From: <tom...@gm...> - 2015-01-07 13:05:04
|
Thanks! Indeed works like charm. Tom January 7 2015 1:52 PM, "Rob Vesse" <rv...@do...> wrote: > Tom > > This query hit a bug in our lazy evaluation implementation, turns out this > query is one that looks like it can be lazily evaluated but can't in > practise which causes us to do lazy evaluation anyway which is very slow > when fully evaluating a BGP > > I have already fixed this to instead fall back to normal BGP evaluation > when this cases are hit and the query now completes fine. > > As a workaround for previous versions you can disable lazy evaluation: > > SparqlOptimiser.RemoveOptimiser(SparqlOptimiser.AlgebraOptimisers.Find(x > => x.GetType().Equals(typeof(LazyBgpOptimiser))); > > Bonus question: > > FROM and FROM NAMED are only allowed by the SPARQL specification at the > top level as they set the dataset for the entire query (including > sub-queries) > > Rob > > On 06/01/2015 20:51, "tom...@gm..." > <tom...@gm...> wrote: > >> Hi Rob >> >> I've been having some problems with a SPARQL query. Actually I extracted >> it from a larger one, where it is a subquery. As a subquery I get some >> inifinite loop and must terminate the process. On its own it times out. I >> get the same results in all version from 0.7.2 onwards. >> >> prefix hydra: <http://www.w3.org/ns/hydra/core#> >> prefix dcterms: <http://purl.org/dc/terms/> >> >> select ?member ?title >> { >> graph <http://localhost:17899/brochures> >> { >> <http://localhost:17899/brochures> hydra:member ?member . >> ?member dcterms:title ?title >> } >> } >> order by ?title >> offset 0 >> limit 10 >> >> I've played with this query to identify the problem and I've got some >> results. It works fine in either one case: >> >> 1. Remove order >> 2. Remove limit >> 3. Replace GRAPH <> with FROM <> >> 4. Remove the first triple pattern >> 5. Replace two triple patterns with property path (and ditch ?member) >> >> Key observation however is a modification to 3. If I use FROM <> but wrap >> the triple pattern in a BGP I'm getting timeouts all the same. >> >> My dataset can be downloaded from http://ge.tt/19Q3B582/v/0?c >> >> As a bonus question, please tell me why is is not possible to use FROM >> (NAMED) in subqueries? SPARQL specification or library limitation? >> >> Thanks, >> Tom >> >> _______________________________ >> >> ---- >> Dive into the World of Parallel Programming! The Go Parallel Website, >> sponsored by Intel and developed in partnership with Slashdot Media, is >> your >> hub for all things parallel software development, from weekly thought >> leadership blogs to news, videos, case studies, tutorials and more. Take a >> look and join the conversation now. http://goparallel.sourceforge.net >> _______________________________ >> >> dotNetRDF-bugs mailing list >> dot...@li... >> https://lists.sourceforge.net/lists/listinfo/dotnetrdf-bugs > > _______________________________ > > Dive into the World of Parallel Programming! The Go Parallel Website, > sponsored by Intel and developed in partnership with Slashdot Media, is your > hub for all things parallel software development, from weekly thought > leadership blogs to news, videos, case studies, tutorials and more. Take a > look and join the conversation now. http://goparallel.sourceforge.net > _______________________________ > > dotNetRDF-bugs mailing list > dot...@li... > https://lists.sourceforge.net/lists/listinfo/dotnetrdf-bugs |