[XML-SIG] DOM API conundrum

uche.ogbuji@fourthought.com uche.ogbuji@fourthought.com
Wed, 22 Nov 2000 17:13:36 -0700

>   As I'm working on the API documentation, I'm finding that I'd like
> to document the nice, Pythonic interface and not the accessors.  Is
> there anyone who would have problems if I deprecate the accessors in
> the documentation, and encourage direct access to the attributes?  The
> only issue for DOM-over-CORBA that I can see is that a wrapper would
> be needed to call the _get_/_set_ methods on the surrogate when the
> attributes are accessed; just a little __getattr__/__setattr__ magic
> would be needed.  Once the Document object is wrapped, the wrapper
> could handle wrapping of other objects from the tree automatically, so
> the infliction of wrappers would not need to permeate client code.
>   Any objections to deprecating the _get_/_set_ methods?

The main reason to have these was to allow users avoid the performance hit of 
__g/setattr__.  I don't have a problem giving this up.  It would be _really_ 
nice if Python had computed attributes (any PEPs out there?).

So +1 on deprecating function-based access in the Python/DOM binding.  I never 
use them anyway.

Uche Ogbuji                               Principal Consultant
uche.ogbuji@fourthought.com               +1 303 583 9900 x 101
Fourthought, Inc.                         http://Fourthought.com 
4735 East Walnut St, Ste. C, Boulder, CO 80301-2537, USA
Software-engineering, knowledge-management, XML, CORBA, Linux, Python