1 2006-01-04 Abel Cheung <maddog@linuxhall.org>
3 * configure.in: Added "ml" "zh_HK" to ALL_LINGUAS.
5 2006-01-03 Martyn Russell <martyn@imendio.com>
7 * docs/reference/glib/glib-sections.txt:
9 * glib/gthreadpool.[ch]:
10 - Added new API g_thread_pool_get_idle_time() and
11 g_thread_pool_set_idle_time(). (#324228).
13 * tests/threadpool-test.c:
14 - Updated test case to do thread pool sorting, thread pool with
15 no sorting and a thread pool with idle thread timeouts.
17 2006-01-03 Matthias Clasen <mclasen@redhat.com>
19 * glib/gmain.h: Add new functions here, too.
21 * glib/glib.symbols: Add new functions.
23 2005-12-20 Michael Meeks <michael.meeks@novell.com>
25 * glib/gmain.c (g_main_context_is_owner): new method
26 to determine if the current thread is the owner of the
29 2006-01-02 Matthias Clasen <mclasen@redhat.com>
33 * glib/gstdio.c (g_rmdir): Don't provide g_rmdir() as a macro
34 expanding to rmdir, since rmdir is not declared in a portable
35 system header we can include in gstdio.h. (#325249, Jani Monoses)
37 2006-01-01 Tor Lillqvist <tml@novell.com>
39 * glib/gspawn-win32.c (g_spawn_sync_utf8): Set the GIOChannels for
40 stdout and stderr to unbuffered. Otherwise the giochannel layer
41 will try to read from them regardless whether the
42 g_io_channel_win32_poll() call here has indicated
43 readability or not. (#325310)
45 2005-12-29 Matthias Clasen <mclasen@redhat.com>
47 * glib/gutils.c (glib_check_version): Fix a copy-and-paste error
48 in the docs. (#325273, Declan Naughton)
50 2005-12-27 Manish Singh <yosh@gimp.org>
52 * tests/run-collate-tests.sh: use LC_ALL instead of LC_COLLATE, to
53 make sure we really override things.
55 2005-12-27 Matthias Clasen <mclasen@redhat.com>
57 Fix #316221, Michal Benes, Stanislav Brabec;
59 * configure.in: Fix a strict aliasing problem in
60 g_static_mutex_get_mutex().
61 * glib/gthread.h: ...and in
62 g_static_mutex_get_mutex_impl_shortcut().
64 * glib/gdatasetprivate.h: Add a cast to silence compiler
65 warnings. (#321978, Andrew Paprocki)
67 Partial fix for bug #32937, Bogdan Nicula.
69 * configure.in: Check for malloc.h
70 * glib/gslice.c: Don't include malloc.h unconditionally.
72 2005-12-27 Manish Singh <yosh@gimp.org>
74 * tests/run-collate-tests.sh: export LC_COLLATE so it takes.
75 (#324950, Dan Yefimov)
77 2005-12-26 Matthias Clasen <mclasen@redhat.com>
79 * glib/gslice.c: On Win32, include process.h (#325015, Kazuki
82 2005-12-25 Matthias Clasen <mclasen@redhat.com>
84 * AUTHORS: Update my email
86 * tests/slice-test.c: Fix C99isms. (#324950, Dan Yefimov)
88 2005-12-21 Matthias Clasen <mclasen@redhat.com>
90 * README.in: Add some notes on when 'make check' may fail.
92 2005-12-20 Matthias Clasen <mclasen@redhat.com>
96 * glib/gthreadpool.c (g_thread_pool_set_sort_function): New function
97 to sort tasks pushed into a threadpool. (#324479, Martyn Russell)
99 * tests/threadpool-test.c: Test this.
101 Tue Dec 20 18:14:14 2005 Tim Janik <timj@imendio.com>
103 * glib/gslice.[hc]: added mem_error() and mem_assert() to test and
104 handle errors without depending on gmessage.c which might not be
105 setup when the error occours.
106 removed G_SLICE_CONFIG_ALWAYS_FREE config option, fixed the code so
107 always freeing can be achieved by adjusting the working set time to
108 0 with G_SLICE_CONFIG_WORKING_SET_MSECS.
109 added G_SLICE_CONFIG_COLOR_INCREMENT to test different color increments
110 (mainly 0 and 1). reduced the minimum block size to 128 bytes, to
111 minimize wastage if small amounts of differently sized structrues are
112 allocated, this does come at a performance cost of roughly 5% though.
113 fixed up block alignment calculation, so it works for varying
114 block sizes. only use strerror() not g_strerror() since the latter
115 depends on working GQuark and GSlice.
116 mem_error(): implemented in terms of fprintf and vfprintf.
118 * tests/slice-color.c: new program to test cache colorization effects.
120 * tests/slice-test.c: trade G_SLICE_CONFIG_ALWAYS_FREE for 0 duration
121 G_SLICE_CONFIG_WORKING_SET_MSECS.
123 2005-12-17 Matthias Clasen <mclasen@redhat.com>
125 * glib/goption.c (parse_short_option): Set an error in all
126 failure cases. (#324332, Tim-Philipp Müller)
128 2005-12-17 Sebastian Wilhelmi <seppi@seppi.de>
130 * glib/gatomic.c: Fix memory barrier position in g_atomic_int_get
131 and g_atomic_pointer_get. Add g_atomic_int_set and
132 g_atomic_pointer_set implementations for the !DEFINE_WITH_MUTEXES &&
133 G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case, as well as defining them
134 as functions (additionally to the macros in the header) for the
135 !G_ATOMIC_OP_MEMORY_BARRIER_NEEDED case.
137 2005-12-16 Matthias Clasen <mclasen@redhat.com>
139 * glib/gmem.c (g_allocator_new): Don't return a pointer to
140 a const struct, since apps expect to be able to modify it.
141 (#324179, J. Ali Harlow)
143 Tue Dec 13 10:13:32 2005 Tim Janik <timj@imendio.com>
145 * glib/gatomic.h: added g_atomic_pointer_set() and g_atomic_int_set()
146 for systems where the initialization of atomic variables requires a
147 write memory barrier.
149 2005-12-09 Matthias Clasen <mclasen@redhat.com>
151 * configure.in: Bump version
153 * === Released 2.9.1 ===
155 2005-12-09 Alexander Larsson <alexl@redhat.com>
159 Use G_GNUC_WARN_UNUSED_RESULT on list functions that return
162 2005-12-08 Matthias Clasen <mclasen@redhat.com>
166 2005-12-07 Martyn Russell <martyn@imendio.com>
168 * glib/gasyncqueue.c:
169 - Call g_queue_insert_sorted() instead of duplicating the code.
170 - Call g_queue_sort() instead of duplicating the code.
171 - Invert sort function results to make sure the same sort function
172 gives the same results across glist, gslist, gqueue and
175 * tests/asyncqueue-test.c:
176 - Updated the sort function to reflect the example in the
177 documentation for gasyncqueue.c.
179 2005-12-07 Martyn Russell <martyn@imendio.com>
181 * docs/reference/glib/glib-sections.txt:
182 * docs/reference/glib/tmpl/linked_lists_double.sgml:
183 * docs/reference/glib/tmpl/linked_lists_single.sgml:
186 - Added g_list_insert_sorted_with_data () and
187 g_slist_insert_sorted_with_data ().
188 - Removed the extra check in g_list_sort() and g_slist_sort() for
189 GCompareDataFunc vs. GCompareFunc.
191 2005-12-07 Tor Lillqvist <tml@novell.com>
193 * glib-gettextize.in: Look up prefix at run-time on Win32,
194 assuming the standard directory structure with glib-gettextize in
197 2005-12-06 Matthias Clasen <mclasen@redhat.com>
199 * glib/gmem.h: Exempt GTK+ from the mem chunk deprecation, since
200 we need GTK+ 2.8 to compile against GLib 2.10.
202 2005-12-05 Manish Singh <yosh@gimp.org>
204 * tests/run-collate-test.sh: set LC_COLLATE, not LANG, to be sure
205 to override any user settings.
207 2005-12-05 Matthias Clasen <mclasen@redhat.com>
209 * configure.in: Revert an accidential commit
211 2005-12-05 Matthias Clasen <mclasen@redhat.com>
213 * glib/gutf8.c: Documentation updates. (#323291, Morten Welinder)
215 * tests/Makefile.am (TESTS_ENVIRONMENT):
216 * tests/gobject/Makefile.am (TESTS_ENVIRONMENT): Set
217 MALLOC_CHECK_ and MALLOC_PERTURB_
219 * tests/run-collate-tests.sh: Run the collation tests explicitly
220 in en_US locale. (#320463)
222 * glib/gmem.h: Really deprecate GMemChunk.
229 * glib/gstrfuncs.c: Add versioned deprecation docs.
231 2005-12-05 Martyn Russell <martyn@imendio.com>
233 * docs/reference/glib/glib-sections.txt:
234 * glib/gasyncqueue.[ch]:
235 - Added support for sorting async queues by with _push_sorted(),
236 _push_sorted_unlocked(), _sort() and _sort_unlocked() (#323047).
239 * tests/asyncqueue-test.c:
240 - Added test case for gasyncqueue.c
242 Mon Dec 5 15:53:20 2005 Tim Janik <timj@imendio.com>
244 * glib/gslice.c: implement chain walking for arbitrary ->next pointer
245 offsets in g_slice_free_chain_with_offset() based on a patch by behdad
246 in bug 323178. moved time consuming logic from g_slice_free() out of
247 the inner loop, so g_slice_free_chain_with_offset() provides a real
248 performance benefit over g_slice_free1() now.
250 * glib/gslice.h: renamed g_slice_free_chain() to
251 g_slice_free_chain_with_offset(). implemented g_slice_free_chain() as
252 a type-safe macro as suggested in bug 323178.
253 simplified the macro implementation of g_slice_free() and implemented
254 it in a type safe manner for all compliers as suggested by Morten
255 Welinder <mortenw@gnome.org>.
260 * glib/glib.symbols: s/g_slice_free_chain/g_slice_free_chain_with_offset/
262 2005-12-05 Matthias Clasen <mclasen@redhat.com>
264 * glib/gasyncqueue.c: Add some docs.
266 * tests/libmoduletestplugin_a.c: Fix compiler warnings.
268 * glib/gatomic.c: In the ia64 implementation, use
269 __sync builtin without _si or _di suffix. (#321229,
270 Stanislav Brabec, patch by Andreas Schwab)
272 2005-12-04 Behdad Esfahbod <behdad@gnome.org>
274 * glib/gslice.h: Remove comma at the end of enum.
276 2005-12-04 Matthias Clasen <mclasen@redhat.com>
278 Handle multiple user names with the same UID better.
279 (#319535, Laszlo Peter)
281 * glib/gutils.c (g_get_any_init_do): When determining user
282 data, first look up $LOGNAME. If the UID doesn't match
283 getuid(), fall back to the current behaviour of looking
284 up the user data based on getuid().
286 2005-12-04 Matthias Clasen <mclasen@redhat.com>
288 * glib/gmacros.h (G_GNUC_WARN_UNUSED_RESULT): Add a macro
289 to make gcc warn if a function result is ignored. (#145466,
290 Arjan van de Ven, Alex Larsson)
292 * glib/gmem.h: Add the new attribute to g_realloc and
295 2005-12-03 Matthias Clasen <mclasen@redhat.com>
298 * glib/gcache.h: Deprecate g_cache_value_foreach. (#322956,
301 * glib/gmappedfile.c: Make mapping of empty files
304 * glib/gfileutils.c: Don't fork a new process just to
305 fix the permissions of the created temp file. (#321318,
306 Alexis S. L. Carvalho)
308 2005-12-02 Matthias Clasen <mclasen@redhat.com>
310 * README.in: Add a note about Solaris threads.
313 * configure.in: Remove support for Solaris threads.
314 (#136971, Sebastian Wilhelmi, patch by Andrew Paprocki)
316 2005-12-02 Matthias Clasen <mclasen@redhat.com>
318 * configure.in: Define G_GUINT64_CONSTANT in analogy to
319 G_GINT64_CONSTANT. (#322568, Andrew Paprocki)
321 2005-12-02 Matthias Clasen <mclasen@redhat.com>
323 * glib/gslice.c: Win32 portability fixes and C99ism removal,
324 pointed out by Kazuki Iwamoto. (#323052)
326 2005-12-02 Matthias Clasen <mclasen@redhat.com>
328 * glib/gatomic.c (g_atomic_int_add): Add a missing volatile for
329 the IA64 implementation.
331 Fri Dec 2 16:18:09 2005 Tim Janik <timj@imendio.com>
333 * glib/gslice.c: some naming and type size fixups.
335 Fri Dec 2 13:08:58 2005 Tim Janik <timj@gtk.org>
337 * glib/gslice.h (g_slice_free): reworked GCC-specific type-safe macro
338 variant into something less verbose, but digestible for gcc-3.4.
340 Fri Dec 2 10:55:07 2005 Tim Janik <timj@imendio.com>
342 * tests/slice-test.c: extended to perform the benchmarking on the old
343 memchunk code if 'O' is selected.
345 * tests/memchunks.c: new file which contains the old GLib mem chunks
346 implementation with prefix old_mem_chunk_.
348 * tests/Makefile.am: added memchunks.c
350 Fri Dec 2 00:16:59 2005 Tim Janik <timj@imendio.com>
352 * glib/gslice.c: improved gettimeofday() timer resolution by maintaining
353 acceptable load distribution of the syscall.
354 (allocator_get_magazine_threshold): reduce minimum/single-thread magazine
355 sizes to approximately page_size/2.5 to avoid excessive startup allocations.
357 2005-12-01 Matthias Clasen <mclasen@redhat.com>
359 * tests/gobject/ifacecheck.c (test_iface_base_init): Don't
360 double-free base interface members.
362 * tests/Makefile.am: Remove duplicate variable.
364 Thu Dec 1 17:32:46 2005 Tim Janik <timj@imendio.com>
366 * glib/gslice.[hc]: new slice allocator implementation.
368 * tests/slice-test.c: added random slice allocation test.
370 * glib/gthread.[hc]: removed newly added private thread mem API.
372 * glib/gthreadinit.h:
375 * glib/gmem.c: divided glib threading initialisation into three phases,
376 initialisation where private keys and messaging are not available (only
377 needed by gmem.c), initialisation without messaging but private keys
378 available (gslice.c, gmessage.c), and full fledged initialisers that
379 server the rest of glib. initialisation functions got renamed to reflect
380 the limitations of their corresponding phases.
382 * glib/gmem.c: removed memchunk code, defer allocations to
385 * glib/gmem.[hc]: removed g_slice_* skeletons.
387 * glib/glib.symbols: added g_slice_* symbols.
389 * configure.in: check for availability of posix_memalign(3), memalign(3)
392 * glib/Makefile.am: added gslice.[hc].
394 2005-12-01 Tor Lillqvist <tml@novell.com>
396 * glib/gstdio.c (g_stat): In the Win32 implementation, strip
397 trailing slash(es) for non-root folders. stat() fails if non-root
398 folders are specified with trailing slashes. It's too much hassle
399 to demand that callers strip such slashes themselves, especially
400 as it is easy to get it wrong and strip the slash of a root
402 (g_rename): On NT-based Windows, use MoveFileEx() with
403 MOVEFILE_REPLACE_EXISTING to better match Unix behaviour.
405 2005-11-28 Matthias Clasen <mclasen@redhat.com>
407 Fix G_STMT_START / G_STMT_END on Solaris. (#321972,
410 * configure.in: Check whether do { } while (0) works.
412 * glib/gmacros.h: Use do { } while (0) for G_STMT_START /
413 G_STMT_END if it works.
415 2005-11-28 Matthias Clasen <mclasen@redhat.com>
417 * glib/gthread.c (g_static_rw_lock_wait, g_static_rw_lock_signal):
418 * glib/gnode.c (g_node_depth_traverse_level):
419 * glib/gmem.c (g_allocator_new):
420 * glib/ghash.c (g_hash_table_unref, g_hash_table_destroy)
421 (g_hash_table_foreach_remove_or_steal): Silence compiler
424 2005-11-27 Matthias Clasen <mclasen@redhat.com>
426 * glib/gunicollate.c (g_utf8_collate_key_for_filename):
427 Don't read beyond len. (#322520, Christian Persch)
429 Wed Nov 23 17:34:01 2005 Tim Janik <timj@imendio.com>
431 * glib/gdataset.c: access datalist flags via atomic pointer access
432 functions, instead of acquiring the dataset lock. this is faster and
433 also matches the atomic pointer readouts in gdatalistprivate.h and
434 direct pointer modifications required by gobject.c.
436 Wed Nov 23 13:35:31 2005 Tim Janik <timj@gtk.org>
438 * glib/gdataset.c: streamlined initialization code somewhat,
439 removed GData node cache.
441 2005-11-22 Matthias Clasen <mclasen@redhat.com>
443 * glib/ghash.c (g_hash_table_ref):
444 (g_hash_table_unref): Mark these as new API.
446 Avoid double locking in g_intern_string (#322133,
449 * glib/gdataset.c (g_quark_from_string_internal): New
450 internal function which factors out common parts of
451 g_quark_from[_static]_string.
452 (g_quark_from_string, g_quark_from_static_string):
453 Use g_quark_from_string_internal.
454 (g_intern_string, g_intern_static_string): Use
455 g_quark_from_string_internal, and only take the
457 (g_quark_new): Don't store the strings shifted by -1
458 in the g_quarks array.
459 (g_quark_to_string): Adapt to the previous change.
461 Tue Nov 22 14:04:26 2005 Tim Janik <timj@imendio.com>
465 g_hash_table_new_full(): create hash tables with a ref count of 1.
466 g_hash_table_ref(): atomically ref_count+=1
467 g_hash_table_unref(): atomically ref_count-=1, destroys hash table
468 when refcount reaches 0.
469 g_hash_table_destroy(): just destroy keys and values, unref by 1.
470 g_hash_table_insert():
471 g_hash_table_replace(): assert ref_count>0.
474 * glib/gatomic.c: added 'volatile' qualifier to all atomic pointer and
477 2005-11-20 Behdad Esfahbod <behdad@gnome.org>
479 * glib/guniprop.c (g_unichar_get_mirror_char): Remove unused
480 variables. (#321984, Andrew Paprocki)
482 2005-11-18 Matthias Clasen <mclasen@redhat.com>
484 * configure.in: Bump version
486 2005-11-17 Matthias Clasen <mclasen@redhat.com>
488 * === Released 2.9.0 ===
490 * glib/glib.symbols: Updates
492 * glib/gutf8.c (g_utf8_offset_to_pointer): Add some performance
493 hints to the docs. (Paolo Borelli)
497 * glib/gthread.c (g_thread_foreach): Mark as new api.
499 * README.in: Updates.
503 * glib/gdate.c (g_date_set_time_t):
504 (g_date_set_time_val): Add functions to set a date from
505 a time_t and from a GTimeVal.
506 (g_date_set_time): Deprecate. (#314067, Roger Leigh)
509 * tests/utf8-pointer.c: Unit tests for g_utf8_pointer_to_offset
510 and g_utf8_offset_to_pointer.
512 * glib/gutf8.c (g_utf8_pointer_to_offset)
513 (g_utf8_offset_to_pointer): Handle negative offsets, and use
514 "stutter stepping" for going backwards. (#320638, Larry
519 * glib/gthread.c: const correctness fixes, found
520 by Arjan van de Ven and gcc.
522 2005-11-16 Behdad Esfahbod <behdad@gnome.org>
524 * docs/reference/glib/tmpl/unicode.sgml: Correct typo on mentioning
525 @G_UNICODE_BREAK_UNKNOWN that should be #G_UNICODE_BREAK_UNKNOWN.
527 2005-11-09 Behdad Esfahbod <behdad@gnome.org>
529 * glib/guniprop.c: Use bit hacks instead when checking a general
530 category value against multiple values.
532 * glib/gutf8.c: Change ISO10646 to Unicode in docs.
534 2005-11-10 Simos Xenitellis <simos@gnome.org>
536 * configure.in: Added tt (Tatar) to ALL_LINGUAS.
538 2005-11-05 Tor Lillqvist <tml@novell.com>
540 * glib/giowin32.c (g_io_win32_set_flags): Don't call g_warning().
541 (#320688) Instead set error and return failure.
543 2005-11-07 Matthias Clasen <mclasen@redhat.com>
545 * glib/gmain.c (g_main_dispatch): Don't call cb_funcs->unref
546 while holding the context lock. (#320886, Andy Wingo)
548 2005-11-07 Billy Biggs <vektor@dumbterm.net>
550 * tests/refcount/Makefile.am: Add a missing reference to libglib
551 in the LDADD for this test.
553 2005-11-07 Matthias Clasen <mclasen@redhat.com>
555 * glib/gmem.c (g_allocator_new): We need to set n_preallocs to a
556 nonzero value, otherwise GTK+ 2.8 breaks when compiled against
557 GLib 2.9. (#320755, Luca Ferretti)
559 2005-11-04 Matthias Clasen <mclasen@redhat.com>
561 * glib/Makefile.am: Apply a patch from Behdad Esfahbod to
562 use a faster lookup table for g_unichar_get_mirror_char().
563 * glib/gmirroringtable.h: The new table.
564 * glib/gunichartables.h: Remove bidi_mirroring_table.
565 * glib/gen-unicode-tables.pl: Don't generate the mirroring
567 * glib/glib-mirroring-tab/*: Sources for the program
568 which generated gmirroringtable.h.
570 * glib/glist.c: Avoid some code duplication.
572 * glib/gscanner.h: Include gdataset.h. (#320322)
574 * glib/gdebug.h: add new GDebugFlag for fatal_criticals
575 * glib/gmessages.c: (_g_debug_init): handle G_DEBUG=fatal_criticals,
576 to help find critical warnings in applications. (#320017,
579 2005-11-02 Tor Lillqvist <tml@novell.com>
581 * glib/glib.symbols: Remove large amount of trailing whitespace
582 from one line. Remove #ifdef G_UNDEFINED from around
583 g_hash_table_new and g_hash_table_new_full symbols, what was the
584 purpose of that? Add the g_slice functions.
586 Tue Nov 1 16:24:20 2005 Tim Janik <timj@imendio.com>
588 * glib/gmem.[hc]: prepared deprecation of GMemChunk and GAllocator.
589 added g_slice_*() API to allocate and cache small bits of memory.
590 an actuall allocator implementation for g_slice_*() is still pending.
592 * glib/gthread.[hc]: changes from a patch by Matthias Clasen.
593 changed GRealThread list to use in-structure *next; fields instead
594 of GSList, in order for thread iteration to not depenend on g_slice_*()
596 _g_thread_mem_private_get():
597 _g_thread_mem_private_set(): added accessors for private memory,
598 needed because the ordinary GPrivate implementation relies on GArray
599 and GSList and therefore indirectly on working g_slice_*() allocations.
602 g_thread_foreach(): new public API function to loop over all existing threads.
616 removed GAllocator and free list usages and accompanying locks.
617 use g_slice_*() API to allocate and cache small bits of memory.
619 * glib/ghook.h: removed GMemChunk field from public API.
622 * glib/glist.h: deprecate allocator API, provide _free1() for consistency.
624 * glib/gnode.h: deprecate allocator API.
626 * glib/gmain.c: reordered GPollRec fields so g_slice_free_chain() can
627 be used for poll rec lists.
629 * glib/grel.c: removed mem chunk usage, and allocated tuples via g_slice_*().
630 g_relation_destroy(): free all tuples from the all_tuples hash table,
631 this effectively maintains the life time track keeping of tuples.
632 g_relation_delete_tuple(): free tuples which are removed from the
633 all_tuples hash table. this fixes a temporary leak that was present
634 in the memchunk code until the destruction of the relation.
636 2005-10-29 Matthias Clasen <mclasen@redhat.com>
638 * tests/convert-test.c: Add some tests for conversions between
639 UTF-8, UCS-4 and UTF-16.
641 * glib/gutf8.c (g_utf8_to_ucs4, g_utf8_to_utf16): Fix handling
642 of len == -1, noticed by Morten Welinder.
644 2005-10-27 Erdal Ronahi <erdal.ronahi@gmail.com>
646 * configure.in: Added ku (Kurdish) to ALL_LINGUAS
648 2005-10-26 Matthias Clasen <mclasen@redhat.com>
650 * glib/gutf8.c (g_ucs4_to_utf8): Don't set items_read twice
651 in the error case, and add some documentation. (#319806, Morten
654 2005-10-19 Manish Singh <yosh@gimp.org>
656 * configure.in: Use AC_CHECK_FUNCS for _NSGetEnviron, to get the
657 config.h symbol automatically. Fixes bug #313731.
659 2005-10-19 Tor Lillqvist <tml@novell.com>
661 * glib/gwin32.c (g_win32_get_package_installation_directory):
662 Return a g_strdup()ed copy of the value stored in the hash table,
663 so that it can be g_free()d without leaving a dangling pointer in
664 the hash table. (#319232)
666 2005-10-06 Matthias Clasen <mclasen@redhat.com>
668 * glib/gunicollate.c (g_utf8_collate_key_for_filename): Handle
669 all-zero sequences correctly. (#317930, Sebastien Bacher)
671 2005-10-05 Matthias Clasen <mclasen@redhat.com>
673 * glib/gmarkup.c (g_markup_escape_text): Doc addition
675 * Makefile.am: Add Collation tests.
677 * tests/collate/*: Inputs and expected outputs for collation tests.
679 * tests/run-collate-tests.sh: Script to run collation tests.
681 * tests/unicode-collate.c (main): Rework slightly to make
682 it usable in unit tests. Also test g_utf8_collate_key_for_filename().
684 2005-10-01 Behdad Esfahbod <behdad@gnome.org>
686 * docs/reference/glib/tmpl/unicode.sgml:
687 * glib/gen-unicode-tables.pl:
689 * glib/gunichartables.h:
691 * tests/casefold.txt:
692 * tests/casemap.txt: Updated to Unicode 4.1. There are five new
693 GUnicodeBreakType types. That may break some applications, like
696 2005-09-26 Matthias Clasen <mclasen@redhat.com>
698 * glib/gstrfuncs.c (g_ascii_strtoull): Add details to the
699 docs. (#314393, Matthew F. Barnes)
702 * glib/gprintf.h: Remove g_snprintf() and g_vsnprintf(), since
703 they are already declared in glib.h. This doesn't break documented
704 use of gprintf.h, but should probably be pointed out in the
705 release notes for 2.10. (#314232, Behdad Esfahbod)
707 Tue Sep 20 13:16:04 2005 Tim Janik <timj@imendio.com>
709 * glib/gpattern.c (g_pattern_ph_match): applied significant recursion
710 complexity optimization, based on a patch from Matthias Clasen.
712 * tests/patterntest.c: more tests, mostly from matthias.
714 2005-09-20 Matthias Clasen <mclasen@redhat.com>
716 * glib/gqueue.c (g_queue_insert_sorted): Correct the docs.
717 (#316703, Mark Drago)
719 Mon Sep 19 17:23:23 2005 Tim Janik <timj@imendio.com>
721 * glib/gpattern.c: applied a patch from matthias which checks on the
722 upper bound of GPatternSpec length to optimize matches.
725 * tests/patterntest.c: added more match cases.
727 2005-09-16 Tor Lillqvist <tml@novell.com>
729 * glib/gstrfuncs.c (g_ascii_strcasecmp, g_ascii_strncasecmp): Add
730 warning to doc comment that these functions should not be used on
731 encodings like CP932.
733 2005-09-14 Matthias Clasen <mclasen@redhat.com>
735 * tests/keyfile-test.c: Add a test for grup names of length 1.
737 * glib/gkeyfile.c (g_key_file_line_is_group): Accept group names
738 of length 1. (#316309)
740 2005-09-12 Matthias Clasen <mclasen@redhat.com>
742 * glib/gmarkup.c (g_markup_escape_text): Clarify docs.
744 2005-09-11 Sebastian Wilhelmi <seppi@seppi.de>
746 * tests/refcount/Makefile.am (INCLUDES): Link the the refcount
747 tests to the system thread library $(G_THREAD_LIBS). Fixes #313744
750 2005-09-11 Kjartan Maraas <kmaraas@gnome.org>
752 * glib/gmain.c: (g_child_watch_prepare), (g_child_watch_check),
753 (child_watch_helper_thread): Remove some dead code. Closes
756 2005-09-07 Tor Lillqvist <tml@novell.com>
758 * glib/Makefile.am: Create also a console version of the
759 gspawn-win32-helper program, gspawn-win32-helper-console.exe.
760 It's otherwise identical to gspawn-win32-helper.exe, except marked
761 as a console application (linked without the -mwindows option).
763 * glib/gspawn-win32.c (do_spawn_directly, do_spawn_with_pipes):
764 Drop the dont_wait parameter. Its truth value correlated 100% with
765 the NULLness of the exit_status parameter anyway, so it's enough
766 to check whether exit_status is NULL. Invert the sense of the
767 dont_return_handle parameter and rename it to do_return_handle, to
768 make the code easier to read by avoiding double negations.
770 (g_spawn_sync_utf8, g_spawn_async_with_pipes_utf8): Modify calls
771 to do_spawn_with_pipes() accordingly.
773 (do_spawn_with_pipes): If we have a console, use the console
774 version of the helper program, otherwise use the GUI one. This
775 avoids extra console windows opening up in some situations. (In
776 case a console application uses the GUI gspawn-win32-helper.exe to
777 spawn another console application we would get a separate console
778 for the spawned console application).
780 * glib-zip.in: Distribute also gspawn-win32-helper-console.exe.
782 2005-09-05 Matthias Clasen <mclasen@redhat.com>
784 * glib/gmappedfile.c (g_mapped_file_new): Report an error
785 if the file is too large. (#315275, Kjartan Maraas)
787 * glib/gkeyfile.c (g_key_file_load_from_fd): The return value
788 of read() is signed. (#315273, Kjartan Maraas)
790 2005-08-31 Tor Lillqvist <tml@novell.com>
792 * glib/gutils.h: Wrapping atexit() is a bad idea on Windows, where
793 the EXE and each DLL have their own atexit function chains.
795 #define g_atexit as atexit instead. This means it has a
796 better chance of doing what the caller wants. For instance,
797 gtkhtml calls g_atexit() registering a function in gtkhtml
798 itself. This caused a crash when g_atexit() was implemented as a
799 function in the GLib DLL. The gtkhtml DLL was already unloaded by
800 the time the GLib DLL got unloaded.
802 * glib/gutils.c: #undef the #define mentioned above, to also get a
803 real g_atexit() into the DLL for backward compatibility. Document
804 the Windows behaviour of g_atexit(), and document the varying ways
805 atexit() can behave in the context of dynamically loaded modules
808 2005-08-31 Matthias Clasen <mclasen@redhat.com>
812 * glib/gdataset.c: Add string interning functions.
814 2005-08-28 Matthias Clasen <mclasen@redhat.com>
816 * glib/giochannel.c: Unify some near-duplicate strings. (#314654,
819 2005-08-26 Matthias Clasen <mclasen@redhat.com>
821 * configure.in: Bump version to 2.9.0
823 2005-08-25 Tor Lillqvist <tml@novell.com>
825 Make also the g_spawn*() functions take parameters in the GLib
826 file name encoding, i.e. UTF-8, on Windows. Has no impact on Unix
827 API or ABI. Like the other GLib API that was earlier changed to
828 use UTF-8 on Windows, the names of the functions that take UTF-8
829 have _utf8 suffixes added by using preprocessor macros in the
830 header file. The old names are kept for functions with the old
831 behaviour, taking parameters in the system codepage, for DLL ABI
834 * glib/gspawn.h: On Win32 add the suffix _utf8 to the names of the
835 g_spawn*() functions.
837 * glib/gspawn-win32.c: Use wide-char API on NT-based
838 Windows. Convert parameters from UTF-8 to wide chars (NT) or
839 system codepage (Win9x) and call the C library _wspawn*() or
840 spawn*() functions respectvely. Add DLL ABI stability versions
841 that take parameters in the system codepage.
843 * glib/gspawn-win32-helper.c: On NT-based Windows use the
844 wide-char versions of argv and envp, and use wide-char API to
845 change directory and spawn the program to run. Remove the verbose
846 debugging output, it was too complex to modify for the wide-char
847 features. (Just add temporary debugging printouts if needed, no
848 need to have them permanently in the source.)
850 * glib/gspawn.c: Corresponding documentation updates.
852 * glib/glib.symbols: Corresponding changes: Mark the ABI stability
853 symbols as PRIVATE, add the new _utf8-suffixed ones.
855 2005-08-24 Stepan Kasal <kasal@ucw.cz>
857 * glib/gtypes.h (G_MININT64): Cast the constant to gint64; it is
858 guint64 otherwise and that can produce warnings about comparison
859 between signed and unsigned.
861 2005-08-23 Matthias Clasen <mclasen@redhat.com>
863 * glib/gutils.c: Fix the crt_externs.h include.
865 2005-08-23 Stepan Kasal <kasal@ucw.cz>
867 * NEWS: Fix spelling of my first name.
869 2005-08-23 Matthias Clasen <mclasen@redhat.com>
873 * === Released 2.8.1 ===
877 2005-08-20 Hans Breuer <hans@breuer.org>
879 * glib/makefile.msc.in : link with ws2_32.lib
881 2005-08-18 Tor Lillqvist <tml@novell.com>
883 * configure.in: Check for <sys/wait.h>
885 * glib/gbacktrace.c: Include <sys/wait.h> on if HAVE_SYS_WAIT_H.
887 2005-08-18 Ross Burton <ross@burtonini.com>
890 Optimise single-character insertions.
895 * tests/string-test.c:
896 Add tests for new optimisation, and fix a leak.
898 2005-08-17 Matthias Clasen <mclasen@redhat.com>
900 * configure.in: Check for crt_externs.h and _NSGetEnviron.
902 * glib/gutils.c: On Darwin, include crt-externs.h and
903 define environ using _NSGetEnviron(). (#313731)
905 2005-08-16 Stepan Kasal <kasal@ucw.cz>
907 * glib/gutils.c (g_get_any_init): Move the body of the big if...
908 (g_get_any_init_do): ... to this new function.
909 (g_get_any_init): Declare as inline.
910 (g_get_any_init_locked): New inline function, does the locking.
911 Make use of these two throughout the code.
913 2005-08-15 Matthias Clasen <mclasen@redhat.com>
915 * glib/gbacktrace.c (g_on_error_stack_trace): Wait for
916 the child process and then simply return. This makes
917 The "S" option work as documented in g_on_error_query().
918 (#313125, Matthew F. Barnes)
920 * glib/gunicode.h: Update the link to Unicode category
921 values. (#313369, Behnam Esfahbod)
923 * glib/gqueue.c (g_queue_find_custom): Clarify docs
924 a little. (#311727, Tristan van Berkom)
926 * glib/abicheck.sh, gobject/abicheck.sh: Make the
927 check work on ia64 too, where some symbols ended up
928 in yet another section.