Helper Flows

There are two kinds of helper interactions.

• Interactive Flow
• Headless Flow

Interactive Flow

The helper follows a logical flow of events, which sandwich a custom routine between a start and stop procedure. We will review the basic flow here, and you can read more about configuration of these events under /helpme/docs-options. The events look like the following:

run --> start --> collect --> submit

And it all starts with “run,” the function that tells the helper to start execution of the flow.

run

For each of the below, the base client has a “run” function that executes the following events:

• start
• collect
• submit

Within each function above, the helper class can subclass an equivalently named (but with an underscore) function that, if it exists, will be called to extend behavior of the step.

start

Start is primarily a step to call self.speak(), where the helper announces itself. If the helper has implemented a self._speak() to add information to show the user here, then this function is called too.

collect

collect is the main guts of the collection and recording process. During this time, the helper section defined in the application helpme.cfg is parsed over, and each performs some collection or recording action. For more detail on these actions, see the documentation on helper options.

submit

On submit, the helper is expected to have a json data structure under self.data that should be parsed to submit the help request to the helper’s specific endpoint. The fields that are specified for collection in the configuration file, typically

• user_prompt_<name>
• record_asciinema
• record_environment

Will be provided in the dictionary under indices of the same name. For the asciinema recording, the fullpath to the file is provided and not the asciinema json itself.

Headless Flow

A headless flow can optionally be implemented for a helper, and it generally goes through the same steps as above, but with minimal user interaction. This means that the calling function usually supplies needed arguments. This flow is most appropriate for integration with other python libraries that might want to automatically submit issues. For an example, see the GitHub headless flow.