Evas language: Added a character script querying function.

This is added until there will be a library that does it; From my
understanding, there are already people working on such library.

SVN revision: 59100

legacy/evas/src/lib/engines/common/Makefile.am

@@ -104,6 +104,7 @@ evas_scale_span.h \
 evas_pipe.h \
 language/evas_bidi_utils.h \
 language/evas_language_utils.h \
+language/evas_script_table.h \
 evas_text_utils.h \
 evas_font_ot.h \
 evas_map_image_internal.c \

legacy/evas/src/lib/engines/common/evas_font_ot.c

@@ -10,6 +10,121 @@
 #include <Eina.h>
 #include "evas_font_private.h"
 
+#ifdef USE_HARFBUZZ
+static const hb_script_t
+_evas_script_to_harfbuzz[] =
+{
+  HB_SCRIPT_COMMON,
+  HB_SCRIPT_INHERITED,
+  HB_SCRIPT_ARABIC,
+  HB_SCRIPT_ARMENIAN,
+  HB_SCRIPT_BENGALI,
+  HB_SCRIPT_BOPOMOFO,
+  HB_SCRIPT_CHEROKEE,
+  HB_SCRIPT_COPTIC,
+  HB_SCRIPT_CYRILLIC,
+  HB_SCRIPT_DESERET,
+  HB_SCRIPT_DEVANAGARI,
+  HB_SCRIPT_ETHIOPIC,
+  HB_SCRIPT_GEORGIAN,
+  HB_SCRIPT_GOTHIC,
+  HB_SCRIPT_GREEK,
+  HB_SCRIPT_GUJARATI,
+  HB_SCRIPT_GURMUKHI,
+  HB_SCRIPT_HAN,
+  HB_SCRIPT_HANGUL,
+  HB_SCRIPT_HEBREW,
+  HB_SCRIPT_HIRAGANA,
+  HB_SCRIPT_KANNADA,
+  HB_SCRIPT_KATAKANA,
+  HB_SCRIPT_KHMER,
+  HB_SCRIPT_LAO,
+  HB_SCRIPT_LATIN,
+  HB_SCRIPT_MALAYALAM,
+  HB_SCRIPT_MONGOLIAN,
+  HB_SCRIPT_MYANMAR,
+  HB_SCRIPT_OGHAM,
+  HB_SCRIPT_OLD_ITALIC,
+  HB_SCRIPT_ORIYA,
+  HB_SCRIPT_RUNIC,
+  HB_SCRIPT_SINHALA,
+  HB_SCRIPT_SYRIAC,
+  HB_SCRIPT_TAMIL,
+  HB_SCRIPT_TELUGU,
+  HB_SCRIPT_THAANA,
+  HB_SCRIPT_THAI,
+  HB_SCRIPT_TIBETAN,
+  HB_SCRIPT_CANADIAN_ABORIGINAL,
+  HB_SCRIPT_YI,
+  HB_SCRIPT_TAGALOG,
+  HB_SCRIPT_HANUNOO,
+  HB_SCRIPT_BUHID,
+  HB_SCRIPT_TAGBANWA,
+
+  /* Unicode-4.0 additions */
+  HB_SCRIPT_BRAILLE,
+  HB_SCRIPT_CYPRIOT,
+  HB_SCRIPT_LIMBU,
+  HB_SCRIPT_OSMANYA,
+  HB_SCRIPT_SHAVIAN,
+  HB_SCRIPT_LINEAR_B,
+  HB_SCRIPT_TAI_LE,
+  HB_SCRIPT_UGARITIC,
+
+  /* Unicode-4.1 additions */
+  HB_SCRIPT_NEW_TAI_LUE,
+  HB_SCRIPT_BUGINESE,
+  HB_SCRIPT_GLAGOLITIC,
+  HB_SCRIPT_TIFINAGH,
+  HB_SCRIPT_SYLOTI_NAGRI,
+  HB_SCRIPT_OLD_PERSIAN,
+  HB_SCRIPT_KHAROSHTHI,
+
+  /* Unicode-5.0 additions */
+  HB_SCRIPT_UNKNOWN,
+  HB_SCRIPT_BALINESE,
+  HB_SCRIPT_CUNEIFORM,
+  HB_SCRIPT_PHOENICIAN,
+  HB_SCRIPT_PHAGS_PA,
+  HB_SCRIPT_NKO,
+
+  /* Unicode-5.1 additions */
+  HB_SCRIPT_KAYAH_LI,
+  HB_SCRIPT_LEPCHA,
+  HB_SCRIPT_REJANG,
+  HB_SCRIPT_SUNDANESE,
+  HB_SCRIPT_SAURASHTRA,
+  HB_SCRIPT_CHAM,
+  HB_SCRIPT_OL_CHIKI,
+  HB_SCRIPT_VAI,
+  HB_SCRIPT_CARIAN,
+  HB_SCRIPT_LYCIAN,
+  HB_SCRIPT_LYDIAN,
+
+  /* Unicode-5.2 additions */
+  HB_SCRIPT_AVESTAN,
+  HB_SCRIPT_BAMUM,
+  HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,
+  HB_SCRIPT_IMPERIAL_ARAMAIC,
+  HB_SCRIPT_INSCRIPTIONAL_PAHLAVI,
+  HB_SCRIPT_INSCRIPTIONAL_PARTHIAN,
+  HB_SCRIPT_JAVANESE,
+  HB_SCRIPT_KAITHI,
+  HB_SCRIPT_TAI_THAM,
+  HB_SCRIPT_LISU,
+  HB_SCRIPT_MEETEI_MAYEK,
+  HB_SCRIPT_OLD_SOUTH_ARABIAN,
+  HB_SCRIPT_OLD_TURKIC,
+  HB_SCRIPT_SAMARITAN,
+  HB_SCRIPT_TAI_VIET,
+
+  /* Unicode-6.0 additions */
+  HB_SCRIPT_BATAK,
+  HB_SCRIPT_BRAHMI,
+  HB_SCRIPT_MANDAIC
+};
+#endif
+
 #ifdef OT_SUPPORT
 /* FIXME: doc. returns #items */
 EAPI int
@@ -233,7 +348,7 @@ evas_common_font_ot_populate_text_props(void *_fn, const Eina_Unicode *text,
    hb_buffer_set_unicode_funcs(buffer, evas_common_language_unicode_funcs_get());
    hb_buffer_set_language(buffer, hb_language_from_string(
             evas_common_language_from_locale_get()));
-   hb_buffer_set_script(buffer, props->script);
+   hb_buffer_set_script(buffer, _evas_script_to_harfbuzz[props->script]);
    hb_buffer_set_direction(buffer,
          (props->bidi.dir == EVAS_BIDI_DIRECTION_RTL) ?
          HB_DIRECTION_RTL : HB_DIRECTION_LTR);

legacy/evas/src/lib/engines/common/language/evas_language_utils.c

@@ -23,6 +23,8 @@
 # include <hb.h>
 #endif
 
+#include "evas_script_table.h"
+
 /* FIXME: rename and move */
 void *
 evas_common_language_unicode_funcs_get(void)
@@ -33,19 +35,38 @@ evas_common_language_unicode_funcs_get(void)
    return NULL;
 }
 
+static Evas_Script_Type
+_evas_common_language_char_script_search(Eina_Unicode unicode)
+{
+   int min = 0;
+   int max  = (sizeof(_evas_script_slow_table) /
+      sizeof(_evas_script_slow_table[0])) - 1;
+   int mid;
+
+   do
+     {
+        mid = (min + max) / 2;
+
+        if (unicode < _evas_script_slow_table[mid].start)
+           max = mid - 1;
+        else if (unicode >= _evas_script_slow_table[mid].start +
+              _evas_script_slow_table[mid].len)
+           min = mid + 1;
+        else
+           return _evas_script_slow_table[mid].script;
+     }
+   while (min <= max);
+
+   return EVAS_SCRIPT_UNKNOWN;
+}
+
 Evas_Script_Type
 evas_common_language_char_script_get(Eina_Unicode unicode)
 {
-#ifdef USE_HARFBUZZ
-   static hb_unicode_funcs_t *funcs;
-   if (!funcs)
-        funcs = evas_common_language_unicode_funcs_get();
-   if (funcs)
-      return hb_unicode_get_script(funcs, unicode);
-#else
-   (void) unicode;
-#endif
-   return EVAS_SCRIPT_COMMON;
+   if (unicode < EVAS_SCRIPT_DIRECT_TABLE_LIMIT)
+      return _evas_script_fast_table[unicode];
+   else
+      return _evas_common_language_char_script_search(unicode);
 }
 
 int