静态爬虫实战之百度百科
2018-06-01 本文已影响0人
zhile_doing
- 目标需求
爬取百度百科一百个词条，保存其相关的链接、标题、摘要信息 - 爬虫架构
想象一下爬虫运行的过程，从计算机启动，从某个特定url开始发出网络请求，得到响应后使用bs4解析，提取出需要保存的数据和url，对新提取出的url重复上述操作，当爬取指定数量url之后，将数据存盘。最明显的我们需要两个功能，获取网络数据，解析网络数据，将这两个功能封装进类中，将数据存盘也放入类中，还需要考虑的问题是url管理，决定进程何时停止。
从以上一般过程，总结爬虫的结构，包括url管理器、url下载器、网页解析器、数据存储器、以及管理主进程的manager。url管理器维护已经下载的url和将要下载的url的集合，保证数据不重复，url下载器根据给出的url下载网络数据，网页解析器解析给出的页面内容，提取必要信息，数据存储器将数据存盘，最后还要有管理器协调以上四者的工作 - 程序开发
- url管理器代码如下
#!/usr/bin/env python # coding:utf-8 class URLManager(object): def __init__(self): self.old_urls = set() self.new_urls = set() # 获取新的url def get_new_url(self): new_url = self.new_urls.pop() self.old_urls.add(new_url) return new_url # 添加待处理url def add_new_url(self, url): if url is None: return elif url not in self.old_urls and url not in self.new_urls: self.new_urls.add(url) else:处理 return # 添加待处理url列表 def add_new_urls(self, urls): if urls is None or len(urls) == 0: return else: for url in urls: self.add_new_url(url) # 返回待处理url数量 def new_url_size(self): return len(self.new_urls) #返回已处理url数量 def old_url_size(self): return len(self.old_urls) # 判断是否有新的url要处理 def has_new_urls(self): return self.new_url_size() != 0 - 下载器
#!/usr/bin/env python # coding:utf-8 import requests class HtmlDownloader(object): def download(self, url): if url is None: return else: user_agent = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/66.0.3359.170 Safari/537.36 OPR/53.0.2907.68' headers = {'User-Agent':user_agent} res = requests.get(url, headers=headers) if res.status_code == 200: res.encoding = 'utf-8' return res.text return None - url解析器
#!/usr/bin/env python # coding:utf-8 from bs4 import BeautifulSoup import re import urlparse class HtmlParser(object): def _get_new_urls(self, page_url, soup): if page_url is None or soup is None: return else: links = [] tag_a = soup.find_all('a', href=re.compile(r'/item/*')) # 经测试第一个和最后一个连接不可用 tag_a.remove(tag_a[0]) tag_a.remove(tag_a[len(tag_a) - 1]) for tag in tag_a: links.append(urlparse.urljoin(page_url, tag['href'])) return set(links) def _get_new_data(self, page_url, soup): if page_url is None or soup is None: return else: data = {} title = soup.find_all('dd', class_='lemmaWgt-lemmaTitle-title') data['title'] = title[0].h1.get_text() summary = soup.find_all('div', class_='lemma-summary') data['summary'] = summary[0].get_text() data['url'] = page_url return {item:data[item] for item in data.keys()} def parser(self, page_url, page_con): if page_url is None or page_con is None: return else: soup = BeautifulSoup(page_con, 'html.parser') data = self._get_new_data(page_url,soup) new_urls = self._get_new_urls(page_url, soup) return new_urls, data - 数据存储
#!/usr/bin/env python # encoding:utf-8 # 编码问题 import sys import codecs reload(sys) sys.setdefaultencoding('utf-8') class DataOutput(object): def __init__(self): self.datas = [] def store_data(self, data): if data is None: return else: self.datas.append(data) def out_data(self): if len(self.datas) == 0: print 'datas contains nothing...' return else: fp = codecs.open('baike.html', 'w',encoding='utf-8') fp.write('<html>') fp.write('<body>') fp.write('<table>') for data in self.datas: fp.write('<tr>') fp.write('<td>{}</td>'.format(data['url'])) fp.write('<td>{}</td>'.format(data['title'])) fp.write('<td>{}</td>'.format(data['summary'])) fp.write('</tr>') fp.write('</table>') fp.write('</body>') fp.write('</html>') fp.close() - 主程序管理
#!/usr/bin/env python # coding:utf-8 from HtmlDownloader import HtmlDownloader from HtmlParser import HtmlParser from URLManager import URLManager from DataOutput import DataOutput class SpiderMan(object): def __init__(self): self.manager = URLManager() self.parser = HtmlParser() self.out_fp = DataOutput() self.downloader = HtmlDownloader() def craw(self, base_url): if base_url is None: return else: self.manager.add_new_url(base_url) while(self.manager.has_new_urls() and self.manager.old_url_size() < 100): try: new_url = self.manager.get_new_url() page = self.downloader.download(new_url) urls, data = self.parser.parser(base_url, page) self.manager.add_new_urls(urls) self.out_fp.store_data(data) print 'already got {} links...'.format(self.manager.old_url_size()) except Exception,e: print 'crawing failed!' self.out_fp.out_data() if __name__ == '__main__': spider = SpiderMan() spider.craw('https://baike.baidu.com/item/Python/407313')
- url管理器代码如下
-
结果
