Processors

Processors in Nuke studio build upon the event system used in ftrack. As such, each processor’s discover and launch method receives a single argument which is an instance of ftrack.Event.

The built-in processors can be extended by creating new processors and placing them in a directory. Then configure the environment by setting the FTRACK_EVENT_PLUGIN_PATH environment variable.

The processors make use of two events ftrack.processor.discover and ftrack.processor.launch that are emitted for each object being created when running Export project.

ftrack.processor.discover

The discover event, ftrack.processor.discover, is emitted to discover which processors are available for a certain object.

The structure of the event is:

ftrack.Event(
    topic='ftrack.processor.discover',
    data=dict(
        name='010',
        object_type='Shot',
        application_object=track_item
    )
)

Where track_item is a core.Hiero.Python.TrackItem corresponding to the object.

To make a processor available for on all shot creations you have to subscribe to the event hub:

def discover(event):
    '''Return processor configuration for *event*.'''
    return dict(
        defaults=dict(
            # Default values.
        ),
        name='Review',
        processor_name='processor.review',
        asset_name='BG'
    )

ftrack.EVENT_HUB.subscribe(
    'topic=ftrack.processor.discover and data.object_type=shot',
    discover
)

The dictionary returned by the discover contains the following values:

  • defaults - Must be a dict and is rendered in the UI to communicate to the end-user any defaults that will be used when running the processor.
  • name - A nice name of the processor displayed in the UI.
  • processor_name - The processor name.
  • asset_name - An optional asset name that is used to group the processors in the UI. A corresponding asset version will be created based on the name.
../_images/processors.png

ftrack.processor.launch

The launch event, ftrack.processor.launch, is emitted to launch the processor when an object has become available on the ftrack server. E.g. a new Shot or Task has been created.

The structure of the event is:

ftrack.Event(
    topic='ftrack.processor.launch',
    data=dict(
        name='Review',
        input=data
    )
)

Where data is a dictionary containing contextual information about the object being processed. The dictionary contains the following information: resolution, source_in, source_out, source_file, destination_in, destination_out, fps, offset, entity_id, entity_type, handles. Where entity_id and entity_type points to the object being created in ftrack.

Optional values are:

  • asset_version_id - Id of the asset version created if asset_name was passed in the defaults dictionary when the processor was discovered.
  • component_name - the suggested name of the component inferred from the processor nice name passed in the discover event.

To make a processor launch on all shot creations you have to subscribe to the event hub:

def launch(event):
    '''Launch processor with *event*.'''
    ...


ftrack.EVENT_HUB.subscribe(
    'topic=ftrack.processor.launch and data.name=processor.review',
    launch
)