29.11.2011 13:17, [email protected] пишет:
On Tue, 29 Nov 2011, Graeme Geldenhuys wrote:
I would even add a FindNode() method which internally uses Node[] -
just to make in very clear to the developer using this class.
I guess if I ever need to work with XML (lets hope that never
happens), I'll implement my (yet again) own classes. Programming is
supposed to solve problems, not create more. The alternative would
have been to modify the FCL DOM code, but that would probably cause
breakages all over existing code (eg: fpdoc, makeskel etc) - so FCL
DOM code is now stuck in near unusable limbo land. :-(
Graeme, please stop bashing code you personally don't understand.
The code works as designed. It is an implementation of a W3 spec.
It passes the W3 test suites, so it is correct code.
I use it in many many projects, and it performs flawlessly.
The fact that we follow the W3 spec allows people to easily switch between a
Delphi or even a Java DOM implementation and a FPC DOM implementation.
If you think the spec sucks, please complain with the W3 consortium.
It's not because you don't agree with this interface that the "FCL DOM code is
now stuck in near
unusable limbo land."
I use tiOPF. I do not agree with many decisions that were taken for it's
implementation, but I do refrain from bashing tiOPF and saying things like
'tiOPF code is stuck in
near unusable limbo land'.
I couldn't have said it better.
To add to this, the core DOM is simply not supposed to do things like path navigation. These tasks
are supposed to be done by other modules, like 'traversal' and 'range'. They are powerful, but
unfortunately not implemented in FCL. The only more-or-less implemented module is XPath.
Regards,
Sergei
--
_______________________________________________
Lazarus mailing list
[email protected]
http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
