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

[fix] Adding country_region to address_parser_train

Browse Source
This commit is contained in:
Al
2016-07-28 16:18:32 -04:00
parent 21bcbd8381
commit 1b09b7f2e5
2 changed files with 6 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
src/address_parser.h
View File

@@ -82,7 +82,8 @@ with the general error-driven averaged perceptron.
#define ADDRESS_COMPONENT_STATE_DISTRICT 1 << 11 #define ADDRESS_COMPONENT_STATE_DISTRICT 1 << 11
#define ADDRESS_COMPONENT_STATE 1 << 12 #define ADDRESS_COMPONENT_STATE 1 << 12
#define ADDRESS_COMPONENT_POSTAL_CODE 1 << 13 #define ADDRESS_COMPONENT_POSTAL_CODE 1 << 13
#define ADDRESS_COMPONENT_COUNTRY 1 << 14 #define ADDRESS_COMPONENT_COUNTRY_REGION 1 << 14
#define ADDRESS_COMPONENT_COUNTRY 1 << 15
typedef enum { typedef enum {
ADDRESS_PARSER_HOUSE, ADDRESS_PARSER_HOUSE,

4
src/address_parser_train.c
View File

@@ -34,6 +34,7 @@ static inline bool is_phrase_component(char *label) {
string_equals(label, ADDRESS_PARSER_LABEL_ISLAND) || string_equals(label, ADDRESS_PARSER_LABEL_ISLAND) ||
string_equals(label, ADDRESS_PARSER_LABEL_STATE) || string_equals(label, ADDRESS_PARSER_LABEL_STATE) ||
string_equals(label, ADDRESS_PARSER_LABEL_POSTAL_CODE) || string_equals(label, ADDRESS_PARSER_LABEL_POSTAL_CODE) ||
string_equals(label, ADDRESS_PARSER_LABEL_COUNTRY_REGION) ||
string_equals(label, ADDRESS_PARSER_LABEL_COUNTRY)); string_equals(label, ADDRESS_PARSER_LABEL_COUNTRY));
} }
@@ -182,6 +183,9 @@ address_parser_t *address_parser_init(char *filename) {
} else if (string_equals(prev_label, ADDRESS_PARSER_LABEL_COUNTRY)) { } else if (string_equals(prev_label, ADDRESS_PARSER_LABEL_COUNTRY)) {
class_id = ADDRESS_PARSER_COUNTRY; class_id = ADDRESS_PARSER_COUNTRY;
component = ADDRESS_COMPONENT_COUNTRY; component = ADDRESS_COMPONENT_COUNTRY;
} else if (string_equals(prev_label, ADDRESS_PARSER_LABEL_COUNTRY_REGION)) {
class_id = ADDRESS_PARSER_COUNTRY_REGION;
component = ADDRESS_COMPONENT_COUNTRY;
} else if (string_equals(prev_label, ADDRESS_PARSER_LABEL_STATE_DISTRICT)) { } else if (string_equals(prev_label, ADDRESS_PARSER_LABEL_STATE_DISTRICT)) {
class_id = ADDRESS_PARSER_STATE_DISTRICT; class_id = ADDRESS_PARSER_STATE_DISTRICT;
component = ADDRESS_COMPONENT_STATE_DISTRICT; component = ADDRESS_COMPONENT_STATE_DISTRICT;