Red Hat Bugzilla – Bug 318071
[kn_IN] consonant + Halant + Consonant is wrongly rendering
Last modified: 2015-04-21 00:34:17 EDT
Description of problem: In almost of all the cases consonant+halant+consanat is
Version-Release number of selected component (if applicable):
How reproducible: Every time
Steps to Reproduce:
1. Open gedit (Or any writer)
2. Activate SCIM
3. Select SCIM-inscript
4. Use the inscript key sequence as shown in the attachement
Actual results: Shown in the attachment
Expected results: Shown in the attachment
Additional info: This was solved for some cases via bug #228976, in which the
second example of the attached image depicts the similar issue and has been
solved. The same needs to be applied for all the combination.
Created attachment 215671 [details]
Few examples of the consonant+halant+consonant
Presumably this is not fixed yet in F8?
No, the problem still persists in the f8 too.
This is not possible in the current opentype framework, also it will be, if
possible, fixed in Layout engine first.
A similar issue exists in Lohit Telugu fonts as well.
I think the attachment 215671 [details] shows the following combinations:
1. U+0C97 U+0CCD U+0CB0 U+0CCD
2. U+0CAA U+0CCD U+0CB5 U+0CCD
3. U+0CAC U+0CCD U+0C97 U+0CCD
Analogously Telugu combinations can be formed which are rendered incorrectly by Lohit Telugu along the same lines:
1. U+0C17 U+0C4D U+0C30 U+0C4D
2. U+0C2A U+0C4D U+0C35 U+0C4D
3. U+0C2C U+0C4D U+0C17 U+0C4D
(subtracting 0x80 from each character)
However there is another OpenType font that is able to render these Telugu combinations correctly -- Pothana2000 from http://www.kavya-nandanam.com/dload.htm (Googling shows A few other Linux distros are currently shipping this font).
Will it be possible to use whatever technique that font uses, to fix the issues in Lohit Kannada and Lohit Telugu? Or does "current opentype framework" in comment 4 mean a restricted framework that precludes fonts like Pothana2000?
Bug 227971 is the same issue. Looks like only some specific combinations which are shown in the attachments to that bug have been fixed (LA+halant+LA+halant, DA+halant+DA+halant, JA+halant+NYA+halant, KA+halant+SSA+halant). May be other combinations could be fixed similarly.
However it looks like that fix defined separate conjunct glyphs for L+LA, D+DA, J+NYA and K+SSA as well as these followed by various vowel signs which brought about other side effects in rendering consonant+halant+consonant+vowel sign (except vowel signs U, UU and AU). For example, in KA+halant+SSA+sign O the SSA is rendered below KA rather than the "O sign" unlike other combinations of consonant+halant+consonant+vowel sign. I am not sure which is more acceptable in Kannada.