def audit_state(db, collection):
states = db_ops.get_values(db, collection, ['address.state'])
# One Document had DE as state. validation showed that Schleswig-Holstein is the correct value
#Be careful if using different data sets
cor_state = {'manual':{'DE':'Schleswig-Holstein', 'NI':'Niedersachsen', 'Lower Saxony':'Niedersachsen', 'HH':'Hamburg'}}
db_ops.update_db(db, collection, 'address.state', cor_state)
states = db_ops.get_values(db, collection, ['address.state'])
for state in states:
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates', 'address.state', state)
def audit_postcode(db, collection):
pcs = db_ops.get_values(db, collection, ['address.postcode'])
for pc in pcs:
#One given postcode did not exist:
if pc == '22701':
pc = '22765' #found at google maps
db[collection].update({'address.postcode':'22701'},{'$set':{'address.postcode':pc}})
#postcodes and states were crossvalidated with the google maps API
url = 'http://maps.googleapis.com/maps/api/geocode/json?address=urlencode($_REQUEST[' + pc + ' Germany])&sensor=false'
attempts = 0
success = False
while success != True and attempts < 3:
page = urllib.urlopen(url)
data = json.loads(page.read())
attempts += 1
if data['status'] == "OVER_QUERY_LIMIT":
time.sleep(2)
continue
success = True
if attempts == 3:
print("Daily limit has been reached")
for i in range(len(data['results'][0]['address_components'])):
if data['results'][0]['address_components'][i]['long_name'] in ['Schleswig-Holstein', 'Niedersachsen','Hamburg', 'Lower Saxony']:
state = data['results'][0]['address_components'][i]['long_name']
if state == 'Lower Saxony':
state = 'Niedersachsen'
#Set the state field, if not already given
if db[collection].find({'address.postcode':pc, 'address.state' : {'$exists' : 'false'}}) != None:
db[collection].update({'address.postcode':pc},{'$set':{'address.state':state}})
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates', 'address.postcode', pc)
def audit_state(db, collection):
states = db_ops.get_values(db, collection, ['address.state'])
# One Document had DE as state. validation showed that Schleswig-Holstein is the correct value
#Be careful if using different data sets
cor_state = {
'manual': {
'DE': 'Schleswig-Holstein',
'NI': 'Niedersachsen',
'Lower Saxony': 'Niedersachsen',
'HH': 'Hamburg'
}
}
db_ops.update_db(db, collection, 'address.state', cor_state)
states = db_ops.get_values(db, collection, ['address.state'])
for state in states:
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates',
'address.state', state)
def audit_streets(db, collection):
streets = db_ops.get_values(db, collection, ['address.street'])
#some streets contain possible problem characters, but most are part of the Name
#The others will be updated here
streets_to_clean = {'1': {'Kandinskyallee, Ecke Feiningerstraße':'Kandinskyallee',
'I. Sandereiweg':'Erster Sandereiweg',
'II. Sandereiweg':'Zweiter Sandereiweg',
'III. Sandereiweg':'Dritter Sandereiweg'}}
db_ops.update_db(db, collection, 'address.street', streets_to_clean)
for street in streets:
t = re.compile('[s|S]tr\.*$').search(street)
if t != None:
print(street)
def audit_postcode(db, collection):
pcs = db_ops.get_values(db, collection, ['address.postcode'])
for pc in pcs:
#One given postcode did not exist:
if pc == '22701':
pc = '22765' #found at google maps
db[collection].update({'address.postcode': '22701'},
{'$set': {
'address.postcode': pc
}})
#postcodes and states were crossvalidated with the google maps API
url = 'http://maps.googleapis.com/maps/api/geocode/json?address=urlencode($_REQUEST[' + pc + ' Germany])&sensor=false'
attempts = 0
success = False
while success != True and attempts < 3:
page = urllib.urlopen(url)
data = json.loads(page.read())
attempts += 1
if data['status'] == "OVER_QUERY_LIMIT":
time.sleep(2)
continue
success = True
if attempts == 3:
print("Daily limit has been reached")
for i in range(len(data['results'][0]['address_components'])):
if data['results'][0]['address_components'][i]['long_name'] in [
'Schleswig-Holstein', 'Niedersachsen', 'Hamburg',
'Lower Saxony'
]:
state = data['results'][0]['address_components'][i][
'long_name']
if state == 'Lower Saxony':
state = 'Niedersachsen'
#Set the state field, if not already given
if db[collection].find({
'address.postcode': pc,
'address.state': {
'$exists': 'false'
}
}) != None:
db[collection].update({'address.postcode': pc},
{'$set': {
'address.state': state
}})
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates',
'address.postcode', pc)
def audit_streets(db, collection):
streets = db_ops.get_values(db, collection, ['address.street'])
#some streets contain possible problem characters, but most are part of the Name
#The others will be updated here
streets_to_clean = {
'1': {
'Kandinskyallee, Ecke Feiningerstraße': 'Kandinskyallee',
'I. Sandereiweg': 'Erster Sandereiweg',
'II. Sandereiweg': 'Zweiter Sandereiweg',
'III. Sandereiweg': 'Dritter Sandereiweg'
}
}
db_ops.update_db(db, collection, 'address.street', streets_to_clean)
for street in streets:
t = re.compile('[s|S]tr\.*$').search(street)
if t != None:
print(street)
def audit_city(db, collection):
cities = db_ops.get_values(db, collection, ['address.city'])
#clean city names shown to have problem characters
cities_to_clean = {'wrong_state': {'Barendorf, Kreis Lüneburg':'Barendorf',
'Wintermoor a. d. Ch.':'Wintermoor an der Chaussee',
'Moisburg/Hollenstedt':'Moisburg',
'Lauenburg/Elbe':'Lauenburg Elbe'}}
db_ops.update_db(db, collection, 'address.city', cities_to_clean)
#cross calidate cities with state and postal codes
for city in cities:
url = 'http://maps.googleapis.com/maps/api/geocode/json?address=urlencode($_REQUEST["' + city.encode('utf8') + '" Germany])&sensor=false&oe=utf-8'
attempts = 0
success = False
while success != True and attempts < 3:
page = urllib.urlopen(url)
data = json.loads(page.read())
attempts += 1
if data['status'] == "OVER_QUERY_LIMIT":
time.sleep(2)
continue
success = True
if attempts == 3:
print("Daily limit has been reached")
if data['status'] == 'ZERO_RESULTS':
#print(city + ' not found!')
with open('Output\\' + collection + '_unknown-cities.txt', 'a') as f:
f.write(unicode(city + '\n').encode("utf-8"))
else:
try:
for i in range(len(data['results'][0]['address_components'])):
if data['results'][0]['address_components'][i]['types'] == ['postal_code']:
pc = data['results'][0]['address_components'][i]['long_name']
if db[collection].find({'address.city':city, 'address.postcode':{'$ne':pc}}) != None:
print(city, pc)
db[collection].update({'address.city':city, 'address.postcode':{'$ne':pc}},{'$set':{'address.postcode':pc}})
if data['results'][0]['address_components'][i]['types'] == [ "administrative_area_level_1", "political" ]:
state = data['results'][0]['address_components'][i]['long_name']
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates', 'address.city', city)
except:
pprint.pprint(data)
def audit_country(db, collection):
countries = db_ops.get_values(db, collection, ['address.country'])
for country in countries:
print(country)
def audit_housenumbers(db, collection):
hns = db_ops.get_values(db, collection, ['address.housenumber'])
unclean_hns = []
#List of non-existing house numbers
entries_to_remove = ['g50', 'g70', 'g30a']
#Dictionary of house numbers, that were cleaned manually, since their scheme
#was rather rare
manual = {
'2-Block D': '2, Block D',
'1a-1': '1, 1A',
'49 - 2.OG': '49',
'71, Tor 2': '71',
'37 (links)': '37',
'44<B': '44B',
'38-': '38',
'7a b': '7A, B',
'38 und a b': '38, 38A, B',
'31 bc': '31B, C',
'7 abc': '7A, B, C',
'6a/b 4a/b': '4A, B, 6A, B',
'48+a, 51,52': '48, 48A, 51, 52',
'57, 59a + b, 61a + b': '57, 59A, B, 61A, B',
'18 + a, 20 + a': '18, 18A, 20, 20A',
'37b-39': '37B-39',
'2 a b': '2A,B',
'103/105/105a': '103,105,105A',
'51a-e; 53 a-e; 55 a-e': '51A-E, 53A-E, 55A-E',
'45, 47 a/b': '45, 47A, B',
'41a (Tal)': '41A',
'36, 36a - 36f': '36, 36A-F',
'32;34:36': '32, 34-36',
'10b-12': '10B-12',
'17anb.': '17',
'10 - 10a - 12': '10, 10A, 12'
}
#Dictionary in the form the db_ops.update_db function needs it. The numbers represent the
#regular expression which detected the respective house number to be able to debug
#possble wrong changes
cleaned = {
'1': {},
'2': {},
'3': {},
'4': {},
'5': {},
'6': {},
'7': {},
'8': {},
'9': {},
'10': {},
'11': {},
'12': {},
'13': {},
'14': {},
'15': {},
'16': {},
'17': {},
'18': {},
'19': {},
'20': {},
'21': {},
'22': {},
'manual': {},
'remove': {},
'not edited': {}
}
for hn in hns:
#e.g. 3 // 3 Hof => 3 // 3 Hof
if re.compile('((^[0-9]+$)|(^[0-9]+\s?Hof))').search(hn) != None:
cleaned['1'][hn] = hn
#e.g. 3a // 3 a // 3-a // 3 - a // 3 und a => 3A
elif re.compile('^[0-9]+(\s|(\s?[-]\s?)|(\s?und\s?))?[a-zA-Z]$'
).search(hn) != None:
num = re.compile('^[0-9]+').search(hn).group(0)
char = re.compile('[A-Za-z]+$').search(hn).group(0)
new_hn = num + char.upper()
cleaned['2'][hn] = new_hn
#e.g. 3/3 // 3 / 3 => 3, Haus 3
elif re.compile('^[0-9]+\s?[/]\s?[0-9]+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
if abs(int(nums[0]) - int(nums[1])) > 3:
new_hn = nums[0] + ', Haus ' + nums[1]
else:
nums = sorted(nums)
new_hn = nums[0] + ', ' + nums[1]
cleaned['3'][hn] = new_hn
#e.g. 3,4 // 3 + 4 => 3, 4
elif re.compile('^[0-9]+\s?([,;+]\s?[0-9]+)+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
new_hn = ', '.join(sorted([num for num in nums]))
cleaned['4'][hn] = new_hn
#e.g. 3 - 4 // 3 - 4a => 3-4 // 3-4A
elif re.compile('^[0-9]+\s?([-:]\s?[0-9]+([a-zA-Z]+)?)+$',
re.UNICODE).search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
sl = sorted([int(num) for num in nums])
sl = [str(x).encode('utf-8') for x in sl]
try:
chars = re.compile('[A-Za-z]+').search(hn).group(0)
new_hn = '-'.join(sl) + chars.upper()
except:
new_hn = '-'.join(sl)
cleaned['5'][hn] = new_hn
#e.g. 3 a-b // 3-a-b => 3A-B
elif re.compile('^[0-9]+(\s|[-])?[a-zA-Z]+(\s?[-]\s?[a-zA-Z]+)+$'
).search(hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + '-'.join(sorted([char.upper() for char in chars]))
cleaned['6'][hn] = new_hn
#e.g. 3a,b,c // 3 a + b + c => 3A,B,C
elif re.compile('^[0-9]+\s?[a-zA-Z]?(\s?[,;/+]\s?[a-zA-Z]?)+$').search(
hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + ','.join(sorted([char.upper() for char in chars]))
cleaned['7'][hn] = new_hn
#e.g. 3a,3b // 3a,4a => 3A,B //3A, 4A
elif re.compile('^[0-9]+\s?[a-zA-Z](\s?[,;/+&]\s?[0-9]+\s?[a-zA-Z]+)+$'
).search(hn) != None:
el = re.split('[,;/+&]', hn)
while '' in el:
el.remove('')
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + ','.join(
sorted([char.upper() for char in num[n]])) + '; '
new_hn = new_hn[:-2]
cleaned['8'][hn] = new_hn
#e.g. 3a - 4b => 3A-4B
elif re.compile('^[0-9]+\s?[a-zA-Z]+(\s?[-]\s?([0-9]+)\s?[a-zA-Z]+)+$'
).search(hn) != None:
el = re.split('-', hn)
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + '-'.join(
sorted([char.upper() for char in num[n]])) + ', '
new_hn = new_hn[:-2]
cleaned['9'][hn] = new_hn
#e.g. 3a 4b 5c => 3A, 4B, 5C
elif re.compile('^[0-9]+[a-zA-Z]+(\s[0-9]+[a-zA-Z]+)+$').search(
hn) != None:
el = re.split(' ', hn)
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + ', '.join(
sorted([char.upper() for char in num[n]])) + ', '
new_hn = new_hn[:-2]
cleaned['10'][hn] = new_hn
#e.g. 3 Haus 5a => 3, Haus 5A
elif re.compile('^[0-9]+\sHaus\s([0-9]+)?[a-zA-Z]?$').search(
hn) != None:
e = re.split(' ', hn)
new_hn = e[0] + ', Haus ' + e[2].upper()
cleaned['11'][hn] = new_hn
#e.g. 3-4 Hs. 9 => 3-4, Haus 9
elif re.compile(
'^[0-9]+([-][0-9]+|[a-z])?\s?((Hs\.)|(HTH)|([.]))\s?[0-9]?\+?[0-9]?$'
).search(hn) != None:
e = re.split('\.?\s?', hn)
if len(e) == 2:
new_hn = e[0] + ', Haus ' + e[1]
elif not e[2] == '':
if re.compile('[a-z]+').search(e[0]) != None:
new_hn = e[0].upper() + ', Haus ' + e[2]
else:
new_hn = '-'.join(re.split(
'-', e[0])).upper() + ', Haus ' + e[2]
else:
new_hn = '-'.join(re.split('-', e[0]))
cleaned['12'][hn] = new_hn
#e.g. 3 A4 => 3, Haus A4
elif re.compile('^[0-9]+\sA\s?[-]?[0-9]+$').search(hn) != None:
e = re.split('[ -]', hn)
new_hn = e[0] + ', Haus A' + e[2]
cleaned['13'][hn] = new_hn
#e.g. 3, 3a, 3b => 3,3A,B
elif re.compile('^[0-9]+\s?([,;/+]\s?[0-9]+\s?[a-zA-Z]+)+$').search(
hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + ', ' + num + ','.join(
sorted([char.upper() for char in chars]))
cleaned['14'][hn] = new_hn
#e.g. 3? => 3
elif re.compile('[?]').search(hn) != None:
new_hn = re.compile('[0-9]+').search(hn).group(0)
cleaned['15'][hn] = new_hn
#e.g. 3a und 3b => 3A,B
elif re.compile('^[0-9]+[a-zA-Z]?(\s|(\s?und\s?))[0-9]+[a-zA-Z]?$'
).search(hn) != None:
e = sorted(re.split(' ', hn))
new_hn = e[0].upper() + ', ' + e[1].upper()
cleaned['16'][hn] = new_hn
#e.g. 3a/b => 3A,B
elif re.compile('^[0-9]+[a-zA-Z][/]?[0-9]+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
char = re.compile('[A-Za-z]+').search(hn).group(0)
new_hn = nums[0] + char.upper() + ', Haus ' + nums[1]
cleaned['17'][hn] = None
#e.g. T3 => 3
elif re.compile('^T[0-9]+').search(hn) != None:
cleaned['18'][hn] = re.compile('[0-9]+').search(hn).group(0)
#e.g. 3a,3,3b => 3,3A,3B
elif re.compile('^([0-9]+[a-zA-Z]?[,]?)+$').search(hn) != None:
n = re.split(',', hn)
new_hn = ', '.join(sorted([x.upper() for x in n]))
cleaned['19'][hn] = None
#3,3a-f => 3, 3A-F
elif re.compile('^[0-9]+[,+]\s?[0-9]+[a-zA-Z][-/][a-zA-Z]$').search(
hn) != None:
n = re.split('[,+]', hn)
new_hn = ', '.join([x.upper() for x in n])
new_hn = re.sub('/', '-', new_hn)
cleaned['20'][hn] = new_hn
#e.g. b => None
elif re.compile('[0-9]+').search(hn) == None:
cleaned['22'][hn] = None
elif hn in entries_to_remove:
cleaned['remove'][hn] = None
elif hn in manual.keys():
cleaned['manual'][hn] = manual[hn]
else:
cleaned['not edited'][hn] = hn
unclean_hns.append(hn)
db_ops.update_db(db, collection, 'address.housenumber', cleaned)
def audit_city(db, collection):
cities = db_ops.get_values(db, collection, ['address.city'])
#clean city names shown to have problem characters
cities_to_clean = {
'wrong_state': {
'Barendorf, Kreis Lüneburg': 'Barendorf',
'Wintermoor a. d. Ch.': 'Wintermoor an der Chaussee',
'Moisburg/Hollenstedt': 'Moisburg',
'Lauenburg/Elbe': 'Lauenburg Elbe'
}
}
db_ops.update_db(db, collection, 'address.city', cities_to_clean)
#cross calidate cities with state and postal codes
for city in cities:
url = 'http://maps.googleapis.com/maps/api/geocode/json?address=urlencode($_REQUEST["' + city.encode(
'utf8') + '" Germany])&sensor=false&oe=utf-8'
attempts = 0
success = False
while success != True and attempts < 3:
page = urllib.urlopen(url)
data = json.loads(page.read())
attempts += 1
if data['status'] == "OVER_QUERY_LIMIT":
time.sleep(2)
continue
success = True
if attempts == 3:
print("Daily limit has been reached")
if data['status'] == 'ZERO_RESULTS':
#print(city + ' not found!')
with open('Output\\' + collection + '_unknown-cities.txt',
'a') as f:
f.write(unicode(city + '\n').encode("utf-8"))
else:
try:
for i in range(len(data['results'][0]['address_components'])):
if data['results'][0]['address_components'][i][
'types'] == ['postal_code']:
pc = data['results'][0]['address_components'][i][
'long_name']
if db[collection].find({
'address.city': city,
'address.postcode': {
'$ne': pc
}
}) != None:
print(city, pc)
db[collection].update(
{
'address.city': city,
'address.postcode': {
'$ne': pc
}
}, {'$set': {
'address.postcode': pc
}})
if data['results'][0]['address_components'][i][
'types'] == [
"administrative_area_level_1", "political"
]:
state = data['results'][0]['address_components'][i][
'long_name']
if state != 'Hamburg':
db_ops.move_db(db, collection, 'SurroundingStates',
'address.city', city)
except:
pprint.pprint(data)
def audit_country(db, collection):
countries = db_ops.get_values(db, collection, ['address.country'])
for country in countries:
print(country)
def audit_housenumbers(db, collection):
hns = db_ops.get_values(db, collection, ['address.housenumber'])
unclean_hns = []
#List of non-existing house numbers
entries_to_remove = ['g50','g70','g30a']
#Dictionary of house numbers, that were cleaned manually, since their scheme
#was rather rare
manual = {'2-Block D':'2, Block D',
'1a-1':'1, 1A',
'49 - 2.OG':'49',
'71, Tor 2':'71',
'37 (links)':'37',
'44<B':'44B',
'38-':'38',
'7a b':'7A, B',
'38 und a b':'38, 38A, B',
'31 bc':'31B, C',
'7 abc':'7A, B, C',
'6a/b 4a/b':'4A, B, 6A, B',
'48+a, 51,52':'48, 48A, 51, 52',
'57, 59a + b, 61a + b':'57, 59A, B, 61A, B',
'18 + a, 20 + a':'18, 18A, 20, 20A',
'37b-39':'37B-39',
'2 a b':'2A,B',
'103/105/105a':'103,105,105A',
'51a-e; 53 a-e; 55 a-e':'51A-E, 53A-E, 55A-E',
'45, 47 a/b':'45, 47A, B',
'41a (Tal)': '41A',
'36, 36a - 36f':'36, 36A-F',
'32;34:36':'32, 34-36',
'10b-12':'10B-12',
'17anb.':'17',
'10 - 10a - 12':'10, 10A, 12'}
#Dictionary in the form the db_ops.update_db function needs it. The numbers represent the
#regular expression which detected the respective house number to be able to debug
#possble wrong changes
cleaned = {'1':{},
'2':{},
'3':{},
'4':{},
'5': {},
'6': {},
'7': {},
'8': {},
'9': {},
'10': {},
'11': {},
'12': {},
'13': {},
'14': {},
'15': {},
'16': {},
'17': {},
'18': {},
'19': {},
'20': {},
'21': {},
'22': {},
'manual': {},
'remove': {},
'not edited': {}}
for hn in hns:
#e.g. 3 // 3 Hof => 3 // 3 Hof
if re.compile('((^[0-9]+$)|(^[0-9]+\s?Hof))').search(hn) != None:
cleaned['1'][hn] = hn
#e.g. 3a // 3 a // 3-a // 3 - a // 3 und a => 3A
elif re.compile('^[0-9]+(\s|(\s?[-]\s?)|(\s?und\s?))?[a-zA-Z]$').search(hn) != None:
num = re.compile('^[0-9]+').search(hn).group(0)
char = re.compile('[A-Za-z]+$').search(hn).group(0)
new_hn = num + char.upper()
cleaned['2'][hn] = new_hn
#e.g. 3/3 // 3 / 3 => 3, Haus 3
elif re.compile('^[0-9]+\s?[/]\s?[0-9]+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
if abs(int(nums[0])-int(nums[1])) > 3:
new_hn = nums[0] + ', Haus ' + nums[1]
else:
nums = sorted(nums)
new_hn = nums[0]+ ', '+ nums[1]
cleaned['3'][hn] = new_hn
#e.g. 3,4 // 3 + 4 => 3, 4
elif re.compile('^[0-9]+\s?([,;+]\s?[0-9]+)+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
new_hn = ', '.join(sorted([num for num in nums]))
cleaned['4'][hn] = new_hn
#e.g. 3 - 4 // 3 - 4a => 3-4 // 3-4A
elif re.compile('^[0-9]+\s?([-:]\s?[0-9]+([a-zA-Z]+)?)+$', re.UNICODE).search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
sl = sorted([int(num) for num in nums])
sl = [str(x).encode('utf-8') for x in sl]
try:
chars = re.compile('[A-Za-z]+').search(hn).group(0)
new_hn = '-'.join(sl) + chars.upper()
except:
new_hn = '-'.join(sl)
cleaned['5'][hn] = new_hn
#e.g. 3 a-b // 3-a-b => 3A-B
elif re.compile('^[0-9]+(\s|[-])?[a-zA-Z]+(\s?[-]\s?[a-zA-Z]+)+$').search(hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + '-'.join(sorted([char.upper() for char in chars]))
cleaned['6'][hn] = new_hn
#e.g. 3a,b,c // 3 a + b + c => 3A,B,C
elif re.compile('^[0-9]+\s?[a-zA-Z]?(\s?[,;/+]\s?[a-zA-Z]?)+$').search(hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + ','.join(sorted([char.upper() for char in chars]))
cleaned['7'][hn] = new_hn
#e.g. 3a,3b // 3a,4a => 3A,B //3A, 4A
elif re.compile('^[0-9]+\s?[a-zA-Z](\s?[,;/+&]\s?[0-9]+\s?[a-zA-Z]+)+$').search(hn) != None:
el = re.split('[,;/+&]', hn)
while '' in el:
el.remove('')
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + ','.join(sorted([char.upper() for char in num[n]])) + '; '
new_hn = new_hn[:-2]
cleaned['8'][hn] = new_hn
#e.g. 3a - 4b => 3A-4B
elif re.compile('^[0-9]+\s?[a-zA-Z]+(\s?[-]\s?([0-9]+)\s?[a-zA-Z]+)+$').search(hn) != None:
el = re.split('-', hn)
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + '-'.join(sorted([char.upper() for char in num[n]])) + ', '
new_hn = new_hn[:-2]
cleaned['9'][hn] = new_hn
#e.g. 3a 4b 5c => 3A, 4B, 5C
elif re.compile('^[0-9]+[a-zA-Z]+(\s[0-9]+[a-zA-Z]+)+$').search(hn) != None:
el = re.split(' ', hn)
num = {}
new_hn = ''
for e in el:
n = re.compile('[0-9]+').search(e).group(0)
c = re.compile('[A-Za-z]+').search(e).group(0)
try:
num[n].add(c)
except:
num[n] = set(c)
for n in num:
new_hn = new_hn + n + ', '.join(sorted([char.upper() for char in num[n]])) + ', '
new_hn = new_hn[:-2]
cleaned['10'][hn] = new_hn
#e.g. 3 Haus 5a => 3, Haus 5A
elif re.compile('^[0-9]+\sHaus\s([0-9]+)?[a-zA-Z]?$').search(hn) != None:
e = re.split(' ', hn)
new_hn = e[0] + ', Haus ' + e[2].upper()
cleaned['11'][hn] = new_hn
#e.g. 3-4 Hs. 9 => 3-4, Haus 9
elif re.compile('^[0-9]+([-][0-9]+|[a-z])?\s?((Hs\.)|(HTH)|([.]))\s?[0-9]?\+?[0-9]?$').search(hn) != None:
e = re.split('\.?\s?', hn)
if len(e) == 2:
new_hn = e[0] + ', Haus ' + e[1]
elif not e[2] == '':
if re.compile('[a-z]+').search(e[0]) != None:
new_hn = e[0].upper() + ', Haus ' + e[2]
else:
new_hn = '-'.join(re.split('-',e[0])).upper() + ', Haus ' + e[2]
else:
new_hn = '-'.join(re.split('-',e[0]))
cleaned['12'][hn] = new_hn
#e.g. 3 A4 => 3, Haus A4
elif re.compile('^[0-9]+\sA\s?[-]?[0-9]+$').search(hn) != None:
e = re.split('[ -]', hn)
new_hn = e[0] + ', Haus A' + e[2]
cleaned['13'][hn] = new_hn
#e.g. 3, 3a, 3b => 3,3A,B
elif re.compile('^[0-9]+\s?([,;/+]\s?[0-9]+\s?[a-zA-Z]+)+$').search(hn) != None:
num = re.compile('[0-9]+').search(hn).group(0)
chars = re.compile('[A-Za-z]+').findall(hn)
new_hn = num + ', ' + num + ','.join(sorted([char.upper() for char in chars]))
cleaned['14'][hn] = new_hn
#e.g. 3? => 3
elif re.compile('[?]').search(hn) != None:
new_hn = re.compile('[0-9]+').search(hn).group(0)
cleaned['15'][hn] = new_hn
#e.g. 3a und 3b => 3A,B
elif re.compile('^[0-9]+[a-zA-Z]?(\s|(\s?und\s?))[0-9]+[a-zA-Z]?$').search(hn) != None:
e = sorted(re.split(' ', hn))
new_hn = e[0].upper() + ', ' + e[1].upper()
cleaned['16'][hn] = new_hn
#e.g. 3a/b => 3A,B
elif re.compile('^[0-9]+[a-zA-Z][/]?[0-9]+$').search(hn) != None:
nums = re.compile('[0-9]+').findall(hn)
char = re.compile('[A-Za-z]+').search(hn).group(0)
new_hn = nums[0] + char.upper() +', Haus ' + nums[1]
cleaned['17'][hn] = None
#e.g. T3 => 3
elif re.compile('^T[0-9]+').search(hn) != None:
cleaned['18'][hn] = re.compile('[0-9]+').search(hn).group(0)
#e.g. 3a,3,3b => 3,3A,3B
elif re.compile('^([0-9]+[a-zA-Z]?[,]?)+$').search(hn) != None:
n = re.split(',', hn)
new_hn = ', '.join(sorted([x.upper() for x in n]))
cleaned['19'][hn] = None
#3,3a-f => 3, 3A-F
elif re.compile('^[0-9]+[,+]\s?[0-9]+[a-zA-Z][-/][a-zA-Z]$').search(hn) != None:
n = re.split('[,+]', hn)
new_hn = ', '.join([x.upper() for x in n])
new_hn = re.sub('/','-', new_hn)
cleaned['20'][hn] = new_hn
#e.g. b => None
elif re.compile('[0-9]+').search(hn) == None:
cleaned['22'][hn] = None
elif hn in entries_to_remove:
cleaned['remove'][hn] = None
elif hn in manual.keys():
cleaned['manual'][hn] = manual[hn]
else:
cleaned['not edited'][hn] = hn
unclean_hns.append(hn)
db_ops.update_db(db, collection, 'address.housenumber', cleaned)