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 1.9.7 (since the 1.9.6 prerelease):
13 ** Complete support for version information in Guile's `module' form
15 Guile modules now have a `#:version' field. They may be loaded by
16 version as well. See "R6RS Version References", "General Information
17 about Modules", "Using Guile Modules", and "Creating Guile Modules" in
18 the manual for more information.
20 ** Support for renaming bindings on module export
22 Wherever Guile accepts a symbol as an argument to specify a binding to
23 export, it now also accepts a pair of symbols, indicating that a binding
24 should be renamed on export. See "Creating Guile Modules" in the manual
27 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
29 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
30 toolkit from Guile-Lib have been imported into Guile proper. See
31 "Standard Library" in the manual for more details.
33 ** Function profiling and tracing at the REPL
35 The `,profile FORM' REPL meta-command can now be used to statistically
36 profile execution of a form, to see which functions are taking the most
37 time. See `,help profile' for more information.
39 Similarly, `,trace FORM' traces all function applications that occur
40 during the execution of `FORM'. See `,help trace' for more information.
44 By default, if an exception is raised at the REPL and not caught by user
45 code, Guile will drop the user into a debugger. The user may request a
46 backtrace, inspect frames, or continue raising the exception. Full
47 documentation is available from within the debugger.
49 ** New function, `truncated-print', with `format' support
51 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
52 will ensure that the output stays within a certain width, truncating the
53 output in what is hopefully an intelligent manner. See the manual for
56 There is a new `format' specifier, `~@y', for doing a truncated
57 print (as opposed to `~y', which does a pretty-print). See the `format'
58 documentation for more details.
60 ** Primitive procedures are now VM trampoline procedures
62 Guile's old evaluator used to have special cases for applying "gsubrs",
63 primitive procedures with specified numbers of required, optional, and
64 rest arguments. Now, however, Guile represents gsubrs as normal VM
65 procedures, with appropriate bytecode to parse out the correct number of
66 arguments, including optional and rest arguments, and then with a
67 special bytecode to apply the gsubr.
69 This allows primitive procedures to appear on the VM stack, allowing
70 them to be accurately counted in profiles. Also they now have more
71 debugging information attached to them -- their number of arguments, for
72 example. In addition, the VM can completely inline the application
73 mechanics, allowing for faster primitive calls.
75 However there are some changes on the C level. There is no more
76 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
77 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
78 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
79 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
80 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
82 Perhaps more significantly, `scm_c_make_subr',
83 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
84 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
85 and there are no more subr typecodes. Use the scm_c_make_gsubr family
88 Normal users of gsubrs should not be affected, though, as the
89 scm_c_make_gsubr family still is the correct way to create primitive
92 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
94 Guile now implements SRFI-4 vectors using bytevectors. Often when you
95 have a numeric vector, you end up wanting to write its bytes somewhere,
96 or have access to the underlying bytes, or read in bytes from somewhere
97 else. Bytevectors are very good at this sort of thing. But the SRFI-4
98 APIs are nicer to use when doing number-crunching, because they are
99 addressed by element and not by byte.
101 So as a compromise, Guile allows all bytevector functions to operate on
102 numeric vectors. They address the underlying bytes in the native
103 endianness, as one would expect.
105 Following the same reasoning, that it's just bytes underneath, Guile
106 also allows uniform vectors of a given type to be accessed as if they
107 were of any type. One can fill a u32vector, and access its elements with
108 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
111 In this way, uniform numeric vectors may be written to and read from
112 input/output ports using the procedures that operate on bytevectors.
114 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
115 inlined to the VM instructions for bytevector access.
117 See "SRFI-4" in the manual, for more information.
119 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
121 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
122 are part of the standard. Complex uniform vectors and the
123 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
125 Guile's default environment imports `(srfi srfi-4)', and probably should
126 import `(srfi srfi-4 gnu)' as well.
128 See "SRFI-4 Extensions" in the manual, for more information.
130 ** Unicode: String normalization, general categories, title case
132 Guile now exports the Unicode string normalization functions
133 `string-normalize-nfc', `string-normalize-nfd', `string-normalize-nfkc',
134 and `string-normalize-nfkd'. See "String Comparison" in the manual for
135 more information. See "Characters" for information on the new
136 `char-general-category', and "Character Case Mapping" for
137 `char-titlecase', `char-locale-titlecase', and
138 `string-locale-titlecase'.
140 ** Faster access to the free variables of a closure
142 Guile's old garbage collector limited the size of Scheme objects to be 2
143 or 4 words. Now with the BDW collector, Guile can allocate
144 variable-sized objects, allowing storage of the free variables of a
145 closure inline to the closure itself.
147 ** New primitive `getsid'
149 Scheme binding for the `getsid' C library call.
151 ** Changed invocation mechanics of applicable SMOBs
153 Guile's old evaluator used to have special cases for applying SMOB
154 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
155 trampoline procedure for it, and use the normal mechanics to apply the
156 trampoline. This simplifies procedure application in the normal,
159 The upshot is that the mechanics used to apply a SMOB are different from
160 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
161 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
162 deprecated. Just use the scm_call_0 family of procedures.
164 ** Remove deprecated array C interfaces
166 Removed the deprecated array functions `scm_i_arrayp',
167 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
168 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
169 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
170 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
172 ** String encoding errors throw to `encoding-error' instead of `misc-error'
174 ** Keyword argument errors uniformly throw to `keyword-argument-error'
176 ** Remove unused snarf macros
178 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
179 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
181 ** Add foreign value wrapper
183 Guile now has a datatype for aliasing "foreign" values, such as native
184 long values. This should be useful for making a proper foreign function
185 interface. Interested hackers should see libguile/foreign.h.
187 ** New functions: `scm_call_n', `scm_c_run_hookn'
189 `scm_call_n' applies to apply a function to an array of arguments.
190 `scm_c_run_hookn' runs a hook with an array of arguments.
192 ** Load path change: search in version-specific paths before site paths
194 When looking for a module, Guile now searches first in Guile's
195 version-specific path (the library path), *then* in the site dir. This
196 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
197 installed. Also it should cut the number of `stat' system calls by half,
200 ** Compile-time warning: -Wunused-toplevel
202 Guile can warn about potentially unused top-level (global)
203 variables. Pass the -Wunused-toplevel on the `guile-tools compile'
205 `#:warnings '(unused-toplevel)' to your `compile' or `compile-file'
208 ** New reader options: `square-brackets' and `r6rs-hex-escapes'
210 The reader supports a new option (changeable via `read-options'),
211 `square-brackets', which instructs it to interpret square brackets as
212 parenthesis. This option is on by default.
214 If this causes problems with your code, make sure to report it to
215 bug-guile@gnu.org so we can change the default.
217 When the new `r6rs-hex-escapes' reader option is enabled, the reader
218 will recognize string escape sequences as defined in R6RS.
220 ** And of course, the usual collection of bugfixes
222 Interested users should see the ChangeLog for more information.
226 Changes in 1.9.x (since the 1.8.x series):
228 * New modules (see the manual for details)
230 ** `(srfi srfi-18)', more sophisticated multithreading support
231 ** `(ice-9 i18n)', internationalization support
232 ** `(rnrs bytevector)', the R6RS bytevector API
233 ** `(rnrs io ports)', a subset of the R6RS I/O port API
234 ** `(system xref)', a cross-referencing facility (FIXME undocumented)
236 ** Imported statprof, SSAX, and texinfo modules from Guile-Lib
238 The statprof statistical profiler, the SSAX XML toolkit, and the texinfo
239 toolkit from Guile-Lib have been imported into Guile proper. See
240 "Standard Library" in the manual for more details.
242 * Changes to the stand-alone interpreter
244 ** Guile now can compile Scheme to bytecode for a custom virtual machine.
246 Compiled code loads much faster than Scheme source code, and runs around
247 3 or 4 times as fast, generating much less garbage in the process.
249 ** The stack limit is now initialized from the environment.
251 If getrlimit(2) is available and a stack limit is set, Guile will set
252 its stack limit to 80% of the rlimit. Otherwise the limit is 160000
253 words, a four-fold increase from the earlier default limit.
255 ** New environment variables: GUILE_LOAD_COMPILED_PATH,
256 GUILE_SYSTEM_LOAD_COMPILED_PATH
258 GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is
259 for source files. It is a different path, however, because compiled
260 files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like
263 ** New read-eval-print loop (REPL) implementation
265 Running Guile with no arguments drops the user into the new REPL. While
266 it is self-documenting to an extent, the new REPL has not yet been
267 documented in the manual. This will be fixed before 2.0.
269 ** New reader options: `square-brackets' and `r6rs-hex-escapes'
271 The reader supports a new option (changeable via `read-options'),
272 `square-brackets', which instructs it to interpret square brackets as
273 parenthesis. This option is on by default.
275 When the new `r6rs-hex-escapes' reader option is enabled, the reader
276 will recognize string escape sequences as defined in R6RS.
278 ** Function profiling and tracing at the REPL
280 The `,profile FORM' REPL meta-command can now be used to statistically
281 profile execution of a form, to see which functions are taking the most
282 time. See `,help profile' for more information.
284 Similarly, `,trace FORM' traces all function applications that occur
285 during the execution of `FORM'. See `,help trace' for more information.
289 By default, if an exception is raised at the REPL and not caught by user
290 code, Guile will drop the user into a debugger. The user may request a
291 backtrace, inspect frames, or continue raising the exception. Full
292 documentation is available from within the debugger.
294 ** New `guile-tools' commands: `compile', `disassemble'
296 Pass the `--help' command-line option to these commands for more
299 ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH
301 Users may now install Guile to nonstandard prefixes and just run
302 `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to
303 include `/path/to/lib'.
305 ** Guile's Emacs integration is now more keyboard-friendly
307 Backtraces may now be disclosed with the keyboard in addition to the
310 ** Load path change: search in version-specific paths before site paths
312 When looking for a module, Guile now searches first in Guile's
313 version-specific path (the library path), *then* in the site dir. This
314 allows Guile's copy of SSAX to override any Guile-Lib copy the user has
315 installed. Also it should cut the number of `stat' system calls by half,
319 * Changes to Scheme functions and syntax
321 ** New implementation of `primitive-eval'
323 Guile's `primitive-eval' is now implemented in Scheme. Actually there is
324 still a C evaluator, used when building a fresh Guile to interpret the
325 compiler, so we can compile eval.scm. Thereafter all calls to
326 primitive-eval are implemented by VM-compiled code.
328 This allows all of Guile's procedures, be they interpreted or compiled,
329 to execute on the same stack, unifying multiple-value return semantics,
330 providing for proper tail recursion between interpreted and compiled
331 code, and simplifying debugging.
333 As part of this change, the evaluator no longer mutates the internal
334 representation of the code being evaluated in a thread-unsafe manner.
336 There are two negative aspects of this change, however. First, Guile
337 takes a lot longer to compile now. Also, there is less debugging
338 information available for debugging interpreted code. We hope to improve
339 both of these situations.
341 There are many changes to the internal C evalator interface, but all
342 public interfaces should be the same. See the ChangeLog for details. If
343 we have inadvertantly changed an interface that you were using, please
344 contact bug-guile@gnu.org.
346 ** Procedure removed: `the-environment'
348 This procedure was part of the interpreter's execution model, and does
349 not apply to the compiler.
351 ** No more `local-eval'
353 `local-eval' used to exist so that one could evaluate code in the
354 lexical context of a function. Since there is no way to get the lexical
355 environment any more, as that concept has no meaning for the compiler,
356 and a different meaning for the interpreter, we have removed the
359 If you think you need `local-eval', you should probably implement your
360 own metacircular evaluator. It will probably be as fast as Guile's
363 ** Files loaded with `primitive-load-path' will now be compiled
366 If a compiled .go file corresponding to a .scm file is not found or is
367 not fresh, the .scm file will be compiled on the fly, and the resulting
368 .go file stored away. An advisory note will be printed on the console.
370 Note that this mechanism depends on preservation of the .scm and .go
371 modification times; if the .scm or .go files are moved after
372 installation, care should be taken to preserve their original
375 Autocompiled files will be stored in the $XDG_CACHE_HOME/guile/ccache
376 directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory
377 will be created if needed.
379 To inhibit autocompilation, set the GUILE_AUTO_COMPILE environment
380 variable to 0, or pass --no-autocompile on the Guile command line.
382 Note that there is currently a bug here: automatic compilation will
383 sometimes be attempted when it shouldn't.
385 For example, the old (lang elisp) modules are meant to be interpreted,
386 not compiled. This bug will be fixed before 2.0. FIXME 2.0: Should say
387 something here about module-transformer called for compile.
389 ** Files loaded with `load' will now be compiled automatically.
391 As with files loaded via `primitive-load-path', `load' will also compile
392 its target if autocompilation is enabled, and a fresh compiled file is
395 There are two points of difference to note, however. First, `load' does
396 not search `GUILE_LOAD_COMPILED_PATH' for the file; it only looks in the
397 autocompilation directory, normally a subdirectory of ~/.cache/guile.
399 Secondly, autocompilation also applies to files loaded via the -l
400 command-line argument -- so the user may experience a slight slowdown
401 the first time they run a Guile script, as the script is autocompiled.
403 ** New POSIX procedures: `getrlimit' and `setrlimit'
405 Note however that the interface of these functions is likely to change
406 in the next prerelease.
408 ** New POSIX procedure: `getsid'
410 Scheme binding for the `getsid' C library call.
412 ** New procedure in `(oops goops)': `method-formals'
414 ** BUG: (procedure-property func 'arity) does not work on compiled
417 This will be fixed one way or another before 2.0.
419 ** New procedures in (ice-9 session): `add-value-help-handler!',
420 `remove-value-help-handler!', `add-name-help-handler!'
421 `remove-name-help-handler!', `procedure-arguments',
423 The value and name help handlers provide some minimal extensibility to
424 the help interface. Guile-lib's `(texinfo reflection)' uses them, for
425 example, to make stexinfo help documentation available. See those
426 procedures' docstrings for more information.
428 `procedure-arguments' describes the arguments that a procedure can take,
429 combining arity and formals. For example:
431 (procedure-arguments resolve-interface)
432 => ((required . (name)) (rest . args))
434 Additionally, `module-commentary' is now publically exported from
437 ** Removed: `procedure->memoizing-macro', `procedure->syntax'
439 These procedures created primitive fexprs for the old evaluator, and are
440 no longer supported. If you feel that you need these functions, you
441 probably need to write your own metacircular evaluator (which will
442 probably be as fast as Guile's, anyway).
444 ** New language: ECMAScript
446 Guile now ships with one other high-level language supported,
447 ECMAScript. The goal is to support all of version 3.1 of the standard,
448 but not all of the libraries are there yet. This support is not yet
449 documented; ask on the mailing list if you are interested.
451 ** New language: Brainfuck
453 Brainfuck is a toy language that closely models Turing machines. Guile's
454 brainfuck compiler is meant to be an example of implementing other
455 languages. See the manual for details, or
456 http://en.wikipedia.org/wiki/Brainfuck for more information about the
457 Brainfuck language itself.
459 ** New language: Elisp
461 Guile now has an experimental Emacs Lisp compiler and runtime. You can
462 now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel
463 Kraft, and all bugs to bug-guile@gnu.org.
465 ** Defmacros may now have docstrings.
467 Indeed, any macro may have a docstring. `object-documentation' from
468 `(ice-9 documentation)' may be used to retrieve the docstring, once you
469 have a macro value -- but see the above note about first-class macros.
470 Docstrings are associated with the syntax transformer procedures.
472 ** The psyntax expander now knows how to interpret the @ and @@ special
475 ** The psyntax expander is now hygienic with respect to modules.
477 Free variables in a macro are scoped in the module that the macro was
478 defined in, not in the module the macro is used in. For example, code
481 (define-module (foo) #:export (bar))
482 (define (helper x) ...)
484 (syntax-rules () ((_ x) (helper x))))
486 (define-module (baz) #:use-module (foo))
489 It used to be you had to export `helper' from `(foo)' as well.
490 Thankfully, this has been fixed.
492 ** Complete support for version information in Guile's `module' form
494 Guile modules now have a `#:version' field. They may be loaded by
495 version as well. See "R6RS Version References", "General Information
496 about Modules", "Using Guile Modules", and "Creating Guile Modules" in
497 the manual for more information.
499 ** Support for renaming bindings on module export
501 Wherever Guile accepts a symbol as an argument to specify a binding to
502 export, it now also accepts a pair of symbols, indicating that a binding
503 should be renamed on export. See "Creating Guile Modules" in the manual
504 for more information.
506 ** `eval-case' has been deprecated, and replaced by `eval-when'.
508 The semantics of `eval-when' are easier to understand. It is still
509 missing documentation, however.
511 ** Guile is now more strict about prohibiting definitions in expression
514 Although previous versions of Guile accepted it, the following
515 expression is not valid, in R5RS or R6RS:
517 (if test (define foo 'bar) (define foo 'baz))
519 In this specific case, it would be better to do:
521 (define foo (if test 'bar 'baz))
523 It is certainly possible to circumvent this resriction with e.g.
524 `(module-define! (current-module) 'foo 'baz)'. We would appreciate
525 feedback about this change (a consequence of using psyntax as the
526 default expander), and may choose to revisit this situation before 2.0
527 in response to user feedback.
529 ** Defmacros must now produce valid Scheme expressions.
531 It used to be that defmacros could unquote in Scheme values, as a way of
532 supporting partial evaluation, and avoiding some hygiene issues. For
535 (define (helper x) ...)
536 (define-macro (foo bar)
539 Assuming this macro is in the `(baz)' module, the direct translation of
542 (define (helper x) ...)
543 (define-macro (foo bar)
544 `((@@ (baz) helper) ,bar))
546 Of course, one could just use a hygienic macro instead:
550 ((_ bar) (helper bar))))
552 ** Guile's psyntax now supports docstrings and internal definitions.
554 The following Scheme is not strictly legal:
561 However its intent is fairly clear. Guile interprets "bar" to be the
562 docstring of `foo', and the definition of `baz' is still in definition
565 ** Macros need to be defined before their first use.
567 It used to be that with lazy memoization, this might work:
571 (define-macro (ref x) x)
574 But now, the body of `foo' is interpreted to mean a call to the toplevel
575 `ref' function, instead of a macro expansion. The solution is to define
576 macros before code that uses them.
578 ** Functions needed by macros at expand-time need to be present at
581 For example, this code will work at the REPL:
583 (define (double-helper x) (* x x))
584 (define-macro (double-literal x) (double-helper x))
585 (double-literal 2) => 4
587 But it will not work when a file is compiled, because the definition of
588 `double-helper' is not present at expand-time. The solution is to wrap
589 the definition of `double-helper' in `eval-when':
591 (eval-when (load compile eval)
592 (define (double-helper x) (* x x)))
593 (define-macro (double-literal x) (double-helper x))
594 (double-literal 2) => 4
596 See the (currently missing) documentation for eval-when for more
599 ** New variable, %pre-modules-transformer
601 Need to document this one some more.
603 ** Temporarily removed functions: `macroexpand', `macroexpand-1'
605 `macroexpand' will be added back before 2.0. It is unclear how to
606 implement `macroexpand-1' with syntax-case, though PLT Scheme does prove
609 ** New reader macros: #' #` #, #,@
611 These macros translate, respectively, to `syntax', `quasisyntax',
612 `unsyntax', and `unsyntax-splicing'. See the R6RS for more information.
613 These reader macros may be overridden by `read-hash-extend'.
615 ** Incompatible change to #'
617 Guile did have a #' hash-extension, by default, which just returned the
618 subsequent datum: #'foo => foo. In the unlikely event that anyone
619 actually used this, this behavior may be reinstated via the
620 `read-hash-extend' mechanism.
622 ** Scheme expresssions may be commented out with #;
624 #; comments out an entire expression. See SRFI-62 or the R6RS for more
627 ** `make-stack' with a tail-called procedural narrowing argument no longer
628 works (with compiled procedures)
630 It used to be the case that a captured stack could be narrowed to select
631 calls only up to or from a certain procedure, even if that procedure
632 already tail-called another procedure. This was because the debug
633 information from the original procedure was kept on the stack.
635 Now with the new compiler, the stack only contains active frames from
636 the current continuation. A narrow to a procedure that is not in the
637 stack will result in an empty stack. To fix this, narrow to a procedure
638 that is active in the current continuation, or narrow to a specific
639 number of stack frames.
641 ** backtraces through compiled procedures only show procedures that are
642 active in the current continuation
644 Similarly to the previous issue, backtraces in compiled code may be
645 different from backtraces in interpreted code. There are no semantic
646 differences, however. Please mail bug-guile@gnu.org if you see any
647 deficiencies with Guile's backtraces.
649 ** syntax-rules and syntax-case macros now propagate source information
650 through to the expanded code
652 This should result in better backtraces.
654 ** The currying behavior of `define' has been removed.
656 Before, `(define ((f a) b) (* a b))' would translate to
658 (define f (lambda (a) (lambda (b) (* a b))))
660 Now a syntax error is signaled, as this syntax is not supported by
661 default. If there is sufficient demand, this syntax can be supported
664 ** New procedure, `define!'
666 `define!' is a procedure that takes two arguments, a symbol and a value,
667 and binds the value to the symbol in the current module. It's useful to
668 programmatically make definitions in the current module, and is slightly
669 less verbose than `module-define!'.
671 ** All modules have names now
673 Before, you could have anonymous modules: modules without names. Now,
674 because of hygiene and macros, all modules have names. If a module was
675 created without a name, the first time `module-name' is called on it, a
676 fresh name will be lazily generated for it.
678 ** Many syntax errors have different texts now
680 Syntax errors still throw to the `syntax-error' key, but the arguments
681 are often different now. Perhaps in the future, Guile will switch to
682 using standard SRFI-35 conditions.
684 ** Returning multiple values to compiled code will silently truncate the
685 values to the expected number
687 For example, the interpreter would raise an error evaluating the form,
688 `(+ (values 1 2) (values 3 4))', because it would see the operands as
689 being two compound "values" objects, to which `+' does not apply.
691 The compiler, on the other hand, receives multiple values on the stack,
692 not as a compound object. Given that it must check the number of values
693 anyway, if too many values are provided for a continuation, it chooses
694 to truncate those values, effectively evaluating `(+ 1 3)' instead.
696 The idea is that the semantics that the compiler implements is more
697 intuitive, and the use of the interpreter will fade out with time.
698 This behavior is allowed both by the R5RS and the R6RS.
700 ** Multiple values in compiled code are not represented by compound
703 This change may manifest itself in the following situation:
705 (let ((val (foo))) (do-something) val)
707 In the interpreter, if `foo' returns multiple values, multiple values
708 are produced from the `let' expression. In the compiler, those values
709 are truncated to the first value, and that first value is returned. In
710 the compiler, if `foo' returns no values, an error will be raised, while
711 the interpreter would proceed.
713 Both of these behaviors are allowed by R5RS and R6RS. The compiler's
714 behavior is more correct, however. If you wish to preserve a potentially
715 multiply-valued return, you will need to set up a multiple-value
716 continuation, using `call-with-values'.
718 ** Defmacros are now implemented in terms of syntax-case.
720 The practical ramification of this is that the `defmacro?' predicate has
721 been removed, along with `defmacro-transformer', `macro-table',
722 `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and
723 `defmacro:transformer'. This is because defmacros are simply macros. If
724 any of these procedures provided useful facilities to you, we encourage
725 you to contact the Guile developers.
727 ** psyntax is now the default expander
729 Scheme code is now expanded by default by the psyntax hygienic macro
730 expander. Expansion is performed completely before compilation or
733 Notably, syntax errors will be signalled before interpretation begins.
734 In the past, many syntax errors were only detected at runtime if the
735 code in question was memoized.
737 As part of its expansion, psyntax renames all lexically-bound
738 identifiers. Original identifier names are preserved and given to the
739 compiler, but the interpreter will see the renamed variables, e.g.,
740 `x432' instead of `x'.
742 Note that the psyntax that Guile uses is a fork, as Guile already had
743 modules before incompatible modules were added to psyntax -- about 10
744 years ago! Thus there are surely a number of bugs that have been fixed
745 in psyntax since then. If you find one, please notify bug-guile@gnu.org.
747 ** syntax-rules and syntax-case are available by default.
749 There is no longer any need to import the `(ice-9 syncase)' module
750 (which is now deprecated). The expander may be invoked directly via
751 `sc-expand', though it is normally searched for via the current module
754 Also, the helper routines for syntax-case are available in the default
755 environment as well: `syntax->datum', `datum->syntax',
756 `bound-identifier=?', `free-identifier=?', `generate-temporaries',
757 `identifier?', and `syntax-violation'. See the R6RS for documentation.
759 ** Tail patterns in syntax-case
761 Guile has pulled in some more recent changes from the psyntax portable
762 syntax expander, to implement support for "tail patterns". Such patterns
763 are supported by syntax-rules and syntax-case. This allows a syntax-case
764 match clause to have ellipses, then a pattern at the end. For example:
768 ((_ val match-clause ... (else e e* ...))
771 Note how there is MATCH-CLAUSE, which is ellipsized, then there is a
772 tail pattern for the else clause. Thanks to Andreas Rottmann for the
773 patch, and Kent Dybvig for the code.
775 ** Lexical bindings introduced by hygienic macros may not be referenced
776 by nonhygienic macros.
778 If a lexical binding is introduced by a hygienic macro, it may not be
779 referenced by a nonhygienic macro. For example, this works:
782 (define-macro (bind-x val body)
783 `(let ((x ,val)) ,body))
784 (define-macro (ref x)
791 (define-syntax bind-x
793 ((_ val body) (let ((x val)) body))))
794 (define-macro (ref x)
798 It is not normal to run into this situation with existing code. However,
799 as code is ported over from defmacros to syntax-case, it is possible to
800 run into situations like this. In the future, Guile will probably port
801 its `while' macro to syntax-case, which makes this issue one to know
804 ** Macros may no longer be referenced as first-class values.
806 In the past, you could evaluate e.g. `if', and get its macro value. Now,
807 expanding this form raises a syntax error.
809 Macros still /exist/ as first-class values, but they must be
810 /referenced/ via the module system, e.g. `(module-ref (current-module)
813 This decision may be revisited before the 2.0 release. Feedback welcome
814 to guile-devel@gnu.org (subscription required) or bug-guile@gnu.org (no
815 subscription required).
817 ** `case-lambda' is now available in the default environment.
819 The binding in the default environment is equivalent to the one from the
820 `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish
821 to maintain compatibility with Guile 1.8 and earlier.
823 ** Compiled procedures may now have more than one arity.
825 This can be the case, for example, in case-lambda procedures. The
826 arities of compiled procedures may be accessed via procedures from the
827 `(system vm program)' module; see "Compiled Procedures", "Optional
828 Arguments", and "Case-lambda" in the manual.
830 ** `lambda*' and `define*' are now available in the default environment
832 As with `case-lambda', `(ice-9 optargs)' continues to be supported, for
833 compatibility purposes. No semantic change has been made (we hope).
834 Optional and keyword arguments now dispatch via special VM operations,
835 without the need to cons rest arguments, making them very fast.
837 ** New function, `truncated-print', with `format' support
839 `(ice-9 pretty-print)' now exports `truncated-print', a printer that
840 will ensure that the output stays within a certain width, truncating the
841 output in what is hopefully an intelligent manner. See the manual for
844 There is a new `format' specifier, `~@y', for doing a truncated
845 print (as opposed to `~y', which does a pretty-print). See the `format'
846 documentation for more details.
848 ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors
850 Guile now implements SRFI-4 vectors using bytevectors. Often when you
851 have a numeric vector, you end up wanting to write its bytes somewhere,
852 or have access to the underlying bytes, or read in bytes from somewhere
853 else. Bytevectors are very good at this sort of thing. But the SRFI-4
854 APIs are nicer to use when doing number-crunching, because they are
855 addressed by element and not by byte.
857 So as a compromise, Guile allows all bytevector functions to operate on
858 numeric vectors. They address the underlying bytes in the native
859 endianness, as one would expect.
861 Following the same reasoning, that it's just bytes underneath, Guile
862 also allows uniform vectors of a given type to be accessed as if they
863 were of any type. One can fill a u32vector, and access its elements with
864 u8vector-ref. One can use f64vector-ref on bytevectors. It's all the
867 In this way, uniform numeric vectors may be written to and read from
868 input/output ports using the procedures that operate on bytevectors.
870 Calls to SRFI-4 accessors (ref and set functions) from Scheme are now
871 inlined to the VM instructions for bytevector access.
873 See "SRFI-4" in the manual, for more information.
875 ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)'
877 Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that
878 are part of the standard. Complex uniform vectors and the
879 `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'.
881 Guile's default environment imports `(srfi srfi-4)', and probably should
882 import `(srfi srfi-4 gnu)' as well.
884 See "SRFI-4 Extensions" in the manual, for more information.
886 ** New syntax: include-from-path.
888 `include-from-path' is like `include', except it looks for its file in
889 the load path. It can be used to compile other files into a file.
891 ** New syntax: quasisyntax.
893 `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS
894 documentation for more information. Thanks to Andre van Tonder for the
897 ** Unicode characters
899 Unicode characters may be entered in octal format via e.g. `#\454', or
900 created via (integer->char 300). A hex external representation will
901 probably be introduced at some point.
905 Internally, strings are now represented either in the `latin-1'
906 encoding, one byte per character, or in UTF-32, with four bytes per
907 character. Strings manage their own allocation, switching if needed.
909 Extended characters may be written in a literal string using the
910 hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit,
911 or 24-bit codepoints, respectively, or entered directly in the native
912 encoding of the port on which the string is read.
916 One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier.
918 ** Support for non-ASCII source code files
920 The default reader now handles source code files for some of the
921 non-ASCII character encodings, such as UTF-8. A non-ASCII source file
922 should have an encoding declaration near the top of the file. Also,
923 there is a new function, `file-encoding', that scans a port for a coding
924 declaration. See the section of the manual entitled, "Character Encoding
927 The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source
928 code. This use is now discouraged.
930 ** Support for locale transcoding when reading from and writing to ports
932 Ports now have an associated character encoding, and port read and write
933 operations do conversion to and from locales automatically. Ports also
934 have an associated strategy for how to deal with locale conversion
937 See the documentation in the manual for the four new support functions,
938 `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!',
939 and `port-conversion-strategy'.
941 ** String and SRFI-13 functions can operate on Unicode strings
943 ** Unicode support for SRFI-14 character sets
945 The default character sets are no longer locale dependent and contain
946 characters from the whole Unicode range. There is a new predefined
947 character set, `char-set:designated', which contains all assigned
948 Unicode characters. There is a new debugging function, `%char-set-dump'.
950 ** Character functions operate on Unicode characters
952 `char-upcase' and `char-downcase' use default Unicode casing rules.
953 Character comparisons such as `char<?' and `char-ci<?' now sort based on
956 ** Global variables `scm_charnames' and `scm_charnums' are removed
958 These variables contained the names of control characters and were
959 used when writing characters. While these were global, they were
960 never intended to be public API. They have been replaced with private
963 ** EBCDIC support is removed
965 There was an EBCDIC compile flag that altered some of the character
966 processing. It appeared that full EBCDIC support was never completed
967 and was unmaintained.
969 ** Compile-time warnings
971 Guile can warn about potentially unbound free variables. Pass the
972 -Wunbound-variable on the `guile-tools compile' command line, or add
973 `#:warnings '(unbound-variable)' to your `compile' or `compile-file'
976 Guile can also warn when you pass the wrong number of arguments to a
977 procedure, with -Warity-mismatch, or `arity-mismatch' in the
978 `#:warnings' as above.
980 Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to
981 warn about unused local or global (top-level) variables.
983 ** A new `memoize-symbol' evaluator trap has been added.
985 This trap can be used for efficiently implementing a Scheme code
988 ** Duplicate bindings among used modules are resolved lazily.
990 This slightly improves program startup times.
992 ** New thread cancellation and thread cleanup API
994 See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'.
996 ** GOOPS dispatch in scheme
998 As an implementation detail, GOOPS dispatch is no longer implemented by
999 special evaluator bytecodes, but rather directly via a Scheme function
1000 associated with an applicable struct. There is some VM support for the
1001 underlying primitives, like `class-of'.
1003 This change will in the future allow users to customize generic function
1004 dispatch without incurring a performance penalty, and allow us to
1005 implement method combinations.
1009 GOOPS had a number of concepts that were relevant to the days of Tcl,
1010 but not any more: operators and entities, mainly. These objects were
1011 never documented, and it is unlikely that they were ever used. Operators
1012 were a kind of generic specific to the Tcl support. Entities were
1013 applicable structures, but were unusable; entities will come back in the
1014 next alpha release, but with a less stupid name.
1016 ** Applicable struct support
1018 One may now make structs from Scheme that may be applied as procedures.
1019 To do so, make a struct whose vtable is `<applicable-struct-vtable>'.
1020 That struct will be the vtable of your applicable structs; instances of
1021 that new struct are assumed to have the procedure in their first slot.
1022 `<applicable-struct-vtable>' is like Common Lisp's
1023 `funcallable-standard-class'. Likewise there is
1024 `<applicable-struct-with-setter-vtable>', which looks for the setter in
1025 the second slot. This needs to be better documented.
1027 ** New struct slot allocation: "hidden"
1029 A hidden slot is readable and writable, but will not be initialized by a
1030 call to make-struct. For example in your layout you would say "ph"
1031 instead of "pw". Hidden slots are useful for adding new slots to a
1032 vtable without breaking existing invocations to make-struct.
1034 ** eqv? not a generic
1036 One used to be able to extend `eqv?' as a primitive-generic, but no
1037 more. Because `eqv?' is in the expansion of `case' (via `memv'), which
1038 should be able to compile to static dispatch tables, it doesn't make
1039 sense to allow extensions that would subvert this optimization.
1041 ** `inet-ntop' and `inet-pton' are always available.
1043 Guile now use a portable implementation of `inet_pton'/`inet_ntop', so
1044 there is no more need to use `inet-aton'/`inet-ntoa'. The latter
1045 functions are deprecated.
1047 ** Fast bit operations.
1049 The bit-twiddling operations `ash', `logand', `logior', and `logxor' now
1050 have dedicated bytecodes. Guile is not just for symbolic computation,
1051 it's for number crunching too.
1053 ** Faster SRFI-9 record access
1055 SRFI-9 records are now implemented directly on top of Guile's structs,
1056 and their accessors are defined in such a way that normal call-sites
1057 inline to special VM opcodes, while still allowing for the general case
1058 (e.g. passing a record accessor to `apply').
1060 ** R6RS block comment support
1062 Guile now supports R6RS nested block comments. The start of a comment is
1063 marked with `#|', and the end with `|#'.
1065 ** `guile-2' cond-expand feature
1067 To test if your code is running under Guile 2.0 (or its alpha releases),
1068 test for the `guile-2' cond-expand feature. Like this:
1070 (cond-expand (guile-2 (eval-when (compile)
1071 ;; This must be evaluated at compile time.
1072 (fluid-set! current-reader my-reader)))
1074 ;; Earlier versions of Guile do not have a
1075 ;; separate compilation phase.
1076 (fluid-set! current-reader my-reader)))
1078 ** Fix bad interaction between `false-if-exception' and stack-call.
1080 Exceptions thrown by `false-if-exception' were erronously causing the
1081 stack to be saved, causing later errors to show the incorrectly-saved
1082 backtrace. This has been fixed.
1084 ** New global variables: %load-compiled-path, %load-compiled-extensions
1086 These are analogous to %load-path and %load-extensions.
1088 ** New procedure, `make-promise'
1090 `(make-promise (lambda () foo))' is equivalent to `(delay foo)'.
1092 ** `defined?' may accept a module as its second argument
1094 Previously it only accepted internal structures from the evaluator.
1096 ** New entry into %guile-build-info: `ccachedir'
1098 ** Fix bug in `module-bound?'.
1100 `module-bound?' was returning true if a module did have a local
1101 variable, but one that was unbound, but another imported module bound
1102 the variable. This was an error, and was fixed.
1104 ** `(ice-9 syncase)' has been deprecated.
1106 As syntax-case is available by default, importing `(ice-9 syncase)' has
1107 no effect, and will trigger a deprecation warning.
1109 ** New readline history functions
1111 The (ice-9 readline) module now provides add-history, read-history,
1112 write-history and clear-history, which wrap the corresponding GNU
1113 History library functions.
1115 ** Removed deprecated uniform array procedures:
1116 dimensions->uniform-array, list->uniform-array, array-prototype
1118 Instead, use make-typed-array, list->typed-array, or array-type,
1121 ** Last but not least, the `λ' macro can be used in lieu of `lambda'
1123 * Changes to the C interface
1125 ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector
1127 The semantics of `scm_gc_malloc ()' have been changed, in a
1128 backward-compatible way. A new allocation routine,
1129 `scm_gc_malloc_pointerless ()', was added.
1131 Libgc is a conservative GC, which we hope will make interaction with C
1132 code easier and less error-prone.
1134 ** New type definitions for `scm_t_intptr' and friends.
1136 `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX',
1137 `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now
1138 available to C. Have fun!
1140 ** The GH interface (deprecated in version 1.6, 2001) was removed.
1142 ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF
1144 This makes these internal functions technically not callable from
1147 ** Functions for handling `scm_option' now no longer require an argument
1148 indicating length of the `scm_t_option' array.
1150 ** Procedures-with-setters are now implemented using applicable structs
1152 From a user's perspective this doesn't mean very much. But if, for some
1153 odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or
1154 SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws
1157 ** Remove old evaluator closures
1159 There used to be ranges of typecodes allocated to interpreted data
1160 structures, but that it no longer the case, given that interpreted
1161 procedure are now just regular VM closures. As a result, there is a
1162 newly free tc3, and a number of removed macros. See the ChangeLog for
1165 ** Primitive procedures are now VM trampoline procedures
1167 It used to be that there were something like 12 different typecodes
1168 allocated to primitive procedures, each with its own calling convention.
1169 Now there is only one, the gsubr. This may affect user code if you were
1170 defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The
1171 solution is to switch to use scm_c_make_gsubr. This solution works well
1172 both with the old 1.8 and and with the current 1.9 branch.
1174 Guile's old evaluator used to have special cases for applying "gsubrs",
1175 primitive procedures with specified numbers of required, optional, and
1176 rest arguments. Now, however, Guile represents gsubrs as normal VM
1177 procedures, with appropriate bytecode to parse out the correct number of
1178 arguments, including optional and rest arguments, and then with a
1179 special bytecode to apply the gsubr.
1181 This allows primitive procedures to appear on the VM stack, allowing
1182 them to be accurately counted in profiles. Also they now have more
1183 debugging information attached to them -- their number of arguments, for
1184 example. In addition, the VM can completely inline the application
1185 mechanics, allowing for faster primitive calls.
1187 However there are some changes on the C level. There is no more
1188 `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as
1189 they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE',
1190 `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and
1191 `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS'
1192 `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'.
1194 Perhaps more significantly, `scm_c_make_subr',
1195 `scm_c_make_subr_with_generic', `scm_c_define_subr', and
1196 `scm_c_define_subr_with_generic'. They all operated on subr typecodes,
1197 and there are no more subr typecodes. Use the scm_c_make_gsubr family
1200 Normal users of gsubrs should not be affected, though, as the
1201 scm_c_make_gsubr family still is the correct way to create primitive
1204 ** Remove deprecated array C interfaces
1206 Removed the deprecated array functions `scm_i_arrayp',
1207 `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v',
1208 `scm_i_array_base', `scm_i_array_dims', and the deprecated macros
1209 `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM',
1210 `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'.
1212 ** Remove unused snarf macros
1214 `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1'
1215 are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead.
1217 ** Add foreign value wrapper
1219 Guile now has a datatype for aliasing "foreign" values, such as native
1220 long values. This should be useful for making a proper foreign function
1221 interface. Interested hackers should see libguile/foreign.h.
1223 ** New functions: `scm_call_n', `scm_c_run_hookn'
1225 `scm_call_n' applies to apply a function to an array of arguments.
1226 `scm_c_run_hookn' runs a hook with an array of arguments.
1228 ** Some SMOB types changed to have static typecodes
1230 Fluids, dynamic states, and hash tables used to be SMOB objects, but now
1231 they have statically allocated tc7 typecodes.
1233 ** Preparations for changing SMOB representation
1235 If things go right, we'll be changing the SMOB representation soon. To
1236 that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when
1237 the code meant SCM_SMOB_DATA_2(x); user code will need similar changes
1238 in the future. Code accessing SMOBs using SCM_CELL macros was never
1239 correct, but until now things still worked. Users should be aware of
1242 ** Changed invocation mechanics of applicable SMOBs
1244 Guile's old evaluator used to have special cases for applying SMOB
1245 objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM
1246 trampoline procedure for it, and use the normal mechanics to apply the
1247 trampoline. This simplifies procedure application in the normal,
1250 The upshot is that the mechanics used to apply a SMOB are different from
1251 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and
1252 `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now
1253 deprecated. Just use the scm_call_0 family of procedures.
1255 ** New C function: scm_module_public_interface
1257 This procedure corresponds to Scheme's `module-public-interface'.
1259 ** Undeprecate `scm_the_root_module ()'
1261 It's useful to be able to get the root module from C without doing a
1264 ** Inline vector allocation
1266 Instead of having vectors point out into the heap for their data, their
1267 data is now allocated inline to the vector object itself. The same is
1268 true for bytevectors, by default, though there is an indirection
1269 available which should allow for making a bytevector from an existing
1272 ** New struct constructors that don't involve making lists
1274 `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array
1275 constructors, respectively, for structs. You might find them useful.
1279 In Guile 1.8, there were debugging frames on the C stack. Now there is
1280 no more need to explicitly mark the stack in this way, because Guile has
1281 a VM stack that it knows how to walk, which simplifies the C API
1282 considerably. See the ChangeLog for details; the relevant interface is
1283 in libguile/stacks.h. The Scheme API has not been changed significantly.
1285 ** Removal of Guile's primitive object system.
1287 There were a number of pieces in `objects.[ch]' that tried to be a
1288 minimal object system, but were never documented, and were quickly
1289 obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object',
1290 `scm_make_subclass_object', `scm_metaclass_standard', and like symbols
1291 from objects.h are no more. In the very unlikely case in which these
1292 were useful to you, we urge you to contact guile-devel.
1296 Actually the future is still in the state that it was, is, and ever
1297 shall be, Amen, except that `futures.c' and `futures.h' are no longer a
1298 part of it. These files were experimental, never compiled, and would be
1299 better implemented in Scheme anyway. In the future, that is.
1301 ** Deprecate trampolines
1303 There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and
1304 so on. The point was to do some precomputation on the type of the
1305 procedure, then return a specialized "call" procedure. However this
1306 optimization wasn't actually an optimization, so it is now deprecated.
1307 Just use `scm_call_0', etc instead.
1309 ** Better support for Lisp `nil'.
1311 The bit representation of `nil' has been tweaked so that it is now very
1312 efficient to check e.g. if a value is equal to Scheme's end-of-list or
1313 Lisp's nil. Additionally there are a heap of new, specific predicates
1314 like scm_is_null_or_nil. Probably in the future we will #define
1315 scm_is_null to scm_is_null_or_nil.
1317 ** Support for static allocation of strings, symbols, and subrs.
1319 Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate
1320 much of their associated data as static variables, reducing Guile's
1323 ** `scm_stat' has an additional argument, `exception_on_error'
1324 ** `scm_primitive_load_path' has an additional argument `exception_on_not_found'
1326 ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type
1328 Previously they would use the `off_t' type, which is fragile since its
1329 definition depends on the application's value for `_FILE_OFFSET_BITS'.
1331 ** The `long_long' C type, deprecated in 1.8, has been removed
1333 ** Removed deprecated uniform array procedures: scm_make_uve,
1334 scm_array_prototype, scm_list_to_uniform_array,
1335 scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref,
1336 scm_ra_set_contp, scm_aind, scm_raprin1
1338 These functions have been deprecated since early 2005.
1340 * Changes to the distribution
1342 ** Guile's license is now LGPLv3+
1344 In other words the GNU Lesser General Public License, version 3 or
1345 later (at the discretion of each person that chooses to redistribute
1348 ** GOOPS documentation folded into Guile reference manual
1350 GOOPS, Guile's object system, used to be documented in separate manuals.
1351 This content is now included in Guile's manual directly.
1353 ** `guile-config' will be deprecated in favor of `pkg-config'
1355 `guile-config' has been rewritten to get its information from
1356 `pkg-config', so this should be a transparent change. Note however that
1357 guile.m4 has yet to be modified to call pkg-config instead of
1360 ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc'
1362 Programs that use `pkg-config' to find Guile or one of its Autoconf
1363 macros should now require `guile-2.0' instead of `guile-1.8'.
1365 ** New installation directory: $(pkglibdir)/1.9/ccache
1367 If $(libdir) is /usr/lib, for example, Guile will install its .go files
1368 to /usr/lib/guile/1.9/ccache. These files are architecture-specific.
1370 ** Dynamically loadable extensions may be placed in a Guile-specific path
1372 Before, Guile only searched the system library paths for extensions
1373 (e.g. /usr/lib), which meant that the names of Guile extensions had to
1374 be globally unique. Installing them to a Guile-specific extensions
1375 directory is cleaner. Use `pkg-config --variable=extensionsdir
1376 guile-2.0' to get the location of the extensions directory.
1378 ** New dependency: libgc
1380 See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information.
1382 ** New dependency: GNU libunistring
1384 See http://www.gnu.org/software/libunistring/, for more information. Our
1385 Unicode support uses routines from libunistring.
1389 Changes in 1.8.8 (since 1.8.7)
1393 ** Fix possible buffer overruns when parsing numbers
1394 ** Avoid clash with system setjmp/longjmp on IA64
1395 ** Fix `wrong type arg' exceptions with IPv6 addresses
1398 Changes in 1.8.7 (since 1.8.6)
1400 * New modules (see the manual for details)
1402 ** `(srfi srfi-98)', an interface to access environment variables
1406 ** Fix compilation with `--disable-deprecated'
1407 ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault
1408 ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion
1409 ** Fix build problem when scm_t_timespec is different from struct timespec
1410 ** Fix build when compiled with -Wundef -Werror
1411 ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64)
1412 ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3)
1413 ** With GCC, always compile with `-mieee' on `alpha*' and `sh*'
1414 ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130)
1415 ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters
1416 ** Fix reading of complex numbers where both parts are inexact decimals
1418 ** Allow @ macro to work with (ice-9 syncase)
1420 Previously, use of the @ macro in a module whose code is being
1421 transformed by (ice-9 syncase) would cause an "Invalid syntax" error.
1422 Now it works as you would expect (giving the value of the specified
1425 ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865)
1428 Changes in 1.8.6 (since 1.8.5)
1430 * New features (see the manual for details)
1432 ** New convenience function `scm_c_symbol_length ()'
1434 ** Single stepping through code from Emacs
1436 When you use GDS to evaluate Scheme code from Emacs, you can now use
1437 `C-u' to indicate that you want to single step through that code. See
1438 `Evaluating Scheme Code' in the manual for more details.
1440 ** New "guile(1)" man page!
1442 * Changes to the distribution
1444 ** Automake's `AM_MAINTAINER_MODE' is no longer used
1446 Thus, the `--enable-maintainer-mode' configure option is no longer
1447 available: Guile is now always configured in "maintainer mode".
1449 ** `ChangeLog' files are no longer updated
1451 Instead, changes are detailed in the version control system's logs. See
1452 the top-level `ChangeLog' files for details.
1457 ** `symbol->string' now returns a read-only string, as per R5RS
1458 ** Fix incorrect handling of the FLAGS argument of `fold-matches'
1459 ** `guile-config link' now prints `-L$libdir' before `-lguile'
1460 ** Fix memory corruption involving GOOPS' `class-redefinition'
1461 ** Fix possible deadlock in `mutex-lock'
1462 ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro)
1463 ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction)
1464 ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r')
1465 ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules")
1466 ** Fix misleading output from `(help rationalize)'
1467 ** Fix build failure on Debian hppa architecture (bad stack growth detection)
1468 ** Fix `gcd' when called with a single, negative argument.
1469 ** Fix `Stack overflow' errors seen when building on some platforms
1470 ** Fix bug when `scm_with_guile ()' was called several times from the
1472 ** The handler of SRFI-34 `with-exception-handler' is now invoked in the
1473 dynamic environment of the call to `raise'
1474 ** Fix potential deadlock in `make-struct'
1475 ** Fix compilation problem with libltdl from Libtool 2.2.x
1476 ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string
1479 Changes in 1.8.5 (since 1.8.4)
1481 * Infrastructure changes
1483 ** Guile repository switched from CVS to Git
1485 The new repository can be accessed using
1486 "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at
1487 http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details.
1489 ** Add support for `pkg-config'
1491 See "Autoconf Support" in the manual for details.
1493 * New modules (see the manual for details)
1497 * New features (see the manual for details)
1499 ** New `postfix' read option, for SRFI-88 keyword syntax
1500 ** Some I/O primitives have been inlined, which improves I/O performance
1501 ** New object-based traps infrastructure
1503 This is a GOOPS-based infrastructure that builds on Guile's low-level
1504 evaluator trap calls and facilitates the development of debugging
1505 features like single-stepping, breakpoints, tracing and profiling.
1506 See the `Traps' node of the manual for details.
1508 ** New support for working on Guile code from within Emacs
1510 Guile now incorporates the `GDS' library (previously distributed
1511 separately) for working on Guile code from within Emacs. See the
1512 `Using Guile In Emacs' node of the manual for details.
1516 ** `scm_add_slot ()' no longer segfaults (fixes bug #22369)
1517 ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)'
1519 Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))'
1520 would trigger an unbound variable error for `match:andmap'.
1522 ** `(oop goops describe)' now properly provides the `describe' feature
1523 ** Fixed `args-fold' from `(srfi srfi-37)'
1525 Previously, parsing short option names of argument-less options would
1526 lead to a stack overflow.
1528 ** `(srfi srfi-35)' is now visible through `cond-expand'
1529 ** Fixed type-checking for the second argument of `eval'
1530 ** Fixed type-checking for SRFI-1 `partition'
1531 ** Fixed `struct-ref' and `struct-set!' on "light structs"
1532 ** Honor struct field access rights in GOOPS
1533 ** Changed the storage strategy of source properties, which fixes a deadlock
1534 ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later
1535 ** Fixed build issue for GNU/Linux on IA64
1536 ** Fixed build issues on NetBSD 1.6
1537 ** Fixed build issue on Solaris 2.10 x86_64
1538 ** Fixed build issue with DEC/Compaq/HP's compiler
1539 ** Fixed `scm_from_complex_double' build issue on FreeBSD
1540 ** Fixed `alloca' build issue on FreeBSD 6
1541 ** Removed use of non-portable makefile constructs
1542 ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation
1543 ** Make sure all tests honor `$TMPDIR'
1546 Changes in 1.8.4 (since 1.8.3)
1550 ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader
1551 ** Fixed a segmentation fault which occurred when displaying the
1552 backtrace of a stack with a promise object (made by `delay') in it.
1553 ** Make `accept' leave guile mode while blocking
1554 ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument
1555 ** Fixed a build problem on AIX (use of func_data identifier)
1556 ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was
1557 called with an associator proc that returns neither a pair nor #f.
1558 ** Secondary threads now always return a valid module for (current-module).
1559 ** Avoid MacOS build problems caused by incorrect combination of "64"
1560 system and library calls.
1561 ** `guile-snarf' now honors `$TMPDIR'
1562 ** `guile-config compile' now reports CPPFLAGS used at compile-time
1563 ** Fixed build with Sun Studio (Solaris 9)
1564 ** Fixed wrong-type-arg errors when creating zero length SRFI-4
1565 uniform vectors on AIX.
1566 ** Fixed a deadlock that occurs upon GC with multiple threads.
1567 ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I)
1568 ** Fixed autotool-derived build problems on AIX 6.1.
1569 ** Fixed NetBSD/alpha support
1570 ** Fixed MacOS build problem caused by use of rl_get_keymap(_name)
1572 * New modules (see the manual for details)
1576 * Documentation fixes and improvements
1578 ** Removed premature breakpoint documentation
1580 The features described are not available in the series of 1.8.x
1581 releases, so the documentation was misleading and has been removed.
1583 ** More about Guile's default *random-state* variable
1585 ** GOOPS: more about how to use `next-method'
1587 * Changes to the distribution
1589 ** Corrected a few files that referred incorrectly to the old GPL + special exception licence
1591 In fact Guile since 1.8.0 has been licensed with the GNU Lesser
1592 General Public License, and the few incorrect files have now been
1593 fixed to agree with the rest of the Guile distribution.
1595 ** Removed unnecessary extra copies of COPYING*
1597 The distribution now contains a single COPYING.LESSER at its top level.
1600 Changes in 1.8.3 (since 1.8.2)
1602 * New modules (see the manual for details)
1609 ** The `(ice-9 slib)' module now works as expected
1610 ** Expressions like "(set! 'x #t)" no longer yield a crash
1611 ** Warnings about duplicate bindings now go to stderr
1612 ** A memory leak in `make-socket-address' was fixed
1613 ** Alignment issues (e.g., on SPARC) in network routines were fixed
1614 ** A threading issue that showed up at least on NetBSD was fixed
1615 ** Build problems on Solaris and IRIX fixed
1617 * Implementation improvements
1619 ** The reader is now faster, which reduces startup time
1620 ** Procedures returned by `record-accessor' and `record-modifier' are faster
1623 Changes in 1.8.2 (since 1.8.1):
1625 * New procedures (see the manual for details)
1627 ** set-program-arguments
1630 * Incompatible changes
1632 ** The body of a top-level `define' no longer sees the binding being created
1634 In a top-level `define', the binding being created is no longer visible
1635 from the `define' body. This breaks code like
1636 "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now
1637 unbound in the body. However, such code was not R5RS-compliant anyway,
1642 ** Fractions were not `equal?' if stored in unreduced form.
1643 (A subtle problem, since printing a value reduced it, making it work.)
1644 ** srfi-60 `copy-bit' failed on 64-bit systems
1645 ** "guile --use-srfi" option at the REPL can replace core functions
1646 (Programs run with that option were ok, but in the interactive REPL
1647 the core bindings got priority, preventing SRFI replacements or
1649 ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg
1650 ** `kill' on mingw throws an error for a PID other than oneself
1651 ** Procedure names are attached to procedure-with-setters
1652 ** Array read syntax works with negative lower bound
1653 ** `array-in-bounds?' fix if an array has different lower bounds on each index
1654 ** `*' returns exact 0 for "(* inexact 0)"
1655 This follows what it always did for "(* 0 inexact)".
1656 ** SRFI-19: Value returned by `(current-time time-process)' was incorrect
1657 ** SRFI-19: `date->julian-day' did not account for timezone offset
1658 ** `ttyname' no longer crashes when passed a non-tty argument
1659 ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address
1660 ** Small memory leaks have been fixed in `make-fluid' and `add-history'
1661 ** GOOPS: Fixed a bug in `method-more-specific?'
1662 ** Build problems on Solaris fixed
1663 ** Build problems on HP-UX IA64 fixed
1664 ** Build problems on MinGW fixed
1667 Changes in 1.8.1 (since 1.8.0):
1669 * LFS functions are now used to access 64-bit files on 32-bit systems.
1671 * New procedures (see the manual for details)
1673 ** primitive-_exit - [Scheme] the-root-module
1674 ** scm_primitive__exit - [C]
1675 ** make-completion-function - [Scheme] (ice-9 readline)
1676 ** scm_c_locale_stringn_to_number - [C]
1677 ** scm_srfi1_append_reverse [C]
1678 ** scm_srfi1_append_reverse_x [C]
1686 ** Build problems have been fixed on MacOS, SunOS, and QNX.
1688 ** `strftime' fix sign of %z timezone offset.
1690 ** A one-dimensional array can now be 'equal?' to a vector.
1692 ** Structures, records, and SRFI-9 records can now be compared with `equal?'.
1694 ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'.
1696 ** `record-accessor' and `record-modifier' now have strict type checks.
1698 Record accessor and modifier procedures now throw an error if the
1699 record type of the record they're given is not the type expected.
1700 (Previously accessors returned #f and modifiers silently did nothing).
1702 ** It is now OK to use both autoload and use-modules on a given module.
1704 ** `apply' checks the number of arguments more carefully on "0 or 1" funcs.
1706 Previously there was no checking on primatives like make-vector that
1707 accept "one or two" arguments. Now there is.
1709 ** The srfi-1 assoc function now calls its equality predicate properly.
1711 Previously srfi-1 assoc would call the equality predicate with the key
1712 last. According to the SRFI, the key should be first.
1714 ** A bug in n-par-for-each and n-for-each-par-map has been fixed.
1716 ** The array-set! procedure no longer segfaults when given a bit vector.
1718 ** Bugs in make-shared-array have been fixed.
1720 ** string<? and friends now follow char<? etc order on 8-bit chars.
1722 ** The format procedure now handles inf and nan values for ~f correctly.
1724 ** exact->inexact should no longer overflow when given certain large fractions.
1726 ** srfi-9 accessor and modifier procedures now have strict record type checks.
1728 This matches the srfi-9 specification.
1730 ** (ice-9 ftw) procedures won't ignore different files with same inode number.
1732 Previously the (ice-9 ftw) procedures would ignore any file that had
1733 the same inode number as a file they had already seen, even if that
1734 file was on a different device.
1737 Changes in 1.8.0 (changes since the 1.6.x series):
1739 * Changes to the distribution
1741 ** Guile is now licensed with the GNU Lesser General Public License.
1743 ** The manual is now licensed with the GNU Free Documentation License.
1745 ** Guile now requires GNU MP (http://swox.com/gmp).
1747 Guile now uses the GNU MP library for arbitrary precision arithmetic.
1749 ** Guile now has separate private and public configuration headers.
1751 That is, things like HAVE_STRING_H no longer leak from Guile's
1754 ** Guile now provides and uses an "effective" version number.
1756 Guile now provides scm_effective_version and effective-version
1757 functions which return the "effective" version number. This is just
1758 the normal full version string without the final micro-version number,
1759 so the current effective-version is "1.8". The effective version
1760 should remain unchanged during a stable series, and should be used for
1761 items like the versioned share directory name
1762 i.e. /usr/share/guile/1.8.
1764 Providing an unchanging version number during a stable release for
1765 things like the versioned share directory can be particularly
1766 important for Guile "add-on" packages, since it provides a directory
1767 that they can install to that won't be changed out from under them
1768 with each micro release during a stable series.
1770 ** Thread implementation has changed.
1772 When you configure "--with-threads=null", you will get the usual
1773 threading API (call-with-new-thread, make-mutex, etc), but you can't
1774 actually create new threads. Also, "--with-threads=no" is now
1775 equivalent to "--with-threads=null". This means that the thread API
1776 is always present, although you might not be able to create new
1779 When you configure "--with-threads=pthreads" or "--with-threads=yes",
1780 you will get threads that are implemented with the portable POSIX
1781 threads. These threads can run concurrently (unlike the previous
1782 "coop" thread implementation), but need to cooperate for things like
1785 The default is "pthreads", unless your platform doesn't have pthreads,
1786 in which case "null" threads are used.
1788 See the manual for details, nodes "Initialization", "Multi-Threading",
1789 "Blocking", and others.
1791 ** There is the new notion of 'discouraged' features.
1793 This is a milder form of deprecation.
1795 Things that are discouraged should not be used in new code, but it is
1796 OK to leave them in old code for now. When a discouraged feature is
1797 used, no warning message is printed like there is for 'deprecated'
1798 features. Also, things that are merely discouraged are nevertheless
1799 implemented efficiently, while deprecated features can be very slow.
1801 You can omit discouraged features from libguile by configuring it with
1802 the '--disable-discouraged' option.
1804 ** Deprecation warnings can be controlled at run-time.
1806 (debug-enable 'warn-deprecated) switches them on and (debug-disable
1807 'warn-deprecated) switches them off.
1809 ** Support for SRFI 61, extended cond syntax for multiple values has
1812 This SRFI is always available.
1814 ** Support for require-extension, SRFI-55, has been added.
1816 The SRFI-55 special form `require-extension' has been added. It is
1817 available at startup, and provides a portable way to load Scheme
1818 extensions. SRFI-55 only requires support for one type of extension,
1819 "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1
1822 ** New module (srfi srfi-26) provides support for `cut' and `cute'.
1824 The (srfi srfi-26) module is an implementation of SRFI-26 which
1825 provides the `cut' and `cute' syntax. These may be used to specialize
1826 parameters without currying.
1828 ** New module (srfi srfi-31)
1830 This is an implementation of SRFI-31 which provides a special form
1831 `rec' for recursive evaluation.
1833 ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have
1834 been merged with the core, making their functionality always
1837 The modules are still available, tho, and you could use them together
1838 with a renaming import, for example.
1840 ** Guile no longer includes its own version of libltdl.
1842 The official version is good enough now.
1844 ** The --enable-htmldoc option has been removed from 'configure'.
1846 Support for translating the documentation into HTML is now always
1847 provided. Use 'make html'.
1849 ** New module (ice-9 serialize):
1851 (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you
1852 don't trust the thread safety of most of your program, but where you
1853 have some section(s) of code which you consider can run in parallel to
1854 other sections. See ice-9/serialize.scm for more information.
1856 ** The configure option '--disable-arrays' has been removed.
1858 Support for arrays and uniform numeric arrays is now always included
1861 * Changes to the stand-alone interpreter
1863 ** New command line option `-L'.
1865 This option adds a directory to the front of the load path.
1867 ** New command line option `--no-debug'.
1869 Specifying `--no-debug' on the command line will keep the debugging
1870 evaluator turned off, even for interactive sessions.
1872 ** User-init file ~/.guile is now loaded with the debugging evaluator.
1874 Previously, the normal evaluator would have been used. Using the
1875 debugging evaluator gives better error messages.
1877 ** The '-e' option now 'read's its argument.
1879 This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to
1880 be used with '-e'. For example, you can now write a script like
1883 exec guile -e '(@ (demo) main)' -s "$0" "$@"
1886 (define-module (demo)
1890 (format #t "Demo: ~a~%" args))
1893 * Changes to Scheme functions and syntax
1895 ** Guardians have changed back to their original semantics
1897 Guardians now behave like described in the paper by Dybvig et al. In
1898 particular, they no longer make guarantees about the order in which
1899 they return objects, and they can no longer be greedy.
1901 They no longer drop cyclic data structures.
1903 The C function scm_make_guardian has been changed incompatibly and no
1904 longer takes the 'greedy_p' argument.
1906 ** New function hashx-remove!
1908 This function completes the set of 'hashx' functions.
1910 ** The concept of dynamic roots has been factored into continuation
1911 barriers and dynamic states.
1913 Each thread has a current dynamic state that carries the values of the
1914 fluids. You can create and copy dynamic states and use them as the
1915 second argument for 'eval'. See "Fluids and Dynamic States" in the
1918 To restrict the influence that captured continuations can have on the
1919 control flow, you can errect continuation barriers. See "Continuation
1920 Barriers" in the manual.
1922 The function call-with-dynamic-root now essentially temporarily
1923 installs a new dynamic state and errects a continuation barrier.
1925 ** The default load path no longer includes "." at the end.
1927 Automatically loading modules from the current directory should not
1928 happen by default. If you want to allow it in a more controlled
1929 manner, set the environment variable GUILE_LOAD_PATH or the Scheme
1930 variable %load-path.
1932 ** The uniform vector and array support has been overhauled.
1934 It now complies with SRFI-4 and the weird prototype based uniform
1935 array creation has been deprecated. See the manual for more details.
1937 Some non-compatible changes have been made:
1938 - characters can no longer be stored into byte arrays.
1939 - strings and bit vectors are no longer considered to be uniform numeric
1941 - array-rank throws an error for non-arrays instead of returning zero.
1942 - array-ref does no longer accept non-arrays when no indices are given.
1944 There is the new notion of 'generalized vectors' and corresponding
1945 procedures like 'generalized-vector-ref'. Generalized vectors include
1946 strings, bitvectors, ordinary vectors, and uniform numeric vectors.
1948 Arrays use generalized vectors as their storage, so that you still
1949 have arrays of characters, bits, etc. However, uniform-array-read!
1950 and uniform-array-write can no longer read/write strings and
1953 ** There is now support for copy-on-write substrings, mutation-sharing
1954 substrings and read-only strings.
1956 Three new procedures are related to this: substring/shared,
1957 substring/copy, and substring/read-only. See the manual for more
1960 ** Backtraces will now highlight the value that caused the error.
1962 By default, these values are enclosed in "{...}", such as in this
1971 <unnamed port>:1:1: In procedure car in expression (car (quote a)):
1972 <unnamed port>:1:1: Wrong type (expecting pair): a
1973 ABORT: (wrong-type-arg)
1975 The prefix and suffix used for highlighting can be set via the two new
1976 printer options 'highlight-prefix' and 'highlight-suffix'. For
1977 example, putting this into ~/.guile will output the bad value in bold
1978 on an ANSI terminal:
1980 (print-set! highlight-prefix "\x1b[1m")
1981 (print-set! highlight-suffix "\x1b[22m")
1984 ** 'gettext' support for internationalization has been added.
1986 See the manual for details.
1988 ** New syntax '@' and '@@':
1990 You can now directly refer to variables exported from a module by
1993 (@ MODULE-NAME VARIABLE-NAME)
1995 For example (@ (ice-9 pretty-print) pretty-print) will directly access
1996 the pretty-print variable exported from the (ice-9 pretty-print)
1997 module. You don't need to 'use' that module first. You can also use
1998 '@' as a target of 'set!', as in (set! (@ mod var) val).
2000 The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@',
2001 but it can also access variables that have not been exported. It is
2002 intended only for kluges and temporary fixes and for debugging, not
2005 ** Keyword syntax has been made more disciplined.
2007 Previously, the name of a keyword was read as a 'token' but printed as
2008 a symbol. Now, it is read as a general Scheme datum which must be a
2019 ERROR: In expression (a b c):
2023 ERROR: Unbound variable: foo
2028 ERROR: Wrong type (expecting symbol): 12
2032 ERROR: Wrong type (expecting symbol): (a b c)
2036 ** The printing of symbols that might look like keywords can be
2039 The new printer option 'quote-keywordish-symbols' controls how symbols
2040 are printed that have a colon as their first or last character. The
2041 default now is to only quote a symbol with #{...}# when the read
2042 option 'keywords' is not '#f'. Thus:
2044 guile> (define foo (string->symbol ":foo"))
2045 guile> (read-set! keywords #f)
2048 guile> (read-set! keywords 'prefix)
2051 guile> (print-set! quote-keywordish-symbols #f)
2055 ** 'while' now provides 'break' and 'continue'
2057 break and continue were previously bound in a while loop, but not
2058 documented, and continue didn't quite work properly. The undocumented
2059 parameter to break which gave a return value for the while has been
2062 ** 'call-with-current-continuation' is now also available under the name
2065 ** The module system now checks for duplicate bindings.
2067 The module system now can check for name conflicts among imported
2070 The behavior can be controlled by specifying one or more 'duplicates'
2071 handlers. For example, to make Guile return an error for every name
2074 (define-module (foo)
2079 The new default behavior of the module system when a name collision
2080 has been detected is to
2082 1. Give priority to bindings marked as a replacement.
2083 2. Issue a warning (different warning if overriding core binding).
2084 3. Give priority to the last encountered binding (this corresponds to
2087 If you want the old behavior back without replacements or warnings you
2090 (default-duplicate-binding-handler 'last)
2092 to your .guile init file.
2094 ** New define-module option: :replace
2096 :replace works as :export, but, in addition, marks the binding as a
2099 A typical example is `format' in (ice-9 format) which is a replacement
2100 for the core binding `format'.
2102 ** Adding prefixes to imported bindings in the module system
2104 There is now a new :use-module option :prefix. It can be used to add
2105 a prefix to all imported bindings.
2107 (define-module (foo)
2108 :use-module ((bar) :prefix bar:))
2110 will import all bindings exported from bar, but rename them by adding
2113 ** Conflicting generic functions can be automatically merged.
2115 When two imported bindings conflict and they are both generic
2116 functions, the two functions can now be merged automatically. This is
2117 activated with the 'duplicates' handler 'merge-generics'.
2119 ** New function: effective-version
2121 Returns the "effective" version number. This is just the normal full
2122 version string without the final micro-version number. See "Changes
2123 to the distribution" above.
2125 ** New threading functions: parallel, letpar, par-map, and friends
2127 These are convenient ways to run calculations in parallel in new
2128 threads. See "Parallel forms" in the manual for details.
2130 ** New function 'try-mutex'.
2132 This function will attempt to lock a mutex but will return immediately
2133 instead of blocking and indicate failure.
2135 ** Waiting on a condition variable can have a timeout.
2137 The function 'wait-condition-variable' now takes a third, optional
2138 argument that specifies the point in time where the waiting should be
2141 ** New function 'broadcast-condition-variable'.
2143 ** New functions 'all-threads' and 'current-thread'.
2145 ** Signals and system asyncs work better with threads.
2147 The function 'sigaction' now takes a fourth, optional, argument that
2148 specifies the thread that the handler should run in. When the
2149 argument is omitted, the handler will run in the thread that called
2152 Likewise, 'system-async-mark' takes a second, optional, argument that
2153 specifies the thread that the async should run in. When it is
2154 omitted, the async will run in the thread that called
2155 'system-async-mark'.
2157 C code can use the new functions scm_sigaction_for_thread and
2158 scm_system_async_mark_for_thread to pass the new thread argument.
2160 When a thread blocks on a mutex, a condition variable or is waiting
2161 for IO to be possible, it will still execute system asyncs. This can
2162 be used to interrupt such a thread by making it execute a 'throw', for
2165 ** The function 'system-async' is deprecated.
2167 You can now pass any zero-argument procedure to 'system-async-mark'.
2168 The function 'system-async' will just return its argument unchanged
2171 ** New functions 'call-with-blocked-asyncs' and
2172 'call-with-unblocked-asyncs'
2174 The expression (call-with-blocked-asyncs PROC) will call PROC and will
2175 block execution of system asyncs for the current thread by one level
2176 while PROC runs. Likewise, call-with-unblocked-asyncs will call a
2177 procedure and will unblock the execution of system asyncs by one
2178 level for the current thread.
2180 Only system asyncs are affected by these functions.
2182 ** The functions 'mask-signals' and 'unmask-signals' are deprecated.
2184 Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs'
2185 instead. Those functions are easier to use correctly and can be
2188 ** New function 'unsetenv'.
2190 ** New macro 'define-syntax-public'.
2192 It works like 'define-syntax' and also exports the defined macro (but
2195 ** There is support for Infinity and NaNs.
2197 Following PLT Scheme, Guile can now work with infinite numbers, and
2200 There is new syntax for numbers: "+inf.0" (infinity), "-inf.0"
2201 (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as
2202 "+nan.0"). These numbers are inexact and have no exact counterpart.
2204 Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the
2205 sign of the dividend. The infinities are integers, and they answer #t
2206 for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is
2207 not '=' to itself, but '+nan.0' is 'eqv?' to itself.
2218 ERROR: Numerical overflow
2220 Two new predicates 'inf?' and 'nan?' can be used to test for the
2223 ** Inexact zero can have a sign.
2225 Guile can now distinguish between plus and minus inexact zero, if your
2226 platform supports this, too. The two zeros are equal according to
2227 '=', but not according to 'eqv?'. For example
2238 ** Guile now has exact rationals.
2240 Guile can now represent fractions such as 1/3 exactly. Computing with
2241 them is also done exactly, of course:
2246 ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers
2247 for exact arguments.
2249 For example: (floor 2) now returns an exact 2 where in the past it
2250 returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1.
2252 ** inexact->exact no longer returns only integers.
2254 Without exact rationals, the closest exact number was always an
2255 integer, but now inexact->exact returns the fraction that is exactly
2256 equal to a floating point number. For example:
2258 (inexact->exact 1.234)
2259 => 694680242521899/562949953421312
2261 When you want the old behavior, use 'round' explicitly:
2263 (inexact->exact (round 1.234))
2266 ** New function 'rationalize'.
2268 This function finds a simple fraction that is close to a given real
2269 number. For example (and compare with inexact->exact above):
2271 (rationalize (inexact->exact 1.234) 1/2000)
2274 Note that, as required by R5RS, rationalize returns only then an exact
2275 result when both its arguments are exact.
2277 ** 'odd?' and 'even?' work also for inexact integers.
2279 Previously, (odd? 1.0) would signal an error since only exact integers
2280 were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0)
2281 returns #f and (odd? 1.5) signals an error.
2283 ** Guile now has uninterned symbols.
2285 The new function 'make-symbol' will return an uninterned symbol. This
2286 is a symbol that is unique and is guaranteed to remain unique.
2287 However, uninterned symbols can not yet be read back in.
2289 Use the new function 'symbol-interned?' to check whether a symbol is
2292 ** pretty-print has more options.
2294 The function pretty-print from the (ice-9 pretty-print) module can now
2295 also be invoked with keyword arguments that control things like
2296 maximum output width. See the manual for details.
2298 ** Variables have no longer a special behavior for `equal?'.
2300 Previously, comparing two variables with `equal?' would recursivly
2301 compare their values. This is no longer done. Variables are now only
2302 `equal?' if they are `eq?'.
2304 ** `(begin)' is now valid.
2306 You can now use an empty `begin' form. It will yield #<unspecified>
2307 when evaluated and simply be ignored in a definition context.
2309 ** Deprecated: procedure->macro
2311 Change your code to use 'define-macro' or r5rs macros. Also, be aware
2312 that macro expansion will not be done during evaluation, but prior to
2315 ** Soft ports now allow a `char-ready?' procedure
2317 The vector argument to `make-soft-port' can now have a length of
2318 either 5 or 6. (Previously the length had to be 5.) The optional 6th
2319 element is interpreted as an `input-waiting' thunk -- i.e. a thunk
2320 that returns the number of characters that can be read immediately
2321 without the soft port blocking.
2323 ** Deprecated: undefine
2325 There is no replacement for undefine.
2327 ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol
2328 have been discouraged.
2330 They are relics from a time where a keyword like #:foo was used
2331 directly as a Tcl option "-foo" and thus keywords were internally
2332 stored as a symbol with a starting dash. We now store a symbol
2335 Use symbol->keyword and keyword->symbol instead.
2337 ** The `cheap' debug option is now obsolete
2339 Evaluator trap calls are now unconditionally "cheap" - in other words,
2340 they pass a debug object to the trap handler rather than a full
2341 continuation. The trap handler code can capture a full continuation
2342 by using `call-with-current-continuation' in the usual way, if it so
2345 The `cheap' option is retained for now so as not to break existing
2346 code which gets or sets it, but setting it now has no effect. It will
2347 be removed in the next major Guile release.
2349 ** Evaluator trap calls now support `tweaking'
2351 `Tweaking' means that the trap handler code can modify the Scheme
2352 expression that is about to be evaluated (in the case of an
2353 enter-frame trap) or the value that is being returned (in the case of
2354 an exit-frame trap). The trap handler code indicates that it wants to
2355 do this by returning a pair whose car is the symbol 'instead and whose
2356 cdr is the modified expression or return value.
2358 * Changes to the C interface
2360 ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer
2361 take a 'delete' function argument.
2363 This argument makes no sense since the delete function is used to
2364 remove a pair from an alist, and this must not be configurable.
2366 This is an incompatible change.
2368 ** The GH interface is now subject to the deprecation mechanism
2370 The GH interface has been deprecated for quite some time but now it is
2371 actually removed from Guile when it is configured with
2372 --disable-deprecated.
2374 See the manual "Transitioning away from GH" for more information.
2376 ** A new family of functions for converting between C values and
2377 Scheme values has been added.
2379 These functions follow a common naming scheme and are designed to be
2380 easier to use, thread-safe and more future-proof than the older
2383 - int scm_is_* (...)
2385 These are predicates that return a C boolean: 1 or 0. Instead of
2386 SCM_NFALSEP, you can now use scm_is_true, for example.
2388 - <type> scm_to_<type> (SCM val, ...)
2390 These are functions that convert a Scheme value into an appropriate
2391 C value. For example, you can use scm_to_int to safely convert from
2394 - SCM scm_from_<type> (<type> val, ...)
2396 These functions convert from a C type to a SCM value; for example,
2397 scm_from_int for ints.
2399 There is a huge number of these functions, for numbers, strings,
2400 symbols, vectors, etc. They are documented in the reference manual in
2401 the API section together with the types that they apply to.
2403 ** New functions for dealing with complex numbers in C have been added.
2405 The new functions are scm_c_make_rectangular, scm_c_make_polar,
2406 scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle.
2407 They work like scm_make_rectangular etc but take or return doubles
2410 ** The function scm_make_complex has been discouraged.
2412 Use scm_c_make_rectangular instead.
2414 ** The INUM macros have been deprecated.
2416 A lot of code uses these macros to do general integer conversions,
2417 although the macros only work correctly with fixnums. Use the
2418 following alternatives.
2420 SCM_INUMP -> scm_is_integer or similar
2421 SCM_NINUMP -> !scm_is_integer or similar
2422 SCM_MAKINUM -> scm_from_int or similar
2423 SCM_INUM -> scm_to_int or similar
2425 SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will
2426 do the validating for you.
2428 ** The scm_num2<type> and scm_<type>2num functions and scm_make_real
2429 have been discouraged.
2431 Use the newer scm_to_<type> and scm_from_<type> functions instead for
2432 new code. The functions have been discouraged since they don't fit
2435 ** The 'boolean' macros SCM_FALSEP etc have been discouraged.
2437 They have strange names, especially SCM_NFALSEP, and SCM_BOOLP
2438 evaluates its argument twice. Use scm_is_true, etc. instead for new
2441 ** The macro SCM_EQ_P has been discouraged.
2443 Use scm_is_eq for new code, which fits better into the naming
2446 ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have
2449 Use the function scm_is_pair or scm_is_null instead.
2451 ** The functions scm_round and scm_truncate have been deprecated and
2452 are now available as scm_c_round and scm_c_truncate, respectively.
2454 These functions occupy the names that scm_round_number and
2455 scm_truncate_number should have.
2457 ** The functions scm_c_string2str, scm_c_substring2str, and
2458 scm_c_symbol2str have been deprecated.
2460 Use scm_to_locale_stringbuf or similar instead, maybe together with
2463 ** New functions scm_c_make_string, scm_c_string_length,
2464 scm_c_string_ref, scm_c_string_set_x, scm_c_substring,
2465 scm_c_substring_shared, scm_c_substring_copy.
2467 These are like scm_make_string, scm_length, etc. but are slightly
2468 easier to use from C.
2470 ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH,
2471 SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated.
2473 They export too many assumptions about the implementation of strings
2474 and symbols that are no longer true in the presence of
2475 mutation-sharing substrings and when Guile switches to some form of
2478 When working with strings, it is often best to use the normal string
2479 functions provided by Guile, such as scm_c_string_ref,
2480 scm_c_string_set_x, scm_string_append, etc. Be sure to look in the
2481 manual since many more such functions are now provided than
2484 When you want to convert a SCM string to a C string, use the
2485 scm_to_locale_string function or similar instead. For symbols, use
2486 scm_symbol_to_string and then work with that string. Because of the
2487 new string representation, scm_symbol_to_string does not need to copy
2488 and is thus quite efficient.
2490 ** Some string, symbol and keyword functions have been discouraged.
2492 They don't fit into the uniform naming scheme and are not explicit
2493 about the character encoding.
2495 Replace according to the following table:
2497 scm_allocate_string -> scm_c_make_string
2498 scm_take_str -> scm_take_locale_stringn
2499 scm_take0str -> scm_take_locale_string
2500 scm_mem2string -> scm_from_locale_stringn
2501 scm_str2string -> scm_from_locale_string
2502 scm_makfrom0str -> scm_from_locale_string
2503 scm_mem2symbol -> scm_from_locale_symboln
2504 scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol
2505 scm_str2symbol -> scm_from_locale_symbol
2507 SCM_SYMBOL_HASH -> scm_hashq
2508 SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p
2510 scm_c_make_keyword -> scm_from_locale_keyword
2512 ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are
2513 now also available to C code.
2515 ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated.
2517 Use scm_is_keyword and scm_keyword_to_symbol instead, but note that
2518 the latter returns the true name of the keyword, not the 'dash name',
2519 as SCM_KEYWORDSYM used to do.
2521 ** A new way to access arrays in a thread-safe and efficient way has
2524 See the manual, node "Accessing Arrays From C".
2526 ** The old uniform vector and bitvector implementations have been
2527 unceremoniously removed.
2529 This implementation exposed the details of the tagging system of
2530 Guile. Use the new C API explained in the manual in node "Uniform
2531 Numeric Vectors" and "Bit Vectors", respectively.
2533 The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE,
2534 SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG,
2535 SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE,
2536 SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH,
2537 SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG,
2538 SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET,
2541 ** The macros dealing with vectors have been deprecated.
2543 Use the new functions scm_is_vector, scm_vector_elements,
2544 scm_vector_writable_elements, etc, or scm_is_simple_vector,
2545 SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the
2546 manual for more details.
2548 Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH,
2549 SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS.
2551 The following macros have been removed: SCM_VECTOR_BASE,
2552 SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH,
2553 SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS.
2555 ** Some C functions and macros related to arrays have been deprecated.
2557 Migrate according to the following table:
2559 scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc.
2560 scm_make_ra -> scm_make_array
2561 scm_shap2ra -> scm_make_array
2562 scm_cvref -> scm_c_generalized_vector_ref
2563 scm_ra_set_contp -> do not use
2564 scm_aind -> scm_array_handle_pos
2565 scm_raprin1 -> scm_display or scm_write
2567 SCM_ARRAYP -> scm_is_array
2568 SCM_ARRAY_NDIM -> scm_c_array_rank
2569 SCM_ARRAY_DIMS -> scm_array_handle_dims
2570 SCM_ARRAY_CONTP -> do not use
2571 SCM_ARRAY_MEM -> do not use
2572 SCM_ARRAY_V -> scm_array_handle_elements or similar
2573 SCM_ARRAY_BASE -> do not use
2575 ** SCM_CELL_WORD_LOC has been deprecated.
2577 Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer
2578 to a SCM, as opposed to a pointer to a scm_t_bits.
2580 This was done to allow the correct use of pointers into the Scheme
2581 heap. Previously, the heap words were of type scm_t_bits and local
2582 variables and function arguments were of type SCM, making it
2583 non-standards-conformant to have a pointer that can point to both.
2585 ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc.
2587 These macros should be used instead of SCM_CELL_WORD_2/3 to access the
2588 second and third words of double smobs. Likewise for
2589 SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3.
2591 Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be
2592 used to get and set the 16 exra bits in the zeroth word of a smob.
2594 And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for
2595 accesing the first immediate word of a smob as a SCM value, and there
2596 is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate
2597 smob word. Like wise for SCM_SMOB_OBJECT_2, etc.
2599 ** New way to deal with non-local exits and re-entries.
2601 There is a new set of functions that essentially do what
2602 scm_internal_dynamic_wind does, but in a way that is more convenient
2603 for C code in some situations. Here is a quick example of how to
2604 prevent a potential memory leak:
2611 scm_dynwind_begin (0);
2613 mem = scm_malloc (100);
2614 scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY);
2616 /* MEM would leak if BAR throws an error.
2617 SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless.
2624 /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by
2625 SCM_DYNWIND_END as well.
2629 For full documentation, see the node "Dynamic Wind" in the manual.
2631 ** New function scm_dynwind_free
2633 This function calls 'free' on a given pointer when a dynwind context
2634 is left. Thus the call to scm_dynwind_unwind_handler above could be
2635 replaced with simply scm_dynwind_free (mem).
2637 ** New functions scm_c_call_with_blocked_asyncs and
2638 scm_c_call_with_unblocked_asyncs
2640 Like scm_call_with_blocked_asyncs etc. but for C functions.
2642 ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs
2644 In addition to scm_c_call_with_blocked_asyncs you can now also use
2645 scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for
2646 scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs.
2648 ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS,
2649 SCM_REALLOW_INTS have been deprecated.
2651 They do no longer fulfill their original role of blocking signal
2652 delivery. Depending on what you want to achieve, replace a pair of
2653 SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a
2654 mutex, blocks asyncs, or both. See node "Critical Sections" in the
2657 ** The value 'scm_mask_ints' is no longer writable.
2659 Previously, you could set scm_mask_ints directly. This is no longer
2660 possible. Use scm_c_call_with_blocked_asyncs and
2661 scm_c_call_with_unblocked_asyncs instead.
2663 ** New way to temporarily set the current input, output or error ports
2665 C code can now use scm_dynwind_current_<foo>_port in a 'dynwind
2666 context' (see above). <foo> is one of "input", "output" or "error".
2668 ** New way to temporarily set fluids
2670 C code can now use scm_dynwind_fluid in a 'dynwind context' (see
2671 above) to temporarily set the value of a fluid.
2673 ** New types scm_t_intmax and scm_t_uintmax.
2675 On platforms that have them, these types are identical to intmax_t and
2676 uintmax_t, respectively. On other platforms, they are identical to
2677 the largest integer types that Guile knows about.
2679 ** The functions scm_unmemocopy and scm_unmemoize have been removed.
2681 You should not have used them.
2683 ** Many public #defines with generic names have been made private.
2685 #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made
2686 private or renamed with a more suitable public name.
2688 ** The macro SCM_TYP16S has been deprecated.
2690 This macro is not intended for public use.
2692 ** The macro SCM_SLOPPY_INEXACTP has been deprecated.
2694 Use scm_is_true (scm_inexact_p (...)) instead.
2696 ** The macro SCM_SLOPPY_REALP has been deprecated.
2698 Use scm_is_real instead.
2700 ** The macro SCM_SLOPPY_COMPLEXP has been deprecated.
2702 Use scm_is_complex instead.
2704 ** Some preprocessor defines have been deprecated.
2706 These defines indicated whether a certain feature was present in Guile
2707 or not. Going forward, assume that the features are always present.
2709 The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS,
2710 DEBUG_EXTENSIONS, DYNAMIC_LINKING.
2712 The following macros have been removed completely: MEMOIZE_LOCALS,
2713 SCM_RECKLESS, SCM_CAUTIOUS.
2715 ** The preprocessor define STACK_DIRECTION has been deprecated.
2717 There should be no need to know about the stack direction for ordinary
2720 ** New function: scm_effective_version
2722 Returns the "effective" version number. This is just the normal full
2723 version string without the final micro-version number. See "Changes
2724 to the distribution" above.
2726 ** The function scm_call_with_new_thread has a new prototype.
2728 Instead of taking a list with the thunk and handler, these two
2729 arguments are now passed directly:
2731 SCM scm_call_with_new_thread (SCM thunk, SCM handler);
2733 This is an incompatible change.
2735 ** New snarfer macro SCM_DEFINE_PUBLIC.
2737 This is like SCM_DEFINE, but also calls scm_c_export for the defined
2738 function in the init section.
2740 ** The snarfer macro SCM_SNARF_INIT is now officially supported.
2742 ** Garbage collector rewrite.
2744 The garbage collector is cleaned up a lot, and now uses lazy
2745 sweeping. This is reflected in the output of (gc-stats); since cells
2746 are being freed when they are allocated, the cells-allocated field
2747 stays roughly constant.
2749 For malloc related triggers, the behavior is changed. It uses the same
2750 heuristic as the cell-triggered collections. It may be tuned with the
2751 environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage
2752 for minimum yield of malloc related triggers. The default is 40.
2753 GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The
2756 Debugging operations for the freelist have been deprecated, along with
2757 the C variables that control garbage collection. The environment
2758 variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2,
2759 GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used.
2761 For understanding the memory usage of a GUILE program, the routine
2762 gc-live-object-stats returns an alist containing the number of live
2763 objects for every type.
2766 ** The function scm_definedp has been renamed to scm_defined_p
2768 The name scm_definedp is deprecated.
2770 ** The struct scm_cell type has been renamed to scm_t_cell
2772 This is in accordance to Guile's naming scheme for types. Note that
2773 the name scm_cell is now used for a function that allocates and
2774 initializes a new cell (see below).
2776 ** New functions for memory management
2778 A new set of functions for memory management has been added since the
2779 old way (scm_must_malloc, scm_must_free, etc) was error prone and
2780 indeed, Guile itself contained some long standing bugs that could
2781 cause aborts in long running programs.
2783 The new functions are more symmetrical and do not need cooperation
2784 from smob free routines, among other improvements.
2786 The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup,
2787 scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc,
2788 scm_gc_free, scm_gc_register_collectable_memory, and
2789 scm_gc_unregister_collectable_memory. Refer to the manual for more
2790 details and for upgrading instructions.
2792 The old functions for memory management have been deprecated. They
2793 are: scm_must_malloc, scm_must_realloc, scm_must_free,
2794 scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free.
2796 ** Declarations of exported features are marked with SCM_API.
2798 Every declaration of a feature that belongs to the exported Guile API
2799 has been marked by adding the macro "SCM_API" to the start of the
2800 declaration. This macro can expand into different things, the most
2801 common of which is just "extern" for Unix platforms. On Win32, it can
2802 be used to control which symbols are exported from a DLL.
2804 If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API
2805 will expand into "__declspec (dllimport) extern", which is needed for
2806 linking to the Guile DLL in Windows.
2808 There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and
2809 SCM_SRFI4_IMPORT, for the corresponding libraries.
2811 ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated.
2813 Use the new functions scm_cell and scm_double_cell instead. The old
2814 macros had problems because with them allocation and initialization
2815 was separated and the GC could sometimes observe half initialized
2816 cells. Only careful coding by the user of SCM_NEWCELL and
2817 SCM_NEWCELL2 could make this safe and efficient.
2819 ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated.
2821 Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p
2824 ** SRCBRKP has been deprecated.
2826 Use scm_c_source_property_breakpoint_p instead.
2828 ** Deprecated: scm_makmacro
2830 Change your code to use either scm_makmmacro or to define macros in
2831 Scheme, using 'define-macro'.
2833 ** New function scm_c_port_for_each.
2835 This function is like scm_port_for_each but takes a pointer to a C
2836 function as the callback instead of a SCM value.
2838 ** The names scm_internal_select, scm_thread_sleep, and
2839 scm_thread_usleep have been discouraged.
2841 Use scm_std_select, scm_std_sleep, scm_std_usleep instead.
2843 ** The GC can no longer be blocked.
2845 The global flags scm_gc_heap_lock and scm_block_gc have been removed.
2846 The GC can now run (partially) concurrently with other code and thus
2847 blocking it is not well defined.
2849 ** Many definitions have been removed that were previously deprecated.
2851 scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify,
2852 scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify,
2853 scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2,
2854 scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY,
2855 SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED,
2856 scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL,
2857 SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL,
2858 SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG,
2859 SCM_NUM_SIGS, scm_top_level_lookup_closure_var,
2860 *top-level-lookup-closure*, scm_system_transformer, scm_eval_3,
2861 scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP,
2862 SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring,
2863 scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP,
2864 SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig,
2865 scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT,
2866 SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET,
2867 SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH,
2868 SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR,
2869 scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0,
2870 scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated,
2871 scm_vector_set_length_x, scm_contregs, scm_debug_info,
2872 scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL,
2873 SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT,
2874 SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING,
2875 SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY,
2876 SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int,
2877 scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo,
2878 scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP,
2879 SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL,
2880 SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable,
2881 SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH.
2883 * Changes to bundled modules
2887 Using the (ice-9 debug) module no longer automatically switches Guile
2888 to use the debugging evaluator. If you want to switch to the
2889 debugging evaluator (which is needed for backtrace information if you
2890 hit an error), please add an explicit "(debug-enable 'debug)" to your
2891 code just after the code to use (ice-9 debug).
2894 Changes since Guile 1.4:
2896 * Changes to the distribution
2898 ** A top-level TODO file is included.
2900 ** Guile now uses a versioning scheme similar to that of the Linux kernel.
2902 Guile now always uses three numbers to represent the version,
2903 i.e. "1.6.5". The first number, 1, is the major version number, the
2904 second number, 6, is the minor version number, and the third number,
2905 5, is the micro version number. Changes in major version number
2906 indicate major changes in Guile.
2908 Minor version numbers that are even denote stable releases, and odd
2909 minor version numbers denote development versions (which may be
2910 unstable). The micro version number indicates a minor sub-revision of
2911 a given MAJOR.MINOR release.
2913 In keeping with the new scheme, (minor-version) and scm_minor_version
2914 no longer return everything but the major version number. They now
2915 just return the minor version number. Two new functions
2916 (micro-version) and scm_micro_version have been added to report the
2917 micro version number.
2919 In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION.
2921 ** New preprocessor definitions are available for checking versions.
2923 version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and
2924 SCM_MICRO_VERSION to the appropriate integer values.
2926 ** Guile now actively warns about deprecated features.
2928 The new configure option `--enable-deprecated=LEVEL' and the
2929 environment variable GUILE_WARN_DEPRECATED control this mechanism.
2930 See INSTALL and README for more information.
2932 ** Guile is much more likely to work on 64-bit architectures.
2934 Guile now compiles and passes "make check" with only two UNRESOLVED GC
2935 cases on Alpha and ia64 based machines now. Thanks to John Goerzen
2936 for the use of a test machine, and thanks to Stefan Jahn for ia64
2939 ** New functions: setitimer and getitimer.
2941 These implement a fairly direct interface to the libc functions of the
2944 ** The #. reader extension is now disabled by default.
2946 For safety reasons, #. evaluation is disabled by default. To
2947 re-enable it, set the fluid read-eval? to #t. For example:
2949 (fluid-set! read-eval? #t)
2951 but make sure you realize the potential security risks involved. With
2952 read-eval? enabled, reading a data file from an untrusted source can
2955 ** New SRFI modules have been added:
2957 SRFI-0 `cond-expand' is now supported in Guile, without requiring
2960 (srfi srfi-1) is a library containing many useful pair- and list-processing
2963 (srfi srfi-2) exports and-let*.
2965 (srfi srfi-4) implements homogeneous numeric vector datatypes.
2967 (srfi srfi-6) is a dummy module for now, since guile already provides
2968 all of the srfi-6 procedures by default: open-input-string,
2969 open-output-string, get-output-string.
2971 (srfi srfi-8) exports receive.
2973 (srfi srfi-9) exports define-record-type.
2975 (srfi srfi-10) exports define-reader-ctor and implements the reader
2978 (srfi srfi-11) exports let-values and let*-values.
2980 (srfi srfi-13) implements the SRFI String Library.
2982 (srfi srfi-14) implements the SRFI Character-Set Library.
2984 (srfi srfi-17) implements setter and getter-with-setter and redefines
2985 some accessor procedures as procedures with getters. (such as car,
2986 cdr, vector-ref etc.)
2988 (srfi srfi-19) implements the SRFI Time/Date Library.
2990 ** New scripts / "executable modules"
2992 Subdirectory "scripts" contains Scheme modules that are packaged to
2993 also be executable as scripts. At this time, these scripts are available:
3002 See README there for more info.
3004 These scripts can be invoked from the shell with the new program
3005 "guile-tools", which keeps track of installation directory for you.
3008 $ guile-tools display-commentary srfi/*.scm
3010 guile-tools is copied to the standard $bindir on "make install".
3012 ** New module (ice-9 stack-catch):
3014 stack-catch is like catch, but saves the current state of the stack in
3015 the fluid the-last-stack. This fluid can be useful when using the
3016 debugger and when re-throwing an error.
3018 ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star)
3020 This has been done to prevent problems on lesser operating systems
3021 that can't tolerate `*'s in file names. The exported macro continues
3022 to be named `and-let*', of course.
3024 On systems that support it, there is also a compatibility module named
3025 (ice-9 and-let*). It will go away in the next release.
3027 ** New modules (oop goops) etc.:
3030 (oop goops describe)
3032 (oop goops active-slot)
3033 (oop goops composite-slot)
3035 The Guile Object Oriented Programming System (GOOPS) has been
3036 integrated into Guile. For further information, consult the GOOPS
3037 manual and tutorial in the `doc' directory.
3039 ** New module (ice-9 rdelim).
3041 This exports the following procedures which were previously defined
3042 in the default environment:
3044 read-line read-line! read-delimited read-delimited! %read-delimited!
3045 %read-line write-line
3047 For backwards compatibility the definitions are still imported into the
3048 default environment in this version of Guile. However you should add:
3050 (use-modules (ice-9 rdelim))
3052 to any program which uses the definitions, since this may change in
3055 Alternatively, if guile-scsh is installed, the (scsh rdelim) module
3056 can be used for similar functionality.
3058 ** New module (ice-9 rw)
3060 This is a subset of the (scsh rw) module from guile-scsh. Currently
3061 it defines two procedures:
3063 *** New function: read-string!/partial str [port_or_fdes [start [end]]]
3065 Read characters from a port or file descriptor into a string STR.
3066 A port must have an underlying file descriptor -- a so-called
3067 fport. This procedure is scsh-compatible and can efficiently read
3070 *** New function: write-string/partial str [port_or_fdes [start [end]]]
3072 Write characters from a string STR to a port or file descriptor.
3073 A port must have an underlying file descriptor -- a so-called
3074 fport. This procedure is mostly compatible and can efficiently
3075 write large strings.
3077 ** New module (ice-9 match)
3079 This module includes Andrew K. Wright's pattern matcher. See
3080 ice-9/match.scm for brief description or
3082 http://www.star-lab.com/wright/code.html
3084 for complete documentation.
3086 ** New module (ice-9 buffered-input)
3088 This module provides procedures to construct an input port from an
3089 underlying source of input that reads and returns its input in chunks.
3090 The underlying input source is a Scheme procedure, specified by the
3091 caller, which the port invokes whenever it needs more input.
3093 This is useful when building an input port whose back end is Readline
3094 or a UI element such as the GtkEntry widget.
3098 The reference and tutorial documentation that was previously
3099 distributed separately, as `guile-doc', is now included in the core
3100 Guile distribution. The documentation consists of the following
3103 - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction
3106 - The Guile Reference Manual (guile.texi) contains (or is intended to
3107 contain) reference documentation on all aspects of Guile.
3109 - The GOOPS Manual (goops.texi) contains both tutorial-style and
3110 reference documentation for using GOOPS, Guile's Object Oriented
3113 - The Revised^5 Report on the Algorithmic Language Scheme
3116 See the README file in the `doc' directory for more details.
3118 ** There are a couple of examples in the examples/ directory now.
3120 * Changes to the stand-alone interpreter
3122 ** New command line option `--use-srfi'
3124 Using this option, SRFI modules can be loaded on startup and be
3125 available right from the beginning. This makes programming portable
3126 Scheme programs easier.
3128 The option `--use-srfi' expects a comma-separated list of numbers,
3129 each representing a SRFI number to be loaded into the interpreter
3130 before starting evaluating a script file or the REPL. Additionally,
3131 the feature identifier for the loaded SRFIs is recognized by
3132 `cond-expand' when using this option.
3135 $ guile --use-srfi=8,13
3136 guile> (receive (x z) (values 1 2) (+ 1 2))
3138 guile> (string-pad "bla" 20)
3141 ** Guile now always starts up in the `(guile-user)' module.
3143 Previously, scripts executed via the `-s' option would run in the
3144 `(guile)' module and the repl would run in the `(guile-user)' module.
3145 Now every user action takes place in the `(guile-user)' module by
3148 * Changes to Scheme functions and syntax
3150 ** Character classifiers work for non-ASCII characters.
3152 The predicates `char-alphabetic?', `char-numeric?',
3153 `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?'
3154 no longer check whether their arguments are ASCII characters.
3155 Previously, a character would only be considered alphabetic when it
3156 was also ASCII, for example.
3158 ** Previously deprecated Scheme functions have been removed:
3160 tag - no replacement.
3161 fseek - replaced by seek.
3162 list* - replaced by cons*.
3164 ** It's now possible to create modules with controlled environments
3168 (use-modules (ice-9 safe))
3169 (define m (make-safe-module))
3170 ;;; m will now be a module containing only a safe subset of R5RS
3171 (eval '(+ 1 2) m) --> 3
3172 (eval 'load m) --> ERROR: Unbound variable: load
3174 ** Evaluation of "()", the empty list, is now an error.
3176 Previously, the expression "()" evaluated to the empty list. This has
3177 been changed to signal a "missing expression" error. The correct way
3178 to write the empty list as a literal constant is to use quote: "'()".
3180 ** New concept of `Guile Extensions'.
3182 A Guile Extension is just a ordinary shared library that can be linked
3183 at run-time. We found it advantageous to give this simple concept a
3184 dedicated name to distinguish the issues related to shared libraries
3185 from the issues related to the module system.
3187 *** New function: load-extension
3189 Executing (load-extension lib init) is mostly equivalent to
3191 (dynamic-call init (dynamic-link lib))
3193 except when scm_register_extension has been called previously.
3194 Whenever appropriate, you should use `load-extension' instead of
3195 dynamic-link and dynamic-call.
3197 *** New C function: scm_c_register_extension
3199 This function registers a initialization function for use by
3200 `load-extension'. Use it when you don't want specific extensions to
3201 be loaded as shared libraries (for example on platforms that don't
3202 support dynamic linking).
3204 ** Auto-loading of compiled-code modules is deprecated.
3206 Guile used to be able to automatically find and link a shared
3207 library to satisfy requests for a module. For example, the module
3208 `(foo bar)' could be implemented by placing a shared library named
3209 "foo/libbar.so" (or with a different extension) in a directory on the
3212 This has been found to be too tricky, and is no longer supported. The
3213 shared libraries are now called "extensions". You should now write a
3214 small Scheme file that calls `load-extension' to load the shared
3215 library and initialize it explicitly.
3217 The shared libraries themselves should be installed in the usual
3218 places for shared libraries, with names like "libguile-foo-bar".
3220 For example, place this into a file "foo/bar.scm"
3222 (define-module (foo bar))
3224 (load-extension "libguile-foo-bar" "foobar_init")
3226 ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER
3228 `eval' is now R5RS, that is it takes two arguments.
3229 The second argument is an environment specifier, i.e. either
3231 (scheme-report-environment 5)
3232 (null-environment 5)
3233 (interaction-environment)
3239 ** The module system has been made more disciplined.
3241 The function `eval' will save and restore the current module around
3242 the evaluation of the specified expression. While this expression is
3243 evaluated, `(current-module)' will now return the right module, which
3244 is the module specified as the second argument to `eval'.
3246 A consequence of this change is that `eval' is not particularly
3247 useful when you want allow the evaluated code to change what module is
3248 designated as the current module and have this change persist from one
3249 call to `eval' to the next. The read-eval-print-loop is an example
3250 where `eval' is now inadequate. To compensate, there is a new
3251 function `primitive-eval' that does not take a module specifier and
3252 that does not save/restore the current module. You should use this
3253 function together with `set-current-module', `current-module', etc
3254 when you want to have more control over the state that is carried from
3255 one eval to the next.
3257 Additionally, it has been made sure that forms that are evaluated at
3258 the top level are always evaluated with respect to the current module.
3259 Previously, subforms of top-level forms such as `begin', `case',
3260 etc. did not respect changes to the current module although these
3261 subforms are at the top-level as well.
3263 To prevent strange behavior, the forms `define-module',
3264 `use-modules', `use-syntax', and `export' have been restricted to only
3265 work on the top level. The forms `define-public' and
3266 `defmacro-public' only export the new binding on the top level. They
3267 behave just like `define' and `defmacro', respectively, when they are
3268 used in a lexical environment.
3270 Also, `export' will no longer silently re-export bindings imported
3271 from a used module. It will emit a `deprecation' warning and will
3272 cease to perform any re-export in the next version. If you actually
3273 want to re-export bindings, use the new `re-export' in place of
3274 `export'. The new `re-export' will not make copies of variables when
3275 rexporting them, as `export' did wrongly.
3277 ** Module system now allows selection and renaming of imported bindings
3279 Previously, when using `use-modules' or the `#:use-module' clause in
3280 the `define-module' form, all the bindings (association of symbols to
3281 values) for imported modules were added to the "current module" on an
3282 as-is basis. This has been changed to allow finer control through two
3283 new facilities: selection and renaming.
3285 You can now select which of the imported module's bindings are to be
3286 visible in the current module by using the `:select' clause. This
3287 clause also can be used to rename individual bindings. For example:
3289 ;; import all bindings no questions asked
3290 (use-modules (ice-9 common-list))
3292 ;; import four bindings, renaming two of them;
3293 ;; the current module sees: every some zonk-y zonk-n
3294 (use-modules ((ice-9 common-list)
3296 (remove-if . zonk-y)
3297 (remove-if-not . zonk-n))))
3299 You can also programmatically rename all selected bindings using the
3300 `:renamer' clause, which specifies a proc that takes a symbol and
3301 returns another symbol. Because it is common practice to use a prefix,
3302 we now provide the convenience procedure `symbol-prefix-proc'. For
3305 ;; import four bindings, renaming two of them specifically,
3306 ;; and all four w/ prefix "CL:";
3307 ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n
3308 (use-modules ((ice-9 common-list)
3310 (remove-if . zonk-y)
3311 (remove-if-not . zonk-n))
3312 :renamer (symbol-prefix-proc 'CL:)))
3314 ;; import four bindings, renaming two of them specifically,
3315 ;; and all four by upcasing.
3316 ;; the current module sees: EVERY SOME ZONK-Y ZONK-N
3317 (define (upcase-symbol sym)
3318 (string->symbol (string-upcase (symbol->string sym))))
3320 (use-modules ((ice-9 common-list)
3322 (remove-if . zonk-y)
3323 (remove-if-not . zonk-n))
3324 :renamer upcase-symbol))
3326 Note that programmatic renaming is done *after* individual renaming.
3327 Also, the above examples show `use-modules', but the same facilities are
3328 available for the `#:use-module' clause of `define-module'.
3330 See manual for more info.
3332 ** The semantics of guardians have changed.
3334 The changes are for the most part compatible. An important criterion
3335 was to keep the typical usage of guardians as simple as before, but to
3336 make the semantics safer and (as a result) more useful.
3338 *** All objects returned from guardians are now properly alive.
3340 It is now guaranteed that any object referenced by an object returned
3341 from a guardian is alive. It's now impossible for a guardian to
3342 return a "contained" object before its "containing" object.
3344 One incompatible (but probably not very important) change resulting
3345 from this is that it is no longer possible to guard objects that
3346 indirectly reference themselves (i.e. are parts of cycles). If you do
3347 so accidentally, you'll get a warning.
3349 *** There are now two types of guardians: greedy and sharing.
3351 If you call (make-guardian #t) or just (make-guardian), you'll get a
3352 greedy guardian, and for (make-guardian #f) a sharing guardian.
3354 Greedy guardians are the default because they are more "defensive".
3355 You can only greedily guard an object once. If you guard an object
3356 more than once, once in a greedy guardian and the rest of times in
3357 sharing guardians, then it is guaranteed that the object won't be
3358 returned from sharing guardians as long as it is greedily guarded
3361 Guardians returned by calls to `make-guardian' can now take one more
3362 optional parameter, which says whether to throw an error in case an
3363 attempt is made to greedily guard an object that is already greedily
3364 guarded. The default is true, i.e. throw an error. If the parameter
3365 is false, the guardian invocation returns #t if guarding was
3366 successful and #f if it wasn't.
3368 Also, since greedy guarding is, in effect, a side-effecting operation
3369 on objects, a new function is introduced: `destroy-guardian!'.
3370 Invoking this function on a guardian renders it unoperative and, if
3371 the guardian is greedy, clears the "greedily guarded" property of the
3372 objects that were guarded by it, thus undoing the side effect.
3374 Note that all this hair is hardly very important, since guardian
3375 objects are usually permanent.
3377 ** Continuations created by call-with-current-continuation now accept
3378 any number of arguments, as required by R5RS.
3380 ** New function `issue-deprecation-warning'
3382 This function is used to display the deprecation messages that are
3383 controlled by GUILE_WARN_DEPRECATION as explained in the README.
3386 (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.")
3390 ;; `id' is deprecated. Use `identity' instead.
3395 ** New syntax `begin-deprecated'
3397 When deprecated features are included (as determined by the configure
3398 option --enable-deprecated), `begin-deprecated' is identical to
3399 `begin'. When deprecated features are excluded, it always evaluates
3400 to `#f', ignoring the body forms.
3402 ** New function `make-object-property'
3404 This function returns a new `procedure with setter' P that can be used
3405 to attach a property to objects. When calling P as
3409 where `obj' is any kind of object, it attaches `val' to `obj' in such
3410 a way that it can be retrieved by calling P as
3414 This function will replace procedure properties, symbol properties and
3415 source properties eventually.
3417 ** Module (ice-9 optargs) now uses keywords instead of `#&'.
3419 Instead of #&optional, #&key, etc you should now use #:optional,
3420 #:key, etc. Since #:optional is a keyword, you can write it as just
3421 :optional when (read-set! keywords 'prefix) is active.
3423 The old reader syntax `#&' is still supported, but deprecated. It
3424 will be removed in the next release.
3426 ** New define-module option: pure
3428 Tells the module system not to include any bindings from the root
3433 (define-module (totally-empty-module)
3436 ** New define-module option: export NAME1 ...
3438 Export names NAME1 ...
3440 This option is required if you want to be able to export bindings from
3441 a module which doesn't import one of `define-public' or `export'.
3445 (define-module (foo)
3447 :use-module (ice-9 r5rs)
3450 ;;; Note that we're pure R5RS below this point!
3455 ** New function: object->string OBJ
3457 Return a Scheme string obtained by printing a given object.
3459 ** New function: port? X
3461 Returns a boolean indicating whether X is a port. Equivalent to
3462 `(or (input-port? X) (output-port? X))'.
3464 ** New function: file-port?
3466 Determines whether a given object is a port that is related to a file.
3468 ** New function: port-for-each proc
3470 Apply PROC to each port in the Guile port table in turn. The return
3471 value is unspecified. More specifically, PROC is applied exactly once
3472 to every port that exists in the system at the time PORT-FOR-EACH is
3473 invoked. Changes to the port table while PORT-FOR-EACH is running
3474 have no effect as far as PORT-FOR-EACH is concerned.
3476 ** New function: dup2 oldfd newfd
3478 A simple wrapper for the `dup2' system call. Copies the file
3479 descriptor OLDFD to descriptor number NEWFD, replacing the
3480 previous meaning of NEWFD. Both OLDFD and NEWFD must be integers.
3481 Unlike for dup->fdes or primitive-move->fdes, no attempt is made
3482 to move away ports which are using NEWFD. The return value is
3485 ** New function: close-fdes fd
3487 A simple wrapper for the `close' system call. Close file
3488 descriptor FD, which must be an integer. Unlike close (*note
3489 close: Ports and File Descriptors.), the file descriptor will be
3490 closed even if a port is using it. The return value is
3493 ** New function: crypt password salt
3495 Encrypts `password' using the standard unix password encryption
3498 ** New function: chroot path
3500 Change the root directory of the running process to `path'.
3502 ** New functions: getlogin, cuserid
3504 Return the login name or the user name of the current effective user
3507 ** New functions: getpriority which who, setpriority which who prio
3509 Get or set the priority of the running process.
3511 ** New function: getpass prompt
3513 Read a password from the terminal, first displaying `prompt' and
3516 ** New function: flock file operation
3518 Set/remove an advisory shared or exclusive lock on `file'.
3520 ** New functions: sethostname name, gethostname
3522 Set or get the hostname of the machine the current process is running
3525 ** New function: mkstemp! tmpl
3527 mkstemp creates a new unique file in the file system and returns a
3528 new buffered port open for reading and writing to the file. TMPL
3529 is a string specifying where the file should be created: it must
3530 end with `XXXXXX' and will be changed in place to return the name
3531 of the temporary file.
3533 ** New function: open-input-string string
3535 Return an input string port which delivers the characters from
3536 `string'. This procedure, together with `open-output-string' and
3537 `get-output-string' implements SRFI-6.
3539 ** New function: open-output-string
3541 Return an output string port which collects all data written to it.
3542 The data can then be retrieved by `get-output-string'.
3544 ** New function: get-output-string
3546 Return the contents of an output string port.
3548 ** New function: identity
3550 Return the argument.
3552 ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses
3553 are represented in Scheme as integers with normal host byte ordering.
3555 ** New function: inet-pton family address
3557 Convert a printable string network address into an integer. Note that
3558 unlike the C version of this function, the result is an integer with
3559 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3562 (inet-pton AF_INET "127.0.0.1") => 2130706433
3563 (inet-pton AF_INET6 "::1") => 1
3565 ** New function: inet-ntop family address
3567 Convert an integer network address into a printable string. Note that
3568 unlike the C version of this function, the input is an integer with
3569 normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'.
3572 (inet-ntop AF_INET 2130706433) => "127.0.0.1"
3573 (inet-ntop AF_INET6 (- (expt 2 128) 1)) =>
3574 ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff
3578 Use `identity' instead.
3584 ** Deprecated: return-it
3588 ** Deprecated: string-character-length
3590 Use `string-length' instead.
3592 ** Deprecated: flags
3594 Use `logior' instead.
3596 ** Deprecated: close-all-ports-except.
3598 This was intended for closing ports in a child process after a fork,
3599 but it has the undesirable side effect of flushing buffers.
3600 port-for-each is more flexible.
3602 ** The (ice-9 popen) module now attempts to set up file descriptors in
3603 the child process from the current Scheme ports, instead of using the
3604 current values of file descriptors 0, 1, and 2 in the parent process.
3606 ** Removed function: builtin-weak-bindings
3608 There is no such concept as a weak binding any more.
3610 ** Removed constants: bignum-radix, scm-line-incrementors
3612 ** define-method: New syntax mandatory.
3614 The new method syntax is now mandatory:
3616 (define-method (NAME ARG-SPEC ...) BODY ...)
3617 (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...)
3619 ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE)
3620 REST-ARG ::= ARG-NAME
3622 If you have old code using the old syntax, import
3623 (oop goops old-define-method) before (oop goops) as in:
3625 (use-modules (oop goops old-define-method) (oop goops))
3627 ** Deprecated function: builtin-variable
3628 Removed function: builtin-bindings
3630 There is no longer a distinction between builtin or other variables.
3631 Use module system operations for all variables.
3633 ** Lazy-catch handlers are no longer allowed to return.
3635 That is, a call to `throw', `error', etc is now guaranteed to not
3638 ** Bugfixes for (ice-9 getopt-long)
3640 This module is now tested using test-suite/tests/getopt-long.test.
3641 The following bugs have been fixed:
3643 *** Parsing for options that are specified to have `optional' args now checks
3644 if the next element is an option instead of unconditionally taking it as the
3647 *** An error is now thrown for `--opt=val' when the option description
3648 does not specify `(value #t)' or `(value optional)'. This condition used to
3649 be accepted w/o error, contrary to the documentation.
3651 *** The error message for unrecognized options is now more informative.
3652 It used to be "not a record", an artifact of the implementation.
3654 *** The error message for `--opt' terminating the arg list (no value), when
3655 `(value #t)' is specified, is now more informative. It used to be "not enough
3658 *** "Clumped" single-char args now preserve trailing string, use it as arg.
3659 The expansion used to be like so:
3661 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz")
3663 Note that the "5d" is dropped. Now it is like so:
3665 ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz")
3667 This enables single-char options to have adjoining arguments as long as their
3668 constituent characters are not potential single-char options.
3670 ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*'
3672 The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure
3673 property `arglist', which can be retrieved by `arity'. The result is that
3674 `arity' can give more detailed information than before:
3678 guile> (use-modules (ice-9 optargs))
3679 guile> (define* (foo #:optional a b c) a)
3681 0 or more arguments in `lambda*:G0'.
3686 3 optional arguments: `a', `b' and `c'.
3687 guile> (define* (bar a b #:key c d #:allow-other-keys) a)
3689 2 required arguments: `a' and `b', 2 keyword arguments: `c'
3690 and `d', other keywords allowed.
3691 guile> (define* (baz a b #:optional c #:rest r) a)
3693 2 required arguments: `a' and `b', 1 optional argument: `c',
3696 * Changes to the C interface
3698 ** Types have been renamed from scm_*_t to scm_t_*.
3700 This has been done for POSIX sake. It reserves identifiers ending
3701 with "_t". What a concept.
3703 The old names are still available with status `deprecated'.
3705 ** scm_t_bits (former scm_bits_t) is now a unsigned type.
3707 ** Deprecated features have been removed.
3711 SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR
3712 SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP
3714 *** C Functions removed
3716 scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo
3717 scm_fseek - replaced by scm_seek.
3718 gc-thunk - replaced by after-gc-hook.
3719 gh_int2scmb - replaced by gh_bool2scm.
3720 scm_tc_dblr - replaced by scm_tc16_real.
3721 scm_tc_dblc - replaced by scm_tc16_complex.
3722 scm_list_star - replaced by scm_cons_star.
3724 ** Deprecated: scm_makfromstr
3726 Use scm_mem2string instead.
3728 ** Deprecated: scm_make_shared_substring
3730 Explicit shared substrings will disappear from Guile.
3732 Instead, "normal" strings will be implemented using sharing
3733 internally, combined with a copy-on-write strategy.
3735 ** Deprecated: scm_read_only_string_p
3737 The concept of read-only strings will disappear in next release of
3740 ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member
3742 Instead, use scm_c_memq or scm_memq, scm_memv, scm_member.
3744 ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3
3746 Call a procedure with the indicated number of arguments. See "Fly
3747 Evaluation" in the manual.
3749 ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3
3751 Call a procedure with the indicated number of arguments and a list of
3752 further arguments. See "Fly Evaluation" in the manual.
3754 ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5
3756 Create a list of the given number of elements. See "List
3757 Constructors" in the manual.
3759 ** Renamed function: scm_listify has been replaced by scm_list_n.
3761 ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4,
3762 SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9.
3764 Use functions scm_list_N instead.
3766 ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size)
3768 Used by an application to read arbitrary number of bytes from a port.
3769 Same semantics as libc read, except that scm_c_read only returns less
3770 than SIZE bytes if at end-of-file.
3772 Warning: Doesn't update port line and column counts!
3774 ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size)
3776 Used by an application to write arbitrary number of bytes to an SCM
3777 port. Similar semantics as libc write. However, unlike libc
3778 write, scm_c_write writes the requested number of bytes and has no
3781 Warning: Doesn't update port line and column counts!
3783 ** New function: scm_init_guile ()
3785 In contrast to scm_boot_guile, scm_init_guile will return normally
3786 after initializing Guile. It is not available on all systems, tho.
3788 ** New functions: scm_str2symbol, scm_mem2symbol
3790 The function scm_str2symbol takes a const char* pointing to a zero-terminated
3791 field of characters and creates a scheme symbol object from that C string.
3792 The function scm_mem2symbol takes a const char* and a number of characters and
3793 creates a symbol from the characters in that memory area.
3795 ** New functions: scm_primitive_make_property
3796 scm_primitive_property_ref
3797 scm_primitive_property_set_x
3798 scm_primitive_property_del_x
3800 These functions implement a new way to deal with object properties.
3801 See libguile/properties.c for their documentation.
3803 ** New function: scm_done_free (long size)
3805 This function is the inverse of scm_done_malloc. Use it to report the
3806 amount of smob memory you free. The previous method, which involved
3807 calling scm_done_malloc with negative argument, was somewhat
3808 unintuitive (and is still available, of course).
3810 ** New function: scm_c_memq (SCM obj, SCM list)
3812 This function provides a fast C level alternative for scm_memq for the case
3813 that the list parameter is known to be a proper list. The function is a
3814 replacement for scm_sloppy_memq, but is stricter in its requirements on its
3815 list input parameter, since for anything else but a proper list the function's
3816 behaviour is undefined - it may even crash or loop endlessly. Further, for
3817 the case that the object is not found in the list, scm_c_memq returns #f which
3818 is similar to scm_memq, but different from scm_sloppy_memq's behaviour.
3820 ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2,
3821 scm_remember_upto_here
3823 These functions replace the function scm_remember.
3825 ** Deprecated function: scm_remember
3827 Use one of the new functions scm_remember_upto_here_1,
3828 scm_remember_upto_here_2 or scm_remember_upto_here instead.
3830 ** New function: scm_allocate_string
3832 This function replaces the function scm_makstr.
3834 ** Deprecated function: scm_makstr
3836 Use the new function scm_allocate_string instead.
3838 ** New global variable scm_gc_running_p introduced.
3840 Use this variable to find out if garbage collection is being executed. Up to
3841 now applications have used scm_gc_heap_lock to test if garbage collection was
3842 running, which also works because of the fact that up to know only the garbage
3843 collector has set this variable. But, this is an implementation detail that
3844 may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use
3845 of this variable is (and has been) not fully safe anyway.
3847 ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH
3849 Use these instead of SCM_LENGTH_MAX.
3851 ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH,
3852 SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH,
3853 SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH.
3855 Use these instead of SCM_LENGTH.
3857 ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH,
3858 SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH,
3859 SCM_SET_BITVECTOR_LENGTH
3861 Use these instead of SCM_SETLENGTH
3863 ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE,
3864 SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM,
3867 Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or
3870 ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS,
3871 SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE,
3874 Use these instead of SCM_SETCHARS.
3876 ** New macro: SCM_BITVECTOR_P
3878 ** New macro: SCM_STRING_COERCE_0TERMINATION_X
3880 Use instead of SCM_COERCE_SUBSTR.
3882 ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN
3884 For directory objects, use these instead of SCM_OPDIRP and SCM_OPN.
3886 ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL,
3887 SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL,
3888 SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD,
3889 SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP,
3890 SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS,
3891 SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY,
3892 SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH,
3893 SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR,
3894 SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS,
3895 SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP,
3896 SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC,
3897 SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG,
3898 SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY,
3899 SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP,
3900 SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR
3902 Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE.
3903 Use scm_memory_error instead of SCM_NALLOC.
3904 Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP.
3905 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR.
3906 Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP
3907 Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS.
3908 Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH.
3909 Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING.
3910 Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR.
3911 Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP.
3912 Use SCM_STRINGP instead of SCM_RWSTRINGP.
3913 Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING.
3914 Use SCM_STRING_CHARS instead of SCM_ROCHARS.
3915 Use SCM_STRING_UCHARS instead of SCM_ROUCHARS.
3916 Use a type specific setter macro instead of SCM_SETLENGTH.
3917 Use a type specific setter macro instead of SCM_SETCHARS.
3918 Use a type specific length macro instead of SCM_LENGTH_MAX.
3919 Use SCM_GCMARKP instead of SCM_GC8MARKP.
3920 Use SCM_SETGCMARK instead of SCM_SETGC8MARK.
3921 Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK.
3922 Use SCM_TYP16 instead of SCM_GCTYP16.
3923 Use SCM_CDR instead of SCM_GCCDR.
3924 Use SCM_DIR_OPEN_P instead of SCM_OPDIRP.
3925 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA.
3926 Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA.
3927 Use SCM_VCELL_INIT instead of SCM_CONST_LONG.
3928 Use SCM_WRONG_NUM_ARGS instead of SCM_WNA.
3929 Use SCM_CONSP instead of SCM_SLOPPY_CONSP.
3930 Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP.
3932 ** Removed function: scm_struct_init
3934 ** Removed variable: scm_symhash_dim
3936 ** Renamed function: scm_make_cont has been replaced by
3937 scm_make_continuation, which has a different interface.
3939 ** Deprecated function: scm_call_catching_errors
3941 Use scm_catch or scm_lazy_catch from throw.[ch] instead.
3943 ** Deprecated function: scm_strhash
3945 Use scm_string_hash instead.
3947 ** Deprecated function: scm_vector_set_length_x
3949 Instead, create a fresh vector of the desired size and copy the contents.
3951 ** scm_gensym has changed prototype
3953 scm_gensym now only takes one argument.
3955 ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols,
3958 There is now only a single symbol type scm_tc7_symbol.
3959 The tag scm_tc7_lvector was not used anyway.
3961 ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe.
3963 Use scm_make_smob_type and scm_set_smob_XXX instead.
3965 ** New function scm_set_smob_apply.
3967 This can be used to set an apply function to a smob type.
3969 ** Deprecated function: scm_strprint_obj
3971 Use scm_object_to_string instead.
3973 ** Deprecated function: scm_wta
3975 Use scm_wrong_type_arg, or another appropriate error signalling function
3978 ** Explicit support for obarrays has been deprecated.
3980 Use `scm_str2symbol' and the generic hashtable functions instead.
3982 ** The concept of `vcells' has been deprecated.
3984 The data type `variable' is now used exclusively. `Vcells' have been
3985 a low-level concept so you are likely not affected by this change.
3987 *** Deprecated functions: scm_sym2vcell, scm_sysintern,
3988 scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0.
3990 Use scm_c_define or scm_c_lookup instead, as appropriate.
3992 *** New functions: scm_c_module_lookup, scm_c_lookup,
3993 scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup,
3994 scm_module_define, scm_define.
3996 These functions work with variables instead of with vcells.
3998 ** New functions for creating and defining `subr's and `gsubr's.
4000 The new functions more clearly distinguish between creating a subr (or
4001 gsubr) object and adding it to the current module.
4003 These new functions are available: scm_c_make_subr, scm_c_define_subr,
4004 scm_c_make_subr_with_generic, scm_c_define_subr_with_generic,
4005 scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic,
4006 scm_c_define_gsubr_with_generic.
4008 ** Deprecated functions: scm_make_subr, scm_make_subr_opt,
4009 scm_make_subr_with_generic, scm_make_gsubr,
4010 scm_make_gsubr_with_generic.
4012 Use the new ones from above instead.
4014 ** C interface to the module system has changed.
4016 While we suggest that you avoid as many explicit module system
4017 operations from C as possible for the time being, the C interface has
4018 been made more similar to the high-level Scheme module system.
4020 *** New functions: scm_c_define_module, scm_c_use_module,
4021 scm_c_export, scm_c_resolve_module.
4023 They mostly work like their Scheme namesakes. scm_c_define_module
4024 takes a function that is called a context where the new module is
4027 *** Deprecated functions: scm_the_root_module, scm_make_module,
4028 scm_ensure_user_module, scm_load_scheme_module.
4030 Use the new functions instead.
4032 ** Renamed function: scm_internal_with_fluids becomes
4035 scm_internal_with_fluids is available as a deprecated function.
4037 ** New function: scm_c_with_fluid.
4039 Just like scm_c_with_fluids, but takes one fluid and one value instead
4042 ** Deprecated typedefs: long_long, ulong_long.
4044 They are of questionable utility and they pollute the global
4047 ** Deprecated typedef: scm_sizet
4049 It is of questionable utility now that Guile requires ANSI C, and is
4052 ** Deprecated typedefs: scm_port_rw_active, scm_port,
4053 scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport,
4054 scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim.
4056 Made more compliant with the naming policy by adding a _t at the end.
4058 ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig,
4059 scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl
4061 With the exception of the mysterious scm_2ulong2big, they are still
4062 available under new names (scm_i_mkbig etc). These functions are not
4063 intended to be used in user code. You should avoid dealing with
4064 bignums directly, and should deal with numbers in general (which can
4067 ** Change in behavior: scm_num2long, scm_num2ulong
4069 The scm_num2[u]long functions don't any longer accept an inexact
4070 argument. This change in behavior is motivated by concordance with
4071 R5RS: It is more common that a primitive doesn't want to accept an
4072 inexact for an exact.
4074 ** New functions: scm_short2num, scm_ushort2num, scm_int2num,
4075 scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short,
4076 scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff,
4079 These are conversion functions between the various ANSI C integral
4080 types and Scheme numbers. NOTE: The scm_num2xxx functions don't
4081 accept an inexact argument.
4083 ** New functions: scm_float2num, scm_double2num,
4084 scm_num2float, scm_num2double.
4086 These are conversion functions between the two ANSI C float types and
4089 ** New number validation macros:
4090 SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF]
4094 ** New functions: scm_gc_protect_object, scm_gc_unprotect_object
4096 These are just nicer-named old scm_protect_object and
4097 scm_unprotect_object.
4099 ** Deprecated functions: scm_protect_object, scm_unprotect_object
4101 ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots
4103 These functions can be used to register pointers to locations that
4106 ** Deprecated function: scm_create_hook.
4108 Its sins are: misleading name, non-modularity and lack of general
4112 Changes since Guile 1.3.4:
4114 * Changes to the distribution
4116 ** Trees from nightly snapshots and CVS now require you to run autogen.sh.
4118 We've changed the way we handle generated files in the Guile source
4119 repository. As a result, the procedure for building trees obtained
4120 from the nightly FTP snapshots or via CVS has changed:
4121 - You must have appropriate versions of autoconf, automake, and
4122 libtool installed on your system. See README for info on how to
4123 obtain these programs.
4124 - Before configuring the tree, you must first run the script
4125 `autogen.sh' at the top of the source tree.
4127 The Guile repository used to contain not only source files, written by
4128 humans, but also some generated files, like configure scripts and
4129 Makefile.in files. Even though the contents of these files could be
4130 derived mechanically from other files present, we thought it would
4131 make the tree easier to build if we checked them into CVS.
4133 However, this approach means that minor differences between
4134 developer's installed tools and habits affected the whole team.
4135 So we have removed the generated files from the repository, and
4136 added the autogen.sh script, which will reconstruct them
4140 ** configure now has experimental options to remove support for certain
4143 --disable-arrays omit array and uniform array support
4144 --disable-posix omit posix interfaces
4145 --disable-networking omit networking interfaces
4146 --disable-regex omit regular expression interfaces
4148 These are likely to become separate modules some day.
4150 ** New configure option --enable-debug-freelist
4152 This enables a debugging version of SCM_NEWCELL(), and also registers
4153 an extra primitive, the setter `gc-set-debug-check-freelist!'.
4155 Configure with the --enable-debug-freelist option to enable
4156 the gc-set-debug-check-freelist! primitive, and then use:
4158 (gc-set-debug-check-freelist! #t) # turn on checking of the freelist
4159 (gc-set-debug-check-freelist! #f) # turn off checking
4161 Checking of the freelist forces a traversal of the freelist and
4162 a garbage collection before each allocation of a cell. This can
4163 slow down the interpreter dramatically, so the setter should be used to
4164 turn on this extra processing only when necessary.
4166 ** New configure option --enable-debug-malloc
4168 Include code for debugging of calls to scm_must_malloc/realloc/free.
4172 1. objects freed by scm_must_free has been mallocated by scm_must_malloc
4173 2. objects reallocated by scm_must_realloc has been allocated by
4175 3. reallocated objects are reallocated with the same what string
4177 But, most importantly, it records the number of allocated objects of
4178 each kind. This is useful when searching for memory leaks.
4180 A Guile compiled with this option provides the primitive
4181 `malloc-stats' which returns an alist with pairs of kind and the
4182 number of objects of that kind.
4184 ** All includes are now referenced relative to the root directory
4186 Since some users have had problems with mixups between Guile and
4187 system headers, we have decided to always refer to Guile headers via
4188 their parent directories. This essentially creates a "private name
4189 space" for Guile headers. This means that the compiler only is given
4190 -I options for the root build and root source directory.
4192 ** Header files kw.h and genio.h have been removed.
4194 ** The module (ice-9 getopt-gnu-style) has been removed.
4196 ** New module (ice-9 documentation)
4198 Implements the interface to documentation strings associated with
4201 ** New module (ice-9 time)
4203 Provides a macro `time', which displays execution time of a given form.
4205 ** New module (ice-9 history)
4207 Loading this module enables value history in the repl.
4209 * Changes to the stand-alone interpreter
4211 ** New command line option --debug
4213 Start Guile with debugging evaluator and backtraces enabled.
4215 This is useful when debugging your .guile init file or scripts.
4217 ** New help facility
4219 Usage: (help NAME) gives documentation about objects named NAME (a symbol)
4220 (help REGEXP) ditto for objects with names matching REGEXP (a string)
4221 (help 'NAME) gives documentation for NAME, even if it is not an object
4222 (help ,EXPR) gives documentation for object returned by EXPR
4223 (help (my module)) gives module commentary for `(my module)'
4224 (help) gives this text
4226 `help' searches among bindings exported from loaded modules, while
4227 `apropos' searches among bindings visible from the "current" module.
4229 Examples: (help help)
4231 (help "output-string")
4233 ** `help' and `apropos' now prints full module names
4235 ** Dynamic linking now uses libltdl from the libtool package.
4237 The old system dependent code for doing dynamic linking has been
4238 replaced with calls to the libltdl functions which do all the hairy
4241 The major improvement is that you can now directly pass libtool
4242 library names like "libfoo.la" to `dynamic-link' and `dynamic-link'
4243 will be able to do the best shared library job you can get, via
4246 The way dynamic libraries are found has changed and is not really
4247 portable across platforms, probably. It is therefore recommended to
4248 use absolute filenames when possible.
4250 If you pass a filename without an extension to `dynamic-link', it will
4251 try a few appropriate ones. Thus, the most platform ignorant way is
4252 to specify a name like "libfoo", without any directories and
4255 ** Guile COOP threads are now compatible with LinuxThreads
4257 Previously, COOP threading wasn't possible in applications linked with
4258 Linux POSIX threads due to their use of the stack pointer to find the
4259 thread context. This has now been fixed with a workaround which uses
4260 the pthreads to allocate the stack.
4262 ** New primitives: `pkgdata-dir', `site-dir', `library-dir'
4264 ** Positions of erring expression in scripts
4266 With version 1.3.4, the location of the erring expression in Guile
4267 scipts is no longer automatically reported. (This should have been
4268 documented before the 1.3.4 release.)
4270 You can get this information by enabling recording of positions of
4271 source expressions and running the debugging evaluator. Put this at
4272 the top of your script (or in your "site" file):
4274 (read-enable 'positions)
4275 (debug-enable 'debug)
4277 ** Backtraces in scripts
4279 It is now possible to get backtraces in scripts.
4283 (debug-enable 'debug 'backtrace)
4285 at the top of the script.
4287 (The first options enables the debugging evaluator.
4288 The second enables backtraces.)
4290 ** Part of module system symbol lookup now implemented in C
4292 The eval closure of most modules is now implemented in C. Since this
4293 was one of the bottlenecks for loading speed, Guile now loads code
4294 substantially faster than before.
4296 ** Attempting to get the value of an unbound variable now produces
4297 an exception with a key of 'unbound-variable instead of 'misc-error.
4299 ** The initial default output port is now unbuffered if it's using a
4300 tty device. Previously in this situation it was line-buffered.
4302 ** New hook: after-gc-hook
4304 after-gc-hook takes over the role of gc-thunk. This hook is run at
4305 the first SCM_TICK after a GC. (Thus, the code is run at the same
4306 point during evaluation as signal handlers.)
4308 Note that this hook should be used only for diagnostic and debugging
4309 purposes. It is not certain that it will continue to be well-defined
4310 when this hook is run in the future.
4312 C programmers: Note the new C level hooks scm_before_gc_c_hook,
4313 scm_before_sweep_c_hook, scm_after_gc_c_hook.
4315 ** Improvements to garbage collector
4317 Guile 1.4 has a new policy for triggering heap allocation and
4318 determining the sizes of heap segments. It fixes a number of problems
4321 1. The new policy can handle two separate pools of cells
4322 (2-word/4-word) better. (The old policy would run wild, allocating
4323 more and more memory for certain programs.)
4325 2. The old code would sometimes allocate far too much heap so that the
4326 Guile process became gigantic. The new code avoids this.
4328 3. The old code would sometimes allocate too little so that few cells
4329 were freed at GC so that, in turn, too much time was spent in GC.
4331 4. The old code would often trigger heap allocation several times in a
4332 row. (The new scheme predicts how large the segments needs to be
4333 in order not to need further allocation.)
4335 All in all, the new GC policy will make larger applications more
4338 The new GC scheme also is prepared for POSIX threading. Threads can
4339 allocate private pools of cells ("clusters") with just a single
4340 function call. Allocation of single cells from such a cluster can
4341 then proceed without any need of inter-thread synchronization.
4343 ** New environment variables controlling GC parameters
4345 GUILE_MAX_SEGMENT_SIZE Maximal segment size
4348 Allocation of 2-word cell heaps:
4350 GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes
4353 GUILE_MIN_YIELD_1 Minimum number of freed cells at each
4354 GC in percent of total heap size
4357 Allocation of 4-word cell heaps
4358 (used for real numbers and misc other objects):
4360 GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2
4362 (See entry "Way for application to customize GC parameters" under
4363 section "Changes to the scm_ interface" below.)
4365 ** Guile now implements reals using 4-word cells
4367 This speeds up computation with reals. (They were earlier allocated
4368 with `malloc'.) There is still some room for optimizations, however.
4370 ** Some further steps toward POSIX thread support have been taken
4372 *** Guile's critical sections (SCM_DEFER/ALLOW_INTS)
4373 don't have much effect any longer, and many of them will be removed in
4377 are only handled at the top of the evaluator loop, immediately after
4378 I/O, and in scm_equalp.
4380 *** The GC can allocate thread private pools of pairs.
4382 * Changes to Scheme functions and syntax
4384 ** close-input-port and close-output-port are now R5RS
4386 These procedures have been turned into primitives and have R5RS behaviour.
4388 ** New procedure: simple-format PORT MESSAGE ARG1 ...
4390 (ice-9 boot) makes `format' an alias for `simple-format' until possibly
4391 extended by the more sophisticated version in (ice-9 format)
4393 (simple-format port message . args)
4394 Write MESSAGE to DESTINATION, defaulting to `current-output-port'.
4395 MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed,
4396 the escapes are replaced with corresponding members of ARGS:
4397 ~A formats using `display' and ~S formats using `write'.
4398 If DESTINATION is #t, then use the `current-output-port',
4399 if DESTINATION is #f, then return a string containing the formatted text.
4400 Does not add a trailing newline."
4402 ** string-ref: the second argument is no longer optional.
4404 ** string, list->string: no longer accept strings in their arguments,
4405 only characters, for compatibility with R5RS.
4407 ** New procedure: port-closed? PORT
4408 Returns #t if PORT is closed or #f if it is open.
4410 ** Deprecated: list*
4412 The list* functionality is now provided by cons* (SRFI-1 compliant)
4414 ** New procedure: cons* ARG1 ARG2 ... ARGn
4416 Like `list', but the last arg provides the tail of the constructed list,
4417 returning (cons ARG1 (cons ARG2 (cons ... ARGn))).
4419 Requires at least one argument. If given one argument, that argument
4420 is returned as result.
4422 This function is called `list*' in some other Schemes and in Common LISP.
4424 ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map!
4426 ** New procedure: object-documentation OBJECT
4428 Returns the documentation string associated with OBJECT. The
4429 procedure uses a caching mechanism so that subsequent lookups are
4432 Exported by (ice-9 documentation).
4434 ** module-name now returns full names of modules
4436 Previously, only the last part of the name was returned (`session' for
4437 `(ice-9 session)'). Ex: `(ice-9 session)'.
4439 * Changes to the gh_ interface
4441 ** Deprecated: gh_int2scmb
4443 Use gh_bool2scm instead.
4445 * Changes to the scm_ interface
4447 ** Guile primitives now carry docstrings!
4449 Thanks to Greg Badros!
4451 ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4453 Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC
4454 macros and must contain a docstring that is extracted into foo.doc using a new
4455 guile-doc-snarf script (that uses guile-doc-snarf.awk).
4457 However, a major overhaul of these macros is scheduled for the next release of
4460 ** Guile primitives use a new technique for validation of arguments
4462 SCM_VALIDATE_* macros are defined to ease the redundancy and improve
4463 the readability of argument checking.
4465 ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents.
4467 ** New macros: SCM_PACK, SCM_UNPACK
4469 Compose/decompose an SCM value.
4471 The SCM type is now treated as an abstract data type and may be defined as a
4472 long, a void* or as a struct, depending on the architecture and compile time
4473 options. This makes it easier to find several types of bugs, for example when
4474 SCM values are treated as integers without conversion. Values of the SCM type
4475 should be treated as "atomic" values. These macros are used when
4476 composing/decomposing an SCM value, either because you want to access
4477 individual bits, or because you want to treat it as an integer value.
4479 E.g., in order to set bit 7 in an SCM value x, use the expression
4481 SCM_PACK (SCM_UNPACK (x) | 0x80)
4483 ** The name property of hooks is deprecated.
4484 Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated.
4486 You can emulate this feature by using object properties.
4488 ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP,
4489 SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP,
4492 These macros will be removed in a future release of Guile.
4494 ** The following types, functions and macros from numbers.h are deprecated:
4495 scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL,
4496 SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG
4498 ** Port internals: the rw_random variable in the scm_port structure
4499 must be set to non-zero in any random access port. In recent Guile
4500 releases it was only set for bidirectional random-access ports.
4502 ** Port internals: the seek ptob procedure is now responsible for
4503 resetting the buffers if required. The change was made so that in the
4504 special case of reading the current position (i.e., seek p 0 SEEK_CUR)
4505 the fport and strport ptobs can avoid resetting the buffers,
4506 in particular to avoid discarding unread chars. An existing port
4507 type can be fixed by adding something like the following to the
4508 beginning of the ptob seek procedure:
4510 if (pt->rw_active == SCM_PORT_READ)
4511 scm_end_input (object);
4512 else if (pt->rw_active == SCM_PORT_WRITE)
4513 ptob->flush (object);
4515 although to actually avoid resetting the buffers and discard unread
4516 chars requires further hacking that depends on the characteristics
4519 ** Deprecated functions: scm_fseek, scm_tag
4521 These functions are no longer used and will be removed in a future version.
4523 ** The scm_sysmissing procedure is no longer used in libguile.
4524 Unless it turns out to be unexpectedly useful to somebody, it will be
4525 removed in a future version.
4527 ** The format of error message strings has changed
4529 The two C procedures: scm_display_error and scm_error, as well as the
4530 primitive `scm-error', now use scm_simple_format to do their work.
4531 This means that the message strings of all code must be updated to use
4532 ~A where %s was used before, and ~S where %S was used before.
4534 During the period when there still are a lot of old Guiles out there,
4535 you might want to support both old and new versions of Guile.
4537 There are basically two methods to achieve this. Both methods use
4540 AC_CHECK_FUNCS(scm_simple_format)
4542 in your configure.in.
4544 Method 1: Use the string concatenation features of ANSI C's
4549 #ifdef HAVE_SCM_SIMPLE_FORMAT
4555 Then represent each of your error messages using a preprocessor macro:
4557 #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!"
4561 (define fmt-s (if (defined? 'simple-format) "~S" "%S"))
4562 (define make-message string-append)
4564 (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!"))
4566 Method 2: Use the oldfmt function found in doc/oldfmt.c.
4570 scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"),
4575 (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!")
4579 ** Deprecated: coop_mutex_init, coop_condition_variable_init
4581 Don't use the functions coop_mutex_init and
4582 coop_condition_variable_init. They will change.
4584 Use scm_mutex_init and scm_cond_init instead.
4586 ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME)
4587 `scm_cond_timedwait' atomically unlocks MUTEX and waits on
4588 COND, as `scm_cond_wait' does, but it also bounds the duration
4589 of the wait. If COND has not been signaled before time ABSTIME,
4590 the mutex MUTEX is re-acquired and `scm_cond_timedwait'
4591 returns the error code `ETIMEDOUT'.
4593 The ABSTIME parameter specifies an absolute time, with the same
4594 origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds
4595 to 00:00:00 GMT, January 1, 1970.
4597 ** New function: scm_cond_broadcast (scm_cond_t *COND)
4598 `scm_cond_broadcast' restarts all the threads that are waiting
4599 on the condition variable COND. Nothing happens if no threads are
4602 ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *))
4603 `scm_key_create' allocates a new TSD key. The key is stored in
4604 the location pointed to by KEY. There is no limit on the number
4605 of keys allocated at a given time. The value initially associated
4606 with the returned key is `NULL' in all currently executing threads.
4608 The DESTR_FUNCTION argument, if not `NULL', specifies a destructor
4609 function associated with the key. When a thread terminates,
4610 DESTR_FUNCTION is called on the value associated with the key in
4611 that thread. The DESTR_FUNCTION is not called if a key is deleted
4612 with `scm_key_delete' or a value is changed with
4613 `scm_setspecific'. The order in which destructor functions are
4614 called at thread termination time is unspecified.
4616 Destructors are not yet implemented.
4618 ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER)
4619 `scm_setspecific' changes the value associated with KEY in the
4620 calling thread, storing the given POINTER instead.
4622 ** New function: scm_getspecific (scm_key_t KEY)
4623 `scm_getspecific' returns the value currently associated with
4624 KEY in the calling thread.
4626 ** New function: scm_key_delete (scm_key_t KEY)
4627 `scm_key_delete' deallocates a TSD key. It does not check
4628 whether non-`NULL' values are associated with that key in the
4629 currently executing threads, nor call the destructor function
4630 associated with the key.
4632 ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE)
4634 Initialize a C level hook HOOK with associated HOOK_DATA and type
4635 TYPE. (See scm_c_hook_run ().)
4637 ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP)
4639 Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP
4640 is true, add it last, otherwise first. The same FUNC can be added
4641 multiple times if FUNC_DATA differ and vice versa.
4643 ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA)
4645 Remove hook function FUNC with associated FUNC_DATA from HOOK. A
4646 function is only removed if both FUNC and FUNC_DATA matches.
4648 ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA)
4650 Run hook HOOK passing DATA to the hook functions.
4652 If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value
4653 returned is undefined.
4655 If TYPE is SCM_C_HOOK_OR, hook functions are run until a function
4656 returns a non-NULL value. This value is returned as the result of
4657 scm_c_hook_run. If all functions return NULL, NULL is returned.
4659 If TYPE is SCM_C_HOOK_AND, hook functions are run until a function
4660 returns a NULL value, and NULL is returned. If all functions returns
4661 a non-NULL value, the last value is returned.
4663 ** New C level GC hooks
4665 Five new C level hooks has been added to the garbage collector.
4667 scm_before_gc_c_hook
4670 are run before locking and after unlocking the heap. The system is
4671 thus in a mode where evaluation can take place. (Except that
4672 scm_before_gc_c_hook must not allocate new cells.)
4674 scm_before_mark_c_hook
4675 scm_before_sweep_c_hook
4676 scm_after_sweep_c_hook
4678 are run when the heap is locked. These are intended for extension of
4679 the GC in a modular fashion. Examples are the weaks and guardians
4682 ** Way for application to customize GC parameters
4684 The application can set up other default values for the GC heap
4685 allocation parameters
4687 GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1,
4688 GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2,
4689 GUILE_MAX_SEGMENT_SIZE,
4693 scm_default_init_heap_size_1, scm_default_min_yield_1,
4694 scm_default_init_heap_size_2, scm_default_min_yield_2,
4695 scm_default_max_segment_size
4697 respectively before callong scm_boot_guile.
4699 (See entry "New environment variables ..." in section
4700 "Changes to the stand-alone interpreter" above.)
4702 ** scm_protect_object/scm_unprotect_object now nest
4704 This means that you can call scm_protect_object multiple times on an
4705 object and count on the object being protected until
4706 scm_unprotect_object has been call the same number of times.
4708 The functions also have better time complexity.
4710 Still, it is usually possible to structure the application in a way
4711 that you don't need to use these functions. For example, if you use a
4712 protected standard Guile list to keep track of live objects rather
4713 than some custom data type, objects will die a natural death when they
4714 are no longer needed.
4716 ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc
4718 Guile does not provide the float representation for inexact real numbers any
4719 more. Now, only doubles are used to represent inexact real numbers. Further,
4720 the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real
4721 and scm_tc16_complex, respectively.
4723 ** Removed deprecated type scm_smobfuns
4725 ** Removed deprecated function scm_newsmob
4727 ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release
4729 There is an ongoing discussion among the developers whether to
4730 deprecate `scm_make_smob_type_mfpe' or not. Please use the current
4731 standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code
4732 until this issue has been settled.
4734 ** Removed deprecated type tag scm_tc16_kw
4736 ** Added type tag scm_tc16_keyword
4738 (This was introduced already in release 1.3.4 but was not documented
4741 ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized
4743 * Changes to system call interfaces:
4745 ** The "select" procedure now tests port buffers for the ability to
4746 provide input or accept output. Previously only the underlying file
4747 descriptors were checked.
4749 ** New variable PIPE_BUF: the maximum number of bytes that can be
4750 atomically written to a pipe.
4752 ** If a facility is not available on the system when Guile is
4753 compiled, the corresponding primitive procedure will not be defined.
4754 Previously it would have been defined but would throw a system-error
4755 exception if called. Exception handlers which catch this case may
4756 need minor modification: an error will be thrown with key
4757 'unbound-variable instead of 'system-error. Alternatively it's
4758 now possible to use `defined?' to check whether the facility is
4761 ** Procedures which depend on the timezone should now give the correct
4762 result on systems which cache the TZ environment variable, even if TZ
4763 is changed without calling tzset.
4765 * Changes to the networking interfaces:
4767 ** New functions: htons, ntohs, htonl, ntohl: for converting short and
4768 long integers between network and host format. For now, it's not
4769 particularly convenient to do this kind of thing, but consider:
4771 (define write-network-long
4772 (lambda (value port)
4773 (let ((v (make-uniform-vector 1 1 0)))
4774 (uniform-vector-set! v 0 (htonl value))
4775 (uniform-vector-write v port))))
4777 (define read-network-long
4779 (let ((v (make-uniform-vector 1 1 0)))
4780 (uniform-vector-read! v port)
4781 (ntohl (uniform-vector-ref v 0)))))
4783 ** If inet-aton fails, it now throws an error with key 'misc-error
4784 instead of 'system-error, since errno is not relevant.
4786 ** Certain gethostbyname/gethostbyaddr failures now throw errors with
4787 specific keys instead of 'system-error. The latter is inappropriate
4788 since errno will not have been set. The keys are:
4789 'host-not-found, 'try-again, 'no-recovery and 'no-data.
4791 ** sethostent, setnetent, setprotoent, setservent: now take an
4792 optional argument STAYOPEN, which specifies whether the database
4793 remains open after a database entry is accessed randomly (e.g., using
4794 gethostbyname for the hosts database.) The default is #f. Previously
4798 Changes since Guile 1.3.2:
4800 * Changes to the stand-alone interpreter
4804 An initial version of the Guile debugger written by Chris Hanson has
4805 been added. The debugger is still under development but is included
4806 in the distribution anyway since it is already quite useful.
4812 after an error to enter the debugger. Type `help' inside the debugger
4813 for a description of available commands.
4815 If you prefer to have stack frames numbered and printed in
4816 anti-chronological order and prefer up in the stack to be down on the
4817 screen as is the case in gdb, you can put
4819 (debug-enable 'backwards)
4821 in your .guile startup file. (However, this means that Guile can't
4822 use indentation to indicate stack level.)
4824 The debugger is autoloaded into Guile at the first use.
4826 ** Further enhancements to backtraces
4828 There is a new debug option `width' which controls the maximum width
4829 on the screen of printed stack frames. Fancy printing parameters
4830 ("level" and "length" as in Common LISP) are adaptively adjusted for
4831 each stack frame to give maximum information while still fitting
4832 within the bounds. If the stack frame can't be made to fit by
4833 adjusting parameters, it is simply cut off at the end. This is marked
4836 ** Some modules are now only loaded when the repl is started
4838 The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9
4839 regex) are now loaded into (guile-user) only if the repl has been
4840 started. The effect is that the startup time for scripts has been
4841 reduced to 30% of what it was previously.
4843 Correctly written scripts load the modules they require at the top of
4844 the file and should not be affected by this change.
4846 ** Hooks are now represented as smobs
4848 * Changes to Scheme functions and syntax
4850 ** Readline support has changed again.
4852 The old (readline-activator) module is gone. Use (ice-9 readline)
4853 instead, which now contains all readline functionality. So the code
4854 to activate readline is now
4856 (use-modules (ice-9 readline))
4859 This should work at any time, including from the guile prompt.
4861 To avoid confusion about the terms of Guile's license, please only
4862 enable readline for your personal use; please don't make it the
4863 default for others. Here is why we make this rather odd-sounding
4866 Guile is normally licensed under a weakened form of the GNU General
4867 Public License, which allows you to link code with Guile without
4868 placing that code under the GPL. This exception is important to some
4871 However, since readline is distributed under the GNU General Public
4872 License, when you link Guile with readline, either statically or
4873 dynamically, you effectively change Guile's license to the strict GPL.
4874 Whenever you link any strictly GPL'd code into Guile, uses of Guile
4875 which are normally permitted become forbidden. This is a rather
4876 non-obvious consequence of the licensing terms.
4878 So, to make sure things remain clear, please let people choose for
4879 themselves whether to link GPL'd libraries like readline with Guile.
4881 ** regexp-substitute/global has changed slightly, but incompatibly.
4883 If you include a function in the item list, the string of the match
4884 object it receives is the same string passed to
4885 regexp-substitute/global, not some suffix of that string.
4886 Correspondingly, the match's positions are relative to the entire
4887 string, not the suffix.
4889 If the regexp can match the empty string, the way matches are chosen
4890 from the string has changed. regexp-substitute/global recognizes the
4891 same set of matches that list-matches does; see below.
4893 ** New function: list-matches REGEXP STRING [FLAGS]
4895 Return a list of match objects, one for every non-overlapping, maximal
4896 match of REGEXP in STRING. The matches appear in left-to-right order.
4897 list-matches only reports matches of the empty string if there are no
4898 other matches which begin on, end at, or include the empty match's
4901 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
4903 ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS]
4905 For each match of REGEXP in STRING, apply PROC to the match object,
4906 and the last value PROC returned, or INIT for the first call. Return
4907 the last value returned by PROC. We apply PROC to the matches as they
4908 appear from left to right.
4910 This function recognizes matches according to the same criteria as
4913 Thus, you could define list-matches like this:
4915 (define (list-matches regexp string . flags)
4916 (reverse! (apply fold-matches regexp string '() cons flags)))
4918 If present, FLAGS is passed as the FLAGS argument to regexp-exec.
4922 *** New function: hook? OBJ
4924 Return #t if OBJ is a hook, otherwise #f.
4926 *** New function: make-hook-with-name NAME [ARITY]
4928 Return a hook with name NAME and arity ARITY. The default value for
4929 ARITY is 0. The only effect of NAME is that it will appear when the
4930 hook object is printed to ease debugging.
4932 *** New function: hook-empty? HOOK
4934 Return #t if HOOK doesn't contain any procedures, otherwise #f.
4936 *** New function: hook->list HOOK
4938 Return a list of the procedures that are called when run-hook is
4941 ** `map' signals an error if its argument lists are not all the same length.
4943 This is the behavior required by R5RS, so this change is really a bug
4944 fix. But it seems to affect a lot of people's code, so we're
4945 mentioning it here anyway.
4947 ** Print-state handling has been made more transparent
4949 Under certain circumstances, ports are represented as a port with an
4950 associated print state. Earlier, this pair was represented as a pair
4951 (see "Some magic has been added to the printer" below). It is now
4952 indistinguishable (almost; see `get-print-state') from a port on the
4955 *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE
4957 Return a new port with the associated print state PRINT-STATE.
4959 *** New function: get-print-state OUTPUT-PORT
4961 Return the print state associated with this port if it exists,
4962 otherwise return #f.
4964 *** New function: directory-stream? OBJECT
4966 Returns true iff OBJECT is a directory stream --- the sort of object
4967 returned by `opendir'.
4969 ** New function: using-readline?
4971 Return #t if readline is in use in the current repl.
4973 ** structs will be removed in 1.4
4975 Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile
4976 and use GOOPS objects as the fundamental record type.
4978 * Changes to the scm_ interface
4980 ** structs will be removed in 1.4
4982 The entire current struct interface (struct.c, struct.h) will be
4983 replaced in Guile 1.4. We will merge GOOPS into libguile and use
4984 GOOPS objects as the fundamental record type.
4986 ** The internal representation of subr's has changed
4988 Instead of giving a hint to the subr name, the CAR field of the subr
4989 now contains an index to a subr entry in scm_subr_table.
4991 *** New variable: scm_subr_table
4993 An array of subr entries. A subr entry contains the name, properties
4994 and documentation associated with the subr. The properties and
4995 documentation slots are not yet used.
4997 ** A new scheme for "forwarding" calls to a builtin to a generic function
4999 It is now possible to extend the functionality of some Guile
5000 primitives by letting them defer a call to a GOOPS generic function on
5001 argument mismatch. This means that there is no loss of efficiency in
5006 (use-modules (oop goops)) ; Must be GOOPS version 0.2.
5007 (define-method + ((x <string>) (y <string>))
5008 (string-append x y))
5010 + will still be as efficient as usual in numerical calculations, but
5011 can also be used for concatenating strings.
5013 Who will be the first one to extend Guile's numerical tower to
5014 rationals? :) [OK, there a few other things to fix before this can
5015 be made in a clean way.]
5017 *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1
5019 New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC)
5021 New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC)
5023 These do the same job as SCM_PROC and SCM_PROC1, but they also define
5024 a variable GENERIC which can be used by the dispatch macros below.
5026 [This is experimental code which may change soon.]
5028 *** New macros for forwarding control to a generic on arg type error
5030 New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR)
5032 New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR)
5034 These correspond to the scm_wta function call, and have the same
5035 behaviour until the user has called the GOOPS primitive
5036 `enable-primitive-generic!'. After that, these macros will apply the
5037 generic function GENERIC to the argument(s) instead of calling
5040 [This is experimental code which may change soon.]
5042 *** New macros for argument testing with generic dispatch
5044 New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR)
5046 New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR)
5048 These correspond to the SCM_ASSERT macro, but will defer control to
5049 GENERIC on error after `enable-primitive-generic!' has been called.
5051 [This is experimental code which may change soon.]
5053 ** New function: SCM scm_eval_body (SCM body, SCM env)
5055 Evaluates the body of a special form.
5057 ** The internal representation of struct's has changed
5059 Previously, four slots were allocated for the procedure(s) of entities
5060 and operators. The motivation for this representation had to do with
5061 the structure of the evaluator, the wish to support tail-recursive
5062 generic functions, and efficiency. Since the generic function
5063 dispatch mechanism has changed, there is no longer a need for such an
5064 expensive representation, and the representation has been simplified.
5066 This should not make any difference for most users.
5068 ** GOOPS support has been cleaned up.
5070 Some code has been moved from eval.c to objects.c and code in both of
5071 these compilation units has been cleaned up and better structured.
5073 *** New functions for applying generic functions
5075 New function: SCM scm_apply_generic (GENERIC, ARGS)
5076 New function: SCM scm_call_generic_0 (GENERIC)
5077 New function: SCM scm_call_generic_1 (GENERIC, ARG1)
5078 New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2)
5079 New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3)
5081 ** Deprecated function: scm_make_named_hook
5083 It is now replaced by:
5085 ** New function: SCM scm_create_hook (const char *name, int arity)
5087 Creates a hook in the same way as make-hook above but also
5088 binds a variable named NAME to it.
5090 This is the typical way of creating a hook from C code.
5092 Currently, the variable is created in the "current" module.
5093 This might change when we get the new module system.
5095 [The behaviour is identical to scm_make_named_hook.]
5099 Changes since Guile 1.3:
5101 * Changes to mailing lists
5103 ** Some of the Guile mailing lists have moved to sourceware.cygnus.com.
5105 See the README file to find current addresses for all the Guile
5108 * Changes to the distribution
5110 ** Readline support is no longer included with Guile by default.
5112 Based on the different license terms of Guile and Readline, we
5113 concluded that Guile should not *by default* cause the linking of
5114 Readline into an application program. Readline support is now offered
5115 as a separate module, which is linked into an application only when
5116 you explicitly specify it.
5118 Although Guile is GNU software, its distribution terms add a special
5119 exception to the usual GNU General Public License (GPL). Guile's
5120 license includes a clause that allows you to link Guile with non-free
5121 programs. We add this exception so as not to put Guile at a
5122 disadvantage vis-a-vis other extensibility packages that support other
5125 In contrast, the GNU Readline library is distributed under the GNU
5126 General Public License pure and simple. This means that you may not
5127 link Readline, even dynamically, into an application unless it is
5128 distributed under a free software license that is compatible the GPL.
5130 Because of this difference in distribution terms, an application that
5131 can use Guile may not be able to use Readline. Now users will be
5132 explicitly offered two independent decisions about the use of these
5135 You can activate the readline support by issuing
5137 (use-modules (readline-activator))
5140 from your ".guile" file, for example.
5142 * Changes to the stand-alone interpreter
5144 ** All builtins now print as primitives.
5145 Previously builtin procedures not belonging to the fundamental subr
5146 types printed as #<compiled closure #<primitive-procedure gsubr-apply>>.
5147 Now, they print as #<primitive-procedure NAME>.
5149 ** Backtraces slightly more intelligible.
5150 gsubr-apply and macro transformer application frames no longer appear
5153 * Changes to Scheme functions and syntax
5155 ** Guile now correctly handles internal defines by rewriting them into
5156 their equivalent letrec. Previously, internal defines would
5157 incrementally add to the innermost environment, without checking
5158 whether the restrictions specified in RnRS were met. This lead to the
5159 correct behaviour when these restriction actually were met, but didn't
5160 catch all illegal uses. Such an illegal use could lead to crashes of
5161 the Guile interpreter or or other unwanted results. An example of
5162 incorrect internal defines that made Guile behave erratically:
5174 The problem with this example is that the definition of `c' uses the
5175 value of `b' directly. This confuses the meoization machine of Guile
5176 so that the second call of `b' (this time in a larger environment that
5177 also contains bindings for `c' and `d') refers to the binding of `c'
5178 instead of `a'. You could also make Guile crash with a variation on
5183 (define (b flag) (if flag a 1))
5184 (define c (1+ (b flag)))
5192 From now on, Guile will issue an `Unbound variable: b' error message
5197 A hook contains a list of functions which should be called on
5198 particular occasions in an existing program. Hooks are used for
5201 A window manager might have a hook before-window-map-hook. The window
5202 manager uses the function run-hooks to call all functions stored in
5203 before-window-map-hook each time a window is mapped. The user can
5204 store functions in the hook using add-hook!.
5206 In Guile, hooks are first class objects.
5208 *** New function: make-hook [N_ARGS]
5210 Return a hook for hook functions which can take N_ARGS arguments.
5211 The default value for N_ARGS is 0.
5213 (See also scm_make_named_hook below.)
5215 *** New function: add-hook! HOOK PROC [APPEND_P]
5217 Put PROC at the beginning of the list of functions stored in HOOK.
5218 If APPEND_P is supplied, and non-false, put PROC at the end instead.
5220 PROC must be able to take the number of arguments specified when the
5223 If PROC already exists in HOOK, then remove it first.
5225 *** New function: remove-hook! HOOK PROC
5227 Remove PROC from the list of functions in HOOK.
5229 *** New function: reset-hook! HOOK
5231 Clear the list of hook functions stored in HOOK.
5233 *** New function: run-hook HOOK ARG1 ...
5235 Run all hook functions stored in HOOK with arguments ARG1 ... .
5236 The number of arguments supplied must correspond to the number given
5237 when the hook was created.
5239 ** The function `dynamic-link' now takes optional keyword arguments.
5240 The only keyword argument that is currently defined is `:global
5241 BOOL'. With it, you can control whether the shared library will be
5242 linked in global mode or not. In global mode, the symbols from the
5243 linked library can be used to resolve references from other
5244 dynamically linked libraries. In non-global mode, the linked
5245 library is essentially invisible and can only be accessed via
5246 `dynamic-func', etc. The default is now to link in global mode.
5247 Previously, the default has been non-global mode.
5249 The `#:global' keyword is only effective on platforms that support
5250 the dlopen family of functions.
5252 ** New function `provided?'
5254 - Function: provided? FEATURE
5255 Return true iff FEATURE is supported by this installation of
5256 Guile. FEATURE must be a symbol naming a feature; the global
5257 variable `*features*' is a list of available features.
5259 ** Changes to the module (ice-9 expect):
5261 *** The expect-strings macro now matches `$' in a regular expression
5262 only at a line-break or end-of-file by default. Previously it would
5263 match the end of the string accumulated so far. The old behaviour
5264 can be obtained by setting the variable `expect-strings-exec-flags'
5267 *** The expect-strings macro now uses a variable `expect-strings-exec-flags'
5268 for the regexp-exec flags. If `regexp/noteol' is included, then `$'
5269 in a regular expression will still match before a line-break or
5270 end-of-file. The default is `regexp/noteol'.
5272 *** The expect-strings macro now uses a variable
5273 `expect-strings-compile-flags' for the flags to be supplied to
5274 `make-regexp'. The default is `regexp/newline', which was previously
5277 *** The expect macro now supplies two arguments to a match procedure:
5278 the current accumulated string and a flag to indicate whether
5279 end-of-file has been reached. Previously only the string was supplied.
5280 If end-of-file is reached, the match procedure will be called an
5281 additional time with the same accumulated string as the previous call
5282 but with the flag set.
5284 ** New module (ice-9 format), implementing the Common Lisp `format' function.
5286 This code, and the documentation for it that appears here, was
5287 borrowed from SLIB, with minor adaptations for Guile.
5289 - Function: format DESTINATION FORMAT-STRING . ARGUMENTS
5290 An almost complete implementation of Common LISP format description
5291 according to the CL reference book `Common LISP' from Guy L.
5292 Steele, Digital Press. Backward compatible to most of the
5293 available Scheme format implementations.
5295 Returns `#t', `#f' or a string; has side effect of printing
5296 according to FORMAT-STRING. If DESTINATION is `#t', the output is
5297 to the current output port and `#t' is returned. If DESTINATION
5298 is `#f', a formatted string is returned as the result of the call.
5299 NEW: If DESTINATION is a string, DESTINATION is regarded as the
5300 format string; FORMAT-STRING is then the first argument and the
5301 output is returned as a string. If DESTINATION is a number, the
5302 output is to the current error port if available by the
5303 implementation. Otherwise DESTINATION must be an output port and
5306 FORMAT-STRING must be a string. In case of a formatting error
5307 format returns `#f' and prints a message on the current output or
5308 error port. Characters are output as if the string were output by
5309 the `display' function with the exception of those prefixed by a
5310 tilde (~). For a detailed description of the FORMAT-STRING syntax
5311 please consult a Common LISP format reference manual. For a test
5312 suite to verify this format implementation load `formatst.scm'.
5313 Please send bug reports to `lutzeb@cs.tu-berlin.de'.
5315 Note: `format' is not reentrant, i.e. only one `format'-call may
5316 be executed at a time.
5319 *** Format Specification (Format version 3.0)
5321 Please consult a Common LISP format reference manual for a detailed
5322 description of the format string syntax. For a demonstration of the
5323 implemented directives see `formatst.scm'.
5325 This implementation supports directive parameters and modifiers (`:'
5326 and `@' characters). Multiple parameters must be separated by a comma
5327 (`,'). Parameters can be numerical parameters (positive or negative),
5328 character parameters (prefixed by a quote character (`''), variable
5329 parameters (`v'), number of rest arguments parameter (`#'), empty and
5330 default parameters. Directive characters are case independent. The
5331 general form of a directive is:
5333 DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER
5335 DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ]
5337 *** Implemented CL Format Control Directives
5339 Documentation syntax: Uppercase characters represent the
5340 corresponding control directive characters. Lowercase characters
5341 represent control directive parameter descriptions.
5344 Any (print as `display' does).
5348 `~MINCOL,COLINC,MINPAD,PADCHARA'
5352 S-expression (print as `write' does).
5356 `~MINCOL,COLINC,MINPAD,PADCHARS'
5362 print number sign always.
5365 print comma separated.
5367 `~MINCOL,PADCHAR,COMMACHARD'
5373 print number sign always.
5376 print comma separated.
5378 `~MINCOL,PADCHAR,COMMACHARX'
5384 print number sign always.
5387 print comma separated.
5389 `~MINCOL,PADCHAR,COMMACHARO'
5395 print number sign always.
5398 print comma separated.
5400 `~MINCOL,PADCHAR,COMMACHARB'
5405 `~N,MINCOL,PADCHAR,COMMACHARR'
5409 print a number as a Roman numeral.
5412 print a number as an "old fashioned" Roman numeral.
5415 print a number as an ordinal English number.
5418 print a number as a cardinal English number.
5423 prints `y' and `ies'.
5426 as `~P but jumps 1 argument backward.'
5429 as `~@P but jumps 1 argument backward.'
5434 prints a character as the reader can understand it (i.e. `#\'
5438 prints a character as emacs does (eg. `^C' for ASCII 03).
5441 Fixed-format floating-point (prints a flonum like MMM.NNN).
5442 `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF'
5444 If the number is positive a plus sign is printed.
5447 Exponential floating-point (prints a flonum like MMM.NNN`E'EE).
5448 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE'
5450 If the number is positive a plus sign is printed.
5453 General floating-point (prints a flonum either fixed or
5455 `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG'
5457 If the number is positive a plus sign is printed.
5460 Dollars floating-point (prints a flonum in fixed with signs
5462 `~DIGITS,SCALE,WIDTH,PADCHAR$'
5464 If the number is positive a plus sign is printed.
5467 A sign is always printed and appears before the padding.
5470 The sign appears before the padding.
5478 print newline if not at the beginning of the output line.
5480 prints `~&' and then N-1 newlines.
5485 print N page separators.
5495 newline is ignored, white space left.
5498 newline is left, white space ignored.
5503 relative tabulation.
5509 Indirection (expects indirect arguments as a list).
5511 extracts indirect arguments from format arguments.
5514 Case conversion (converts by `string-downcase').
5516 converts by `string-capitalize'.
5519 converts by `string-capitalize-first'.
5522 converts by `string-upcase'.
5525 Argument Jumping (jumps 1 argument forward).
5527 jumps N arguments forward.
5530 jumps 1 argument backward.
5533 jumps N arguments backward.
5536 jumps to the 0th argument.
5539 jumps to the Nth argument (beginning from 0)
5541 `~[STR0~;STR1~;...~;STRN~]'
5542 Conditional Expression (numerical clause conditional).
5544 take argument from N.
5547 true test conditional.
5550 if-else-then conditional.
5556 default clause follows.
5559 Iteration (args come from the next argument (a list)).
5561 at most N iterations.
5564 args from next arg (a list of lists).
5567 args from the rest of arguments.
5570 args from the rest args (lists).
5581 aborts if N <= M <= K
5583 *** Not Implemented CL Format Control Directives
5586 print `#f' as an empty list (see below).
5589 print `#f' as an empty list (see below).
5595 (sorry I don't understand its semantics completely)
5597 *** Extended, Replaced and Additional Control Directives
5599 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD'
5600 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX'
5601 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO'
5602 `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB'
5603 `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR'
5604 COMMAWIDTH is the number of characters between two comma
5608 print a R4RS complex number as `~F~@Fi' with passed parameters for
5612 Pretty print formatting of an argument for scheme code lists.
5618 Flushes the output if format DESTINATION is a port.
5621 Print a `#\space' character
5623 print N `#\space' characters.
5626 Print a `#\tab' character
5628 print N `#\tab' characters.
5631 Takes N as an integer representation for a character. No arguments
5632 are consumed. N is converted to a character by `integer->char'. N
5633 must be a positive decimal number.
5636 Print out readproof. Prints out internal objects represented as
5637 `#<...>' as strings `"#<...>"' so that the format output can always
5638 be processed by `read'.
5641 Print out readproof. Prints out internal objects represented as
5642 `#<...>' as strings `"#<...>"' so that the format output can always
5643 be processed by `read'.
5646 Prints information and a copyright notice on the format
5649 prints format version.
5652 may also print number strings, i.e. passing a number as a string
5653 and format it accordingly.
5655 *** Configuration Variables
5657 The format module exports some configuration variables to suit the
5658 systems and users needs. There should be no modification necessary for
5659 the configuration that comes with Guile. Format detects automatically
5660 if the running scheme system implements floating point numbers and
5663 format:symbol-case-conv
5664 Symbols are converted by `symbol->string' so the case type of the
5665 printed symbols is implementation dependent.
5666 `format:symbol-case-conv' is a one arg closure which is either
5667 `#f' (no conversion), `string-upcase', `string-downcase' or
5668 `string-capitalize'. (default `#f')
5670 format:iobj-case-conv
5671 As FORMAT:SYMBOL-CASE-CONV but applies for the representation of
5672 implementation internal objects. (default `#f')
5675 The character prefixing the exponent value in `~E' printing.
5678 *** Compatibility With Other Format Implementations
5684 Downward compatible except for padding support and `~A', `~S',
5685 `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style
5686 `printf' padding support which is completely replaced by the CL
5687 `format' padding style.
5690 Downward compatible except for `~', which is not documented
5691 (ignores all characters inside the format string up to a newline
5692 character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%',
5693 numerical and variable parameters and `:/@' modifiers in the CL
5697 Downward compatible except for `~A' and `~S' which print in
5698 uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no
5699 directive parameters or modifiers)).
5702 Downward compatible except for an optional destination parameter:
5703 S2C accepts a format call without a destination which returns a
5704 formatted string. This is equivalent to a #f destination in S2C.
5705 (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive
5706 parameters or modifiers)).
5709 ** Changes to string-handling functions.
5711 These functions were added to support the (ice-9 format) module, above.
5713 *** New function: string-upcase STRING
5714 *** New function: string-downcase STRING
5716 These are non-destructive versions of the existing string-upcase! and
5717 string-downcase! functions.
5719 *** New function: string-capitalize! STRING
5720 *** New function: string-capitalize STRING
5722 These functions convert the first letter of each word in the string to
5725 (string-capitalize "howdy there")
5728 As with the other functions, string-capitalize! modifies the string in
5729 place, while string-capitalize returns a modified copy of its argument.
5731 *** New function: string-ci->symbol STRING
5733 Return a symbol whose name is STRING, but having the same case as if
5734 the symbol had be read by `read'.
5736 Guile can be configured to be sensitive or insensitive to case
5737 differences in Scheme identifiers. If Guile is case-insensitive, all
5738 symbols are converted to lower case on input. The `string-ci->symbol'
5739 function returns a symbol whose name in STRING, transformed as Guile
5740 would if STRING were input.
5742 *** New function: substring-move! STRING1 START END STRING2 START
5744 Copy the substring of STRING1 from START (inclusive) to END
5745 (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same
5746 string, and the source and destination areas may overlap; in all
5747 cases, the function behaves as if all the characters were copied
5750 *** Extended functions: substring-move-left! substring-move-right!
5752 These functions now correctly copy arbitrarily overlapping substrings;
5753 they are both synonyms for substring-move!.
5756 ** New module (ice-9 getopt-long), with the function `getopt-long'.
5758 getopt-long is a function for parsing command-line arguments in a
5759 manner consistent with other GNU programs.
5761 (getopt-long ARGS GRAMMAR)
5762 Parse the arguments ARGS according to the argument list grammar GRAMMAR.
5764 ARGS should be a list of strings. Its first element should be the
5765 name of the program; subsequent elements should be the arguments
5766 that were passed to the program on the command line. The
5767 `program-arguments' procedure returns a list of this form.
5769 GRAMMAR is a list of the form:
5770 ((OPTION (PROPERTY VALUE) ...) ...)
5772 Each OPTION should be a symbol. `getopt-long' will accept a
5773 command-line option named `--OPTION'.
5774 Each option can have the following (PROPERTY VALUE) pairs:
5776 (single-char CHAR) --- Accept `-CHAR' as a single-character
5777 equivalent to `--OPTION'. This is how to specify traditional
5779 (required? BOOL) --- If BOOL is true, the option is required.
5780 getopt-long will raise an error if it is not found in ARGS.
5781 (value BOOL) --- If BOOL is #t, the option accepts a value; if
5782 it is #f, it does not; and if it is the symbol
5783 `optional', the option may appear in ARGS with or
5785 (predicate FUNC) --- If the option accepts a value (i.e. you
5786 specified `(value #t)' for this option), then getopt
5787 will apply FUNC to the value, and throw an exception
5788 if it returns #f. FUNC should be a procedure which
5789 accepts a string and returns a boolean value; you may
5790 need to use quasiquotes to get it into GRAMMAR.
5792 The (PROPERTY VALUE) pairs may occur in any order, but each
5793 property may occur only once. By default, options do not have
5794 single-character equivalents, are not required, and do not take
5797 In ARGS, single-character options may be combined, in the usual
5798 Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option
5799 accepts values, then it must be the last option in the
5800 combination; the value is the next argument. So, for example, using
5801 the following grammar:
5802 ((apples (single-char #\a))
5803 (blimps (single-char #\b) (value #t))
5804 (catalexis (single-char #\c) (value #t)))
5805 the following argument lists would be acceptable:
5806 ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values
5807 for "blimps" and "catalexis")
5808 ("-ab" "bang" "-c" "couth") (same)
5809 ("-ac" "couth" "-b" "bang") (same)
5810 ("-abc" "couth" "bang") (an error, since `-b' is not the
5811 last option in its combination)
5813 If an option's value is optional, then `getopt-long' decides
5814 whether it has a value by looking at what follows it in ARGS. If
5815 the next element is a string, and it does not appear to be an
5816 option itself, then that string is the option's value.
5818 The value of a long option can appear as the next element in ARGS,
5819 or it can follow the option name, separated by an `=' character.
5820 Thus, using the same grammar as above, the following argument lists
5822 ("--apples" "Braeburn" "--blimps" "Goodyear")
5823 ("--apples=Braeburn" "--blimps" "Goodyear")
5824 ("--blimps" "Goodyear" "--apples=Braeburn")
5826 If the option "--" appears in ARGS, argument parsing stops there;
5827 subsequent arguments are returned as ordinary arguments, even if
5828 they resemble options. So, in the argument list:
5829 ("--apples" "Granny Smith" "--" "--blimp" "Goodyear")
5830 `getopt-long' will recognize the `apples' option as having the
5831 value "Granny Smith", but it will not recognize the `blimp'
5832 option; it will return the strings "--blimp" and "Goodyear" as
5833 ordinary argument strings.
5835 The `getopt-long' function returns the parsed argument list as an
5836 assocation list, mapping option names --- the symbols from GRAMMAR
5837 --- onto their values, or #t if the option does not accept a value.
5838 Unused options do not appear in the alist.
5840 All arguments that are not the value of any option are returned
5841 as a list, associated with the empty list.
5843 `getopt-long' throws an exception if:
5844 - it finds an unrecognized option in ARGS
5845 - a required option is omitted
5846 - an option that requires an argument doesn't get one
5847 - an option that doesn't accept an argument does get one (this can
5848 only happen using the long option `--opt=value' syntax)
5849 - an option predicate fails
5854 `((lockfile-dir (required? #t)
5857 (predicate ,file-is-directory?))
5858 (verbose (required? #f)
5861 (x-includes (single-char #\x))
5862 (rnet-server (single-char #\y)
5863 (predicate ,string?))))
5865 (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include"
5866 "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3")
5868 => ((() "foo1" "-fred" "foo2" "foo3")
5869 (rnet-server . "lamprod")
5870 (x-includes . "/usr/include")
5871 (lockfile-dir . "/tmp")
5874 ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long).
5876 It will be removed in a few releases.
5878 ** New syntax: lambda*
5879 ** New syntax: define*
5880 ** New syntax: define*-public
5881 ** New syntax: defmacro*
5882 ** New syntax: defmacro*-public
5883 Guile now supports optional arguments.
5885 `lambda*', `define*', `define*-public', `defmacro*' and
5886 `defmacro*-public' are identical to the non-* versions except that
5887 they use an extended type of parameter list that has the following BNF
5888 syntax (parentheses are literal, square brackets indicate grouping,
5889 and `*', `+' and `?' have the usual meaning):
5891 ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]?
5892 [#&key [ext-var-decl]+ [#&allow-other-keys]?]?
5893 [[#&rest identifier]|[. identifier]]? ) | [identifier]
5895 ext-var-decl ::= identifier | ( identifier expression )
5897 The semantics are best illustrated with the following documentation
5898 and examples for `lambda*':
5901 lambda extended for optional and keyword arguments
5903 lambda* creates a procedure that takes optional arguments. These
5904 are specified by putting them inside brackets at the end of the
5905 paramater list, but before any dotted rest argument. For example,
5906 (lambda* (a b #&optional c d . e) '())
5907 creates a procedure with fixed arguments a and b, optional arguments c
5908 and d, and rest argument e. If the optional arguments are omitted
5909 in a call, the variables for them are unbound in the procedure. This
5910 can be checked with the bound? macro.
5912 lambda* can also take keyword arguments. For example, a procedure
5914 (lambda* (#&key xyzzy larch) '())
5915 can be called with any of the argument lists (#:xyzzy 11)
5916 (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments
5917 are given as keywords are bound to values.
5919 Optional and keyword arguments can also be given default values
5920 which they take on when they are not present in a call, by giving a
5921 two-item list in place of an optional argument, for example in:
5922 (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz))
5923 foo is a fixed argument, bar is an optional argument with default
5924 value 42, and baz is a keyword argument with default value 73.
5925 Default value expressions are not evaluated unless they are needed
5926 and until the procedure is called.
5928 lambda* now supports two more special parameter list keywords.
5930 lambda*-defined procedures now throw an error by default if a
5931 keyword other than one of those specified is found in the actual
5932 passed arguments. However, specifying #&allow-other-keys
5933 immediately after the kyword argument declarations restores the
5934 previous behavior of ignoring unknown keywords. lambda* also now
5935 guarantees that if the same keyword is passed more than once, the
5936 last one passed is the one that takes effect. For example,
5937 ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails)))
5938 #:heads 37 #:tails 42 #:heads 99)
5939 would result in (99 47) being displayed.
5941 #&rest is also now provided as a synonym for the dotted syntax rest
5942 argument. The argument lists (a . b) and (a #&rest b) are equivalent in
5943 all respects to lambda*. This is provided for more similarity to DSSSL,
5944 MIT-Scheme and Kawa among others, as well as for refugees from other
5947 Further documentation may be found in the optargs.scm file itself.
5949 The optional argument module also exports the macros `let-optional',
5950 `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These
5951 are not documented here because they may be removed in the future, but
5952 full documentation is still available in optargs.scm.
5954 ** New syntax: and-let*
5955 Guile now supports the `and-let*' form, described in the draft SRFI-2.
5957 Syntax: (land* (<clause> ...) <body> ...)
5958 Each <clause> should have one of the following forms:
5959 (<variable> <expression>)
5962 Each <variable> or <bound-variable> should be an identifier. Each
5963 <expression> should be a valid expression. The <body> should be a
5964 possibly empty sequence of expressions, like the <body> of a
5967 Semantics: A LAND* expression is evaluated by evaluating the
5968 <expression> or <bound-variable> of each of the <clause>s from
5969 left to right. The value of the first <expression> or
5970 <bound-variable> that evaluates to a false value is returned; the
5971 remaining <expression>s and <bound-variable>s are not evaluated.
5972 The <body> forms are evaluated iff all the <expression>s and
5973 <bound-variable>s evaluate to true values.
5975 The <expression>s and the <body> are evaluated in an environment
5976 binding each <variable> of the preceding (<variable> <expression>)
5977 clauses to the value of the <expression>. Later bindings
5978 shadow earlier bindings.
5980 Guile's and-let* macro was contributed by Michael Livshin.
5982 ** New sorting functions
5984 *** New function: sorted? SEQUENCE LESS?
5985 Returns `#t' when the sequence argument is in non-decreasing order
5986 according to LESS? (that is, there is no adjacent pair `... x y
5987 ...' for which `(less? y x)').
5989 Returns `#f' when the sequence contains at least one out-of-order
5990 pair. It is an error if the sequence is neither a list nor a
5993 *** New function: merge LIST1 LIST2 LESS?
5994 LIST1 and LIST2 are sorted lists.
5995 Returns the sorted list of all elements in LIST1 and LIST2.
5997 Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal"
5998 in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2},
5999 and that a < b1 in LIST1. Then a < b1 < b2 in the result.
6000 (Here "<" should read "comes before".)
6002 *** New procedure: merge! LIST1 LIST2 LESS?
6003 Merges two lists, re-using the pairs of LIST1 and LIST2 to build
6004 the result. If the code is compiled, and LESS? constructs no new
6005 pairs, no pairs at all will be allocated. The first pair of the
6006 result will be either the first pair of LIST1 or the first pair of
6009 *** New function: sort SEQUENCE LESS?
6010 Accepts either a list or a vector, and returns a new sequence
6011 which is sorted. The new sequence is the same type as the input.
6012 Always `(sorted? (sort sequence less?) less?)'. The original
6013 sequence is not altered in any way. The new sequence shares its
6014 elements with the old one; no elements are copied.
6016 *** New procedure: sort! SEQUENCE LESS
6017 Returns its sorted result in the original boxes. No new storage is
6018 allocated at all. Proper usage: (set! slist (sort! slist <))
6020 *** New function: stable-sort SEQUENCE LESS?
6021 Similar to `sort' but stable. That is, if "equal" elements are
6022 ordered a < b in the original sequence, they will have the same order
6025 *** New function: stable-sort! SEQUENCE LESS?
6026 Similar to `sort!' but stable.
6027 Uses temporary storage when sorting vectors.
6029 *** New functions: sort-list, sort-list!
6030 Added for compatibility with scsh.
6032 ** New built-in random number support
6034 *** New function: random N [STATE]
6035 Accepts a positive integer or real N and returns a number of the
6036 same type between zero (inclusive) and N (exclusive). The values
6037 returned have a uniform distribution.
6039 The optional argument STATE must be of the type produced by
6040 `copy-random-state' or `seed->random-state'. It defaults to the value
6041 of the variable `*random-state*'. This object is used to maintain the
6042 state of the pseudo-random-number generator and is altered as a side
6043 effect of the `random' operation.
6045 *** New variable: *random-state*
6046 Holds a data structure that encodes the internal state of the
6047 random-number generator that `random' uses by default. The nature
6048 of this data structure is implementation-dependent. It may be
6049 printed out and successfully read back in, but may or may not
6050 function correctly as a random-number state object in another
6053 *** New function: copy-random-state [STATE]
6054 Returns a new object of type suitable for use as the value of the
6055 variable `*random-state*' and as a second argument to `random'.
6056 If argument STATE is given, a copy of it is returned. Otherwise a
6057 copy of `*random-state*' is returned.
6059 *** New function: seed->random-state SEED
6060 Returns a new object of type suitable for use as the value of the
6061 variable `*random-state*' and as a second argument to `random'.
6062 SEED is a string or a number. A new state is generated and
6063 initialized using SEED.
6065 *** New function: random:uniform [STATE]
6066 Returns an uniformly distributed inexact real random number in the
6067 range between 0 and 1.
6069 *** New procedure: random:solid-sphere! VECT [STATE]
6070 Fills VECT with inexact real random numbers the sum of whose
6071 squares is less than 1.0. Thinking of VECT as coordinates in
6072 space of dimension N = `(vector-length VECT)', the coordinates are
6073 uniformly distributed within the unit N-shere. The sum of the
6074 squares of the numbers is returned. VECT can be either a vector
6075 or a uniform vector of doubles.
6077 *** New procedure: random:hollow-sphere! VECT [STATE]
6078 Fills VECT with inexact real random numbers the sum of whose squares
6079 is equal to 1.0. Thinking of VECT as coordinates in space of
6080 dimension n = `(vector-length VECT)', the coordinates are uniformly
6081 distributed over the surface of the unit n-shere. VECT can be either
6082 a vector or a uniform vector of doubles.
6084 *** New function: random:normal [STATE]
6085 Returns an inexact real in a normal distribution with mean 0 and
6086 standard deviation 1. For a normal distribution with mean M and
6087 standard deviation D use `(+ M (* D (random:normal)))'.
6089 *** New procedure: random:normal-vector! VECT [STATE]
6090 Fills VECT with inexact real random numbers which are independent and
6091 standard normally distributed (i.e., with mean 0 and variance 1).
6092 VECT can be either a vector or a uniform vector of doubles.
6094 *** New function: random:exp STATE
6095 Returns an inexact real in an exponential distribution with mean 1.
6096 For an exponential distribution with mean U use (* U (random:exp)).
6098 ** The range of logand, logior, logxor, logtest, and logbit? have changed.
6100 These functions now operate on numbers in the range of a C unsigned
6103 These functions used to operate on numbers in the range of a C signed
6104 long; however, this seems inappropriate, because Guile integers don't
6107 ** New function: make-guardian
6108 This is an implementation of guardians as described in
6109 R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a
6110 Generation-Based Garbage Collector" ACM SIGPLAN Conference on
6111 Programming Language Design and Implementation, June 1993
6112 ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz
6114 ** New functions: delq1!, delv1!, delete1!
6115 These procedures behave similar to delq! and friends but delete only
6116 one object if at all.
6118 ** New function: unread-string STRING PORT
6119 Unread STRING to PORT, that is, push it back onto the port so that
6120 next read operation will work on the pushed back characters.
6122 ** unread-char can now be called multiple times
6123 If unread-char is called multiple times, the unread characters will be
6124 read again in last-in first-out order.
6126 ** the procedures uniform-array-read! and uniform-array-write! now
6127 work on any kind of port, not just ports which are open on a file.
6129 ** Now 'l' in a port mode requests line buffering.
6131 ** The procedure truncate-file now works on string ports as well
6132 as file ports. If the size argument is omitted, the current
6133 file position is used.
6135 ** new procedure: seek PORT/FDES OFFSET WHENCE
6136 The arguments are the same as for the old fseek procedure, but it
6137 works on string ports as well as random-access file ports.
6139 ** the fseek procedure now works on string ports, since it has been
6140 redefined using seek.
6142 ** the setvbuf procedure now uses a default size if mode is _IOFBF and
6143 size is not supplied.
6145 ** the newline procedure no longer flushes the port if it's not
6146 line-buffered: previously it did if it was the current output port.
6148 ** open-pipe and close-pipe are no longer primitive procedures, but
6149 an emulation can be obtained using `(use-modules (ice-9 popen))'.
6151 ** the freopen procedure has been removed.
6153 ** new procedure: drain-input PORT
6154 Drains PORT's read buffers (including any pushed-back characters)
6155 and returns the contents as a single string.
6157 ** New function: map-in-order PROC LIST1 LIST2 ...
6158 Version of `map' which guarantees that the procedure is applied to the
6159 lists in serial order.
6161 ** Renamed `serial-array-copy!' and `serial-array-map!' to
6162 `array-copy-in-order!' and `array-map-in-order!'. The old names are
6163 now obsolete and will go away in release 1.5.
6165 ** New syntax: collect BODY1 ...
6166 Version of `begin' which returns a list of the results of the body
6167 forms instead of the result of the last body form. In contrast to
6168 `begin', `collect' allows an empty body.
6170 ** New functions: read-history FILENAME, write-history FILENAME
6171 Read/write command line history from/to file. Returns #t on success
6172 and #f if an error occured.
6174 ** `ls' and `lls' in module (ice-9 ls) now handle no arguments.
6176 These procedures return a list of definitions available in the specified
6177 argument, a relative module reference. In the case of no argument,
6178 `(current-module)' is now consulted for definitions to return, instead
6179 of simply returning #f, the former behavior.
6181 ** The #/ syntax for lists is no longer supported.
6183 Earlier versions of Scheme accepted this syntax, but printed a
6186 ** Guile no longer consults the SCHEME_LOAD_PATH environment variable.
6188 Instead, you should set GUILE_LOAD_PATH to tell Guile where to find
6191 * Changes to the gh_ interface
6195 Now takes a second argument which is the result array. If this
6196 pointer is NULL, a new array is malloced (the old behaviour).
6198 ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars,
6199 gh_scm2shorts, gh_scm2longs, gh_scm2floats
6203 * Changes to the scm_ interface
6205 ** Function: scm_make_named_hook (char* name, int n_args)
6207 Creates a hook in the same way as make-hook above but also
6208 binds a variable named NAME to it.
6210 This is the typical way of creating a hook from C code.
6212 Currently, the variable is created in the "current" module. This
6213 might change when we get the new module system.
6215 ** The smob interface
6217 The interface for creating smobs has changed. For documentation, see
6218 data-rep.info (made from guile-core/doc/data-rep.texi).
6220 *** Deprecated function: SCM scm_newsmob (scm_smobfuns *)
6222 >>> This function will be removed in 1.3.4. <<<
6226 *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size)
6227 This function adds a new smob type, named NAME, with instance size
6228 SIZE to the system. The return value is a tag that is used in
6229 creating instances of the type. If SIZE is 0, then no memory will
6230 be allocated when instances of the smob are created, and nothing
6231 will be freed by the default free function.
6233 *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM))
6234 This function sets the smob marking procedure for the smob type
6235 specified by the tag TC. TC is the tag returned by
6236 `scm_make_smob_type'.
6238 *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM))
6239 This function sets the smob freeing procedure for the smob type
6240 specified by the tag TC. TC is the tag returned by
6241 `scm_make_smob_type'.
6243 *** Function: void scm_set_smob_print (tc, print)
6245 - Function: void scm_set_smob_print (long tc,
6246 scm_sizet (*print) (SCM,
6250 This function sets the smob printing procedure for the smob type
6251 specified by the tag TC. TC is the tag returned by
6252 `scm_make_smob_type'.
6254 *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM))
6255 This function sets the smob equality-testing predicate for the
6256 smob type specified by the tag TC. TC is the tag returned by
6257 `scm_make_smob_type'.
6259 *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data)
6260 Make VALUE contain a smob instance of the type with type code TC and
6261 smob data DATA. VALUE must be previously declared as C type `SCM'.
6263 *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data)
6264 This macro expands to a block of code that creates a smob instance
6265 of the type with type code TC and smob data DATA, and returns that
6266 `SCM' value. It should be the last piece of code in a block.
6268 ** The interfaces for using I/O ports and implementing port types
6269 (ptobs) have changed significantly. The new interface is based on
6270 shared access to buffers and a new set of ptob procedures.
6272 *** scm_newptob has been removed
6276 *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush)
6278 - Function: SCM scm_make_port_type (char *type_name,
6279 int (*fill_buffer) (SCM port),
6280 void (*write_flush) (SCM port));
6282 Similarly to the new smob interface, there is a set of function
6283 setters by which the user can customize the behaviour of his port
6284 type. See ports.h (scm_set_port_XXX).
6286 ** scm_strport_to_string: New function: creates a new string from
6287 a string port's buffer.
6289 ** Plug in interface for random number generators
6290 The variable `scm_the_rng' in random.c contains a value and three
6291 function pointers which together define the current random number
6292 generator being used by the Scheme level interface and the random
6293 number library functions.
6295 The user is free to replace the default generator with the generator
6298 *** Variable: size_t scm_the_rng.rstate_size
6299 The size of the random state type used by the current RNG
6302 *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE)
6303 Given the random STATE, return 32 random bits.
6305 *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N)
6306 Seed random state STATE using string S of length N.
6308 *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE)
6309 Given random state STATE, return a malloced copy.
6312 The default RNG is the MWC (Multiply With Carry) random number
6313 generator described by George Marsaglia at the Department of
6314 Statistics and Supercomputer Computations Research Institute, The
6315 Florida State University (http://stat.fsu.edu/~geo).
6317 It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and
6318 passes all tests in the DIEHARD test suite
6319 (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits
6320 costs one multiply and one add on platforms which either supports long
6321 longs (gcc does this on most systems) or have 64 bit longs. The cost
6322 is four multiply on other systems but this can be optimized by writing
6323 scm_i_uniform32 in assembler.
6325 These functions are provided through the scm_the_rng interface for use
6326 by libguile and the application.
6328 *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE)
6329 Given the random STATE, return 32 random bits.
6330 Don't use this function directly. Instead go through the plugin
6331 interface (see "Plug in interface" above).
6333 *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N)
6334 Initialize STATE using SEED of length N.
6336 *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE)
6337 Return a malloc:ed copy of STATE. This function can easily be re-used
6338 in the interfaces to other RNGs.
6340 ** Random number library functions
6341 These functions use the current RNG through the scm_the_rng interface.
6342 It might be a good idea to use these functions from your C code so
6343 that only one random generator is used by all code in your program.
6345 The default random state is stored in:
6347 *** Variable: SCM scm_var_random_state
6348 Contains the vcell of the Scheme variable "*random-state*" which is
6349 used as default state by all random number functions in the Scheme
6354 double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state)));
6356 *** Function: scm_rstate *scm_c_default_rstate (void)
6357 This is a convenience function which returns the value of
6358 scm_var_random_state. An error message is generated if this value
6359 isn't a random state.
6361 *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH)
6362 Make a new random state from the string SEED of length LENGTH.
6364 It is generally not a good idea to use multiple random states in a
6365 program. While subsequent random numbers generated from one random
6366 state are guaranteed to be reasonably independent, there is no such
6367 guarantee for numbers generated from different random states.
6369 *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE)
6370 Return 32 random bits.
6372 *** Function: double scm_c_uniform01 (scm_rstate *STATE)
6373 Return a sample from the uniform(0,1) distribution.
6375 *** Function: double scm_c_normal01 (scm_rstate *STATE)
6376 Return a sample from the normal(0,1) distribution.
6378 *** Function: double scm_c_exp1 (scm_rstate *STATE)
6379 Return a sample from the exp(1) distribution.
6381 *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M)
6382 Return a sample from the discrete uniform(0,M) distribution.
6384 *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M)
6385 Return a sample from the discrete uniform(0,M) distribution.
6386 M must be a bignum object. The returned value may be an INUM.
6390 Changes in Guile 1.3 (released Monday, October 19, 1998):
6392 * Changes to the distribution
6394 ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH.
6395 To avoid conflicts, programs should name environment variables after
6396 themselves, except when there's a common practice establishing some
6399 For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH,
6400 giving the former precedence, and printing a warning message if the
6401 latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all.
6403 ** The header files related to multi-byte characters have been removed.
6404 They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code
6405 which referred to these explicitly will probably need to be rewritten,
6406 since the support for the variant string types has been removed; see
6409 ** The header files append.h and sequences.h have been removed. These
6410 files implemented non-R4RS operations which would encourage
6411 non-portable programming style and less easy-to-read code.
6413 * Changes to the stand-alone interpreter
6415 ** New procedures have been added to implement a "batch mode":
6417 *** Function: batch-mode?
6419 Returns a boolean indicating whether the interpreter is in batch
6422 *** Function: set-batch-mode?! ARG
6424 If ARG is true, switches the interpreter to batch mode. The `#f'
6425 case has not been implemented.
6427 ** Guile now provides full command-line editing, when run interactively.
6428 To use this feature, you must have the readline library installed.
6429 The Guile build process will notice it, and automatically include
6432 The readline library is available via anonymous FTP from any GNU
6433 mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu".
6435 ** the-last-stack is now a fluid.
6437 * Changes to the procedure for linking libguile with your programs
6439 ** You can now use the `guile-config' utility to build programs that use Guile.
6441 Guile now includes a command-line utility called `guile-config', which
6442 can provide information about how to compile and link programs that
6445 *** `guile-config compile' prints any C compiler flags needed to use Guile.
6446 You should include this command's output on the command line you use
6447 to compile C or C++ code that #includes the Guile header files. It's
6448 usually just a `-I' flag to help the compiler find the Guile headers.
6451 *** `guile-config link' prints any linker flags necessary to link with Guile.
6453 This command writes to its standard output a list of flags which you
6454 must pass to the linker to link your code against the Guile library.
6455 The flags include '-lguile' itself, any other libraries the Guile
6456 library depends upon, and any `-L' flags needed to help the linker
6457 find those libraries.
6459 For example, here is a Makefile rule that builds a program named 'foo'
6460 from the object files ${FOO_OBJECTS}, and links them against Guile:
6463 ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo
6465 Previous Guile releases recommended that you use autoconf to detect
6466 which of a predefined set of libraries were present on your system.
6467 It is more robust to use `guile-config', since it records exactly which
6468 libraries the installed Guile library requires.
6470 This was originally called `build-guile', but was renamed to
6471 `guile-config' before Guile 1.3 was released, to be consistent with
6472 the analogous script for the GTK+ GUI toolkit, which is called
6476 ** Use the GUILE_FLAGS macro in your configure.in file to find Guile.
6478 If you are using the GNU autoconf package to configure your program,
6479 you can use the GUILE_FLAGS autoconf macro to call `guile-config'
6480 (described above) and gather the necessary values for use in your
6483 The GUILE_FLAGS macro expands to configure script code which runs the
6484 `guile-config' script, to find out where Guile's header files and
6485 libraries are installed. It sets two variables, marked for
6486 substitution, as by AC_SUBST.
6488 GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build
6489 code that uses Guile header files. This is almost always just a
6492 GUILE_LDFLAGS --- flags to pass to the linker to link a
6493 program against Guile. This includes `-lguile' for the Guile
6494 library itself, any libraries that Guile itself requires (like
6495 -lqthreads), and so on. It may also include a -L flag to tell the
6496 compiler where to find the libraries.
6498 GUILE_FLAGS is defined in the file guile.m4, in the top-level
6499 directory of the Guile distribution. You can copy it into your
6500 package's aclocal.m4 file, and then use it in your configure.in file.
6502 If you are using the `aclocal' program, distributed with GNU automake,
6503 to maintain your aclocal.m4 file, the Guile installation process
6504 installs guile.m4 where aclocal will find it. All you need to do is
6505 use GUILE_FLAGS in your configure.in file, and then run `aclocal';
6506 this will copy the definition of GUILE_FLAGS into your aclocal.m4
6510 * Changes to Scheme functions and syntax
6512 ** Multi-byte strings have been removed, as have multi-byte and wide
6513 ports. We felt that these were the wrong approach to
6514 internationalization support.
6516 ** New function: readline [PROMPT]
6517 Read a line from the terminal, and allow the user to edit it,
6518 prompting with PROMPT. READLINE provides a large set of Emacs-like
6519 editing commands, lets the user recall previously typed lines, and
6520 works on almost every kind of terminal, including dumb terminals.
6522 READLINE assumes that the cursor is at the beginning of the line when
6523 it is invoked. Thus, you can't print a prompt yourself, and then call
6524 READLINE; you need to package up your prompt as a string, pass it to
6525 the function, and let READLINE print the prompt itself. This is
6526 because READLINE needs to know the prompt's screen width.
6528 For Guile to provide this function, you must have the readline
6529 library, version 2.1 or later, installed on your system. Readline is
6530 available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from
6531 any GNU mirror site.
6533 See also ADD-HISTORY function.
6535 ** New function: add-history STRING
6536 Add STRING as the most recent line in the history used by the READLINE
6537 command. READLINE does not add lines to the history itself; you must
6538 call ADD-HISTORY to make previous input available to the user.
6540 ** The behavior of the read-line function has changed.
6542 This function now uses standard C library functions to read the line,
6543 for speed. This means that it doesn not respect the value of
6544 scm-line-incrementors; it assumes that lines are delimited with
6547 (Note that this is read-line, the function that reads a line of text
6548 from a port, not readline, the function that reads a line from a
6549 terminal, providing full editing capabilities.)
6551 ** New module (ice-9 getopt-gnu-style): Parse command-line arguments.
6553 This module provides some simple argument parsing. It exports one
6556 Function: getopt-gnu-style ARG-LS
6557 Parse a list of program arguments into an alist of option
6560 Each item in the list of program arguments is examined to see if
6561 it meets the syntax of a GNU long-named option. An argument like
6562 `--MUMBLE' produces an element of the form (MUMBLE . #t) in the
6563 returned alist, where MUMBLE is a keyword object with the same
6564 name as the argument. An argument like `--MUMBLE=FROB' produces
6565 an element of the form (MUMBLE . FROB), where FROB is a string.
6567 As a special case, the returned alist also contains a pair whose
6568 car is the symbol `rest'. The cdr of this pair is a list
6569 containing all the items in the argument list that are not options
6570 of the form mentioned above.
6572 The argument `--' is treated specially: all items in the argument
6573 list appearing after such an argument are not examined, and are
6574 returned in the special `rest' list.
6576 This function does not parse normal single-character switches.
6577 You will need to parse them out of the `rest' list yourself.
6579 ** The read syntax for byte vectors and short vectors has changed.
6581 Instead of #bytes(...), write #y(...).
6583 Instead of #short(...), write #h(...).
6585 This may seem nutty, but, like the other uniform vectors, byte vectors
6586 and short vectors want to have the same print and read syntax (and,
6587 more basic, want to have read syntax!). Changing the read syntax to
6588 use multiple characters after the hash sign breaks with the
6589 conventions used in R5RS and the conventions used for the other
6590 uniform vectors. It also introduces complexity in the current reader,
6591 both on the C and Scheme levels. (The Right solution is probably to
6592 change the syntax and prototypes for uniform vectors entirely.)
6595 ** The new module (ice-9 session) provides useful interactive functions.
6597 *** New procedure: (apropos REGEXP OPTION ...)
6599 Display a list of top-level variables whose names match REGEXP, and
6600 the modules they are imported from. Each OPTION should be one of the
6603 value --- Show the value of each matching variable.
6604 shadow --- Show bindings shadowed by subsequently imported modules.
6605 full --- Same as both `shadow' and `value'.
6609 guile> (apropos "trace" 'full)
6610 debug: trace #<procedure trace args>
6611 debug: untrace #<procedure untrace args>
6612 the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>>
6613 the-scm-module: before-backtrace-hook ()
6614 the-scm-module: backtrace #<primitive-procedure backtrace>
6615 the-scm-module: after-backtrace-hook ()
6616 the-scm-module: has-shown-backtrace-hint? #f
6619 ** There are new functions and syntax for working with macros.
6621 Guile implements macros as a special object type. Any variable whose
6622 top-level binding is a macro object acts as a macro. The macro object
6623 specifies how the expression should be transformed before evaluation.
6625 *** Macro objects now print in a reasonable way, resembling procedures.
6627 *** New function: (macro? OBJ)
6628 True iff OBJ is a macro object.
6630 *** New function: (primitive-macro? OBJ)
6631 Like (macro? OBJ), but true only if OBJ is one of the Guile primitive
6632 macro transformers, implemented in eval.c rather than Scheme code.
6634 Why do we have this function?
6635 - For symmetry with procedure? and primitive-procedure?,
6636 - to allow custom print procedures to tell whether a macro is
6637 primitive, and display it differently, and
6638 - to allow compilers and user-written evaluators to distinguish
6639 builtin special forms from user-defined ones, which could be
6642 *** New function: (macro-type OBJ)
6643 Return a value indicating what kind of macro OBJ is. Possible return
6646 The symbol `syntax' --- a macro created by procedure->syntax.
6647 The symbol `macro' --- a macro created by procedure->macro.
6648 The symbol `macro!' --- a macro created by procedure->memoizing-macro.
6649 The boolean #f --- if OBJ is not a macro object.
6651 *** New function: (macro-name MACRO)
6652 Return the name of the macro object MACRO's procedure, as returned by
6655 *** New function: (macro-transformer MACRO)
6656 Return the transformer procedure for MACRO.
6658 *** New syntax: (use-syntax MODULE ... TRANSFORMER)
6660 Specify a new macro expander to use in the current module. Each
6661 MODULE is a module name, with the same meaning as in the `use-modules'
6662 form; each named module's exported bindings are added to the current
6663 top-level environment. TRANSFORMER is an expression evaluated in the
6664 resulting environment which must yield a procedure to use as the
6665 module's eval transformer: every expression evaluated in this module
6666 is passed to this function, and the result passed to the Guile
6669 *** macro-eval! is removed. Use local-eval instead.
6671 ** Some magic has been added to the printer to better handle user
6672 written printing routines (like record printers, closure printers).
6674 The problem is that these user written routines must have access to
6675 the current `print-state' to be able to handle fancy things like
6676 detection of circular references. These print-states have to be
6677 passed to the builtin printing routines (display, write, etc) to
6678 properly continue the print chain.
6680 We didn't want to change all existing print code so that it
6681 explicitly passes thru a print state in addition to a port. Instead,
6682 we extented the possible values that the builtin printing routines
6683 accept as a `port'. In addition to a normal port, they now also take
6684 a pair of a normal port and a print-state. Printing will go to the
6685 port and the print-state will be used to control the detection of
6686 circular references, etc. If the builtin function does not care for a
6687 print-state, it is simply ignored.
6689 User written callbacks are now called with such a pair as their
6690 `port', but because every function now accepts this pair as a PORT
6691 argument, you don't have to worry about that. In fact, it is probably
6692 safest to not check for these pairs.
6694 However, it is sometimes necessary to continue a print chain on a
6695 different port, for example to get a intermediate string
6696 representation of the printed value, mangle that string somehow, and
6697 then to finally print the mangled string. Use the new function
6699 inherit-print-state OLD-PORT NEW-PORT
6701 for this. It constructs a new `port' that prints to NEW-PORT but
6702 inherits the print-state of OLD-PORT.
6704 ** struct-vtable-offset renamed to vtable-offset-user
6706 ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer
6708 ** There is now a third optional argument to make-vtable-vtable
6709 (and fourth to make-struct) when constructing new types (vtables).
6710 This argument initializes field vtable-index-printer of the vtable.
6712 ** The detection of circular references has been extended to structs.
6713 That is, a structure that -- in the process of being printed -- prints
6714 itself does not lead to infinite recursion.
6716 ** There is now some basic support for fluids. Please read
6717 "libguile/fluid.h" to find out more. It is accessible from Scheme with
6718 the following functions and macros:
6720 Function: make-fluid
6722 Create a new fluid object. Fluids are not special variables or
6723 some other extension to the semantics of Scheme, but rather
6724 ordinary Scheme objects. You can store them into variables (that
6725 are still lexically scoped, of course) or into any other place you
6726 like. Every fluid has a initial value of `#f'.
6728 Function: fluid? OBJ
6730 Test whether OBJ is a fluid.
6732 Function: fluid-ref FLUID
6733 Function: fluid-set! FLUID VAL
6735 Access/modify the fluid FLUID. Modifications are only visible
6736 within the current dynamic root (that includes threads).
6738 Function: with-fluids* FLUIDS VALUES THUNK
6740 FLUIDS is a list of fluids and VALUES a corresponding list of
6741 values for these fluids. Before THUNK gets called the values are
6742 installed in the fluids and the old values of the fluids are
6743 saved in the VALUES list. When the flow of control leaves THUNK
6744 or reenters it, the values get swapped again. You might think of
6745 this as a `safe-fluid-excursion'. Note that the VALUES list is
6746 modified by `with-fluids*'.
6748 Macro: with-fluids ((FLUID VALUE) ...) FORM ...
6750 The same as `with-fluids*' but with a different syntax. It looks
6751 just like `let', but both FLUID and VALUE are evaluated. Remember,
6752 fluids are not special variables but ordinary objects. FLUID
6753 should evaluate to a fluid.
6755 ** Changes to system call interfaces:
6757 *** close-port, close-input-port and close-output-port now return a
6758 boolean instead of an `unspecified' object. #t means that the port
6759 was successfully closed, while #f means it was already closed. It is
6760 also now possible for these procedures to raise an exception if an
6761 error occurs (some errors from write can be delayed until close.)
6763 *** the first argument to chmod, fcntl, ftell and fseek can now be a
6766 *** the third argument to fcntl is now optional.
6768 *** the first argument to chown can now be a file descriptor or a port.
6770 *** the argument to stat can now be a port.
6772 *** The following new procedures have been added (most use scsh
6775 *** procedure: close PORT/FD
6776 Similar to close-port (*note close-port: Closing Ports.), but also
6777 works on file descriptors. A side effect of closing a file
6778 descriptor is that any ports using that file descriptor are moved
6779 to a different file descriptor and have their revealed counts set
6782 *** procedure: port->fdes PORT
6783 Returns the integer file descriptor underlying PORT. As a side
6784 effect the revealed count of PORT is incremented.
6786 *** procedure: fdes->ports FDES
6787 Returns a list of existing ports which have FDES as an underlying
6788 file descriptor, without changing their revealed counts.
6790 *** procedure: fdes->inport FDES
6791 Returns an existing input port which has FDES as its underlying
6792 file descriptor, if one exists, and increments its revealed count.
6793 Otherwise, returns a new input port with a revealed count of 1.
6795 *** procedure: fdes->outport FDES
6796 Returns an existing output port which has FDES as its underlying
6797 file descriptor, if one exists, and increments its revealed count.
6798 Otherwise, returns a new output port with a revealed count of 1.
6800 The next group of procedures perform a `dup2' system call, if NEWFD
6801 (an integer) is supplied, otherwise a `dup'. The file descriptor to be
6802 duplicated can be supplied as an integer or contained in a port. The
6803 type of value returned varies depending on which procedure is used.
6805 All procedures also have the side effect when performing `dup2' that
6806 any ports using NEWFD are moved to a different file descriptor and have
6807 their revealed counts set to zero.
6809 *** procedure: dup->fdes PORT/FD [NEWFD]
6810 Returns an integer file descriptor.
6812 *** procedure: dup->inport PORT/FD [NEWFD]
6813 Returns a new input port using the new file descriptor.
6815 *** procedure: dup->outport PORT/FD [NEWFD]
6816 Returns a new output port using the new file descriptor.
6818 *** procedure: dup PORT/FD [NEWFD]
6819 Returns a new port if PORT/FD is a port, with the same mode as the
6820 supplied port, otherwise returns an integer file descriptor.
6822 *** procedure: dup->port PORT/FD MODE [NEWFD]
6823 Returns a new port using the new file descriptor. MODE supplies a
6824 mode string for the port (*note open-file: File Ports.).
6826 *** procedure: setenv NAME VALUE
6827 Modifies the environment of the current process, which is also the
6828 default environment inherited by child processes.
6830 If VALUE is `#f', then NAME is removed from the environment.
6831 Otherwise, the string NAME=VALUE is added to the environment,
6832 replacing any existing string with name matching NAME.
6834 The return value is unspecified.
6836 *** procedure: truncate-file OBJ SIZE
6837 Truncates the file referred to by OBJ to at most SIZE bytes. OBJ
6838 can be a string containing a file name or an integer file
6839 descriptor or port open for output on the file. The underlying
6840 system calls are `truncate' and `ftruncate'.
6842 The return value is unspecified.
6844 *** procedure: setvbuf PORT MODE [SIZE]
6845 Set the buffering mode for PORT. MODE can be:
6853 block buffered, using a newly allocated buffer of SIZE bytes.
6854 However if SIZE is zero or unspecified, the port will be made
6857 This procedure should not be used after I/O has been performed with
6860 Ports are usually block buffered by default, with a default buffer
6861 size. Procedures e.g., *Note open-file: File Ports, which accept a
6862 mode string allow `0' to be added to request an unbuffered port.
6864 *** procedure: fsync PORT/FD
6865 Copies any unwritten data for the specified output file descriptor
6866 to disk. If PORT/FD is a port, its buffer is flushed before the
6867 underlying file descriptor is fsync'd. The return value is
6870 *** procedure: open-fdes PATH FLAGS [MODES]
6871 Similar to `open' but returns a file descriptor instead of a port.
6873 *** procedure: execle PATH ENV [ARG] ...
6874 Similar to `execl', but the environment of the new process is
6875 specified by ENV, which must be a list of strings as returned by
6876 the `environ' procedure.
6878 This procedure is currently implemented using the `execve' system
6879 call, but we call it `execle' because of its Scheme calling
6882 *** procedure: strerror ERRNO
6883 Returns the Unix error message corresponding to ERRNO, an integer.
6885 *** procedure: primitive-exit [STATUS]
6886 Terminate the current process without unwinding the Scheme stack.
6887 This is would typically be useful after a fork. The exit status
6888 is STATUS if supplied, otherwise zero.
6890 *** procedure: times
6891 Returns an object with information about real and processor time.
6892 The following procedures accept such an object as an argument and
6893 return a selected component:
6896 The current real time, expressed as time units relative to an
6900 The CPU time units used by the calling process.
6903 The CPU time units used by the system on behalf of the
6907 The CPU time units used by terminated child processes of the
6908 calling process, whose status has been collected (e.g., using
6912 Similarly, the CPU times units used by the system on behalf of
6913 terminated child processes.
6915 ** Removed: list-length
6916 ** Removed: list-append, list-append!
6917 ** Removed: list-reverse, list-reverse!
6919 ** array-map renamed to array-map!
6921 ** serial-array-map renamed to serial-array-map!
6923 ** catch doesn't take #f as first argument any longer
6925 Previously, it was possible to pass #f instead of a key to `catch'.
6926 That would cause `catch' to pass a jump buffer object to the procedure
6927 passed as second argument. The procedure could then use this jump
6928 buffer objekt as an argument to throw.
6930 This mechanism has been removed since its utility doesn't motivate the
6931 extra complexity it introduces.
6933 ** The `#/' notation for lists now provokes a warning message from Guile.
6934 This syntax will be removed from Guile in the near future.
6936 To disable the warning message, set the GUILE_HUSH environment
6937 variable to any non-empty value.
6939 ** The newline character now prints as `#\newline', following the
6940 normal Scheme notation, not `#\nl'.
6942 * Changes to the gh_ interface
6944 ** The gh_enter function now takes care of loading the Guile startup files.
6945 gh_enter works by calling scm_boot_guile; see the remarks below.
6947 ** Function: void gh_write (SCM x)
6949 Write the printed representation of the scheme object x to the current
6950 output port. Corresponds to the scheme level `write'.
6952 ** gh_list_length renamed to gh_length.
6954 ** vector handling routines
6956 Several major changes. In particular, gh_vector() now resembles
6957 (vector ...) (with a caveat -- see manual), and gh_make_vector() now
6958 exists and behaves like (make-vector ...). gh_vset() and gh_vref()
6959 have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing
6960 vector-related gh_ functions have been implemented.
6962 ** pair and list routines
6964 Implemented several of the R4RS pair and list functions that were
6967 ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect
6969 New function. Converts double arrays back and forth between Scheme
6972 * Changes to the scm_ interface
6974 ** The function scm_boot_guile now takes care of loading the startup files.
6976 Guile's primary initialization function, scm_boot_guile, now takes
6977 care of loading `boot-9.scm', in the `ice-9' module, to initialize
6978 Guile, define the module system, and put together some standard
6979 bindings. It also loads `init.scm', which is intended to hold
6980 site-specific initialization code.
6982 Since Guile cannot operate properly until boot-9.scm is loaded, there
6983 is no reason to separate loading boot-9.scm from Guile's other
6984 initialization processes.
6986 This job used to be done by scm_compile_shell_switches, which didn't
6987 make much sense; in particular, it meant that people using Guile for
6988 non-shell-like applications had to jump through hoops to get Guile
6989 initialized properly.
6991 ** The function scm_compile_shell_switches no longer loads the startup files.
6992 Now, Guile always loads the startup files, whenever it is initialized;
6993 see the notes above for scm_boot_guile and scm_load_startup_files.
6995 ** Function: scm_load_startup_files
6996 This new function takes care of loading Guile's initialization file
6997 (`boot-9.scm'), and the site initialization file, `init.scm'. Since
6998 this is always called by the Guile initialization process, it's
6999 probably not too useful to call this yourself, but it's there anyway.
7001 ** The semantics of smob marking have changed slightly.
7003 The smob marking function (the `mark' member of the scm_smobfuns
7004 structure) is no longer responsible for setting the mark bit on the
7005 smob. The generic smob handling code in the garbage collector will
7006 set this bit. The mark function need only ensure that any other
7007 objects the smob refers to get marked.
7009 Note that this change means that the smob's GC8MARK bit is typically
7010 already set upon entry to the mark function. Thus, marking functions
7011 which look like this:
7014 if (SCM_GC8MARKP (ptr))
7016 SCM_SETGC8MARK (ptr);
7017 ... mark objects to which the smob refers ...
7020 are now incorrect, since they will return early, and fail to mark any
7021 other objects the smob refers to. Some code in the Guile library used
7024 ** The semantics of the I/O port functions in scm_ptobfuns have changed.
7026 If you have implemented your own I/O port type, by writing the
7027 functions required by the scm_ptobfuns and then calling scm_newptob,
7028 you will need to change your functions slightly.
7030 The functions in a scm_ptobfuns structure now expect the port itself
7031 as their argument; they used to expect the `stream' member of the
7032 port's scm_port_table structure. This allows functions in an
7033 scm_ptobfuns structure to easily access the port's cell (and any flags
7034 it its CAR), and the port's scm_port_table structure.
7036 Guile now passes the I/O port itself as the `port' argument in the
7037 following scm_ptobfuns functions:
7039 int (*free) (SCM port);
7040 int (*fputc) (int, SCM port);
7041 int (*fputs) (char *, SCM port);
7042 scm_sizet (*fwrite) SCM_P ((char *ptr,
7046 int (*fflush) (SCM port);
7047 int (*fgetc) (SCM port);
7048 int (*fclose) (SCM port);
7050 The interfaces to the `mark', `print', `equalp', and `fgets' methods
7053 If you have existing code which defines its own port types, it is easy
7054 to convert your code to the new interface; simply apply SCM_STREAM to
7055 the port argument to yield the value you code used to expect.
7057 Note that since both the port and the stream have the same type in the
7058 C code --- they are both SCM values --- the C compiler will not remind
7059 you if you forget to update your scm_ptobfuns functions.
7062 ** Function: int scm_internal_select (int fds,
7066 struct timeval *timeout);
7068 This is a replacement for the `select' function provided by the OS.
7069 It enables I/O blocking and sleeping to happen for one cooperative
7070 thread without blocking other threads. It also avoids busy-loops in
7071 these situations. It is intended that all I/O blocking and sleeping
7072 will finally go through this function. Currently, this function is
7073 only available on systems providing `gettimeofday' and `select'.
7075 ** Function: SCM scm_internal_stack_catch (SCM tag,
7076 scm_catch_body_t body,
7078 scm_catch_handler_t handler,
7081 A new sibling to the other two C level `catch' functions
7082 scm_internal_catch and scm_internal_lazy_catch. Use it if you want
7083 the stack to be saved automatically into the variable `the-last-stack'
7084 (scm_the_last_stack_var) on error. This is necessary if you want to
7085 use advanced error reporting, such as calling scm_display_error and
7086 scm_display_backtrace. (They both take a stack object as argument.)
7088 ** Function: SCM scm_spawn_thread (scm_catch_body_t body,
7090 scm_catch_handler_t handler,
7093 Spawns a new thread. It does a job similar to
7094 scm_call_with_new_thread but takes arguments more suitable when
7095 spawning threads from application C code.
7097 ** The hook scm_error_callback has been removed. It was originally
7098 intended as a way for the user to install his own error handler. But
7099 that method works badly since it intervenes between throw and catch,
7100 thereby changing the semantics of expressions like (catch #t ...).
7101 The correct way to do it is to use one of the C level catch functions
7102 in throw.c: scm_internal_catch/lazy_catch/stack_catch.
7104 ** Removed functions:
7106 scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x,
7107 scm_list_reverse, scm_list_reverse_x
7109 ** New macros: SCM_LISTn where n is one of the integers 0-9.
7111 These can be used for pretty list creation from C. The idea is taken
7112 from Erick Gallesio's STk.
7114 ** scm_array_map renamed to scm_array_map_x
7116 ** mbstrings are now removed
7118 This means that the type codes scm_tc7_mb_string and
7119 scm_tc7_mb_substring has been removed.
7121 ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed.
7123 Since we no longer support multi-byte strings, these I/O functions
7124 have been simplified, and renamed. Here are their old names, and
7125 their new names and arguments:
7127 scm_gen_putc -> void scm_putc (int c, SCM port);
7128 scm_gen_puts -> void scm_puts (char *s, SCM port);
7129 scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port);
7130 scm_gen_getc -> void scm_getc (SCM port);
7133 ** The macros SCM_TYP7D and SCM_TYP7SD has been removed.
7135 ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D
7137 SCM_TYP7S now masks away the bit which distinguishes substrings from
7140 ** scm_catch_body_t: Backward incompatible change!
7142 Body functions to scm_internal_catch and friends do not any longer
7143 take a second argument. This is because it is no longer possible to
7144 pass a #f arg to catch.
7146 ** Calls to scm_protect_object and scm_unprotect now nest properly.
7148 The function scm_protect_object protects its argument from being freed
7149 by the garbage collector. scm_unprotect_object removes that
7152 These functions now nest properly. That is, for every object O, there
7153 is a counter which scm_protect_object(O) increments and
7154 scm_unprotect_object(O) decrements, if the counter is greater than
7155 zero. Every object's counter is zero when it is first created. If an
7156 object's counter is greater than zero, the garbage collector will not
7157 reclaim its storage.
7159 This allows you to use scm_protect_object in your code without
7160 worrying that some other function you call will call
7161 scm_unprotect_object, and allow it to be freed. Assuming that the
7162 functions you call are well-behaved, and unprotect only those objects
7163 they protect, you can follow the same rule and have confidence that
7164 objects will be freed only at appropriate times.
7167 Changes in Guile 1.2 (released Tuesday, June 24 1997):
7169 * Changes to the distribution
7171 ** Nightly snapshots are now available from ftp.red-bean.com.
7172 The old server, ftp.cyclic.com, has been relinquished to its rightful
7175 Nightly snapshots of the Guile development sources are now available via
7176 anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz.
7178 Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz
7179 For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz
7181 ** To run Guile without installing it, the procedure has changed a bit.
7183 If you used a separate build directory to compile Guile, you'll need
7184 to include the build directory in SCHEME_LOAD_PATH, as well as the
7185 source directory. See the `INSTALL' file for examples.
7187 * Changes to the procedure for linking libguile with your programs
7189 ** The standard Guile load path for Scheme code now includes
7190 $(datadir)/guile (usually /usr/local/share/guile). This means that
7191 you can install your own Scheme files there, and Guile will find them.
7192 (Previous versions of Guile only checked a directory whose name
7193 contained the Guile version number, so you had to re-install or move
7194 your Scheme sources each time you installed a fresh version of Guile.)
7196 The load path also includes $(datadir)/guile/site; we recommend
7197 putting individual Scheme files there. If you want to install a
7198 package with multiple source files, create a directory for them under
7201 ** Guile 1.2 will now use the Rx regular expression library, if it is
7202 installed on your system. When you are linking libguile into your own
7203 programs, this means you will have to link against -lguile, -lqt (if
7204 you configured Guile with thread support), and -lrx.
7206 If you are using autoconf to generate configuration scripts for your
7207 application, the following lines should suffice to add the appropriate
7208 libraries to your link command:
7210 ### Find Rx, quickthreads and libguile.
7211 AC_CHECK_LIB(rx, main)
7212 AC_CHECK_LIB(qt, main)
7213 AC_CHECK_LIB(guile, scm_shell)
7215 The Guile 1.2 distribution does not contain sources for the Rx
7216 library, as Guile 1.0 did. If you want to use Rx, you'll need to
7217 retrieve it from a GNU FTP site and install it separately.
7219 * Changes to Scheme functions and syntax
7221 ** The dynamic linking features of Guile are now enabled by default.
7222 You can disable them by giving the `--disable-dynamic-linking' option
7225 (dynamic-link FILENAME)
7227 Find the object file denoted by FILENAME (a string) and link it
7228 into the running Guile application. When everything works out,
7229 return a Scheme object suitable for representing the linked object
7230 file. Otherwise an error is thrown. How object files are
7231 searched is system dependent.
7233 (dynamic-object? VAL)
7235 Determine whether VAL represents a dynamically linked object file.
7237 (dynamic-unlink DYNOBJ)
7239 Unlink the indicated object file from the application. DYNOBJ
7240 should be one of the values returned by `dynamic-link'.
7242 (dynamic-func FUNCTION DYNOBJ)
7244 Search the C function indicated by FUNCTION (a string or symbol)
7245 in DYNOBJ and return some Scheme object that can later be used
7246 with `dynamic-call' to actually call this function. Right now,
7247 these Scheme objects are formed by casting the address of the
7248 function to `long' and converting this number to its Scheme
7251 (dynamic-call FUNCTION DYNOBJ)
7253 Call the C function indicated by FUNCTION and DYNOBJ. The
7254 function is passed no arguments and its return value is ignored.
7255 When FUNCTION is something returned by `dynamic-func', call that
7256 function and ignore DYNOBJ. When FUNCTION is a string (or symbol,
7257 etc.), look it up in DYNOBJ; this is equivalent to
7259 (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f)
7261 Interrupts are deferred while the C function is executing (with
7262 SCM_DEFER_INTS/SCM_ALLOW_INTS).
7264 (dynamic-args-call FUNCTION DYNOBJ ARGS)
7266 Call the C function indicated by FUNCTION and DYNOBJ, but pass it
7267 some arguments and return its return value. The C function is
7268 expected to take two arguments and return an `int', just like
7271 int c_func (int argc, char **argv);
7273 ARGS must be a list of strings and is converted into an array of
7274 `char *'. The array is passed in ARGV and its size in ARGC. The
7275 return value is converted to a Scheme number and returned from the
7276 call to `dynamic-args-call'.
7278 When dynamic linking is disabled or not supported on your system,
7279 the above functions throw errors, but they are still available.
7281 Here is a small example that works on GNU/Linux:
7283 (define libc-obj (dynamic-link "libc.so"))
7284 (dynamic-args-call 'rand libc-obj '())
7286 See the file `libguile/DYNAMIC-LINKING' for additional comments.
7288 ** The #/ syntax for module names is depreciated, and will be removed
7289 in a future version of Guile. Instead of
7297 The latter syntax is more consistent with existing Lisp practice.
7299 ** Guile now does fancier printing of structures. Structures are the
7300 underlying implementation for records, which in turn are used to
7301 implement modules, so all of these object now print differently and in
7302 a more informative way.
7304 The Scheme printer will examine the builtin variable *struct-printer*
7305 whenever it needs to print a structure object. When this variable is
7306 not `#f' it is deemed to be a procedure and will be applied to the
7307 structure object and the output port. When *struct-printer* is `#f'
7308 or the procedure return `#f' the structure object will be printed in
7309 the boring #<struct 80458270> form.
7311 This hook is used by some routines in ice-9/boot-9.scm to implement
7312 type specific printing routines. Please read the comments there about
7315 One of the more specific uses of structs are records. The printing
7316 procedure that could be passed to MAKE-RECORD-TYPE is now actually
7317 called. It should behave like a *struct-printer* procedure (described
7320 ** Guile now supports a new R4RS-compliant syntax for keywords. A
7321 token of the form #:NAME, where NAME has the same syntax as a Scheme
7322 symbol, is the external representation of the keyword named NAME.
7323 Keyword objects print using this syntax as well, so values containing
7324 keyword objects can be read back into Guile. When used in an
7325 expression, keywords are self-quoting objects.
7327 Guile suports this read syntax, and uses this print syntax, regardless
7328 of the current setting of the `keyword' read option. The `keyword'
7329 read option only controls whether Guile recognizes the `:NAME' syntax,
7330 which is incompatible with R4RS. (R4RS says such token represent
7333 ** Guile has regular expression support again. Guile 1.0 included
7334 functions for matching regular expressions, based on the Rx library.
7335 In Guile 1.1, the Guile/Rx interface was removed to simplify the
7336 distribution, and thus Guile had no regular expression support. Guile
7337 1.2 again supports the most commonly used functions, and supports all
7338 of SCSH's regular expression functions.
7340 If your system does not include a POSIX regular expression library,
7341 and you have not linked Guile with a third-party regexp library such as
7342 Rx, these functions will not be available. You can tell whether your
7343 Guile installation includes regular expression support by checking
7344 whether the `*features*' list includes the `regex' symbol.
7346 *** regexp functions
7348 By default, Guile supports POSIX extended regular expressions. That
7349 means that the characters `(', `)', `+' and `?' are special, and must
7350 be escaped if you wish to match the literal characters.
7352 This regular expression interface was modeled after that implemented
7353 by SCSH, the Scheme Shell. It is intended to be upwardly compatible
7354 with SCSH regular expressions.
7356 **** Function: string-match PATTERN STR [START]
7357 Compile the string PATTERN into a regular expression and compare
7358 it with STR. The optional numeric argument START specifies the
7359 position of STR at which to begin matching.
7361 `string-match' returns a "match structure" which describes what,
7362 if anything, was matched by the regular expression. *Note Match
7363 Structures::. If STR does not match PATTERN at all,
7364 `string-match' returns `#f'.
7366 Each time `string-match' is called, it must compile its PATTERN
7367 argument into a regular expression structure. This operation is
7368 expensive, which makes `string-match' inefficient if the same regular
7369 expression is used several times (for example, in a loop). For better
7370 performance, you can compile a regular expression in advance and then
7371 match strings against the compiled regexp.
7373 **** Function: make-regexp STR [FLAGS]
7374 Compile the regular expression described by STR, and return the
7375 compiled regexp structure. If STR does not describe a legal
7376 regular expression, `make-regexp' throws a
7377 `regular-expression-syntax' error.
7379 FLAGS may be the bitwise-or of one or more of the following:
7381 **** Constant: regexp/extended
7382 Use POSIX Extended Regular Expression syntax when interpreting
7383 STR. If not set, POSIX Basic Regular Expression syntax is used.
7384 If the FLAGS argument is omitted, we assume regexp/extended.
7386 **** Constant: regexp/icase
7387 Do not differentiate case. Subsequent searches using the
7388 returned regular expression will be case insensitive.
7390 **** Constant: regexp/newline
7391 Match-any-character operators don't match a newline.
7393 A non-matching list ([^...]) not containing a newline matches a
7396 Match-beginning-of-line operator (^) matches the empty string
7397 immediately after a newline, regardless of whether the FLAGS
7398 passed to regexp-exec contain regexp/notbol.
7400 Match-end-of-line operator ($) matches the empty string
7401 immediately before a newline, regardless of whether the FLAGS
7402 passed to regexp-exec contain regexp/noteol.
7404 **** Function: regexp-exec REGEXP STR [START [FLAGS]]
7405 Match the compiled regular expression REGEXP against `str'. If
7406 the optional integer START argument is provided, begin matching
7407 from that position in the string. Return a match structure
7408 describing the results of the match, or `#f' if no match could be
7411 FLAGS may be the bitwise-or of one or more of the following:
7413 **** Constant: regexp/notbol
7414 The match-beginning-of-line operator always fails to match (but
7415 see the compilation flag regexp/newline above) This flag may be
7416 used when different portions of a string are passed to
7417 regexp-exec and the beginning of the string should not be
7418 interpreted as the beginning of the line.
7420 **** Constant: regexp/noteol
7421 The match-end-of-line operator always fails to match (but see the
7422 compilation flag regexp/newline above)
7424 **** Function: regexp? OBJ
7425 Return `#t' if OBJ is a compiled regular expression, or `#f'
7428 Regular expressions are commonly used to find patterns in one string
7429 and replace them with the contents of another string.
7431 **** Function: regexp-substitute PORT MATCH [ITEM...]
7432 Write to the output port PORT selected contents of the match
7433 structure MATCH. Each ITEM specifies what should be written, and
7434 may be one of the following arguments:
7436 * A string. String arguments are written out verbatim.
7438 * An integer. The submatch with that number is written.
7440 * The symbol `pre'. The portion of the matched string preceding
7441 the regexp match is written.
7443 * The symbol `post'. The portion of the matched string
7444 following the regexp match is written.
7446 PORT may be `#f', in which case nothing is written; instead,
7447 `regexp-substitute' constructs a string from the specified ITEMs
7450 **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...]
7451 Similar to `regexp-substitute', but can be used to perform global
7452 substitutions on STR. Instead of taking a match structure as an
7453 argument, `regexp-substitute/global' takes two string arguments: a
7454 REGEXP string describing a regular expression, and a TARGET string
7455 which should be matched against this regular expression.
7457 Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following
7460 * A function may be supplied. When this function is called, it
7461 will be passed one argument: a match structure for a given
7462 regular expression match. It should return a string to be
7463 written out to PORT.
7465 * The `post' symbol causes `regexp-substitute/global' to recurse
7466 on the unmatched portion of STR. This *must* be supplied in
7467 order to perform global search-and-replace on STR; if it is
7468 not present among the ITEMs, then `regexp-substitute/global'
7469 will return after processing a single match.
7471 *** Match Structures
7473 A "match structure" is the object returned by `string-match' and
7474 `regexp-exec'. It describes which portion of a string, if any, matched
7475 the given regular expression. Match structures include: a reference to
7476 the string that was checked for matches; the starting and ending
7477 positions of the regexp match; and, if the regexp included any
7478 parenthesized subexpressions, the starting and ending positions of each
7481 In each of the regexp match functions described below, the `match'
7482 argument must be a match structure returned by a previous call to
7483 `string-match' or `regexp-exec'. Most of these functions return some
7484 information about the original target string that was matched against a
7485 regular expression; we will call that string TARGET for easy reference.
7487 **** Function: regexp-match? OBJ
7488 Return `#t' if OBJ is a match structure returned by a previous
7489 call to `regexp-exec', or `#f' otherwise.
7491 **** Function: match:substring MATCH [N]
7492 Return the portion of TARGET matched by subexpression number N.
7493 Submatch 0 (the default) represents the entire regexp match. If
7494 the regular expression as a whole matched, but the subexpression
7495 number N did not match, return `#f'.
7497 **** Function: match:start MATCH [N]
7498 Return the starting position of submatch number N.
7500 **** Function: match:end MATCH [N]
7501 Return the ending position of submatch number N.
7503 **** Function: match:prefix MATCH
7504 Return the unmatched portion of TARGET preceding the regexp match.
7506 **** Function: match:suffix MATCH
7507 Return the unmatched portion of TARGET following the regexp match.
7509 **** Function: match:count MATCH
7510 Return the number of parenthesized subexpressions from MATCH.
7511 Note that the entire regular expression match itself counts as a
7512 subexpression, and failed submatches are included in the count.
7514 **** Function: match:string MATCH
7515 Return the original TARGET string.
7517 *** Backslash Escapes
7519 Sometimes you will want a regexp to match characters like `*' or `$'
7520 exactly. For example, to check whether a particular string represents
7521 a menu entry from an Info node, it would be useful to match it against
7522 a regexp like `^* [^:]*::'. However, this won't work; because the
7523 asterisk is a metacharacter, it won't match the `*' at the beginning of
7524 the string. In this case, we want to make the first asterisk un-magic.
7526 You can do this by preceding the metacharacter with a backslash
7527 character `\'. (This is also called "quoting" the metacharacter, and
7528 is known as a "backslash escape".) When Guile sees a backslash in a
7529 regular expression, it considers the following glyph to be an ordinary
7530 character, no matter what special meaning it would ordinarily have.
7531 Therefore, we can make the above example work by changing the regexp to
7532 `^\* [^:]*::'. The `\*' sequence tells the regular expression engine
7533 to match only a single asterisk in the target string.
7535 Since the backslash is itself a metacharacter, you may force a
7536 regexp to match a backslash in the target string by preceding the
7537 backslash with itself. For example, to find variable references in a
7538 TeX program, you might want to find occurrences of the string `\let\'
7539 followed by any number of alphabetic characters. The regular expression
7540 `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp
7541 each match a single backslash in the target string.
7543 **** Function: regexp-quote STR
7544 Quote each special character found in STR with a backslash, and
7545 return the resulting string.
7547 *Very important:* Using backslash escapes in Guile source code (as
7548 in Emacs Lisp or C) can be tricky, because the backslash character has
7549 special meaning for the Guile reader. For example, if Guile encounters
7550 the character sequence `\n' in the middle of a string while processing
7551 Scheme code, it replaces those characters with a newline character.
7552 Similarly, the character sequence `\t' is replaced by a horizontal tab.
7553 Several of these "escape sequences" are processed by the Guile reader
7554 before your code is executed. Unrecognized escape sequences are
7555 ignored: if the characters `\*' appear in a string, they will be
7556 translated to the single character `*'.
7558 This translation is obviously undesirable for regular expressions,
7559 since we want to be able to include backslashes in a string in order to
7560 escape regexp metacharacters. Therefore, to make sure that a backslash
7561 is preserved in a string in your Guile program, you must use *two*
7562 consecutive backslashes:
7564 (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*"))
7566 The string in this example is preprocessed by the Guile reader before
7567 any code is executed. The resulting argument to `make-regexp' is the
7568 string `^\* [^:]*', which is what we really want.
7570 This also means that in order to write a regular expression that
7571 matches a single backslash character, the regular expression string in
7572 the source code must include *four* backslashes. Each consecutive pair
7573 of backslashes gets translated by the Guile reader to a single
7574 backslash, and the resulting double-backslash is interpreted by the
7575 regexp engine as matching a single backslash character. Hence:
7577 (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*"))
7579 The reason for the unwieldiness of this syntax is historical. Both
7580 regular expression pattern matchers and Unix string processing systems
7581 have traditionally used backslashes with the special meanings described
7582 above. The POSIX regular expression specification and ANSI C standard
7583 both require these semantics. Attempting to abandon either convention
7584 would cause other kinds of compatibility problems, possibly more severe
7585 ones. Therefore, without extending the Scheme reader to support
7586 strings with different quoting conventions (an ungainly and confusing
7587 extension when implemented in other languages), we must adhere to this
7588 cumbersome escape syntax.
7590 * Changes to the gh_ interface
7592 * Changes to the scm_ interface
7594 * Changes to system call interfaces:
7596 ** The value returned by `raise' is now unspecified. It throws an exception
7599 *** A new procedure `sigaction' can be used to install signal handlers
7601 (sigaction signum [action] [flags])
7603 signum is the signal number, which can be specified using the value
7606 If action is omitted, sigaction returns a pair: the CAR is the current
7607 signal hander, which will be either an integer with the value SIG_DFL
7608 (default action) or SIG_IGN (ignore), or the Scheme procedure which
7609 handles the signal, or #f if a non-Scheme procedure handles the
7610 signal. The CDR contains the current sigaction flags for the handler.
7612 If action is provided, it is installed as the new handler for signum.
7613 action can be a Scheme procedure taking one argument, or the value of
7614 SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore
7615 whatever signal handler was installed before sigaction was first used.
7616 Flags can optionally be specified for the new handler (SA_RESTART is
7617 always used if the system provides it, so need not be specified.) The
7618 return value is a pair with information about the old handler as
7621 This interface does not provide access to the "signal blocking"
7622 facility. Maybe this is not needed, since the thread support may
7623 provide solutions to the problem of consistent access to data
7626 *** A new procedure `flush-all-ports' is equivalent to running
7627 `force-output' on every port open for output.
7629 ** Guile now provides information on how it was built, via the new
7630 global variable, %guile-build-info. This variable records the values
7631 of the standard GNU makefile directory variables as an assocation
7632 list, mapping variable names (symbols) onto directory paths (strings).
7633 For example, to find out where the Guile link libraries were
7634 installed, you can say:
7636 guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)"
7639 * Changes to the scm_ interface
7641 ** The new function scm_handle_by_message_noexit is just like the
7642 existing scm_handle_by_message function, except that it doesn't call
7643 exit to terminate the process. Instead, it prints a message and just
7644 returns #f. This might be a more appropriate catch-all handler for
7645 new dynamic roots and threads.
7648 Changes in Guile 1.1 (released Friday, May 16 1997):
7650 * Changes to the distribution.
7652 The Guile 1.0 distribution has been split up into several smaller
7654 guile-core --- the Guile interpreter itself.
7655 guile-tcltk --- the interface between the Guile interpreter and
7656 Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk
7657 is a toolkit for building graphical user interfaces.
7658 guile-rgx-ctax --- the interface between Guile and the Rx regular
7659 expression matcher, and the translator for the Ctax
7660 programming language. These are packaged together because the
7661 Ctax translator uses Rx to parse Ctax source code.
7663 This NEWS file describes the changes made to guile-core since the 1.0
7666 We no longer distribute the documentation, since it was either out of
7667 date, or incomplete. As soon as we have current documentation, we
7672 * Changes to the stand-alone interpreter
7674 ** guile now accepts command-line arguments compatible with SCSH, Olin
7675 Shivers' Scheme Shell.
7677 In general, arguments are evaluated from left to right, but there are
7678 exceptions. The following switches stop argument processing, and
7679 stash all remaining command-line arguments as the value returned by
7680 the (command-line) function.
7681 -s SCRIPT load Scheme source code from FILE, and exit
7682 -c EXPR evalute Scheme expression EXPR, and exit
7683 -- stop scanning arguments; run interactively
7685 The switches below are processed as they are encountered.
7686 -l FILE load Scheme source code from FILE
7687 -e FUNCTION after reading script, apply FUNCTION to
7688 command line arguments
7689 -ds do -s script at this point
7690 --emacs enable Emacs protocol (experimental)
7691 -h, --help display this help and exit
7692 -v, --version display version information and exit
7693 \ read arguments from following script lines
7695 So, for example, here is a Guile script named `ekko' (thanks, Olin)
7696 which re-implements the traditional "echo" command:
7698 #!/usr/local/bin/guile -s
7701 (map (lambda (arg) (display arg) (display " "))
7705 (main (command-line))
7707 Suppose we invoke this script as follows:
7709 ekko a speckled gecko
7711 Through the magic of Unix script processing (triggered by the `#!'
7712 token at the top of the file), /usr/local/bin/guile receives the
7713 following list of command-line arguments:
7715 ("-s" "./ekko" "a" "speckled" "gecko")
7717 Unix inserts the name of the script after the argument specified on
7718 the first line of the file (in this case, "-s"), and then follows that
7719 with the arguments given to the script. Guile loads the script, which
7720 defines the `main' function, and then applies it to the list of
7721 remaining command-line arguments, ("a" "speckled" "gecko").
7723 In Unix, the first line of a script file must take the following form:
7725 #!INTERPRETER ARGUMENT
7727 where INTERPRETER is the absolute filename of the interpreter
7728 executable, and ARGUMENT is a single command-line argument to pass to
7731 You may only pass one argument to the interpreter, and its length is
7732 limited. These restrictions can be annoying to work around, so Guile
7733 provides a general mechanism (borrowed from, and compatible with,
7734 SCSH) for circumventing them.
7736 If the ARGUMENT in a Guile script is a single backslash character,
7737 `\', Guile will open the script file, parse arguments from its second
7738 and subsequent lines, and replace the `\' with them. So, for example,
7739 here is another implementation of the `ekko' script:
7741 #!/usr/local/bin/guile \
7745 (for-each (lambda (arg) (display arg) (display " "))
7749 If the user invokes this script as follows:
7751 ekko a speckled gecko
7753 Unix expands this into
7755 /usr/local/bin/guile \ ekko a speckled gecko
7757 When Guile sees the `\' argument, it replaces it with the arguments
7758 read from the second line of the script, producing:
7760 /usr/local/bin/guile -e main -s ekko a speckled gecko
7762 This tells Guile to load the `ekko' script, and apply the function
7763 `main' to the argument list ("a" "speckled" "gecko").
7765 Here is how Guile parses the command-line arguments:
7766 - Each space character terminates an argument. This means that two
7767 spaces in a row introduce an empty-string argument.
7768 - The tab character is not permitted (unless you quote it with the
7769 backslash character, as described below), to avoid confusion.
7770 - The newline character terminates the sequence of arguments, and will
7771 also terminate a final non-empty argument. (However, a newline
7772 following a space will not introduce a final empty-string argument;
7773 it only terminates the argument list.)
7774 - The backslash character is the escape character. It escapes
7775 backslash, space, tab, and newline. The ANSI C escape sequences
7776 like \n and \t are also supported. These produce argument
7777 constituents; the two-character combination \n doesn't act like a
7778 terminating newline. The escape sequence \NNN for exactly three
7779 octal digits reads as the character whose ASCII code is NNN. As
7780 above, characters produced this way are argument constituents.
7781 Backslash followed by other characters is not allowed.
7783 * Changes to the procedure for linking libguile with your programs
7785 ** Guile now builds and installs a shared guile library, if your
7786 system support shared libraries. (It still builds a static library on
7787 all systems.) Guile automatically detects whether your system
7788 supports shared libraries. To prevent Guile from buildisg shared
7789 libraries, pass the `--disable-shared' flag to the configure script.
7791 Guile takes longer to compile when it builds shared libraries, because
7792 it must compile every file twice --- once to produce position-
7793 independent object code, and once to produce normal object code.
7795 ** The libthreads library has been merged into libguile.
7797 To link a program against Guile, you now need only link against
7798 -lguile and -lqt; -lthreads is no longer needed. If you are using
7799 autoconf to generate configuration scripts for your application, the
7800 following lines should suffice to add the appropriate libraries to
7803 ### Find quickthreads and libguile.
7804 AC_CHECK_LIB(qt, main)
7805 AC_CHECK_LIB(guile, scm_shell)
7807 * Changes to Scheme functions
7809 ** Guile Scheme's special syntax for keyword objects is now optional,
7810 and disabled by default.
7812 The syntax variation from R4RS made it difficult to port some
7813 interesting packages to Guile. The routines which accepted keyword
7814 arguments (mostly in the module system) have been modified to also
7815 accept symbols whose names begin with `:'.
7817 To change the keyword syntax, you must first import the (ice-9 debug)
7819 (use-modules (ice-9 debug))
7821 Then you can enable the keyword syntax as follows:
7822 (read-set! keywords 'prefix)
7824 To disable keyword syntax, do this:
7825 (read-set! keywords #f)
7827 ** Many more primitive functions accept shared substrings as
7828 arguments. In the past, these functions required normal, mutable
7829 strings as arguments, although they never made use of this
7832 ** The uniform array functions now operate on byte vectors. These
7833 functions are `array-fill!', `serial-array-copy!', `array-copy!',
7834 `serial-array-map', `array-map', `array-for-each', and
7837 ** The new functions `trace' and `untrace' implement simple debugging
7838 support for Scheme functions.
7840 The `trace' function accepts any number of procedures as arguments,
7841 and tells the Guile interpreter to display each procedure's name and
7842 arguments each time the procedure is invoked. When invoked with no
7843 arguments, `trace' returns the list of procedures currently being
7846 The `untrace' function accepts any number of procedures as arguments,
7847 and tells the Guile interpreter not to trace them any more. When
7848 invoked with no arguments, `untrace' untraces all curretly traced
7851 The tracing in Guile has an advantage over most other systems: we
7852 don't create new procedure objects, but mark the procedure objects
7853 themselves. This means that anonymous and internal procedures can be
7856 ** The function `assert-repl-prompt' has been renamed to
7857 `set-repl-prompt!'. It takes one argument, PROMPT.
7858 - If PROMPT is #f, the Guile read-eval-print loop will not prompt.
7859 - If PROMPT is a string, we use it as a prompt.
7860 - If PROMPT is a procedure accepting no arguments, we call it, and
7861 display the result as a prompt.
7862 - Otherwise, we display "> ".
7864 ** The new function `eval-string' reads Scheme expressions from a
7865 string and evaluates them, returning the value of the last expression
7866 in the string. If the string contains no expressions, it returns an
7869 ** The new function `thunk?' returns true iff its argument is a
7870 procedure of zero arguments.
7872 ** `defined?' is now a builtin function, instead of syntax. This
7873 means that its argument should be quoted. It returns #t iff its
7874 argument is bound in the current module.
7876 ** The new syntax `use-modules' allows you to add new modules to your
7877 environment without re-typing a complete `define-module' form. It
7878 accepts any number of module names as arguments, and imports their
7879 public bindings into the current module.
7881 ** The new function (module-defined? NAME MODULE) returns true iff
7882 NAME, a symbol, is defined in MODULE, a module object.
7884 ** The new function `builtin-bindings' creates and returns a hash
7885 table containing copies of all the root module's bindings.
7887 ** The new function `builtin-weak-bindings' does the same as
7888 `builtin-bindings', but creates a doubly-weak hash table.
7890 ** The `equal?' function now considers variable objects to be
7891 equivalent if they have the same name and the same value.
7893 ** The new function `command-line' returns the command-line arguments
7894 given to Guile, as a list of strings.
7896 When using guile as a script interpreter, `command-line' returns the
7897 script's arguments; those processed by the interpreter (like `-s' or
7898 `-c') are omitted. (In other words, you get the normal, expected
7899 behavior.) Any application that uses scm_shell to process its
7900 command-line arguments gets this behavior as well.
7902 ** The new function `load-user-init' looks for a file called `.guile'
7903 in the user's home directory, and loads it if it exists. This is
7904 mostly for use by the code generated by scm_compile_shell_switches,
7905 but we thought it might also be useful in other circumstances.
7907 ** The new function `log10' returns the base-10 logarithm of its
7910 ** Changes to I/O functions
7912 *** The functions `read', `primitive-load', `read-and-eval!', and
7913 `primitive-load-path' no longer take optional arguments controlling
7914 case insensitivity and a `#' parser.
7916 Case sensitivity is now controlled by a read option called
7917 `case-insensitive'. The user can add new `#' syntaxes with the
7918 `read-hash-extend' function (see below).
7920 *** The new function `read-hash-extend' allows the user to change the
7921 syntax of Guile Scheme in a somewhat controlled way.
7923 (read-hash-extend CHAR PROC)
7924 When parsing S-expressions, if we read a `#' character followed by
7925 the character CHAR, use PROC to parse an object from the stream.
7926 If PROC is #f, remove any parsing procedure registered for CHAR.
7928 The reader applies PROC to two arguments: CHAR and an input port.
7930 *** The new functions read-delimited and read-delimited! provide a
7931 general mechanism for doing delimited input on streams.
7933 (read-delimited DELIMS [PORT HANDLE-DELIM])
7934 Read until we encounter one of the characters in DELIMS (a string),
7935 or end-of-file. PORT is the input port to read from; it defaults to
7936 the current input port. The HANDLE-DELIM parameter determines how
7937 the terminating character is handled; it should be one of the
7940 'trim omit delimiter from result
7941 'peek leave delimiter character in input stream
7942 'concat append delimiter character to returned value
7943 'split return a pair: (RESULT . TERMINATOR)
7945 HANDLE-DELIM defaults to 'peek.
7947 (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END])
7948 A side-effecting variant of `read-delimited'.
7950 The data is written into the string BUF at the indices in the
7951 half-open interval [START, END); the default interval is the whole
7952 string: START = 0 and END = (string-length BUF). The values of
7953 START and END must specify a well-defined interval in BUF, i.e.
7954 0 <= START <= END <= (string-length BUF).
7956 It returns NBYTES, the number of bytes read. If the buffer filled
7957 up without a delimiter character being found, it returns #f. If the
7958 port is at EOF when the read starts, it returns the EOF object.
7960 If an integer is returned (i.e., the read is successfully terminated
7961 by reading a delimiter character), then the HANDLE-DELIM parameter
7962 determines how to handle the terminating character. It is described
7963 above, and defaults to 'peek.
7965 (The descriptions of these functions were borrowed from the SCSH
7966 manual, by Olin Shivers and Brian Carlstrom.)
7968 *** The `%read-delimited!' function is the primitive used to implement
7969 `read-delimited' and `read-delimited!'.
7971 (%read-delimited! DELIMS BUF GOBBLE? [PORT START END])
7973 This returns a pair of values: (TERMINATOR . NUM-READ).
7974 - TERMINATOR describes why the read was terminated. If it is a
7975 character or the eof object, then that is the value that terminated
7976 the read. If it is #f, the function filled the buffer without finding
7977 a delimiting character.
7978 - NUM-READ is the number of characters read into BUF.
7980 If the read is successfully terminated by reading a delimiter
7981 character, then the gobble? parameter determines what to do with the
7982 terminating character. If true, the character is removed from the
7983 input stream; if false, the character is left in the input stream
7984 where a subsequent read operation will retrieve it. In either case,
7985 the character is also the first value returned by the procedure call.
7987 (The descriptions of this function was borrowed from the SCSH manual,
7988 by Olin Shivers and Brian Carlstrom.)
7990 *** The `read-line' and `read-line!' functions have changed; they now
7991 trim the terminator by default; previously they appended it to the
7992 returned string. For the old behavior, use (read-line PORT 'concat).
7994 *** The functions `uniform-array-read!' and `uniform-array-write!' now
7995 take new optional START and END arguments, specifying the region of
7996 the array to read and write.
7998 *** The `ungetc-char-ready?' function has been removed. We feel it's
7999 inappropriate for an interface to expose implementation details this
8002 ** Changes to the Unix library and system call interface
8004 *** The new fcntl function provides access to the Unix `fcntl' system
8007 (fcntl PORT COMMAND VALUE)
8008 Apply COMMAND to PORT's file descriptor, with VALUE as an argument.
8009 Values for COMMAND are:
8011 F_DUPFD duplicate a file descriptor
8012 F_GETFD read the descriptor's close-on-exec flag
8013 F_SETFD set the descriptor's close-on-exec flag to VALUE
8014 F_GETFL read the descriptor's flags, as set on open
8015 F_SETFL set the descriptor's flags, as set on open to VALUE
8016 F_GETOWN return the process ID of a socket's owner, for SIGIO
8017 F_SETOWN set the process that owns a socket to VALUE, for SIGIO
8018 FD_CLOEXEC not sure what this is
8020 For details, see the documentation for the fcntl system call.
8022 *** The arguments to `select' have changed, for compatibility with
8023 SCSH. The TIMEOUT parameter may now be non-integral, yielding the
8024 expected behavior. The MILLISECONDS parameter has been changed to
8025 MICROSECONDS, to more closely resemble the underlying system call.
8026 The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the
8027 corresponding return set will be the same.
8029 *** The arguments to the `mknod' system call have changed. They are
8032 (mknod PATH TYPE PERMS DEV)
8033 Create a new file (`node') in the file system. PATH is the name of
8034 the file to create. TYPE is the kind of file to create; it should
8035 be 'fifo, 'block-special, or 'char-special. PERMS specifies the
8036 permission bits to give the newly created file. If TYPE is
8037 'block-special or 'char-special, DEV specifies which device the
8038 special file refers to; its interpretation depends on the kind of
8039 special file being created.
8041 *** The `fork' function has been renamed to `primitive-fork', to avoid
8042 clashing with various SCSH forks.
8044 *** The `recv' and `recvfrom' functions have been renamed to `recv!'
8045 and `recvfrom!'. They no longer accept a size for a second argument;
8046 you must pass a string to hold the received value. They no longer
8047 return the buffer. Instead, `recv' returns the length of the message
8048 received, and `recvfrom' returns a pair containing the packet's length
8049 and originating address.
8051 *** The file descriptor datatype has been removed, as have the
8052 `read-fd', `write-fd', `close', `lseek', and `dup' functions.
8053 We plan to replace these functions with a SCSH-compatible interface.
8055 *** The `create' function has been removed; it's just a special case
8058 *** There are new functions to break down process termination status
8059 values. In the descriptions below, STATUS is a value returned by
8062 (status:exit-val STATUS)
8063 If the child process exited normally, this function returns the exit
8064 code for the child process (i.e., the value passed to exit, or
8065 returned from main). If the child process did not exit normally,
8066 this function returns #f.
8068 (status:stop-sig STATUS)
8069 If the child process was suspended by a signal, this function
8070 returns the signal that suspended the child. Otherwise, it returns
8073 (status:term-sig STATUS)
8074 If the child process terminated abnormally, this function returns
8075 the signal that terminated the child. Otherwise, this function
8078 POSIX promises that exactly one of these functions will return true on
8079 a valid STATUS value.
8081 These functions are compatible with SCSH.
8083 *** There are new accessors and setters for the broken-out time vectors
8084 returned by `localtime', `gmtime', and that ilk. They are:
8086 Component Accessor Setter
8087 ========================= ============ ============
8088 seconds tm:sec set-tm:sec
8089 minutes tm:min set-tm:min
8090 hours tm:hour set-tm:hour
8091 day of the month tm:mday set-tm:mday
8092 month tm:mon set-tm:mon
8093 year tm:year set-tm:year
8094 day of the week tm:wday set-tm:wday
8095 day in the year tm:yday set-tm:yday
8096 daylight saving time tm:isdst set-tm:isdst
8097 GMT offset, seconds tm:gmtoff set-tm:gmtoff
8098 name of time zone tm:zone set-tm:zone
8100 *** There are new accessors for the vectors returned by `uname',
8101 describing the host system:
8104 ============================================== ================
8105 name of the operating system implementation utsname:sysname
8106 network name of this machine utsname:nodename
8107 release level of the operating system utsname:release
8108 version level of the operating system utsname:version
8109 machine hardware platform utsname:machine
8111 *** There are new accessors for the vectors returned by `getpw',
8112 `getpwnam', `getpwuid', and `getpwent', describing entries from the
8113 system's user database:
8116 ====================== =================
8117 user name passwd:name
8118 user password passwd:passwd
8121 real name passwd:gecos
8122 home directory passwd:dir
8123 shell program passwd:shell
8125 *** There are new accessors for the vectors returned by `getgr',
8126 `getgrnam', `getgrgid', and `getgrent', describing entries from the
8127 system's group database:
8130 ======================= ============
8131 group name group:name
8132 group password group:passwd
8134 group members group:mem
8136 *** There are new accessors for the vectors returned by `gethost',
8137 `gethostbyaddr', `gethostbyname', and `gethostent', describing
8141 ========================= ===============
8142 official name of host hostent:name
8143 alias list hostent:aliases
8144 host address type hostent:addrtype
8145 length of address hostent:length
8146 list of addresses hostent:addr-list
8148 *** There are new accessors for the vectors returned by `getnet',
8149 `getnetbyaddr', `getnetbyname', and `getnetent', describing internet
8153 ========================= ===============
8154 official name of net netent:name
8155 alias list netent:aliases
8156 net number type netent:addrtype
8157 net number netent:net
8159 *** There are new accessors for the vectors returned by `getproto',
8160 `getprotobyname', `getprotobynumber', and `getprotoent', describing
8164 ========================= ===============
8165 official protocol name protoent:name
8166 alias list protoent:aliases
8167 protocol number protoent:proto
8169 *** There are new accessors for the vectors returned by `getserv',
8170 `getservbyname', `getservbyport', and `getservent', describing
8174 ========================= ===============
8175 official service name servent:name
8176 alias list servent:aliases
8177 port number servent:port
8178 protocol to use servent:proto
8180 *** There are new accessors for the sockaddr structures returned by
8181 `accept', `getsockname', `getpeername', `recvfrom!':
8184 ======================================== ===============
8185 address format (`family') sockaddr:fam
8186 path, for file domain addresses sockaddr:path
8187 address, for internet domain addresses sockaddr:addr
8188 TCP or UDP port, for internet sockaddr:port
8190 *** The `getpwent', `getgrent', `gethostent', `getnetent',
8191 `getprotoent', and `getservent' functions now return #f at the end of
8192 the user database. (They used to throw an exception.)
8194 Note that calling MUMBLEent function is equivalent to calling the
8195 corresponding MUMBLE function with no arguments.
8197 *** The `setpwent', `setgrent', `sethostent', `setnetent',
8198 `setprotoent', and `setservent' routines now take no arguments.
8200 *** The `gethost', `getproto', `getnet', and `getserv' functions now
8201 provide more useful information when they throw an exception.
8203 *** The `lnaof' function has been renamed to `inet-lnaof'.
8205 *** Guile now claims to have the `current-time' feature.
8207 *** The `mktime' function now takes an optional second argument ZONE,
8208 giving the time zone to use for the conversion. ZONE should be a
8209 string, in the same format as expected for the "TZ" environment variable.
8211 *** The `strptime' function now returns a pair (TIME . COUNT), where
8212 TIME is the parsed time as a vector, and COUNT is the number of
8213 characters from the string left unparsed. This function used to
8214 return the remaining characters as a string.
8216 *** The `gettimeofday' function has replaced the old `time+ticks' function.
8217 The return value is now (SECONDS . MICROSECONDS); the fractional
8218 component is no longer expressed in "ticks".
8220 *** The `ticks/sec' constant has been removed, in light of the above change.
8222 * Changes to the gh_ interface
8224 ** gh_eval_str() now returns an SCM object which is the result of the
8227 ** gh_scm2str() now copies the Scheme data to a caller-provided C
8230 ** gh_scm2newstr() now makes a C array, copies the Scheme data to it,
8231 and returns the array
8233 ** gh_scm2str0() is gone: there is no need to distinguish
8234 null-terminated from non-null-terminated, since gh_scm2newstr() allows
8235 the user to interpret the data both ways.
8237 * Changes to the scm_ interface
8239 ** The new function scm_symbol_value0 provides an easy way to get a
8240 symbol's value from C code:
8242 SCM scm_symbol_value0 (char *NAME)
8243 Return the value of the symbol named by the null-terminated string
8244 NAME in the current module. If the symbol named NAME is unbound in
8245 the current module, return SCM_UNDEFINED.
8247 ** The new function scm_sysintern0 creates new top-level variables,
8248 without assigning them a value.
8250 SCM scm_sysintern0 (char *NAME)
8251 Create a new Scheme top-level variable named NAME. NAME is a
8252 null-terminated string. Return the variable's value cell.
8254 ** The function scm_internal_catch is the guts of catch. It handles
8255 all the mechanics of setting up a catch target, invoking the catch
8256 body, and perhaps invoking the handler if the body does a throw.
8258 The function is designed to be usable from C code, but is general
8259 enough to implement all the semantics Guile Scheme expects from throw.
8261 TAG is the catch tag. Typically, this is a symbol, but this function
8262 doesn't actually care about that.
8264 BODY is a pointer to a C function which runs the body of the catch;
8265 this is the code you can throw from. We call it like this:
8266 BODY (BODY_DATA, JMPBUF)
8268 BODY_DATA is just the BODY_DATA argument we received; we pass it
8269 through to BODY as its first argument. The caller can make
8270 BODY_DATA point to anything useful that BODY might need.
8271 JMPBUF is the Scheme jmpbuf object corresponding to this catch,
8272 which we have just created and initialized.
8274 HANDLER is a pointer to a C function to deal with a throw to TAG,
8275 should one occur. We call it like this:
8276 HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS)
8278 HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the
8279 same idea as BODY_DATA above.
8280 THROWN_TAG is the tag that the user threw to; usually this is
8281 TAG, but it could be something else if TAG was #t (i.e., a
8282 catch-all), or the user threw to a jmpbuf.
8283 THROW_ARGS is the list of arguments the user passed to the THROW
8286 BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA
8287 is just a pointer we pass through to HANDLER. We don't actually
8288 use either of those pointers otherwise ourselves. The idea is
8289 that, if our caller wants to communicate something to BODY or
8290 HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and
8291 HANDLER can then use. Think of it as a way to make BODY and
8292 HANDLER closures, not just functions; MUMBLE_DATA points to the
8295 Of course, it's up to the caller to make sure that any data a
8296 MUMBLE_DATA needs is protected from GC. A common way to do this is
8297 to make MUMBLE_DATA a pointer to data stored in an automatic
8298 structure variable; since the collector must scan the stack for
8299 references anyway, this assures that any references in MUMBLE_DATA
8302 ** The new function scm_internal_lazy_catch is exactly like
8303 scm_internal_catch, except:
8305 - It does not unwind the stack (this is the major difference).
8306 - If handler returns, its value is returned from the throw.
8307 - BODY always receives #f as its JMPBUF argument (since there's no
8308 jmpbuf associated with a lazy catch, because we don't unwind the
8311 ** scm_body_thunk is a new body function you can pass to
8312 scm_internal_catch if you want the body to be like Scheme's `catch'
8313 --- a thunk, or a function of one argument if the tag is #f.
8315 BODY_DATA is a pointer to a scm_body_thunk_data structure, which
8316 contains the Scheme procedure to invoke as the body, and the tag
8317 we're catching. If the tag is #f, then we pass JMPBUF (created by
8318 scm_internal_catch) to the body procedure; otherwise, the body gets
8321 ** scm_handle_by_proc is a new handler function you can pass to
8322 scm_internal_catch if you want the handler to act like Scheme's catch
8323 --- call a procedure with the tag and the throw arguments.
8325 If the user does a throw to this catch, this function runs a handler
8326 procedure written in Scheme. HANDLER_DATA is a pointer to an SCM
8327 variable holding the Scheme procedure object to invoke. It ought to
8328 be a pointer to an automatic variable (i.e., one living on the stack),
8329 or the procedure object should be otherwise protected from GC.
8331 ** scm_handle_by_message is a new handler function to use with
8332 `scm_internal_catch' if you want Guile to print a message and die.
8333 It's useful for dealing with throws to uncaught keys at the top level.
8335 HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a
8336 message header to print; if zero, we use "guile" instead. That
8337 text is followed by a colon, then the message described by ARGS.
8339 ** The return type of scm_boot_guile is now void; the function does
8340 not return a value, and indeed, never returns at all.
8342 ** The new function scm_shell makes it easy for user applications to
8343 process command-line arguments in a way that is compatible with the
8344 stand-alone guile interpreter (which is in turn compatible with SCSH,
8347 To use the scm_shell function, first initialize any guile modules
8348 linked into your application, and then call scm_shell with the values
8349 of ARGC and ARGV your `main' function received. scm_shell will add
8350 any SCSH-style meta-arguments from the top of the script file to the
8351 argument vector, and then process the command-line arguments. This
8352 generally means loading a script file or starting up an interactive
8353 command interpreter. For details, see "Changes to the stand-alone
8356 ** The new functions scm_get_meta_args and scm_count_argv help you
8357 implement the SCSH-style meta-argument, `\'.
8359 char **scm_get_meta_args (int ARGC, char **ARGV)
8360 If the second element of ARGV is a string consisting of a single
8361 backslash character (i.e. "\\" in Scheme notation), open the file
8362 named by the following argument, parse arguments from it, and return
8363 the spliced command line. The returned array is terminated by a
8366 For details of argument parsing, see above, under "guile now accepts
8367 command-line arguments compatible with SCSH..."
8369 int scm_count_argv (char **ARGV)
8370 Count the arguments in ARGV, assuming it is terminated by a null
8373 For an example of how these functions might be used, see the source
8374 code for the function scm_shell in libguile/script.c.
8376 You will usually want to use scm_shell instead of calling this
8379 ** The new function scm_compile_shell_switches turns an array of
8380 command-line arguments into Scheme code to carry out the actions they
8381 describe. Given ARGC and ARGV, it returns a Scheme expression to
8382 evaluate, and calls scm_set_program_arguments to make any remaining
8383 command-line arguments available to the Scheme code. For example,
8384 given the following arguments:
8386 -e main -s ekko a speckled gecko
8388 scm_set_program_arguments will return the following expression:
8390 (begin (load "ekko") (main (command-line)) (quit))
8392 You will usually want to use scm_shell instead of calling this
8395 ** The function scm_shell_usage prints a usage message appropriate for
8396 an interpreter that uses scm_compile_shell_switches to handle its
8397 command-line arguments.
8399 void scm_shell_usage (int FATAL, char *MESSAGE)
8400 Print a usage message to the standard error output. If MESSAGE is
8401 non-zero, write it before the usage message, followed by a newline.
8402 If FATAL is non-zero, exit the process, using FATAL as the
8403 termination status. (If you want to be compatible with Guile,
8404 always use 1 as the exit status when terminating due to command-line
8407 You will usually want to use scm_shell instead of calling this
8410 ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no
8411 expressions. It used to return SCM_EOL. Earth-shattering.
8413 ** The macros for declaring scheme objects in C code have been
8414 rearranged slightly. They are now:
8416 SCM_SYMBOL (C_NAME, SCHEME_NAME)
8417 Declare a static SCM variable named C_NAME, and initialize it to
8418 point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should
8419 be a C identifier, and SCHEME_NAME should be a C string.
8421 SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME)
8422 Just like SCM_SYMBOL, but make C_NAME globally visible.
8424 SCM_VCELL (C_NAME, SCHEME_NAME)
8425 Create a global variable at the Scheme level named SCHEME_NAME.
8426 Declare a static SCM variable named C_NAME, and initialize it to
8427 point to the Scheme variable's value cell.
8429 SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME)
8430 Just like SCM_VCELL, but make C_NAME globally visible.
8432 The `guile-snarf' script writes initialization code for these macros
8433 to its standard output, given C source code as input.
8435 The SCM_GLOBAL macro is gone.
8437 ** The scm_read_line and scm_read_line_x functions have been replaced
8438 by Scheme code based on the %read-delimited! procedure (known to C
8439 code as scm_read_delimited_x). See its description above for more
8442 ** The function scm_sys_open has been renamed to scm_open. It now
8443 returns a port instead of an FD object.
8445 * The dynamic linking support has changed. For more information, see
8446 libguile/DYNAMIC-LINKING.
8451 User-visible changes from Thursday, September 5, 1996 until Guile 1.0
8454 * Changes to the 'guile' program:
8456 ** Guile now loads some new files when it starts up. Guile first
8457 searches the load path for init.scm, and loads it if found. Then, if
8458 Guile is not being used to execute a script, and the user's home
8459 directory contains a file named `.guile', Guile loads that.
8461 ** You can now use Guile as a shell script interpreter.
8463 To paraphrase the SCSH manual:
8465 When Unix tries to execute an executable file whose first two
8466 characters are the `#!', it treats the file not as machine code to
8467 be directly executed by the native processor, but as source code
8468 to be executed by some interpreter. The interpreter to use is
8469 specified immediately after the #! sequence on the first line of
8470 the source file. The kernel reads in the name of the interpreter,
8471 and executes that instead. It passes the interpreter the source
8472 filename as its first argument, with the original arguments
8473 following. Consult the Unix man page for the `exec' system call
8474 for more information.
8476 Now you can use Guile as an interpreter, using a mechanism which is a
8477 compatible subset of that provided by SCSH.
8479 Guile now recognizes a '-s' command line switch, whose argument is the
8480 name of a file of Scheme code to load. It also treats the two
8481 characters `#!' as the start of a comment, terminated by `!#'. Thus,
8482 to make a file of Scheme code directly executable by Unix, insert the
8483 following two lines at the top of the file:
8485 #!/usr/local/bin/guile -s
8488 Guile treats the argument of the `-s' command-line switch as the name
8489 of a file of Scheme code to load, and treats the sequence `#!' as the
8490 start of a block comment, terminated by `!#'.
8492 For example, here's a version of 'echo' written in Scheme:
8494 #!/usr/local/bin/guile -s
8496 (let loop ((args (cdr (program-arguments))))
8499 (display (car args))
8500 (if (pair? (cdr args))
8502 (loop (cdr args)))))
8505 Why does `#!' start a block comment terminated by `!#', instead of the
8506 end of the line? That is the notation SCSH uses, and although we
8507 don't yet support the other SCSH features that motivate that choice,
8508 we would like to be backward-compatible with any existing Guile
8509 scripts once we do. Furthermore, if the path to Guile on your system
8510 is too long for your kernel, you can start the script with this
8514 exec /really/long/path/to/guile -s "$0" ${1+"$@"}
8517 Note that some very old Unix systems don't support the `#!' syntax.
8520 ** You can now run Guile without installing it.
8522 Previous versions of the interactive Guile interpreter (`guile')
8523 couldn't start up unless Guile's Scheme library had been installed;
8524 they used the value of the environment variable `SCHEME_LOAD_PATH'
8525 later on in the startup process, but not to find the startup code
8526 itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme
8529 To run Guile without installing it, build it in the normal way, and
8530 then set the environment variable `SCHEME_LOAD_PATH' to a
8531 colon-separated list of directories, including the top-level directory
8532 of the Guile sources. For example, if you unpacked Guile so that the
8533 full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then
8536 export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3
8539 ** Guile's read-eval-print loop no longer prints #<unspecified>
8540 results. If the user wants to see this, she can evaluate the
8541 expression (assert-repl-print-unspecified #t), perhaps in her startup
8544 ** Guile no longer shows backtraces by default when an error occurs;
8545 however, it does display a message saying how to get one, and how to
8546 request that they be displayed by default. After an error, evaluate
8548 to see a backtrace, and
8549 (debug-enable 'backtrace)
8550 to see them by default.
8554 * Changes to Guile Scheme:
8556 ** Guile now distinguishes between #f and the empty list.
8558 This is for compatibility with the IEEE standard, the (possibly)
8559 upcoming Revised^5 Report on Scheme, and many extant Scheme
8562 Guile used to have #f and '() denote the same object, to make Scheme's
8563 type system more compatible with Emacs Lisp's. However, the change
8564 caused too much trouble for Scheme programmers, and we found another
8565 way to reconcile Emacs Lisp with Scheme that didn't require this.
8568 ** Guile's delq, delv, delete functions, and their destructive
8569 counterparts, delq!, delv!, and delete!, now remove all matching
8570 elements from the list, not just the first. This matches the behavior
8571 of the corresponding Emacs Lisp functions, and (I believe) the Maclisp
8572 functions which inspired them.
8574 I recognize that this change may break code in subtle ways, but it
8575 seems best to make the change before the FSF's first Guile release,
8579 ** The compiled-library-path function has been deleted from libguile.
8581 ** The facilities for loading Scheme source files have changed.
8583 *** The variable %load-path now tells Guile which directories to search
8584 for Scheme code. Its value is a list of strings, each of which names
8587 *** The variable %load-extensions now tells Guile which extensions to
8588 try appending to a filename when searching the load path. Its value
8589 is a list of strings. Its default value is ("" ".scm").
8591 *** (%search-load-path FILENAME) searches the directories listed in the
8592 value of the %load-path variable for a Scheme file named FILENAME,
8593 with all the extensions listed in %load-extensions. If it finds a
8594 match, then it returns its full filename. If FILENAME is absolute, it
8595 returns it unchanged. Otherwise, it returns #f.
8597 %search-load-path will not return matches that refer to directories.
8599 *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP)
8600 uses %seach-load-path to find a file named FILENAME, and loads it if
8601 it finds it. If it can't read FILENAME for any reason, it throws an
8604 The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the
8607 *** load uses the same searching semantics as primitive-load.
8609 *** The functions %try-load, try-load-with-path, %load, load-with-path,
8610 basic-try-load-with-path, basic-load-with-path, try-load-module-with-
8611 path, and load-module-with-path have been deleted. The functions
8612 above should serve their purposes.
8614 *** If the value of the variable %load-hook is a procedure,
8615 `primitive-load' applies its value to the name of the file being
8616 loaded (without the load path directory name prepended). If its value
8617 is #f, it is ignored. Otherwise, an error occurs.
8619 This is mostly useful for printing load notification messages.
8622 ** The function `eval!' is no longer accessible from the scheme level.
8623 We can't allow operations which introduce glocs into the scheme level,
8624 because Guile's type system can't handle these as data. Use `eval' or
8625 `read-and-eval!' (see below) as replacement.
8627 ** The new function read-and-eval! reads an expression from PORT,
8628 evaluates it, and returns the result. This is more efficient than
8629 simply calling `read' and `eval', since it is not necessary to make a
8630 copy of the expression for the evaluator to munge.
8632 Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as
8633 for the `read' function.
8636 ** The function `int?' has been removed; its definition was identical
8637 to that of `integer?'.
8639 ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should
8640 use the R4RS names for these functions.
8642 ** The function object-properties no longer returns the hash handle;
8643 it simply returns the object's property list.
8645 ** Many functions have been changed to throw errors, instead of
8646 returning #f on failure. The point of providing exception handling in
8647 the language is to simplify the logic of user code, but this is less
8648 useful if Guile's primitives don't throw exceptions.
8650 ** The function `fileno' has been renamed from `%fileno'.
8652 ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0.
8655 * Changes to Guile's C interface:
8657 ** The library's initialization procedure has been simplified.
8658 scm_boot_guile now has the prototype:
8660 void scm_boot_guile (int ARGC,
8662 void (*main_func) (),
8665 scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV.
8666 MAIN_FUNC should do all the work of the program (initializing other
8667 packages, reading user input, etc.) before returning. When MAIN_FUNC
8668 returns, call exit (0); this function never returns. If you want some
8669 other exit value, MAIN_FUNC may call exit itself.
8671 scm_boot_guile arranges for program-arguments to return the strings
8672 given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call
8673 scm_set_program_arguments with the final list, so Scheme code will
8674 know which arguments have been processed.
8676 scm_boot_guile establishes a catch-all catch handler which prints an
8677 error message and exits the process. This means that Guile exits in a
8678 coherent way when system errors occur and the user isn't prepared to
8679 handle it. If the user doesn't like this behavior, they can establish
8680 their own universal catcher in MAIN_FUNC to shadow this one.
8682 Why must the caller do all the real work from MAIN_FUNC? The garbage
8683 collector assumes that all local variables of type SCM will be above
8684 scm_boot_guile's stack frame on the stack. If you try to manipulate
8685 SCM values after this function returns, it's the luck of the draw
8686 whether the GC will be able to find the objects you allocate. So,
8687 scm_boot_guile function exits, rather than returning, to discourage
8688 people from making that mistake.
8690 The IN, OUT, and ERR arguments were removed; there are other
8691 convenient ways to override these when desired.
8693 The RESULT argument was deleted; this function should never return.
8695 The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more
8699 ** Guile's header files should no longer conflict with your system's
8702 In order to compile code which #included <libguile.h>, previous
8703 versions of Guile required you to add a directory containing all the
8704 Guile header files to your #include path. This was a problem, since
8705 Guile's header files have names which conflict with many systems'
8708 Now only <libguile.h> need appear in your #include path; you must
8709 refer to all Guile's other header files as <libguile/mumble.h>.
8710 Guile's installation procedure puts libguile.h in $(includedir), and
8711 the rest in $(includedir)/libguile.
8714 ** Two new C functions, scm_protect_object and scm_unprotect_object,
8715 have been added to the Guile library.
8717 scm_protect_object (OBJ) protects OBJ from the garbage collector.
8718 OBJ will not be freed, even if all other references are dropped,
8719 until someone does scm_unprotect_object (OBJ). Both functions
8722 Note that calls to scm_protect_object do not nest. You can call
8723 scm_protect_object any number of times on a given object, and the
8724 next call to scm_unprotect_object will unprotect it completely.
8726 Basically, scm_protect_object and scm_unprotect_object just
8727 maintain a list of references to things. Since the GC knows about
8728 this list, all objects it mentions stay alive. scm_protect_object
8729 adds its argument to the list; scm_unprotect_object remove its
8730 argument from the list.
8733 ** scm_eval_0str now returns the value of the last expression
8736 ** The new function scm_read_0str reads an s-expression from a
8737 null-terminated string, and returns it.
8739 ** The new function `scm_stdio_to_port' converts a STDIO file pointer
8740 to a Scheme port object.
8742 ** The new function `scm_set_program_arguments' allows C code to set
8743 the value returned by the Scheme `program-arguments' function.
8748 * Guile no longer includes sophisticated Tcl/Tk support.
8750 The old Tcl/Tk support was unsatisfying to us, because it required the
8751 user to link against the Tcl library, as well as Tk and Guile. The
8752 interface was also un-lispy, in that it preserved Tcl/Tk's practice of
8753 referring to widgets by names, rather than exporting widgets to Scheme
8754 code as a special datatype.
8756 In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk
8757 maintainers described some very interesting changes in progress to the
8758 Tcl/Tk internals, which would facilitate clean interfaces between lone
8759 Tk and other interpreters --- even for garbage-collected languages
8760 like Scheme. They expected the new Tk to be publicly available in the
8763 Since it seems that Guile might soon have a new, cleaner interface to
8764 lone Tk, and that the old Guile/Tk glue code would probably need to be
8765 completely rewritten, we (Jim Blandy and Richard Stallman) have
8766 decided not to support the old code. We'll spend the time instead on
8767 a good interface to the newer Tk, as soon as it is available.
8769 Until then, gtcltk-lib provides trivial, low-maintenance functionality.
8772 Copyright information:
8774 Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc.
8776 Permission is granted to anyone to make or distribute verbatim copies
8777 of this document as received, in any medium, provided that the
8778 copyright notice and this permission notice are preserved,
8779 thus giving the recipient permission to redistribute in turn.
8781 Permission is granted to distribute modified versions
8782 of this document, or of portions of it,
8783 under the above conditions, provided also that they
8784 carry prominent notices stating who last changed them.
8789 paragraph-separate: "[
\f]*$"