lines 72 - 88: new method for conversion of byte stings into Unicode strings using "utf-8" encoding. It should work in Python v2 also (do nothing, returning the initial object), but in Python v3 it performs a real conversion.
lines 110 and 122: using this method (otherwise the Python v3 produces an exception in these places).
line 124: fixing another bug: use a case-insensitive string search in Subject. The only google.com sends e-mails with line "Report domain:" in subject; most of other providers use the "Report Domain:" phrase ("Domain" from uppercase). So, only part of mailbox was processed.
3 lines near the end of file (old numbers 255 - 257) have been deleted as a duplicates (probably, just a misprint).
Proposed modifications (see issue #34):