* bzrlib/tests/__init__.py: (TestCase.run_bzr): Add a check to catch errors in error_regexes usage where people use a a string that it then iterated char by char for matching (which is most certainly not the intent of the test writer).
* bzrlib/tests/blackbox/test_log.py: Fix run_bzr_error calls. Fix some white spaces issues (some expected strings were using TABs and were wrongly caught by test_coding_style(), but fixing the strings wasn't really a problem either.