option to skip .copiedFiles in regression test

[thorade]

The file .copiedFiles.txt contains a list of all files that were copied to a library from the IBPSA library. Assuming the copied models work just fine, I sometimes would like to run only the tests that were added in the extending library, to save some time.
If that sounds reasonable, I can implement it and submit a Pull Request in ~3 weeks.

[mwetter]

Is this (optional setting I assume) saving much time? For Buildings, the majority of the testing time comes from examples/validations of Buildings that are not in IBPSA. Also, as tests are run on travis and this merge is not done that frequent, it is for us not an issue.

[thorade]

I would probably use the option only when testing on my own computer.
For us, it is ~350 tests from IBPSA and ~60 tests that we added.
Plus, we have a couple of examples that take very long and should also be excluded from testing:
UdK-VPT/BuildingSystems#78

[mwetter]

Then such an option makes sense for you. Just make sure the default run that checks all models is not changed.

[thorade]

There seems to be everything prepared for this feature: the function _includeFile is called for every mos script to test whether it should be included, it compares the mos script name against the global list self._excludeMos. I added a function that adds the copied files to that exclude list, but they are still included in the regression tests, so either my list format is wrong, or the function _includeFile does not work as I expected.

[thorade]

I changed the _includeFile function
https://github.com/lbl-srg/BuildingsPy/commits/issue143_excludeMos
and the scripts from the _excludeMos list now seem to be excluded.

It works well if I only add a few pathes to the file.

But when excluding all files from .copiedFiles the regression tests fail with KeyError: 'ResultFile'
Edit: I opened a new issue (#153) because that error seems to be independent of the excludeMos list thing.