tests: avoid implicit conversion of str to unicode
On Python 2, str.encode('ascii') implicitly converts the string to unicode and
then back to str.
Since the converted value is a path, _bytespath can be used instead.
Our full contribution guidelines are in our wiki, please see:
https://www.mercurial-scm.org/wiki/ContributingChanges
If you just want a checklist to follow, you can go straight to
https://www.mercurial-scm.org/wiki/ContributingChanges#Submission_checklist
If you can't run the entire testsuite for some reason (it can be
difficult on Windows), please at least run `contrib/check-code.py` on
any files you've modified and run `python contrib/check-commit` on any
commits you've made (for example, `python contrib/check-commit
273ce12ad8f1` will report some style violations on a very old commit).