@nesk/puphpeteer

PuPHPeteer

A Puppeteer bridge for PHP, supporting the full API. Based on Rialto, a package to manage Node resources from PHP.

Here are some examples borrowed from Puppeteer's documentation and adapted to PHP's syntax:

Example - navigating to https://example.com and saving a screenshot as example.png:

use Nesk\Puphpeteer\Puppeteer;

$puppeteer = new Puppeteer;
$browser = $puppeteer->launch();

$page = $browser->newPage();
$page->goto('https://example.com');
$page->screenshot(['path' => 'example.png']);

$browser->close();

Example - evaluate a script in the context of the page:

use Nesk\Puphpeteer\Puppeteer;
use Nesk\Rialto\Data\JsFunction;

$puppeteer = new Puppeteer;

$browser = $puppeteer->launch();
$page = $browser->newPage();
$page->goto('https://example.com');

// Get the "viewport" of the page, as reported by the page.
$dimensions = $page->evaluate(JsFunction::create("
    return {
        width: document.documentElement.clientWidth,
        height: document.documentElement.clientHeight,
        deviceScaleFactor: window.devicePixelRatio
    };
"));

printf('Dimensions: %s', print_r($dimensions, true));

$browser->close();

Requirements and installation

This package requires PHP >= 7.1 and Node >= 8.

Install it with these two command lines:

composer require nesk/puphpeteer
npm install @nesk/puphpeteer

Notable differences between PuPHPeteer and Puppeteer

Puppeteer's class must be instanciated

Instead of requiring Puppeteer:

const puppeteer = require('puppeteer');

You have to instanciate the Puppeteer class:

$puppeteer = new Puppeteer;

This will create a new Node process controlled by PHP.

You can also pass some options to the constructor, see Rialto's documentation.

Note: If you use some timeouts higher than 30 seconds in Puppeteer's API, you will have to set a higher value for the read_timeout option (default: 35):

$puppeteer = new Puppeteer([
    'read_timeout' => 65, // In seconds
]);

$puppeteer->launch()->newPage()->goto($url, [
    'timeout' => 60000, // In milliseconds
]);

No need to use the await keyword

With PuPHPeteer, every method call or property getting/setting is synchronous.

Some methods have been aliased

The following methods have been aliased because PHP doesn't support the $ character in method names:

• $ => querySelector
• $$ => querySelectorAll
• $x => querySelectorXPath
• $eval => querySelectorEval
• $$eval => querySelectorAllEval

Use these aliases just like you would have used the original methods:

$divs = $page->querySelectorAll('div');

Evaluated functions must be created with JsFunction

Functions evaluated in the context of the page must be written with the JsFunction class, the body of these functions must be written in JavaScript instead of PHP.

use Nesk\Rialto\Data\JsFunction;

$pageFunction = JsFunction::create(['element'], "
    return element.textContent;
");

Exceptions must be catched with ->tryCatch

If an error occurs in Node, a Node\FatalException will be thrown and the process closed, you will have to create a new instance of Puppeteer.

To avoid that, you can ask Node to catch these errors by prepending your instruction with ->tryCatch:

use Nesk\Rialto\Exceptions\Node;

try {
    $page->tryCatch->goto('invalid_url');
} catch (Node\Exception $exception) {
    // Handle the exception...
}

Instead, a Node\Exception will be thrown, the Node process will stay alive and usable.

License

The MIT License (MIT). Please see License File for more information.

Changelog

[1.0.0] - 2018-06-05

Changed

• Change PHP's vendor name from extractr-io to nesk
• Change NPM's scope name from @extractr-io to @nesk
• Upgrade to Rialto v1