PHP extracts web page titles and removes irrelevant seo keywords_PHP tutorial

场景描述:

  过往我们在抽取网页标题的时候，都会直接抽取 之间的内容. 但实际情况是这样,例如javaeye 的一篇文章 http://www.iteye.com/news/21643 ,  的内容为 "10年软件开发教会我最重要的10件事 - 非技术 - ITeye资讯", 但实际引用中我们期望的标题应该为 "10年软件开发教会我最重要的10件事". 所以标题后面堆砌了很多不相关的关键字(应该是为了  seo 吧). 所以我们希望过滤掉这些关键字. 有下面的方法可以参考:

1. 查找 h1 等标签.(分析sina news 一些网站之后, 觉得不可行,会有很多干扰)

2. 从全文去标题后，将 之间的内容切割(按 _ | -)为 a1,a2,a3,a4，然后从最长的词组a3开始从全文查找. 如果查找成功,那么开始向左边迭代查询 a2,a1,直到查询失败为止 。左侧失败后，再继续向右迭代，同理. (这里我采用的是这种方法)

Php代码 
/**
* @author pqcc
* @date: 2011-06-18
* Description: Given a web page content, extract the title of the web page. The extracted title does not Including seo keywords.
* e.g: The result of a news title extracted directly from /i";  
        preg_match($preg, $contents, $matches);  
        if(count($matches)<=1){  
            return "标题抽取失败";  
        }  
        $title = $matches[1];  
        return $this->trimTitle($title, $contents);  
    }/*}}}*/ 
 
    function trimMeta($contents){/*{{{*/ 
        // 首先去除 /i";  
        $contents   = preg_replace($preg, , $contents);  
        $preg       = "/]*>/i";  
        $contents   = preg_replace($preg, , $contents);  
        return $contents;  
    }/*}}}*/ 
 
 
    // 获取长度最长的 item 所处的index.  
    function getMaxIndex($titles){/*{{{*/ 
        $maxItemIndex   = 0;  
        $maxLength      = 0;  
        $loop           = 0;  
        foreach($titles as $item){  
            if(strlen($item)>$maxLength){  
                $maxLength      = strlen($item);  
                $maxItemIndex   = $loop;  
            }          
            $loop++;  
        }  
        return $maxItemIndex;  
    }/*}}}*/

function trim($title, $titles, $contents, $maxItemIndex){/*{{{*/
//@todo: Contents can be optimized here
// If Search successful. result = tempTitle.                                                                                             current index Iterate to the left (stop until the first one is reached or the match fails).
$leftIndex = $maxItemIndex-1;
while(true && $leftIndex>=0){
$leftIndex = $maxItemIndex-1; // tempTitle+left one . 
                    preg_match("/({$this->matches_preg}+{$tempTitle})/i", $title, $matches); 🎜>                                                                                                                                             .$matches[ 1];
// Continue to match with tempTitle.
preg_match("/$tempTitle/i", $contents, $matches);
// If the search fails....
if(count($matches)<1){                                                                                                                    $result = $tempTitle;                                            { // Under normal circumstances, this situation will not occur.
                                                                                                 


http://www.bkjia.com/PHPjc/478770.html

www.bkjia.com

true

http: //www.bkjia.com/PHPjc/478770.html

TechArticle

Scene description: In the past, when we extracted the title of the web page, we would directly extract the content between them. But the actual situation is In this way, for example, an article by javaeye http://www.iteye.com/news/2164...