Entrypoints

class q2_sdk.core.entrypoint.BaseEntryPoint

This is largely based off of the way Django organizes its Management Commands.

We do not need the power nor the flexibility that Django provides here, but it’s such a great pattern, it seems a shame to reinvent the wheel.

Basic idea: When running q2 at the command line, each file in the entrypoints folder of the SDK that ends in .py and does not start with _ is loaded. The parser is built off of what lives in add_arguments, while the action when run is determined by what lives in the handle function.

name: The command line parameter to be called (name=inspect means q2 inspect) help: Help message at the command line run_forever: If True, the asyncio loop will not exit after completion of handle function

property cache: Q2CacheClient

Instantiates a communication object between this extension and the pymemcache library. Cache can be configured via configuration.settings.CACHE.

Returns:

Instantiated pymemcache.cache.Client object

async async_run_forever_cleanup()

Hook as the run forever loop is terminating

run_forever_cleanup()

Hook for any cleanup necessary after run forever loop is terminated

add_arguments(parser)

Hook for subclassed EntryPoints to add custom arguments.

setup(logger)

Registers a logger in self.logger with all appropriate filters

generate_completion()

Works with q2 install_completion to create completion file in <root>/.antilles/entrypoints/<self.name>

async handle(*args, **kwargs)

The actual logic of the command. Subclasses must implement this method.

class q2_sdk.core.entrypoint_periodic.PeriodicJobEntryPoint

Same as BaseEntryPoint but specialized for registering with a Cron-like system. Adds more logging and defined success/failure hooks for easier debugging when run overnight, for instance.

setup(logger)

Ensures periodic jobs will log in the format of request handlers

async run()

Code here will be executed when periodic job starts

async handle(*args, **kwargs)

The actual logic of the command. Subclasses must implement this method.

async on_pre_run()

Called before run()

async on_success(run_result)

Called after run() if it did not raise an Exception.

If run() returns a value, it will be passed as the first parameter here.

async on_failure(err, failure_dict)

Called after run() if it raised an Exception.

By default, all local variables (failure_dict) will be logged on failure. This hook exists if additional failure processing is desired.