python从爬虫基础到爬取网络小说实例

一.爬虫基础

1.1 requests类

　　1.1.1 request的7个方法

requests.request() 实例化一个对象,拥有以下方法

requests.get(url, *args)

requests.head() 头信息

requests.post()

requests.put()

requests.patch() 修改一部分内容

requests.delete()

url = "http://quanben5.com/n/doupocangqiong/6.html"

headers = {
    "User-Agent":"Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.81 Safari/537.36"
}

r = requests.get(url, headers=headers)

data = {
"pinyin": "doupocangqiong",
"content_id": "",
}

r = requests.post(url, data=data, headers=headers)

　　1.1.2*arg里面的参数

params 字典或者字节序列,作为参数增加到url中
data 字典字节序列文件对象, 放在url里面对应的地方
json 作为requests的内容
headers 字典 模拟服务头
cookies 字典 cookieJar
auth 元组
files 字典类型,传输文件
timeout 设定的超时时间
proxies 字典类型,设定访问代理服务器,可以增加登录认证 pxs={"http":"http://user:pass@10.10.1.1234"}
allow_redirects 默认True 是否允许重定向
stream 默认TRUE 获得数据立刻下载
verity 默认True 认证SSL证书开关
cert 本地SSL证书路径

1.2 BeautifulSoup类

soup = BeautifulSoup("","html.parser")

soup.prettify()

soup.find_all(name, attrs, recursive, string, **kwargs)

name: 标签名称
attrs: 属性
string: 检索字符串
soup.head

soup.head.contents 列表
soup.body.contents[1]
soup.body.children
soup.body.descendants
.parent

.parents
.next_sibling 下一个
.previous_sibling 上一个
.next_siblings 下一个所有 迭代
.previous_siblings 上一个所有

1.3 selenium

 from selenium import webdriver
 from selenium.webdriver.chrome.options import Options
 import time

 chrome_options = Options()
 chrome_options.add_argument('--headless')
 chrome_options.add_argument('--disable-gpu')

 d = webdriver.Chrome(chrome_options=chrome_options)  # 设置成不显示浏览器的模式
 d.get('http://quanben5.com/n/dazhuzai/23241.html')
 time.sleep(2)
 print(d.page_source)  # 显示出的是加载完之后的内容

查找单个和多个元素

d.find_element_by_id

d.find_elements_by_id

元素交互

from selenium import webdriver

d = webdriver.Chrome()
d.get('http://www.baidu.com')
inputText = d.find_element_by_id("kw")
inputText.send_keys("萝莉")
button = d.find_element_by_id("su")
button.click()

二.实战

首先要找到可以在线看小说的网页

这里我随便百度了一下,首先选择了一个全本5200小说网("https://www.qb5200.tw")

打开某小说章节目录表

https://www.qb5200.tw/xiaoshuo/0/357/

查看源代码

发现正文卷是在class为listmain的div下面的第二个dt标签里

之后的路径标签为a

     url_text = soup.find_all('div', 'listmain')[0]  # 找到第一个class=listmain的div标签
     main_text = url_text.find_all('dt')[1]  # 找到下面的第二个dt标签
     for tag in main_text.next_siblings:
         try:
             url = ''.join(['https://', host, tag.a.attrs['href']])
             print('parsering', url)
         except:
             continue

找到每个章节的url

在随便打开一章

查看源代码为:

 def getHTMLText(url, headers={}):
     """
     获取网站源码
     :param url:
     :return: class response
     """
     if headers != {}:
         headers = {"User-Agent":"Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.81 Safari/537.36"}

     # proxies = get_random_ip()
     proxies = {}
     try:
         # print("start", url)
         r = requests.get(url, proxies=proxies, headers=headers)
         r.raise_for_status()
         # r.encoding = r.apparent_encoding
         # print('end', url)
         return r
     except:
         return r.status_code

 def parseByQB5200(soup, host, f):
     """
     在全本小说网的源码下爬取小说
     :param soup, host:
     :param f:
     :return:
     """
     url_text = soup.find_all('div', 'listmain')[0]
     main_text = url_text.find_all('dt')[1]
     x = 0
     for tag in main_text.next_siblings:
         time.sleep(1)
         try:
             url = ''.join(['https://', host, tag.a.attrs['href']])
             print('parsering', url)
             soup = BeautifulSoup(getHTMLText(url).text, "html.parser")
             passage = soup.find_all("div", "content")[0]
             title = passage.h1.string
             f.writelines(''.join([title, '\n']))
             passage = soup.find_all("div", "showtxt")[0]
             for i in passage.descendants:
                 if i.name != "br":
                     st = i.string
                     if st.strip().startswith('http') or st.strip().startswith('请记住本书'):
                         continue
                     f.writelines(''.join(['  ', st.strip(), '\n']))
             x += 1
             print('%d.%s 下载完成' % (x, title))
         except:
             continue

 def getNovelUrls(url):
     """
     通过小说的目录网址判断小说所在的网站
     并调用属于该网站的爬虫语句
     :param url:
     :return:
     """

     response = getHTMLText(url)
     host = url.split('//')[1].split('/')[0]
     host_list = {
         "www.qb5200.tw": parseByQB5200,
         # "www.qu.la": parseByQuLa,
         "quanben5.com": parseByQB5
     }
     print(response)
     soup = BeautifulSoup(response.text, 'html.parser')
     with open('1.txt', 'w', encoding='utf8') as f:
         host_list[host](soup, host, f)

 if __name__ == '__main__':
     getNovelUrls("https://www.qb5200.tw/xiaoshuo/0/357/")

在全本5200爬取小说txt

问题在于

全本小说网("www.qb5200.tw")在这样的暴力获取下只允许爬3次,之后就403错误

本来以为是同一IP限制访问次数, 使用了IP代理之后发现问题依旧

猜测应该是跟请求头有关

因此加上了访问该网站时浏览器的所有请求头,并且把User-Agent设置为随机

USER_AGENT_LIST = [
        'MSIE (MSIE 6.0; X11; Linux; i686) Opera 7.23',
        'Opera/9.20 (Macintosh; Intel Mac OS X; U; en)',
        'Opera/9.0 (Macintosh; PPC Mac OS X; U; en)',
        'iTunes/9.0.3 (Macintosh; U; Intel Mac OS X 10_6_2; en-ca)',
        'Mozilla/4.76 [en_jp] (X11; U; SunOS 5.8 sun4u)',
        'iTunes/4.2 (Macintosh; U; PPC Mac OS X 10.2)',
        'Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:5.0) Gecko/20100101 Firefox/5.0',
        'Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:9.0) Gecko/20100101 Firefox/9.0',
        'Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:16.0) Gecko/20120813 Firefox/16.0',
        'Mozilla/4.77 [en] (X11; I; IRIX;64 6.5 IP30)',
        'Mozilla/4.8 [en] (X11; U; SunOS; 5.7 sun4u)'
    ]
    headers = {"User-Agent": random.choice(USER_AGENT_LIST),
"Host": "www.qb5200.tw",
"Connection": "keep-alive",
"Pragma": "no-cache",
"Cache-Control": "no-cache",
"Upgrade-Insecure-Requests": "",
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8",
"Referer": "https://www.qb5200.tw/xiaoshuo/0/355/",
"Accept-Encoding": "gzip, deflate, br",
"Accept-Language": "zh-CN,zh;q=0.9",
"Cookie": "UM_distinctid=16736c47b1d8b-04bd85d2b5e8ab-50422618-144000-16736c47b1e1f2; bcolor=; font=; size=; fontcolor=; width=; CNZZDATA1260750615=1497982366-1542811927-%7C1542882690; fikker-7ZUK-qXIL=qDaKEBihatSNEFgtMlGVIKubCHYLi89J; fikker-7ZUK-qXIL=qDaKEBihatSNEFgtMlGVIKubCHYLi89J; fikker-Sbwr-GN9F=GbMX3HOhwvoDMxopLMGt3VWliXQIK0SP; fikker-Sbwr-GN9F=GbMX3HOhwvoDMxopLMGt3VWliXQIK0SP; fikker-yJ3O-W61D=UfinETMnCR38ADWZEl1KNHQRU2m81Fwb; fikker-yJ3O-W61D=UfinETMnCR38ADWZEl1KNHQRU2m81Fwb; fikker-rWK3-6KHs=T9T5b7lYVJReTQviPm2IdLPyHu83RwFM; fikker-rWK3-6KHs=T9T5b7lYVJReTQviPm2IdLPyHu83RwFM"

               }

随机请求头

解决问题.

三.使用ajax动态加载的实例:

全本5小说网("quanben5.com")

同样的方法搜索源代码

然而发现了问题

给出的html页面只有一半的源码

因此按F12打开检查

发现所有的文本存在这个xhr里

点击查看请求头信息

发现是post请求

请求的url是/index.php?c=book&a=ajax_content

请求的数据在最下面的form表单里

打开网页源文件和js文件,搜索这些表单信息

分别在ajax.js里和源文件里找到了这些

源文件里面的可以直接生成data数据表单

在ajax.js里可以知道rndval字段是当前时间,精确到毫秒

四.优化

采用了gvent进行异步IO处理,每一张网页保存在temp里面,最后将文件合成一个txt

加入了搜索功能,目前仅支持一个小说网站

代码如下:

 #!/usr/bin/env python
 # encoding: utf-8

 """
 @version:
 @author: Wish Chen
 @contact: 986859110@qq.com
 @file: get_novels.py
 @time: 2018/11/21 19:43

 """
 import gevent
 from gevent import monkey
 monkey.patch_all()
 import requests, time, random, re, os
 from bs4 import BeautifulSoup

 dir = os.path.dirname(os.path.abspath(__file__))

 def getHTMLText(url, data=[], method='get'):
     """
     获取网站的源代码
     请求默认为get
     :param url:
     :param data:
     :param method:
     :return:
     """
     headers = {"User-Agent":"Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.81 Safari/537.36"}
     proxies = {}
     try:
         # print("start", url)
         r = requests.request(method, url, proxies=proxies, headers=headers, data=data)
         r.raise_for_status()
         # r.encoding = r.apparent_encoding
         # print('end', url)
         return r
     except:
         return r.status_code

 def fetch_async(x, url):
     """
     异步IO所需要执行的操作
     获取源文件
     模拟向ajax请求获取完整文字
     每一章输入到temp文件夹下
     :param x:
     :param url:
     :return:
     """
     url_main = "http://quanben5.com/index.php?c=book&a=ajax_content"
     r = getHTMLText(url)  # 获取每一章的源文件
     title = re.search(r'<h1 class="title1">(.*)</h1>', r.text).group(1)
     result = re.search(r'<script type="text/javascript">ajax_post\((.*)\)</script>', r.text).group(1)
     num_list = result.split("','")
     num_list[9] = num_list[9][:-1]
     content = {}  # 开始模拟post请求发送的表单
     for i in range(1, 5):
         content[num_list[i * 2]] = num_list[i * 2 + 1]
     content['_type'] = "ajax"
     content['rndval'] = int(time.time() * 1000)
     r = getHTMLText(url_main, data=content, method='post')  # 模拟post请求
     soup = BeautifulSoup(r.text, "lxml")
     with open(os.path.join(dir, 'temp', "%s.txt" % x), "w", encoding='utf8') as f:
         f.writelines(''.join([str(x), '.', title, '\n\n']))
         for tag in soup.body.children:
             if tag.name == 'p':
                 f.writelines(''.join(['    ', tag.string.strip(), '\n\n']))
     print('%d.%s 下载完成' % (x, title))

 def get_together(name, author, x):
     """
     将temp目录下的各网页下载下来的txt
     合并在一起
     并删除temp文件
     :param name:
     :param author:
     :return:
     """
     with open(os.path.join(dir, "%s.txt" % name), "w", encoding='utf8') as f:
         f.writelines(''.join([name, '\n\n作者:', author, '\n\n']))

         for i in range(x):
             try:
                 f.write(open(os.path.join(dir, 'temp', "%s.txt" % (i+1)), "r", encoding='utf8').read())
                 f.write('\n\n')
                 # os.remove(os.path.join(dir, 'temp', "%s.txt" % (i+1)))
             except:
                 continue

 def parseByQB5(response, host):
     """
     在全本5小说网的源码下爬取小说
     获得书名和作者
     采用gevent异步IO优化
     :param response:
     :param host:
     :return:
     """
     soup = BeautifulSoup(response.text, 'html.parser')
     url_text = soup.find_all('div', 'box')[2]
     main_text = url_text.find_all('h2')[0].next_sibling
     url_list = []
     for tag in main_text.descendants:
         if tag.name == 'li':
             url = ''.join(['http://', host, tag.a.attrs['href']])
             url_list.append(url)
     from gevent.pool import Pool
     pool = Pool(100)

     gevent.joinall([pool.spawn(fetch_async, i+1, url=url_list[i]) for i in range(len(url_list))])

     name = re.search(r"<h3><span>(.*)</span></h3>", response.text).group(1)
     author = re.search(r'<span class="author">(.*)</span></p>', response.text).group(1)
     print("%d文档已下载,正在合并..." % len(url_list))
     get_together(name, author, len(url_list))

 def getNovelUrls(url):
     """
     通过小说的目录网址判断小说所在的网站
     并调用属于该网站的爬虫语句
     :param url:
     :return:
     """

     response = getHTMLText(url)
     host = url.split('//')[1].split('/')[0]
     host_list = {
         "quanben5.com": parseByQB5
     }
     host_list[host](response, host)

 def get_url():
     input_name = input('>>')
     r = getHTMLText("http://quanben5.com//index.php?c=book&a=search&keywords=%s" % input_name)
     soup = BeautifulSoup(r.text, "html.parser")
     main_book = soup.find_all("div", "pic_txt_list")
     for i in range(len(main_book)):
         tag = main_book[i].h3
         print("%s.%s %s" %(i, tag.span.text, tag.next_sibling.next_sibling.text))
     choice = int(input(">>"))
     if choice in range(len(main_book)):
         return ''.join(["http://quanben5.com", main_book[choice].h3.a["href"], "xiaoshuo.html"])

 if __name__ == '__main__':
     # url_list = [
     #     "https://www.qu.la/book/365/",
     #     "https://www.qb5200.tw/xiaoshuo/0/357/",
     #     "http://quanben5.com/n/doupocangqiong/xiaoshuo.html",
     #     "http://quanben5.com/n/dazhuzai/xiaoshuo.html",
     #     "http://quanben5.com/n/douluodalu/xiaoshuo.html",
     #     "http://quanben5.com/n/renxingderuodian/xiaoshuo.html"
     # ]
     # if not os.path.exists('temp'):
     #     os.mkdir('temp')
     # getNovelUrls(url_list[5])
     while True:
         url = get_url()
         time_start = time.time()
         getNovelUrls(url)
         print("成功爬取! 用时:%ds" % (int((time.time()-time_start)*100)/100))

异步IO+搜索

封装性还不够好

最好一个网站用一个类来封装

采用scrapy框架

正在设计中...

五.未解决问题:

代理IP问题:

目前只有免费的代理IP网

生成随机IP并使用代理IP访问

多网站定制

要观察各个网站的目录源代码结构以及文章源代码结构

每一个网站都可以用一个parse函数来解析其内容