| Summary: | Incorrect calculated width for mspace. | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Frédéric Wang (:fredw) <fred.wang> | ||||||||||
| Component: | MathML | Assignee: | zalan <zalan> | ||||||||||
| Status: | RESOLVED FIXED | ||||||||||||
| Severity: | Normal | CC: | cfleizach, commit-queue, dbarton, esprehn+autocc, glenn, mrobinson, slewis, zalan | ||||||||||
| Priority: | P2 | ||||||||||||
| Version: | 528+ (Nightly build) | ||||||||||||
| Hardware: | All | ||||||||||||
| OS: | All | ||||||||||||
| Bug Depends on: | |||||||||||||
| Bug Blocks: | 118812 | ||||||||||||
| Attachments: |
|
||||||||||||
Created attachment 206745 [details]
Patch
@Zalan I've cc'ed to two related bugs on MathML metrics. I think this one can be fixed independently but I just want to make you aware of Martin Robinson's refactoring of the MathML code that removes the "preferred heights" computation. (In reply to comment #2) > @Zalan I've cc'ed to two related bugs on MathML metrics. I think this one can be fixed independently but I just want to make you aware of Martin Robinson's refactoring of the MathML code that removes the "preferred heights" computation. good to know, thanks! In the test: <div> and </span> do not match. I'm not sure why you need to use a margin style. What about using <span style="display: inline-block; width: ...; height: ...; background: ...;"> as in the other mspace reftests? Also do you really need a platform-specific reference file for a reftest (see bug 118599)? (In reply to comment #4) > In the test: <div> and </span> do not match. I'm not sure why you need to use a margin style. What about using <span style="display: inline-block; width: ...; height: ...; background: ...;"> as in the other mspace reftests? Also do you really need a platform-specific reference file for a reftest (see bug 118599)? I'll look into those. I also thought i didnt need the platform specific ref file for the reftest, but since the test framework was complaining, I added. I'll check again and will remove it. Thanks. Created attachment 206772 [details]
Patch
Skipped Crashing tests in http://trac.webkit.org/changeset/152822 while waiting for a fix Comment on attachment 206772 [details]
Patch
OK
Just a note: if bug 118812 is taken before that one, then you won't need to add the platform-specific reference for fractions. Of course that reference can also just be removed later. (In reply to comment #9) > Just a note: if bug 118812 is taken before that one, then you won't need to add the platform-specific reference for fractions. Of course that reference can also just be removed later. I'll just commit this and bug 118812 will take care of the cleanup of the platform specific files. Created attachment 206963 [details]
Patch for landing
(In reply to comment #10) > I'll just commit this and bug 118812 will take care of the cleanup of the platform specific files. ok, sounds good to me. Comment on attachment 206963 [details] Patch for landing Clearing flags on attachment: 206963 Committed r152840: <http://trac.webkit.org/changeset/152840> All reviewed patches have been landed. Closing bug. |
Created attachment 206536 [details] testcase See the testcase. I guess RenderMathMLSpace::computePreferredLogicalWidths() should just set the specified width and should not call RenderMathMLBlock::computePreferredLogicalWidths()