037 ** The uniform vector and array support has been overhauled.
2039 It now complies with SRFI-4 and the weird prototype based uniform
2040 array creation has been deprecated. See the manual for more details.
2042 Some non-compatible changes have been made:
2043 - characters can no longer be stored into byte arrays.
2044 - strings and bit vectors are no longer considered to be uniform numeric
2046 - array-rank throws an error for non-arrays instead of returning zero.
2047 - array-ref does no longer accept non-arrays when no indices are given.
2049 There is the new notion of 'generalized vectors' and corresponding
2050 procedures like 'generalized-vector-ref'. Generalized vectors include
2051 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
2053 Arrays use generalized vectors as their storage, so that you still
2054 have arrays of characters, bits, etc. However, uniform-array-read!
2055 and uniform-array-write can no longer read/write strings and
2058 ** There is now support for copy-on-write substrings, mutation-sharing
2059 substrings and read-only strings.
2061 Three new procedures are related to this: substring/shared,
2062 substring/copy, and substring/read-only. See the manual for more
2065 ** Backtraces will now highlight the value that caused the error.
2067 By default, these values are enclosed in "{...}", such as in this
2076 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
2077 <unnamed port>:1:1: Wrong type (expecting pair): a
2078 ABORT: (wrong-type-arg)
2080 The prefix and suffix used for highlighting can be set via the two new
2081 printer options 'highlight-prefix' and 'highlight-suffix'. For
2082 example, putting this into ~/.guile will output the bad value in bold
2083 on an ANSI terminal:
2085 (print-set! highlight-prefix "\x1b[1m")
2086 (print-set! highlight-suffix "\x1b[22m")
2089 ** 'gettext' support for internationalization has been added.
2091 See the manual for details.
2093 ** New syntax '@' and '@@':
2095 You can now directly refer to variables exported from a module by
2098 (@ MODULE-NAME VARIABLE-NAME)
2100 For example (@ (ice-9 pretty-print) pretty-print) will directly access
2101 the pretty-print variable exported from the (ice-9 pretty-print)
2102 module. You don't need to 'use' that module first. You can also use
2103 '@' as a target of 'set!', as in (set! (@ mod var) val).
2105 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
2106 but it can also access variables that have not been exported. It is
2107 intended only for kluges and temporary fixes and for debugging, not
2110 ** Keyword syntax has been made more disciplined.
2112 Previously, the name of a keyword was read as a 'token' but printed as
2113 a symbol. Now, it is read as a general Scheme datum which must be a
2124 ERROR: In expression (a b c):
2128 ERROR: Unbound variable: foo
2133 ERROR: Wrong type (expecting symbol): 12
2137 ERROR: Wrong type (expecting symbol): (a b c)
2141 ** The printing of symbols that might look like keywords can be
2144 The new printer option 'quote-keywordish-symbols' controls how symbols
2145 are printed that have a colon as their first or last character. The
2146 default now is to only quote a symbol with #{...}# when the read
2147 option 'keywords' is not '#f'. Thus:
2149 guile> (define foo (string->symbol ":foo"))
2150 guile> (read-set! keywords #f)
2153 guile> (read-set! keywords 'prefix)
2156 guile> (print-set! quote-keywordish-symbols #f)
2160 ** 'while' now provides 'break' and 'continue'
2162 break and continue were previously bound in a while loop, but not
2163 documented, and continue didn't quite work properly. The undocumented
2164 parameter to break which gave a return value for the while has been
2167 ** 'call-with-current-continuation' is now also available under the name
2170 ** The module system now checks for duplicate bindings.
2172 The module system now can check for name conflicts among imported
2175 The behavior can be controlled by specifying one or more 'duplicates'
2176 handlers. For example, to make Guile return an error for every name
2179 (define-module (foo)
2184 The new default behavior of the module system when a name collision
2185 has been detected is to
2187 1. Give priority to bindings marked as a replacement.
2188 2. Issue a warning (different warning if overriding core binding).
2189 3. Give priority to the last encountered binding (this corresponds to
2192 If you want the old behavior back without replacements or warnings you
2195 (default-duplicate-binding-handler 'last)
2197 to your .guile init file.
2199 ** New define-module option: :replace
2201 :replace works as :export, but, in addition, marks the binding as a
2204 A typical example is `format' in (ice-9 format) which is a replacement
2205 for the core binding `format'.
2207 ** Adding prefixes to imported bindings in the module system
2209 There is now a new :use-module option :prefix. It can be used to add
2210 a prefix to all imported bindings.
2212 (define-module (foo)
2213 :use-module ((bar) :prefix bar:))
2215 will import all bindings exported from bar, but rename them by adding
2218 ** Conflicting generic functions can be automatically merged.
2220 When two imported bindings conflict and they are both generic
2221 functions, the two functions can now be merged automatically. This is
2222 activated with the 'duplicates' handler 'merge-generics'.
2224 ** New function: effective-version
2226 Returns the "effective" version number. This is just the normal full
2227 version string without the final micro-version number. See "Changes
2228 to the distribution" above.
2230 ** New threading functions: parallel, letpar, par-map, and friends
2232 These are convenient ways to run calculations in parallel in new
2233 threads. See "Parallel forms" in the manual for details.
2235 ** New function 'try-mutex'.
2237 This function will attempt to lock a mutex but will return immediately
2238 instead of blocking and indicate failure.
2240 ** Waiting on a condition variable can have a timeout.
2242 The function 'wait-condition-variable' now takes a third, optional
2243 argument that specifies the point in time where the waiting should be
2246 ** New function 'broadcast-condition-variable'.
2248 ** New functions 'all-threads' and 'current-thread'.
2250 ** Signals and system asyncs work better with threads.
2252 The function 'sigaction' now takes a fourth, optional, argument that
2253 specifies the thread that the handler should run in. When the
2254 argument is omitted, the handler will run in the thread that called
2257 Likewise, 'system-async-mark' takes a second, optional, argument that
2258 specifies the thread that the async should run in. When it is
2259 omitted, the async will run in the thread that called
2260 'system-async-mark'.
2262 C code can use the new functions scm_sigaction_for_thread and
2263 scm_system_async_mark_for_thread to pass the new thread argument.
2265 When a thread blocks on a mutex, a condition variable or is waiting
2266 for IO to be possible, it will still execute system asyncs. This can
2267 be used to interrupt such a thread by making it execute a 'throw', for
2270 ** The function 'system-async' is deprecated.
2272 You can now pass any zero-argument procedure to 'system-async-mark'.
2273 The function 'system-async' will just return its argument unchanged
2276 ** New functions 'call-with-blocked-asyncs' and
2277 'call-with-unblocked-asyncs'
2279 The expression (call-with-blocked-asyncs PROC) will call PROC and will
2280 block execution of system asyncs for the current thread by one level
2281 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
2282 procedure and will unblock the execution of system asyncs by one
2283 level for the current thread.
2285 Only system asyncs are affected by these functions.
2287 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
2289 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
2290 instead. Those functions are easier to use correctly and can be
2293 ** New function 'unsetenv'.
2295 ** New macro 'define-syntax-public'.
2297 It works like 'define-syntax' and also exports the defined macro (but
2300 ** There is support for Infinity and NaNs.
2302 Following PLT Scheme, Guile can now work with infinite numbers, and
2305 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
2306 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
2307 "+nan.0"). These numbers are inexact and have no exact counterpart.
2309 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
2310 sign of the dividend. The infinities are integers, and they answer #t
2311 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
2312 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
2323 ERROR: Numerical overflow
2325 Two new predicates 'inf?' and 'nan?' can be used to test for the
2328 ** Inexact zero can have a sign.
2330 Guile can now distinguish between plus and minus inexact zero, if your
2331 platform supports this, too. The two zeros are equal according to
2332 '=', but not according to 'eqv?'. For example
2343 ** Guile now has exact rationals.
2345 Guile can now represent fractions such as 1/3 exactly. Computing with
2346 them is also done exactly, of course:
2351 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
2352 for exact arguments.
2354 For example: (floor 2) now returns an exact 2 where in the past it
2355 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
2357 ** inexact->exact no longer returns only integers.
2359 Without exact rationals, the closest exact number was always an
2360 integer, but now inexact->exact returns the fraction that is exactly
2361 equal to a floating point number. For example:
2363 (inexact->exact 1.234)
2364 => 694680242521899/562949953421312
2366 When you want the old behavior, use 'round' explicitly:
2368 (inexact->exact (round 1.234))
2371 ** New function 'rationalize'.
2373 This function finds a simple fraction that is close to a given real
2374 number. For example (and compare with inexact->exact above):
2376 (rationalize (inexact->exact 1.234) 1/2000)
2379 Note that, as required by R5RS, rationalize returns only then an exact
2380 result when both its arguments are exact.
2382 ** 'odd?' and 'even?' work also for inexact integers.
2384 Previously, (odd? 1.0) would signal an error since only exact integers
2385 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
2386 returns #f and (odd? 1.5) signals an error.
2388 ** Guile now has uninterned symbols.
2390 The new function 'make-symbol' will return an uninterned symbol. This
2391 is a symbol that is unique and is guaranteed to remain unique.
2392 However, uninterned symbols can not yet be read back in.
2394 Use the new function 'symbol-interned?' to check whether a symbol is
2397 ** pretty-print has more options.
2399 The function pretty-print from the (ice-9 pretty-print) module can now
2400 also be invoked with keyword arguments that control things like
2401 maximum output width. See the manual for details.
2403 ** Variables have no longer a special behavior for `equal?'.
2405 Previously, comparing two variables with `equal?' would recursivly
2406 compare their values. This is no longer done. Variables are now only
2407 `equal?' if they are `eq?'.
2409 ** `(begin)' is now valid.
2411 You can now use an empty `begin' form. It will yield #<unspecified>
2412 when evaluated and simply be ignored in a definition context.
2414 ** Deprecated: procedure->macro
2416 Change your code to use 'define-macro' or r5rs macros. Also, be aware
2417 that macro expansion will not be done during evaluation, but prior to
2420 ** Soft ports now allow a `char-ready?' procedure
2422 The vector argument to `make-soft-port' can now have a length of
2423 either 5 or 6. (Previously the length had to be 5.) The optional 6th
2424 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
2425 that returns the number of characters that can be read immediately
2426 without the soft port blocking.
2428 ** Deprecated: undefine
2430 There is no replacement for undefine.
2432 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
2433 have been discouraged.
2435 They are relics from a time where a keyword like #:foo was used
2436 directly as a Tcl option "-foo" and thus keywords were internally
2437 stored as a symbol with a starting dash. We now store a symbol
2440 Use symbol->keyword and keyword->symbol instead.
2442 ** The `cheap' debug option is now obsolete
2444 Evaluator trap calls are now unconditionally "cheap" - in other words,
2445 they pass a debug object to the trap handler rather than a full
2446 continuation. The trap handler code can capture a full continuation
2447 by using `call-with-current-continuation' in the usual way, if it so
2450 The `cheap' option is retained for now so as not to break existing
2451 code which gets or sets it, but setting it now has no effect. It will
2452 be removed in the next major Guile release.
2454 ** Evaluator trap calls now support `tweaking'
2456 `Tweaking' means that the trap handler code can modify the Scheme
2457 expression that is about to be evaluated (in the case of an
2458 enter-frame trap) or the value that is being returned (in the case of
2459 an exit-frame trap). The trap handler code indicates that it wants to
2460 do this by returning a pair whose car is the symbol 'instead and whose
2461 cdr is the modified expression or return value.
2463 * Changes to the C interface
2465 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
2466 take a 'delete' function argument.
2468 This argument makes no sense since the delete function is used to
2469 remove a pair from an alist, and this must not be configurable.
2471 This is an incompatible change.
2473 ** The GH interface is now subject to the deprecation mechanism
2475 The GH interface has been deprecated for quite some time but now it is
2476 actually removed from Guile when it is configured with
2477 --disable-deprecated.
2479 See the manual "Transitioning away from GH" for more information.
2481 ** A new family of functions for converting between C values and
2482 Scheme values has been added.
2484 These functions follow a common naming scheme and are designed to be
2485 easier to use, thread-safe and more future-proof than the older
2488 - int scm_is_* (...)
2490 These are predicates that return a C boolean: 1 or 0. Instead of
2491 SCM_NFALSEP, you can now use scm_is_true, for example.
2493 - <type> scm_to_<type> (SCM val, ...)
2495 These are functions that convert a Scheme value into an appropriate
2496 C value. For example, you can use scm_to_int to safely convert from
2499 - SCM scm_from_<type> (<type> val, ...)
2501 These functions convert from a C type to a SCM value; for example,
2502 scm_from_int for ints.
2504 There is a huge number of these functions, for numbers, strings,
2505 symbols, vectors, etc. They are documented in the reference manual in
2506 the API section together with the types that they apply to.
2508 ** New functions for dealing with complex numbers in C have been added.
2510 The new functions are scm_c_make_rectangular, scm_c_make_polar,
2511 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
2512 They work like scm_make_rectangular etc but take or return doubles
2515 ** The function scm_make_complex has been discouraged.
2517 Use scm_c_make_rectangular instead.
2519 ** The INUM macros have been deprecated.
2521 A lot of code uses these macros to do general integer conversions,
2522 although the macros only work correctly with fixnums. Use the
2523 following alternatives.
2525 SCM_INUMP -> scm_is_integer or similar
2526 SCM_NINUMP -> !scm_is_integer or similar
2527 SCM_MAKINUM -> scm_from_int or similar
2528 SCM_INUM -> scm_to_int or similar
2530 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
2531 do the validating for you.
2533 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
2534 have been discouraged.
2536 Use the newer scm_to_<type> and scm_from_<type> functions instead for
2537 new code. The functions have been discouraged since they don't fit
2540 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
2542 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
2543 evaluates its argument twice. Use scm_is_true, etc. instead for new
2546 ** The macro SCM_EQ_P has been discouraged.
2548 Use scm_is_eq for new code, which fits better into the naming
2551 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
2554 Use the function scm_is_pair or scm_is_null instead.
2556 ** The functions scm_round and scm_truncate have been deprecated and
2557 are now available as scm_c_round and scm_c_truncate, respectively.
2559 These functions occupy the names that scm_round_number and
2560 scm_truncate_number should have.
2562 ** The functions scm_c_string2str, scm_c_substring2str, and
2563 scm_c_symbol2str have been deprecated.
2565 Use scm_to_locale_stringbuf or similar instead, maybe together with
2568 ** New functions scm_c_make_string, scm_c_string_length,
2569 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
2570 scm_c_substring_shared, scm_c_substring_copy.
2572 These are like scm_make_string, scm_length, etc. but are slightly
2573 easier to use from C.
2575 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
2576 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
2578 They export too many assumptions about the implementation of strings
2579 and symbols that are no longer true in the presence of
2580 mutation-sharing substrings and when Guile switches to some form of
2583 When working with strings, it is often best to use the normal string
2584 functions provided by Guile, such as scm_c_string_ref,
2585 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
2586 manual since many more such functions are now provided than
2589 When you want to convert a SCM string to a C string, use the
2590 scm_to_locale_string function or similar instead. For symbols, use
2591 scm_symbol_to_string and then work with that string. Because of the
2592 new string representation, scm_symbol_to_string does not need to copy
2593 and is thus quite efficient.
2595 ** Some string, symbol and keyword functions have been discouraged.
2597 They don't fit into the uniform naming scheme and are not explicit
2598 about the character encoding.
2600 Replace according to the following table:
2602 scm_allocate_string -> scm_c_make_string
2603 scm_take_str -> scm_take_locale_stringn
2604 scm_take0str -> scm_take_locale_string
2605 scm_mem2string -> scm_from_locale_stringn
2606 scm_str2string -> scm_from_locale_string
2607 scm_makfrom0str -> scm_from_locale_string
2608 scm_mem2symbol -> scm_from_locale_symboln
2609 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
2610 scm_str2symbol -> scm_from_locale_symbol
2612 SCM_SYMBOL_HASH -> scm_hashq
2613 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
2615 scm_c_make_keyword -> scm_from_locale_keyword
2617 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
2618 now also available to C code.
2620 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
2622 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
2623 the latter returns the true name of the keyword, not the 'dash name',
2624 as SCM_KEYWORDSYM used to do.
2626 ** A new way to access arrays in a thread-safe and efficient way has
2629 See the manual, node "Accessing Arrays From C".
2631 ** The old uniform vector and bitvector implementations have been
2632 unceremoniously removed.
2634 This implementation exposed the details of the tagging system of
2635 Guile. Use the new C API explained in the manual in node "Uniform
2636 Numeric Vectors" and "Bit Vectors", respectively.
2638 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
2639 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
2640 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
2641 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
2642 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
2643 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
2646 ** The macros dealing with vectors have been deprecated.
2648 Use the new functions scm_is_vector, scm_vector_elements,
2649 scm_vector_writable_elements, etc, or scm_is_simple_vector,
2650 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
2651 manual for more details.
2653 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
2654 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
2656 The following macros have been removed: SCM_VECTOR_BASE,
2657 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
2658 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
2660 ** Some C functions and macros related to arrays have been deprecated.
2662 Migrate according to the following table:
2664 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
2665 scm_make_ra -> scm_make_array
2666 scm_shap2ra -> scm_make_array
2667 scm_cvref -> scm_c_generalized_vector_ref
2668 scm_ra_set_contp -> do not use
2669 scm_aind -> scm_array_handle_pos
2670 scm_raprin1 -> scm_display or scm_write
2672 SCM_ARRAYP -> scm_is_array
2673 SCM_ARRAY_NDIM -> scm_c_array_rank
2674 SCM_ARRAY_DIMS -> scm_array_handle_dims
2675 SCM_ARRAY_CONTP -> do not use
2676 SCM_ARRAY_MEM -> do not use
2677 SCM_ARRAY_V -> scm_array_handle_elements or similar
2678 SCM_ARRAY_BASE -> do not use
2680 ** SCM_CELL_WORD_LOC has been deprecated.
2682 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
2683 to a SCM, as opposed to a pointer to a scm_t_bits.
2685 This was done to allow the correct use of pointers into the Scheme
2686 heap. Previously, the heap words were of type scm_t_bits and local
2687 variables and function arguments were of type SCM, making it
2688 non-standards-conformant to have a pointer that can point to both.
2690 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
2692 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
2693 second and third words of double smobs. Likewise for
2694 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
2696 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
2697 used to get and set the 16 exra bits in the zeroth word of a smob.
2699 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
2700 accesing the first immediate word of a smob as a SCM value, and there
2701 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
2702 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
2704 ** New way to deal with non-local exits and re-entries.
2706 There is a new set of functions that essentially do what
2707 scm_internal_dynamic_wind does, but in a way that is more convenient
2708 for C code in some situations. Here is a quick example of how to
2709 prevent a potential memory leak:
2716 scm_dynwind_begin (0);
2718 mem = scm_malloc (100);
2719 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
2721 /* MEM would leak if BAR throws an error.
2722 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
2729 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
2730 SCM_DYNWIND_END as well.
2734 For full documentation, see the node "Dynamic Wind" in the manual.
2736 ** New function scm_dynwind_free
2738 This function calls 'free' on a given pointer when a dynwind context
2739 is left. Thus the call to scm_dynwind_unwind_handler above could be
2740 replaced with simply scm_dynwind_free (mem).
2742 ** New functions scm_c_call_with_blocked_asyncs and
2743 scm_c_call_with_unblocked_asyncs
2745 Like scm_call_with_blocked_asyncs etc. but for C functions.
2747 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
2749 In addition to scm_c_call_with_blocked_asyncs you can now also use
2750 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
2751 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
2753 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
2754 SCM_REALLOW_INTS have been deprecated.
2756 They do no longer fulfill their original role of blocking signal
2757 delivery. Depending on what you want to achieve, replace a pair of
2758 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
2759 mutex, blocks asyncs, or both. See node "Critical Sections" in the
2762 ** The value 'scm_mask_ints' is no longer writable.
2764 Previously, you could set scm_mask_ints directly. This is no longer
2765 possible. Use scm_c_call_with_blocked_asyncs and
2766 scm_c_call_with_unblocked_asyncs instead.
2768 ** New way to temporarily set the current input, output or error ports
2770 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
2771 context' (see above). <foo> is one of "input", "output" or "error".
2773 ** New way to temporarily set fluids
2775 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
2776 above) to temporarily set the value of a fluid.
2778 ** New types scm_t_intmax and scm_t_uintmax.
2780 On platforms that have them, these types are identical to intmax_t and
2781 uintmax_t, respectively. On other platforms, they are identical to
2782 the largest integer types that Guile knows about.
2784 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
2786 You should not have used them.
2788 ** Many public #defines with generic names have been made private.
2790 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
2791 private or renamed with a more suitable public name.
2793 ** The macro SCM_TYP16S has been deprecated.
2795 This macro is not intended for public use.
2797 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
2799 Use scm_is_true (scm_inexact_p (...)) instead.
2801 ** The macro SCM_SLOPPY_REALP has been deprecated.
2803 Use scm_is_real instead.
2805 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
2807 Use scm_is_complex instead.
2809 ** Some preprocessor defines have been deprecated.
2811 These defines indicated whether a certain feature was present in Guile
2812 or not. Going forward, assume that the features are always present.
2814 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
2815 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
2817 The following macros have been removed completely: MEMOIZE_LOCALS,
2818 SCM_RECKLESS, SCM_CAUTIOUS.
2820 ** The preprocessor define STACK_DIRECTION has been deprecated.
2822 There should be no need to know about the stack direction for ordinary
2825 ** New function: scm_effective_version
2827 Returns the "effective" version number. This is just the normal full
2828 version string without the final micro-version number. See "Changes
2829 to the distribution" above.
2831 ** The function scm_call_with_new_thread has a new prototype.
2833 Instead of taking a list with the thunk and handler, these two
2834 arguments are now passed directly:
2836 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
2838 This is an incompatible change.
2840 ** New snarfer macro SCM_DEFINE_PUBLIC.
2842 This is like SCM_DEFINE, but also calls scm_c_export for the defined
2843 function in the init section.
2845 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
2847 ** Garbage collector rewrite.
2849 The garbage collector is cleaned up a lot, and now uses lazy
2850 sweeping. This is reflected in the output of (gc-stats); since cells
2851 are being freed when they are allocated, the cells-allocated field
2852 stays roughly constant.
2854 For malloc related triggers, the behavior is changed. It uses the same
2855 heuristic as the cell-triggered collections. It may be tuned with the
2856 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
2857 for minimum yield of malloc related triggers. The default is 40.
2858 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
2861 Debugging operations for the freelist have been deprecated, along with
2862 the C variables that control garbage collection. The environment
2863 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
2864 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
2866 For understanding the memory usage of a GUILE program, the routine
2867 gc-live-object-stats returns an alist containing the number of live
2868 objects for every type.
2871 ** The function scm_definedp has been renamed to scm_defined_p
2873 The name scm_definedp is deprecated.
2875 ** The struct scm_cell type has been renamed to scm_t_cell
2877 This is in accordance to Guile's naming scheme for types. Note that
2878 the name scm_cell is now used for a function that allocates and
2879 initializes a new cell (see below).
2881 ** New functions for memory management
2883 A new set of functions for memory management has been added since the
2884 old way (scm_must_malloc, scm_must_free, etc) was error prone and
2885 indeed, Guile itself contained some long standing bugs that could
2886 cause aborts in long running programs.
2888 The new functions are more symmetrical and do not need cooperation
2889 from smob free routines, among other improvements.
2891 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
2892 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
2893 scm_gc_free, scm_gc_register_collectable_memory, and
2894 scm_gc_unregister_collectable_memory. Refer to the manual for more
2895 details and for upgrading instructions.
2897 The old functions for memory management have been deprecated. They
2898 are: scm_must_malloc, scm_must_realloc, scm_must_free,
2899 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
2901 ** Declarations of exported features are marked with SCM_API.
2903 Every declaration of a feature that belongs to the exported Guile API
2904 has been marked by adding the macro "SCM_API" to the start of the
2905 declaration. This macro can expand into different things, the most
2906 common of which is just "extern" for Unix platforms. On Win32, it can
2907 be used to control which symbols are exported from a DLL.
2909 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
2910 will expand into "__declspec (dllimport) extern", which is needed for
2911 linking to the Guile DLL in Windows.
2913 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
2914 SCM_SRFI4_IMPORT, for the corresponding libraries.
2916 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
2918 Use the new functions scm_cell and scm_double_cell instead. The old
2919 macros had problems because with them allocation and initialization
2920 was separated and the GC could sometimes observe half initialized
2921 cells. Only careful coding by the user of SCM_NEWCELL and
2922 SCM_NEWCELL2 could make this safe and efficient.
2924 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
2926 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
2929 ** SRCBRKP has been deprecated.
2931 Use scm_c_source_property_breakpoint_p instead.
2933 ** Deprecated: scm_makmacro
2935 Change your code to use either scm_makmmacro or to define macros in
2936 Scheme, using 'define-macro'.
2938 ** New function scm_c_port_for_each.
2940 This function is like scm_port_for_each but takes a pointer to a C
2941 function as the callback instead of a SCM value.
2943 ** The names scm_internal_select, scm_thread_sleep, and
2944 scm_thread_usleep have been discouraged.
2946 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
2948 ** The GC can no longer be blocked.
2950 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
2951 The GC can now run (partially) concurrently with other code and thus
2952 blocking it is not well defined.
2954 ** Many definitions have been removed that were previously deprecated.
2956 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
2957 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
2958 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
2959 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
2960 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
2961 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
2962 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
2963 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
2964 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
2965 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
2966 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
2967 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
2968 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
2969 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
2970 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
2971 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
2972 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
2973 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
2974 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
2975 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
2976 scm_vector_set_length_x, scm_contregs, scm_debug_info,
2977 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
2978 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
2979 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
2980 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
2981 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
2982 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
2983 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
2984 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
2985 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
2986 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
2988 * Changes to bundled modules
2992 Using the (ice-9 debug) module no longer automatically switches Guile
2993 to use the debugging evaluator. If you want to switch to the
2994 debugging evaluator (which is needed for backtrace information if you
2995 hit an error), please add an explicit "(debug-enable 'debug)" to your
2996 code just after the code to use (ice-9 debug).
2999 Changes since Guile 1.4:
3001 * Changes to the distribution
3003 ** A top-level TODO file is included.
3005 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
3007 Guile now always uses three numbers to represent the version,
3008 i.e. "1.6.5". The first number, 1, is the major version number, the
3009 second number, 6, is the minor version number, and the third number,
3010 5, is the micro version number. Changes in major version number
3011 indicate major changes in Guile.
3013 Minor version numbers that are even denote stable releases, and odd
3014 minor version numbers denote development versions (which may be
3015 unstable). The micro version number indicates a minor sub-revision of
3016 a given MAJOR.MINOR release.
3018 In keeping with the new scheme, (minor-version) and scm_minor_version
3019 no longer return everything but the major version number. They now
3020 just return the minor version number. Two new functions
3021 (micro-version) and scm_micro_version have been added to report the
3022 micro version number.
3024 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
3026 ** New preprocessor definitions are available for checking versions.
3028 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
3029 SCM_MICRO_VERSION to the appropriate integer values.
3031 ** Guile now actively warns about deprecated features.
3033 The new configure option `--enable-deprecated=LEVEL' and the
3034 environment variable GUILE_WARN_DEPRECATED control this mechanism.
3035 See INSTALL and README for more information.
3037 ** Guile is much more likely to work on 64-bit architectures.
3039 Guile now compiles and passes "make check" with only two UNRESOLVED GC
3040 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
3041 for the use of a test machine, and thanks to Stefan Jahn for ia64
3044 ** New functions: setitimer and getitimer.
3046 These implement a fairly direct interface to the libc functions of the
3049 ** The #. reader extension is now disabled by default.
3051 For safety reasons, #. evaluation is disabled by default. To
3052 re-enable it, set the fluid read-eval? to #t. For example:
3054 (fluid-set! read-eval? #t)
3056 but make sure you realize the potential security risks involved. With
3057 read-eval? enabled, reading a data file from an untrusted source can
3060 ** New SRFI modules have been added:
3062 SRFI-0 `cond-expand' is now supported in Guile, without requiring
3065 (srfi srfi-1) is a library containing many useful pair- and list-processing
3068 (srfi srfi-2) exports and-let*.
3070 (srfi srfi-4) implements homogeneous numeric vector datatypes.
3072 (srfi srfi-6) is a dummy module for now, since guile already provides
3073 all of the srfi-6 procedures by default: open-input-string,
3074 open-output-string, get-output-string.
3076 (srfi srfi-8) exports receive.
3078 (srfi srfi-9) exports define-record-type.
3080 (srfi srfi-10) exports define-reader-ctor and implements the reader
3083 (srfi srfi-11) exports let-values and let*-values.
3085 (srfi srfi-13) implements the SRFI String Library.
3087 (srfi srfi-14) implements the SRFI Character-Set Library.
3089 (srfi srfi-17) implements setter and getter-with-setter and redefines
3090 some accessor procedures as procedures with getters. (such as car,
3091 cdr, vector-ref etc.)
3093 (srfi srfi-19) implements the SRFI Time/Date Library.
3095 ** New scripts / "executable modules"
3097 Subdirectory "scripts" contains Scheme modules that are packaged to
3098 also be executable as scripts. At this time, these scripts are available:
3107 See README there for more info.
3109 These scripts can be invoked from the shell with the new program
3110 "guile-tools", which keeps track of installation directory for you.
3113 $ guile-tools display-commentary srfi/*.scm
3115 guile-tools is copied to the standard $bindir on "make install".
3117 ** New module (ice-9 stack-catch):
3119 stack-catch is like catch, but saves the current state of the stack in
3120 the fluid the-last-stack. This fluid can be useful when using the
3121 debugger and when re-throwing an error.
3123 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
3125 This has been done to prevent problems on lesser operating systems
3126 that can't tolerate `*'s in file names. The exported macro continues
3127 to be named `and-let*', of course.
3129 On systems that support it, there is also a compatibility module named
3130 (ice-9 and-let*). It will go away in the next release.
3132 ** New modules (oop goops) etc.:
3135 (oop goops describe)
3137 (oop goops active-slot)
3138 (oop goops composite-slot)
3140 The Guile Object Oriented Programming System (GOOPS) has been
3141 integrated into Guile. For further information, consult the GOOPS
3142 manual and tutorial in the `doc' directory.
3144 ** New module (ice-9 rdelim).
3146 This exports the following procedures which were previously defined
3147 in the default environment:
3149 read-line read-line! read-delimited read-delimited! %read-delimited!
3150 %read-line write-line
3152 For backwards compatibility the definitions are still imported into the
3153 default environment in this version of Guile. However you should add:
3155 (use-modules (ice-9 rdelim))
3157 to any program which uses the definitions, since this may change in
3160 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
3161 can be used for similar functionality.
3163 ** New module (ice-9 rw)
3165 This is a subset of the (scsh rw) module from guile-scsh. Currently
3166 it defines two procedures:
3168 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
3170 Read characters from a port or file descriptor into a string STR.
3171 A port must have an underlying file descriptor -- a so-called
3172 fport. This procedure is scsh-compatible and can efficiently read
3175 *** New function: write-string/partial str [port_or_fdes [start [end]]]
3177 Write characters from a string STR to a port or file descriptor.
3178 A port must have an underlying file descriptor -- a so-called
3179 fport. This procedure is mostly compatible and can efficiently
3180 write large strings.
3182 ** New module (ice-9 match)
3184 This module includes Andrew K. Wright's pattern matcher. See
3185 ice-9/match.scm for brief description or
3187 http://www.star-lab.com/wright/code.html
3189 for complete documentation.
3191 ** New module (ice-9 buffered-input)
3193 This module provides procedures to construct an input port from an
3194 underlying source of input that reads and returns its input in chunks.
3195 The underlying input source is a Scheme procedure, specified by the
3196 caller, which the port invokes whenever it needs more input.
3198 This is useful when building an input port whose back end is Readline
3199 or a UI element such as the GtkEntry widget.
3203 The reference and tutorial documentation that was previously
3204 distributed separately, as `guile-doc', is now included in the core
3205 Guile distribution. The documentation consists of the following
3208 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
3211 - The Guile Reference Manual (guile.texi) contains (or is intended to
3212 contain) reference documentation on all aspects of Guile.
3214 - The GOOPS Manual (goops.texi) contains both tutorial-style and
3215 reference documentation for using GOOPS, Guile's Object Oriented
3218 - The Revised^5 Report on the Algorithmic Language Scheme
3221 See the README file in the `doc' directory for more details.
3223 ** There are a couple of examples in the examples/ directory now.
3225 * Changes to the stand-alone interpreter
3227 ** New command line option `--use-srfi'
3229 Using this option, SRFI modules can be loaded on startup and be
3230 available right from the beginning. This makes programming portable
3231 Scheme programs easier.
3233 The option `--use-srfi' expects a comma-separated list of numbers,
3234 each representing a SRFI number to be loaded into the interpreter
3235 before starting evaluating a script file or the REPL. Additionally,
3236 the feature identifier for the loaded SRFIs is recognized by
3237 `cond-expand' when using this option.
3240 $ guile --use-srfi=8,13
3241 guile> (receive (x z) (values 1 2) (+ 1 2))
3243 guile> (string-pad "bla" 20)
3246 ** Guile now always starts up in the `(guile-user)' module.
3248 Previously, scripts executed via the `-s' option would run in the
3249 `(guile)' module and the repl would run in the `(guile-user)' module.
3250 Now every user action takes place in the `(guile-user)' module by
3253 * Changes to Scheme functions and syntax
3255 ** Character classifiers work for non-ASCII characters.
3257 The predicates `char-alphabetic?', `char-numeric?',
3258 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
3259 no longer check whether their arguments are ASCII characters.
3260 Previously, a character would only be considered alphabetic when it
3261 was also ASCII, for example.
3263 ** Previously deprecated Scheme functions have been removed:
3265 tag - no replacement.
3266 fseek - replaced by seek.
3267 list* - replaced by cons*.
3269 ** It's now possible to create modules with controlled environments
3273 (use-modules (ice-9 safe))
3274 (define m (make-safe-module))
3275 ;;; m will now be a module containing only a safe subset of R5RS
3276 (eval '(+ 1 2) m) --> 3
3277 (eval 'load m) --> ERROR: Unbound variable: load
3279 ** Evaluation of "()", the empty list, is now an error.
3281 Previously, the expression "()" evaluated to the empty list. This has
3282 been changed to signal a "missing expression" error. The correct way
3283 to write the empty list as a literal constant is to use quote: "'()".
3285 ** New concept of `Guile Extensions'.
3287 A Guile Extension is just a ordinary shared library that can be linked
3288 at run-time. We found it advantageous to give this simple concept a
3289 dedicated name to distinguish the issues related to shared libraries
3290 from the issues related to the module system.
3292 *** New function: load-extension
3294 Executing (load-extension lib init) is mostly equivalent to
3296 (dynamic-call init (dynamic-link lib))
3298 except when scm_register_extension has been called previously.
3299 Whenever appropriate, you should use `load-extension' instead of
3300 dynamic-link and dynamic-call.
3302 *** New C function: scm_c_register_extension
3304 This function registers a initialization function for use by
3305 `load-extension'. Use it when you don't want specific extensions to
3306 be loaded as shared libraries (for example on platforms that don't
3307 support dynamic linking).
3309 ** Auto-loading of compiled-code modules is deprecated.
3311 Guile used to be able to automatically find and link a shared
3312 library to satisfy requests for a module. For example, the module
3313 `(foo bar)' could be implemented by placing a shared library named
3314 "foo/libbar.so" (or with a different extension) in a directory on the
3317 This has been found to be too tricky, and is no longer supported. The
3318 shared libraries are now called "extensions". You should now write a
3319 small Scheme file that calls `load-extension' to load the shared
3320 library and initialize it explicitly.
3322 The shared libraries themselves should be installed in the usual
3323 places for shared libraries, with names like "libguile-foo-bar".
3325 For example, place this into a file "foo/bar.scm"
3327 (define-module (foo bar))
3329 (load-extension "libguile-foo-bar" "foobar_init")
3331 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
3333 `eval' is now R5RS, that is it takes two arguments.
3334 The second argument is an environment specifier, i.e. either
3336 (scheme-report-environment 5)
3337 (null-environment 5)
3338 (interaction-environment)
3344 ** The module system has been made more disciplined.
3346 The function `eval' will save and restore the current module around
3347 the evaluation of the specified expression. While this expression is
3348 evaluated, `(current-module)' will now return the right module, which
3349 is the module specified as the second argument to `eval'.
3351 A consequence of this change is that `eval' is not particularly
3352 useful when you want allow the evaluated code to change what module is
3353 designated as the current module and have this change persist from one
3354 call to `eval' to the next. The read-eval-print-loop is an example
3355 where `eval' is now inadequate. To compensate, there is a new
3356 function `primitive-eval' that does not take a module specifier and
3357 that does not save/restore the current module. You should use this
3358 function together with `set-current-module', `current-module', etc
3359 when you want to have more control over the state that is carried from
3360 one eval to the next.
3362 Additionally, it has been made sure that forms that are evaluated at
3363 the top level are always evaluated with respect to the current module.
3364 Previously, subforms of top-level forms such as `begin', `case',
3365 etc. did not respect changes to the current module although these
3366 subforms are at the top-level as well.
3368 To prevent strange behavior, the forms `define-module',
3369 `use-modules', `use-syntax', and `export' have been restricted to only
3370 work on the top level. The forms `define-public' and
3371 `defmacro-public' only export the new binding on the top level. They
3372 behave just like `define' and `defmacro', respectively, when they are
3373 used in a lexical environment.
3375 Also, `export' will no longer silently re-export bindings imported
3376 from a used module. It will emit a `deprecation' warning and will
3377 cease to perform any re-export in the next version. If you actually
3378 want to re-export bindings, use the new `re-export' in place of
3379 `export'. The new `re-export' will not make copies of variables when
3380 rexporting them, as `export' did wrongly.
3382 ** Module system now allows selection and renaming of imported bindings
3384 Previously, when using `use-modules' or the `#:use-module' clause in
3385 the `define-module' form, all the bindings (association of symbols to
3386 values) for imported modules were added to the "current module" on an
3387 as-is basis. This has been changed to allow finer control through two
3388 new facilities: selection and renaming.
3390 You can now select which of the imported module's bindings are to be
3391 visible in the current module by using the `:select' clause. This
3392 clause also can be used to rename individual bindings. For example:
3394 ;; import all bindings no questions asked
3395 (use-modules (ice-9 common-list))
3397 ;; import four bindings, renaming two of them;
3398 ;; the current module sees: every some zonk-y zonk-n
3399 (use-modules ((ice-9 common-list)
3401 (remove-if . zonk-y)
3402 (remove-if-not . zonk-n))))
3404 You can also programmatically rename all selected bindings using the
3405 `:renamer' clause, which specifies a proc that takes a symbol and
3406 returns another symbol. Because it is common practice to use a prefix,
3407 we now provide the convenience procedure `symbol-prefix-proc'. For
3410 ;; import four bindings, renaming two of them specifically,
3411 ;; and all four w/ prefix "CL:";
3412 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
3413 (use-modules ((ice-9 common-list)
3415 (remove-if . zonk-y)
3416 (remove-if-not . zonk-n))
3417 :renamer (symbol-prefix-proc 'CL:)))
3419 ;; import four bindings, renaming two of them specifically,
3420 ;; and all four by upcasing.
3421 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
3422 (define (upcase-symbol sym)
3423 (string->symbol (string-upcase (symbol->string sym))))
3425 (use-modules ((ice-9 common-list)
3427 (remove-if . zonk-y)
3428 (remove-if-not . zonk-n))
3429 :renamer upcase-symbol))
3431 Note that programmatic renaming is done *after* individual renaming.
3432 Also, the above examples show `use-modules', but the same facilities are
3433 available for the `#:use-module' clause of `define-module'.
3435 See manual for more info.
3437 ** The semantics of guardians have changed.
3439 The changes are for the most part compatible. An important criterion
3440 was to keep the typical usage of guardians as simple as before, but to
3441 make the semantics safer and (as a result) more useful.
3443 *** All objects returned from guardians are now properly alive.
3445 It is now guaranteed that any object referenced by an object returned
3446 from a guardian is alive. It's now impossible for a guardian to
3447 return a "contained" object before its "containing" object.
3449 One incompatible (but probably not very important) change resulting
3450 from this is that it is no longer possible to guard objects that
3451 indirectly reference themselves (i.e. are parts of cycles). If you do
3452 so accidentally, you'll get a warning.
3454 *** There are now two types of guardians: greedy and sharing.
3456 If you call (make-guardian #t) or just (make-guardian), you'll get a
3457 greedy guardian, and for (make-guardian #f) a sharing guardian.
3459 Greedy guardians are the default because they are more "defensive".
3460 You can only greedily guard an object once. If you guard an object
3461 more than once, once in a greedy guardian and the rest of times in
3462 sharing guardians, then it is guaranteed that the object won't be
3463 returned from sharing guardians as long as it is greedily guarded
3466 Guardians returned by calls to `make-guardian' can now take one more
3467 optional parameter, which says whether to throw an error in case an
3468 attempt is made to greedily guard an object that is already greedily
3469 guarded. The default is true, i.e. throw an error. If the parameter
3470 is false, the guardian invocation returns #t if guarding was
3471 successful and #f if it wasn't.
3473 Also, since greedy guarding is, in effect, a side-effecting operation
3474 on objects, a new function is introduced: `destroy-guardian!'.
3475 Invoking this function on a guardian renders it unoperative and, if
3476 the guardian is greedy, clears the "greedily guarded" property of the
3477 objects that were guarded by it, thus undoing the side effect.
3479 Note that all this hair is hardly very important, since guardian
3480 objects are usually permanent.
3482 ** Continuations created by call-with-current-continuation now accept
3483 any number of arguments, as required by R5RS.
3485 ** New function `issue-deprecation-warning'
3487 This function is used to display the deprecation messages that are
3488 controlled by GUILE_WARN_DEPRECATION as explained in the README.
3491 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
3495 ;; `id' is deprecated. Use `identity' instead.
3500 ** New syntax `begin-deprecated'
3502 When deprecated features are included (as determined by the configure
3503 option --enable-deprecated), `begin-deprecated' is identical to
3504 `begin'. When deprecated features are excluded, it always evaluates
3505 to `#f', ignoring the body forms.
3507 ** New function `make-object-property'
3509 This function returns a new `procedure with setter' P that can be used
3510 to attach a property to objects. When calling P as
3514 where `obj' is any kind of object, it attaches `val' to `obj' in such
3515 a way that it can be retrieved by calling P as
3519 This function will replace procedure properties, symbol properties and
3520 source properties eventually.
3522 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
3524 Instead of #&optional, #&key, etc you should now use #:optional,
3525 #:key, etc. Since #:optional is a keyword, you can write it as just
3526 :optional when (read-set! keywords 'prefix) is active.
3528 The old reader syntax `#&' is still supported, but deprecated. It
3529 will be removed in the next release.
3531 ** New define-module option: pure
3533 Tells the module system not to include any bindings from the root
3538 (define-module (totally-empty-module)
3541 ** New define-module option: export NAME1 ...
3543 Export names NAME1 ...
3545 This option is required if you want to be able to export bindings from
3546 a module which doesn't import one of `define-public' or `export'.
3550 (define-module (foo)
3552 :use-module (ice-9 r5rs)
3555 ;;; Note that we're pure R5RS below this point!
3560 ** New function: object->string OBJ
3562 Return a Scheme string obtained by printing a given object.
3564 ** New function: port? X
3566 Returns a boolean indicating whether X is a port. Equivalent to
3567 `(or (input-port? X) (output-port? X))'.
3569 ** New function: file-port?
3571 Determines whether a given object is a port that is related to a file.
3573 ** New function: port-for-each proc
3575 Apply PROC to each port in the Guile port table in turn. The return
3576 value is unspecified. More specifically, PROC is applied exactly once
3577 to every port that exists in the system at the time PORT-FOR-EACH is
3578 invoked. Changes to the port table while PORT-FOR-EACH is running
3579 have no effect as far as PORT-FOR-EACH is concerned.
3581 ** New function: dup2 oldfd newfd
3583 A simple wrapper for the `dup2' system call. Copies the file
3584 descriptor OLDFD to descriptor number NEWFD, replacing the
3585 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
3586 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
3587 to move away ports which are using NEWFD. The return value is
3590 ** New function: close-fdes fd
3592 A simple wrapper for the `close' system call. Close file
3593 descriptor FD, which must be an integer. Unlike close (*note
3594 close: Ports and File Descriptors.), the file descriptor will be
3595 closed even if a port is using it. The return value is
3598 ** New function: crypt password salt
3600 Encrypts `password' using the standard unix password encryption
3603 ** New function: chroot path
3605 Change the root directory of the running process to `path'.
3607 ** New functions: getlogin, cuserid
3609 Return the login name or the user name of the current effective user
3612 ** New functions: getpriority which who, setpriority which who prio
3614 Get or set the priority of the running process.
3616 ** New function: getpass prompt
3618 Read a password from the terminal, first displaying `prompt' and
3621 ** New function: flock file operation
3623 Set/remove an advisory shared or exclusive lock on `file'.
3625 ** New functions: sethostname name, gethostname
3627 Set or get the hostname of the machine the current process is running
3630 ** New function: mkstemp! tmpl
3632 mkstemp creates a new unique file in the file system and returns a
3633 new buffered port open for reading and writing to the file. TMPL
3634 is a string specifying where the file should be created: it must
3635 end with `XXXXXX' and will be changed in place to return the name
3636 of the temporary file.
3638 ** New function: open-input-string string
3640 Return an input string port which delivers the characters from
3641 `string'. This procedure, together with `open-output-string' and
3642 `get-output-string' implements SRFI-6.
3644 ** New function: open-output-string
3646 Return an output string port which collects all data written to it.
3647 The data can then be retrieved by `get-output-string'.
3649 ** New function: get-output-string
3651 Return the contents of an output string port.
3653 ** New function: identity
3655 Return the argument.
3657 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
3658 are represented in Scheme as integers with normal host byte ordering.
3660 ** New function: inet-pton family address
3662 Convert a printable string network address into an integer. Note that
3663 unlike the C version of this function, the result is an integer with
3664 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3667 (inet-pton AF_INET "127.0.0.1") => 2130706433
3668 (inet-pton AF_INET6 "::1") => 1
3670 ** New function: inet-ntop family address
3672 Convert an integer network address into a printable string. Note that
3673 unlike the C version of this function, the input is an integer with
3674 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3677 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
3678 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
3679 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
3683 Use `identity' instead.
3689 ** Deprecated: return-it
3693 ** Deprecated: string-character-length
3695 Use `string-length' instead.
3697 ** Deprecated: flags
3699 Use `logior' instead.
3701 ** Deprecated: close-all-ports-except.
3703 This was intended for closing ports in a child process after a fork,
3704 but it has the undesirable side effect of flushing buffers.
3705 port-for-each is more flexible.
3707 ** The (ice-9 popen) module now attempts to set up file descriptors in
3708 the child process from the current Scheme ports, instead of using the
3709 current values of file descriptors 0, 1, and 2 in the parent process.
3711 ** Removed function: builtin-weak-bindings
3713 There is no such concept as a weak binding any more.
3715 ** Removed constants: bignum-radix, scm-line-incrementors
3717 ** define-method: New syntax mandatory.
3719 The new method syntax is now mandatory:
3721 (define-method (NAME ARG-SPEC ...) BODY ...)
3722 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
3724 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
3725 REST-ARG ::= ARG-NAME
3727 If you have old code using the old syntax, import
3728 (oop goops old-define-method) before (oop goops) as in:
3730 (use-modules (oop goops old-define-method) (oop goops))
3732 ** Deprecated function: builtin-variable
3733 Removed function: builtin-bindings
3735 There is no longer a distinction between builtin or other variables.
3736 Use module system operations for all variables.
3738 ** Lazy-catch handlers are no longer allowed to return.
3740 That is, a call to `throw', `error', etc is now guaranteed to not
3743 ** Bugfixes for (ice-9 getopt-long)
3745 This module is now tested using test-suite/tests/getopt-long.test.
3746 The following bugs have been fixed:
3748 *** Parsing for options that are specified to have `optional' args now checks
3749 if the next element is an option instead of unconditionally taking it as the
3752 *** An error is now thrown for `--opt=val' when the option description
3753 does not specify `(value #t)' or `(value optional)'. This condition used to
3754 be accepted w/o error, contrary to the documentation.
3756 *** The error message for unrecognized options is now more informative.
3757 It used to be "not a record", an artifact of the implementation.
3759 *** The error message for `--opt' terminating the arg list (no value), when
3760 `(value #t)' is specified, is now more informative. It used to be "not enough
3763 *** "Clumped" single-char args now preserve trailing string, use it as arg.
3764 The expansion used to be like so:
3766 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
3768 Note that the "5d" is dropped. Now it is like so:
3770 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
3772 This enables single-char options to have adjoining arguments as long as their
3773 constituent characters are not potential single-char options.
3775 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
3777 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
3778 property `arglist', which can be retrieved by `arity'. The result is that
3779 `arity' can give more detailed information than before:
3783 guile> (use-modules (ice-9 optargs))
3784 guile> (define* (foo #:optional a b c) a)
3786 0 or more arguments in `lambda*:G0'.
3791 3 optional arguments: `a', `b' and `c'.
3792 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
3794 2 required arguments: `a' and `b', 2 keyword arguments: `c'
3795 and `d', other keywords allowed.
3796 guile> (define* (baz a b #:optional c #:rest r) a)
3798 2 required arguments: `a' and `b', 1 optional argument: `c',
3801 * Changes to the C interface
3803 ** Types have been renamed from scm_*_t to scm_t_*.
3805 This has been done for POSIX sake. It reserves identifiers ending
3806 with "_t". What a concept.
3808 The old names are still available with status `deprecated'.
3810 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
3812 ** Deprecated features have been removed.
3816 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
3817 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
3819 *** C Functions removed
3821 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
3822 scm_fseek - replaced by scm_seek.
3823 gc-thunk - replaced by after-gc-hook.
3824 gh_int2scmb - replaced by gh_bool2scm.
3825 scm_tc_dblr - replaced by scm_tc16_real.
3826 scm_tc_dblc - replaced by scm_tc16_complex.
3827 scm_list_star - replaced by scm_cons_star.
3829 ** Deprecated: scm_makfromstr
3831 Use scm_mem2string instead.
3833 ** Deprecated: scm_make_shared_substring
3835 Explicit shared substrings will disappear from Guile.
3837 Instead, "normal" strings will be implemented using sharing
3838 internally, combined with a copy-on-write strategy.
3840 ** Deprecated: scm_read_only_string_p
3842 The concept of read-only strings will disappear in next release of
3845 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
3847 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
3849 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
3851 Call a procedure with the indicated number of arguments. See "Fly
3852 Evaluation" in the manual.
3854 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
3856 Call a procedure with the indicated number of arguments and a list of
3857 further arguments. See "Fly Evaluation" in the manual.
3859 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
3861 Create a list of the given number of elements. See "List
3862 Constructors" in the manual.
3864 ** Renamed function: scm_listify has been replaced by scm_list_n.
3866 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
3867 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
3869 Use functions scm_list_N instead.
3871 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
3873 Used by an application to read arbitrary number of bytes from a port.
3874 Same semantics as libc read, except that scm_c_read only returns less
3875 than SIZE bytes if at end-of-file.
3877 Warning: Doesn't update port line and column counts!
3879 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
3881 Used by an application to write arbitrary number of bytes to an SCM
3882 port. Similar semantics as libc write. However, unlike libc
3883 write, scm_c_write writes the requested number of bytes and has no
3886 Warning: Doesn't update port line and column counts!
3888 ** New function: scm_init_guile ()
3890 In contrast to scm_boot_guile, scm_init_guile will return normally
3891 after initializing Guile. It is not available on all systems, tho.
3893 ** New functions: scm_str2symbol, scm_mem2symbol
3895 The function scm_str2symbol takes a const char* pointing to a zero-terminated
3896 field of characters and creates a scheme symbol object from that C string.
3897 The function scm_mem2symbol takes a const char* and a number of characters and
3898 creates a symbol from the characters in that memory area.
3900 ** New functions: scm_primitive_make_property
3901 scm_primitive_property_ref
3902 scm_primitive_property_set_x
3903 scm_primitive_property_del_x
3905 These functions implement a new way to deal with object properties.
3906 See libguile/properties.c for their documentation.
3908 ** New function: scm_done_free (long size)
3910 This function is the inverse of scm_done_malloc. Use it to report the
3911 amount of smob memory you free. The previous method, which involved
3912 calling scm_done_malloc with negative argument, was somewhat
3913 unintuitive (and is still available, of course).
3915 ** New function: scm_c_memq (SCM obj, SCM list)
3917 This function provides a fast C level alternative for scm_memq for the case
3918 that the list parameter is known to be a proper list. The function is a
3919 replacement for scm_sloppy_memq, but is stricter in its requirements on its
3920 list input parameter, since for anything else but a proper list the function's
3921 behaviour is undefined - it may even crash or loop endlessly. Further, for
3922 the case that the object is not found in the list, scm_c_memq returns #f which
3923 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
3925 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
3926 scm_remember_upto_here
3928 These functions replace the function scm_remember.
3930 ** Deprecated function: scm_remember
3932 Use one of the new functions scm_remember_upto_here_1,
3933 scm_remember_upto_here_2 or scm_remember_upto_here instead.
3935 ** New function: scm_allocate_string
3937 This function replaces the function scm_makstr.
3939 ** Deprecated function: scm_makstr
3941 Use the new function scm_allocate_string instead.
3943 ** New global variable scm_gc_running_p introduced.
3945 Use this variable to find out if garbage collection is being executed. Up to
3946 now applications have used scm_gc_heap_lock to test if garbage collection was
3947 running, which also works because of the fact that up to know only the garbage
3948 collector has set this variable. But, this is an implementation detail that
3949 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
3950 of this variable is (and has been) not fully safe anyway.
3952 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
3954 Use these instead of SCM_LENGTH_MAX.
3956 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
3957 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
3958 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
3960 Use these instead of SCM_LENGTH.
3962 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
3963 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
3964 SCM_SET_BITVECTOR_LENGTH
3966 Use these instead of SCM_SETLENGTH
3968 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
3969 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
3972 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
3975 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
3976 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
3979 Use these instead of SCM_SETCHARS.
3981 ** New macro: SCM_BITVECTOR_P
3983 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
3985 Use instead of SCM_COERCE_SUBSTR.
3987 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
3989 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
3991 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
3992 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
3993 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
3994 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
3995 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
3996 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
3997 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
3998 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
3999 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
4000 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
4001 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
4002 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
4003 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
4004 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
4005 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
4007 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
4008 Use scm_memory_error instead of SCM_NALLOC.
4009 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
4010 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
4011 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
4012 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
4013 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
4014 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
4015 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
4016 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
4017 Use SCM_STRINGP instead of SCM_RWSTRINGP.
4018 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
4019 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
4020 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
4021 Use a type specific setter macro instead of SCM_SETLENGTH.
4022 Use a type specific setter macro instead of SCM_SETCHARS.
4023 Use a type specific length macro instead of SCM_LENGTH_MAX.
4024 Use SCM_GCMARKP instead of SCM_GC8MARKP.
4025 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
4026 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
4027 Use SCM_TYP16 instead of SCM_GCTYP16.
4028 Use SCM_CDR instead of SCM_GCCDR.
4029 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
4030 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
4031 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
4032 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
4033 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
4034 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
4035 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
4037 ** Removed function: scm_struct_init
4039 ** Removed variable: scm_symhash_dim
4041 ** Renamed function: scm_make_cont has been replaced by
4042 scm_make_continuation, which has a different interface.
4044 ** Deprecated function: scm_call_catching_errors
4046 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
4048 ** Deprecated function: scm_strhash
4050 Use scm_string_hash instead.
4052 ** Deprecated function: scm_vector_set_length_x
4054 Instead, create a fresh vector of the desired size and copy the contents.
4056 ** scm_gensym has changed prototype
4058 scm_gensym now only takes one argument.
4060 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
4063 There is now only a single symbol type scm_tc7_symbol.
4064 The tag scm_tc7_lvector was not used anyway.
4066 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
4068 Use scm_make_smob_type and scm_set_smob_XXX instead.
4070 ** New function scm_set_smob_apply.
4072 This can be used to set an apply function to a smob type.
4074 ** Deprecated function: scm_strprint_obj
4076 Use scm_object_to_string instead.
4078 ** Deprecated function: scm_wta
4080 Use scm_wrong_type_arg, or another appropriate error signalling function
4083 ** Explicit support for obarrays has been deprecated.
4085 Use `scm_str2symbol' and the generic hashtable functions instead.
4087 ** The concept of `vcells' has been deprecated.
4089 The data type `variable' is now used exclusively. `Vcells' have been
4090 a low-level concept so you are likely not affected by this change.
4092 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
4093 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
4095 Use scm_c_define or scm_c_lookup instead, as appropriate.
4097 *** New functions: scm_c_module_lookup, scm_c_lookup,
4098 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
4099 scm_module_define, scm_define.
4101 These functions work with variables instead of with vcells.
4103 ** New functions for creating and defining `subr's and `gsubr's.
4105 The new functions more clearly distinguish between creating a subr (or
4106 gsubr) object and adding it to the current module.
4108 These new functions are available: scm_c_make_subr, scm_c_define_subr,
4109 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
4110 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
4111 scm_c_define_gsubr_with_generic.
4113 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
4114 scm_make_subr_with_generic, scm_make_gsubr,
4115 scm_make_gsubr_with_generic.
4117 Use the new ones from above instead.
4119 ** C interface to the module system has changed.
4121 While we suggest that you avoid as many explicit module system
4122 operations from C as possible for the time being, the C interface has
4123 been made more similar to the high-level Scheme module system.
4125 *** New functions: scm_c_define_module, scm_c_use_module,
4126 scm_c_export, scm_c_resolve_module.
4128 They mostly work like their Scheme namesakes. scm_c_define_module
4129 takes a function that is called a context where the new module is
4132 *** Deprecated functions: scm_the_root_module, scm_make_module,
4133 scm_ensure_user_module, scm_load_scheme_module.
4135 Use the new functions instead.
4137 ** Renamed function: scm_internal_with_fluids becomes
4140 scm_internal_with_fluids is available as a deprecated function.
4142 ** New function: scm_c_with_fluid.
4144 Just like scm_c_with_fluids, but takes one fluid and one value instead
4147 ** Deprecated typedefs: long_long, ulong_long.
4149 They are of questionable utility and they pollute the global
4152 ** Deprecated typedef: scm_sizet
4154 It is of questionable utility now that Guile requires ANSI C, and is
4157 ** Deprecated typedefs: scm_port_rw_active, scm_port,
4158 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
4159 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
4161 Made more compliant with the naming policy by adding a _t at the end.
4163 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
4164 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
4166 With the exception of the mysterious scm_2ulong2big, they are still
4167 available under new names (scm_i_mkbig etc). These functions are not
4168 intended to be used in user code. You should avoid dealing with
4169 bignums directly, and should deal with numbers in general (which can
4172 ** Change in behavior: scm_num2long, scm_num2ulong
4174 The scm_num2[u]long functions don't any longer accept an inexact
4175 argument. This change in behavior is motivated by concordance with
4176 R5RS: It is more common that a primitive doesn't want to accept an
4177 inexact for an exact.
4179 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
4180 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
4181 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
4184 These are conversion functions between the various ANSI C integral
4185 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
4186 accept an inexact argument.
4188 ** New functions: scm_float2num, scm_double2num,
4189 scm_num2float, scm_num2double.
4191 These are conversion functions between the two ANSI C float types and
4194 ** New number validation macros:
4195 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
4199 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
4201 These are just nicer-named old scm_protect_object and
4202 scm_unprotect_object.
4204 ** Deprecated functions: scm_protect_object, scm_unprotect_object
4206 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
4208 These functions can be used to register pointers to locations that
4211 ** Deprecated function: scm_create_hook.
4213 Its sins are: misleading name, non-modularity and lack of general
4217 Changes since Guile 1.3.4:
4219 * Changes to the distribution
4221 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
4223 We've changed the way we handle generated files in the Guile source
4224 repository. As a result, the procedure for building trees obtained
4225 from the nightly FTP snapshots or via CVS has changed:
4226 - You must have appropriate versions of autoconf, automake, and
4227 libtool installed on your system. See README for info on how to
4228 obtain these programs.
4229 - Before configuring the tree, you must first run the script
4230 `autogen.sh' at the top of the source tree.
4232 The Guile repository used to contain not only source files, written by
4233 humans, but also some generated files, like configure scripts and
4234 Makefile.in files. Even though the contents of these files could be
4235 derived mechanically from other files present, we thought it would
4236 make the tree easier to build if we checked them into CVS.
4238 However, this approach means that minor differences between
4239 developer's installed tools and habits affected the whole team.
4240 So we have removed the generated files from the repository, and
4241 added the autogen.sh script, which will reconstruct them
4245 ** configure now has experimental options to remove support for certain
4248 --disable-arrays omit array and uniform array support
4249 --disable-posix omit posix interfaces
4250 --disable-networking omit networking interfaces
4251 --disable-regex omit regular expression interfaces
4253 These are likely to become separate modules some day.
4255 ** New configure option --enable-debug-freelist
4257 This enables a debugging version of SCM_NEWCELL(), and also registers
4258 an extra primitive, the setter `gc-set-debug-check-freelist!'.
4260 Configure with the --enable-debug-freelist option to enable
4261 the gc-set-debug-check-freelist! primitive, and then use:
4263 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
4264 (gc-set-debug-check-freelist! #f) # turn off checking
4266 Checking of the freelist forces a traversal of the freelist and
4267 a garbage collection before each allocation of a cell. This can
4268 slow down the interpreter dramatically, so the setter should be used to
4269 turn on this extra processing only when necessary.
4271 ** New configure option --enable-debug-malloc
4273 Include code for debugging of calls to scm_must_malloc/realloc/free.
4277 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
4278 2. objects reallocated by scm_must_realloc has been allocated by
4280 3. reallocated objects are reallocated with the same what string
4282 But, most importantly, it records the number of allocated objects of
4283 each kind. This is useful when searching for memory leaks.
4285 A Guile compiled with this option provides the primitive
4286 `malloc-stats' which returns an alist with pairs of kind and the
4287 number of objects of that kind.
4289 ** All includes are now referenced relative to the root directory
4291 Since some users have had problems with mixups between Guile and
4292 system headers, we have decided to always refer to Guile headers via
4293 their parent directories. This essentially creates a "private name
4294 space" for Guile headers. This means that the compiler only is given
4295 -I options for the root build and root source directory.
4297 ** Header files kw.h and genio.h have been removed.
4299 ** The module (ice-9 getopt-gnu-style) has been removed.
4301 ** New module (ice-9 documentation)
4303 Implements the interface to documentation strings associated with
4306 ** New module (ice-9 time)
4308 Provides a macro `time', which displays execution time of a given form.
4310 ** New module (ice-9 history)
4312 Loading this module enables value history in the repl.
4314 * Changes to the stand-alone interpreter
4316 ** New command line option --debug
4318 Start Guile with debugging evaluator and backtraces enabled.
4320 This is useful when debugging your .guile init file or scripts.
4322 ** New help facility
4324 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
4325 (help REGEXP) ditto for objects with names matching REGEXP (a string)
4326 (help 'NAME) gives documentation for NAME, even if it is not an object
4327 (help ,EXPR) gives documentation for object returned by EXPR
4328 (help (my module)) gives module commentary for `(my module)'
4329 (help) gives this text
4331 `help' searches among bindings exported from loaded modules, while
4332 `apropos' searches among bindings visible from the "current" module.
4334 Examples: (help help)
4336 (help "output-string")
4338 ** `help' and `apropos' now prints full module names
4340 ** Dynamic linking now uses libltdl from the libtool package.
4342 The old system dependent code for doing dynamic linking has been
4343 replaced with calls to the libltdl functions which do all the hairy
4346 The major improvement is that you can now directly pass libtool
4347 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
4348 will be able to do the best shared library job you can get, via
4351 The way dynamic libraries are found has changed and is not really
4352 portable across platforms, probably. It is therefore recommended to
4353 use absolute filenames when possible.
4355 If you pass a filename without an extension to `dynamic-link', it will
4356 try a few appropriate ones. Thus, the most platform ignorant way is
4357 to specify a name like "libfoo", without any directories and
4360 ** Guile COOP threads are now compatible with LinuxThreads
4362 Previously, COOP threading wasn't possible in applications linked with
4363 Linux POSIX threads due to their use of the stack pointer to find the
4364 thread context. This has now been fixed with a workaround which uses
4365 the pthreads to allocate the stack.
4367 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
4369 ** Positions of erring expression in scripts
4371 With version 1.3.4, the location of the erring expression in Guile
4372 scipts is no longer automatically reported. (This should have been
4373 documented before the 1.3.4 release.)
4375 You can get this information by enabling recording of positions of
4376 source expressions and running the debugging evaluator. Put this at
4377 the top of your script (or in your "site" file):
4379 (read-enable 'positions)
4380 (debug-enable 'debug)
4382 ** Backtraces in scripts
4384 It is now possible to get backtraces in scripts.
4388 (debug-enable 'debug 'backtrace)
4390 at the top of the script.
4392 (The first options enables the debugging evaluator.
4393 The second enables backtraces.)
4395 ** Part of module system symbol lookup now implemented in C
4397 The eval closure of most modules is now implemented in C. Since this
4398 was one of the bottlenecks for loading speed, Guile now loads code
4399 substantially faster than before.
4401 ** Attempting to get the value of an unbound variable now produces
4402 an exception with a key of 'unbound-variable instead of 'misc-error.
4404 ** The initial default output port is now unbuffered if it's using a
4405 tty device. Previously in this situation it was line-buffered.
4407 ** New hook: after-gc-hook
4409 after-gc-hook takes over the role of gc-thunk. This hook is run at
4410 the first SCM_TICK after a GC. (Thus, the code is run at the same
4411 point during evaluation as signal handlers.)
4413 Note that this hook should be used only for diagnostic and debugging
4414 purposes. It is not certain that it will continue to be well-defined
4415 when this hook is run in the future.
4417 C programmers: Note the new C level hooks scm_before_gc_c_hook,
4418 scm_before_sweep_c_hook, scm_after_gc_c_hook.
4420 ** Improvements to garbage collector
4422 Guile 1.4 has a new policy for triggering heap allocation and
4423 determining the sizes of heap segments. It fixes a number of problems
4426 1. The new policy can handle two separate pools of cells
4427 (2-word/4-word) better. (The old policy would run wild, allocating
4428 more and more memory for certain programs.)
4430 2. The old code would sometimes allocate far too much heap so that the
4431 Guile process became gigantic. The new code avoids this.
4433 3. The old code would sometimes allocate too little so that few cells
4434 were freed at GC so that, in turn, too much time was spent in GC.
4436 4. The old code would often trigger heap allocation several times in a
4437 row. (The new scheme predicts how large the segments needs to be
4438 in order not to need further allocation.)
4440 All in all, the new GC policy will make larger applications more
4443 The new GC scheme also is prepared for POSIX threading. Threads can
4444 allocate private pools of cells ("clusters") with just a single
4445 function call. Allocation of single cells from such a cluster can
4446 then proceed without any need of inter-thread synchronization.
4448 ** New environment variables controlling GC parameters
4450 GUILE_MAX_SEGMENT_SIZE Maximal segment size
4453 Allocation of 2-word cell heaps:
4455 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
4458 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
4459 GC in percent of total heap size
4462 Allocation of 4-word cell heaps
4463 (used for real numbers and misc other objects):
4465 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
4467 (See entry "Way for application to customize GC parameters" under
4468 section "Changes to the scm_ interface" below.)
4470 ** Guile now implements reals using 4-word cells
4472 This speeds up computation with reals. (They were earlier allocated
4473 with `malloc'.) There is still some room for optimizations, however.
4475 ** Some further steps toward POSIX thread support have been taken
4477 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
4478 don't have much effect any longer, and many of them will be removed in
4482 are only handled at the top of the evaluator loop, immediately after
4483 I/O, and in scm_equalp.
4485 *** The GC can allocate thread private pools of pairs.
4487 * Changes to Scheme functions and syntax
4489 ** close-input-port and close-output-port are now R5RS
4491 These procedures have been turned into primitives and have R5RS behaviour.
4493 ** New procedure: simple-format PORT MESSAGE ARG1 ...
4495 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
4496 extended by the more sophisticated version in (ice-9 format)
4498 (simple-format port message . args)
4499 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
4500 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
4501 the escapes are replaced with corresponding members of ARGS:
4502 ~A formats using `display' and ~S formats using `write'.
4503 If DESTINATION is #t, then use the `current-output-port',
4504 if DESTINATION is #f, then return a string containing the formatted text.
4505 Does not add a trailing newline."
4507 ** string-ref: the second argument is no longer optional.
4509 ** string, list->string: no longer accept strings in their arguments,
4510 only characters, for compatibility with R5RS.
4512 ** New procedure: port-closed? PORT
4513 Returns #t if PORT is closed or #f if it is open.
4515 ** Deprecated: list*
4517 The list* functionality is now provided by cons* (SRFI-1 compliant)
4519 ** New procedure: cons* ARG1 ARG2 ... ARGn
4521 Like `list', but the last arg provides the tail of the constructed list,
4522 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
4524 Requires at least one argument. If given one argument, that argument
4525 is returned as result.
4527 This function is called `list*' in some other Schemes and in Common LISP.
4529 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
4531 ** New procedure: object-documentation OBJECT
4533 Returns the documentation string associated with OBJECT. The
4534 procedure uses a caching mechanism so that subsequent lookups are
4537 Exported by (ice-9 documentation).
4539 ** module-name now returns full names of modules
4541 Previously, only the last part of the name was returned (`session' for
4542 `(ice-9 session)'). Ex: `(ice-9 session)'.
4544 * Changes to the gh_ interface
4546 ** Deprecated: gh_int2scmb
4548 Use gh_bool2scm instead.
4550 * Changes to the scm_ interface
4552 ** Guile primitives now carry docstrings!
4554 Thanks to Greg Badros!
4556 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4558 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4559 macros and must contain a docstring that is extracted into foo.doc using a new
4560 guile-doc-snarf script (that uses guile-doc-snarf.awk).
4562 However, a major overhaul of these macros is scheduled for the next release of
4565 ** Guile primitives use a new technique for validation of arguments
4567 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
4568 the readability of argument checking.
4570 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
4572 ** New macros: SCM_PACK, SCM_UNPACK
4574 Compose/decompose an SCM value.
4576 The SCM type is now treated as an abstract data type and may be defined as a
4577 long, a void* or as a struct, depending on the architecture and compile time
4578 options. This makes it easier to find several types of bugs, for example when
4579 SCM values are treated as integers without conversion. Values of the SCM type
4580 should be treated as "atomic" values. These macros are used when
4581 composing/decomposing an SCM value, either because you want to access
4582 individual bits, or because you want to treat it as an integer value.
4584 E.g., in order to set bit 7 in an SCM value x, use the expression
4586 SCM_PACK (SCM_UNPACK (x) | 0x80)
4588 ** The name property of hooks is deprecated.
4589 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
4591 You can emulate this feature by using object properties.
4593 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
4594 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
4597 These macros will be removed in a future release of Guile.
4599 ** The following types, functions and macros from numbers.h are deprecated:
4600 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
4601 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
4603 ** Port internals: the rw_random variable in the scm_port structure
4604 must be set to non-zero in any random access port. In recent Guile
4605 releases it was only set for bidirectional random-access ports.
4607 ** Port internals: the seek ptob procedure is now responsible for
4608 resetting the buffers if required. The change was made so that in the
4609 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
4610 the fport and strport ptobs can avoid resetting the buffers,
4611 in particular to avoid discarding unread chars. An existing port
4612 type can be fixed by adding something like the following to the
4613 beginning of the ptob seek procedure:
4615 if (pt->rw_active == SCM_PORT_READ)
4616 scm_end_input (object);
4617 else if (pt->rw_active == SCM_PORT_WRITE)
4618 ptob->flush (object);
4620 although to actually avoid resetting the buffers and discard unread
4621 chars requires further hacking that depends on the characteristics
4624 ** Deprecated functions: scm_fseek, scm_tag
4626 These functions are no longer used and will be removed in a future version.
4628 ** The scm_sysmissing procedure is no longer used in libguile.
4629 Unless it turns out to be unexpectedly useful to somebody, it will be
4630 removed in a future version.
4632 ** The format of error message strings has changed
4634 The two C procedures: scm_display_error and scm_error, as well as the
4635 primitive `scm-error', now use scm_simple_format to do their work.
4636 This means that the message strings of all code must be updated to use
4637 ~A where %s was used before, and ~S where %S was used before.
4639 During the period when there still are a lot of old Guiles out there,
4640 you might want to support both old and new versions of Guile.
4642 There are basically two methods to achieve this. Both methods use
4645 AC_CHECK_FUNCS(scm_simple_format)
4647 in your configure.in.
4649 Method 1: Use the string concatenation features of ANSI C's
4654 #ifdef HAVE_SCM_SIMPLE_FORMAT
4660 Then represent each of your error messages using a preprocessor macro:
4662 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
4666 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
4667 (define make-message string-append)
4669 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
4671 Method 2: Use the oldfmt function found in doc/oldfmt.c.
4675 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
4680 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
4684 ** Deprecated: coop_mutex_init, coop_condition_variable_init
4686 Don't use the functions coop_mutex_init and
4687 coop_condition_variable_init. They will change.
4689 Use scm_mutex_init and scm_cond_init instead.
4691 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
4692 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
4693 COND, as `scm_cond_wait' does, but it also bounds the duration
4694 of the wait. If COND has not been signaled before time ABSTIME,
4695 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
4696 returns the error code `ETIMEDOUT'.
4698 The ABSTIME parameter specifies an absolute time, with the same
4699 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
4700 to 00:00:00 GMT, January 1, 1970.
4702 ** New function: scm_cond_broadcast (scm_cond_t *COND)
4703 `scm_cond_broadcast' restarts all the threads that are waiting
4704 on the condition variable COND. Nothing happens if no threads are
4707 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
4708 `scm_key_create' allocates a new TSD key. The key is stored in
4709 the location pointed to by KEY. There is no limit on the number
4710 of keys allocated at a given time. The value initially associated
4711 with the returned key is `NULL' in all currently executing threads.
4713 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
4714 function associated with the key. When a thread terminates,
4715 DESTR_FUNCTION is called on the value associated with the key in
4716 that thread. The DESTR_FUNCTION is not called if a key is deleted
4717 with `scm_key_delete' or a value is changed with
4718 `scm_setspecific'. The order in which destructor functions are
4719 called at thread termination time is unspecified.
4721 Destructors are not yet implemented.
4723 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
4724 `scm_setspecific' changes the value associated with KEY in the
4725 calling thread, storing the given POINTER instead.
4727 ** New function: scm_getspecific (scm_key_t KEY)
4728 `scm_getspecific' returns the value currently associated with
4729 KEY in the calling thread.
4731 ** New function: scm_key_delete (scm_key_t KEY)
4732 `scm_key_delete' deallocates a TSD key. It does not check
4733 whether non-`NULL' values are associated with that key in the
4734 currently executing threads, nor call the destructor function
4735 associated with the key.
4737 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
4739 Initialize a C level hook HOOK with associated HOOK_DATA and type
4740 TYPE. (See scm_c_hook_run ().)
4742 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
4744 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
4745 is true, add it last, otherwise first. The same FUNC can be added
4746 multiple times if FUNC_DATA differ and vice versa.
4748 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
4750 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
4751 function is only removed if both FUNC and FUNC_DATA matches.
4753 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
4755 Run hook HOOK passing DATA to the hook functions.
4757 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
4758 returned is undefined.
4760 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
4761 returns a non-NULL value. This value is returned as the result of
4762 scm_c_hook_run. If all functions return NULL, NULL is returned.
4764 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
4765 returns a NULL value, and NULL is returned. If all functions returns
4766 a non-NULL value, the last value is returned.
4768 ** New C level GC hooks
4770 Five new C level hooks has been added to the garbage collector.
4772 scm_before_gc_c_hook
4775 are run before locking and after unlocking the heap. The system is
4776 thus in a mode where evaluation can take place. (Except that
4777 scm_before_gc_c_hook must not allocate new cells.)
4779 scm_before_mark_c_hook
4780 scm_before_sweep_c_hook
4781 scm_after_sweep_c_hook
4783 are run when the heap is locked. These are intended for extension of
4784 the GC in a modular fashion. Examples are the weaks and guardians
4787 ** Way for application to customize GC parameters
4789 The application can set up other default values for the GC heap
4790 allocation parameters
4792 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
4793 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
4794 GUILE_MAX_SEGMENT_SIZE,
4798 scm_default_init_heap_size_1, scm_default_min_yield_1,
4799 scm_default_init_heap_size_2, scm_default_min_yield_2,
4800 scm_default_max_segment_size
4802 respectively before callong scm_boot_guile.
4804 (See entry "New environment variables ..." in section
4805 "Changes to the stand-alone interpreter" above.)
4807 ** scm_protect_object/scm_unprotect_object now nest
4809 This means that you can call scm_protect_object multiple times on an
4810 object and count on the object being protected until
4811 scm_unprotect_object has been call the same number of times.
4813 The functions also have better time complexity.
4815 Still, it is usually possible to structure the application in a way
4816 that you don't need to use these functions. For example, if you use a
4817 protected standard Guile list to keep track of live objects rather
4818 than some custom data type, objects will die a natural death when they
4819 are no longer needed.
4821 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
4823 Guile does not provide the float representation for inexact real numbers any
4824 more. Now, only doubles are used to represent inexact real numbers. Further,
4825 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
4826 and scm_tc16_complex, respectively.
4828 ** Removed deprecated type scm_smobfuns
4830 ** Removed deprecated function scm_newsmob
4832 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
4834 There is an ongoing discussion among the developers whether to
4835 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
4836 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
4837 until this issue has been settled.
4839 ** Removed deprecated type tag scm_tc16_kw
4841 ** Added type tag scm_tc16_keyword
4843 (This was introduced already in release 1.3.4 but was not documented
4846 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
4848 * Changes to system call interfaces:
4850 ** The "select" procedure now tests port buffers for the ability to
4851 provide input or accept output. Previously only the underlying file
4852 descriptors were checked.
4854 ** New variable PIPE_BUF: the maximum number of bytes that can be
4855 atomically written to a pipe.
4857 ** If a facility is not available on the system when Guile is
4858 compiled, the corresponding primitive procedure will not be defined.
4859 Previously it would have been defined but would throw a system-error
4860 exception if called. Exception handlers which catch this case may
4861 need minor modification: an error will be thrown with key
4862 'unbound-variable instead of 'system-error. Alternatively it's
4863 now possible to use `defined?' to check whether the facility is
4866 ** Procedures which depend on the timezone should now give the correct
4867 result on systems which cache the TZ environment variable, even if TZ
4868 is changed without calling tzset.
4870 * Changes to the networking interfaces:
4872 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
4873 long integers between network and host format. For now, it's not
4874 particularly convenient to do this kind of thing, but consider:
4876 (define write-network-long
4877 (lambda (value port)
4878 (let ((v (make-uniform-vector 1 1 0)))
4879 (uniform-vector-set! v 0 (htonl value))
4880 (uniform-vector-write v port))))
4882 (define read-network-long
4884 (let ((v (make-uniform-vector 1 1 0)))
4885 (uniform-vector-read! v port)
4886 (ntohl (uniform-vector-ref v 0)))))
4888 ** If inet-aton fails, it now throws an error with key 'misc-error
4889 instead of 'system-error, since errno is not relevant.
4891 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
4892 specific keys instead of 'system-error. The latter is inappropriate
4893 since errno will not have been set. The keys are:
4894 'host-not-found, 'try-again, 'no-recovery and 'no-data.
4896 ** sethostent, setnetent, setprotoent, setservent: now take an
4897 optional argument STAYOPEN, which specifies whether the database
4898 remains open after a database entry is accessed randomly (e.g., using
4899 gethostbyname for the hosts database.) The default is #f. Previously
4903 Changes since Guile 1.3.2:
4905 * Changes to the stand-alone interpreter
4909 An initial version of the Guile debugger written by Chris Hanson has
4910 been added. The debugger is still under development but is included
4911 in the distribution anyway since it is already quite useful.
4917 after an error to enter the debugger. Type `help' inside the debugger
4918 for a description of available commands.
4920 If you prefer to have stack frames numbered and printed in
4921 anti-chronological order and prefer up in the stack to be down on the
4922 screen as is the case in gdb, you can put
4924 (debug-enable 'backwards)
4926 in your .guile startup file. (However, this means that Guile can't
4927 use indentation to indicate stack level.)
4929 The debugger is autoloaded into Guile at the first use.
4931 ** Further enhancements to backtraces
4933 There is a new debug option `width' which controls the maximum width
4934 on the screen of printed stack frames. Fancy printing parameters
4935 ("level" and "length" as in Common LISP) are adaptively adjusted for
4936 each stack frame to give maximum information while still fitting
4937 within the bounds. If the stack frame can't be made to fit by
4938 adjusting parameters, it is simply cut off at the end. This is marked
4941 ** Some modules are now only loaded when the repl is started
4943 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
4944 regex) are now loaded into (guile-user) only if the repl has been
4945 started. The effect is that the startup time for scripts has been
4946 reduced to 30% of what it was previously.
4948 Correctly written scripts load the modules they require at the top of
4949 the file and should not be affected by this change.
4951 ** Hooks are now represented as smobs
4953 * Changes to Scheme functions and syntax
4955 ** Readline support has changed again.
4957 The old (readline-activator) module is gone. Use (ice-9 readline)
4958 instead, which now contains all readline functionality. So the code
4959 to activate readline is now
4961 (use-modules (ice-9 readline))
4964 This should work at any time, including from the guile prompt.
4966 To avoid confusion about the terms of Guile's license, please only
4967 enable readline for your personal use; please don't make it the
4968 default for others. Here is why we make this rather odd-sounding
4971 Guile is normally licensed under a weakened form of the GNU General
4972 Public License, which allows you to link code with Guile without
4973 placing that code under the GPL. This exception is important to some
4976 However, since readline is distributed under the GNU General Public
4977 License, when you link Guile with readline, either statically or
4978 dynamically, you effectively change Guile's license to the strict GPL.
4979 Whenever you link any strictly GPL'd code into Guile, uses of Guile
4980 which are normally permitted become forbidden. This is a rather
4981 non-obvious consequence of the licensing terms.
4983 So, to make sure things remain clear, please let people choose for
4984 themselves whether to link GPL'd libraries like readline with Guile.
4986 ** regexp-substitute/global has changed slightly, but incompatibly.
4988 If you include a function in the item list, the string of the match
4989 object it receives is the same string passed to
4990 regexp-substitute/global, not some suffix of that string.
4991 Correspondingly, the match's positions are relative to the entire
4992 string, not the suffix.
4994 If the regexp can match the empty string, the way matches are chosen
4995 from the string has changed. regexp-substitute/global recognizes the
4996 same set of matches that list-matches does; see below.
4998 ** New function: list-matches REGEXP STRING [FLAGS]
5000 Return a list of match objects, one for every non-overlapping, maximal
5001 match of REGEXP in STRING. The matches appear in left-to-right order.
5002 list-matches only reports matches of the empty string if there are no
5003 other matches which begin on, end at, or include the empty match's
5006 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5008 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
5010 For each match of REGEXP in STRING, apply PROC to the match object,
5011 and the last value PROC returned, or INIT for the first call. Return
5012 the last value returned by PROC. We apply PROC to the matches as they
5013 appear from left to right.
5015 This function recognizes matches according to the same criteria as
5018 Thus, you could define list-matches like this:
5020 (define (list-matches regexp string . flags)
5021 (reverse! (apply fold-matches regexp string '() cons flags)))
5023 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
5027 *** New function: hook? OBJ
5029 Return #t if OBJ is a hook, otherwise #f.
5031 *** New function: make-hook-with-name NAME [ARITY]
5033 Return a hook with name NAME and arity ARITY. The default value for
5034 ARITY is 0. The only effect of NAME is that it will appear when the
5035 hook object is printed to ease debugging.
5037 *** New function: hook-empty? HOOK
5039 Return #t if HOOK doesn't contain any procedures, otherwise #f.
5041 *** New function: hook->list HOOK
5043 Return a list of the procedures that are called when run-hook is
5046 ** `map' signals an error if its argument lists are not all the same length.
5048 This is the behavior required by R5RS, so this change is really a bug
5049 fix. But it seems to affect a lot of people's code, so we're
5050 mentioning it here anyway.
5052 ** Print-state handling has been made more transparent
5054 Under certain circumstances, ports are represented as a port with an
5055 associated print state. Earlier, this pair was represented as a pair
5056 (see "Some magic has been added to the printer" below). It is now
5057 indistinguishable (almost; see `get-print-state') from a port on the
5060 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
5062 Return a new port with the associated print state PRINT-STATE.
5064 *** New function: get-print-state OUTPUT-PORT
5066 Return the print state associated with this port if it exists,
5067 otherwise return #f.
5069 *** New function: directory-stream? OBJECT
5071 Returns true iff OBJECT is a directory stream --- the sort of object
5072 returned by `opendir'.
5074 ** New function: using-readline?
5076 Return #t if readline is in use in the current repl.
5078 ** structs will be removed in 1.4
5080 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
5081 and use GOOPS objects as the fundamental record type.
5083 * Changes to the scm_ interface
5085 ** structs will be removed in 1.4
5087 The entire current struct interface (struct.c, struct.h) will be
5088 replaced in Guile 1.4. We will merge GOOPS into libguile and use
5089 GOOPS objects as the fundamental record type.
5091 ** The internal representation of subr's has changed
5093 Instead of giving a hint to the subr name, the CAR field of the subr
5094 now contains an index to a subr entry in scm_subr_table.
5096 *** New variable: scm_subr_table
5098 An array of subr entries. A subr entry contains the name, properties
5099 and documentation associated with the subr. The properties and
5100 documentation slots are not yet used.
5102 ** A new scheme for "forwarding" calls to a builtin to a generic function
5104 It is now possible to extend the functionality of some Guile
5105 primitives by letting them defer a call to a GOOPS generic function on
5106 argument mismatch. This means that there is no loss of efficiency in
5111 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
5112 (define-method + ((x <string>) (y <string>))
5113 (string-append x y))
5115 + will still be as efficient as usual in numerical calculations, but
5116 can also be used for concatenating strings.
5118 Who will be the first one to extend Guile's numerical tower to
5119 rationals? :) [OK, there a few other things to fix before this can
5120 be made in a clean way.]
5122 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
5124 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
5126 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
5128 These do the same job as SCM_PROC and SCM_PROC1, but they also define
5129 a variable GENERIC which can be used by the dispatch macros below.
5131 [This is experimental code which may change soon.]
5133 *** New macros for forwarding control to a generic on arg type error
5135 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
5137 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
5139 These correspond to the scm_wta function call, and have the same
5140 behaviour until the user has called the GOOPS primitive
5141 `enable-primitive-generic!'. After that, these macros will apply the
5142 generic function GENERIC to the argument(s) instead of calling
5145 [This is experimental code which may change soon.]
5147 *** New macros for argument testing with generic dispatch
5149 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
5151 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
5153 These correspond to the SCM_ASSERT macro, but will defer control to
5154 GENERIC on error after `enable-primitive-generic!' has been called.
5156 [This is experimental code which may change soon.]
5158 ** New function: SCM scm_eval_body (SCM body, SCM env)
5160 Evaluates the body of a special form.
5162 ** The internal representation of struct's has changed
5164 Previously, four slots were allocated for the procedure(s) of entities
5165 and operators. The motivation for this representation had to do with
5166 the structure of the evaluator, the wish to support tail-recursive
5167 generic functions, and efficiency. Since the generic function
5168 dispatch mechanism has changed, there is no longer a need for such an
5169 expensive representation, and the representation has been simplified.
5171 This should not make any difference for most users.
5173 ** GOOPS support has been cleaned up.
5175 Some code has been moved from eval.c to objects.c and code in both of
5176 these compilation units has been cleaned up and better structured.
5178 *** New functions for applying generic functions
5180 New function: SCM scm_apply_generic (GENERIC, ARGS)
5181 New function: SCM scm_call_generic_0 (GENERIC)
5182 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
5183 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
5184 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
5186 ** Deprecated function: scm_make_named_hook
5188 It is now replaced by:
5190 ** New function: SCM scm_create_hook (const char *name, int arity)
5192 Creates a hook in the same way as make-hook above but also
5193 binds a variable named NAME to it.
5195 This is the typical way of creating a hook from C code.
5197 Currently, the variable is created in the "current" module.
5198 This might change when we get the new module system.
5200 [The behaviour is identical to scm_make_named_hook.]
5204 Changes since Guile 1.3:
5206 * Changes to mailing lists
5208 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
5210 See the README file to find current addresses for all the Guile
5213 * Changes to the distribution
5215 ** Readline support is no longer included with Guile by default.
5217 Based on the different license terms of Guile and Readline, we
5218 concluded that Guile should not *by default* cause the linking of
5219 Readline into an application program. Readline support is now offered
5220 as a separate module, which is linked into an application only when
5221 you explicitly specify it.
5223 Although Guile is GNU software, its distribution terms add a special
5224 exception to the usual GNU General Public License (GPL). Guile's
5225 license includes a clause that allows you to link Guile with non-free
5226 programs. We add this exception so as not to put Guile at a
5227 disadvantage vis-a-vis other extensibility packages that support other
5230 In contrast, the GNU Readline library is distributed under the GNU
5231 General Public License pure and simple. This means that you may not
5232 link Readline, even dynamically, into an application unless it is
5233 distributed under a free software license that is compatible the GPL.
5235 Because of this difference in distribution terms, an application that
5236 can use Guile may not be able to use Readline. Now users will be
5237 explicitly offered two independent decisions about the use of these
5240 You can activate the readline support by issuing
5242 (use-modules (readline-activator))
5245 from your ".guile" file, for example.
5247 * Changes to the stand-alone interpreter
5249 ** All builtins now print as primitives.
5250 Previously builtin procedures not belonging to the fundamental subr
5251 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
5252 Now, they print as #<primitive-procedure NAME>.
5254 ** Backtraces slightly more intelligible.
5255 gsubr-apply and macro transformer application frames no longer appear
5258 * Changes to Scheme functions and syntax
5260 ** Guile now correctly handles internal defines by rewriting them into
5261 their equivalent letrec. Previously, internal defines would
5262 incrementally add to the innermost environment, without checking
5263 whether the restrictions specified in RnRS were met. This lead to the
5264 correct behaviour when these restriction actually were met, but didn't
5265 catch all illegal uses. Such an illegal use could lead to crashes of
5266 the Guile interpreter or or other unwanted results. An example of
5267 incorrect internal defines that made Guile behave erratically:
5279 The problem with this example is that the definition of `c' uses the
5280 value of `b' directly. This confuses the meoization machine of Guile
5281 so that the second call of `b' (this time in a larger environment that
5282 also contains bindings for `c' and `d') refers to the binding of `c'
5283 instead of `a'. You could also make Guile crash with a variation on
5288 (define (b flag) (if flag a 1))
5289 (define c (1+ (b flag)))
5297 From now on, Guile will issue an `Unbound variable: b' error message
5302 A hook contains a list of functions which should be called on
5303 particular occasions in an existing program. Hooks are used for
5306 A window manager might have a hook before-window-map-hook. The window
5307 manager uses the function run-hooks to call all functions stored in
5308 before-window-map-hook each time a window is mapped. The user can
5309 store functions in the hook using add-hook!.
5311 In Guile, hooks are first class objects.
5313 *** New function: make-hook [N_ARGS]
5315 Return a hook for hook functions which can take N_ARGS arguments.
5316 The default value for N_ARGS is 0.
5318 (See also scm_make_named_hook below.)
5320 *** New function: add-hook! HOOK PROC [APPEND_P]
5322 Put PROC at the beginning of the list of functions stored in HOOK.
5323 If APPEND_P is supplied, and non-false, put PROC at the end instead.
5325 PROC must be able to take the number of arguments specified when the
5328 If PROC already exists in HOOK, then remove it first.
5330 *** New function: remove-hook! HOOK PROC
5332 Remove PROC from the list of functions in HOOK.
5334 *** New function: reset-hook! HOOK
5336 Clear the list of hook functions stored in HOOK.
5338 *** New function: run-hook HOOK ARG1 ...
5340 Run all hook functions stored in HOOK with arguments ARG1 ... .
5341 The number of arguments supplied must correspond to the number given
5342 when the hook was created.
5344 ** The function `dynamic-link' now takes optional keyword arguments.
5345 The only keyword argument that is currently defined is `:global
5346 BOOL'. With it, you can control whether the shared library will be
5347 linked in global mode or not. In global mode, the symbols from the
5348 linked library can be used to resolve references from other
5349 dynamically linked libraries. In non-global mode, the linked
5350 library is essentially invisible and can only be accessed via
5351 `dynamic-func', etc. The default is now to link in global mode.
5352 Previously, the default has been non-global mode.
5354 The `#:global' keyword is only effective on platforms that support
5355 the dlopen family of functions.
5357 ** New function `provided?'
5359 - Function: provided? FEATURE
5360 Return true iff FEATURE is supported by this installation of
5361 Guile. FEATURE must be a symbol naming a feature; the global
5362 variable `*features*' is a list of available features.
5364 ** Changes to the module (ice-9 expect):
5366 *** The expect-strings macro now matches `$' in a regular expression
5367 only at a line-break or end-of-file by default. Previously it would
5368 match the end of the string accumulated so far. The old behaviour
5369 can be obtained by setting the variable `expect-strings-exec-flags'
5372 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
5373 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
5374 in a regular expression will still match before a line-break or
5375 end-of-file. The default is `regexp/noteol'.
5377 *** The expect-strings macro now uses a variable
5378 `expect-strings-compile-flags' for the flags to be supplied to
5379 `make-regexp'. The default is `regexp/newline', which was previously
5382 *** The expect macro now supplies two arguments to a match procedure:
5383 the current accumulated string and a flag to indicate whether
5384 end-of-file has been reached. Previously only the string was supplied.
5385 If end-of-file is reached, the match procedure will be called an
5386 additional time with the same accumulated string as the previous call
5387 but with the flag set.
5389 ** New module (ice-9 format), implementing the Common Lisp `format' function.
5391 This code, and the documentation for it that appears here, was
5392 borrowed from SLIB, with minor adaptations for Guile.
5394 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
5395 An almost complete implementation of Common LISP format description
5396 according to the CL reference book `Common LISP' from Guy L.
5397 Steele, Digital Press. Backward compatible to most of the
5398 available Scheme format implementations.
5400 Returns `#t', `#f' or a string; has side effect of printing
5401 according to FORMAT-STRING. If DESTINATION is `#t', the output is
5402 to the current output port and `#t' is returned. If DESTINATION
5403 is `#f', a formatted string is returned as the result of the call.
5404 NEW: If DESTINATION is a string, DESTINATION is regarded as the
5405 format string; FORMAT-STRING is then the first argument and the
5406 output is returned as a string. If DESTINATION is a number, the
5407 output is to the current error port if available by the
5408 implementation. Otherwise DESTINATION must be an output port and
5411 FORMAT-STRING must be a string. In case of a formatting error
5412 format returns `#f' and prints a message on the current output or
5413 error port. Characters are output as if the string were output by
5414 the `display' function with the exception of those prefixed by a
5415 tilde (~). For a detailed description of the FORMAT-STRING syntax
5416 please consult a Common LISP format reference manual. For a test
5417 suite to verify this format implementation load `formatst.scm'.
5418 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
5420 Note: `format' is not reentrant, i.e. only one `format'-call may
5421 be executed at a time.
5424 *** Format Specification (Format version 3.0)
5426 Please consult a Common LISP format reference manual for a detailed
5427 description of the format string syntax. For a demonstration of the
5428 implemented directives see `formatst.scm'.
5430 This implementation supports directive parameters and modifiers (`:'
5431 and `@' characters). Multiple parameters must be separated by a comma
5432 (`,'). Parameters can be numerical parameters (positive or negative),
5433 character parameters (prefixed by a quote character (`''), variable
5434 parameters (`v'), number of rest arguments parameter (`#'), empty and
5435 default parameters. Directive characters are case independent. The
5436 general form of a directive is:
5438 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
5440 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
5442 *** Implemented CL Format Control Directives
5444 Documentation syntax: Uppercase characters represent the
5445 corresponding control directive characters. Lowercase characters
5446 represent control directive parameter descriptions.
5449 Any (print as `display' does).
5453 `~MINCOL,COLINC,MINPAD,PADCHARA'
5457 S-expression (print as `write' does).
5461 `~MINCOL,COLINC,MINPAD,PADCHARS'
5467 print number sign always.
5470 print comma separated.
5472 `~MINCOL,PADCHAR,COMMACHARD'
5478 print number sign always.
5481 print comma separated.
5483 `~MINCOL,PADCHAR,COMMACHARX'
5489 print number sign always.
5492 print comma separated.
5494 `~MINCOL,PADCHAR,COMMACHARO'
5500 print number sign always.
5503 print comma separated.
5505 `~MINCOL,PADCHAR,COMMACHARB'
5510 `~N,MINCOL,PADCHAR,COMMACHARR'
5514 print a number as a Roman numeral.
5517 print a number as an "old fashioned" Roman numeral.
5520 print a number as an ordinal English number.
5523 print a number as a cardinal English number.
5528 prints `y' and `ies'.
5531 as `~P but jumps 1 argument backward.'
5534 as `~@P but jumps 1 argument backward.'
5539 prints a character as the reader can understand it (i.e. `#\'
5543 prints a character as emacs does (eg. `^C' for ASCII 03).
5546 Fixed-format floating-point (prints a flonum like MMM.NNN).
5547 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
5549 If the number is positive a plus sign is printed.
5552 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
5553 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
5555 If the number is positive a plus sign is printed.
5558 General floating-point (prints a flonum either fixed or
5560 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
5562 If the number is positive a plus sign is printed.
5565 Dollars floating-point (prints a flonum in fixed with signs
5567 `~DIGITS,SCALE,WIDTH,PADCHAR$'
5569 If the number is positive a plus sign is printed.
5572 A sign is always printed and appears before the padding.
5575 The sign appears before the padding.
5583 print newline if not at the beginning of the output line.
5585 prints `~&' and then N-1 newlines.
5590 print N page separators.
5600 newline is ignored, white space left.
5603 newline is left, white space ignored.
5608 relative tabulation.
5614 Indirection (expects indirect arguments as a list).
5616 extracts indirect arguments from format arguments.
5619 Case conversion (converts by `string-downcase').
5621 converts by `string-capitalize'.
5624 converts by `string-capitalize-first'.
5627 converts by `string-upcase'.
5630 Argument Jumping (jumps 1 argument forward).
5632 jumps N arguments forward.
5635 jumps 1 argument backward.
5638 jumps N arguments backward.
5641 jumps to the 0th argument.
5644 jumps to the Nth argument (beginning from 0)
5646 `~[STR0~;STR1~;...~;STRN~]'
5647 Conditional Expression (numerical clause conditional).
5649 take argument from N.
5652 true test conditional.
5655 if-else-then conditional.
5661 default clause follows.
5664 Iteration (args come from the next argument (a list)).
5666 at most N iterations.
5669 args from next arg (a list of lists).
5672 args from the rest of arguments.
5675 args from the rest args (lists).
5686 aborts if N <= M <= K
5688 *** Not Implemented CL Format Control Directives
5691 print `#f' as an empty list (see below).
5694 print `#f' as an empty list (see below).
5700 (sorry I don't understand its semantics completely)
5702 *** Extended, Replaced and Additional Control Directives
5704 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
5705 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
5706 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
5707 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
5708 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
5709 COMMAWIDTH is the number of characters between two comma
5713 print a R4RS complex number as `~F~@Fi' with passed parameters for
5717 Pretty print formatting of an argument for scheme code lists.
5723 Flushes the output if format DESTINATION is a port.
5726 Print a `#\space' character
5728 print N `#\space' characters.
5731 Print a `#\tab' character
5733 print N `#\tab' characters.
5736 Takes N as an integer representation for a character. No arguments
5737 are consumed. N is converted to a character by `integer->char'. N
5738 must be a positive decimal number.
5741 Print out readproof. Prints out internal objects represented as
5742 `#<...>' as strings `"#<...>"' so that the format output can always
5743 be processed by `read'.
5746 Print out readproof. Prints out internal objects represented as
5747 `#<...>' as strings `"#<...>"' so that the format output can always
5748 be processed by `read'.
5751 Prints information and a copyright notice on the format
5754 prints format version.
5757 may also print number strings, i.e. passing a number as a string
5758 and format it accordingly.
5760 *** Configuration Variables
5762 The format module exports some configuration variables to suit the
5763 systems and users needs. There should be no modification necessary for
5764 the configuration that comes with Guile. Format detects automatically
5765 if the running scheme system implements floating point numbers and
5768 format:symbol-case-conv
5769 Symbols are converted by `symbol->string' so the case type of the
5770 printed symbols is implementation dependent.
5771 `format:symbol-case-conv' is a one arg closure which is either
5772 `#f' (no conversion), `string-upcase', `string-downcase' or
5773 `string-capitalize'. (default `#f')
5775 format:iobj-case-conv
5776 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
5777 implementation internal objects. (default `#f')
5780 The character prefixing the exponent value in `~E' printing.
5783 *** Compatibility With Other Format Implementations
5789 Downward compatible except for padding support and `~A', `~S',
5790 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
5791 `printf' padding support which is completely replaced by the CL
5792 `format' padding style.
5795 Downward compatible except for `~', which is not documented
5796 (ignores all characters inside the format string up to a newline
5797 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
5798 numerical and variable parameters and `:/@' modifiers in the CL
5802 Downward compatible except for `~A' and `~S' which print in
5803 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
5804 directive parameters or modifiers)).
5807 Downward compatible except for an optional destination parameter:
5808 S2C accepts a format call without a destination which returns a
5809 formatted string. This is equivalent to a #f destination in S2C.
5810 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
5811 parameters or modifiers)).
5814 ** Changes to string-handling functions.
5816 These functions were added to support the (ice-9 format) module, above.
5818 *** New function: string-upcase STRING
5819 *** New function: string-downcase STRING
5821 These are non-destructive versions of the existing string-upcase! and
5822 string-downcase! functions.
5824 *** New function: string-capitalize! STRING
5825 *** New function: string-capitalize STRING
5827 These functions convert the first letter of each word in the string to
5830 (string-capitalize "howdy there")
5833 As with the other functions, string-capitalize! modifies the string in
5834 place, while string-capitalize returns a modified copy of its argument.
5836 *** New function: string-ci->symbol STRING
5838 Return a symbol whose name is STRING, but having the same case as if
5839 the symbol had be read by `read'.
5841 Guile can be configured to be sensitive or insensitive to case
5842 differences in Scheme identifiers. If Guile is case-insensitive, all
5843 symbols are converted to lower case on input. The `string-ci->symbol'
5844 function returns a symbol whose name in STRING, transformed as Guile
5845 would if STRING were input.
5847 *** New function: substring-move! STRING1 START END STRING2 START
5849 Copy the substring of STRING1 from START (inclusive) to END
5850 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
5851 string, and the source and destination areas may overlap; in all
5852 cases, the function behaves as if all the characters were copied
5855 *** Extended functions: substring-move-left! substring-move-right!
5857 These functions now correctly copy arbitrarily overlapping substrings;
5858 they are both synonyms for substring-move!.
5861 ** New module (ice-9 getopt-long), with the function `getopt-long'.
5863 getopt-long is a function for parsing command-line arguments in a
5864 manner consistent with other GNU programs.
5866 (getopt-long ARGS GRAMMAR)
5867 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
5869 ARGS should be a list of strings. Its first element should be the
5870 name of the program; subsequent elements should be the arguments
5871 that were passed to the program on the command line. The
5872 `program-arguments' procedure returns a list of this form.
5874 GRAMMAR is a list of the form:
5875 ((OPTION (PROPERTY VALUE) ...) ...)
5877 Each OPTION should be a symbol. `getopt-long' will accept a
5878 command-line option named `--OPTION'.
5879 Each option can have the following (PROPERTY VALUE) pairs:
5881 (single-char CHAR) --- Accept `-CHAR' as a single-character
5882 equivalent to `--OPTION'. This is how to specify traditional
5884 (required? BOOL) --- If BOOL is true, the option is required.
5885 getopt-long will raise an error if it is not found in ARGS.
5886 (value BOOL) --- If BOOL is #t, the option accepts a value; if
5887 it is #f, it does not; and if it is the symbol
5888 `optional', the option may appear in ARGS with or
5890 (predicate FUNC) --- If the option accepts a value (i.e. you
5891 specified `(value #t)' for this option), then getopt
5892 will apply FUNC to the value, and throw an exception
5893 if it returns #f. FUNC should be a procedure which
5894 accepts a string and returns a boolean value; you may
5895 need to use quasiquotes to get it into GRAMMAR.
5897 The (PROPERTY VALUE) pairs may occur in any order, but each
5898 property may occur only once. By default, options do not have
5899 single-character equivalents, are not required, and do not take
5902 In ARGS, single-character options may be combined, in the usual
5903 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
5904 accepts values, then it must be the last option in the
5905 combination; the value is the next argument. So, for example, using
5906 the following grammar:
5907 ((apples (single-char #\a))
5908 (blimps (single-char #\b) (value #t))
5909 (catalexis (single-char #\c) (value #t)))
5910 the following argument lists would be acceptable:
5911 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
5912 for "blimps" and "catalexis")
5913 ("-ab" "bang" "-c" "couth") (same)
5914 ("-ac" "couth" "-b" "bang") (same)
5915 ("-abc" "couth" "bang") (an error, since `-b' is not the
5916 last option in its combination)
5918 If an option's value is optional, then `getopt-long' decides
5919 whether it has a value by looking at what follows it in ARGS. If
5920 the next element is a string, and it does not appear to be an
5921 option itself, then that string is the option's value.
5923 The value of a long option can appear as the next element in ARGS,
5924 or it can follow the option name, separated by an `=' character.
5925 Thus, using the same grammar as above, the following argument lists
5927 ("--apples" "Braeburn" "--blimps" "Goodyear")
5928 ("--apples=Braeburn" "--blimps" "Goodyear")
5929 ("--blimps" "Goodyear" "--apples=Braeburn")
5931 If the option "--" appears in ARGS, argument parsing stops there;
5932 subsequent arguments are returned as ordinary arguments, even if
5933 they resemble options. So, in the argument list:
5934 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
5935 `getopt-long' will recognize the `apples' option as having the
5936 value "Granny Smith", but it will not recognize the `blimp'
5937 option; it will return the strings "--blimp" and "Goodyear" as
5938 ordinary argument strings.
5940 The `getopt-long' function returns the parsed argument list as an
5941 assocation list, mapping option names --- the symbols from GRAMMAR
5942 --- onto their values, or #t if the option does not accept a value.
5943 Unused options do not appear in the alist.
5945 All arguments that are not the value of any option are returned
5946 as a list, associated with the empty list.
5948 `getopt-long' throws an exception if:
5949 - it finds an unrecognized option in ARGS
5950 - a required option is omitted
5951 - an option that requires an argument doesn't get one
5952 - an option that doesn't accept an argument does get one (this can
5953 only happen using the long option `--opt=value' syntax)
5954 - an option predicate fails
5959 `((lockfile-dir (required? #t)
5962 (predicate ,file-is-directory?))
5963 (verbose (required? #f)
5966 (x-includes (single-char #\x))
5967 (rnet-server (single-char #\y)
5968 (predicate ,string?))))
5970 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
5971 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
5973 => ((() "foo1" "-fred" "foo2" "foo3")
5974 (rnet-server . "lamprod")
5975 (x-includes . "/usr/include")
5976 (lockfile-dir . "/tmp")
5979 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
5981 It will be removed in a few releases.
5983 ** New syntax: lambda*
5984 ** New syntax: define*
5985 ** New syntax: define*-public
5986 ** New syntax: defmacro*
5987 ** New syntax: defmacro*-public
5988 Guile now supports optional arguments.
5990 `lambda*', `define*', `define*-public', `defmacro*' and
5991 `defmacro*-public' are identical to the non-* versions except that
5992 they use an extended type of parameter list that has the following BNF
5993 syntax (parentheses are literal, square brackets indicate grouping,
5994 and `*', `+' and `?' have the usual meaning):
5996 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
5997 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
5998 [[#&rest identifier]|[. identifier]]? ) | [identifier]
6000 ext-var-decl ::= identifier | ( identifier expression )
6002 The semantics are best illustrated with the following documentation
6003 and examples for `lambda*':
6006 lambda extended for optional and keyword arguments
6008 lambda* creates a procedure that takes optional arguments. These
6009 are specified by putting them inside brackets at the end of the
6010 paramater list, but before any dotted rest argument. For example,
6011 (lambda* (a b #&optional c d . e) '())
6012 creates a procedure with fixed arguments a and b, optional arguments c
6013 and d, and rest argument e. If the optional arguments are omitted
6014 in a call, the variables for them are unbound in the procedure. This
6015 can be checked with the bound? macro.
6017 lambda* can also take keyword arguments. For example, a procedure
6019 (lambda* (#&key xyzzy larch) '())
6020 can be called with any of the argument lists (#:xyzzy 11)
6021 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
6022 are given as keywords are bound to values.
6024 Optional and keyword arguments can also be given default values
6025 which they take on when they are not present in a call, by giving a
6026 two-item list in place of an optional argument, for example in:
6027 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
6028 foo is a fixed argument, bar is an optional argument with default
6029 value 42, and baz is a keyword argument with default value 73.
6030 Default value expressions are not evaluated unless they are needed
6031 and until the procedure is called.
6033 lambda* now supports two more special parameter list keywords.
6035 lambda*-defined procedures now throw an error by default if a
6036 keyword other than one of those specified is found in the actual
6037 passed arguments. However, specifying #&allow-other-keys
6038 immediately after the kyword argument declarations restores the
6039 previous behavior of ignoring unknown keywords. lambda* also now
6040 guarantees that if the same keyword is passed more than once, the
6041 last one passed is the one that takes effect. For example,
6042 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
6043 #:heads 37 #:tails 42 #:heads 99)
6044 would result in (99 47) being displayed.
6046 #&rest is also now provided as a synonym for the dotted syntax rest
6047 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
6048 all respects to lambda*. This is provided for more similarity to DSSSL,
6049 MIT-Scheme and Kawa among others, as well as for refugees from other
6052 Further documentation may be found in the optargs.scm file itself.
6054 The optional argument module also exports the macros `let-optional',
6055 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
6056 are not documented here because they may be removed in the future, but
6057 full documentation is still available in optargs.scm.
6059 ** New syntax: and-let*
6060 Guile now supports the `and-let*' form, described in the draft SRFI-2.
6062 Syntax: (land* (<clause> ...) <body> ...)
6063 Each <clause> should have one of the following forms:
6064 (<variable> <expression>)
6067 Each <variable> or <bound-variable> should be an identifier. Each
6068 <expression> should be a valid expression. The <body> should be a
6069 possibly empty sequence of expressions, like the <body> of a
6072 Semantics: A LAND* expression is evaluated by evaluating the
6073 <expression> or <bound-variable> of each of the <clause>s from
6074 left to right. The value of the first <expression> or
6075 <bound-variable> that evaluates to a false value is returned; the
6076 remaining <expression>s and <bound-variable>s are not evaluated.
6077 The <body> forms are evaluated iff all the <expression>s and
6078 <bound-variable>s evaluate to true values.
6080 The <expression>s and the <body> are evaluated in an environment
6081 binding each <variable> of the preceding (<variable> <expression>)
6082 clauses to the value of the <expression>. Later bindings
6083 shadow earlier bindings.
6085 Guile's and-let* macro was contributed by Michael Livshin.
6087 ** New sorting functions
6089 *** New function: sorted? SEQUENCE LESS?
6090 Returns `#t' when the sequence argument is in non-decreasing order
6091 according to LESS? (that is, there is no adjacent pair `... x y
6092 ...' for which `(less? y x)').
6094 Returns `#f' when the sequence contains at least one out-of-order
6095 pair. It is an error if the sequence is neither a list nor a
6098 *** New function: merge LIST1 LIST2 LESS?
6099 LIST1 and LIST2 are sorted lists.
6100 Returns the sorted list of all elements in LIST1 and LIST2.
6102 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
6103 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
6104 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
6105 (Here "<" should read "comes before".)
6107 *** New procedure: merge! LIST1 LIST2 LESS?
6108 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
6109 the result. If the code is compiled, and LESS? constructs no new
6110 pairs, no pairs at all will be allocated. The first pair of the
6111 result will be either the first pair of LIST1 or the first pair of
6114 *** New function: sort SEQUENCE LESS?
6115 Accepts either a list or a vector, and returns a new sequence
6116 which is sorted. The new sequence is the same type as the input.
6117 Always `(sorted? (sort sequence less?) less?)'. The original
6118 sequence is not altered in any way. The new sequence shares its
6119 elements with the old one; no elements are copied.
6121 *** New procedure: sort! SEQUENCE LESS
6122 Returns its sorted result in the original boxes. No new storage is
6123 allocated at all. Proper usage: (set! slist (sort! slist <))
6125 *** New function: stable-sort SEQUENCE LESS?
6126 Similar to `sort' but stable. That is, if "equal" elements are
6127 ordered a < b in the original sequence, they will have the same order
6130 *** New function: stable-sort! SEQUENCE LESS?
6131 Similar to `sort!' but stable.
6132 Uses temporary storage when sorting vectors.
6134 *** New functions: sort-list, sort-list!
6135 Added for compatibility with scsh.
6137 ** New built-in random number support
6139 *** New function: random N [STATE]
6140 Accepts a positive integer or real N and returns a number of the
6141 same type between zero (inclusive) and N (exclusive). The values
6142 returned have a uniform distribution.
6144 The optional argument STATE must be of the type produced by
6145 `copy-random-state' or `seed->random-state'. It defaults to the value
6146 of the variable `*random-state*'. This object is used to maintain the
6147 state of the pseudo-random-number generator and is altered as a side
6148 effect of the `random' operation.
6150 *** New variable: *random-state*
6151 Holds a data structure that encodes the internal state of the
6152 random-number generator that `random' uses by default. The nature
6153 of this data structure is implementation-dependent. It may be
6154 printed out and successfully read back in, but may or may not
6155 function correctly as a random-number state object in another
6158 *** New function: copy-random-state [STATE]
6159 Returns a new object of type suitable for use as the value of the
6160 variable `*random-state*' and as a second argument to `random'.
6161 If argument STATE is given, a copy of it is returned. Otherwise a
6162 copy of `*random-state*' is returned.
6164 *** New function: seed->random-state SEED
6165 Returns a new object of type suitable for use as the value of the
6166 variable `*random-state*' and as a second argument to `random'.
6167 SEED is a string or a number. A new state is generated and
6168 initialized using SEED.
6170 *** New function: random:uniform [STATE]
6171 Returns an uniformly distributed inexact real random number in the
6172 range between 0 and 1.
6174 *** New procedure: random:solid-sphere! VECT [STATE]
6175 Fills VECT with inexact real random numbers the sum of whose
6176 squares is less than 1.0. Thinking of VECT as coordinates in
6177 space of dimension N = `(vector-length VECT)', the coordinates are
6178 uniformly distributed within the unit N-shere. The sum of the
6179 squares of the numbers is returned. VECT can be either a vector
6180 or a uniform vector of doubles.
6182 *** New procedure: random:hollow-sphere! VECT [STATE]
6183 Fills VECT with inexact real random numbers the sum of whose squares
6184 is equal to 1.0. Thinking of VECT as coordinates in space of
6185 dimension n = `(vector-length VECT)', the coordinates are uniformly
6186 distributed over the surface of the unit n-shere. VECT can be either
6187 a vector or a uniform vector of doubles.
6189 *** New function: random:normal [STATE]
6190 Returns an inexact real in a normal distribution with mean 0 and
6191 standard deviation 1. For a normal distribution with mean M and
6192 standard deviation D use `(+ M (* D (random:normal)))'.
6194 *** New procedure: random:normal-vector! VECT [STATE]
6195 Fills VECT with inexact real random numbers which are independent and
6196 standard normally distributed (i.e., with mean 0 and variance 1).
6197 VECT can be either a vector or a uniform vector of doubles.
6199 *** New function: random:exp STATE
6200 Returns an inexact real in an exponential distribution with mean 1.
6201 For an exponential distribution with mean U use (* U (random:exp)).
6203 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
6205 These functions now operate on numbers in the range of a C unsigned
6208 These functions used to operate on numbers in the range of a C signed
6209 long; however, this seems inappropriate, because Guile integers don't
6212 ** New function: make-guardian
6213 This is an implementation of guardians as described in
6214 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
6215 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
6216 Programming Language Design and Implementation, June 1993
6217 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
6219 ** New functions: delq1!, delv1!, delete1!
6220 These procedures behave similar to delq! and friends but delete only
6221 one object if at all.
6223 ** New function: unread-string STRING PORT
6224 Unread STRING to PORT, that is, push it back onto the port so that
6225 next read operation will work on the pushed back characters.
6227 ** unread-char can now be called multiple times
6228 If unread-char is called multiple times, the unread characters will be
6229 read again in last-in first-out order.
6231 ** the procedures uniform-array-read! and uniform-array-write! now
6232 work on any kind of port, not just ports which are open on a file.
6234 ** Now 'l' in a port mode requests line buffering.
6236 ** The procedure truncate-file now works on string ports as well
6237 as file ports. If the size argument is omitted, the current
6238 file position is used.
6240 ** new procedure: seek PORT/FDES OFFSET WHENCE
6241 The arguments are the same as for the old fseek procedure, but it
6242 works on string ports as well as random-access file ports.
6244 ** the fseek procedure now works on string ports, since it has been
6245 redefined using seek.
6247 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
6248 size is not supplied.
6250 ** the newline procedure no longer flushes the port if it's not
6251 line-buffered: previously it did if it was the current output port.
6253 ** open-pipe and close-pipe are no longer primitive procedures, but
6254 an emulation can be obtained using `(use-modules (ice-9 popen))'.
6256 ** the freopen procedure has been removed.
6258 ** new procedure: drain-input PORT
6259 Drains PORT's read buffers (including any pushed-back characters)
6260 and returns the contents as a single string.
6262 ** New function: map-in-order PROC LIST1 LIST2 ...
6263 Version of `map' which guarantees that the procedure is applied to the
6264 lists in serial order.
6266 ** Renamed `serial-array-copy!' and `serial-array-map!' to
6267 `array-copy-in-order!' and `array-map-in-order!'. The old names are
6268 now obsolete and will go away in release 1.5.
6270 ** New syntax: collect BODY1 ...
6271 Version of `begin' which returns a list of the results of the body
6272 forms instead of the result of the last body form. In contrast to
6273 `begin', `collect' allows an empty body.
6275 ** New functions: read-history FILENAME, write-history FILENAME
6276 Read/write command line history from/to file. Returns #t on success
6277 and #f if an error occured.
6279 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
6281 These procedures return a list of definitions available in the specified
6282 argument, a relative module reference. In the case of no argument,
6283 `(current-module)' is now consulted for definitions to return, instead
6284 of simply returning #f, the former behavior.
6286 ** The #/ syntax for lists is no longer supported.
6288 Earlier versions of Scheme accepted this syntax, but printed a
6291 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
6293 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
6296 * Changes to the gh_ interface
6300 Now takes a second argument which is the result array. If this
6301 pointer is NULL, a new array is malloced (the old behaviour).
6303 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
6304 gh_scm2shorts, gh_scm2longs, gh_scm2floats
6308 * Changes to the scm_ interface
6310 ** Function: scm_make_named_hook (char* name, int n_args)
6312 Creates a hook in the same way as make-hook above but also
6313 binds a variable named NAME to it.
6315 This is the typical way of creating a hook from C code.
6317 Currently, the variable is created in the "current" module. This
6318 might change when we get the new module system.
6320 ** The smob interface
6322 The interface for creating smobs has changed. For documentation, see
6323 data-rep.info (made from guile-core/doc/data-rep.texi).
6325 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
6327 >>> This function will be removed in 1.3.4. <<<
6331 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
6332 This function adds a new smob type, named NAME, with instance size
6333 SIZE to the system. The return value is a tag that is used in
6334 creating instances of the type. If SIZE is 0, then no memory will
6335 be allocated when instances of the smob are created, and nothing
6336 will be freed by the default free function.
6338 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
6339 This function sets the smob marking procedure for the smob type
6340 specified by the tag TC. TC is the tag returned by
6341 `scm_make_smob_type'.
6343 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
6344 This function sets the smob freeing procedure for the smob type
6345 specified by the tag TC. TC is the tag returned by
6346 `scm_make_smob_type'.
6348 *** Function: void scm_set_smob_print (tc, print)
6350 - Function: void scm_set_smob_print (long tc,
6351 scm_sizet (*print) (SCM,
6355 This function sets the smob printing procedure for the smob type
6356 specified by the tag TC. TC is the tag returned by
6357 `scm_make_smob_type'.
6359 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
6360 This function sets the smob equality-testing predicate for the
6361 smob type specified by the tag TC. TC is the tag returned by
6362 `scm_make_smob_type'.
6364 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
6365 Make VALUE contain a smob instance of the type with type code TC and
6366 smob data DATA. VALUE must be previously declared as C type `SCM'.
6368 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
6369 This macro expands to a block of code that creates a smob instance
6370 of the type with type code TC and smob data DATA, and returns that
6371 `SCM' value. It should be the last piece of code in a block.
6373 ** The interfaces for using I/O ports and implementing port types
6374 (ptobs) have changed significantly. The new interface is based on
6375 shared access to buffers and a new set of ptob procedures.
6377 *** scm_newptob has been removed
6381 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
6383 - Function: SCM scm_make_port_type (char *type_name,
6384 int (*fill_buffer) (SCM port),
6385 void (*write_flush) (SCM port));
6387 Similarly to the new smob interface, there is a set of function
6388 setters by which the user can customize the behaviour of his port
6389 type. See ports.h (scm_set_port_XXX).
6391 ** scm_strport_to_string: New function: creates a new string from
6392 a string port's buffer.
6394 ** Plug in interface for random number generators
6395 The variable `scm_the_rng' in random.c contains a value and three
6396 function pointers which together define the current random number
6397 generator being used by the Scheme level interface and the random
6398 number library functions.
6400 The user is free to replace the default generator with the generator
6403 *** Variable: size_t scm_the_rng.rstate_size
6404 The size of the random state type used by the current RNG
6407 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
6408 Given the random STATE, return 32 random bits.
6410 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
6411 Seed random state STATE using string S of length N.
6413 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
6414 Given random state STATE, return a malloced copy.
6417 The default RNG is the MWC (Multiply With Carry) random number
6418 generator described by George Marsaglia at the Department of
6419 Statistics and Supercomputer Computations Research Institute, The
6420 Florida State University (http://stat.fsu.edu/~geo).
6422 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
6423 passes all tests in the DIEHARD test suite
6424 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
6425 costs one multiply and one add on platforms which either supports long
6426 longs (gcc does this on most systems) or have 64 bit longs. The cost
6427 is four multiply on other systems but this can be optimized by writing
6428 scm_i_uniform32 in assembler.
6430 These functions are provided through the scm_the_rng interface for use
6431 by libguile and the application.
6433 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
6434 Given the random STATE, return 32 random bits.
6435 Don't use this function directly. Instead go through the plugin
6436 interface (see "Plug in interface" above).
6438 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
6439 Initialize STATE using SEED of length N.
6441 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
6442 Return a malloc:ed copy of STATE. This function can easily be re-used
6443 in the interfaces to other RNGs.
6445 ** Random number library functions
6446 These functions use the current RNG through the scm_the_rng interface.
6447 It might be a good idea to use these functions from your C code so
6448 that only one random generator is used by all code in your program.
6450 The default random state is stored in:
6452 *** Variable: SCM scm_var_random_state
6453 Contains the vcell of the Scheme variable "*random-state*" which is
6454 used as default state by all random number functions in the Scheme
6459 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
6461 *** Function: scm_rstate *scm_c_default_rstate (void)
6462 This is a convenience function which returns the value of
6463 scm_var_random_state. An error message is generated if this value
6464 isn't a random state.
6466 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
6467 Make a new random state from the string SEED of length LENGTH.
6469 It is generally not a good idea to use multiple random states in a
6470 program. While subsequent random numbers generated from one random
6471 state are guaranteed to be reasonably independent, there is no such
6472 guarantee for numbers generated from different random states.
6474 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
6475 Return 32 random bits.
6477 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
6478 Return a sample from the uniform(0,1) distribution.
6480 *** Function: double scm_c_normal01 (scm_rstate *STATE)
6481 Return a sample from the normal(0,1) distribution.
6483 *** Function: double scm_c_exp1 (scm_rstate *STATE)
6484 Return a sample from the exp(1) distribution.
6486 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
6487 Return a sample from the discrete uniform(0,M) distribution.
6489 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
6490 Return a sample from the discrete uniform(0,M) distribution.
6491 M must be a bignum object. The returned value may be an INUM.
6495 Changes in Guile 1.3 (released Monday, October 19, 1998):
6497 * Changes to the distribution
6499 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
6500 To avoid conflicts, programs should name environment variables after
6501 themselves, except when there's a common practice establishing some
6504 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
6505 giving the former precedence, and printing a warning message if the
6506 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
6508 ** The header files related to multi-byte characters have been removed.
6509 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
6510 which referred to these explicitly will probably need to be rewritten,
6511 since the support for the variant string types has been removed; see
6514 ** The header files append.h and sequences.h have been removed. These
6515 files implemented non-R4RS operations which would encourage
6516 non-portable programming style and less easy-to-read code.
6518 * Changes to the stand-alone interpreter
6520 ** New procedures have been added to implement a "batch mode":
6522 *** Function: batch-mode?
6524 Returns a boolean indicating whether the interpreter is in batch
6527 *** Function: set-batch-mode?! ARG
6529 If ARG is true, switches the interpreter to batch mode. The `#f'
6530 case has not been implemented.
6532 ** Guile now provides full command-line editing, when run interactively.
6533 To use this feature, you must have the readline library installed.
6534 The Guile build process will notice it, and automatically include
6537 The readline library is available via anonymous FTP from any GNU
6538 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
6540 ** the-last-stack is now a fluid.
6542 * Changes to the procedure for linking libguile with your programs
6544 ** You can now use the `guile-config' utility to build programs that use Guile.
6546 Guile now includes a command-line utility called `guile-config', which
6547 can provide information about how to compile and link programs that
6550 *** `guile-config compile' prints any C compiler flags needed to use Guile.
6551 You should include this command's output on the command line you use
6552 to compile C or C++ code that #includes the Guile header files. It's
6553 usually just a `-I' flag to help the compiler find the Guile headers.
6556 *** `guile-config link' prints any linker flags necessary to link with Guile.
6558 This command writes to its standard output a list of flags which you
6559 must pass to the linker to link your code against the Guile library.
6560 The flags include '-lguile' itself, any other libraries the Guile
6561 library depends upon, and any `-L' flags needed to help the linker
6562 find those libraries.
6564 For example, here is a Makefile rule that builds a program named 'foo'
6565 from the object files ${FOO_OBJECTS}, and links them against Guile:
6568 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
6570 Previous Guile releases recommended that you use autoconf to detect
6571 which of a predefined set of libraries were present on your system.
6572 It is more robust to use `guile-config', since it records exactly which
6573 libraries the installed Guile library requires.
6575 This was originally called `build-guile', but was renamed to
6576 `guile-config' before Guile 1.3 was released, to be consistent with
6577 the analogous script for the GTK+ GUI toolkit, which is called
6581 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
6583 If you are using the GNU autoconf package to configure your program,
6584 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
6585 (described above) and gather the necessary values for use in your
6588 The GUILE_FLAGS macro expands to configure script code which runs the
6589 `guile-config' script, to find out where Guile's header files and
6590 libraries are installed. It sets two variables, marked for
6591 substitution, as by AC_SUBST.
6593 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
6594 code that uses Guile header files. This is almost always just a
6597 GUILE_LDFLAGS --- flags to pass to the linker to link a
6598 program against Guile. This includes `-lguile' for the Guile
6599 library itself, any libraries that Guile itself requires (like
6600 -lqthreads), and so on. It may also include a -L flag to tell the
6601 compiler where to find the libraries.
6603 GUILE_FLAGS is defined in the file guile.m4, in the top-level
6604 directory of the Guile distribution. You can copy it into your
6605 package's aclocal.m4 file, and then use it in your configure.in file.
6607 If you are using the `aclocal' program, distributed with GNU automake,
6608 to maintain your aclocal.m4 file, the Guile installation process
6609 installs guile.m4 where aclocal will find it. All you need to do is
6610 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
6611 this will copy the definition of GUILE_FLAGS into your aclocal.m4
6615 * Changes to Scheme functions and syntax
6617 ** Multi-byte strings have been removed, as have multi-byte and wide
6618 ports. We felt that these were the wrong approach to
6619 internationalization support.
6621 ** New function: readline [PROMPT]
6622 Read a line from the terminal, and allow the user to edit it,
6623 prompting with PROMPT. READLINE provides a large set of Emacs-like
6624 editing commands, lets the user recall previously typed lines, and
6625 works on almost every kind of terminal, including dumb terminals.
6627 READLINE assumes that the cursor is at the beginning of the line when
6628 it is invoked. Thus, you can't print a prompt yourself, and then call
6629 READLINE; you need to package up your prompt as a string, pass it to
6630 the function, and let READLINE print the prompt itself. This is
6631 because READLINE needs to know the prompt's screen width.
6633 For Guile to provide this function, you must have the readline
6634 library, version 2.1 or later, installed on your system. Readline is
6635 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
6636 any GNU mirror site.
6638 See also ADD-HISTORY function.
6640 ** New function: add-history STRING
6641 Add STRING as the most recent line in the history used by the READLINE
6642 command. READLINE does not add lines to the history itself; you must
6643 call ADD-HISTORY to make previous input available to the user.
6645 ** The behavior of the read-line function has changed.
6647 This function now uses standard C library functions to read the line,
6648 for speed. This means that it doesn not respect the value of
6649 scm-line-incrementors; it assumes that lines are delimited with
6652 (Note that this is read-line, the function that reads a line of text
6653 from a port, not readline, the function that reads a line from a
6654 terminal, providing full editing capabilities.)
6656 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
6658 This module provides some simple argument parsing. It exports one
6661 Function: getopt-gnu-style ARG-LS
6662 Parse a list of program arguments into an alist of option
6665 Each item in the list of program arguments is examined to see if
6666 it meets the syntax of a GNU long-named option. An argument like
6667 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
6668 returned alist, where MUMBLE is a keyword object with the same
6669 name as the argument. An argument like `--MUMBLE=FROB' produces
6670 an element of the form (MUMBLE . FROB), where FROB is a string.
6672 As a special case, the returned alist also contains a pair whose
6673 car is the symbol `rest'. The cdr of this pair is a list
6674 containing all the items in the argument list that are not options
6675 of the form mentioned above.
6677 The argument `--' is treated specially: all items in the argument
6678 list appearing after such an argument are not examined, and are
6679 returned in the special `rest' list.
6681 This function does not parse normal single-character switches.
6682 You will need to parse them out of the `rest' list yourself.
6684 ** The read syntax for byte vectors and short vectors has changed.
6686 Instead of #bytes(...), write #y(...).
6688 Instead of #short(...), write #h(...).
6690 This may seem nutty, but, like the other uniform vectors, byte vectors
6691 and short vectors want to have the same print and read syntax (and,
6692 more basic, want to have read syntax!). Changing the read syntax to
6693 use multiple characters after the hash sign breaks with the
6694 conventions used in R5RS and the conventions used for the other
6695 uniform vectors. It also introduces complexity in the current reader,
6696 both on the C and Scheme levels. (The Right solution is probably to
6697 change the syntax and prototypes for uniform vectors entirely.)
6700 ** The new module (ice-9 session) provides useful interactive functions.
6702 *** New procedure: (apropos REGEXP OPTION ...)
6704 Display a list of top-level variables whose names match REGEXP, and
6705 the modules they are imported from. Each OPTION should be one of the
6708 value --- Show the value of each matching variable.
6709 shadow --- Show bindings shadowed by subsequently imported modules.
6710 full --- Same as both `shadow' and `value'.
6714 guile> (apropos "trace" 'full)
6715 debug: trace #<procedure trace args>
6716 debug: untrace #<procedure untrace args>
6717 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
6718 the-scm-module: before-backtrace-hook ()
6719 the-scm-module: backtrace #<primitive-procedure backtrace>
6720 the-scm-module: after-backtrace-hook ()
6721 the-scm-module: has-shown-backtrace-hint? #f
6724 ** There are new functions and syntax for working with macros.
6726 Guile implements macros as a special object type. Any variable whose
6727 top-level binding is a macro object acts as a macro. The macro object
6728 specifies how the expression should be transformed before evaluation.
6730 *** Macro objects now print in a reasonable way, resembling procedures.
6732 *** New function: (macro? OBJ)
6733 True iff OBJ is a macro object.
6735 *** New function: (primitive-macro? OBJ)
6736 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
6737 macro transformers, implemented in eval.c rather than Scheme code.
6739 Why do we have this function?
6740 - For symmetry with procedure? and primitive-procedure?,
6741 - to allow custom print procedures to tell whether a macro is
6742 primitive, and display it differently, and
6743 - to allow compilers and user-written evaluators to distinguish
6744 builtin special forms from user-defined ones, which could be
6747 *** New function: (macro-type OBJ)
6748 Return a value indicating what kind of macro OBJ is. Possible return
6751 The symbol `syntax' --- a macro created by procedure->syntax.
6752 The symbol `macro' --- a macro created by procedure->macro.
6753 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
6754 The boolean #f --- if OBJ is not a macro object.
6756 *** New function: (macro-name MACRO)
6757 Return the name of the macro object MACRO's procedure, as returned by
6760 *** New function: (macro-transformer MACRO)
6761 Return the transformer procedure for MACRO.
6763 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
6765 Specify a new macro expander to use in the current module. Each
6766 MODULE is a module name, with the same meaning as in the `use-modules'
6767 form; each named module's exported bindings are added to the current
6768 top-level environment. TRANSFORMER is an expression evaluated in the
6769 resulting environment which must yield a procedure to use as the
6770 module's eval transformer: every expression evaluated in this module
6771 is passed to this function, and the result passed to the Guile
6774 *** macro-eval! is removed. Use local-eval instead.
6776 ** Some magic has been added to the printer to better handle user
6777 written printing routines (like record printers, closure printers).
6779 The problem is that these user written routines must have access to
6780 the current `print-state' to be able to handle fancy things like
6781 detection of circular references. These print-states have to be
6782 passed to the builtin printing routines (display, write, etc) to
6783 properly continue the print chain.
6785 We didn't want to change all existing print code so that it
6786 explicitly passes thru a print state in addition to a port. Instead,
6787 we extented the possible values that the builtin printing routines
6788 accept as a `port'. In addition to a normal port, they now also take
6789 a pair of a normal port and a print-state. Printing will go to the
6790 port and the print-state will be used to control the detection of
6791 circular references, etc. If the builtin function does not care for a
6792 print-state, it is simply ignored.
6794 User written callbacks are now called with such a pair as their
6795 `port', but because every function now accepts this pair as a PORT
6796 argument, you don't have to worry about that. In fact, it is probably
6797 safest to not check for these pairs.
6799 However, it is sometimes necessary to continue a print chain on a
6800 different port, for example to get a intermediate string
6801 representation of the printed value, mangle that string somehow, and
6802 then to finally print the mangled string. Use the new function
6804 inherit-print-state OLD-PORT NEW-PORT
6806 for this. It constructs a new `port' that prints to NEW-PORT but
6807 inherits the print-state of OLD-PORT.
6809 ** struct-vtable-offset renamed to vtable-offset-user
6811 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
6813 ** There is now a third optional argument to make-vtable-vtable
6814 (and fourth to make-struct) when constructing new types (vtables).
6815 This argument initializes field vtable-index-printer of the vtable.
6817 ** The detection of circular references has been extended to structs.
6818 That is, a structure that -- in the process of being printed -- prints
6819 itself does not lead to infinite recursion.
6821 ** There is now some basic support for fluids. Please read
6822 "libguile/fluid.h" to find out more. It is accessible from Scheme with
6823 the following functions and macros:
6825 Function: make-fluid
6827 Create a new fluid object. Fluids are not special variables or
6828 some other extension to the semantics of Scheme, but rather
6829 ordinary Scheme objects. You can store them into variables (that
6830 are still lexically scoped, of course) or into any other place you
6831 like. Every fluid has a initial value of `#f'.
6833 Function: fluid? OBJ
6835 Test whether OBJ is a fluid.
6837 Function: fluid-ref FLUID
6838 Function: fluid-set! FLUID VAL
6840 Access/modify the fluid FLUID. Modifications are only visible
6841 within the current dynamic root (that includes threads).
6843 Function: with-fluids* FLUIDS VALUES THUNK
6845 FLUIDS is a list of fluids and VALUES a corresponding list of
6846 values for these fluids. Before THUNK gets called the values are
6847 installed in the fluids and the old values of the fluids are
6848 saved in the VALUES list. When the flow of control leaves THUNK
6849 or reenters it, the values get swapped again. You might think of
6850 this as a `safe-fluid-excursion'. Note that the VALUES list is
6851 modified by `with-fluids*'.
6853 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
6855 The same as `with-fluids*' but with a different syntax. It looks
6856 just like `let', but both FLUID and VALUE are evaluated. Remember,
6857 fluids are not special variables but ordinary objects. FLUID
6858 should evaluate to a fluid.
6860 ** Changes to system call interfaces:
6862 *** close-port, close-input-port and close-output-port now return a
6863 boolean instead of an `unspecified' object. #t means that the port
6864 was successfully closed, while #f means it was already closed. It is
6865 also now possible for these procedures to raise an exception if an
6866 error occurs (some errors from write can be delayed until close.)
6868 *** the first argument to chmod, fcntl, ftell and fseek can now be a
6871 *** the third argument to fcntl is now optional.
6873 *** the first argument to chown can now be a file descriptor or a port.
6875 *** the argument to stat can now be a port.
6877 *** The following new procedures have been added (most use scsh
6880 *** procedure: close PORT/FD
6881 Similar to close-port (*note close-port: Closing Ports.), but also
6882 works on file descriptors. A side effect of closing a file
6883 descriptor is that any ports using that file descriptor are moved
6884 to a different file descriptor and have their revealed counts set
6887 *** procedure: port->fdes PORT
6888 Returns the integer file descriptor underlying PORT. As a side
6889 effect the revealed count of PORT is incremented.
6891 *** procedure: fdes->ports FDES
6892 Returns a list of existing ports which have FDES as an underlying
6893 file descriptor, without changing their revealed counts.
6895 *** procedure: fdes->inport FDES
6896 Returns an existing input port which has FDES as its underlying
6897 file descriptor, if one exists, and increments its revealed count.
6898 Otherwise, returns a new input port with a revealed count of 1.
6900 *** procedure: fdes->outport FDES
6901 Returns an existing output port which has FDES as its underlying
6902 file descriptor, if one exists, and increments its revealed count.
6903 Otherwise, returns a new output port with a revealed count of 1.
6905 The next group of procedures perform a `dup2' system call, if NEWFD
6906 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
6907 duplicated can be supplied as an integer or contained in a port. The
6908 type of value returned varies depending on which procedure is used.
6910 All procedures also have the side effect when performing `dup2' that
6911 any ports using NEWFD are moved to a different file descriptor and have
6912 their revealed counts set to zero.
6914 *** procedure: dup->fdes PORT/FD [NEWFD]
6915 Returns an integer file descriptor.
6917 *** procedure: dup->inport PORT/FD [NEWFD]
6918 Returns a new input port using the new file descriptor.
6920 *** procedure: dup->outport PORT/FD [NEWFD]
6921 Returns a new output port using the new file descriptor.
6923 *** procedure: dup PORT/FD [NEWFD]
6924 Returns a new port if PORT/FD is a port, with the same mode as the
6925 supplied port, otherwise returns an integer file descriptor.
6927 *** procedure: dup->port PORT/FD MODE [NEWFD]
6928 Returns a new port using the new file descriptor. MODE supplies a
6929 mode string for the port (*note open-file: File Ports.).
6931 *** procedure: setenv NAME VALUE
6932 Modifies the environment of the current process, which is also the
6933 default environment inherited by child processes.
6935 If VALUE is `#f', then NAME is removed from the environment.
6936 Otherwise, the string NAME=VALUE is added to the environment,
6937 replacing any existing string with name matching NAME.
6939 The return value is unspecified.
6941 *** procedure: truncate-file OBJ SIZE
6942 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
6943 can be a string containing a file name or an integer file
6944 descriptor or port open for output on the file. The underlying
6945 system calls are `truncate' and `ftruncate'.
6947 The return value is unspecified.
6949 *** procedure: setvbuf PORT MODE [SIZE]
6950 Set the buffering mode for PORT. MODE can be:
6958 block buffered, using a newly allocated buffer of SIZE bytes.
6959 However if SIZE is zero or unspecified, the port will be made
6962 This procedure should not be used after I/O has been performed with
6965 Ports are usually block buffered by default, with a default buffer
6966 size. Procedures e.g., *Note open-file: File Ports, which accept a
6967 mode string allow `0' to be added to request an unbuffered port.
6969 *** procedure: fsync PORT/FD
6970 Copies any unwritten data for the specified output file descriptor
6971 to disk. If PORT/FD is a port, its buffer is flushed before the
6972 underlying file descriptor is fsync'd. The return value is
6975 *** procedure: open-fdes PATH FLAGS [MODES]
6976 Similar to `open' but returns a file descriptor instead of a port.
6978 *** procedure: execle PATH ENV [ARG] ...
6979 Similar to `execl', but the environment of the new process is
6980 specified by ENV, which must be a list of strings as returned by
6981 the `environ' procedure.
6983 This procedure is currently implemented using the `execve' system
6984 call, but we call it `execle' because of its Scheme calling
6987 *** procedure: strerror ERRNO
6988 Returns the Unix error message corresponding to ERRNO, an integer.
6990 *** procedure: primitive-exit [STATUS]
6991 Terminate the current process without unwinding the Scheme stack.
6992 This is would typically be useful after a fork. The exit status
6993 is STATUS if supplied, otherwise zero.
6995 *** procedure: times
6996 Returns an object with information about real and processor time.
6997 The following procedures accept such an object as an argument and
6998 return a selected component:
7001 The current real time, expressed as time units relative to an
7005 The CPU time units used by the calling process.
7008 The CPU time units used by the system on behalf of the
7012 The CPU time units used by terminated child processes of the
7013 calling process, whose status has been collected (e.g., using
7017 Similarly, the CPU times units used by the system on behalf of
7018 terminated child processes.
7020 ** Removed: list-length
7021 ** Removed: list-append, list-append!
7022 ** Removed: list-reverse, list-reverse!
7024 ** array-map renamed to array-map!
7026 ** serial-array-map renamed to serial-array-map!
7028 ** catch doesn't take #f as first argument any longer
7030 Previously, it was possible to pass #f instead of a key to `catch'.
7031 That would cause `catch' to pass a jump buffer object to the procedure
7032 passed as second argument. The procedure could then use this jump
7033 buffer objekt as an argument to throw.
7035 This mechanism has been removed since its utility doesn't motivate the
7036 extra complexity it introduces.
7038 ** The `#/' notation for lists now provokes a warning message from Guile.
7039 This syntax will be removed from Guile in the near future.
7041 To disable the warning message, set the GUILE_HUSH environment
7042 variable to any non-empty value.
7044 ** The newline character now prints as `#\newline', following the
7045 normal Scheme notation, not `#\nl'.
7047 * Changes to the gh_ interface
7049 ** The gh_enter function now takes care of loading the Guile startup files.
7050 gh_enter works by calling scm_boot_guile; see the remarks below.
7052 ** Function: void gh_write (SCM x)
7054 Write the printed representation of the scheme object x to the current
7055 output port. Corresponds to the scheme level `write'.
7057 ** gh_list_length renamed to gh_length.
7059 ** vector handling routines
7061 Several major changes. In particular, gh_vector() now resembles
7062 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
7063 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
7064 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
7065 vector-related gh_ functions have been implemented.
7067 ** pair and list routines
7069 Implemented several of the R4RS pair and list functions that were
7072 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
7074 New function. Converts double arrays back and forth between Scheme
7077 * Changes to the scm_ interface
7079 ** The function scm_boot_guile now takes care of loading the startup files.
7081 Guile's primary initialization function, scm_boot_guile, now takes
7082 care of loading `boot-9.scm', in the `ice-9' module, to initialize
7083 Guile, define the module system, and put together some standard
7084 bindings. It also loads `init.scm', which is intended to hold
7085 site-specific initialization code.
7087 Since Guile cannot operate properly until boot-9.scm is loaded, there
7088 is no reason to separate loading boot-9.scm from Guile's other
7089 initialization processes.
7091 This job used to be done by scm_compile_shell_switches, which didn't
7092 make much sense; in particular, it meant that people using Guile for
7093 non-shell-like applications had to jump through hoops to get Guile
7094 initialized properly.
7096 ** The function scm_compile_shell_switches no longer loads the startup files.
7097 Now, Guile always loads the startup files, whenever it is initialized;
7098 see the notes above for scm_boot_guile and scm_load_startup_files.
7100 ** Function: scm_load_startup_files
7101 This new function takes care of loading Guile's initialization file
7102 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
7103 this is always called by the Guile initialization process, it's
7104 probably not too useful to call this yourself, but it's there anyway.
7106 ** The semantics of smob marking have changed slightly.
7108 The smob marking function (the `mark' member of the scm_smobfuns
7109 structure) is no longer responsible for setting the mark bit on the
7110 smob. The generic smob handling code in the garbage collector will
7111 set this bit. The mark function need only ensure that any other
7112 objects the smob refers to get marked.
7114 Note that this change means that the smob's GC8MARK bit is typically
7115 already set upon entry to the mark function. Thus, marking functions
7116 which look like this:
7119 if (SCM_GC8MARKP (ptr))
7121 SCM_SETGC8MARK (ptr);
7122 ... mark objects to which the smob refers ...
7125 are now incorrect, since they will return early, and fail to mark any
7126 other objects the smob refers to. Some code in the Guile library used
7129 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
7131 If you have implemented your own I/O port type, by writing the
7132 functions required by the scm_ptobfuns and then calling scm_newptob,
7133 you will need to change your functions slightly.
7135 The functions in a scm_ptobfuns structure now expect the port itself
7136 as their argument; they used to expect the `stream' member of the
7137 port's scm_port_table structure. This allows functions in an
7138 scm_ptobfuns structure to easily access the port's cell (and any flags
7139 it its CAR), and the port's scm_port_table structure.
7141 Guile now passes the I/O port itself as the `port' argument in the
7142 following scm_ptobfuns functions:
7144 int (*free) (SCM port);
7145 int (*fputc) (int, SCM port);
7146 int (*fputs) (char *, SCM port);
7147 scm_sizet (*fwrite) SCM_P ((char *ptr,
7151 int (*fflush) (SCM port);
7152 int (*fgetc) (SCM port);
7153 int (*fclose) (SCM port);
7155 The interfaces to the `mark', `print', `equalp', and `fgets' methods
7158 If you have existing code which defines its own port types, it is easy
7159 to convert your code to the new interface; simply apply SCM_STREAM to
7160 the port argument to yield the value you code used to expect.
7162 Note that since both the port and the stream have the same type in the
7163 C code --- they are both SCM values --- the C compiler will not remind
7164 you if you forget to update your scm_ptobfuns functions.
7167 ** Function: int scm_internal_select (int fds,
7171 struct timeval *timeout);
7173 This is a replacement for the `select' function provided by the OS.
7174 It enables I/O blocking and sleeping to happen for one cooperative
7175 thread without blocking other threads. It also avoids busy-loops in
7176 these situations. It is intended that all I/O blocking and sleeping
7177 will finally go through this function. Currently, this function is
7178 only available on systems providing `gettimeofday' and `select'.
7180 ** Function: SCM scm_internal_stack_catch (SCM tag,
7181 scm_catch_body_t body,
7183 scm_catch_handler_t handler,
7186 A new sibling to the other two C level `catch' functions
7187 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
7188 the stack to be saved automatically into the variable `the-last-stack'
7189 (scm_the_last_stack_var) on error. This is necessary if you want to
7190 use advanced error reporting, such as calling scm_display_error and
7191 scm_display_backtrace. (They both take a stack object as argument.)
7193 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
7195 scm_catch_handler_t handler,
7198 Spawns a new thread. It does a job similar to
7199 scm_call_with_new_thread but takes arguments more suitable when
7200 spawning threads from application C code.
7202 ** The hook scm_error_callback has been removed. It was originally
7203 intended as a way for the user to install his own error handler. But
7204 that method works badly since it intervenes between throw and catch,
7205 thereby changing the semantics of expressions like (catch #t ...).
7206 The correct way to do it is to use one of the C level catch functions
7207 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
7209 ** Removed functions:
7211 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
7212 scm_list_reverse, scm_list_reverse_x
7214 ** New macros: SCM_LISTn where n is one of the integers 0-9.
7216 These can be used for pretty list creation from C. The idea is taken
7217 from Erick Gallesio's STk.
7219 ** scm_array_map renamed to scm_array_map_x
7221 ** mbstrings are now removed
7223 This means that the type codes scm_tc7_mb_string and
7224 scm_tc7_mb_substring has been removed.
7226 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
7228 Since we no longer support multi-byte strings, these I/O functions
7229 have been simplified, and renamed. Here are their old names, and
7230 their new names and arguments:
7232 scm_gen_putc -> void scm_putc (int c, SCM port);
7233 scm_gen_puts -> void scm_puts (char *s, SCM port);
7234 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
7235 scm_gen_getc -> void scm_getc (SCM port);
7238 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
7240 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
7242 SCM_TYP7S now masks away the bit which distinguishes substrings from
7245 ** scm_catch_body_t: Backward incompatible change!
7247 Body functions to scm_internal_catch and friends do not any longer
7248 take a second argument. This is because it is no longer possible to
7249 pass a #f arg to catch.
7251 ** Calls to scm_protect_object and scm_unprotect now nest properly.
7253 The function scm_protect_object protects its argument from being freed
7254 by the garbage collector. scm_unprotect_object removes that
7257 These functions now nest properly. That is, for every object O, there
7258 is a counter which scm_protect_object(O) increments and
7259 scm_unprotect_object(O) decrements, if the counter is greater than
7260 zero. Every object's counter is zero when it is first created. If an
7261 object's counter is greater than zero, the garbage collector will not
7262 reclaim its storage.
7264 This allows you to use scm_protect_object in your code without
7265 worrying that some other function you call will call
7266 scm_unprotect_object, and allow it to be freed. Assuming that the
7267 functions you call are well-behaved, and unprotect only those objects
7268 they protect, you can follow the same rule and have confidence that
7269 objects will be freed only at appropriate times.
7272 Changes in Guile 1.2 (released Tuesday, June 24 1997):
7274 * Changes to the distribution
7276 ** Nightly snapshots are now available from ftp.red-bean.com.
7277 The old server, ftp.cyclic.com, has been relinquished to its rightful
7280 Nightly snapshots of the Guile development sources are now available via
7281 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
7283 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
7284 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
7286 ** To run Guile without installing it, the procedure has changed a bit.
7288 If you used a separate build directory to compile Guile, you'll need
7289 to include the build directory in SCHEME_LOAD_PATH, as well as the
7290 source directory. See the `INSTALL' file for examples.
7292 * Changes to the procedure for linking libguile with your programs
7294 ** The standard Guile load path for Scheme code now includes
7295 $(datadir)/guile (usually /usr/local/share/guile). This means that
7296 you can install your own Scheme files there, and Guile will find them.
7297 (Previous versions of Guile only checked a directory whose name
7298 contained the Guile version number, so you had to re-install or move
7299 your Scheme sources each time you installed a fresh version of Guile.)
7301 The load path also includes $(datadir)/guile/site; we recommend
7302 putting individual Scheme files there. If you want to install a
7303 package with multiple source files, create a directory for them under
7306 ** Guile 1.2 will now use the Rx regular expression library, if it is
7307 installed on your system. When you are linking libguile into your own
7308 programs, this means you will have to link against -lguile, -lqt (if
7309 you configured Guile with thread support), and -lrx.
7311 If you are using autoconf to generate configuration scripts for your
7312 application, the following lines should suffice to add the appropriate
7313 libraries to your link command:
7315 ### Find Rx, quickthreads and libguile.
7316 AC_CHECK_LIB(rx, main)
7317 AC_CHECK_LIB(qt, main)
7318 AC_CHECK_LIB(guile, scm_shell)
7320 The Guile 1.2 distribution does not contain sources for the Rx
7321 library, as Guile 1.0 did. If you want to use Rx, you'll need to
7322 retrieve it from a GNU FTP site and install it separately.
7324 * Changes to Scheme functions and syntax
7326 ** The dynamic linking features of Guile are now enabled by default.
7327 You can disable them by giving the `--disable-dynamic-linking' option
7330 (dynamic-link FILENAME)
7332 Find the object file denoted by FILENAME (a string) and link it
7333 into the running Guile application. When everything works out,
7334 return a Scheme object suitable for representing the linked object
7335 file. Otherwise an error is thrown. How object files are
7336 searched is system dependent.
7338 (dynamic-object? VAL)
7340 Determine whether VAL represents a dynamically linked object file.
7342 (dynamic-unlink DYNOBJ)
7344 Unlink the indicated object file from the application. DYNOBJ
7345 should be one of the values returned by `dynamic-link'.
7347 (dynamic-func FUNCTION DYNOBJ)
7349 Search the C function indicated by FUNCTION (a string or symbol)
7350 in DYNOBJ and return some Scheme object that can later be used
7351 with `dynamic-call' to actually call this function. Right now,
7352 these Scheme objects are formed by casting the address of the
7353 function to `long' and converting this number to its Scheme
7356 (dynamic-call FUNCTION DYNOBJ)
7358 Call the C function indicated by FUNCTION and DYNOBJ. The
7359 function is passed no arguments and its return value is ignored.
7360 When FUNCTION is something returned by `dynamic-func', call that
7361 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
7362 etc.), look it up in DYNOBJ; this is equivalent to
7364 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
7366 Interrupts are deferred while the C function is executing (with
7367 SCM_DEFER_INTS/SCM_ALLOW_INTS).
7369 (dynamic-args-call FUNCTION DYNOBJ ARGS)
7371 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
7372 some arguments and return its return value. The C function is
7373 expected to take two arguments and return an `int', just like
7376 int c_func (int argc, char **argv);
7378 ARGS must be a list of strings and is converted into an array of
7379 `char *'. The array is passed in ARGV and its size in ARGC. The
7380 return value is converted to a Scheme number and returned from the
7381 call to `dynamic-args-call'.
7383 When dynamic linking is disabled or not supported on your system,
7384 the above functions throw errors, but they are still available.
7386 Here is a small example that works on GNU/Linux:
7388 (define libc-obj (dynamic-link "libc.so"))
7389 (dynamic-args-call 'rand libc-obj '())
7391 See the file `libguile/DYNAMIC-LINKING' for additional comments.
7393 ** The #/ syntax for module names is depreciated, and will be removed
7394 in a future version of Guile. Instead of
7402 The latter syntax is more consistent with existing Lisp practice.
7404 ** Guile now does fancier printing of structures. Structures are the
7405 underlying implementation for records, which in turn are used to
7406 implement modules, so all of these object now print differently and in
7407 a more informative way.
7409 The Scheme printer will examine the builtin variable *struct-printer*
7410 whenever it needs to print a structure object. When this variable is
7411 not `#f' it is deemed to be a procedure and will be applied to the
7412 structure object and the output port. When *struct-printer* is `#f'
7413 or the procedure return `#f' the structure object will be printed in
7414 the boring #<struct 80458270> form.
7416 This hook is used by some routines in ice-9/boot-9.scm to implement
7417 type specific printing routines. Please read the comments there about
7420 One of the more specific uses of structs are records. The printing
7421 procedure that could be passed to MAKE-RECORD-TYPE is now actually
7422 called. It should behave like a *struct-printer* procedure (described
7425 ** Guile now supports a new R4RS-compliant syntax for keywords. A
7426 token of the form #:NAME, where NAME has the same syntax as a Scheme
7427 symbol, is the external representation of the keyword named NAME.
7428 Keyword objects print using this syntax as well, so values containing
7429 keyword objects can be read back into Guile. When used in an
7430 expression, keywords are self-quoting objects.
7432 Guile suports this read syntax, and uses this print syntax, regardless
7433 of the current setting of the `keyword' read option. The `keyword'
7434 read option only controls whether Guile recognizes the `:NAME' syntax,
7435 which is incompatible with R4RS. (R4RS says such token represent
7438 ** Guile has regular expression support again. Guile 1.0 included
7439 functions for matching regular expressions, based on the Rx library.
7440 In Guile 1.1, the Guile/Rx interface was removed to simplify the
7441 distribution, and thus Guile had no regular expression support. Guile
7442 1.2 again supports the most commonly used functions, and supports all
7443 of SCSH's regular expression functions.
7445 If your system does not include a POSIX regular expression library,
7446 and you have not linked Guile with a third-party regexp library such as
7447 Rx, these functions will not be available. You can tell whether your
7448 Guile installation includes regular expression support by checking
7449 whether the `*features*' list includes the `regex' symbol.
7451 *** regexp functions
7453 By default, Guile supports POSIX extended regular expressions. That
7454 means that the characters `(', `)', `+' and `?' are special, and must
7455 be escaped if you wish to match the literal characters.
7457 This regular expression interface was modeled after that implemented
7458 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
7459 with SCSH regular expressions.
7461 **** Function: string-match PATTERN STR [START]
7462 Compile the string PATTERN into a regular expression and compare
7463 it with STR. The optional numeric argument START specifies the
7464 position of STR at which to begin matching.
7466 `string-match' returns a "match structure" which describes what,
7467 if anything, was matched by the regular expression. *Note Match
7468 Structures::. If STR does not match PATTERN at all,
7469 `string-match' returns `#f'.
7471 Each time `string-match' is called, it must compile its PATTERN
7472 argument into a regular expression structure. This operation is
7473 expensive, which makes `string-match' inefficient if the same regular
7474 expression is used several times (for example, in a loop). For better
7475 performance, you can compile a regular expression in advance and then
7476 match strings against the compiled regexp.
7478 **** Function: make-regexp STR [FLAGS]
7479 Compile the regular expression described by STR, and return the
7480 compiled regexp structure. If STR does not describe a legal
7481 regular expression, `make-regexp' throws a
7482 `regular-expression-syntax' error.
7484 FLAGS may be the bitwise-or of one or more of the following:
7486 **** Constant: regexp/extended
7487 Use POSIX Extended Regular Expression syntax when interpreting
7488 STR. If not set, POSIX Basic Regular Expression syntax is used.
7489 If the FLAGS argument is omitted, we assume regexp/extended.
7491 **** Constant: regexp/icase
7492 Do not differentiate case. Subsequent searches using the
7493 returned regular expression will be case insensitive.
7495 **** Constant: regexp/newline
7496 Match-any-character operators don't match a newline.
7498 A non-matching list ([^...]) not containing a newline matches a
7501 Match-beginning-of-line operator (^) matches the empty string
7502 immediately after a newline, regardless of whether the FLAGS
7503 passed to regexp-exec contain regexp/notbol.
7505 Match-end-of-line operator ($) matches the empty string
7506 immediately before a newline, regardless of whether the FLAGS
7507 passed to regexp-exec contain regexp/noteol.
7509 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
7510 Match the compiled regular expression REGEXP against `str'. If
7511 the optional integer START argument is provided, begin matching
7512 from that position in the string. Return a match structure
7513 describing the results of the match, or `#f' if no match could be
7516 FLAGS may be the bitwise-or of one or more of the following:
7518 **** Constant: regexp/notbol
7519 The match-beginning-of-line operator always fails to match (but
7520 see the compilation flag regexp/newline above) This flag may be
7521 used when different portions of a string are passed to
7522 regexp-exec and the beginning of the string should not be
7523 interpreted as the beginning of the line.
7525 **** Constant: regexp/noteol
7526 The match-end-of-line operator always fails to match (but see the
7527 compilation flag regexp/newline above)
7529 **** Function: regexp? OBJ
7530 Return `#t' if OBJ is a compiled regular expression, or `#f'
7533 Regular expressions are commonly used to find patterns in one string
7534 and replace them with the contents of another string.
7536 **** Function: regexp-substitute PORT MATCH [ITEM...]
7537 Write to the output port PORT selected contents of the match
7538 structure MATCH. Each ITEM specifies what should be written, and
7539 may be one of the following arguments:
7541 * A string. String arguments are written out verbatim.
7543 * An integer. The submatch with that number is written.
7545 * The symbol `pre'. The portion of the matched string preceding
7546 the regexp match is written.
7548 * The symbol `post'. The portion of the matched string
7549 following the regexp match is written.
7551 PORT may be `#f', in which case nothing is written; instead,
7552 `regexp-substitute' constructs a string from the specified ITEMs
7555 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
7556 Similar to `regexp-substitute', but can be used to perform global
7557 substitutions on STR. Instead of taking a match structure as an
7558 argument, `regexp-substitute/global' takes two string arguments: a
7559 REGEXP string describing a regular expression, and a TARGET string
7560 which should be matched against this regular expression.
7562 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
7565 * A function may be supplied. When this function is called, it
7566 will be passed one argument: a match structure for a given
7567 regular expression match. It should return a string to be
7568 written out to PORT.
7570 * The `post' symbol causes `regexp-substitute/global' to recurse
7571 on the unmatched portion of STR. This *must* be supplied in
7572 order to perform global search-and-replace on STR; if it is
7573 not present among the ITEMs, then `regexp-substitute/global'
7574 will return after processing a single match.
7576 *** Match Structures
7578 A "match structure" is the object returned by `string-match' and
7579 `regexp-exec'. It describes which portion of a string, if any, matched
7580 the given regular expression. Match structures include: a reference to
7581 the string that was checked for matches; the starting and ending
7582 positions of the regexp match; and, if the regexp included any
7583 parenthesized subexpressions, the starting and ending positions of each
7586 In each of the regexp match functions described below, the `match'
7587 argument must be a match structure returned by a previous call to
7588 `string-match' or `regexp-exec'. Most of these functions return some
7589 information about the original target string that was matched against a
7590 regular expression; we will call that string TARGET for easy reference.
7592 **** Function: regexp-match? OBJ
7593 Return `#t' if OBJ is a match structure returned by a previous
7594 call to `regexp-exec', or `#f' otherwise.
7596 **** Function: match:substring MATCH [N]
7597 Return the portion of TARGET matched by subexpression number N.
7598 Submatch 0 (the default) represents the entire regexp match. If
7599 the regular expression as a whole matched, but the subexpression
7600 number N did not match, return `#f'.
7602 **** Function: match:start MATCH [N]
7603 Return the starting position of submatch number N.
7605 **** Function: match:end MATCH [N]
7606 Return the ending position of submatch number N.
7608 **** Function: match:prefix MATCH
7609 Return the unmatched portion of TARGET preceding the regexp match.
7611 **** Function: match:suffix MATCH
7612 Return the unmatched portion of TARGET following the regexp match.
7614 **** Function: match:count MATCH
7615 Return the number of parenthesized subexpressions from MATCH.
7616 Note that the entire regular expression match itself counts as a
7617 subexpression, and failed submatches are included in the count.
7619 **** Function: match:string MATCH
7620 Return the original TARGET string.
7622 *** Backslash Escapes
7624 Sometimes you will want a regexp to match characters like `*' or `$'
7625 exactly. For example, to check whether a particular string represents
7626 a menu entry from an Info node, it would be useful to match it against
7627 a regexp like `^* [^:]*::'. However, this won't work; because the
7628 asterisk is a metacharacter, it won't match the `*' at the beginning of
7629 the string. In this case, we want to make the first asterisk un-magic.
7631 You can do this by preceding the metacharacter with a backslash
7632 character `\'. (This is also called "quoting" the metacharacter, and
7633 is known as a "backslash escape".) When Guile sees a backslash in a
7634 regular expression, it considers the following glyph to be an ordinary
7635 character, no matter what special meaning it would ordinarily have.
7636 Therefore, we can make the above example work by changing the regexp to
7637 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
7638 to match only a single asterisk in the target string.
7640 Since the backslash is itself a metacharacter, you may force a
7641 regexp to match a backslash in the target string by preceding the
7642 backslash with itself. For example, to find variable references in a
7643 TeX program, you might want to find occurrences of the string `\let\'
7644 followed by any number of alphabetic characters. The regular expression
7645 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
7646 each match a single backslash in the target string.
7648 **** Function: regexp-quote STR
7649 Quote each special character found in STR with a backslash, and
7650 return the resulting string.
7652 *Very important:* Using backslash escapes in Guile source code (as
7653 in Emacs Lisp or C) can be tricky, because the backslash character has
7654 special meaning for the Guile reader. For example, if Guile encounters
7655 the character sequence `\n' in the middle of a string while processing
7656 Scheme code, it replaces those characters with a newline character.
7657 Similarly, the character sequence `\t' is replaced by a horizontal tab.
7658 Several of these "escape sequences" are processed by the Guile reader
7659 before your code is executed. Unrecognized escape sequences are
7660 ignored: if the characters `\*' appear in a string, they will be
7661 translated to the single character `*'.
7663 This translation is obviously undesirable for regular expressions,
7664 since we want to be able to include backslashes in a string in order to
7665 escape regexp metacharacters. Therefore, to make sure that a backslash
7666 is preserved in a string in your Guile program, you must use *two*
7667 consecutive backslashes:
7669 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
7671 The string in this example is preprocessed by the Guile reader before
7672 any code is executed. The resulting argument to `make-regexp' is the
7673 string `^\* [^:]*', which is what we really want.
7675 This also means that in order to write a regular expression that
7676 matches a single backslash character, the regular expression string in
7677 the source code must include *four* backslashes. Each consecutive pair
7678 of backslashes gets translated by the Guile reader to a single
7679 backslash, and the resulting double-backslash is interpreted by the
7680 regexp engine as matching a single backslash character. Hence:
7682 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
7684 The reason for the unwieldiness of this syntax is historical. Both
7685 regular expression pattern matchers and Unix string processing systems
7686 have traditionally used backslashes with the special meanings described
7687 above. The POSIX regular expression specification and ANSI C standard
7688 both require these semantics. Attempting to abandon either convention
7689 would cause other kinds of compatibility problems, possibly more severe
7690 ones. Therefore, without extending the Scheme reader to support
7691 strings with different quoting conventions (an ungainly and confusing
7692 extension when implemented in other languages), we must adhere to this
7693 cumbersome escape syntax.
7695 * Changes to the gh_ interface
7697 * Changes to the scm_ interface
7699 * Changes to system call interfaces:
7701 ** The value returned by `raise' is now unspecified. It throws an exception
7704 *** A new procedure `sigaction' can be used to install signal handlers
7706 (sigaction signum [action] [flags])
7708 signum is the signal number, which can be specified using the value
7711 If action is omitted, sigaction returns a pair: the CAR is the current
7712 signal hander, which will be either an integer with the value SIG_DFL
7713 (default action) or SIG_IGN (ignore), or the Scheme procedure which
7714 handles the signal, or #f if a non-Scheme procedure handles the
7715 signal. The CDR contains the current sigaction flags for the handler.
7717 If action is provided, it is installed as the new handler for signum.
7718 action can be a Scheme procedure taking one argument, or the value of
7719 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
7720 whatever signal handler was installed before sigaction was first used.
7721 Flags can optionally be specified for the new handler (SA_RESTART is
7722 always used if the system provides it, so need not be specified.) The
7723 return value is a pair with information about the old handler as
7726 This interface does not provide access to the "signal blocking"
7727 facility. Maybe this is not needed, since the thread support may
7728 provide solutions to the problem of consistent access to data
7731 *** A new procedure `flush-all-ports' is equivalent to running
7732 `force-output' on every port open for output.
7734 ** Guile now provides information on how it was built, via the new
7735 global variable, %guile-build-info. This variable records the values
7736 of the standard GNU makefile directory variables as an assocation
7737 list, mapping variable names (symbols) onto directory paths (strings).
7738 For example, to find out where the Guile link libraries were
7739 installed, you can say:
7741 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
7744 * Changes to the scm_ interface
7746 ** The new function scm_handle_by_message_noexit is just like the
7747 existing scm_handle_by_message function, except that it doesn't call
7748 exit to terminate the process. Instead, it prints a message and just
7749 returns #f. This might be a more appropriate catch-all handler for
7750 new dynamic roots and threads.
7753 Changes in Guile 1.1 (released Friday, May 16 1997):
7755 * Changes to the distribution.
7757 The Guile 1.0 distribution has been split up into several smaller
7759 guile-core --- the Guile interpreter itself.
7760 guile-tcltk --- the interface between the Guile interpreter and
7761 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
7762 is a toolkit for building graphical user interfaces.
7763 guile-rgx-ctax --- the interface between Guile and the Rx regular
7764 expression matcher, and the translator for the Ctax
7765 programming language. These are packaged together because the
7766 Ctax translator uses Rx to parse Ctax source code.
7768 This NEWS file describes the changes made to guile-core since the 1.0
7771 We no longer distribute the documentation, since it was either out of
7772 date, or incomplete. As soon as we have current documentation, we
7777 * Changes to the stand-alone interpreter
7779 ** guile now accepts command-line arguments compatible with SCSH, Olin
7780 Shivers' Scheme Shell.
7782 In general, arguments are evaluated from left to right, but there are
7783 exceptions. The following switches stop argument processing, and
7784 stash all remaining command-line arguments as the value returned by
7785 the (command-line) function.
7786 -s SCRIPT load Scheme source code from FILE, and exit
7787 -c EXPR evalute Scheme expression EXPR, and exit
7788 -- stop scanning arguments; run interactively
7790 The switches below are processed as they are encountered.
7791 -l FILE load Scheme source code from FILE
7792 -e FUNCTION after reading script, apply FUNCTION to
7793 command line arguments
7794 -ds do -s script at this point
7795 --emacs enable Emacs protocol (experimental)
7796 -h, --help display this help and exit
7797 -v, --version display version information and exit
7798 \ read arguments from following script lines
7800 So, for example, here is a Guile script named `ekko' (thanks, Olin)
7801 which re-implements the traditional "echo" command:
7803 #!/usr/local/bin/guile -s
7806 (map (lambda (arg) (display arg) (display " "))
7810 (main (command-line))
7812 Suppose we invoke this script as follows:
7814 ekko a speckled gecko
7816 Through the magic of Unix script processing (triggered by the `#!'
7817 token at the top of the file), /usr/local/bin/guile receives the
7818 following list of command-line arguments:
7820 ("-s" "./ekko" "a" "speckled" "gecko")
7822 Unix inserts the name of the script after the argument specified on
7823 the first line of the file (in this case, "-s"), and then follows that
7824 with the arguments given to the script. Guile loads the script, which
7825 defines the `main' function, and then applies it to the list of
7826 remaining command-line arguments, ("a" "speckled" "gecko").
7828 In Unix, the first line of a script file must take the following form:
7830 #!INTERPRETER ARGUMENT
7832 where INTERPRETER is the absolute filename of the interpreter
7833 executable, and ARGUMENT is a single command-line argument to pass to
7836 You may only pass one argument to the interpreter, and its length is
7837 limited. These restrictions can be annoying to work around, so Guile
7838 provides a general mechanism (borrowed from, and compatible with,
7839 SCSH) for circumventing them.
7841 If the ARGUMENT in a Guile script is a single backslash character,
7842 `\', Guile will open the script file, parse arguments from its second
7843 and subsequent lines, and replace the `\' with them. So, for example,
7844 here is another implementation of the `ekko' script:
7846 #!/usr/local/bin/guile \
7850 (for-each (lambda (arg) (display arg) (display " "))
7854 If the user invokes this script as follows:
7856 ekko a speckled gecko
7858 Unix expands this into
7860 /usr/local/bin/guile \ ekko a speckled gecko
7862 When Guile sees the `\' argument, it replaces it with the arguments
7863 read from the second line of the script, producing:
7865 /usr/local/bin/guile -e main -s ekko a speckled gecko
7867 This tells Guile to load the `ekko' script, and apply the function
7868 `main' to the argument list ("a" "speckled" "gecko").
7870 Here is how Guile parses the command-line arguments:
7871 - Each space character terminates an argument. This means that two
7872 spaces in a row introduce an empty-string argument.
7873 - The tab character is not permitted (unless you quote it with the
7874 backslash character, as described below), to avoid confusion.
7875 - The newline character terminates the sequence of arguments, and will
7876 also terminate a final non-empty argument. (However, a newline
7877 following a space will not introduce a final empty-string argument;
7878 it only terminates the argument list.)
7879 - The backslash character is the escape character. It escapes
7880 backslash, space, tab, and newline. The ANSI C escape sequences
7881 like \n and \t are also supported. These produce argument
7882 constituents; the two-character combination \n doesn't act like a
7883 terminating newline. The escape sequence \NNN for exactly three
7884 octal digits reads as the character whose ASCII code is NNN. As
7885 above, characters produced this way are argument constituents.
7886 Backslash followed by other characters is not allowed.
7888 * Changes to the procedure for linking libguile with your programs
7890 ** Guile now builds and installs a shared guile library, if your
7891 system support shared libraries. (It still builds a static library on
7892 all systems.) Guile automatically detects whether your system
7893 supports shared libraries. To prevent Guile from buildisg shared
7894 libraries, pass the `--disable-shared' flag to the configure script.
7896 Guile takes longer to compile when it builds shared libraries, because
7897 it must compile every file twice --- once to produce position-
7898 independent object code, and once to produce normal object code.
7900 ** The libthreads library has been merged into libguile.
7902 To link a program against Guile, you now need only link against
7903 -lguile and -lqt; -lthreads is no longer needed. If you are using
7904 autoconf to generate configuration scripts for your application, the
7905 following lines should suffice to add the appropriate libraries to
7908 ### Find quickthreads and libguile.
7909 AC_CHECK_LIB(qt, main)
7910 AC_CHECK_LIB(guile, scm_shell)
7912 * Changes to Scheme functions
7914 ** Guile Scheme's special syntax for keyword objects is now optional,
7915 and disabled by default.
7917 The syntax variation from R4RS made it difficult to port some
7918 interesting packages to Guile. The routines which accepted keyword
7919 arguments (mostly in the module system) have been modified to also
7920 accept symbols whose names begin with `:'.
7922 To change the keyword syntax, you must first import the (ice-9 debug)
7924 (use-modules (ice-9 debug))
7926 Then you can enable the keyword syntax as follows:
7927 (read-set! keywords 'prefix)
7929 To disable keyword syntax, do this:
7930 (read-set! keywords #f)
7932 ** Many more primitive functions accept shared substrings as
7933 arguments. In the past, these functions required normal, mutable
7934 strings as arguments, although they never made use of this
7937 ** The uniform array functions now operate on byte vectors. These
7938 functions are `array-fill!', `serial-array-copy!', `array-copy!',
7939 `serial-array-map', `array-map', `array-for-each', and
7942 ** The new functions `trace' and `untrace' implement simple debugging
7943 support for Scheme functions.
7945 The `trace' function accepts any number of procedures as arguments,
7946 and tells the Guile interpreter to display each procedure's name and
7947 arguments each time the procedure is invoked. When invoked with no
7948 arguments, `trace' returns the list of procedures currently being
7951 The `untrace' function accepts any number of procedures as arguments,
7952 and tells the Guile interpreter not to trace them any more. When
7953 invoked with no arguments, `untrace' untraces all curretly traced
7956 The tracing in Guile has an advantage over most other systems: we
7957 don't create new procedure objects, but mark the procedure objects
7958 themselves. This means that anonymous and internal procedures can be
7961 ** The function `assert-repl-prompt' has been renamed to
7962 `set-repl-prompt!'. It takes one argument, PROMPT.
7963 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
7964 - If PROMPT is a string, we use it as a prompt.
7965 - If PROMPT is a procedure accepting no arguments, we call it, and
7966 display the result as a prompt.
7967 - Otherwise, we display "> ".
7969 ** The new function `eval-string' reads Scheme expressions from a
7970 string and evaluates them, returning the value of the last expression
7971 in the string. If the string contains no expressions, it returns an
7974 ** The new function `thunk?' returns true iff its argument is a
7975 procedure of zero arguments.
7977 ** `defined?' is now a builtin function, instead of syntax. This
7978 means that its argument should be quoted. It returns #t iff its
7979 argument is bound in the current module.
7981 ** The new syntax `use-modules' allows you to add new modules to your
7982 environment without re-typing a complete `define-module' form. It
7983 accepts any number of module names as arguments, and imports their
7984 public bindings into the current module.
7986 ** The new function (module-defined? NAME MODULE) returns true iff
7987 NAME, a symbol, is defined in MODULE, a module object.
7989 ** The new function `builtin-bindings' creates and returns a hash
7990 table containing copies of all the root module's bindings.
7992 ** The new function `builtin-weak-bindings' does the same as
7993 `builtin-bindings', but creates a doubly-weak hash table.
7995 ** The `equal?' function now considers variable objects to be
7996 equivalent if they have the same name and the same value.
7998 ** The new function `command-line' returns the command-line arguments
7999 given to Guile, as a list of strings.
8001 When using guile as a script interpreter, `command-line' returns the
8002 script's arguments; those processed by the interpreter (like `-s' or
8003 `-c') are omitted. (In other words, you get the normal, expected
8004 behavior.) Any application that uses scm_shell to process its
8005 command-line arguments gets this behavior as well.
8007 ** The new function `load-user-init' looks for a file called `.guile'
8008 in the user's home directory, and loads it if it exists. This is
8009 mostly for use by the code generated by scm_compile_shell_switches,
8010 but we thought it might also be useful in other circumstances.
8012 ** The new function `log10' returns the base-10 logarithm of its
8015 ** Changes to I/O functions
8017 *** The functions `read', `primitive-load', `read-and-eval!', and
8018 `primitive-load-path' no longer take optional arguments controlling
8019 case insensitivity and a `#' parser.
8021 Case sensitivity is now controlled by a read option called
8022 `case-insensitive'. The user can add new `#' syntaxes with the
8023 `read-hash-extend' function (see below).
8025 *** The new function `read-hash-extend' allows the user to change the
8026 syntax of Guile Scheme in a somewhat controlled way.
8028 (read-hash-extend CHAR PROC)
8029 When parsing S-expressions, if we read a `#' character followed by
8030 the character CHAR, use PROC to parse an object from the stream.
8031 If PROC is #f, remove any parsing procedure registered for CHAR.
8033 The reader applies PROC to two arguments: CHAR and an input port.
8035 *** The new functions read-delimited and read-delimited! provide a
8036 general mechanism for doing delimited input on streams.
8038 (read-delimited DELIMS [PORT HANDLE-DELIM])
8039 Read until we encounter one of the characters in DELIMS (a string),
8040 or end-of-file. PORT is the input port to read from; it defaults to
8041 the current input port. The HANDLE-DELIM parameter determines how
8042 the terminating character is handled; it should be one of the
8045 'trim omit delimiter from result
8046 'peek leave delimiter character in input stream
8047 'concat append delimiter character to returned value
8048 'split return a pair: (RESULT . TERMINATOR)
8050 HANDLE-DELIM defaults to 'peek.
8052 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
8053 A side-effecting variant of `read-delimited'.
8055 The data is written into the string BUF at the indices in the
8056 half-open interval [START, END); the default interval is the whole
8057 string: START = 0 and END = (string-length BUF). The values of
8058 START and END must specify a well-defined interval in BUF, i.e.
8059 0 <= START <= END <= (string-length BUF).
8061 It returns NBYTES, the number of bytes read. If the buffer filled
8062 up without a delimiter character being found, it returns #f. If the
8063 port is at EOF when the read starts, it returns the EOF object.
8065 If an integer is returned (i.e., the read is successfully terminated
8066 by reading a delimiter character), then the HANDLE-DELIM parameter
8067 determines how to handle the terminating character. It is described
8068 above, and defaults to 'peek.
8070 (The descriptions of these functions were borrowed from the SCSH
8071 manual, by Olin Shivers and Brian Carlstrom.)
8073 *** The `%read-delimited!' function is the primitive used to implement
8074 `read-delimited' and `read-delimited!'.
8076 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
8078 This returns a pair of values: (TERMINATOR . NUM-READ).
8079 - TERMINATOR describes why the read was terminated. If it is a
8080 character or the eof object, then that is the value that terminated
8081 the read. If it is #f, the function filled the buffer without finding
8082 a delimiting character.
8083 - NUM-READ is the number of characters read into BUF.
8085 If the read is successfully terminated by reading a delimiter
8086 character, then the gobble? parameter determines what to do with the
8087 terminating character. If true, the character is removed from the
8088 input stream; if false, the character is left in the input stream
8089 where a subsequent read operation will retrieve it. In either case,
8090 the character is also the first value returned by the procedure call.
8092 (The descriptions of this function was borrowed from the SCSH manual,
8093 by Olin Shivers and Brian Carlstrom.)
8095 *** The `read-line' and `read-line!' functions have changed; they now
8096 trim the terminator by default; previously they appended it to the
8097 returned string. For the old behavior, use (read-line PORT 'concat).
8099 *** The functions `uniform-array-read!' and `uniform-array-write!' now
8100 take new optional START and END arguments, specifying the region of
8101 the array to read and write.
8103 *** The `ungetc-char-ready?' function has been removed. We feel it's
8104 inappropriate for an interface to expose implementation details this
8107 ** Changes to the Unix library and system call interface
8109 *** The new fcntl function provides access to the Unix `fcntl' system
8112 (fcntl PORT COMMAND VALUE)
8113 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
8114 Values for COMMAND are:
8116 F_DUPFD duplicate a file descriptor
8117 F_GETFD read the descriptor's close-on-exec flag
8118 F_SETFD set the descriptor's close-on-exec flag to VALUE
8119 F_GETFL read the descriptor's flags, as set on open
8120 F_SETFL set the descriptor's flags, as set on open to VALUE
8121 F_GETOWN return the process ID of a socket's owner, for SIGIO
8122 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
8123 FD_CLOEXEC not sure what this is
8125 For details, see the documentation for the fcntl system call.
8127 *** The arguments to `select' have changed, for compatibility with
8128 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
8129 expected behavior. The MILLISECONDS parameter has been changed to
8130 MICROSECONDS, to more closely resemble the underlying system call.
8131 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
8132 corresponding return set will be the same.
8134 *** The arguments to the `mknod' system call have changed. They are
8137 (mknod PATH TYPE PERMS DEV)
8138 Create a new file (`node') in the file system. PATH is the name of
8139 the file to create. TYPE is the kind of file to create; it should
8140 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
8141 permission bits to give the newly created file. If TYPE is
8142 'block-special or 'char-special, DEV specifies which device the
8143 special file refers to; its interpretation depends on the kind of
8144 special file being created.
8146 *** The `fork' function has been renamed to `primitive-fork', to avoid
8147 clashing with various SCSH forks.
8149 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
8150 and `recvfrom!'. They no longer accept a size for a second argument;
8151 you must pass a string to hold the received value. They no longer
8152 return the buffer. Instead, `recv' returns the length of the message
8153 received, and `recvfrom' returns a pair containing the packet's length
8154 and originating address.
8156 *** The file descriptor datatype has been removed, as have the
8157 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
8158 We plan to replace these functions with a SCSH-compatible interface.
8160 *** The `create' function has been removed; it's just a special case
8163 *** There are new functions to break down process termination status
8164 values. In the descriptions below, STATUS is a value returned by
8167 (status:exit-val STATUS)
8168 If the child process exited normally, this function returns the exit
8169 code for the child process (i.e., the value passed to exit, or
8170 returned from main). If the child process did not exit normally,
8171 this function returns #f.
8173 (status:stop-sig STATUS)
8174 If the child process was suspended by a signal, this function
8175 returns the signal that suspended the child. Otherwise, it returns
8178 (status:term-sig STATUS)
8179 If the child process terminated abnormally, this function returns
8180 the signal that terminated the child. Otherwise, this function
8183 POSIX promises that exactly one of these functions will return true on
8184 a valid STATUS value.
8186 These functions are compatible with SCSH.
8188 *** There are new accessors and setters for the broken-out time vectors
8189 returned by `localtime', `gmtime', and that ilk. They are:
8191 Component Accessor Setter
8192 ========================= ============ ============
8193 seconds tm:sec set-tm:sec
8194 minutes tm:min set-tm:min
8195 hours tm:hour set-tm:hour
8196 day of the month tm:mday set-tm:mday
8197 month tm:mon set-tm:mon
8198 year tm:year set-tm:year
8199 day of the week tm:wday set-tm:wday
8200 day in the year tm:yday set-tm:yday
8201 daylight saving time tm:isdst set-tm:isdst
8202 GMT offset, seconds tm:gmtoff set-tm:gmtoff
8203 name of time zone tm:zone set-tm:zone
8205 *** There are new accessors for the vectors returned by `uname',
8206 describing the host system:
8209 ============================================== ================
8210 name of the operating system implementation utsname:sysname
8211 network name of this machine utsname:nodename
8212 release level of the operating system utsname:release
8213 version level of the operating system utsname:version
8214 machine hardware platform utsname:machine
8216 *** There are new accessors for the vectors returned by `getpw',
8217 `getpwnam', `getpwuid', and `getpwent', describing entries from the
8218 system's user database:
8221 ====================== =================
8222 user name passwd:name
8223 user password passwd:passwd
8226 real name passwd:gecos
8227 home directory passwd:dir
8228 shell program passwd:shell
8230 *** There are new accessors for the vectors returned by `getgr',
8231 `getgrnam', `getgrgid', and `getgrent', describing entries from the
8232 system's group database:
8235 ======================= ============
8236 group name group:name
8237 group password group:passwd
8239 group members group:mem
8241 *** There are new accessors for the vectors returned by `gethost',
8242 `gethostbyaddr', `gethostbyname', and `gethostent', describing
8246 ========================= ===============
8247 official name of host hostent:name
8248 alias list hostent:aliases
8249 host address type hostent:addrtype
8250 length of address hostent:length
8251 list of addresses hostent:addr-list
8253 *** There are new accessors for the vectors returned by `getnet',
8254 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
8258 ========================= ===============
8259 official name of net netent:name
8260 alias list netent:aliases
8261 net number type netent:addrtype
8262 net number netent:net
8264 *** There are new accessors for the vectors returned by `getproto',
8265 `getprotobyname', `getprotobynumber', and `getprotoent', describing
8269 ========================= ===============
8270 official protocol name protoent:name
8271 alias list protoent:aliases
8272 protocol number protoent:proto
8274 *** There are new accessors for the vectors returned by `getserv',
8275 `getservbyname', `getservbyport', and `getservent', describing
8279 ========================= ===============
8280 official service name servent:name
8281 alias list servent:aliases
8282 port number servent:port
8283 protocol to use servent:proto
8285 *** There are new accessors for the sockaddr structures returned by
8286 `accept', `getsockname', `getpeername', `recvfrom!':
8289 ======================================== ===============
8290 address format (`family') sockaddr:fam
8291 path, for file domain addresses sockaddr:path
8292 address, for internet domain addresses sockaddr:addr
8293 TCP or UDP port, for internet sockaddr:port
8295 *** The `getpwent', `getgrent', `gethostent', `getnetent',
8296 `getprotoent', and `getservent' functions now return #f at the end of
8297 the user database. (They used to throw an exception.)
8299 Note that calling MUMBLEent function is equivalent to calling the
8300 corresponding MUMBLE function with no arguments.
8302 *** The `setpwent', `setgrent', `sethostent', `setnetent',
8303 `setprotoent', and `setservent' routines now take no arguments.
8305 *** The `gethost', `getproto', `getnet', and `getserv' functions now
8306 provide more useful information when they throw an exception.
8308 *** The `lnaof' function has been renamed to `inet-lnaof'.
8310 *** Guile now claims to have the `current-time' feature.
8312 *** The `mktime' function now takes an optional second argument ZONE,
8313 giving the time zone to use for the conversion. ZONE should be a
8314 string, in the same format as expected for the "TZ" environment variable.
8316 *** The `strptime' function now returns a pair (TIME . COUNT), where
8317 TIME is the parsed time as a vector, and COUNT is the number of
8318 characters from the string left unparsed. This function used to
8319 return the remaining characters as a string.
8321 *** The `gettimeofday' function has replaced the old `time+ticks' function.
8322 The return value is now (SECONDS . MICROSECONDS); the fractional
8323 component is no longer expressed in "ticks".
8325 *** The `ticks/sec' constant has been removed, in light of the above change.
8327 * Changes to the gh_ interface
8329 ** gh_eval_str() now returns an SCM object which is the result of the
8332 ** gh_scm2str() now copies the Scheme data to a caller-provided C
8335 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
8336 and returns the array
8338 ** gh_scm2str0() is gone: there is no need to distinguish
8339 null-terminated from non-null-terminated, since gh_scm2newstr() allows
8340 the user to interpret the data both ways.
8342 * Changes to the scm_ interface
8344 ** The new function scm_symbol_value0 provides an easy way to get a
8345 symbol's value from C code:
8347 SCM scm_symbol_value0 (char *NAME)
8348 Return the value of the symbol named by the null-terminated string
8349 NAME in the current module. If the symbol named NAME is unbound in
8350 the current module, return SCM_UNDEFINED.
8352 ** The new function scm_sysintern0 creates new top-level variables,
8353 without assigning them a value.
8355 SCM scm_sysintern0 (char *NAME)
8356 Create a new Scheme top-level variable named NAME. NAME is a
8357 null-terminated string. Return the variable's value cell.
8359 ** The function scm_internal_catch is the guts of catch. It handles
8360 all the mechanics of setting up a catch target, invoking the catch
8361 body, and perhaps invoking the handler if the body does a throw.
8363 The function is designed to be usable from C code, but is general
8364 enough to implement all the semantics Guile Scheme expects from throw.
8366 TAG is the catch tag. Typically, this is a symbol, but this function
8367 doesn't actually care about that.
8369 BODY is a pointer to a C function which runs the body of the catch;
8370 this is the code you can throw from. We call it like this:
8371 BODY (BODY_DATA, JMPBUF)
8373 BODY_DATA is just the BODY_DATA argument we received; we pass it
8374 through to BODY as its first argument. The caller can make
8375 BODY_DATA point to anything useful that BODY might need.
8376 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
8377 which we have just created and initialized.
8379 HANDLER is a pointer to a C function to deal with a throw to TAG,
8380 should one occur. We call it like this:
8381 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
8383 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
8384 same idea as BODY_DATA above.
8385 THROWN_TAG is the tag that the user threw to; usually this is
8386 TAG, but it could be something else if TAG was #t (i.e., a
8387 catch-all), or the user threw to a jmpbuf.
8388 THROW_ARGS is the list of arguments the user passed to the THROW
8391 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
8392 is just a pointer we pass through to HANDLER. We don't actually
8393 use either of those pointers otherwise ourselves. The idea is
8394 that, if our caller wants to communicate something to BODY or
8395 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
8396 HANDLER can then use. Think of it as a way to make BODY and
8397 HANDLER closures, not just functions; MUMBLE_DATA points to the
8400 Of course, it's up to the caller to make sure that any data a
8401 MUMBLE_DATA needs is protected from GC. A common way to do this is
8402 to make MUMBLE_DATA a pointer to data stored in an automatic
8403 structure variable; since the collector must scan the stack for
8404 references anyway, this assures that any references in MUMBLE_DATA
8407 ** The new function scm_internal_lazy_catch is exactly like
8408 scm_internal_catch, except:
8410 - It does not unwind the stack (this is the major difference).
8411 - If handler returns, its value is returned from the throw.
8412 - BODY always receives #f as its JMPBUF argument (since there's no
8413 jmpbuf associated with a lazy catch, because we don't unwind the
8416 ** scm_body_thunk is a new body function you can pass to
8417 scm_internal_catch if you want the body to be like Scheme's `catch'
8418 --- a thunk, or a function of one argument if the tag is #f.
8420 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
8421 contains the Scheme procedure to invoke as the body, and the tag
8422 we're catching. If the tag is #f, then we pass JMPBUF (created by
8423 scm_internal_catch) to the body procedure; otherwise, the body gets
8426 ** scm_handle_by_proc is a new handler function you can pass to
8427 scm_internal_catch if you want the handler to act like Scheme's catch
8428 --- call a procedure with the tag and the throw arguments.
8430 If the user does a throw to this catch, this function runs a handler
8431 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
8432 variable holding the Scheme procedure object to invoke. It ought to
8433 be a pointer to an automatic variable (i.e., one living on the stack),
8434 or the procedure object should be otherwise protected from GC.
8436 ** scm_handle_by_message is a new handler function to use with
8437 `scm_internal_catch' if you want Guile to print a message and die.
8438 It's useful for dealing with throws to uncaught keys at the top level.
8440 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
8441 message header to print; if zero, we use "guile" instead. That
8442 text is followed by a colon, then the message described by ARGS.
8444 ** The return type of scm_boot_guile is now void; the function does
8445 not return a value, and indeed, never returns at all.
8447 ** The new function scm_shell makes it easy for user applications to
8448 process command-line arguments in a way that is compatible with the
8449 stand-alone guile interpreter (which is in turn compatible with SCSH,
8452 To use the scm_shell function, first initialize any guile modules
8453 linked into your application, and then call scm_shell with the values
8454 of ARGC and ARGV your `main' function received. scm_shell will add
8455 any SCSH-style meta-arguments from the top of the script file to the
8456 argument vector, and then process the command-line arguments. This
8457 generally means loading a script file or starting up an interactive
8458 command interpreter. For details, see "Changes to the stand-alone
8461 ** The new functions scm_get_meta_args and scm_count_argv help you
8462 implement the SCSH-style meta-argument, `\'.
8464 char **scm_get_meta_args (int ARGC, char **ARGV)
8465 If the second element of ARGV is a string consisting of a single
8466 backslash character (i.e. "\\" in Scheme notation), open the file
8467 named by the following argument, parse arguments from it, and return
8468 the spliced command line. The returned array is terminated by a
8471 For details of argument parsing, see above, under "guile now accepts
8472 command-line arguments compatible with SCSH..."
8474 int scm_count_argv (char **ARGV)
8475 Count the arguments in ARGV, assuming it is terminated by a null
8478 For an example of how these functions might be used, see the source
8479 code for the function scm_shell in libguile/script.c.
8481 You will usually want to use scm_shell instead of calling this
8484 ** The new function scm_compile_shell_switches turns an array of
8485 command-line arguments into Scheme code to carry out the actions they
8486 describe. Given ARGC and ARGV, it returns a Scheme expression to
8487 evaluate, and calls scm_set_program_arguments to make any remaining
8488 command-line arguments available to the Scheme code. For example,
8489 given the following arguments:
8491 -e main -s ekko a speckled gecko
8493 scm_set_program_arguments will return the following expression:
8495 (begin (load "ekko") (main (command-line)) (quit))
8497 You will usually want to use scm_shell instead of calling this
8500 ** The function scm_shell_usage prints a usage message appropriate for
8501 an interpreter that uses scm_compile_shell_switches to handle its
8502 command-line arguments.
8504 void scm_shell_usage (int FATAL, char *MESSAGE)
8505 Print a usage message to the standard error output. If MESSAGE is
8506 non-zero, write it before the usage message, followed by a newline.
8507 If FATAL is non-zero, exit the process, using FATAL as the
8508 termination status. (If you want to be compatible with Guile,
8509 always use 1 as the exit status when terminating due to command-line
8512 You will usually want to use scm_shell instead of calling this
8515 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
8516 expressions. It used to return SCM_EOL. Earth-shattering.
8518 ** The macros for declaring scheme objects in C code have been
8519 rearranged slightly. They are now:
8521 SCM_SYMBOL (C_NAME, SCHEME_NAME)
8522 Declare a static SCM variable named C_NAME, and initialize it to
8523 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
8524 be a C identifier, and SCHEME_NAME should be a C string.
8526 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
8527 Just like SCM_SYMBOL, but make C_NAME globally visible.
8529 SCM_VCELL (C_NAME, SCHEME_NAME)
8530 Create a global variable at the Scheme level named SCHEME_NAME.
8531 Declare a static SCM variable named C_NAME, and initialize it to
8532 point to the Scheme variable's value cell.
8534 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
8535 Just like SCM_VCELL, but make C_NAME globally visible.
8537 The `guile-snarf' script writes initialization code for these macros
8538 to its standard output, given C source code as input.
8540 The SCM_GLOBAL macro is gone.
8542 ** The scm_read_line and scm_read_line_x functions have been replaced
8543 by Scheme code based on the %read-delimited! procedure (known to C
8544 code as scm_read_delimited_x). See its description above for more
8547 ** The function scm_sys_open has been renamed to scm_open. It now
8548 returns a port instead of an FD object.
8550 * The dynamic linking support has changed. For more information, see
8551 libguile/DYNAMIC-LINKING.
8556 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
8559 * Changes to the 'guile' program:
8561 ** Guile now loads some new files when it starts up. Guile first
8562 searches the load path for init.scm, and loads it if found. Then, if
8563 Guile is not being used to execute a script, and the user's home
8564 directory contains a file named `.guile', Guile loads that.
8566 ** You can now use Guile as a shell script interpreter.
8568 To paraphrase the SCSH manual:
8570 When Unix tries to execute an executable file whose first two
8571 characters are the `#!', it treats the file not as machine code to
8572 be directly executed by the native processor, but as source code
8573 to be executed by some interpreter. The interpreter to use is
8574 specified immediately after the #! sequence on the first line of
8575 the source file. The kernel reads in the name of the interpreter,
8576 and executes that instead. It passes the interpreter the source
8577 filename as its first argument, with the original arguments
8578 following. Consult the Unix man page for the `exec' system call
8579 for more information.
8581 Now you can use Guile as an interpreter, using a mechanism which is a
8582 compatible subset of that provided by SCSH.
8584 Guile now recognizes a '-s' command line switch, whose argument is the
8585 name of a file of Scheme code to load. It also treats the two
8586 characters `#!' as the start of a comment, terminated by `!#'. Thus,
8587 to make a file of Scheme code directly executable by Unix, insert the
8588 following two lines at the top of the file:
8590 #!/usr/local/bin/guile -s
8593 Guile treats the argument of the `-s' command-line switch as the name
8594 of a file of Scheme code to load, and treats the sequence `#!' as the
8595 start of a block comment, terminated by `!#'.
8597 For example, here's a version of 'echo' written in Scheme:
8599 #!/usr/local/bin/guile -s
8601 (let loop ((args (cdr (program-arguments))))
8604 (display (car args))
8605 (if (pair? (cdr args))
8607 (loop (cdr args)))))
8610 Why does `#!' start a block comment terminated by `!#', instead of the
8611 end of the line? That is the notation SCSH uses, and although we
8612 don't yet support the other SCSH features that motivate that choice,
8613 we would like to be backward-compatible with any existing Guile
8614 scripts once we do. Furthermore, if the path to Guile on your system
8615 is too long for your kernel, you can start the script with this
8619 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
8622 Note that some very old Unix systems don't support the `#!' syntax.
8625 ** You can now run Guile without installing it.
8627 Previous versions of the interactive Guile interpreter (`guile')
8628 couldn't start up unless Guile's Scheme library had been installed;
8629 they used the value of the environment variable `SCHEME_LOAD_PATH'
8630 later on in the startup process, but not to find the startup code
8631 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
8634 To run Guile without installing it, build it in the normal way, and
8635 then set the environment variable `SCHEME_LOAD_PATH' to a
8636 colon-separated list of directories, including the top-level directory
8637 of the Guile sources. For example, if you unpacked Guile so that the
8638 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
8641 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
8644 ** Guile's read-eval-print loop no longer prints #<unspecified>
8645 results. If the user wants to see this, she can evaluate the
8646 expression (assert-repl-print-unspecified #t), perhaps in her startup
8649 ** Guile no longer shows backtraces by default when an error occurs;
8650 however, it does display a message saying how to get one, and how to
8651 request that they be displayed by default. After an error, evaluate
8653 to see a backtrace, and
8654 (debug-enable 'backtrace)
8655 to see them by default.
8659 * Changes to Guile Scheme:
8661 ** Guile now distinguishes between #f and the empty list.
8663 This is for compatibility with the IEEE standard, the (possibly)
8664 upcoming Revised^5 Report on Scheme, and many extant Scheme
8667 Guile used to have #f and '() denote the same object, to make Scheme's
8668 type system more compatible with Emacs Lisp's. However, the change
8669 caused too much trouble for Scheme programmers, and we found another
8670 way to reconcile Emacs Lisp with Scheme that didn't require this.
8673 ** Guile's delq, delv, delete functions, and their destructive
8674 counterparts, delq!, delv!, and delete!, now remove all matching
8675 elements from the list, not just the first. This matches the behavior
8676 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
8677 functions which inspired them.
8679 I recognize that this change may break code in subtle ways, but it
8680 seems best to make the change before the FSF's first Guile release,
8684 ** The compiled-library-path function has been deleted from libguile.
8686 ** The facilities for loading Scheme source files have changed.
8688 *** The variable %load-path now tells Guile which directories to search
8689 for Scheme code. Its value is a list of strings, each of which names
8692 *** The variable %load-extensions now tells Guile which extensions to
8693 try appending to a filename when searching the load path. Its value
8694 is a list of strings. Its default value is ("" ".scm").
8696 *** (%search-load-path FILENAME) searches the directories listed in the
8697 value of the %load-path variable for a Scheme file named FILENAME,
8698 with all the extensions listed in %load-extensions. If it finds a
8699 match, then it returns its full filename. If FILENAME is absolute, it
8700 returns it unchanged. Otherwise, it returns #f.
8702 %search-load-path will not return matches that refer to directories.
8704 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
8705 uses %seach-load-path to find a file named FILENAME, and loads it if
8706 it finds it. If it can't read FILENAME for any reason, it throws an
8709 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
8712 *** load uses the same searching semantics as primitive-load.
8714 *** The functions %try-load, try-load-with-path, %load, load-with-path,
8715 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
8716 path, and load-module-with-path have been deleted. The functions
8717 above should serve their purposes.
8719 *** If the value of the variable %load-hook is a procedure,
8720 `primitive-load' applies its value to the name of the file being
8721 loaded (without the load path directory name prepended). If its value
8722 is #f, it is ignored. Otherwise, an error occurs.
8724 This is mostly useful for printing load notification messages.
8727 ** The function `eval!' is no longer accessible from the scheme level.
8728 We can't allow operations which introduce glocs into the scheme level,
8729 because Guile's type system can't handle these as data. Use `eval' or
8730 `read-and-eval!' (see below) as replacement.
8732 ** The new function read-and-eval! reads an expression from PORT,
8733 evaluates it, and returns the result. This is more efficient than
8734 simply calling `read' and `eval', since it is not necessary to make a
8735 copy of the expression for the evaluator to munge.
8737 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
8738 for the `read' function.
8741 ** The function `int?' has been removed; its definition was identical
8742 to that of `integer?'.
8744 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
8745 use the R4RS names for these functions.
8747 ** The function object-properties no longer returns the hash handle;
8748 it simply returns the object's property list.
8750 ** Many functions have been changed to throw errors, instead of
8751 returning #f on failure. The point of providing exception handling in
8752 the language is to simplify the logic of user code, but this is less
8753 useful if Guile's primitives don't throw exceptions.
8755 ** The function `fileno' has been renamed from `%fileno'.
8757 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
8760 * Changes to Guile's C interface:
8762 ** The library's initialization procedure has been simplified.
8763 scm_boot_guile now has the prototype:
8765 void scm_boot_guile (int ARGC,
8767 void (*main_func) (),
8770 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
8771 MAIN_FUNC should do all the work of the program (initializing other
8772 packages, reading user input, etc.) before returning. When MAIN_FUNC
8773 returns, call exit (0); this function never returns. If you want some
8774 other exit value, MAIN_FUNC may call exit itself.
8776 scm_boot_guile arranges for program-arguments to return the strings
8777 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
8778 scm_set_program_arguments with the final list, so Scheme code will
8779 know which arguments have been processed.
8781 scm_boot_guile establishes a catch-all catch handler which prints an
8782 error message and exits the process. This means that Guile exits in a
8783 coherent way when system errors occur and the user isn't prepared to
8784 handle it. If the user doesn't like this behavior, they can establish
8785 their own universal catcher in MAIN_FUNC to shadow this one.
8787 Why must the caller do all the real work from MAIN_FUNC? The garbage
8788 collector assumes that all local variables of type SCM will be above
8789 scm_boot_guile's stack frame on the stack. If you try to manipulate
8790 SCM values after this function returns, it's the luck of the draw
8791 whether the GC will be able to find the objects you allocate. So,
8792 scm_boot_guile function exits, rather than returning, to discourage
8793 people from making that mistake.
8795 The IN, OUT, and ERR arguments were removed; there are other
8796 convenient ways to override these when desired.
8798 The RESULT argument was deleted; this function should never return.
8800 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
8804 ** Guile's header files should no longer conflict with your system's
8807 In order to compile code which #included <libguile.h>, previous
8808 versions of Guile required you to add a directory containing all the
8809 Guile header files to your #include path. This was a problem, since
8810 Guile's header files have names which conflict with many systems'
8813 Now only <libguile.h> need appear in your #include path; you must
8814 refer to all Guile's other header files as <libguile/mumble.h>.
8815 Guile's installation procedure puts libguile.h in $(includedir), and
8816 the rest in $(includedir)/libguile.
8819 ** Two new C functions, scm_protect_object and scm_unprotect_object,
8820 have been added to the Guile library.
8822 scm_protect_object (OBJ) protects OBJ from the garbage collector.
8823 OBJ will not be freed, even if all other references are dropped,
8824 until someone does scm_unprotect_object (OBJ). Both functions
8827 Note that calls to scm_protect_object do not nest. You can call
8828 scm_protect_object any number of times on a given object, and the
8829 next call to scm_unprotect_object will unprotect it completely.
8831 Basically, scm_protect_object and scm_unprotect_object just
8832 maintain a list of references to things. Since the GC knows about
8833 this list, all objects it mentions stay alive. scm_protect_object
8834 adds its argument to the list; scm_unprotect_object remove its
8835 argument from the list.
8838 ** scm_eval_0str now returns the value of the last expression
8841 ** The new function scm_read_0str reads an s-expression from a
8842 null-terminated string, and returns it.
8844 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
8845 to a Scheme port object.
8847 ** The new function `scm_set_program_arguments' allows C code to set
8848 the value returned by the Scheme `program-arguments' function.
8853 * Guile no longer includes sophisticated Tcl/Tk support.
8855 The old Tcl/Tk support was unsatisfying to us, because it required the
8856 user to link against the Tcl library, as well as Tk and Guile. The
8857 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
8858 referring to widgets by names, rather than exporting widgets to Scheme
8859 code as a special datatype.
8861 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
8862 maintainers described some very interesting changes in progress to the
8863 Tcl/Tk internals, which would facilitate clean interfaces between lone
8864 Tk and other interpreters --- even for garbage-collected languages
8865 like Scheme. They expected the new Tk to be publicly available in the
8868 Since it seems that Guile might soon have a new, cleaner interface to
8869 lone Tk, and that the old Guile/Tk glue code would probably need to be
8870 completely rewritten, we (Jim Blandy and Richard Stallman) have
8871 decided not to support the old code. We'll spend the time instead on
8872 a good interface to the newer Tk, as soon as it is available.
8874 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
8877 Copyright information:
8879 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
8881 Permission is granted to anyone to make or distribute verbatim copies
8882 of this document as received, in any medium, provided that the
8883 copyright notice and this permission notice are preserved,
8884 thus giving the recipient permission to redistribute in turn.
8886 Permission is granted to distribute modified versions
8887 of this document, or of portions of it,
8888 under the above conditions, provided also that they
8889 carry prominent notices stating who last changed them.
8894 paragraph-separate: "[
\f]*$"