PHP采集www.php.cn的文章,并存入数据库。

[PHP]代码

<?php

class Fork36kr
{
    private $start;
    private $end;
    private $number = 0;
    private $dsn;
    private $user;
    private $password;
    private $pdo;

    /**
     * @param int $start    采集起点文章id
     * @param int $end   采集终点文章id
     * @param string $dsn   PDO数据源
     * @param string $user  数据库用户名
     * @param string $password  数据库密码
     */
    public function __construct($start=200100, $end=206670,$dsn='',$user='',$password='')
    {
        $this->start = $start;
        $this->end = $end;
        $this->dsn = $dsn;
        $this->user = $user;
        $this->password = $password;
        if($dsn)
        {
            $this->pdo = new PDO($this->dsn,$this->user,$this->password);
        }
    }

    public function fork()
    {
        $ch = curl_init();
        curl_setopt($ch, CURLOPT_HEADER, 0);
        curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);

        echo "<<<<< OK. Start Fork 36kr >>>>>\n";
        for($i=$this->start;$i<=$this->end;$i++)
        {
            $url = "http://www.36kr.com/p/".$i.".html";
            curl_setopt($ch, CURLOPT_URL, $url);
            $page = curl_exec($ch);

            if(curl_getinfo($ch)['http_code']==200)
            {
                $t = preg_match('#<h1 class="entry-title sep10">.*</h1>#', $page, $title);
                $c = preg_match('#<p class="mainContent sep-10">.*</p>#Us', $page, $content);
                if($t&&$c)
                {
                    $title = strip_tags($title[0]);
                    $content = strip_tags($content[0]);
                    //$content = strip_tags($content[0],'<p><a>');  //保留<p>和<a>标记
                    echo $url.','.$title."\n";
                    $this->number++;
                }
                if($this->dsn)
                {
                    $this->save($title,$content,$url);
                }
            }
        }
        echo '<<<< Fork Over! Total： '.$this->number.'  >>>>';
    }

    private function save($title,$content,$url)
    {
        $sql = "INSERT INTO `36kr` (`id`,`title`,`content`,`url`) VALUES (null,:title,:content,:url)";
        $stmt = $this->pdo->prepare($sql);
        $stmt->bindParam(':title',$title);
        $stmt->bindParam(':content',$content);
        $stmt->bindParam(':url',$url);
        $stmt->execute();
    }
}

$dsn = 'mysql:host=localhost;dbname=test';
$user = 'root';
$password= 'root';

$kr = new Fork36kr(200100,206670,$dsn,$user,$password);
$kr->fork();

[文件] phpcn.sql

-- phpMyAdmin SQL Dump
-- version 4.0.5
-- http://www.php.cn/
--
-- 主机: localhost
-- 生成日期: 2013 �?10 �?03 �?00:36
-- 服务器版本: 5.6.12-log
-- PHP 版本: 5.5.3

SET SQL_MODE = "NO_AUTO_VALUE_ON_ZERO";
SET time_zone = "+00:00";


/*!40101 SET @OLD_CHARACTER_SET_CLIENT=@@CHARACTER_SET_CLIENT */;
/*!40101 SET @OLD_CHARACTER_SET_RESULTS=@@CHARACTER_SET_RESULTS */;
/*!40101 SET @OLD_COLLATION_CONNECTION=@@COLLATION_CONNECTION */;
/*!40101 SET NAMES utf8 */;

--
-- 数据库: `test`
--

-- --------------------------------------------------------

--
-- 表的结构 `36kr`
--

CREATE TABLE IF NOT EXISTS `36kr` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `title` varchar(128) NOT NULL,
  `content` text NOT NULL,
  `url` varchar(128) NOT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 AUTO_INCREMENT=1 ;

/*!40101 SET CHARACTER_SET_CLIENT=@OLD_CHARACTER_SET_CLIENT */;
/*!40101 SET CHARACTER_SET_RESULTS=@OLD_CHARACTER_SET_RESULTS */;
/*!40101 SET COLLATION_CONNECTION=@OLD_COLLATION_CONNECTION */;