爬取《战狼2》电影短评论，生成图云

最新推荐文章于 2021-07-21 20:48:27 发布

DaXian-Will

最新推荐文章于 2021-07-21 20:48:27 发布

阅读量3k

点赞数 1

分类专栏：【Python开发】

【Python开发】专栏收录该内容

0 篇文章 0 订阅

订阅专栏

在网上看到用Python做的小demo，copy下来学习学习！

模拟登陆豆瓣

第一次登陆需要验证码，之后的登陆可以隐去 “login(”username’,’password’）”,因为使用session保存了必要的登陆信息，代码如下:

import requests
try:
    import cookielib
except:
    import http.cookiejar as cookielib
import re
import time
import os.path
import json
from bs4 import BeautifulSoup
try:
    from PIL import Image
except:
    pass

from mywordCloud import save_jieba_result
from mywordCloud import draw_wordcloud
import threading
import codecs

构造 Request headers

agent = 'Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36'
headers = {
    "Host": "www.douban.com",
    "Referer": "https://www.douban.com/",
    'User-Agent': agent,
}

使用cookie登录信息

session=requests.session()
session.cookies=cookielib.LWPCookieJar(filename='cookies')

try:
    session.cookies.load(ignore_discard=True)
    print('成功加载cookie')
except:
    print("cookie 未能加载")

获取验证码

def get_captcha(url):
    #获取验证码
    print('获取验证码',url)
    captcha_url = url
    r = session.get(captcha_url, headers=headers)
    print('test')
    with open('captcha.jpg', 'wb') as f:
        f.write(r.content)
        f.close()
    # 用pillow 的 Image 显示验证码
    # 如果没有安装 pillow 到源代码所在的目录去找到验证码然后手动输入
    try:
        im = Image.open('captcha.jpg')
        im.show()
        im.close()
    except:
        print(u'请到 %s 目录找到captcha.jpg 手动输入' % os.path.abspath('captcha.jpg'))
    captcha = input("please input the captcha\n>")
    return captcha

def isLogin():
    #登录个人主页，查看是否登录成功
    url='https://www.douban.com/people/151607908/'
    login_code=session.get(url,headers=headers,allow_redirects=False).status_code
    if login_code==200:
        return True
    else:
        return False

def login(acount,secret):
    douban="https://www.douban.com/"
    htmlcha=session.get(douban,headers=headers).text
    patterncha=r'id="captcha_image" src="(.*?)" alt="captcha"'
    httpcha=re.findall(patterncha,htmlcha)
    pattern2=r'type="hidden" name="captcha-id" value="(.*?)"'
    hidden_value=re.findall(pattern2,htmlcha)
    print(hidden_value)

    post_data = {
        "source": "index_nav",
        'form_email': acount,
        'form_password': secret
    }
    if len(httpcha)>0:
        print('验证码连接',httpcha)
        capcha=get_captcha(httpcha[0])
        post_data['captcha-solution']=capcha
        post_data['captcha-id']=hidden_value[0]

    print (post_data)
    post_url='https://www.douban.com/accounts/login'
    login_page=session.post(post_url,data=post_data,headers=headers)
    #保存cookies
    session.cookies.save()

    if isLogin():
        print('登录成功')
    else:
        print('登录失败')


def get_movie_sort():
    time.sleep(1)
    movie_url='https://movie.douban.com/chart'
    html=session.get(movie_url,headers=headers)
    soup=BeautifulSoup(html.text,'html.parser')
    result=soup.find_all('a',{'class':'nbg'})
    print(result)

爬取短评论

def get_comment(filename):  #filename为爬取得内容保存的文件
    begin=1
    comment_url = 'https://movie.douban.com/subject/11600078/comments'
    next_url='?start=20&limit=20&sort=new_score&status=P'
    headers2 = {
            "Host": "movie.douban.com",
            "Referer": "https://www.douban.com/",
            'User-Agent': agent,
            'Connection': 'keep-alive',
        }
    f=open(filename,'w+',encoding='utf-8')
    while(True):
        time.sleep(6)
        html=session.get(url=comment_url+next_url,headers=headers2)
        soup=BeautifulSoup(html.text,'html.parser')

        #爬取当前页面的所有评论
        result=soup.find_all('div',{'class':'comment'}) #爬取得所有的短评
        pattern4 = r'<p class=""> (.*?)' \
                   r'</p>'
        for item in result:
            s=str(item)
            count2=s.find('<p class="">')
            count3=s.find('</p>')
            s2=s[count2+12:count3]  #抽取字符串中的评论
            if 'class' not in s2:
                f.write(s2)

        #获取下一页的链接
        next_url=soup.find_all('div',{'id':'paginator'})
        pattern3=r'href="(.*?)">后页'
        if(len(next_url)==0):
            break
        next_url=re.findall(pattern3,str(next_url[0]))  #得到后页的链接
        if(len(next_url)==0): #如果没有后页的链接跳出循环
            break
        next_url=next_url[0]
        print('%d爬取下一页评论...'%begin)
        begin=begin+1
        #如果爬取了5次则多休息2秒
        if(begin%6==0):
            time.sleep(40)
            print('休息...')
        print(next_url)
    f.close()

多线程爬虫，爬取豆瓣影评

def thread_get_comment(filename):
    next_url = '?start=19&limit=20&sort=new_score&status=P'
    headers2 = {
        "Host": "movie.douban.com",
        "Referer": "https://www.douban.com/",
        'User-Agent': agent,
        'Connection': 'keep-alive',
    }
    f = open(filename, 'w+', encoding='utf-8')
    comment_url = 'https://movie.douban.com/subject/26363254/comments'
    crawl_queue=[comment_url+next_url]
    crawl_queue.append('https://movie.douban.com/subject/26363254/comments?start=144&limit=20&sort=new_score&status=P')
    seen=set(crawl_queue)

    def process_queue():
        begin = 1
        while True:
            try:
                url=crawl_queue.pop()
            except  IndexError:
                break
            else:
                time.sleep(5)
                html = session.get(url=url,headers=headers2)
                soup = BeautifulSoup(html.text, 'html.parser')

                # 爬取当前页面的所有评论
                result = soup.find_all('div', {'class': 'comment'})  # 爬取得所有的短评
                pattern4 = r'<p class=""> (.*?)' \
                           r'</p>'
                for item in result:
                    s = str(item)
                    count2 = s.find('<p class="">')
                    count3 = s.find('</p>')
                    s2 = s[count2 + 12:count3]  # 抽取字符串中的评论
                    f.write(s2)

                # 获取下一页的链接
                next_url = soup.find_all('div', {'id': 'paginator'})
                pattern3 = r'href="(.*?)">后页'
                if (len(next_url) == 0):
                    break
                next_url = re.findall(pattern3, str(next_url[0]))  # 得到后页的链接
                if (len(next_url) == 0):  # 如果没有后页的链接跳出循环
                    break
                next_url = next_url[0]
                print('%d爬取下一页评论...' % begin)
                begin = begin + 1
                # 如果爬取了6次则多休息2秒
                if (begin % 6 == 0):
                    print('休息...')
                    time.sleep(30)

                print(next_url)
                if comment_url+next_url not in seen:
                    seen.add(comment_url+next_url)
                    crawl_queue.append(comment_url+next_url)

    threads=[]
    max_threads=5
    while threads or crawl_queue:
        for thread in threads:
            if not thread.is_alive():
                threads.remove(thread)
        while len(threads)< max_threads and crawl_queue:
            thread=threading.Thread(target=process_queue)
            print('--------下一个线程----------')
            thread.setDaemon(True) # set daemon so main thread can exit when receive ctrl + C
            thread.start()
            threads.append(thread)
        time.sleep(2)


    f.close()

if __name__=='__main__':
    if isLogin():
        print('您已经登录')
    else:
        print('xs')
        login('dsdz@qq.com','5sdfsd6')

    file_name='key3.txt'
    get_comment(file_name)        #单线程爬虫
    #thread_get_comment(file_name)  #多线程爬虫
    save_jieba_result(file_name)
    draw_wordcloud('pjl_jieba.txt')

爬取得评论保存在key3.txt 文本文件中：
这里写图片描述

生成图云

第一步需要：
安装必要的Python 库，其中需要的有生成图云scipy 、wordcloud。python库的安装方法，可以参考笔者的博客安装第三方库。一切准备就绪之后，就可以使用jieba分词对得到的所有评论进行分词，分词时候就可以绘制图云。
其中主要的代码 mywordCloud.py

from scipy.misc import  imread
import codecs
from os import  path
import jieba
from wordcloud import WordCloud

暂时没有用到

def get_all_keywords(file_name):
    word_lists=[]  #关键词列表
    with codecs.open(file_name,'r',encoding='utf-8') as f:
        Lists=f.readlines()
        for li in Lists:
            cut_list=list(jieba.cut(li))
            for word in cut_list:
                word_lists.append(word)

    word_lists_set=set(word_lists)  #去除相同的元素
    sort_count=[]
    word_lists_set=list(word_lists_set)

    length=len(word_lists_set)
    print(u'共有%d个关键词'%length)
    k = 1
    for w in word_lists_set:
        sort_count.append(w + u':' + str(word_lists.count(w)) + u"次\n")
        print(u"%d---" % k + w + u":" + str(word_lists.count(w)) + u"次")
        k += 1
    with codecs.open('count_word.txt', 'w', encoding='utf-8') as f:
        f.writelines(sort_count)

def save_jieba_result(file_name):
    #设置多线程切割
    #jieba.enable_parallel(4)
    dirs=path.join(path.dirname(__file__),file_name)
    print(dirs)
    with codecs.open(dirs,encoding='utf-8') as f:
        comment_text=f.read()
    cut_text=" ".join(jieba.cut(comment_text))
    with codecs.open('pjl_jieba.txt','w',encoding='utf-8') as f:
        f.write(cut_text)

def draw_wordcloud(file_name):
    with codecs.open(file_name,encoding='utf-8') as f:
        comment_text=f.read()
    color_mask=imread('timg.jpg') #读取背景图片
    stopwords = ['png','douban','com','href','https','img','img3','class','source','icon','shire',u'有点',u'真的',u'觉得',u'还是',u'一个',u'就是', u'电影', u'你们', u'这么', u'不过', u'但是', u'什么', u'没有', u'这个', u'那个', u'大家', u'比较', u'看到', u'真是',
                 u'除了', u'时候', u'已经', u'可以']
    font = r'C:\Windows\Fonts\simfang.ttf'
    cloud=WordCloud(font_path=font,background_color='white',max_words=20000,max_font_size=200,min_font_size=10,mask=color_mask,stopwords=stopwords)
    word_cloud=cloud.generate(comment_text)  #产生词云
    word_cloud.to_file('mycloud.jpg')