The script for finding order dependency test failures
was dependent on inputs from Jenkins logs which we no
longer use.
This quick fix was to just use a list of tests with the
right format. A future iteration might process a new type
of output, like that from (pytest -v).
