import re
import time
import sys

combined_format_re = re.compile(r'''(?P<client>.*?) (?P<ident>.*?) (?P<authuser>.*?) \[(?P<date>.*?)\] "(?P<method>.*?) (?P<url>.*?)(?P<querystr>\?.*?)? (?P<proto>.*?)" (?P<status>\d*) (?P<bytes>.*?) "(?P<referer>.*?)" "(?P<agent>.*?)"''')

class bunch(object):
    pass

month_list = {"Jan": 1,
              "Feb": 2,
              "Mar": 3,
              "Apr": 4,
              "May": 5,
              "Jun": 6,
              "Jul": 7,
              "Aug": 8,
              "Sep": 9,
              "Oct": 10,
              "Nov": 11,
              "Dec": 12 
              }

MAX_LINE_SIZE = 1200

class Parser(object):

    _url_norm_re = re.compile('//+')

    def __init__(self, file_obj):
        self._fd = file_obj

    def __iter__(self):
        fd = self._fd
        url_norm_re = self._url_norm_re
        cursor = bunch()
        cursor._fd = fd
        attrs = '''
        client ident authuser date method proto
        referer agent'''.split()
        while 1:
            line = self._fd.readline(MAX_LINE_SIZE)
            if not line:
                break
            match = combined_format_re.search(line)
            if not match:
                if line.strip():
                    sys.stderr.write('Bad line: %s' % line)
                continue
            for attr in attrs:
                setattr(cursor, attr, match.group(attr))
            cursor.utime = self.make_time(match.group('date'))
            cursor.status = int(match.group('status'))
            b = match.group('bytes')
            if b == '-':
                cursor.bytes = 0
            else:
                cursor.bytes = int(b)
            url = url_norm_re.sub('/', match.group('url'))
            if url.endswith('/index.html'):
                url = url[:-10]
            cursor.url = url
            yield cursor
        
    def make_time(self, date):
        return time.mktime(time.strptime(date, '%d/%b/%Y:%H:%M:%S +0000'))