I re-read your initial post ;-) How about: //element(*)[...@prefix='<path prefix extracted from the nodes path>']
If your subscription nodes have a specific nodetype, e.g. 'myblog:subscription': //element(*, myblog:subscription)[...@prefix='<path prefix extracted from the nodes path>'] -----Ursprüngliche Nachricht----- Von: Torgeir Veimo [mailto:[email protected]] Gesendet: Mittwoch, 17. November 2010 14:49 An: [email protected] Betreff: Re: 'reverse path prefix' xpath query On 17 November 2010 23:40, Clemens Wyss <[email protected]> wrote: > Where are the subscribers located/persisted? > Beneath /blog/<category>/subscribers (e.g. /blog/wildlife/subscribers)? If so > you could take the node's path, apply the following regexp /(\/.*\/.*\/)/ on > it (->/blog/wildlife/) and then get the childnodes of the regexp result > concatenated with 'subscribers' (-> /blog/wildlife/subscribers)... So the only solution is to store the subscriber list local to the data being subscribed to? -- -Tor
