citations ending in 'i', continue part 1

[funderburkjim]

This is analogous to #17.

Here, we extend the identification of stems ending in 'i' to those lexnorm-all2 cases where

• the citation ends in 'i'
• the lexnorm is exactly one of
  • f:n 3 cases
  • m:n 11 cases
  • m:f 122
  • m:f:n 1689

This handles about 1800 additional records of lexnorm-all2

After removing duplications, there are additional stems added:

• 1559 additional stems with m_i model
• 1820 additional stems with f_i model
• 1663 additional stems with n_i model.

for a total of 5042 additional stem-model cases ending in 'i'.

There remain 158 records of lexnorm-all2 with stems ending in 'i' which will be accounted for later. Some examples:

25903   Ayuzpati    Ayuz-pati   m:f#tnI:n
28537   i   i   LEXID=pron,STEM=idam
232753  saptati saptati LEXID=card,STEM=saptati

The declension of these additional 5042 cases requires no modification to the declension algorithms.

[gasyoun]

@funderburkjim Jim, you do remember https://github.com/drdhaval2785/SanskritSubanta

and

https://github.com/drdhaval2785/SanskritVerb, right?

[funderburkjim]

I had forgotten, or not known of, the SanskritSubanta. Thanks for mentioning.

Have worked with @drdhaval2785 in comparing his SanskritVerb verb forms to those generated by my code. This work done back in 2016. See https://github.com/funderburkjim/elispsanskrit/issues. Will no doubt revisit this work in this MWinflect project when I get to verb conjugations.

[drdhaval2785]

I will also need to revisit SanskritSubanta. It was merged with SanskritVerb at one point of time.

[gasyoun]

also need to revisit

Yes, indeed. MSS will not run away if you take some rest with us, will they?