From 142bc04f56a2bf85113d72f10fa837106bd00b9d Mon Sep 17 00:00:00 2001 From: Romain Bignon Date: Tue, 28 Jun 2011 17:09:39 +0200 Subject: [PATCH] support roadmaps in transilien (#681) --- weboob/applications/traveloob/traveloob.py | 7 ++ weboob/backends/transilien/backend.py | 33 +++++--- weboob/backends/transilien/browser.py | 37 +++++++-- .../pages/{route.py => departures.py} | 6 +- weboob/backends/transilien/pages/roadmap.py | 81 +++++++++++++++++++ 5 files changed, 147 insertions(+), 17 deletions(-) rename weboob/backends/transilien/pages/{route.py => departures.py} (96%) create mode 100644 weboob/backends/transilien/pages/roadmap.py diff --git a/weboob/applications/traveloob/traveloob.py b/weboob/applications/traveloob/traveloob.py index b02df326..5b0c9ed1 100644 --- a/weboob/applications/traveloob/traveloob.py +++ b/weboob/applications/traveloob/traveloob.py @@ -72,3 +72,10 @@ class Traveloob(ReplApplication): for backend, departure in self.do('iter_station_departures', station_id, arrival_id, backends=backends): self.format(departure) self.flush() + + def do_roadmap(self, line): + departure, arrival = self.parse_command_args(line, 2, 2) + + for backend, route in self.do('iter_roadmap', departure, arrival): + self.format(route) + self.flush() diff --git a/weboob/backends/transilien/backend.py b/weboob/backends/transilien/backend.py index b25af715..b75aebac 100644 --- a/weboob/backends/transilien/backend.py +++ b/weboob/backends/transilien/backend.py @@ -18,7 +18,7 @@ # along with weboob. If not, see . -from weboob.capabilities.travel import ICapTravel, Station, Departure +from weboob.capabilities.travel import ICapTravel, Station, Departure, RoadStep from weboob.tools.backend import BaseBackend from .browser import Transilien @@ -40,11 +40,26 @@ class TransilienBackend(BaseBackend, ICapTravel): yield Station(_id, name) def iter_station_departures(self, station_id, arrival_id=None): - for i, d in enumerate(self.browser.iter_station_departures(station_id, arrival_id)): - departure = Departure(i, d['type'], d['time']) - departure.departure_station = d['departure'] - departure.arrival_station = d['arrival'] - departure.late = d['late'] - departure.information = d['late_reason'] - departure.plateform = d['plateform'] - yield departure + with self.browser: + for i, d in enumerate(self.browser.iter_station_departures(station_id, arrival_id)): + departure = Departure(i, d['type'], d['time']) + departure.departure_station = d['departure'] + departure.arrival_station = d['arrival'] + departure.late = d['late'] + departure.information = d['late_reason'] + departure.plateform = d['plateform'] + yield departure + + def iter_roadmap(self, departure, arrival): + with self.browser: + roadmap = self.browser.get_roadmap(departure, arrival) + + for s in roadmap['steps']: + step = RoadStep(s['id']) + step.line = s['line'] + step.start_time = s['start_time'] + step.end_time = s['end_time'] + step.departure = s['departure'] + step.arrival = s['arrival'] + step.duration = s['duration'] + yield step diff --git a/weboob/backends/transilien/browser.py b/weboob/backends/transilien/browser.py index 6ec1dfcd..47905895 100644 --- a/weboob/backends/transilien/browser.py +++ b/weboob/backends/transilien/browser.py @@ -18,18 +18,31 @@ # along with weboob. If not, see . -from weboob.tools.browser import BaseBrowser +from weboob.tools.browser import BaseBrowser, BasePage, BrowserUnavailable -from .pages.route import RoutePage +from .pages.departures import DeparturesPage +from .pages.roadmap import RoadmapSearchPage, RoadmapConfirmPage, RoadmapPage + +class UnavailablePage(BasePage): + def on_loaded(self): + raise BrowserUnavailable('Website is currently unavailable') class Transilien(BaseBrowser): DOMAIN = 'www.transilien.com' PROTOCOL = 'https' USER_AGENT = BaseBrowser.USER_AGENTS['microb'] - PAGES = {'https://www\.transilien\.com/web/ITProchainsTrainsAvecDest\.do\?.*': RoutePage, - 'https://www\.transilien\.com/web/ITProchainsTrains\.do\?.*': RoutePage, + PAGES = {'https://www\.transilien\.com/web/ITProchainsTrainsAvecDest\.do\?.*': DeparturesPage, + 'https://www\.transilien\.com/web/ITProchainsTrains\.do\?.*': DeparturesPage, + 'https://www\.transilien\.com/web/site.*': RoadmapSearchPage, + 'https://www\.transilien\.com/web/RedirectHI.do.*': RoadmapConfirmPage, + 'https://www\.transilien\.com/web/RedirectHIIntermediaire.do.*': RoadmapPage, + 'https://www\.transilien\.com/transilien_sncf_maintenance_en_cours.htm': UnavailablePage, } + def is_logged(self): + """ Do not need to be logged """ + return True + def iter_station_search(self, pattern): pass @@ -41,6 +54,16 @@ class Transilien(BaseBrowser): return self.page.iter_routes() - def is_logged(self): - """ Do not need to be logged """ - return True + def get_roadmap(self, departure, arrival): + self.location('/web/site/accueil/etat-trafic/chercher-itineraire/lang/en') + + assert self.is_on_page(RoadmapSearchPage) + self.page.search(departure, arrival) + + assert self.is_on_page(RoadmapConfirmPage) + self.page.confirm() + + assert self.is_on_page(RoadmapPage) + roadmap = {} + roadmap['steps'] = list(self.page.get_steps()) + return roadmap diff --git a/weboob/backends/transilien/pages/route.py b/weboob/backends/transilien/pages/departures.py similarity index 96% rename from weboob/backends/transilien/pages/route.py rename to weboob/backends/transilien/pages/departures.py index f069c688..aa26d008 100644 --- a/weboob/backends/transilien/pages/route.py +++ b/weboob/backends/transilien/pages/departures.py @@ -22,10 +22,14 @@ import datetime from weboob.tools.misc import to_unicode from weboob.tools.browser import BasePage, BrokenPageError + +__all__ = ['StationNotFound', 'DeparturesPage'] + + class StationNotFound(Exception): pass -class RoutePage(BasePage): +class DeparturesPage(BasePage): def iter_routes(self): try: table = self.parser.select(self.document.getroot(), 'table.horaires3', 1) diff --git a/weboob/backends/transilien/pages/roadmap.py b/weboob/backends/transilien/pages/roadmap.py new file mode 100644 index 00000000..18da4a7a --- /dev/null +++ b/weboob/backends/transilien/pages/roadmap.py @@ -0,0 +1,81 @@ +# -*- coding: utf-8 -*- + +# Copyright(C) 2010-2011 Romain Bignon +# +# This file is part of weboob. +# +# weboob is free software: you can redistribute it and/or modify +# it under the terms of the GNU Affero General Public License as published by +# the Free Software Foundation, either version 3 of the License, or +# (at your option) any later version. +# +# weboob is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU Affero General Public License for more details. +# +# You should have received a copy of the GNU Affero General Public License +# along with weboob. If not, see . + + +import re +import datetime + +from weboob.tools.browser import BasePage +from weboob.tools.misc import to_unicode + + +__all__ = ['RoadmapPage'] + + +class RoadmapSearchPage(BasePage): + def search(self, departure, arrival): + self.browser.select_form('formHiRecherche') + self.browser['lieuDepart'] = departure.encode('utf-8') + self.browser['lieuArrivee'] = arrival.encode('utf-8') + self.browser.submit() + +class RoadmapConfirmPage(BasePage): + def select(self, name, num): + try: + self.browser[name] = str(num) + except TypeError: + self.browser[name] = [str(num)] + + def confirm(self): + self.browser.select_form('form1') + self.browser.set_all_readonly(False) + self.select('idDepart', 1) + self.select('idArrivee', 1) + self.browser.submit() + +class RoadmapPage(BasePage): + def get_steps(self): + current_step = None + for tr in self.parser.select(self.document.getroot(), 'table.horaires2 tbody tr'): + if not 'class' in tr.attrib: + continue + elif tr.attrib['class'] == 'trHautTroncon': + current_step = {} + current_step['id'] = 0 + current_step['start_time'] = self.parse_time(self.parser.select(tr, 'td.formattedHeureDepart p', 1).text.strip()) + current_step['line'] = self.parser.select(tr, 'td.rechercheResultatColumnMode img')[-1].attrib['alt'] + current_step['departure'] = to_unicode(self.parser.select(tr, 'td.descDepart p strong', 1).text.strip()) + current_step['duration'] = self.parse_duration(self.parser.select(tr, 'td.rechercheResultatVertAlign', 1).text.strip()) + elif tr.attrib['class'] == 'trBasTroncon': + current_step['end_time'] = self.parse_time(self.parser.select(tr, 'td.formattedHeureArrivee p', 1).text.strip()) + current_step['arrival'] = to_unicode(self.parser.select(tr, 'td.descArrivee p strong', 1).text.strip()) + yield current_step + + def parse_time(self, time): + h, m = time.split('h') + return datetime.time(int(h), int(m)) + + def parse_duration(self, dur): + m = re.match('(\d+)min.', dur) + if m: + return datetime.timedelta(minutes=int(m.group(1))) + m = re.match('(\d+)h(\d+)', dur) + if m: + return datetime.timedelta(hours=int(m.group(1)), + minutes=int(m.group(2)))