Shape Syriac, stretch with stch, and fall back for fonts without Arabic forms

The rest of HarfBuzz 10.2's Arabic shaper. Joining types now come from
HarfBuzz's own table (Unicode 16, with Syriac's Alaph and Dalath-Rish
groups the state machine's last two columns use), the general category
deciding where it has none, so Syriac shapes with fin2, fin3 and med2.
What stch multiplies is recorded and, after positioning, stretched under
the word before it, repeating pieces copied as often as fit. A font with
presentation forms but no Arabic features gets HarfBuzz's fallback
lookups, which the build tool writes as OpenType and the runtime applies
at their pause like any GSUB lookup.

The generator now reads HarfBuzz's generated tables through their lookup
functions' ranges and offsets: rows whose range starts mid-row have
blank cells in front, which the row comments alone miscounted (Arabic
Supplement); the Indic table, read again this way, did not change.

Noto Sans Syriac, FreeSans's Syriac, Unifont's Arabic and Noto Sans
Arabic shape as HarfBuzz does at 3,000 runs a plan; no installed font
is refused any more (213 fonts, 1,822 plans).

1 files changed, 17 insertions, 4 deletions

diff --git a/text/shape.h b/text/shape.h
index 6e56819..9381754 100644
--- a/text/shape.h
+++ b/text/shape.h
@@ -31,7 +31,12 @@
  *                                          y_bearing, width, height; 0 when
  *                                          no plan positions marks without
  *                                          GPOS; x_bearing INT16_MIN: none),
- *   u32 0
+ *   u32 fallback_offset                   (Arabic fallback shaping, 0 when no
+ *                                          plan falls back: u8 lookups, u8
+ *                                          feature[7] each is for (init,
+ *                                          medi, fina, isol, rlig x3), u32
+ *                                          size, then a GSUB table of just
+ *                                          those lookups)
  * Offsets are from the section's start. A plan (VEGA_PLAN_BYTES):
  *   u32 script (ISO 15924 as a tag: 'Latn'), u8 direction, u8 native
  *   direction (of its script; 0 when it has none), u8 shaper, u8 flags,
@@ -67,18 +72,22 @@ enum { VEGA_MASK_RPHF = VEGA_MASK_SHAPER, VEGA_MASK_PREF, VEGA_MASK_BLWF, VEGA_M
 enum { VEGA_MASK_KHMER_PREF = VEGA_MASK_SHAPER, VEGA_MASK_KHMER_BLWF, VEGA_MASK_KHMER_ABVF, VEGA_MASK_KHMER_PSTF,
        VEGA_MASK_KHMER_CFAR };                                                /* Khmer */
 enum { VEGA_MASK_USE_RPHF = VEGA_MASK_INIT + 1 };         /* USE: that, and Arabic's for its joining forms */
+enum { VEGA_MASK_ARABIC_RLIG = VEGA_MASK_INIT + 1 };      /* Arabic, for fallback shaping */
 /* The stages a plan names: Indic's, whose features decide a consonant's
  * position by whether they would substitute it. */
 enum { VEGA_STAGE_RPHF, VEGA_STAGE_PREF, VEGA_STAGE_BLWF, VEGA_STAGE_PSTF, VEGA_STAGE_VATU };
 /* Shaper flags: Indic's font follows the old specification (deva, not
  * dev2); Thai's font has no Thai script of its own, so its marks take the
- * font's private-use glyphs where it has them. */
-enum { VEGA_INDIC_OLD_SPEC = 1, VEGA_THAI_PUA = 2 };
+ * font's private-use glyphs where it has them; Arabic's font lacks Arabic's
+ * forms, so lookups made from its presentation forms stand in; Arabic's
+ * font stretches (stch) what the stretching feature multiplied. */
+enum { VEGA_INDIC_OLD_SPEC = 1, VEGA_THAI_PUA = 2, VEGA_ARABIC_FALLBACK = 4, VEGA_ARABIC_STCH = 8 };
 /* Step flags; pauses, which the shaper tells apart. */
 enum { VEGA_STEP_AUTO_ZWNJ = 1, VEGA_STEP_AUTO_ZWJ = 2, VEGA_STEP_RANDOM = 4, VEGA_STEP_PER_SYLLABLE = 8 };
 enum { VEGA_STEP_PAUSE = 0xFFFF };
 enum { VEGA_PAUSE_SYLLABLES = 1, VEGA_PAUSE_INITIAL_REORDERING, VEGA_PAUSE_FINAL_REORDERING, VEGA_PAUSE_CLEAR_SYLLABLES,
-       VEGA_PAUSE_CLEAR_SUBSTITUTED, VEGA_PAUSE_RECORD_RPHF, VEGA_PAUSE_RECORD_PREF };
+       VEGA_PAUSE_CLEAR_SUBSTITUTED, VEGA_PAUSE_RECORD_RPHF, VEGA_PAUSE_RECORD_PREF, VEGA_PAUSE_ARABIC_FALLBACK,
+       VEGA_PAUSE_RECORD_STCH };
 
 typedef struct { const unsigned char *p; uint32_t size; } VegaTable;
 /* Which glyphs may be in a set, as three 64-bit patterns of glyph bits
@@ -103,6 +112,10 @@ typedef struct {
     uint32_t cmap_count, plan_count;
     const unsigned char *cmap, *advances, *plans, *steps, *extents;
     VegaTable gdef, gsub, gpos, kern;
+    VegaTable fallback;             /* Arabic fallback shaping's lookups, a GSUB table */
+    uint32_t fallback_count;
+    uint8_t fallback_features[7];   /* what each is for: init, medi, fina, isol, rlig x3 (0..6) */
+    VegaLookup *fallback_lookups;
     uint32_t glyph_classes, attach_classes, mark_sets; /* in GDEF; 0 when absent */
     VegaLookup *lookups[2];         /* GSUB, GPOS */
     uint32_t lookup_count[2];