On Oct 16, 2014, at 7:33 AM, Andy Seaborne <[email protected]> wrote: > On 15/10/14 04:18, Ian Emmons wrote: >> At [1], the ARQ documentation describes how to create your own filter and >> property functions and have those dynamically loaded. That page also says, >> in the first paragraph, “ARQ also has syntactic extensions for GROUP BY ... >> these are described elsewhere.” Where are they described? The location >> isn’t linked from that sentence, and I couldn’t find it. >> >> What I am looking for is this: Is there a mechanism, similar to that for >> filter and property functions, to define and dynamically load my own >> aggregate functions in ARQ? >> >> Thanks, >> >> Ian >> >> [1] https://jena.apache.org/documentation/query/extension.html > > Ian, > > It is possible but hard to add aggregates to ARQ. What aggregates were you > wanting? > > That text is pre-SPARQL 1.1 and refers to the fact ARQ had GROUP BY and > sub-SELECT at all. I've removed it as it is out of date. > > Andy
Andy, Thanks for letting me know. I was asking on behalf of a friend who wanted to create custom aggregates for his own use case. In other words, not the sort of thing that you would want in ARQ itself. I’m not sure what his real goal is, but the example he gave is an aggregate that would compute the maximum score of a grouping of hotels based on multiple properties of the hotels. That isn’t actually what he wants, but he was being a bit cagey about his true goals. Perhaps a better example (that is harder to do with the existing facilities of SPARQL) would be an aggregate that computes the standard deviation of a grouping. Anyway, I was hoping that there was a clever trick with the “java:" URL scheme that would enable him to inject his own Java aggregator implementation into a running ARQ processor, but it would seem that this is not supported. Thanks, Ian
