Rewritten using the tokenize module, which gives us a real tokenizer
rather than a number of approximating regular expressions. Alas, it is 3-4 times slower. Let that be a challenge for the tokenize module.
Showing
This diff is collapsed.
Please register or sign in to comment