Regenerate generated code as part of the build

There are a number of cases where people don't realize that the schemas
are owned by Redfish, and that some of the generated files (registries,
indexes

Regenerate generated code as part of the build

There are a number of cases where people don't realize that the schemas
are owned by Redfish, and that some of the generated files (registries,
indexes) are in fact generated, and shouldn't be modified by hand.

This commit forces a regeneration of all those files within Jenkins,
each time we do a local build, then does a git diff to compare if the
upstream schema files and generated files are different than what is
checked in, it will print the diff immediately in the Jenkins build.
This should hopefully give much faster feedback to users that their
change needs to modify the right files.

As part of this commit, we also need to take the version string out of
clang-format. CI/the system should own the clang-format version.

Related commit:
https://gerrit.openbmc-project.xyz/c/openbmc/openbmc-build-scripts/+/51041]

Tested: Ran script, build passed. Removed the schema fixes from patch
series, saw build failed. Did the same with running the
openbmc-build-scripts unit test script.

Signed-off-by: Ed Tanous <edtanous@google.com>
Change-Id: I17eb0e106be267e8dbba7053b0b07ec617893c86

show more ...