Lists of GROUP_CONCAT are shown incorrectly

[graue70]

Take the following query for example:

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
SELECT ?predicate (GROUP_CONCAT(?alias) AS ?aliases) WHERE {
  { SELECT ?predicate
    WHERE { ?x ql:has-predicate ?predicate }
    GROUP BY ?predicate }
  ?entity ?tmp ?predicate .
  ?entity @en@rdfs:label|@en@skos:altLabel ?alias .
  FILTER REGEX(?predicate, "^<http://www\\.wikidata\\.org/prop/direct")
}
GROUP BY ?predicate
ORDER BY ASC(?predicate)

The aliases column shows an unedited (correct) tooltip. For the actual contents of the column, the first " and the last "@en is cut off, which leaves strings like ZDB ID"@en "ZDB identifier or in other words a list of elements separated by the string "@en ".

It would be preferable to iterate over all elements of the list (separated by a space) and cut off the beginning and end from every one of them instead of from the whole thing. A new separator would be needed in that case. Using semicolon, this would yield ZDB ID; ZDB identifier for the mentioned example.

This might lead to problems if the actual content contains the separator (e.g. semicolon), so it should be chosen carefully. I suggest ; or |.