Irregexp:

* Fix UC16 character classes on ASCII subjects.
* Fix sign problem in Irregexp interpreter.
* Make passes over text nodes more readable.
Review URL: http://codereview.chromium.org/21450

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1304 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

src/bytecodes-irregexp.h

@@ -33,7 +33,10 @@ namespace v8 { namespace internal {
 
 
 static const int BYTECODE_MASK = 0xff;
-static const unsigned int MAX_FIRST_ARG = 0xffffffu;
+// The first argument is packed in with the byte code in one word, but so it
+// has 24 bits, but it can be positive and negative so only use 23 bits for
+// positive values.
+static const unsigned int MAX_FIRST_ARG = 0x7fffffu;
 static const int BYTECODE_SHIFT = 8;
 
 #define BYTECODE_ITERATOR(V)                                                   \

src/jsregexp.h

@@ -824,11 +824,15 @@ class TextNode: public SeqRegExpNode {
 
  private:
   enum TextEmitPassType {
-    NON_ASCII_MATCH,
-    CHARACTER_MATCH,
-    CASE_CHARACTER_MATCH,
-    CHARACTER_CLASS_MATCH
+    NON_ASCII_MATCH,             // Check for characters that can't match.
+    SIMPLE_CHARACTER_MATCH,      // Case-dependent single character check.
+    NON_LETTER_CHARACTER_MATCH,  // Check characters that have no case equivs.
+    CASE_CHARACTER_MATCH,        // Case-independent single character check.
+    CHARACTER_CLASS_MATCH        // Character class.
   };
+  static bool SkipPass(int pass, bool ignore_case);
+  static const int kFirstRealPass = SIMPLE_CHARACTER_MATCH;
+  static const int kLastPass = CHARACTER_CLASS_MATCH;
   void TextEmitPass(RegExpCompiler* compiler,
                     TextEmitPassType pass,
                     bool preloaded,

src/objects.h

@@ -3240,6 +3240,7 @@ class String: public HeapObject {
 
   // Max ascii char code.
   static const int kMaxAsciiCharCode = unibrow::Utf8::kMaxOneByteChar;
+  static const unsigned kMaxAsciiCharCodeU = unibrow::Utf8::kMaxOneByteChar;
   static const int kMaxUC16CharCode = 0xffff;
 
   // Minimum length for a cons or sliced string.

test/mjsunit/regexp-UC16.js

@@ -41,3 +41,7 @@ var longUC16String = "x\u03a3\u03c2\u039b\u03c2\u03c3\u03bb\u03c3\u03a3\u03bb";
 assertEquals(longUC16String + "," + longUC16String.substring(1,4),
              String(/x(...)\1\1/i.exec(longUC16String)),
              "backref-UC16-twice");
+
+assertFalse(/\xc1/i.test('fooA'), "quickcheck-uc16-pattern-ascii-subject");
+assertFalse(/[\xe9]/.test('i'), "charclass-uc16-pattern-ascii-subject");
+assertFalse(/\u5e74|\u6708/.test('t'), "alternation-uc16-pattern-ascii-subject");