gecko-dev/js
brendan%mozilla.org 26c06ddc42 - String.prototype.match, .search, and .replace all use match_or_replace as
a common subroutine, but only replace might need to dereference the regexp
  created from a non-regexp-type argument, which is converted to string and
  compiled into a regexp data struct unowned by any rooted object.  This case
  was broken: a dangling GlobData.regexp pointer could be dereferenced in the
  "lambda replace" code under str_replace, in find_replen, *after* the call
  to match_or_replace had destroyed the regexp and returned to str_replace.
  Bug 167658 reported the symptoms.
- Consolidate GlobData members into a flags word, avoiding yet another ad hoc
  paramter to m_or_r (forceFlat).
- Use a local root in str_match to avoid adding and removing a global root.
2002-09-23 18:33:17 +00:00
..
benchmarks
docs apply scole@planetweb.com's fix for JS_InitClass (bug 162946) 2002-08-16 05:11:36 +00:00
jsd Removing old nmake build makefiles. Bug #158528 r=pavlov 2002-08-10 07:55:43 +00:00
jsdj Bug 106386 rid source of misspellings 2002-03-19 04:30:17 +00:00
jsj
macbuild Fix for smoketest blocker 135697. Bump up version on the compiler test to force JS to be compiled with opt level 2. a=leaf. 2002-04-05 19:06:29 +00:00
ref
rhino Reduce "new X" usage by inlining few classes or packing their members into long field 2002-09-18 12:57:09 +00:00
src - String.prototype.match, .search, and .replace all use match_or_replace as 2002-09-23 18:33:17 +00:00
tests Initial add. Regression test for bug 170193. 2002-09-22 22:11:40 +00:00
.cvsignore
landbranch.pl
Makefile.in bug 76050, build js/jsd from toplevel makefile so we're sure we've got an xpidl compiler. r=cls 2001-05-15 20:38:26 +00:00
makefile.win Adding placeholders to silence cvs warnings 2002-09-20 09:14:42 +00:00