This is a foma-based morphological analyzer for Hungarian, which I developed under the instruction of Dr. Mans Hulden during the 2016 San Sebastián (Donostia) NLP summer school "Codefest."
The morphological analyzer consists of two parts: a lexicon (HUN.lexc) and a foma script (HUN.foma), the latter including a number of alternation rules (or rewrite rules) and the subgrammar composition. Thanks to the bidirectional feature of finite-state networks, this linguistic tool is simultaneously a word analyzer (i.e., it can analyze an inflected Hungarian word into its morphological components) and a word generator (i.e., it can convert an abstract morphological representation into an orthographically correct Hungarian word).
For example:
(1) Word analysis: ablakon => ablak[N][SG][SUP] (i.e., ablakon is the superessive case form of the singular noun ablak 'window', which literally means 'on [the] window')
(2) Word generation: fekete[ADJ][PL] => feketék (i.e., the plural form of the adjective fekete 'black' is feketék)
To run the morphological analyzer, you need to install foma, a finite-state toolkit designed for natural language processing developed by Dr. Mans Hulden. For installation instruction see its GitHub repo.
After installing foma, simply enter the foma environment in the terminal and run source HUN.foma (within the directory where you have downloaded the foma scripts) to compile. When the compilation is completed, either enter the up mode and do word analyses or enter the down mode and do word generation.
The morphological analyzer in its current state has a basic lexicon of 9654 nouns, 2238 verbs, and 1637 adjectives. It may very well make occasional mistakes as the summer school was only five days and I have not had time to further work on the project since then. But I am likely to return to it in the future, so any suggestions are welcome! Köszönöm szépen!