[
https://issues.apache.org/jira/browse/JENA-1544?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16465998#comment-16465998
]
Andy Seaborne commented on JENA-1544:
-------------------------------------
There is one other case - the data is in the query through `BIND` or `VALUES`.
The command line SPARQL execution tool puts in an empty dataset if there is no
dataset description and no data, making `VALUES` etc.
For programmatic use, I propose that "no data, no dataset description" is
treated as an error. The application can provide an empty dataset, and iot
probably meant to provide a dataset with data.
The command line tool works the way it does for simple examples and quick
experiments (see stackoverflow).
> FROM handling is inconsistent
> -----------------------------
>
> Key: JENA-1544
> URL: https://issues.apache.org/jira/browse/JENA-1544
> Project: Apache Jena
> Issue Type: Bug
> Affects Versions: Jena 3.7.0
> Reporter: Andy Seaborne
> Assignee: Andy Seaborne
> Priority: Major
> Fix For: Jena 3.8.0
>
>
> Dataset description (FROM and FROM NAMED) handling seems to be missing for
> memory datasets.
> For persistent ones, the action is to take graph from the dataset itself.
> The proposal is to add this for memory datasets.
> The ability to load graphs from the web or from files using FROM / FROM NAMED
> will be available when a query is executed with no dataset.
> Executing a query with no dataset and no dataset description is an error.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
