4）点击【搜索按钮】，进行评论数据搜索，寻找trul_url
当出现如下界面后，按照如图所示操作即可。
紧接着，查看该请求所对应的Request URL，就是我们最终要找的true_url。信不信吗？下面可以试试。
6、进行第三次请求测试
首先我们在上述图的基础上，顺着Request URL往下面找，获取Request Headers中user-agent、referer、cokie这3样东西。
反正都在Request Headers中，我们将这3样东西，一一粘贴到headers中，形成一个字典格式的键值对。然后我们发起其三次请求。
true_url = "https://rate.tmall.com/list_detail_rate.htm?itemId=43751299764&spuId=864684242&sellerId=2380958892&order=3¤tPage=1&append=0&content=1&tagId=&posi=&picture=&groupId=&ua=098%23E1hvmvvPvBvvUvCkvvvvvjiPn25UQjlHnLs9ljD2PmPUtjEmPFsUzjiUPsFh0jYR9phv2nsGdj2qzYswMjBt7u6CvvDvpiQZN9CC0Jujvpvhphhvv2yCvvBvpvvvKphv8hCvvvvvvhjPphvhqvvvpkxvpCBXvvC2p6CvHHyvvh84phvWipvvpS0EvpvVpyUUCCQOuphvmhCvCW780nVlmphvLhm6DQmFdcHmjobWlCkfJy3Qp%2BoQRqwiL7CpqU0QKoZHaysEDLuTWD19C7zhdixdBAyZgn94e7Q4b64B9CkaU6UsxI2wjLyDZLxjKOvCvpvVphhvvvvv2QhvCvvvMM%2F5vpvhphvhH86CvCHUooppwvCvws1B34ZzKsELW86CvvDvp3%2BpDQCCTzurvpvEphHi7WWvpORURphvChCvvvv%3D&needFold=0&_ksTS=1585207007017_726&callback=jsonp727" headers ={ # 用的哪个浏览器 'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.149 Safari/537.36', # 从哪个页面发出的数据申请，每个网站可能略有不同 'referer': 'https://detail.tmall.com/item.htm?spm=a220m.1000858.1000725.1.744840c2FKDkFG&id=43751299764&skuId=4493124079453&areaId=421300&user_id=2380958892&cat_id=2&is_b=1&rn=388ceadeefb8d85e5bae2d83bd0b732a', # 哪个用户想要看数据，是游客还是注册用户,建议使用登录后的cookie 'cookie': 'tk_trace=1; cna=qzxtFlkIdkICARsvgIh8ftNm; t=972672ef4a0555634bb4c53147d9c209; _tb_token_=f9ed83467e7ab; cookie2=11c397b08781b52815002215ea5d1ad4; dnk=huang%5Cu81F3%5Cu5C0A; tracknick=huang%5Cu81F3%5Cu5C0A; lid=huang%E8%87%B3%E5%B0%8A; lgc=huang%5Cu81F3%5Cu5C0A; uc1=cookie16=UIHiLt3xCS3yM2h4eKHS9lpEOw%3D%3D&pas=0&existShop=false&cookie15=UtASsssmOIJ0bQ%3D%3D&cookie14=UoTUP2D4F2IHjA%3D%3D&cookie21=VFC%2FuZ9aiKCaj7AzMHh1; uc3=id2=UU8BrRJJcs7Z0Q%3D%3D&lg2=VT5L2FSpMGV7TQ%3D%3D&vt3=F8dBxd9hhEzOWS%2BU9Dk%3D&nk2=CzhMCY1UcRnL; _l_g_=Ug%3D%3D; uc4=id4=0%40U22GV4QHIgHvC14BqrCleMrzYb3K&nk4=0%40CX8JzNJ900MInLAoQ2Z33x1zsSo%3D; unb=2791663324; cookie1=BxeNCqlvVZOUgnKrsmThRXrLiXfQF7m%2FKvrURubODpk%3D; login=true; cookie17=UU8BrRJJcs7Z0Q%3D%3D; _nk_=huang%5Cu81F3%5Cu5C0A; sgcookie=E53NoUsJWtrYT7Pyx14Px; sg=%E5%B0%8A41; csg=8d6d2aae; enc=VZMEO%2BOI3U59DBFwyF9LE3kQNM84gfIKeZFLokEQSzC5TubpmVCJlS8olhYmgHiBe15Rvd8rsOeqeC1Em9GfWA%3D%3D; l=dBLKMV6rQcVJihfaBOfgSVrsTkQ9UIRb8sPrQGutMICP9ZCwNsyFWZ4Kb-8eCnGVHsMvR3oGfmN0BDTHXyIVokb4d_BkdlkmndC..; isg=BK2tcrfNj3CNMWubo5GaxlajvEknCuHcPbxLgO-yO8QhZswYt1ujrPVwUDqAZvmU' } data = requests.get(true_url,headers=headers).text data
结果如下：
结果分析：经过一番波折，我们最终找到了我们想要获取的数据，接下来的话，就是我们进行页面解析的工作了。其实在真实的爬虫环境中，可能会遇到更多的反爬措施，真正难得不是解析网页，而是分析网页和反爬。
7、获取网页中的评论数据
result = re.findall('rateContent":"(.*?)"fromMall"',data) result
结果如下：
8、翻页爬取(最终代码)
我们的目的肯定不只是爬取一个页面的评论数据，而是进行翻页爬取，我们需要仔细观察true_url中，有一个【currentPage=1】参数，当这个数字变化的时候，对应的页面也就发生的变化，基于此，我们将完整的爬虫代码写在下面。
import pandas as pd import requests import re import time data_list = [] for i in range(1,300,1): print("正在爬取第" + str(i) + "页") url = first + str(i) + last headers ={ # 用的哪个浏览器 'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.149 Safari/537.36', # 从哪个页面发出的数据申请，每个网站可能略有不同 'referer': 'https://detail.tmall.com/item.htm?spm=a220m.1000858.1000725.1.744840c2FKDkFG&id=43751299764&skuId=4493124079453&areaId=421300&user_id=2380958892&cat_id=2&is_b=1&rn=388ceadeefb8d85e5bae2d83bd0b732a', # 哪个用户想要看数据，是游客还是注册用户,建议使用登录后的cookie 'cookie': 'tk_trace=1; cna=qzxtFlkIdkICARsvgIh8ftNm; t=972672ef4a0555634bb4c53147d9c209; _tb_token_=f9ed83467e7ab; cookie2=11c397b08781b52815002215ea5d1ad4; dnk=huang%5Cu81F3%5Cu5C0A; tracknick=huang%5Cu81F3%5Cu5C0A; lid=huang%E8%87%B3%E5%B0%8A; lgc=huang%5Cu81F3%5Cu5C0A; uc1=cookie16=UIHiLt3xCS3yM2h4eKHS9lpEOw%3D%3D&pas=0&existShop=false&cookie15=UtASsssmOIJ0bQ%3D%3D&cookie14=UoTUP2D4F2IHjA%3D%3D&cookie21=VFC%2FuZ9aiKCaj7AzMHh1; uc3=id2=UU8BrRJJcs7Z0Q%3D%3D&lg2=VT5L2FSpMGV7TQ%3D%3D&vt3=F8dBxd9hhEzOWS%2BU9Dk%3D&nk2=CzhMCY1UcRnL; _l_g_=Ug%3D%3D; uc4=id4=0%40U22GV4QHIgHvC14BqrCleMrzYb3K&nk4=0%40CX8JzNJ900MInLAoQ2Z33x1zsSo%3D; unb=2791663324; cookie1=BxeNCqlvVZOUgnKrsmThRXrLiXfQF7m%2FKvrURubODpk%3D; login=true; cookie17=UU8BrRJJcs7Z0Q%3D%3D; _nk_=huang%5Cu81F3%5Cu5C0A; sgcookie=E53NoUsJWtrYT7Pyx14Px; sg=%E5%B0%8A41; csg=8d6d2aae; enc=VZMEO%2BOI3U59DBFwyF9LE3kQNM84gfIKeZFLokEQSzC5TubpmVCJlS8olhYmgHiBe15Rvd8rsOeqeC1Em9GfWA%3D%3D; l=dBLKMV6rQcVJihfaBOfgSVrsTkQ9UIRb8sPrQGutMICP9ZCwNsyFWZ4Kb-8eCnGVHsMvR3oGfmN0BDTHXyIVokb4d_BkdlkmndC..; isg=BK2tcrfNj3CNMWubo5GaxlajvEknCuHcPbxLgO-yO8QhZswYt1ujrPVwUDqAZvmU' } try: data = requests.get(url,headers = headers).text time.sleep(10) result = re.findall('rateContent":"(.*?)"fromMall"',data) data_list.extend(result) except: print("本页爬取失败") df = pd.DataFrame() df["评论"] = data_list df.to_excel("评论_汇总.xlsx")
结果如下：
9、词云图的制作
import numpy as np import pandas as pd import jieba from wordcloud import WordCloud import matplotlib.pyplot as plt from imageio import imread import warnings warnings.filterwarnings("ignore") # 读取数据 df = pd.read_excel("评论_汇总.xlsx") df.head() # 利用jieba进行分析操作 df["评论"] = df["评论"].apply(jieba.lcut) df.head() # 去除停用词操作 with open("stopword.txt","r",encoding="gbk") as f: stop = f.read() # 返回的是一个字符串 stop = stop.split() # 这里得到的是一个列表.split()会将空格，\n，\t进行切分，因此我们可以将这些加到停用词当中 stop = stop + [" ","\n","\t"] df_after = df["评论"].apply(lambda x: [i for i in x if i not in stop]) df_after.head() # 词频统计 all_words = [] for i in df_after: all_words.extend(i) word_count = pd.Series(all_words).value_counts() word_count[:10] # 绘制词云图 # 1、读取背景图片 back_picture = imread(r"G:\6Tipdm\wordcloud\alice_color.png") # 2、设置词云参数 wc = WordCloud(font_path="G:\\6Tipdm\\wordcloud\\simhei.ttf", background_color="white", max_words=2000, mask=back_picture, max_font_size=200, random_state=42 ) wc2 = wc.fit_words(word_count) # 3、绘制词云图 plt.figure(figsize=(16,8)) plt.imshow(wc2) plt.axis("off") plt.show() wc.to_file("ciyun.png")
