635286 - ":not()" pseudo class in ":-moz-any()" selector doesn't work correctly

Status: VERIFIED FIXED

(Core :: CSS Parsing and Computation, defect, P1)

Description

[YUKI "Piro" Hiroshi]

Attachment: simple testcase

Build identifier: Mozilla/5.0 (Windows NT 6.0; rv:2.0b12pre) Gecko/20110218 Firefox/4.0b12pre

When an element <div class="before"> becomes to <div class="after"> by changing its class, following four CSS selectors don't match to the element anymore:

  /* 1. any element which has a class "before" */
  :-moz-any(.before) { background: red; }
  /* 2. any element which doesn't have a class "after" */
  :not(.after) { background: red; }
  /* 3. equals to the case 2 but wrapped by :-moz-any */
  :-moz-any(:not(.after)) { background: red; }
  /* 4. equals to the case 2 but the inner selector is wrapped by :-moz-any */
  :not(:-moz-any(.after)) { background: red; }

However, actually the case 3 seems to be still applied to <div class="after">.  The attached HTML is a simple testcase based on this scenario. All four lines should become green, but actually the third line is still red on my Minefield.

Moreover, when I inspected actually-applied style declarations to the third line by the DOM Inspector addon, then the line suddenly became to green, even if I did no change on the DOM. It is very odd behavior.

Comment 1

[YUKI "Piro" Hiroshi]

Attachment: mochitest version

mochitest version

Comment 2

[YUKI "Piro" Hiroshi]

I've verified that this problem appears on Windows 2000 with clean profile.

Build identifier: Mozilla/5.0 (Windows NT 5.0; rv:2.0b12pre) Gecko/20110218 Firefox/4.0b12pre

Comment 3

[Boris Zbarsky [:bzbarsky]]

Yeah, the recurrence through :-moz-any in AddSelector needs to look at mNegations too.

Patch coming up.

Comment 4

[Boris Zbarsky [:bzbarsky]]

Attachment: Proposed fix

Comment 5

[Boris Zbarsky [:bzbarsky]]

Attachment: diff -w of same

Comment 6

[Boris Zbarsky [:bzbarsky]]

Attachment: Less invasive patch, if desired

Comment 7

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Comment on attachment 513538 [details] [diff] [review]
Proposed fix

r+a2.0=dbaron

Comment 8

[Boris Zbarsky [:bzbarsky]]

Pushed http://hg.mozilla.org/mozilla-central/rev/66167f6eccce

SHIMODA Hiroshi, thank you very much for the bug report and test!

Comment 9

[YUKI "Piro" Hiroshi]

Verified on Mozilla/5.0 (Windows NT 6.0; rv:2.0b13pre) Gecko/20110225 Firefox/4.0b13pre

Thanks a lot!