[Python-3000] Adaptation and type declarations

[Guido van Rossum]

On 4/11/06, Jim Jewett <jimjjewett at gmail.com> wrote:
> On 4/10/06, Guido van Rossum <guido at python.org> wrote:
> > [Guido]
> > > > It should mean neither. As I tried to say, the *only*
> > > > semantics is that the value of the expression is
> > > > accessible through the __signature__ attribute of
> > > > the function. It is then up to the decorator to decide
> > > > what it means.
>
> Rephrasing for verification:
>
> In and of itself, the type specifier does nothing.  It does add a
> (virtual) annotation to the argument, so that anything which *does*
> act (such as a decorator) can be written in a more generic manner.

Right.

> > [Jim Jewett]
> > > So type annotation is only for use with decorators,
> > > not for general documentation or compiler speedups?
>
> > Those are two entirely different purposes. It's fine for
> > documentation, for use by IDEs, and for use by e.g.
> > pychecker.
>
> So (besides decorators), it is for (exactly) introspective uses?

It is for no specific purpose. :-)

> > However, using it for compiler speedups is unlikely
> > given the run-time semantics and Python's extremely
> > dynamic nature.
>
> There are several PEPs (and some PyPy progress) towards more stable
> namespaces.  As PEP 266 points out, these namespaces typically never
> change, and it might be reasonable to make these changes more
> expensive if they sped up the common "from now on, it is read-only"
> case.
>
> I can easily imagine a slightly modified __import__ that sealed the
> module after creation.  Even for a seal that *can* be broken, it might
> be reasonable to say "but if you do, these precompiled objects that
> relied on the seal will take the slow path until they're recompiled."

That is all pretty much uncharted territory (for Python). If and when
it comes to fruition I hope it will come with a set of recommended
best practices for using argument types and decorators to tell the
compiler more about a function.

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