拉勾网爬取全国python职位并数据分析薪资，工作经验，学历等信息

首先前往拉勾网“爬虫”职位相关页面
确定网页的加载方式是JavaScript加载
通过谷歌浏览器开发者工具分析和寻找网页的真实请求，确定真实数据在position.Ajax开头的链接里，请求方式是POST
使用requests的post方法获取数据，发现并没有返回想要的数据，说明需要加上headers和每隔多长时间爬取
我们可以看到拉勾网列表页的信息一般js加载的都在xhr和js中，通过发送ajax加载POST请求，获取页面信息。
这个是ajax的头信息，通过Form Data中的的信息获取页面
下面是scrapy爬虫的代码部分

 1 import scrapy
 2 import json
 3 from lagou.items import LagouItem
 4 class LagoupositionSpider(scrapy.Spider):
 5     name = 'lagouposition'
 6     allowed_domains = ['lagou.com']
 7     kd = input('请输入你要搜索的职位信息:')
 8     ct =input('请输入要搜索的城市信息')
 9     page=1
10     start_urls = ["https://www.lagou.com/jobs/list_"+str(kd)+"&city="+str(ct)]
11     headers={"User-Agent": "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36",
12              'Referer': 'https://www.lagou.com/jobs/list_'+str(kd)+'?labelWords=&fromSearch=true&suginput=',
13              'Cookie':' _ga=GA1.2.1036647455.1532143907; user_trace_token=20180721113217-aacd6291-8c96-11e8-a020-525400f775ce; LGUID=20180721113217-aacd667e-8c96-11e8-a020-525400f775ce; index_location_city=%E5%8C%97%E4%BA%AC; _gid=GA1.2.1320510576.1532272161; WEBTJ-ID=20180723084204-164c4960832159-09bf89fcd2732e-5e442e19-1049088-164c496083348; JSESSIONID=ABAAABAABEEAAJAC7D58B57D1CAE4616ED47AACF945615E; _gat=1; LGSID=20180723203627-04b27de6-8e75-11e8-9ee6-5254005c3644; PRE_UTM=; PRE_HOST=www.baidu.com; PRE_SITE=https%3A%2F%2Fwww.baidu.com%2Flink%3Furl%3DYhfCtaCVlOHCdncJxMCMMS3PB1wGlwfw9Yt2c_FXqgu%26wd%3D%26eqid%3D8f013ed00002f4c7000000035b55cbc4; PRE_LAND=https%3A%2F%2Fwww.lagou.com%2F; Hm_lvt_4233e74dff0ae5bd0a3d81c6ccf756e6=1532306722,1532306725,1532306732,1532349358; SEARCH_ID=cdd7822cf3e2429fbc654720657d5873; LGRID=20180723203743-3221dec8-8e75-11e8-a35a-525400f775ce; Hm_lpvt_4233e74dff0ae5bd0a3d81c6ccf756e6=1532349434; TG-TRACK-CODE=search_code'
14              }
15 
16 
17     def parse(self, response):
18         with open('lagou.html','w') as f:
19             f.write(response.text)
20         url="https://www.lagou.com/jobs/positionAjax.json?needAddtionalResult=false"
21         formdata={'first':'true','kd':str(self.kd),'pn':'1','city':str(self.ct)}
22         yield scrapy.FormRequest(url,formdata=formdata,callback=self.parse_detail,headers=self.headers)
23 
24     def parse_detail(self,response):
25         text=json.loads(response.text)
26         res=[]
27         try:
28             res = text["content"]["positionResult"]["result"]
29             print(res)
30         except:
31             pass
32         if len(res)>0:
33             item = LagouItem()
34             for position in res:
35                 try:
36                     item['title']=position['positionName']
37                     item['education']=position['education']
38                     item['company']=position['companyFullName']
39                     item['experience']=position['workYear']
40                     item['location']=position['city']
41                     item['salary'] = position['salary']
42                     print(item)
43                 except:
44                     pass
45                 yield item
46             self.page+=1
47             url='https://www.lagou.com/jobs/positionAjax.json?needAddtionalResult=false
48             formdata={'first':'false','kd':str(self.kd),'pn':str(self.page),'city':str(self.ct)}
49             print('===========================',formdata)
50             yield scrapy.FormRequest(url, callback=self.parse_detail, formdata=formdata,headers=self.headers)
51         else:
52             print("爬取结束！")

注意拉钩网有反爬措施，我们在Formreqest提交POST请求消息必须携带kd等键值对，在setting中也许设置

 1 DOWNLOAD_DELAY = 20
 2 #设置爬取时间
 3 ROBOTSTXT_OBEY = False
 4 #是否遵循发爬虫协议
 5 DEFAULT_REQUEST_HEADERS = {
 6     'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
 7     'Accept-Language': 'zh-CN,zh;q=0.8',
 8     'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
 9     'Host': 'www.lagou.com',
10     'Origin': 'https://www.lagou.com',
11     'Referer': 'https://www.lagou.com/jobs',
12     'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/60.0.3112.113 Safari/537.36',
13     'X-Anit-Forge-Code': '0',
14     'X-Anit-Forge-Token': 'None',
15     'X-Requested-With': 'XMLHttpRequest'
16 }
17 #请求头信息headers

接下来就是在items中设置爬取信息的字段

 1 import scrapy
 2 
 3 
 4 class LagouItem(scrapy.Item):
 5     # define the fields for your item here like:
 6     # name = scrapy.Field()
 7     # pass
 8 
 9     education= scrapy.Field()
10     company= scrapy.Field()
11     experience= scrapy.Field()
12     location= scrapy.Field()
13     salary= scrapy.Field()
14     title= scrapy.Field()

在Pipeline.py文件中设置保存爬取文件的格式等

 1 import json
 2 class LagouPipeline(object):
 3     def open_spider(self,spider):
 4         self.file=open('pythonposition.json','w',encoding='utf-8')
 5     def process_item(self, item, spider):
 6         python_dict=dict(item)
 7         content=json.dumps(python_dict,ensure_ascii=False)+'\n'
 8         self.file.write(content)
 9         return item
10     def close_spider(self,spider):
11         self.file.close()

注意一定要把setting中的ITEM_PIPELINES解注释，接下来就是跑起我们的项目，通过input输入想要爬取的职位和城市，

上面就是爬取到的信息总共是855条招聘消息，接下来就是用jumpter-notebook打开爬取到的csv文件用pandas,numpy,和mupltlib进行分析

1 import pandas as pd
2 import numpy as np
3 import seaborn as sns
4 lagou=pd.read_csv('./examples/lagou.csv')
5 lagou.info()
6 #查看缺失值情况

通过读取文件并显示出855条招聘信息是否有缺失值

1 city=lagou['location']
2 city=pd.DataFrame(city.unique())
3 city

通过上面可以看到招聘python职位的城市，总共有38城市

 1 education=lagou['education']
 2 education=pd.DataFrame(education.unique())
 3 lagou['education'] = lagou['education'].replace('不限','unlimited')
 4 lagou['education'] = lagou['education'].replace('大专','junior')
 5 lagou['education'] = lagou['education'].replace('本科','regular')
 6 lagou['education'] = lagou['education'].replace('硕士','master')
 7 lagou['education'] = lagou['education'].replace('博士','doctor')
 8 #seaborn不支持中文需将对应的中文替换
 9 import seaborn as sns
10 sns.set_style('whitegrid')
11 sns.countplot(x='education',data=lagou,palette='Greens_d')

通过上图可以看到大多数的Python职位招聘还是本科学历为主

 1 experience=lagou['experience']
 2 experience=pd.DataFrame(experience.unique())
 3 lagou['experience'] = lagou['experience'].replace('不限','unlimited')
 4 lagou['experience'] = lagou['experience'].replace('3-5年','3-5')
 5 lagou['experience'] = lagou['experience'].replace('1-3年','1-3')
 6 lagou['experience'] = lagou['experience'].replace('5-10年','5-10')
 7 lagou['experience'] = lagou['experience'].replace('1年以下','<1')
 8 lagou['experience'] = lagou['experience'].replace('应届毕业生','intern')
 9 experience
10 sns.countplot(x="experience", data=lagou,palette="Blues_d")

上图是招聘的工作经验的人数分布图，可以看到3-5年的Python工程师比较抢手，其次就是1-3年工作经验的

 1 import matplotlib.pyplot as plt
 2 %matplotlib inline
 3 f, ax1= plt.subplots(figsize=(20,20))
 4 sns.countplot(y='salary', data=lagou, ax=ax1)
 5 ax1.set_title('Python salary distribute ',fontsize=15)
 6 #薪资分布
 7 ax1.set_xlabel('salary')
 8 #薪资
 9 ax1.set_ylabel('level')             
10 plt.show()