Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
<?xml version="1.0" encoding="UTF-8"?>
|
2019-05-13 13:12:46 +01:00
|
|
|
<ttFont sfntVersion="OTTO" ttLibVersion="3.41">
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
|
|
|
|
|
<fvar>
|
|
|
|
|
|
|
|
|
|
<!-- Weight -->
|
|
|
|
|
<Axis>
|
|
|
|
|
<AxisTag>wght</AxisTag>
|
|
|
|
|
<Flags>0x0</Flags>
|
|
|
|
|
<MinValue>200.0</MinValue>
|
|
|
|
|
<DefaultValue>200.0</DefaultValue>
|
|
|
|
|
<MaxValue>900.0</MaxValue>
|
|
|
|
|
<AxisNameID>256</AxisNameID>
|
|
|
|
|
</Axis>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w0.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w0.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="258" subfamilyNameID="257">
|
|
|
|
|
<coord axis="wght" value="200.0"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w239.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w239.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="260" subfamilyNameID="259">
|
|
|
|
|
<coord axis="wght" value="324.6875"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w240.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w240.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="262" subfamilyNameID="261">
|
|
|
|
|
<coord axis="wght" value="325.0"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w439.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w439.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="264" subfamilyNameID="263">
|
|
|
|
|
<coord axis="wght" value="428.82353"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w440.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w440.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="266" subfamilyNameID="265">
|
|
|
|
|
<coord axis="wght" value="429.41176"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w499.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w499.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="268" subfamilyNameID="267">
|
|
|
|
|
<coord axis="wght" value="464.11765"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w500.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w500.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="270" subfamilyNameID="269">
|
|
|
|
|
<coord axis="wght" value="464.70589"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w599.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w599.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="272" subfamilyNameID="271">
|
|
|
|
|
<coord axis="wght" value="535.45454"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w600.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w600.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="274" subfamilyNameID="273">
|
|
|
|
|
<coord axis="wght" value="536.36363"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w669.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w669.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="276" subfamilyNameID="275">
|
|
|
|
|
<coord axis="wght" value="599.09091"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w670.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w670.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="278" subfamilyNameID="277">
|
|
|
|
|
<coord axis="wght" value="600.0"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w699.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w699.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="280" subfamilyNameID="279">
|
|
|
|
|
<coord axis="wght" value="626.36363"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w700.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w700.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="282" subfamilyNameID="281">
|
|
|
|
|
<coord axis="wght" value="627.27272"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-799.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-799.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="284" subfamilyNameID="283">
|
|
|
|
|
<coord axis="wght" value="717.27272"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w800.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w800.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="286" subfamilyNameID="285">
|
|
|
|
|
<coord axis="wght" value="718.18182"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w889.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w889.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="288" subfamilyNameID="287">
|
|
|
|
|
<coord axis="wght" value="799.09091"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w890.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w890.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="290" subfamilyNameID="289">
|
|
|
|
|
<coord axis="wght" value="800.0"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
|
|
|
|
|
<!-- Kanji-w1000.00 -->
|
|
|
|
|
<!-- PostScript: SHSansJPVFTest-Kanji-w1000.00 -->
|
|
|
|
|
<NamedInstance flags="0x0" postscriptNameID="292" subfamilyNameID="291">
|
|
|
|
|
<coord axis="wght" value="900.0"/>
|
|
|
|
|
</NamedInstance>
|
|
|
|
|
</fvar>
|
|
|
|
|
|
|
|
|
|
<CFF2>
|
|
|
|
|
<major value="2"/>
|
|
|
|
|
<minor value="0"/>
|
|
|
|
|
<CFFFont name="CFF2Font">
|
|
|
|
|
<FontMatrix value="0.001 0 0 0.001 0 0"/>
|
|
|
|
|
<FDSelect format="3"/>
|
|
|
|
|
<FDArray>
|
|
|
|
|
<FontDict index="0">
|
|
|
|
|
<Private>
|
|
|
|
|
<BlueValues value="-250 -250 1100 1100"/>
|
2019-06-24 17:10:36 -07:00
|
|
|
<FamilyBlues value="-250 -240 1100 1110"/>
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
<BlueScale value="0.039625"/>
|
|
|
|
|
<BlueShift value="7"/>
|
|
|
|
|
<BlueFuzz value="0"/>
|
|
|
|
|
<StdHW value="1"/>
|
|
|
|
|
<StdVW value="1"/>
|
2021-03-30 21:24:14 +09:00
|
|
|
<LanguageGroup value="1"/>
|
|
|
|
|
<ExpansionFactor value="0.06"/>
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</Private>
|
|
|
|
|
</FontDict>
|
|
|
|
|
<FontDict index="1">
|
|
|
|
|
<Private>
|
|
|
|
|
<BlueValues value="-250 -250 1100 1100"/>
|
|
|
|
|
<BlueScale value="0.039625"/>
|
|
|
|
|
<BlueShift value="7"/>
|
|
|
|
|
<BlueFuzz value="0"/>
|
|
|
|
|
<StdHW value="1"/>
|
|
|
|
|
<StdVW value="1"/>
|
2021-03-30 21:24:14 +09:00
|
|
|
<LanguageGroup value="1"/>
|
|
|
|
|
<ExpansionFactor value="0.06"/>
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</Private>
|
|
|
|
|
</FontDict>
|
|
|
|
|
<FontDict index="2">
|
|
|
|
|
<Private>
|
|
|
|
|
<BlueValues value="0 0"/>
|
|
|
|
|
<BlueScale value="0.039625"/>
|
|
|
|
|
<BlueShift value="7"/>
|
|
|
|
|
<BlueFuzz value="1"/>
|
2021-03-30 21:24:14 +09:00
|
|
|
<LanguageGroup value="0"/>
|
|
|
|
|
<ExpansionFactor value="0.06"/>
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</Private>
|
|
|
|
|
</FontDict>
|
|
|
|
|
</FDArray>
|
|
|
|
|
<CharStrings>
|
|
|
|
|
<CharString name=".notdef" fdSelectIndex="0">
|
|
|
|
|
-120 50 859 91 -50 50 hstemhm
|
|
|
|
|
100 50 700 50 vstemhm
|
|
|
|
|
hintmask 10111000
|
|
|
|
|
100 -120 rmoveto
|
|
|
|
|
800 1000 -800 hlineto
|
|
|
|
|
400 -459 rmoveto
|
|
|
|
|
-318 409 rlineto
|
|
|
|
|
636 hlineto
|
|
|
|
|
-286 -450 rmoveto
|
|
|
|
|
hintmask 11011000
|
|
|
|
|
318 409 rlineto
|
|
|
|
|
-818 vlineto
|
|
|
|
|
-668 -41 rmoveto
|
|
|
|
|
318 409 318 -409 rlineto
|
|
|
|
|
-668 859 rmoveto
|
|
|
|
|
318 -409 -318 -409 rlineto
|
|
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid00001" fdSelectIndex="2">
|
|
|
|
|
</CharString>
|
|
|
|
|
<CharString name="cid00002" fdSelectIndex="2">
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
-12 83 1 126 2 blend
|
|
|
|
|
hstemhm
|
|
|
|
|
74 73 -57 40 -26 129 -125 122 4 blend
|
|
|
|
|
vstemhm
|
|
|
|
|
hintmask 10100000
|
|
|
|
|
134 755 107 18 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-48 -137 1 blend
|
|
|
|
|
hlineto
|
|
|
|
|
8 -565 16 50 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
32 106 1 blend
|
|
|
|
|
hlineto
|
|
|
|
|
hintmask 11000000
|
|
|
|
|
-17 -202 -52 -67 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
24 14 18 22 25 -14 18 -22 -23 -14 -21 -24 -18 13 -20 22 38 25 26 38 37 -28 25 -37 -34 -30 -22 -38 -40 27 -26 39 16 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid01177" fdSelectIndex="1">
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
1 vsindex
|
|
|
|
|
-72 30 253 30 94 30 92 30 65 30 131 45 -30 112 -99 17 -8 0 -12 59 -2 85 -64 2 -92 39 -1 56 -44 2 -63 35 -1 51 -43 1 -62 39 -1 56 -21 0 -31 56 -2 81 -56 2 -81 44 -2 63 -57 3 -81 31 -1 45 -18 0 -27 44 -2 63 16 blend
|
|
|
|
|
hstemhm
|
|
|
|
|
193 30 83 30 147 30 173 30 66 30 -20 1 -28 75 -3 107 -74 3 -106 78 -4 111 -99 5 -141 81 -3 116 -79 2 -114 64 -2 92 -81 4 -115 80 -4 114 10 blend
|
|
|
|
|
vstemhm
|
|
|
|
|
hintmask 1111100011111000
|
|
|
|
|
306 142 -19 1 -27 7 0 10 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-156 45 -2 64 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-50 22 -8 79 -42 2 -59 8 -1 11 -18 0 -27 43 -1 62 4 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
17 186 8 -1 11 -68 3 -97 2 blend
|
|
|
|
|
0 18 8 0 12 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
70 13 25 114 5 22 -1 31 17 -1 24 2 1 4 0 -1 -1 7 0 10 5 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
-9 3 -12 4 -9 6 -20 1 -28 3 0 4 -31 1 -45 10 0 15 -13 0 -19 9 -1 13 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
-109 -4 -7 -13 -49 -38 -156 33 -1 47 -1 0 -1 1 0 1 2 0 3 10 0 15 9 0 13 60 -3 85 7 blend
|
|
|
|
|
0 -27 5 0 8 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-59 -10 5 22 11 0 16 2 -1 2 -1 0 -2 4 0 6 4 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
157 -47 2 -67 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
63 34 -74 3 -106 -25 1 -36 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
65 -30 74 -47 37 -37 -7 1 -10 5 -1 7 -3 0 -4 5 0 7 -3 0 -4 6 0 9 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
20 22 -37 36 -75 47 -65 28 48 -2 68 47 -2 67 -1 0 -1 -6 1 -8 2 0 3 -8 0 -11 8 -1 11 -6 0 -9 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
320 -64 -49 3 -69 -32 1 -46 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
76 -49 83 -75 38 -12 0 -18 -6 1 -8 -13 0 -19 -4 0 -6 -9 1 -12 5 blend
|
|
|
|
|
-54 rrcurveto
|
|
|
|
|
23 19 -38 54 -84 73 -76 49 69 -3 98 35 -2 50 5 0 8 2 0 3 11 0 16 2 0 2 13 -1 18 2 0 4 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-557 -5 -85 4 -121 -6 0 -9 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-28 -68 -50 -72 -77 -40 5 -1 6 1 0 1 3 1 5 6 0 9 13 -1 18 1 0 1 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
24 -17 79 42 47 74 31 71 62 -3 89 -42 2 -60 -7 1 -10 5 -1 7 -6 0 -8 1 0 1 -2 -1 -4 4 1 7 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-117 625 -26 2 -36 42 -3 59 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 775 30 -57 3 -81 -3 0 -5 57 -3 81 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-818 -176 -1 0 -2 12 0 18 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 869 30 -56 2 -81 3 0 5 56 -2 81 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 0000001000100000
|
|
|
|
|
-455 258 -40 2 -57 -38 2 -54 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
hintmask 0000000100100000
|
|
|
|
|
-99 30 -18 0 -27 81 -3 116 2 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 0000001000100000
|
|
|
|
|
99 18 0 27 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 0111010010001000
|
|
|
|
|
-236 -127 -60 2 -86 -18 0 -27 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
26 -40 25 -53 9 -36 -12 1 -17 10 0 15 -10 -1 -15 13 0 19 -4 0 -6 11 -1 15 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
29 12 -10 35 -25 53 -27 39 76 -3 109 12 0 18 3 1 5 -10 0 -15 10 -1 14 -15 1 -21 11 0 16 -11 0 -16 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
393 2 -112 4 -161 -1 0 -2 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-16 -38 -31 -57 -23 -35 7 0 11 12 -1 17 13 -1 18 19 0 28 10 0 14 8 -1 11 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
27 -12 24 36 26 48 23 46 70 -2 101 -10 1 -14 -8 0 -12 -13 1 -18 -6 -1 -9 -17 0 -25 -1 1 -1 -10 1 -14 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-504 -378 27 -1 39 -8 0 -12 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
559 -94 -559 -110 5 -157 44 -2 63 110 -5 157 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
216 -52 2 -74 1 blend
|
|
|
|
|
vmoveto
|
|
|
|
|
559 -92 -559 -110 5 -157 43 -1 62 110 -5 157 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 122 -75 3 -107 -4 0 -6 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-276 619 276 -26 0 -38 45 -2 64 26 0 38 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid06449" fdSelectIndex="1">
|
|
|
|
|
2 vsindex
|
2021-03-04 13:01:09 -07:00
|
|
|
-60 30 203 30 -9 9 67 7 -7 14 -14 30 -20 20 80 30 59 30 121 30 18 93 -30 30 -30 108 -23 0 -26 67 2 76 -98 -2 -111 42 0 47 -13 0 -14 13 0 14 -33 0 -37 11 0 13 -11 0 -13 8 0 8 -8 0 -8 53 0 60 -32 0 -36 32 0 36 -52 0 -59 57 1 65 -33 0 -38 53 0 60 -83 -1 -93 54 0 60 -6 -19 -24 33 19 55 -76 -1 -86 76 1 86 -76 -1 -86 59 1 67 26 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
77 30 42 30 139 30 23 30 71 10 74 30 15 30 16 30 158 30 28 30 -4 29 -14 0 -16 88 1 99 -82 -1 -92 87 1 98 -130 -1 -146 102 1 114 -73 -1 -82 74 2 84 -112 -2 -126 27 0 30 13 0 15 90 1 101 -126 -1 -142 75 1 84 -68 -1 -76 102 1 115 -144 -1 -162 94 1 105 -79 -1 -88 95 1 106 -81 -1 -91 74 1 83 22 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 110001011101011101101101
|
|
|
|
|
53 761 -3 0 -3 31 0 35 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 896 30 -76 -1 -86 5 0 5 76 1 86 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-802 -461 2 0 2 -23 0 -26 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 703 30 -53 0 -60 3 0 4 53 0 60 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000000100100000000
|
|
|
|
|
-532 539 -58 -1 -65 6 0 7 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000010000100000000
|
|
|
|
|
-171 30 -16 -19 -36 102 1 114 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000000100100001000
|
|
|
|
|
171 16 19 36 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
299 -100 -1 -112 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
hintmask 000000000010000000001000
|
|
|
|
|
-171 30 -16 -19 -36 102 1 115 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000000100000001000
|
|
|
|
|
171 16 19 36 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000111100011010010100
|
|
|
|
|
-46 -219 -34 0 -39 -64 -1 -72 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
204 -121 -204 -110 -1 -123 83 1 93 110 1 123 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-230 121 33 1 38 -83 -1 -93 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
200 -121 -200 -108 -2 -122 83 1 93 108 2 122 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-222 121 27 -1 30 -83 -1 -93 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
192 -121 -192 -101 -1 -114 83 1 93 101 1 114 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 151 -87 -1 -98 -29 0 -33 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-181 716 181 -24 0 -27 11 0 12 24 0 27 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-788 -240 -17 0 -19 9 0 11 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-130 30 100 -37 0 -42 88 1 99 -20 0 -23 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000110000000000000010
|
|
|
|
|
786 -100 30 130 -150 -1 -168 20 0 23 95 1 106 37 0 42 4 blend
|
|
|
|
|
hlineto
|
|
|
|
|
hintmask 000010000000000100000000
|
|
|
|
|
-610 -123 -56 -1 -63 -44 0 -50 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-50 -62 -93 -73 -118 -54 8 -4 10 -9 6 -7 9 0 11 13 0 15 19 0 21 29 0 32 9 0 10 22 0 25 12 0 14 -11 0 -12 19 0 21 -26 0 -30 7 0 8 -16 0 -18 12 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
hintmask 010000000000000001000000
|
|
|
|
|
121 58 92 75 59 70 3 0 3 -13 0 -14 -10 0 -11 -19 0 -21 -2 0 -2 8 0 8 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
124 -78 -89 -1 -100 32 0 36 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-7 -6 0 -6 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-65 -139 -176 -81 -162 -31 6 -6 8 -12 3 -8 16 0 17 30 0 34 36 0 41 26 0 29 -7 0 -8 12 0 13 12 0 14 -16 0 -18 15 0 16 -30 0 -33 5 0 6 -18 0 -21 12 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
hintmask 001000000000000001000000
|
|
|
|
|
168 37 178 84 72 154 26 0 29 -5 0 -5 -23 0 -26 -12 0 -14 -5 0 -5 6 0 7 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
hintmask 110100000000000001100001
|
|
|
|
|
-19 11 -6 -2 -47 0 -53 13 0 15 -13 0 -15 0 0 -1 4 blend
|
|
|
|
|
rlineto
|
|
|
|
|
-333 -72 75 1 85 -55 0 -61 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
65 -25 75 -46 38 -35 -35 0 -40 8 0 9 -38 0 -42 15 0 17 -18 0 -21 14 0 15 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
26 19 -39 34 -76 45 -64 25 49 0 56 31 0 35 19 0 21 -14 0 -16 39 0 44 -18 0 -20 32 0 36 -9 0 -10 8 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rlinecurve
|
2019-05-13 13:12:46 +01:00
|
|
|
72 55 -55 0 -62 28 0 31 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 -30 -42 0 -47 -42 0 -47 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
269 30 -14 0 -16 42 0 47 2 blend
|
|
|
|
|
hlineto
|
|
|
|
|
74 74 13 0 15 -22 0 -24 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-276 80 1 90 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-52 21 -9 77 -48 0 -54 8 0 9 -21 0 -24 44 0 49 4 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
16 182 8 0 9 -90 -1 -101 2 blend
|
|
|
|
|
0 18 8 0 9 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
62 12 21 88 4 25 0 28 20 0 22 6 0 7 10 0 11 9 0 10 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-9 2 -12 5 -8 6 -24 0 -26 4 0 5 -34 0 -39 12 0 13 -16 0 -18 10 0 11 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-81 25 0 28 1 blend
|
|
|
|
|
-4 -6 -11 -41 -37 -154 -1 0 -1 0 1 1 11 -1 12 15 1 17 79 1 89 5 blend
|
|
|
|
|
0 -26 9 0 10 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-56 -9 6 25 17 0 19 2 0 3 -1 -1 -2 4 0 5 4 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
276 -81 -1 -91 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
278 -62 -114 -1 -128 32 0 36 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-66 -32 -126 -33 -107 -23 5 -7 5 -10 2 -7 110 22 126 32 81 36 10 0 11 7 0 8 30 0 34 11 0 12 21 0 23 9 0 10 7 0 8 -14 0 -16 9 0 10 -27 0 -30 3 0 4 -15 0 -17 -15 0 -17 -10 0 -11 -12 0 -14 -11 0 -12 3 0 4 -3 0 -4 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
|
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid06821" fdSelectIndex="1">
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
3 vsindex
|
2019-05-13 13:12:46 +01:00
|
|
|
-58 30 100 30 70 22 -22 30 94 30 19 31 -17 28 152 20 -20 30 -12 12 66 30 -30 89 -5 30 -30 121 -11 0 -24 36 0 81 -32 0 -74 22 0 52 -17 0 -39 16 1 37 -16 -1 -37 21 0 48 -27 0 -63 21 0 49 -11 0 -26 41 0 93 -47 0 -107 24 0 56 -34 0 -78 11 0 26 -11 0 -26 17 0 39 -15 0 -35 15 0 35 -19 0 -43 12 0 26 -12 0 -26 4 0 8 -5 0 -11 28 0 65 -28 0 -65 23 0 52 28 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
127 30 -18 18 199 30 -20 20 -20 30 -24 14 97 30 -11 11 72 31 202 30 87 29 -12 0 -27 44 1 101 -19 -1 -45 19 1 45 -46 -1 -106 37 0 85 -31 0 -71 31 0 71 -31 0 -71 40 0 91 -27 0 -62 18 0 42 -47 0 -108 51 0 117 -27 0 -62 27 0 62 -53 0 -122 43 0 99 -60 -1 -138 52 1 120 -32 0 -73 32 0 72 22 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00011000000000000000000100000000
|
|
|
|
|
193 296 41 0 93 -8 0 -19 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
625 -94 -625 -84 -1 -192 27 0 63 84 1 192 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 124 -48 0 -110 -6 0 -14 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-154 685 154 -15 0 -34 16 0 38 15 0 34 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00100000000000000000100000000000
|
|
|
|
|
-365 -132 -33 0 -76 1 1 3 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-232 -7 -1 -16 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
30 -5 51 0 117 -11 0 -27 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
237 18 1 43 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 01000000000010010000010000000000
|
|
|
|
|
-11 -92 -27 0 -62 1 -1 2 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 397 30 -22 0 -52 -12 0 -27 22 0 52 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-760 647 25 0 56 -4 0 -9 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 811 30 -28 0 -65 -12 0 -27 28 0 65 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000010100000000000000000
|
|
|
|
|
-823 -13 0 -29 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
-143 12 0 27 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-83 -13 -107 -75 -82 4 0 9 3 0 6 5 1 12 -1 0 -1 5 -1 11 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
7 -4 11 -9 5 -6 10 0 21 -5 0 -12 20 0 46 -17 0 -38 6 0 15 -8 0 -18 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
79 5 0 11 1 blend
|
|
|
|
|
85 16 118 88 1 1 3 9 0 19 6 0 15 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
143 -11 0 -25 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000010100001000000000000000
|
|
|
|
|
199 -25 -46 -1 -106 -23 0 -54 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-167 vlineto
|
|
|
|
|
hintmask 00000000010100000100000000000000
|
|
|
|
|
30 37 0 85 1 blend
|
|
|
|
|
167 hlineto
|
|
|
|
|
hintmask 00000000101000000001000000000000
|
|
|
|
|
-14 -59 -18 0 -42 8 0 18 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 185 30 -12 0 -26 -4 0 -9 12 0 26 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-365 -96 10 0 22 7 0 17 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 392 30 -17 0 -39 -4 0 -9 17 0 39 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000011000000000100000000000000
|
|
|
|
|
-218 -10 -15 0 -33 -6 0 -13 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-160 23 0 51 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-8 -2 0 0 -1 1 blend
|
|
|
|
|
-3 -11 -1 1 0 3 0 0 1 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-11 -1 -30 2 0 4 1 0 1 4 0 10 3 blend
|
|
|
|
|
0 -47 13 0 30 1 blend
|
|
|
|
|
1 5 -9 6 -10 2 -9 4 0 8 -6 0 -13 6 0 13 -11 0 -25 2 0 6 -8 0 -19 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000011000001000010001000000000
|
|
|
|
|
50 30 -5 0 -11 1 0 2 2 blend
|
|
|
|
|
0 6 17 3 0 8 5 1 12 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
17 5 4 9 21 6 -1 12 4 0 9 1 0 3 4 0 8 11 0 25 5 blend
|
|
|
|
|
vvcurveto
|
|
|
|
|
159 -21 0 -46 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-132 -50 -39 0 -88 1 0 1 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-25 -42 -40 -39 -44 -30 8 -4 13 -10 5 -4 41 6 0 12 3 0 8 7 0 16 3 0 5 5 0 13 1 0 4 6 0 13 -3 0 -8 10 0 22 -6 0 -14 5 0 12 -5 0 -10 -3 0 -8 13 blend
|
|
|
|
|
30 45 -7 0 -14 1 blend
|
|
|
|
|
47 26 45 -3 0 -8 1 0 1 2 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
153 -7 -13 0 -30 -1 0 -2 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
35 -27 38 -39 18 -28 -8 3 -11 3 -5 -3 -9 3 -14 6 -7 -3 -5 1 -9 4 -4 0 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
24 18 -18 27 -39 39 -34 25 23 1 55 6 -1 12 4 -1 8 -3 4 1 9 -3 13 -6 7 2 7 -3 9 -4 5 4 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
115 330 -53 -1 -124 9 1 21 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
hintmask 10000101000001000000001010000000
|
|
|
|
|
14 -286 131 -209 160 0 50 1 18 34 6 108 -9 3 -11 5 -9 7 -4 -92 -9 -34 -31 -1 -137 -2 -126 185 -12 281 3 0 8 6 0 14 5 0 10 -10 0 -22 -3 0 -6 0 0 -1 14 0 33 -1 0 -1 11 0 23 -3 0 -8 5 0 12 10 0 24 -10 0 -23 3 0 7 -14 0 -32 8 0 18 -8 0 -17 8 0 17 0 0 -1 11 0 26 0 0 1 4 0 9 5 0 11 1 0 1 29 0 67 0 1 2 8 0 17 0 -1 -1 -2 0 -4 -37 0 -85 30 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
207 -169 -37 0 -85 -4 0 -9 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-61 -129 -111 -108 -121 -69 7 -5 12 -11 5 -6 119 74 113 110 66 136 4 15 19 8 14 33 4 28 29 8 5 22 2 28 27 6 2 17 8 1 20 -6 0 -14 14 1 34 -13 -1 -31 6 0 15 -7 0 -17 0 -28 -23 -4 -2 -10 0 -27 -20 -1 -3 -5 -1 -16 -12 -1 -15 -19 18 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
-156 153 -20 -2 -49 -2 0 -3 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
52 -15 63 -26 34 -1 0 -3 -1 0 -1 0 0 1 0 0 -2 0 0 -2 5 blend
|
|
|
|
|
-21 rrcurveto
|
|
|
|
|
15 27 -34 20 -64 24 -51 14 21 0 48 20 0 47 -1 0 -1 1 0 1 0 0 -1 0 0 1 1 0 3 -1 0 -2 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-453 -763 1 0 2 12 0 27 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-25 -16 -31 0 -71 -7 0 -17 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
-100 89 146 -18 233 -21 0 -46 -5 0 -12 -13 0 -29 -4 0 -9 -8 0 -18 5 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
249 23 0 53 1 blend
|
|
|
|
|
hlineto
|
|
|
|
|
2 8 6 14 6 8 -35 0 -207 2 -1 3 11 0 25 5 1 12 17 0 38 4 0 10 8 0 18 -16 0 -37 -1 0 -3 -1 0 -2 9 blend
|
|
|
|
|
0 -22 -14 0 -32 1 blend
|
|
|
|
|
0 -214 0 -150 15 -78 89 24 0 55 0 0 1 18 0 40 -2 0 -5 12 0 28 -1 0 -2 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
5 62 -50 0 -114 -10 0 -22 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 -97 -92 -60 -107 -36 8 -6 12 -11 4 -6 105 41 99 65 32 106 5 0 12 7 0 15 15 0 34 1 0 3 7 0 16 1 0 2 10 0 22 -6 0 -15 18 0 41 -17 0 -37 8 0 18 -8 0 -19 -2 0 -5 4 0 9 -12 0 -27 7 0 16 -1 0 -2 6 0 14 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
|
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid07253" fdSelectIndex="1">
|
|
|
|
|
1 vsindex
|
|
|
|
|
-80 27 95 49 -48 48 -45 45 -30 30 -16 16 -13 13 49 30 48 30 47 19 -19 30 53 30 -18 18 51 11 -11 30 -22 22 62 30 60 30 15 81 -30 30 -30 102 -10 1 -14 41 -2 59 -53 2 -76 27 -1 38 -26 1 -37 26 -1 37 -27 1 -39 27 -1 39 -27 1 -39 27 -1 39 -13 0 -19 13 0 19 -14 0 -20 14 0 20 -19 1 -27 13 -1 19 -18 1 -26 13 0 19 -18 0 -26 18 0 26 -18 0 -26 23 -1 33 -21 1 -30 42 -2 60 -29 1 -42 29 -1 42 -19 1 -27 7 0 10 -7 0 -10 26 -1 37 -24 1 -34 24 -1 34 -27 1 -39 24 -1 34 -26 1 -37 26 -1 37 -40 1 -45 53 -2 66 -44 2 -62 44 -2 62 -44 2 -62 18 0 23 42 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
193 30 -1 30 -15 15 106 29 96 30 142 30 109 30 5 10 -28 1 -40 71 -2 102 -56 2 -80 75 -4 106 -21 2 -29 21 -2 29 -104 5 -148 55 -3 78 -42 3 -59 69 -4 98 -84 4 -120 79 -3 113 -94 3 -135 76 -3 109 -51 2 -73 25 -1 36 16 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 10000011101100101101000101110000
|
|
|
|
|
55 767 2 0 3 37 -2 55 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 892 30 -44 2 -62 -6 0 -9 44 -2 62 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000000000000100000000000
|
|
|
|
|
-637 72 -28 1 -40 -26 2 -39 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000000000010000000000000
|
|
|
|
|
-153 30 -27 0 -27 77 -2 111 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000000000000100000100000
|
|
|
|
|
153 27 0 27 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
315 -89 3 -128 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000000000010000000100000
|
|
|
|
|
-153 30 -27 0 -27 79 -3 113 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000000000100101100110000110000
|
|
|
|
|
153 27 0 27 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-462 -288 8 0 12 -11 0 -16 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
571 -62 -571 -102 3 -147 27 -1 39 102 -3 147 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
152 -29 1 -42 1 blend
|
|
|
|
|
vmoveto
|
|
|
|
|
571 -60 -571 -102 3 -147 26 -1 37 102 -3 147 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 -71 2 -102 1 blend
|
|
|
|
|
90 rmoveto
|
|
|
|
|
-212 631 212 -23 1 -32 45 -2 64 23 -1 32 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-776 -263 -22 1 -31 -4 0 -5 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 905 30 -42 2 -60 10 0 14 42 -2 60 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 00000001100000000000000100000000
|
|
|
|
|
-716 -160 36 -1 52 -26 2 -37 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 554 30 -13 0 -19 -59 2 -85 13 0 19 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-554 -78 59 -2 85 5 -1 7 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 563 30 -13 1 -19 -56 1 -81 13 -1 19 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 00000010000000000000001000000000
|
|
|
|
|
-578 -79 2 1 4 6 0 8 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00001000000000000000001000001000
|
|
|
|
|
-30 617 -27 1 -39 4 -1 5 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000010000001000000000000001000
|
|
|
|
|
30 27 -1 39 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-477 382 -24 2 -34 8 0 12 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-46 -92 -113 -104 -167 -65 7 -5 10 -9 5 -8 6 -1 8 -5 -1 -8 17 0 25 11 0 16 -3 -1 -5 6 0 9 12 0 18 -11 0 -16 18 0 26 -27 1 -39 6 -1 8 -16 1 -23 12 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000100010010010000000001000000
|
|
|
|
|
172 70 111 106 55 101 14 0 20 3 0 5 -6 0 -8 1 0 1 3 0 4 28 -1 41 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
298 -65 -24 0 -35 3 0 4 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-25 -12 -55 2 -79 -15 0 -22 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
62 -80 121 -81 100 -38 5 8 9 11 7 6 -101 33 -119 76 -59 77 2 0 3 -14 1 -20 -10 1 -14 2 0 3 20 0 29 1 0 2 9 -1 13 18 -1 25 20 -1 28 26 -1 38 14 0 21 14 -1 19 -13 0 -19 -7 0 -10 9 0 13 -18 2 -25 4 -1 5 -7 0 -10 18 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
-211 -88 -39 3 -55 -12 1 -17 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-239 30 239 -2 -1 -4 69 -4 98 2 1 4 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 10000010000000000000000000001000
|
|
|
|
|
316 -223 -74 3 -106 11 -1 15 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-6 -4 0 -6 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-8 -87 -7 -34 -10 -10 2 0 3 24 -1 35 -1 1 -1 6 0 9 1 -1 1 1 0 1 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-6 -1 0 -1 1 blend
|
|
|
|
|
-6 -6 -1 -12 2 0 3 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-11 -31 1 0 1 10 0 15 2 blend
|
|
|
|
|
1 3 -34 0 -1 -1 9 0 13 2 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
5 -8 3 -13 6 -1 8 -11 1 -16 5 0 8 -19 1 -26 4 blend
|
|
|
|
|
1 -8 28 -2 30 -1 14 1 -14 1 -20 7 -1 9 1 0 1 2 0 3 2 -1 2 3 1 5 0 1 1 7 blend
|
|
|
|
|
21 0 10 4 10 9 16 15 7 35 2 -1 2 8 -1 11 2 0 3 5 0 7 5 0 8 3 -1 4 3 0 4 2 1 4 3 0 4 9 blend
|
|
|
|
|
9 89 -15 1 -21 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
7 1 1 12 6 -1 8 1 0 1 1 -1 1 9 0 13 4 blend
|
|
|
|
|
0 hhcurveto
|
|
|
|
|
-660 -34 -57 3 -82 -8 0 -11 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-17 -46 1 0 2 7 0 10 2 blend
|
|
|
|
|
-32 -46 -46 5 0 7 5 0 7 2 blend
|
|
|
|
|
-23 20 -21 56 -2 81 -24 0 -35 2 blend
|
|
|
|
|
rcurveline
|
|
|
|
|
hintmask 10010000000000000000000000000000
|
|
|
|
|
52 28 31 51 17 46 -4 -1 -7 0 1 1 -4 1 -5 -7 0 -10 1 -1 1 0 0 -1 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00100000000000000000000010000000
|
|
|
|
|
110 -3 -67 3 -96 1 0 2 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
13 -38 10 -49 0 -32 -3 0 -4 4 -1 5 -2 0 -3 2 1 4 -1 1 -1 3 0 4 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
29 6 55 -3 78 8 -1 11 2 blend
|
|
|
|
|
-1 31 -10 50 -15 37 -3 0 -4 0 1 1 -4 0 -6 3 -1 4 -4 1 -5 5 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
hintmask 01000000000000000000000000100000
|
|
|
|
|
113 -6 -56 3 -80 -7 0 -10 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
22 -32 20 -44 7 -30 2 0 3 1 -1 1 3 -1 3 1 0 1 2 0 3 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
28 10 -8 29 -21 44 -23 32 48 -2 69 15 0 22 -2 1 -2 -1 0 -2 0 -1 -1 -5 1 -6 -1 1 -1 -4 0 -6 8 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rlinecurve
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00010000001000000000001000000000
|
|
|
|
|
117 -5 -45 1 -65 -17 1 -24 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
25 -23 -1 0 -1 2 -1 2 2 blend
|
|
|
|
|
27 -32 13 -23 -2 1 -2 1 0 2 2 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
21 14 -12 44 -2 63 20 -1 28 0 0 -1 3 blend
|
|
|
|
|
22 -27 32 -26 22 -2 0 -2 -2 0 -3 1 0 1 -2 1 -2 4 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-381 267 39 -1 56 7 -1 10 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-16 -30 -33 1 -47 -23 1 -33 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
498 30 -42 1 -61 23 -1 33 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-516 -23 21 0 31 -14 0 -21 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000010000000000000001000000000
|
|
|
|
|
-224 6 0 9 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 00000010001000000000000100000000
|
|
|
|
|
30 247 75 -4 106 10 0 14 2 blend
|
|
|
|
|
hlineto
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid13393" fdSelectIndex="1">
|
|
|
|
|
4 vsindex
|
|
|
|
|
-50 30 -19 19 114 30 44 30 23 30 -30 114 35 30 316 30 -10 10 37 12 -21 0 -26 66 0 82 -29 21 -10 29 -21 10 -64 0 -80 55 0 69 -79 0 -99 75 0 94 -46 0 -58 56 0 71 -56 0 -71 26 21 59 -18 -25 -54 54 0 68 -76 8 -85 58 0 73 -24 0 -31 24 0 31 -46 -4 -63 30 0 37 20 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
82 30 197 30 -26 8 317 30 168 13 -13 0 -16 77 0 96 -109 -1 -136 78 0 97 -77 0 -96 29 0 36 -10 0 -12 84 0 105 -86 0 -108 21 0 27 10 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 1010101101110110
|
|
|
|
|
529 746 23 0 29 30 4 43 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 320 30 -58 0 -73 -29 0 -36 58 0 73 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-397 -495 15 0 18 12 -4 10 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 442 30 -56 0 -71 21 0 27 56 0 71 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-420 149 -6 0 -8 6 -4 2 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 374 30 -54 0 -68 -25 0 -31 54 0 68 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-514 -420 34 0 42 -3 4 1 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 626 30 -66 0 -82 -29 0 -36 66 0 82 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-531 144 15 0 19 -9 0 -11 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 460 30 -55 0 -69 -4 0 -5 55 0 69 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-53 622 -42 0 -53 -6 4 -2 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-7 -9 0 -12 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-86 -171 -222 -118 -188 -45 7 -7 8 -11 3 -8 14 0 18 37 0 46 27 0 34 19 0 24 -7 0 -9 5 0 7 15 0 18 -16 0 -20 17 0 22 -32 0 -40 9 0 11 -19 0 -24 12 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0000000010000010
|
|
|
|
|
192 51 224 119 94 187 21 0 26 3 0 3 -17 0 -21 -9 0 -11 2 0 2 -3 0 -4 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0100010100000110
|
|
|
|
|
-19 12 -6 -2 -55 0 -68 27 0 34 -12 0 -15 -3 0 -3 4 blend
|
|
|
|
|
rlineto
|
|
|
|
|
-323 -32 55 0 69 -25 0 -32 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-25 -11 -68 0 -86 -23 0 -28 2 blend
|
|
|
|
|
rlineto
|
|
|
|
|
83 -154 177 -116 201 -44 4 8 9 12 7 6 -200 39 -177 113 -79 147 11 0 14 12 0 15 -18 0 -22 21 0 26 -1 0 -1 4 0 5 11 0 13 21 0 26 21 0 27 32 0 40 17 0 21 16 0 20 9 0 11 -10 0 -12 17 0 21 -36 0 -45 1 0 2 -37 0 -47 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
59 127 -46 0 -58 9 -4 6 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-40 -82 -80 -104 -112 -75 8 -4 10 -9 6 -7 115 80 2 0 2 8 0 10 7 0 9 23 0 29 2 0 3 16 0 20 16 0 20 -12 0 -15 26 0 32 -30 0 -37 10 0 13 -18 0 -23 8 0 10 -4 0 -5 14 blend
|
|
|
|
|
80 106 47 90 -13 0 -16 11 0 13 14 0 17 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-129 -493 -106 -5 -137 21 6 34 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-27 -73 -43 -71 -51 -50 8 -5 13 -9 5 -5 49 52 47 77 29 77 6 0 8 11 0 14 7 0 8 8 0 10 5 0 7 8 0 10 16 0 20 -8 0 -10 28 0 35 -17 -1 -22 15 0 18 -11 0 -14 -3 0 -4 -4 0 -5 -2 0 -2 -1 0 -1 -3 0 -4 -3 1 -3 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
124 -1 -66 4 -77 10 15 31 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-374 30 374 4 0 5 84 0 105 -4 0 -5 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 0000000000101000
|
|
|
|
|
-586 460 -72 0 -90 2 -21 -24 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-875 30 845 209 30 -27 0 -33 77 0 96 -53 0 -66 -79 0 -99 80 0 99 5 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-8 -29 0 -36 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
-7 -29 0 -36 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-28 -75 -43 -102 -46 -95 14 0 17 10 0 13 11 0 14 -41 0 -51 17 0 22 4 0 5 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0001000000010000
|
|
|
|
|
89 -91 24 -74 -63 -32 0 -40 23 0 28 -11 0 -14 10 0 13 17 0 21 5 blend
|
|
|
|
|
vvcurveto
|
|
|
|
|
-33 -6 -35 -19 -13 3 0 4 1 0 1 15 0 18 7 0 9 4 0 5 5 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
-10 -6 -12 3 0 3 0 0 1 1 0 2 3 blend
|
|
|
|
|
-3 -14 -1 -20 -2 -26 1 -29 4 0 5 1 0 1 7 0 9 2 0 2 13 0 16 -1 0 -1 11 0 14 7 blend
|
|
|
|
|
2 7 -9 4 -13 11 0 13 -21 0 -26 5 0 6 -33 0 -41 4 blend
|
|
|
|
|
1 -8 22 -2 27 0 22 -21 0 -27 1 0 2 1 0 1 -3 0 -4 0 0 1 -4 0 -5 6 blend
|
|
|
|
|
2 19 2 17 5 12 9 3 0 4 2 0 2 3 0 3 2 0 2 4 0 5 3 0 4 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
25 17 10 7 0 9 5 0 7 4 0 5 3 blend
|
|
|
|
|
43 44 22 0 27 1 blend
|
|
|
|
|
vvcurveto
|
|
|
|
|
67 -22 76 -86 89 -8 0 -10 9 0 12 -6 0 -8 24 0 30 -11 0 -13 5 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
hintmask 0000000001001000
|
|
|
|
|
39 84 42 98 33 81 -10 0 -13 -4 0 -5 -8 0 -10 14 0 17 -6 0 -8 7 0 9 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0000000000001000
|
|
|
|
|
-20 14 -6 -2 -60 0 -75 32 0 40 -12 0 -14 -2 0 -3 4 blend
|
|
|
|
|
rlineto
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid17290" fdSelectIndex="1">
|
|
|
|
|
5 vsindex
|
|
|
|
|
121 30 -22 22 148 30 -30 136 23 30 129 30 116 30 -21 4 -29 52 3 92 -32 23 -21 32 -23 21 -54 9 -83 50 4 90 -50 -4 -90 22 27 62 -2 -43 -47 41 0 69 -44 0 -74 37 0 62 -50 0 -84 36 0 61 14 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
167 30 129 30 -16 16 123 30 48 30 -6 29 -29 111 -30 30 -16 16 201 30 1 29 -29 0 -49 64 0 108 -34 0 -57 51 0 85 -29 0 -48 29 0 48 -72 -2 -123 60 2 103 -69 0 -115 46 0 77 -42 0 -70 42 0 70 -42 0 -70 67 0 111 -51 0 -85 51 0 85 -29 0 -48 29 0 48 -79 0 -132 47 0 79 -45 0 -75 42 0 70 22 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 011011111011001010000000
|
|
|
|
|
326 793 1 0 2 17 0 29 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-280 24 0 40 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-47 16 -8 59 -31 0 -53 6 0 10 -13 0 -21 20 0 33 4 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
hintmask 000010000000100000000000
|
|
|
|
|
13 120 4 0 6 -46 0 -76 2 blend
|
|
|
|
|
0 13 4 0 7 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
49 10 20 82 4 12 0 19 12 0 20 3 0 5 2 0 3 4 0 8 5 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
hintmask 101010101000010000000000
|
|
|
|
|
-10 2 -11 5 -8 6 -12 0 -21 3 0 5 -21 0 -35 6 0 11 -9 0 -14 7 0 11 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
-75 19 0 32 1 blend
|
|
|
|
|
-3 -5 -10 -29 -24 -102 1 0 1 1 0 2 7 0 12 9 0 14 42 0 70 5 blend
|
|
|
|
|
0 -18 6 0 10 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-38 -6 4 21 10 0 18 2 0 3 0 0 -1 4 0 7 4 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
280 -25 0 -41 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-41 -464 -40 -8 -74 10 20 41 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 617 30 -50 -4 -90 -5 12 5 50 4 90 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-661 -178 11 -4 12 4 -13 -7 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 689 30 -52 -3 -92 -11 0 -18 52 3 92 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 010101100111001000000000
|
|
|
|
|
-481 284 -27 -2 -48 -32 36 -21 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-306 30 306 0 -13 0 60 2 103 0 13 0 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
218 0 -61 0 -102 -1 0 -1 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-306 30 306 0 -13 0 61 1 104 0 13 0 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-417 358 -17 -1 -30 19 -43 -12 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 217 -116 -217 -30 247 176 -36 0 -61 -52 0 -87 50 0 84 52 0 87 -37 0 -62 -6 0 -10 23 0 39 7 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
75 -26 0 -44 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
hintmask 000010100000001001000000
|
|
|
|
|
-280 24 0 40 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-47 17 -8 60 -31 0 -53 5 0 9 -13 0 -21 20 0 33 4 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
12 125 5 0 8 -47 0 -78 2 blend
|
|
|
|
|
0 14 4 0 7 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
49 11 20 82 3 12 0 20 12 0 19 3 1 6 2 1 6 5 0 9 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-9 2 -12 4 -8 7 -14 1 -22 3 0 5 -19 -1 -34 7 0 12 -9 0 -14 6 0 10 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-75 19 -1 29 1 blend
|
|
|
|
|
-3 -5 -10 -30 -25 -105 1 -1 1 8 0 13 8 0 14 42 0 70 4 blend
|
|
|
|
|
0 -18 6 0 9 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-40 -6 4 21 11 0 19 2 0 3 0 0 -1 4 1 8 4 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
280 -25 -1 -42 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 000001110000000110000000
|
|
|
|
|
-16 -29 0 -48 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
-30 217 -116 -217 -30 247 176 -36 0 -61 -50 0 -84 50 0 84 50 0 84 -37 0 -62 -3 0 -5 23 0 39 7 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-424 -714 -19 0 -32 -12 0 -21 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-52 -54 -91 -49 -81 -33 8 -5 11 -13 4 -6 80 36 94 56 56 58 7 0 11 9 0 15 5 0 9 11 0 18 -2 0 -3 9 0 15 13 0 22 -11 0 -18 24 0 39 -22 0 -36 11 0 19 -12 0 -21 4 0 7 -4 0 -6 2 0 2 -2 0 -4 -1 0 -1 3 0 5 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
200 -7 -92 0 -154 -5 0 -8 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
76 -41 90 -62 46 -42 -6 0 -10 5 0 8 -5 0 -7 6 0 10 -4 0 -7 4 0 7 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
22 23 -46 42 -91 60 -75 39 60 0 100 29 0 48 0 0 -1 -3 0 -5 3 0 5 -7 0 -11 6 0 11 -7 0 -11 8 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rlinecurve
|
2019-05-13 13:12:46 +01:00
|
|
|
-499 750 -48 0 -81 6 0 10 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-54 -167 -87 -164 -96 -108 7 -6 11 -12 4 -6 98 116 88 165 58 175 7 0 13 15 0 25 10 0 16 14 0 22 11 0 19 10 0 17 9 0 15 -20 0 -33 15 0 24 -44 0 -73 4 0 7 -18 0 -30 4 0 6 4 0 6 3 0 6 19 0 32 0 0 -1 1 0 1 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-113 -214 -60 0 -100 -23 0 -37 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-691 30 718 20 0 33 64 0 108 43 0 72 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-1 -1 0 -3 1 blend
|
|
|
|
|
2 rlineto
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
</CharString>
|
2019-05-13 13:12:46 +01:00
|
|
|
<CharString name="cid17852" fdSelectIndex="1">
|
|
|
|
|
5 vsindex
|
|
|
|
|
-67 29 219 30 154 30 -16 16 150 30 -30 122 -85 30 -18 18 87 30 -30 140 -122 12 -14 0 -22 46 0 78 -59 -3 -106 46 0 77 -53 -9 -92 46 2 81 -18 20 -1 18 -20 1 -54 13 -80 46 2 81 -46 -2 -81 25 31 61 -14 -34 -48 60 0 100 -64 0 -107 64 0 107 -55 0 -92 54 0 90 -54 0 -90 36 0 59 -19 0 -31 37 0 62 22 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
51 188 -30 30 -30 149 21 30 -18 18 -13 13 66 30 -12 12 135 30 41 30 172 30 -6 28 -8 0 -14 30 0 50 -62 0 -103 62 0 103 -62 0 -103 32 0 53 -5 0 -7 59 0 98 -24 0 -41 24 0 41 -16 0 -27 16 0 27 -32 0 -53 53 0 88 -33 0 -56 33 0 56 -87 0 -146 63 0 106 -42 0 -70 54 0 90 -99 0 -165 55 0 91 -42 0 -70 45 0 75 24 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vstemhm
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000100001000000000000
|
|
|
|
|
51 612 -8 0 -14 29 0 49 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 -60 0 -100 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000100000010000000000
|
|
|
|
|
307 30 60 0 100 1 blend
|
|
|
|
|
hlineto
|
|
|
|
|
hintmask 000000010010100100000000
|
|
|
|
|
-149 228 -32 0 -53 -20 0 -34 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-918 30 918 -19 0 -32 62 0 103 19 0 32 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
-36 -238 -55 0 -91 -32 0 -53 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-31 -160 -74 -193 -68 -95 7 -5 10 -11 6 -8 70 101 74 203 33 160 6 0 10 25 0 42 13 0 21 23 0 37 4 0 7 1 0 2 8 0 14 -18 0 -30 13 0 21 -27 0 -44 4 0 7 -19 0 -32 1 0 2 6 0 10 -12 0 -20 -2 0 -3 -2 0 -4 -1 0 -2 18 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
4 -143 19 0 32 77 0 128 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-21 -16 25 -26 72 -92 21 -33 -23 0 -38 -34 0 -57 1 0 2 -15 0 -24 -12 0 -21 -6 0 -11 2 0 3 -18 0 -29 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
24 24 -18 25 -81 96 -22 22 28 0 48 63 0 105 2 0 2 -1 0 -2 1 0 3 10 0 16 1 0 1 1 0 2 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
157 278 1 0 1 -14 0 -23 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
hintmask 000000001000000100000000
|
|
|
|
|
-30 559 -54 0 -90 -17 3 -23 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 010000000010000000100000
|
|
|
|
|
30 54 0 90 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-457 -518 29 -3 43 -9 -3 -20 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 176 30 -46 0 -77 -17 0 -27 46 0 77 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000100000001010000
|
|
|
|
|
-194 120 -3 0 -5 -42 37 -35 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-365 30 365 38 -29 45 53 0 88 -38 29 -45 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
135 508 -87 0 -146 33 -34 24 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000000000010000000010000
|
|
|
|
|
-122 30 -19 0 -31 63 0 106 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 000101000100000000010000
|
|
|
|
|
122 19 0 31 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-115 -172 -60 0 -100 -27 34 -19 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-288 30 288 11 -24 18 50 0 83 -11 24 -18 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
148 -62 -2 -106 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
-288 30 288 11 -24 18 50 0 83 -11 24 -18 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
156 -394 -30 2 -47 19 -34 6 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-52 -36 -89 -48 -61 -29 7 0 12 2 0 4 14 0 23 3 0 4 11 0 18 4 0 8 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
15 -21 62 28 86 41 57 44 25 0 42 -39 0 -66 -10 0 -17 -4 0 -6 -12 0 -19 -3 0 -5 -6 0 -11 -5 0 -9 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
hintmask 101010000000000010001100
|
|
|
|
|
-541 323 10 0 17 44 5 84 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 517 -150 -517 -30 547 210 -46 -2 -81 -74 0 -123 54 -13 80 74 0 123 -46 -2 -81 -19 0 -32 38 17 82 7 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-232 -242 -10 0 -16 -28 29 -27 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-344 58 -32 71 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-47 15 -9 54 -33 -2 -58 3 0 4 -15 0 -25 22 0 37 4 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
hintmask 100000000010001000001010
|
|
|
|
|
12 100 3 0 5 -47 0 -78 2 blend
|
|
|
|
|
0 12 4 0 6 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
48 10 25 102 3 12 0 20 11 0 19 4 1 9 11 -1 16 5 0 8 5 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
-9 3 -11 4 -8 6 -14 0 -23 3 -1 5 -23 1 -37 8 1 15 -8 -1 -15 8 0 12 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-97 11 1 20 1 blend
|
|
|
|
|
-3 -4 -14 -29 -21 -84 0 0 1 1 -1 1 10 0 16 10 1 17 43 -1 71 5 blend
|
|
|
|
|
0 -16 7 0 12 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-33 -6 5 22 13 0 22 3 0 5 -1 0 -2 4 0 7 4 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
344 -59 34 -71 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-346 -371 -24 0 -41 65 -34 78 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
10 -31 77 16 100 22 99 21 3 0 5 -54 0 -90 -2 0 -3 -3 0 -5 -9 0 -15 -6 0 -10 -10 0 -17 -5 0 -8 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-2 29 -108 -22 -104 -22 -72 -13 -3 0 -5 52 0 86 9 0 16 6 0 10 8 0 13 6 0 11 4 0 6 4 0 6 8 blend
|
|
|
|
|
rlinecurve
|
|
|
|
|
-16 767 -44 0 -72 -13 0 -21 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-316 -6 0 -11 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-142 -7 -194 -74 -141 2 0 2 -2 0 -2 2 0 4 6 0 9 4 blend
|
|
|
|
|
vhcurveto
|
|
|
|
|
8 -3 13 -7 5 -6 13 0 21 -7 0 -11 25 0 43 -20 0 -34 11 0 17 -10 0 -17 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
75 143 10 205 145 4 0 7 3 0 5 2 0 4 21 0 35 9 0 15 5 blend
|
|
|
|
|
vvcurveto
|
|
|
|
|
316 6 0 11 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
</CharString>
|
|
|
|
|
<CharString name="cid18480" fdSelectIndex="1">
|
|
|
|
|
3 vsindex
|
|
|
|
|
-71 30 427 30 153 30 33 111 -30 30 -30 126 -6 0 -13 45 0 102 -58 0 -132 38 0 87 -48 0 -111 38 0 87 -4 -2 -13 21 2 53 -43 0 -99 43 0 99 -43 0 -99 24 0 55 12 blend
|
|
|
|
|
hstemhm
|
|
|
|
|
159 30 -19 19 126 30 -6 30 281 30 160 30 18 31 -7 0 -16 50 0 114 -18 0 -42 18 0 42 -71 0 -161 50 0 114 -26 0 -61 48 0 111 -66 0 -150 51 0 115 -68 0 -154 50 0 114 -36 -1 -84 44 1 101 14 blend
|
|
|
|
|
vstemhm
|
|
|
|
|
hintmask 1110100101110000
|
|
|
|
|
58 743 -1 0 -2 26 0 60 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
-30 887 30 -43 0 -99 2 0 5 43 0 99 3 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0000010010000000
|
|
|
|
|
-630 96 -29 0 -66 -19 0 -44 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0001000010000000
|
|
|
|
|
-207 30 -2 -2 -9 50 0 114 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0000010010100000
|
|
|
|
|
207 2 2 9 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
305 -44 0 -100 1 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
hmoveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0001000000100000
|
|
|
|
|
-207 30 -2 -2 -9 51 0 115 2 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vlineto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 0010011000100000
|
|
|
|
|
207 2 2 9 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-521 -240 -36 0 -82 2 0 4 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-206 -5 0 -10 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-137 -15 -184 -109 -136 5 0 11 3 0 6 5 0 10 -1 0 -1 8 0 19 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
7 -3 12 -9 5 -6 12 0 27 -6 0 -13 22 0 51 -15 0 -35 10 0 21 -8 0 -19 6 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
hintmask 1110000101010000
|
|
|
|
|
112 139 18 194 141 3 0 7 -4 0 -8 1 0 3 11 0 24 4 0 10 5 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vvcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
207 5 0 11 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-19 -18 0 -42 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
-30 670 -153 -670 -30 700 213 -38 0 -87 -64 0 -144 48 0 111 64 0 144 -38 0 -87 -14 0 -30 28 0 63 7 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-531 -249 -15 0 -36 -23 0 -51 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-343 50 0 112 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-66 31 -12 105 -29 0 -66 6 0 14 -13 0 -28 29 0 66 4 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
vhcurveto
|
2019-05-13 13:12:46 +01:00
|
|
|
23 278 5 0 12 -59 0 -134 2 blend
|
|
|
|
|
0 24 6 0 14 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
hintmask 1000000001001000
|
|
|
|
|
96 15 31 123 8 20 0 44 11 0 26 4 0 8 14 0 32 5 0 11 5 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
-9 3 -13 4 -9 7 -13 0 -30 2 0 5 -21 0 -48 8 0 17 -10 -1 -23 6 0 15 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
-117 -6 -11 -21 -69 -56 -236 8 0 18 -1 1 -1 1 -1 1 3 0 7 3 1 9 7 0 15 49 0 112 7 blend
|
|
|
|
|
0 -41 4 0 8 1 blend
|
|
|
|
|
hhcurveto
|
|
|
|
|
-84 -16 11 37 4 0 10 2 0 5 -3 0 -7 1 0 2 4 blend
|
|
|
|
|
hvcurveto
|
|
|
|
|
343 -51 0 -115 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
444 -47 -59 0 -135 26 0 59 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-101 -52 -195 -56 -169 -40 4 -7 5 -10 3 -7 172 40 193 54 120 56 4 0 8 3 0 7 18 0 43 9 0 19 12 0 26 8 0 19 5 0 12 -10 0 -22 7 0 15 -18 0 -41 1 0 3 -11 0 -25 -8 0 -19 -9 0 -21 -8 0 -18 -8 0 -19 5 0 11 0 0 1 18 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
</CharString>
|
|
|
|
|
<CharString name="cid22370" fdSelectIndex="1">
|
|
|
|
|
2 vsindex
|
|
|
|
|
64 30 77 30 76 30 74 30 72 30 109 30 25 84 -30 30 -30 108 -2 0 -2 42 0 47 -48 0 -54 38 0 43 -48 0 -54 38 0 43 -46 0 -52 42 0 47 -43 0 -48 56 1 63 -72 -1 -81 57 1 64 -8 -32 -41 30 32 65 -65 -1 -73 65 1 73 -65 -1 -73 43 0 49 18 blend
|
|
|
|
|
hstemhm
|
|
|
|
|
135 30 21 30 102 30 14 30 205 30 17 30 113 30 19 30 -19 0 -21 87 2 98 -86 -2 -97 99 1 111 -125 -1 -141 98 1 111 -79 -1 -89 75 1 84 -99 -1 -111 75 1 84 -77 -1 -86 100 1 112 -127 -1 -143 105 1 118 -102 -1 -114 94 1 105 16 blend
|
|
|
|
|
vstemhm
|
|
|
|
|
hintmask 111111010011001100000000
|
|
|
|
|
53 761 -3 0 -3 36 0 40 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-30 896 30 -65 -1 -73 5 0 5 65 1 73 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 000000001001000000000000
|
|
|
|
|
-631 78 -46 0 -52 -22 0 -24 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
hintmask 000000100001000000000000
|
|
|
|
|
-162 30 -8 -32 -41 98 1 111 2 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 000000001001001000000000
|
|
|
|
|
162 8 32 41 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
296 -105 -1 -118 1 blend
|
|
|
|
|
hmoveto
|
|
|
|
|
hintmask 000000100000001000000000
|
|
|
|
|
-162 30 -8 -32 -41 100 1 112 2 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 000000001000001000000000
|
|
|
|
|
162 8 32 41 1 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 000011000100110010000000
|
|
|
|
|
-47 -217 -23 0 -26 -57 -1 -64 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
209 -109 -209 -101 -1 -113 72 1 81 101 1 113 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-235 109 24 0 27 -72 -1 -81 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
205 -109 -205 -99 -1 -111 72 1 81 99 1 111 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-227 109 18 1 21 -72 -1 -81 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
197 -109 -197 -93 -2 -105 72 1 81 93 2 105 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 139 -87 -2 -98 -15 0 -17 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-169 731 169 -41 0 -46 38 0 42 41 0 46 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
hintmask 111100000010000100000000
|
|
|
|
|
-650 -375 62 1 70 -32 0 -36 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
571 -76 -571 -159 -1 -179 48 0 54 159 1 179 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 -38 0 -43 1 blend
|
|
|
|
|
vmoveto
|
|
|
|
|
571 -77 -571 -159 -1 -179 48 0 54 159 1 179 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
287 -66 -1 -74 1 blend
|
|
|
|
|
vmoveto
|
|
|
|
|
571 -74 -571 -159 -1 -179 46 0 52 159 1 179 3 blend
|
|
|
|
|
hlineto
|
|
|
|
|
-30 104 -99 -1 -111 -4 0 -5 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-347 631 347 -18 0 -20 45 0 50 18 0 20 3 blend
|
|
|
|
|
vlineto
|
|
|
|
|
-216 -389 -86 -1 -96 -31 0 -35 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
127 -34 121 -39 72 -31 -17 0 -19 2 0 2 -13 0 -15 -2 0 -2 -13 0 -15 3 0 3 6 blend
|
|
|
|
|
rrcurveto
|
|
|
|
|
31 22 -78 32 -126 39 -121 136 1 153 39 0 44 1 0 1 -3 0 -3 -8 0 -9 4 0 5 9 0 10 7 blend
|
|
|
|
|
31 rlinecurve
|
|
|
|
|
-258 -1 -67 -1 -75 0 0 -1 2 blend
|
|
|
|
|
rmoveto
|
|
|
|
|
-81 -39 -128 -36 -107 -23 8 -6 12 -12 5 -6 103 25 130 41 86 43 9 0 10 6 0 7 3 0 4 7 0 8 -4 0 -5 7 0 8 19 0 22 -14 0 -16 32 0 36 -32 0 -36 17 0 19 -19 0 -21 3 0 3 -1 0 -1 5 0 6 2 0 2 1 0 1 4 0 5 18 blend
|
Sparse cff2vf support (#1591)
* Added getter (in the form of a property decorator) for T2Charstring.vsindex. Fixes endless compile loop in some circumstances.
Fixed bug in mutator: need to remove vsindex from snapshotted charstrings, plus formatting clean up
* Fix for subsetting HVAR tables that have an AdvanceWidthMap when the --retain-gid option is used. Needed to make subset_test.py::test_retain_gids_cff2 tests pass.
* in varLib/cffLib.py, add support for sparse sources, and sources with more than one model, and hence more than one VarData element in the VarStore.
CFF2 source fonts with multiple FontDicts in the FDArray need some extra work. With sparse fonts, some of the source fonts may have a fewer FontDicts than the default font. The getfd_map function() builds a map from the FontDict indices in the default font to those in each region font. This is needed when building up the blend value lists in the master font FontDict PrivateDicts, in order to fetch PrivateDict values from the correct FontDict in each region font.
In specializer.py, add support for CFF2 CharStrings with blend operators. 1) In generalizeCommands, convert a blend op to a list of args that are blend lists for the following regular operator. A blend list as a default font value, followed by the delta tuple. 2) In specializeCommands(), convert these back to blend ops, combining as many successive blend lists as allowed by the stack limit.
Add test case for sparse CFF2 sources.
The test font has 55 glyphs. 2 glyphs use only 2 sources (weight = 0 and 100). The rest use 4 source fonts: the two end points of the weight axis, and two intermediate masters. The intermediate masters are only 1 design space unit apart, and are used to change glyph design at the point in design space. For the rest, at most 2 glyphs use the same set of source fonts. There are 12 source fonts.
Add test case for specializer programToCommands() and commandsToProgram by converting each CharString.program in the font to a command list, and back again, and comparing original and final versions.
2019-04-26 09:33:52 -07:00
|
|
|
rrcurveto
|
|
|
|
|
</CharString>
|
|
|
|
|
</CharStrings>
|
|
|
|
|
<VarStore Format="1">
|
|
|
|
|
<Format value="1"/>
|
|
|
|
|
<VarRegionList>
|
|
|
|
|
<!-- RegionAxisCount=1 -->
|
|
|
|
|
<!-- RegionCount=16 -->
|
|
|
|
|
<Region index="0">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="1">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="0.669"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="2">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.669"/>
|
|
|
|
|
<PeakCoord value="0.67"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="3">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.67"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="4">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="0.889"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="5">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.889"/>
|
|
|
|
|
<PeakCoord value="0.89"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="6">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.89"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="7">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="0.439"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="8">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.439"/>
|
|
|
|
|
<PeakCoord value="0.44"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="9">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.44"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="10">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="0.799"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="11">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.799"/>
|
|
|
|
|
<PeakCoord value="0.8"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="12">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.8"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="13">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.0"/>
|
|
|
|
|
<PeakCoord value="0.599"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="14">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.599"/>
|
|
|
|
|
<PeakCoord value="0.6"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
<Region index="15">
|
|
|
|
|
<VarRegionAxis index="0">
|
|
|
|
|
<StartCoord value="0.6"/>
|
|
|
|
|
<PeakCoord value="1.0"/>
|
|
|
|
|
<EndCoord value="1.0"/>
|
|
|
|
|
</VarRegionAxis>
|
|
|
|
|
</Region>
|
|
|
|
|
</VarRegionList>
|
|
|
|
|
<!-- VarDataCount=6 -->
|
|
|
|
|
<VarData index="0">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=1 -->
|
|
|
|
|
<VarRegionIndex index="0" value="0"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
<VarData index="1">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=3 -->
|
|
|
|
|
<VarRegionIndex index="0" value="1"/>
|
|
|
|
|
<VarRegionIndex index="1" value="2"/>
|
|
|
|
|
<VarRegionIndex index="2" value="3"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
<VarData index="2">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=3 -->
|
|
|
|
|
<VarRegionIndex index="0" value="4"/>
|
|
|
|
|
<VarRegionIndex index="1" value="5"/>
|
|
|
|
|
<VarRegionIndex index="2" value="6"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
<VarData index="3">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=3 -->
|
|
|
|
|
<VarRegionIndex index="0" value="7"/>
|
|
|
|
|
<VarRegionIndex index="1" value="8"/>
|
|
|
|
|
<VarRegionIndex index="2" value="9"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
<VarData index="4">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=3 -->
|
|
|
|
|
<VarRegionIndex index="0" value="10"/>
|
|
|
|
|
<VarRegionIndex index="1" value="11"/>
|
|
|
|
|
<VarRegionIndex index="2" value="12"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
<VarData index="5">
|
|
|
|
|
<!-- ItemCount=0 -->
|
|
|
|
|
<NumShorts value="0"/>
|
|
|
|
|
<!-- VarRegionCount=3 -->
|
|
|
|
|
<VarRegionIndex index="0" value="13"/>
|
|
|
|
|
<VarRegionIndex index="1" value="14"/>
|
|
|
|
|
<VarRegionIndex index="2" value="15"/>
|
|
|
|
|
</VarData>
|
|
|
|
|
</VarStore>
|
|
|
|
|
</CFFFont>
|
|
|
|
|
|
|
|
|
|
<GlobalSubrs>
|
|
|
|
|
<!-- The 'index' attribute is only for humans; it is ignored when parsed. -->
|
|
|
|
|
</GlobalSubrs>
|
|
|
|
|
</CFF2>
|
|
|
|
|
|
|
|
|
|
</ttFont>
|
