CLDR-16595 Add DisplayAndInputProcessor to normalize Kashmiri YEH WIT…

…H TAIL -> KASHMIRI YEH
unicode-org · Nov 3, 2023 · 81e2656 · 81e2656
1 parent 12380e8
commit 81e2656
Showing 1 changed file with 5 additions and 0 deletions.
diff --git a/tools/cldr-code/src/main/java/org/unicode/cldr/test/DisplayAndInputProcessor.java b/tools/cldr-code/src/main/java/org/unicode/cldr/test/DisplayAndInputProcessor.java
@@ -180,6 +180,7 @@ public class DisplayAndInputProcessor {
     private static final CLDRLocale GERMAN_SWITZERLAND = CLDRLocale.getInstance("de_CH");
     private static final CLDRLocale SWISS_GERMAN = CLDRLocale.getInstance("gsw");
     private static final CLDRLocale FF_ADLAM = CLDRLocale.getInstance("ff_Adlm");
+    private static final CLDRLocale KASHMIRI = CLDRLocale.getInstance("ks");
     public static final Set<String> LANGUAGES_USING_MODIFIER_APOSTROPHE =
             new HashSet<>(
                     Arrays.asList(
@@ -223,6 +224,8 @@ public class DisplayAndInputProcessor {
 
     private static final char[][] URDU_PLUS_CONVERSIONS = {{'\u0643', '\u06A9'}}; //  wrong char
 
+    private static final char[][] KASHMIRI_CONVERSIONS = {{'ۍ', 'ؠ'}}; //  wrong char (see CLDR-16595)
+
     private static final ZawgyiDetector detector = new ZawgyiDetector();
     private static final Transliterator zawgyiUnicodeTransliterator =
             Transliterator.getInstance("Zawgyi-my");
@@ -598,6 +601,8 @@ private String processLocaleSpecificInput(String path, String value, boolean isU
             value = replaceChars(path, value, URDU_PLUS_CONVERSIONS, true);
         } else if (locale.childOf(FF_ADLAM) && !isUnicodeSet) {
             value = fixAdlamNasalization(value);
+        } else if (locale.childOf(KASHMIRI)) {
+            value = replaceChars(path, value, KASHMIRI_CONVERSIONS, false);
         }
         return value;
     }