Created attachment 76920 [details] test case Run the attached test case to see the crash.
Created attachment 76921 [details] Patch
Comment on attachment 76921 [details] Patch Do we do the right thing (i.e. remove blockquote) with your patch? Although if the blockquote is inline, then the text isn't really indented so outdenting doesn't really have any visible effect. Also I'd like to see a test case where ul/ol and li are inline.
(In reply to comment #2) > (From update of attachment 76921 [details]) > Do we do the right thing (i.e. remove blockquote) with your patch? Although if the blockquote is inline, then the text isn't really indented so outdenting doesn't really have any visible effect. Also I'd like to see a test case where ul/ol and li are inline. I believe the answer is yes. My change is meant to take into account the fact that startOfEnclosingBlock doesn't handle properly a blockquote with inline style and returns a node outside the blockquote. What we want here, is to be able to fall into the first if statement (case in which the block quote is in its own paragraph). The other scenario you mention, is handled by the code after the if statement where we do the splitting. The ordered/unordered list case is handled by completely different code.
Comment on attachment 76921 [details] Patch We should add more test cases as Ryosuke suggests.
Added additional tests, per Ryosuke's comments. Committed revision 74369.
(In reply to comment #5) > Added additional tests, per Ryosuke's comments. > Committed revision 74369. Thanks for the test case.
http://trac.webkit.org/changeset/74369 might have broken GTK Linux 64-bit Debug The following tests are not passing: inspector/extensions-audits.html