Behdad Esfahbod
5515355367
[ot-font] Don't pre-load cmap table
...
Now that we have get_h_advances() and get_nominal_glyphs() implemented, the
overhead of doing a proper atomic load would be once per run, NOT once per
glyph. So, no need to pre-load the tables to avoid that overhead.
As such, hb_ot_font_set_funcs() has become really cheap. Can *finally* make
it be default font functions on all newly created fonts!
2018-10-09 15:12:09 -04:00
Behdad Esfahbod
ec84460e46
[ot/ft] Implement get_nominal_glyphs() callback
...
Some more measurable speedup. The recent commits' speedups are as follows:
Testing with Roboto, ****when disabling kern and liga****:
Before:
FT --features=-kern,-liga
user↦ 0m0.521s
OT --features=-liga,-kern
user↦ 0m0.568s
After:
FT --features=-liga,-kern
user↦ 0m0.428s
OT --features=-liga,-kern
user↦ 0m0.470s
So, 17% speedup.
Note that FT callbacks are faster than OT these days since we added an advance
cache to FT. I don't think the difference is enough to justify adding a cache
to OT.
When not disabling kern, the thing is three times slower, so the speedups
are three times less impressive... Still, 5% not bad for a codebase that I
otherwise thought is optimized out.
Note that, because of this and other optimiztions in our main shaper,
disabling kern and liga, the OT shaper is now *faster* than the fallback
shaper. So, that's my recommendation to clients that need the absolute
fastest...
2018-10-09 15:07:04 -04:00
Behdad Esfahbod
bee93e2697
Add const to get_*_advances API
...
Ouch!
2018-10-09 08:02:25 -04:00
Behdad Esfahbod
47030b1855
[ft/ot] Remove implementation of deprecated kerning funcs
2018-10-09 00:30:45 -04:00
Behdad Esfahbod
13da3be0b3
[ot-font] Load hmtx/vmtx lazily
...
Since we have get_*_advanes() API now, the overhead is once per shape,
not once per glyph.
Only cmap is warmed-up at set_funcs() time now.
2018-10-07 18:25:41 +02:00
Behdad Esfahbod
856db4c918
Minor
2018-10-07 18:21:15 +02:00
Behdad Esfahbod
8a31e40629
[font] Make *_advance() fallback to *_advances
...
And remove redundant implementations.
2018-09-30 06:08:11 -04:00
Behdad Esfahbod
cbea7d49ab
[ot-font] Rename
2018-09-11 12:57:41 +02:00
Behdad Esfahbod
d8a67dac2a
[ot-font] Add advances() callbacks
2018-09-11 12:55:54 +02:00
Behdad Esfahbod
a268068010
[ot-face] Remove tables_t. The face_data_t is what tables_t was
2018-08-26 15:18:46 -07:00
Behdad Esfahbod
3a0b3a29cc
Free up hb-ot-face.hh from includes
...
There might be a better way to do this, but I couldn't find...
2018-08-26 15:11:24 -07:00
Behdad Esfahbod
e6cb938065
[ot-face] Unify accelerators and tables
2018-08-26 00:21:29 -07:00
Behdad Esfahbod
59bca0d2cb
[ot-font] Rename
2018-08-26 00:06:12 -07:00
Behdad Esfahbod
1c48a7ef06
[ot-font] Cache tables on face, not font
2018-08-25 23:57:16 -07:00
Behdad Esfahbod
71f8893eff
[ot-font] Use init0
2018-08-25 23:27:45 -07:00
Behdad Esfahbod
f36ae6f8a8
[ot-font] Make everything (cmap/metrics) lazy
2018-08-25 23:26:45 -07:00
Behdad Esfahbod
62fc54725a
[ot-font] Fallback vertical advance to UPEM
...
Used to be ascent+descent. Whatever. This is more convenient for
upcoming changes.
2018-08-25 23:15:52 -07:00
Behdad Esfahbod
c77ae40852
Rename hb-*private.hh to hb-*.hh
...
Sorry for the noise, downstream custom builders. Please adjust.
2018-08-25 22:36:36 -07:00
Emil A Eklund
7112c6413c
Guard free_static function definitions with macro
...
Add HB_USE_ATEXIT macros around free_static function definitions to
avoid unused function compiler warnings/errors.
2018-08-14 14:04:10 -07:00
Behdad Esfahbod
7bd508a0c4
[lazy] Rename free()
2018-08-12 17:19:55 -07:00
Behdad Esfahbod
bb905e9afc
[lazy] Minor
2018-08-12 16:44:44 -07:00
Behdad Esfahbod
53442be1ed
[lazy] Use for ft_library
2018-08-12 16:20:11 -07:00
Behdad Esfahbod
7a8d480378
[lazy] Add .free()
2018-08-12 16:00:13 -07:00
Behdad Esfahbod
c7ca30a533
[ot/ft] Port font_funcs statis to lazy-loader
2018-08-12 13:46:53 -07:00
Behdad Esfahbod
1f7380944d
[atomic] Add hb_atomic_ptr_t<> and port all uses
...
Found and fixed a couple bugs.
Found a couple multithreading issues. Marked them with "XXX-MT-bug".
2018-08-09 00:27:01 -07:00
Behdad Esfahbod
30ff8ac865
Rename
2018-08-03 17:33:13 -07:00
Behdad Esfahbod
c988b11999
Move code around
2018-08-02 22:55:52 -07:00
Behdad Esfahbod
574b37e933
Minor
2018-08-02 02:22:38 -07:00
Behdad Esfahbod
ff7826e90b
Reduce storage by sharing face amongst lazy_loaders
2018-08-02 01:29:05 -07:00
Behdad Esfahbod
ed7b2e58fc
Remove OT namespace from hb-machinery-private.hh
2018-08-02 00:08:08 -07:00
Ebrahim Byagowi
24b8b9b227
Resolve clang's used-but-marked-unused warnings
2018-04-23 11:38:45 -07:00
Ebrahim Byagowi
70d36543aa
Make atexit callbacks threadsafe ( #930 )
2018-03-29 17:30:28 -07:00
Ebrahim Byagowi
0ef6ab25ab
[ot-color] Move cbdt to the namespace ( #860 )
2018-03-04 02:47:26 +03:30
Behdad Esfahbod
7b40876a58
[ot] Remove unneeded TODO item
...
Looks like hb-ot-font is complete after all!
CFF font names and CFF glyph metrics don't work though...
2017-11-14 20:22:05 -08:00
Behdad Esfahbod
eab4feb5e2
[hmtx] Use curiously recurring template pattern
...
https://en.wikipedia.org/wiki/Curiously_recurring_template_pattern
2017-11-14 20:17:15 -08:00
Behdad Esfahbod
977ddff1f1
[ot] Move hb_ot_face_cmap_accelerator_t
2017-11-14 20:06:19 -08:00
Behdad Esfahbod
a85d7ead04
[ot] Move hb_ot_face_metrics_accelerator_t
2017-11-14 20:00:34 -08:00
Behdad Esfahbod
7e2839c438
[ot] Move hb_ot_face_glyf_accelerator_t
2017-11-14 19:52:29 -08:00
Behdad Esfahbod
c4e18e54dc
[ot] Move hb_ot_face_cbdt_accelerator_t
2017-11-14 19:52:26 -08:00
Behdad Esfahbod
9b04b0384f
[ot] Fold hb_ot_face_post_accelerator_t
2017-11-14 19:31:50 -08:00
Behdad Esfahbod
702d86ba6b
[ot] Fold hb_ot_face_kern_accelerator_t
2017-11-14 19:25:51 -08:00
Behdad Esfahbod
93f7c1652a
Revert "[glib/ucdn/icu/ft/ot] Make returned funcs inert"
...
This reverts commit 5daf3bd449
.
If other atexit callbacks try to destruct the objects we destruct
in atexit callbacks, bad things will happen.
I'll come up with some other way to catch premature destruction
of HB-owned objects.
Fixes https://github.com/behdad/harfbuzz/issues/618
2017-11-14 10:59:54 -08:00
Behdad Esfahbod
82a38d1f7a
[kern] Allow subtables longer than 64kb
...
Apparently calibri.ttf does this:
https://github.com/fonttools/fonttools/pull/1094#discussion_r148933791
2017-11-06 15:28:24 -05:00
Behdad Esfahbod
4b3278ef8d
[ot] Hook up horizontal kerning to kern table
...
Seems to work. Yay!
Still to do: run kerning if GPOS doesn't have 'kern' feature.
2017-11-01 19:41:29 -06:00
Behdad Esfahbod
77acc1106e
[kern] More...
...
Almost there.
2017-11-01 19:40:51 -06:00
Behdad Esfahbod
2a16f647ae
[kern] Start implementing kern table
...
Pushing this out early to catch bot errors since I'm using template tricks
we were not using in HarfBuzz before.
2017-11-01 17:32:34 -06:00
Behdad Esfahbod
b0e33da02d
Add BinSearchArrayOf<>
2017-10-31 20:05:37 -06:00
Behdad Esfahbod
5de83fab94
[ot] Speed up get_glyph_name()
...
get_glyph_from_name() coming soon.
2017-10-28 19:54:04 -06:00
Behdad Esfahbod
5daf3bd449
[glib/ucdn/icu/ft/ot] Make returned funcs inert
...
Such that client cannot accidentally destroy them, even though that
will be a bug in their code...
2017-10-27 16:37:11 -06:00
Behdad Esfahbod
543bd58d7a
[post] Touch up and fixes
...
The get_glyph_from_name() is incorrect; needs fixing.
2017-10-25 18:01:03 -06:00