shadowfacts/fonttools
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
10,538 Commits 1 Branch 0 Tags
Commit Graph

204 Commits

Author SHA1 Message Date
Cosimo Lupo
d9f9acd5c2
copy OTTableReader before copying/pickling lazily loaded table 
Fixes https://github.com/fonttools/fonttools/issues/2965
 2023-01-26 16:52:13 +00:00
Nikolaus Waxweiler
d584daa8fd Blacken code 2022-12-13 11:26:36 +00:00
Garret Rieger
a33972da62 Use new uharfbuzz repack_with_tag if available. 2022-07-29 21:48:25 +00:00
Cosimo Lupo
32aa8eafcb [otBase] fix AttributeError when uharfbuzz < 0.23.0 and 'repack' missing 
Fixes #2671
 2022-07-19 14:46:50 +02:00
Cosimo Lupo
2a07518b70
Merge pull request #2660 from fonttools/variable-colr 
[varLib] add support for building variable COLR from set of master COLRv1 tables
 2022-07-06 10:55:50 +01:00
Garret Rieger
10140f9df3 Remove TODO. 2022-06-27 18:32:04 +00:00
Garret Rieger
8196166e53 Add some additional comments about packing state machine. 2022-06-27 18:28:02 +00:00
Garret Rieger
d92fd2665c Use enum's for repacker state. 2022-06-27 18:18:20 +00:00
Cosimo Lupo
fcd2a67f52 Merge remote-tracking branch 'origin/main' into variable-colr 2022-06-24 10:16:51 +01:00
Garret Rieger
c63b84db7e Update how table packing falls back to fontTools from harfbuzz. 
Introduce three compilation modes:
1. Pure python: only uses the existing fonttools packing and overflow resolution.
2. Harfbuzz+python: uses harfbuzz packing and python overflow resolution. Extensions are allowed to be shared.
3. Python fallback: if harfbuzz+python runs out of resolution options, this disables extension sharing and only uses python packing. Once it succeeds control is passed back to the harfbuzz packer to produce the final packing with extension sharing enabled.
 2022-06-23 19:39:56 +00:00
Cosimo Lupo
0e6a4ec581 [otBase] temporarily disable extension sharing during harfbuzz repacking 
If extension sharing is enabled some fonts may fail to repack due to fontTools repacker getting stuck in a unrecoverable error.

https://github.com/fonttools/fonttools/issues/2661
 2022-06-23 16:34:12 +01:00
Cosimo Lupo
ca2f9e0a94 otBase: use OptionalValue.DEFAULT in populateDefaults() 2022-06-21 17:27:12 +01:00
Cosimo Lupo
97bcda9d35 otBase: define method to get variable attrs with associated VarIndexBase + {offset} 2022-06-21 17:27:11 +01:00
Cosimo Lupo
0cbdc8536b [otBase/varLib.merger] don't recurse in ensureDecompiled, merger is already working recursively 2022-06-21 17:27:11 +01:00
Garret Rieger
086799357a Enable sharing across subtables of extension for harfbuzz packing. 
Harfbuzz repacker can handle subtables shared across extension lookups so there's no need to disallow such sharing. Add a parameter to _doneWriting to allow sharing to be enabled when preparing the object graph for passing to harfbuzz.
 2022-05-25 23:55:43 +00:00
Cosimo Lupo
350f893e17 [otBase] demote repacker ERROR to WARNING, only 1 per loop; don't exit at firstfail 
Fixes https://github.com/fonttools/fonttools/issues/2594
 2022-04-22 19:05:54 +01:00
Cosimo Lupo
f68ce14d7a [otBase] log the name of uharfbuzz exception 2022-04-22 15:31:34 +01:00
Cosimo Lupo
af6804bed5 make USE_HARFBUZZ_REPACKER a 3-state option, defaults to auto 
if explicitly enabled, it will raise ImportError if uharfbuzz is not found, and will propagate the uharfbuzz error instead of silently falling back to the pure-python serializer
 2022-04-21 18:11:20 +01:00
Cosimo Lupo
c204537073 [otBase] check USE_HARFBUZZ_REPACKER in TTFont.cfg, add more logging 2022-04-21 15:46:12 +01:00
Qunxin Liu
fe196bf88e Use Harfbuzz Repacker for packing GSUB/GPOS tables 2022-04-14 10:42:50 -07:00
Cosimo Lupo
25746a3600 add clarifying comment as per review 2022-03-18 09:23:07 +00:00
Cosimo Lupo
ab8fc321a7 otBase: also return name and index from iterSubTables 
makes it more useful for constructing generic traversals of trees of otTables
 2022-03-17 17:47:41 +00:00
Cosimo Lupo
43d2ee2822 Revert "return self so one can do font=TTFont(file).ensureDecompiled()" 
This reverts commit 61e7b294483cbe729b1437804406b4571ae57cfb.
 2022-03-17 15:55:20 +00:00
Cosimo Lupo
61e7b29448 return self so one can do font=TTFont(file).ensureDecompiled() 2022-03-17 15:44:02 +00:00
Cosimo Lupo
ee27b73d7c BaseTTXConverter: add ensureDecompiled method 
to unlazify a whole tree of otTables, recursively
 2022-03-17 12:54:22 +00:00
Cosimo Lupo
64dc37fc01 otBase: add iterSubTables method to iterate over all BaseTables 
can be useful to traverse a tree of otTables
 2022-03-17 12:53:06 +00:00
Simon Cozens
c04639af70 Add comment to demystify deduplication 2021-12-06 11:15:12 +00:00
Simon Cozens
00f37ea6b6
[docs] Split table docs into individual pages (#2444) 2021-11-18 09:01:27 +00:00
Khaled Hosny
232b2ccbc4 Move the rest of py23 module to textTools 
Change all imports to use textTools module, except the test_py23.py test
which is kept until we decide to remove the module (if ever).
 2021-08-20 01:29:45 +02:00
Behdad Esfahbod
7af43123d4 [otBase] Tiniest cleanup 2021-05-24 14:52:34 -06:00
Behdad Esfahbod
f21091b6ff [otBase] Assert array.array('i').itemsize 2021-05-07 15:52:35 -06:00
Behdad Esfahbod
b8963256fd [otBase/otConverters] Add back array index to exception in writeArray() 
db6171df14 (r628543432)
 2021-05-07 15:31:53 -06:00
justvanrossum
747f9f49b7 fix struct vs array mismatch: for array 'l' is 8 bytes, not 4. I'm not 100% sure this is correct for all platforms. 2021-05-04 18:54:26 +02:00
Behdad Esfahbod
db6171df14 [otBase] Actually call conv.writeArray() 
Huh. Somehow the writeArray() was never wired up. We lose the failing
array index in the exception, but is fine to me.
 2021-04-30 16:28:28 -06:00
Behdad Esfahbod
81ca053039 [otBase] Add array reader/writer for UInt24 2021-04-30 16:28:08 -06:00
Behdad Esfahbod
36dd271cd5 [otBase/otConverters] Add array-writers for int types 2021-04-30 15:58:45 -06:00
Behdad Esfahbod
1fd2a44bbf [otBase] fix array-reader to return list, not array.array 
Was not noticed because it was for the most part unused.
 2021-04-30 15:57:53 -06:00
Behdad Esfahbod
df0d6bbadf [otBase] Add array version of various int readers 2021-04-30 15:26:32 -06:00
Behdad Esfahbod
9959916c64
Merge pull request #2275 from fonttools/subset-device 
[subset] Fix hint-dropping
 2021-04-22 12:37:48 -06:00
Behdad Esfahbod
830cff6253 [ttLib] Add Value.getEffectiveFormat() 
This ignores 0/None values that don't need to be encoded.
 2021-04-21 14:10:27 -06:00
Garret Rieger
e0f1c15ad0 Include record size adjustment. 2021-04-21 12:15:29 -07:00
Just van Rossum
5fc65d7168
Misc py23 cleanups (#2243) 
* Replaced all from ...py23 import * with explicit name imports, or removed completely when possible.
* Replaced tounicode() with tostr()
* Changed all BytesIO ans StringIO imports to from io import ..., replaced all UnicodeIO with StringIO.
* Replaced all unichr() with chr()
* Misc minor tweaks and fixes
 2021-03-29 11:45:58 +02:00
Behdad Esfahbod
6243a24a09 [otBase] Delete .Format if was introduced by .preWrite() 
https://github.com/fonttools/fonttools/pull/2238#issuecomment-805192631
 2021-03-23 18:13:07 -07:00
Cosimo Lupo
c7e3a692e1
otBase: support FormatSwitchingBaseTable with uint8 Format type 2020-10-08 12:39:56 +01:00
Cosimo Lupo
d60d84d3fc
otBase/otConverters: add support for 3-byte offsets 2020-10-08 12:39:54 +01:00
Cosimo Lupo
690080b14d
use 32-bit GlyphIDs in COLRv1 baseGlyphs and layers 
For forward-compatibility, see https://github.com/googlefonts/colr-gradients-spec/issues/8
 2020-07-01 19:50:48 +01:00
Cosimo Lupo
b32a1072b5
otBase: propagate pre-initialized CountReferences so nested subtable see them 2020-02-04 17:14:12 +00:00
Cosimo Lupo
ebadcd96e6
set pre-initialized CountReference in VarRegionList.preWrite with fvar.axisCount 
so we can reuse CountReference class, as suggested in https://github.com/fonttools/fonttools/pull/1752#issuecomment-547113944

The patch is no shorter though.
 2019-10-29 12:52:42 +00:00
Cosimo Lupo
e8f5bbcc6a
renamed LiteralCount to StaticCount 
makes it clearer what it is
 2019-10-24 15:42:09 +01:00
Cosimo Lupo
adc5b2997e
[otBase|otTables] enforce VarStore RegionAxisCount == fvar.AxisCount 
even when there are no Regions and thus we can't take the length of VarRegionAxis array.
This is to appease older versions of OTS which blindly enforce this rule and
reject a VF that has, e.g., an empty HVAR table with no regions if the
HVAR.VarStore.VarRegionList.RegionAxisCount != fvar.AxisCount.

Fixes https://github.com/fonttools/fonttools/issues/1670

Related:
https://github.com/fonttools/fonttools/pull/1671
https://github.com/googlefonts/fontmake/issues/565
https://github.com/khaledhosny/ots/pull/192
 2019-10-24 12:54:16 +01:00