The DOKK graph.

zPlus b1b6cecb47 Add "documentation" source.		5 years ago
fuseki	8c82fea3a1 Add configuration template for Fuseki.	5 years ago
query	b1b6cecb47 Add "documentation" source.	5 years ago
source	b1b6cecb47 Add "documentation" source.	5 years ago
.gitignore	b1b6cecb47 Add "documentation" source.	5 years ago
Makefile	b1b6cecb47 Add "documentation" source.	5 years ago
README.md	1e2385a388 README.md: Fix description and add list of dependencies.	5 years ago
dataset.ttl	b1b6cecb47 Add "documentation" source.	5 years ago
vocab.json	a7b43cebc3 Move community datasets to another repository. Here we keep only the	6 years ago

This repository contains the code to generate the DOKK graph using the data repository.

Requirements

Tools used for creating the graph:

java
jena

$ wget $ tar xzf ... $ mv jena-directory /usr/local/bin
Append PATH="$HOME/jena/bin:$PATH" to your shell: $ vim ~/.profile
jq
rapper
wget

SPARQL endpoint

These are the instructions to setup your own Fuseki server if you want to have a local copy of the DOKK database. You are not required to follow these steps unless you want your own local copy of the DOKK database.

Install Fuseki

Follow the instructions from the official website. to install Fuseki, either as a service or a standalone server. Also read Fuseki File System Layout.

Compile and install hdt-java libraries

    $ git clone https://github.com/rdfhdt/hdt-java.git

    Compile hdt-java
    --file: Specify a POM file as if we were executing Maven from that directory.
    $ mvn --file hdt-java/pom.xml clean install package

    Delete sources and only keep hdt-java .jar files
    $ cp hdt-java/hdt-api/target/hdt-api-*-SNAPSHOT.jar "${FUSEKI_HOME}/hdt-api.jar"
    $ cp hdt-java/hdt-java-core/target/hdt-java-core-*-SNAPSHOT.jar "${FUSEKI_HOME}/hdt-java-core.jar"
    $ cp hdt-java/hdt-jena/target/hdt-jena-*-SNAPSHOT.jar "${FUSEKI_HOME}/hdt-jena.jar"

    $ rm -rf hdt-java/

Customize Fuseki configuration

Copy fuseki/config.ttl and fuseki/log4j.properties into your $FUSEKI_BASE folder
Edit the startup script ${FUSEKI_HOME}/fuseki-server like this:
- Adjust JVM_ARGS=${JVM_ARGS:--Xmx1G} to increase the JVM memory. This is required to accomodate HDT indices. If the memory is too low, you'll get an error when starting the server
- Replace exec $JAVA $JVM_ARGS -jar "$JAR" "$@" with this Path to HDT jars to add to the class path. $ export FUSEKI_CLASSES="${FUSEKI_HOME}hdt-api.jar:${FUSEKI_HOME}hdt-java-core.jar:${FUSEKI_HOME}hdt-jena.jar" $ exec $JAVA $JVM_ARGS -classpath "$JAR:$FUSEKI_CLASSES" org.apache.jena.fuseki.cmd.FusekiCmd "$@"`

Run server

Create the fuseki user and group:

    $ sudo groupadd fuseki
    $ sudo useradd -g fuseki fuseki

As a standalone server

From $FUSEKI_HOME run ./fuseki-server. Make sure you have $FUSEKI_HOME and $FUSEKI_BASE defined in your environment.

As a service (systemd)

Follow instructions in fuseki.service file in the Fuseki repository. The server should be available at http://localhost:3030.

Add data (graphs) to the database

DOKK uses HDT files, so you can add data to the database simply by copying HDT files into ${FUSEKI_BASE}/databases/hdt/<graph-name>/. When you've copied all your HDT files, run ./write-configuration.sh to update the Fuseki configuration such that it will load the new graph at startup.

Contact

https://peers.community

contact@peers.community

#peers on freenode

README.md