Recognize identifiers containing unicode escape sequences.

Task-number: QTBUG-8108
Reviewed-by: Olivier Goffart

2 files changed, 43 insertions, 3 deletions

diff --git a/src/declarative/qml/parser/qdeclarativejslexer.cpp b/src/declarative/qml/parser/qdeclarativejslexer.cpp
index a616146..7c1c30c 100644
--- a/src/declarative/qml/parser/qdeclarativejslexer.cpp
+++ b/src/declarative/qml/parser/qdeclarativejslexer.cpp
@@ -484,6 +484,8 @@ int Lexer::lex()
         stackToken = -1;
     }
 
+    bool identifierWithEscapedUnicode = false;
+
     while (!done) {
         switch (state) {
         case Start:
@@ -523,7 +525,26 @@ int Lexer::lex()
                 state = InString;
                 multiLineString = false;
                 stringType = current;
+            } else if (current == '\\' && next1 == 'u') {
+                identifierWithEscapedUnicode = true;
+                recordStartPos();
+
+                shift(2); // skip the unicode escape prefix `\u'
+
+                if (isHexDigit(current) && isHexDigit(next1) &&
+                     isHexDigit(next2) && isHexDigit(next3)) {
+                    record16(convertUnicode(current, next1, next2, next3));
+                    shift(3);
+                    state = InIdentifier;
+                } else {
+                    setDone(Bad);
+                    err = IllegalUnicodeEscapeSequence;
+                    errmsg = QCoreApplication::translate("QDeclarativeParser", "Illegal unicode escape sequence");
+                    break;
+                }
+
             } else if (isIdentLetter(current)) {
+                identifierWithEscapedUnicode = false;
                 recordStartPos();
                 record16(current);
                 state = InIdentifier;
@@ -683,6 +704,21 @@ int Lexer::lex()
             if (isIdentLetter(current) || isDecimalDigit(current)) {
                 record16(current);
                 break;
+            } else if (current == '\\' && next1 == 'u') {
+                identifierWithEscapedUnicode = true;
+                shift(2); // skip the unicode escape prefix `\u'
+
+                if (isHexDigit(current) && isHexDigit(next1) &&
+                     isHexDigit(next2) && isHexDigit(next3)) {
+                    record16(convertUnicode(current, next1, next2, next3));
+                    shift(3);
+                    break;
+                } else {
+                    setDone(Bad);
+                    err = IllegalUnicodeEscapeSequence;
+                    errmsg = QCoreApplication::translate("QDeclarativeParser", "Illegal unicode escape sequence");
+                    break;
+                }
             }
             setDone(Identifier);
             break;
@@ -825,7 +861,11 @@ int Lexer::lex()
             delimited = true;
         return token;
     case Identifier:
-        if ((token = findReservedWord(buffer16, pos16)) < 0) {
+        token = -1;
+        if (! identifierWithEscapedUnicode)
+            token = findReservedWord(buffer16, pos16);
+
+        if (token < 0) {
             /* TODO: close leak on parse error. same holds true for String */
             if (driver)
                 qsyylval.ustr = driver->intern(buffer16, pos16);
diff --git a/tests/auto/declarative/parserstress/tst_parserstress.cpp b/tests/auto/declarative/parserstress/tst_parserstress.cpp
index 971496d..f61ca9f 100644
--- a/tests/auto/declarative/parserstress/tst_parserstress.cpp
+++ b/tests/auto/declarative/parserstress/tst_parserstress.cpp
@@ -131,8 +131,8 @@ void tst_parserstress::ecmascript()
     QDeclarativeComponent component(&engine);
     component.setData(qmlData, QUrl::fromLocalFile(SRCDIR + QString("/dummy.qml")));
     QSet<QString> failingTests;
-    failingTests << "uc-003.js" << "uc-005.js" << "regress-352044-02-n.js"
-                 << "regress-334158.js" << "regress-58274.js";
+    failingTests << "regress-352044-02-n.js"
+                 << "regress-334158.js";
     QFileInfo info(file);
     foreach (const QString &failing, failingTests) {
         if (info.fileName().endsWith(failing)) {