dultik

this is a (still very bad) xextan parser. it returns all valid parses.

"why is it very bad" i hear you ask.

• it is not uptodate with current xextan (though neither is the PEG its based on)
• also it doesnt keep track of the parse structure yet which is kind of silly

good news is now there is a tokenizer which makes it not ridiculously slow. yay

what to do

1. clone
2. cd dultik/src
3. either
   • ghci ../app/Main.hs and do something like test "xoi len sin jo tak xextan"
   • edit app/Main.hs to parse whatever you want and then compile+run with ghc -O2 ../app/Main.hs && ../app/Main