On Mon, Oct 24, 2016 at 5:56 PM, Chris Angelico <rosuav@gmail.com> wrote:

On Tue, Oct 25, 2016 at 4:48 AM, Chris Barker <chris.barker@noaa.gov> wrote:
> On Mon, Oct 24, 2016 at 8:21 AM, Michel Desmoulin
> <desmoulinmichel@gmail.com> wrote:
>>
>> This actually could be implemented directly in str.replace() without
>> breaking the API by accepting:
>>
>> "stuff".replace('a', '')
>> "stuff".replace(('a', 'b', 'c'), '')
>> "stuff".replace(('a', 'b', 'c'), ('?', '*', ''))
>
>
> +1 -- I have found I Need to do this often enough that I've wondered why
> it's not there.
>
> making three calls to replace() isn't too bad, but is klunky and has
> performance issues.

And it may not be semantically identical. In the examples above, three
separate replace calls would work, but a syntax like this ought to be
capable of an exchange - "aabbccdd".replace(('b', 'd'), ('d', 'b')) ==
"aaddccbb".

What would be the expected behavior of "aabbccdd".replace(('a', 'aa'), ('x', 'y'))? It's not obvious to me whether longer replacement strings ('aa') or earlier replacement strings ('a') should take priority.

Or is the proposal to only support this for replacements of single characters?

Nathan

ChrisA

_______________________________________________
Python-ideas mailing list
Python-ideas@python.org
https://mail.python.org/mailman/listinfo/python-ideas
Code of Conduct: http://python.org/psf/codeofconduct/