If there is a specific version of Python that you need and you don't want to worry about any potential breaking changes due to patch updates (going from `3.7.5` to `3.7.6` for example), you should specify the **exact major, minor, and patch version** (such as `3.7.5`):
- The only downside to this is that setup may take a little longer. If the exact version is not already installed on the runner due to more recent versions, the exact version will have to be downloaded.
- MSI installers are used on Windows for this, so runs will take a little longer to set up vs macOS and Linux.
- There will be a single patch version already installed on each runner for every minor version of Python that is supported.
- The patch version that will be preinstalled, will generally be the latest and every time there is a new patch released, the older version that is preinstalled will be replaced.
- Using the most recent patch version will result in a very quick setup since no downloads will be required since a locally installed version of Python on the runner will be used.
It's also possible to use **x.y-dev syntax** to download and set up the latest patch version of Python, alpha, beta and rc (release candidate) releases included. (for specified major & minor versions):
- **[hyphen ranges](https://github.com/npm/node-semver#hyphen-ranges-xyz---abc)** to download and set up the latest available version of Python (includes both pre-release and stable versions):
Please refer to the [Advanced range syntax section](https://github.com/npm/node-semver#advanced-range-syntax) of the [semver](https://github.com/npm/node-semver) to check other available range syntaxes.
Using `setup-python` it's possible to use [matrix syntax](https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#jobsjob_idstrategymatrix) to install several versions of Python or PyPy:
`setup-python` action can read Python or PyPy version from a version file. `python-version-file` input is used for specifying the path to the version file. If the file that was supplied to `python-version-file` input doesn't exist, the action will fail with error.
>In case both `python-version` and `python-version-file` inputs are supplied, the `python-version-file` input will be ignored due to its lower priority.
The `check-latest` flag defaults to `false`. Use the default or set `check-latest` to `false` if you prefer stability and if you want to ensure a specific `Python or PyPy` version is always used.
If `check-latest` is set to `true`, the action first checks if the cached version is the latest one. If the locally cached version is not the most up-to-date, a `Python or PyPy` version will then be downloaded. Set `check-latest` to `true` if you want the most up-to-date `Python or PyPy` version to always be used.
Using **python-version** output it's possible to get the installed by action Python or PyPy version. This output is useful when the input `python-version` is given as a range (e.g. 3.8.0 - 3.10.0 ), but down in a workflow you need to operate with the exact installed version (e.g. 3.10.1).
With this setting, the action will add/update environment variables (e.g. `PATH`, `PKG_CONFIG_PATH`, `pythonLocation`) for Python or PyPy to just work out of the box.
This can prove useful if you want the only side-effect to be to ensure Python or PyPy is installed and rely on the `python-path` output to run executable.
- For detailed information regarding the available versions of Python that are installed, see [Supported software](https://docs.github.com/en/actions/reference/specifications-for-github-hosted-runners#supported-software).
- For every minor version of Python, expect only the latest patch to be preinstalled.
- If the exact patch version doesn't matter to you, specifying just the major and minor versions will get you the latest preinstalled patch version. In the previous example, the version spec `3.8` will use the `3.8.2` Python version found in the cache.
- Use `-dev` instead of a patch number (e.g., `3.11-dev`) to install the latest patch version release for a given minor version, *alpha and beta releases included*.
- Downloadable Python versions from GitHub Releases ([actions/python-versions](https://github.com/actions/python-versions/releases)).
- All available versions are listed in the [version-manifest.json](https://github.com/actions/python-versions/blob/main/versions-manifest.json) file.
- If there is a specific version of Python that is not available, you can open an issue here
>**Note:** Python versions used in this action are generated in the [python-versions](https://github.com/actions/python-versions) repository. For macOS and Ubuntu images, python versions are built from the source code. For Windows, the python-versions repository uses installation executable. For more information please refer to the [python-versions](https://github.com/actions/python-versions) repository.
- For detailed information regarding the available versions of PyPy that are installed, see [Supported software](https://docs.github.com/en/actions/reference/specifications-for-github-hosted-runners#supported-software).
- For the latest PyPy release, all versions of Python are cached.
- Cache is updated with a 1-2 week delay. If you specify the PyPy version as `pypy3.7` or `pypy-3.7`, the cached version will be used although a newer version is available. If you need to start using the recently released version right after release, you should specify the exact PyPy version using `pypy3.7-v7.3.3` or `pypy-3.7-v7.3.3`.
- Downloadable PyPy versions from the [official PyPy site](https://downloads.python.org/pypy/).
- All available versions that we can download are listed in [versions.json](https://downloads.python.org/pypy/versions.json) file.
- PyPy <7.3.3arenotavailabletoinstallon-flight.
- If some versions are not available, you can open an issue in https://foss.heptapod.net/pypy/pypy/
GitHub hosted runners have a tool cache that comes with a few versions of Python + PyPy already installed. This tool cache helps speed up runs and tool setup by not requiring any new downloads. There is an environment variable called `RUNNER_TOOL_CACHE` on each runner that describes the location of the tool cache with Python and PyPy installed. `setup-python` works by taking a specific version of Python or PyPy from this tool cache and adding it to PATH.
GitHub runner images are set up in [actions/runner-images](https://github.com/actions/runner-images). During the setup, the available versions of Python and PyPy are automatically downloaded, set up and documented.
- Tool cache setup for Ubuntu: [Install-Toolset.ps1](https://github.com/actions/runner-images/blob/main/images/linux/scripts/installers/Install-Toolset.ps1) [Configure-Toolset.ps1](https://github.com/actions/runner-images/blob/main/images/linux/scripts/installers/Configure-Toolset.ps1)
- Tool cache setup for Windows: [Install-Toolset.ps1](https://github.com/actions/runner-images/blob/main/images/win/scripts/Installers/Install-Toolset.ps1) [Configure-Toolset.ps1](https://github.com/actions/runner-images/blob/main/images/win/scripts/Installers/Configure-Toolset.ps1)
Python distributions are only available for the same [environments](https://github.com/actions/runner-images#available-images) that GitHub Actions hosted environments are available for. If you are using an unsupported version of Ubuntu such as `19.04` or another Linux distribution such as Fedora, `setup-python` may not work.
If you have a supported self-hosted runner and you would like to use `setup-python`, there are a few extra things you need to make sure are set up so that new versions of Python can be downloaded and configured on your runner.
- Your runner needs to be running with administrator privileges so that the appropriate directories and files can be set up when downloading and installing a new version of Python for the first time.
- If your runner is configured as a service, make sure the account that is running the service has the appropriate write permissions so that Python can get installed. The default `NT AUTHORITY\NETWORK SERVICE` should be sufficient.
- You need `7zip` installed and added to your `PATH` so that the downloaded versions of Python files can be extracted properly during the first-time setup.
- MSI installers are used when setting up Python on Windows. A word of caution as MSI installers update registry settings.
- The 3.8 MSI installer for Windows will not let you install another 3.8 version of Python. If `setup-python` fails for a 3.8 version of Python, make sure any previously installed versions are removed by going to "Apps & Features" in the Settings app.
> By default runner downloads and installs tools into the folder set up by `RUNNER_TOOL_CACHE` environment variable. The environment variable called `AGENT_TOOLSDIRECTORY` can be set to change this location for Windows self-hosted runners.
>If you are experiencing problems while configuring Python on your self-hosted runner, turn on [step debugging](https://github.com/actions/toolkit/blob/main/docs/action-debugging.md#step-debug-logs) to see additional logs.
By default runner downloads and installs tools into the folder set up by `RUNNER_TOOL_CACHE` environment variable. The environment variable called `AGENT_TOOLSDIRECTORY` can be set to change this location for Linux self-hosted runners:
- In the same shell that your runner is using, type `export AGENT_TOOLSDIRECTORY=/path/to/folder`.
- More permanent way of setting the environment variable is to create an `.env` file in the same directory as your runner and to add `AGENT_TOOLSDIRECTORY=/path/to/folder`. This ensures the variable is always set if your runner is configured as a service.
If you're using a non-default tool cache directory be sure that the user starting the runner has write permission to the new tool cache directory. To check the current user and group that the runner belongs type `ls -l` inside the runner's root directory.
> If your runner is configured as a service and you run into problems, make sure the user that the service is running as is correct. For more information, you can [check the status of your self-hosted runner](https://docs.github.com/en/actions/hosting-your-own-runners/configuring-the-self-hosted-runner-application-as-a-service#checking-the-status-of-the-service).
The Python packages for macOS that are downloaded from `actions/python-versions` are originally compiled from the source in `/Users/runner/hostedtoolcache`. Due to the fixed shared library path, these Python packages are non-relocatable and require to be installed only in `/Users/runner/hostedtoolcache`. Before the use of `setup-python` on the macOS self-hosted runner:
> If your runner is configured as a service and you run into problems, make sure the user that the service is running as is correct. For more information, you can [check the status of your self-hosted runner](https://docs.github.com/en/actions/hosting-your-own-runners/configuring-the-self-hosted-runner-application-as-a-service#checking-the-status-of-the-service).
`setup-python` comes pre-installed on the appliance with GHES if Actions is enabled. When dynamically downloading Python distributions, `setup-python` downloads distributions from [`actions/python-versions`](https://github.com/actions/python-versions) on github.com (outside of the appliance). These calls to `actions/python-versions` are by default made via unauthenticated requests, which are limited to [60 requests per hour per IP](https://docs.github.com/en/rest/overview/resources-in-the-rest-api#rate-limiting). If more requests are made within the time frame, then you will start to see rate-limit errors during downloading that look like this:
##[error]API rate limit exceeded for YOUR_IP. (But here's the good news: Authenticated requests get a higher rate limit. Check out the documentation for more details.)
To get a higher rate limit, you can [generate a personal access token (PAT) on github.com](https://github.com/settings/tokens/new) and pass it as the `token` input for the action. It is important to understand that this needs to be a token from github.com and _not_ from your GHES instance. If you or your colleagues do not yet have a github.com account, you might need to create one.
Here are the steps you need to follow to avoid the rate limit:
1. Create a PAT on any github.com account by using [this link](https://github.com/settings/tokens/new) after logging into github.com (not your Enterprise instance). This PAT does _not_ need any rights, so make sure all the boxes are unchecked.
2. Store this PAT in the repository / organization where you run your workflow, e.g. as `GH_GITHUB_COM_TOKEN`. You can do this by navigating to your repository -> **Settings** -> **Secrets** -> **Actions** -> **New repository secret**.
3. To use this functionality, you need to use any version newer than `v4.3`. Also, change _python-version_ as needed.
Requests should now be authenticated. To verify that you are getting the higher rate limit, you can call GitHub's [rate limit API](https://docs.github.com/en/rest/rate-limit) from within your workflow ([example](https://github.com/actions/setup-python/pull/443#issuecomment-1206776401)).
If the runner is not able to access github.com, any Python versions requested during a workflow run must come from the runner's tool cache. See "[Setting up the tool cache on self-hosted runners without internet access](https://docs.github.com/en/enterprise-server@3.2/admin/github-actions/managing-access-to-actions-from-githubcom/setting-up-the-tool-cache-on-self-hosted-runners-without-internet-access)" for more information.
