5. Pywbemtools development

This section only needs to be read by developers of the pywbemtools package. People that want to make a fix or develop some extension, and people that want to test the project are also considered developers for the purpose of this section.

Generally development users will install pywbemtools by cloning the pywbemtools GitHub package and using the Make utility to handle the installation of pywbemtools and its prerequisites. This provides the user with all of the source of pywbemtools and in addition, the test environment and the documentation files.

5.1. Repository

The repository for pywbemtools is on GitHub:

https://github.com/pywbem/pywbemtools

5.2. Setting up the development environment

It is recommended to use Linux as the development environment for pywbemtools. OS-X should work as well; Windows requires additional manual setup steps.

The pywbemtools development environment is based on a Makefile and therefore requires the GNUmake utility to execute most of the build steps. For Linux and the Mac environments, the GNU make utility normally exists but that is not the case for Windows. GNU make can be installed on native windows from tools such as chocolatey or winget for native windows or window environments such as cygwin or WSL(Windows Subsystem for Linux) enabled to provide a suitable GNUmake utility.

Clone the Git repo of this project and switch to its working directory:

$ git clone git@github.com:pywbem/pywbemtools.git
$ cd pywbemtools

It is recommended that you set up a Python virtual environment. See section Using Python virtual environments. Have the virtual Python environment active for all remaining steps.
Install pywbemtools and its prerequisites for installing and running it as described in Installation. This will install Python packages into the active Python environment, and OS-level packages.
Unix-like environments on Windows (such as CygWin, MinGW, Babun, Gow, or WSL(Windows subsystem for Linux)) may bring their own Python, so double check that the active Python environment is the one you want to use.
Install the prerequisites for pywbemtools development. This will install Python packages into the active Python environment, and OS-level packages:
```
$ make develop
```
This project uses make to do things in the currently active Python environment. The command:
```
$ make
```
displays a list of valid Make targets for pywbemtools installation and test and a short description of what each target does.

5.2.1. Using Python virtual environments

It is beneficial to set up a virtual Python environment for your project, because that leaves your system Python installation unchanged, it does not require sudo rights, and gives you better control about the installed packages and their versions. In effect it isolates the environment for your pywbemtools installation from the rest of the OS environment.

There are a number of different virtual environment tools available for python depending on your OS and Python including venv, virtualenv, etc. More information on virtual environments can be found at sites like virtual Python environment, virtual Python environment tutorial or Python venv

The pywbem development team extensively uses virtualenv and virtualenvwrapper in multiple OS and python version environments.

5.3. Building the documentation

The ReadTheDocs (RTD) site is used to publish the documentation for the pywbemtools package at https://pywbemtools.readthedocs.io/

This page is automatically updated whenever the Git repo for this package changes the branch from which this documentation is built.

In order to build the documentation locally from the Git work directory, execute:

$ make builddoc

The top-level document to open with a web browser will be build_doc/html/docs/index.html.

5.4. Testing

All of the following make commands run the tests in the currently active Python environment. Depending on how the pywbemtools package is installed in that Python environment, either the pywbemtools directory in the main repository directory is used, or the installed pywbemtools package. The test case files and any utility functions they use are always used from the tests directory in the main repository directory.

The tests directory has the following subdirectory structure:

tests
 +-- unit                Unit tests
 |    +-- utils               Utility functions used by unit tests
 +-- manual              Manual tests
 +-- schema              The CIM schema MOF files used by some tests

There are multiple types of tests in pywbemtools:

Unit tests and function tests

Today, the unit tests and function tests are contained in the single directory unit.

The distinction between unit tests and function tests as used in pywbemtools is that function tests exercise the entire pywbemcli client component or entire scripts using the pywbem_mock module and mock CIM model definitions to emulate a WBEM server, while unit tests exercise single modules without using access to a WBEM server.

Generally, the function tests are organized by the command group so that for example the function tests for the class command group are in the file tests\unit\pywbemcli\test_class_subcmd.py.

Tests are run by executing:
```
$ make test
```
Test execution can be modified by a number of environment variables, as documented in the make help (execute make help).
Individual test files

Individual test files in the tests/unit/ environment can be executed by executing pytest <test_file_path>. Note that the tests require some dependencies to locate the pywbemtools code and to set the terminal width for the tests to 120 characters (many command outputs depend on the terminal width to format output). Thus the following would execute the tests on the class command group contained in tests\unit\pywbemcli\test_class_cmds.py.
```
$ PYTHONPATH=. PYWBEMTOOLS_TERMWIDTH=120 pytest tests/unit/pywbemcli/test_class_cmds.py
```
Manual tests

There are several Python scripts and shell scripts that can be run manually. The results need to be validated manually.

These scripts are in the directory:
```
tests/manual/
```
and are executed by simply invoking them from within the main directory of the repository, e.g.:
```
tests/manual/test_pegasus.py
```
Some of the scripts support a --help option that informs about their usage.

Some tests depend on the existence of a DMTF Schema defining the classes and qualifier declarations in a particular release
Running Tox

To run the unit and function tests in all supported Python environments, the Tox tool can be used. It creates the necessary virtual Python environments and executes make test (i.e. the unit and function tests) in each of them.

For running Tox, it does not matter which Python environment is currently active, as long as the Python tox package is installed in it:
```
$ tox                              # Run tests on all supported Python versions
$ tox -e py313                     # Run tests on Python 3.13
```

5.5. Disabling the spinner when debugging

Subcommands normally display a spinner (a character-based spinning wheel) while waiting for completion.

For debugging, it is useful to disable that spinner. This can be done by setting the PYWBEM_SPINNER environment variable to ‘false’, ‘0’, or the empty string. For example:

$ export PYWBEM_SPINNER=false

. .. Git workflow:

5.6. Git workflow

Long-lived branches:
- master - for next functional version
- stable_M.N - for fix stream of released version M.N.
We use topic branches for everything!
- Based upon the intended long-lived branch, if no dependencies
- Based upon an earlier topic branch, in case of dependencies
- It is valid to rebase topic branches and force-push them.
We use pull requests to review the branches.
- Use the correct long-lived branch (e.g. master or stable_0.8) as a merge target!
- Review happens as comments on the pull requests.
- At least two +1 are required for merging.
GitHub meanwhile offers different ways to merge pull requests. We merge pull requests by creating merge commits, so the single commits of a topic branch remain unchanged, and we see the title line of the pull request in the merge commit message, which is often the only place that tells the issue that was fixed.

5.7. Releasing a version

This section describes how to release a version of pywbemtools to PyPI.

It covers all variants of versions that can be released:

Releasing a new major version (Mnew.0.0) based on the master branch
Releasing a new minor version (M.Nnew.0) based on the master branch
Releasing a new update version (M.N.Unew) based on the stable branch of its minor version

This description assumes that you are authorized to push to the remote repo at https://github.com/pywbem/pywbemtools and that the remote repo has the remote name origin in your local clone.

Any commands in the following steps are executed in the main directory of your local clone of the pywbem/pywbemtools Git repo.

On GitHub, verify open items in milestone M.N.U.

Verify that milestone M.N.U has no open issues or PRs anymore. If there are open PRs or open issues, make a decision for each of those whether or not it should go into version M.N.U you are about to release.

If there are open issues or PRs that should go into this version, abandon the release process.

If none of the open issues or PRs should go into this version, change their milestones to a future version, and proceed with the release process. You may need to create the milestone for the future version.
Run the Safety tool:
```
make safety
```
If any of the two safety runs fails, fix the safety issues that are reported, in a separate branch/PR.

Roll back the PR into any maintained stable branches.
Check for any dependabot alerts.

If there are any dependabot alerts, fix them in a separate branch/PR.

Roll back the PR into any maintained stable branches.
Create and push the release branch (replace M,N,U accordingly):
```
VERSION=M.N.U make release_branch
```
This uses the default branch determined from VERSION: For M.N.0, the master branch is used, otherwise the stable_M.N branch is used. That covers for all cases except if you want to release a new minor version based on an earlier stable branch. In that case, you need to specify that branch:
```
VERSION=M.N.0 BRANCH=stable_M.N make release_branch
```
This includes the following steps:
- create the release branch (release_M.N.U), if it does not yet exist
- make sure the AUTHORS.md file is up to date
- update the change log from the change fragment files, and delete those
- commit the changes to the release branch
- push the release branch
If this command fails, the fix can be committed to the release branch and the command above can be retried.
On GitHub, create a Pull Request for branch release_M.N.U.

Important: When creating Pull Requests, GitHub by default targets the master branch. When releasing based on a stable branch, you need to change the target branch of the Pull Request to stable_M.N.

Set the milestone of that PR to version M.N.U.

This PR should normally be set to be reviewed by at least one of the maintainers.

The PR creation will cause the “test” workflow to run. That workflow runs tests for all defined environments, since it discovers by the branch name that this is a PR for a release.
On GitHub, once the checks for that Pull Request have succeeded, merge the Pull Request (no review is needed). This automatically deletes the branch on GitHub.

If the PR did not succeed, fix the issues.
On GitHub, close milestone M.N.U.

Verify that the milestone has no open items anymore. If it does have open items, investigate why and fix (probably step 1 was not performed).
Publish the package (replace M,N,U accordingly):
```
VERSION=M.N.U make release_publish
```
or (see step 4):
```
VERSION=M.N.0 BRANCH=stable_M.N make release_publish
```
This includes the following steps:
- create and push the release tag
- clean up the release branch
Pushing the release tag will cause the “publish” workflow to run. That workflow builds the package, publishes it on PyPI, creates a release for it on GitHub, and finally creates a new stable branch on GitHub if the master branch was released.

Verify the publishing

Wait for the “publish” workflow for the new release to have completed: https://github.com/pywbem/pywbemtools/actions/workflows/publish.yml

Then, perform the following verifications:
- Verify that the new version is available on PyPI at https://pypi.python.org/pypi/pywbemtools/
- Verify that the new version has a release on Github at https://github.com/pywbem/pywbemtools/releases
- Verify that the new version has documentation on ReadTheDocs at https://pywbemtools.readthedocs.io/en/stable/changes.html
  
  The new version M.N.U should be automatically active on ReadTheDocs, causing the documentation for the new version to be automatically built and published.
  
  If you cannot see the new version after some minutes, log in to https://readthedocs.org/projects/pywbemtools/versions/ and activate the new version.

5.8. Starting a new version

This section shows the steps for starting development of a new version.

This section covers all variants of new versions:

Starting a new major version (Mnew.0.0) based on the master branch
Starting a new minor version (M.Nnew.0) based on the master branch
Starting a new update version (M.N.Unew) based on the stable branch of its minor version

This description assumes that you are authorized to push to the remote repo at https://github.com/pywbem/pywbemtools and that the remote repo has the remote name origin in your local clone.

Any commands in the following steps are executed in the main directory of your local clone of the pywbem/pywbemtools Git repo.

Create and push the start branch (replace M,N,U accordingly):
```
VERSION=M.N.U make start_branch
```
This uses the default branch determined from VERSION: For M.N.0, the master branch is used, otherwise the stable_M.N branch is used. That covers for all cases except if you want to start a new minor version based on an earlier stable branch. In that case, you need to specify that branch:
```
VERSION=M.N.0 BRANCH=stable_M.N make start_branch
```
This includes the following steps:
- create the start branch (start_M.N.U), if it does not yet exist
- create a dummy change
- commit and push the start branch (start_M.N.U)
On GitHub, create a milestone for the new version M.N.U.

You can create a milestone in GitHub via Issues -> Milestones -> New Milestone.
On GitHub, create a Pull Request for branch start_M.N.U.

Important: When creating Pull Requests, GitHub by default targets the master branch. When starting a version based on a stable branch, you need to change the target branch of the Pull Request to stable_M.N.

No review is needed for this PR.

Set the milestone of that PR to the new version M.N.U.
On GitHub, go through all open issues and pull requests that still have milestones for previous releases set, and either set them to the new milestone, or to have no milestone.

Note that when the release process has been performed as described, there should not be any such issues or pull requests anymore. So this step here is just an additional safeguard.
On GitHub, once the checks for the Pull Request for branch start_M.N.U have succeeded, merge the Pull Request (no review is needed). This automatically deletes the branch on GitHub.
Update and clean up the local repo (replace M,N,U accordingly):
```
VERSION=M.N.U make start_tag
```
or (see step 1):
```
VERSION=M.N.0 BRANCH=stable_M.N make start_tag
```
This includes the following steps:
- checkout and pull the branch that was started (master or stable_M.N)
- delete the start branch (start_M.N.U) locally and remotely
- create and push the start tag (M.N.Ua0)

5.9. Contributing

Third party contributions to this project are welcome!

In order to contribute, create a Git pull request, considering this:

Test is required.
Each commit should only contain one “logical” change.
A “logical” change should be put into one commit, and not split over multiple commits.
Large new features should be split into stages.
The commit message should not only summarize what you have done, but explain why the change is useful.
The commit message must follow the format explained below.

What comprises a “logical” change is subject to sound judgement. Sometimes, it makes sense to produce a set of commits for a feature (even if not large). For example, a first commit may introduce a (presumably) compatible API change without exploitation of that feature. With only this commit applied, it should be demonstrable that everything is still working as before. The next commit may be the exploitation of the feature in other components.

For further discussion of good and bad practices regarding commits, see:

5.10. Creating and submitting a change to pywbemtools

All changes to pywbemtools are made through Github with PRs created on topic branches and merged with the current master after successful group review.

To make a change, create a topic branch. You can assume that you are the only one using that branch, so force-pushes to that branch and rebasing that branch is fine.

When you are ready to push your change, describe the change for users of the package in a change fragment file. That is a small file in RST format with just a single change. For more background, read the [towncrier concept](https://towncrier.readthedocs.io/en/stable/markdown.html) (which uses Markdown format in that description and calls these files ‘news fragment files’).

To create a change fragment file, execute:

For changes that have a corresponding issue:

towncrier create <issue>.<type>.rst --edit

For changes that have no corresponding issue:

towncrier create noissue.<number>.<type>.rst --edit

For changes where you do not want to create a change log entry:

towncrier create noissue.<number>.notshown.rst --edit
# The file content will be ignored - it can also be empty

where:

<issue> - The issue number of the issue that is addressed by the change. If the change addresses more than one issue, copy the new change fragment file after its content has been edited, using the other issue number in the file name. It is important that the file content is exactly the same, so that towncrier can create a single change log entry from the two (or more) files.

If the change has no related issue, use the noissue.<number>.<type>.rst file name format, where <number> is any number that results in a file name that does not yet exist in the changes directory.
<type> - The type of the change, using one of the following values:
- incompatible - An incompatible change. This will show up in the “Incompatible Changes” section of the change log. The text should include a description of the incompatibility from a user perspective and if possible, how to mitigate the change or what replacement functionality can be used instead.
- deprecation - An externally visible functionality is being deprecated in this release. This will show up in the “Deprecations” section of the change log. The deprecated functionality still works in this release, but may go away in a future release. If there is a replacement functionality, the text should mention it.
- fix - A bug fix in the code, documentation or development environment. This will show up in the “Bug fixes” section of the change log.
- feature - A feature or enhancement in the code, documentation or development environment. This will show up in the “Enhancements” section of the change log.
- cleanup - A cleanup in the code, documentation or development environment, that does not fix a bug and is not an enhanced functionality. This will show up in the “Cleanup” section of the change log.
- notshown - The change will not be shown in the change log.

This command will create a new change fragment file in the changes directory and will bring up your editor (usually vim).

If your change does multiple things of different types listed above, create a separate change fragment file for each type.

If you need to modify an existing change log entry as part of your change, edit the existing corresponding change fragment file.

Add the new or changed change fragment file(s) to your commit. The test workflow running on your Pull Request will check whether your change adds or modifies change fragment files.

You can review how your changes will show up in the final change log for the upcoming release by running:

towncrier build --draft

Always make sure that your pushed branch has either just one commit, or if you do multiple things, one commit for each logical change. What is not OK is to submit for review a PR with the multiple commits it took you to get to the final result for the change.

5.11. Core Development Team

Anyone can contribute to pywbemtools via pull requests as described in the previous section.

The pywbemtools project has a core development team that holds regular web conferences and that is using Slack for offline communication, on the Slack workspace: https://pywbem.slack.com.

The web conference and the Slack workspace are by invitation, and if you want to participate in the core team, please open a pywbem issue to let us know.