FuncExtensionBase Class

An abstract class defines the life-cycle hooks which to be implemented by customer's extension.

Everytime when a new extension is initialized in customer function scripts, the ExtensionManager._func_exts field records the extension to this specific function name.

Constructor for extension. This needs to be implemented and ensure super().init(file_path) is called.

The initializer serializes the extension to a tree. This speeds up the worker lookup and reduce the overhead on each invocation. _func_exts[<trigger_name>].<hook_name>.(ext_name, ext_impl) (e.g. _func_exts['HttpTrigger'].pre_invocation.ext_impl)

Inheritance

builtins.object

FuncExtensionBase

Constructor

FuncExtensionBase(*args, **kwargs)

Parameters

Name	Description
file_path Required	str The name of trigger the extension attaches to (e.g. file).

Methods

post_function_load	This hook will be called right after a customer's function loaded. In this stage, the customer's logger is not fully initialized, so it is not provided. Please use print() to emit message if necessary.
post_invocation	This hook will be called right after a customer's function is executed.
pre_invocation	This hook will be called right before customer's function is being executed.

post_function_load

This hook will be called right after a customer's function loaded. In this stage, the customer's logger is not fully initialized, so it is not provided. Please use print() to emit message if necessary.

post_function_load(function_name: str, function_directory: str, *args, **kwargs) -> None

Parameters

Name	Description
function_name Required	str The name of customer's function (e.g. HttpTrigger)
function_directory Required	str The path to customer's function directory (e.g. /home/site/wwwroot/HttpTrigger)

post_invocation

This hook will be called right after a customer's function is executed.

post_invocation(logger: Logger, context: Context, func_args: Dict[str, object] | None = None, func_ret: object | None = None, *args, **kwargs) -> None

Parameters

Name	Description
logger Required	Logger A logger provided by Python worker. Extension developer should use this logger to emit telemetry to Azure Functions customers.
context Required	Context This will include the function_name, function_directory and an invocation_id of this specific invocation.
func_args Required	Dict[str, object] Arguments that are passed into the Azure Functions. The name of each parameter is defined in function.json. Extension developers may also want to do isinstance() check if you want to apply operations to specific trigger types or input binding types. Default value: None
func_ret Required	Optional[object] Return value from Azure Functions. This is usually the value defined in function.json $return section. Extension developers may also want to do isinstance() check if you want to apply operations to specific types or input binding types. Default value: None

pre_invocation

This hook will be called right before customer's function is being executed.

pre_invocation(logger: Logger, context: Context, func_args: Dict[str, object] | None = None, *args, **kwargs) -> None

Parameters

Name	Description
logger Required	Logger A logger provided by Python worker. Extension developer should use this logger to emit telemetry to Azure Functions customers.
context Required	Context This will include the function_name, function_directory and an invocation_id of this specific invocation.
func_args Required	Dict[str, object] Arguments that are passed into the Azure Functions. The name of each parameter is defined in function.json. Extension developers may also want to do isinstance() check if you want to apply operations to specific trigger types or input binding types. Default value: None