CloudFactory is a nice and powerful library for building crawlers in PHP. It provides a simple, human readable API while keeping all power of cURL under the hood. Key feature of CloudFactory is Request Validation. When using proxies in multi-threaded environment you will often get lots of problems with timeouts, response spoofing, HTTP/proxy errors and invalid content.

The goal of CloudFactory is to minimize the effort needed for 100% valid content delivery, and you can achieve that by using this library.

###Requirements

• PHP 7.0
• php-curl
• mbstring
• guzzle

composer require exfriend/cloudfactory

<?php

require 'vendor/autoload.php';

$engine = new \Exfriend\CloudFactory\Engine();
$request = new \Exfriend\CloudFactory\Request('http://httpbin.org/get');
$engine->run($request);

print_r($request->response);

$request = (new \Exfriend\CloudFactory\Request('http://httpbin.org/get'))->setOpt(CURLOPT_FOLLOWLOCATION, true);

You can set raw curl options as well as use some helpers available on Request object.

$request = (new \Exfriend\CloudFactory\Request('http://httpbin.org/post'))
           ->setOpt(CURLOPT_FOLLOWLOCATION, true)
           ->withProxy('127.0.0.1:8080', CURLPROXY_SOCKS5)
           ->sendHeaders(['Foo' => 'Bar'])
           ->post(['user' => 'admin', 'pass' => 'secret'])
           ->withTimeouts(10, 5)
           ->withSsl();

You can set almost all of the options directly on the Engine instance:

$engine = new Engine()->withUserAgent()->withSsl()->withCookies('cookie.txt');

$engine = (new \Exfriend\CloudFactory\Engine())->setThreads(25);

for ($i = 0; $i < 100; $i++) {
    $request = new \Exfriend\CloudFactory\Request('http://httpbin.org/get?i=' . $i);
    $engine->addRequest($request);
}

$engine->run();

foreach ($engine->requests->processed() as $request) {
    print_r($request->response);
}

You can set priority for each request by using ->withPriority(int). This way when it comes to pulling the request from the queue, they will be sorted by priority.

The more priority you set, the higher your request will be in the queue.

$request = (new Request('http://httpbin.org/get'))
    ->maxTries(3)
    ->store('i', $i)// to pass through to the callback
    ->validateUsing(function (Request $request) { // using closure
        // must return boolean
        return strpos($request->response, 'origin') !== false;
    })
    ->onSuccess('request_success')// using string
    ->onSuccess(function ($r) {
        echo 'You can stack callbacks of one type' . PHP_EOL;
    })
    ->onFail(function ($r) {
        echo "Request failed {$r->tries_current} times of {$r->tries_max}: " . $r->url;
    })
    ->onLastFail(function ($r) {
        echo "Request failed last time: " . $r->url;
    });

You can validate your responses using validateWith( callable ) when building a request. Your callable should accept the instance of Request as a parameter and return boolean true or false as a validity indicator.

Depending on the validity, engine will call callbacks provided with onSuccess or onFail or onLastFail.

You can set maxTries( int ) to inform the engine how many times the request should be repeated before it fails the last time.

Note that on the last failed try both onFail and onLastFail callback groups will be called.

You can use store(key,value) method when creating the request to pass any additional data on the storage bag of your request. You can access it from callback like this:

...
  ->store( 'user', $user )
  ->onSuccess(function($request)){
      $user = $request->storage->get('user');
  })
...

You can add new requests while engine is still running e.g. from callback.

   ->onSuccess(function($r)use($engine){
       $newRequest = new Request('http://site.com/simething_else')
                        ->withSsl()
                        ->onSuccess('some_other_callback');
       $engine->addRequest($newRequest);
   })

When you add a request with $engine->addRequest() it gets pushed to $engine->requests collection. It extends Illuminate\Support\Collection which is the powerful collection pattern implementation used in Laravel. This means you have all the perks like $htmls = $engine->requests()->processed()->pluck('response') to get all responses as array.

Full documentation on collections can be found on the Official Laravel Documentation pages.

You can process the results using $engine->requests collection or/and using callbacks.

You can add multiple callbacks for each state.

   ...
     ->onSuccess('parse_content_and_save_it_to_db')
     ->onSuccess('notify_other_class')
     ->onFail('show_notification')
     ->onFail('change_proxy')
     ->onLastFail([$this,'send_some_email'])
   ...

This gives you a possibility to easily write for example proxy rotating trait for your own descendant of Request class.

For more examples check out the ./examples directory.

###Contributing

This package is work-in-progress.