- From: Sebastian Hammer <quinn@indexdata.dk>
- Date: Mon, 22 Apr 2002 12:35:52 +0200
- To: Dan Brickley <danbri@w3.org>, Matthew Dovey <matthew.dovey@las.ox.ac.uk>
- Cc: Robert Sanderson <azaroth@liverpool.ac.uk>, <www-zig@w3.org>
At 06:31 22-04-2002 -0400, Dan Brickley wrote: > > I think my view (and this was something we investigated in some of the > > precursor work for SRW) would be that specifying an XSLT as an e-spec > > would be a better proposal that just specifying a list of XPATHs. This > > would also have the advantage of have toolkits available for XSLT > > transforms that could plug in. > >Isn't XSLT turing complete? queries could take a while to complete unless >there was some filtering put in place to subset things to sensible >XSLTs... (also we'd need to watch that extension functions weren't left >open). Oh, but watch out here! Matthew is talking about Element specifications, which is a transformation that is carried out on each retrieval record prior to sending it to the client. It has *no* bearing on the search. I agree about extension functions though. Please, let's keep this discussion a tad separate from the discussion about searching. --Sebastian -- Sebastian Hammer, Index Data <http://www.indexdata.dk/> Ph: +45 3341 0100, Fax: +45 3341 0101
Received on Monday, 22 April 2002 06:35:05 UTC