使用python3去爬取汽车之家的汽车数据(源码分享)

时间:2018-11-24
本文章向大家介绍使用python3去爬取汽车之家的汽车数据(源码分享),需要的朋友可以参考一下

  题记:

  互联网上关于使用python3去爬取汽车之家的汽车数据(主要是汽车基本参数,配置参数,颜色参数,内饰参数)的教程已经非常多了,但大体的方案分两种:

  1.解析出汽车之家某个车型的网页,然后正则表达式匹配出混淆后的数据对象与混淆后的js,并对混淆后的js使用pyv8进行解析返回正常字符,然后通过字符与数据对象进行匹配,具体方法见这位园友,传送门:https://www.cnblogs.com/my8100/p/js_qichezhijia.html (感谢这位大神前半部分的思路)

       2.解析出汽车之家某个车型的网页,然后正则表达式匹配出混淆后的数据对象与混淆后的js,针对混淆后的js进行进行手动匹配,因为混淆的js大概分为8大类(无参数 返回常量,无参数 返回函数,参数等于返回值函数,无参数 返回常量,无参数 返回常量中间无混淆代码,字符串拼接时使无参常量,字符串拼接时使用返回参数的函数),然后通过正则表达式进行解析出8类内容并进行逐个替换,最终也会返回一个带有顺序的字符串,将这个字符串与前边的数据对象再次替换,最终数据对象中的所有span都会被替换成中文,具体操作见园友的地址,传送门:https://www.cnblogs.com/dyfblog/p/6753251.html (感谢这位大神前半部分的思路)

不过鉴于作者技术有限,上述的两种方案,我都没有完整的执行完成,哪怕花了一周的时间也没有,但是没有办法,谁让我是一个很爱钻牛角尖的人呢,下一步提出我自己琢磨出来的方案,流程上稍微有点复杂,但是稳打稳扎,还是可以爬出来的,好了话不多说了,贴出步骤;

1.获取所有车型的网页,保存到本地: 

 1 import bs4
 2 import requests as req
 3 '''
 4 第一步,下载出所有车型的网页。
 5 '''
 6 def mainMethod():
 7     '''
 8     解析汽车之家所有车型数据保存到D盘
 9     '''
10     li = [chr(i) for i in range(ord("A"),ord("Z")+1)]
11     firstSite="https://www.autohome.com.cn/grade/carhtml/"
12     firstSiteSurfixe=".html"
13     secondSite = "https://car.autohome.com.cn/config/series/"
14     secondSiteSurfixe = ".html"
15 
16     for a in li:
17         if a is not None:
18             requestUrl = firstSite+a+firstSiteSurfixe
19             print(requestUrl)
20             #开始获取每个品牌的车型
21             resp = req.get(requestUrl)
22             # print(str(resp.content,"gbk"))
23             bs = bs4.BeautifulSoup(str(resp.content,"gbk"),"html.parser")
24             bss = bs.find_all("li")
25             con = 0
26             for b in bss:
27                 d = b.h4
28                 if d is not None:
29                     her = str(d.a.attrs['href'])
30                     her = her.split("#")[0]
31                     her = her[her.index(".cn")+3:].replace("/",'')
32                     if her is not None:
33                         secSite = secondSite +her + secondSiteSurfixe
34                         print("secSite="+secSite)
35                         # print(secSite)
36                         #奥迪A3
37                         if her is not None:
38                             resp = req.get(secSite)
39                             text = str(resp.content,encoding="utf-8")
40                             print(a)
41                             fil = open("d:\\autoHome\\html\\"+str(her),"a",encoding="utf-8")
42                             fil.write(text)
43                     con = (con+1)
44             else:
45                 print(con)
46 if __name__ =="__main__":
47     mainMethod()

2.解析出每个车型的关键js并拼装成一个html,保存到本地。

 1 import os
 2 import re
 3 '''
 4 第二步,解析出每个车型的关键js拼装成一个html
 5 '''
 6 if __name__=="__main__":
 7     print("Start...")
 8     rootPath = "D:\\autoHome\\html\\"
 9     files = os.listdir(rootPath)
10     for file in files:
11         print("fileName=="+file.title())
12         text = ""
13         for fi in open(rootPath+file,'r',encoding="utf-8"):
14             text = text+fi
15         else:
16             print("fileName=="+file.title())
17         #解析数据的json
18         alljs = ("var rules = '2';"
19                  "var document = {};"
20                  "function getRules(){return rules}"
21                  "document.createElement = function() {"
22                  "      return {"
23                  "              sheet: {"
24                  "                      insertRule: function(rule, i) {"
25                  "                              if (rules.length == 0) {"
26                  "                                      rules = rule;"
27                  "                              } else {"
28                  "                                      rules = rules + '#' + rules;"
29                  "                              }"
30                  "                      }"
31                  "              }"
32                  "      }"
33                  "};"
34                 
37                  "document.head = {};"
38                  "document.head.appendChild = function() {};"
39 
40                  "var window = {};"
41                  "window.decodeURIComponent = decodeURIComponent;")
42         try:
43             js = re.findall('(\(function\([a-zA-Z]{2}.*?_\).*?\(document\);)', text)
44             for item in js:
45                 alljs = alljs + item
46         except Exception as e:
47             print('makejs function exception')
48 
49 
50         newHtml = "<html><meta http-equiv='Content-Type' content='text/html; charset=utf-8' /><head></head><body>    <script type='text/javascript'>"
51         alljs = newHtml + alljs+" document.write(rules)</script></body></html>"
52         f = open("D:\\autoHome\\newhtml\\"+file+".html","a",encoding="utf-8")
53         f.write(alljs)
54         f.close()

3.解析出每个车型的数据json,比如var config  ,var option , var bag  var innerbag..但我就解析了基本信息跟配置信息,其他的无所谓。

 1 import os
 2 import re
 3 '''
 4 解析出每个车型的数据json,保存到本地。
 5 '''
 6 if __name__=="__main__":
 7     print("Start...")
 8     rootPath = "D:\\autoHome\\html\\"
 9     files = os.listdir(rootPath)
10     for file in files:
11         print("fileName=="+file.title())
12         text = ""
13         for fi in open(rootPath+file,'r',encoding="utf-8"):
14             text = text+fi
15         else:
16             print("fileName=="+file.title())
17         #解析数据的json
18         jsonData = ""
19         config = re.search('var config = (.*?){1,};',text)
20         if config!= None:
21             print(config.group(0))
22             jsonData = jsonData+ config.group(0)
23         option = re.search('var option = (.*?)};',text)
24         if option != None:
25             print(option.group(0))
26             jsonData = jsonData+ option.group(0)
27         bag = re.search('var bag = (.*?);',text)
28         if bag != None:
29             print(bag.group(0))
30             jsonData = jsonData+ bag.group(0)
31         # print(jsonData)
32         f = open("D:\\autoHome\\json\\"+file,"a",encoding="utf-8")
33         f.write(jsonData)
34         f.close()

4.生成样式文件,保存 到本地。

 1 import os
 2 from selenium import webdriver
 3 
 4 '''
 5     第四步,浏览器执行第二步生成的html文件,抓取执行结果,保存到本地。
 6 '''
 7 class Crack():
 8     def __init__(self,keyword,username,passod):
 9         self.url = 'https://www.baidu.com'
10         self.browser = webdriver.Chrome('E:\work\ChromePortable\App\Google Chrome\chromedriver.exe')
11 
12 if __name__=="__main__":
13     lists = os.listdir("D:/autoHome/newHtml/")
14     for fil in lists:
15         file = os.path.exists("D:\\autoHome\\content\\"+fil)
16         if file :
17             print('文件已经解析。。。'+str(file))
18             continue
19         crack = Crack('测试公司','17610177519','17610177519')
20 
21         print(fil)
22         crack.browser.get("file:///D:/autoHome/newHtml/"+fil+"")
23         text = crack.browser.find_element_by_tag_name('body')
24         print(text.text)
25         f = open("D:\\autoHome\\content\\"+fil,"a",encoding="utf-8")
26         f.write(text.text)
27         f.close()
28         crack.browser.close()

5.读取样式文件,匹配数据文件,生成正常数据文件

 1 import os
 2 import re
 3 '''
 4 匹配样式文件与json数据文件,生成正常的数据文件。
 5 '''
 6 if __name__ =="__main__":
 7     rootPath = "D:\\autoHome\\json\\"
 8     listdir = os.listdir(rootPath)
 9     for json_s in listdir:
10         print(json_s.title())
11         jso = ""
12         #读取json数据文件
13         for fi in open(rootPath+json_s,'r',encoding="utf-8"):
14             jso = jso+fi
15         content = ""
16         #读取样式文件
17         spansPath = "D:\\autoHome\\content\\"+json_s.title()+".html"
18         # print(spansPath)
19         for spans in  open(spansPath,"r",encoding="utf-8"):
20             content = content+ spans
21         print(content)
22         #获取所有span对象
23         jsos = re.findall("<span(.*?)></span>",jso)
24         num = 0
25         for js in jsos:
26             print("匹配到的span=>>"+js)
27             num = num +1
28             #获取class属性值
29             sea = re.search("'(.*?)'",js)
30             print("匹配到的class==>"+sea.group(1))
31             spanContent = str(sea.group(1))+"::before { content:(.*?)}"
32             #匹配样式值
33             spanContentRe = re.search(spanContent,content)
34             if spanContentRe != None:
35                 if sea.group(1) != None:
36                     print("匹配到的样式值="+spanContentRe.group(1))
37                     jso = jso.replace(str("<span class='"+sea.group(1)+"'></span>"),re.search("\"(.*?)\"",spanContentRe.group(1)).group(1))
38         print(jso)
39         fi = open("D:\\autoHome\\newJson\\"+json_s.title(),"a",encoding="utf-8")
40         fi.write(jso)
41         fi.close()

6.到前五步已经可以看到json数据文件都已经是混淆前的了,说明已经爬取成功了。

7.读取数据文件,生成excel

 1 import json
 2 import os
 3 import re
 4 import xlwt
 5 '''
 6 读取数据文件,生成excel
 7 '''
 8 if __name__ == "__main__":
 9     rootPath = "D:\\autoHome\\newJson\\"
10     workbook = xlwt.Workbook(encoding = 'ascii')#创建一个文件
11     worksheet = workbook.add_sheet('汽车之家')#创建一个表
12     files = os.listdir(rootPath)
13     startRow = 0
14     isFlag = True #默认记录表头
15     for file in files:
16         list = []
17         carItem = {}
18         print("fileName=="+file.title())
19         text = ""
20         for fi in open(rootPath+file,'r',encoding="utf-8"):
21             text = text+fi
22         # else:
23             # print("文件内容=="+text)
24         #解析基本参数配置参数,颜色三种参数,其他参数
25         config = "var config = (.*?);"
26         option = "var option = (.*?);var"
27         bag = "var bag = (.*?);"
28 
29         configRe = re.findall(config,text)
30         optionRe = re.findall(option,text)
31         bagRe = re.findall(bag,text)
32         for a in configRe:
33             config = a
34         print("++++++++++++++++++++++\n")
35         for b in optionRe:
36             option = b
37             print("---------------------\n")
38         for c in bagRe:
39             bag = c
40         # print(config)
41         # print(option)
42         # print(bag)
43         config = json.loads(config)
44         option = json.loads(option)
45         # print(bag)
46         try:
47             bag = json.loads(bag)
48             # print(config)
49             # print(option)
50             # print(bag)
51             path = "D:\\autoHome\\autoHome.xls"
52 
53             configItem = config['result']['paramtypeitems'][0]['paramitems']
54             optionItem = option['result']['configtypeitems'][0]['configitems']
55             optionItem = option['result']['configtypeitems'][0]['configitems']
56         except Exception as e:
57             f =  open("D:\\autoHome\\异常数据\\exception.txt","a",encoding="utf-8")
58             f.write(file.title()+"\n")
59 
60         #解析基本参数
61         for car in configItem:
62             carItem[car['name']]=[]
63             for ca in car['valueitems']:
64                 carItem[car['name']].append(ca['value'])
65         # print(carItem)
66         #解析配置参数
67         for car in optionItem:
68             carItem[car['name']]=[]
69             for ca in car['valueitems']:
70                 carItem[car['name']].append(ca['value'])
71 
72         if isFlag:
73             co1s = 0
74 
75             for co in carItem:
76                 co1s = co1s +1
77                 worksheet.write(startRow,co1s,co)
78             else:
79                 startRow = startRow+1
80                 isFlag = False
81 
82         #计算起止行号
83         endRowNum = startRow + len(carItem['最大扭矩(N·m)']) #车辆款式记录数
84         for row in range(startRow,endRowNum):
85             print(row)
86             colNum = 0
87             for col in carItem:
88 
89                 colNum = colNum +1
90                 print(str(carItem[col][row-startRow]),end='|')
91                 worksheet.write(row,colNum,str(carItem[col][row-startRow]))
92 
93         else:
94             startRow  = endRowNum
95     workbook.save('d:\\autoHome\\Mybook.xls')

8.最后打开excel文件,给你们看看。

数据量大概有8300的样子。以后买车就用这个参考了。