Commit Graph

8401 Commits

Author SHA1 Message Date
Behdad Esfahbod 9d3677899f Use a function template instead of struct template for Null 2010-04-21 01:55:12 -04:00
Behdad Esfahbod ffd321afd9 Simplify Tag struct 2010-04-21 01:55:09 -04:00
Behdad Esfahbod 00e23fcc6f Cosmetic 2010-04-21 01:55:03 -04:00
Behdad Esfahbod a87072db5d Fix Class operator return type 2010-04-21 01:54:57 -04:00
Behdad Esfahbod f9b3772798 Add couple consts to operators 2010-04-21 01:54:53 -04:00
Behdad Esfahbod 53d237ec63 Remove ASSERT_SIZE_DATA 2010-04-21 01:54:49 -04:00
Behdad Esfahbod 06558d2a74 Round instead of trunc 2010-04-21 01:50:15 -04:00
Behdad Esfahbod 0e206de986 Fix warnings 2010-04-21 01:48:41 -04:00
Behdad Esfahbod d5943407a4 Fix debug build 2010-04-21 01:42:11 -04:00
Behdad Esfahbod 12b27ed91d Fix leak when duplicating blob 2010-04-21 01:41:53 -04:00
Behdad Esfahbod 4b8487d83e Fix the mystery bug!
A couple bugs joined forces to exhibit the mystery behavior of
crashes / infinite loops on OS X / wrong kerning / invalid memory
access.  Pooh!

The bugs were involved:

  - Wrong pointer math with ValueRecord in PairPosFormat1

  - Fallout from avoiding flex arrays, code not correctly updated
    to remove sizeof() usage.

We strictly never use sizeof() directly now.  And the PairPos code
is cleaned up.  Should fix them all.  Bugs are:

  Bug 605655 - Pango 1.26.2 introduces kerning bug
  Bug 611229 - Pango reads from uninitialized memory
  Bug 593240 - (pangoosx) Crash / infinite loop with Mac OS X

We were also doing wrong math converting Device adjustments to
hb_position_t.  Fallout from FreeType days.  Should shift 16, not
6.  Fixed that too.

There's still another bug: we don't sanitize Device records
referenced from value records.  Fixing that also.
2010-04-21 01:41:29 -04:00
Behdad Esfahbod 9b39755d10 Typo 2010-04-15 14:00:25 -04:00
Behdad Esfahbod f85ec1c7da Cosmetic 2010-03-10 04:14:40 -05:00
Behdad Esfahbod e48ed72230 [ft] Check stream->read instead of stream->base
The former is more robust.  See discussion on freetype-devel.
2010-03-01 22:33:45 -05:00
Behdad Esfahbod 917c227500 Make blob unlocking 64bit-safe
GNOME Bug 604128 - Applications crash when displaying Hebrew characters
2010-02-23 16:47:51 -05:00
Behdad Esfahbod ecd2e996d2 Improve the alignment for NullPool 2010-02-23 02:42:00 -05:00
Behdad Esfahbod bc7830e425 Use __attribute__((unused)) only with gcc 4 and later
See discussion at:
https://bugzilla.gnome.org/show_bug.cgi?id=610183
2010-02-17 15:14:57 -05:00
Behdad Esfahbod 555d11273e [GDEF] Fix bug in building synthetic GDEF 2010-01-26 12:58:59 -05:00
Behdad Esfahbod 70834d89c3 [TODO] Add kern/GPOS interaction 2010-01-26 12:17:37 -05:00
Behdad Esfahbod cd11a98fa1 Fix compile with older FreeType 2009-12-20 23:05:02 +01:00
Behdad Esfahbod 7f7448a354 Indent 2009-12-20 21:42:14 +01:00
Behdad Esfahbod e70f45eb52 Revert "Fallback to 'kern' if no GPOS applied"
This reverts commit d740c8f78b.

The change is wrong.  If there is a GPOS table, we should not use 'kern'.
2009-12-20 21:26:24 +01:00
Behdad Esfahbod d740c8f78b Fallback to 'kern' if no GPOS applied 2009-12-20 21:23:56 +01:00
Behdad Esfahbod 2f78c17197 Remove glibism! 2009-12-20 21:03:11 +01:00
Behdad Esfahbod 907e67da13 Oops, add file. 2009-12-20 20:59:51 +01:00
Behdad Esfahbod 2014b8d110 Hook OpenType shaping up
Default features only for now.
2009-12-20 20:58:26 +01:00
Behdad Esfahbod 196610ba4c Pass features down 2009-12-20 19:02:23 +01:00
Behdad Esfahbod 51f141a7f3 Avoid overflow 2009-12-20 18:22:52 +01:00
Behdad Esfahbod 26d7a75752 Refactor hb_shape a bit 2009-12-20 17:58:25 +01:00
Behdad Esfahbod 001fc2d2aa Add TrueType kern support 2009-12-20 17:24:05 +01:00
Behdad Esfahbod 2c1b85cf66 Direct unicode->get_mirroring directly 2009-12-20 16:29:17 +01:00
Behdad Esfahbod 6a2ef5aa54 Do mirroring 2009-12-20 16:28:01 +01:00
Behdad Esfahbod 0465e69832 Protect against NULL funcs 2009-12-20 16:25:18 +01:00
Behdad Esfahbod 5ceefa1d8d Add hb_unicode_get_*() functions 2009-12-20 15:29:16 +01:00
Behdad Esfahbod b8a53e44ce We'll have to link to libstdc++ if linking to ICU, so disable test for now
We have to get rid of the ICU in main lib.  Still thinking about best way
to do it.
2009-12-20 14:56:25 +01:00
Behdad Esfahbod 314905d754 Explicitly track whether the buffer has positions 2009-12-20 14:50:42 +01:00
Behdad Esfahbod 314b460d8a Add HB_DIRECTION_IS_FORWARD/BACKWARD 2009-12-20 13:58:50 +01:00
Behdad Esfahbod bdd0ff5290 Make main.cc compile without glib 2009-12-15 04:07:40 -05:00
Behdad Esfahbod 8a7d168082 Distribute autogen.sh 2009-12-15 03:53:45 -05:00
Behdad Esfahbod 807b8aa486 Another C++ strictness fix
Pango Bug 602408 - Invalid C++ code breaks compile with Sun C++ Compiler
(Error: A union member cannot have a user-defined assignment operator)

According to the bug:

C++ Programming Language by Bjarne Stroustrup: Chapter 10.4.12 forbids
explicitly using of union members with constructors, destructors or assignment
operations.

So we use a set() method instead of the assignment operator.  Ugly, but hey,
that's life.
2009-11-19 20:35:01 -05:00
Behdad Esfahbod 8b9b866d3e Fix warning 2009-11-19 20:27:57 -05:00
Behdad Esfahbod c65b26acf2 Use autoconf FLEXIBLE_ARRAY_MEMBER when available 2009-11-18 11:27:33 -05:00
Behdad Esfahbod d33f674cb7 Add Mongolian variation selectors 2009-11-18 09:47:44 -05:00
Behdad Esfahbod d6387757de Fix sanitize 2009-11-11 17:15:03 -05:00
Behdad Esfahbod 636f017e38 Remove obsolete TODO item 2009-11-10 12:56:35 -05:00
Behdad Esfahbod bf2b25dd85 Remove done item 2009-11-06 19:52:47 -05:00
Behdad Esfahbod 4a86053159 Reverse buffer at the end if RTL 2009-11-06 19:52:01 -05:00
Behdad Esfahbod ff44f88df2 Handle shaping in non-native direction 2009-11-06 19:48:16 -05:00
Behdad Esfahbod b4b4272c8d Oops, wrong change 2009-11-06 17:46:33 -05:00
Behdad Esfahbod 52ea47767c Change order of font and face for API consistency 2009-11-06 17:45:38 -05:00