Are there any tools out there for introspecting a collection of Bazel build files to run queries against a codebase with? I'm thinking of a simple case of collecting all defined tags used in a codebase. Some sort of bazel metaquery sort of capability that will let me scope out the conventions and usages across a repo with a substantial amount of build files.
It would even be nice to be able to do a cross tabulation of cc_test and py_test rules against their collective tags. Ideally there'd be a python client to introspect the bazel files.
bazel query provides information about your target dependency graph, with a highly expressive query language. It can output into various formats like DOT, XML, Protobuf, and the text representation of the expanded BUILD files themselves (if there are macros) for post-processing. See: Bazel query how-to, Bazel query reference.
bazel cquery does the same as query, but also performs the analysis phase, which computes information about configurations (e.g. CPU, API levels) over the target dependency graph. This takes slightly longer, but gives you a more accurate representation of the graph that Bazel brings the execution phase. See: Bazel cquery reference.
bazel aquery is not directly related to BUILD file introspections in that it presents information about executable actions, which is a few layers of computation after BUILD file parsing and analysis. See: Bazel aquery reference
query, cquery and aquery don't operate on the syntax of the BUILD files. If you want to work with the Starlark syntax / AST, check out the buildozer and buildifier tooling in the bazelbuild/buildtools repository.
If there are information about your build graph that cannot be retrieved using these mechanisms, please file a feature request on the Bazel GitHub project.