Ooops, correction:
public static Dataset loadData() {
Dataset ds = DatasetFactory.createTxnMem();
Dataset indexDS = createIndex(ds);
FileUtil.loadDataGraphs(indexDS, dataGraphs);
return indexDS;
}
public static Dataset createIndex(Dataset ds) {
Resource res =
ResourceFactory.createProperty("http://ao.com/name/";);
EntityDefinition entDef = new EntityDefinition("uri", "text", res) ;
Directory dir = new RAMDirectory();
Dataset indexDataset = TextDatasetFactory.createLucene(ds, dir,
entDef, null);
return indexDataset;
}
On 4/7/17, 2:12 PM, "Dimov, Stefan" <[email protected]> wrote:
Thanks Andy,
I’ve changed the code:
public static Dataset loadData() {
Dataset ds = DatasetFactory.createTxnMem();
Dataset indexDS = createIndex(ds);
FileUtil.loadDataGraphs(indexDS, dataGraphs);
return ds;
}
public static Dataset createIndex(Dataset ds) {
Resource res =
ResourceFactory.createProperty("http://ao.com/name/";);
EntityDefinition entDef = new EntityDefinition("uri", "text", res) ;
Directory dir = new RAMDirectory();
Dataset indexDataset = TextDatasetFactory.createLucene(ds, dir,
entDef, null);
return indexDataset;
}
but after that, still text query returns empty result set.
What am I missing?
S.
On 4/7/17, 1:48 PM, "Andy Seaborne" <[email protected]> wrote:
> Dataset ds = loadData();
^^^^^^^^^^
Not shown.
> …
> Resource res =
ResourceFactory.createProperty("http://ao.com/name/";);
> EntityDefinition entDef = new EntityDefinition("uri", "text",
res) ;
> Directory dir = new RAMDirectory();
> Dataset indexDataset = TextDatasetFactory.createLucene(ds, dir,
entDef, null);
Stefan,
You have to load the data through "indexDataset" - I guess you are
loading it into a plain dataset in loadData.
The Lucene index is built as data is added to the dataset, not when the
dataset is wrapper later.
Andy
On 07/04/17 19:40, Dimov, Stefan wrote:
> Thanks Rob,
>
> The problem is, for someone like me, who is new in Jena, it’s hard to
follow the documentation – not enough examples and the explanations are kind of
ambiguous (again – from my point of view).
>
> I was hoping on comment and snippet, specifically about my example.
>
>
> S.
>
>
>
> On 4/7/17, 2:25 AM, "Rob Vesse" <[email protected]> wrote:
>
> What these lines are doing is creating a blank index, in
particular the last line simply associates a pre-existing index with a dataset
i.e. It doesn’t actually cause indexing to happen, typically you would first
build a persistent Lucene index using the command line tool
>
>
http://jena.apache.org/documentation/query/text-query.html#building-a-text-index
>
> if you want to explicitly built the index from code your best
bet is to refer to the code for the commandline tool:
>
>
https://github.com/apache/jena/blob/master/jena-text/src/main/java/jena/textindexer.java
>
> Rob
>
> On 07/04/2017 04:07, "Dimov, Stefan" <[email protected]> wrote:
>
> Resource res =
ResourceFactory.createProperty("http://ao.com/name/";);
> EntityDefinition entDef = new EntityDefinition("uri", "text",
res) ;
> Directory dir = new RAMDirectory();
> Dataset indexDataset = TextDatasetFactory.createLucene(ds,
dir, entDef, null);
>
>
>
>
>
>
