RESOLVED FIXED 164240
[CSS Parser] Support the shadow DOM
https://bugs.webkit.org/show_bug.cgi?id=164240
Summary [CSS Parser] Support the shadow DOM
Dave Hyatt
Reported 2016-10-31 14:51:16 PDT
[CSS Parser] Support the shadow DOM
Attachments
Patch (14.40 KB, patch)
2016-10-31 14:55 PDT, Dave Hyatt
no flags
Patch (15.40 KB, patch)
2016-11-01 17:28 PDT, Dave Hyatt
dino: review+
Dave Hyatt
Comment 1 2016-10-31 14:55:50 PDT
Dave Hyatt
Comment 2 2016-10-31 15:11:20 PDT
Landed in r208180.
Ryan Haddad
Comment 3 2016-10-31 17:04:34 PDT
This change caused fast/shadow-dom/slotted-pseudo-element-css-text.html to crash on macOS and iOS on every test run: https://build.webkit.org/results/Apple%20Yosemite%20Release%20WK2%20(Tests)/r208180%20(19203)/results.html
Ryan Haddad
Comment 4 2016-10-31 17:04:55 PDT
Is there a reason this was landed before EWS had a chance to complete?
Ryan Haddad
Comment 5 2016-10-31 17:13:22 PDT
Reverted r208180 for reason: This change caused fast/shadow-dom/slotted-pseudo-element-css-text.html to crash on macOS and iOS on every test run. Committed r208198: <http://trac.webkit.org/changeset/208198>
Dave Hyatt
Comment 6 2016-11-01 17:28:02 PDT
Dave Hyatt
Comment 7 2016-11-01 18:02:58 PDT
Landed in r208267.
Ryosuke Niwa
Comment 8 2016-11-02 01:08:48 PDT
I'm not certain if it makes sense for us to be treating ::slotted(~) as an "implicit shadow combinator" because unlike ::-webkit-~ and other pseudo element that goes into a shadow tree, ::slotted goes out of the shadow tree to access a slot. That is, ::~webbkit- wants to match against the nodes inside a shadow tree from outside the shadow tree and ::slotted wants to match the nodes outside a shadow tree from inside the shadow tree. We might want to differentiate the two by having CSSSelector::ShadowDescendant and, let's say, CSSSelector::SlottedDescendant.
Note You need to log in before you can comment on or make changes to this bug.