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

[addresses] doing hyphenation for existing components in component expansion (i.e. OSM training data)

Browse Source
This commit is contained in:
Al
2016-10-21 22:02:15 -04:00
parent 0216a991c6
commit d51a1d6196
1 changed files with 9 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
scripts/geodata/addresses/components.py
View File

@@ -1392,7 +1392,7 @@ class AddressComponents(object):
def expanded(self, address_components, latitude, longitude, language=None, def expanded(self, address_components, latitude, longitude, language=None,
dropout_places=True, population=None, population_from_city=False, dropout_places=True, population=None, population_from_city=False,
add_sub_building_components=True, add_sub_building_components=True, hyphenation=True,
num_floors=None, num_basements=None, zone=None): num_floors=None, num_basements=None, zone=None):
''' '''
Expanded components Expanded components
@@ -1411,6 +1411,14 @@ class AddressComponents(object):
except Exception: except Exception:
return None, None, None return None, None, None
if hyphenation:
for component in address_components:
if component in place_config.ADMIN_COMPONENTS:
value = address_components[component]
value_hyphens = self.name_hyphens(value)
if value_hyphens != value:
address_components[component] = value_hyphens
osm_components = self.osm_reverse_geocoded_components(latitude, longitude) osm_components = self.osm_reverse_geocoded_components(latitude, longitude)
country, candidate_languages = self.osm_country_and_languages(osm_components) country, candidate_languages = self.osm_country_and_languages(osm_components)
if not (country and candidate_languages): if not (country and candidate_languages):