Bug 99976

Summary: Rebaseline some sputnik unicode failed cases.
Product: WebKit Reporter: Kangil Han <kangil.han>
Component: WebKit EFLAssignee: Kangil Han <kangil.han>
Status: RESOLVED FIXED    
Severity: Normal CC: aroben, barraclough, dpranke, gyuyoung.kim, lucas.de.marchi, oliver, rakuco, roger_fong, webkit.review.bot
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Other   
OS: Linux   
Attachments:
Description Flags
patch
none
patch
none
patch
none
patch
none
patch none

Kangil Han
Reported 2012-10-22 03:27:08 PDT
Rebaseline expected files as other ports, e.g. mac, chromium and gtk+, have done.
Attachments
patch (3.24 KB, patch)
2012-10-22 03:30 PDT, Kangil Han 		no flags
DetailsFormatted DiffDiff
patch (13.01 KB, patch)
2012-10-22 04:21 PDT, Kangil Han 		no flags
DetailsFormatted DiffDiff
patch (10.83 KB, patch)
2012-10-22 05:25 PDT, Kangil Han 		no flags
DetailsFormatted DiffDiff
patch (1.71 MB, patch)
2012-10-22 17:01 PDT, Kangil Han 		no flags
DetailsFormatted DiffDiff
patch (10.85 KB, patch)
2012-10-24 18:07 PDT, Kangil Han 		no flags
DetailsFormatted DiffDiff
Show Obsolete (4) View All Add attachment proposed patch, testcase, etc.
Kangil Han
Comment 1 2012-10-22 03:30:07 PDT
Created attachment 169868 [details] patch
Kenneth Rohde Christiansen
Comment 2 2012-10-22 03:32:29 PDT
Comment on attachment 169868 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=169868&action=review > LayoutTests/platform/efl/sputnik/Unicode/Unicode_510/S7.6_A3.1-expected.txt:3 > +FAIL SputnikError: #0524-0527 So these are expeced to fail?
Kangil Han
Comment 3 2012-10-22 03:36:39 PDT
(In reply to comment #2) > So these are expeced to fail? It seems so since mac, chromium and gtk+ ports use same expected files.
Kangil Han
Comment 4 2012-10-22 04:21:02 PDT
Created attachment 169873 [details] patch As discussed on irc, I've removed expected results reside on each port and merged those to common place.
Raphael Kubo da Costa (:rakuco)
Comment 5 2012-10-22 05:10:37 PDT
chromium still maintains its port on SnowLeopard, so you should probably only remove the mac-snowleopard results.
Kangil Han
Comment 6 2012-10-22 05:25:37 PDT
Created attachment 169882 [details] patch Done. :)
Alexey Proskuryakov
Comment 7 2012-10-22 11:13:31 PDT
Perhaps we should just remove all those Unicode tests. Sputnik tests are obsolete, and these tests in particular seem to be more pain than they are worth.
Dirk Pranke
Comment 8 2012-10-22 16:52:49 PDT
should we just remove all of the sputnik tests? There are like 7000 of them ...
Dirk Pranke
Comment 9 2012-10-22 16:53:13 PDT
(oh, and Chromium skips them)
Kangil Han
Comment 10 2012-10-22 17:01:09 PDT
Created attachment 170030 [details] patch Done. :)
Kangil Han
Comment 11 2012-10-22 17:14:05 PDT
(In reply to comment #8) > should we just remove all of the sputnik tests? There are like 7000 of them ... Currently, just(?) sputnik/Unicode cases have been discussed. :)
Kangil Han
Comment 12 2012-10-23 17:24:55 PDT
@gbarra, @olliej : Would you please give me your opinion on this bug?
Gavin Barraclough
Comment 13 2012-10-24 01:04:23 PDT
Many of these test cases are valid and passing. Even those that fail are also guarding correct behaviour (reversion to ES3 behaviour would show as an unexpected pass). If we remove these without replacing them, we'll be reducing test coverage. It would make sense to me to remove these tests if we had a strategy to replace them with the updated (ES5 correct) tests from Test-262 - otherwise I'd be keen to not lose coverage.
Dirk Pranke
Comment 14 2012-10-24 12:04:09 PDT
Okay, I'm certainly happy to keep them if there's value.
Kangil Han
Comment 15 2012-10-24 18:07:41 PDT
Created attachment 170528 [details] patch Just rebaseline the expected result as gbarra suggested. :)
WebKit Review Bot
Comment 16 2012-10-24 20:08:12 PDT
Comment on attachment 170528 [details] patch Clearing flags on attachment: 170528 Committed r132442: <http://trac.webkit.org/changeset/132442>
WebKit Review Bot
Comment 17 2012-10-24 20:08:17 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.