Hey Jeni,



Sorry for the late reply!  Was out yesterday for the holiday event here 
on the other side of the pond. :)



Jeni Tennison wrote:



Hi David,



You mean "foo[number(@bar)]" (which selects the <foo> element children

of the current node whose bar attributes match their position amongst

other <foo> elements within the current node) or, more feasibly,

"foo[number(current()/@bar)]" (which selects the nth <foo> element

child of the current node where n is the value of the bar attribute of

the current node).

 



Hmmm... I'm not sure I understand.  The way I interpreted what David 
Carlisle had stated was that, for example, if @bar had the value of "5" 
that you could use then use the number function to tell the XSLT 
processor to use the literal number value of @bar to determine the 
position of an element in relation to the document order of an element 
in relation to its siblings.  Is that how it is or did I misinterpret 
what David was saying?




Right -- that's part of why I tend to advise people to use matching

templates with modes rather than named templates: with a matching

template, you have some hope of knowing what the context node is by

looking at the pattern, with a named template you can't tell without

finding the call to that template, which makes debugging that much

harder.

 



Ah yes, very good advice.  In fact I have found that the only time I 
even use named templates anymore is for recursively seeking out 
particular content within a string and processing it further or as a way 
to specify the number of times the named template should be called and 
the instruction set processed before moving forward.  And even then I 
try to make minimal use of this type of processing for the simple fact 
that if used in the wrong situation you can find yourself spending a lot 
of CPU cycles and bottleneck an entire application because of it.




I'll certainly try to pop in.

 



That would be great!  I know that we would all appreciate any further 
insight you have on particular topics.  We are going to use Dr. Kay's 
XSLT and XPath (1.0 and 2.0) Programmer References as a way of further 
referencing a topic. But to have your insight to further understand a 
topic of interest would be invaluable and an outstanding asset to the 
project.




No; a link to Amazon is fine.

 



Great, I will leave it as is then.



Thanks again Jeni!  I've added a couple of entries on the site today, 
one which is specific to copyrights and the policy I will be 
implementing when it comes to anything that could be construed as a 
copyright infringement.  If you have a chance to take a look at the 
policy and have anything additional you would like for me to consider, 
please let me know.  The last thing I want to have happen is to infringe 
upon copyrights and will be making a specific effort to not use the 
literal text from the book beyond simple quotes and, of course, all of 
these will be properly marked and attributed.  I will be asking for 
those who make comments to follow these same guidelines and I will be 
monitoring conformance, deleting anything that could even possibly be 
considered an infringement.  If at anytime you see anything on the site 
that I have missed that you or APress feel a bit uneasy about please let 
me know and I will make sure it is taken off the site immediatelly!



Cheers!



<M:D/>



Cheers,



Jeni



---
Jeni Tennison
http://www.jenitennison.com/