Merge remote-tracking branch 'upstream/master'

[nominatim.git] / test / python / test_tokenizer_legacy.py

diff --git a/test/python/test_tokenizer_legacy.py b/test/python/test_tokenizer_legacy.py
index 801471723c4b72b5be4c4f6938c188b059639408..53d45c1c93a0f44f375ab11b675902497789b73e 100644 (file)
--- a/test/python/test_tokenizer_legacy.py
+++ b/test/python/test_tokenizer_legacy.py
@@ -5,6 +5,7 @@ import shutil
 
 import pytest
 
 
 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
 from nominatim.tokenizer import legacy_tokenizer
 from nominatim.db import properties
 from nominatim.errors import UsageError


@@ -46,7 +47,7 @@ def tokenizer_factory(dsn, tmp_path, property_table):
 
 @pytest.fixture
 def tokenizer_setup(tokenizer_factory, test_config, monkeypatch, sql_preprocessor):
 
 @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)
 
     tok = tokenizer_factory()
     tok.init_new_db(test_config)
 


@@ -60,7 +61,7 @@ def analyzer(tokenizer_factory, test_config, monkeypatch, sql_preprocessor,
           RETURNS INTEGER AS $$ SELECT 342; $$ LANGUAGE SQL;
         """)
 
           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)
     monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', ':: lower();')
     tok = tokenizer_factory()
     tok.init_new_db(test_config)


@@ -73,26 +74,16 @@ def analyzer(tokenizer_factory, test_config, monkeypatch, sql_preprocessor,
 @pytest.fixture
 def make_standard_name(temp_db_cursor):
     temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION make_standard_name(name TEXT)
 @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 $$ SELECT ' ' || name; $$ LANGUAGE SQL""")
+                              RETURNS TEXT AS $$ SELECT '#' || lower(name) || '#'; $$ LANGUAGE SQL""")

 
 
 @pytest.fixture
 
 
 @pytest.fixture

-def create_postcode_id(table_factory, temp_db_cursor):
-    table_factory('out_postcode_table', 'postcode TEXT')
-
+def create_postcode_id(temp_db_cursor):

     temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_postcode_id(postcode TEXT)
                               RETURNS BOOLEAN AS $$
     temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_postcode_id(postcode TEXT)
                               RETURNS BOOLEAN AS $$

-                              INSERT INTO out_postcode_table VALUES (postcode) RETURNING True;
-                              $$ LANGUAGE SQL""")
-
-
-@pytest.fixture
-def create_housenumbers(temp_db_cursor):
-    temp_db_cursor.execute("""CREATE OR REPLACE FUNCTION create_housenumbers(
-                                  housenumbers TEXT[],
-                                  OUT tokens TEXT, OUT normtext TEXT)
-                              AS $$
-                              SELECT housenumbers::TEXT, array_to_string(housenumbers, ';')
+                              INSERT INTO word (word_token, word, class, type)
+                                VALUES (' ' || postcode, postcode, 'place', 'postcode')
+                              RETURNING True;

                               $$ LANGUAGE SQL""")
 
 
                               $$ LANGUAGE SQL""")
 
 


@@ -105,7 +96,7 @@ def make_keywords(temp_db_cursor, temp_db_with_extensions):
 def test_init_new(tokenizer_factory, test_config, monkeypatch,
                   temp_db_conn, sql_preprocessor):
     monkeypatch.setenv('NOMINATIM_TERM_NORMALIZATION', 'xxvv')
 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)
 
     tok = tokenizer_factory()
     tok.init_new_db(test_config)


@@ -119,8 +110,7 @@ def test_init_new(tokenizer_factory, test_config, monkeypatch,
     assert outfile.stat().st_mode == 33261
 
 
     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):
     tok = tokenizer_factory()
 
     with pytest.raises(UsageError):


@@ -134,7 +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))
     (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)
 
     tok = tokenizer_factory()
     tok.init_new_db(test_config)


@@ -142,10 +132,10 @@ def test_init_module_custom(tokenizer_factory, test_config,
     assert not (test_config.project_dir / 'module').exists()
 
 
     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 = tokenizer_factory()
 

-    tok.init_from_project()
+    tok.init_from_project(test_config)

 
     assert tok.normalization is not None
 
 
     assert tok.normalization is not None
 


@@ -154,7 +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')
                               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()
     tok = tokenizer_factory()
     tok.init_new_db(test_config)
     monkeypatch.undo()


@@ -174,7 +164,7 @@ def test_update_sql_functions(sql_preprocessor, temp_db_conn,
 
 
 def test_migrate_database(tokenizer_factory, test_config, temp_db_conn, monkeypatch):
 
 
 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)
 
     tok = tokenizer_factory()
     tok.migrate_database(test_config)
 


@@ -192,121 +182,161 @@ def test_normalize(analyzer):
     assert analyzer.normalize('TEsT') == 'test'
 
 
     assert analyzer.normalize('TEsT') == 'test'
 
 

-def test_add_postcodes_from_db(analyzer, table_factory, temp_db_cursor,
-                               create_postcode_id):
+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',)))
 
     table_factory('location_postcode', 'postcode TEXT',
                   content=(('1234',), ('12 34',), ('AB23',), ('1234',)))
 

-    analyzer.add_postcodes_from_db()
+    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 temp_db_cursor.row_set("SELECT * from out_postcode_table") \
-               == set((('1234', ), ('12 34', ), ('AB23',)))
+    assert word_table.count() == 3
+    assert word_table.get_postcodes() == {'1234', '45BC', 'XX45'}

 
 
 
 

-def test_update_special_phrase_empty_table(analyzer, word_table, temp_db_cursor,
-                                           make_standard_name):
+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", "-"),
     analyzer.update_special_phrases([
         ("König bei", "amenity", "royal", "near"),
         ("Könige", "amenity", "royal", "-"),

+        ("könige", "amenity", "royal", "-"),

         ("strasse", "highway", "primary", "in")
     ], True)
 
         ("strasse", "highway", "primary", "in")
     ], True)
 

-    assert temp_db_cursor.row_set("""SELECT word_token, word, class, type, operator
-                                     FROM word WHERE class != 'place'""") \
-               == set(((' könig bei', 'könig bei', 'amenity', 'royal', 'near'),
-                       (' könige', 'könige', 'amenity', 'royal', None),
-                       (' strasse', 'strasse', 'highway', 'primary', 'in')))
+    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, temp_db_cursor,
-                                          make_standard_name):
-    temp_db_cursor.execute("""INSERT INTO word (word_token, word, class, type, operator)
-                              VALUES (' foo', 'foo', 'amenity', 'prison', 'in'),
-                                     (' bar', 'bar', 'highway', 'road', null)""")
+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 2 == temp_db_cursor.scalar("SELECT count(*) FROM word WHERE class != 'place'""")
+    assert word_table.count_special() == 2

 
     analyzer.update_special_phrases([], True)
 
 
     analyzer.update_special_phrases([], True)
 

-    assert 0 == temp_db_cursor.scalar("SELECT count(*) FROM word WHERE class != 'place'""")
+    assert word_table.count_special() == 0

 
 
 
 

-def test_update_special_phrases_no_replace(analyzer, word_table, temp_db_cursor,
-                                          make_standard_name):
-    temp_db_cursor.execute("""INSERT INTO word (word_token, word, class, type, operator)
-                              VALUES (' foo', 'foo', 'amenity', 'prison', 'in'),
-                                     (' bar', 'bar', 'highway', 'road', null)""")
+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 2 == temp_db_cursor.scalar("SELECT count(*) FROM word WHERE class != 'place'""")
+    assert word_table.count_special() == 2

 
     analyzer.update_special_phrases([], False)
 
 
     analyzer.update_special_phrases([], False)
 

-    assert 2 == temp_db_cursor.scalar("SELECT count(*) FROM word WHERE class != 'place'""")
+    assert word_table.count_special() == 2

 
 
 
 

-def test_update_special_phrase_modify(analyzer, word_table, temp_db_cursor,
-                                      make_standard_name):
-    temp_db_cursor.execute("""INSERT INTO word (word_token, word, class, type, operator)
-                              VALUES (' foo', 'foo', 'amenity', 'prison', 'in'),
-                                     (' bar', 'bar', 'highway', 'road', null)""")
+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 2 == temp_db_cursor.scalar("SELECT count(*) FROM word WHERE class != 'place'""")
+    assert word_table.count_special() == 2

 
     analyzer.update_special_phrases([
 
     analyzer.update_special_phrases([

-      ('prison', 'amenity', 'prison', 'in'),
-      ('bar', 'highway', 'road', '-'),
-      ('garden', 'leisure', 'garden', 'near')
+        ('prison', 'amenity', 'prison', 'in'),
+        ('bar', 'highway', 'road', '-'),
+        ('garden', 'leisure', 'garden', 'near')

     ], True)
 
     ], True)
 

-    assert temp_db_cursor.row_set("""SELECT word_token, word, class, type, operator
-                                     FROM word WHERE class != 'place'""") \
-               == set(((' prison', 'prison', 'amenity', 'prison', 'in'),
-                       (' bar', 'bar', 'highway', 'road', None),
-                       (' garden', 'garden', 'leisure', 'garden', 'near')))
+    assert word_table.get_special() \
+               == set(((' #prison#', 'prison', 'amenity', 'prison', 'in'),
+                       (' #bar#', 'bar', 'highway', 'road', None),
+                       (' #garden#', 'garden', 'leisure', 'garden', 'near')))

 
 
 
 

-def test_process_place_names(analyzer, make_keywords):
+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'})

 
 

-    info = analyzer.process_place({'name' : {'name' : 'Soft bAr', 'ref': '34'}})
+    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}'
 
 
 
     assert info['names'] == '{1,2,3}'
 
 

-@pytest.mark.parametrize('pc', ['12345', 'AB 123', '34-345'])
-def test_process_place_postcode(analyzer, temp_db_cursor, create_postcode_id, pc):
+@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, }

 
 

-    info = analyzer.process_place({'address': {'postcode' : pc}})

 
 

-    assert temp_db_cursor.row_set("SELECT * from out_postcode_table") \
-               == set(((pc, ),))
+@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()

 
 

-@pytest.mark.parametrize('pc', ['12:23', 'ab;cd;f', '123;836'])
-def test_process_place_bad_postcode(analyzer, temp_db_cursor, create_postcode_id,
-                                    pc):

 
 

-    info = analyzer.process_place({'address': {'postcode' : pc}})
+class TestHousenumberName:

 
 

-    assert 0 == temp_db_cursor.scalar("SELECT count(*) from out_postcode_table")
+    @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""")

 
 
 
 

-@pytest.mark.parametrize('hnr', ['123a', '1', '101'])
-def test_process_place_housenumbers_simple(analyzer, create_housenumbers, hnr):
-    info = analyzer.process_place({'address': {'housenumber' : hnr}})
+    @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("{")
+        assert info['hnr'] == hnr
+        assert info['hnr_tokens'].startswith("{")

 
 
 
 

-def test_process_place_housenumbers_lists(analyzer, create_housenumbers):
-    info = analyzer.process_place({'address': {'conscriptionnumber' : '1; 2;3'}})
+    @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'))
+        assert set(info['hnr'].split(';')) == set(('1', '2', '3'))

 
 
 
 

-def test_process_place_housenumbers_duplicates(analyzer, create_housenumbers):
-    info = analyzer.process_place({'address': {'housenumber' : '134',
-                                               'conscriptionnumber' : '134',
-                                               'streetnumber' : '99a'}})
+    @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'))
+        assert set(info['hnr'].split(';')) == set(('134', '99a'))