vcs.maemo.org Git - pywienerlinien/blob - gotovienna/routing.py

   1 #!/usr/bin/env python
   2 # -*- coding: UTF-8 -*-
   3
   4 from BeautifulSoup import BeautifulSoup, NavigableString
   5 from urllib2 import urlopen
   6 from urllib import urlencode
   7 from datetime import datetime, time, timedelta
   8 from textwrap import wrap
   9 import argparse
  10 import sys
  11 import os.path
  12 import re
  13
  14 from gotovienna import defaults
  15
  16 POSITION_TYPES = ('stop', 'address', 'poi')
  17 TIMEFORMAT = '%H:%M'
  18 DEBUGLOG = os.path.expanduser('~/gotoVienna.debug')
  19
  20 class ParserError(Exception):
  21
  22     def __init__(self, msg='Parser error'):
  23         self.message = msg
  24
  25 class PageType:
  26     UNKNOWN, CORRECTION, RESULT = range(3)
  27
  28
  29 def extract_city(station):
  30     """ Extract city from string if present,
  31     else return default city
  32
  33     >>> extract_city('Karlsplatz, Wien')
  34     'Wien'
  35     """
  36     if len(station.split(',')) > 1:
  37         return station.split(',')[-1].strip()
  38     else:
  39         return 'Wien'
  40
  41 def extract_station(station):
  42     """ Remove city from string
  43
  44     >>> extract_station('Karlsplatz, Wien')
  45     'Karlsplatz'
  46     """
  47     if len(station.split(',')) > 1:
  48         return station[:station.rindex(',')].strip()
  49     else:
  50         return station
  51
  52 def split_station(station):
  53     """ >>> split_station('Karlsplatz, Wien')
  54     ('Karlsplatz', 'Wien')
  55     >>> split_station('Karlsplatz')
  56     ('Karlsplatz', 'Wien')
  57     """
  58     if len(station.split(',')) > 1:
  59         return (station[:station.rindex(',')].strip(), station.split(',')[-1].strip())
  60     else:
  61         return (station, 'Wien')
  62
  63 def guess_location_type(location):
  64     """Guess type (stop, address, poi) of a location
  65
  66     >>> guess_location_type('pilgramgasse')
  67     'stop'
  68
  69     >>> guess_location_type('karlsplatz 14')
  70     'address'
  71
  72     >>> guess_location_type('reumannplatz 12/34')
  73     'address'
  74     """
  75     parts = location.split()
  76     first_part = parts[0]
  77     last_part = parts[-1]
  78
  79     # Assume all single-word locations are stops
  80     if len(parts) == 1:
  81         return 'stop'
  82
  83     # If the last part is numeric, assume address
  84     if last_part.isdigit() and len(parts) > 1:
  85         return 'address'
  86
  87     # Addresses with door number (e.g. "12/34")
  88     if all(x.isdigit() or x == '/' for x in last_part):
  89         return 'address'
  90
  91     # Sane default - assume it's a stop/station name
  92     return 'stop'
  93
  94 def search(origin_tuple, destination_tuple, dtime=None):
  95     """ build route request
  96     returns html result (as urllib response)
  97     """
  98     if not dtime:
  99         dtime = datetime.now()
 100
 101     origin, origin_type = origin_tuple
 102     origin, origin_city = split_station(origin)
 103
 104     destination, destination_type = destination_tuple
 105     destination, destination_city = split_station(destination)
 106
 107
 108     if origin_type is None:
 109         origin_type = guess_location_type(origin)
 110         print 'Guessed origin type:', origin_type
 111
 112     if destination_type is None:
 113         destination_type = guess_location_type(destination)
 114         print 'Guessed destination type:', destination_type
 115
 116     if (origin_type not in POSITION_TYPES or
 117             destination_type not in POSITION_TYPES):
 118         raise ParserError('Invalid position type')
 119
 120     post = defaults.search_post
 121     post['name_origin'] = origin
 122     post['type_origin'] = origin_type
 123     post['name_destination'] = destination
 124     post['type_destination'] = destination_type
 125     post['itdDateDayMonthYear'] = dtime.strftime('%d.%m.%Y')
 126     post['itdTime'] = dtime.strftime('%H:%M')
 127     post['place_origin'] = origin_city
 128     post['place_destination'] = destination_city
 129     params = urlencode(post)
 130     url = '%s?%s' % (defaults.action, params)
 131
 132     try:
 133         f = open(DEBUGLOG, 'a')
 134         f.write(url + '\n')
 135         f.close()
 136     except:
 137         print 'Unable to write to DEBUGLOG: %s' % DEBUGLOG
 138
 139     return urlopen(url)
 140
 141
 142 class sParser:
 143     """ Parser for search response
 144     """
 145
 146     def __init__(self, html):
 147         self.soup = BeautifulSoup(html)
 148
 149     def check_page(self):
 150         if self.soup.find('form', {'id': 'form_efaresults'}):
 151             return PageType.RESULT
 152
 153         if self.soup.find('div', {'class':'form_error'}):
 154             return PageType.CORRECTION
 155
 156         return PageType.UNKNOWN
 157
 158     state = property(check_page)
 159
 160     def get_correction(self):
 161         names_origin = self.soup.find('select', {'id': 'nameList_origin'})
 162         names_destination = self.soup.find('select', {'id': 'nameList_destination'})
 163         places_origin = self.soup.find('select', {'id': 'placeList_origin'})
 164         places_destination = self.soup.find('select', {'id': 'placeList_destination'})
 165
 166
 167         if names_origin or names_destination or places_origin or places_destination:
 168             dict = {}
 169
 170             if names_origin:
 171                 dict['origin'] = map(lambda x: x.text, names_origin.findAll('option'))
 172             if names_destination:
 173                 dict['destination'] = map(lambda x: x.text, names_destination.findAll('option'))
 174
 175             if places_origin:
 176                 dict['place_origin'] = map(lambda x: x.text, names_origin.findAll('option'))
 177             if names_destination:
 178                 dict['place_destination'] = map(lambda x: x.text, names_destination.findAll('option'))
 179
 180             return dict
 181
 182         else:
 183             raise ParserError('Unable to parse html')
 184
 185     def get_result(self):
 186         return rParser(str(self.soup))
 187
 188
 189
 190 class rParser:
 191     """ Parser for routing results
 192     """
 193
 194     def __init__(self, html):
 195         self.soup = BeautifulSoup(html)
 196         self._overview = None
 197         self._details = None
 198
 199     @classmethod
 200     def get_tdtext(cls, x, cl):
 201             return x.find('td', {'class': cl}).text
 202
 203     @classmethod
 204     def get_change(cls, x):
 205         y = rParser.get_tdtext(x, 'col_change')
 206         if y:
 207             return int(y)
 208         else:
 209             return 0
 210
 211     @classmethod
 212     def get_price(cls, x):
 213         y = rParser.get_tdtext(x, 'col_price')
 214         if y == '*':
 215             return 0.0
 216         if y.find(','):
 217             return float(y.replace(',', '.'))
 218         else:
 219             return 0.0
 220
 221     @classmethod
 222     def get_date(cls, x):
 223         y = rParser.get_tdtext(x, 'col_date')
 224         if y:
 225             return datetime.strptime(y, '%d.%m.%Y').date()
 226         else:
 227             return None
 228
 229     @classmethod
 230     def get_datetime(cls, x):
 231         y = rParser.get_tdtext(x, 'col_time')
 232         if y:
 233             if (y.find("-") > 0):
 234                 # overview mode
 235                 times = map(lambda z: time(*map(int, z.split(':'))), y.split('-'))
 236                 d = rParser.get_date(x)
 237                 from_dtime = datetime.combine(d, times[0])
 238                 if times[0] > times[1]:
 239                     # dateline crossing
 240                     to_dtime = datetime.combine(d + timedelta(1), times[1])
 241                 else:
 242                     to_dtime = datetime.combine(d, times[1])
 243
 244                 return [from_dtime, to_dtime]
 245
 246             else:
 247                 dtregex = {'date' : '\d\d\.\d\d',
 248                            'time': '\d\d:\d\d'}
 249
 250                 regex = "\s*(?P<date1>{date})?\s*(?P<time1>{time})\s*(?P<date2>{date})?\s*(?P<time2>{time})\s*".format(**dtregex)
 251                 ma = re.match(regex, y)
 252
 253                 if not ma:
 254                     return []
 255
 256                 gr = ma.groupdict()
 257
 258                 def extract_datetime(gr, n):
 259                     if 'date%d' % n in gr and gr['date%d' % n]:
 260                         from_dtime = datetime.strptime(str(datetime.today().year) + gr['date%d' % n] + gr['time%d' % n], '%Y%d.%m.%H:%M')
 261                     else:
 262                         t = datetime.strptime(gr['time%d' % n], '%H:%M').time()
 263                         d = datetime.today().date()
 264                         return datetime.combine(d, t)
 265
 266                 # detail mode
 267                 from_dtime = extract_datetime(gr, 1)
 268                 to_dtime = extract_datetime(gr, 2)
 269
 270                 return [from_dtime, to_dtime]
 271
 272         else:
 273             return []
 274
 275     def __iter__(self):
 276         for detail in self.details():
 277             yield detail
 278
 279     def _parse_details(self):
 280         tours = self.soup.findAll('div', {'class': 'data_table tourdetail'})
 281
 282         trips = map(lambda x: map(lambda y: {
 283                         'timespan': rParser.get_datetime(y),
 284                         'station': map(lambda z: z[2:].strip(),
 285                                        filter(lambda x: type(x) == NavigableString, y.find('td', {'class': 'col_station'}).contents)), # filter non NaviStrings
 286                         'info': map(lambda x: x.strip(),
 287                                     filter(lambda z: type(z) == NavigableString, y.find('td', {'class': 'col_info'}).contents)),
 288                     }, x.find('tbody').findAll('tr')),
 289                     tours) # all routes
 290         return trips
 291
 292     @property
 293     def details(self):
 294         """returns list of trip details
 295         [ [ { 'time': [datetime.time, datetime.time] if time else [],
 296               'station': [u'start', u'end'] if station else [],
 297               'info': [u'start station' if station else u'details for walking', u'end station' if station else u'walking duration']
 298             }, ... # next trip step
 299           ], ... # next trip possibility
 300         ]
 301         """
 302         if not self._details:
 303             self._details = self._parse_details()
 304
 305         return self._details
 306
 307     def _parse_overview(self):
 308
 309         # get overview table
 310         table = self.soup.find('table', {'id': 'tbl_fahrten'})
 311
 312         # check if there is an overview table
 313         if table and table.findAll('tr'):
 314             # get rows
 315             rows = table.findAll('tr')[1:] # cut off headline
 316
 317             overview = map(lambda x: {
 318                                'timespan': rParser.get_datetime(x),
 319                                'change': rParser.get_change(x),
 320                                'price': rParser.get_price(x),
 321                            },
 322                            rows)
 323         else:
 324             raise ParserError('Unable to parse overview')
 325
 326         return overview
 327
 328     @property
 329     def overview(self):
 330         """dict containing
 331         date: datetime
 332         time: [time, time]
 333         duration: time
 334         change: int
 335         price: float
 336         """
 337         if not self._overview:
 338             try:
 339                 self._overview = self._parse_overview()
 340             except AttributeError:
 341                 f = open(DEBUGLOG, 'w')
 342                 f.write(str(self.soup))
 343                 f.close()
 344
 345         return self._overview
 346