tommy/libpostal
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[geoplanet] only add short postal codes to GeoPlanet data set if they match the Google regexes

Browse Source
This commit is contained in:
Al
2016-12-13 17:03:26 -05:00
parent 34db27b80c
commit 09f808ca47
1 changed files with 13 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
scripts/geodata/geoplanet/geoplanet_training_data.py
View File

@@ -18,6 +18,7 @@ from geodata.address_expansions.gazetteers import *
from geodata.address_formatting.formatter import AddressFormatter from geodata.address_formatting.formatter import AddressFormatter
from geodata.countries.names import country_names from geodata.countries.names import country_names
from geodata.i18n.google import postcode_regexes
from geodata.names.normalization import name_affixes from geodata.names.normalization import name_affixes
from geodata.places.config import place_config from geodata.places.config import place_config
@@ -199,6 +200,18 @@ class GeoPlanetFormatter(object):
language = self.language_codes[language] language = self.language_codes[language]
if len(postal_code) <= 3:
postcode_regex = postcode_regexes.get(country)
valid_postcode = False
if postcode_regex:
match = postcode_regex.match(postal_code)
if match and match.end() == len(postal_code):
valid_postcode = True
if not valid_postcode:
continue
# If the county/state is coterminous with a city and contains only one place, # If the county/state is coterminous with a city and contains only one place,
# set the parent_id to the city instead # set the parent_id to the city instead
if parent_id in self.coterminous_admins: if parent_id in self.coterminous_admins: