[
https://issues.apache.org/jira/browse/JENA-721?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14040759#comment-14040759
]
Andy Seaborne commented on JENA-721:
------------------------------------
There is a difference in the kind of validation going on; it can be layered.
Validation of RDF and RDF terms (and RDFS) can be done on the input stream.
There is then validation that might be required that is whether the RDF "shape"
is as expected for the application domain that can't be stream processed.
W3C are quite likely to start a working group on the shape validation
http://lists.w3.org/Archives/Public/public-rdf-shapes/
Is your work aligned with that?
> Inline literals, source types are discarded
> -------------------------------------------
>
> Key: JENA-721
> URL: https://issues.apache.org/jira/browse/JENA-721
> Project: Apache Jena
> Issue Type: Improvement
> Components: TDB
> Affects Versions: TDB 1.0.1
> Reporter: Marek Kowalczyk
> Priority: Minor
> Attachments: JENA-721.patch
>
>
> NodeId.inline$ changes the actual type of literals from subtypes of
> xsd:integer to xsd:integer, for instance: literals of type
> xsd:positiveInteger are stored ad inline type INTEGER and during read decoded
> as xsd:integer in NodeId.extract(NodeId)
> {code}
> case INTEGER:
> {
> long val = IntegerNode.unpack(v) ;
> Node n = NodeFactory.createLiteral(Long.toString(val), null,
> XSDDatatype.XSDinteger) ;
> return n ;
> }
> {code}
> It would be nice to add support for various xsd:types and use the 7 bits
> reserved for inline type to represent more than 7 types.
> As a fastest workaround I've disabled the inline literals in NodeId( using
> reflection) , but it would be great If it would be configurable via context
> or parameter in StoreConnection.make(); or via SystemParams
--
This message was sent by Atlassian JIRA
(v6.2#6252)
