iostream-like lib?

[Anton Muhin]

Max Khesin wrote:
> The trouble is that readline() reads more than i have to in the first place,
> even before I call split().
> I did hack it along the lines you suggested with a generator (limiting
> readline() to a number of bytes and accounting for the last character being
> possibly whitespace). I was just wondering if (and why not) there is/is not
> direct support for whitespace-delimited input.

I don't know :)

The only thig I want to add: I see no reason why you should limit 
readline with number of bytes? Doesn't the following code (untested) work?

def read_tokens(filename):
     f = file(filename)
     for l in f:
          for token in l.split():
               yield token
     close(f)

Best regards,
anton.