Tim wrote:
> Hi, I need to generate a list of file names that increment, like
this:
> fname1
> fname2
> fname3 and so on.
>
> I don't know how many I'll need until runtime so I figure a
generator is
> called for.
>
> def fname_gen(stem):
> i = 0
> while True:
> i = i+1
> yield '%s%d' % (stem,i)
>
> blarg = fname_gen('blarg')
> boo = fname_gen('boo')
>
> n = 3
> for w in range(0,n):
> in_name = blarg.next()
> out_name = boo.next()
>
>
> This works, but is it the 'normal' way to accomplish the task when
you
> don't know 'n' until run-time? thanks,
It's kind of overkill in the toy demo example, but if the main loop is
a little more abstract, e.g.
for task in task_supplier():
in_name = blarg.next()
out_name = boo.next()
handle_task (task, in_name, out_name)
then it's obviously a good thing.
One quibble (that Peter Otten also suggested): if your business rules
expect that files blarg25 and boo25 (for example) work together, then
you'd be better off generating them together as a pair in a single
generator call. As it is, there's a chance of the blarg and boo
generators getting out of step and supplying mismatched names.
Mel.
--
http://mail.python.org/mailman/listinfo/python-list
