Add 1-element caches to RegExp.exec and String.replace. We...

Add 1-element caches to RegExp.exec and String.replace.  We
probably want to remove this again if and when Dromaeo is
fixed so this strategy doesn't pay off.

Committed: http://code.google.com/p/v8/source/detail?r=4083

[Mads Ager (chromium), 2010-03-10 08:12:22]

LGTM

http://codereview.chromium.org/817001/diff/1/2
File src/regexp-delay.js (right):

http://codereview.chromium.org/817001/diff/1/2#newcode164
src/regexp-delay.js:164: cachedSubject === string) {
I would use %_ObjectEquals for the string and regexp comparisons here.

http://codereview.chromium.org/817001/diff/1/3
File src/string.js (right):

http://codereview.chromium.org/817001/diff/1/3#newcode250
src/string.js:250: subject === cachedReplaceSubject &&
%_ObjectEquals for these?

[Lasse Reichstein, 2010-03-10 08:53:24]

http://codereview.chromium.org/817001/diff/1/2
File src/regexp-delay.js (right):

http://codereview.chromium.org/817001/diff/1/2#newcode162
src/regexp-delay.js:162: if (cachedLastIndex === this.lastIndex &&
lastIndex is only relevant if the regexp is global.

http://codereview.chromium.org/817001/diff/1/2#newcode163
src/regexp-delay.js:163: cachedRegexp === this &&
Checking the regexp itself is unsafe. Its behavior can be overwritten by
RegExp.prototype.compile.
You should check the underlying FixedArray instead (which is good, because
that's what the RegExp compile cache reuses too).

http://codereview.chromium.org/817001/diff/1/2#newcode166
src/regexp-delay.js:166: if (last == null) {
Use ===?

http://codereview.chromium.org/817001/diff/1/3
File src/string.js (right):

http://codereview.chromium.org/817001/diff/1/3#newcode250
src/string.js:250: subject === cachedReplaceSubject &&
Test FixedArray of subject.

[Lasse Reichstein, 2010-03-10 08:55:24]

See also http://codereview.chromium.org/652224/
(which didn't give as impressive results, for some reason).