[Python-Dev] Informal educator feedback on PEP 572 (was Re: 2018 Python Language Summit coverage, last part)

Steven D'Aprano steve at pearwood.info
Sun Jun 24 01:56:47 EDT 2018

On Sun, Jun 24, 2018 at 02:33:59PM +1000, Nick Coghlan wrote:

> Given that PEP 572 *is* proposing implicit comprehension state export,

"Implicit" and "explicit" are two terms which often get misused to mean 
"I don't like it" and "I do like it".

Making the intentional choice to use an assignment expression is not 
really "implicit" in any meaningful sense. One might as well complain 
that "import this" implicitly creates a local variable "this". Well, 
yes, it does, in a very loose sense, but that's what imports are 
defined as do and it is the whole purpose for making them.

If PEP 572's proposal goes ahead, the behaviour of assignment 
expressions will be *defined* as creating assignments in the local scope 
rather than the sublocal comprehension scope. To call that "implicit" 
is rather like saying that regular assignment is implicit.

> though, then I think it's important to make the case that seeing the
> proposed semantics as intuitive is only going to be the case for folks
> that have used Python 2 style comprehensions extensively - anyone
> that's never encountered the old state-leaking behaviour for iteration
> variables is going to be surprised when assignment expressions ignore
> the existence of the comprehension scope (even though the iteration
> variable pays attention to it).

You are making the assumption that most people are even aware of 
"comprehension scope". I don't think that is the case.

In my experience, scoping in Python is still typically seen as the LGB 
rule (locals/globals/builtins). See for example this StackOverflow post 
from 2016:


Sometimes people remember the E/N (enclosing function/nonlocal) part. 
Hardly anyone remembers the C (class) part unless they are actively 
thinking in terms of code running inside a class definition, and even if 
they do, they typically aren't sure of exactly how it interacts with the 

And I predict that even fewer think of comprehensions as a separate 
scope, except by ommission: they *don't think about* the scope of the 
loop variable until it bites them.

But as Tim Peters has previously discussed, the loop variable is 
special, and is especially prone to accidental shadowing. That won't be 
the case for assignment expressions. If there's shadowing going on, it 
will be deliberate.

Aside: I've said before that I'm not a fan of sublocal comprehension 
scoping, since I personally found it helpful on occasion for the loop 
variable to be visible outside of the comprehension. But given that the 
only experience most people apparently had with comprehension scoping 
was to be bitten by it, I grudgingly accept that encapsulating the loop 
variable was the right decision to make, even if it personally 
inconvenienced me more than it saved me.

Nor was I the only one: others have been bitten by the change to 
comprehension scope, see for example:


There is no consensus that the change to comprehensions was a good thing 
or justified.

The bottom line is that I don't think people will be surprised by 
assignment expression scope being local instead of sublocal. Rather I 
expect that they won't even think about it, until they do, and then 
*whatever* behaviour we pick, we'll annoy somebody.


More information about the Python-Dev mailing list