Augmented Assignement (was: Re: PEP scepticism)

[Rainer Deyke]

"Tim Peters" <tim.one at home.com> wrote in message
news:mailman.993952381.32699.python-list at python.org...
> As usual in this debate, the participants aren't *themselves* confused
about
> it, they're worried about people much feebler than they are.  In this
> particular case, don't be so sure that isn't an empty set <wink>.

You misunderstand my position.  I am only concerned about myself.  Newbies
mean nothing to me, maintainers mean nothing to me, and I'll happily use
code like "([a] or [b] and [c])[0]".  I'm worried about conceptual overhead
and correctness.

Conceptual overhead: I can learn simple things and I can learn complex
things, but I've found that I am more productive when using simple and
flexible tools than I am when working with complex tools.  This is one of
the reasons why I prefer Python over C++.

Correctness: Consider the following function:

def f(x):
  x += (1, 2, 3)
  print x

This function is just plain incorrect.  If I want a function that mutates
its argument, it should not accept immutable objects.  If I want a function
that does not mutate its argument, the above is simply incorrect.  There is
no conceivable case where I would want a function that mutates or does not
mutate its object based on the mutability of the object.

In fact, it is quite possible that the function above fails to mutate a
mutable argument if the type/class of the argument was not written to
support augmented assignment.  Once this oversight is corrected, my
function's semantics change and my program breaks.


--
Rainer Deyke (root at rainerdeyke.com)
Shareware computer games           -           http://rainerdeyke.com
"In ihren Reihen zu stehen heisst unter Feinden zu kaempfen" - Abigor