Commit abb2f7305465e2e0de32cdeac81540224c428e84
2014-01-01T08:00:16
[autofit] Fix style assignments to glyphs. * src/autofit/hbshim.c (af_get_coverage) [FT_CONFIG_OPTION_USE_HARFBUZZ]: Scan GPOS coverage of features also so that we can skip glyphs that have both GSUB and GPOS data.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
diff --git a/ChangeLog b/ChangeLog
index 3ecde6e..ee5825f 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,13 @@
2014-01-01 Werner Lemberg <wl@gnu.org>
+ [autofit] Fix style assignments to glyphs.
+
+ * src/autofit/hbshim.c (af_get_coverage)
+ [FT_CONFIG_OPTION_USE_HARFBUZZ]: Scan GPOS coverage of features also
+ so that we can skip glyphs that have both GSUB and GPOS data.
+
+2014-01-01 Werner Lemberg <wl@gnu.org>
+
* src/autofit/hbshim.c: s/{lookups,glyphs}/gsub_{lookups,glyphs}/.
2014-01-01 Werner Lemberg <wl@gnu.org>
diff --git a/src/autofit/hbshim.c b/src/autofit/hbshim.c
index c8077ba..aec545b 100644
--- a/src/autofit/hbshim.c
+++ b/src/autofit/hbshim.c
@@ -102,6 +102,8 @@
hb_set_t* gsub_lookups; /* GSUB lookups for a given script */
hb_set_t* gsub_glyphs; /* glyphs covered by GSUB lookups */
+ hb_set_t* gpos_lookups; /* GPOS lookups for a given script */
+ hb_set_t* gpos_glyphs; /* glyphs covered by GPOS lookups */
hb_tag_t script;
const hb_tag_t* coverage_tags;
@@ -123,6 +125,8 @@
gsub_lookups = hb_set_create();
gsub_glyphs = hb_set_create();
+ gpos_lookups = hb_set_create();
+ gpos_glyphs = hb_set_create();
coverage_tags = coverages[style_class->coverage];
script = scripts[style_class->script];
@@ -162,6 +166,12 @@
NULL,
coverage_tags,
gsub_lookups );
+ hb_ot_layout_collect_lookups( face,
+ HB_OT_TAG_GPOS,
+ script_tags,
+ NULL,
+ coverage_tags,
+ gpos_lookups );
FT_TRACE4(( "GSUB lookups (style `%s'):\n"
" ",
@@ -178,6 +188,7 @@
count++;
#endif
+ /* get output coverage of GSUB feature */
hb_ot_layout_lookup_collect_glyphs( face,
HB_OT_TAG_GSUB,
idx,
@@ -191,9 +202,84 @@
if ( !count )
FT_TRACE4(( " (none)" ));
FT_TRACE4(( "\n\n" ));
+#endif
+
+ FT_TRACE4(( "GPOS lookups (style `%s'):\n"
+ " ",
+ af_style_names[style_class->style] ));
- FT_TRACE4(( " glyphs (`*' means already assigned)" ));
+#ifdef FT_DEBUG_LEVEL_TRACE
+ count = 0;
+#endif
+ for ( idx = -1; hb_set_next( gpos_lookups, &idx ); )
+ {
+#ifdef FT_DEBUG_LEVEL_TRACE
+ FT_TRACE4(( " %d", idx ));
+ count++;
+#endif
+
+ /* get input coverage of GPOS feature */
+ hb_ot_layout_lookup_collect_glyphs( face,
+ HB_OT_TAG_GPOS,
+ idx,
+ NULL,
+ gpos_glyphs,
+ NULL,
+ NULL );
+ }
+
+#ifdef FT_DEBUG_LEVEL_TRACE
+ if ( !count )
+ FT_TRACE4(( " (none)" ));
+ FT_TRACE4(( "\n\n" ));
+#endif
+
+ /*
+ * Various OpenType features might use the same glyphs at different
+ * vertical positions; for example, superscript and subscript glyphs
+ * could be the same. However, FreeType's auto-hinting is completely
+ * agnostic of OpenType features after the feature analysis has been
+ * completed: FreeType then simply receives a glyph index and returns a
+ * hinted and usually rendered glyph.
+ *
+ * Consider the superscript feature of font `pala.ttf': Some of the
+ * glyphs are `real', this is, they have a zero vertical offset, but
+ * most of them are small caps glyphs shifted up to the superscript
+ * position (this is, the `sups' feature is present in both the GSUB and
+ * GPOS tables). The code for blue zones computation actually uses a
+ * feature's y offset so that the `real' glyphs get correct hints. But
+ * later on it is impossible to decide whether a glyph index belongs to,
+ * say, the small caps or superscript feature.
+ *
+ * For this reason, we don't assign a style to a glyph if the current
+ * feature covers the glyph in both the GSUB and the GPOS tables. This
+ * is quite a broad condition, assuming that
+ *
+ * (a) glyphs that get used in multiple features are present in a
+ * feature without vertical shift,
+ *
+ * and
+ *
+ * (b) a feature's GPOS data really moves the glyph vertically.
+ *
+ * Not fulfilling condition (a) makes a font larger; it would also
+ * reduce the number of glyphs that could be addressed directly without
+ * using OpenType features, so this assumption is rather strong.
+ *
+ * Condition (b) is much weaker, and there might be glyphs which get
+ * missed. However, the OpenType features we are going to handle are
+ * primarily located in GSUB, and HarfBuzz doesn't provide an API to
+ * directly get the necessary information from the GPOS table. A
+ * possible solution might be to directly parse the GPOS table to find
+ * out whether a glyph gets shifted vertically, but this is something I
+ * would like to avoid if not really necessary.
+ *
+ */
+ hb_set_subtract( gsub_glyphs, gpos_glyphs );
+
+#ifdef FT_DEBUG_LEVEL_TRACE
+ FT_TRACE4(( " glyphs without GPOS data (`*' means already assigned)" ));
count = 0;
#endif
@@ -225,6 +311,8 @@
hb_set_destroy( gsub_lookups );
hb_set_destroy( gsub_glyphs );
+ hb_set_destroy( gpos_lookups );
+ hb_set_destroy( gpos_glyphs );
return FT_Err_Ok;
}