[Gif Spider] Fetch the Gif images on the page of http://www.ik123.com/q/tuku/gaoxiao/.

composer.json

{
    "require": {
        "bupt1987/html-parser": "dev-master"
    }
}

PHP_Spider.php

<?php

require __DIR__ . 'vendor/autoload.php';
define('IMG_FOLDER', '/tmp/img/');

function show($msg)
{
    echo $msg, PHP_EOL;
}

function fetch($url, $reffer)
{
    $header = [
        'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/52.0.2743.116 Safari/537.36',
        'Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
        'Referer: ' . $reffer,
        'Accept-Encoding: gzip, deflate, sdch',
        'Accept-Language: zh-CN,en-US;q=0.8,en;q=0.6',
    ];

    $ch = curl_init();
    curl_setopt ($ch, CURLOPT_URL, $url);
    curl_setopt($ch,CURLOPT_HTTPHEADER, $header);
    curl_setopt ($ch, CURLOPT_RETURNTRANSFER, 1);
    $file_contents = curl_exec($ch);
    curl_close($ch);

    return $file_contents;
}

function getDom($url)
{
    $html = file_get_contents($url);
    $html_dom = new \HtmlParser\ParserDom($html);

    return $html_dom;
}

function download($url)
{
    $html_dom = getDom($url);
    $arr = $html_dom->find('img');

    foreach ($arr as $img){
        $img_url = $img->getAttr('src');

        show($img_url);

        $tmp = explode('/', $img_url);
        $filename = IMG_FOLDER . end($tmp);

        if (file_exists($filename)) {
            show('Existed!');
            continue;
        }

        $data = fetch($img_url, $url);
        file_put_contents($filename, $data);
    }
}

function main()
{
    foreach (range(1, 86) as $i) {
        $url_pool = [];

        $url = "http://www.ik123.com/q/tuku/gaoxiao/list_14_{$i}.html";
        $dom = getDom($url);
        $arr = $dom->find('.img a');

        foreach ($arr as $e) {
            $url = $e->getAttr('href');
            $url_pool[] = $url;
        }

        $url_pool = array_unique($url_pool);

        foreach ($url_pool as $url) {
            download($url);
        }
    }
}

main();

# eof