RE: [xsl] Newbie: Extracting and evaluating partial values from elements

["Michael Kay" <mike@xxxxxxxxxxxx>]

It's easy enough even in XSLT 1.0. Assuming the relevant bit_position is the
context node,

<td colspan="{number(substring-after(., ':')) - number(substring-before(.,
':'))}">

If you're using Saxon 6.5.5, then there's really no reason not to use Saxon
8.9 and take advantage of XSLT 2.0, which makes a lot of things much easier.
Though it doesn't make much difference in this case.

Michael Kay
http://www.saxonica.com/ 

> -----Original Message-----
> From: Jeff Hooker [mailto:jeff@xxxxxxxxxxxx] 
> Sent: 22 October 2007 17:32
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject: [xsl] Newbie: Extracting and evaluating partial 
> values from elements
> 
> Hi all,
> 
> I'm trying to take 32-bit registers that have been documented 
> as follows:
> 
> <?xml version="1.0" encoding="UTF-8"?>
> <register>
>     <bit_postion>31:12</bit_postion>
>     <bit_postiont>11:9</bit_postion>
>     <bit_postion>8</bit_postion>
>     <bit_postion>7</bit_postion>
>     <bit_postion>6:1</bit_postion>
>     <bit_postion>0</bit_postion>
> </register>
> 
> and turn them into a 32-column table where the number of bits 
> in each <bit_position> element is popluated into the a 
> colspan tag. This means that the first bit element would 
> result in <td colspan="19">, the third bit_position element 
> would result in <colspan='1'>, etc. 
> 
> Getting the values out of the single-entry elements is easy; 
> if there's no colon, I know it only contains a single bit. 
> Getting the values when there IS a colon is driving me nuts. 
> I need to to extract the values, determine the difference, 
> and populate the colspan, and XSL's handling of 
> variables/parameters is very new and unfamiliar to me. Any 
> pointers on a good approach would be much apprecitated; 
> everything I've tried so far has ended up chasing its tail, 
> logically speaking. 
> 
> I'm using Saxon 6.5.5, but would prefer to avoid using 
> extensions if possible.
> 
> Thanks in advance for any guidance.
> 
> (Yes, I know that that could have been avoided by adding 
> something like optional <high_value> and <low_value> elements 
> to the <bit_position> element, but the horses left the barn 
> on that one 6 years ago.)