| Nikki added a comment. |
In T197161#4306966, @MusikAnimal wrote:XTools uses wb_terms to report if a page is missing a label and description on Wikidata for the corresponding language. Performance is paramount so if SPARQL is indeed that slow (T197161#4302203) I don't think it would be an option for XTools. If we can efficiently query for it on the replicas via some other table(s) this is perfectly fine for us :)
Not all queries for terms in SPARQL are slow. The ones which are problematic are ones which involve checking a large number (i.e. millions) of terms to see if they match some criteria (which is not fast in SQL either :)). Because of the way the data is modelled in RDF, this includes queries which try to select all terms in a particular language even when there aren't many terms for that language: The text and the language are both part of the value and have to be extracted separately using functions (similar to how you might extract the year from a date), they're not indexed seperately. Selecting terms for small numbers of items should work fine (e.g. selecting English labels/descriptions for Q1 and Q2 and selecting English labels/descriptions for cats both take less than a second).
Cc: MusikAnimal, Daniel_Mietchen, Nikki, Magnus, Mahir256, Nikerabbit, GoranSMilovanovic, Lea_Lacroix_WMDE, WMDE-leszek, Aklapper, Lahi, Gq86, QZanden, LawExplorer, Wikidata-bugs, aude, Mbch331
_______________________________________________ Wikidata-bugs mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs
