On Wednesday 12 February 2003 06:33 am, Matt Sergeant wrote:
> On Wed, 12 Feb 2003, Robin Berjon wrote:
> > > Also let me know now if there's stuff you'd like to see go into
> > > PerForm. I may be hacking a bit with it in the next week or so, so now
> > > is the time to ask for features.
> >
> > Full XForms integration? [ducks!]
> >
> :-) Seriously though this might be awesome, but I think it's a lot more
>
> work, and might even not fit into the XSP model. I haven't had time to
> investigate it thoroughly though.

I looked at XForms and its very 'Cocoony' in its structure to my thinking. It 
certainly would make sense to utilize their syntax at the very least if its 
got the necessary features though. 

As for the way they deal with object model binding and such, I'm not really 
sure that it makes a lot of sense. My experience, which parlty comes out of 
being on the OWASP input filtering project, was that a seperate XSP taglib 
that lets you specify 'assertions' about input is better. I have a simple one 
(pitifully simple, but I'd be happy to have some people comment on it) which 
does this. So for input validation you just have XSP tags like

<assert:numeric><param:mynumberonlyfield/></assert:numeric>

which returns either its input, or throws an assertion error. Other tags do 
similar things, and you can easily build tags that do transforms as well 
(thus I have one that strips tags from input, etc.).

Neither the old Perform callback structure nore (I think) the way XForms works 
would play with that. Of course we COULD always come up with some other way 
of doing it, but I like the concept of having an input testing/filtering 
package.

-- 
Tod Harter
Giant Electronic Brain

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to