Re: [xsl] Ranking Random Nodes from Top to Bottom

[Abel Braaksma <abel.online@xxxxxxxxx>]

J. Zhang wrote:
> I was thinking of the determining the string length from root to a 
> particular node, and store that as output in the output file using 
> XSLT. Then post-process it in Python to rank the paths into document 
> order, where the lowest number appears on top and the highest number 
> at the bottom. Does that make any sense?

an example would help a great deal here. You've talked of unstructured 
text in your first post, now I understand that it is actually about real 
nodes. It seems to me that you have strings of XPaths and that for some 
reason (why?) you want to show them to the user in the order equal to 
what the XPaths would select from the (a?) document. If so, what about 
an XPath that selects multiple nodes?

To know whether one node is before the other, you can use the 
preceding:: axis (but this is an expensive operation). But you'll first 
have to write an XPath parser in XSLT if you want the xpaths to be given 
as strings. If you can (or want to) rely on extensions, you can use 
saxon:evaluate() when you run your stylesheets with Saxon.


Cheers,
-- Abel Braaksma