Merge remote-tracking branch 'upstream/master'

[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 a4ab16f99cf1eb7e5c1df7154b03649b514fd98f..eda88903a9733598455b5c013804b929e5f2731f 100644 (file)
--- a/test/python/test_tools_database_import.py
+++ b/test/python/test_tools_database_import.py
@@ -1,96 +1,158 @@
 """
 Tests for functions to import a new database.
 """
 """
 Tests for functions to import a new database.
 """

+from pathlib import Path
+from contextlib import closing
+

 import pytest
 import psycopg2
 import pytest
 import psycopg2

-import sys

 
 from nominatim.tools import database_import
 from nominatim.errors import UsageError
 
 
 from nominatim.tools import database_import
 from nominatim.errors import UsageError
 

-@pytest.fixture
-def nonexistant_db():
-    dbname = 'test_nominatim_python_unittest'
+class TestDatabaseSetup:
+    DBNAME = 'test_nominatim_python_unittest'
+
+    @pytest.fixture(autouse=True)
+    def setup_nonexistant_db(self):
+        conn = psycopg2.connect(database='postgres')
+
+        try:
+            conn.set_isolation_level(0)
+            with conn.cursor() as cur:
+                cur.execute(f'DROP DATABASE IF EXISTS {self.DBNAME}')
+
+            yield True
+
+            with conn.cursor() as cur:
+                cur.execute(f'DROP DATABASE IF EXISTS {self.DBNAME}')
+        finally:
+            conn.close()
+
+    @pytest.fixture
+    def cursor(self):
+        conn = psycopg2.connect(database=self.DBNAME)
+
+        try:
+            with conn.cursor() as cur:
+                yield cur
+        finally:
+            conn.close()
+
+
+    def conn(self):
+        return closing(psycopg2.connect(database=self.DBNAME))

 
 

-    conn = psycopg2.connect(database='postgres')

 
 

-    conn.set_isolation_level(0)
-    with conn.cursor() as cur:
-        cur.execute('DROP DATABASE IF EXISTS {}'.format(dbname))
+    def test_setup_skeleton(self):
+        database_import.setup_database_skeleton(f'dbname={self.DBNAME}')

 
 

-    yield dbname
+        # Check that all extensions are set up.
+        with self.conn() as conn:
+            with conn.cursor() as cur:
+                cur.execute('CREATE TABLE t (h HSTORE, geom GEOMETRY(Geometry, 4326))')

 
 

-    with conn.cursor() as cur:
-        cur.execute('DROP DATABASE IF EXISTS {}'.format(dbname))

 
 

+    def test_unsupported_pg_version(self, monkeypatch):
+        monkeypatch.setattr(database_import, 'POSTGRESQL_REQUIRED_VERSION', (100, 4))

 
 

-def test_create_db_success(nonexistant_db):
-    database_import.create_db('dbname=' + nonexistant_db, rouser='www-data')
+        with pytest.raises(UsageError, match='PostgreSQL server is too old.'):
+            database_import.setup_database_skeleton(f'dbname={self.DBNAME}')

 
 

-    conn = psycopg2.connect(database=nonexistant_db)
-    conn.close()

 
 

+    def test_create_db_missing_ro_user(self):
+        with pytest.raises(UsageError, match='Missing read-only user.'):
+            database_import.setup_database_skeleton(f'dbname={self.DBNAME}',
+                                                    rouser='sdfwkjkjgdugu2;jgsafkljas;')

 
 

-def test_create_db_already_exists(temp_db):
+
+    def test_setup_extensions_old_postgis(self, monkeypatch):
+        monkeypatch.setattr(database_import, 'POSTGIS_REQUIRED_VERSION', (50, 50))
+
+        with pytest.raises(UsageError, match='PostGIS is too old.'):
+            database_import.setup_database_skeleton(f'dbname={self.DBNAME}')
+
+
+def test_setup_skeleton_already_exists(temp_db):

     with pytest.raises(UsageError):
     with pytest.raises(UsageError):

-        database_import.create_db('dbname=' + temp_db)
+        database_import.setup_database_skeleton(f'dbname={temp_db}')
+
+
+def test_import_osm_data_simple(table_factory, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))
+
+    database_import.import_osm_data(Path('file.pbf'), osm2pgsql_options)
+
+
+def test_import_osm_data_multifile(table_factory, tmp_path, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))
+    osm2pgsql_options['osm2pgsql_cache'] = 0

 
 

+    files = [tmp_path / 'file1.osm', tmp_path / 'file2.osm']
+    for f in files:
+        f.write_text('test')

 
 

-def test_create_db_unsupported_version(nonexistant_db, monkeypatch):
-    monkeypatch.setattr(database_import, 'POSTGRESQL_REQUIRED_VERSION', (100, 4))
+    database_import.import_osm_data(files, osm2pgsql_options)

 
 

-    with pytest.raises(UsageError, match='PostgreSQL server is too old.'):
-        database_import.create_db('dbname=' + nonexistant_db)

 
 

+def test_import_osm_data_simple_no_data(table_factory, osm2pgsql_options):
+    table_factory('place')

 
 

-def test_create_db_missing_ro_user(nonexistant_db):
-    with pytest.raises(UsageError, match='Missing read-only user.'):
-        database_import.create_db('dbname=' + nonexistant_db, rouser='sdfwkjkjgdugu2;jgsafkljas;')
+    with pytest.raises(UsageError, match='No data.*'):
+        database_import.import_osm_data(Path('file.pbf'), osm2pgsql_options)

 
 
 
 

-def test_setup_extensions(temp_db_conn, temp_db_cursor):
-    database_import.setup_extensions(temp_db_conn)
+def test_import_osm_data_drop(table_factory, temp_db_conn, tmp_path, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))
+    table_factory('planet_osm_nodes')

 
 

-    temp_db_cursor.execute('CREATE TABLE t (h HSTORE, geom GEOMETRY(Geometry, 4326))')
+    flatfile = tmp_path / 'flatfile'
+    flatfile.write_text('touch')

 
 

+    osm2pgsql_options['flatnode_file'] = str(flatfile.resolve())

 
 

-def test_setup_extensions_old_postgis(temp_db_conn, monkeypatch):
-    monkeypatch.setattr(database_import, 'POSTGIS_REQUIRED_VERSION', (50, 50))
+    database_import.import_osm_data(Path('file.pbf'), osm2pgsql_options, drop=True)

 
 

-    with pytest.raises(UsageError, match='PostGIS version is too old.'):
-        database_import.setup_extensions(temp_db_conn)
+    assert not flatfile.exists()
+    assert not temp_db_conn.table_exists('planet_osm_nodes')

 
 
 
 

-def test_install_module(tmp_path):
-    src_dir = tmp_path / 'source'
-    src_dir.mkdir()
-    (src_dir / 'nominatim.so').write_text('TEST nomiantim.so')
+def test_import_osm_data_default_cache(table_factory, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))

 
 

-    project_dir = tmp_path / 'project'
-    project_dir.mkdir()
+    osm2pgsql_options['osm2pgsql_cache'] = 0

 
 

-    database_import.install_module(src_dir, project_dir, '')
+    database_import.import_osm_data(Path(__file__), osm2pgsql_options)

 
 

-    outfile = project_dir / 'module' / 'nominatim.so'

 
 

-    assert outfile.exists()
-    assert outfile.read_text() == 'TEST nomiantim.so'
-    assert outfile.stat().st_mode == 33261
+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)

 
 

-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')
-    database_import.import_base_data('dbname=' + temp_db, src_dir / 'data')
+    for table in tables:
+        assert temp_db_cursor.table_rows(table) == 0

 
 

-    assert temp_db_cursor.scalar('SELECT count(*) FROM country_name') > 0

 
 

+@pytest.mark.parametrize("threads", (1, 5))
+def test_load_data(dsn, 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)')

 
 

-def test_import_base_data_ignore_partitions(src_dir, temp_db, temp_db_cursor):
-    temp_db_cursor.execute('CREATE EXTENSION hstore')
-    temp_db_cursor.execute('CREATE EXTENSION postgis')
-    database_import.import_base_data('dbname=' + temp_db, src_dir / 'data',
-                                     ignore_partitions=True)
+    database_import.load_data(dsn, threads)

 
 

-    assert temp_db_cursor.scalar('SELECT count(*) FROM country_name') > 0
-    assert temp_db_cursor.scalar('SELECT count(*) FROM country_name WHERE partition != 0') == 0
+    assert temp_db_cursor.table_rows('placex') == 30
+    assert temp_db_cursor.table_rows('location_property_osmline') == 1