README.rst

Python-Capellambse

A Python 3 headless implementation of the Capella modeling tool

Intro

capellambse allows you reading and writing Capella models from Python without Java or the Capella tool on any (reasonable) platform. We wanted to "talk" to Capella models from Python, but without any Java on the way. We thought this project will help individuals and organisations getting through the MBSE adoption journey with Capella faster and so we made it public and open-source.

With capellambse you can access all (or almost all) Capella model elements, and render diagrams as SVG and PNG. We made it for automation of Systems Engineering work, so it integrates nicely into most CI/CD toolchains. We also found it at the core of our artifact generation pipelines (model to documents, model to SW interfaces).

The library works with PVMT and Requirements extensions without any additional efforts.

It started as a basic library somewhere mid 2019. Since then it was re-architected a few times and now has a full read/write capability for most of the present Capella ontology. We are continuously improving the API (introducing shortcuts), increasing the meta-model coverage and have more engineering automations and improvements in the pipeline to share.

Documentation and examples

The library is designed to be easy to use and discover, especially in an interactive environment such as JupyterLab. Additionally, API documentation is automatically generated and published whenever new features and bug fixes are added.

You are encouraged to explore our test models and demo notebooks. Click on the button below to launch a Jupyter notebook server on the public myBinder service, and get started in seconds:

Warning: Do not enter confidential information, such as passwords for non-public models, into a notebook hosted on myBinder. If you want to try out capellambse with those models, please install and run it in a local, trusted environment!

The docs/source/examples directory contains several hands-on example notebooks that you can immediately run and start experimenting with. Below is a short summary of each notebook's goal. If you are in the JupyterLab environment, you can click the notebook names to directly open them in a new lab tab. On Github, you will be shown a statically rendered preview of the notebook.

• 01 Introduction.ipynb provides a high-level overview of the library features. It visualizes examples like a Component - Function allocation table by leveraging Jupyter's and IPython's rich display functionality.
• 02 Intro to Physical Architecture.ipynb explores some more advanced concepts on the example of the Physical Architecture Layer. It shows how to derive tabular data, like a Bill of Materials or a Software to Hardware allocation table, by using pandas dataframes.
• 03 Data Values.ipynb shows how the API can be used to explore classes, class instances and other objects related to data modeling.
• 04 Intro to Jinja templating.ipynb demonstrates how to effectively combine capellambse with the powerful Jinja templating engine. This enables the creation of all sorts of model-derived documents and artifacts, including interactive web pages, PDF documents and any other textual representations of your models.
• 05 Introduction to Libraries.ipynb shows how to use Capella Library Projects within capellambse. In this example you'll learn how the API can be used to open a project that is based on a library and find objects in both models.
• 06 Introduction to Requirement access and management.ipynb shows how the API can be used to work with requirements objects, introduced by the Capella Requirements Viewpoint. In this example you'll see how to find requirements in the model, see which objects requirements are linked / traced to and even export requirements to Excel or ReqIF formats.
• 07 Code Generation.ipynb shows how to generate code from class diagrams. In particular, we focus on Interface Descriptive Languages with concrete examples for Class to ROS2 IDL and Google Protocol Buffers. We also show how simple Python stubs could be generated given a Class object.

We are constantly working on improving everything shown here, as well as adding even more useful functionality and helpful demos. If you have any new ideas that were not mentioned yet, don't hesitate to contribute!

Installation

In order to use private models that are not publicly available, please install and use capellambse in a local, trusted environment.

You can install the latest released version directly from PyPI.

pip install capellambse

To set up a development environment, clone the project and install it into a virtual environment.

git clone https://github.com/DSD-DBS/py-capellambse
cd capellambse
python -m venv .venv

source .venv/bin/activate.sh  # for Linux / Mac
.venv\Scripts\activate  # for Windows

pip install -U pip pre-commit
pip install -e '.[docs,test]'
pre-commit install

We recommend developing within a local Jupyter notebook server environment. In order to install and run it in the same virtual environment, execute the following additional commands:

pip install jupyter capellambse
cd docs/source/examples
jupyter-notebook

If your browser did not open automatically, follow the instructions in the terminal to start it manually.

Once in the browser, simply click on the 01 Introduction.ipynb notebook to start!

Current limitations

We are continuously improving coverage of Capella onthology with our high-level API (the current coverage map is available here), however it is still incomplete. It covers most of the commonly used paths but when you need to get to an ontology element that isnt covered yet you may do so by using the low-level API.

Also, as we started in mid 2019 and there was no such thing as Python4Capella yet, we are not API compatible with that project. However, we intend to add API compatibility with Python4Capella in later releases.

The generated diagrams are currently not persisted in .aird files, and currently there is no plan to implement this. If there is a genuine usecase for that we may re-consider it - feel free to create an issue or add comments to an existing one.

Render diagrams in untrusted jupyter notebooks

The SVG format relies on a stylesheet, however in untrusted notebooks any stylesheets are stripped. As a workaround we support the PNG format. For this the cairosvg library is needed which requires the following additional setup steps on windows:

• Download and execute the latest GTK installer.
• Reboot to add the path to the installed compiled libraries into your system environment PATH

Contributing

We'd love to see your bug reports and improvement suggestions! Please take a look at our guidelines for contributors for details.

Licenses

This project is compliant with the REUSE Specification Version 3.0.

Copyright DB Netz AG, licensed under Apache 2.0 (see full text in LICENSES/Apache-2.0.txt)

Dot-files are licensed under CC0-1.0 (see full text in LICENSES/CC0-1.0.txt)

To provide the same look and feel across platforms, we distribute our library bundled with the OpenSans font (capellambse/OpenSans-Regular.ttf). The OpenSans font is Copyright 2020 The Open Sans Project Authors, licensed under OFL-1.1 (see full text in LICENSES/OFL-1.1.txt).