move index creation for word table to tokenizer

[nominatim.git] / test / bdd / steps / steps_db_ops.py

diff --git a/test/bdd/steps/steps_db_ops.py b/test/bdd/steps/steps_db_ops.py
index 9d443b434da263654ed0f40c4f5e843bf2a0d433..52a50a511c8a095c0c9ce5bd8bd9cb9f7e0783fb 100644 (file)
--- a/test/bdd/steps/steps_db_ops.py
+++ b/test/bdd/steps/steps_db_ops.py
@@ -1,3 +1,4 @@
+import logging
 from itertools import chain
 
 import psycopg2.extras
@@ -5,7 +6,8 @@ import psycopg2.extras
 from place_inserter import PlaceColumn
 from table_compare import NominatimID, DBRow
 
-from nominatim.indexer.indexer import Indexer
+from nominatim.indexer import indexer
+from nominatim.tokenizer import factory as tokenizer_factory
 
 def check_database_integrity(context):
     """ Check some generic constraints on the tables.
@@ -85,12 +87,29 @@ def add_data_to_planet_ways(context):
 def import_and_index_data_from_place_table(context):
     """ Import data previously set up in the place table.
     """
+    nctx = context.nominatim
+
+    tokenizer = tokenizer_factory.create_tokenizer(nctx.get_test_config())
     context.nominatim.copy_from_place(context.db)
-    context.nominatim.run_setup_script('calculate-postcodes')
+
+    # XXX use tool function as soon as it is ported
+    with context.db.cursor() as cur:
+        with (context.nominatim.src_dir / 'lib-sql' / 'postcode_tables.sql').open('r') as fd:
+            cur.execute(fd.read())
+        cur.execute("""
+            INSERT INTO location_postcode
+             (place_id, indexed_status, country_code, postcode, geometry)
+            SELECT nextval('seq_place'), 1, country_code,
+                   upper(trim (both ' ' from address->'postcode')) as pc,
+                   ST_Centroid(ST_Collect(ST_Centroid(geometry)))
+              FROM placex
+             WHERE address ? 'postcode' AND address->'postcode' NOT SIMILAR TO '%(,|;)%'
+                   AND geometry IS NOT null
+             GROUP BY country_code, pc""")
 
     # Call directly as the refresh function does not include postcodes.
-    indexer = Indexer(context.nominatim.test_env['NOMINATIM_DATABASE_DSN'][6:], 1)
-    indexer.index_full(analyse=False)
+    indexer.LOG.setLevel(logging.ERROR)
+    indexer.Indexer(context.nominatim.get_libpq_dsn(), tokenizer, 1).index_full(analyse=False)
 
     check_database_integrity(context)
 
@@ -215,7 +234,7 @@ def check_search_name_contents(context, exclude):
                                 if exclude:
                                     assert not present, "Found term for {}/{}: {}".format(row['object'], name, wid[1])
                                 else:
-                                    assert present, "Missing term for {}/{}: {}".fromat(row['object'], name, wid[1])
+                                    assert present, "Missing term for {}/{}: {}".format(row['object'], name, wid[1])
                     elif name != 'object':
                         assert db_row.contains(name, value), db_row.assert_msg(name, value)