X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/3eb4d8805700ba12bd601e552c3bc48064423083..8c54a274e7180b0e06f9581bc16305f87c8ae412:/test/python/test_tokenizer_legacy.py diff --git a/test/python/test_tokenizer_legacy.py b/test/python/test_tokenizer_legacy.py index 45c400fe..53d45c1c 100644 --- a/test/python/test_tokenizer_legacy.py +++ b/test/python/test_tokenizer_legacy.py @@ -5,6 +5,7 @@ import shutil import pytest +from nominatim.indexer.place_info import PlaceInfo from nominatim.tokenizer import legacy_tokenizer from nominatim.db import properties from nominatim.errors import UsageError @@ -35,7 +36,7 @@ def test_config(def_config, tmp_path): @pytest.fixture -def tokenizer_factory(dsn, tmp_path, monkeypatch, property_table): +def tokenizer_factory(dsn, tmp_path, property_table): (tmp_path / 'tokenizer').mkdir() def _maker(): @@ -43,9 +44,10 @@ def tokenizer_factory(dsn, tmp_path, monkeypatch, property_table): return _maker + @pytest.fixture def tokenizer_setup(tokenizer_factory, test_config, monkeypatch, sql_preprocessor): - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) tok = tokenizer_factory() tok.init_new_db(test_config) @@ -59,7 +61,7 @@ def analyzer(tokenizer_factory, test_config, monkeypatch, sql_preprocessor, RETURNS INTEGER AS $$ SELECT 342; $$ LANGUAGE SQL; """) - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', ':: lower();') tok = tokenizer_factory() tok.init_new_db(test_config) @@ -69,10 +71,32 @@ def analyzer(tokenizer_factory, test_config, monkeypatch, sql_preprocessor, yield analyzer +@pytest.fixture +def make_standard_name(temp_db_cursor): + temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION make_standard_name(name TEXT) + RETURNS TEXT AS $$ SELECT '#' || lower(name) || '#'; $$ LANGUAGE SQL""") + + +@pytest.fixture +def create_postcode_id(temp_db_cursor): + temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_postcode_id(postcode TEXT) + RETURNS BOOLEAN AS $$ + INSERT INTO word (word_token, word, class, type) + VALUES (' ' || postcode, postcode, 'place', 'postcode') + RETURNING True; + $$ LANGUAGE SQL""") + + +@pytest.fixture +def make_keywords(temp_db_cursor, temp_db_with_extensions): + temp_db_cursor.execute( + """CREATE OR REPLACE FUNCTION make_keywords(names HSTORE) + RETURNS INTEGER[] AS $$ SELECT ARRAY[1, 2, 3] $$ LANGUAGE SQL""") + def test_init_new(tokenizer_factory, test_config, monkeypatch, temp_db_conn, sql_preprocessor): monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', 'xxvv') - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) tok = tokenizer_factory() tok.init_new_db(test_config) @@ -86,8 +110,7 @@ def test_init_new(tokenizer_factory, test_config, monkeypatch, assert outfile.stat().st_mode == 33261 -def test_init_module_load_failed(tokenizer_factory, test_config, - monkeypatch, temp_db_conn): +def test_init_module_load_failed(tokenizer_factory, test_config): tok = tokenizer_factory() with pytest.raises(UsageError): @@ -101,7 +124,7 @@ def test_init_module_custom(tokenizer_factory, test_config, (module_dir/ 'nominatim.so').write_text('CUSTOM nomiantim.so') monkeypatch.setenv('NOMINATIM_DATABASE_MODULE_PATH', str(module_dir)) - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) tok = tokenizer_factory() tok.init_new_db(test_config) @@ -109,10 +132,10 @@ def test_init_module_custom(tokenizer_factory, test_config, assert not (test_config.project_dir / 'module').exists() -def test_init_from_project(tokenizer_setup, tokenizer_factory): +def test_init_from_project(tokenizer_setup, tokenizer_factory, test_config): tok = tokenizer_factory() - tok.init_from_project() + tok.init_from_project(test_config) assert tok.normalization is not None @@ -121,7 +144,7 @@ def test_update_sql_functions(sql_preprocessor, temp_db_conn, tokenizer_factory, test_config, table_factory, monkeypatch, temp_db_cursor): monkeypatch.setenv('NOMINATIM_MAX_WORD_FREQUENCY', '1133') - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) tok = tokenizer_factory() tok.init_new_db(test_config) monkeypatch.undo() @@ -141,7 +164,7 @@ def test_update_sql_functions(sql_preprocessor, temp_db_conn, def test_migrate_database(tokenizer_factory, test_config, temp_db_conn, monkeypatch): - monkeypatch.setattr(legacy_tokenizer, '_check_module' , lambda m, c: None) + monkeypatch.setattr(legacy_tokenizer, '_check_module', lambda m, c: None) tok = tokenizer_factory() tok.migrate_database(test_config) @@ -157,3 +180,163 @@ def test_migrate_database(tokenizer_factory, test_config, temp_db_conn, monkeypa def test_normalize(analyzer): assert analyzer.normalize('TEsT') == 'test' + + +def test_update_postcodes_from_db_empty(analyzer, table_factory, word_table, + create_postcode_id): + table_factory('location_postcode', 'postcode TEXT', + content=(('1234',), ('12 34',), ('AB23',), ('1234',))) + + analyzer.update_postcodes_from_db() + + assert word_table.count() == 3 + assert word_table.get_postcodes() == {'1234', '12 34', 'AB23'} + + +def test_update_postcodes_from_db_add_and_remove(analyzer, table_factory, word_table, + create_postcode_id): + table_factory('location_postcode', 'postcode TEXT', + content=(('1234',), ('45BC', ), ('XX45', ))) + word_table.add_postcode(' 1234', '1234') + word_table.add_postcode(' 5678', '5678') + + analyzer.update_postcodes_from_db() + + assert word_table.count() == 3 + assert word_table.get_postcodes() == {'1234', '45BC', 'XX45'} + + +def test_update_special_phrase_empty_table(analyzer, word_table, make_standard_name): + analyzer.update_special_phrases([ + ("König bei", "amenity", "royal", "near"), + ("Könige", "amenity", "royal", "-"), + ("könige", "amenity", "royal", "-"), + ("strasse", "highway", "primary", "in") + ], True) + + assert word_table.get_special() \ + == set(((' #könig bei#', 'könig bei', 'amenity', 'royal', 'near'), + (' #könige#', 'könige', 'amenity', 'royal', None), + (' #strasse#', 'strasse', 'highway', 'primary', 'in'))) + + +def test_update_special_phrase_delete_all(analyzer, word_table, make_standard_name): + word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in') + word_table.add_special(' #bar#', 'bar', 'highway', 'road', None) + + assert word_table.count_special() == 2 + + analyzer.update_special_phrases([], True) + + assert word_table.count_special() == 0 + + +def test_update_special_phrases_no_replace(analyzer, word_table, make_standard_name): + word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in') + word_table.add_special(' #bar#', 'bar', 'highway', 'road', None) + + assert word_table.count_special() == 2 + + analyzer.update_special_phrases([], False) + + assert word_table.count_special() == 2 + + +def test_update_special_phrase_modify(analyzer, word_table, make_standard_name): + word_table.add_special(' #foo#', 'foo', 'amenity', 'prison', 'in') + word_table.add_special(' #bar#', 'bar', 'highway', 'road', None) + + assert word_table.count_special() == 2 + + analyzer.update_special_phrases([ + ('prison', 'amenity', 'prison', 'in'), + ('bar', 'highway', 'road', '-'), + ('garden', 'leisure', 'garden', 'near') + ], True) + + assert word_table.get_special() \ + == set(((' #prison#', 'prison', 'amenity', 'prison', 'in'), + (' #bar#', 'bar', 'highway', 'road', None), + (' #garden#', 'garden', 'leisure', 'garden', 'near'))) + + +def test_add_country_names(analyzer, word_table, make_standard_name): + analyzer.add_country_names('de', {'name': 'Germany', + 'name:de': 'Deutschland', + 'short_name': 'germany'}) + + assert word_table.get_country() \ + == {('de', ' #germany#'), + ('de', ' #deutschland#')} + + +def test_add_more_country_names(analyzer, word_table, make_standard_name): + word_table.add_country('fr', ' #france#') + word_table.add_country('it', ' #italy#') + word_table.add_country('it', ' #itala#') + + analyzer.add_country_names('it', {'name': 'Italy', 'ref': 'IT'}) + + assert word_table.get_country() \ + == {('fr', ' #france#'), + ('it', ' #italy#'), + ('it', ' #itala#'), + ('it', ' #it#')} + + +def test_process_place_names(analyzer, make_keywords): + info = analyzer.process_place(PlaceInfo({'name' : {'name' : 'Soft bAr', 'ref': '34'}})) + + assert info['names'] == '{1,2,3}' + + +@pytest.mark.parametrize('pcode', ['12345', 'AB 123', '34-345']) +def test_process_place_postcode(analyzer, create_postcode_id, word_table, pcode): + analyzer.process_place(PlaceInfo({'address': {'postcode' : pcode}})) + + assert word_table.get_postcodes() == {pcode, } + + +@pytest.mark.parametrize('pcode', ['12:23', 'ab;cd;f', '123;836']) +def test_process_place_bad_postcode(analyzer, create_postcode_id, word_table, pcode): + analyzer.process_place(PlaceInfo({'address': {'postcode' : pcode}})) + + assert not word_table.get_postcodes() + + +class TestHousenumberName: + + @staticmethod + @pytest.fixture(autouse=True) + def setup_create_housenumbers(temp_db_cursor): + temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_housenumbers( + housenumbers TEXT[], + OUT tokens TEXT, OUT normtext TEXT) + AS $$ + SELECT housenumbers::TEXT, array_to_string(housenumbers, ';') + $$ LANGUAGE SQL""") + + + @staticmethod + @pytest.mark.parametrize('hnr', ['123a', '1', '101']) + def test_process_place_housenumbers_simple(analyzer, hnr): + info = analyzer.process_place(PlaceInfo({'address': {'housenumber' : hnr}})) + + assert info['hnr'] == hnr + assert info['hnr_tokens'].startswith("{") + + + @staticmethod + def test_process_place_housenumbers_lists(analyzer): + info = analyzer.process_place(PlaceInfo({'address': {'conscriptionnumber' : '1; 2;3'}})) + + assert set(info['hnr'].split(';')) == set(('1', '2', '3')) + + + @staticmethod + def test_process_place_housenumbers_duplicates(analyzer): + info = analyzer.process_place(PlaceInfo({'address': {'housenumber' : '134', + 'conscriptionnumber' : '134', + 'streetnumber' : '99a'}})) + + assert set(info['hnr'].split(';')) == set(('134', '99a'))