[arabic-table] Use segmented table

No functional change.

src/gen-arabic-table.py

@@ -71,42 +71,76 @@ def print_joining_table(f):
 	for value,short in short_value.items():
 		print "#define %s	%s" % (short, value)
 
-	keys = values.keys()
+	uu = sorted(values.keys())
-	min_u = min(keys)
-	max_u = max(keys)
 	num = len(values)
+
+	last = -1
+	ranges = []
+	for u in uu:
+		if u - last <= 1+16*3:
+			ranges[-1][-1] = u
+		else:
+			ranges.append([u,u])
+		last = u
+
 	print
 	print "static const uint8_t joining_table[] ="
 	print "{"
 	last_block = None
-	for u in range(min_u, max_u+1):
+	offset = 0
+	for start,end in ranges:
 
-		value = values.get(u, "JOINING_TYPE_X")
+		print
+		print "#define joining_offset_0x%04x %d" % (start, offset)
+
+		for u in range(start, end+1):
 
 			block = blocks.get(u, last_block)
-		if block != last_block:
+			value = values.get(u, "JOINING_TYPE_X")
-			print "\n\n  /* %s */" % block
+
+			if block != last_block or u == start:
+				if u != start:
+					print
+				print "\n  /* %s */" % block
 				last_block = block
 				if u % 32 != 0:
 					print
-				print "  /* %04X */" % u, "  " * (u % 32),
+					print "  /* %04X */" % (u//32*32), "  " * (u % 32),
 
 			if u % 32 == 0:
 				print
 				print "  /* %04X */ " % u,
 			sys.stdout.write("%s," % short_value[value])
 		print
-	print "};"
+
+		offset += end - start + 1
 	print
-	print "#define JOINING_TABLE_FIRST	0x%04X" % min_u
+	occupancy = num * 100. / offset
-	print "#define JOINING_TABLE_LAST	0x%04X" % max_u
+	print "}; /* Table items: %d; occupancy: %d%% */" % (offset, occupancy)
 	print
 
-	occupancy = num * 100 / (max_u - min_u + 1)
+	page_bits = 8
-	# Maintain at least 40% occupancy in the table */
+	print
-	if occupancy < 40:
+	print "static unsigned int"
-		raise Exception ("Table too sparse, please investigate: ", occupancy)
+	print "joining_type (hb_codepoint_t u)"
-
+	print "{"
+	print "  switch (u >> %d)" % page_bits
+	print "  {"
+	pages = set([u>>page_bits for u in [s for s,e in ranges]+[e for s,e in ranges]])
+	for p in sorted(pages):
+		print "    case 0x%0X:" % p
+		for (start,end) in ranges:
+			if p not in [start>>page_bits, end>>page_bits]: continue
+			offset = "joining_offset_0x%04x" % start
+			print "      if (0x%04X <= u && u <= 0x%04X) return joining_table[u - 0x%04X + %s];" % (start, end, start, offset)
+		print "      break;"
+		print ""
+	print "    default:"
+	print "      break;"
+	print "  }"
+	print "  return X;"
+	print "}"
+	print
 	for value,short in short_value.items():
 		print "#undef %s" % (short)
 	print

src/hb-ot-shape-complex-arabic-table.hh

@@ -28,6 +28,7 @@
 static const uint8_t joining_table[] =
 {
 
+#define joining_offset_0x0600 0
 
   /* Arabic */
 
@@ -48,37 +49,55 @@ static const uint8_t joining_table[] =
 
   /* Arabic Supplement */
 
-  /* 0750 */                                 D,D,D,D,D,D,D,D,D,R,R,R,D,D,D,D,
+  /* 0740 */                                 D,D,D,D,D,D,D,D,D,R,R,R,D,D,D,D,
   /* 0760 */ D,D,D,D,D,D,D,D,D,D,D,R,R,D,D,D,D,R,D,R,R,D,D,D,R,R,D,D,D,D,D,D,
 
-  /* Thaana */
+#define joining_offset_0x07ca 384
-
-  /* 0780 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
-  /* 07A0 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
 
   /* NKo */
 
-  /* 07C0 */ X,X,X,X,X,X,X,X,X,X,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
+  /* 07C0 */                     D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,D,
-  /* 07E0 */ D,D,D,D,D,D,D,D,D,D,D,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,C,X,X,X,X,X,
+  /* 07E0 */ D,D,D,D,D,D,D,D,D,D,D,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,C,
 
-  /* Samaritan */
+#define joining_offset_0x0840 433
-
-  /* 0800 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
-  /* 0820 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
 
   /* Mandaic */
 
-  /* 0840 */ R,D,D,D,D,D,R,D,D,R,D,D,D,D,D,R,D,D,D,D,R,D,U,U,U,X,X,X,X,X,X,X,
+  /* 0840 */ R,D,D,D,D,D,R,D,D,R,D,D,D,D,D,R,D,D,D,D,R,D,U,U,U,
-  /* 0860 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+
-  /* 0880 */ X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,X,
+#define joining_offset_0x08a0 458
 
   /* Arabic Extended-A */
 
   /* 08A0 */ D,X,D,D,D,D,D,D,D,D,R,R,R,
-};
 
-#define JOINING_TABLE_FIRST	0x0600
+}; /* Table items: 471; occupancy: 66% */
-#define JOINING_TABLE_LAST	0x08AC
+
+
+static unsigned int
+joining_type (hb_codepoint_t u)
+{
+  switch (u >> 8)
+  {
+    case 0x6:
+      if (0x0600 <= u && u <= 0x077F) return joining_table[u - 0x0600 + joining_offset_0x0600];
+      break;
+
+    case 0x7:
+      if (0x0600 <= u && u <= 0x077F) return joining_table[u - 0x0600 + joining_offset_0x0600];
+      if (0x07CA <= u && u <= 0x07FA) return joining_table[u - 0x07CA + joining_offset_0x07ca];
+      break;
+
+    case 0x8:
+      if (0x0840 <= u && u <= 0x0858) return joining_table[u - 0x0840 + joining_offset_0x0840];
+      if (0x08A0 <= u && u <= 0x08AC) return joining_table[u - 0x08A0 + joining_offset_0x08a0];
+      break;
+
+    default:
+      break;
+  }
+  return X;
+}
 
 #undef X
 #undef R

src/hb-ot-shape-complex-arabic.cc

@@ -57,11 +57,9 @@ enum {
 
 static unsigned int get_joining_type (hb_codepoint_t u, hb_unicode_general_category_t gen_cat)
 {
-  if (likely (hb_in_range<hb_codepoint_t> (u, JOINING_TABLE_FIRST, JOINING_TABLE_LAST))) {
+  unsigned int j_type = joining_type(u);
-    unsigned int j_type = joining_table[u - JOINING_TABLE_FIRST];
   if (likely (j_type != JOINING_TYPE_X))
     return j_type;
-  }
 
   /* Mongolian joining data is not in ArabicJoining.txt yet. */
   if (unlikely (hb_in_range<hb_codepoint_t> (u, 0x1800, 0x18AF)))