Shape text as HarfBuzz does, from a symbol file, checked against HarfBuzz

The build tool compiles a font's shaping plans the way HarfBuzz's planner
does (features, stages, masks, lookups per script and direction) and
carries them with the font's character map, advances and GDEF, GSUB and
GPOS tables in a section of the symbol file (VEGASYM4); glyphs shaping
reaches become symbols of their own. A small engine ports HarfBuzz 10's
default, Arabic and Hebrew shaping step for step and reads the tables in
place, every read bounds-checked. The build refuses a file whose shaping
differs from HarfBuzz's on random runs: DejaVu Sans agrees on 600,000
runs in six scripts both ways, and every installed font the tool accepts
agrees too, at 1.05-1.18 times HarfBuzz's time.

1 files changed, 18 insertions, 5 deletions

diff --git a/Makefile b/Makefile
index 29bbc27..119e116 100644
--- a/Makefile
+++ b/Makefile
@@ -39,24 +39,37 @@ $(BUILD)/vega-surface.so: $(SURFACES) surfaces/vega.h symbols/symbols.c symbols/
 		-fdata-sections -DVEGA_SURFACE -I$(PURE)/core -I$(PURE)/serve -Isymbols $(SURFACES) symbols/symbols.c \
 		$(LDFLAGS) -Wl,--gc-sections -lm $(LDLIBS) -o $@
 
-# Building symbol files reads fonts through FreeType; reading them does not.
+# Building symbol files reads fonts through FreeType, and compiles and checks
+# their shaping with HarfBuzz; reading them needs neither.
 FREETYPE_CFLAGS ?= $(shell pkg-config --cflags freetype2)
 FREETYPE_LIBS ?= $(shell pkg-config --libs freetype2)
+HARFBUZZ_CFLAGS ?= $(shell pkg-config --cflags harfbuzz)
+HARFBUZZ_LIBS ?= $(shell pkg-config --libs harfbuzz)
 # Drawing them on the GPU takes OpenGL ES 3.1 through EGL.
 GL_LIBS ?= $(shell pkg-config --libs egl glesv2)
 SYMBOLS = symbols/symbols.c symbols/symbols-gl.c
-$(BUILD)/vega-symbols: symbols/symbols-tool.c $(SYMBOLS) symbols/symbols.h symbols/symbols-gl.h | $(BUILD)
-	$(CC) $(CPPFLAGS) $(CFLAGS) $(FREETYPE_CFLAGS) symbols/symbols-tool.c $(SYMBOLS) $(LDFLAGS) \
-		$(FREETYPE_LIBS) $(GL_LIBS) -lm $(LDLIBS) -o $@
+SHAPE = text/shape.c text/unicode-data.c
+$(BUILD)/vega-symbols: symbols/symbols-tool.c $(SYMBOLS) symbols/symbols.h symbols/symbols-gl.h $(SHAPE) text/shape-build.c \
+		text/shape.h text/shape-build.h text/text.h | $(BUILD)
+	$(CC) $(CPPFLAGS) $(CFLAGS) $(FREETYPE_CFLAGS) $(HARFBUZZ_CFLAGS) -Itext symbols/symbols-tool.c $(SYMBOLS) $(SHAPE) \
+		text/shape-build.c $(LDFLAGS) $(FREETYPE_LIBS) $(HARFBUZZ_LIBS) $(GL_LIBS) -lm $(LDLIBS) -o $@
 
 TEXT = text/bidi.c text/normalize.c text/unicode-data.c
 $(BUILD)/test-text: text/test-text.c $(TEXT) text/text.h | $(BUILD)
 	$(CC) $(CPPFLAGS) $(CFLAGS) -Itext text/test-text.c $(TEXT) $(LDFLAGS) $(LDLIBS) -o $@
 
-# Unicode's conformance tests, where the UCD is installed (Debian: unicode-data).
+# Unicode's conformance tests, where the UCD is installed (Debian: unicode-data);
+# and shaping, where DejaVu Sans is: the build refuses a file whose shaping
+# differs from HarfBuzz's on random runs of Latin, Greek, Cyrillic, Hebrew,
+# Arabic and Common text, both ways.
 UCD ?= /usr/share/unicode
+SHAPED_FONT ?= /usr/share/fonts/truetype/dejavu/DejaVuSans.ttf
+SHAPED_CODES = 0x20-0x7e,0xa0-0x24f,0x300-0x36f,0x370-0x3ff,0x400-0x4ff,0x591-0x5f4,0x600-0x6ff,0x2000-0x206f
 check: $(BUILD)/vega $(BUILD)/vega-symbols $(BUILD)/vega-surface.so $(BUILD)/test-text
 	if [ -f $(UCD)/BidiTest.txt ]; then $(BUILD)/test-text $(UCD); else echo "no UCD at $(UCD): text tests skipped"; fi
+	if [ -f $(SHAPED_FONT) ]; then $(BUILD)/vega-symbols build --font $(SHAPED_FONT) --codes $(SHAPED_CODES) \
+		--shape Latn,Grek,Cyrl,Hebr,Arab,Zyyy --runs 5000 -o $(BUILD)/shaped.vsym; \
+		else echo "no $(SHAPED_FONT): shaping tests skipped"; fi
 	$(MAKE) -C $(AUP) build/aup
 	$(MAKE) -C $(PURE) build/pure-serve
 	python3 tests/check.py --vega $(BUILD)/vega --symbols $(BUILD)/vega-symbols --aup $(AUP)/build/aup --serve $(PURE)/build/pure-serve