elections2014

elections2014.py

import os
import json

import requests
from lxml import html


def num_to_state(n):
    assert n <= 35
    if n < 10:
        return 'S0{}'.format(n)
    if n <= 28:
        return 'S{}'.format(n)
    return 'U0{}'.format(n-28)


def state_const_map():
    e = html.parse('http://eciresults.nic.in/ConstituencywiseS011.htm?ac=1')
    cts = e.xpath("//input[@type='hidden']/@value")
    states = {s.attrib['value']: s.text for s in 
              e.xpath("//select[@name='ctl00$ContentPlaceHolder1$ACWise1$ddlState']/option")[1:]}    
    sc_map = ((num_to_state(i+1), dict([tuple(x.strip().strip(';').split(','))
                                        for x in c.split('\r\n')]))
              for i, c in enumerate(cts))
    sc_map = {states[s]: {'code': s, 'constituencies': cs}
              for s, cs in sc_map}
    filepath = '/home/vineet/elections2014/state-const-map2.json'
    with open(filepath, 'w') as f:
        json.dump(sc_map, f, indent=4)
    return sc_map


def ensure_dir(p):
    if not os.path.exists(p):
        os.makedirs(p)
    return p


def constituency_wise():
    dirpath = ensure_dir('/home/vineet/elections2014/constituency-wise')
    state_const_map = json.load(open('/home/vineet/elections2014/state-const-map2.json'))
    for k, v in state_const_map.iteritems():
        scode = v['code']
        for c, n in v['constituencies'].iteritems():
            url = 'http://eciresults.nic.in/Constituencywise{}{}.htm?ac={}'.format(scode, c, c)
            r = requests.get(url)
            if r.status_code == 200:
                p = ensure_dir(os.path.join(dirpath, k))
                with open(os.path.join(p, '{}.html'.format(n)), 'w') as f:
                    f.write(r.content)


def write_html(filepath, content):
    with open(filepath, 'w') as f:
        f.write(content)


if __name__ == '__main__':
    # state_const_map()
    constituency_wise()