贴一下抓到的评论，最后几条的

本渣是在校学生(非计算机相关)，今年上半年接触Python，然后暑假开始学着写爬虫。

新手，本文如有错误，谢谢指正。

这个爬虫开始思路有点乱，现在比较清晰了

思路：requests库获取html

在html中找到翻页方法和评论总条数(评论总条数方便我查看爬取进度)

然后再写一个函数获取到了评论信息

这里多说一句，爬取豆瓣实现翻页得在html中找到下一页url的一部分

最后成功了，但是打开txt文件，发现有10%左右的评论都是乱码，其实我设置了编码处理的，最后文件也是用的utf-8写入：

#网页编码

r.encoding = r.apparent_encoding

#写入文件

with open(fpath, 'a', encoding='utf-8') as f:

在网上找了很多文档和分享，明白了requests库设计初衷就是希望用户自己来修改编码，然后我直接把编码改了

r.encoding = 'utf-8'

最后出现的问题是，翻页的函数总是报错，找不到对应的标签。最后我想了一下，如果是获取评论，就按照utf-8的编码来；如果是翻页，还是用r.encoding = r.apparent_encoding

try:

if flag == 1: #如果传入的是获取评论信息的url

r.encoding = 'utf-8' #尝试改变编码

elif flag == 2:

r.encoding = r.apparent_encoding

return r.text

except:

return ""

最后中文乱码的问题解决了。

然而，又碰到了验证码了。。。终于，数次之后，我的ip好像被封了

我找了个ip代理网站，弄代理ip，这里参考了Python爬虫技巧之设置代理IP

但是疏忽了一点，不能一直访问这个网站获取ip，然后访问次数过多，导致ip被代理ip网站给封了，4000+评论在2000+的时候报错了。

然后我在那个代理网站里随便找个一个ip，再去获取ip，生成列表

url = 'http://www.xicidaili.com/nn/'

ipAgent = {

'http':'http://221.214.181.98:53281'

}

headers = {

'User-Agent': "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526.106 Safari/537.36"

}

web_data = requests.get(url, headers=headers, proxies = ipAgent)

最后终于可以愉快地玩耍了。

爬取评论的全部代码如下(两个文件)，header是我用一个浏览器插件生成的，所以很长，不过也很方便，是参考了知乎一位大佬的Python爬虫传送post请求要携带哪些参数?

#第一个文件

#CrawDoubanComments.py

#下面的from cons import headerChange是我写的一个随机取user-agent的代码文件

import requests

import os

from bs4 import BeautifulSoup

import bs4

from cons import headerChange

import random

def get_ip_list():

url = 'http://www.xicidaili.com/nn/'

ipAgent = {

'http':'http://221.214.181.98:53281'

}

headers = {

'User-Agent': "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526.106 Safari/537.36"

}

web_data = requests.get(url, headers=headers, proxies = ipAgent)

soup = BeautifulSoup(web_data.text, 'lxml')

ips = soup.find_all('tr')

ip_list = []

for i in range(1, len(ips)):

ip_info = ips[i]

tds = ip_info.find_all('td')

ip_list.append(tds[1].text + ':' + tds[2].text)

return ip_list

def get_random_ip(ip_list):

proxy_list = []

for ip in ip_list:

proxy_list.append('http://' + ip)

proxy_ip = random.choice(proxy_list)

proxies = {'http': proxy_ip}

return proxies

ipList = get_ip_list()

print(ipList)

def getHTMLTxt(url, flag):

querystring = {"status":"P"}

headers = {

'accept': "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8",

'accept-encoding': "gzip, deflate, br",

'accept-language': "zh-CN,zh;q=0.8",

'cache-control': "no-cache",

'connection': "keep-alive",

'cookie': "ll=\"118161\"; bid=hvtlFYn8N2g; _pk_ref.100001.4cf6=%5B%22%22%2C%22%22%2C1502935063%2C%22https%3A%2F%2Fwww.google.com%2F%22%5D; __yadk_uid=4glVfZfkxqS5u5wHymY2aa0dEnqhZ1GL; ps=y; dbcl2=\"142710097:FLM/J8N4I1w\"; ck=12Nj; _vwo_uuid_v2=3474CBA32AC36EAA99D3D339FC358DBA|16192af0fbe498c7433958350c7be868; __utma=30149280.810866807.1502935064.1502935064.1502935064.1; __utmb=30149280.3.9.1502935124997; __utmc=30149280; __utmz=30149280.1502935064.1.1.utmcsr=google|utmccn=(organic)|utmcmd=organic|utmctr=(not%20provided); __utmv=30149280.14271; __utma=223695111.302797075.1502935064.1502935064.1502935064.1; __utmb=223695111.0.10.1502935064; __utmc=223695111; __utmz=223695111.1502935064.1.1.utmcsr=google|utmccn=(organic)|utmcmd=organic|utmctr=(not%20provided); _pk_id.100001.4cf6=12c01e44b4266fee.1502935063.1.1502936044.1502935063.; _pk_ses.100001.4cf6=*; push_noty_num=0; push_doumail_num=0",

'host': "movie.douban.com",

'referer': "https://movie.douban.com/subject/26392671/",

'upgrade-insecure-requests': "1",

'user-agent': headerChange(),

'postman-token': "e1715f99-6029-83ce-5574-3f4e6148b96c"

}

proxies = get_random_ip(ipList)

r = requests.request("GET", url, headers=headers, params=querystring, proxies = proxies)

try:

if flag == 1: #如果传入的是获取评论信息的url

r.encoding = 'utf-8' #尝试改变编码

elif flag == 2:

r.encoding = r.apparent_encoding

return r.text

except:

return ""

def getComments(url, cmt): #获取一整页的评论

count = 0

html = getHTMLTxt(url, 1)

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

for tagDiv in soup.find_all('div', attrs={'class':'comment'}):

if isinstance(tagDiv, bs4.element.Tag):

infoComt = tagDiv.find('p')

cmt.append(infoComt.string)

count = count + 1

return count

def nextPageURL(urlPre): #根据前一页的url返回的html中找到后一页的url

html = getHTMLTxt(urlPre, 2)

try:

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

divTag = soup.find('div', attrs={'id':'paginator', 'class':'center'}) #找到存有下一页url的div标签

aTag = divTag('a') #找到存有信息的a标签

index = len(aTag) #下一页的url信息在最后一个标签内

return aTag[index - 1].get('href')

except:

return -1

def countOfComments(urlStart): #根据首页返回的html拿到评论的总条数

try:

html = getHTMLTxt(urlStart, 1)

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

#下面代码在各标签中查找到评论的总条数

liTag = soup.find('li', attrs={'class':'is-active'})

spanTag = liTag('span')

strPage = spanTag[0].string

return eval(strPage[-5:-1]) #这里的切片根据评论总条数的位数来定

except:

return 0

def main():

urlStart = "https://movie.douban.com/subject/10477598/comments?" #此处comments后的都不要有，包括'?'

depth = countOfComments(urlStart)

#depth = 1000 #手动设置评论条数，控制时间

fpath = "D:\PY\CrawDoubanComments\comments.txt"

cmt = []

count = 0

if os.path.exists('D:\PY\CrawDoubanComments\comments.txt'):

os.remove('D:\PY\CrawDoubanComments\comments.txt') #因为需要测试，懒得手动删除，所以程序开始时对文件进行删除操作

urlNext = ""

print("爬取评论")

while(count < depth and len(cmt) < depth):

url = urlStart + urlNext

count = count + getComments(url, cmt)

urlNext = nextPageURL(url)

if urlNext == -1:

print("意外终止！")

break

else:

print('\r正在爬取评论，进度：{:.2f}%'.format((count * 100) / depth), end = "")

print("\n爬取成功，进行储存。。。")

for i in range(len(cmt)):

try:

with open(fpath, 'a', encoding='utf-8') as f:

f.write(cmt[i] + '\n')

print('\r正在储存评论，进度：第{}条，共{}条'.format(i + 1, len(cmt)), end = "")

except:

continue

print("\n评论储存完毕！")

main()

#第二个文件，获取user-agent

#cons.py

import random

#在网上找了多个user-agent，然后每次访问时利用随机库在其中随机选择一个

#

headerstr = '''Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.0; Trident/4.0)Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0)Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:2.0.1) Gecko/20100101 Firefox/4.0.1Mozilla/5.0 (Windows NT 6.1; rv:2.0.1) Gecko/20100101 Firefox/4.0.1Opera/9.80 (Macintosh; Intel Mac OS X 10.6.8; U; en) Presto/2.8.131 Version/11.11Opera/9.80 (Windows NT 6.1; U; en) Presto/2.8.131 Version/11.11Mozilla/5.0 (Macintosh; Intel Mac OS X 10_7_0) AppleWebKit/535.11 (KHTML, like Gecko) Chrome/17.0.963.56 Safari/535.11Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; Maxthon 2.0)Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.101 Safari/537.36'''

def headerChange():

headerList = headerstr.split('\n')

length = len(headerList)

return headerList[random.randint(0,length - 1)]

最后还有生成词云的文件

import matplotlib.pyplot as plt

from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator

from scipy.misc import imread

import jieba

from os import path

text = open("comments.txt","rb").read()

#结巴分词

wordlist = jieba.cut(text,cut_all=True)

wl = " ".join(wordlist)

coloring = imread(path.join("罗小黑.png"))

#设置词云

wc = WordCloud(background_color = "white", #设置背景颜色

mask = coloring, #设置背景图片

max_words = 10000, #设置最大显示的字数

#stopwords = "", #设置停用词

font_path = "C:\\Windows\Fonts\\simhei.ttf",

#设置中文字体，使得词云可以显示(词云默认字体是“DroidSansMono.ttf字体库”，不支持中文)

max_font_size = 40, #设置字体最大值

random_state = 42, #设置有多少种随机生成状态，即有多少种配色方案

)

myword = wc.generate(wl)#生成词云

image_colors = ImageColorGenerator(coloring) #得到图片

#展示词云图

#plt.imshow(myword) #随机的颜色，没有与背景一样

# 绘制词云

# recolor wordcloud and show

# we could also give color_func=image_colors directly in the constructor

plt.imshow(wc.recolor(color_func=image_colors))

plt.axis("off")

# 绘制背景图片为颜色的图片

plt.figure()

plt.imshow(coloring, cmap=plt.cm.gray)

plt.axis("off")

plt.show()