Error processing query on imdb index

[Stiksels]

Context

• Trying to retrieve imdb movie id's based on the event titles in our dataset
• it looks like the qlever server interprets the escape character \ in a imdb movie title and throws an error

Query

PREFIX imdb: <https://www.imdb.com/>
SELECT ?event ?udb_title ?cp_type ?cp_thema ?imdb_movie ?imdb_title WHERE {
  {
    SELECT DISTINCT ?event ?udb_title ?cp_type ?cp_thema { 
      ?event <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.cidoc-crm.org/cidoc-crm/E7_Activity> .
      ?event @nl@<http://purl.org/dc/terms/title> ?udb_title . 
      ?event <http://purl.org/dc/terms/type> ?udb_type FILTER(?udb_type = <https://taxonomy.uitdatabank.be/terms/0.50.6.0.0> ) .
      ?udb_type ^<http://www.w3.org/2002/07/owl#sameAs> ?cp_type .
      OPTIONAL {
        ?event <https://data.vlaanderen.be/ns/cultuurparticipatie#thema> ?udb_thema .
        ?udb_thema ^<http://www.w3.org/2002/07/owl#sameAs> ?cp_thema .
      }
    }
    LIMIT 10
  }
  MINUS {
    ?event <http://www.w3.org/ns/prov#invalidatedAtTime> ?deleted . 
    }
    SERVICE <https://qlever.cs.uni-freiburg.de/api/imdb> {
      SELECT ?imdb_movie ?imdb_title WHERE {
        ?imdb_movie  imdb:title ?imdb_title .
        ?imdb_movie imdb:numVotes ?votes .
        ?imdb_movie imdb:averageRating ?rating .
#        FILTER CONTAINS(?imdb_title, "De Buurtpolitie: De Perfecte Overval")
        FILTER (CONTAINS(LCASE(STR(?imdb_title)), LCASE(STR(?udb_title))))
     }
    }
}
GROUP BY ?event ?udb_title ?cp_type ?cp_thema ?imdb_movie ?imdb_title

curl request

curl -s https://qlever-api.publiq.services/uit-kg/ -H "Accept: text/tab-separated-values" -H "Content-type: application/sparql-query" --data "PREFIX imdb: <https://www.imdb.com/> SELECT ?event ?udb_title ?cp_type ?cp_thema ?imdb_movie ?imdb_title WHERE { { SELECT DISTINCT ?event ?udb_title ?cp_type ?cp_thema { ?event <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.cidoc-crm.org/cidoc-crm/E7_Activity> . ?event @nl@<http://purl.org/dc/terms/title> ?udb_title . ?event <http://purl.org/dc/terms/type> ?udb_type FILTER(?udb_type = <https://taxonomy.uitdatabank.be/terms/0.50.6.0.0> ) . ?udb_type ^<http://www.w3.org/2002/07/owl#sameAs> ?cp_type . OPTIONAL { ?event <https://data.vlaanderen.be/ns/cultuurparticipatie#thema> ?udb_thema . ?udb_thema ^<http://www.w3.org/2002/07/owl#sameAs> ?cp_thema } } LIMIT 10 } MINUS { ?event <http://www.w3.org/ns/prov#invalidatedAtTime> ?deleted } SERVICE <https://qlever.cs.uni-freiburg.de/api/imdb> { SELECT ?imdb_movie ?imdb_title WHERE { ?imdb_movie imdb:title ?imdb_title . ?imdb_movie imdb:numVotes ?votes . ?imdb_movie imdb:averageRating ?rating . FILTER (CONTAINS(LCASE(STR(?imdb_title)), LCASE(STR(?udb_title)))) } } } GROUP BY ?event ?udb_title ?cp_type ?cp_thema ?imdb_movie ?imdb_title"

Error message

Error processing query
Assertion `false` failed. Unsupported escape sequence found in RDF literal "Re \ Entry". Please report this to the developers. In file "/app/src/parser/RdfEscaping.cpp " at line 144

[joka921]

Hi @Stiksels I a just merged the fix for this, it will take some time until the new version is rolled out in our public demos. Thanks again for reporting this subtle bug.

[Stiksels]

hi @joka921 , is this new version available in the public demo at SERVICE https://qlever.cs.uni-freiburg.de/api/imdb ? I still get the same error as mentioned above:

Assertion `false` failed. Unsupported escape sequence found in RDF literal "Re \ Entry". Please report this to the developers. In file "/app/src/parser/RdfEscaping.cpp " at line 144

[joka921]

Hi @Stiksels The bug was on the receiving side of the SERVICE clause. So if you post the queries on a QLever instance that you host yourself (it looks like it, I don't recognize your dataset), then you have to update your own QLever to actually benefit from this. If you start Qlever via Docker, a docker pull ad-freiburg/qlever should do the trick. Unfortunately the qlever-control script (pip install qleverl) currently doesn't automatically update the QLever binaries/Docker containers. This is something where we are currently discussing how we want to keep these tools in sync.