Jani Nikula <jani at nikula.org> writes: > This is v2 of id:cover.1381948853.git.jani at nikula.org with more polish. > > Patches 1-4 do prep work to fix some of the differences in the parsers > in advance. Arguably they are not that bad regardless of the parser > change. > > Patches 5-6 actually make the change. Having two patches is a somewhat > artificial division, but perhaps makes it easier to review. >
I had a quick look at these changes, and nothing jumped out at me. I'd appreciate a second pair of eyes on them. I ran the performance suite, and there is only one message (in version 0.4 of the corpus) newly classified as non-mail. Of course I did clean up the corpus a bunch from 0.3 to 0.4. I didn't see any shocking changes in performance before and after the patches. I only had patience enough to run twice in both cases. d