Skip to content

Latest commit

 

History

History
138 lines (84 loc) · 3.49 KB

_UseCaseExecutorAPI.md

File metadata and controls

138 lines (84 loc) · 3.49 KB
Raw
id title
usecaseexecutor-api
UseCaseExecutor

Interface

export declare class UseCaseExecutorImpl<T extends UseCaseLike> extends Dispatcher implements UseCaseExecutor<T> {
    useCase: T;
    onReleaseOnce(handler: () => void): void;
    executor(executor: (useCase: Pick<T, "execute">) => any): Promise<void>;
    execute<P extends Arguments<T["execute"]>>(...args: P): Promise<void>;

export declare class UseCaseExecutorImpl<T extends UseCaseLike> extends Dispatcher implements UseCaseExecutor<T> {

UseCaseExecutor is a helper class for executing UseCase.

You can not create the instance of UseCaseExecutor directory. You can get the instance by Context#useCase(useCase),

useCase: T;

A executable useCase

onReleaseOnce(handler: () => void): void;

Call handler when this UseCaseExecutor will be released

param handler

executor(executor: (useCase: Pick<T, "execute">) => any): Promise<void>;

Stability: Deprecated(Previously: experimental)

  • This feature is subject to change. It may change or be removed in future versions.
  • If you inserting in this, please see #193

Similar to execute(arguments), but it accept an executor function insteadof arguments executor(useCase => useCase.execute()) return a Promise object that resolved with undefined.

This method is type-safe. It is useful for TypeScript.

Example

context.useCase(new MyUseCase())
 .executor(useCase => useCase.execute("value"))
 .then(() => {
   console.log("test");
 });

Notes

What is difference between executor(executor) and execute(arguments)?

The execute(arguments) is a alias of following codes:

context.useCase(new MyUseCase())
 .execute("value")
// ===
context.useCase(new MyUseCase())
 .executor(useCase => useCase.execute("value"))

Why executor's result always to be undefined?

execute() return a Promise that will resolve undefined by design. In CQRS, the command always have returned void type.

So, execute() only return command result that is success or failure. You should not relay on the result value of the useCase#execute.

deprecated Use execute() instead of executor() Almin 0.18+ make execute type complete. It will be remove in the future.

Please apply migration scripts: #356

execute<P extends Arguments<T["execute"]>>(...args: P): Promise<void>;

Execute UseCase instance. UseCase is a executable object that has execute method.

This method invoke UseCase's execute method and return a promise. The promise will be resolved when the UseCase is completed finished.

Why executor's result always to be undefined?

execute() return a Promise that will resolve undefined by design. In CQRS, the command always have returned void type.

So, execute() only return command result that is success or failure. You should not relay on the result value of the useCase#execute.

Notes

Added: Almin 0.17.0+

execute() support type check in Almin 0.17.0. However, it has a limitation about argument lengths. For more details, please see #107 (comment)

Added: Almin 0.18.0+

execute() support type check completely. No more need to use executor() for typing.