At the moment, I do not parse semicolon-separated values, but store them as strings. In theory, it should be possible to split them into URIs and use them for matching, but it will be painful. On the other hand, it should be fairly easy to parse them during upload - RDF DBs work well with multiple values per statement. I will adjust my parser and next time I do a big import, I will update it. But I am still looking for a server to host it, so might take a bit of time.
SPARQL language took me a bit to understand, until I went back and simply read the very few basics - then it all made sense all of a sudden. At the end, its just a single giant table with 3 columns. :) On Mon, Sep 18, 2017 at 3:44 AM, Jo <winfi...@gmail.com> wrote: > That is wonderful news! It will take a while before I get used to that > query language though. > > Does it also work if an object has a semicolon separated list of wikidata > items in for example subject:wikidata? A statue with more than one person > in it, for example? > > Polyglot > > 2017-09-18 7:28 GMT+02:00 Yuri Astrakhan <yuriastrak...@gmail.com>: > >> The "not yet fully named" service is now accessible directly from JOSM - >> just like OT. Simply install or update Wikipedia plugin, and it will show >> up in the download data screen (expert mode). >> >> Documentation: >> https://wiki.openstreetmap.org/wiki/Wikidata%2BOSM_SPARQL_qu >> ery_service#Using_from_JOSM >> >> _______________________________________________ >> talk mailing list >> talk@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/talk >> >> >
_______________________________________________ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
