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

[languages] Adding several different types of dictionaries for name expansion/abbreviation in OSM

Browse Source
This commit is contained in:
Al
2016-01-22 00:51:32 -05:00
parent 89aa039692
commit f4995d4f0f
1 changed files with 49 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

72
scripts/geodata/language_id/disambiguation.py
View File

@@ -144,11 +144,11 @@ class DictionaryPhraseFilter(PhraseFilter):
suffix_search, suffix_len = self.search_suffix(token) suffix_search, suffix_len = self.search_suffix(token)
if suffix_search and self.trie.get(token[(token_len - suffix_len):].rstrip('.')): if suffix_search and self.trie.get(token[(token_len - suffix_len):].rstrip('.')):
yield (t, PHRASE, suffix_len, map(safe_decode, suffix_search)) yield ([(t, c)], PHRASE, suffix_len, map(safe_decode, suffix_search))
continue continue
prefix_search, prefix_len = self.search_prefix(token) prefix_search, prefix_len = self.search_prefix(token)
if prefix_search and self.trie.get(token[:prefix_len]): if prefix_search and self.trie.get(token[:prefix_len]):
yield (t, PHRASE, prefix_len, map(safe_decode, prefix_search)) yield ([(t, c)], PHRASE, prefix_len, map(safe_decode, prefix_search))
continue continue
else: else:
c = PHRASE c = PHRASE
@@ -159,6 +159,11 @@ STREET_TYPES_DICTIONARIES = ('street_types.txt',
'concatenated_suffixes_separable.txt', 'concatenated_suffixes_separable.txt',
'concatenated_suffixes_inseparable.txt', 'concatenated_suffixes_inseparable.txt',
'concatenated_prefixes_separable.txt', 'concatenated_prefixes_separable.txt',
'organizations.txt',
'people.txt',
'personal_suffixes.txt',
'personal_titles.txt',
'qualifiers.txt',
'stopwords.txt',) 'stopwords.txt',)
GIVEN_NAME_DICTIONARY = 'given_names.txt' GIVEN_NAME_DICTIONARY = 'given_names.txt'
@@ -167,28 +172,46 @@ SURNAME_DICTIONARY = 'surnames.txt'
NAME_DICTIONARIES = (GIVEN_NAME_DICTIONARY, NAME_DICTIONARIES = (GIVEN_NAME_DICTIONARY,
SURNAME_DICTIONARY,) SURNAME_DICTIONARY,)
ALL_ABBREVIATION_DICTIONARIES = STREET_TYPES_DICTIONARIES + ('academic_degrees.txt',
'building_types.txt',
'company_types.txt',
'directionals.txt',
'level_types.txt',
'no_number.txt',
'nulls.txt',
'organizations.txt',
'people.txt',
'personal_suffixes.txt',
'personal_titles.txt',
'place_names.txt',
'post_office.txt',
'qualifiers.txt',
'synonyms.txt',
'toponyms.txt',
'unit_types.txt',
)
street_types_gazetteer = DictionaryPhraseFilter(*STREET_TYPES_DICTIONARIES)
abbreviations_gazetteer = DictionaryPhraseFilter(*ALL_ABBREVIATION_DICTIONARIES) NAME_ABBREVIATION_DICTIONARIES = STREET_TYPES_DICTIONARIES + ('academic_degrees.txt',
given_name_gazetteer = DictionaryPhraseFilter(GIVEN_NAME_DICTIONARY) 'building_types.txt',
'company_types.txt',
'place_names.txt',
'qualifiers.txt',
'synonyms.txt',
'toponyms.txt',
)
UNIT_ABBREVIATION_DICTIONARIES = ('level_types.txt',
'post_office.txt',
'unit_types.txt',
)
ALL_ABBREVIATION_DICTIONARIES = STREET_TYPES_DICTIONARIES + \
NAME_ABBREVIATION_DICTIONARIES + \
UNIT_ABBREVIATION_DICTIONARIES + \
('no_number.txt', 'nulls.txt',)
gazetteers = []
def create_gazetteer(*dictionaries):
g = DictionaryPhraseFilter(*dictionaries)
gazetteers.append(g)
return g
street_types_gazetteer = create_gazetteer(*STREET_TYPES_DICTIONARIES)
names_gazetteer = create_gazetteer(*NAME_ABBREVIATION_DICTIONARIES)
unit_types_gazetteer = create_gazetteer(*UNIT_ABBREVIATION_DICTIONARIES)
street_and_unit_types_gazetteer = create_gazetteer(*(STREET_TYPES_DICTIONARIES + UNIT_ABBREVIATION_DICTIONARIES))
abbreviations_gazetteer = create_gazetteer(*ALL_ABBREVIATION_DICTIONARIES)
given_name_gazetteer = create_gazetteer(GIVEN_NAME_DICTIONARY)
char_scripts = [] char_scripts = []
script_languages = {} script_languages = {}
@@ -200,6 +223,9 @@ def init_disambiguation():
char_scripts.extend(get_chars_by_script()) char_scripts.extend(get_chars_by_script())
script_languages.update({script: set(langs) for script, langs in get_script_languages().iteritems()}) script_languages.update({script: set(langs) for script, langs in get_script_languages().iteritems()})
for g in gazetteers:
g.configure()
UNKNOWN_SCRIPT = 'Unknown' UNKNOWN_SCRIPT = 'Unknown'
COMMON_SCRIPT = 'Common' COMMON_SCRIPT = 'Common'
MAX_ASCII = 127 MAX_ASCII = 127