- From: Mike Shaver <mike.shaver@gmail.com>
- Date: Wed, 23 Jun 2004 13:34:45 -0400
On Wed, 16 Jun 2004 11:27:43 +0100, Jim Ley <jim.ley at gmail.com> wrote: > I'm not actually confident it is non-conformant (ECMAScript has lots > of get outs allowing all sorts of extensions.) but it's on the > interpretation of \b - Mozilla is more generous than the standard on > what concerns a character, it's better IMO, and I wouldn't bother > filing a bug as I'm sure it'll resolve as wont fix. I don't think the behaviour you describe is non-conformant, but I'll review the standard to be sure. There are many cases in which both IE and Mozilla treat input which would generate an error in a no-extension ECMA-262 implementation as valid, but this doesn't sound quite like the same case (example: treating 08 as a valid literal for the number 8, rather than an invalid octal literal; people get those values out of date strings all the time, and it's been permitted for aeons on the web). If you would file a bug about this case -- presuming none already exists -- the Mozilla JS team would appreciate it. There are ways for a page author to ask for stricter ECMA interpretations, in some cases, and this might be a suitable change for that mode. It's also _quite_ possible, sadly, that this is just a bug in the RegExp implementation in Mozilla. Mike
Received on Wednesday, 23 June 2004 10:34:45 UTC