Instead of strictly reverting the sha1_lookup change, let's leave it in pack.c, that way we don't have two additional files in the tree for a single function. We can provide it internally for the library and name it git_pack__lookup_sha1 or something.

Instead of strictly reverting the sha1_lookup change, let's leave it in pack.c, that way we don't have two additional files in the tree for a single function. We can provide it internally for the library and name it git_pack__lookup_sha1 or something.

good point! done! (also took the opportunity to add a comment to the function and rename the arguments).

Sorry about the delay here. Let's move forward with this. Would you mind rebasing this on master so that we can get an up-to-date CI pass?

Sorry about the delay here. Let's move forward with this. Would you mind rebasing this on master so that we can get an up-to-date CI pass?

Done!

One thought that I have is that maybe this should be called git_midx instead of git_multipack_index. a) midx is significantly shorter 😀 but b) the multipack index is the atomic unit here, if I understand correctly. git_multipack_index makes it sound like there's an index on an atomic unit called a git_multipack. But there is not.

Since this is an internal-only type, I don't necessarily think that it's a hard and fast requirement to make this change, but I think that the update would be an improvement. What do you think @lhchavez?

One thought that I have is that maybe this should be called git_midx instead of git_multipack_index. a) midx is significantly shorter but b) the multipack index is the atomic unit here, if I understand correctly. git_multipack_index makes it sound like there's an index on an atomic unit called a git_multipack. But there is not.

Since this is an internal-only type, I don't necessarily think that it's a hard and fast requirement to make this change, but I think that the update would be an improvement. What do you think @lhchavez?

sounds reasonable. Done!

Cooooool. I'm excited to see this. I'm going to chip away at your other PRs, I apologize for making you wait so long in the review queue - I haven't had much spare time in the last few months.