Support for CSS selectors Level 3 and 4

[tadast]

Hi Annand,

deadweight breaks when CSS files have selectors like these:

a.button:not(:disabled):hover { ... }
@-webkit-keyframes appear { ... }

This patch fixes it.

[aanand]

Looks nice. Could you perhaps use the /x flag to split the big regex out onto multiple lines and document it line-by-line?

I feel like something that hairy should have some more test cases, too :)

On 17 Feb 2012, at 18:34, Tadas Tamosauskasreply@reply.github.com wrote:

Hi Annand,

deadweight breaks when CSS files have selectors like these:

a.button:not(:disabled):hover { ... } @-webkit-keyframes appear { ... }

This patch fixes it.

You can merge this Pull Request by running:

git pull https://github.com/tadast/deadweight master

Or you can view, comment on it, or merge it online at:

https://github.com/aanand/deadweight/pull/15

-- Commit Summary --

• Support more CSS3/CSS4 selectors

-- File Changes --

M lib/deadweight.rb (2) M test/deadweight_test.rb (6) M test/fixtures/index.html (1) M test/fixtures/style.css (8)

-- Patch Links --

https://github.com/aanand/deadweight/pull/15.patch https://github.com/aanand/deadweight/pull/15.diff

---

Reply to this email directly or view it on GitHub: https://github.com/aanand/deadweight/pull/15

[tadast]

I've updated my commit. I agree about hairiness, but I don't want to copy the expression to the comment and document it line by line, because it is not DRY. Also, one does not simply comment in the middle of the multi-line regex... So I just split it into multiple lines.

Regarding tests, I can write some isolated unit tests for the #split method if you want, but it wouldn't follow the current style. I also don't want to clutter the global test setup with different tests for each case, instead I've added a single worst edge case I could think of :)

[pboling]

Does this fix the problem with CSS3 substring matchers support?

[aanand]

@pboling: Doesn't look like it I'm afraid - @tadast's patch only strips the : suffix from selectors.

@tadast: After reading over this again, would there be any danger in simply matching against /:.*/?

[tadast]

@aanand I actually had the same though in the shower and forgot it later on. I can't think of a case where matching against /:.*/ would break anything. Well of course we still need to strip all @-webkit, and @media print, and friends. Should I go for it?

[aanand]

I say yes.

On 23 Feb 2012, at 17:23, Tadas Tamosauskasreply@reply.github.com wrote:

@aanand I actually had the same though in the shower and forgot it later on. I can't think of a case where matching against /:.*/ would break anything. Well of course we still need to strip all @-webkit, and @media print, and friends. Should I go for it?

---

Reply to this email directly or view it on GitHub: https://github.com/aanand/deadweight/pull/15#issuecomment-4140684