Greek uppercase transforms fail for some characters

https://bugs.webkit.org/show_bug.cgi?id=237953 rdar://problem/90364897 Reviewed by Chris Dumez. Skip the fast path when uppercasing in the "el" locale. * LayoutTests/TestExpectations: Expect more tests to pass. * Source/WTF/wtf/text/StringImpl.cpp: (WTF::needsTurkishCasingRules): Tweaked comment. (WTF::needsGreekUppercasingRules): Added. (WTF::StringImpl::convertToUppercaseWithLocale): Don't use fast path when needsGreekUppercasingRules returns true. Canonical link: https://commits.webkit.org/274036@main
WebKit · Feb 3, 2024 · de609ec · de609ec
1 parent 674f7d1
commit de609ec
Show file tree

Hide file tree

Showing 2 changed files with 23 additions and 21 deletions.
diff --git a/LayoutTests/TestExpectations b/LayoutTests/TestExpectations
@@ -2999,23 +2999,11 @@ webkit.org/b/214290 imported/w3c/web-platform-tests/css/css-text/text-encoding/s
 webkit.org/b/214290 imported/w3c/web-platform-tests/css/css-text/text-indent/text-indent-tab-positions-001.html [ ImageOnlyFailure ]
 webkit.org/b/240837 imported/w3c/web-platform-tests/css/css-text/text-indent/anonymous-flex-item-001.html [ ImageOnlyFailure ]
 webkit.org/b/240837 imported/w3c/web-platform-tests/css/css-text/text-indent/anonymous-grid-item-001.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-018.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-020.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-026.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-028.html [ ImageOnlyFailure Pass ]
+webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-026.html [ ImageOnlyFailure ]
+webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-028.html [ ImageOnlyFailure ]
 webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-tailoring-001.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-tailoring-002.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-tailoring-002a.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-tailoring-003.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-tailoring-005.html [ ImageOnlyFailure ]
 webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-006.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-018.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-019.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-020.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-021.html [ ImageOnlyFailure Pass ]
 webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-039.html [ ImageOnlyFailure ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-103.html [ ImageOnlyFailure Pass ]
-webkit.org/b/183258 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-upperlower-104.html [ ImageOnlyFailure Pass ]
 webkit.org/b/195275 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-capitalize-033.html [ ImageOnlyFailure ]
 webkit.org/b/195275 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-shaping-003.html [ ImageOnlyFailure ]
 webkit.org/b/195275 imported/w3c/web-platform-tests/css/css-text/text-transform/text-transform-shaping-002.html [ ImageOnlyFailure ]

diff --git a/Source/WTF/wtf/text/StringImpl.cpp b/Source/WTF/wtf/text/StringImpl.cpp
@@ -519,14 +519,23 @@ Ref<StringImpl> StringImpl::convertToUppercaseWithoutLocale()
 
 static inline bool needsTurkishCasingRules(const AtomString& localeIdentifier)
 {
-    // Either "tr" or "az" locale, with case sensitive comparison and allowing for an ignored subtag.
+    // Either "tr" or "az" locale, with ASCII case insensitive comparison and allowing for an ignored subtag.
     UChar first = localeIdentifier[0];
     UChar second = localeIdentifier[1];
     return ((isASCIIAlphaCaselessEqual(first, 't') && isASCIIAlphaCaselessEqual(second, 'r'))
         || (isASCIIAlphaCaselessEqual(first, 'a') && isASCIIAlphaCaselessEqual(second, 'z')))
         && (localeIdentifier.length() == 2 || localeIdentifier[2] == '-');
 }
 
+static inline bool needsGreekUppercasingRules(const AtomString& localeIdentifier)
+{
+    // The "el" locale, with ASCII case insensitive comparison and allowing for an ignored subtag.
+    UChar first = localeIdentifier[0];
+    UChar second = localeIdentifier[1];
+    return (isASCIIAlphaCaselessEqual(first, 'e') && isASCIIAlphaCaselessEqual(second, 'l'))
+        && (localeIdentifier.length() == 2 || localeIdentifier[2] == '-');
+}
+
 Ref<StringImpl> StringImpl::convertToLowercaseWithLocale(const AtomString& localeIdentifier)
 {
     // Use the more-optimized code path most of the time.
@@ -568,27 +577,32 @@ Ref<StringImpl> StringImpl::convertToUppercaseWithLocale(const AtomString& local
     // Use the more-optimized code path most of the time.
     // Assuming here that the only locale-specific lowercasing is the Turkish casing rules,
     // and that the only affected character is lowercase "i".
-    if (!needsTurkishCasingRules(localeIdentifier) || find('i') == notFound)
+    const char* locale;
+    if (needsTurkishCasingRules(localeIdentifier) && find('i') != notFound) {
+        // Passing in the hardcoded locale "tr" is more efficient than
+        // allocating memory just to turn localeIdentifier into a C string, and we assume
+        // there is no difference between the uppercasing for "tr" and "az" locales.
+        locale = "tr";
+    } else if (needsGreekUppercasingRules(localeIdentifier))
+        locale = "el";
+    else
         return convertToUppercaseWithoutLocale();
 
     if (m_length > MaxLength)
         CRASH();
     int length = m_length;
 
-    // Below, we pass in the hardcoded locale "tr". Passing that is more efficient than
-    // allocating memory just to turn localeIdentifier into a C string, and we assume
-    // there is no difference between the uppercasing for "tr" and "az" locales.
     auto upconvertedCharacters = StringView(*this).upconvertedCharacters();
     const UChar* source16 = upconvertedCharacters;
     UChar* data16;
     auto newString = createUninitialized(length, data16);
     UErrorCode status = U_ZERO_ERROR;
-    int realLength = u_strToUpper(data16, length, source16, length, "tr", &status);
+    int realLength = u_strToUpper(data16, length, source16, length, locale, &status);
     if (U_SUCCESS(status) && realLength == length)
         return newString;
     newString = createUninitialized(realLength, data16);
     status = U_ZERO_ERROR;
-    u_strToUpper(data16, realLength, source16, length, "tr", &status);
+    u_strToUpper(data16, realLength, source16, length, locale, &status);
     if (U_FAILURE(status))
         return *this;
     return newString;