爬取70城房价到oracle数据库并6合1

学习数据分析，然后没有合适的数据源，从国家统计局的网页上抓取一页数据来玩玩（没有发现robots协议，也仅仅发出一次连接请求，不对网站造成任何负荷）

运行效果

源码

python代码

'''
本脚本旨在爬取70城房价进入oracle数据库以供学习
code by 九命猫幺

网页中有6个表格

最终爬取到数据库中形成6合1报表
'''
import requests
from bs4 import BeautifulSoup
import numpy as np
import pandas as pd
from sqlalchemy import create_engine

#爬取网页
def getHTMLText(url):
    try:
        headers={'User-Agent':'Baiduspider'}
        r = requests.get(url,headers=headers,timeout=30)
        r.raise_for_status()
        r.encoding = r.apparent_encoding
        return r.text
    except:
        return '产生异常'

#解析出列表
def getTrText(tbody,tnum):
    uinfo1 = []
    uinfo2 = []
    for i in tbody.strings:
        if i != ' ':
            uinfo1.append(str(i.string).replace('\u3000','').replace(' ',''))
    for i in uinfo1:
        if i not in ['皇','岛', '家','庄','丹','江','尔','滨','顶','山']:
            uinfo2.append(i.replace('秦','秦皇岛').replace('石','石家庄').replace('牡','牡丹江').replace('哈','哈尔滨').replace('平','平顶山'))
    uinfo2 = uinfo2[{1:-280,2:-280,3:-350,4:-350,5:-350,6:-350}[tnum]::]
    return uinfo2

#将解析出的列表加工转换传入oracle库
def toSql(uinfo,tnum):
    if tnum in [1,2]:
        df = pd.DataFrame(np.array(uinfo).reshape(70,4),columns=['city','mom','yoy','fbr'])
    else:
        df = pd.DataFrame(np.array(uinfo).reshape(35,10),columns=['city','mom_90l','yoy_90l','fbr_90l','mom_90t144','yoy_90t144','fbr_90t144','mom_144u','yoy_144u','fbr_144u'])
    con = create_engine('oracle+cx_oracle://edw:oracle@192.168.168.5:1521/?service_name=edw')
    df.to_sql('tb_fj_70city_t'+str(tnum),con,if_exists='replace',index=False)

if __name__ == "__main__":
    uinfo = []
    url = 'http://www.stats.gov.cn/tjsj/zxfb/201911/t20191115_1709560.html'

    #爬网页
    html = getHTMLText(url)
    soup = BeautifulSoup(html,'html.parser')
    tbody = soup.select('table.MsoNormalTable tbody')
    #解析存储
    for i in range(6):
        #解析表
        uinfo = getTrText(tbody[i],i+1)
        #存表入数据库
        toSql(uinfo,i+1)

数据库代码

--70个大中城市商品住宅销售价格变动情况
CREATE TABLE tb_fj_70city_201910 AS
WITH tmp1 AS(
SELECT to_char(a.city) city,to_number(a.mom) new_mom,to_number(a.yoy) new_yoy,to_number(a.fbr) new_fbr
FROM tb_fj_70city_t1 a),
tmp2 AS(
SELECT to_char(a.city) city,to_number(a.mom) old_mom,to_number(a.yoy) old_yoy,to_number(a.fbr) old_fbr
FROM tb_fj_70city_t2 a),
tmp3 AS(
SELECT to_char(a.city) city,to_number(a.mom_90l) new_mom_90l,to_number(a.yoy_90l) new_yoy_90l,to_number(a.fbr_90l) new_fbr_90l,
to_number(a.mom_90t144) new_mom_90t144,to_number(a.yoy_90t144) new_yoy_90t144,to_number(a.fbr_90t144) new_fbr_90t144,
to_number(a.mom_144u) new_mom_144u,to_number(a.yoy_144u) new_yoy_144u,to_number(a.fbr_144u) new_fbr_144u
FROM tb_fj_70city_t3 a
UNION
SELECT to_char(a.city) city,to_number(a.mom_90l) new_mom_90l,to_number(a.yoy_90l) new_yoy_90l,to_number(a.fbr_90l) new_fbr_90l,
to_number(a.mom_90t144) new_mom_90t144,to_number(a.yoy_90t144) new_yoy_90t144,to_number(a.fbr_90t144) new_fbr_90t144,
to_number(a.mom_144u) new_mom_144u,to_number(a.yoy_144u) new_yoy_144u,to_number(a.fbr_144u) new_fbr_144u
FROM tb_fj_70city_t4 a),
tmp4 AS(
SELECT to_char(a.city) city,to_number(a.mom_90l) old_mom_90l,to_number(a.yoy_90l) old_yoy_90l,to_number(a.fbr_90l) old_fbr_90l,
to_number(a.mom_90t144) old_mom_90t144,to_number(a.yoy_90t144) old_yoy_90t144,to_number(a.fbr_90t144) old_fbr_90t144,
to_number(a.mom_144u) old_mom_144u,to_number(a.yoy_144u) old_yoy_144u,to_number(a.fbr_144u) old_fbr_144u
FROM tb_fj_70city_t5 a
UNION
SELECT to_char(a.city) city,to_number(a.mom_90l) old_mom_90l,to_number(a.yoy_90l) old_yoy_90l,to_number(a.fbr_90l) old_fbr_90l,
to_number(a.mom_90t144) old_mom_90t144,to_number(a.yoy_90t144) old_yoy_90t144,to_number(a.fbr_90t144) old_fbr_90t144,
to_number(a.mom_144u) old_mom_144u,to_number(a.yoy_144u) old_yoy_144u,to_number(a.fbr_144u) old_fbr_144u
FROM tb_fj_70city_t6 a)
SELECT 201910 month,aa.city,aa.new_mom,aa.new_yoy,aa.new_fbr,bb. old_mom,bb.old_yoy,bb.old_fbr,
cc.new_mom_90l,cc.new_yoy_90l,cc.new_fbr_90l,
cc.new_mom_90t144,cc.new_yoy_90t144,cc.new_fbr_90t144,
cc.new_mom_144u,cc.new_yoy_144u,cc.new_fbr_144u,
dd.old_mom_90l,dd.old_yoy_90l,dd.old_fbr_90l,
dd.old_mom_90t144,dd.old_yoy_90t144,dd.old_fbr_90t144,
dd.old_mom_144u,dd.old_yoy_144u,dd.old_fbr_144u
FROM tmp1 aa
JOIN tmp2 bb ON aa.city=bb.city
JOIN tmp3 cc ON aa.city=cc.city
JOIN tmp4 dd ON aa.city=dd.city;

CALL p_drop_table_if_exist('tb_fj_70city_t1');
CALL p_drop_table_if_exist('tb_fj_70city_t2');
CALL p_drop_table_if_exist('tb_fj_70city_t3');
CALL p_drop_table_if_exist('tb_fj_70city_t4');
CALL p_drop_table_if_exist('tb_fj_70city_t5');
CALL p_drop_table_if_exist('tb_fj_70city_t6');

SELECT * FROM tb_fj_70city_201910;

就这样，表名中列名，取英文首字母：

mom:month on month ，环比

yoy:year on year,同比

fbr:fixed base ratio,定基比

90l:90 lower,90平米以下

144u:144 upper,144平米以上

90t144:90 to 144,90到144平米之间

优化后

上述脚本只能爬取一个月的，并且6表合1操作在数据库中执行，现在优化为批量爬取多个月份的数据

'''
本脚本旨在爬取70城房价进入oracle数据库以供学习
code by 九命猫幺

网页中有6个表格

最终爬取到数据库中形成6合1报表

网址：
'''
import requests
from bs4 import BeautifulSoup
import numpy as np
import pandas as pd
from sqlalchemy import create_engine
import cx_Oracle

#爬取网页
def getHTMLText(url):
    try:
        headers={'User-Agent':'Baiduspider'}
        r = requests.get(url,headers=headers,timeout=30)
        r.raise_for_status()
        r.encoding = r.apparent_encoding
        return r.text
    except:
        return '产生异常'

#解析出列表
def getTrText(tbody,tnum):
    uinfo1 = []
    uinfo2 = []
    for i in tbody.strings:
        if i != ' ':
            uinfo1.append(str(i.string).replace('\u3000','').replace(' ',''))
    for i in uinfo1:
        if i not in ['皇','岛', '家','庄','丹','江','尔','滨','顶','山']:
            uinfo2.append(i.replace('秦','秦皇岛').replace('石','石家庄').replace('牡','牡丹江').replace('哈','哈尔滨').replace('平','平顶山'))
    uinfo2 = uinfo2[{1:-280,2:-280,3:-350,4:-350,5:-350,6:-350}[tnum]::]
    return uinfo2

#将解析出的列表加工转换传入oracle库
def toSql(uinfo,tnum):
    if tnum in [1,2]:
        df = pd.DataFrame(np.array(uinfo).reshape(70,4),columns=['city','mom','yoy','fbr'])
    else:
        df = pd.DataFrame(np.array(uinfo).reshape(35,10),columns=['city','mom_90l','yoy_90l','fbr_90l','mom_90t144','yoy_90t144','fbr_90t144','mom_144u','yoy_144u','fbr_144u'])
    con = create_engine('oracle+cx_oracle://edw:oracle@192.168.168.5:1521/?service_name=edw')
    df.to_sql('tb_fj_70city_t'+str(tnum),con,if_exists='replace',index=False)

#6合1 并插入历史宽表
def intoWideTable(month):
    con = cx_Oracle.connect('edw','oracle','192.168.168.5:1521/edw')
    cur = con.cursor()
    cur.execute("CALL p_drop_table_if_exist('tb_fj_70city_"+str(month)+"')")
    cur.execute('''CREATE TABLE tb_fj_70city_'''+str(month)+''' AS
WITH tmp1 AS(
SELECT to_char(a.city) city,to_number(a.mom) new_mom,to_number(a.yoy) new_yoy,to_number(a.fbr) new_fbr
FROM tb_fj_70city_t1 a),
tmp2 AS(
SELECT to_char(a.city) city,to_number(a.mom) old_mom,to_number(a.yoy) old_yoy,to_number(a.fbr) old_fbr
FROM tb_fj_70city_t2 a),
tmp3 AS(
SELECT to_char(a.city) city,to_number(a.mom_90l) new_mom_90l,to_number(a.yoy_90l) new_yoy_90l,to_number(a.fbr_90l) new_fbr_90l,
to_number(a.mom_90t144) new_mom_90t144,to_number(a.yoy_90t144) new_yoy_90t144,to_number(a.fbr_90t144) new_fbr_90t144,
to_number(a.mom_144u) new_mom_144u,to_number(a.yoy_144u) new_yoy_144u,to_number(a.fbr_144u) new_fbr_144u
FROM tb_fj_70city_t3 a
UNION
SELECT to_char(a.city) city,to_number(a.mom_90l) new_mom_90l,to_number(a.yoy_90l) new_yoy_90l,to_number(a.fbr_90l) new_fbr_90l,
to_number(a.mom_90t144) new_mom_90t144,to_number(a.yoy_90t144) new_yoy_90t144,to_number(a.fbr_90t144) new_fbr_90t144,
to_number(a.mom_144u) new_mom_144u,to_number(a.yoy_144u) new_yoy_144u,to_number(a.fbr_144u) new_fbr_144u
FROM tb_fj_70city_t4 a),
tmp4 AS(
SELECT to_char(a.city) city,to_number(a.mom_90l) old_mom_90l,to_number(a.yoy_90l) old_yoy_90l,to_number(a.fbr_90l) old_fbr_90l,
to_number(a.mom_90t144) old_mom_90t144,to_number(a.yoy_90t144) old_yoy_90t144,to_number(a.fbr_90t144) old_fbr_90t144,
to_number(a.mom_144u) old_mom_144u,to_number(a.yoy_144u) old_yoy_144u,to_number(a.fbr_144u) old_fbr_144u
FROM tb_fj_70city_t5 a
UNION
SELECT to_char(a.city) city,to_number(a.mom_90l) old_mom_90l,to_number(a.yoy_90l) old_yoy_90l,to_number(a.fbr_90l) old_fbr_90l,
to_number(a.mom_90t144) old_mom_90t144,to_number(a.yoy_90t144) old_yoy_90t144,to_number(a.fbr_90t144) old_fbr_90t144,
to_number(a.mom_144u) old_mom_144u,to_number(a.yoy_144u) old_yoy_144u,to_number(a.fbr_144u) old_fbr_144u
FROM tb_fj_70city_t6 a)
SELECT '''+str(month)+''' month,aa.city,aa.new_mom,aa.new_yoy,aa.new_fbr,bb. old_mom,bb.old_yoy,bb.old_fbr,
cc.new_mom_90l,cc.new_yoy_90l,cc.new_fbr_90l,
cc.new_mom_90t144,cc.new_yoy_90t144,cc.new_fbr_90t144,
cc.new_mom_144u,cc.new_yoy_144u,cc.new_fbr_144u,
dd.old_mom_90l,dd.old_yoy_90l,dd.old_fbr_90l,
dd.old_mom_90t144,dd.old_yoy_90t144,dd.old_fbr_90t144,
dd.old_mom_144u,dd.old_yoy_144u,dd.old_fbr_144u
FROM tmp1 aa
JOIN tmp2 bb ON aa.city=bb.city
JOIN tmp3 cc ON aa.city=cc.city
JOIN tmp4 dd ON aa.city=dd.city''')
    cur.close()
    con.close()

if __name__ == "__main__":
    uinfo = []
    urls = {201910:'http://www.stats.gov.cn/tjsj/zxfb/201911/t20191115_1709560.html',
            201909:'http://www.stats.gov.cn/tjsj/zxfb/201910/t20191021_1704063.html',
            201908:'http://www.stats.gov.cn/tjsj/zxfb/201909/t20190917_1697943.html',
            201907:'http://www.stats.gov.cn/statsinfo/auto2074/201908/t20190815_1691536.html',
            201906:'http://www.stats.gov.cn/tjsj/zxfb/201907/t20190715_1676000.html',
            201905:'http://www.stats.gov.cn/tjsj/zxfb/201906/t20190618_1670960.html',
            201904:'http://www.stats.gov.cn/tjsj/zxfb/201905/t20190516_1665286.html',
            201903:'http://www.stats.gov.cn/tjsj/zxfb/201904/t20190416_1659682.html'
            }
    for key in urls:
    #爬网页
        html = getHTMLText(urls[key])
        soup = BeautifulSoup(html,'html.parser')
        tbody = soup.select('table.MsoNormalTable tbody')
        #解析存储
        for i in range(6):
            #解析表
            uinfo = getTrText(tbody[i],i+1)
            #存表入数据库
            toSql(uinfo,i+1)
        #存入宽表
        intoWideTable(key)

数据库中同时得到了多个月份的

再优化单一月份爬取的代码

import requests
from bs4 import BeautifulSoup
import numpy as np
import pandas as pd
from sqlalchemy import create_engine
import cx_Oracle

#爬取网页
def getHTMLText(url):
    try:
        headers={'User-Agent':'Baiduspider'}
        r = requests.get(url,headers=headers,timeout=30)
        r.raise_for_status()
        r.encoding = r.apparent_encoding
        return r.text
    except:
        return '产生异常'
#解析出列表
def getTrText(tbody,tnum):
    uinfo1 = []
    uinfo2 = ['...']
    for i in tbody.strings:
        if i not in ['　　',' ']:
            uinfo1.append(str(i.string).replace('　　',''))
    for i in uinfo1:
        if '\u4e00'

ok了