Events and Actions

Events

A pipeline is defined by creating an event table, linking events to the corresponding class or code (see pipelines_).

To define an event, we need to specify the following four elements: name, action, state, and next event.

The name is arbitrary and is the only connection between the pipeline, the framework and the corresponding code. This means that there is no way to directly call a function or a class other than by the name associated with it.

The state is arbitrary. While not widely used yet, it can be used to control the flow of the pipeline. For now, it is sufficient to know that once an event is triggered, the special variable context.state will be set to the value specified in this field.

The next event array element is used to create automatic chains of events, if that is desired. For example, if you are creating a basic CCD reduction pipeline, you could write events like this:

event_table = {
   "correct_bias":    ("subtract_bias", "bias_processing", "correct_overscan"),
   "correct_overscan: ("fit_and_sub_overscan", "overscan_processing", "correct_flat"),
   "correct_flat":    ("fit_and_div_flat", "flat_processing", None)
   }

If correct_bias is triggered, the framework would automatically proceed to trigger correct_overscan, and continue with correct_flat. At this point, the framework would encounter the event None and would not proceed further. The variable context.state would change value from bias_processing to overscan_processing to flat_processing.

Actions

The Base_pipeline offers a method to convert an Event into an Action.

This operation searches the namespace for classes or functions that match the action field of the event that has been triggered, and sends the resulting code to the framework for execution.

The actual execution depends on how the code is defined. If the code is contained in a class the framework would look for pre and post conditions and run the apply method if it is defined.

See primitives_ for further information.