What does the list_folders() method of mailbox.Maildir actually ??do (if anything)?

tinnews at isbd.co.uk tinnews at isbd.co.uk
Sat Sep 26 17:48:01 CEST 2009


Jeff McNeil <jeff at jmcneil.net> wrote:
> > > The Maildir++ spec states that folders need to begin with a period.
> > > The list_folders method enforces that:
> >
> > >     def list_folders(self):
> > >         """Return a list of folder names."""
> > >         result = []
> > >         for entry in os.listdir(self._path):
> > >             if len(entry) > 1 and entry[0] == '.' and \
> > >                os.path.isdir(os.path.join(self._path, entry)):
> > >                 result.append(entry[1:])
> > >         return result
> >
> > > The above example is from 2.6.  Your structure is simply a list of
> > > Maildir compliant directories below '/home/chris/Mail/apex.' They're
> > > not, in the Maildir++ sense of the word, folders.
> >
> > So where does it say in the Python documentation that list_folders()
> > works only with Maildir++?  It's a big and non-obvious limitation to
> > my mind.
> >
> > My maildir hierarchy is created by mutt which is a *very* standards
> > compliant MUA, surely standard python libraries should work with
> > standard maildirs not some wierd extension thereof.
> >
> > --
> > Chris Green
> 
> 
> The doc says that "Folders of the style introduced by the Courier mail
> transfer agent are also supported. Any subdirectory of the main

"... are also supported." says to me that 'standard' ones are
supported as well and they're *not*.

> mailbox is considered a folder if '.' is the first character in its
> name."  It's not an explicit endorsement of Maildir++, but it touches
> on what it considers a folder definition. I'm treading on hazy memory
> here, but I believe Maildir++ is the definition provided by the
> Courier folks.
> 
So why isn't the class called mailbox.Maildir++ ?   :-)

-- 
Chris Green




More information about the Python-list mailing list