[Python-Dev] RE: PySequence_Contains

[Guido van Rossum]

> [Guido]
> > Actually, instance_contains checks for AttributeError only after
> > calling instance_getattr(), whose only purpose is to return the
> > requested attribute or raise AttributeError, so here it is safe: the
> > __contains__ function hasn't been called yet.

[Tim]
> I'd say "safer", but not "safe":  at that point we only know that *some*
> attribute didn't exist, somewhere, while attempting to look up
> "__contains__".  Ignoring it could, e.g., be masking a bug in a __getattr__
> hook, like
> 
>     def __getattr__(self, attr):
>         return global_resolver.resolve(self, attr)
> 
> where global_resolver has lost its "resolve" attr.  "except" clauses aren't
> more bulletproof in C than in Python <0.9 wink>.

Yes, but attribute errors inside __getattr__ hooks are *always* a
problem to debug, since raising AttributeError is part of its job.  So
this is not new.  I should have said "as safe as it gets."

> > With previous behavior of 'x in instance'.  Before we had
> > __contains__, 'x in y' *always* iterated over the items of y as a
> > sequence, comparing them to x one at a time.
> 
> I don't believe I ever knew that!  Thanks.  I erronesouly assumed that the
> looping behavior was *introduced* when __contains__ was added.

Surely you knew that "x in y" looped over the items of y?  What else
could it have done?  It was only defined on sequences!

> > ...
> > No, that was probably just an oversight -- clearly it should have used
> > rich comparisons.  (I guess this is a disadvantage of the approach I'm
> > recommending here: if the default behavior changes, the
> > reimplementation of the default behavior in the class must be changed
> > too.)
> 
> I factored out the new iterator-based __contains__ logic into a new private
> API function, called when appropriate by both PySequence_Contains() and
> instance_contains().  So any future changes to what iterator-based
> __contains__ means will only need to be made in one place.

Cool.

--Guido van Rossum (home page: http://www.python.org/~guido/)