Brendan O'Connor

Hm. Maybe the locale has to be set as a java flag or property? Or it would be better to use non-locale-dependent, floating point number parsing code. I have no...

It would be great to see a diff of tokenization under twokenize's current rules, versus what it is when using twitter-text's rules.

what's the intended input to runString()? is it intended for just one tweet? or many tweets? it looks like it only gets one tweet worth of output. that's not acceptable...

i'm not using it anyways, i just wanted it to be more robust and ready to be useful before merging it.

OK, that's weird. it should be copying those from the lib/ directory. I have no idea why or how maven works. It's always a mystery to me. Some people complained...

ah, the conll output format isn't very well tested i'm afraid...

Hm. Do you have other examples? Is it always with the double underscore? Please send us a pull request with your fix if you can. To test a fix to...

Yeah it's one of the regexes for sure... On Sun, Oct 21, 2012 at 11:10 PM, haijieg [email protected] wrote: > This is not the only bad example. Also I don't...

If you could assemble as many as you can find, it would be helpful. I'm trying to create minimal test cases and it's very subtle. (The problem is in non-determinancy...

actually no don't bother i think i see the problem (the eastern emoticon system)