Fix a few bugs with the new ID cursors.

 - stop once end of collection reached (empty result).
 - don't raise ValueError if ResultSet is empty when
   trying to get max/min ID of the set (max() and min() don't
   accept empty lists)

tweepy/cursor.py

@@ -91,6 +91,8 @@ class IdIterator(BaseIterator):
         # to avoid requesting duplicate items.
         max_id = self.since_id - 1 if self.max_id else None
         data = self.method(max_id = max_id, *self.args, **self.kargs)
+        if len(data) == 0:
+            raise StopIteration
         self.max_id = data.max_id
         self.since_id = data.since_id
         return data
@@ -99,6 +101,8 @@ class IdIterator(BaseIterator):
         """Fetch a set of items with IDs greater than current set."""
         since_id = self.max_id
         data = self.method(since_id = since_id, *self.args, **self.kargs)
+        if len(data) == 0:
+            raise StopIteration
         self.max_id = data.max_id
         self.since_id = data.since_id
         return data

tweepy/models.py

@@ -16,11 +16,17 @@ class ResultSet(list):
 
     @property
     def max_id(self):
-        return self._max_id or max(self.ids())
+        if self._max_id:
+            return self._max_id
+        ids = self.ids()
+        return max(ids) if ids else None
 
     @property
     def since_id(self):
-        return self._since_id or min(self.ids())
+        if self._since_id:
+            return self._since_id
+        ids = self.ids()
+        return min(ids) if ids else None
 
     def ids(self):
         return [item.id for item in self if hasattr(item, 'id')]