Fix bug 486, Cyrillic character ranges in case independent regexps.

http://code.google.com/p/v8/issues/detail?id=486
Review URL: http://codereview.chromium.org/361033

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3236 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

src/jsregexp.cc

@@ -2440,8 +2440,8 @@ void TextNode::MakeCaseIndependent() {
       RegExpCharacterClass* cc = elm.data.u_char_class;
       ZoneList<CharacterRange>* ranges = cc->ranges();
       int range_count = ranges->length();
-      for (int i = 0; i < range_count; i++) {
-        ranges->at(i).AddCaseEquivalents(ranges);
+      for (int j = 0; j < range_count; j++) {
+        ranges->at(j).AddCaseEquivalents(ranges);
       }
     }
   }
@@ -3961,7 +3961,7 @@ void CharacterRange::AddCaseEquivalents(ZoneList<CharacterRange>* ranges) {
     } else {
       start = pos;
     }
-    // Then we add the ranges on at a time, incrementing the current
+    // Then we add the ranges one at a time, incrementing the current
     // position to be after the last block each time.  The position
     // always points to the start of a block.
     while (pos < to()) {
@@ -3987,8 +3987,45 @@ void CharacterRange::AddCaseEquivalents(ZoneList<CharacterRange>* ranges) {
       }
       start = pos = block_end + 1;
     }
-  } else {
-    // TODO(plesner) when we've fixed the 2^11 bug in unibrow.
+  } else if (from() > 0 || to() < String::kMaxUC16CharCode) {
+    // Unibrow ranges don't work for high characters due to the "2^11 bug".
+    // Therefore we do something dumber for these ranges.  We don't bother
+    // if the range is 0-max (as encountered at the start of an unanchored
+    // regexp).
+    ZoneList<unibrow::uchar> *characters = new ZoneList<unibrow::uchar>(100);
+    int bottom = from();
+    int top = to();
+    for (int i = bottom; i <= top; i++) {
+      int length = uncanonicalize.get(i, '\0', chars);
+      for (int j = 0; j < length; j++) {
+        uc32 chr = chars[j];
+        if (chr != i && chr < bottom || chr > top) {
+          characters->Add(chr);
+        }
+      }
+    }
+    if (characters->length() > 0) {
+      int new_from = characters->at(0);
+      int new_to = new_from;
+      for (int i = 1; i < characters->length(); i++) {
+        int chr = characters->at(i);
+        if (chr == new_to + 1) {
+          new_to++;
+        } else {
+          if (new_to == new_from) {
+            ranges->Add(CharacterRange::Singleton(new_from));
+          } else {
+            ranges->Add(CharacterRange(new_from, new_to));
+          }
+          new_from = new_to = chr;
+        }
+      }
+      if (new_to == new_from) {
+        ranges->Add(CharacterRange::Singleton(new_from));
+      } else {
+        ranges->Add(CharacterRange(new_from, new_to));
+      }
+    }
   }
 }
 

test/mjsunit/cyrillic.js

+// Copyright 2009 the V8 project authors. All rights reserved.
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//     * Redistributions of source code must retain the above copyright
+//       notice, this list of conditions and the following disclaimer.
+//     * Redistributions in binary form must reproduce the above
+//       copyright notice, this list of conditions and the following
+//       disclaimer in the documentation and/or other materials provided
+//       with the distribution.
+//     * Neither the name of Google Inc. nor the names of its
+//       contributors may be used to endorse or promote products derived
+//       from this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+// Test Unicode character ranges in regexps.
+
+
+// Cyrillic.
+var cyrillic = {
+  FIRST: "\u0410",   // A
+  first: "\u0430",   // a
+  LAST: "\u042f",    // YA
+  last: "\u044f",    // ya
+  MIDDLE: "\u0427",  // CHE
+  middle: "\u0447",   // che
+  // Actually no characters are between the cases in Cyrillic.
+  BetweenCases: false};
+
+var SIGMA = "\u03a3";
+var sigma = "\u03c3";
+var alternative_sigma = "\u03c2";
+
+// Greek.
+var greek = {
+  FIRST: "\u0391",     // ALPHA
+  first: "\u03b1",     // alpha
+  LAST: "\u03a9",      // OMEGA
+  last: "\u03c9",      // omega
+  MIDDLE: SIGMA,       // SIGMA
+  middle: sigma,       // sigma
+  // Epsilon acute is between ALPHA-OMEGA and alpha-omega, ie it
+  // is between OMEGA and alpha.
+  BetweenCases: "\u03ad"};
+
+
+function Range(from, to, flags) {
+  return new RegExp("[" + from + "-" + to + "]", flags);
+}
+
+for (var lang = 0; lang < 2; lang++) {
+  var chars = (lang == 0) ? cyrillic : greek;
+
+  for (var i = 0; i < 2; i++) {
+    var lc = (i == 0);  // Lower case.
+    var first = lc ? chars.first : chars.FIRST;
+    var middle = lc ? chars.middle : chars.MIDDLE;
+    var last = lc ? chars.last : chars.LAST;
+    var first_other_case = lc ? chars.FIRST : chars.first;
+    var middle_other_case = lc ? chars.MIDDLE : chars.middle;
+    var last_other_case = lc ? chars.LAST : chars.last;
+
+    assertTrue(Range(first, last).test(first), 1);
+    assertTrue(Range(first, last).test(middle), 2);
+    assertTrue(Range(first, last).test(last), 3);
+
+    assertFalse(Range(first, last).test(first_other_case), 4);
+    assertFalse(Range(first, last).test(middle_other_case), 5);
+    assertFalse(Range(first, last).test(last_other_case), 6);
+
+    assertTrue(Range(first, last, "i").test(first), 7);
+    assertTrue(Range(first, last, "i").test(middle), 8);
+    assertTrue(Range(first, last, "i").test(last), 9);
+
+    assertTrue(Range(first, last, "i").test(first_other_case), 10);
+    assertTrue(Range(first, last, "i").test(middle_other_case), 11);
+    assertTrue(Range(first, last, "i").test(last_other_case), 12);
+
+    if (chars.BetweenCases) {
+      assertFalse(Range(first, last).test(chars.BetweenCases), 13);
+      assertFalse(Range(first, last, "i").test(chars.BetweenCases), 14);
+    }
+  }
+  if (chars.BetweenCases) {
+    assertTrue(Range(chars.FIRST, chars.last).test(chars.BetweenCases), 15);
+    assertTrue(Range(chars.FIRST, chars.last, "i").test(chars.BetweenCases), 16);
+  }
+}
+
+for (key in greek) {
+  assertTrue(Range(greek.FIRST, cyrillic.last).test(greek[key]), 17 + key);
+  if (cyrillic[key]) {
+    assertTrue(Range(greek.FIRST, cyrillic.last).test(cyrillic[key]), 18 + key);
+  }
+}
+
+
+for (var i = 0; i < 2; i++) {
+  var ignore_case = (i == 0);
+  var flag = ignore_case ? "i" : "";
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(greek.first), 19);
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(greek.middle), 20);
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(greek.last), 21);
+
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(cyrillic.FIRST), 22);
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(cyrillic.MIDDLE), 23);
+  assertTrue(Range(greek.first, cyrillic.LAST, flag).test(cyrillic.LAST), 24);
+
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(greek.FIRST), 25);
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(greek.MIDDLE), 26);
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(greek.LAST), 27);
+
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(cyrillic.first), 28);
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(cyrillic.middle), 29);
+  assertEquals(ignore_case, Range(greek.first, cyrillic.LAST, flag).test(cyrillic.last), 30);
+}
+
+
+for (var i = 0; i < 2; i++) {
+  var simple = (i != 0);
+  var name = simple ? "" : "[]";
+  var regex = simple ? SIGMA : "[" + SIGMA + "]";
+
+  assertFalse(new RegExp(regex).test(sigma), 31 + name);
+  assertFalse(new RegExp(regex).test(alternative_sigma), 32 + name);
+  assertTrue(new RegExp(regex).test(SIGMA), 33 + name);
+
+  assertTrue(new RegExp(regex, "i").test(sigma), 34 + name);
+  // JSC and Tracemonkey fail this one.
+  assertTrue(new RegExp(regex, "i").test(alternative_sigma), 35 + name);
+  assertTrue(new RegExp(regex, "i").test(SIGMA), 36 + name);
+
+  regex = simple ? sigma : "[" + sigma + "]";
+
+  assertTrue(new RegExp(regex).test(sigma), 41 + name);
+  assertFalse(new RegExp(regex).test(alternative_sigma), 42 + name);
+  assertFalse(new RegExp(regex).test(SIGMA), 43 + name);
+
+  assertTrue(new RegExp(regex, "i").test(sigma), 44 + name);
+  // JSC and Tracemonkey fail this one.
+  assertTrue(new RegExp(regex, "i").test(alternative_sigma), 45 + name);
+  assertTrue(new RegExp(regex, "i").test(SIGMA), 46 + name);
+
+  regex = simple ? alternative_sigma : "[" + alternative_sigma + "]";
+
+  assertFalse(new RegExp(regex).test(sigma), 51 + name);
+  assertTrue(new RegExp(regex).test(alternative_sigma), 52 + name);
+  assertFalse(new RegExp(regex).test(SIGMA), 53 + name);
+
+  // JSC and Tracemonkey fail this one.
+  assertTrue(new RegExp(regex, "i").test(sigma), 54 + name);
+  assertTrue(new RegExp(regex, "i").test(alternative_sigma), 55 + name);
+  // JSC and Tracemonkey fail this one.
+  assertTrue(new RegExp(regex, "i").test(SIGMA), 56 + name);
+}
+
+print("ok");

test/mjsunit/regress/regress-486.js

+// Copyright 2009 the V8 project authors. All rights reserved.
+// Redistribution and use in source and binary forms, with or without
+// modification, are permitted provided that the following conditions are
+// met:
+//
+//     * Redistributions of source code must retain the above copyright
+//       notice, this list of conditions and the following disclaimer.
+//     * Redistributions in binary form must reproduce the above
+//       copyright notice, this list of conditions and the following
+//       disclaimer in the documentation and/or other materials provided
+//       with the distribution.
+//     * Neither the name of Google Inc. nor the names of its
+//       contributors may be used to endorse or promote products derived
+//       from this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+var st = "\u0422\u0435\u0441\u0442";  // Test in Cyrillic characters.
+var cyrillicMatch = /^[\u0430-\u044fa-z]+$/i.test(st);  // a-ja a-z.
+assertTrue(cyrillicMatch);