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

[dedupe] allow near-dupe hashes if only a small containing boundary is present (e.g. county/state district).

Browse Source
This commit is contained in:
Al
2019-02-16 22:26:13 -05:00
parent 3610ffaa05
commit b7052caf6b
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/near_dupe.c
View File

@@ -144,6 +144,7 @@ bool cstring_array_add_string_no_whitespace(cstring_array *strings, char *str) {
cstring_array *expanded_component_combined(char *input, libpostal_normalize_options_t options, bool remove_spaces, size_t *n) { cstring_array *expanded_component_combined(char *input, libpostal_normalize_options_t options, bool remove_spaces, size_t *n) {
char *expansion;
size_t num_expansions = 0; size_t num_expansions = 0;
cstring_array *expansions = expand_address(input, options, &num_expansions); cstring_array *expansions = expand_address(input, options, &num_expansions);
@@ -160,7 +161,6 @@ cstring_array *expanded_component_combined(char *input, libpostal_normalize_opti
return root_expansions; return root_expansions;
} else { } else {
khash_t(str_set) *unique_strings = kh_init(str_set); khash_t(str_set) *unique_strings = kh_init(str_set);
char *expansion;
khiter_t k; khiter_t k;
int ret; int ret;
@@ -640,7 +640,7 @@ static inline void add_string_hash_permutations(cstring_array *near_dupe_hashes,
cstring_array *near_dupe_hashes_languages(size_t num_components, char **labels, char **values, libpostal_near_dupe_hash_options_t options, size_t num_languages, char **languages) { cstring_array *near_dupe_hashes_languages(size_t num_components, char **labels, char **values, libpostal_near_dupe_hash_options_t options, size_t num_languages, char **languages) {
if (!options.with_latlon && !options.with_city_or_equivalent && !options.with_postal_code) return NULL; if (!options.with_latlon && !options.with_city_or_equivalent && !options.with_small_containing_boundaries && !options.with_postal_code) return NULL;
place_t *place = place_from_components(num_components, labels, values); place_t *place = place_from_components(num_components, labels, values);
log_debug("created place\n"); log_debug("created place\n");