homepage

Test equality with None/True/False singletons should be done
by "is" rather than "==" to be on the safe side. Otherwise
objects overriding __eq__ could compare equal to one of those
singletons.

You are right of course, but just out of curiosity, do you really have
objects that compare equal to None?

I'm in favor of this patch. Not only is "is" faster here, but it is also
way more idiomatic.

Yes, PEP8 says::
 Comparisons to singletons like None should always be done with
 'is' or 'is not', never the equality operators.
Reading the patch:
- a change modifies "x == False" into "not x", another moves some lines.
I checked that they are OK (x is already the result of a comparison).
- some occurrences of "x != None" are not replaced. Why? (ex. in
test_ast.py)

Amaury, I never saw an object comparing equal to None.
I think the most likely case is a buggy x.__eq__() implementation. Then
the "if x == None" statement gets triggered, and somebody has a hard
time with bug hunting.
Just a note: I used an adapted source checker for this patch, which
initially came from the Sphinx documentation project, found under
tools/check_sources.py. So the credits for this go to Georg Brandl.
That is also why '!=' did not get replaced (the source checker only
searched for '=='. I will post an updated patch.

Despite the title, the patch replaces "result == False" with "not 
result" rather than "result is False". While probably ok in this 
particular context, this changes the logic. For example,
>>> result = ""
>>> result == False
False
>>> not result
True

Here is an updated patch, using "is False" to be consistent, and also
replacing the "!=" occurences.

This patch is fine.
Before applying, check the code in PyShell to see if the "if response 
is False" line can be simplified to "if not response".

Benjamin, do you want to apply this? Add a Misc/NEWS item as well.

Don't think changes like this warrant a NEWS entry. It's a code clean-
up, not a semantic change.

Patch was committed in r62043. Wummel, thanks for the patch! Georg,
thanks for the practice.