launch a tdbquery on multiple location

[Steven Blanchard <st...@microbiome.studio>]

Hello,

I would like to know if it is possible to perform a query with 
tdb2.tdbquery on multiple tdb folder?

It is possible to put several times the --loc argument in the command 
line but in the results, I have only the triple matches to the 2nd 
folder.

example with Order 1 :
```bash
apache-jena-4.7.0/bin/tdb2.tdbquery --loc Demo_2/ --loc Demo_1/ --query 
demo.sparql

?s	?p	?o
<<http://example.org/#spiderman>>	<<http://www.perceive.net/schemas/relationship/enemyOf>>	<<http://example.org/#green-goblin>>
```

example with Order 2:
```bash
apache-jena-4.7.0/bin/tdb2.tdbquery --loc Demo_1/ --loc Demo_2/ --query 
demo.sparql

?s	?p	?o
<<http://example.org/#spiderman>>	<<http://www.perceive.net/schemas/relationship/loveOf>> 
<<http://example.org/#lois_lane>>
```

demo.sparql:
```sparql
SELECT *
WHERE {?s ?p ?o}
```

I would like that the data available in both folder are merged to form 
a new default graph for the query.

Thanks,
Regards,

Steven

Re: launch a tdbquery on multiple location

[Andy Seaborne <an...@apache.org>]

On 21/01/2023 09:11, Steven Blanchard wrote:
> Hello,
> 
> I would like to know if it is possible to perform a query with 
> tdb2.tdbquery on multiple tdb folder?

It is not possible.

> 
> It is possible to put several times the --loc argument in the command 
> line but in the results, I have only the triple matches to the 2nd folder.

That should be an error.
Could you file an issue please?

> 
> example with Order 1 :
> ```bash
> apache-jena-4.7.0/bin/tdb2.tdbquery --loc Demo_2/ --loc Demo_1/ --query 
> demo.sparql
> 
> ?s    ?p    ?o
> <<http://example.org/#spiderman>>    
> <<http://www.perceive.net/schemas/relationship/enemyOf>>    
> <<http://example.org/#green-goblin>>
> ```
> 
> example with Order 2:
> ```bash
> apache-jena-4.7.0/bin/tdb2.tdbquery --loc Demo_1/ --loc Demo_2/ --query 
> demo.sparql
> 
> ?s    ?p    ?o
> <<http://example.org/#spiderman>>    
> <<http://www.perceive.net/schemas/relationship/loveOf>> 
> <<http://example.org/#lois_lane>>
> ```
> 
> demo.sparql:
> ```sparql
> SELECT *
> WHERE {?s ?p ?o}
> ```
> 
> I would like that the data available in both folder are merged to form a 
> new default graph for the query.

There is tdb2:unionDefaultGraph true ; for all named graphs in one 
database as the default graph.

You can also define a general dataset and include (union) graphs from 
several other storage units.

Another way is to use federated query (SERVICE)
https://jena.apache.org/documentation/query/service_enhancer.html

(The two TDB folders can't simply be used together because the internal 
ids were allocated in such a way that they will overlap).

     Andy

> 
> Thanks,
> Regards,
> 
> Steven
> 
>