WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
13675
Focus rings don't repaint properly on editable continuations
https://bugs.webkit.org/show_bug.cgi?id=13675
Summary
Focus rings don't repaint properly on editable continuations
Dave Hyatt
Reported
2007-05-11 01:48:03 PDT
Reduction forthcoming.
Attachments
Demo of the problem... just start typing and hitting enter.
(91 bytes, text/html)
2007-05-11 01:48 PDT
,
Dave Hyatt
no flags
Details
Patch to improve painting
(8.80 KB, patch)
2007-05-11 01:51 PDT
,
Dave Hyatt
no flags
Details
Formatted Diff
Diff
Patch with layout test
(110.10 KB, patch)
2007-05-11 02:05 PDT
,
Dave Hyatt
oliver
: review+
Details
Formatted Diff
Diff
Show Obsolete
(1)
View All
Add attachment
proposed patch, testcase, etc.
Dave Hyatt
Comment 1
2007-05-11 01:48:35 PDT
Created
attachment 14483
[details]
Demo of the problem... just start typing and hitting enter.
Dave Hyatt
Comment 2
2007-05-11 01:51:06 PDT
Created
attachment 14484
[details]
Patch to improve painting This patch makes continuation painting be done by the containing block of the entire continuation run. If any component of the continuation needs to paint an outline, it will schedule the paint with the enclosing block who will then paint it. This patch preserves the old paint path when a layer is present. I also patched absoluteOutlineBox and absoluteClippedOverflowRect so that blocks in the continuation will assume they have outlines if the inline that generated them does.
Dave Hyatt
Comment 3
2007-05-11 02:05:28 PDT
Created
attachment 14485
[details]
Patch with layout test Added a layer paint test. I will also turn the attachment into an editing layout test.
mitz
Comment 4
2007-05-11 02:17:55 PDT
+<p><span>Content before<h3>Middle</h3>No content after.</span></p> "No content after." is... content ;-)
Dave Hyatt
Comment 5
2007-05-11 02:24:44 PDT
Details.... :)
Oliver Hunt
Comment 6
2007-05-11 02:27:14 PDT
Comment on
attachment 14485
[details]
Patch with layout test + if (table->isEmpty()) + return; + + RenderFlowSequencedSet* continuations = table->get(this); + if (!continuations) + return; is the isEmpty() check necessary given we also check the result of get? Looks good, i'd like you to run the pixel tests though, and (you'll hate me for this) manually check any pixel tests that fail
Dave Hyatt
Comment 7
2007-05-11 02:58:27 PDT
Fixed in
r21397
.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug