jumpinjackie
commented
5 years ago We might need a blob id computation step to compute all the possible sha1s for these extraneous docs
https://stackoverflow.com/questions/21142986/remove-filenames-from-specific-path
According to the current imported git repo, 44% of the "source" is HTML. When you break this down, it is mostly API documentation our internal copy of dbxml.
Since we're already cleaning out stuff, we should be omitting these docs as well as based on analytics they contribute a whole bunch of pointless bloat.