move index creation for word table to tokenizer

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

diff --git a/test/python/test_tools_database_import.py b/test/python/test_tools_database_import.py
index 597fdfc126f0838f75b8690d04b6fb5ff11b411b..ceac7a2421dc43afd23ecc1cf3a4c3066c7b02f1 100644 (file)
--- a/test/python/test_tools_database_import.py
+++ b/test/python/test_tools_database_import.py
@@ -24,6 +24,24 @@ def nonexistant_db():
     with conn.cursor() as cur:
         cur.execute('DROP DATABASE IF EXISTS {}'.format(dbname))
 
+@pytest.mark.parametrize("no_partitions", (True, False))
+def test_setup_skeleton(src_dir, nonexistant_db, no_partitions):
+    database_import.setup_database_skeleton('dbname=' + nonexistant_db,
+                                            src_dir / 'data', no_partitions)
+
+    conn = psycopg2.connect(database=nonexistant_db)
+
+    try:
+        with conn.cursor() as cur:
+            cur.execute("SELECT distinct partition FROM country_name")
+            partitions = set([r[0] for r in list(cur)])
+            if no_partitions:
+                assert partitions == set([0])
+            else:
+                assert len(partitions) > 10
+    finally:
+        conn.close()
+
 
 def test_create_db_success(nonexistant_db):
     database_import.create_db('dbname=' + nonexistant_db, rouser='www-data')
@@ -62,23 +80,6 @@ def test_setup_extensions_old_postgis(temp_db_conn, monkeypatch):
         database_import.setup_extensions(temp_db_conn)
 
 
-def test_install_module(tmp_path):
-    src_dir = tmp_path / 'source'
-    src_dir.mkdir()
-    (src_dir / 'nominatim.so').write_text('TEST nomiantim.so')
-
-    project_dir = tmp_path / 'project'
-    project_dir.mkdir()
-
-    database_import.install_module(src_dir, project_dir, '')
-
-    outfile = project_dir / 'module' / 'nominatim.so'
-
-    assert outfile.exists()
-    assert outfile.read_text() == 'TEST nomiantim.so'
-    assert outfile.stat().st_mode == 33261
-
-
 def test_import_base_data(src_dir, temp_db, temp_db_cursor):
     temp_db_cursor.execute('CREATE EXTENSION hstore')
     temp_db_cursor.execute('CREATE EXTENSION postgis')
@@ -134,3 +135,62 @@ def test_import_osm_data_default_cache(temp_db_cursor,osm2pgsql_options):
     osm2pgsql_options['osm2pgsql_cache'] = 0
 
     database_import.import_osm_data(Path(__file__), osm2pgsql_options)
+
+
+def test_truncate_database_tables(temp_db_conn, temp_db_cursor, table_factory):
+    tables = ('placex', 'place_addressline', 'location_area',
+              'location_area_country',
+              'location_property_tiger', 'location_property_osmline',
+              'location_postcode', 'search_name', 'location_road_23')
+    for table in tables:
+        table_factory(table, content=((1, ), (2, ), (3, )))
+        assert temp_db_cursor.table_rows(table) == 3
+
+    database_import.truncate_data_tables(temp_db_conn)
+
+    for table in tables:
+        assert temp_db_cursor.table_rows(table) == 0
+
+
+@pytest.mark.parametrize("threads", (1, 5))
+def test_load_data(dsn, src_dir, place_row, placex_table, osmline_table, word_table,
+                   temp_db_cursor, threads):
+    for func in ('precompute_words', 'getorcreate_housenumber_id', 'make_standard_name'):
+        temp_db_cursor.execute("""CREATE FUNCTION {} (src TEXT)
+                                  RETURNS TEXT AS $$ SELECT 'a'::TEXT $$ LANGUAGE SQL
+                               """.format(func))
+    for oid in range(100, 130):
+        place_row(osm_id=oid)
+    place_row(osm_type='W', osm_id=342, cls='place', typ='houses',
+              geom='SRID=4326;LINESTRING(0 0, 10 10)')
+
+    database_import.load_data(dsn, threads)
+
+    assert temp_db_cursor.table_rows('placex') == 30
+    assert temp_db_cursor.table_rows('location_property_osmline') == 1
+
+
+@pytest.mark.parametrize("languages", (None, ' fr,en'))
+def test_create_country_names(temp_db_with_extensions, temp_db_conn, temp_db_cursor,
+                              table_factory, tokenizer_mock, languages):
+
+    table_factory('country_name', 'country_code varchar(2), name hstore',
+                  content=(('us', '"name"=>"us1","name:af"=>"us2"'),
+                           ('fr', '"name"=>"Fra", "name:en"=>"Fren"')))
+
+    assert temp_db_cursor.scalar("SELECT count(*) FROM country_name") == 2
+
+    tokenizer = tokenizer_mock()
+
+    database_import.create_country_names(temp_db_conn, tokenizer, languages)
+
+    assert len(tokenizer.analyser_cache['countries']) == 2
+
+    result_set = {k: set(v) for k, v in tokenizer.analyser_cache['countries']}
+
+    if languages:
+        assert result_set == {'us' : set(('us', 'us1', 'United States')),
+                              'fr' : set(('fr', 'Fra', 'Fren'))}
+    else:
+        assert result_set == {'us' : set(('us', 'us1', 'us2', 'United States')),
+                              'fr' : set(('fr', 'Fra', 'Fren'))}