airflow.providers.standard.utils.python_virtualenv

Utilities for creating a virtual environment.

Module Contents

Functions

prepare_virtualenv(venv_directory, python_bin, ...[, ...])

Create a virtual environment and install the additional python packages.

write_python_script(jinja_context, filename[, ...])

Render the python script to a file to execute in the virtual environment.
airflow.providers.standard.utils.python_virtualenv.prepare_virtualenv(venv_directory, python_bin, system_site_packages, requirements=None, requirements_file_path=None, pip_install_options=None, index_urls=None)[source]

Create a virtual environment and install the additional python packages.

Parameters

  • venv_directory (str) – The path for directory where the environment will be created.

  • python_bin (str) – Path to the Python executable.

  • system_site_packages (bool) – Whether to include system_site_packages in your virtualenv. See virtualenv documentation for more information.

  • requirements (list[str] | None) – List of additional python packages.

  • requirements_file_path (str | None) – Path to the requirements.txt file.

  • pip_install_options (list[str] | None) – a list of pip install options when installing requirements See ‘pip install -h’ for available options

  • index_urls (list[str] | None) – an optional list of index urls to load Python packages from. If not provided the system pip conf will be used to source packages from.

Returns

Path to a binary file with Python in a virtual environment.

Return type

str

airflow.providers.standard.utils.python_virtualenv.write_python_script(jinja_context, filename, render_template_as_native_obj=False)[source]

Render the python script to a file to execute in the virtual environment.

Parameters

  • jinja_context (dict) – The jinja context variables to unpack and replace with its placeholders in the template file.

  • filename (str) – The name of the file to dump the rendered script to.

  • render_template_as_native_obj (bool) – If True, rendered Jinja template would be converted to a native Python object

Was this entry helpful?