def find_location_in_string(text):
# try to return one location from a short string
result = None
tokens = [x.lower().strip(".,") for x in text.split()][::-1]
cursor = data.get_database_connection()
select = "SELECT * FROM cities WHERE last_word = ?"
possible_country = None
possible_region = None
city = None
possible_cities = []
for token in tokens:
if token in countries_cache:
possible_country = countries_cache[token][0]['country_code'].lower()
elif token in regions_cache:
possible_region = regions_cache[token][0]['region_code'].lower()
else:
possible_cities.append(token)
for token in possible_cities:
cities = data.get_cities(token, token, possible_country, possible_region)
if len(cities) > 0:
city = sorted(cities.values(), key=lambda x: x['population'])[0]
break
if city is not None:
current_result = {
'found_tokens': [],
}
lat = city['lat']
lon = city['lon']
current_result['found_tokens'].append({
'type': 'CITY',
'lat': lat,
'lon': lon,
'matched_string': text,
})
result = current_result
return result
def city_data():
city_obj = db.get_cities()
for c in city_obj:
try:
c_data = b.getCityData(c.city, c.state_id, c.best_places_url)
except:
print "ERROR could not get data for {}".format(c.city)
continue
for d in c_data:
db.set_cities(
state_id = d['state_id'],
city = d['city'],
population = d['population'],
unemployment = d['unemployment'],
avg_commute = d['avg_commute'],
median_age = d['median_age'],
household_size = d['household_size'],
median_home_price = d['median_home_price'],
msa = d['msa'],
zipcodes = d['zipcodes']
)
print "imported {}".format(c.city)
return
def is_city(text, text_starting_index, previous_result):
# If we're part of a sequence, then use any country or region information to narrow down our search
country_code = None
region_code = None
if previous_result is not None:
found_tokens = previous_result["found_tokens"]
for found_token in found_tokens:
type = found_token["type"]
if type == "COUNTRY":
country_code = found_token["code"]
elif type == "REGION":
region_code = found_token["code"]
current_word = ""
current_index = text_starting_index
pulled_word_count = 0
found_row = None
while pulled_word_count < geodict_config.word_max:
pulled_word, current_index, end_skipped = pull_word_from_end(text, current_index)
pulled_word_count += 1
if current_word == "":
current_word = pulled_word
word_end_index = text_starting_index - end_skipped
name_map = data.get_cities(pulled_word, current_word, country_code, region_code)
# print candidate_rows
if len(name_map) < 1:
break
else:
current_word = pulled_word + " " + current_word
if current_word == "":
return None
if current_word[0:1].islower():
continue
name_key = current_word.lower()
if name_key in name_map:
found_row = name_map[name_key]
if found_row is not None:
break
if current_index < 0:
break
if found_row is None:
return None
if previous_result is None:
current_result = {"found_tokens": []}
else:
current_result = previous_result
lat = found_row["lat"]
lon = found_row["lon"]
current_result["found_tokens"].insert(
0,
{
"type": "CITY",
"lat": lat,
"lon": lon,
"matched_string": current_word,
"start_index": (current_index + 1),
"end_index": word_end_index,
},
)
return current_result
def is_city( text, text_starting_index, previous_result):
# If we're part of a sequence, then use any country or region information to narrow down our search
country_code = None
region_code = None
if previous_result is not None:
found_tokens = previous_result['found_tokens']
for found_token in found_tokens:
type = found_token['type']
if type == 'COUNTRY':
country_code = found_token['code']
elif type == 'REGION':
region_code = found_token['code']
current_word = ''
current_index = text_starting_index
pulled_word_count = 0
found_row = None
while pulled_word_count < geodict_config.word_max:
pulled_word, current_index, end_skipped = pull_word_from_end(text, current_index)
pulled_word_count += 1
if current_word == '':
current_word = pulled_word
word_end_index = (text_starting_index-end_skipped)
name_map = data.get_cities(pulled_word,current_word,country_code,region_code)
#print candidate_rows
if len(name_map) < 1:
break
else:
current_word = pulled_word+' '+current_word
if current_word == '':
return None
if current_word[0:1].islower():
continue
name_key = current_word.lower()
if name_key in name_map:
found_row = name_map[name_key]
if found_row is not None:
break
if current_index < 0:
break
if found_row is None:
return None
if previous_result is None:
current_result = {
'found_tokens': [],
}
else:
current_result = previous_result
lat = found_row['lat']
lon = found_row['lon']
current_result['found_tokens'].insert(0, {
'type': 'CITY',
'lat': lat,
'lon': lon,
'matched_string': current_word,
'start_index': (current_index+1),
'end_index': word_end_index
})
return current_result