The latest jar copy of SANSA core on Maven central may have a problem, specifically with the rdf_loader.conf file.
Exception in thread "main" com.typesafe.config.ConfigException$Parse: rdf_loader.conf @ jar:file:/Users/amadkour/.m2/repository/net/sansa-stack/sansa-rdf-spark-core_2.11/0.3.0/sansa-rdf-spark-core_2.11-0.3.0.jar!/rdf_loader.conf: 5: Key 'rdf.parser.skipinvalid' may not be followed by token: end of file
Iv checked the Github repo version of the configuration file and it seems to be correct (on the development branch):
rdf.parser.skipinvalid = true
while in my maven-downloaded local jar is as follows:
Essentially, its missing the assignment "= true" and that's why the ConfigParser is complaining.
This error can be re-created by running the following two lines
val lang: Lang = Lang.NTRIPLES
val triplesDataset = spark.read.rdf(lang)(Metadata.datasetPath)
I think the solution is just to update the maven central Jars for the master version.
This is the link for the master version problem:
https://github.com/SANSA-Stack/SANSA-RDF/blob/bbb5838dedb7137e03dd130deb1ae6d38b200f05/sansa-rdf-spark-parent/sansa-rdf-spark-core/src/main/resources/rdf_loader.conf#L5
Iv created a PR to fix this issue #29.
The latest jar copy of SANSA core on Maven central may have a problem, specifically with the rdf_loader.conf file.
Iv checked the Github repo version of the configuration file and it seems to be correct (on the development branch):
while in my maven-downloaded local jar is as follows:
Essentially, its missing the assignment "= true" and that's why the ConfigParser is complaining.
This error can be re-created by running the following two lines
I think the solution is just to update the maven central Jars for the master version.
This is the link for the master version problem:
https://github.com/SANSA-Stack/SANSA-RDF/blob/bbb5838dedb7137e03dd130deb1ae6d38b200f05/sansa-rdf-spark-parent/sansa-rdf-spark-core/src/main/resources/rdf_loader.conf#L5
Iv created a PR to fix this issue #29.