Closed
Bug 933044
Opened 11 years ago
Closed 11 years ago
AutoAncestorPusher should push style scopes regardless of whether we have an ancestor filter
Categories
(Core :: CSS Parsing and Computation, defect, P4)
Core
CSS Parsing and Computation
Tracking
()
RESOLVED
FIXED
mozilla28
People
(Reporter: heycam, Assigned: heycam)
Details
(Whiteboard: [qa-])
Attachments
(1 file)
17.56 KB,
patch
|
bzbarsky
:
review+
|
Details | Diff | Splinter Review |
There are some uses of AutoAncestorPusher, such as this one: https://hg.mozilla.org/mozilla-central/annotate/tip/layout/base/nsCSSFrameConstructor.cpp#l3569 which does the ancestor pushing only if there is an ancestor filter. But I think we still want to push the style scope element.
Assignee | ||
Updated•11 years ago
|
Summary: AutoAncestorPusher should be used regardless of whether we have an ancestor filter → AutoAncestorPusher should push style scopes regardless of whether we have an ancestor filter
Assignee | ||
Updated•11 years ago
|
Priority: -- → P3
Assignee | ||
Updated•11 years ago
|
Priority: P3 → P4
Assignee | ||
Comment 1•11 years ago
|
||
This makes AutoAncestorPusher no longer take a boolean that says whether to push or not. Instead, we call separate functions on the AutoAncestorPusher to either push the ancestor and the style scope, or only the style scope. If we don't want to push anything, we can either pass nullptr as the aElement (as we can now), or just not call Push* on the object. All the uses of AutoAncestorPusher have been updated to push the style scope element when we were going to skip pushing altogether, due to not having an ancestor filter on the TreeMatchContext.
Assignee: nobody → cam
Status: NEW → UNCONFIRMED
Ever confirmed: false
Attachment #825684 -
Flags: review?(bzbarsky)
Assignee | ||
Comment 2•11 years ago
|
||
Mats this might affect your display:contents patches.
Status: UNCONFIRMED → ASSIGNED
Ever confirmed: true
Comment 3•11 years ago
|
||
Comment on attachment 825684 [details] [diff] [review] patch r=me. I'm not sure what I think of the two-stage init thing, but I agree that trying to do all that in the ctor would just make the code unreadable...
Attachment #825684 -
Flags: review?(bzbarsky) → review+
Assignee | ||
Comment 4•11 years ago
|
||
https://tbpl.mozilla.org/?tree=Try&rev=2f08027f2376
Assignee | ||
Comment 5•11 years ago
|
||
https://hg.mozilla.org/integration/mozilla-inbound/rev/7d219bb72548
Comment 6•11 years ago
|
||
https://hg.mozilla.org/mozilla-central/rev/7d219bb72548
Status: ASSIGNED → RESOLVED
Closed: 11 years ago
Flags: in-testsuite?
Resolution: --- → FIXED
Target Milestone: --- → mozilla28
Updated•10 years ago
|
Whiteboard: [qa-]
You need to log in
before you can comment on or make changes to this bug.
Description
•