Make --incompatible filing smarter about deps

[thomasvl]

rules_swift/rules_apple got a few issues filed for --incompatible-* things, but in starting to look at them a few of them seems to be because of things that the rules in turn depend on. It would be nice if the auto handling was smarter and realized when when there is a dependency with an issue, and included that info in the issue. Because until that dependency is fixed, the rules set can't update its dependency to avoid the problem.

[achew22]

+1. I just attempted to look into an issue for bazel-watcher but I'm not sure how to address an issue in @bazel_tools, isn't that a part of the main part of Bazel?

$ bazel build --incompatible_no_implicit_file_export //...
INFO: Invocation ID: 721ba4ac-8f79-4a7a-9fdb-9790952f809d
ERROR: /home/achew22/.cache/bazel/_bazel_achew22/378856566cc5ab2d8fea39cb4a335e93/external/bazel_tools/tools/test/CoverageOutputGenerator/java/com/google/devtools/coverageoutputgenerator/BUILD:4:1: in java_import rule @bazel_tools//tools/test/CoverageOutputGenerator/java/com/google/devtools/coverageoutputgenerator:all_lcov_merger_lib: target '@remote_coverage_tools//:all_lcov_merger_tools_deploy.jar' is not visible from target '@bazel_tools//tools/test/CoverageOutputGenerator/java/com/google/devtools/coverageoutputgenerator:all_lcov_merger_lib'. Check the visibility declaration of the former target if you think the dependency is legitimate
ERROR: Analysis of target '//bazel:go_default_test' failed; build aborted: Analysis of target '@bazel_tools//tools/test/CoverageOutputGenerator/java/com/google/devtools/coverageoutputgenerator:all_lcov_merger_lib' failed; build aborted
INFO: Elapsed time: 0.444s
INFO: 0 processes.
FAILED: Build did NOT complete successfully (0 packages loaded, 0 targets configured)

Would it be possible to compute a DAG of project dependencies and notify you only once your dependencies have finished migration on that front?

[drigz]

@achew22 FYI, I hit the same error and commented on the migration issue: bazelbuild/bazel#10225 (comment)

[ittaiz]

+1