2015년 자료 ( 인구 ) 파이썬 크롤링

#______________________________________________
import requests as req
from bs4 import BeautifulSoup
import pandas as pd
from pandas import Series, DataFrame
import pprint as ppr
import matplotlib.pyplot as plt
from matplotlib import font_manager, rc
#______________________________________________
class WebCawling():
    def __init__(self):
        self.urlTarget   = "https://ko.wikipedia.org/wiki/대한민국의_인구순_도시_목록"
        self.htmlNode    = None
        self.bsObject    = None
        self.infor_korea = dict()

        self.barGraph    = dict()
    def servRequest(self):
        self.htmlNode = req.get(url = self.urlTarget)
        if self.htmlNode.status_code == 200:
            self.bsObject = BeautifulSoup(self.htmlNode.text, "html.parser")
            # 행정구역
            qury      = '#mw-content-text > div > table > tbody > tr > td:nth-of-type(2)'
            # 지역
            sub_qury  = '#mw-content-text > div > table > tbody > tr > td:nth-of-type(3)'
            # 인구명
            ssub_qury = '#mw-content-text > div > table > tbody > tr > td:nth-of-type(4)'
            a_tag, b_tag, c_tag = \
                self.bsObject.select(qury), self.bsObject.select(sub_qury), self.bsObject.select(ssub_qury)
            for i in zip(a_tag, b_tag, c_tag):
                local_name = ""
                for j in i[0]:
                    #print (j.string, end='')
                    local_name += j.string
                    # 지역
                    #mw-content-text > div > table > tbody > tr > td:nth-of-type(3)
                man_count = int(str(i[2].string).replace(',',''))
                self.infor_korea[local_name] = {"지역":i[1].string, "인구":man_count}
                #print ("\n==============================")
        #ppr.pprint (self.infor_korea)

    def graph(self):
        font_name = font_manager.FontProperties(fname='C:\\Windows\\Fonts\\malgun.ttf').get_name()
        rc ('font', family = font_name)
        for i in self.infor_korea.values():
            if i['지역'] not in self.barGraph.keys():
                self.barGraph[i['지역']]  = i['인구']
            else:
                self.barGraph[i['지역']] += i['인구']
        ppr.pprint(self.barGraph)

        y_col = [y for y in self.barGraph.values()]
        x_col = [x for x in self.barGraph.keys()]

        plt.bar(x_col, y_col)
        plt.show()
#_____________________________

def main():
    # 객체 생성
    webInstance = WebCawling()
    webInstance.servRequest()
    webInstance.graph()
if __name__ == "__main__":
    main()