CLDR-17847 Fix NPE in CodePointEscaper

unicode-org · Jul 30, 2024 · 28011cb · 28011cb
1 parent 31cdf11
commit 28011cb
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 1 deletion.
diff --git a/tools/cldr-code/src/main/java/org/unicode/cldr/util/CodePointEscaper.java b/tools/cldr-code/src/main/java/org/unicode/cldr/util/CodePointEscaper.java
@@ -228,9 +228,10 @@ public static String toUnescaped(String escaped) {
     public static String toExample(int codePoint) {
         CodePointEscaper cpe = _fromCodePoint.get(codePoint);
         if (cpe == null) { // hex
+            final String name = UCharacter.getExtendedName(codePoint);
             return codePointToEscaped(codePoint)
                     + HAS_NAME
-                    + UCharacter.getName(codePoint).toLowerCase();
+                    + (name != null ? name.toLowerCase() : "");
         } else {
             return CodePointEscaper.codePointToEscaped(cpe.codePoint)
                     + HAS_NAME

diff --git a/tools/cldr-code/src/test/java/org/unicode/cldr/unittest/UnicodeSetPrettyPrinterTest.java b/tools/cldr-code/src/test/java/org/unicode/cldr/unittest/UnicodeSetPrettyPrinterTest.java
@@ -439,4 +439,19 @@ public void TestStringEscaper() {
             assertEquals(expected, expectedRoundtrip, actualRoundtrip);
         }
     }
+
+    public void test17847() {
+        String path =
+                "//ldml/characters/parseLenients[@scope=\"general\"][@level=\"lenient\"]/parseLenient[@sample=\"$\"]";
+        String value = "[\\$＄﹩ \\uFFFF]";
+        UnicodeSet valueSet = new UnicodeSet(value).freeze();
+        UnicodeSet examples = new UnicodeSet();
+
+        // throws exception
+        if (valueSet.containsSome(CodePointEscaper.FORCE_ESCAPE)) {
+            for (String nsm : new UnicodeSet(valueSet).retainAll(CodePointEscaper.FORCE_ESCAPE)) {
+                assertNotNull(Utility.hex(nsm), CodePointEscaper.toExample(nsm.codePointAt(0)));
+            }
+        }
+    }
 }