pylint woes

[MRAB]

On 2016-05-10 23:36, DFS wrote:
[snip]
>
> If lists are still being created:
>
> * at every moment in time, len(list1) returns a length that doesn't
> change even if data is added to the list after the call to len().
>
> Example: If the list has 100 items in it at the point len(list) is called:
>
> for i in range(len(list1))
>
> will never iterate more than 100x, no matter how large list1 grows to.
>
> Caveat: since list1 may be bigger or smaller than the other lists at
> that moment in time, an error may occur when using list2[i], list3[i].
>
>
> Is that all correct as you understand it?
>
Yes.
>
>
> * at every moment in time, zip(list1, list2, etc) will return a fixed,
> same-length lists of tuples, which doesn't change even if data is added
> to any of the lists after the call to zip().
>
> Example: if the lists have 100, 97 and 102 items in them at the point
> zip(lists) is called:
>
> for item1, item2, item3 in zip(list1, list2, list3)
>
> will never iterate beyond 97x, even if the lists grow while the
> enumeration is occurring.
>
> Caveat: since zip() possibly truncates lists, the results - the usage of
> the data - could be completely invalid.
>
>
> Is that all correct as you understand it?
>
In Python 2, zip iterates through the arguments immediately and returns 
a list of tuples, so the answer is yes.

In Python 3, zip returns a lazy iterator (like itertools.izip in Python 
2) that gets the values from the arguments _on demand_, so the answer is no.

[snip]