[x] commit-and-push-repos.sh : Catalog.git_add_commit_push_all_repos, used as python -m astrocats supervovae --push
[x] repo-status.sh
[ ] pull-repos.sh : not sure how/if this should be done yet
[x] reset-repos.sh
[x] add tests for these methods
These methods are definitely going to take some trial and error to figure out the right way to do things. Especially resolving merge conflicts (e.g. do we need to pull before pushing, in general?).
For the moment, instead of a pull method, for example, I'm adding a reset-origin command which will effectively run git reset --hard origin/master. The reasoning is that we don't actually want to pull/merge the different sources, we just want to update to the remote copy.
Scripts:
Catalog.git_add_commit_push_all_repos
, used aspython -m astrocats supervovae --push
pull-repos.sh: not sure how/if this should be done yetThese methods are definitely going to take some trial and error to figure out the right way to do things. Especially resolving merge conflicts (e.g. do we need to pull before pushing, in general?).
For the moment, instead of a
pull
method, for example, I'm adding areset-origin
command which will effectively rungit reset --hard origin/master
. The reasoning is that we don't actually want to pull/merge the different sources, we just want to update to the remote copy.Test data repositories were created for the catalog test task:
catalog-test-output
andcatalog-test-input
. The method determining where entries are saved (_get_save_path
)[https://github.com/astrocatalogs/astrocats/blob/2ec101e9604a3a2e37b73055b79387a76b3e5fd8/astrocats/catalog/entry.py#L120] now requires that an output data repository exists before a file can be saved. Previously the method just returned theoutput/
directory itself.