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

[parser] Fixing config keys so OSM streets/venues get abbreviated. Selecting namespaced address fields in cases like Brussels or Hong Kong where everything is bilingual. Adding the ability to pass a known language into address component expansion

Browse Source
This commit is contained in:
Al
2016-05-26 12:05:46 -04:00
parent 206a471732
commit 5daa64faef
2 changed files with 35 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
scripts/geodata/addresses/components.py
View File

@@ -554,8 +554,8 @@ class AddressComponents(object):
States
------
Primarily for the US, Canada and Australia, OSM tends to use the abbreviated state name
whereas we'd like to include both forms, so wtih some probability, replace the abbreviated
Primarily for the US, Canada and Australia, OSM addr:state tags tend to use the abbreviated
state name whereas we'd like to include both forms. With some probability, replace the abbreviated
name with the unabbreviated one e.g. CA => California
'''
address_state = address_components.get(AddressFormatter.STATE)
@@ -1029,7 +1029,7 @@ class AddressComponents(object):
return True
return False
def expanded(self, address_components, latitude, longitude,
def expanded(self, address_components, latitude, longitude, language=None,
dropout_places=True, add_sub_building_components=True,
num_floors=None, num_basements=None, zone=None):
'''
@@ -1053,11 +1053,10 @@ class AddressComponents(object):
if not (country and candidate_languages):
return None, None, None
language = None
more_than_one_official_language = len(candidate_languages) > 1
language = self.address_language(address_components, candidate_languages)
if not language:
language = self.address_language(address_components, candidate_languages)
non_local_language = self.non_local_language()
# If a country was already specified