This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author giampaolo.rodola
Recipients belopolsky, eric.araujo, ezio.melotti, giampaolo.rodola, pitrou, stuaxo, tarek
Date 2010-05-09.19:25:43
SpamBayes Score 0.00010556052
Marked as misclassified No
Message-id <1273433145.21.0.463298069236.issue4972@psf.upfronthosting.co.za>
In-reply-to
Content
> Magic methods usually don’t have docstrings, since they implement one 
> operation that is described in one place (think __len__ vs. len).

Agreed, I only left it since it was there in the first place.

> What did you add a cmd_noop method that does the same thing as
> cmd_user for?

Basically all DummyFTPHandler commands are no-ops except pasv, port, retr, stor and some others. There's no real reason why I added NOOP except for consistency with the FTP procol.
What I wanted to do in the new tests was just sending a command and receive a response and the "correct" way to do that in FTP is using NOOP.

> Will the change from handler to handler_instance not break 
> third-party code?

Not sure what you mean by "third party code" but it shouldn't break the existing tests if that's what you're worried about.
I did that because the original server behavior was to accept only one connection and I needed to connect more than one client in the same test.
History
Date User Action Args
2010-05-09 19:25:45giampaolo.rodolasetrecipients: + giampaolo.rodola, belopolsky, pitrou, tarek, ezio.melotti, eric.araujo, stuaxo
2010-05-09 19:25:45giampaolo.rodolasetmessageid: <1273433145.21.0.463298069236.issue4972@psf.upfronthosting.co.za>
2010-05-09 19:25:43giampaolo.rodolalinkissue4972 messages
2010-05-09 19:25:43giampaolo.rodolacreate