The following code does not work like expected, it triggers a NameError.

class C:
    a = 42
    list(a for _ in 'x')

>>> NameError: global name 'a' is not defined 

This issue was discussed on comp.lang.python in the thread "Is this a
bug, or is it me?".

It's a weakness, not a bug.

Just use a different variable name besides 'a'.

I don't follow what you mean.  Can you post a working version of the
code fragment?

Don't follow you:

>>> class C:
	a = 42
	list(a for _ in 'x')

>>> 

Works here! (Python 2.5.1 on win32)

Facundo, are your sure that your output starts from a fresh environment?
I get:

C:\Python25>python
Python 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.1310 32 bit
(Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> class C:
...     a = 42
...     list(a for _ in 'x')
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 3, in C
  File "<stdin>", line 3, in <genexpr>
NameError: global name 'a' is not defined

This error is actually approximately "documented" in an obscure sentence
in http://docs.python.org/dev/reference/executionmodel.html#naming

"""
The scope of names defined in a class block is limited to the class
block; it does not extend to the code blocks of methods.
"""

Well, I'm not sure that the genexpr can be considered as a "method", but
it is certainly a nested code block.

Yes, something was bad with my test. Now I have the same behaviour.

Sorry for the noise.

Amaury Forgeot d'Arc schrieb:

> Well, I'm not sure that the genexpr can be considered as a "method", but
> it is certainly a nested code block.

Technically it is a method, that's why this happens. I added a note to the
docs in r60051; closing this as "won't fix".