def test_normalize_simple():
item = ' foo: BAR ,sna '
res = address.normalize_string(item)
eq(res, 'foo bar sna')
def _match_with_place(event, place_ids, places_coll, database, _log=None):
if _log is None:
_log = log
venue = _event_venue(event)
matches = DefaultOrderedDict(OrderedDict)
seen = []
places = itertools.chain(
# Always check nearby places first so that coord information
# is not skipped if the same place is found twice
_nearby_places(venue=venue, event_id=event["_id"], places_coll=places_coll, database=database, _log=_log),
_place_ids_places(place_ids=place_ids, places_coll=places_coll, _log=_log),
_normalized_places(event=event, places_coll=places_coll, _log=_log),
)
for place in places:
place_info = place["info"]
place_id = place_info["_id"]
search_type = place["type"]
if place_id in seen:
continue
seen.append(place_id)
matched = OrderedDict([(item, False) for item in match_fields])
distance = None
venue_name = None
score = 0
if "distance" in place:
distance = place["distance"]
matched["coord"] = True
if venue["address"]:
matched["address"] = streets_equal(venue["address"], place_info["address"]["address"])
if venue["locality"]:
matched["locality"] = cities_equal(venue["locality"], place_info["address"]["locality"])
for name in venue["names"]:
name_ = normalize_string(name)
address_name = normalize_string(place_info["address"]["name"])
match_ = address_name in name_ or name_ in address_name
if match_:
venue_name = name
matched["name"] = match_
break
if venue["region"] and "region" in place_info["address"]:
matched["region"] = states_equal(venue["region"], place_info["address"]["region"])
if venue["country"] and "country" in place_info["address"]:
matched["country"] = countries_equal(venue["country"], place_info["address"]["country"])
if venue["postcode"] and "postcode" in place_info["address"]:
matched["postcode"] = zipcodes_equal(venue["postcode"], place_info["address"]["postcode"])
if place_info["page_ids"] and venue["page_id"]:
if venue["page_id"] in place_info["page_ids"]:
matched["page"] = True
# A score of 100 or more denotes a successful match
# TODO places from place_ids should be probably be
# treated differently since there is a higher chance
# they will be a match
if matched["coord"] and matched["name"]:
score += 100
if matched["coord"] and matched["address"]:
score += 100
if matched["address"] and matched["name"]:
score += 100
if matched["address"] and matched["locality"]:
score += 100
if matched["postcode"] and matched["address"]:
score += 100
if matched["page"]:
score += 100
current = matches[place_id]
current["ubernear"] = OrderedDict(
[
("score", score),
("place_id", place_id),
("source", place_info["source"]),
("location", place_info["coords"]),
("search_type", search_type),
]
)
current["ubernear"]["matched"] = matched
if distance is not None:
current["ubernear"]["distance"] = distance
# Store address information in the match so no places_coll
# look-ups are necessary when serving the event
current["place"] = place_info["address"]
# Some places will match on the address but the name
# of the places are actually different, e.g., two
# businesses with the same address but different unit
# numbers. So, it's safer to use the facebook venue name
if venue_name:
current["place"]["name"] = venue_name
elif venue["names"]:
current["place"]["name"] = venue["names"][0]
highest = _highest_match(matches)
match = None
if highest is not None:
match = matches[highest]
# Store a list of only those parts that have matched
highest_matched = [k for k, v in match["ubernear"]["matched"].iteritems() if v]
match["ubernear"]["matched"] = highest_matched
return match