On Sat, 16 Jul 2005 23:28:02 -0400, "George Sakkis" <[EMAIL PROTECTED]> wrote:
>"Peter Hansen" <[EMAIL PROTECTED]> wrote:
>
>> George Sakkis wrote:
>> > "Peter Hansen" <[EMAIL PROTECTED]> wrote:
>> >>>> Where did you learn that, George?
>> >
>> > Actually I first read about this in the Cookbook; there are two or three
>> > recipes related to string.translate. As for string.maketrans, it
>> > doesn't do anything special for empty string arguments: ...
>>
>> I guess so. I was going to offer to suggest a new paragraph on that
>> usage for the docs, but as you and Jp both seem to think the behaviour
>> is obvious, I conclude "it's just me" so I suppose I shouldn't bother.
>
>It's only obvious in the sense that _after_ you see this idiom, you can go
>back to the docs and
>realize it's not doing something special; OTOH if you haven't seen it, it's
>not at all the obvious
>solution to "how do I get the first 256 characters". So IMO it should be
>mentioned, given that
>string.translate often operates on the identity table. I think a single
>sentence is adequate for the
>reference docs.
>
I would suggest changing
"""
maketrans(from, to)
Return a translation table suitable for passing to translate() or
regex.compile(),
that will map each character in from into the character at the same
position in to;
from and to must have the same length.
"""
to something that would make the idiom more easily inferrable, e.g.,
"""
maketrans(from, to)
Return a translation table suitable for passing to translate() or
regex.compile(),
that will map each character in from into the character at the same
position in to,
while leaving all characters other than those in from unchanged;
from and to must have the same length.
"""
Meanwhile, if my python feature request #1193128 on sourceforge gets
implemented,
we'll be able to write s.translate(None, badchars) instead of having to build
an identity table to pass as the first argument. Maybe 2.5? (Not being pushy ;-)
Regards,
Bengt Richter
--
http://mail.python.org/mailman/listinfo/python-list
