The buffer size FT_MAX_GRAY_SPANS is set to 10 spans, which should be
enough to cover the entire scanline for simple glyphs in most cases:
each slightly slanted edge needs up to two spans, plus a filling span
in-between. This is not new, we used to do it before cb4388783c.
* src/smooth/ftgrays.c (gray_TWorker): Add `spans' and `num_spans'.
(gray_hline, gray_sweep): Implement the span buffering.
(gray_raster_render): Use negative `num_spans' to avoid the direct
mode.
Up to now, only the unscaled CVT values were varied; in other words,
the `CVAR' data was never used for bytecode hinting.
* src/truetype/ttgxvar.c (tt_cvt_ready_iterator): New auxiliary
function.
(tt_face_vary_cvt): Use it to trigger rescaling of CVT values.
If `CVAR' data is applied to variation fonts, fractional values are
possible.
* include/freetype/internal/tttypes.h (TT_FaceRec): Change type of
`cvt' from `FT_Short' to `FT_Int32'.
* src/truetype/ttgxvar.c (FT_fdot6ToFixed): New macro.
(tt_face_vary_cvt): Use it to update code to 26.6 format.
* src/truetype/ttobjs.c (tt_size_run_prep): Update code to 26.6
format.
* src/truetype/ttpload.c (tt_face_load_cvt): Stora data in 26.6
format.
Very embarassing :-)
Reported as
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14701https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14705https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14710
* src/truetype/ttgload.c (IS_DEFAULT_INSTANCE): Move up and add
argument; update all callers.
(TT_Process_Simple_Glyph): Use it. The `unrounded' array is active
for variation fonts only, thus also enclose related code with
`#ifdef TT_CONFIG_OPTION_GX_VAR_SUPPORT ... #endif' where
necessary.
Revert commit a113e5d from 2019-05-09, and don't use `extra_points2'
but allocate a temporary array.
Speed up the scaling of the `unrounded' array.
* src/truetype/ttgxvar.c (FT_fixedToInt, FT_FixedToFdot6): Fix type
conversions and rounding. The unsigned type must have more or equal
bits to the signed type.
This patch make FreeType use font units in 26.6 format internally
instead of integers.
* src/truetype/ttgxvar.c (FT_fixedToFdot6): New macro.
(TT_Vary_Apply_Glyph_Deltas): Add argument to output unrounded font
coordinates.
* src/truetype/ttgxvar.h: Updated.
* src/truetype/ttgload.c (TT_Process_Simple_Glyph): Use
`extra_points2' array to temporarily hold unrounded point
coordinates; use them to compute scaled coordinates and linear
advance width and height.
(load_truetype_code): Adjust similarly.
Reported by Sender Ghost <lightside@gmx.com>.
* src/autofit/afcjk.c (af_cjk_hints_compute_edges): Catch a corner
case where `edge->first' could be NULL.
* src/pfr/pfrobjs.c (pfr_slot_load): Remove unnecessary test of
`size'.
* src/raster/ftraster.c (Draw_Sweep): Catch a corner case where
`draw_right' might be NULL.
* src/sfnt/ttmtx.c (tt_face_get_metrics): Fix limit test for
`aadvance'.
Ensure `abearing' always hold a meaningful result.
* src/truetype/ttgload.c (load_truetype_glyph): Ensure `subglyph' is
not NULL before accessing it.
* src/truetype/ttgxvar.c (TT_Set_Named_Instance): Remove unnecessary
test of `namedstyle'.
* src/type42/t42parse.c (t42_parser_done): Ensure
`parser->root.funcs.done' is not NULL before accessing it.
The previous implementation is correct but it is too complex.
The revised algorithm is based on the fact that each split moves
the control points closer to the trisection points on the chord.
The corresponding distances are good surrogates for the curve
deviation from the straight line.
This cubic flattening algorithm is somewhat similar to the conic
algorithm based the distance from the control point to the middle of
the chord. The cubic distances, however, decrease less predictably
but are easy enough to calculate on each step.
* src/smooth/ftgrays.c (gray_render_cubic): Replace the split
condition.
Alexei Podtelezhnikov