• Remove the tokenizer’s pre-processing step.

    @SimonSapin
    SimonSapin committed Dec 23, 2014
    98c1e58