[jira] [Commented] (JENA-2005) SPARQL does not correctly handle xsd:decimal datatypes

["Andy Seaborne (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/JENA-2005?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17240372#comment-17240372 ] 

Andy Seaborne commented on JENA-2005:
-------------------------------------

Matching triple patterns is by RDF term - not by value. 

{{"4"^^xsd:decimal}} does not match {{"4.0"^^xsd:decimal}} (it would under D-entailment but not normal triple matching).

The usual approach is represent decimals in canonical form and the Jena parsers provide this with 

{noformat}
    RDFParser.create().canonicalValues(true).source(...).parse(...);
{noformat}

of put a FILTER in the MINUS because comparison by "=" is value sensitive (and not by "seameTerm").

Note: The canonical representation of a decimal changed between XML Schema 1 and 1.1:
https://www.w3.org/TR/xmlschema11-2/#decimal

It is now "4.0".


> SPARQL does not correctly handle xsd:decimal datatypes
> ------------------------------------------------------
>
>                 Key: JENA-2005
>                 URL: https://issues.apache.org/jira/browse/JENA-2005
>             Project: Apache Jena
>          Issue Type: Bug
>          Components: Fuseki
>    Affects Versions: Jena 3.16.0
>            Reporter: Steve Baskauf
>            Priority: Minor
>
> In performing a federated query at the Wikidata Query Service ([https://query.wikidata.org/sparql),] I bound the object of the triple:
>  
> <http://www.wikidata.org/entity/Q97446840>  <http://www.wikidata.org/prop/direct/P2896>  "4"^^xsd:decimal.
>  
> as it was provided by the SERVICE endpoint. I also bound the value of the triple:
>  
> <http://www.wikidata.org/entity/Q97446840>  <http://www.wikidata.org/prop/direct/P2896>  "4.0"^^xsd:decimal.
>  
> which was present in the local dataset. (The value was originally loaded from Turtle as the number 4.0 (not a datatyped literal).
> When I performed a MINUS operation between the bound values from the SERVICE endpoint and the bound values from the local triplestore, this value should have been removed from the result set because according to the XML schema Recommendation [https://www.w3.org/TR/xmlschema-2/#decimal] both "4" and "4.0" are valid lexical representations ("If the fractional part is zero, the period and following zero(es) can be omitted.") of the same decimal number. However, Fuseki did not recognize the two values as identical, which is an error.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)