import urllib
import re
from BeautifulSoup import BeautifulSoup

class ParseError(Exception):
    """Couldn't parse document"""

def makeurl(fromplace, toplace, fromnum = None, tonum = None):
    params = {'keya': fromplace,
              'keyb': toplace,
              'vm': '1',
              'vp': '2',
              'vs': '3',
              'vr': '1',
              'vn': '5'}
    
    if fromnum is not None:
        params['a'] = str(fromnum)
    
    if tonum is not None:
        params['b'] = str(tonum)
    
    return "http://atlas.tripplanner.fi/tkl/fi/?" + urllib.urlencode(params)

def getroutes(html):
    b = BeautifulSoup(html)
    
    for table in b.findAll("table"):
        cells = table.findAll("td")
        if len(cells) >= 2 and "Matkaan" in str(cells[1]):
            break
    else:
        raise ParseError # Didn't find the table
    
    timepat = re.compile("^\s*[0-9]?[0-9][.:][0-9][0-9]\s*$")
    routepat = re.compile("line=([0-9a-zA-Z]{1,3})")
    results = []
    
    for row in table.findAll("tr"):
        columns = row.findAll("td")
        if len(columns) < 7:
            continue
        
        leaves = str(columns[2].find(text = timepat))
        arrives = str(columns[3].find(text = timepat))
        route = routepat.findall(str(columns[6]))
        
        if leaves and arrives and route:
            results.append((leaves, arrives, route))
    
    return results

def cgimain(toplace, tonum = None):
    print "Content-type: text/plain\r\n\r"

    url = makeurl("Opiskelijankatu 4", toplace, tonum = tonum)
    data = urllib.urlopen(url).read()
    
    for leaves, arrives, route in getroutes(data):
        print leaves, arrives, ','.join(route)

if __name__ == "__main__":
    u = makeurl("Opiskelijankatu 4", "Ammattikorkeakoulu")
    d = urllib.urlopen(u).read()
    print getroutes(d)