Contributing
Issues
You can find our open issues in the project's issue tracker. Please let us know if you find any issues or have any feature requests there.
Contributing
If you want to contribute to the project, your help is very welcome. Just fork the project, make your changes and send us a pull request. You can find the detailed description of how to do this in Github's guide to contributing to projects.
Our Roadmap page is a
comprehensive list of tasks we want to do in the future. It is a good place to
start if you want to contribute to Vizzu
. In case you have something else in
mind, that's awesome and we are very interested in hearing about it.
CI-CD
Development environment
For contributing to the project, it is recommended to use Python
3.10
as the
primary programming language for most parts of the source code. However, a
specific portion of the codebase is written in JavaScript
. If you plan to
contribute to this JavaScript
part or the documentation, you will need
Node.js
, preferably version 18
.
The following steps demonstrate how to set up the development environment on an
Ubuntu
22.04
operating system. However, the process can be adapted for other
operating systems as well.
To start using the ipyvizzu
development environment, you need to create a
virtual environment and install pdm
within it.
python3.10 -m venv ".venv"
source .venv/bin/activate
pip install pdm==2.8.0
Once set up, you can install development dependencies:
pdm install
pdm run deploy
For contributing to the documentation you need to install npm
development
dependencies:
npm install
The development requirements are installed based on the pdm.lock
and
package-lock.json
files. To update the development requirements, you can use
the command pdm run lock
.
Note: For all available pdm
scripts, run pdm run --list
.
For better development practices, you can set up pre-commit
and pre-push
hooks in your local Git
repository. The pre-commit
hook will format the code
automatically, and the pre-push
hook will run the CI steps before pushing your
changes.
pre-commit install --hook-type pre-commit --hook-type pre-push -c ./tools/ci/.pre-commit-ubuntu.yaml
Note: The provided .pre-commit-ubuntu.yaml
configuration file is tailored
for Ubuntu
22.04
. If you intend to use another operating system, you may
need to create a custom configuration file suitable for that environment.
CI
The CI pipeline includes code formatting checks, code analysis, typing
validation, and unit tests for the ipyvizzu
project.
To run the entire CI pipeline, execute the following pdm
script:
pdm run ci
However, if you want to run the CI steps on specific parts of the project, you
can use the following scripts: ci-src-py
, ci-src
, ci-docs
, or ci-tools
.
Formatting
You can check the code's formatting using the format
script:
pdm run format
If you need to fix any formatting issues, you can use the fix-format
script:
pdm run fix-format
If you wish to format specific parts of the project, you can use the following
scripts: format-src-py
, format-src
, format-docs
, format-tools
, or
fix-format-src-py
, fix-format-src
, fix-format-docs
, fix-format-tools
.
Code analyses
To perform code analyses, you can use the lint
script:
pdm run lint
If you need to run code analyses for specific parts of the project, you can
utilize the following scripts: lint-src-py
, lint-src
, lint-docs
, or
lint-tools
.
Typing
For type checking, you can use the type
script:
pdm run type
If you want to check specific parts of the project, you can use the following
scripts: type-src
or type-tools
.
Testing
The project is tested using the unittest
testing framework and tox
. To run
the tests, you can use the test
script:
pdm run test
Documentation
Note: The preset, static, animated, and analytical operation examples are generated from the vizzu-lib repository. If you wish to build them as well, run the following command before building the site.
git clone --depth 1 https://github.com/vizzuhq/vizzu-lib.git
To build the documentation, you can use the build-docs
script:
pdm run build-docs
You can read the online version at ipyvizzu.com.
Release
ipyvizzu
is distributed on pypi.
Note: You need to be an administrator to release the project.
To release ipyvizzu
, follow the steps below:
- Increase the version number in
__version__.py
. The version bump should be in a separate commit.
- Generate the release notes and publish the new release on Releases.
Note: Publishing a new release will automatically trigger the cd
workflow,
which builds, checks, and uploads the ipyvizzu
package to
pypi.
Before making a release, you can build and check the package using the release
script:
pdm run release