• Perl6 必应抓取(1):测试版代码


    一个相当丑漏的代码, 以后有时间再优化了。

    默认所有查找都是15页, 如果结果没有15页这么多估计会有重复。速度还是很快的。

    sub MAIN() {
      my $fp = open 'bin_result.txt', :w;
      my $number = 15;
      print 'String:';
      my $string = get;
      $string  = do given $string {S:g/s/+/};
      use HTTP::UserAgent;
      my $url = 'http://cn.bing.com/search?q=';
      my $ua = HTTP::UserAgent.new;
      my $check = rx/'<'cite'>'(.*?)'</cite>'/;#要查的内容
      my @number = '';
      @number.append(0..$number);
      my $page='';
      my $html;
      my $target = $url~$string~'&first=20&FROM=FERE'~$page;
      $html = $ua.get($target).content;
      loop {
        say '===============> '~$target;
        $html ~~ $check;
        $html = $/.postmatch;
        #$0 = do given ~$0 {S:g/'<strong>'//;}
        if not $0 {
          #当是null时, 说明这一页已全部提取, 构造下一页
          $page = Int($page);
          my $page_next = $string~'&first='~$page~'0&FROM=FERE'~$page;
          $target = $url~$page_next;
          $html = $ua.get($target).content;
          $page++;
        #/search?q=123&first=10&FORM=PERE
        #/search?q=123&first=20&FORM=PERE1
        #/search?q=123&first=30&FORM=PERE2
        #/search?q=123&first=30&FORM=PERE2
        #last;
         $html ~~ $check;
            $html = $/.postmatch;
            if ($page > $number) {last;}
        }
        my $ok_check = $0.Str;
        my $result = $ok_check;
        $result = do given $result {S:g/'<strong>'//;}
        $result = do given $result {S:g/'</strong>'//;}
        say $result;
        $fp.say($result);
    
      }
      #$fp.print($html);
    }

    下次代码优化:

    总结一下必应的规律, 如下:

    http://cn.bing.com/search?q=123456789&first=1&FORM=PERE
    http://cn.bing.com/search?q=123456789&first=11&FORM=PERE
    http://cn.bing.com/search?q=123456789&first=21&FORM=PERE1
    http://cn.bing.com/search?q=123456789&first=31&FORM=PERE2
    http://cn.bing.com/search?q=123456789&first=41&FORM=PERE3
    http://cn.bing.com/search?q=123456789&first=51&FORM=PERE4
    http://cn.bing.com/search?q=123456789&first=61&FORM=PERE4
    http://cn.bing.com/search?q=123456789&first=71&FORM=PERE4
    http://cn.bing.com/search?q=123456789&first=81&FORM=PERE4
    http://cn.bing.com/search?q=123456789&first=91&FORM=PERE4

    在页面上测试, 参数只虽两个即可:

    q=查询字符串&first=起始帐号

  • 相关阅读:
    初学三种神经网络(前馈,竞争,递归联想存储网络)
    链表操作,空间复杂度要求为O(1)
    单链表的回文判断(O(n)时间复杂度和O(1)的空间复杂度)
    Floyd算法小结
    排序算法Java实现(基数排序)
    排序算法Java实现(归并排序)
    排序算法Java实现(堆排序)
    排序算法Java实现(选择排序)
    排序算法Java实现(快速排序)
    排序算法Java实现(冒泡排序)
  • 原文地址:https://www.cnblogs.com/perl6/p/7425641.html
Copyright © 2020-2023  润新知