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 ceac7a2421dc43afd23ecc1cf3a4c3066c7b02f1..eda88903a9733598455b5c013804b929e5f2731f 100644 (file)
--- a/test/python/test_tools_database_import.py
+++ b/test/python/test_tools_database_import.py
@@ -1,136 +1,124 @@
 """
 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 pathlib import Path

 
 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'
-
-    conn = psycopg2.connect(database='postgres')
-
-    conn.set_isolation_level(0)
-    with conn.cursor() as cur:
-        cur.execute('DROP DATABASE IF EXISTS {}'.format(dbname))
+class TestDatabaseSetup:
+    DBNAME = 'test_nominatim_python_unittest'

 
 

-    yield dbname
+    @pytest.fixture(autouse=True)
+    def setup_nonexistant_db(self):
+        conn = psycopg2.connect(database='postgres')

 
 

-    with conn.cursor() as cur:
-        cur.execute('DROP DATABASE IF EXISTS {}'.format(dbname))
+        try:
+            conn.set_isolation_level(0)
+            with conn.cursor() as cur:
+                cur.execute(f'DROP DATABASE IF EXISTS {self.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)
+            yield True

 
 

-    conn = psycopg2.connect(database=nonexistant_db)
+            with conn.cursor() as cur:
+                cur.execute(f'DROP DATABASE IF EXISTS {self.DBNAME}')
+        finally:
+            conn.close()

 
 

-    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()
+    @pytest.fixture
+    def cursor(self):
+        conn = psycopg2.connect(database=self.DBNAME)

 
 

+        try:
+            with conn.cursor() as cur:
+                yield cur
+        finally:
+            conn.close()

 
 

-def test_create_db_success(nonexistant_db):
-    database_import.create_db('dbname=' + nonexistant_db, rouser='www-data')

 
 

-    conn = psycopg2.connect(database=nonexistant_db)
-    conn.close()
+    def conn(self):
+        return closing(psycopg2.connect(database=self.DBNAME))

 
 
 
 

-def test_create_db_already_exists(temp_db):
-    with pytest.raises(UsageError):
-        database_import.create_db('dbname=' + temp_db)
-
-
-def test_create_db_unsupported_version(nonexistant_db, monkeypatch):
-    monkeypatch.setattr(database_import, 'POSTGRESQL_REQUIRED_VERSION', (100, 4))
+    def test_setup_skeleton(self):
+        database_import.setup_database_skeleton(f'dbname={self.DBNAME}')

 
 

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

 
 
 
 

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

 
 

+        with pytest.raises(UsageError, match='PostgreSQL server is too old.'):
+            database_import.setup_database_skeleton(f'dbname={self.DBNAME}')

 
 

-def test_setup_extensions(temp_db_conn, temp_db_cursor):
-    database_import.setup_extensions(temp_db_conn)

 
 

-    temp_db_cursor.execute('CREATE TABLE t (h HSTORE, geom GEOMETRY(Geometry, 4326))')
+    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_setup_extensions_old_postgis(temp_db_conn, monkeypatch):
-    monkeypatch.setattr(database_import, 'POSTGIS_REQUIRED_VERSION', (50, 50))
+    def test_setup_extensions_old_postgis(self, monkeypatch):
+        monkeypatch.setattr(database_import, 'POSTGIS_REQUIRED_VERSION', (50, 50))

 
 

-    with pytest.raises(UsageError, match='PostGIS version is too old.'):
-        database_import.setup_extensions(temp_db_conn)
+        with pytest.raises(UsageError, match='PostGIS is too old.'):
+            database_import.setup_database_skeleton(f'dbname={self.DBNAME}')

 
 
 
 

-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')
+def test_setup_skeleton_already_exists(temp_db):
+    with pytest.raises(UsageError):
+        database_import.setup_database_skeleton(f'dbname={temp_db}')

 
 

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

 
 

+def test_import_osm_data_simple(table_factory, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))

 
 

-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.import_osm_data(Path('file.pbf'), osm2pgsql_options)

 
 

-    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

 
 

+def test_import_osm_data_multifile(table_factory, tmp_path, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))
+    osm2pgsql_options['osm2pgsql_cache'] = 0

 
 

-def test_import_osm_data_simple(temp_db_cursor,osm2pgsql_options):
-    temp_db_cursor.execute('CREATE TABLE place (id INT)')
-    temp_db_cursor.execute('INSERT INTO place values (1)')
+    files = [tmp_path / 'file1.osm', tmp_path / 'file2.osm']
+    for f in files:
+        f.write_text('test')

 
 

-    database_import.import_osm_data('file.pdf', osm2pgsql_options)
+    database_import.import_osm_data(files, osm2pgsql_options)

 
 
 
 

-def test_import_osm_data_simple_no_data(temp_db_cursor,osm2pgsql_options):
-    temp_db_cursor.execute('CREATE TABLE place (id INT)')
+def test_import_osm_data_simple_no_data(table_factory, osm2pgsql_options):
+    table_factory('place')

 
     with pytest.raises(UsageError, match='No data.*'):
 
     with pytest.raises(UsageError, match='No data.*'):

-        database_import.import_osm_data('file.pdf', osm2pgsql_options)
+        database_import.import_osm_data(Path('file.pbf'), osm2pgsql_options)

 
 
 
 

-def test_import_osm_data_drop(temp_db_conn, temp_db_cursor, tmp_path, osm2pgsql_options):
-    temp_db_cursor.execute('CREATE TABLE place (id INT)')
-    temp_db_cursor.execute('CREATE TABLE planet_osm_nodes (id INT)')
-    temp_db_cursor.execute('INSERT INTO place values (1)')
+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')

 
     flatfile = tmp_path / 'flatfile'
     flatfile.write_text('touch')
 
     osm2pgsql_options['flatnode_file'] = str(flatfile.resolve())
 
 
     flatfile = tmp_path / 'flatfile'
     flatfile.write_text('touch')
 
     osm2pgsql_options['flatnode_file'] = str(flatfile.resolve())
 

-    database_import.import_osm_data('file.pdf', osm2pgsql_options, drop=True)
+    database_import.import_osm_data(Path('file.pbf'), osm2pgsql_options, drop=True)

 
     assert not flatfile.exists()
     assert not temp_db_conn.table_exists('planet_osm_nodes')
 
 
 
     assert not flatfile.exists()
     assert not temp_db_conn.table_exists('planet_osm_nodes')
 
 

-def test_import_osm_data_default_cache(temp_db_cursor,osm2pgsql_options):
-    temp_db_cursor.execute('CREATE TABLE place (id INT)')
-    temp_db_cursor.execute('INSERT INTO place values (1)')
+def test_import_osm_data_default_cache(table_factory, osm2pgsql_options):
+    table_factory('place', content=((1, ), ))

 
     osm2pgsql_options['osm2pgsql_cache'] = 0
 
 
     osm2pgsql_options['osm2pgsql_cache'] = 0
 


@@ -153,8 +141,8 @@ def test_truncate_database_tables(temp_db_conn, temp_db_cursor, table_factory):
 
 
 @pytest.mark.parametrize("threads", (1, 5))
 
 
 @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):
+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
     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


@@ -168,29 +156,3 @@ def test_load_data(dsn, src_dir, place_row, placex_table, osmline_table, word_ta
 
     assert temp_db_cursor.table_rows('placex') == 30
     assert temp_db_cursor.table_rows('location_property_osmline') == 1
 
     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'))}