PEP 284, Integer for-loops

[John Machin]

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:<mailman.1015439892.7298.python-list at python.org>...
> > From: David Eppstein [mailto:eppstein at ics.uci.edu] 
> > Subject: PEP 284, Integer for-loops
> 
> Thanks for writing the PEP.
> 
> I see this as a general proposal for a range syntax, and I disagree with
> it because:
> 
>   - it doesn't handle the general range issues, i.e. step value,
>     and if we had to add yet another syntax to get general ranges
>     it would be too confusing.

Agreed. Let's generalize the existing syntax.

Existing: for <var> in <sequence>:

Proposal: For a splendid blast from the past, let's take a leaf out of
ALGOL 60's book.

for <var> in <for-list>:

where the elements of <for list> can be of 3 different kinds:

(1) <arithmetic expression>
(2) <step-until-element> ::= <start-value> [step <step-value>] until
<last-value>
(3) <while-element> ::= <arithmetic expression> while <boolean
expression>

Lets you do things like

for k in 0,1,2, 5 step 5 until 25, k * 2 while k <=1024:

> 
>   - I would like range objects to be first class, ie. I would want
>     to be able to pass ranges to functions, return them from 
>     functions, and do all the other things one can do with first
>     class objects.

I don't understand this. range and xrange *are* first-class objects.
You can do all of those things.

> 
>   - It doesn't make the general case trivial, i.e. 
>     for 0 <= i < len(mySequence) is not much better than 
>     for i in range(len(mySequence) and is more verbose than
>     for i in len(mySequence).
> 

for i in 0, i+1 while i < len(mySequence):

for i in 0 until len(mySequence) - 1: 

Hmmm, ... maybe not.