在网上找了很多关于爬取百度POI的文章，但是对“全量”的做法并没有得到最终的解决方案，自己写了一个，但还是不能实现全量POI抓取，能够达到至少50%的信息抓取。
注意：这里所指“全量”是能够达到100%的POI信息获取。
以下是自己写的代码可直接复制粘贴使用,只针对重庆主城区的。

# -*- coding: utf-8 -*-
#-----------------采集全重庆范围内的POI兴趣点-------------------
import requests
import json
import xlrd
import pandas as pd
import pymysql
import time

def sleeptime(day=1): # 默认参数智能放最后
return day * 24 * 3600

def get_Region():
#'重庆市渝中区',
region = ['重庆市渝中区','重庆市江北区','重庆市渝北区','重庆市沙坪坝区','重庆市九龙坡区',
'重庆市南岸区','重庆市巴南区','重庆市北碚区','重庆市大渡口区']

lng_lat = {'重庆市渝中区':'29.525375,106.494032,29.582745,106.602727', #重庆市渝中区的矩形区域左下角和右上角的经纬度（可参看百度文档对矩形区域的解释）
'重庆市江北区':'29.571878,106.440282,29.689841,106.896587', #左下：106.440282,29.571878
'重庆市渝北区':'29.575192,106.434211,30.051651,106.993128', #右上106.993128,30.051651
'重庆市沙坪坝区':'29.497572,106.261671,29.755899,106.51607', #106.275683,29.497572
'重庆市九龙坡区': '29.248442,106.275683,29.574961,106.553261',
'重庆市南岸区':'29.476615,106.542511,29.614757,106.803882',
'重庆市巴南区':'29.118686,106.520931,29.735054,107.062741',
'重庆市北碚区':'29.66131,106.29192,30.109513,106.720972',
'重庆市大渡口区':'29.344094,106.403174,29.509104,106.532859'}

#'重庆市渝中区':'29.525375,106.494032,29.582745,106.602727', #右上 106.597337,29.578724
# '重庆市江北区':'29.571878,106.440282,29.689841,106.896587', #左下：106.440282,29.571878
# '重庆市渝北区':'29.575192,106.434211,30.051651,106.993128', #右上106.993128,30.051651
# '重庆市沙坪坝区':'29.497572,106.261671,29.755899,106.51607', #106.275683,29.497572
#'重庆市九龙坡区':'29.248442,106.275683,29.574961,106.553261'#, #106.275683,29.248442,106.553261,29.574961
# '重庆市南岸区':'29.476615,106.542511,29.614757,106.803882',
# '重庆市巴南区':'29.118686,106.520931,29.735054,107.062741',
# '重庆市北碚区':'29.66131,106.29192,30.109513,106.720972',
# '重庆市大渡口区':'29.344094,106.403174,29.509104,106.532859'}#各行政区左下右上经纬度
return region,lng_lat

#
# 左下角： [29.547183, 106.545379]
# 右上角: [29.5560735, 106.559716]

def get_Keyword():
#path =r'E:\Project\电子围栏\Data\兴趣点.xlsx'
#data = pd.read_excel(path,sheet_name='keyword2')
data = ['酒店','公司']
return data

def get_ak():#改成你自己的ak
ak = ['xYhFWBgDCF12nPU5v2pMue5uN14d****',
'nryXXcwxdwGws64qmB9jFE29yY6P****',
'tP4RTTuWu0Rds81Hvm0Mn9QXajBs****',
'GP12Wh7S1pM1oa4oNyib7xByol5X****']
return ak

def get_Url(data,region,lng_lat,aks,connect):

url_POI = 'http://api.map.baidu.com/place/v2/search?'
headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2785.143 Safari/537.36'
}
ak_num = 0
ak = aks[ak_num]
payload = {'output': 'json'}
payload['ak'] = ak
payload['city_limit'] = 'true' #区域数据限制（只返回指定区域的信息）
payload['scope'] = '2' # 返回详细信息，1或空为基本信息
payload['page_size'] = '20' #每一页最多返回20个兴趣点
for i in range(len(region)):
payload['region'] = region[i] #指定行政区域名称
trapeze = lng_lat[region[i]].split(',') #指定行政区的左下右上经纬度 trapeze 经纬度格

#以下是将一个行政区（大矩形）划分为多个矩形网格，获取大于400个兴趣点的返回值
left_bottom = [trapeze[0],trapeze[1]] #左下角坐标
right_top = [trapeze[2],trapeze[3]] #右上角坐标
part_n = 20; #网格切割数量，20*20=400 可根据行政区域的大小调整网格数量

x_item = (float(right_top[0]) - float(left_bottom[0])) / part_n; #纬度平均分割大小
print(x_item)
y_item = (float(right_top[1]) - float(left_bottom[1])) / part_n; #经度平均分割大小
print(y_item)
for key in range(len(data['keyword'])):
payload['query'] = data['keyword'][key] #指定查询关键字
#print(data['keyword'][key])
for ii in range(part_n): # 遍历横向网格
for jj in range(part_n): # 遍历纵向网格
#以下left_bottom_part、right_top_part认真理解
left_bottom_part = [float(left_bottom[0]) + ii * x_item, float(left_bottom[1]) + jj * y_item]; #切片的左下角坐标
right_top_part = [float(right_top[0]) - (part_n-1-ii) * x_item, float(right_top[1]) - (part_n-1-jj) * y_item]; # 切片的右上角坐标
for page_num in range(20): #接口最多返回20页
payload['page_num'] = page_num
payload['bounds']=str(left_bottom_part[0]) + ',' + str(left_bottom_part[1]) + ',' + \
str(right_top_part[0])+','+str(right_top_part[1])
#print(payload)
try:
content = requests.get(url_POI,params=payload,headers=headers).json()
#print(content)
if content['status'] == 0: #能够正常返回值
if content['results']:
storeTomysql(content, data['keyword'][key], connect,region[i])#存放数据库
else:
break
else:
if content['status'] == 302: # 如果ak配额用完
print('第' + str(ak_num) + '个ak配额用完，自动调用下一个····')
ak_num = ak_num+1
if ak_num>(len(aks)-1):
print('今天ak已用完，明天再继续吧。。。')
time.sleep(sleeptime()) #程序休眠24小时
ak_num = 0
ak = aks[ak_num]
payload['ak']=ak
content = requests.get(url_POI, params=payload, headers=headers).json()
if content['status']==0:
if content['results']:
storeTomysql(content,data['keyword'][key],connect,region[i])
else:
break #为空则跳出本层循环
else:pass
else:
ak = aks[ak_num]
payload['ak'] = ak
content = requests.get(url_POI, params=payload, headers=headers).json()
if content['status']==0:
if content['results']:
storeTomysql(content,data['keyword'][key],connect,region[i])
else:
break
else:pass
else:
pass #其他错误
# print(content)

except Exception as e:
print(e)

def mysql_Detail():
# 使用的mysql数据库
connect = pymysql.Connect(host='localhost', port=3306, user='root', passwd='更改为你的密码',
db='addr_trans', charset='utf8')
return connect

def storeTomysql(content,Key_word,connect,region):
#在 content 中提取详细信息
# print('test01')
cursor = connect.cursor()
# if content['results']:
for h in range(len(content['results'])): # 遍历返回地点
name = content['results'][h]['name']
lng = content['results'][h]['location']['lng']
lat = content['results'][h]['location']['lat']

if 'area' in content['results'][h].keys():
area = content['results'][h]['area']
else:
area = None
if 'address' in content['results'][h].keys():
address = content['results'][h]['address']
else:
address = None
if 'province' in content['results'][h].keys():
province = content['results'][h]['province']
else:
province = None
if 'city' in content['results'][h].keys():
city = content['results'][h]['city']
else:
city = None
if 'telephone' in content['results'][h].keys():
telephone = content['results'][h]['telephone']
else:
telephone = None
if 'uid' in content['results'][h].keys():
uid = content['results'][h]['uid']
else:
uid = None
if 'street_id' in content['results'][h].keys():
street_id = content['results'][h]['street_id']
else:
street_id = None
if content['results'][h]['detail'] == 1:
if 'type' in content['results'][h]['detail_info'].keys():
type0 = content['results'][h]['detail_info']['type']
else:
type0 = None
if 'tag' in content['results'][h]['detail_info'].keys():
tag = content['results'][h]['detail_info']['tag']
else:
tag = None
else:
type0 = None
tag = None

if area == region[3:]:#区域相符才存数据
try:
sql = "insert into poi_full(keyword,name0,lng,lat,address,province,city,area,telephone,uid,street_id,type0,tag) " \
"VALUES ('%s','%s','%s','%s','%s','%s','%s','%s','%s','%s','%s','%s','%s')"
data0 = (Key_word,name,lng,lat,address,province,city,area,telephone,uid,street_id,type0,tag)
cursor.execute(sql % data0)
connect.commit()
except Exception as e:
print(e)
else:pass
cursor.close()

if __name__ == "__main__":
aks = get_ak()
keyword = get_Keyword()
region,lng_lat = get_Region()
connect = mysql_Detail()
get_Url(keyword, region, lng_lat,aks,connect)
connect.close()

---------------------
作者：Seven_047
来源：CSDN
原文：https://blog.csdn.net/Seven_047/article/details/81476924
版权声明：本文为博主原创文章，转载请附上博文链接！