And assumes that it is is run in the source repo. It makes
more sense to run it in the destination repository and specify
both the source repo and the commit ids in which it comes from.
Change the CLI argument parsing to use options rather than trying to
guess the intention from the number of CLI options presented. Also
add checking to ensure the options are presented correctly, add more
meaningful error messages and check that the source/destination
repositories are recognised as either kernel or libxfs repositories.
The resulting usage is:
$ libxfs-apply
Need to specify both source repo and commit id