Only case-fold Cherokee under unicode flag (matches v8's behavior).

chakra-core · Jan 12, 2017 · dc3e750 · dc3e750
1 parent fbfb953
commit dc3e750
Show file tree

Hide file tree

Showing 2 changed files with 29 additions and 28 deletions.
diff --git a/lib/Parser/CaseInsensitive.cpp b/lib/Parser/CaseInsensitive.cpp
@@ -440,9 +440,9 @@ END {
             1, MappingSource::UnicodeData, 0x10a0, 0x10c5, 0, 7264, 7264, 7264,
             1, MappingSource::UnicodeData, 0x10c7, 0x10c7, 0, 7264, 7264, 7264,
             1, MappingSource::UnicodeData, 0x10cd, 0x10cd, 0, 7264, 7264, 7264,
-            1, MappingSource::UnicodeData, 0x13a0, 0x13ef, 0, 38864, 38864, 38864,
-            1, MappingSource::UnicodeData, 0x13f0, 0x13f5, 0, 8, 8, 8,
-            1, MappingSource::UnicodeData, 0x13f8, 0x13fd, -8, 0, 0, 0,
+            1, MappingSource::CaseFolding, 0x13a0, 0x13ef, 0, 38864, 38864, 38864,
+            1, MappingSource::CaseFolding, 0x13f0, 0x13f5, 0, 8, 8, 8,
+            1, MappingSource::CaseFolding, 0x13f8, 0x13fd, -8, 0, 0, 0,
             1, MappingSource::UnicodeData, 0x1d79, 0x1d79, 0, 35332, 35332, 35332,
             1, MappingSource::UnicodeData, 0x1d7d, 0x1d7d, 0, 3814, 3814, 3814,
             2, MappingSource::UnicodeData, 0x1e00, 0x1e5f, -1, 1, 1, 1,
@@ -565,7 +565,7 @@ END {
             1, MappingSource::UnicodeData, 0xa7b3, 0xa7b3, 0, 928, 928, 928,
             2, MappingSource::UnicodeData, 0xa7b4, 0xa7b7, -1, 1, 1, 1,
             1, MappingSource::UnicodeData, 0xab53, 0xab53, -928, 0, 0, 0,
-            1, MappingSource::UnicodeData, 0xab70, 0xabbf, -38864, 0, 0, 0,
+            1, MappingSource::CaseFolding, 0xab70, 0xabbf, -38864, 0, 0, 0,
             1, MappingSource::UnicodeData, 0xff21, 0xff3a, 0, 32, 32, 32,
             1, MappingSource::UnicodeData, 0xff41, 0xff5a, -32, 0, 0, 0,
             1, MappingSource::CaseFolding, 0x10400, 0x10427, 0, 40, 40, 40,

diff --git a/test/es6/regex-unicode-CaseInsensitive.js b/test/es6/regex-unicode-CaseInsensitive.js
@@ -268,7 +268,7 @@ assertMatches(/\ua69a/i, 0xa69b, '\ua69b');
 assertMatches(/\ua69b/i, 0xa69a, '\ua69a');
 assertMatches(/\ua69b/i, 0xa69b, '\ua69b');
 
-// New Cherokee uppercase-lowercase mappings
+// New Cherokee uppercase-lowercase mappings and case-mapping pairs.
 assertMatches(/\u13a0/iu, 0xab70, '\uab70');
 assertMatches(/\uab70/iu, 0x13a0, '\u13a0');
 assertMatches(/\u13a1/iu, 0xab71, '\uab71');
@@ -279,17 +279,6 @@ assertMatches(/\uabbe/iu, 0x13ee, '\u13ee');
 assertMatches(/\u13ef/iu, 0xabbf, '\uabbf');
 assertMatches(/\uabbf/iu, 0x13ef, '\u13ef');
 
-assertMatches(/\u13a0/i, 0xab70, '\uab70');
-assertMatches(/\uab70/i, 0x13a0, '\u13a0');
-assertMatches(/\u13a1/i, 0xab71, '\uab71');
-assertMatches(/\uab71/i, 0x13a1, '\u13a1');
-// ...
-assertMatches(/\u13ee/i, 0xabbe, '\uabbe');
-assertMatches(/\uabbe/i, 0x13ee, '\u13ee');
-assertMatches(/\u13ef/i, 0xabbf, '\uabbf');
-assertMatches(/\uabbf/i, 0x13ef, '\u13ef');
-
-// New Cherokee case-mapping pairs
 assertMatches(/\u13f0/iu, 0x13f8, '\u13f8');
 assertMatches(/\u13f8/iu, 0x13f0, '\u13f0');
 assertMatches(/\u13f1/iu, 0x13f9, '\u13f9');
@@ -303,18 +292,30 @@ assertMatches(/\u13fc/iu, 0x13f4, '\u13f4');
 assertMatches(/\u13f5/iu, 0x13fd, '\u13fd');
 assertMatches(/\u13fd/iu, 0x13f5, '\u13f5');
 
-assertMatches(/\u13f0/i, 0x13f8, '\u13f8');
-assertMatches(/\u13f8/i, 0x13f0, '\u13f0');
-assertMatches(/\u13f1/i, 0x13f9, '\u13f9');
-assertMatches(/\u13f9/i, 0x13f1, '\u13f1');
-assertMatches(/\u13f2/i, 0x13fa, '\u13fa');
-assertMatches(/\u13fa/i, 0x13f2, '\u13f2');
-assertMatches(/\u13f3/i, 0x13fb, '\u13fb');
-assertMatches(/\u13fb/i, 0x13f3, '\u13f3');
-assertMatches(/\u13f4/i, 0x13fc, '\u13fc');
-assertMatches(/\u13fc/i, 0x13f4, '\u13f4');
-assertMatches(/\u13f5/i, 0x13fd, '\u13fd');
-assertMatches(/\u13fd/i, 0x13f5, '\u13f5');
+// REVIEW: Not matching without /u is compat with v8, but not what was expected from the UCD data.
+// REVIEW: Need to confirm with someone more familiar with spec in this area.
+assertDoesNotMatch(/\u13a0/i, 0xab70, '\uab70');
+assertDoesNotMatch(/\uab70/i, 0x13a0, '\u13a0');
+assertDoesNotMatch(/\u13a1/i, 0xab71, '\uab71');
+assertDoesNotMatch(/\uab71/i, 0x13a1, '\u13a1');
+// ...
+assertDoesNotMatch(/\u13ee/i, 0xabbe, '\uabbe');
+assertDoesNotMatch(/\uabbe/i, 0x13ee, '\u13ee');
+assertDoesNotMatch(/\u13ef/i, 0xabbf, '\uabbf');
+assertDoesNotMatch(/\uabbf/i, 0x13ef, '\u13ef');
+
+assertDoesNotMatch(/\u13f0/i, 0x13f8, '\u13f8');
+assertDoesNotMatch(/\u13f8/i, 0x13f0, '\u13f0');
+assertDoesNotMatch(/\u13f1/i, 0x13f9, '\u13f9');
+assertDoesNotMatch(/\u13f9/i, 0x13f1, '\u13f1');
+assertDoesNotMatch(/\u13f2/i, 0x13fa, '\u13fa');
+assertDoesNotMatch(/\u13fa/i, 0x13f2, '\u13f2');
+assertDoesNotMatch(/\u13f3/i, 0x13fb, '\u13fb');
+assertDoesNotMatch(/\u13fb/i, 0x13f3, '\u13f3');
+assertDoesNotMatch(/\u13f4/i, 0x13fc, '\u13fc');
+assertDoesNotMatch(/\u13fc/i, 0x13f4, '\u13f4');
+assertDoesNotMatch(/\u13f5/i, 0x13fd, '\u13fd');
+assertDoesNotMatch(/\u13fd/i, 0x13f5, '\u13f5');
 
 // Latin extensions added in Unicode 7.0
 assertMatches(/\ua796/iu, 0xa796, '\ua796');