[PEAK] How do I resolve AmbiguousMethod?

[Daniel Nouri]

I've played around a bit with RuleDispatch today.  Basically, I wanted to
implement Zope's adaptation with RuleDispatch, which was pretty straight
forward, until I came to the point where I had two adapters that would both
fit my criteria.

My first generic function factory looked like this:

def provideAdapter(adapter):
    @_do_adapt.when(
        "issubclass(adapter.implements, interface) and "
        "issubclass(obj.implements, adapter.adapts)")
    def adaptor(obj, interface):
        return adapter(obj)

However, if I register two adapters (by calling provideAdapter), where one
provides a more specialized interface (interface B(A)) than the other
(interface A), and I request adaptation to A, I want the other adapter, i.e.
the one for interface A, to be returned.

I went along and implemented a function base_distance that returns the
distance between a class A and a class B in the class hierarchy.  Now I'd
like to be able to express this:

        "issubclass(adapter.implements, interface) and "
        "base_distance(obj.implements, adapter.adapts) >= 0"

and somehow imply that the smaller base_distance, the better.  That is, if
base_distance is 0 for one function and 3 for another, the first one should
be picked.  I tried to look into RuleDispatch's interfaces.py, but got stuck.

What's the best way to do what I want to do?

You can the source code of my experiment here:
http://danielnouri.org/dev/file/scratch/experiments/adaptation_with_gf.py

TIA,
Daniel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: daniel.nouri.vcf
Type: text/x-vcard
Size: 138 bytes
Desc: not available
Url : http://www.eby-sarna.com/pipermail/peak/attachments/20070126/591f48b9/daniel.nouri.vcf