Description of problem: Rendering problem with [Consonant] + u0d4d+0d31 Character should be "Un"Join Version-Release number of selected component (if applicable): libicu-3.8-5.fc9 How reproducible: Everytime Steps to Reproduce: 1. open oowriter 2. type 0d15 (any consonant except 0d31] + 0d4d+0d31 3. Actual results: it combined Expected results: should not combined to make a glyph (like QT or Pango) Additional info: 1) QT/Pango working fine 2) Screenshot
Created attachment 288581 [details] oowriter with kedit and gedit
i.e. U+0D15 U+0D4D U+0D31 yes ? But that's test case 153 at http://batman.brisbane.redhat.com/~indic/IndicTC/lang/ml_IN/font/IndicFontTestCaseGSUB-Malayalam.html which looks just like the OOo rendering. Now I'm totally confused. How is "working" and "not working" being determined. If there is something wrong here can I just e.g. get the above testcase expanded for the new thing that doesn't work, or a pointer like "test case #102 from IndicFontTestCaseGSUB-Malayalam.html" doesn't work.
Those Test Cases Need to update (as per Confirmed with Malayalam Translator) (which our Team will update soon). (#153-179 Test Case need to update) regards A S Alam
Just let me know when we have a new testsuite I can work towards. We changed the original behaviour which gave the results being asked for here to match those test-cases which are now apparently not what is desired.
I'll set this to NEEDINFO because I can't really work on undoing the change we did to make the test-cases correct for this new requirement unless I can see what the new test-cases should be to see if we need to remove the original change totally or modify it for some sub-cases.
So all the 0d31 test-cases are now the other way around. Reverted the hack that we had in place in icu to get the original results. So should be good in icu-3.8.1-4.fc9. The same hack for 0d30 test-cases remains.
Hi Caolan, Working perfect in icu-3.8.1-4.fc9. Tested and confirmed. Thanks Ani