Commit Graph

861 Commits

Author SHA1 Message Date
Alan Coopersmith 5b8380d52e Avoid memory leak when NULL path passed to FcStrBuildFilename
Reported by parfait 1.3:
   Memory leak of pointer sset allocated with FcStrSetCreate()
        at line 933 of src/fcstr.c in function 'FcStrBuildFilename'.
          sset allocated at line 927 with FcStrSetCreate().
          sset leaks when sset != NULL at line 932.

Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
2013-11-05 11:43:41 +09:00
Akira TAGOH aa22e6e639 Warn if no <test> nor <edit> elements in <match>
This corrects an error message being reported at
  https://bugs.freedesktop.org/show_bug.cgi?id=71085
  Bug 71085 - "out of memory" errors on empty match element in fonts.conf

and somewhat works as a workaround for
  https://bugs.freedesktop.org/show_bug.cgi?id=59438
  Bug 59438 - Fix <alias> inside <match>
2013-10-31 21:48:27 +09:00
Akira TAGOH 525a135ccf Change the default weight on match to FC_WEIGHT_NORMAL 2013-10-28 11:54:04 +09:00
Akira TAGOH 06b388523d Fix build issue on Debian/kFreeBSD 7.0
There are posix_fadvise(2) but not POSIX_FADV_WILLNEED.
Patch from Ryo ONODERA.
2013-10-21 11:50:55 +09:00
Akira TAGOH 0203055520 Workaround the race condition issue on updating cache 2013-10-02 16:34:34 +09:00
Akira TAGOH 43f768b53f avoid reading config.h twice
config.h is read from fcint.h now so having a line of the sort of #include "config.h"
is duplicate.

Bug 69833 - Incorrect SIZEOF_VOID_P and ALIGNOF_DOUBLE definitions causes nasty warnings on MacOSX when building fat libraries
2013-09-26 17:51:15 +09:00
Akira TAGOH 8a174b6c51 Fix a crash when FcPattern is set to null on FcFontSetList() and FcFontList() 2013-09-24 11:14:57 +09:00
Akira TAGOH 5e6b8894ea Copy all values from the font to the pattern if the pattern doesn't have the element 2013-09-18 17:33:45 +09:00
Akira TAGOH a61e145304 Fix memory leaks in FcFreeTypeQueryFace 2013-09-09 19:59:31 +09:00
Akira TAGOH 3e5f70a16a Do not create a config dir for migration when no config files nor dirs 2013-09-02 20:51:46 +09:00
Akira TAGOH 272a99217b Fix a crash 2013-08-31 10:43:13 +09:00
Akira TAGOH fba9efecd2 Fix a wrong edit position when 'kind' is different 2013-08-26 12:47:07 +09:00
Akira TAGOH 223c1384c9 Fix a crash when non-builtin objects are edited 2013-08-24 13:46:44 +09:00
Akira TAGOH 6c664d533d Fix a typo 2013-08-24 13:46:40 +09:00
Behdad Esfahbod f3bb3f19c9 Fix assertion
Apparently some AIX versions have 64bit pointers yet 32bit double
alignment.  Fix assertion.
2013-08-21 14:31:55 -04:00
Behdad Esfahbod ec3f0af6f0 Minor 2013-08-21 14:27:16 -04:00
Akira TAGOH 084cf7c44e Bug 16818 - fontformat in match pattern is not respected? 2013-08-14 12:51:26 +09:00
Akira TAGOH 45221ab12f Bug 67845 - Match on FC_SCALABLE 2013-08-07 12:19:33 +09:00
Akira TAGOH 041deb0cc5 warn deprecated only when migration failed 2013-08-07 12:05:32 +09:00
Akira TAGOH d52daa0024 Bug 67809 - Invalid read/write with valgrind when assigning something twice 2013-08-06 15:09:23 +09:00
Akira TAGOH a51d2767ce Fix wrong edit position 2013-08-06 13:05:53 +09:00
Akira TAGOH 04bd904632 trivial code optimization 2013-07-09 16:43:26 +09:00
Akira TAGOH d420e1df98 Rework to apply the intermixed test and edit elements in one-pass 2013-07-04 19:51:03 +09:00
Akira TAGOH 1162515a98 Add FC_UNUSED to FC_ASSERT_STATIC macro to avoid compiler warning 2013-07-03 11:56:58 +09:00
Akira TAGOH cbf06d7d3c Use INT_MAX instead of unreliable hardcoding value 2013-07-02 19:04:36 +09:00
Akira TAGOH 0907589a79 Fix the behavior of intermixed tests end edits in match
to get the following recipe working:

<match>
  <test1 .../>
  <edit1 .../>
  <test2 .../>
  <edit2 .../>
</match>

as:

<match>
  <test1 .../>
  </edit1 .../>
</match>
<match>
  <test1 .../>
  <test2 .../>
  <edit2 .../>
</match>
2013-06-28 15:54:38 +09:00
Akira TAGOH 197d06c49b Add FcTypeUnknown to FcType to avoid comparison of constant -1
This change reverts 9acc14c34a
because it doesn't work as expected when building
with -fshort-enums which is default for older arms ABIs

Thanks for pointing this out, Thomas Klausner, Valery Ushakov, and Martin Husemann
2013-06-28 15:04:11 +09:00
Akira TAGOH 38ab7ab2fb Fix a incompatible pointer warning on NetBSD 2013-06-27 13:10:27 +09:00
Akira TAGOH 8603e58695 Fix a shift count overflow on 32bit box 2013-06-27 12:30:56 +09:00
Akira TAGOH 9acc14c34a Fix a comparison of constant warning with clang 2013-06-26 12:03:38 +09:00
Akira TAGOH cd9b1033a6 Bug 64906 - FcNameParse() should ignore leading whitespace in parameters
After this change, the following works as expected:
$ FC_DEBUG=4 fc-match ":family=foo bar, sans-serif"
...
FcConfigSubstitute Pattern has 3 elts (size 16)
        family: "foo bar"(s) "sans-serif"(s)
...
2013-05-24 13:57:23 +09:00
Akira TAGOH 1cad82cde2 Bug 63922 - FcFreeTypeQueryFace fails on postscripts fonts loaded from memory
Workaround to not failing even when the hash is unable to generate from fonts.
This change also contains to ignore the case if the hash isn't in either both
patterns.
2013-05-16 16:49:14 +09:00
Akira TAGOH 0f9aa8759d Fix missing OSAtomicCompareAndSwapPtrBarrier() on Mac OS X 10.4
based on hb-atomic-private.hh in harfbuzz
2013-05-16 14:58:35 +09:00
Akira TAGOH f6244d2cf2 Use the glob matching for filename
Regex is expensive to compare filenames. we already have the glob matching
and it works enough in this case.

Prior to this change, renaming FcConfigGlobMatch() to FcStrGlobMatch() and moving to fcstr.c
2013-05-08 11:57:49 +09:00
Akira TAGOH 03216ccf4c Bug 63329 - make check fails: .. contents:: :depth: 2
Add back FcHashGetSHA256DigestFromFile() and fall back to it
when font isn't SFNT-based font because FT_Load_Sfnt_Table
fails with FT_Err_Invalid_Face_Handle.
2013-04-10 18:41:22 +09:00
Akira TAGOH fc5a589aba Revert the previous change and rework to not export freetype API outside fcfreetype.c 2013-04-09 17:18:43 +09:00
Akira TAGOH c93a8b8b54 Obtain fonts data via FT_Face instead of opening a file directly 2013-04-09 12:46:30 +09:00
Akira TAGOH 9299155b52 Ensure closing fp on error 2013-04-09 11:34:35 +09:00
Sebastian Freundt 18bf57c70a build-chain, replace INCLUDES directive by AM_CPPFLAGS
As of automake-13.1 the INCLUDES directive is no longer supported.
An automake run will return with an error.

This changeset simply follows automake's advice to replace INCLUDES
by AM_CPPFLAGS.
2013-04-08 11:40:58 +09:00
Akira TAGOH 8fd0ed60a6 Bug 62980 - matching native fonts with even :lang=en
Fix the matcher modified by 4eab908c86
to deal with both strong and weak of FC_LANG as the same location in the score
2013-04-01 18:16:28 +09:00
Akira TAGOH b561ff2016 Bug 38737 - Wishlist: support FC_POSTSCRIPT_NAME
Add the PostScript name into the cache and the matcher.
Scoring the better font against the PostScript name by
the forward-matching.
2013-03-21 16:34:42 +09:00
Akira TAGOH c758206e8c Fix a SIGSEGV on FcPatternGet* with NULL pattern 2013-03-21 11:58:06 +09:00
Behdad Esfahbod bdf1581e3d Fix crash with FcConfigSetCurrent(NULL) 2013-03-08 05:53:27 -05:00
Akira TAGOH aad4d6f6c6 Do not copy FC_*LANG_OBJECT even if it's not available on the pattern
those objects are linked to the corresponding string objects.
this may causes inconsistency that those objects has more values than them.
2013-03-07 13:20:31 +09:00
Akira TAGOH e96d776088 Bug 59456 - Adding a --sysroot like option to fc-cache
Add an ability to set the system root to generate the caches.
In order to do this, new APIs, FcConfigGetSysRoot() and
FcConfigSetSysRoot() is available.
2013-03-05 18:38:50 +09:00
Akira TAGOH 569657a24c Fix a memory leak 2013-03-05 12:46:01 +09:00
Akira TAGOH 612ee2a5c9 Fix broken sort order with FcFontSort()
which was introduced by 4eab908c86
2013-03-01 22:21:25 +09:00
Akira TAGOH ea4ebd5937 Fix a crash when the object is non-builtin object 2013-03-01 19:38:21 +09:00
Behdad Esfahbod 83f679ce55 Accept digits as part of OpenType script tags
They've been used since 2005.
2013-02-15 09:48:56 -05:00
Akira TAGOH 72b0480a21 Add Culmus foundry to the vendor list
Maps fonts produced by the Culmus project <http://culmus.sourceforge.net>
to the XLFD foundry name culmus.

For TrueType fonts, maps the vendor code CLM from the TrueType vendor id field.

For Type1 fonts, which use heuristics to guess mappings to XLFD foundries from
words in the copyright notice, add the names of the main contributors to
the Culmus product to recognize the fonts under their copyright.

Patch from Maxim Iorsh
2013-02-07 17:56:36 +09:00