You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jena.apache.org by Daniel Maatari Okouya <ok...@yahoo.fr> on 2013/12/28 20:19:32 UTC

DataSet vs DataSource

I saw the previous answer to the question posted in 2011 about the difference between a dataSet and a dataSource. I would like to further clarify that point:

You have the TDB and the ARQ API both can create DataSet but only ARQ can create DataSource.

I want to use AQR to query my dataset/Source. Currently my dataSet is one that was created with the TDB API because i learned it before ARQ.

But in learning ARQ i saw this dataSource which they say is updatable. This is confusing to me for the following reasons.

I have been so far able programmatically to create NamedModels with TDB.DataSet. When it comes to the defaultGraph i can not set it. Nevertheless to add triples to it, i just get the defaultModel, and add(..) triples from another in memory model.

Hence i would like to know what would be the advantage of using, ARQ dataSet and most importantly DataSource ?

Many thanks,

Maatary


-- 
Daniel Maatari Okouya
Sent with Airmail

Re: DataSet vs DataSource

Posted by Andy Seaborne <an...@apache.org>.
On 28/12/13 19:19, Daniel Maatari Okouya wrote:
> I saw the previous answer to the question posted in 2011 about the
> difference between a dataSet and a dataSource. I would like to
> further clarify that point:
>
> You have the TDB and the ARQ API both can create DataSet but only ARQ
> can create DataSource.

DataSource no longer exists - all the functionality has been absorbed 
into Dataset.

> I want to use AQR to query my dataset/Source. Currently my dataSet is
> one that was created with the TDB API because i learned it before
> ARQ.
>
> But in learning ARQ i saw this dataSource which they say is
> updatable. This is confusing to me for the following reasons.
>
> I have been so far able programmatically to create NamedModels with
> TDB.DataSet. When it comes to the defaultGraph i can not set it.
> Nevertheless to add triples to it, i just get the defaultModel, and
> add(..) triples from another in memory model.

That will do fine.

You never have to add or create the default graph/model.

For named graphs, it's only storage systems (e.g. in-memory, or datasets 
are that combinations of models from different other storage systems) 
where when you add a graph you are also adding the storage.

In TDB, Dataset.getNamedModel(uri) will always return a Model you can 
update.  For TDB, addNamedModel is a copy of the data into the database.

	Andy

> Hence i would like to know what would be the advantage of using, ARQ
> dataSet and most importantly DataSource ?
>
> Many thanks,
>
> Maatary
>
>
> -- Daniel Maatari Okouya Sent with Airmail
>