|
edf40148
|
2016-12-15T21:56:44
|
|
[cff] Implement CFF2 support (2/2).
The font variation code. All parts dependent on the GX code in the
`truetype' module are guarded with TT_CONFIG_OPTION_GX_VAR_SUPPORT.
In other words, you can still compile the `cff' module without
defining TT_CONFIG_OPTION_GX_VAR_SUPPORT (which brings you CFF2
support without font variation).
* src/cff/cf2font.c (cf2_font_setup): Add support for font
variation.
* src/cff/cf2font.h (CF2_Font): Add fields for variation data.
* src/cff/cf2ft.c (cf2_free_instance): Free blend data.
(cf2_getVStore, cf2_getNormalizedVector): New functions.
* src/cff/cf2ft.h: Updated.
* src/cff/cf2intrp.c: Include `cffload.h'.
(cf2_cmdRESERVED_15, cf2_cmdRESERVED_16): Replace with...
(cf2_cmdVSINDEX, cf2_cmdBLEND): ... this new enum values.
(cf2_doBlend): New function.
(cf2_interpT2CharString): Handle `vsindex' and `blend' opcodes.
* src/cff/cffload.c (FT_fdot14ToFixed): New macro.
(cff_vstore_done, cff_vstore_load): New functions.
(cff_blend_clear, cff_blend_doBlend, cff_blend_build_vector,
cff_blend_check_vector): New functions.
(cff_load_private_dict): Add arguments for blend vector.
Handle blend data.
(cff_subfont_load, cff_subfont_done): Updated.
(cff_font_load): Handle CFF2 variation store data.
(cff_font_done): Updated.
* src/cff/cffload.h: Include `cffparse.h'.
Updated.
* src/cff/cffobjs.c (cff_face_done): Updated.
* src/cff/cffparse.c: Include `cffload.h'.
(cff_parse_num): Handle internal value 255.
(cff_parse_vsindex, cff_parse_blend): New functions.
(CFF_FIELD_BLEND): New macro.
(cff_parser_run): Updated.
* src/cff/cffparse.h (cff_kind_blend): New enum value.
* src/cff/cfftoken.h: Handle `vstore', `vsindex', and `blend'
dictionary values.
* src/cff/cfftypes.h (CFF_VarData, CFF_AxisCoords, CFF_VarRegion,
CFF_VStore, CFF_Blend): New structures.
(CFF_FontRecDict): Add `vstore_offset' field.
(CFF_Private): Add `vsindex' field.
(CFF_SubFont): Add fields for blend data.
(CFF_Font): Add `vstore' field.
* src/truetype/ttgxvar.c (TT_Get_MM_Var): `CFF2' is equal to `gvar',
since glyph variation data is directly embedded.
(TT_Set_MM_Blend): Don't load `gvar' table for CFF2 fonts.
|
|
|
9f62d2ca
|
2016-12-15T20:27:47
|
|
[cff] Implement CFF2 support (1/2).
This commit does not contain the blend code for font variation
support, which follows in another commit.
You should ignore whitespace while inspecting this commit.
* include/freetype/internal/tttypes.h (TT_Face): Add `isCFF2'
member.
* src/cff/cf2font.h (CF2_Font): Add `isCFF2' member.
* src/cff/cf2ft.c (cf2_decoder_parse_charstrings): Handle `isCFF2'
flag.
(cf2_getMaxstack): New function.
* src/cff/cf2ft.h: Updated.
* src/cff/cf2intrp.c (cf2_escRESERVED_38): New enum.
(cf2_interpT2CharString): Handle CFF2 differences.
Add tracing message for errors.
* src/cff/cffdrivr.c (cff_get_glyph_name, cff_get_name_index):
Update for CFF2.
* src/cff/cffload.c (FT_FIXED_ONE): New macro.
(cff_index_init, cff_index_load_offsets, cff_index_access_element,
cff_index_get_name, cff_ft_select_get, cff_load_private_dict,
cff_subfont_load, cff_font_load): Handle CFF2.
* src/cff/cffload.h: Updated.
* src/cff/cffobjs.c (cff_face_init): Handle CFF2.
* src/cff/cffparse.c (cff_parse_maxstack): New function.
(CFFCODE_TOPDICT, CFFCODE_PRIVATE): Removed
* src/cff/cffparse.h (CFF2_MAX_STACK, CFF2_DEFAULT_STACK): New
macros.
(CFF2_CODE_TOPDICT, CFF2_CODE_FONTDICT, CFF2_CODE_PRIVATE): New
macros.
* src/cff/cfftoken.h: Add fields for CFF2 dictionaries (but no blend
stuff).
* src/cff/cfftypes.h (CFF_Index): Add `hdr_size' field.
(CFF_FontRecDict): Add `maxstack' field.
(CFF_Private): Add `subfont' field.
(CFF_Font): Add `top_dict_length' and `cff2' fields.
* src/sfnt/sfobjs.c (sfnt_load_face): Handle `CFF2' table.
|
|
37412ff9
|
2016-01-12T21:37:13
|
|
Don't use macro names that contain `__' [1/2].
Such macro names are reserved for both C and C++.
*/*: Replace macros of the form `__XXX_H__' with `XXX_H_'.
|
|
|
f1c93439
|
2015-10-22T10:11:23
|
|
[cff] Avoid overflow/module arithmetic.
This modifies the addition of subroutine number to subroutine bias
from unsigned to signed, but does not change any results.
* src/cff/cf2ft.c (cf2_initGlobalRegionBuffer,
cf2_initLocalRegionBuffer): Change variable names from (unsigned)
`idx' to (signed) `subrNum', since it is not an index until after
the bias is added.
* src/cff/cf2ft.h: Updated.
* src/cff/cf2intrp.c (cf2_interpT2CharString) <cf2_cmdCALLSUBR>:
Updated similarly.
|
|
|
eb05bfbe
|
2015-02-20T08:37:35
|
|
[cff] Signedness fixes for new engine.
* src/cff/cf2arrst.c, src/cff/cf2fixed.h, src/cff/cf2ft.c,
src/cff/cf2ft.h, src/cff/cf2hints.c, src/cff/cf2intrp.c: Apply.
|
|
|
77c39b1d
|
2013-05-04T18:04:07
|
|
[cff] Make Adobe CFF engine work correctly on 64bit hosts.
Reported by numerous people on the `freetype-devel' list. Without
this fix, glyphs aren't properly aligned on a common baseline.
On 64bit systems, `FT_Pos' expands to `long int', having a width of
64bit. `CF2_Fixed' expands to `int' which is normally 32bit wide on
64bit hosts also. Wrong casts filled up the blues arrays with
incorrect values. Note that all blues values are accessed with the
`cf2_blueToFixed' macro which handles the 64bit to 32bit conversion.
* src/cff/cf2ft.h (cf2_getBlueValues, cf2_getOtherBlues,
cf2_getFamilyBlues, cf2_getFamilyOtherBlues): Use `FT_Pos' for
`data', not `CF2_Fixed'.
* src/cff/cf2ft.c (cf2_getBlueValues, cf2_getOtherBlues,
cf2_getFamilyBlues, cf2_getFamilyOtherBlues): Updated.
* src/cff/cf2blues.c (cf2_blues_init): Updated.
|
|
|
283c8ed8
|
2013-04-13T15:02:31
|
|
[cff] New files for Adobe's Type 2 interpreter and hinting engine.
|