Using XML Events and Actions outside XForms

> Hi,
>
> AFAIK, the implementation of XML Events and Actions in OrbeonForms is
> tightly coupled to XForms, but I'd better ask the question  
> anyway ;) ...
>
> What I'd like to do is trigger pipelines (or processors) based on the
> modifications done on XML documents posted through OrbeonForms.
>
> One of the idea would be to do that through a pipeline that would  
> use an
> "XML Events and Actions" processor. The processor would take as  
> inputs:
> the XML Events definitions, the document before update and the  
> document
> after update and apply the processing defined in the XML Events
> definitions based on the differences applied to the document.
>
> Would that make sense? Are there better ways to do this kind of  
> things?
>
> Would there be anything to reuse in the implementation that you've  
> done
> for XForms or (like I would tend to think) is it too tightly coupled
> with XForms?
>
> Thanks,
>
> Eric
> --
> GPG-PGP: 2A528005
> If you have a XML document, you have its schema.
>                                                  http://
> examplotron.org
> ------------------------------------------------------------------------
> Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
> (ISO) RELAX NG   ISBN:0-596-00421-4 http://oreilly.com/catalog/relax
> (W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
> ------------------------------------------------------------------------

> >> I am not 100% I understand the idea though. Are you saying you would
> >> like the processor to diff the two documents, and then dispatch DOM
> >> mutation events?
> >
> > Yes, something like that.
> >
> > It might be more efficient to compare only the XPath expressions
> > specifying the event targets than doing a full diff, but that would be
> > logically equivalent.
>
> Could you produce a simple example showing how that would look like?

>
> Le mardi 25 mars 2008 à 23:38 -0700, Erik Bruchez a écrit :
>>>> I am not 100% I understand the idea though. Are you saying you  
>>>> would
>>>> like the processor to diff the two documents, and then dispatch DOM
>>>> mutation events?
>>>
>>> Yes, something like that.
>>>
>>> It might be more efficient to compare only the XPath expressions
>>> specifying the event targets than doing a full diff, but that  
>>> would be
>>> logically equivalent.
>>
>> Could you produce a simple example showing how that would look like?
>
> An "XmlEventsAndActions" processor could have a config entry with
> declarations such as:
>
> <listener
>   observer="status"
>   event="DOMCharacterDataModified"
>   if="event('newValue') = 'approved' "
>   handler="oxf:sendNotificationMail.xpl"/>
>
> Together with two other inputs (one for the previous version and one  
> for
> the new version of the document).
>
> In that case, to check if the handler needs to be called, it could  
> just
> evaluate if the node with ID "status" has been changed rather than  
> do a
> full difference between the two versions.
>
> The effect of this declaration would be to call
> oxf:sendNotificationMail.xpl if the node with an id "status" has seen
> its text node updated and if the new value is equal to "approved"  
> and I
> think that this would be fairly conform to the rec.
>
> Of course, that's just the basic idea! We would have to wok on details
> such as how we can specify the inputs and outputs of
> oxf:sendNotificationMail.xpl.
>
> Also, the fact that the observer and targetid attributes accept only  
> id
> references instead of XPath expressions and that the spec assumes that
> the event definitions are embedded in the document itself isn't really
> convenient for what we want to do, however I think that we could  
> derive
> something interesting if we keep a liberal interpretation of the  
> spec...
>
> Eric
>
>>
>> -Erik
>
> --
> GPG-PGP: 2A528005
> Curious about Relax NG? Read my book online.
>                                   http://books.xmlschemata.org/ 
> relaxng/
> ------------------------------------------------------------------------
> Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
> (ISO) RELAX NG   ISBN:0-596-00421-4 http://oreilly.com/catalog/relax
> (W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
> ------------------------------------------------------------------------

> Hi Erik,
>
> Le mercredi 26 mars 2008 à 19:13 -0700, Erik Bruchez a écrit :
>> Eric,
>>
>> Thanks for the example, I understand much better now.
>>
>> Are you thinking about implementing something like this?
>
> I am not 100% sure, but it's something I am seriously considering,  
> yes.
>
> Before that, I wonder if I couldn't prototype the feature (at least  
> for
> simple cases) in XSLT: it should be possible and even quite simple to
> generate a pipeline based on listener declarations and the before and
> after images of a document.
>
> I'll keep this mailing list informed.