是否可以使用SPARQL和RDF4J批处理选择查询？

我正在使用一个很大的数据集（大约500Mio-Triples）存储在graphDB Free中，并在我的本地开发人员计算机上运行。

我想对带有RDF4J的数据集进行一些操作，并且必须选择或多或少地选择整个数据集。要进行测试，我只需选择所需的元组。对于前百万个元组，该代码运行良好，此后，由于graphDB继续分配更多的RAM，它变得非常慢。

是否可以对非常大的数据集进行SELECT查询并批量获取它们？

基本上，我只想“迭代”某些选定的三元组，因此不必使用来自graphDB的那么多RAM。我可以看到我已经准备好在查询完成之前在RDF4J中获取数据，因为它仅在大约1.4 Mio读取元组时崩溃（HeapSpaceError）。不幸的是，graphDB不能以某种方式释放所有就绪读取元组的内存。我想念什么吗？

非常感谢你的帮助。

ps。我已经将graphDB的可用heapSpace设置为20GB。

RDF4J（Java）代码如下所示：

package ch.test;


import org.eclipse.rdf4j.query.*;
import org.eclipse.rdf4j.repository.RepositoryConnection;
import org.eclipse.rdf4j.repository.http.HTTPRepository;

import java.io.File;
import java.io.IOException;
import java.nio.charset.StandardCharsets;
import java.nio.file.Files;
import java.nio.file.Paths;

public class RDF2RDF {

    public static void main(String[] args) {
        System.out.println("Running RDF2RDF");

        HTTPRepository sourceRepo = new HTTPRepository("http://localhost:7200/repositories/datatraining");
        try {
            String path = new File("").getAbsolutePath();
            String sparqlCommand= Files.readString(Paths.get(path + "/src/main/resources/sparql/select.sparql"), StandardCharsets.ISO_8859_1);

            int chunkSize = 10000;
            int positionInChunk = 0;
            long loadedTuples = 0;

            RepositoryConnection sourceConnection = sourceRepo.getConnection();
            TupleQuery query = sourceConnection.prepareTupleQuery(sparqlCommand);

            try (TupleQueryResult result = query.evaluate()) {
                for (BindingSet solution:result) {
                    loadedTuples++;
                    positionInChunk++;

                    if (positionInChunk >= chunkSize) {
                        System.out.println("Got " + loadedTuples + " Tuples");
                        positionInChunk = 0;
                    }
                }
            }

        } catch (IOException err) {
            err.printStackTrace();
        }
    }
}

select.sparql：

PREFIX XXX_meta_schema: <http://schema.XXX.ch/meta/>
PREFIX XXX_post_schema: <http://schema.XXX.ch/post/>
PREFIX XXX_post_tech_schema: <http://schema.XXX.ch/post/tech/>

PREFIX XXX_geo_schema: <http://schema.XXX.ch/geo/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX XXX_raw_schema: <http://schema.XXX.ch/raw/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT * WHERE {

    BIND(<http://data.XXX.ch/raw/Table/XXX.csv> as ?table).

    ?row XXX_raw_schema:isDefinedBy ?table.

    ?cellStreetAdress XXX_raw_schema:isDefinedBy ?row;
        XXX_raw_schema:ofColumn <http://data.XXX.ch/raw/Column/Objektadresse>;
        rdf:value ?valueStreetAdress.

    ?cellOrt mobi_raw_schema:isDefinedBy ?row;
        XXX_raw_schema:ofColumn <http://XXX.mobi.ch/raw/Column/Ort>;
        rdf:value ?valueOrt.

    ?cellPlz mobi_raw_schema:isDefinedBy ?row;
        XXX_raw_schema:ofColumn <http://XXX.mobi.ch/raw/Column/PLZ>;
        rdf:value ?valuePLZ.

    BIND (URI(concat("http://data.XXX.ch/post/tech/Adress/", MD5(STR(?cellStreetAdress)))) as ?iri_tech_Adress).
}

我的解决方案：使用子选择statemant首先获取所有“行”。

PREFIX mobi_post_schema: <http://schema.mobi.ch/post/>
PREFIX mobi_post_tech_schema: <http://schema.mobi.ch/post/tech/>

PREFIX mobi_geo_schema: <http://schema.mobi.ch/geo/>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX mobi_raw_schema: <http://schema.mobi.ch/raw/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

SELECT * WHERE {

    {
        SELECT ?row WHERE
        {
            BIND(<http://data.mobi.ch/raw/Table/Gebaeudeobjekte_August2020_ARA_Post.csv> as ?table).

            ?row mobi_raw_schema:isDefinedBy ?table.
        }
    }


    ?cellStreetAdress mobi_raw_schema:isDefinedBy ?row;
        mobi_raw_schema:ofColumn <http://data.mobi.ch/raw/Column/Objektadresse>;
        rdf:value ?valueStreetAdress.

    ?cellOrt mobi_raw_schema:isDefinedBy ?row;
        mobi_raw_schema:ofColumn <http://data.mobi.ch/raw/Column/Ort>;
        rdf:value ?valueOrt.

    ?cellPlz mobi_raw_schema:isDefinedBy ?row;
        mobi_raw_schema:ofColumn <http://data.mobi.ch/raw/Column/PLZ>;
        rdf:value ?valuePLZ.

    BIND (URI(concat("http://data.mobi.ch/post/tech/Adress/", MD5(STR(?cellStreetAdress)))) as ?iri_tech_Adress).
}