Skip to content

Contributing to the App

The project is packaged with a light development environment based on docker-compose to help with the local development of the project and to run tests.

The project is following Network to Code software development guidelines and is leveraging the following:

  • Python linting and formatting: black, pylint, bandit, flake8, and pydocstyle.
  • YAML linting is done with yamllint.
  • Django unit test to ensure the plugin is working properly.

Documentation is built using mkdocs. The Docker based development environment automatically starts a container hosting a live version of the documentation website on http://localhost:8001 that auto-refreshes when you make any changes to your local files.

Creating Changelog Fragments

All pull requests to next or develop must include a changelog fragment file in the ./changes directory. To create a fragment, use your GitHub issue number and fragment type as the filename. For example, 2362.added. Valid fragment types are added, changed, deprecated, fixed, removed, and security. The change summary is added to the file in plain text. Change summaries should be complete sentences, starting with a capital letter and ending with a period, and be in past tense. Each line of the change fragment will generate a single change entry in the release notes. Use multiple lines in the same file if your change needs to generate multiple release notes in the same category. If the change needs to create multiple entries in separate categories, create multiple files.

Example

Wrong

changes/1234.fixed
fix critical bug in documentation

Right

changes/1234.fixed
Fixed critical bug in documentation.

Multiple Entry Example

This will generate 2 entries in the fixed category and one entry in the changed category.

changes/1234.fixed
Fixed critical bug in documentation.
Fixed release notes generation.
changes/1234.changed
Changed release notes generation.

Adding a new top-level command

First, you should be familiar with the design goals and constraints involved in Nautobot (design.md). Be sure that this is really what you want to do, versus adding a sub-command instead.

We recommend that each command exist as its own submodule under nautobot_chatops/workers/ (or, as a separate package entirely, such as nautobot_chatops_mycommand/worker.py, using the entrypoint/plugin capability described in design.md) to keep code files to a reasonable size and complexity. This submodule or package should implement a celery worker function(s). In general this worker function shouldn't need to do much more than call the handle_subcommands helper function provided:

# nautobot_chatops/workers/mycommand.py

from nautobot_chatops.workers import handle_subcommands, subcommand_of


def mycommand(subcommand, **kwargs)
    """Perform mycommand and its subcommands."""
    return handle_subcommands("mycommand", subcommand, **kwargs)

By using handle_subcommands, the top-level command worker will automatically recognize the sub-command "help", as well as any sub-commands registered using the @subcommand_of decorator.

You shouldn't need to make any changes to the views or dispatchers modules in this scenario.

For usability, you should use the App Studio app in the Microsoft Teams client to update the bot settings (Nautobot_ms_teams.zip) to include this new top-level command as a documented command supported by the bot. You will probably then need to delete the bot deployment from your team and re-deploy it for the new command to appear.

You will also need to log in to api.slack.com and add the new slash-command to your bot's configuration.

Adding a new sub-command

First, you should be familiar with the design goals and constraints involved in Nautobot (design.md).

To register a sub-command, write a function whose name matches the sub-command's name (any _ in the function name will be automatically converted to - for the sub-command name), and decorate it with the @subcommand_of decorator. This function must take dispatcher (an instance of any Dispatcher subclass) as its first argument; any additional positional arguments become arguments in the chat app UI. The docstring of this function will become the help text displayed for this sub-command when a user invokes <command> help, so it should be concise and to the point.

from nautobot_chatops.workers import subcommand_of

# ...

@subcommand_of("mycommand")
def do_something(dispatcher, first_arg, second_arg):
    """Do something with two arguments."""
    # ...

With the above code, the command mycommand do_something [first_arg] [second_arg] will now be available.

You shouldn't need to make any changes to the views or dispatchers modules in this scenario.

A sub-command worker function should always return one of the following:

return False

This indicates that the function did not do anything meaningful, and it so should not be logged in Nautobot's command log. Typically, this is only returned when not all required parameters have been provided by the user and so the function needs to prompt the user for additional inputs, for example:

@subcommand_of("nautobot")
def get_rack(dispatcher, site_key, rack_id):
    """Get information about a specific rack from Nautobot."""
    site_lt = LocationType.objects.get(name="Site")
    if not site_key:
        site_options = [(site.name, site.composite_key) for site in Location.objects.filter(location_type=site_lt)]
        dispatcher.prompt_from_menu("nautobot get-rack", "Select a site (location)", site_options)
        return False  # command did not run to completion and therefore should not be logged
    ...

return CommandStatusChoices.STATUS_SUCCEEDED

This indicates that the command was successful, and no further details are necessary in the logging. You could return another status code besides STATUS_SUCCEEDED in this pattern, but in general any other status code should be accompanied by an explanatory message:

return (CommandStatusChoices.STATUS_FAILED, details_str)

This indicates that the command failed for some reason, which is provided for logging purposes. You could also use other status codes (including STATUS_SUCCEEDED) for any other outcome that also requires explanation.

The provided details_str will be stored in the Nautobot command log history.

Adding support for a new chat platform (Webhooks)

First, you should be familiar with the design goals and constraints involved in Nautobot (design.md).

You'll need to add a new nautobot_chatops.views.<platform> submodule that provides any necessary API endpoints.

You'll also need to add a new nautobot_chatops.dispatchers.<platform> submodule that implements an appropriate subclass of Dispatcher. This new dispatcher class will need to implement any abstract methods of the base class and override any other methods where platform-specific behavior is required (which will probably be most of them).

You shouldn't need to make any changes to the workers module in this scenario.

Adding support for a new chat platform (WebSockets)

First, you should be familiar with the design goals and constraints involved in Nautobot (design.md).

You'll need to add a new nautobot_chatops.sockets.<platform> submodule that provides the necessary WebSockets connection to the Platform.

You'll also need to add a new nautobot_chatops.dispatchers.<platform> submodule that implements an appropriate subclass of Dispatcher. This new dispatcher class will need to implement any abstract methods of the base class and override any other methods where platform-specific behavior is required (which will probably be most of them).

Finally, you will need to add a new nautobot_chatops.management.start_<platform>_socket management command that will start the WebSockets asynchronous loop. In 2.0 these will likely be condensed to use a single base command with arguments to select the platform.

You shouldn't need to make any changes to the workers module in this scenario.

Submitting Pull Requests

  • It is recommended to open an issue before starting work on a pull request, and discuss your idea with the Nautobot maintainers before beginning work. This will help prevent wasting time on something that we might not be able to implement. When suggesting a new feature, also make sure it won't conflict with any work that's already in progress.

  • Once you've opened or identified an issue you'd like to work on, ask that it be assigned to you so that others are aware it's being worked on. A maintainer will then mark the issue as "accepted."

  • If you followed the project guidelines, have ample tests, code quality, you will first be acknowledged for your work. So, thank you in advance! After that, the PR will be quickly reviewed to ensure that it makes sense as a contribution to the project, and to gauge the work effort or issues with merging into current. If the effort required by the core team isn’t trivial, it’ll likely still be a few weeks before it gets thoroughly reviewed and merged, thus it won't be uncommon to move it to near term with a near-term label. It will just depend on the current backlog.

  • All code submissions should meet the following criteria (CI will enforce these checks):

    • Python syntax is valid
    • All unit tests pass successfully
    • PEP 8 compliance is enforced, with the exception that lines may be greater than 80 characters in length
    • At least one changelog fragment has been included in the feature branch

Branching Policy

The branching policy includes the following tenets:

  • The develop branch is the primary branch to develop off of.
  • PRs intended to add new features should be sourced from the develop branch.
  • PRs intended to address bug fixes and security patches should be sourced from the develop branch.
  • PRs intended to add new features that break backward compatibility should be discussed before a PR is created.

Nautobot ChatOps app will observe semantic versioning, as of 1.0. This may result in a quick turn around in minor versions to keep pace with an ever-growing feature set.

Release Policy

Nautobot ChatOps currently has no intended scheduled release schedule, and will release new features in minor versions.

When a new release of any kind (e.g. from develop to main, or a release of a stable-<major>.<minor>) is created the following should happen.

  • A release PR is created:
    • Add and/or update to the changelog in docs/admin/release_notes/version_<major>.<minor>.md file to reflect the changes.
    • Update the mkdocs.yml file to include updates when adding a new release_notes version file.
    • Change the version from <major>.<minor>.<patch>-beta to <major>.<minor>.<patch> in pyproject.toml.
    • Set the PR to the proper branch, e.g. either main or stable-<major>.<minor>.
  • Ensure the tests for the PR pass.
  • Merge the PR.
  • Create a new tag:
    • The tag should be in the form of v<major>.<minor>.<patch>.
    • The title should be in the form of v<major>.<minor>.<patch>.
    • The description should be the changes that were added to the version_<major>.<minor>.md document.
  • If merged into main, then push from main to develop, in order to retain the merge commit created when the PR was merged.
  • If there is a new <major>.<minor>, create a stable-<major>.<minor> for the previous version, so that security updates to old versions may be applied more easily.
  • A post release PR is created:
    • Change the version from <major>.<minor>.<patch> to <major>.<minor>.<patch + 1>-beta in pyproject.toml.
    • Set the PR to the proper branch, e.g. either develop or stable-<major>.<minor>.
    • Once tests pass, merge.