Skip to main content

Subprocess Management extension for the Kabaret framework

Project description


Subprocess Manager for Kabaret.

Provides an Actor, Views and Flow Objects to manage collections of environment + command-lines, as well as running & controling them.



You implement your Runners by subclassing kabaret.subprocess_manager.runner_factory.Runner. Runners have a name, some tags, a liste of supported versions, an icon, etc...

At runtime, you add your Runners to a Factory. A Factory has a name, can lookup Runners based on name and tags, etc... You will probably have several factories.

SubprocessManager Actor

In order to use them, you will Add them to the SubprocessManager Actor. There is 2 places where you may want to do it:

  • In your Session, when creating the Action. This is nice for generic stuff (explorer, shell, text editors, ...)
  • In your Flow, when a project is instaciated. This is nice for production specific tools where the environment and the version are precisely defined (scene editing, baking, render, ...)

LaunchToolBar and SubprocessView

If you add the LaunchToolBar to your Session, you will be able to run your Runner by clicking on the toolbar buttons.

Those run take no parameters.

If you add the SubprocessView (and it get implemented some day), you will see a list of process you ran, with their outputs, and you will have opportinities to remove/kill/rerun...

Flow RunAction

A big majority of your Runner will be ran from your flows though.

The kabaret.subprocess_manager.flow.RunAction Action will let the user trigger your Runners with the usual flow.Action fonctionnalities.

By using the RunAction, you will be able to access your flow to configure your Runners runs. For example using project settings to get the project resolution and frame rate...


Define your process Runners:

from kabaret.subprocess_manager.runner_factory import Runner

class MyStuffToRun(Runner):

    ICON = ('icons.gui', 'team-lead')
    TAGS = ['Team', 'Lead', 'Blah']

    def supported_versions(self):
        return ['v1', 'v2]

    def executable(self):
        return '/path/to/{}/executable'.format(self.version)

    def argv(self):
        index = ''
        if self.version == 'v2':
            index = ''

        argv = ['--user', current_user, '--index', index]

        # To allow configured run, you must
        # use self.extra_argv:
        return argv+self.extra_argv

    def env(self):
        env = os.environ.copy()
        env['MY_STUFF_CONFIGVAR] = 'my_stuff_config_value'

        # To allow configured run, you must
        # use self.extra_argv:

        return env

There are more tunning options, like terminal hidding etc...

Everything is procedural (method call, no static/declarative thingy) so you're in full control ;)

If you want the Runner icon or tags to depend on the date or the current user your can override the classmethods runner_icon() and runner_tags()

Configure your Session:

In your session class, add the SubprocessView and the LaunchToolBar to the registered view types.

class MySession(gui.KabaretStandaloneGUISession):

    def register_view_types(self):
        super(MySession, self).register_view_types()

        type_name = self.register_view_type(SubprocessView)
        self.add_view(type_name, hidden=True)

        type_name = self.register_view_type(LauncherToolBar)
            type_name, hidden=False, 

In the _create_actors() method, you will create the SubprocessManager Actor and configure it with your generic runners:

    def _create_actors(self):
        # Add the Actor to the session:
        subprocess_manager = SubprocessManager(self)

        # Create your factory:
        site_factory = factories.create_new_factory('Studio Tools')

        # Ensure your Runners are known there:

        # Register your factory to the Actor:

Now you have an awesome toolbar with buttons to launch each version of the runner you registered \o/ (They appear grouped by the first tag of each Runner)

Configure your Project

Some Runners will be specific to flow type. For example, not all pipelines will supported USDView...

So you need to define Runners inside your flow and have them show up in the toolbar only if you entered a project using this flow.

One way to do so is to use your Project's _fill_ui() as a trigger for installation:

class MyProject(flow.Object):

    episodes = flow.Child(Episodes).ui(show_filter=True)
    admin = flow.Child(Admin)


    def ensure_runners_loaded(self):
        session = self.root().session()
        subprocess_manager = session.get_actor('SubprocessManager')

        if self._RUNNERS_FACTORY is None:
            self._RUNNERS_FACTORY = subprocess_manager.create_new_factory(
                'Best Pipeline Eva'

    def _fill_ui(self, ui):
        # We can't setup runner in the constructor because
        # it needs access to self.session() which is set
        # *after* constructor :/
        # So we set them up the first time the project is displayed:
        if self._RUNNERS_FACTORY is None:

Run Subprocesses from the Flow

This is the most important since running stuff within a context is what we want 90% of time :p

In order to use your Runners from the flow and to have them use some values from your flow (filename, env var, flag, etc...), you will use the Action kabaret.subprocess_manager.flow.RunAction:

Here we define a RunAction that uses the parent's filename Param as a command line argument to a TextEdit runner:

class EditAction(RunAction):

    _scene = flow.Parent()

    def runner_name_and_tags(self):
        # Specifying None for tags means 'accept all'
        return 'TextEdit', None

    def needs_dialog(self):
        return False

    def extra_argv(self):
        # All strings returned here will be
        # appended to the runner command line:
        return [self._scene_.filename.get()]

You can also provide extra_env, let the user select the version of the runner, etc...

You can ask for more examples and support in the kabaret discord:

Happy Running ! :)

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kabaret.subprocess_manager-1.0.7.tar.gz (30.1 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page