[issue9124] Mailbox module should use binary I/O, not text I/O

[Steffen Daode Nurpmeso]

Steffen Daode Nurpmeso <sdaoden at googlemail.com> added the comment:

This message will not help anyone.
And it's not a chat, but because it seems i really made the horses gone
grazy (direct translation of a german proverb):
- msg127002: RDM, i didn't know all of that and i am really sorry.
  Now I am looking forward to do something good for montys,
  but the truth is: today i ain't able to (yet), because i am so god
  damned re-new to python(1) (16 days today including weekends)!
  My local fix *was indeed already* just using encoding=latin1, which,
  by the way, decreased CPU usage of parsing a 150KB archive mbox from
  4 minutes (!) to about 50 seconds (dumping is a tenth in both cases).
  And it did not fail due to conversion errors, too.
msg127078: haypo, thank you!  I couldn't have done all of this, being
  so far away from the great picture of python encoding internals.
msg127036: RH, "people like Steffen" do the camel-walk!
  *This* is slow!!
  But once i've realized that mailbox.py is a dead-end on Py3K
  (for shipping out of my really useless, brutally simple current
  S-Postman script thing), i would have been much happier if a simple
  sentence like "python3 uptake process in transition" would have
  warned me (in whatever RST output).
Thanks for reading this book in a bug tracker.

----------

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue9124>
_______________________________________