[ 
https://issues.apache.org/jira/browse/JENA-2282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17492893#comment-17492893
 ] 

Nicholas commented on JENA-2282:
--------------------------------

> A related feature is server query templates

Yes, I agree that query storage is a necessary first step to query templating 
which would be another level of functionality altogether since it would allow 
for parameterised calls. 

> For storage, would an RDF dataset in the server be suitable?

Would the storage of queries be able to follow a pattern elsewhere established 
in Fuseki? I know config is saved in RDF files but is it loaded into a Named 
Graph somewhere secretly too? I don't know the internals of Fuseki/Jena to 
know. I think back to previous work of mine with MySQL where all config for any 
databases created is just stored in another admin DB. Could it be the case here 
where stored queries (and an ID for them, name, description and perhaps 
eventually input vars for templating) are all stored as RDF in a Dataset that 
is loaded into an administrative Named Graph?

If we could store queries in a special Named Graph, I would later propose to 
store other things in other special NGs, such as NGs that are closures of other 
NGs by way of importing them. This is really just a convenience around SPARQL 
queries using FROM NAMED but it's very useful when you have lots (50+) NGs.

Happy to work out the pattern of storage in line with existing patterns.

 

> Fuseki2 Query Store
> -------------------
>
>                 Key: JENA-2282
>                 URL: https://issues.apache.org/jira/browse/JENA-2282
>             Project: Apache Jena
>          Issue Type: Wish
>          Components: Fuseki UI
>    Affects Versions: Jena 4.4.0
>            Reporter: Nicholas
>            Priority: Minor
>   Original Estimate: 72h
>  Remaining Estimate: 72h
>
> Many triplestore applications have a way to store SPARQL queries. These sorts 
> of application parts are really useful: you can keep coming back to useful 
> queries. If the queries can be named, then you can build up a query library.
> Not hard to make and super useful. Unless there is already an extension for 
> this or plans for such, I'm happy to (have my staff) give this a go!



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

Reply via email to