Aldo, I've reproduced the bug, will fix it. Thank you for the report (and for your patience while I was unable to understand the problem).
Best Regards, Ivan Mikhailov OpenLink Software http://virtuoso.openlinksw.com On Thu, 2009-10-15 at 06:30 -0300, Aldo Bucchi wrote: > Hi Ivan, > > On Thu, Oct 15, 2009 at 6:08 AM, Ivan Mikhailov > <imikhai...@openlinksw.com> wrote: > > Hello Aldo, > > > >> Isn't there any way that the compiler can detect which variables are > > being referenced in the select projections part? > > > > It detects, but you write "select ?text", not "select ?s ?text" in > > subqueries. > > But try it: > > select distinct ?text where { > { select ?text where {?s rdfs:label ?text} } union { select ?text > where {?s rdfs:comment ?text} } > } limit 10 > > You'll get an error as well: > http://bit.ly/3vNARm > > > > > Like in SQL, Datalog and other similar languages, variables inside > > subexpressions are not in the surrounding scope, you should place ?s to > > the inner SELECT result-list to refer to it at the upper level. > > Similarly, ?text is usable not because it appears in { ?s > > rdfs:label ?text } but because it is found after SELECT. It could be, > > say, > > { select (?inner_s) as ?s (?inner_text) as ?text > > where {?inner_s rdfs:label ?inner_text} } > > to provide values for ?s and ?text even if there are no such variables > > in subquery at all. > > Yes, this part is clear. > The way I see it is that all variables projected ( select'ed ) by the > inner expression become "potentially" available for the outer > expression. > This only happens IF the outer expression binds them as well in the > conditions clause. ( it doesn't suffice for it to "select" them ). > > It is that IF that makes me mad every now and then. > > > > > Next version of SPARQL compiler may contain special check for > > "namesakes" in different scopes, to keep the author warned. > > The author says: "All warnings are welcome!" ;) > > Regards, > A > > > > > Best Regards, > > Ivan. > > > > On Thu, 2009-10-15 at 05:33 -0300, Aldo Bucchi wrote: > >> Hi, > >> > >> There is a general caveat when building SPARQL Unions of Subqueries in > >> Virtuoso which I believe is unnecessary. > >> > >> The following query works fine. > >> > >> select distinct ?s ?text where { > >> { ?s rdfs:label ?text } union { ?s rdfs:comment ?text } > >> } limit 10 > >> > >> But assume I want to complicate it a bit by using subqueries ( in this > >> case it makes no sense, but there are definitely use cases for this, > >> like combing results of complex queries ) > >> > >> select distinct ?s ?text where { > >> { select ?text where {?s rdfs:label ?text} } union { select ?text > >> where {?s rdfs:comment ?text} } > >> } limit 10 > >> > >> The latter query will yield an error: > >> > >> 37000 Error SP031: SPARQL compiler: Variable 's' is used in the query > >> result set but not assigned > >> > >> So, it follows that I need to bind the variables at least once in the > >> outer pattern. > >> Fine, I can do that: > >> > >> select distinct ?s ?text where { > >> ?s ?p ?text . > >> { select ?text where {?s rdfs:label ?text} } union { select ?text > >> where {?s rdfs:comment ?text} } > >> } limit 10 > >> > >> And it works ;) > >> > >> But, here's the observation: Why do I HAVE to declare the variables > >> outside? > >> Isn't there any way that the compiler can detect which variables are > >> being referenced in the select projections part? > >> > >> When dealing with complex queries I usually find myself writing dummy > >> BGPs, for example: > >> > >> ?interestingSubject ?xx1 ?xx2 . > >> ?xx3 ?xx4 ?interestingObject . > >> { subquery1... } union { subquery2...} union {subquery3...} > >> > >> Will enabling this "automatically" introduce ambiguity? > >> > >> Regards, > >> A > >> > >> > > > > > > > > -- > Aldo Bucchi > skype:aldo.bucchi > http://www.univrz.com/ > http://aldobucchi.com/ > > PRIVILEGED AND CONFIDENTIAL INFORMATION > This message is only for the use of the individual or entity to which it is > addressed and may contain information that is privileged and confidential. If > you are not the intended recipient, please do not distribute or copy this > communication, by e-mail or otherwise. Instead, please notify us immediately > by > return e-mail. > > ------------------------------------------------------------------------------ > Come build with us! The BlackBerry(R) Developer Conference in SF, CA > is the only developer event you need to attend this year. Jumpstart your > developing skills, take BlackBerry mobile applications to market and stay > ahead of the curve. Join us from November 9 - 12, 2009. Register now! > http://p.sf.net/sfu/devconference > _______________________________________________ > Virtuoso-users mailing list > Virtuoso-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/virtuoso-users
