Fix miyêw:miý and phonological rule.

[eddieantonio]

See: UAlbertaALTLab/morphodict#929 (comment)

miyêw:miý VTA ;

[fbanados]

I believe this entry has already been incorporated into https://github.com/giellalt/lang-crk/blame/main/src/fst/morphology/stems/verb_stems.lexc#L7902. Blame to 04374ed

[fbanados]

@aarppe I believe we should close this.

[aarppe]

It may be that the phonological rules that would apply on y would not apply forý as it is not part of the glides that can contract. But the way that non-collapsing glides are denoted generally is by having a special character for that purpose, which we might want to implement for ý in this particular stem, i.e. ý2

[fbanados]

Ok, I was assuming that since the observable behaviour in UAlbertaALTLab/morphodict#929 has been already fixed by the change I previously mentioned, that we did not need to do anything else. If you believe there's extra linguist work here to be done we can keep the issue open.

[aarppe]

I updated the altlab2lexc.sh script in altlab/crk/bin/ to deal with coding of single-syllable stem-final ý->ý2 as one of the non-collapsing stem-final glides, which should match the morphophonological rules.

This should result in desirable behavior:

lang-crk % hfst-lookup -q src/fst/generator-gt-strict.hfst 

ýahki-miýêw+V+TA+Ind+1Sg+2SgO	kiyahki-miyitin	0.000000

What yet needs to be added is the optional y (<-) ý on the input side, which can be done as follows:

hfst[0]: load src/fst/generator-gt-strict.hfst 
warning: converting transducer type from foma to openfst-tropical when reading from file 'src/fst/generator-gt-strict.hfst'
? bytes. 90583 states, 160253 arcs, ? paths

hfst[1]: define GEN
Defined 'GEN'

hfst[0]: define yRule [ y (<-) ý ];
Defined 'yRule'

hfst[0]: regex yRule .o. GEN;
? bytes. 90795 states, 163357 arcs, ? paths

hfst[1]: down yahki-miyêw+V+TA+Ind+1Sg+2SgO
kiyahki-miyitin

hfst[1]: down ýahki-miýêw+V+TA+Ind+1Sg+2SgO
kiyahki-miyitin