Hello please rebase your current codebase to the updated codebase as a lot of regex has been modified. Secondly lets modify this implementation of yours a little bit.

There is only one time when the full hash is required and it is when fetching the commit from remote.

What we should do instead is as follows, generally prefer the shortened hash (i think it is trimmed to like 7 characters) for everything, but when building from source explicitly ask for the full hash.

hi !

yup, i've seen the other PR getting merged :)

the 2 commits from my PR have been successfully rebased without any conflict.

as for your requests, that makes sense => i'll switch to short hashes then ;)

@MordechaiHadad, could you assign me to this PR, so that i can see the status changes with gh status (i'm working on a TTY-based workflow on my Raspberry Pi for fun !)

@antoineeestevaaan - A fellow terminal-only enjoyer; clearly 😄 . You should be able to watch PR check runs via gh pr checks --watch N (315 - N would be the PR number for whatever you're after).

I may perhaps be able to assist after working on/around the hash stuff in the last PR (Though not too extensively).
Given the amount of semantic meaning that some of the fields for the ParsedVersion have (tag_name & others like target_commitish); perhaps it would be worth moving them to smaller struct(s) and implementing FromStr/From.
Would make validation easier once it's got its own type (instead of stringly checking), then only leaves the more open-ended one for heavy checking and parsing paths.

Perhaps @MordechaiHadad you can weigh in on where/how the split between ParsedVersion and UpstreamVersion happens and if it would be worth simply consolidating under an enum or tuple struct with trait impl's to act as guide/states for data flow?

Ill read this tomorrow sorry, i was away on shabat.

@MordechaiHadad, just updated the PR !

mostly reverted all of it xD
then the patch is really simple !

please tell me if you like it more ;)

Hmm 5 commits but only 1 line of code changed? lol

these can be squashed if you want.

do i need to change something else?

these can be squashed if you want.

do i need to change something else?

Yeah nothing else, but squashing. This is just to handle hashes in the ls command correct?

Yeah nothing else, but squashing.

roger that, i'll do that soon.

This is just to handle hashes in the ls command correct?

yes, this PR only changes the bob ls command by less strictly checking the currently used version.

Alright waiting for squashing and then ill merge

@MordechaiHadad, i think this is good to go 😉