GNOME Bugzilla – Bug 569180
Kannada nukta (U+0CBC) support
Last modified: 2009-10-01 14:43:08 UTC
In http://svn.gnome.org/svn/pango/trunk/modules/indic/indic-ot-class-tables.c the character class of U+0CBC is set to _xx, which is different from the way nukta is handled for other scripts where it is set to _nu. Was there a conscious decision taken to handle it this way?
Created attachment 127250 [details] Rendering of "za zaa zi zu zuu ze zau z fa faa fi fu fuu fe fau f zra fra" Attached is text as rendered by the vanilla Pango version from Mandriva Linux's pango-1.10.0-3mdk using Red Hat's Lohit Kannada font from lohit-fonts-2.3.1-1.fc10. The top 2 rows use the character order "consonant nukta matra/halanth" as is done in Devanagari for "za zaa ... z fa faa ... f." The next row uses "consonant nukta halanth ra" to attempt to get a conjunct consonant. The bottom 3 rows use the order "consonant matra/halanth nukta" instead. Note however that chapter 9 of The Unicode Standard says for a dead consonant the nukta should precede the halanth. I do not know Kannada but it is perhaps obvious that in this case "za fa zi fi ze fe" are the only characters rendered pleasingly (fourth & fifth rows). Looking at the character "fa" with the two dots of the nukta on both sides of the horizontal stroke, I doubt putting nukta below the vowel matra instead of below the consonant for "zaa zu zuu zau z faa fu fuu fau f" is correct.
Created attachment 127251 [details] Rendering with character class = _nu Attached is text as rendered using Pango compiled from pango-1.10.0-3mdk.src.rpm with the character class of U+0CBC set to _nu using Red Hat's Lohit Kannada font from lohit-fonts-2.3.1-1.fc10. I do not know Kannada but using the logic that the nukta should be below the consonant in all cases, I think in this case "za fa zu zuu fu fuu zra fra" are the only characters rendered pleasingly (top 2 rows). The matras and halanths in "zaa zi ze zau z faa fi fe fau f" do not ligate but I think this could be handled in the font just as bug 516947. Bug 516947 exists only in Red Hat's Lohit Telugu font but is handled by the Pothana2000 font from http://www.kavya-nandanam.com/dload.htm by declaring the OTL classes of all below-base forms as "mark" and by specifying that consonant+halanth ligation ignore these marks. See comments 4 & 6 in https://bugzilla.redhat.com/show_bug.cgi?id=476427#c4. Similarly the nukta can be declared as a "mark" and consonant+matra ligation rules can specify ignoring any marks. Of course, if the correct way to present the nukta is different in Kannada, it might need to be handled differently.
Thanks Padmanabhan for your bug report. I will include this bug fix in patch to 594101 bug.
*** This bug has been marked as a duplicate of bug 593904 ***