[Webkit-unassigned] [Bug 17427] Line breaking opportunities at the end of a text node are missed

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=17427


Alex Henrie <alexhenrie24 at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #164327|review?                     |review-
               Flag|                            |




--- Comment #10 from Alex Henrie <alexhenrie24 at gmail.com>  2012-09-17 20:24:14 PST ---
(From update of attachment 164327)
Okay, a few notes. text-combine-image-test.html is the test with the string "20日". I looked at this test again and without any markup there is no line break, so the line break my patch causes must be incorrect. However, I wasn't able to investigate it further because I spent all my free time today trying to recompile WebKit and when I run the GTK launcher all I get is a segmentation fault. GDB reports:

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff4645a91 in JSC::ArrayConstructor::finishCreation(JSC::ExecState*, JSC::ArrayPrototype*) ()
   from /home/alex/webkit/WebKitBuild/Release/.libs/libjavascriptcoregtk-3.0.so.0

I used four characters because I thought that that was always going to be enough context to determine line breaking, and in everything I've tested, it is. However, reading UAX14 more closely it looks like Southeast Asian languages (such as Thai) require dictionary-based line-breaking algorithms, which would require more context. Do you know of a way to provide this context to the line-breaking algorithm without allocating a new string and copying the contents of the previous and current text nodes into it? Will Thai require more context than just the first previous text node? Any tips on how to fix the segfault?

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.