mirror of
https://github.com/danog/amp.git
synced 2024-12-11 00:49:48 +01:00
165 lines
7.4 KiB
Markdown
165 lines
7.4 KiB
Markdown
---
|
|
layout: docs
|
|
title: Promises
|
|
permalink: /promises/
|
|
---
|
|
A `Promise` is an object representing the eventual result of an asynchronous operation.
|
|
There are three states:
|
|
|
|
- **Success**: The promise resolved successfully.
|
|
- **Failure**: The promise failed.
|
|
- **Pending**: The promise has not been resolved yet.
|
|
|
|
A successful resolution is like returning a value in synchronous code while failing a promise is like throwing an exception.
|
|
|
|
Promises are the basic unit of concurrency in asynchronous applications.
|
|
In Amp they implement the `Amp\Promise` interface.
|
|
These objects should be thought of as placeholders for values or tasks that might not be complete immediately.
|
|
|
|
|
|
Another way to approach asynchronous APIs is using callbacks that are passed when the operation is started.
|
|
|
|
```php
|
|
doSomething(function ($error, $value) {
|
|
if ($error) {
|
|
/* ... */
|
|
} else {
|
|
/* ... */
|
|
}
|
|
});
|
|
```
|
|
|
|
The callback approach has several drawbacks.
|
|
|
|
- Passing callbacks and doing further actions in them that depend on the result of the first action gets messy really quickly.
|
|
- An explicit callback is required as input parameter to the function, and the return value is simply unused. There's no way to use this API without involving a callback.
|
|
|
|
That's where promises come into play.
|
|
They're simple placeholders that are returned and allow a callback (or several callbacks) to be registered.
|
|
|
|
```php
|
|
doSomething()->onResolve(function ($error, $value) {
|
|
if ($error) {
|
|
/* ... */
|
|
} else {
|
|
/* ... */
|
|
}
|
|
});
|
|
```
|
|
|
|
This doesn't seem a lot better at first sight, we have just moved the callback.
|
|
But in fact this enabled a lot.
|
|
We can now write helper functions like [`Amp\Promise\all()`](https://amphp.org/amp/promises/combinators#all) which subscribe to several of those placeholders and combine them. We don't have to write any complicated code to combine the results of several callbacks.
|
|
|
|
But the most important improvement of promises is that they allow writing [coroutines](https://amphp.org/amp/coroutines/), which completely eliminate the need for _any_ callbacks.
|
|
|
|
Coroutines make use of PHP's generators.
|
|
Every time a promise is `yield`ed, the coroutine subscribes to the promise and automatically continues it once the promise resolved.
|
|
On successful resolution the coroutine will send the resolution value into the generator using [`Generator::send()`](https://secure.php.net/generator.send).
|
|
On failure it will throw the exception into the generator using [`Generator::throw()`](https://secure.php.net/generator.throw).
|
|
This allows writing asynchronous code almost like synchronous code.
|
|
|
|
{:.note}
|
|
> Amp's `Promise` interface **does not** conform to the "Thenables" abstraction common in JavaScript promise implementations. Chaining `.then()` calls is a suboptimal method for avoiding callback hell in a world with generator coroutines. Instead, Amp utilizes PHP generators as described above.
|
|
>
|
|
> However, as ReactPHP is another wide-spread implementation, we also accept any `React\Promise\PromiseInterface` where we accept instances of `Amp\Promise`. In case of custom implementations not implementing `React\Promise\PromiseInterface`, `Amp\Promise\adapt()` can be used to adapt any object having a `then` or `done` method.
|
|
|
|
## Promise Consumption
|
|
|
|
```php
|
|
interface Promise {
|
|
public function onResolve(callable $onResolve);
|
|
}
|
|
```
|
|
|
|
In its simplest form the `Amp\Promise` aggregates callbacks for dealing with results once they eventually resolve. While most code will not interact with this API directly thanks to [coroutines](../coroutines/), let's take a quick look at the one simple API method exposed on `Amp\Promise` implementations:
|
|
|
|
| Parameter | Callback Signature |
|
|
| ------------ | ------------------------------------------ |
|
|
| `$onResolve` | `function ($error = null, $result = null)` |
|
|
|
|
`Amp\Promise::onResolve()` accepts an error-first callback. This callback is responsible for reacting to the eventual result represented by the promise placeholder. For example:
|
|
|
|
```php
|
|
<?php
|
|
|
|
$promise = someFunctionThatReturnsAPromise();
|
|
$promise->onResolve(function (Throwable $error = null, $result = null) {
|
|
if ($error) {
|
|
printf(
|
|
"Something went wrong:\n%s\n",
|
|
$error->getMessage()
|
|
);
|
|
} else {
|
|
printf(
|
|
"Hurray! Our result is:\n%s\n",
|
|
print_r($result, true)
|
|
);
|
|
}
|
|
});
|
|
```
|
|
|
|
Those familiar with JavaScript code generally reflect that the above interface quickly devolves into ["callback hell"](http://callbackhell.com/), and they're correct. We will shortly see how to avoid this problem in the [coroutines](../coroutines/README.md) section.
|
|
|
|
## Promise Creation
|
|
|
|
Promises can be created in several different ways. Most code will use [`Amp\call()`](https://amphp.org/amp/coroutines/helpers#call) which takes a function and runs it as coroutine if it returns a `Generator`.
|
|
|
|
### Success and Failure
|
|
|
|
Sometimes values are immediately available. This might be due to them being cached, but can also be the case if an interface mandates a promise to be returned to allow for async I/O but the specific implementation always having the result directly available. In these cases `Amp\Success` and `Amp\Failure` can be used to construct an immediately resolved promise. `Amp\Success` accepts a resolution value. `Amp\Failure` accepts an exception as failure reason.
|
|
|
|
### Deferred
|
|
|
|
{:.note}
|
|
> The `Deferred` API described below is an advanced API that many applications probably don't need. Use [`Amp\call()`](https://amphp.org/amp/coroutines/helpers#call) or [promise combinators](https://amphp.org/amp/promises/combinators) instead where possible.
|
|
|
|
`Amp\Deferred` is the abstraction responsible for resolving future values once they become available. A library that resolves values asynchronously creates an `Amp\Deferred` and uses it to return an `Amp\Promise` to API consumers. Once the async library determines that the value is ready it resolves the promise held by the API consumer using methods on the linked promisor.
|
|
|
|
```php
|
|
final class Deferred
|
|
{
|
|
public function promise(): Promise;
|
|
public function resolve($result = null);
|
|
public function fail(Throwable $error);
|
|
}
|
|
```
|
|
|
|
#### `promise()`
|
|
|
|
Returns the corresponding `Promise` instance. `Deferred` and `Promise` are separated, so the consumer of the promise can't fulfill it. You should always return `$deferred->promise()` to API consumers. If you're passing `Deferred` objects around, you're probably doing something wrong.
|
|
|
|
#### `resolve()`
|
|
|
|
Resolves the promise with the first parameter as value, otherwise `null`. If a `Amp\Promise` is passed, the resolution will wait until the passed promise has been resolved. Invokes all registered `Promise::onResolve()` callbacks.
|
|
|
|
#### `fail()`
|
|
|
|
Makes the promise fail. Invokes all registered `Promise::onResolve()` callbacks with the passed `Throwable` as `$error` argument.
|
|
|
|
Here's a simple example of an async value producer `asyncMultiply()` creating a deferred and returning the associated promise to its API consumer.
|
|
|
|
```php
|
|
<?php // Example async producer using promisor
|
|
|
|
use Amp\Loop;
|
|
|
|
function asyncMultiply($x, $y)
|
|
{
|
|
// Create a new promisor
|
|
$deferred = new Amp\Deferred;
|
|
|
|
// Resolve the async result one second from now
|
|
Loop::delay($msDelay = 1000, function () use ($deferred, $x, $y) {
|
|
$deferred->resolve($x * $y);
|
|
});
|
|
|
|
return $deferred->promise();
|
|
}
|
|
|
|
$promise = asyncMultiply(6, 7);
|
|
$result = Amp\Promise\wait($promise);
|
|
|
|
var_dump($result); // int(42)
|
|
```
|