OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   Re: [xml-dev] XSLT vs. CSS (Re: Indexing)

[ Lists Home | Date Index | Thread Index ]

David Carlisle wrote:

>>That makes perfect sense if you are looking at the XML as a document
>>with no semantic other than presentation. 
> 
> 
> Actually many of my documents contain mathematics that i hope doesn't
> just have presentational semantics.

Granted, in one sense. (I'd rather not get into a tit-for-tat on the meaning of the word semantics. )

> But I just really meant the point that you re-inforced:
> 
> 
>>Placing simplifying constraints on the transforms seems to make it a
>>plausible approach...  
> 
> 
> It's clearly possible, if you know what you are doing, to write specific
> transforms for specific documents that have a reliable inverse mapping,
> but the nature of XSLT makes it a highly unsuitable language for that
> use. 

Well, I won't argue that, but leave that one for the XSLT engine developers out there. 

> You can't simply constrain the usage by restricting to a subset of
> the language: the default behaviour of an XSLT stylesheet is to discard
> all element markup and return a character string. 

If you are a tool developer and wish to utilize XSLT as an internal component, then you can indeed constrain the usage, as well as the run-time context. How to achieve that may be much less straightforward than with CSS.

Isn't the default behavior to construct a result tree? If I were designing such a system, it would probably provide meaningful defaults, and a set of small, modular transforms rather than a single monolithic stylesheet. 

Having said that, I'm not currently designing such a system and on consideration I don't see any tools out there attempting to use XSLT in an interactive manner. What I see is XSLT used to construct editors' user interfaces, but this is probably not the mode of operation that you meant.

Regards,

Mitch

> Since that presumably
> isn't much use if you want to interact with the original document,
> you'll need to specify constraints on the result having the right
> properties rather than constraints on the way the XSLT is used.
> That's a rather hard thing to enforce. I would have thought a system
> designed for editing would have been better designed to use a language
> that ensures that there is a well defined mapping back from  the result
> to the source.
 
> David
> 
> 
> ________________________________________________________________________
> This e-mail has been scanned for all viruses by Star Internet. The
> service is powered by MessageLabs. For more information on a proactive
> anti-virus service working around the clock, around the globe, visit:
> http://www.star.net.uk
> ________________________________________________________________________
> 






 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS