jelly-cli

Fast and convenient CLI utility for working with the Jelly knowledge graph streaming protocol.

Quick start

If you are using Linux (x86_64, ARM64), macOS (ARM64), or Windows (x86_64), the recommended way run jelly-cli is to use a pre-built binary. Go to the releases page and download the binary built for your platform.

You can then run it like so:

$ chmod +x jelly-cli
$ ./jelly-cli --help

Usage

Convert RDF to Jelly

To convert an RDF file (e.g., Turtle) to Jelly, simply run:

$ ./jelly-cli rdf to-jelly input.ttl > output.jelly

Convert Jelly to RDF

To convert from Jelly to RDF run:

$ ./jelly-cli rdf from-jelly input.jelly > output.nq

By default, jelly-cli will translate files to NQuads. But you can also specify the output format with --out-format, for example:

$ ./jelly-cli rdf from-jelly input.jelly --out-format=ttl > output.ttl

You can specify most well-known formats supported by Apache Jena, but also a custom Jelly-Text format. Jelly-Text is a human-readable translation of Jelly binary. It's not meant for machine consumption. It is useful for debugging and inspecting Jelly files.

Transcode Jelly files

The rdf transcode command turns one or more input Jelly streams into a single output stream. It's extremely fast, using a dedicated transcoding algorithm. However, the numerical values for each of the options in the output stream must be greater than or equal to those in the input stream(s).

$ ./jelly-cli rdf transcode input.jelly > output.jelly

Inspect Jelly files

To inspect a Jelly file and get basic information describing its contents, such as stream options or number of triples in the file, run

$ ./jelly-cli rdf inspect input.jelly

You can also compute the statistics separately for each stream frame with the --per-frame option:

$ ./jelly-cli rdf inspect input.jelly --per-frame

In both cases, you will get the output as a valid YAML.

Validate Jelly files

To validate a Jelly file, run

$ ./jelly-cli rdf validate input.jelly

You can also check whether the Jelly file has been encoded using specific stream options or is equivalent to another RDF file, with the use of additional options to this command.

General tips

Use the --help option to learn more about all the available settings:

$ ./jelly-cli rdf to-jelly --help
$ ./jelly-cli rdf from-jelly --help
$ ./jelly-cli rdf transcode --help
$ ./jelly-cli rdf inspect --help
$ ./jelly-cli rdf validate --help

And use the --debug option to get more information about any exceptions you encounter.

Alternative installation: JAR builds

If for some reason the binaries wouldn't work for you, or you want to get better performance for large files (see Performance considerations below), you can use the JAR build. This build runs on any platform, as long as you have Java (min. version 17). Go to the releases page and download the jelly-cli.jar file. Then, run it like so:

java -jar jelly-cli.jar --help

For most use cases, we recommend using the binary distribution, because it has way faster startup times and doesn't require you to install Java.

Performance considerations

If you are bulk-converting large amounts of RDF data (>10M triples), you may want to use the JAR build instead of the pre-compiled binary. With the JAR, your JVM will perform just-in-time compilation, resulting in a better optimized code. The application will take longer to start, but the overall throughput will be better for large files.

For maximum performance, we recommend using a recent JVM (e.g., GraalVM or OpenJDK). In some cases we saw up to 4x better throughput when running the JAR with GraalVM 24 in JIT mode, as compared to the pre-compiled binary.

GitHub Action

If you're using jelly-cli in your GitHub Action CI/CD workflows, consider using the jelly-rdf/setup-cli action (Marketplace link). The action will automatically download and install the appropriate binary. Simply run the action before jelly-cli usage:

steps:
  - uses: Jelly-RDF/setup-cli@v1
    
  - run: jelly-cli rdf to-jelly input.ttl > output.jelly

Developer notes

Run sbt fixAll before committing. Your code should be formatted and free of warnings. The CI checks will not pass if this is not the case.

Building from source

Ahead-of-time compilation (single binary)

• Ensure you have GraalVM installed and the native-image utility is available in your PATH.
• Clone the repository.
• Run sbt GraalVMNativeImage/packageBin
• The binary will be available at ./target/graalvm-native-image/jelly-cli.

Über-JAR build (just-in-time)

• Run sbt assembly
• The resulting JAR will be in ./target/scala-3.*.*/jelly-cli-assembly-*.jar
• Run it like: java -jar <path-to-jar>

---

The development of the Jelly protocol, its implementations, and supporting tooling was co-funded by the European Union. More details.