首页 > 代码库 > Python 网络爬虫(新闻采集脚本)

Python 网络爬虫(新闻采集脚本)

=====================爬虫原理=====================

通过Python访问新闻首页,获取首页所有新闻链接,并存放至URL集合中。

逐一取出集合中的URL,并访问链接获取源码,解析出新的URL链接添加到集合中。

为防止重复访问,设置一个历史访问,用于对新添加的URL进行过滤。

解析DOM树,获取文章相关信息,并将信息存储到Article对象中。

将Article对象中的数据通过pymysql保存到数据库中。

每完成一次数据的存储,计数器增加并打印文章标题,否则打印错误信息。

如果集合中的URL全部读取完或数据数量达到设定值,程序结束。

=====================存储结构=====================

CREATE TABLE `news` (  `id` int(6) unsigned NOT NULL AUTO_INCREMENT,  `url` varchar(255) NOT NULL,  `title` varchar(45) NOT NULL,  `author` varchar(12) NOT NULL,  `date` varchar(12) NOT NULL,  `about` varchar(255) NOT NULL,  `content` text NOT NULL,  PRIMARY KEY (`id`),  UNIQUE KEY `url_UNIQUE` (`url`)) ENGINE=InnoDB DEFAULT CHARSET=utf8;

=====================脚本代码=====================

‘‘‘百度百家新闻收集‘‘‘import re  # 网络连接模块import bs4  # DOM解析模块import pymysql  # 数据库连接模块import urllib.request  # 网络访问模块# 配置参数maxcount = 1000  # 数据数量home = http://baijia.baidu.com/  # 起始位置# 数据库连接参数db_config = {    host: localhost,    port: 3310,    username: woider,    password: 3243,    database: python,    charset: utf8}url_set = set()  # url集合url_old = set()  # 过期url# 获取首页链接html = urllib.request.urlopen(home).read().decode(utf8)soup = bs4.BeautifulSoup(html, html.parser)pattern = http://\w+\.baijia\.baidu\.com/article/\w+links = soup.find_all(a, href=http://www.mamicode.com/re.compile(pattern))for link in links:    url_set.add(link[href])# 文章类定义class Article(object):    def __init__(self):        self.url = None        self.title = None        self.author = None        self.date = None        self.about = None        self.content = None# 连接数据库connect = pymysql.Connect(    host=db_config[host],    port=int(db_config[port]),    user=db_config[username],    passwd=db_config[password],    db=db_config[database],    charset=db_config[charset])cursor = connect.cursor()# 处理URL信息count = 0while len(url_set) != 0:    try:        # 获取链接        url = url_set.pop()        url_old.add(url)        # 获取代码        html = urllib.request.urlopen(url).read().decode(utf8)
# DOM解析 soup = bs4.BeautifulSoup(html, html.parser) pattern = http://\w+\.baijia\.baidu\.com/article/\w+ # 链接匹配规则 links = soup.find_all(a, href=http://www.mamicode.com/re.compile(pattern)) # 获取URL for link in links: if link[href] not in url_old: url_set.add(link[href]) # 数据防重 sql = "SELECT id FROM news WHERE url = ‘%s‘ " data = (url,) cursor.execute(sql % data) if cursor.rowcount != 0: raise Exception(Data Repeat Exception: + url) # 获取信息 article = Article() article.url = url # URL信息 page = soup.find(div, {id: page}) article.title = page.find(h1).get_text() # 标题信息 info = page.find(div, {class: article-info}) article.author = info.find(a, {class: name}).get_text() # 作者信息 article.date = info.find(span, {class: time}).get_text() # 日期信息 article.about = page.find(blockquote).get_text() pnode = page.find(div, {class: article-detail}).find_all(p) article.content = ‘‘ for node in pnode: # 获取文章段落 article.content += node.get_text() + \n # 追加段落信息 # 存储数据 sql = "INSERT INTO news( url, title, author, date, about, content ) " sql = sql + " VALUES (‘%s‘, ‘%s‘, ‘%s‘, ‘%s‘, ‘%s‘, ‘%s‘) " data = (article.url, article.title, article.author, article.date, article.about, article.content) cursor.execute(sql % data) connect.commit() except Exception as e: print(e) continue else: print(article.title) count += 1 finally: # 判断数据是否收集完成 if count == maxcount: break# 关闭数据库连接cursor.close()connect.close()

=====================运行结果=====================

设置参数 maxcount = 10 , home = ‘http://baijia.baidu.com/‘

技术分享

查询数据 SELECT title, author FROM python.news;

技术分享

 

Python 网络爬虫(新闻采集脚本)