It'd be kind of cool if re-running ripoff didn't run an upsert on every row, even if those upserts are relatively fast.
To accomplish this, we could store a map of row keys to hashes, and if the hash in the ripoff file is the same as the lockfile then don't add that upsert query.
It'd be kind of cool if re-running
ripoffdidn't run an upsert on every row, even if those upserts are relatively fast.To accomplish this, we could store a map of row keys to hashes, and if the hash in the ripoff file is the same as the lockfile then don't add that upsert query.