net.sansa-stack:sansa-query-spark_2.11

SANSA Query Library

License

License

Categories

Categories

Net
GroupId

GroupId

net.sansa-stack
ArtifactId

ArtifactId

sansa-query-spark_2.11
Last Version

Last Version

0.7.1
Release Date

Release Date

Type

Type

jar
Description

Description

SANSA Query Library
Project Organization

Project Organization

Smart Data Analytics (SDA) research group

Download sansa-query-spark_2.11

How to add to project

<!-- https://jarcasting.com/artifacts/net.sansa-stack/sansa-query-spark_2.11/ -->
<dependency>
    <groupId>net.sansa-stack</groupId>
    <artifactId>sansa-query-spark_2.11</artifactId>
    <version>0.7.1</version>
</dependency>
// https://jarcasting.com/artifacts/net.sansa-stack/sansa-query-spark_2.11/
implementation 'net.sansa-stack:sansa-query-spark_2.11:0.7.1'
// https://jarcasting.com/artifacts/net.sansa-stack/sansa-query-spark_2.11/
implementation ("net.sansa-stack:sansa-query-spark_2.11:0.7.1")
'net.sansa-stack:sansa-query-spark_2.11:jar:0.7.1'
<dependency org="net.sansa-stack" name="sansa-query-spark_2.11" rev="0.7.1">
  <artifact name="sansa-query-spark_2.11" type="jar" />
</dependency>
@Grapes(
@Grab(group='net.sansa-stack', module='sansa-query-spark_2.11', version='0.7.1')
)
libraryDependencies += "net.sansa-stack" % "sansa-query-spark_2.11" % "0.7.1"
[net.sansa-stack/sansa-query-spark_2.11 "0.7.1"]

Dependencies

compile (11)

Group / Artifact Type Version
net.sansa-stack : sansa-rdf-common_2.11 jar 0.7.1
net.sansa-stack : sansa-rdf-spark_2.11 jar 0.7.1
org.aksw.jena-sparql-api : jena-sparql-api-server-standalone jar 3.13.1-1
org.apache.tinkerpop : spark-gremlin jar 3.3.0
io.github.litmus-benchmark-suite : gremlinator jar 1.0
net.sansa-stack : sansa-datalake-spark_2.11 jar 0.1.2
org.apache.jena : apache-jena-libs pom 3.13.1
org.scala-lang : scala-library jar 2.11.12
it.unibz.inf.ontop : ontop-owlapi jar 3.0.0-beta-2
it.unibz.inf.ontop : ontop-system-sql-owlapi jar 3.0.0-beta-2
io.netty : netty-all jar 4.1.42.Final

provided (3)

Group / Artifact Type Version
org.apache.spark : spark-core_2.11 jar 2.4.4
org.apache.spark : spark-sql_2.11 jar 2.4.4
org.apache.spark : spark-graphx_2.11 jar 2.4.4

test (3)

Group / Artifact Type Version
org.aksw.bsbm : bsbm-core jar 3.1.4
org.scalatest : scalatest_2.11 jar 3.0.5
com.holdenkarau : spark-testing-base_2.11 jar 2.4.3_0.12.0

Project Modules

There are no modules declared in this project.

Archived Repository - Do not use this repository anymore!

SANSA got easier to use! All its code has been consolidated into a single repository at https://github.com/SANSA-Stack/SANSA-Stack

SANSA Query

Maven Central Build Status License Twitter

Description

SANSA Query is a library to perform SPARQL queries over RDF data using big data engines Spark and Flink. It allows to query RDF data that resides both in HDFS and in a local file system. Queries are executed distributed and in parallel across Spark RDDs/DataFrames or Flink DataSets. Further, SANSA-Query can query non-RDF data stored in databases e.g., MongoDB, Cassandra, MySQL or file format Parquet, using Spark.

For RDF data, SANSA uses vertical partitioning (VP) approach and is designed to support extensible partitioning of RDF data. Instead of dealing with a single triple table (s, p, o), data is partitioned into multiple tables based on the used RDF predicates, RDF term types and literal datatypes. The first column of these tables is always a string representing the subject. The second column always represents the literal value as a Scala/Java datatype. Tables for storing literals with language tags have an additional third string column for the language tag. Its uses Sparqlify as a scalable SPARQL-SQL rewriter.

For heterogeneous data sources (data lake), SANSA uses virtual property tables (PT) partitioning, whereby data relevant to a query is loaded on the fly into Spark DataFrames composed of attributes corresponding to the properties of the query.

SANSA Query SPARK - RDF

On SANSA Query Spark for RDF the method for partitioning an RDD[Triple] is located in RdfPartitionUtilsSpark. It uses an RdfPartitioner which maps a Triple to a single RdfPartition instance.

  • RdfPartition - as the name suggests, represents a partition of the RDF data and defines two methods:
    • matches(Triple): Boolean: This method is used to test whether a triple fits into a partition.
    • layout: TripleLayout: This method returns the TripleLayout associated with the partition, as explained below.
    • Furthermore, RdfPartitions are expected to be serializable, and to define equals and hash code.
  • TripleLayout instances are used to obtain framework-agnostic compact tabular representations of triples according to a partition. For this purpose it defines the two methods:
    • fromTriple(triple: Triple): Product: This method must, for a given triple, return its representation as a Product (this is the super class of all Scala tuples)
    • schema: Type: This method must return the exact Scala type of the objects returned by fromTriple, such as typeOf[Tuple2[String,Double]]. Hence, layouts are expected to only yield instances of one specific type.

See the available layouts for details.

SANSA Query SPARK - Heterogeneous Data Sources

SANSA Query Spark for heterogeneous data sources (data data) is composed of three main components:

  • Anlyser: it extracts SPARQL triple patters and groups them by subject, it also extracts any operation on subjects like filters, group by, order by, distinct, limit.
  • ِPlanner: it extracts joins between subject-based triple patter groups and generates join plan accordingly. The join order followed is left-deep.
  • Mapper: it access (RML) mappings and matches properties of a subject-based triples patter group against the attributes of individual data sources. If a match exists of every property of the triple pattern, the respective data source is declared relavant and loaded into Spark DataFrame. The loading into DataFrame is performed using Spark Connectors.
  • Executor: it analyses SPARQL query and generates equivalent Spark SQL functions over DataFrames, for SELECT, WHERE, GROUP-BY, ORDER-BY, LIMIT. Connection between subject-based triple pattern groups are translated into JOINs between relevant Spark DataFrames.

Usage

The following Scala code shows how to query an RDF file SPARQL syntax (be it a local file or a file residing in HDFS):

val spark: SparkSession = ...

val lang = Lang.NTRIPLES
val triples = spark.rdf(lang)("path/to/rdf.nt")


val partitions = RdfPartitionUtilsSpark.partitionGraph(triples)
val rewriter = SparqlifyUtils3.createSparqlSqlRewriter(spark, partitions)

val qef = new QueryExecutionFactorySparqlifySpark(spark, rewriter)

val port = 7531
val server = FactoryBeanSparqlServer.newInstance.setSparqlServiceFactory(qef).setPort(port).create()
server.join()

An overview is given in the FAQ section of the SANSA project page. Further documentation about the builder objects can also be found on the ScalaDoc page.

For querying heterogeneous data sources, refer to the documentation of the dedicated SANSA-DatLake component.

How to Contribute

We always welcome new contributors to the project! Please see our contribution guide for more details on how to get started contributing to SANSA.

net.sansa-stack
Semantic Analytics Stack - Open source platform for distributed data processing for large-scale RDF datasets

Versions

Version
0.7.1
0.6.0
0.5.0
0.4.0