turning off sentence detection fails about ucto

[kosloot]

OK, not quit clear what the problem is.
The input file is already tokenized correctly.
If i run ucto on it with the -n -m options, then the produced output file is exactly the same as the input file. So what is the problem?

[proycon]

I use this quite often, but I don't have empty lines in my input files so not sure how ucto handles those. If there are none, then -n and -m should do the job already, the latter it implies -S.

[kosloot]

closing due to lack of response