- refactor: move import_journal to hyperdb .-- it was reimplemented in back_anydbm and rdbms_common and was already inconsistent -- and it doen't have any backend dependencies itself. Interestingly this now fails a test for memorydb (but passes for anydbm: Huh?)

#!/usr/bin/env python
import zlib, base64, sys
print base64.encodestring(zlib.compress(sys.stdin.read()))