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

[osm] removing postcode phrase from place training data and adding CLDR countries only after all the other normalizations

Browse Source
This commit is contained in:
Al
2016-08-02 14:52:12 -04:00
parent 5117fb21d3
commit c40ad99ec7
1 changed files with 4 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
scripts/geodata/osm/formatter.py
View File

@@ -557,11 +557,6 @@ class OSMAddressFormatter(object):
revised_place_tags = []
for address_components, language, is_default in place_tags:
if (AddressFormatter.COUNTRY in address_components or place_config.include_component(AddressFormatter.COUNTRY, containing_ids, country=country)) and random.random() < cldr_country_prob:
address_country = self.components.cldr_country_name(country, language)
if address_country:
address_components[AddressFormatter.COUNTRY] = address_country
revised_address_components = place_config.dropout_components(address_components, osm_components, country=country, population=population)
revised_address_components[component_name] = address_components[component_name]
self.components.drop_invalid_components(revised_address_components)
@@ -573,7 +568,10 @@ class OSMAddressFormatter(object):
self.components.remove_numeric_boundary_names(revised_address_components)
self.components.cleanup_boundary_names(revised_address_components)
self.components.add_postcode_phrase(revised_address_components, language, country=country)
if (AddressFormatter.COUNTRY in address_components or place_config.include_component(AddressFormatter.COUNTRY, containing_ids, country=country)) and random.random() < cldr_country_prob:
address_country = self.components.cldr_country_name(country, language)
if address_country:
address_components[AddressFormatter.COUNTRY] = address_country
if revised_address_components:
revised_place_tags.append((revised_address_components, language, is_default))