Lists Home |
Date Index |
- From: Glenn Vanderburg <email@example.com>
- To: John Cowan <firstname.lastname@example.org>
- Date: Tue, 09 Mar 1999 13:46:00 -0600
> We should take this off-line. I'll simply say: exceptions are
> suitable for reporting exceptional conditions. Having an object
> request its own replacement is certainly exceptional.
Well, yes and no. But I'd prefer to go the cleaner route of not
allowing the object to request its own replacement.
> The idea here is that an application may request a feature which
> a parser does not itself support, but can be adapted to support
> by pushing a filter between itself and the application.
Yes, I understand.
> of course requires that the application now talk to the filter
> instead. (In principle, the parser could act as an adapter for
> the filter, but that would complicated the bejesus out of it.)
It's not complicated at all --- merely a little tedious. It would
be easy to provide a class in the helpers package that would make it
My primary objection to the idea is precisely what you mentioned above:
that it is an extremely unusual thing to happen. Programmers will be
surprised by this behavior. Coupled with the fact that it's very easy
to make it all transparent, I think exposing the parser's internal
tricks is a bad idea.
xml-dev: A list for W3C XML Developers. To post, mailto:email@example.com
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:firstname.lastname@example.org the following message;
To subscribe to the digests, mailto:email@example.com the following message;
List coordinator, Henry Rzepa (mailto:firstname.lastname@example.org)