- From: Arnold deVos <adv@langdale.com.au>
- Date: Fri, 8 Oct 1999 09:43:27 +1000
- To: <www-rdf-interest@w3.org>
- Cc: <franklin.reynolds@nokia.com>
From: <franklin.reynolds@nokia.com> > Is the query service intended to be similar to SQL? No, there is no query language as such. Each method in the query interface is a canned query, parameterized by the method arguments. > > Are the functions get_child_values() and get_descendent_values() > the primary methods for traversing a graph? > More or less. Single-valued properties are traversed by get_values(), many-valued properties are traversed by get_child_values(). The word "child" is perhaps missused, since we are not traversing a tree. The get_extent_values() method traverses whatever the inverse property of rdf:type is. The get_descendent_values() method is a short form for repeated get_child_values() invokations. The interface is a compromise between simplicity and efficiency. In particular, we kept the fixed cost of each call in mind (e.g network latency). It must be possible for an application to obtain a reasonable amount of information in each operation. Thus each method returns values for many properties and, usually, many resources. I hope that helps, Arnold. --- Arnold deVos Langdale Consultants adv@langdale.com.au
Received on Thursday, 7 October 1999 19:46:09 UTC