Hi, I am new to use kythe and have some questions about using kythe with bazel.
1. Is there anyway to do the whole process from extraction to build the serving_table using bazel? So far I just noticed the extraction can use bazel, and the rest of jobs was done by script. This is fine, just want to see any better solution:)
2. I have a bazel repo with go and proto. And I tried to build the cross_reference for it. The command I used is like:
2.1 extraction
bazel --bazelrc=$KYTHE_DIR/extractors.bazelrc \
build --override_repository kythe_release=$KYTHE_DIR \
//...
2.2 indexing
find -L bazel-out -name '*.go.kzip' | sort > units
time cat units | parallel -t -L1 /opt/kythe/indexers/go_indexer | /opt/kythe/tools/dedup_stream >> entries
2.3 graphstore and serving table
/opt/kythe/tools/write_entries --graphstore leveldb:/tmp/gs < entries
/opt/kythe/tools/write_tables --graphstore /tmp/gs --out /tmp/ts
So far, everything can run, but all of the import were missing. errors like that.
I have no idea what's going on, I tried a go project without bazel and it works fine. Is there any special concern when we using bazel for our system?
3. another question is is there any better way to index a repo with multiple language? Suppose we have C++, Java, Go, and proto. Is there a way to using one script to do all the thing?