Switch to using jelly-cli

Author: Ostrzyciel

Dockerfile

@@ -38,16 +38,8 @@ WORKDIR /app
 # Download robot.jar
 RUN wget https://github.com/ontodev/robot/releases/download/v1.9.7/robot.jar
 
-# Download Apache Jena
-# The path to Jena RIOT CLI tool is /app/jena/bin/riot
-RUN wget https://dlcdn.apache.org/jena/binaries/apache-jena-5.2.0.tar.gz -O apache-jena.tar.gz && \
-    tar -xzf apache-jena.tar.gz && \
-    rm apache-jena.tar.gz && \
-    mv apache-jena-* jena
-
-# Download the Jelly-JVM plugin for Jena
-RUN wget https://github.com/Jelly-RDF/jelly-jvm/releases/download/v2.3.0/jelly-jena-plugin.jar \
-    -O jena/lib/jelly-jena-plugin.jar
+# Install jelly-cli
+RUN bash -c "$(curl -sSfL https://w3id.org/jelly/setup-cli.sh)"
 
 # Copy virtual env from python-deps stage
 COPY --from=python-deps /.venv /app/.venv
@@ -56,4 +48,4 @@ ENV PATH="/app/.venv/bin:$PATH"
 # Install application into container
 COPY . .
 
-ENTRYPOINT []
+ENTRYPOINT []

README.md

@@ -4,7 +4,7 @@ CI worker for processing the RDF Stream Taxonomy (RDF-STaX).
 
 ## Usage
 
-Use the provided Docker image. The container includes the Python scripts, a JRE, the [ROBOT tool](http://robot.obolibrary.org/), and an [Apache Jena CLI distribution](https://jena.apache.org/documentation/tools/index.html) with the [Jelly-JVM plugin](https://w3id.org/jelly/jelly-jvm/dev/getting-started-plugins/) installed.
+Use the provided Docker image. The container includes the Python scripts, a JRE, the [ROBOT tool](http://robot.obolibrary.org/), and [jelly-cli](https://github.com/Jelly-RDF/cli) for converting between various RDF formats, including [Jelly](https://w3id.org/jelly).
 
 In the container you can invoke the commands:
 

tasks/nanopub_fetch.py

@@ -70,10 +70,10 @@ def main():
         d.parse(Path(cache_dir, source_file), format='trig')
     d.serialize(Path(output_dir, 'nanopubs.trig'), format='trig')
     d.serialize(Path(output_dir, 'nanopubs.nq'), format='nquads')
-    # Convert it to Jelly as well, using Apache Jena RIOT
+    # Convert it to Jelly as well, using jelly-cli
     with open(Path(output_dir, 'nanopubs.jelly'), 'wb') as f:
         subprocess.run([
-            str(parent_dir / 'jena/bin/riot'), '--stream=jelly', str(Path(output_dir, 'nanopubs.nq'))
+            'jelly-cli', 'rdf', 'to-jelly', str(Path(output_dir, 'nanopubs.nq'))
         ], check=True, stdout=f, stderr=subprocess.STDOUT)
 
     print(f'Writing download links...')

tasks/package.py

@@ -137,13 +137,13 @@ def serialize(g: Graph, output_dir: Path, filename: str, parent_dir: Path):
     g.serialize(destination=output_dir / f'{filename}.rdf', format='xml', encoding='utf-8')
     g.serialize(destination=output_dir / f'{filename}.jsonld', format='json-ld', encoding='utf-8')
     g.serialize(destination=output_dir / f'{filename}.nt', format='nt', encoding='utf-8')
-    # Convert it to Jelly as well, using Apache Jena RIOT
+    # Convert it to Jelly as well, using jelly-cli
     with open(output_dir / f'{filename}.jelly', 'wb') as f:
         subprocess.run([
-            str(parent_dir / 'jena/bin/riot'),
-            '--stream=jelly',
+            'jelly-cli', 'rdf', 'to-jelly',
             # Disable generalized RDF and RDF-star in the metadata, because we are not using that
-            '--set="https://ostrzyciel.eu/jelly/riot/symbols#preset=SMALL_STRICT"',
+            '--opt.rdf-star=false',
+            '--opt.generalized-statements=false',
             str(output_dir / f'{filename}.nt')
         ], check=True, stdout=f, stderr=subprocess.STDOUT)