shadowfacts/fonttools
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
4,061 Commits 1 Branch 0 Tags
Commit Graph

4061 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Masayuki Takeda
c4fcb2da3d add test font files to MANIFEST 2017-05-22 11:53:19 +09:00
Behdad Esfahbod
81498e5baf [varLib] Fix bug with recombining PairPosClass2 subtables that don't 
Ouch.  I have no idea why I made two separate if's to begin with, but the
continue should have also been aligned with the inner block...

Fixes root cause of https://github.com/fonttools/fonttools/issues/914
but now we hit an unimplemented feature there...
 2017-05-20 23:03:25 -07:00
Behdad Esfahbod
462b7f8683 [cffLib.specializer] When copying iterables, cast to list specifically 
See 41445b8449
 2017-05-20 18:34:20 -07:00
Cosimo Lupo
0ff0d022e0
Bump version: 3.12.1 → 3.12.2.dev0 2017-05-18 11:38:34 +01:00
Cosimo Lupo
7545539bf5
Release 3.12.1 2017-05-18 11:37:31 +01:00
Behdad Esfahbod
7ade0096de [pens.t2CharStringPen] Call BasePen constructor 
Fixes https://github.com/fonttools/fonttools/issues/965
 2017-05-17 17:06:00 -07:00
Cosimo Lupo
1ced13124a
Bump version: 3.12.0 → 3.12.1.dev0 2017-05-17 15:16:33 +01:00
Cosimo Lupo
f798f52236
Release 3.12.0 2017-05-17 15:15:16 +01:00
Cosimo Lupo
6accfa1776
Update changelog 2017-05-17 15:14:48 +01:00
Cosimo Lupo
faa899f9e9
[specializer_test] temporarily disable failing tests 
Let's appease the bots for now.
Later on we can add tests for invalid args number for each operator so specializer.py module can reach 100 coverage.
 2017-05-16 23:01:00 +01:00
Cosimo Lupo
ced61f787c Merge pull request #964 from anthrotype/coords-as-double 
[_g_l_y_f] fix float precision loss in GlyphCoordinates
 2017-05-16 17:16:15 +01:00
Cosimo Lupo
71486bc0a9
[_g_l_y_f] use 'd' instead of 'f' as array.array typecode in GlyphCoordinates 
Fixes #963
 2017-05-16 10:50:56 +01:00
Cosimo Lupo
1c9198e3f0
[_g_l_y_f] add test for GlyphCoordinates' float precision loss issue (#964) 2017-05-16 10:34:37 +01:00
Behdad Esfahbod
4fd719159a [cffLib.specializer] Enforce minimum number of arguments in all ops 2017-05-08 14:09:02 -06:00
Behdad Esfahbod
0cbe43483d [cffLib.specializer] Fix width-extraction code when stack is empty 2017-05-08 14:06:26 -06:00
Behdad Esfahbod
a3743828ce [cffLib.specializer] Default ignoreErrors to False 
Ouch.  I meant False.  I think I changed it to True for testing and accidentally committed it.
 2017-05-08 14:05:10 -06:00
Behdad Esfahbod
a71639f8b6 [cffLib.specializer] Add minor TODO items 2017-05-08 13:01:43 -06:00
Cosimo Lupo
41445b8449
t2CharStringPen: specializeCommands expects command args to be list, not tuples 
In the docstring of programToCommands, it says that:

> Each command is a two-tuple of commandname,arg-list

Previously the T2CharStringPen was passing command args as tuples instead
of lists to the specializeCommands function with option generalizeFirst=False,
which would only make a shallow copy of the input commands to modify them
in place. The problem is that it attempted to call list-only methods, leading
to errors like:

File "fontTools/cffLib/specializer.py", line 432, in specializeCommands
    args.insert(pos, 0)
AttributeError: 'tuple' object has no attribute 'insert'

Since the expectation of the code here and elsewhere is that args is a
list, it makes sense that the T2 pen passes lists instead of tuples to the
specializeCommands function.
 2017-05-08 10:51:39 +01:00
Behdad Esfahbod
30b804003e [cffLib.specializer] Add programToString and stringToProgram from _test.py 2017-05-07 22:12:19 -06:00
Behdad Esfahbod
a8b841a4a9 [cffLib.specizalier_test] Add a few more tests 2017-05-07 22:04:17 -06:00
Behdad Esfahbod
041384a0e8 [cffLib.specializer_test] Add tests for peephole optimization 2017-05-07 13:08:41 -06:00
Behdad Esfahbod
3720c6e020 [cffLib.specializer] Fix combining of more ops into rlinecurve / rcurveline 
Fixes failing tests.
 2017-05-07 12:54:31 -06:00
Miguel Sousa
2f14200a89 [cffLib.specializer_test] specializeProgram tests 
The failing tests are due to missed optimization opportunities.
The current tests are not hitting some edge cases.
 2017-05-07 03:14:24 -07:00
Miguel Sousa
2f7f36e2f4 [cffLib.specializer_test] charstring extras 2017-05-07 00:23:12 -07:00
Miguel Sousa
3f9eeec58d [cffLib.specializer] Fix DeprecationWarnings in py3 2017-05-06 23:20:11 -07:00
Miguel Sousa
9617f5f26b [cffLib.specializer_test] generalizeProgram tests 2017-05-06 22:57:37 -07:00
Behdad Esfahbod
c017063d40 Merge pull request #948 from fonttools/cffOpSpecializer 
CFF operator specializer
 2017-05-06 17:55:10 -06:00
Behdad Esfahbod
8432c04bb1 [cffLib.specializer] Remove -3 hack that was to match AFDKO 2017-05-06 17:54:05 -06:00
Behdad Esfahbod
2fd2e4a7cf [cffLib.specializer] Add TODO item 2017-05-06 14:14:54 -06:00
Behdad Esfahbod
0e94f909a6 [pens.t2CharStringPen] Disallow glyph width in CFF2 2017-05-06 13:16:07 -06:00
Behdad Esfahbod
2abacb9366 [pens.t2CharString] Bump CFF2 stack size from 193 to 513 
That's happening in the next version of OpenType spec (and maxstack
operator deprecated.)
 2017-05-06 13:14:25 -06:00
Behdad Esfahbod
0bf5044d12 [cffLib.specializer] Implement width extraction in programToCommands() 
Test:

./fonttools cffLib.specializer 7 0 8 rmoveto 1 2 3 4 5 0 rrcurvet
 2017-05-06 13:12:58 -06:00
Behdad Esfahbod
bef46993f6 [cffLib.specializer] Use bool() cast instead of comparison to 0 
Makes it work on custom types better.
 2017-05-06 05:02:22 -06:00
Behdad Esfahbod
d1d36ce66b [cffLib.specializer] Minor 2017-05-06 04:58:59 -06:00
Behdad Esfahbod
4b5c1be29b [pens.t2CharStringPen] Add CFF2 mode 2017-05-06 04:54:07 -06:00
Behdad Esfahbod
72ec474f74 [pens.t2CharStringPen] Simplify drastically 2017-05-06 04:50:24 -06:00
Behdad Esfahbod
d3947b86d2 [cffLib.specializer] Fix bug introduced in c9653e98c6da7770a13faebd4a218f5d1d362536 
Test case (which apparently is not covered by our current tests!):

./fonttools cffLib.specializer 1 2 3 4 5 0 rrcurveto
 2017-05-06 04:28:45 -06:00
Behdad Esfahbod
050bcc0c8a [cffLib.specializer] Minor adjustments to previous commit 2017-05-06 03:49:06 -06:00
Behdad Esfahbod
c9653e98c6 [cffLib.specializer] Fix curve merging logic! 
Found that elegant model I've been looking for, yay!  My previous modeling
was overly complicating the problem, and failing at solving all corner cases.
New one should fix them all!
 2017-05-06 03:23:00 -06:00
Behdad Esfahbod
e35976900a [cffLib.specializer] Fix typo in peephole optimization 2017-05-06 03:00:39 -06:00
Behdad Esfahbod
d8fcb3a12a [cffLib.specializer] Fix orientation propagation through a pizza spline 2017-05-06 02:06:58 -06:00
Behdad Esfahbod
a5ca6c49f2 [cffLib.specializer] Fix hvcurveto/vhcurveto argument swapping logic 
Fixes remaining 2 failing tests.
 2017-05-06 01:44:19 -06:00
Behdad Esfahbod
df7f36e234 [cffLib.specializer] Fix logic for choosing vhcurveto vs hvcurveto 
Fixes 2 of the 4 failing tests.
 2017-05-06 01:41:07 -06:00
Behdad Esfahbod
b81250b10d [cffLib.specializer] Fix typo 2017-05-06 01:25:57 -06:00
Miguel Sousa
2d02c54d1a [t2CharStringPen_test] Update test results 2017-05-06 00:11:27 -07:00
Behdad Esfahbod
c8cf233e26 [pens.t2CharStringPen] Hook up to cffLib.specializer 
Tests are failing now, as cffLib.specializer combines operations whereas
previous pen didn't.  I checked all failures and they all look like the
expectation needs to be updated to me.

Fixes https://github.com/fonttools/fonttools/issues/403
 2017-05-05 21:21:30 -06:00
Behdad Esfahbod
c3a94fd4e5 [cffLib.specializer] Finish specializing curves 
This is complete now.  Unhooked and lightly tested.
 2017-05-05 21:07:43 -06:00
Behdad Esfahbod
86a398b5c3 [cffLib.specializer] Implement rlinecurve and rcurveline 2017-05-05 17:05:04 -06:00
Behdad Esfahbod
a3c1ebbfce [cffLib.specializer] Make progress on specialization 
The core is in.  The real fun is not yet.
 2017-05-05 16:56:00 -06:00
Behdad Esfahbod
f32ad88953 [cffLib.specializer] Making progress on specialization 2017-05-05 16:56:00 -06:00