distance.py

#!/usr/bin/python
import csv
import sys
import urllib
import BeautifulSoup

def main():
    reader = csv.reader(sys.stdin)
    writer = csv.writer(sys.stdout)
    writer.writerow(reader.next())
    for row in reader:
        if len(row) < 9:
            row += [''] * (9 - len(row))
        if row[7].strip() and row[8].strip():
            writer.writerow(row)
            continue
        #print list(enumerate(row))

        distance, time = get_time_and_distance(row[5], row[6])
        row[7] = distance
        row[8] = time
        writer.writerow(row)

def get_time_and_distance(saddr, daddr):
    url = 'http://maps.google.com/m/directions?ie=UTF8&hl=en&f=d&dirflg=&' +  \
            urllib.urlencode({'saddr': saddr, 'daddr': daddr})
    response = urllib.urlopen(url).read()
    root = BeautifulSoup.BeautifulSoup(response)
    try:
        time = root.find('b').text.strip()
    except:
        time = '#NA'
    try:
        distance = str(root.findAll('p')[2]).split('- about')[0].split('<br />')[1].strip()
    except:
        distance = '#NA'
    return distance, time

if __name__ == '__main__':
    main()