----- Original Message ----- 
From: Matthew Bentley 

> In XSLT, at the moment, if you want to check if the preceding-sibling is a
> particular element type, lets say, 'heading', then the following code is
> necessary:
> 
> test="preceding-sibling::*[self::heading and position() = 1]"
> 
> It seems to me that this situation (I.e testing the type of an immediately
> neighbouring element) is fairly common and there should be some abbreviation
> for it for preceding, following, preceding-sibling and following-sibling.
> Anybody agree?, disagree?

I think you are not the only one who feels that current 
Xpath is more verbose than it could be.

I was thinking about 

some[  -1 ]  for preceding-sibling of element 'some'
some[ +1 ]  for following-sibling of element 'some'
some[  1 ]   for first child of element 'some'

Rgds.Paul.

<offtopic>
PS. Considering attribute to be ordinary child 
element also simplifies a model ( and 'everything' ).
The only drawback ( scientifically valid, 
but almost not existent in the real life ) is that there 
should be no  attribute with the name equal to 
the name of child element. Considering 
PCDATA to be a special element <_> also 
simplifies some things. In fact this all is MinML 
model, not XML model, so sorry for offtopic - 
just wanted to point out some other view 
which sometimes makes things easier.
</offtopic>



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list