See project homepage.

tldr:

cargo make --cwd crates/pipeline_manager/ python_test

To avoid installing the library every time you make updates and run the test scripts, you can set up a local development environment using a virtual environment and then install your library in editable mode. This allows you to make changes to the library code and have those changes immediately reflected when running your test scripts. Here's how you can do it:

1. Create a virtual environment: Open a terminal/command prompt and navigate to your project directory. Then, create a virtual environment by running the following command in the dbsp base directory:

python3 -m venv venv

2. Activate the virtual environment: Activate the virtual environment using the appropriate command for your operating system:

• On macOS/Linux:

source venv/bin/activate

• On Windows:

venv\Scripts\activate.bat

3. Install your library in editable mode: While the virtual environment is active, navigate to the directory where your library's pyproject.toml file is located. Then, install the library in editable mode using pip as follows:

cd python/feldera-api-client
pip install -e .
cd ..
pip install -e .

This will create a symlink from your virtual environment to your library code, allowing changes to be immediately reflected.

4. Run your test scripts: You can now run your test scripts without having to install the library again. The virtual environment will use the locally installed version of your library.

cd demo
bash demo.sh

Whenever you make changes to your library code, you don't need to reinstall it. Simply rerun your test scripts, and the changes will be picked up automatically.

Remember to activate the virtual environment every time you work on your project or want to run the test scripts.

In case you make changes to the OpenAPI spec (by modifying crates/pipeline_manager), you'll need to regenerated the feldera-api-client library. To do so, run the following command:

cargo make --cwd crates/pipeline_manager/ openapi_python