refactor: use the proposed implementation for redirects #12
Conversation
Signed-off-by: Antonio Nuno Monteiro <[email protected]>
…caml#10288) Signed-off-by: Antonio Nuno Monteiro <[email protected]>
…ocaml#10286) * refactor: add labeled arguments to `Dep_rules.immediate_deps_of` Signed-off-by: Antonio Nuno Monteiro <[email protected]> * fix(melange): track immediate `.cmj` dependents as dependencies of JS rules Signed-off-by: Antonio Nuno Monteiro <[email protected]> * test(melange): show outdated paths for include_subdirs and melange.emit Signed-off-by: Antonio Nuno Monteiro <[email protected]> * fix(melange): track immediate `.cmj` dependents in `melange.emit` too Signed-off-by: Antonio Nuno Monteiro <[email protected]> * chore: add a changelog entry Signed-off-by: Antonio Nuno Monteiro <[email protected]> * fix: symlink `.impl.d` files for virtual library modules too Signed-off-by: Antonio Nuno Monteiro <[email protected]> --------- Signed-off-by: Antonio Nuno Monteiro <[email protected]>
Signed-off-by: Marek Kubica <[email protected]>
Signed-off-by: Rudi Grinberg <[email protected]>
Signed-off-by: Etienne Millon <[email protected]>
…s.t during emission (ocaml#10297) * test: show melange.emit regression attempting to read wrong ocamldep result Signed-off-by: Antonio Nuno Monteiro <[email protected]> * fix: read the processed file in Ocamldep.read_immediate_deps_of Signed-off-by: Antonio Nuno Monteiro <[email protected]> * fix(virtual_lib_compilation_test): only add dependency if impl exists Signed-off-by: Antonio Nuno Monteiro <[email protected]> --------- Signed-off-by: Antonio Nuno Monteiro <[email protected]>
Signed-off-by: Antonio Nuno Monteiro <[email protected]>
There was a problem hiding this comment.
Thanks so much for your help, I really appreciate it 🤗 I leave the PR opened so it's easier to review for @rgrinberg, but I would merge it otherwise.
I couldn't understand how to make the suggested
Idmodule API work
I wasn't sure about this either. I started trying to implement it but changing the returning value of find with an option led to multiple changes in the codebase (I prob was applying the change wrongly).
| in | ||
| if enabled | ||
| then Lib.DB.Resolve_result.redirect_in_the_same_db lib | ||
| else Lib.DB.Resolve_result.not_found |
There was a problem hiding this comment.
@rgrinberg mentioned that in this case we should redirect to the parent db (I can apply this change in a follow up commit)
| and enabled = lib.enabled_if in | ||
| { lib_name; enabled } | ||
| in | ||
| { loc; new_public_name; old_name; project = lib.project } |
There was a problem hiding this comment.
Just an idea for the future, but I wonder if old_name should be renamed to lib_info or target_info or something like that.
There was a problem hiding this comment.
I had it old_info in an earlier draft but then I changed it back to avoid polluting the diff.
Signed-off-by: Rudi Grinberg <[email protected]>
Signed-off-by: Antonio Nuno Monteiro <[email protected]>
anmonteiro
force-pushed
the
anmonteiro/try-public-libs
branch
from
c1d1c32 to
7434bcb
Compare
Signed-off-by: Javier Chávarri <[email protected]>
jchavarri
force-pushed
the
anmonteiro/try-public-libs
branch
from
fb00000 to
3eb7e79
Compare
Signed-off-by: Javier Chávarri <[email protected]>
This nests the pages too deeply and requires extra clicks. Instead we put this directly in the reference hierarchy and will highlight them in a separate section when using cards. Signed-off-by: Etienne Millon <[email protected]>
Signed-off-by: Javier Chávarri <[email protected]>
Signed-off-by: Javier Chávarri <[email protected]>
Signed-off-by: Javier Chávarri <[email protected]>
Signed-off-by: Javier Chávarri <[email protected]>
|
I have to stop midway through the bigarray fix. One thing I found out and that is strange is that while in |
|
Continues in ocaml#10307. |
Idmodule API work, though, and unless I'm missing something, it looks like we might not need it?cc @jchavarri @rgrinberg