日期:2014-05-17  浏览次数:20687 次

share 一段小说的采集代码。
最近用火车头,ET采集小说,按他们的规则经常配不出来。碰到像小说520 里面的用iframe 的直接挂掉,只有自己写了一个,刚开始觉得就两正则解决[列表,内容]的事,写着写着便复杂起来了。
好好的改了几版,碰到最大的麻烦,,如何封装代码,就是采不同站时改动成本较小。这里小小的用了一个决策者模式。然后把该封装的功能一封。以及再次采集时,对已采集的章节的跳过机制,避免重采[毕竟一个小说站好几w篇文章,中断一次,接不上去,是很郁闷的事]

PHP code


class grep extends Controller {
    var $tableName = 'grep';
    var $pagesize =31;
    var $order_string = "grep_order desc,grep_id desc";
    var $filter_field = "grep_title";
    var $check_repeat_field = "grep_title";
     var $buttons = array(
     );
     var $description = "[爬取小说]";
    


function index()
{
    //get the story list    
    $story_model = "story_model";
    $this->load->model($story_model);
    $where = array("story_id < 445");
    $rows_story = $this->$story_model->get($where);
        
    foreach ($rows_story as $key=>$val_story):
    if($key < 237) continue;
          $url = "http://www.xiaoxiaoshuo.net/".$val_story->storycate_vtitle."/".$val_story->story_vtitle;
        $src_content = file_get_contents($url);
        $src_content = iconv("GBK","utf-8//IGNORE",$src_content);
                
            $src_content = str_replace("/style=\"border-width:0px\s*1px\s*1px\s*0px;border-color:#C8D8B8;border-style:solid;padding:3px;float:left;width:313px;\"/i","",$src_content);
            $src_content = str_replace("style=\"BORDER-RIGHT: #c8d8b8 1px solid; PADDING-RIGHT: 3px; BORDER-TOP: #c8d8b8 0px solid; PADDING-LEFT: 3px; FLOAT: left; PADDING-BOTTOM: 3px; BORDER-LEFT: #c8d8b8 0px solid; WIDTH: 313px; PADDING-TOP: 3px; BORDER-BOTTOM: #c8d8b8 1px solid\"","",$src_content);
            
              $src_content = preg_replace("/title=\"[^\"]*\"/iU","",$src_content);
              $src_content = preg_replace("/title=\"[^\"]*\"/iU","",$src_content);
            $src_content = preg_replace("/<LI[^>]*>/iU","",$src_content);
            $src_content = preg_replace("/<\/LI[^>]*>/iU","",$src_content);
            $src_content = preg_replace("/<a(?!href)[\d\D]*href/iU","<a href",$src_content);
            $src_content = preg_replace('/\s(?=\s)/', '', $src_content);
            $src_content = preg_replace('/[\n\r\t]/', ' ', $src_content);
            $src_content = preg_replace('/[\n\r\t]/', ' ', $src_content);
            $src_content = str_replace("http://www.xiaoxiaoshuo.net/yanqingxiaoshuo2/tijiaxinniang/","",$src_content);
            
            preg_match_all("/<td\s*bgcolor=\"#EDF5EA\"([\d\D]*)<\/ul>/iU",$src_content,$arr_dstorycate);
        
              
        
            $dstorycate_arr = $arr_dstorycate[1];
    
            foreach ($dstorycate_arr as $key_dstorycate => $val_dstory_cate)
        {
                  preg_match_all("/<font\s*color=\"#000000\">([^<]*)<\/font>/i",$val_dstory_cate,$dcate_title);
                    $datacate["dstorycate_pid"]   = $val_story->story_id;
                    $datacate["dstorycate_title"] = $dcate_title[1][0];
                    
                //获取类别对象,记将之前的类别标置为已下载
                 $dtitle =$datacate["dstorycate_title"];
                 $obj_storycate = $this->check_dcate($dtitle,$val_story );
                 
                 
                 
                 //pr($obj_storycate);
                 if($obj_storycate->dstorycate_ishot == 1)
                 {
                     $this->log( "<font color = gray>已此章节已抓取完 $val_story->story_title - $dtitle </font>,跳过");