php 获取页面中指定内容的实现类

<?php
/** grep class
* date: 2013-06-15
* author: fdipzone
* ver: 1.0
*
* func:
*
* set: 设置内容
* get: 返回指定的内容
* replace: 返回替换后的内容
* get_pattern 根据type返回pattern
*/

class grep{ // class start

private $_pattern = array(
'url' => '/<a.*?href="((http(s)?:\/\/).*?)".*?/si',
'email' => '/([\w\-\.]+@[\w\-\.]+(\.\w+))/',
'image' => '/<img.*?src=\"(http:\/\/.+\.(jpg|jpeg|gif|bmp|png))\">/i'
);

private $_content = ''; // 源内容


/* 設置搜尋的內容
* @param string $content
*/
public function set($content=''){
$this->_content = $content;
}


/* 获取指定内容
* @param string $type
* @param int $unique 0:all 1:unique
* @return array
*/
public function get($type='', $unique=0){

$type = strtolower($type);

if($this->_content=='' || !in_array($type, array_keys($this->_pattern))){
return array();
}

$pattern = $this->get_pattern($type); // 获取pattern

preg_match_all($pattern, $this->_content, $matches);

return isset($matches[1])? ( $unique==0? $matches[1] : array_unique($matches[1]) ) : array();

}


/* 获取替换后的内容
* @param string $type
* @param string $callback
* @return string
*/
public function replace($type='', $callback=''){

$type = strtolower($type);

if($this->_content=='' || !in_array($type, array_keys($this->_pattern)) || $callback==''){
return $this->_content;
}

$pattern = $this->get_pattern($type);

return preg_replace_callback($pattern, $callback, $this->_content);

}


/* 根据type获取pattern
* @param string $type
* @return string
*/
private function get_pattern($type){
return $this->_pattern[$type];
}
} // class end

?>

<?php
header('content-type:text/htm;charset=utf8');

require('grep.class.php');

$content = file_get_contents('http://www.test.com/');

$obj = new grep();
$obj->set($content);

$url = $obj->get('url', 0);
$email = $obj->get('email', 1);
$image = $obj->get('image', 1);

print_r($url);
print_r($email);
print_r($image);

$url_new = $obj->replace('url', 'replace_url');
echo $url_new;

function replace_url($matches){
return isset($matches[1])? '[url]'.$matches[1].'[/url]' : '';
}
?>