Bug 122837

Summary: [MathML] Poor spacing around delimiters in MathML Torture Test 14
Product: WebKit Reporter: Martin Robinson <mrobinson>
Component: MathMLAssignee: Martin Robinson <mrobinson>
Status: RESOLVED FIXED    
Severity: Normal CC: bfulgham, buildbot, cfleizach, commit-queue, davidc, dbarton, esprehn+autocc, fred.wang, glenn, kondapallykalyan, rniwa, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
URL: https://developer.mozilla.org/en-US/docs/Mozilla/MathML_Project/MathML_Torture_Test
Bug Depends on: 123820    
Bug Blocks: 123543    
Attachments:
Description Flags
Patch
none
Adjust the test to use &Verbar; instead of &DoubleVerticalLine;
none
Archive of layout-test-results from webkit-ews-13 for mac-mountainlion-wk2
none
Archive of layout-test-results from webkit-ews-05 for mac-mountainlion
none
Patch with Mac results from archive
none
Uploading patch again to get new Mac results
buildbot: commit-queue-
Archive of layout-test-results from webkit-ews-13 for mac-mountainlion-wk2 none

Description Martin Robinson 2013-10-15 09:24:18 PDT
In the attached test case and number 14 on the MathML torture test, the delimiters have too much spacing around them.
Comment 1 Frédéric Wang (:fredw) 2013-10-15 09:57:46 PDT
Adding bug 99620, since having proper operator spacing will help here.

I suspect we will end up doing something similar as Gecko to get accurate ink/logical metrics and do width correction.
Comment 2 Martin Robinson 2013-10-30 16:12:10 PDT
It looks like the issue here is very specific to the delimiter. In STIX, the stretchable delimiter is thicker with a very large left and right bearing. For us, this means that when we consider the stretchable delimiter as a possibility for drawing an operator we will always create a large amount of horizontal space for that operator. I think the fix here is to copy Gecko and ignore the stretchable delimiter.
Comment 3 Martin Robinson 2013-10-31 15:43:35 PDT
Created attachment 215684 [details]
Patch
Comment 4 Martin Robinson 2013-10-31 16:10:00 PDT
Created attachment 215688 [details]
Adjust the test to use &Verbar; instead of &DoubleVerticalLine;
Comment 5 Martin Robinson 2013-10-31 16:35:34 PDT
Bug 123543 also changes the mo-stretch test so it might makes sense to create Mac baselines when both these patches have landed.
Comment 6 Build Bot 2013-10-31 17:02:04 PDT
Comment on attachment 215688 [details]
Adjust the test to use &Verbar; instead of &DoubleVerticalLine;

Attachment 215688 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/19298017

New failing tests:
mathml/presentation/mo-stretch.html
Comment 7 Build Bot 2013-10-31 17:02:06 PDT
Created attachment 215693 [details]
Archive of layout-test-results from webkit-ews-13 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-13  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5
Comment 8 Build Bot 2013-10-31 19:01:35 PDT
Comment on attachment 215688 [details]
Adjust the test to use &Verbar; instead of &DoubleVerticalLine;

Attachment 215688 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/19358066

New failing tests:
mathml/presentation/mo-stretch.html
Comment 9 Build Bot 2013-10-31 19:01:37 PDT
Created attachment 215711 [details]
Archive of layout-test-results from webkit-ews-05 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-05  Port: mac-mountainlion  Platform: Mac OS X 10.8.5
Comment 10 Martin Robinson 2013-11-01 09:47:16 PDT
Tests fail for Mac because they need a rebaseline.
Comment 11 chris fleizach 2013-11-04 09:48:49 PST
Martin can you grab the results from the layout-test-results in the attached archives and update your patch again?
Comment 12 Martin Robinson 2013-11-04 10:46:15 PST
Created attachment 215930 [details]
Patch with Mac results from archive
Comment 13 Martin Robinson 2013-11-04 12:42:27 PST
Looks like tests are passing now. :)
Comment 14 Brent Fulgham 2013-11-05 09:43:22 PST
Comment on attachment 215930 [details]
Patch with Mac results from archive

r=me
Comment 15 Radar WebKit Bug Importer 2013-11-05 09:46:34 PST
<rdar://problem/15394108>
Comment 16 Martin Robinson 2013-11-05 10:04:13 PST
Committed r158678: <http://trac.webkit.org/changeset/158678>
Comment 17 WebKit Commit Bot 2013-11-05 13:11:13 PST
Re-opened since this is blocked by bug 123820
Comment 18 Martin Robinson 2013-11-09 18:02:46 PST
Created attachment 216503 [details]
Uploading patch again to get new Mac results
Comment 19 Build Bot 2013-11-09 19:10:09 PST
Comment on attachment 216503 [details]
Uploading patch again to get new Mac results

Attachment 216503 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/22638963

New failing tests:
mathml/presentation/mo-stretch.html
Comment 20 Build Bot 2013-11-09 19:10:14 PST
Created attachment 216504 [details]
Archive of layout-test-results from webkit-ews-13 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-13  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5
Comment 21 Martin Robinson 2013-11-09 19:37:49 PST
Committed r159007: <http://trac.webkit.org/changeset/159007>