Default page

Edit File: ChangeLog

2017-10-25  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* NEWS:
	=== Version 0.1.27 ===

2017-10-25  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-geo.txt:
	* data/tdict-history.txt:
	* data/tdict-ict.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	  - Add words.

* data/tdict-proper.txt:
	  - Correct order for some entries.

2017-08-07  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Strip prefix for C source files in doc

* doc/Doxyfile.in:
	  - Append "@top_srcdir@/src" and "@top_srcdir@/tests"
	    to STRIP_FROM_PATH, so that references to source directories
	    in the generated document do not capture full build path.

Defect caught by Debian Reproducible

2017-06-29  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Remove duplicated dict entry.

* data/tdict-common.txt:
	  - Remove 'ดัมพ์' which is already in tdict-std.
	* data/tdict-common.txt:
	* data/tdict-spell.txt:
	  - Move 'ดั๊มพ์' from tdict-common to tdict-spell.

2017-06-29  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	  - Add words.

2016-12-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* NEWS:
	=== Version 0.1.26 ===

2016-12-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Include git-version-gen in tarball

* Makefile.am:
	  - Add build-aux/git-version-gen to EXTRA_DIST.

2016-12-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Move word.

* data/tdict-common.txt, data/tdict-ict.txt:
	  - Move 'แอดมิน' from tdict-common to tdict-ict.

2016-12-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	  - Add words.

2016-07-09  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Move some uncommon compound names out of tdict-std dictionary.

* data/tdict-std.txt, data/tdict-std-compound.txt:
	  - Move 'กระดองหาย', 'กระดาดขาว', 'นางรำ', 'นางหงส์'
	    to tdict-std-compound.

2016-07-08  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Move compounds out of tdict-std dictionary.

* data/tdict-std.txt:
	  - Remove long compound 'ตะลึงพรึงเพริด'.
	  - Keep only 'สะพรึง' instead of 'สะพรึงกลัว'.
	* data/tdict-std.txt, data/tdict-std-compound.txt:
	  - Move compounds 'ตะลุ่มนก', 'ตากวาง', 'ตานนกกด' to tdict-std-compound.

2016-07-08  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Match git tag with version pattern.

* build-aux/git-version-gen:
	  - Apply --match pattern to git-describe to prevent confusion
	    with other tags.
	  - Drop unused LF var.

2016-07-07  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Use versioning based on Git snapshot.

* Makefile.am:
	  - Add dist-hook to generate VERSION file on tarball generation.
	* +build-aux/git-version-gen:
	  - Add script to generate version based on 'git describe'
	    if in git tree, or using VERSION file if in release tarball.
	* configure.ac:
	  - Call git-version-gen to get package version.

2016-06-28  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac, NEWS:
	=== Version 0.1.25 ===

2016-06-28  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	- Add words.

2016-06-27  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Document about multi-thread use.

* src/thbrk/thbrk.c (th_brk_new):
	  - Add documentation about deleting after use, and about
	    creating/destroying word breaker instance in critical sections.

2016-06-26  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Use 'wc' prefix instead of suffix for wide-char word break APIs.

To be consistent with other modules, wide-char APIs should be
	prefixed, not suffixed with 'wc'.

* include/thai/thwbrk.h:
	* src/libthai.c:
	* src/libthai.def:
	* src/libthai.map:
	* src/thwbrk/thwbrk.c:
	* tests/test_thwbrk.c:
	  - Rename th_brk_find_breaks_wc() -> th_brk_wc_find_breaks().
	  - Rename th_brk_insert_breaks_wc() -> th_brk_wc_insert_breaks().

2016-06-25  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Rename the new APIs to be more meaningful.

The methods for ThBrk I proposed to Mark Brown in our discussion
	were too confusing. Before they get public, let's pick more
	meaningful names instead:
	  - th_brk_brk()       -> th_brk_find_breaks()
	  - th_brk_brk_line()  -> th_brk_insert_breaks()
	  - th_brk_wbrk()      -> th_brk_find_breaks_wc()
	  - th_brk_wbrk_line() -> th_brk_insert_breaks_wc()

* include/thai/thbrk.h:
	* include/thai/thwbrk.h:
	* src/libthai.c:
	* src/libthai.def:
	* src/libthai.map:
	* src/thbrk/thbrk.c:
	* src/thwbrk/thwbrk.c:
	* tests/test_thbrk.c:
	* tests/test_thwbrk.c:
	  - Rename functions as listed above.
	  - Rename the 'n' argument in the functions to indicate
	    whose size it describes.

2016-06-24  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Add function deprecation warnings.

* include/thai/thailib.h:
	  - Add TH_DEPRECATED and TH_DEPRECATED_FOR() macros,
	    taken from GLib.
	* include/thai/thbrk.h (th_brk, th_brk_line):
	* include/thai/thwbrk.h (th_wbrk, th_wbrk_line):
	  - Apply TH_DEPRECATED_FOR() to deprecated functions,
	    with suggestions for alternatives.

2016-06-24  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update tests to test new APIs.

* tests/test_thbrk.c (main), tests/test_thwbk.c (main):
	  - Create, use, and free ThBrk instance with the new APIs
	    instead of the deprecated ones.

2016-06-23  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update documentation for API deprecation.

* src/thbrk/thbrk.c, src/thwbrk/thwbrk.c:
	  - Add info for deprecated functions.
	  - Add availability info for new functions.

* src/thinp/thinp.c:
	  - Add availability info for th_validate_leveled().

* src/libthai.c:
	  - Add missing link to th_validate_leveled().

2016-06-23  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update library versioning.

* configure.ac:
	  - Bump library versions to reflect the added API.

2016-06-22  Mark Brown  <mark@mercurylang.org>

Implement a new thread-safe interface for word break.

To achieve more thread-safety without depending on mutex mechanisms,
	a new set of APIs is added so that the client can create
	a shared instance of word break engine by him/herself
	under appropriate mutex. Then, the word break functions
	can be safely called in parallel using the shared engine.

* include/thai/thbrk.h:
	* include/thai/thwbrk.h:
	* src/libthai.c:
	* src/libthai.def:
	* src/libthai.map:
	  - Add new exported APIs:
	    th_brk_new(), th_brk_delete(),
	    th_brk_brk(), th_brk_brk_line(),
	    th_brk_wbrk(), th_brk_wbrk_line().

* src/thbrk/brk-common.h, src/thbrk/brk-common.c
	  (-brk_on_unload, -brk_get_dict, +brk_load_default_dict):
	  - Remove old shared dict management. It's to be as part of
	    ThBrk implementation in ThBrk layer instead.
	  - The logic for finding and loading dictionary at
	    default paths is still retained here.

* src/thbrk/thbrk.c (th_brk_new, th_brk_delete):
	  - Implement ThBrk (de)allocation, with dictionary loading
	    at specified path or at default paths if not specified.

* src/thbrk/brk-maximal.h, src/thbrk/brk-maximal.c
	  (struct _BrkEnv, brk_env_new):
	  - Add ThBrk engine as a member of BrkEnv.
	* src/thbrk/brk-maximal.c (brk_root_pool):
	  - Access dict trie from ThBrk in BrkEnv instead of getting
	    shared dict directly.

* src/thbrk/thbrk.c
	    (th_brk -> th_brk_brk, th_brk_line -> th_brk_brk_line):
	* src/thwbrk/thwbrk.c
	    (th_wbrk -> th_brk_wbrk, th_wbrk_line -> th_brk_wbrk_line):
	  - Modify old functions to new ones by adding ThBrk* parameter.

* src/thbrk/Makefile.am, +src/thbrk/thbrk-priv.h,
	  src/thbrk/thbrk.c (brk_get_shared_brk, brk_free_shared_brk):
	  - Add functions for managing the shared engine to preserve
	    old behavior.
	* src/libthai.c (_libthai_on_unload):
	  - Call brk_free_shared_brk() on unload.
	* src/thbrk/thbrk.c (th_brk, th_brk_line):
	* src/thwbrk/thwbrk.c (th_wbrk, th_wbrk_line):
	  - Provide old APIs as wrappers to the new APIs,
	    for backward compatibility.

Merging pull request #1.

2016-05-02  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-common.txt:
	* data/tdict-spell.txt: Add words.

2016-03-28  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Do not test word breaking if dict is disabled.

* tests/Makefile.am:
	  - Only add test-thbrk.sh and test-thwbrk.sh to TESTS
	    when dictionary is enabled.

2016-03-03  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Get rid of unused variable.

* src/thrend/thrend.c (tis620_0_shape_table_):
	  - Comment out unused variable reported by GCC 6.

2016-01-28  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Make source more readable.

* src/thbrk/brk-common.c (brk_get_dict):
	  - Break assignment and comparison chain on reading
	    LIBTHAI_DICTDIR environment.

2016-01-20  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix compilation error with GCC 6

* include/thai/thctype.h (_th_bitmsk):
	  - Use unsigned int instead of unsigned short for bitmask base,
	    as GCC 6 takes ~0 as -1 for unsigned short,
	    and causes compilation error for scim-thai.

Thanks Martin Michlmayr for the report via Debian #811690
	http://bugs.debian.org/811690

2016-01-20  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Post-release version suffix added.

2015-11-22  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac:
	  - Bump library revision to reflect code changes.

* configure.ac, NEWS:
	=== Version 0.1.24 ===

2015-11-22  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2015-11-22  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix infinite loop introduced by recent optimization.

* src/thbrk/brk-maximal.c (brk_recover_try):
	  - Update 'pool' correctly when deleting matched BrkPool node.
	    brk_pool_delete_node() cannot be started at pool_tail indeed.

Thanks Zack Weinberg for the report via Debian #805703
	http://bugs.debian.org/805703

2015-11-22  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Post-release version suffix added.

2015-10-22  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac:
	  - Bump library revision to reflect code changes.

* configure.ac, NEWS:
	=== Version 0.1.23 ===

2015-10-21  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-country.txt:
	* data/tdict-ict.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt: Add words.

2015-10-16  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Share BrkEnv longer.

The code before the threadsafe commit used to initialize the
	free list earlier in th_brk() and then called brk_maximal_do()
	multiple times. This allowed the free list to be shared among
	all such calls. The threadsafe commit did it deep in
	brk_maximal_do_impl(), limiting the share within the single
	brk_maximal_do() call only. We should widen the use to be
	the same as previously done.

* src/thbrk/brk-maximal.h (+brk_env_new, +brk_env_free):
	  - Re-add initialization/quit functions in terms of BrkEnv.
	  - Add BrkEnv* arg to brk_maximal_do().
	* src/thbrk/brk-maximal.c (brk_env_init -> brk_env_new,
	  brk_env_destroy -> brk_env_free):
	  - Turn the init/destroy functions into new/free version.
	* src/thbrk/brk-maximal.c (brk_maximal_do, brk_maximal_do_impl):
	  - Add the BrkEnv* arg and use it instead of the locally created
	    one.
	* src/thbrk/thbrk.c (th_brk):
	  - Create BrkEnv object at start and free it at end.
	  - Add the env arg to brk_maximal_do() calls.

This appears to save more 2.4% of the total runtime and makes
	it on par with the code before the threadsafe commit again.

2015-10-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Call brk_pool_match() linearly.

* src/thbrk/brk-maximal.c (brk_maximal_do_impl, brk_recover_try):
	  - Start next brk_pool_match() on node next to previously matched
	    one, instead of the pool head, making the match O(n) instead of
	    O(n^2).
	  - [brk_maximal_do_impl] Also prevent using the deleted node to
	    start the next round.
	  - [brk_recover_try] Also start brk_pool_delete_node() on the
	    pool tail, making it also linear. This is not possible for
	    brk_maximal_do_impl, as the node to be deleted is not fixed.

This appears to save time by 8.85% on brk_pool_match() calls,
	lowering its bottleneck rank by 2 positions, and saves time by
	0.0388% of total run time.

2015-10-14  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Declare variable inside block.

* src/thbrk/brk-maximal.c (brk_maximal_do_impl, brk_recover_try):
	  - Move "BrkPool *match" declaration into the block it's used.

2015-10-12  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Declare const arg for brk_pool_match().

* src/thbrk/brk-maximal.c (brk_pool_match):
	  - Add const-ness to the read-only 'node' arg.

2015-10-11  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Minor optimization on brk_pool_match().

Callgrind reports brk_pool_match() among top bottlenecks.
	So, try to optimize it as we can.

* src/thbrk/brk-maximal.c (brk_pool_match):
	  - Instead of sharing the loop for two different cases,
	    check the case once and run loop for each case separately,
	    to minimize branching down to a single time.
	  - Evaluate the unchanged 'node->shot.brk_pos[node_cur_pos - 1]'
	    expression only once, instead of on every round.

This appears to save time by 9.3% for total brk_pool_match() calls,
	and by 0.067% of total run time.

2015-10-11  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix 'make check' error on long path names.

* tests/thsort.c (main):
	  - Increase file name buffers from 64 to 512, to afford long path
	    names in command-line arguments.

2015-10-10  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Replace static global free-list with local one.

The static global free-list in brk-maximal.c makes it not
	thread-safe.

* src/thbrk/brk-maximal.c:
	  - Add type BrkEnv to keep break pool free list, with
	    brk_env_init() and brk_env_destroy() methods.
	  - (brk_pool_node_new, brk_pool_node_free, brk_pool_free,
	     brk_pool_delete, brk_root_pool, brk_recover_try,
	     brk_recover):
	    Add (BrkEnv *) arg to relevant functions and use it properly.
	  - (brk_maximal_do):
	    Create and destroy BrkEnv instance for use locally.
	    Pass it to the required function calls.
	  - (brk_pool_delete -> brk_pool_delete_node):
	    Rename function to avoid confusion with brk_pool_free().

* src/thbrk/brk-maximal.c:
	* src/thbrk/brk-maximal.h:
	* src/thbrk/thbrk.c:
	  - Get rid of now-unnecessary brk_maximal_init() and
	    brk_maximal_quit().

Thanks Behdad Esfahbod for the report.

2015-09-07  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix doxygen version checking.

* configure.ac:
	  - Correctly compare doxygen versions. Simple expr comparison
	    didn't work with version 1.8.10.

Taken from Petr Gajdos <pgajdos@suse.cz>'s patch for libdatrie.

2015-06-04  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Protect against invalid array access.

* src/thbrk/thbrk.c (th_brk):
	  - Add a condition to protect against accessing pos[-1]
	    which may be caused by a failure in brk_maximal_do()
	    on the first Thai chunk, which makes cur_pos remain 0.

Thanks Behdad Esfahbod for the report.

2015-06-04  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Post-release version suffix added.

2015-05-08  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac, NEWS:
	=== Version 0.1.22 ===

2015-05-08  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Bump library revision, due to code change.

2015-05-08  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix 'make distcheck' failure.

* doc/Makefile.am:
	  - Remove doxygen db file on clean.

2015-05-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2015-05-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Use my Gmail e-mail address everywhere.

* AUTHORS:
	* include/thai/thailib.h:
	* include/thai/thbrk.h:
	* include/thai/thcell.h:
	* include/thai/thcoll.h:
	* include/thai/thctype.h:
	* include/thai/thinp.h:
	* include/thai/thrend.h:
	* include/thai/thstr.h:
	* include/thai/thwbrk.h:
	* include/thai/thwchar.h:
	* include/thai/thwcoll.h:
	* include/thai/thwctype.h:
	* include/thai/thwinp.h:
	* include/thai/thwrend.h:
	* include/thai/thwstr.h:
	* include/thai/tis.h:
	* include/thai/wtt.h:
	* man/man3/libthai.3:
	* man/man3/thctype.3:
	* man/man3/wtt.3:
	* man/template.3:
	* src/libthai.c:
	* src/thbrk/brk-common.c:
	* src/thbrk/brk-common.h:
	* src/thbrk/brk-ctype.c:
	* src/thbrk/brk-ctype.h:
	* src/thbrk/brk-maximal.c:
	* src/thbrk/brk-maximal.h:
	* src/thbrk/thbrk.c:
	* src/thcell/thcell.c:
	* src/thcoll/cweight.c:
	* src/thcoll/cweight.h:
	* src/thcoll/thcoll.c:
	* src/thctype/thctype.c:
	* src/thctype/wtt.c:
	* src/thinp/thinp.c:
	* src/thrend/thrend.c:
	* src/thstr/thstr.c:
	* src/thwbrk/thwbrk.c:
	* src/thwchar/thwchar.c:
	* src/thwctype/thwctype.c:
	* src/thwstr/thwstr.c:
	* tests/thsort.c:
	  - Replace my e-mail address with the Gmail one.

2015-05-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

More LIKELY/UNLIKELY hints.

* src/thbrk/brk-maximal.c (brk_recover):
	  - Add UNLIKELY to recovery history hits, reducing time by 0.03%

2015-05-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Split LIKELY/UNLIKELY to a separate header.

This prevents modules from depending on brk-common.h,
	which provides low-level APIs for breaker backends.

* src/thbrk/Makefile.am, src/thbrk/brk-common.h
	  +src/thbrk/thbrk-utils.h:
	  - Move LIKELY/UNLIKELY macro definitions from brk-common.h
	    to a new header thbrk-utils.h
	* src/thbrk/brk-common.c, src/thbrk/brk-maximal.c, src/thbrk/thbrk.c:
	  - Include thbrk-utils.h instead of brk-common.h for LIKELY/UNLIKELY.

2015-05-04  Theppitak Karoonboonyanan  <theppitak@gmail.com>

More LIKELY/UNLIKELY hints.

* src/thbrk/thbrk.c (th_brk_line):
	  - Add UNLIKELY() on error conditions.

Note: adding UNLIKELY() to th_brk() first if-statement makes it
	slower, which is weird. So, not added there.

2015-05-02  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Bump library revision, due to code change.

2015-05-02  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt: Add words.

2015-04-30  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Catch error on break hints allocation.

* src/thbrk/brk-maximal.c (brk_maximal_do):
	  - Do not continue if brkpos_hints allocation fails.

2015-04-30  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Do TIS-to-Unicode conversion only once on word breaking.

* src/thbrk/brk-maximal.c (brk_maximal_do):
	  - Convert input string to Unicode before passing to
	    brk_maximal_do_impl().
	* src/thbrk/brk-maximal.c (brk_maximal_do_impl, brk_recover,
	  brk_recover_try):
	  - Redeclare to accept wide char string.
	  - Do not call th_tis2uni() on every input char.

This appears to reduce word breaking time by 0.28% on test case.
	Thanks edgehogapp@gmail.com for the suggestion.
	https://groups.google.com/forum/#!topic/thai-linux-foss-devel/Be-OLMRYF7M

2015-04-29  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Bump doxygen required version.

* configure.ac:
	  - Bump doxygen required version to 1.8.8, according to recent
	    Doxyfile update.

2015-03-07  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Also check for 'trietool' (without -0.2 suffix)

* configure.ac:
	  - Check for both trietool-0.2 (for libdatrie < 0.2.9) and
	    trietool (for libdatrie >= 0.2.9) utility.

2015-03-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update Doxyfile.

* doc/Doxyfile.in:
	  - Updated for doxygen 1.8.8 with 'doxygen -u'.

2015-03-06  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix compiler warning.

* src/thbrk/brk-maximal.c (brk_recover):
	  - Initialize last_brk_pos on declaration.

2015-02-05  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Disable timestamp in Doxygen-generated doc.

* doc/Doxyfile.in:
	  - Set HTML_TIMESTAMP to NO to make the document reproducible.
	    (reported by Debian Reproducible)

2015-02-04  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Micro-optimize with likely/unlikely hints.

* src/thbrk/brk-common.h:
	  - Add LIKELY() and UNLIKELY() macros based on compiler extension.
	* src/thbrk/brk-common.c (brk_get_dict):
	* src/thbrk/brk-maximal.c (brk_maximal_do_impl, brk_root_pool,
	  brk_shot_init, brk_pool_node_new, best_brk_new):
	  - Use LIKELY() and UNLIKELY() where it is known to be so, mostly
	    for one-time initialization and failure handling.

Callgrind says it does help speed up a little bit.

2014-12-16  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Update word break dictionary.

* data/tdict-city.txt:
	* data/tdict-collection.txt:
	* data/tdict-common.txt:
	* data/tdict-country.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt: Add words.

2014-08-23  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac: Post-release version suffix added.

2014-08-21  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* configure.ac, NEWS:
	=== Version 0.1.21 ===

2014-08-19  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Be more careful about CFLAGS.

* src/Makefile.am:
	  - Add $(DATRIE_CFLAGS) to CFLAGS so that thbrk/brk-common.h
	    included from libthai.c can safely find datrie/trie.h header.

2014-08-19  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Fix build failure due to missing header.

* src/Makefile.am:
	  - Add CFLAGS so that thbrk/brk-common.h included from libthai.c
	    can find thai/thctype.h header.

2014-08-18  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2014-08-09  Theppitak Karoonboonyanan  <theppitak@gmail.com>

thbrk: Be greedy when recovering from error.

* src/thbrk/brk-maximal.c (brk_recover_try):
	  - Loop to match as many words as possible on every node picked
	    from pool, to increase the chance of early sufficient words
	    matched, and to avoid too much possibility exploration.

2014-08-09  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Minor optimization.

Callgrind says brk_recover_try() is the most frequently called.
	So, try to optimize it as we can.

* src/thbrk/brk-maximal.c (brk_recover, brk_recover_try):
	  - Only return the last break position instead of all.
	* src/thbrk/brk-maximal.c (brk_recover_try):
	  - Remove unneeded assignment on is_terminal for the case in which
	    its value is irrelevant.
	  - Remove unneeded check on is_terminal where its value is known.

2014-08-09  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* src/thbrk/brk-maximal.c:
	  - Add missing #include <limits.h>.

2014-08-05  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Code clean up.

* src/thbrk/brk-maximal.c (brk_shot_init):
	  - Split assignment and if-condition.
	* src/thbrk/brk-maximal.c (brk_pool_free_node -> brk_pool_node_free):
	  - Rename the function to obj-action form, in accordance with
	    brk_pool_node_new().
	  - Add forward declaration for it.

2014-08-04  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Refactor common thbrk code into brk-common.

This prevents mutual dependency between brk-* and thbrk.

* src/thbrk/Makefile.am:
	* src/thbrk/thbrk-private.h -> src/thbrk/brk-common.h:
	  - Rename file, as thbrk-private.h contains only functions for
	    common brk-* codes.
	  - Rename thbrk_* functions back to brk_*.
	* src/thbrk/Makefile.am:
	* src/thbrk/thbrk.c:
	* +src/thbrk/brk-common.c:
	  - Move common thbrk_* functions implementations to brk-common.c.
	  - Rename the functions according to changes in brk-common.h.
	* src/thbrk/brk-maximal.c:
	  - Update #include for thbrk-private.h to brk-common.h.
	  - Update calls to thbrk_* to brk_*.
	  - Remove #include's previously used by the common code.
	* src/libthai.c (_libthai_on_unload):
	  - Update #include for thbrk-private.h to brk-common.h.
	  - Update call to thbrk_on_unload() to brk_on_unload().

2014-08-03  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Move brk_get_dict() to common thbrk code.

* src/thbrk/brk-maximal.c, src/thbrk/thbrk-private.h:
	  - Move brk_get_dict() declaration to thbrk-private.h.
	  - Rename the function to thbrk_get_dict() for consistency.
	* src/thbrk/brk-maximal.c, src/thbrk/thbrk.c:
	  - Move implementation of brk_get_dict() to thbrk.c.
	  - Rename function & variable accordingly.
	* src/thbrk/brk-maximal.c (brk_root_pool):
	  - Rename function call to thbrk_get_dict().
	* src/thbrk/brk-maximal.h, src/thbrk/brk-maximal.c,
	  src/thbrk/thbrk.c:
	  - Move implementation of brk_maximal_on_unload() to
	    thbrk_on_unload(). 
	  - Remove brk_maximal_on_unload() which is no longer needed.

2014-08-03  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Rename th_brkpos_hints() to thbrk_brkpos_hints().

* src/thbrk/thbrk-private.h, src/thbrk/thbrk.c,
	  src/thbrk/brk-maximal.c:
	  - Change function prefix for consistency.

2014-08-03  Theppitak Karoonboonyanan  <theppitak@gmail.com>

Move th_brkpos_hints() to common thbrk code.

This is a preparation for alternative word break algorithms.
	
	* src/thbrk/brk-maximal.c, src/thbrk/thbrk-private.h:
	  - Move declaration of th_brkpos_hints() to thbrk-private.h.
	* src/thbrk/brk-maximal.c:
	  - Include thbrk-private.h to get th_brkpos_hints() declaration.
	* src/thbrk/brk-maximal.c, src/thbrk/thbrk.c:
	  - Move implementation of th_brkpos_hints() to thbrk.c.

2014-07-31  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-lang-ethnic.txt: Add words.

2014-06-30  Theppitak Karoonboonyanan  <theppitak@gmail.com>

* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-geo.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

* data/tdict-std.txt:
	  - Remove rare word 'จอแส' which causes wrong segmentation for 'จอแสดงผล'
	    as 'จอแส|ดง|ผล'.

2014-03-06  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-spell.txt: Add words.

* data/tdict-common.txt:
	  - Remove 'บ๊วย' which duplicates an entry in tdict-std.txt.
	* data/tdict-proper.txt, data/tdict-district.txt:
	  - Move 'กรือเซะ' from tdict-proper.txt to tdict-district.txt.
	* data/tdict-proper.txt, data/tdict-geo.txt:
	  - Move 'แอตแลนติส' from tdict-proper.txt to tdict-geo.txt.

2014-01-11  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix warning in test suite.

* tests/test_thbrk.c (main):
	  - Fix compiler warning on unused return value from fgets().
	    Caught by debian package building with test suites enabled.

2013-12-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-std-compound.txt, data/tdict-std.txt:
	  - Move 'ชุติมา' from -std-compound to -std.

2013-12-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2013-12-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2013-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.ac, NEWS:
	=== Version 0.1.20 ===

2013-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix warnings in tests.

* tests/test_thwbrk.c (main):
	  - Do some useful checks for 'unicodeCutCodeLength' unused var.
	  - Cast (size_t) returned value of wcslen() to (long) and use "%ld"
	    format to print it.
	  - Remove trailing spaces in source.

2013-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix warnings in tests.

* tests/test_thwchar.c (main):
	  - Cast (thchar_t *) to (char *) to fix signedness differences.
	  - Cast (thwchar_t) to (unsigned long) for "%lx" string format.
	  - Cast (size_t) return values from wcslen() and strlen() to long,
	    and use "%ld" format to print them.
	  - Replace 'outputLength' var with 'tisLength' and 'uniLength'
	    and do some useful checks on return values of th_tis2uni_line()
	    and th_uni2tis_line().

2013-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix warnings in tests.

* tests/test_thinp.c (test_th_validate):
	  - Cast (thchar_t *) to (char *), to fix signedness differences.
	* tests/thsort.c (readData):
	  - Cast (size_t) to (long) and use "%ld" format to print it.
	  - Redeclare maxData from int to size_t, to match that of nData.

2013-10-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* README: Remove version-specific text.

2013-10-22  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-geo.txt:
	* data/tdict-history.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt: Add words.

2013-09-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2013-09-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Check for doxygen required version.

* configure.ac:
	  - When doxygen-doc is enabled, also check doxygen version.

2013-09-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Update Doxyfile.

* doc/Doxyfile.in:
	  - Updated with 'doxygen -u'.

2013-09-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix automake warnings.

* src/thbrk/Makefile.am:
	* src/thcell/Makefile.am:
	* src/thcoll/Makefile.am:
	* src/thctype/Makefile.am:
	* src/thinp/Makefile.am:
	* src/thrend/Makefile.am:
	* src/thstr/Makefile.am:
	* src/thwbrk/Makefile.am:
	* src/thwchar/Makefile.am:
	* src/thwctype/Makefile.am:
	* src/thwstr/Makefile.am:
	* tests/Makefile.am:
	  - Replace deprecated INCLUDES with AM_CPPFLAGS.

2013-08-02  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt: Add words.

* data/tdict-std.txt: Remove 'พาส' whose use is rare and which
	potentially causes ambiguity like in 'พาสมองเสื่อม'.

2013-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2013-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Remove compound words that potentially cause ambiguity.

2013-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2013-01-29  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.ac, NEWS:
	=== Version 0.1.19 ===

2013-01-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Modernize autoconf & switch to XZ tarball compression.

* configure.in -> configure.ac:
	  - Rename file for modern autoconf.
	  - Replace deprecated AC_INIT() form with PACKAGE, VERSION form.
	  - Use AC_CONFIG_SRCDIR() instead for the old AC_INIT() form.
	  - Add AC_CONFIG_MACRO_DIR().
	  - Replace deprecated AC_INIT_AUTOMAKE() form with OPTIONS form,
	    passing "dist-xz no-dist-gzip" options to switch to XZ tarball.

* Makefile.am:
	  - Add "-I m4" to ACLOCAL_AMFLAGS, as suggested by libtool.

2013-01-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Change th_validate() default level back to ISC_STRICT.

* src/thinp/thinp.c (th_validate):
	  - Change ISC level back from ISC_BASICCHECK to ISC_STRICT.
	    ISC_BASICCHECK fails 'make check', and the previous value
	    ISC_STRICT should be better for general input with correction.

2013-01-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-science.txt: Fix typo.

2013-01-27  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-science.txt:
	* data/tdict-geo.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-district.txt: Add words.

* data/tdict-std-compound.txt: Drop loose compound 'สว่างไสว'.

2012-11-07  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Add input method API for correction with levels.

* include/thai/thinp.h:
	* src/libthai.def:
	* src/libthai.map (+th_validate_leveled):
	  - Add API for input validation with strictness arg.
	* src/thinp/thinp.c (th_validate, th_validate_leveled):
	  - Convert old th_validate() code to th_validate_leveled() and replace
	    the hard-coded ISC_STRICT with the arg.
	  - Make the old th_validate() a wrapper to th_validate_leveled() with
	    default param (ISC_BASICCHECK, not ISC_STRICT, though).
	* configure.in:
	  - Bump library versions to reflect the added API.

2012-11-07  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-ict.txt:
	* data/tdict-lang-ethnic.txt: Add words.

2012-10-16  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-district.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt: Add words.

2012-07-30  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* doc/Doxyfile.in: Upgrade to doxygen 1.8.1.2 format.

2012-07-29  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt: Add words.

2012-07-30  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* doc/Doxyfile.in: Upgrade to doxygen 1.8.1.2 format.

2012-06-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2012-06-12  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in, NEWS:
	=== Version 0.1.18 ===

2012-06-12  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-lang-ethnic.txt:
	* data/tdict-proper.txt: Add words.

2012-06-11  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-std.txt: Remove some prefix-words as found during
	  working on thailatex hyphenation.

2012-06-10  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-ict.txt: Add words.

2012-06-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Belated post-release version suffix added.

2012-06-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-city.txt:
	* data/tdict-common.txt:
	* data/tdict-ict.txt:
	* data/tdict-proper.txt:
	* data/tdict-spell.txt: Add words.

2012-03-07  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-std-compound.txt, data/tdict-std.txt: Fix typos.

2012-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt: Add words.

2012-02-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in, NEWS:
	=== Version 0.1.17 ===

2012-02-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Bump library revision, due to code change.

2012-02-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* tests/test_thbrk.c, tests/test_thwbrk.c: Add acronym to the tests.

2012-02-21  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-spell.txt:
	* data/tdict-ict.txt:
	* data/tdict-common.txt: Add words.

2012-02-20  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/thbrk.c (th_brk): Do not break last Thai chunk with
	dictionary if it's acronym.

2012-02-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-spell.txt:
	* data/tdict-common.txt: Add words.

* data/tdict-geo.txt: Fix typo.

2012-02-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-spell.txt:
	* data/tdict-district.txt:
	* data/tdict-common.txt: Add words.

2012-02-07  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/thbrk.c (th_brk): Handle Thai acronyms by detecting
	groups of <= 3 Thai characters + '.'. Thanks John Tapsell for
	the report and initial patch.

2012-02-07  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thrend/thrend.c (th_render_cell_): Reformat source a little bit.

2012-01-18  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-ict.txt:
	* data/tdict-common.txt: Add words.

* data/tdic-proper.txt, data/tdict-history.txt: Move "ศรีวิชัย" to
	-history.

2012-01-18  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in, NEWS:
	=== Version 0.1.16 ===

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in:
	  - Remove remaining '$Id'.
	  - Remove unnecessary AC_SUBST(VERSION).

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* -TODO: Removed, the plan is obsolete.

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* README: Update project URL & e-mail.
	* configure.in: Use googlegroups as bug report address.

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Add bug report address.

2011-11-05  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-spell.txt:
	* data/tdict-common.txt: Add words.

* data/tdict-country.txt: Use more popular short name for Saudi Arabia.

2011-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-common.txt, data/tdict-city.txt: "ไดฟุกุ" is city name.

2011-10-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-science.txt:
	* data/tdict-common.txt: Add words.

2011-09-03  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-common.txt: Add words.

* data/tdict-std.txt: Remove "มาย" which is rare but can potentially
	cause ambiguity (e.g. "มายกร่าง"). A valid compound, "เมามาย" is
	already in tdict-std-compound.txt.

2011-07-19  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-common.txt: Add words.

* data/tdict-std-compound.txt: Remove loose compound.

* data/tdict-std.txt:
	  - Remove rare words which potentially cause ambiguities.
	  - Remove a duplicated entry.

2011-06-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-geo.txt:
	* data/tdict-ict.txt:
	* data/tdict-spell.txt:
	* data/tdict-common.txt:
	* data/tdict-lang-ethnic.txt: Add words.

2011-06-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2011-03-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in, NEWS:
	=== Version 0.1.15 ===

2011-03-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	* data/tdict-district.txt:
	* data/tdict-common.txt: Add words.

2011-03-22  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-science.txt:
	* data/tdict-spell.txt:
	* data/tdict-district.txt:
	* data/tdict-common.txt: Add words.

2011-03-22  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-district.txt:
	Move 2 names from proper, district to history.

* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-district.txt:
	* data/tdict-common.txt: Add words.

2011-03-20  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-district.txt:
	* data/tdict-history.txt:
	Move "หริภุญชัย" from district to history list.

* data/tdict-city.txt:
	Change "ฟูกุโอกะ" to "ฟูกูโอกะ", which is statistically more popular.

* data/tdict-history.txt:
	* data/tdict-proper.txt:
	* data/tdict-city.txt:
	* data/tdict-science.txt:
	* data/tdict-geo.txt:
	* data/tdict-spell.txt:
	* data/tdict-common.txt:
	Add words.

2011-03-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Split historical names out of tdict-city.txt and tdict-district.txt.

* data/Makefile.am:
	* +data/tdict-history.txt:
	* data/tdict-city.txt:
	* data/tdict-district.txt:
	Split historical names into tdict-history.txt.

* data/tdict-city.txt:
	* data/tdict-geo.txt:
	* data/tdict-lang-ethnic.txt:
	Move 2 names from tdict-city.txt to tdict-geo.txt and
	tdict-lang-ethnic.txt.

2011-03-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Split tdict-country.txt and tdict-city.txt out of tdict-geo.txt
	and let tdict-geo.txt keeps only geographical names.

* data/Makefile.am:
	* +data/tdict-country.txt:
	* +data/tdict-city.txt:
	* data/tdict-geo.txt:
	  - Split country names into tdict-country.txt
	  - Split city names into tdict-city.txt

* data/tdict-geo.txt:
	* data/tdict-lang-ethnic.txt:
	  - Split language/etchnic names into tdict-lang-ethnic.txt

2011-03-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-proper.txt:
	* data/tdict-common.txt:
	* data/tdict-geo.txt: Add words.

2011-03-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-spell.txt:
	* data/tdict-proper.txt:
	* data/tdict-common.txt:
	* data/tdict-geo.txt: Add words.

2011-03-09  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-ict.txt:
	* data/tdict-district.txt:
	* data/tdict-common.txt:
	* data/tdict-science.txt: Add words.

2010-12-27  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Adjust dictionary, as per Widhaya's feedback:
	http://groups.google.com/group/thai-linux-foss-devel/browse_thread/
	thread/986f74d786c9bace

* data/tdict-std-compound.txt:
	* data/tdict-std-common.txt:
	  - Remove some compounds that may cause potential ambiguities.
	* data/tdict-spell.txt:
	* data/tdict-common.txt:
	  - Add words.

2010-12-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{common,geo}.txt: Add words.

2010-05-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{district,common,science}.txt: Add words.

2010-03-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

Turn off manpages generation
	* doc/Doxyfile.in: Turn GENERATE_MAN off.
	* doc/Makefile.am: Remove man targets.

2010-02-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Bump library revision.

* configure.in, NEWS:
	=== Version 0.1.14 ===

2010-02-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{ict,proper,common}.txt: Add words.

2010-02-28  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thcoll/cweight.c (char_weight_tbl_): Replace TIS-620 characters
	in comments with character names.
	* src/thctype/thctype.c (_th_ctype_tbl, _th_chlevel_tbl): Remove
	non-TIS-620 characters from comments. Convert source to UTF-8.

2010-02-27  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_get_dict): Limit path[] scope.

2010-02-27  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_get_dict): Print some warning when
	dictionary opening fails.

2010-02-27  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_get_dict): Do not try to open dictionary
	again if once tried. Multiple failing trials degraded performance.

2010-02-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* include/thai/*.h, src/*.c, src/*/*.c: Add my e-mail address to
	license header as the contact address.

* src/thctype/wtt.c,
	* src/thctype/thctype.c,
	* src/thcoll/cweight.h,
	* src/thcoll/cweight.c,
	* src/thwctype/thwctype.c:
	Add editor modeline.

2010-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thinp/thinp.c,
	* src/thrend/thrend.c,
	* src/thcoll/thcoll.c,
	* src/thcoll/cweight.c,
	* src/thwstr/thwstr.c,
	* src/thwchar/thwchar.c,
	* src/thwbrk/thwbrk.c,
	* src/thstr/thstr.c,
	* src/thcell/thcell.c:
	Reformat source. Add editor modeline.

2010-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* tests/thsort.c: Yet another GPL header leftover removal.
	Update my e-mail address.

2010-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Replace <thai/tis.h> with a rewritten one, so we can enforce the
	license.

* include/thai/tis.h:
	Rewritten based on Unicode names, with TIS_ prefix.

* src/thinp/thinp.c (corrections, th_validate),
	* src/thrend/thrend.c (shiftdown_tone_ad, shiftdownleft_tone_ad,
	  shiftleft_tone_ad, shiftdown_bv_bd, tailcutcons, th_render_cell_,
	  th_render_cell_tis),
	* src/thbrk/brk-maximal.c (th_brkpos_hints),
	* src/thcell/thcell.c (th_next_cell, th_prev_cell),
	* tests/test_thcell.c (test_ans_nodecomp_am, test_ans_decomp_am):
	Update character names to the new TIS_* ones.

2010-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* include/thai/*.h, src/*.c, src/*/*.c,
	* Makefile.am, include/Makefile.am, include/thai/Makefile.am,
	* src/Makefile.am, src/*/Makefile.am,
	* tests/*.c, tests/*.sh,
	* tests/Makefile.am,
	* data/Makefile.am,
	* man/template.3, man/man3/*.3,
	* man/Makefile.am, man/man3/Makefile.am:
	Remove CVS $Id in all files.

* src/thctype/wtt.c: Fix file name in comment.

2010-02-24  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* include/thai/*.h, src/*.c, src/*/*.c: Add license header to every
	source file.

2010-02-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* doc/Makefile.am: The added *.c dependencies must be from src/*/*.c,
	not include/thai/*.c. Silly copy-and-paste error.

2010-02-23  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* doc/Makefile.am: Add *.c to doxygen.stamp dependency, as the doc
	comments have been moved there.

2010-02-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_root_pool): Fail if brk_get_dict()
	returns NULL.

2010-02-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{ict,proper,common,science}.txt: Add words.

2010-01-30  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thcoll/cweight.c, src/thcoll/cweight.h, src/thcoll/thcoll.c:
	Remove GPL header left over. The library is LGPL.

2010-01-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/brk-maximal.c (brk_root_pool): Fail if brk_pool_node_new()
	returns NULL.
	* src/brk-maximal.c (brk_maximal_do_impl): Fail if best_brk_new()
	returns NULL.

2010-01-25  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

2010-01-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Bump library revision.

* configure.in, NEWS:
	=== Version 0.1.13 ===

2010-01-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Add 'AC_CONFIG_MACRO_DIR([m4])' for the new libtool.

2010-01-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thwchar/thwchar.c (th_tis2uni_line, th_uni2tis_line):
	Fix type of 'result' argument to match that in header file, so
	doxygen recognizes and generates doc for it.

2010-01-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{proper,common}.txt: Add words.

2010-01-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* include/thai/thctype.h, src/thctype/thctype.c: Replace non-portable
	isascii() with most-significant bit check.

2010-01-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{common,geo}.txt: Add words.

2010-01-14  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* include/thai/thailib.h, src/libthai.c: Move more doc comments.

2010-01-09  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Move documentation from *.h to *.c, so libthai developers have the doc
	at hand. Users can still read the doxygen-generated doc BTW.

* include/thai/*.h, src/*/*.c: Move doc comments from *.h to *.c.
	* doc/Doxyfile.in: Set INPUT to @top_srcdir@.

2010-01-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-{proper,common,geo}.txt: Add words.

2010-01-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_shot_init):
	Return failure/success status.
	* src/thbrk/brk-maximal.c (brk_pool_node_new):
	Check & handle return code from brk_shot_init().

2010-01-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thwbrk/thwbrk.c (th_wbrk, th_wbrk_line):
	Rewritten, for harmony with the rests of the source, plus potential
	integer overflow vulnerability removed.

2010-01-08  Theppitak Karoonboonyanan  <thep@linux.thai.net>

Fix potential integer overflow vulnerabilities.
	Thanks Tim Starling <tstarling@wikimedia.org> for pointing out.

* src/thbrk/thbrk.c (th_brk_line): Return zero if malloc size overflows
	or malloc fails.
	* src/thbrk/brk-maximal.c (best_brk_new): Return NULL if malloc size
	overflows.

2009-08-22  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Post-release version suffix added.

* data/tdict-std-compound.txt: Remove loose compounds.
	* data/tdict-{ict,district,proper,common,science,lang-ethnic,geo}.txt:
	Add words.

2009-06-18  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* configure.in: Bump library revision.

* configure.in, NEWS:
	=== Version 0.1.12 ===

2009-06-18  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-std-compound.txt: Remove loose compounds.

2009-06-18  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/tdict-std-compound.txt: Remove loose compounds.
	* data/tdict-{ict,district,proper,common,science,geo}.txt: Add words.

2009-06-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_maximal_do_impl, best_brk_contest): 
	Back out weighted scoring for solution contest, as it sometimes tries
	too hard to minimize word count, even though another solution with more
	known words can match more text and results in smaller unknown zone.
	Accidental words problem, which this scheme tried to address, has
	already been alleviated with increased recovery threshold.

2009-06-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_maximal_do_impl): Add break pos at
	string end and let it enter the contest, instead of backing it out by
	moving back str_pos as done in previous commit. That commit just
	missed the point. Doing the right thing also guarantees consistency
	with the case when some text of valid words is appended to the string.

2009-06-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (brk_maximal_do_impl): Set str_pos for cases
	that fail at string end back to the last progress position, to prevent
	it from wrongly winning the contest over some good candidates.

2009-06-17  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/tkbrk/brk-maximal.c (RECOVERED_WORDS): Increase recovery
	threshold, 2 is too few and can trigger wrong recovery position too
	easily.

2009-06-15  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* data/Makefile.am, +data/tdict-lang-ethnic.txt: Add a new category.
	* data/tdict-{common,geo}.txt, data/tdict-lang-ethnic.txt:
	Move some words to the new category.
	* data/tdict-{lang-ethnic,ict,spell,proper,common,science,geo}.txt:
	Add words.
	* data/tdict-std.txt: Remove a duplicated compound.

2009-06-13  Theppitak Karoonboonyanan  <thep@linux.thai.net>

* src/thbrk/brk-maximal.c (best_brk_contest): Compare penalties with
	'<=' like in previous code, instead of '<', to retain longest-matching
	effect in case of equal scores.