Python开发轻量级爬虫

这两天自学了python写爬虫,总结一下:

开发目的:抓取百度百科python词条页面的1000个网页

设计思路:

1,了解简单的爬虫架构:

2,动态的执行流程:

 

3,各部分的实现:

URL管理器:python内存

网页下载器:python3自带的urllib模块

网页解析器:使用第三方插BeautifulSoup

4,开发思路:

入口页:http://baike.baidu.com/view/21087.htm

URL格式:

                 词条页面URL:/view/125370.htm

数据格式:

                 标题:<dd class=”lemma Wgt-lemmaTitle-title”><h1>***</h1></dd>

                                 

                 简介:<div class=”lemma-summary”>***</div>

页面编码:UTF-8

5,开发过程:

①首先建立一个Pydev工程,新建一个web_spider库,新建一个主函数spider_main,在主函数中调用:URL管理器:url_manager,网页下载器:html_downloader,网页解析器:html_paeser.,网页输出:html_outputer来完成相应的功能

②URL管理器的建立:

新建已读取和未读取的URL集合,用来存放相应的URL地址,定义三个函数:get_new_url(self), add_new_urls(self,urls),add_new_url(self,url)来完成相应的功能

③ 网页下载器的建立:

网页下载的方法:

                 import urllib.request

       response=urllib.request.urlopen(url)

       if response.getcode()!=200:

            return None

       return response.read().decode('UTF-8')

④网页解析器的建立:

导入BeautifulSoup模块,

创建BeautifulSoup对象,调用find方法搜索并访问节点

⑤网页输出的实现:

将抓取到的数据写入到一个HTML网页中

6,遇到的问题及解决方案:

准备工作:

一、1,BeautifulSoup下载地址:https://www.crummy.com/software/BeautifulSoup/

        2.下载完成之后需要解压缩,假设放到D:/python下。

        3.运行cmd,切换到D:/python/beautifulsoup4-4.1.3/目录下(根据自己解压缩后的目录      和下载的版本号修改),

        4.运行命令:

                 setup.py build

                 setup.py install

        5.在IDE下from bs4 import BeautifulSoup,没有报错说明安装成功。

二、在ecplise中安装PyDev插件后却无法建立项目,后来发现是没有配置python3的解释器,具体步骤为:在 Eclipse 菜单栏中,选择 Window > Preferences > Pydev > Interpreter - (Python/Jython),在这里配置 Python/Jython 解释器

实例代码如下

1spider_main.py

# coding:utf8

from baike_spider import url_manager, html_downloader, html_outputer,\

    html_parser

class SpiderMain(object):

    def __init__(self):

        self.urls=url_manager.UrlManager()

        self.downloader=html_downloader.HtmlDownloader()

        self.parser=html_parser.HtmlParser()

        self.outputer=html_outputer.HtmlOutputer()

 

    def craw(self, root_url):

        count=1

        self.urls.add_new_url(root_url)

        while self.urls.has_new_url(): 

            try:

                new_url=self.urls.get_new_url()

                print("craw %d :%s"%(count,new_url))

                html_cont=self.downloader.download(new_url)

                new_urls,new_data=self.parser.parse(new_url,html_cont)

                self.urls.add_new_urls(new_urls)

                self.outputer.collect_data(new_data)

                if count==1000:

                    break

                count=count+1

            except:

                print("craw failed")          

        self.outputer.output_html()

if __name__=="__main__":

    root_url="http://baike.baidu.com/view/21087.htm"

    obj_spider=SpiderMain()

obj_spider.craw(root_url)

 

 (2)url_manager.py

# coding:utf8

class UrlManager(object):

    def __init__(self):

        self.new_urls=set()

        self.old_urls=set()  

    def has_new_url(self):

        return len(self.new_urls)!=0

   

      

    def get_new_url(self):

        new_url=self.new_urls.pop()

        self.old_urls.add(new_url)

        return new_url

 

   

    def add_new_urls(self,urls):

        if urls is None or len(urls)==0:

            return

        for url in urls:

            self.add_new_url(url)

 

   

    def add_new_url(self,url):

        if url is None:

            return

        if url not in self.new_urls and url not in self.old_urls:

            self.new_urls.add(url)

 

(3)html_downloader.py

# coding:utf8

import urllib.request

class HtmlDownloader(object):

    def download(self,url):

        if url is None:

            return None

        response=urllib.request.urlopen(url)

        if response.getcode()!=200:

            return None

        return response.read().decode('UTF-8')

       

       

 (4)html_paeser.py

# coding:utf8

from bs4 import BeautifulSoup

import re

import urllib

class HtmlParser(object):

   

   

    def _get_new_urls(self, page_url, soup):

        new_urls=set()

        links=soup.find_all('a',href=re.compile(r"/view/\d+\.htm"))

        for link in links:

            new_url=link['href']

            new_full_url=urllib.parse.urljoin(page_url,new_url)

            new_urls.add(new_full_url)

        return new_urls

   

   

    def _get_new_data(self, page_url, soup):

        #<dd class="lemmaWgt-lemmaTitle-title"><h1>Python</h1>

        res_data={}

        res_data['url']=page_url

        title_node=soup.find('dd',class_="lemmaWgt-lemmaTitle-title").find("h1")

        res_data['title']=title_node.get_text()

        #<div class="lemma-summary" label-module="lemmaSummary">

        summary_node=soup.find('div',class_="lemma-summary")

        res_data['summary']=summary_node.get_text()

        return res_data

   

    def parse(self,page_url,html_cont):

        if page_url is None or html_cont is None:

            return None

        soup=BeautifulSoup(html_cont,'html.parser',from_encoding='utf8')

        new_urls=self._get_new_urls(page_url,soup)

        new_data=self._get_new_data(page_url,soup)

        return new_urls,new_data

 

(5)html_outputer.py

# coding:utf8

class HtmlOutputer(object):

    def __init__(self):

        self.datas=[]  

    def collect_data(self,data):

        if data is None:

            return

        self.datas.append(data)

    def output_html(self):

        fout=open("output.html",'w',encoding='utf8')   

        fout.write("<html>")

        fout.write("<body>")

        fout.write("<table>")

        for data  in self.datas:

            fout.write("<tr>")

            fout.write("<td>%s</td>"%str(data['url']))

            fout.write("<td>%s</td>"%str(data['title']))           

            fout.write("<td>%s</td>"%str(data['summary']))

            fout.write("</tr>")

        fout.write("</table>")

        fout.write("</body>")

        fout.write("</html>")   

        fout.close()

 

运行结果如图:

 

 

 

 

posted @ 2016-05-22 16:40  左岸繁华右岸殇  阅读(2475)  评论(0编辑  收藏  举报