Gang, I am working on the Named Query support and need a bit of feedback.
I have a bit of a problem whether a "Named Native Query" should be a "full", "modifiable" or "sub" query of the underlying language. "full" --> The query is written exactly as to be executed, and will not be altered in any way. The query author is responsible for correctness and developer responsible to use it in the right manner. If this is chosen, then it is no point that the QueryBuilderFactory returns a Query<type> from a newNamedQuery() method call, but a Iterable<type> since all the access within the Query are not relevant. "modifiable" --> The query is written as to be executed, but since org.qi4j.api.query.Query is a representation of a re-executable query with 'variables', 'first', 'max' and 'orderBy' alterations possible, I could locate those clauses in the query and replace the values. There seems to be no support in Sesame for the equivalent of PreparedStatement (in SQL) which can be re-used over and over again with different variables. This is clearly the most complicated case, since the 'replacer' may misunderstand the query, unless a full SparQL parser is made. "semi" --> The query is only written as 'variableList', 'datasetClause' and 'whereClause', whereas the so called 'solutionModifier' is created via Query execution. 'solutionModifier' must come at the end, making this fairly easy. 'dataSet' refers to the RDF sources, and would allow NAMED sources other than the actual Repository used for Qi4j indexing. Anybody has any thoughts on this? Cheers -- Niclas Hedhman, Software Developer http://www.qi4j.org - New Energy for Java I live here; http://tinyurl.com/2qq9er I work here; http://tinyurl.com/2ymelc I relax here; http://tinyurl.com/2cgsug _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
