lberki
lberki
That said, the `rdeps()` clauses could be executed in parallel, so on second thought, I'll reopen this issue and reframe it as "please make `bazel query` faster".
Probably neither? If we wanted to implement the above Buck feature, it would be much easier to do as a feature of `bazel query` instead of making it possible to...
I took a look at this and with the impending deprecation of WORKSPACE files and the fact that `.bazelignore` has always been a somewhat awkward interface (it's yet another configuration...
Good call. I forgot about the existence of `REPO.bazel`, which is just the perfect place for this functionality. Let me see if I can crank out the code today...
Ok, I give up (for today). This ended up being _way_ more involved than I had expected. I'll regroup and maybe try again tomorrow.
I have a change out that *almost* works. The "almost" part is because apparently requring the evaluation of `REPO.bazel` to determine the ignored package prefixes when the `WORKSPACE` file is...
@fmeum You are technically correct, but `REPO.bazel` does deal in Labels as parameters of `repo()` (e.g. `default_compatible_with=` and the like). So yes, that dependency edge could in theory be cut...
I don't know. I assumed that the reason why the main repo mapping is passed in is so that the labels reported on the debug output are consistent, no matter...
https://github.com/bazelbuild/bazel/pull/24022 it is.
(paging @Wyverald )