X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/185d369404bbb6ebffbd4bd24100797a407ed8ac..0ba93e5ba97c96aee788dccd20a6f7f6e308548f:/nominatim/tools/database_import.py diff --git a/nominatim/tools/database_import.py b/nominatim/tools/database_import.py index 3618ed28..400ce7a5 100644 --- a/nominatim/tools/database_import.py +++ b/nominatim/tools/database_import.py @@ -5,11 +5,9 @@ import logging import os import selectors import subprocess -import shutil from pathlib import Path import psutil -import psycopg2 from nominatim.db.connection import connect, get_pg_env from nominatim.db import utils as db_utils @@ -89,49 +87,6 @@ def setup_extensions(conn): raise UsageError('PostGIS version is too old.') -def install_module(src_dir, project_dir, module_dir, conn=None): - """ Copy the normalization module from src_dir into the project - directory under the '/module' directory. If 'module_dir' is set, then - use the module from there instead and check that it is accessible - for Postgresql. - - The function detects when the installation is run from the - build directory. It doesn't touch the module in that case. - - If 'conn' is given, then the function also tests if the module - can be access via the given database. - """ - if not module_dir: - module_dir = project_dir / 'module' - - if not module_dir.exists() or not src_dir.samefile(module_dir): - - if not module_dir.exists(): - module_dir.mkdir() - - destfile = module_dir / 'nominatim.so' - shutil.copy(str(src_dir / 'nominatim.so'), str(destfile)) - destfile.chmod(0o755) - - LOG.info('Database module installed at %s', str(destfile)) - else: - LOG.info('Running from build directory. Leaving database module as is.') - else: - LOG.info("Using custom path for database module at '%s'", module_dir) - - if conn is not None: - with conn.cursor() as cur: - try: - cur.execute("""CREATE FUNCTION nominatim_test_import_func(text) - RETURNS text AS '{}/nominatim.so', 'transliteration' - LANGUAGE c IMMUTABLE STRICT; - DROP FUNCTION nominatim_test_import_func(text) - """.format(module_dir)) - except psycopg2.DatabaseError as err: - LOG.fatal("Error accessing database module: %s", err) - raise UsageError("Database module cannot be accessed.") from err - - def import_base_data(dsn, sql_dir, ignore_partitions=False): """ Create and populate the tables with basic static data that provides the background for geocoding. Data is assumed to not yet exist. @@ -205,11 +160,10 @@ def create_partition_tables(conn, config): sql.run_sql_file(conn, 'partition-tables.src.sql') -def truncate_data_tables(conn, max_word_frequency=None): +def truncate_data_tables(conn): """ Truncate all data tables to prepare for a fresh load. """ with conn.cursor() as cur: - cur.execute('TRUNCATE word') cur.execute('TRUNCATE placex') cur.execute('TRUNCATE place_addressline') cur.execute('TRUNCATE location_area') @@ -228,23 +182,13 @@ def truncate_data_tables(conn, max_word_frequency=None): for table in [r[0] for r in list(cur)]: cur.execute('TRUNCATE ' + table) - if max_word_frequency is not None: - # Used by getorcreate_word_id to ignore frequent partial words. - cur.execute("""CREATE OR REPLACE FUNCTION get_maxwordfreq() - RETURNS integer AS $$ - SELECT {} as maxwordfreq; - $$ LANGUAGE SQL IMMUTABLE - """.format(max_word_frequency)) - conn.commit() + conn.commit() _COPY_COLUMNS = 'osm_type, osm_id, class, type, name, admin_level, address, extratags, geometry' -def load_data(dsn, data_dir, threads): +def load_data(dsn, threads): """ Copy data into the word and placex table. """ - # Pre-calculate the most important terms in the word list. - db_utils.execute_file(dsn, data_dir / 'words.sql') - sel = selectors.DefaultSelector() # Then copy data from place to placex in chunks. place_threads = max(1, threads - 1)