Shape Thai and Lao as HarfBuzz does

The Thai shaper, from HarfBuzz 10.2's source: SARA AM comes apart into
NIKHAHIT and SARA AA, the NIKHAHIT moving back over the above-base marks
(and counting as a mark when advances are zeroed). A font with no Thai
script of its own (only DFLT or latn in GSUB) gets its marks shifted to
its private-use glyphs, Windows' before the Mac's, as the stacking
around the base asks; the planner records, as HarfBuzz's does, whether
the font had the script under its own tag.

Noto Sans Thai and Lao, TLWG Loma, FreeSerif's Thai, DejaVu Sans's Lao
and Unifont shape as HarfBuzz does at 3,000 runs a plan, and Loma with
its thai script taken out (so the private-use path runs) at 5,000.

1 files changed, 8 insertions, 4 deletions

diff --git a/text/shape-build.c b/text/shape-build.c
index a3b924e..96bc748 100644
--- a/text/shape-build.c
+++ b/text/shape-build.c
@@ -33,6 +33,7 @@ typedef struct {
     hb_face_t *face;
     unsigned script_index[2], language_index[2];
     hb_tag_t chosen_script[2];
+    bool found_script[2]; /* one of the script's own tags, not DFLT or latn */
     FeatureInfo infos[MAX_FEATURES];
     unsigned info_count, current_stage[2];
     unsigned pause[2][MAX_STAGES]; /* what ends each stage: VEGA_PAUSE_..., or 0 */
@@ -332,7 +333,8 @@ static int categorize(uint32_t script, hb_tag_t gsub_script) {
     if (script == TAG("Hebr")) return VEGA_SHAPER_HEBREW;
     if (script == TAG("Hang")) return VEGA_SHAPER_HANGUL;
     if (script == TAG("Khmr")) return VEGA_SHAPER_KHMER;
-    if (script == TAG("Syrc") || script == TAG("Thai") || script == TAG("Laoo") || script == TAG("Qaag")) return -1;
+    if (script == TAG("Thai") || script == TAG("Laoo")) return VEGA_SHAPER_THAI;
+    if (script == TAG("Syrc") || script == TAG("Qaag")) return -1;
     if (listed(indic_scripts, sizeof(indic_scripts) / sizeof(*indic_scripts), script))
         return dflt ? VEGA_SHAPER_DEFAULT : (gsub_script & 0xFF) == '3' ? VEGA_SHAPER_USE : VEGA_SHAPER_INDIC;
     if (listed(use_scripts, sizeof(use_scripts) / sizeof(*use_scripts), script))
@@ -430,8 +432,8 @@ static bool plan_for(hb_face_t *face, uint32_t script_tag, int direction, Plan *
     hb_ot_tags_from_script_and_language(script, HB_LANGUAGE_INVALID, &script_count, script_tags, &language_count,
                                         language_tags);
     for (int t = 0; t < 2; ++t) {
-        hb_ot_layout_table_select_script(face, table_tags[t], script_count, script_tags, &m->script_index[t],
-                                         &m->chosen_script[t]);
+        m->found_script[t] = hb_ot_layout_table_select_script(face, table_tags[t], script_count, script_tags,
+                                                              &m->script_index[t], &m->chosen_script[t]);
         hb_ot_layout_script_select_language(face, table_tags[t], m->script_index[t], language_count, language_tags,
                                             &m->language_index[t]);
     }
@@ -610,7 +612,7 @@ static bool plan_for(hb_face_t *face, uint32_t script_tag, int direction, Plan *
     /* Hangul's, Indic's and Khmer's shapers zero no marks (Myanmar's does,
      * before GPOS); none of these places marks without GPOS. */
     bool zeroes = shaper != VEGA_SHAPER_HANGUL && shaper != VEGA_SHAPER_INDIC && shaper != VEGA_SHAPER_KHMER;
-    bool script_marks = zeroes && shaper != VEGA_SHAPER_MYANMAR && shaper != VEGA_SHAPER_USE;
+    bool script_marks = zeroes && shaper != VEGA_SHAPER_MYANMAR && shaper != VEGA_SHAPER_USE && shaper != VEGA_SHAPER_THAI;
     bool adjust = !apply_gpos && (!apply_kern || !kern_table.cross), fallback_marks = adjust && script_marks;
     if (shaper == VEGA_SHAPER_ARABIC) {
         if (one_mask(m, "stch")) FAIL("%s: the font stretches (stch)", name);
@@ -655,6 +657,8 @@ static bool plan_for(hb_face_t *face, uint32_t script_tag, int direction, Plan *
     }
     if (shaper == VEGA_SHAPER_KHMER)
         for (int i = 0; i < 5; ++i) plan->masks[VEGA_MASK_KHMER_PREF + i] = one_mask(m, khmer_basic[i]);
+    if (shaper == VEGA_SHAPER_THAI && script == HB_SCRIPT_THAI && !m->found_script[0])
+        plan->shaper_flags |= VEGA_THAI_PUA;
     free(m);
     return true;
 fail: