You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
That can probably be done in script.sh, if I had to guess. If script.sh tells update.py that a repo contains certain trees/blobs, update.py won't know any different.
Hmm, yes, I think that's true. Will probably require overwriting quite a bit of stuff, get_type(), get_blob(), get_file(), get_dir(); and of course list_blobs() ...
Actually much more, but I guess things like tokenize_file(), parse_defs_*() and parse_docs() really could use get_blob() instead of open-coding git cat-file blob.
Someone related to the draft of #394. Zephyr does not use submodules but instead requires running a magical west command that pulls additional source code into the directory.
We are extending Elixir to let each project replace git fetch by its more complex behavior. Those steps will be responsible for generating tags that only contain Git blobs. Zephyr will be this pseudocode:
git fetch --tags
west ...
git commit -a ...
git tag resolved-...
Why do it that way? Because all the Elixir code assumes a Git repo is a list of blobs tied to filepaths. Supporting submodules would introduce too much complexity. It is much more straight-forward to generate a fake Git repo that only contains blobs.
tleb
added
the
indexing
Related to the index content — missing definitions/references, lexer bugs, new ctags features...
label
Feb 25, 2025
For a project that I'm considering using elixir for, submodule support would be rather necessary.
Again, not sure how it can be done, but leaving it here in case somebody has ideas.
The text was updated successfully, but these errors were encountered: