From 182d489f306c5fb1b0a16df52a9c851b702f231e Mon Sep 17 00:00:00 2001
From: Michael Smith <michael@hurts.ca>
Date: Tue, 7 Jun 2005 11:42:44 +0000
Subject: [PATCH] mcyar;

---
 Itinerary.py                   | 112 +++++++++++++++++++++++++++++++++++++++--
 tests/test_itinerary_parser.py |  36 +++++++++++++
 2 files changed, 144 insertions(+), 4 deletions(-)
 create mode 100755 tests/test_itinerary_parser.py

diff --git a/Itinerary.py b/Itinerary.py
index e4c737e..e3e5db3 100644
--- a/Itinerary.py
+++ b/Itinerary.py
@@ -4,6 +4,7 @@
 """mcPlan"""
 
 import sre
+from HTMLParser import HTMLParser, HTMLParseError
 
 import Planner
 from PlannerExceptions import *
@@ -14,11 +15,12 @@ class Itinerary:
         self.end = end
         self.html = html
 
-        self._parseIt(html)
+        itinParser = ItineraryParser()
 
-    def _parseIt(self, html):
-        # Extract the HTML chunk for the actual itinerary.
-        itin = self._extractItin(html)
+        # Extract the HTML chunk for the actual itinerary, and feed it to
+        # the sub-parser.
+        itinParser.feed(self._extractItin(html))
+        itinParser.close()
 
     def _extractItin(self, html):
         match = _itin_rx.search(html)
@@ -32,3 +34,105 @@ _itin_re = ('<!-- begin graphical itin.*?>\s*'
                 '(?s)(?P<itin><table.*</table>)\s*'
                 '<!-- end graphical itin')
 _itin_rx = sre.compile(_itin_re)
+
+
+
+class ItineraryParser(HTMLParser):
+    def __init__(self):
+        HTMLParser.__init__(self)
+        self.state = STATE_NOTHING
+
+        # List of ItinEntry
+        self.entries = []
+
+        # Most recent ItinEntry, to which a bus stop number may be added
+        self.lastIE = None
+
+    def handle_starttag(self, tag, attrs):
+        # The directions are contained in <td class="itinText...">.
+        if tag == "td" and attrStarts(attrs, "class", "itinText"):
+
+            # start saving text
+            self.data = ""
+            self.state = STATE_IN_ITIN_TEXT
+
+        # The stop numbers are inside a block like this:
+        # <span class="itinBusStop">O-TRAIN CARLETON S.</span>
+        # <span class="itinBusStop"><a href="#" onClick="MM_openBrWindow('get.stop.timetable.oci?sptDate=2005-06-05&stop560=3062&stopLabel=CG995&stopName=O-TRAIN%2520CARLETON%2520S.','_stopWindow','scrollbars=yes,resizable=yes,width=800,height=600')">(3062)</a></span>
+        #
+        # We'll save the data in the <a> tag.
+        elif self.state == STATE_AFTER_ITIN_TEXT            \
+             and tag == "span" and attrStarts(attrs, "class", "itinBusStop"):
+
+            self.data = ""
+            self.state = STATE_IN_BUS_STOP
+
+        elif self.state == STATE_IN_ITIN_TEXT and tag == "br":
+            self.data += "\n"
+
+    def handle_endtag(self, tag):
+        if self.state == STATE_IN_ITIN_TEXT and tag == "td":
+
+            self.state = STATE_AFTER_ITIN_TEXT
+            self._saveTextEntry()
+
+        elif self.state == STATE_IN_BUS_STOP and tag == "span":
+            if self._saveBusStop():
+                self.state = STATE_NOTHING
+            else:
+                # Usually there are two itinBusStop spans in a row, and
+                # this was probably the first one.
+                self.state = STATE_AFTER_ITIN_TEXT
+
+    def handle_data(self, data):
+        if self.state in (STATE_IN_ITIN_TEXT, STATE_IN_BUS_STOP):
+            self.data += data
+
+    def _saveTextEntry(self):
+        ie = ItinEntry(self.data.strip())
+        self.entries.append(ie)
+        self.lastIE = ie
+
+    def _saveBusStop(self):
+        match = _bus_stop_rx.search(self.data)
+        if match:
+            self.lastIE.busStop = match.group("stopnum")
+        return bool(match)
+
+# ItineraryParser is an FSM.
+STATE_NOTHING = 0
+STATE_IN_ITIN_TEXT = 1
+STATE_AFTER_ITIN_TEXT = 2
+STATE_IN_BUS_STOP = 3
+
+
+def attrStarts(attrs, name, val):
+    """Searches for attribute "name" in the attrs list. If it finds it,
+       returns true if its value starts with "val". Case insensitive."""
+
+    name = name.lower()
+    val = val.lower()
+
+    for kv in attrs:
+        # The attribute name is already lowercased by HTMLParser;
+        # make the value lowercase, too.
+        if kv[0] == name and kv[1].lower().startswith(val):
+            return True
+    return False
+
+_bus_stop_re = '\((?P<stopnum>\d{4})\)'
+_bus_stop_rx = sre.compile(_bus_stop_re)
+
+class ItinEntry:
+    def __init__(self, text):
+        self.text = text
+        self.busStop = None
+
+    def __str__(self):
+        if self.busStop:
+            return "<%s (%s)>" % (self.text, self.busStop)
+        else:
+            return "<%s>" % self.text
+
+    def __repr__(self):
+        return self.__str__()
diff --git a/tests/test_itinerary_parser.py b/tests/test_itinerary_parser.py
new file mode 100755
index 0000000..842906c
--- /dev/null
+++ b/tests/test_itinerary_parser.py
@@ -0,0 +1,36 @@
+#
+# vi: set softtabstop=4 shiftwidth=4 tabstop=8 expandtab:
+
+import unittest
+import Itinerary
+
+class TestItinGrab(unittest.TestCase):
+
+    def setUp(self):
+        fp = open(self.filename, "r")
+        html = fp.read()
+        fp.close()
+
+        itinp = Itinerary.ItineraryParser()
+        itinp.feed(html)
+        itinp.close()
+
+        self.entries = itinp.entries
+
+
+class TestChampagneItin(TestItinGrab):
+
+    filename = "grabs/champagne-itin.html"
+
+    def testEntries(self):
+        print self.entries
+
+class TestLongItin(TestItinGrab):
+
+    filename = "grabs/long-itin.html"
+
+    def testEntries(self):
+        print self.entries
+
+if __name__ == '__main__':
+    unittest.main()
-- 
2.11.4.GIT