celery-types

The celery-types package provides type stubs for the following projects:

• celery
• django-celery-results
• amqp
• kombu
• billiard
• vine
• ephem

Install the type stubs for Celery and its related ecosystem using pip.

pip install celery-types

Celery's CLI is built on top of click. The following classes define the base behavior for commands and options within the Celery ecosystem:

• CeleryCommand: A subclass of click.Command that holds a reference to a Celery app instance (app: Celery | None).
• CeleryDaemonCommand: A specialized CeleryCommand designed for daemonized processes.
• CeleryOption: A subclass of click.Option used for Celery-specific command options.
• DaemonOption: A specialized click.Option that includes a daemon_setting method for managing daemon configurations.

The MultiTool class provides high-level methods to control a _Cluster object (representing a group of Celery nodes):

The TermLogger class (which MultiTool inherits) manages how the CLI communicates with the user. You can configure the terminal environment using setup_terminal to control verbosity, color, and output streams.

Configuration Options for setup_terminal:

• stdout: The output stream for standard messages.
• stderr: The output stream for errors.
• nosplash: Boolean to disable the splash text.
• quiet: Boolean to suppress non-essential output.
• verbose: Boolean to enable detailed output.
• no_color: Boolean to disable colored terminal output.

# Configuring terminal behavior
logger.setup_terminal(
    stdout=None, 
    stderr=None, 
    quiet=False, 
    verbose=True, 
    no_color=False
)

The following utility functions are available for handling CLI-related logic:

• get_current_app(): Returns the current active Celery app instance.
• pformat(object, ...): A highly configurable pretty-printer for Python objects, supporting indentation, width, depth, and dictionary sorting.
• mlevel(level): Converts a log level (string or integer) into its integer representation.
• maybe_iso8601(dt): Attempts to parse a datetime or string into a datetime object using ISO8601 format.
• handle_preload_options(f): A decorator used to handle option preloading logic.

The CLIContext object provides utilities for managing the CLI environment and formatting output. It is associated with a Celery app instance and includes methods for styled printing and data formatting.

Key methods include:

• echo(message, **kwargs): Prints a message.
• error(message, **kwargs): Prints an error message.
• secho(message, **kwargs): Prints a message with ANSI color/style.
• pretty(n): Returns a formatted string representation of an object.
• pretty_dict_ok_error(n): Formats a dictionary for success or error states.
• pretty_list(n): Formats a list for display.
• say_chat(direction, title, body, show_body): Formats a chat-like message block.

When running the worker command, you can use several flags to control its behavior. Key configuration options include:

• --autoscale: Accepts a value that converts to a tuple of two integers (max_concurrency, min_concurrency) to manage worker process scaling.
• --celery-beat: A boolean flag to enable/disable Celery Beat functionality.
• --hostname: Sets the hostname for the worker.
• --workers-pool: Specifies the pool type (using WorkersPool choices).
• --fake-fork: (Represented by C_FAKEFORK) A string option for fork behavior.

Additionally, the underlying detach mechanism allows for daemonization with options like --logfile, --pidfile, --uid, --gid, --umask, and --workdir.

worker --autoscale 10,2 --hostname my-worker --workers-pool solo

The Celery CLI uses several specialized parameter types for parsing command-line arguments. When building or extending Celery commands, you can use these types to ensure correct validation and parsing of inputs.

COMMA_SEPARATED_LIST: CommaSeparatedList
LOG_LEVEL: LogLevel
ISO8601: ISO8601DateTime
ISO8601_OR_FLOAT: ISO8601DateTimeOrFloat
JSON_ARRAY: JsonArray
JSON_OBJECT: JsonObject