Move enum types to hb-common.h

src/hb-common.c

@@ -141,4 +141,410 @@ hb_language_to_string (hb_language_t language)
 }
 
 
+/* hb_script_t */
+
+static const hb_tag_t script_to_iso15924_tag[] =
+{
+  HB_TAG('Z','y','y','y'),	/* HB_SCRIPT_COMMON */
+  HB_TAG('Q','a','a','i'),	/* HB_SCRIPT_INHERITED */
+  HB_TAG('A','r','a','b'),	/* HB_SCRIPT_ARABIC */
+  HB_TAG('A','r','m','n'),	/* HB_SCRIPT_ARMENIAN */
+  HB_TAG('B','e','n','g'),	/* HB_SCRIPT_BENGALI */
+  HB_TAG('B','o','p','o'),	/* HB_SCRIPT_BOPOMOFO */
+  HB_TAG('C','h','e','r'),	/* HB_SCRIPT_CHEROKEE */
+  HB_TAG('Q','a','a','c'),	/* HB_SCRIPT_COPTIC */
+  HB_TAG('C','y','r','l'),	/* HB_SCRIPT_CYRILLIC */
+  HB_TAG('D','s','r','t'),	/* HB_SCRIPT_DESERET */
+  HB_TAG('D','e','v','a'),	/* HB_SCRIPT_DEVANAGARI */
+  HB_TAG('E','t','h','i'),	/* HB_SCRIPT_ETHIOPIC */
+  HB_TAG('G','e','o','r'),	/* HB_SCRIPT_GEORGIAN */
+  HB_TAG('G','o','t','h'),	/* HB_SCRIPT_GOTHIC */
+  HB_TAG('G','r','e','k'),	/* HB_SCRIPT_GREEK */
+  HB_TAG('G','u','j','r'),	/* HB_SCRIPT_GUJARATI */
+  HB_TAG('G','u','r','u'),	/* HB_SCRIPT_GURMUKHI */
+  HB_TAG('H','a','n','i'),	/* HB_SCRIPT_HAN */
+  HB_TAG('H','a','n','g'),	/* HB_SCRIPT_HANGUL */
+  HB_TAG('H','e','b','r'),	/* HB_SCRIPT_HEBREW */
+  HB_TAG('H','i','r','a'),	/* HB_SCRIPT_HIRAGANA */
+  HB_TAG('K','n','d','a'),	/* HB_SCRIPT_KANNADA */
+  HB_TAG('K','a','n','a'),	/* HB_SCRIPT_KATAKANA */
+  HB_TAG('K','h','m','r'),	/* HB_SCRIPT_KHMER */
+  HB_TAG('L','a','o','o'),	/* HB_SCRIPT_LAO */
+  HB_TAG('L','a','t','n'),	/* HB_SCRIPT_LATIN */
+  HB_TAG('M','l','y','m'),	/* HB_SCRIPT_MALAYALAM */
+  HB_TAG('M','o','n','g'),	/* HB_SCRIPT_MONGOLIAN */
+  HB_TAG('M','y','m','r'),	/* HB_SCRIPT_MYANMAR */
+  HB_TAG('O','g','a','m'),	/* HB_SCRIPT_OGHAM */
+  HB_TAG('I','t','a','l'),	/* HB_SCRIPT_OLD_ITALIC */
+  HB_TAG('O','r','y','a'),	/* HB_SCRIPT_ORIYA */
+  HB_TAG('R','u','n','r'),	/* HB_SCRIPT_RUNIC */
+  HB_TAG('S','i','n','h'),	/* HB_SCRIPT_SINHALA */
+  HB_TAG('S','y','r','c'),	/* HB_SCRIPT_SYRIAC */
+  HB_TAG('T','a','m','l'),	/* HB_SCRIPT_TAMIL */
+  HB_TAG('T','e','l','u'),	/* HB_SCRIPT_TELUGU */
+  HB_TAG('T','h','a','a'),	/* HB_SCRIPT_THAANA */
+  HB_TAG('T','h','a','i'),	/* HB_SCRIPT_THAI */
+  HB_TAG('T','i','b','t'),	/* HB_SCRIPT_TIBETAN */
+  HB_TAG('C','a','n','s'),	/* HB_SCRIPT_CANADIAN_ABORIGINAL */
+  HB_TAG('Y','i','i','i'),	/* HB_SCRIPT_YI */
+  HB_TAG('T','g','l','g'),	/* HB_SCRIPT_TAGALOG */
+  HB_TAG('H','a','n','o'),	/* HB_SCRIPT_HANUNOO */
+  HB_TAG('B','u','h','d'),	/* HB_SCRIPT_BUHID */
+  HB_TAG('T','a','g','b'),	/* HB_SCRIPT_TAGBANWA */
+
+  /* Unicode-4.0 additions */
+  HB_TAG('B','r','a','i'),	/* HB_SCRIPT_BRAILLE */
+  HB_TAG('C','p','r','t'),	/* HB_SCRIPT_CYPRIOT */
+  HB_TAG('L','i','m','b'),	/* HB_SCRIPT_LIMBU */
+  HB_TAG('O','s','m','a'),	/* HB_SCRIPT_OSMANYA */
+  HB_TAG('S','h','a','w'),	/* HB_SCRIPT_SHAVIAN */
+  HB_TAG('L','i','n','b'),	/* HB_SCRIPT_LINEAR_B */
+  HB_TAG('T','a','l','e'),	/* HB_SCRIPT_TAI_LE */
+  HB_TAG('U','g','a','r'),	/* HB_SCRIPT_UGARITIC */
+
+  /* Unicode-4.1 additions */
+  HB_TAG('T','a','l','u'),	/* HB_SCRIPT_NEW_TAI_LUE */
+  HB_TAG('B','u','g','i'),	/* HB_SCRIPT_BUGINESE */
+  HB_TAG('G','l','a','g'),	/* HB_SCRIPT_GLAGOLITIC */
+  HB_TAG('T','f','n','g'),	/* HB_SCRIPT_TIFINAGH */
+  HB_TAG('S','y','l','o'),	/* HB_SCRIPT_SYLOTI_NAGRI */
+  HB_TAG('X','p','e','o'),	/* HB_SCRIPT_OLD_PERSIAN */
+  HB_TAG('K','h','a','r'),	/* HB_SCRIPT_KHAROSHTHI */
+
+  /* Unicode-5.0 additions */
+  HB_TAG('Z','z','z','z'),	/* HB_SCRIPT_UNKNOWN */
+  HB_TAG('B','a','l','i'),	/* HB_SCRIPT_BALINESE */
+  HB_TAG('X','s','u','x'),	/* HB_SCRIPT_CUNEIFORM */
+  HB_TAG('P','h','n','x'),	/* HB_SCRIPT_PHOENICIAN */
+  HB_TAG('P','h','a','g'),	/* HB_SCRIPT_PHAGS_PA */
+  HB_TAG('N','k','o','o'),	/* HB_SCRIPT_NKO */
+
+  /* Unicode-5.1 additions */
+  HB_TAG('K','a','l','i'),	/* HB_SCRIPT_KAYAH_LI */
+  HB_TAG('L','e','p','c'),	/* HB_SCRIPT_LEPCHA */
+  HB_TAG('R','j','n','g'),	/* HB_SCRIPT_REJANG */
+  HB_TAG('S','u','n','d'),	/* HB_SCRIPT_SUNDANESE */
+  HB_TAG('S','a','u','r'),	/* HB_SCRIPT_SAURASHTRA */
+  HB_TAG('C','h','a','m'),	/* HB_SCRIPT_CHAM */
+  HB_TAG('O','l','c','k'),	/* HB_SCRIPT_OL_CHIKI */
+  HB_TAG('V','a','i','i'),	/* HB_SCRIPT_VAI */
+  HB_TAG('C','a','r','i'),	/* HB_SCRIPT_CARIAN */
+  HB_TAG('L','y','c','i'),	/* HB_SCRIPT_LYCIAN */
+  HB_TAG('L','y','d','i'),	/* HB_SCRIPT_LYDIAN */
+
+  /* Unicode-5.2 additions */
+  HB_TAG('A','v','s','t'),	/* HB_SCRIPT_AVESTAN */
+  HB_TAG('B','a','m','u'),	/* HB_SCRIPT_BAMUM */
+  HB_TAG('E','g','y','p'),	/* HB_SCRIPT_EGYPTIAN_HIEROGLYPHS */
+  HB_TAG('A','r','m','i'),	/* HB_SCRIPT_IMPERIAL_ARAMAIC */
+  HB_TAG('P','h','l','i'),	/* HB_SCRIPT_INSCRIPTIONAL_PAHLAVI */
+  HB_TAG('P','r','t','i'),	/* HB_SCRIPT_INSCRIPTIONAL_PARTHIAN */
+  HB_TAG('J','a','v','a'),	/* HB_SCRIPT_JAVANESE */
+  HB_TAG('K','t','h','i'),	/* HB_SCRIPT_KAITHI */
+  HB_TAG('L','i','s','u'),	/* HB_SCRIPT_LISU */
+  HB_TAG('M','t','e','i'),	/* HB_SCRIPT_MEETEI_MAYEK */
+  HB_TAG('S','a','r','b'),	/* HB_SCRIPT_OLD_SOUTH_ARABIAN */
+  HB_TAG('O','r','k','h'),	/* HB_SCRIPT_OLD_TURKIC */
+  HB_TAG('S','a','m','r'),	/* HB_SCRIPT_SAMARITAN */
+  HB_TAG('L','a','n','a'),	/* HB_SCRIPT_TAI_THAM */
+  HB_TAG('T','a','v','t'),	/* HB_SCRIPT_TAI_VIET */
+
+  /* Unicode-6.0 additions */
+  HB_TAG('B','a','t','k'),	/* HB_SCRIPT_BATAK */
+  HB_TAG('B','r','a','h'),	/* HB_SCRIPT_BRAHMI */
+  HB_TAG('M','a','n','d') 	/* HB_SCRIPT_MANDAIC */
+};
+
+struct tag_script_pair {
+  hb_tag_t tag;
+  hb_script_t script;
+};
+static const struct tag_script_pair script_from_iso15924_tag[] =
+{
+  {HB_TAG('A','r','a','b'), HB_SCRIPT_ARABIC},
+  {HB_TAG('A','r','m','i'), HB_SCRIPT_IMPERIAL_ARAMAIC},
+  {HB_TAG('A','r','m','n'), HB_SCRIPT_ARMENIAN},
+  {HB_TAG('A','v','s','t'), HB_SCRIPT_AVESTAN},
+  {HB_TAG('B','a','l','i'), HB_SCRIPT_BALINESE},
+  {HB_TAG('B','a','m','u'), HB_SCRIPT_BAMUM},
+  {HB_TAG('B','a','t','k'), HB_SCRIPT_BATAK},
+  {HB_TAG('B','e','n','g'), HB_SCRIPT_BENGALI},
+  {HB_TAG('B','o','p','o'), HB_SCRIPT_BOPOMOFO},
+  {HB_TAG('B','r','a','h'), HB_SCRIPT_BRAHMI},
+  {HB_TAG('B','r','a','i'), HB_SCRIPT_BRAILLE},
+  {HB_TAG('B','u','g','i'), HB_SCRIPT_BUGINESE},
+  {HB_TAG('B','u','h','d'), HB_SCRIPT_BUHID},
+  {HB_TAG('C','a','n','s'), HB_SCRIPT_CANADIAN_ABORIGINAL},
+  {HB_TAG('C','a','r','i'), HB_SCRIPT_CARIAN},
+  {HB_TAG('C','h','a','m'), HB_SCRIPT_CHAM},
+  {HB_TAG('C','h','e','r'), HB_SCRIPT_CHEROKEE},
+  {HB_TAG('C','p','r','t'), HB_SCRIPT_CYPRIOT},
+  {HB_TAG('C','y','r','l'), HB_SCRIPT_CYRILLIC},
+  {HB_TAG('C','y','r','s'), HB_SCRIPT_CYRILLIC},
+  {HB_TAG('D','e','v','a'), HB_SCRIPT_DEVANAGARI},
+  {HB_TAG('D','s','r','t'), HB_SCRIPT_DESERET},
+  {HB_TAG('E','g','y','p'), HB_SCRIPT_EGYPTIAN_HIEROGLYPHS},
+  {HB_TAG('E','t','h','i'), HB_SCRIPT_ETHIOPIC},
+  {HB_TAG('G','e','o','a'), HB_SCRIPT_GEORGIAN},
+  {HB_TAG('G','e','o','n'), HB_SCRIPT_GEORGIAN},
+  {HB_TAG('G','e','o','r'), HB_SCRIPT_GEORGIAN},
+  {HB_TAG('G','l','a','g'), HB_SCRIPT_GLAGOLITIC},
+  {HB_TAG('G','o','t','h'), HB_SCRIPT_GOTHIC},
+  {HB_TAG('G','r','e','k'), HB_SCRIPT_GREEK},
+  {HB_TAG('G','u','j','r'), HB_SCRIPT_GUJARATI},
+  {HB_TAG('G','u','r','u'), HB_SCRIPT_GURMUKHI},
+  {HB_TAG('H','a','n','g'), HB_SCRIPT_HANGUL},
+  {HB_TAG('H','a','n','i'), HB_SCRIPT_HAN},
+  {HB_TAG('H','a','n','o'), HB_SCRIPT_HANUNOO},
+  {HB_TAG('H','e','b','r'), HB_SCRIPT_HEBREW},
+  {HB_TAG('H','i','r','a'), HB_SCRIPT_HIRAGANA},
+  {HB_TAG('I','t','a','l'), HB_SCRIPT_OLD_ITALIC},
+  {HB_TAG('J','a','v','a'), HB_SCRIPT_JAVANESE},
+  {HB_TAG('K','a','l','i'), HB_SCRIPT_KAYAH_LI},
+  {HB_TAG('K','a','n','a'), HB_SCRIPT_KATAKANA},
+  {HB_TAG('K','h','a','r'), HB_SCRIPT_KHAROSHTHI},
+  {HB_TAG('K','h','m','r'), HB_SCRIPT_KHMER},
+  {HB_TAG('K','n','d','a'), HB_SCRIPT_KANNADA},
+  {HB_TAG('K','t','h','i'), HB_SCRIPT_KAITHI},
+  {HB_TAG('L','a','n','a'), HB_SCRIPT_TAI_THAM},
+  {HB_TAG('L','a','o','o'), HB_SCRIPT_LAO},
+  {HB_TAG('L','a','t','f'), HB_SCRIPT_LATIN},
+  {HB_TAG('L','a','t','g'), HB_SCRIPT_LATIN},
+  {HB_TAG('L','a','t','n'), HB_SCRIPT_LATIN},
+  {HB_TAG('L','e','p','c'), HB_SCRIPT_LEPCHA},
+  {HB_TAG('L','i','m','b'), HB_SCRIPT_LIMBU},
+  {HB_TAG('L','i','n','b'), HB_SCRIPT_LINEAR_B},
+  {HB_TAG('L','i','s','u'), HB_SCRIPT_LISU},
+  {HB_TAG('L','y','c','i'), HB_SCRIPT_LYCIAN},
+  {HB_TAG('L','y','d','i'), HB_SCRIPT_LYDIAN},
+  {HB_TAG('M','a','n','d'), HB_SCRIPT_MANDAIC},
+  {HB_TAG('M','l','y','m'), HB_SCRIPT_MALAYALAM},
+  {HB_TAG('M','o','n','g'), HB_SCRIPT_MONGOLIAN},
+  {HB_TAG('M','t','e','i'), HB_SCRIPT_MEETEI_MAYEK},
+  {HB_TAG('M','y','m','r'), HB_SCRIPT_MYANMAR},
+  {HB_TAG('N','k','o','o'), HB_SCRIPT_NKO},
+  {HB_TAG('O','g','a','m'), HB_SCRIPT_OGHAM},
+  {HB_TAG('O','l','c','k'), HB_SCRIPT_OL_CHIKI},
+  {HB_TAG('O','r','k','h'), HB_SCRIPT_OLD_TURKIC},
+  {HB_TAG('O','r','y','a'), HB_SCRIPT_ORIYA},
+  {HB_TAG('O','s','m','a'), HB_SCRIPT_OSMANYA},
+  {HB_TAG('P','h','a','g'), HB_SCRIPT_PHAGS_PA},
+  {HB_TAG('P','h','l','i'), HB_SCRIPT_INSCRIPTIONAL_PAHLAVI},
+  {HB_TAG('P','h','n','x'), HB_SCRIPT_PHOENICIAN},
+  {HB_TAG('P','r','t','i'), HB_SCRIPT_INSCRIPTIONAL_PARTHIAN},
+  {HB_TAG('Q','a','a','c'), HB_SCRIPT_COPTIC},
+  {HB_TAG('Q','a','a','i'), HB_SCRIPT_INHERITED},
+  {HB_TAG('R','j','n','g'), HB_SCRIPT_REJANG},
+  {HB_TAG('R','u','n','r'), HB_SCRIPT_RUNIC},
+  {HB_TAG('S','a','m','r'), HB_SCRIPT_SAMARITAN},
+  {HB_TAG('S','a','r','b'), HB_SCRIPT_OLD_SOUTH_ARABIAN},
+  {HB_TAG('S','a','u','r'), HB_SCRIPT_SAURASHTRA},
+  {HB_TAG('S','h','a','w'), HB_SCRIPT_SHAVIAN},
+  {HB_TAG('S','i','n','h'), HB_SCRIPT_SINHALA},
+  {HB_TAG('S','u','n','d'), HB_SCRIPT_SUNDANESE},
+  {HB_TAG('S','y','l','o'), HB_SCRIPT_SYLOTI_NAGRI},
+  {HB_TAG('S','y','r','c'), HB_SCRIPT_SYRIAC},
+  {HB_TAG('S','y','r','e'), HB_SCRIPT_SYRIAC},
+  {HB_TAG('S','y','r','n'), HB_SCRIPT_SYRIAC},
+  {HB_TAG('T','a','g','b'), HB_SCRIPT_TAGBANWA},
+  {HB_TAG('T','a','l','e'), HB_SCRIPT_TAI_LE},
+  {HB_TAG('T','a','l','u'), HB_SCRIPT_NEW_TAI_LUE},
+  {HB_TAG('T','a','m','l'), HB_SCRIPT_TAMIL},
+  {HB_TAG('T','a','v','t'), HB_SCRIPT_TAI_VIET},
+  {HB_TAG('T','e','l','u'), HB_SCRIPT_TELUGU},
+  {HB_TAG('T','f','n','g'), HB_SCRIPT_TIFINAGH},
+  {HB_TAG('T','g','l','g'), HB_SCRIPT_TAGALOG},
+  {HB_TAG('T','h','a','a'), HB_SCRIPT_THAANA},
+  {HB_TAG('T','h','a','i'), HB_SCRIPT_THAI},
+  {HB_TAG('T','i','b','t'), HB_SCRIPT_TIBETAN},
+  {HB_TAG('U','g','a','r'), HB_SCRIPT_UGARITIC},
+  {HB_TAG('V','a','i','i'), HB_SCRIPT_VAI},
+  {HB_TAG('X','p','e','o'), HB_SCRIPT_OLD_PERSIAN},
+  {HB_TAG('X','s','u','x'), HB_SCRIPT_CUNEIFORM},
+  {HB_TAG('Y','i','i','i'), HB_SCRIPT_YI},
+  {HB_TAG('Z','y','y','y'), HB_SCRIPT_COMMON},
+  {HB_TAG('Z','z','z','z'), HB_SCRIPT_UNKNOWN}
+};
+
+static int
+_tag_cmp (hb_tag_t *pa, hb_tag_t *pb)
+{
+  hb_tag_t a = *pa, b = *pb;
+  return a < b ? -1 : a == b ? 0 : +1;
+}
+
+
+hb_script_t
+hb_script_from_iso15924_tag (hb_tag_t tag)
+{
+  const struct tag_script_pair *pair;
+
+  /* Be lenient, adjust case (one capital letter followed by three small letters) */
+  tag = (tag & 0xDFDFDFDF) | 0x00202020;
+
+  pair = (const struct tag_script_pair *) bsearch (&tag,
+						   script_from_iso15924_tag,
+						   ARRAY_LENGTH (script_from_iso15924_tag),
+						   sizeof (script_from_iso15924_tag[0]),
+						   (hb_compare_func_t) _tag_cmp);
+
+  if (pair)
+    return pair->script;
+
+  /* If it looks right, just use the tag as a script */
+  if (((uint32_t) tag & 0xE0E0E0E0) == 0x40606060)
+    return (hb_script_t) tag;
+
+  /* Otherwise, return unknown */
+  return HB_SCRIPT_UNKNOWN;
+}
+
+hb_script_t
+hb_script_from_string (const char *s)
+{
+  return hb_script_from_iso15924_tag (hb_tag_from_string (s));
+}
+
+hb_tag_t
+hb_script_to_iso15924_tag (hb_script_t script)
+{
+  if (likely ((unsigned int) script < ARRAY_LENGTH (script_to_iso15924_tag)))
+    return script_to_iso15924_tag[script];
+
+  /* if script is of the right shape (one capital letter followed by three small letters),
+   * return as is. */
+  if (((uint32_t) script & 0xE0E0E0E0) == 0x40606060)
+    return (hb_tag_t) script;
+
+  /* Otherwise, we don't know what that is */
+  return script_to_iso15924_tag[HB_SCRIPT_UNKNOWN];
+}
+
+
+#define LTR HB_DIRECTION_LTR
+#define RTL HB_DIRECTION_RTL
+const hb_direction_t horiz_dir[] =
+{
+  LTR,	/* Zyyy */
+  LTR,	/* Qaai */
+  RTL,	/* Arab */
+  LTR,	/* Armn */
+  LTR,	/* Beng */
+  LTR,	/* Bopo */
+  LTR,	/* Cher */
+  LTR,	/* Qaac */
+  LTR,	/* Cyrl (Cyrs) */
+  LTR,	/* Dsrt */
+  LTR,	/* Deva */
+  LTR,	/* Ethi */
+  LTR,	/* Geor (Geon, Geoa) */
+  LTR,	/* Goth */
+  LTR,	/* Grek */
+  LTR,	/* Gujr */
+  LTR,	/* Guru */
+  LTR,	/* Hani */
+  LTR,	/* Hang */
+  RTL,	/* Hebr */
+  LTR,	/* Hira */
+  LTR,	/* Knda */
+  LTR,	/* Kana */
+  LTR,	/* Khmr */
+  LTR,	/* Laoo */
+  LTR,	/* Latn (Latf, Latg) */
+  LTR,	/* Mlym */
+  LTR,	/* Mong */
+  LTR,	/* Mymr */
+  LTR,	/* Ogam */
+  LTR,	/* Ital */
+  LTR,	/* Orya */
+  LTR,	/* Runr */
+  LTR,	/* Sinh */
+  RTL,	/* Syrc (Syrj, Syrn, Syre) */
+  LTR,	/* Taml */
+  LTR,	/* Telu */
+  RTL,	/* Thaa */
+  LTR,	/* Thai */
+  LTR,	/* Tibt */
+  LTR,	/* Cans */
+  LTR,	/* Yiii */
+  LTR,	/* Tglg */
+  LTR,	/* Hano */
+  LTR,	/* Buhd */
+  LTR,	/* Tagb */
+
+  /* Unicode-4.0 additions */
+  LTR,	/* Brai */
+  RTL,	/* Cprt */
+  LTR,	/* Limb */
+  LTR,	/* Osma */
+  LTR,	/* Shaw */
+  LTR,	/* Linb */
+  LTR,	/* Tale */
+  LTR,	/* Ugar */
+
+  /* Unicode-4.1 additions */
+  LTR,	/* Talu */
+  LTR,	/* Bugi */
+  LTR,	/* Glag */
+  LTR,	/* Tfng */
+  LTR,	/* Sylo */
+  LTR,	/* Xpeo */
+  LTR,	/* Khar */
+
+  /* Unicode-5.0 additions */
+  LTR,	/* Zzzz */
+  LTR,	/* Bali */
+  LTR,	/* Xsux */
+  RTL,	/* Phnx */
+  LTR,	/* Phag */
+  RTL,	/* Nkoo */
+
+  /* Unicode-5.1 additions */
+  LTR,	/* Kali */
+  LTR,	/* Lepc */
+  LTR,	/* Rjng */
+  LTR,	/* Sund */
+  LTR,	/* Saur */
+  LTR,	/* Cham */
+  LTR,	/* Olck */
+  LTR,	/* Vaii */
+  LTR,	/* Cari */
+  LTR,	/* Lyci */
+  LTR,	/* Lydi */
+
+  /* Unicode-5.2 additions */
+  RTL,	/* Avst */
+  LTR,	/* Bamu */
+  LTR,	/* Egyp */
+  RTL,	/* Armi */
+  RTL,	/* Phli */
+  RTL,	/* Prti */
+  LTR,	/* Java */
+  LTR,	/* Kthi */
+  LTR,	/* Lisu */
+  LTR,	/* Mtei */
+  RTL,	/* Sarb */
+  RTL,	/* Orkh */
+  RTL,	/* Samr */
+  LTR,	/* Lana */
+  LTR,	/* Tavt */
+
+  /* Unicode-6.0 additions */
+  LTR,	/* Batk */
+  LTR,	/* Brah */
+  RTL 	/* Mand */
+};
+#undef LTR
+#undef RTL
+
+hb_direction_t
+hb_script_get_horizontal_direction (hb_script_t script)
+{
+  if (unlikely ((unsigned int) script >= ARRAY_LENGTH (horiz_dir)))
+    return HB_DIRECTION_LTR;
+
+  return horiz_dir[script];
+}
+
+
 HB_END_DECLS

src/hb-common.h

@@ -108,6 +108,176 @@ const char *
 hb_language_to_string (hb_language_t language);
 
 
+/* hb_unicode_general_category_t */
+
+typedef enum
+{
+  HB_UNICODE_GENERAL_CATEGORY_CONTROL,			/* Cc */
+  HB_UNICODE_GENERAL_CATEGORY_FORMAT,			/* Cf */
+  HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED,		/* Cn */
+  HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE,		/* Co */
+  HB_UNICODE_GENERAL_CATEGORY_SURROGATE,		/* Cs */
+  HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER,		/* Ll */
+  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER,		/* Lm */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER,		/* Lo */
+  HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER,		/* Lt */
+  HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER,		/* Lu */
+  HB_UNICODE_GENERAL_CATEGORY_COMBINING_MARK,		/* Mc */
+  HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK,		/* Me */
+  HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK,		/* Mn */
+  HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER,		/* Nd */
+  HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER,		/* Nl */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER,		/* No */
+  HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION,	/* Pc */
+  HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION,		/* Pd */
+  HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION,	/* Pe */
+  HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION,	/* Pf */
+  HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION,	/* Pi */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION,	/* Po */
+  HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION,		/* Ps */
+  HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL,		/* Sc */
+  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL,		/* Sk */
+  HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL,		/* Sm */
+  HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL,		/* So */
+  HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR,		/* Zl */
+  HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR,	/* Zp */
+  HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR		/* Zs */
+} hb_unicode_general_category_t;
+
+
+/* hb_script_t */
+
+typedef enum
+{                               /* ISO 15924 code */
+  HB_SCRIPT_INVALID = -1,
+
+  HB_SCRIPT_COMMON       = 0,   /* Zyyy */
+  HB_SCRIPT_INHERITED,          /* Qaai */
+  HB_SCRIPT_ARABIC,             /* Arab */
+  HB_SCRIPT_ARMENIAN,           /* Armn */
+  HB_SCRIPT_BENGALI,            /* Beng */
+  HB_SCRIPT_BOPOMOFO,           /* Bopo */
+  HB_SCRIPT_CHEROKEE,           /* Cher */
+  HB_SCRIPT_COPTIC,             /* Qaac */
+  HB_SCRIPT_CYRILLIC,           /* Cyrl (Cyrs) */
+  HB_SCRIPT_DESERET,            /* Dsrt */
+  HB_SCRIPT_DEVANAGARI,         /* Deva */
+  HB_SCRIPT_ETHIOPIC,           /* Ethi */
+  HB_SCRIPT_GEORGIAN,           /* Geor (Geon, Geoa) */
+  HB_SCRIPT_GOTHIC,             /* Goth */
+  HB_SCRIPT_GREEK,              /* Grek */
+  HB_SCRIPT_GUJARATI,           /* Gujr */
+  HB_SCRIPT_GURMUKHI,           /* Guru */
+  HB_SCRIPT_HAN,                /* Hani */
+  HB_SCRIPT_HANGUL,             /* Hang */
+  HB_SCRIPT_HEBREW,             /* Hebr */
+  HB_SCRIPT_HIRAGANA,           /* Hira */
+  HB_SCRIPT_KANNADA,            /* Knda */
+  HB_SCRIPT_KATAKANA,           /* Kana */
+  HB_SCRIPT_KHMER,              /* Khmr */
+  HB_SCRIPT_LAO,                /* Laoo */
+  HB_SCRIPT_LATIN,              /* Latn (Latf, Latg) */
+  HB_SCRIPT_MALAYALAM,          /* Mlym */
+  HB_SCRIPT_MONGOLIAN,          /* Mong */
+  HB_SCRIPT_MYANMAR,            /* Mymr */
+  HB_SCRIPT_OGHAM,              /* Ogam */
+  HB_SCRIPT_OLD_ITALIC,         /* Ital */
+  HB_SCRIPT_ORIYA,              /* Orya */
+  HB_SCRIPT_RUNIC,              /* Runr */
+  HB_SCRIPT_SINHALA,            /* Sinh */
+  HB_SCRIPT_SYRIAC,             /* Syrc (Syrj, Syrn, Syre) */
+  HB_SCRIPT_TAMIL,              /* Taml */
+  HB_SCRIPT_TELUGU,             /* Telu */
+  HB_SCRIPT_THAANA,             /* Thaa */
+  HB_SCRIPT_THAI,               /* Thai */
+  HB_SCRIPT_TIBETAN,            /* Tibt */
+  HB_SCRIPT_CANADIAN_ABORIGINAL, /* Cans */
+  HB_SCRIPT_YI,                 /* Yiii */
+  HB_SCRIPT_TAGALOG,            /* Tglg */
+  HB_SCRIPT_HANUNOO,            /* Hano */
+  HB_SCRIPT_BUHID,              /* Buhd */
+  HB_SCRIPT_TAGBANWA,           /* Tagb */
+
+  /* Unicode-4.0 additions */
+  HB_SCRIPT_BRAILLE,            /* Brai */
+  HB_SCRIPT_CYPRIOT,            /* Cprt */
+  HB_SCRIPT_LIMBU,              /* Limb */
+  HB_SCRIPT_OSMANYA,            /* Osma */
+  HB_SCRIPT_SHAVIAN,            /* Shaw */
+  HB_SCRIPT_LINEAR_B,           /* Linb */
+  HB_SCRIPT_TAI_LE,             /* Tale */
+  HB_SCRIPT_UGARITIC,           /* Ugar */
+
+  /* Unicode-4.1 additions */
+  HB_SCRIPT_NEW_TAI_LUE,        /* Talu */
+  HB_SCRIPT_BUGINESE,           /* Bugi */
+  HB_SCRIPT_GLAGOLITIC,         /* Glag */
+  HB_SCRIPT_TIFINAGH,           /* Tfng */
+  HB_SCRIPT_SYLOTI_NAGRI,       /* Sylo */
+  HB_SCRIPT_OLD_PERSIAN,        /* Xpeo */
+  HB_SCRIPT_KHAROSHTHI,         /* Khar */
+
+  /* Unicode-5.0 additions */
+  HB_SCRIPT_UNKNOWN,            /* Zzzz */
+  HB_SCRIPT_BALINESE,           /* Bali */
+  HB_SCRIPT_CUNEIFORM,          /* Xsux */
+  HB_SCRIPT_PHOENICIAN,         /* Phnx */
+  HB_SCRIPT_PHAGS_PA,           /* Phag */
+  HB_SCRIPT_NKO,                /* Nkoo */
+
+  /* Unicode-5.1 additions */
+  HB_SCRIPT_KAYAH_LI,           /* Kali */
+  HB_SCRIPT_LEPCHA,             /* Lepc */
+  HB_SCRIPT_REJANG,             /* Rjng */
+  HB_SCRIPT_SUNDANESE,          /* Sund */
+  HB_SCRIPT_SAURASHTRA,         /* Saur */
+  HB_SCRIPT_CHAM,               /* Cham */
+  HB_SCRIPT_OL_CHIKI,           /* Olck */
+  HB_SCRIPT_VAI,                /* Vaii */
+  HB_SCRIPT_CARIAN,             /* Cari */
+  HB_SCRIPT_LYCIAN,             /* Lyci */
+  HB_SCRIPT_LYDIAN,             /* Lydi */
+
+  /* Unicode-5.2 additions */
+  HB_SCRIPT_AVESTAN,                /* Avst */
+  HB_SCRIPT_BAMUM,                  /* Bamu */
+  HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,   /* Egyp */
+  HB_SCRIPT_IMPERIAL_ARAMAIC,       /* Armi */
+  HB_SCRIPT_INSCRIPTIONAL_PAHLAVI,  /* Phli */
+  HB_SCRIPT_INSCRIPTIONAL_PARTHIAN, /* Prti */
+  HB_SCRIPT_JAVANESE,               /* Java */
+  HB_SCRIPT_KAITHI,                 /* Kthi */
+  HB_SCRIPT_LISU,                   /* Lisu */
+  HB_SCRIPT_MEETEI_MAYEK,           /* Mtei */
+  HB_SCRIPT_OLD_SOUTH_ARABIAN,      /* Sarb */
+  HB_SCRIPT_OLD_TURKIC,             /* Orkh */
+  HB_SCRIPT_SAMARITAN,              /* Samr */
+  HB_SCRIPT_TAI_THAM,               /* Lana */
+  HB_SCRIPT_TAI_VIET,               /* Tavt */
+
+  /* Unicode-6.0 additions */
+  HB_SCRIPT_BATAK,                  /* Batk */
+  HB_SCRIPT_BRAHMI,                 /* Brah */
+  HB_SCRIPT_MANDAIC                 /* Mand */
+} hb_script_t;
+
+
+/* Script functions */
+
+hb_script_t
+hb_script_from_iso15924_tag (hb_tag_t tag);
+
+/* suger for tag_from_string() then script_from_iso15924_tag */
+hb_script_t
+hb_script_from_string (const char *s);
+
+hb_tag_t
+hb_script_to_iso15924_tag (hb_script_t script);
+
+hb_direction_t
+hb_script_get_horizontal_direction (hb_script_t script);
+
+
 HB_END_DECLS
 
 #endif /* HB_COMMON_H */

src/hb-unicode.c

@@ -234,410 +234,4 @@ hb_unicode_get_eastasian_width (hb_unicode_funcs_t *ufuncs,
 }
 
 
-/* hb_script_t */
-
-static const hb_tag_t script_to_iso15924_tag[] =
-{
-  HB_TAG('Z','y','y','y'),	/* HB_SCRIPT_COMMON */
-  HB_TAG('Q','a','a','i'),	/* HB_SCRIPT_INHERITED */
-  HB_TAG('A','r','a','b'),	/* HB_SCRIPT_ARABIC */
-  HB_TAG('A','r','m','n'),	/* HB_SCRIPT_ARMENIAN */
-  HB_TAG('B','e','n','g'),	/* HB_SCRIPT_BENGALI */
-  HB_TAG('B','o','p','o'),	/* HB_SCRIPT_BOPOMOFO */
-  HB_TAG('C','h','e','r'),	/* HB_SCRIPT_CHEROKEE */
-  HB_TAG('Q','a','a','c'),	/* HB_SCRIPT_COPTIC */
-  HB_TAG('C','y','r','l'),	/* HB_SCRIPT_CYRILLIC */
-  HB_TAG('D','s','r','t'),	/* HB_SCRIPT_DESERET */
-  HB_TAG('D','e','v','a'),	/* HB_SCRIPT_DEVANAGARI */
-  HB_TAG('E','t','h','i'),	/* HB_SCRIPT_ETHIOPIC */
-  HB_TAG('G','e','o','r'),	/* HB_SCRIPT_GEORGIAN */
-  HB_TAG('G','o','t','h'),	/* HB_SCRIPT_GOTHIC */
-  HB_TAG('G','r','e','k'),	/* HB_SCRIPT_GREEK */
-  HB_TAG('G','u','j','r'),	/* HB_SCRIPT_GUJARATI */
-  HB_TAG('G','u','r','u'),	/* HB_SCRIPT_GURMUKHI */
-  HB_TAG('H','a','n','i'),	/* HB_SCRIPT_HAN */
-  HB_TAG('H','a','n','g'),	/* HB_SCRIPT_HANGUL */
-  HB_TAG('H','e','b','r'),	/* HB_SCRIPT_HEBREW */
-  HB_TAG('H','i','r','a'),	/* HB_SCRIPT_HIRAGANA */
-  HB_TAG('K','n','d','a'),	/* HB_SCRIPT_KANNADA */
-  HB_TAG('K','a','n','a'),	/* HB_SCRIPT_KATAKANA */
-  HB_TAG('K','h','m','r'),	/* HB_SCRIPT_KHMER */
-  HB_TAG('L','a','o','o'),	/* HB_SCRIPT_LAO */
-  HB_TAG('L','a','t','n'),	/* HB_SCRIPT_LATIN */
-  HB_TAG('M','l','y','m'),	/* HB_SCRIPT_MALAYALAM */
-  HB_TAG('M','o','n','g'),	/* HB_SCRIPT_MONGOLIAN */
-  HB_TAG('M','y','m','r'),	/* HB_SCRIPT_MYANMAR */
-  HB_TAG('O','g','a','m'),	/* HB_SCRIPT_OGHAM */
-  HB_TAG('I','t','a','l'),	/* HB_SCRIPT_OLD_ITALIC */
-  HB_TAG('O','r','y','a'),	/* HB_SCRIPT_ORIYA */
-  HB_TAG('R','u','n','r'),	/* HB_SCRIPT_RUNIC */
-  HB_TAG('S','i','n','h'),	/* HB_SCRIPT_SINHALA */
-  HB_TAG('S','y','r','c'),	/* HB_SCRIPT_SYRIAC */
-  HB_TAG('T','a','m','l'),	/* HB_SCRIPT_TAMIL */
-  HB_TAG('T','e','l','u'),	/* HB_SCRIPT_TELUGU */
-  HB_TAG('T','h','a','a'),	/* HB_SCRIPT_THAANA */
-  HB_TAG('T','h','a','i'),	/* HB_SCRIPT_THAI */
-  HB_TAG('T','i','b','t'),	/* HB_SCRIPT_TIBETAN */
-  HB_TAG('C','a','n','s'),	/* HB_SCRIPT_CANADIAN_ABORIGINAL */
-  HB_TAG('Y','i','i','i'),	/* HB_SCRIPT_YI */
-  HB_TAG('T','g','l','g'),	/* HB_SCRIPT_TAGALOG */
-  HB_TAG('H','a','n','o'),	/* HB_SCRIPT_HANUNOO */
-  HB_TAG('B','u','h','d'),	/* HB_SCRIPT_BUHID */
-  HB_TAG('T','a','g','b'),	/* HB_SCRIPT_TAGBANWA */
-
-  /* Unicode-4.0 additions */
-  HB_TAG('B','r','a','i'),	/* HB_SCRIPT_BRAILLE */
-  HB_TAG('C','p','r','t'),	/* HB_SCRIPT_CYPRIOT */
-  HB_TAG('L','i','m','b'),	/* HB_SCRIPT_LIMBU */
-  HB_TAG('O','s','m','a'),	/* HB_SCRIPT_OSMANYA */
-  HB_TAG('S','h','a','w'),	/* HB_SCRIPT_SHAVIAN */
-  HB_TAG('L','i','n','b'),	/* HB_SCRIPT_LINEAR_B */
-  HB_TAG('T','a','l','e'),	/* HB_SCRIPT_TAI_LE */
-  HB_TAG('U','g','a','r'),	/* HB_SCRIPT_UGARITIC */
-
-  /* Unicode-4.1 additions */
-  HB_TAG('T','a','l','u'),	/* HB_SCRIPT_NEW_TAI_LUE */
-  HB_TAG('B','u','g','i'),	/* HB_SCRIPT_BUGINESE */
-  HB_TAG('G','l','a','g'),	/* HB_SCRIPT_GLAGOLITIC */
-  HB_TAG('T','f','n','g'),	/* HB_SCRIPT_TIFINAGH */
-  HB_TAG('S','y','l','o'),	/* HB_SCRIPT_SYLOTI_NAGRI */
-  HB_TAG('X','p','e','o'),	/* HB_SCRIPT_OLD_PERSIAN */
-  HB_TAG('K','h','a','r'),	/* HB_SCRIPT_KHAROSHTHI */
-
-  /* Unicode-5.0 additions */
-  HB_TAG('Z','z','z','z'),	/* HB_SCRIPT_UNKNOWN */
-  HB_TAG('B','a','l','i'),	/* HB_SCRIPT_BALINESE */
-  HB_TAG('X','s','u','x'),	/* HB_SCRIPT_CUNEIFORM */
-  HB_TAG('P','h','n','x'),	/* HB_SCRIPT_PHOENICIAN */
-  HB_TAG('P','h','a','g'),	/* HB_SCRIPT_PHAGS_PA */
-  HB_TAG('N','k','o','o'),	/* HB_SCRIPT_NKO */
-
-  /* Unicode-5.1 additions */
-  HB_TAG('K','a','l','i'),	/* HB_SCRIPT_KAYAH_LI */
-  HB_TAG('L','e','p','c'),	/* HB_SCRIPT_LEPCHA */
-  HB_TAG('R','j','n','g'),	/* HB_SCRIPT_REJANG */
-  HB_TAG('S','u','n','d'),	/* HB_SCRIPT_SUNDANESE */
-  HB_TAG('S','a','u','r'),	/* HB_SCRIPT_SAURASHTRA */
-  HB_TAG('C','h','a','m'),	/* HB_SCRIPT_CHAM */
-  HB_TAG('O','l','c','k'),	/* HB_SCRIPT_OL_CHIKI */
-  HB_TAG('V','a','i','i'),	/* HB_SCRIPT_VAI */
-  HB_TAG('C','a','r','i'),	/* HB_SCRIPT_CARIAN */
-  HB_TAG('L','y','c','i'),	/* HB_SCRIPT_LYCIAN */
-  HB_TAG('L','y','d','i'),	/* HB_SCRIPT_LYDIAN */
-
-  /* Unicode-5.2 additions */
-  HB_TAG('A','v','s','t'),	/* HB_SCRIPT_AVESTAN */
-  HB_TAG('B','a','m','u'),	/* HB_SCRIPT_BAMUM */
-  HB_TAG('E','g','y','p'),	/* HB_SCRIPT_EGYPTIAN_HIEROGLYPHS */
-  HB_TAG('A','r','m','i'),	/* HB_SCRIPT_IMPERIAL_ARAMAIC */
-  HB_TAG('P','h','l','i'),	/* HB_SCRIPT_INSCRIPTIONAL_PAHLAVI */
-  HB_TAG('P','r','t','i'),	/* HB_SCRIPT_INSCRIPTIONAL_PARTHIAN */
-  HB_TAG('J','a','v','a'),	/* HB_SCRIPT_JAVANESE */
-  HB_TAG('K','t','h','i'),	/* HB_SCRIPT_KAITHI */
-  HB_TAG('L','i','s','u'),	/* HB_SCRIPT_LISU */
-  HB_TAG('M','t','e','i'),	/* HB_SCRIPT_MEETEI_MAYEK */
-  HB_TAG('S','a','r','b'),	/* HB_SCRIPT_OLD_SOUTH_ARABIAN */
-  HB_TAG('O','r','k','h'),	/* HB_SCRIPT_OLD_TURKIC */
-  HB_TAG('S','a','m','r'),	/* HB_SCRIPT_SAMARITAN */
-  HB_TAG('L','a','n','a'),	/* HB_SCRIPT_TAI_THAM */
-  HB_TAG('T','a','v','t'),	/* HB_SCRIPT_TAI_VIET */
-
-  /* Unicode-6.0 additions */
-  HB_TAG('B','a','t','k'),	/* HB_SCRIPT_BATAK */
-  HB_TAG('B','r','a','h'),	/* HB_SCRIPT_BRAHMI */
-  HB_TAG('M','a','n','d') 	/* HB_SCRIPT_MANDAIC */
-};
-
-struct tag_script_pair {
-  hb_tag_t tag;
-  hb_script_t script;
-};
-static const struct tag_script_pair script_from_iso15924_tag[] =
-{
-  {HB_TAG('A','r','a','b'), HB_SCRIPT_ARABIC},
-  {HB_TAG('A','r','m','i'), HB_SCRIPT_IMPERIAL_ARAMAIC},
-  {HB_TAG('A','r','m','n'), HB_SCRIPT_ARMENIAN},
-  {HB_TAG('A','v','s','t'), HB_SCRIPT_AVESTAN},
-  {HB_TAG('B','a','l','i'), HB_SCRIPT_BALINESE},
-  {HB_TAG('B','a','m','u'), HB_SCRIPT_BAMUM},
-  {HB_TAG('B','a','t','k'), HB_SCRIPT_BATAK},
-  {HB_TAG('B','e','n','g'), HB_SCRIPT_BENGALI},
-  {HB_TAG('B','o','p','o'), HB_SCRIPT_BOPOMOFO},
-  {HB_TAG('B','r','a','h'), HB_SCRIPT_BRAHMI},
-  {HB_TAG('B','r','a','i'), HB_SCRIPT_BRAILLE},
-  {HB_TAG('B','u','g','i'), HB_SCRIPT_BUGINESE},
-  {HB_TAG('B','u','h','d'), HB_SCRIPT_BUHID},
-  {HB_TAG('C','a','n','s'), HB_SCRIPT_CANADIAN_ABORIGINAL},
-  {HB_TAG('C','a','r','i'), HB_SCRIPT_CARIAN},
-  {HB_TAG('C','h','a','m'), HB_SCRIPT_CHAM},
-  {HB_TAG('C','h','e','r'), HB_SCRIPT_CHEROKEE},
-  {HB_TAG('C','p','r','t'), HB_SCRIPT_CYPRIOT},
-  {HB_TAG('C','y','r','l'), HB_SCRIPT_CYRILLIC},
-  {HB_TAG('C','y','r','s'), HB_SCRIPT_CYRILLIC},
-  {HB_TAG('D','e','v','a'), HB_SCRIPT_DEVANAGARI},
-  {HB_TAG('D','s','r','t'), HB_SCRIPT_DESERET},
-  {HB_TAG('E','g','y','p'), HB_SCRIPT_EGYPTIAN_HIEROGLYPHS},
-  {HB_TAG('E','t','h','i'), HB_SCRIPT_ETHIOPIC},
-  {HB_TAG('G','e','o','a'), HB_SCRIPT_GEORGIAN},
-  {HB_TAG('G','e','o','n'), HB_SCRIPT_GEORGIAN},
-  {HB_TAG('G','e','o','r'), HB_SCRIPT_GEORGIAN},
-  {HB_TAG('G','l','a','g'), HB_SCRIPT_GLAGOLITIC},
-  {HB_TAG('G','o','t','h'), HB_SCRIPT_GOTHIC},
-  {HB_TAG('G','r','e','k'), HB_SCRIPT_GREEK},
-  {HB_TAG('G','u','j','r'), HB_SCRIPT_GUJARATI},
-  {HB_TAG('G','u','r','u'), HB_SCRIPT_GURMUKHI},
-  {HB_TAG('H','a','n','g'), HB_SCRIPT_HANGUL},
-  {HB_TAG('H','a','n','i'), HB_SCRIPT_HAN},
-  {HB_TAG('H','a','n','o'), HB_SCRIPT_HANUNOO},
-  {HB_TAG('H','e','b','r'), HB_SCRIPT_HEBREW},
-  {HB_TAG('H','i','r','a'), HB_SCRIPT_HIRAGANA},
-  {HB_TAG('I','t','a','l'), HB_SCRIPT_OLD_ITALIC},
-  {HB_TAG('J','a','v','a'), HB_SCRIPT_JAVANESE},
-  {HB_TAG('K','a','l','i'), HB_SCRIPT_KAYAH_LI},
-  {HB_TAG('K','a','n','a'), HB_SCRIPT_KATAKANA},
-  {HB_TAG('K','h','a','r'), HB_SCRIPT_KHAROSHTHI},
-  {HB_TAG('K','h','m','r'), HB_SCRIPT_KHMER},
-  {HB_TAG('K','n','d','a'), HB_SCRIPT_KANNADA},
-  {HB_TAG('K','t','h','i'), HB_SCRIPT_KAITHI},
-  {HB_TAG('L','a','n','a'), HB_SCRIPT_TAI_THAM},
-  {HB_TAG('L','a','o','o'), HB_SCRIPT_LAO},
-  {HB_TAG('L','a','t','f'), HB_SCRIPT_LATIN},
-  {HB_TAG('L','a','t','g'), HB_SCRIPT_LATIN},
-  {HB_TAG('L','a','t','n'), HB_SCRIPT_LATIN},
-  {HB_TAG('L','e','p','c'), HB_SCRIPT_LEPCHA},
-  {HB_TAG('L','i','m','b'), HB_SCRIPT_LIMBU},
-  {HB_TAG('L','i','n','b'), HB_SCRIPT_LINEAR_B},
-  {HB_TAG('L','i','s','u'), HB_SCRIPT_LISU},
-  {HB_TAG('L','y','c','i'), HB_SCRIPT_LYCIAN},
-  {HB_TAG('L','y','d','i'), HB_SCRIPT_LYDIAN},
-  {HB_TAG('M','a','n','d'), HB_SCRIPT_MANDAIC},
-  {HB_TAG('M','l','y','m'), HB_SCRIPT_MALAYALAM},
-  {HB_TAG('M','o','n','g'), HB_SCRIPT_MONGOLIAN},
-  {HB_TAG('M','t','e','i'), HB_SCRIPT_MEETEI_MAYEK},
-  {HB_TAG('M','y','m','r'), HB_SCRIPT_MYANMAR},
-  {HB_TAG('N','k','o','o'), HB_SCRIPT_NKO},
-  {HB_TAG('O','g','a','m'), HB_SCRIPT_OGHAM},
-  {HB_TAG('O','l','c','k'), HB_SCRIPT_OL_CHIKI},
-  {HB_TAG('O','r','k','h'), HB_SCRIPT_OLD_TURKIC},
-  {HB_TAG('O','r','y','a'), HB_SCRIPT_ORIYA},
-  {HB_TAG('O','s','m','a'), HB_SCRIPT_OSMANYA},
-  {HB_TAG('P','h','a','g'), HB_SCRIPT_PHAGS_PA},
-  {HB_TAG('P','h','l','i'), HB_SCRIPT_INSCRIPTIONAL_PAHLAVI},
-  {HB_TAG('P','h','n','x'), HB_SCRIPT_PHOENICIAN},
-  {HB_TAG('P','r','t','i'), HB_SCRIPT_INSCRIPTIONAL_PARTHIAN},
-  {HB_TAG('Q','a','a','c'), HB_SCRIPT_COPTIC},
-  {HB_TAG('Q','a','a','i'), HB_SCRIPT_INHERITED},
-  {HB_TAG('R','j','n','g'), HB_SCRIPT_REJANG},
-  {HB_TAG('R','u','n','r'), HB_SCRIPT_RUNIC},
-  {HB_TAG('S','a','m','r'), HB_SCRIPT_SAMARITAN},
-  {HB_TAG('S','a','r','b'), HB_SCRIPT_OLD_SOUTH_ARABIAN},
-  {HB_TAG('S','a','u','r'), HB_SCRIPT_SAURASHTRA},
-  {HB_TAG('S','h','a','w'), HB_SCRIPT_SHAVIAN},
-  {HB_TAG('S','i','n','h'), HB_SCRIPT_SINHALA},
-  {HB_TAG('S','u','n','d'), HB_SCRIPT_SUNDANESE},
-  {HB_TAG('S','y','l','o'), HB_SCRIPT_SYLOTI_NAGRI},
-  {HB_TAG('S','y','r','c'), HB_SCRIPT_SYRIAC},
-  {HB_TAG('S','y','r','e'), HB_SCRIPT_SYRIAC},
-  {HB_TAG('S','y','r','n'), HB_SCRIPT_SYRIAC},
-  {HB_TAG('T','a','g','b'), HB_SCRIPT_TAGBANWA},
-  {HB_TAG('T','a','l','e'), HB_SCRIPT_TAI_LE},
-  {HB_TAG('T','a','l','u'), HB_SCRIPT_NEW_TAI_LUE},
-  {HB_TAG('T','a','m','l'), HB_SCRIPT_TAMIL},
-  {HB_TAG('T','a','v','t'), HB_SCRIPT_TAI_VIET},
-  {HB_TAG('T','e','l','u'), HB_SCRIPT_TELUGU},
-  {HB_TAG('T','f','n','g'), HB_SCRIPT_TIFINAGH},
-  {HB_TAG('T','g','l','g'), HB_SCRIPT_TAGALOG},
-  {HB_TAG('T','h','a','a'), HB_SCRIPT_THAANA},
-  {HB_TAG('T','h','a','i'), HB_SCRIPT_THAI},
-  {HB_TAG('T','i','b','t'), HB_SCRIPT_TIBETAN},
-  {HB_TAG('U','g','a','r'), HB_SCRIPT_UGARITIC},
-  {HB_TAG('V','a','i','i'), HB_SCRIPT_VAI},
-  {HB_TAG('X','p','e','o'), HB_SCRIPT_OLD_PERSIAN},
-  {HB_TAG('X','s','u','x'), HB_SCRIPT_CUNEIFORM},
-  {HB_TAG('Y','i','i','i'), HB_SCRIPT_YI},
-  {HB_TAG('Z','y','y','y'), HB_SCRIPT_COMMON},
-  {HB_TAG('Z','z','z','z'), HB_SCRIPT_UNKNOWN}
-};
-
-static int
-_tag_cmp (hb_tag_t *pa, hb_tag_t *pb)
-{
-  hb_tag_t a = *pa, b = *pb;
-  return a < b ? -1 : a == b ? 0 : +1;
-}
-
-
-hb_script_t
-hb_script_from_iso15924_tag (hb_tag_t tag)
-{
-  const struct tag_script_pair *pair;
-
-  /* Be lenient, adjust case (one capital letter followed by three small letters) */
-  tag = (tag & 0xDFDFDFDF) | 0x00202020;
-
-  pair = (const struct tag_script_pair *) bsearch (&tag,
-						   script_from_iso15924_tag,
-						   ARRAY_LENGTH (script_from_iso15924_tag),
-						   sizeof (script_from_iso15924_tag[0]),
-						   (hb_compare_func_t) _tag_cmp);
-
-  if (pair)
-    return pair->script;
-
-  /* If it looks right, just use the tag as a script */
-  if (((uint32_t) tag & 0xE0E0E0E0) == 0x40606060)
-    return (hb_script_t) tag;
-
-  /* Otherwise, return unknown */
-  return HB_SCRIPT_UNKNOWN;
-}
-
-hb_script_t
-hb_script_from_string (const char *s)
-{
-  return hb_script_from_iso15924_tag (hb_tag_from_string (s));
-}
-
-hb_tag_t
-hb_script_to_iso15924_tag (hb_script_t script)
-{
-  if (likely ((unsigned int) script < ARRAY_LENGTH (script_to_iso15924_tag)))
-    return script_to_iso15924_tag[script];
-
-  /* if script is of the right shape (one capital letter followed by three small letters),
-   * return as is. */
-  if (((uint32_t) script & 0xE0E0E0E0) == 0x40606060)
-    return (hb_tag_t) script;
-
-  /* Otherwise, we don't know what that is */
-  return script_to_iso15924_tag[HB_SCRIPT_UNKNOWN];
-}
-
-
-#define LTR HB_DIRECTION_LTR
-#define RTL HB_DIRECTION_RTL
-const hb_direction_t horiz_dir[] =
-{
-  LTR,	/* Zyyy */
-  LTR,	/* Qaai */
-  RTL,	/* Arab */
-  LTR,	/* Armn */
-  LTR,	/* Beng */
-  LTR,	/* Bopo */
-  LTR,	/* Cher */
-  LTR,	/* Qaac */
-  LTR,	/* Cyrl (Cyrs) */
-  LTR,	/* Dsrt */
-  LTR,	/* Deva */
-  LTR,	/* Ethi */
-  LTR,	/* Geor (Geon, Geoa) */
-  LTR,	/* Goth */
-  LTR,	/* Grek */
-  LTR,	/* Gujr */
-  LTR,	/* Guru */
-  LTR,	/* Hani */
-  LTR,	/* Hang */
-  RTL,	/* Hebr */
-  LTR,	/* Hira */
-  LTR,	/* Knda */
-  LTR,	/* Kana */
-  LTR,	/* Khmr */
-  LTR,	/* Laoo */
-  LTR,	/* Latn (Latf, Latg) */
-  LTR,	/* Mlym */
-  LTR,	/* Mong */
-  LTR,	/* Mymr */
-  LTR,	/* Ogam */
-  LTR,	/* Ital */
-  LTR,	/* Orya */
-  LTR,	/* Runr */
-  LTR,	/* Sinh */
-  RTL,	/* Syrc (Syrj, Syrn, Syre) */
-  LTR,	/* Taml */
-  LTR,	/* Telu */
-  RTL,	/* Thaa */
-  LTR,	/* Thai */
-  LTR,	/* Tibt */
-  LTR,	/* Cans */
-  LTR,	/* Yiii */
-  LTR,	/* Tglg */
-  LTR,	/* Hano */
-  LTR,	/* Buhd */
-  LTR,	/* Tagb */
-
-  /* Unicode-4.0 additions */
-  LTR,	/* Brai */
-  RTL,	/* Cprt */
-  LTR,	/* Limb */
-  LTR,	/* Osma */
-  LTR,	/* Shaw */
-  LTR,	/* Linb */
-  LTR,	/* Tale */
-  LTR,	/* Ugar */
-
-  /* Unicode-4.1 additions */
-  LTR,	/* Talu */
-  LTR,	/* Bugi */
-  LTR,	/* Glag */
-  LTR,	/* Tfng */
-  LTR,	/* Sylo */
-  LTR,	/* Xpeo */
-  LTR,	/* Khar */
-
-  /* Unicode-5.0 additions */
-  LTR,	/* Zzzz */
-  LTR,	/* Bali */
-  LTR,	/* Xsux */
-  RTL,	/* Phnx */
-  LTR,	/* Phag */
-  RTL,	/* Nkoo */
-
-  /* Unicode-5.1 additions */
-  LTR,	/* Kali */
-  LTR,	/* Lepc */
-  LTR,	/* Rjng */
-  LTR,	/* Sund */
-  LTR,	/* Saur */
-  LTR,	/* Cham */
-  LTR,	/* Olck */
-  LTR,	/* Vaii */
-  LTR,	/* Cari */
-  LTR,	/* Lyci */
-  LTR,	/* Lydi */
-
-  /* Unicode-5.2 additions */
-  RTL,	/* Avst */
-  LTR,	/* Bamu */
-  LTR,	/* Egyp */
-  RTL,	/* Armi */
-  RTL,	/* Phli */
-  RTL,	/* Prti */
-  LTR,	/* Java */
-  LTR,	/* Kthi */
-  LTR,	/* Lisu */
-  LTR,	/* Mtei */
-  RTL,	/* Sarb */
-  RTL,	/* Orkh */
-  RTL,	/* Samr */
-  LTR,	/* Lana */
-  LTR,	/* Tavt */
-
-  /* Unicode-6.0 additions */
-  LTR,	/* Batk */
-  LTR,	/* Brah */
-  RTL 	/* Mand */
-};
-#undef LTR
-#undef RTL
-
-hb_direction_t
-hb_script_get_horizontal_direction (hb_script_t script)
-{
-  if (unlikely ((unsigned int) script >= ARRAY_LENGTH (horiz_dir)))
-    return HB_DIRECTION_LTR;
-
-  return horiz_dir[script];
-}
-
-
 HB_END_DECLS

src/hb-unicode.h

@@ -32,157 +32,6 @@
 HB_BEGIN_DECLS
 
 
-/* Unicode General Category property */
-typedef enum
-{
-  HB_UNICODE_GENERAL_CATEGORY_CONTROL,			/* Cc */
-  HB_UNICODE_GENERAL_CATEGORY_FORMAT,			/* Cf */
-  HB_UNICODE_GENERAL_CATEGORY_UNASSIGNED,		/* Cn */
-  HB_UNICODE_GENERAL_CATEGORY_PRIVATE_USE,		/* Co */
-  HB_UNICODE_GENERAL_CATEGORY_SURROGATE,		/* Cs */
-  HB_UNICODE_GENERAL_CATEGORY_LOWERCASE_LETTER,		/* Ll */
-  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_LETTER,		/* Lm */
-  HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER,		/* Lo */
-  HB_UNICODE_GENERAL_CATEGORY_TITLECASE_LETTER,		/* Lt */
-  HB_UNICODE_GENERAL_CATEGORY_UPPERCASE_LETTER,		/* Lu */
-  HB_UNICODE_GENERAL_CATEGORY_COMBINING_MARK,		/* Mc */
-  HB_UNICODE_GENERAL_CATEGORY_ENCLOSING_MARK,		/* Me */
-  HB_UNICODE_GENERAL_CATEGORY_NON_SPACING_MARK,		/* Mn */
-  HB_UNICODE_GENERAL_CATEGORY_DECIMAL_NUMBER,		/* Nd */
-  HB_UNICODE_GENERAL_CATEGORY_LETTER_NUMBER,		/* Nl */
-  HB_UNICODE_GENERAL_CATEGORY_OTHER_NUMBER,		/* No */
-  HB_UNICODE_GENERAL_CATEGORY_CONNECT_PUNCTUATION,	/* Pc */
-  HB_UNICODE_GENERAL_CATEGORY_DASH_PUNCTUATION,		/* Pd */
-  HB_UNICODE_GENERAL_CATEGORY_CLOSE_PUNCTUATION,	/* Pe */
-  HB_UNICODE_GENERAL_CATEGORY_FINAL_PUNCTUATION,	/* Pf */
-  HB_UNICODE_GENERAL_CATEGORY_INITIAL_PUNCTUATION,	/* Pi */
-  HB_UNICODE_GENERAL_CATEGORY_OTHER_PUNCTUATION,	/* Po */
-  HB_UNICODE_GENERAL_CATEGORY_OPEN_PUNCTUATION,		/* Ps */
-  HB_UNICODE_GENERAL_CATEGORY_CURRENCY_SYMBOL,		/* Sc */
-  HB_UNICODE_GENERAL_CATEGORY_MODIFIER_SYMBOL,		/* Sk */
-  HB_UNICODE_GENERAL_CATEGORY_MATH_SYMBOL,		/* Sm */
-  HB_UNICODE_GENERAL_CATEGORY_OTHER_SYMBOL,		/* So */
-  HB_UNICODE_GENERAL_CATEGORY_LINE_SEPARATOR,		/* Zl */
-  HB_UNICODE_GENERAL_CATEGORY_PARAGRAPH_SEPARATOR,	/* Zp */
-  HB_UNICODE_GENERAL_CATEGORY_SPACE_SEPARATOR		/* Zs */
-} hb_unicode_general_category_t;
-
-/* Unicode Script property */
-typedef enum
-{                               /* ISO 15924 code */
-  HB_SCRIPT_INVALID = -1,
-
-  HB_SCRIPT_COMMON       = 0,   /* Zyyy */
-  HB_SCRIPT_INHERITED,          /* Qaai */
-  HB_SCRIPT_ARABIC,             /* Arab */
-  HB_SCRIPT_ARMENIAN,           /* Armn */
-  HB_SCRIPT_BENGALI,            /* Beng */
-  HB_SCRIPT_BOPOMOFO,           /* Bopo */
-  HB_SCRIPT_CHEROKEE,           /* Cher */
-  HB_SCRIPT_COPTIC,             /* Qaac */
-  HB_SCRIPT_CYRILLIC,           /* Cyrl (Cyrs) */
-  HB_SCRIPT_DESERET,            /* Dsrt */
-  HB_SCRIPT_DEVANAGARI,         /* Deva */
-  HB_SCRIPT_ETHIOPIC,           /* Ethi */
-  HB_SCRIPT_GEORGIAN,           /* Geor (Geon, Geoa) */
-  HB_SCRIPT_GOTHIC,             /* Goth */
-  HB_SCRIPT_GREEK,              /* Grek */
-  HB_SCRIPT_GUJARATI,           /* Gujr */
-  HB_SCRIPT_GURMUKHI,           /* Guru */
-  HB_SCRIPT_HAN,                /* Hani */
-  HB_SCRIPT_HANGUL,             /* Hang */
-  HB_SCRIPT_HEBREW,             /* Hebr */
-  HB_SCRIPT_HIRAGANA,           /* Hira */
-  HB_SCRIPT_KANNADA,            /* Knda */
-  HB_SCRIPT_KATAKANA,           /* Kana */
-  HB_SCRIPT_KHMER,              /* Khmr */
-  HB_SCRIPT_LAO,                /* Laoo */
-  HB_SCRIPT_LATIN,              /* Latn (Latf, Latg) */
-  HB_SCRIPT_MALAYALAM,          /* Mlym */
-  HB_SCRIPT_MONGOLIAN,          /* Mong */
-  HB_SCRIPT_MYANMAR,            /* Mymr */
-  HB_SCRIPT_OGHAM,              /* Ogam */
-  HB_SCRIPT_OLD_ITALIC,         /* Ital */
-  HB_SCRIPT_ORIYA,              /* Orya */
-  HB_SCRIPT_RUNIC,              /* Runr */
-  HB_SCRIPT_SINHALA,            /* Sinh */
-  HB_SCRIPT_SYRIAC,             /* Syrc (Syrj, Syrn, Syre) */
-  HB_SCRIPT_TAMIL,              /* Taml */
-  HB_SCRIPT_TELUGU,             /* Telu */
-  HB_SCRIPT_THAANA,             /* Thaa */
-  HB_SCRIPT_THAI,               /* Thai */
-  HB_SCRIPT_TIBETAN,            /* Tibt */
-  HB_SCRIPT_CANADIAN_ABORIGINAL, /* Cans */
-  HB_SCRIPT_YI,                 /* Yiii */
-  HB_SCRIPT_TAGALOG,            /* Tglg */
-  HB_SCRIPT_HANUNOO,            /* Hano */
-  HB_SCRIPT_BUHID,              /* Buhd */
-  HB_SCRIPT_TAGBANWA,           /* Tagb */
-
-  /* Unicode-4.0 additions */
-  HB_SCRIPT_BRAILLE,            /* Brai */
-  HB_SCRIPT_CYPRIOT,            /* Cprt */
-  HB_SCRIPT_LIMBU,              /* Limb */
-  HB_SCRIPT_OSMANYA,            /* Osma */
-  HB_SCRIPT_SHAVIAN,            /* Shaw */
-  HB_SCRIPT_LINEAR_B,           /* Linb */
-  HB_SCRIPT_TAI_LE,             /* Tale */
-  HB_SCRIPT_UGARITIC,           /* Ugar */
-
-  /* Unicode-4.1 additions */
-  HB_SCRIPT_NEW_TAI_LUE,        /* Talu */
-  HB_SCRIPT_BUGINESE,           /* Bugi */
-  HB_SCRIPT_GLAGOLITIC,         /* Glag */
-  HB_SCRIPT_TIFINAGH,           /* Tfng */
-  HB_SCRIPT_SYLOTI_NAGRI,       /* Sylo */
-  HB_SCRIPT_OLD_PERSIAN,        /* Xpeo */
-  HB_SCRIPT_KHAROSHTHI,         /* Khar */
-
-  /* Unicode-5.0 additions */
-  HB_SCRIPT_UNKNOWN,            /* Zzzz */
-  HB_SCRIPT_BALINESE,           /* Bali */
-  HB_SCRIPT_CUNEIFORM,          /* Xsux */
-  HB_SCRIPT_PHOENICIAN,         /* Phnx */
-  HB_SCRIPT_PHAGS_PA,           /* Phag */
-  HB_SCRIPT_NKO,                /* Nkoo */
-
-  /* Unicode-5.1 additions */
-  HB_SCRIPT_KAYAH_LI,           /* Kali */
-  HB_SCRIPT_LEPCHA,             /* Lepc */
-  HB_SCRIPT_REJANG,             /* Rjng */
-  HB_SCRIPT_SUNDANESE,          /* Sund */
-  HB_SCRIPT_SAURASHTRA,         /* Saur */
-  HB_SCRIPT_CHAM,               /* Cham */
-  HB_SCRIPT_OL_CHIKI,           /* Olck */
-  HB_SCRIPT_VAI,                /* Vaii */
-  HB_SCRIPT_CARIAN,             /* Cari */
-  HB_SCRIPT_LYCIAN,             /* Lyci */
-  HB_SCRIPT_LYDIAN,             /* Lydi */
-
-  /* Unicode-5.2 additions */
-  HB_SCRIPT_AVESTAN,                /* Avst */
-  HB_SCRIPT_BAMUM,                  /* Bamu */
-  HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,   /* Egyp */
-  HB_SCRIPT_IMPERIAL_ARAMAIC,       /* Armi */
-  HB_SCRIPT_INSCRIPTIONAL_PAHLAVI,  /* Phli */
-  HB_SCRIPT_INSCRIPTIONAL_PARTHIAN, /* Prti */
-  HB_SCRIPT_JAVANESE,               /* Java */
-  HB_SCRIPT_KAITHI,                 /* Kthi */
-  HB_SCRIPT_LISU,                   /* Lisu */
-  HB_SCRIPT_MEETEI_MAYEK,           /* Mtei */
-  HB_SCRIPT_OLD_SOUTH_ARABIAN,      /* Sarb */
-  HB_SCRIPT_OLD_TURKIC,             /* Orkh */
-  HB_SCRIPT_SAMARITAN,              /* Samr */
-  HB_SCRIPT_TAI_THAM,               /* Lana */
-  HB_SCRIPT_TAI_VIET,               /* Tavt */
-
-  /* Unicode-6.0 additions */
-  HB_SCRIPT_BATAK,                  /* Batk */
-  HB_SCRIPT_BRAHMI,                 /* Brah */
-  HB_SCRIPT_MANDAIC                 /* Mand */
-} hb_script_t;
-
-
 /*
  * hb_unicode_funcs_t
  */
@@ -290,21 +139,6 @@ hb_unicode_get_eastasian_width (hb_unicode_funcs_t *ufuncs,
 				hb_codepoint_t unicode);
 
 
-/* Script functions */
-
-hb_script_t
-hb_script_from_iso15924_tag (hb_tag_t tag);
-
-/* suger for tag_from_string() then script_from_iso15924_tag */
-hb_script_t
-hb_script_from_string (const char *s);
-
-hb_tag_t
-hb_script_to_iso15924_tag (hb_script_t script);
-
-hb_direction_t
-hb_script_get_horizontal_direction (hb_script_t script);
-
 HB_END_DECLS
 
 #endif /* HB_UNICODE_H */