RESOLVED FIXED 89081
[Shadow] Assertion in VisibleSelection::adjustSelectionToAvoidCrossingBoundaries() is triggered 
https://bugs.webkit.org/show_bug.cgi?id=89081
Summary [Shadow] Assertion in VisibleSelection::adjustSelectionToAvoidCrossingBoundar...
Shinya Kawanaka
Reported 2012-06-14 02:14:16 PDT
This is related to Bug 88500, but another case. Select from "fuga 4" to "hoge 1". This will trigger ASSERT_NOT_REACHED().
Attachments
Repro (3.76 KB, text/html)
2012-06-14 02:19 PDT, Shinya Kawanaka 		no flags
Details
Patch (6.63 KB, patch)
2012-06-21 15:05 PDT, Shinya Kawanaka 		no flags
DetailsFormatted DiffDiff
Patch for landing (6.71 KB, patch)
2012-06-21 15:42 PDT, Shinya Kawanaka 		no flags
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Shinya Kawanaka
Comment 1 2012-06-14 02:19:46 PDT
Created attachment 147526 [details] Repro
Shinya Kawanaka
Comment 2 2012-06-20 10:56:21 PDT
Non-content-editable <table> became content-editable here again... We should fix that problem before this. I believe that problem causes this problem.
Shinya Kawanaka
Comment 3 2012-06-20 10:57:02 PDT
See also Bug 89077.
Shinya Kawanaka
Comment 4 2012-06-20 17:41:49 PDT
> Non-content-editable <table> became content-editable here again... According to the spec, <table> should be contenteditable... So this sentent is wrong, I'm sorry.
Shinya Kawanaka
Comment 5 2012-06-21 10:41:39 PDT
(In reply to comment #4) > > Non-content-editable <table> became content-editable here again... > > According to the spec, <table> should be contenteditable... > So this sentent is wrong, I'm sorry. NO... <table> in Shadow DOM should not be contenteditable here. This is not the same as another cases. This is a real BUG. Let's investigate this more.
Shinya Kawanaka
Comment 6 2012-06-21 11:45:08 PDT
(In reply to comment #5) > (In reply to comment #4) > > > Non-content-editable <table> became content-editable here again... > > > > According to the spec, <table> should be contenteditable... > > So this sentent is wrong, I'm sorry. > > NO... <table> in Shadow DOM should not be contenteditable here. > This is not the same as another cases. This is a real BUG. > > Let's investigate this more. Oh sorry.... it's wrong... not a real bug... sorry for confusing...
Shinya Kawanaka
Comment 7 2012-06-21 15:05:30 PDT
Created attachment 148895 [details] Patch
Ryosuke Niwa
Comment 8 2012-06-21 15:13:19 PDT
Again, please cc darin & enrica on these bugs. Also there's no need to add [Editing] since this bug in the editing component.
Ryosuke Niwa
Comment 9 2012-06-21 15:17:08 PDT
Comment on attachment 148895 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=148895&action=review > Source/WebCore/editing/htmlediting.cpp:267 > + Why the blank line? I'd rather have blank lines in both functions or not have in either one. Doing different things in firstEditablePositionAfterPositionInRoot and lastEditablePositionBeforePositionInRoot is confusing. > LayoutTests/editing/shadow/adjusting-editing-boundary-with-table-in-shadow-expected.txt:1 > +PASS It's odd to see PASS before the description. > LayoutTests/editing/shadow/adjusting-editing-boundary-with-table-in-shadow.html:15 > +<p>Placing this sentence before table won't trigger an asertion.</p> > +<p>Selecting around a table which is distributed from shadow subtree to nested shadow subtree will trigger an assertion. To try manually, select from "shadow 2" to around "after" and confirm a crash does not occur.</p> Please place this description before the container.
Shinya Kawanaka
Comment 10 2012-06-21 15:42:01 PDT
Created attachment 148903 [details] Patch for landing
WebKit Review Bot
Comment 11 2012-06-21 19:02:56 PDT
Comment on attachment 148903 [details] Patch for landing Clearing flags on attachment: 148903 Committed r120991: <http://trac.webkit.org/changeset/120991>
WebKit Review Bot
Comment 12 2012-06-21 19:03:02 PDT
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.