[ZPT] zope3-dev proposal on changing implicitly calling

[Steve Alexander]

Lynn Ranen wrote:
> I was under the impression that the TALES was seperate from Zope3.

Yes, in the sense that Zope3 depends on TALES, not the other way around.


> [1] I disagree that the TALES should be modified to work with Zope3
> -- shouldn't it work the other way around?

I am not proposing that TALES be modified to work with Zope 3.

I am proposing that TALES drop its special case for when the result of a
path expression has a __call__ attribute.

Guido has indicated approval of this. Seeing as Python 2.x has new-style 
classes with properties, there is no compelling need to have a special 
case for callable attributes.

   http://mail.zope.org/pipermail/zope3-dev/2003-June/007373.html


> [2] To me, it looks like the behaviour (__call__ at the end of path 
> traversal) is similiar to the behaviour of the publishing system.

Which publishing system?


> This appears to be consistent design.

What do you think is consistent with what else?


> In fact, I would go 1 step
> further, and say that what the TALES needs is a 'mapply' -- just like
> the publishing system has.

The policy for evaluating path expressions can be plugged into TALES as 
needed. Zope 3 uses this plug-point to make TALES expressions use zope's 
traversal mechanisms.


> Then, it will be possible/easy to pass in arguments.

I don't follow how this enables passing in arguments. I'm not sure that 
passing in arguments to TALES path expressions is desirable. A lot of 
the advantages of TALES comes from its simplicity.
Can you explain what you mean about passing in arguments?

--
Steve Alexander