critpath.py: drop exception handling in `expand_dnf_critpath`
I don't think this is appropriate. If we hit an exception we
should just raise it. If we catch it, print something, and
continue, best case we generate an incomplete critpath, worst
case we generate an empty one. Why would we ever want to do
either of those?
Signed-off-by: Adam Williamson <awilliam@redhat.com>