Re: [legalxml-econtracts] Caption Numbers

[Jason Harrop <jharrop@speedlegal.com>]

John McClure wrote:

>Jason said:
>  
>
>>To recap what i put in my FAQ document, i think it is highly desirable
>>to have a clause model which has one and only one thing at each level in
>>the hierachy.  As soon as you depart from this, not only do you make
>>things needlessly complex for the user, but for stylesheet purposes, you
>>can't infer the level of something (and therefore how to number it) from
>>its name (eg Section).
>>    
>>
>
>  
>
:

>Here's the corresponding XSL match conditions; within these templates, one can
>generate caption numbers without too much problem.
>  
>
I'm not suggesting it is impossible.  I'm saying it is undesirable and 
unnecessary.

>  
>
><xsl:template match='*[parent::Contract]'>
><xsl:template match='Paragraph[parent::Contract]'>
><xsl:template match='Article[parent::Contract]'>
>
look at the 2 templates you have created for Section.  

><xsl:template match='Section[parent::Contract]'>
><xsl:template match='Section[parent::Article[parent::Contract]]'>
>
With a seven level hierarchy where anything is allowed anywhere else, 
well, seven factorial equals 720.  Now, i know you won't catch each 
possibility with a distinct template.

But the problem is that users will put a Paragraph directly in an 
Article _by mistake_, and using your approach, your styles would need to 
cater for that.  And I don't want to impose this burden on 
administrators or support staff who are not XPath experts.

><xsl:template match='Paragraph[parent::Article[parent::Contract]]'>
>  
>