ES6 unicode escapes, part 2: Regexps.

test/mjsunit/harmony/unicode-escapes-in-regexps.js

Index: test/mjsunit/harmony/unicode-escapes-in-regexps.js
diff --git a/test/mjsunit/harmony/unicode-escapes-in-regexps.js b/test/mjsunit/harmony/unicode-escapes-in-regexps.js
new file mode 100644
index 0000000000000000000000000000000000000000..41abcb78961058b8d01b689134f9feab33f15adc
--- /dev/null
+++ b/test/mjsunit/harmony/unicode-escapes-in-regexps.js
@@ -0,0 +1,212 @@
+// Copyright 2014 the V8 project authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+// ES6 extends the \uxxxx escape and also allows \u{xxxxx}.
+
+// Flags: --harmony-unicode --harmony-regexps
+
+function testRegexpHelper(r) {
+  assertTrue(r.test("foo"));
+  assertTrue(r.test("boo"));
+  assertFalse(r.test("moo"));
+}
+
+
+(function TestUnicodeEscapes() {
+  testRegexpHelper(/(\u0066|\u0062)oo/);
+  testRegexpHelper(/(\u0066|\u0062)oo/u);
+  testRegexpHelper(/(\u{0066}|\u{0062})oo/u);
+  testRegexpHelper(/(\u{66}|\u{000062})oo/u);
+
+  // Note that we need \\ inside a string, otherwise it's interpreted as a
+  // unicode escape inside a string.
+  testRegexpHelper(new RegExp("(\\u0066|\\u0062)oo"));
+  testRegexpHelper(new RegExp("(\\u0066|\\u0062)oo", "u"));
+  testRegexpHelper(new RegExp("(\\u{0066}|\\u{0062})oo", "u"));
+  testRegexpHelper(new RegExp("(\\u{66}|\\u{000062})oo", "u"));
+
+  // Though, unicode escapes via strings should work too.
+  testRegexpHelper(new RegExp("(\u0066|\u0062)oo"));
+  testRegexpHelper(new RegExp("(\u0066|\u0062)oo", "u"));
+  testRegexpHelper(new RegExp("(\u{0066}|\u{0062})oo", "u"));
+  testRegexpHelper(new RegExp("(\u{66}|\u{000062})oo", "u"));
+})();
+
+
+(function TestUnicodeEscapesInCharacterClasses() {
+  testRegexpHelper(/[\u0062-\u0066]oo/);
+  testRegexpHelper(/[\u0062-\u0066]oo/u);
+  testRegexpHelper(/[\u{0062}-\u{0066}]oo/u);
+  testRegexpHelper(/[\u{62}-\u{00000066}]oo/u);
+
+  // Note that we need \\ inside a string, otherwise it's interpreted as a
+  // unicode escape inside a string.
+  testRegexpHelper(new RegExp("[\\u0062-\\u0066]oo"));
+  testRegexpHelper(new RegExp("[\\u0062-\\u0066]oo", "u"));
+  testRegexpHelper(new RegExp("[\\u{0062}-\\u{0066}]oo", "u"));
+  testRegexpHelper(new RegExp("[\\u{62}-\\u{00000066}]oo", "u"));
+
+  // Though, unicode escapes via strings should work too.
+  testRegexpHelper(new RegExp("[\u0062-\u0066]oo"));
+  testRegexpHelper(new RegExp("[\u0062-\u0066]oo", "u"));
+  testRegexpHelper(new RegExp("[\u{0062}-\u{0066}]oo", "u"));
+  testRegexpHelper(new RegExp("[\u{62}-\u{00000066}]oo", "u"));
+})();
+
+
+(function TestBraceEscapesWithoutUnicodeFlag() {
+  // \u followed by illegal escape will be parsed as u. {x} will be the
+  // character count.
+  function helper1(r) {
+    assertFalse(r.test("fbar"));
+    assertFalse(r.test("fubar"));
+    assertTrue(r.test("fuubar"));
+    assertFalse(r.test("fuuubar"));
+  }
+  helper1(/f\u{2}bar/);
+  helper1(new RegExp("f\\u{2}bar"));
+
+  function helper2(r) {
+    assertFalse(r.test("fbar"));
+    assertTrue(r.test("fubar"));
+    assertTrue(r.test("fuubar"));
+    assertFalse(r.test("fuuubar"));
+  }
+
+  helper2(/f\u{1,2}bar/);
+  helper2(new RegExp("f\\u{1,2}bar"));
+
+  function helper3(r) {
+    assertTrue(r.test("u"));
+    assertTrue(r.test("{"));
+    assertTrue(r.test("2"));
+    assertTrue(r.test("}"));
+    assertFalse(r.test("q"));
+    assertFalse(r.test("("));
+    assertFalse(r.test(")"));
+  }
+  helper3(/[\u{2}]/);
+  helper3(new RegExp("[\\u{2}]"));
+})();
+
+
+(function TestInvalidEscapes() {
+  // Without the u flag, invalid unicode escapes and other invalid escapes are
+  // treated as identity escapes.
+  function helper1(r) {
+    assertTrue(r.test("firstuxz89second"));
+  }
+  helper1(/first\u\x\z\8\9second/);
+  helper1(new RegExp("first\\u\\x\\z\\8\\9second"));
+
+  function helper2(r) {
+    assertTrue(r.test("u"));
+    assertTrue(r.test("x"));
+    assertTrue(r.test("z"));
+    assertTrue(r.test("8"));
+    assertTrue(r.test("9"));
+    assertFalse(r.test("q"));
+    assertFalse(r.test("7"));
+  }
+  helper2(/[\u\x\z\8\9]/);
+  helper2(new RegExp("[\\u\\x\\z\\8\\9]"));
+
+  // However, with the u flag, these are treated as invalid escapes.
+  assertThrows("/\\u/u", SyntaxError);
+  assertThrows("/\\u12/u", SyntaxError);
+  assertThrows("/\\ufoo/u", SyntaxError);
+  assertThrows("/\\x/u", SyntaxError);
+  assertThrows("/\\xfoo/u", SyntaxError);
+  assertThrows("/\\z/u", SyntaxError);
+  assertThrows("/\\8/u", SyntaxError);
+  assertThrows("/\\9/u", SyntaxError);
+
+  assertThrows("new RegExp('\\\\u', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\u12', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\ufoo', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\x', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\xfoo', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\z', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\8', 'u')", SyntaxError);
+  assertThrows("new RegExp('\\\\9', 'u')", SyntaxError);
+})();
+
+
+(function TestTooBigHexEscape() {
+  // The hex number inside \u{} has a maximum value.
+  /\u{10ffff}/u
+  new RegExp("\\u{10ffff}", "u")
+  assertThrows("/\\u{110000}/u", SyntaxError);
+  assertThrows("new RegExp('\\\\u{110000}', 'u')", SyntaxError);
+
+  // Without the u flag, they're of course fine ({x} is the count).
+  /\u{110000}/
+  new RegExp("\\u{110000}")

[rossberg, 2015/01/08 12:47:45]

Perhaps add the case

assertThrows{"new RegExp('\\\\u{1f}')", SyntaxError)

[marja, 2015/01/08 13:42:19]

That doesn't throw even w/ the current implementation.

Looks like the thing inside { } doesn't need to be a number, if it's not, it's
just not treated as a count specifier.

[rossberg, 2015/01/08 14:11:45]

Hm, that looks like a spec deviation then, as I don't see anything in the
grammar allowing it. But that's clearly a separate issue then.

+})();
+
+
+(function TestSyntaxEscapes() {
+  // Syntax escapes work the same with or without the u flag.
+  function helper(r) {
+    assertTrue(r.test("foo[bar"));
+    assertFalse(r.test("foo]bar"));
+  }
+  helper(/foo\[bar/);
+  helper(new RegExp("foo\\[bar"));
+  helper(/foo\[bar/u);
+  helper(new RegExp("foo\\[bar", "u"));
+})();
+
+
+(function TestUnicodeSurrogates() {
+  // U+10E6D corresponds to the surrogate pair [U+D803, U+DE6D].
+  function helper(r) {
+    assertTrue(r.test("foo\u{10e6d}bar"));
+  }
+  helper(/foo\ud803\ude6dbar/u);
+  helper(new RegExp("foo\\ud803\\ude6dbar", "u"));
+})();
+
+
+(function AllFlags() {
+  // Test that we can pass all possible regexp flags and they work properly.
+  function helper1(r) {
+    assertTrue(r.global);
+    assertTrue(r.ignoreCase);
+    assertTrue(r.multiline);
+    assertTrue(r.sticky);
+    assertTrue(r.unicode);
+  }
+
+  helper1(/foo/gimyu);
+  helper1(new RegExp("foo", "gimyu"));
+
+  function helper2(r) {
+    assertFalse(r.global);
+    assertFalse(r.ignoreCase);
+    assertFalse(r.multiline);
+    assertFalse(r.sticky);
+    assertFalse(r.unicode);
+  }
+
+  helper2(/foo/);
+  helper2(new RegExp("foo"));
+})();
+
+
+(function DuplicatedFlags() {
+  // Test that duplicating the u flag is not allowed.
+  assertThrows("/foo/ugu");
+  assertThrows("new RegExp('foo', 'ugu')");
+})();
+
+
+(function ToString() {
+  // Test that the u flag is included in the string representation of regexps.
+  function helper(r) {
+    assertEquals(r.toString(), "/foo/u");
+  }
+  helper(/foo/u);
+  helper(new RegExp("foo", "u"));
+})();

[mathias, 2015/01/08 12:29:07]

How about adding a test for `/[a-c\-]/u`, given the above comment?

Also you may want to update `test/mjsunit/harmony/regexp-flags.js` so it
includes `u`.

[marja, 2015/01/08 13:42:19]

I'll do the first part if the spec gets updated, second part done.