국회의원 목록 크롤링 (python2) 테스트 코드

assembly_member_list_py2.py

# -*- coding: utf-8 -*-
from __future__ import unicode_literals
import sys
import re
import requests
from bs4 import BeautifulSoup
import time

url = "http://www.assembly.go.kr/assm/memact/congressman/memCond/memCondListAjax.do?currentPage=1&rowPerPage=1000"

resp = requests.get(url)
resp.encoding = 'utf8'
html = resp.text
html = re.sub(r'\s{1,2}[0-9a-f]{3,4}\s{1,2}', '', html)

soup = BeautifulSoup(html, "html.parser")

for member_tag in soup.select('.memberna_list dl dt a'):
    name = member_tag.text
    link = member_tag['href']

    matched = re.search(r'\d+', link)
    if matched:
        member_id = matched.group(0)
    else:
        member_id = None

    print('{} - {}'.format(name, member_id))

print(sys.version)