no side effects

[holger krekel]

"Martin v. L?wis" wrote:
> holger krekel wrote:
> >     for i in [1,2,3]:
> >         print i
> >         i=3
> > 
> > the for-loop will iteratively bind the (local) name to  
> > 1, then 2, then 3.  It never looks at the binding of
> > the name 'i'.  With 'i=3' you only change the binding
> > of 'i' to the object '3' but the for-loop will blindly
> > change the binding again. 
> 
> This has nothing to do with namespaces: i is not local to the for loop, 
> but might be global, or local to the function.

'i' might not be local but usually it is.  When answering 
newcomer questions i try to refrain from talking in language
lawyer details (as much as possible).  I only tried to explain the 
for-loop from the viewpoint of names bound to objects.  I didn't 
want to compare iteration protocols between python and C++ just 
to explain that. 

    holger