* benim/bizim is not analyzed as genitive
* auto generate analyzer/generator for segmentation
* -ki added to (time) adverbs is not properly handled
* define analysis/surface/intermediate symbols clearly
  --> phonological part may not need all the symbols
* better analysis of question particles
* n/y change in Yusufeli'ne/Yusufeli'ye (some of these are already marked
  in the lexicon.)
* mark "already plural" lexical items in the lexicon
* numeric forms currently do not inflect 10(')da
* proper handling of proper nouns (apostrophe, no consonant changes)

* expressions like yarın-ki dün-kü, bu yıl-ki
* -de followed by other morphemes: gide debilirim
* derivation attached to inflected words
  bizimki-leş-me 
  (http://sozluk.sourtimes.org/show.asp?t=akp+nin+kuresel+ekonomik+kriz+icin+aldigi+onlemler)
  bizden-leş-me 
  (http://www.kafkasya.net/site/modules.php?name=Forums&file=viewtopic&p=2535)
  kaçınılmazlığı (metu corpus)
  (oku-ma-mış-lık, biz-den-leş-miş-lik, 
     duyarsızlaşmışlık)
* better/wider coverage of derivational morphemes.
* Review and convert all analysis strings to correct/principled/detailed 
  descriptions.

DONE:

* can reflexive and reciprocal co-occur?
* bana -> ben + a ?
* question suffix
* s-drop: camii
* gemination: üssu, hakkı ...
* Doble copula cases: gidiyor-muş-sa, gidiyor-du-ysa (also gidiyor-sa-ydı)
* 3p agrement, -lar, variation. 
* compound noun forms: ağustosböcegi, ağustosböcekleri
* `denildi': dobule passive or??
