标签:wan htm img 发送 spider span pre https lock
爬取链家深圳全部二手房的详细信息,并将爬取的数据存储到CSV文件中
class LianjiaSpider(object):
def __init__(self):
def getMaxPage(self, url): # 获取maxPage
def parsePage(self, url): # 解析每个page,获取每个huose的Link
def parseDetail(self, url): # 根据Link,获取每个house的详细信息
· hearders用到了fake_useragent库,用来随机生成请求头。
· datas空列表,用于保存爬取的数据。
def __init__(self):
self.headers = {"User-Agent": UserAgent().random}
self.datas = list()
主要用来获取二手房页面的最大页数.
def getMaxPage(self, url):
response = requests.get(url, headers = self.headers)
if response.status_code == 200:
source = response.text
soup = BeautifulSoup(source, "html.parser")
pageData = soup.find("div", class_ = "page-box house-lst-page-box")["page-data"]
# pageData = ‘{"totalPage":100,"curPage":1}‘,通过eval()函数把字符串转换为字典
maxPage = eval(pageData)["totalPage"]
return maxPage
else:
print("Fail status: {}".format(response.status_code))
return None
主要是用来进行翻页的操作,得到每一页的所有二手房的Links链接。它通过利用一个for循环来重构 url实现翻页操作,而循环最大页数就是通过上面的 getMaxPage() 来获取到。
def parsePage(self, url):
maxPage = self.getMaxPage(url)
# 解析每个page,获取每个二手房的链接
for pageNum in range(1, maxPage+1 ):
url = "https://sz.lianjia.com/ershoufang/pg{}/".format(pageNum)
print("当前正在爬取: {}".format(url))
response = requests.get(url, headers = self.headers)
soup = BeautifulSoup(response.text, "html.parser")
links = soup.find_all("div", class_ = "info clear")
for i in links:
link = i.find("a")["href"] #每个<info clear>标签有很多<a>,而我们只需要第一个,所以用find
detail = self.parseDetail(link)
self.datas.append(detail)
根据parsePage()函数获取的二手房Link链接,向该链接发送请求,获取出详细页面信息。
def parseDetail(self, url):
response = requests.get(url, headers = self.headers)
detail = {}
if response.status_code == 200:
soup = BeautifulSoup(response.text, "html.parser")
detail["价格"] = soup.find("span", class_ = "total").text
detail["单价"] = soup.find("span", class_ = "unitPriceValue").text
detail["小区"] = soup.find("div", class_ = "communityName").find("a", class_ = "info").text
detail["位置"] = soup.find("div", class_="areaName").find("span", class_="info").text
detail["地铁"] = soup.find("div", class_="areaName").find("a", class_="supplement").text
base = soup.find("div", class_ = "base").find_all("li") # 基本信息
detail["户型"] = base[0].text[4:]
detail["面积"] = base[2].text[4:]
detail["朝向"] = base[6].text[4:]
detail["电梯"] = base[10].text[4:]
return detail
else:
return None
这里用到了 pandas 库的 DataFrame() 方法,它默认的是按照列名的字典顺序排序的。想要自定义列的顺序,可以加columns字段。
# 将所有爬取的二手房数据存储到csv文件中
data = pd.DataFrame(self.datas)
# columns字段:自定义列的顺序(DataFrame默认按列名的字典序排序)
columns = ["小区", "户型", "面积", "价格", "单价", "朝向", "电梯", "位置", "地铁"]
data.to_csv(".\Lianjia_II.csv", encoding=‘utf_8_sig‘, index=False, columns=columns)
# -* coding: utf-8 *-
#author: wangshx6
#data: 2018-11-07
#descriptinon: 爬取链家深圳全部二手房的详细信息,并将爬取的数据存储到CSV文
import requests
from bs4 import BeautifulSoup
import pandas as pd
from fake_useragent import UserAgent
class LianjiaSpider(object):
def __init__(self):
self.headers = {"User-Agent": UserAgent().random}
self.datas = list()
def getMaxPage(self, url):
response = requests.get(url, headers = self.headers)
if response.status_code == 200:
source = response.text
soup = BeautifulSoup(source, "html.parser")
pageData = soup.find("div", class_ = "page-box house-lst-page-box")["page-data"]
# pageData = ‘{"totalPage":100,"curPage":1}‘,通过eval()函数把字符串转换为字典
maxPage = eval(pageData)["totalPage"]
return maxPage
else:
print("Fail status: {}".format(response.status_code))
return None
def parsePage(self, url):
maxPage = self.getMaxPage(url)
# 解析每个page,获取每个二手房的链接
for pageNum in range(1, maxPage+1 ):
url = "https://sz.lianjia.com/ershoufang/pg{}/".format(pageNum)
print("当前正在爬取: {}".format(url))
response = requests.get(url, headers = self.headers)
soup = BeautifulSoup(response.text, "html.parser")
links = soup.find_all("div", class_ = "info clear")
for i in links:
link = i.find("a")["href"] #每个<info clear>标签有很多<a>,而我们只需要第一个,所以用find
detail = self.parseDetail(link)
self.datas.append(detail)
# 将所有爬取的二手房数据存储到csv文件中
data = pd.DataFrame(self.datas)
# columns字段:自定义列的顺序(DataFrame默认按列名的字典序排序)
columns = ["小区", "户型", "面积", "价格", "单价", "朝向", "电梯", "位置", "地铁"]
data.to_csv(".\Lianjia_II.csv", encoding=‘utf_8_sig‘, index=False, columns=columns)
def parseDetail(self, url):
response = requests.get(url, headers = self.headers)
detail = {}
if response.status_code == 200:
soup = BeautifulSoup(response.text, "html.parser")
detail["价格"] = soup.find("span", class_ = "total").text
detail["单价"] = soup.find("span", class_ = "unitPriceValue").text
detail["小区"] = soup.find("div", class_ = "communityName").find("a", class_ = "info").text
detail["位置"] = soup.find("div", class_="areaName").find("span", class_="info").text
detail["地铁"] = soup.find("div", class_="areaName").find("a", class_="supplement").text
base = soup.find("div", class_ = "base").find_all("li") # 基本信息
detail["户型"] = base[0].text[4:]
detail["面积"] = base[2].text[4:]
detail["朝向"] = base[6].text[4:]
detail["电梯"] = base[10].text[4:]
return detail
else:
return None
if __name__ == "__main__":
Lianjia = LianjiaSpider()
Lianjia.parsePage("https://sz.lianjia.com/ershoufang/")
标签:wan htm img 发送 spider span pre https lock
原文地址:https://www.cnblogs.com/wangshx6/p/9923553.html