cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Fagerstrom <>
Subject Re: [RT] do me a favor, don't call them taglibs (TAL-like two-step templating?)
Date Sat, 04 Dec 2004 16:11:09 GMT
Bertrand Delacretaz wrote:
> Le 4 déc. 04, à 14:14, Daniel Fagerstrom a écrit :
>> Bertrand Delacretaz wrote:
>>> ...b) "presentation templates" downstream to convert the XML to HTML 
>>> or other presentation markup (currently covered by XSLT, far from ideal)
>>> ...For b), being dreamweaver-compatible would be a big plus, allowing 
>>> less technical people to create templates themselves....
>> b) would be nice but it is quite complicated to implement in an 
>> efficient way. One can of course stream the transformer input to a DOM 
>> tree that is made available in the context object and then let a JXTG 
>> like script work on that. But the question is if we would be happy wih 
>> the performance...
> It's hard to say without trying it, but IIUC we can essentially get this 
> (i.e. reuse the data-to-XML templating mechanism for downstream 
> transformations) for free, if taken into acount in the design. Just a 
> simple stream-to-DOM implementation would allow people to experiment.

Yes, we would get that for free AFAIU. The design discussed this far 
should handle that without any problems. We need a convention for where 
to find the input DOM. Either implicitly like xslt or some special 
variable "$input".

> And it might make a big difference, from the users point of view, by 
> making it possible to use a single tool for the whole presentation 
> pipeline, and making presentation templates way easier than raw XSLT, 
> which is a major stumbling block for many people.
> We might want to make it work first, to see how much people like it. If 
> they like it, I'm sure someone will find a way to make it fast, if needed.

You're right no need to optimize until we know if it is needed. AFAICS 
the only reasonable way of accessing the XML input is things like XPath 
that in general need access to the complete XML-tree. And in that case 
we can reuse XSLT processor technology if needed. Going for something 
like STX-style access would be far to complecated for the intended audience.


View raw message