As I've tried to detect russian language, it failed. Removed serialized arrays and also removed _next_char PHP based implementation, as its not needed in JS.
Also added string cleaning before processing.
More work to be done :)
I've fixed tests in .js file only, as I do not see any need in CS here..
Converted serialized PHP arrays to JSON. Removed PHP based character …
Commited .json data files. Fixed test with trigram rank
Hmm, I've changed algorithm a bit, Travis build are failing, I'll into it later
Thanks! I'll merge this when all test pass (you can update the tests accordingly).
How about removing .coffee files? As for now project is on JS, and this files for tests just add complexity :) I think I'll support this project and will try to improve some parts, as they are from port from PHP and some things just could be done better )
Fixed failed tests
Yes, please remove .coffee files, that's fine with me (at that time I was testing coffee capabilities).
I just added you as a contributor on github, I will wait a little before giving you the npm right as well. Is that okay with you?
Hey, yes sure, this is nice :) I will do my best to improve it and I hope this becomes useful to more people :)