Sysl (pronounced "sizzle") is a system specification language. Using Sysl, you can specify systems, endpoints, endpoint behaviour, data models and data transformations. The Sysl compiler automatically generates sequence diagrams, integrations, and other views. It also offers a range of code generation options, all from one common Sysl specification.
The set of outputs is open-ended and allows for your own extensions. Sysl has been created with extensibility in mind and it will grow to support other representations over time.
Windows users can download the sysl-bundle-windows.zip, containing sysl.exe and reljam.exe, from our release page.
Users on other operating systems need to work with Python or Docker.
Install Python 2.7. If your specific environment causes problems you might find our guide helpful.
Install Sysl with
> pip install sysl
Now you can execute Sysl on the command line with
> sysl textpb -o out/petshop.txt /demo/petshop/petshop > reljam model /demo/petshop/petshop PetShopModel
See sysl --help and reljam --help for more options.
Install Docker and pull the Docker Image with
> docker pull anzbank/sysl
Consider tagging the docker image to make commands shorter
> docker tag anzbank/sysl sysl
Try the following commands
> docker run sysl > docker run sysl sysl -h > docker run sysl reljam -h
See https://hub.docker.com/r/anzbank/sysl/ for more details.
Install dependencies and the sysl package with symlinks
> pip install pytest flake8 -e .
Sysl depends upon PlantUML for diagram generation. Some of the automated tests require a PlantUML dependency. Provide PlantUML access either via local installation or URL to remote service. Warning, for sensitive data the public service at www.plantuml.com is not suitable. You can use one of the following options to set up your environment:
- execute
SYSL_PLANTUML=http://www.plantuml.com/plantuml - add
export SYSL_PLANTUML=http://www.plantuml.com/plantumlto you.bashrcor similar - install PlantUML locally and run on port 8080
Test and lint the source code and your changes with
> pytest > flake8
Consider using virtualenv and virtualenvwrapper to isolate your development environment.
For Java tests install Java 8 and gradle and run
> gradle test -b test/java/build.gradle
If your corporate environment restricts access to jcenter our environment guide might hold the answer for you. It also includes tips on using virtualenv with gradle test.
We encourage contributions to this project! Please have a look at the contributing guide for more information.
If you need to create a release follow the release documentation.
./run-travis.sh runs a local Travis CI build. This is intended primarily to
test Travis builds offline.
In order to easily reuse and extend Sysl across systems, the Sysl compiler translates Sysl files into an intermediate representation expressed as protocol buffer messages. These protobuf messages can be consumed in your favorite programming language and transformed to your desired output. In this way you are creating your own Sysl exporter.
Using the protoc compiler you can translate the definition file of the intermediate representation src/proto/sysl.proto into your preferred programming language in a one-off step or on every build. You can then easily consume Sysl models in your programming language of choice in a typesafe way without having to write a ton of mapping
boilerplate. With that you can create your own tailored output diagrams, source code, views, integrations or other desired outputs.
In this project, several Python based exporters exist under src/sysl/exporters and the relevant Python protobuf definitions sysl_pb2.py have been created from sysl.proto with
> protoc --python_out=src/sysl/proto --proto_path=src/proto sysl.proto
If sysl.proto is updated, the above command needs to be re-run to update the corresponding Python definitions in sysl_pb2.py.
Sysl is currently targeted at early adopters. The current focus is to improve documentation and usability, especially error messages and warnings.