add internal api: _unicode_normalize_script

src/font/unicode-script.c

@@ -289,5 +289,30 @@ int GUIAPI UCharScriptTypeFromISO15924 (Uint32 iso15924)
     return UCHAR_SCRIPT_UNKNOWN;
 }
 
+/*
+ * https://www.w3.org/TR/css-text-3/#script-tagging
+ */
+UCharScriptType _unicode_normalize_script(LanguageCode cl,
+        UCharScriptType ws)
+{
+    if (cl == LANGCODE_zh && ws == UCHAR_SCRIPT_INVALID_CODE) {
+        return UCHAR_SCRIPT_HAN;
+    }
+    else if (ws == UCHAR_SCRIPT_BOPOMOFO) {
+        return UCHAR_SCRIPT_HAN;
+    }
+    else if (cl == LANGCODE_ja && ws == UCHAR_SCRIPT_INVALID_CODE) {
+        return UCHAR_SCRIPT_KATAKANA;
+    }
+    else if (cl == LANGCODE_ko && ws == UCHAR_SCRIPT_INVALID_CODE) {
+        return UCHAR_SCRIPT_HANGUL;
+    }
+    else if (cl == LANGCODE_unknown || ws == UCHAR_SCRIPT_INVALID_CODE) {
+        return UCHAR_SCRIPT_INVALID_CODE;
+    }
+
+    return ws;
+}
+
 #endif /* _MGCHARSET_UNICODE */
 

src/include/unicode-ops.h

@@ -58,6 +58,9 @@ BOOL _unicode_is_emoji_emoji_default(Uchar32 ch);
 BOOL _unicode_is_emoji_keycap_base(Uchar32 ch);
 BOOL _unicode_is_regional_indicator(Uchar32 ch);
 
+UCharScriptType _unicode_normalize_script(LanguageCode content_language,
+        UCharScriptType writing_system);
+
 #ifdef __cplusplus
 }
 #endif  /* __cplusplus */

src/newgdi/shaped-glyph.c

@@ -104,6 +104,8 @@ int GUIAPI GetShapedGlyphsBasic(LOGFONT* logfont,
         return 0;
     }
 
+    writing_system = _unicode_normalize_script(content_language, writing_system);
+
     /*
      * TODO: tailor break opportunties according to the content languages.
      */
@@ -170,7 +172,7 @@ int GUIAPI GetShapedGlyphsBasic(LOGFONT* logfont,
             pos_v2l[i] = i;
     }
 
-    if (content_language == LANGCODE_ar) {
+    if (writing_system == UCHAR_SCRIPT_ARABIC) {
         /* Arabic joining */
         if (nr_ucs < LOCAL_ARRAY_SIZE)
             ar_props = local_ar_props;