畫江湖之 PHP 多程式開發 【多工採集圖片】

<?php

include __DIR__ . '/vendor/autoload.php';//引入自動載入類

use Goutte\Client;//使用第三方採集庫

$client = new Client();

$links = [
    'http://www.nipic.com/topic/show_27192_1.html',
    'http://www.nipic.com/topic/show_27054_1.html',
    'http://www.nipic.com/topic/show_27085_1.html',
];//要爬的三個分類下的圖片

$pids = [];//定義一個子程式的陣列

foreach ($links as $url) {
    $pid = pcntl_fork();//fork 一個子程式
    switch ($pid) {
    case -1:
        die("Fork failed\n");
    case 0:

        $id = posix_getpid();//得到子程式id
        $pids[$id] = $id;//存入到子程式陣列id 中
        $data = [];

        $crawler = $client->request('GET', $url);
        $crawler->filter('.search-works-thumb')->each(function($node) use ($client, $id, &$data) {
            $url = $node->link()->getUri();

            $crawler = $client->request('GET', $url);
            $crawler->filter('#J_worksImg')->each(function($node) use ($id, &$data) {
                $src = $node->image()->getUri();

                $data[$id][] = $src;//存入訊息佇列 或者redis 都可以
                //echo $src . PHP_EOL;
            });
        });

        print_r($data);

        exit;//結束子程式

        break;
    default:
        sleep(2);
        break;
    }
}

while ( count($pids) ) {
    if (($id = pcntl_wait($status, WUNTRACED)) > 0) {//主程式 阻塞等待子程式完成任務
        unset($pids[$id]);
    }
}

echo "Done\n";