archive: explicitly treat text we got from tarfile as encoded UTF-8 (Closes: MAT#9831).

Without this, I see test failures caused by:

  test_remove_bz2 (libtest.TestArchiveProcessing)
  Test MAT's ability to process .tar.bz2 ... /usr/lib/python2.7/tarfile.py:1993: UnicodeWarning: Unicode equal comparison failed to convert both arguments to Unicode - interpreting them as being unequal
    if self.name is not None and os.path.abspath(name) == self.name:

This change seems to make sense because we're creating our tarin and tarout
objects with encoding='utf-8' in the first place, so it's no big surprise that
what we get from it is UTF-8 -encoded text.

0 files changed, 0 insertions, 0 deletions