2 Test for legacy tokenizer.
8 from nominatim.indexer.place_info import PlaceInfo
9 from nominatim.tokenizer import legacy_tokenizer
10 from nominatim.db import properties
11 from nominatim.errors import UsageError
14 def test_config(def_config, tmp_path):
15 def_config.project_dir = tmp_path / 'project'
16 def_config.project_dir.mkdir()
18 module_dir = tmp_path / 'module_src'
20 (module_dir / 'nominatim.so').write_text('TEST nomiantim.so')
22 def_config.lib_dir.module = module_dir
24 sqldir = tmp_path / 'sql'
26 (sqldir / 'tokenizer').mkdir()
27 (sqldir / 'tokenizer' / 'legacy_tokenizer.sql').write_text("SELECT 'a'")
28 (sqldir / 'words.sql').write_text("SELECT 'a'")
29 shutil.copy(str(def_config.lib_dir.sql / 'tokenizer' / 'legacy_tokenizer_tables.sql'),
30 str(sqldir / 'tokenizer' / 'legacy_tokenizer_tables.sql'))
32 def_config.lib_dir.sql = sqldir
33 def_config.lib_dir.data = sqldir
39 def tokenizer_factory(dsn, tmp_path, property_table):
40 (tmp_path / 'tokenizer').mkdir()
43 return legacy_tokenizer.create(dsn, tmp_path / 'tokenizer')
49 def tokenizer_setup(tokenizer_factory, test_config, monkeypatch, sql_preprocessor):
50 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
51 tok = tokenizer_factory()
52 tok.init_new_db(test_config)
56 def analyzer(tokenizer_factory, test_config, monkeypatch, sql_preprocessor,
57 word_table, temp_db_with_extensions, tmp_path):
58 sql = tmp_path / 'sql' / 'tokenizer' / 'legacy_tokenizer.sql'
60 CREATE OR REPLACE FUNCTION getorcreate_housenumber_id(lookup_word TEXT)
61 RETURNS INTEGER AS $$ SELECT 342; $$ LANGUAGE SQL;
64 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
65 monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', ':: lower();')
66 tok = tokenizer_factory()
67 tok.init_new_db(test_config)
70 with tok.name_analyzer() as analyzer:
75 def make_standard_name(temp_db_cursor):
76 temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION make_standard_name(name TEXT)
77 RETURNS TEXT AS $$ SELECT '#' || lower(name) || '#'; $$ LANGUAGE SQL""")
81 def create_postcode_id(temp_db_cursor):
82 temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_postcode_id(postcode TEXT)
84 INSERT INTO word (word_token, word, class, type)
85 VALUES (' ' || postcode, postcode, 'place', 'postcode')
91 def make_keywords(temp_db_cursor, temp_db_with_extensions):
92 temp_db_cursor.execute(
93 """CREATE OR REPLACE FUNCTION make_keywords(names HSTORE)
94 RETURNS INTEGER[] AS $$ SELECT ARRAY[1, 2, 3] $$ LANGUAGE SQL""")
96 def test_init_new(tokenizer_factory, test_config, monkeypatch,
97 temp_db_conn, sql_preprocessor):
98 monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', 'xxvv')
99 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
101 tok = tokenizer_factory()
102 tok.init_new_db(test_config)
104 assert properties.get_property(temp_db_conn, legacy_tokenizer.DBCFG_NORMALIZATION) == 'xxvv'
106 outfile = test_config.project_dir / 'module' / 'nominatim.so'
108 assert outfile.exists()
109 assert outfile.read_text() == 'TEST nomiantim.so'
110 assert outfile.stat().st_mode == 33261
113 def test_init_module_load_failed(tokenizer_factory, test_config):
114 tok = tokenizer_factory()
116 with pytest.raises(UsageError):
117 tok.init_new_db(test_config)
120 def test_init_module_custom(tokenizer_factory, test_config,
121 monkeypatch, tmp_path, sql_preprocessor):
122 module_dir = (tmp_path / 'custom').resolve()
124 (module_dir/ 'nominatim.so').write_text('CUSTOM nomiantim.so')
126 monkeypatch.setenv('NOMINATIM_DATABASE_MODULE_PATH', str(module_dir))
127 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
129 tok = tokenizer_factory()
130 tok.init_new_db(test_config)
132 assert not (test_config.project_dir / 'module').exists()
135 def test_init_from_project(tokenizer_setup, tokenizer_factory, test_config):
136 tok = tokenizer_factory()
138 tok.init_from_project(test_config)
140 assert tok.normalization is not None
143 def test_update_sql_functions(sql_preprocessor, temp_db_conn,
144 tokenizer_factory, test_config, table_factory,
145 monkeypatch, temp_db_cursor):
146 monkeypatch.setenv('NOMINATIM_MAX_WORD_FREQUENCY', '1133')
147 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
148 tok = tokenizer_factory()
149 tok.init_new_db(test_config)
152 assert properties.get_property(temp_db_conn, legacy_tokenizer.DBCFG_MAXWORDFREQ) == '1133'
154 table_factory('test', 'txt TEXT')
156 func_file = test_config.lib_dir.sql / 'tokenizer' / 'legacy_tokenizer.sql'
157 func_file.write_text("""INSERT INTO test VALUES ('{{max_word_freq}}'),
158 ('{{modulepath}}')""")
160 tok.update_sql_functions(test_config)
162 test_content = temp_db_cursor.row_set('SELECT * FROM test')
163 assert test_content == set((('1133', ), (str(test_config.project_dir / 'module'), )))
166 def test_migrate_database(tokenizer_factory, test_config, temp_db_conn, monkeypatch):
167 monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None)
168 tok = tokenizer_factory()
169 tok.migrate_database(test_config)
171 assert properties.get_property(temp_db_conn, legacy_tokenizer.DBCFG_MAXWORDFREQ) is not None
172 assert properties.get_property(temp_db_conn, legacy_tokenizer.DBCFG_NORMALIZATION) is not None
174 outfile = test_config.project_dir / 'module' / 'nominatim.so'
176 assert outfile.exists()
177 assert outfile.read_text() == 'TEST nomiantim.so'
178 assert outfile.stat().st_mode == 33261
181 def test_normalize(analyzer):
182 assert analyzer.normalize('TEsT') == 'test'
185 def test_update_postcodes_from_db_empty(analyzer, table_factory, word_table,
187 table_factory('location_postcode', 'postcode TEXT',
188 content=(('1234',), ('12 34',), ('AB23',), ('1234',)))
190 analyzer.update_postcodes_from_db()
192 assert word_table.count() == 3
193 assert word_table.get_postcodes() == {'1234', '12 34', 'AB23'}
196 def test_update_postcodes_from_db_add_and_remove(analyzer, table_factory, word_table,
198 table_factory('location_postcode', 'postcode TEXT',
199 content=(('1234',), ('45BC', ), ('XX45', )))
200 word_table.add_postcode(' 1234', '1234')
201 word_table.add_postcode(' 5678', '5678')
203 analyzer.update_postcodes_from_db()
205 assert word_table.count() == 3
206 assert word_table.get_postcodes() == {'1234', '45BC', 'XX45'}
209 def test_update_special_phrase_empty_table(analyzer, word_table, make_standard_name):
210 analyzer.update_special_phrases([
211 ("König bei", "amenity", "royal", "near"),
212 ("Könige", "amenity", "royal", "-"),
213 ("könige", "amenity", "royal", "-"),
214 ("strasse", "highway", "primary", "in")
217 assert word_table.get_special() \
218 == set(((' #könig bei#', 'könig bei', 'amenity', 'royal', 'near'),
219 (' #könige#', 'könige', 'amenity', 'royal', None),
220 (' #strasse#', 'strasse', 'highway', 'primary', 'in')))
223 def test_update_special_phrase_delete_all(analyzer, word_table, make_standard_name):
224 word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in')
225 word_table.add_special(' #bar#', 'bar', 'highway', 'road', None)
227 assert word_table.count_special() == 2
229 analyzer.update_special_phrases([], True)
231 assert word_table.count_special() == 0
234 def test_update_special_phrases_no_replace(analyzer, word_table, make_standard_name):
235 word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in')
236 word_table.add_special(' #bar#', 'bar', 'highway', 'road', None)
238 assert word_table.count_special() == 2
240 analyzer.update_special_phrases([], False)
242 assert word_table.count_special() == 2
245 def test_update_special_phrase_modify(analyzer, word_table, make_standard_name):
246 word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in')
247 word_table.add_special(' #bar#', 'bar', 'highway', 'road', None)
249 assert word_table.count_special() == 2
251 analyzer.update_special_phrases([
252 ('prison', 'amenity', 'prison', 'in'),
253 ('bar', 'highway', 'road', '-'),
254 ('garden', 'leisure', 'garden', 'near')
257 assert word_table.get_special() \
258 == set(((' #prison#', 'prison', 'amenity', 'prison', 'in'),
259 (' #bar#', 'bar', 'highway', 'road', None),
260 (' #garden#', 'garden', 'leisure', 'garden', 'near')))
263 def test_add_country_names(analyzer, word_table, make_standard_name):
264 analyzer.add_country_names('de', {'name': 'Germany',
265 'name:de': 'Deutschland',
266 'short_name': 'germany'})
268 assert word_table.get_country() \
269 == {('de', ' #germany#'),
270 ('de', ' #deutschland#')}
273 def test_add_more_country_names(analyzer, word_table, make_standard_name):
274 word_table.add_country('fr', ' #france#')
275 word_table.add_country('it', ' #italy#')
276 word_table.add_country('it', ' #itala#')
278 analyzer.add_country_names('it', {'name': 'Italy', 'ref': 'IT'})
280 assert word_table.get_country() \
281 == {('fr', ' #france#'),
287 def test_process_place_names(analyzer, make_keywords):
288 info = analyzer.process_place(PlaceInfo({'name' : {'name' : 'Soft bAr', 'ref': '34'}}))
290 assert info['names'] == '{1,2,3}'
293 @pytest.mark.parametrize('pcode', ['12345', 'AB 123', '34-345'])
294 def test_process_place_postcode(analyzer, create_postcode_id, word_table, pcode):
295 analyzer.process_place(PlaceInfo({'address': {'postcode' : pcode}}))
297 assert word_table.get_postcodes() == {pcode, }
300 @pytest.mark.parametrize('pcode', ['12:23', 'ab;cd;f', '123;836'])
301 def test_process_place_bad_postcode(analyzer, create_postcode_id, word_table, pcode):
302 analyzer.process_place(PlaceInfo({'address': {'postcode' : pcode}}))
304 assert not word_table.get_postcodes()
307 class TestHousenumberName:
310 @pytest.fixture(autouse=True)
311 def setup_create_housenumbers(temp_db_cursor):
312 temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_housenumbers(
314 OUT tokens TEXT, OUT normtext TEXT)
316 SELECT housenumbers::TEXT, array_to_string(housenumbers, ';')
321 @pytest.mark.parametrize('hnr', ['123a', '1', '101'])
322 def test_process_place_housenumbers_simple(analyzer, hnr):
323 info = analyzer.process_place(PlaceInfo({'address': {'housenumber' : hnr}}))
325 assert info['hnr'] == hnr
326 assert info['hnr_tokens'].startswith("{")
330 def test_process_place_housenumbers_lists(analyzer):
331 info = analyzer.process_place(PlaceInfo({'address': {'conscriptionnumber' : '1; 2;3'}}))
333 assert set(info['hnr'].split(';')) == set(('1', '2', '3'))
337 def test_process_place_housenumbers_duplicates(analyzer):
338 info = analyzer.process_place(PlaceInfo({'address': {'housenumber' : '134',
339 'conscriptionnumber' : '134',
340 'streetnumber' : '99a'}}))
342 assert set(info['hnr'].split(';')) == set(('134', '99a'))