PEP 284, Integer for-loops

[phil hunt]

On Thu, 07 Mar 2002 09:36:10 -0800, David Eppstein <eppstein at ics.uci.edu> wrote:
>In article <slrna8dd1d.d3t.philh at comuno.freeserve.co.uk>,
> philh at comuno.freeserve.co.uk (phil hunt) wrote:
>
>> How about:
>> 
>>    for i = 1 to 10:
>> 
>> or
>> 
>>    for i is 1 to 10:
>> 
>> I think these are both equally good and both much preferable over 
>> the horrible
>> 
>>    for 1 <= i <= 10:
>> 
>> which I can only assume is someone's sick idea of a joke.
>
>Your "1 to 10" syntax (and most other ideas floated previously on this 
>group) assumes a fixed type of interval (closed, I'm guessing, but it's 
>not clear), and not even the most common type (half-open).

Not sure what you mean here.

> The proposed 
>syntax makes explicit what would happen at the ends of the intervals.
>
>One alternative proposal, that didn't seem to have as much support but 
>might still be worth looking at, was to allow the creation of ranges by
>"1 <= ... <= 10" (possible without much modification of the compiler via 
>a clever enough definition of "...") so then one could do
>"for i in 1 <= ... <= 10".
>
>I would appreciate it if you could give me a more constructive criticism 
>of PEP 284 than "horrible", "someone's sick idea of a joke", which don't 
>give me much idea why you don't like it.

Thev reasdon i don't like it is that the syntax already has a 
meaning: to compare the magnitude of values. Now you are proposing 
another, totally different, meaning.

In a way what you are saying is (a <= b <= c) means return a list of
all the integer values of (b) that satisfy this statement.

But that isn't really true, because you can't do things like:

   a <= b*2 <= c

or:

   2 <= b and b*b <= c

etc.

So the system isn't fully general. In fact, its expressiveness is 
*less* than the existing range/xrange system, in which you can 
specify a step.

The onlyt problem with range/xrange, as I see it is that it always 
excludes the to-value, ie.

   range(0,3)

returns [0, 1, 2]

So perhaps it would be better if there were rangei/xrangei 
functions, such that:

   rangei(0,3)

returns [0, 1, 2, 3]

(Or has this been fixed in 2.2?)

-- 
<"><"><"> Philip Hunt <philh at comuno.freeserve.co.uk> <"><"><">
"I would guess that he really believes whatever is politically 
advantageous for him to believe." 
                        -- Alison Brooks, referring to Michael
                              Portillo, on soc.history.what-if