rewrite transilien to use lxmlparser (closes #271)

weboob/backends/transilien/pages/route.py

@@ -17,9 +17,40 @@
 # You should have received a copy of the GNU Affero General Public License
 # along with weboob. If not, see <http://www.gnu.org/licenses/>.
 
+import datetime
 
-from weboob.tools.browser import BasePage
+from weboob.tools.misc import to_unicode
+from weboob.tools.browser import BasePage, BrokenPageError
+
+class StationNotFound(Exception):
+    pass
 
 class RoutePage(BasePage):
-    def on_loaded(self):
-        return
+    def iter_routes(self):
+        try:
+            table = self.parser.select(self.document.getroot(), 'table.horaires3', 1)
+        except BrokenPageError:
+            raise StationNotFound('Station not found')
+
+        departure = self.parser.select(table, 'td.caption strong', 1).text
+        for tr in table.findall('tr'):
+            if len(tr.findall('td')) != 4:
+                continue
+
+            code_mission = self.parser.select(tr, 'td[headers=Code_de_mission] a', 1).text.strip()
+            time = self.parser.select(tr, 'td[headers=Heure_de_passage]', 1).text.strip()
+            destination = self.parser.select(tr, 'td[headers=Destination]', 1).text.strip()
+            plateform = self.parser.select(tr, 'td[headers=Voie]', 1).text.strip()
+
+            try :
+                time = datetime.datetime.combine(datetime.date.today(), datetime.time(*[int(x) for x in time.split(':')]))
+            except ValueError:
+                self.logger.warning('Unable to parse datetime')
+
+            yield {'type':        to_unicode(code_mission),
+                   'time':        time,
+                   'departure':   to_unicode(departure),
+                   'arrival':     to_unicode(destination),
+                   'late':        datetime.time(),
+                   'late_reason': None,
+                   'plateform':   to_unicode(plateform)}