The parser has passed with IE6/7/8. Comment-free could be concrete now I think ;)
Experimental CSS media query parser.
Added string and comment awareness. Smaller function name as per paul…
Was overcomplicating the string handling.
Remove unneeded code. If there is an unexpected end of file, return a…
…ll the queries we found up to that point.
Make sure while loops can hang if CSS is malformed.
Added stylesheet full of bullshit to test the parser.
Use the namespace that respond already creates.
Unit tests started. YAY
Merge branch 'master' of https://github.com/pifantastic/Respond into …
All tests are passing now in IE 6-8. More tests needed!
integrating @pifantasic's comment-free media query parser
removed the separate parser file
removed parser js and comment tokens - no longer needed
updated min file
updated readme - no more comments needed
IE<=7 doesn't catch () statements in a regex. No problems, the regex …
…is not shorter and faster, thanks IE to points me that :D
Moreover, IE8 doesn't crash anymore. A new hope?
awesome. Have you tried this parser on a media query with embedded @ rules? I just did and it crashed IE8 (same problem as the latest regexp approach). Thoughts?
It's the same for me for IE6/7/8: hard crash ^^
Maybe it comes from another part of Respond... Have you some ideas about this?
I'm going to make some tests on my side ;)
Confirmed that the parser match embedded rules in @media as expected. My final thoughts are: since your regexp matched everything well too, it definitely comes from another part of Respond.
yeah. In fact, you mentioned having the same problem with the version in master too anyway. I'll look around!
Regexp in split is useless
Okay, cool. This is in the parser-integration branch now. https://github.com/scottjehl/Respond/tree/parser-integration
I'll do some comparing of the two functioning approaches and see which makes the most sense to keep. The file size difference isn't really enough to matter, so it'll come down to performance I think. You'd mentioned seeing better results with this one, so that's encouraging. (Though I do appreciate the simplicity of the regexp version). Anyway, Thanks so much, this is an awesome help.
It was a pleasure to me, I love Respond :) The next step (but not now) will be to analyze if Sheethub could be integrated or not, to have a centralize way to work with all the other CSS polyfills.
As a note, I did my benchmarks with benchmark.js against my (very) big stylesheet ;)