How to implement a web scraper in PHP?

Scraping generally encompasses 3 steps:

• first you GET or POST your request to a specified URL
• next you receive the html that is returned as the response
• finally you parse out of that html the text you'd like to scrape.

To accomplish steps 1 and 2, below is a simple php class which uses Curl to fetch webpages using either GET or POST. After you get the HTML back, you just use Regular Expressions to accomplish step 3 by parsing out the text you'd like to scrape.

For regular expressions, my favorite tutorial site is the following: Regular Expressions Tutorial

My Favorite program for working with RegExs is Regex Buddy. I would advise you to try the demo of that product even if you have no intention of buying it. It is an invaluable tool and will even generate code for your regexs you make in your language of choice (including php).

Usage:

$curl = new Curl();
$html = $curl->get("http://www.google.com");

// now, do your regex work against $html

PHP Class:

<?php

class Curl
{       

    public $cookieJar = "";

    public function __construct($cookieJarFile = 'cookies.txt') {
        $this->cookieJar = $cookieJarFile;
    }

    function setup()
    {


        $header = array();
        $header[0] = "Accept: text/xml,application/xml,application/xhtml+xml,";
        $header[0] .= "text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5";
        $header[] =  "Cache-Control: max-age=0";
        $header[] =  "Connection: keep-alive";
        $header[] = "Keep-Alive: 300";
        $header[] = "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7";
        $header[] = "Accept-Language: en-us,en;q=0.5";
        $header[] = "Pragma: "; // browsers keep this blank.


        curl_setopt($this->curl, CURLOPT_USERAGENT, 'Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.8.1.7) Gecko/20070914 Firefox/2.0.0.7');
        curl_setopt($this->curl, CURLOPT_HTTPHEADER, $header);
        curl_setopt($this->curl,CURLOPT_COOKIEJAR, $this->cookieJar); 
        curl_setopt($this->curl,CURLOPT_COOKIEFILE, $this->cookieJar);
        curl_setopt($this->curl,CURLOPT_AUTOREFERER, true);
        curl_setopt($this->curl,CURLOPT_FOLLOWLOCATION, true);
        curl_setopt($this->curl,CURLOPT_RETURNTRANSFER, true);  
    }


    function get($url)
    { 
        $this->curl = curl_init($url);
        $this->setup();

        return $this->request();
    }

    function getAll($reg,$str)
    {
        preg_match_all($reg,$str,$matches);
        return $matches[1];
    }

    function postForm($url, $fields, $referer='')
    {
        $this->curl = curl_init($url);
        $this->setup();
        curl_setopt($this->curl, CURLOPT_URL, $url);
        curl_setopt($this->curl, CURLOPT_POST, 1);
        curl_setopt($this->curl, CURLOPT_REFERER, $referer);
        curl_setopt($this->curl, CURLOPT_POSTFIELDS, $fields);
        return $this->request();
    }

    function getInfo($info)
    {
        $info = ($info == 'lasturl') ? curl_getinfo($this->curl, CURLINFO_EFFECTIVE_URL) : curl_getinfo($this->curl, $info);
        return $info;
    }

    function request()
    {
        return curl_exec($this->curl);
    }
}

?>

If you need something that is easy to maintain, rather than fast to execute, it could help to use a scriptable browser, such as SimpleTest's.

file_get_contents() can take a remote URL and give you the source. You can then use regular expressions (with the Perl-compatible functions) to grab what you need.

Out of curiosity, what are you trying to scrape?

I'd either use libcurl or Perl's LWP (libwww for perl). Is there a libwww for php?

Scraper class from my framework:

<?php

/*
    Example:

    $site = $this->load->cls('scraper', 'http://www.anysite.com');
    $excss = $site->getExternalCSS();
    $incss = $site->getInternalCSS();
    $ids = $site->getIds();
    $classes = $site->getClasses();
    $spans = $site->getSpans(); 

    print '<pre>';
    print_r($excss);
    print_r($incss);
    print_r($ids);
    print_r($classes);
    print_r($spans);        

*/

class scraper
{
    private $url = '';

    public function __construct($url)
    {
        $this->url = file_get_contents("$url");
    }

    public function getInternalCSS()
    {
        $tmp = preg_match_all('/(style=")(.*?)(")/is', $this->url, $patterns);
        $result = array();
        array_push($result, $patterns[2]);
        array_push($result, count($patterns[2]));
        return $result;
    }

    public function getExternalCSS()
    {
        $tmp = preg_match_all('/(href=")(\w.*\.css)"/i', $this->url, $patterns);
        $result = array();
        array_push($result, $patterns[2]);
        array_push($result, count($patterns[2]));
        return $result;
    }

    public function getIds()
    {
        $tmp = preg_match_all('/(id="(\w*)")/is', $this->url, $patterns);
        $result = array();
        array_push($result, $patterns[2]);
        array_push($result, count($patterns[2]));
        return $result;
    }

    public function getClasses()
    {
        $tmp = preg_match_all('/(class="(\w*)")/is', $this->url, $patterns);
        $result = array();
        array_push($result, $patterns[2]);
        array_push($result, count($patterns[2]));
        return $result;
    }

    public function getSpans(){
        $tmp = preg_match_all('/(<span>)(.*)(<\/span>)/', $this->url, $patterns);
        $result = array();
        array_push($result, $patterns[2]);
        array_push($result, count($patterns[2]));
        return $result;
    }

}
?>

The curl library allows you to download web pages. You should look into regular expressions for doing the scraping.