[smc-discuss] Permanent solution for ഖ്ര

[santhosh.thottingal at gmail.com]

Hi,

I was experimenting with the ഖ്ര issue and possibly all such cases like
പ്പ്ര ന്റ്ര etc.

Please read this carefully and let me know what do you think.

Add a prebase form lookup

+Lookup: 4 0 0 "'pref' Pre Base Forms in Malay+AQEA-lam lookup 8"
{"'pref' Pre Base Forms in Malay+AQEA-lam lookup 8-1"  } ['pref' ('DFLT'
<'dflt' > 'mlym' <'dflt' > )

let us take the case of മന്ത്രി case.

For ്ര I am removing

-Ligature2: "'pstf' Post Base Forms in Malay+AQEA-lam lookup 7 subtable"
r3 xx

psts for this case did not make any sense to me. As per my understanding
of OT spec it should be pres. Lohit Malayalam uses pres.

And adding

+Ligature2: "'pref' Pre Base Forms in Malay+AQEA-lam lookup 8-1" r3 xx


For ന്ത്ര:

Remove these existing rules
-Ligature2: "'pres' Pre Base Substitutions in Malay+AQEA-lam lookup 4
subtable" n1th1 r4
-Ligature2: "'pres' Pre Base Substitutions in Malay+AQEA-lam lookup 4
subtable" n1th1 xx r3

And add this
+Ligature2: "'pres' Pre Base Substitutions in Malay+AQEA-lam lookup 4
subtable" r4 n1th1

that means
ന്ത്ര <=  ്ര +  ന്ത  because this is what we get after reordering.

With this all ് ര cases fallbacks to correct shape with  ്ര in left side
and ന്ത്ര is rendered correctly.

Font used Meera(master) with harfbuzz-ng


Thanks
Santhosh