Improved performance of the postcodes query and some code cleaning

[nominatim.git] / nominatim / clicmd / replication.py

diff --git a/nominatim/clicmd/replication.py b/nominatim/clicmd/replication.py
index 4852c8b1390a45382c70fe79645037f707dbb523..242b0f6a0b00c80bacc05ef01b97823e6bed2dfd 100644 (file)
--- a/nominatim/clicmd/replication.py
+++ b/nominatim/clicmd/replication.py
@@ -6,9 +6,9 @@ import logging
 import socket
 import time
 
 import socket
 import time
 

-from ..db import status
-from ..db.connection import connect
-from ..errors import UsageError
+from nominatim.db import status
+from nominatim.db.connection import connect
+from nominatim.errors import UsageError

 
 LOG = logging.getLogger()
 
 
 LOG = logging.getLogger()
 


@@ -17,17 +17,6 @@ LOG = logging.getLogger()
 # Using non-top-level imports to make pyosmium optional for replication only.
 # pylint: disable=E0012,C0415
 
 # Using non-top-level imports to make pyosmium optional for replication only.
 # pylint: disable=E0012,C0415
 

-def _osm2pgsql_options_from_args(args, default_cache, default_threads):
-    """ Set up the standard osm2pgsql from the command line arguments.
-    """
-    return dict(osm2pgsql=args.osm2pgsql_path,
-                osm2pgsql_cache=args.osm2pgsql_cache or default_cache,
-                osm2pgsql_style=args.config.get_import_style_file(),
-                threads=args.threads or default_threads,
-                dsn=args.config.get_libpq_dsn(),
-                flatnode_file=args.config.FLATNODE_FILE)
-
-

 class UpdateReplication:
     """\
     Update the database using an online replication service.
 class UpdateReplication:
     """\
     Update the database using an online replication service.


@@ -40,17 +29,17 @@ class UpdateReplication:
                            help='Initialise the update process')
         group.add_argument('--no-update-functions', dest='update_functions',
                            action='store_false',
                            help='Initialise the update process')
         group.add_argument('--no-update-functions', dest='update_functions',
                            action='store_false',

-                           help="""Do not update the trigger function to
-                                   support differential updates.""")
+                           help=("Do not update the trigger function to "
+                                 "support differential updates."))

         group = parser.add_argument_group('Arguments for updates')
         group.add_argument('--check-for-updates', action='store_true',
                            help='Check if new updates are available and exit')
         group.add_argument('--once', action='store_true',
         group = parser.add_argument_group('Arguments for updates')
         group.add_argument('--check-for-updates', action='store_true',
                            help='Check if new updates are available and exit')
         group.add_argument('--once', action='store_true',

-                           help="""Download and apply updates only once. When
-                                   not set, updates are continuously applied""")
+                           help=("Download and apply updates only once. When "
+                                 "not set, updates are continuously applied"))

         group.add_argument('--no-index', action='store_false', dest='do_index',
         group.add_argument('--no-index', action='store_false', dest='do_index',

-                           help="""Do not index the new data. Only applicable
-                                   together with --once""")
+                           help=("Do not index the new data. Only applicable "
+                                 "together with --once"))

         group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
                            help='Size of cache to be used by osm2pgsql (in MB)')
         group = parser.add_argument_group('Download parameters')
         group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
                            help='Size of cache to be used by osm2pgsql (in MB)')
         group = parser.add_argument_group('Download parameters')


@@ -62,13 +51,11 @@ class UpdateReplication:
         from ..tools import replication, refresh
 
         LOG.warning("Initialising replication updates")
         from ..tools import replication, refresh
 
         LOG.warning("Initialising replication updates")

-        conn = connect(args.config.get_libpq_dsn())
-        replication.init_replication(conn, base_url=args.config.REPLICATION_URL)
-        if args.update_functions:
-            LOG.warning("Create functions")
-            refresh.create_functions(conn, args.config, args.sqllib_dir,
-                                     True, False)
-        conn.close()
+        with connect(args.config.get_libpq_dsn()) as conn:
+            replication.init_replication(conn, base_url=args.config.REPLICATION_URL)
+            if args.update_functions:
+                LOG.warning("Create functions")
+                refresh.create_functions(conn, args.config, True, False)

         return 0
 
 
         return 0
 
 


@@ -76,10 +63,8 @@ class UpdateReplication:
     def _check_for_updates(args):
         from ..tools import replication
 
     def _check_for_updates(args):
         from ..tools import replication
 

-        conn = connect(args.config.get_libpq_dsn())
-        ret = replication.check_for_updates(conn, base_url=args.config.REPLICATION_URL)
-        conn.close()
-        return ret
+        with connect(args.config.get_libpq_dsn()) as conn:
+            return replication.check_for_updates(conn, base_url=args.config.REPLICATION_URL)

 
     @staticmethod
     def _report_update(batchdate, start_import, start_index):
 
     @staticmethod
     def _report_update(batchdate, start_import, start_index):


@@ -98,8 +83,9 @@ class UpdateReplication:
     def _update(args):
         from ..tools import replication
         from ..indexer.indexer import Indexer
     def _update(args):
         from ..tools import replication
         from ..indexer.indexer import Indexer

+        from ..tokenizer import factory as tokenizer_factory

 
 

-        params = _osm2pgsql_options_from_args(args, 2000, 1)
+        params = args.osm2pgsql_options(default_cache=2000, default_threads=1)

         params.update(base_url=args.config.REPLICATION_URL,
                       update_interval=args.config.get_int('REPLICATION_UPDATE_INTERVAL'),
                       import_file=args.project_dir / 'osmosischange.osc',
         params.update(base_url=args.config.REPLICATION_URL,
                       update_interval=args.config.get_int('REPLICATION_UPDATE_INTERVAL'),
                       import_file=args.project_dir / 'osmosischange.osc',


@@ -121,26 +107,28 @@ class UpdateReplication:
                 raise UsageError("Bad argument '--no-index'.")
             recheck_interval = args.config.get_int('REPLICATION_RECHECK_INTERVAL')
 
                 raise UsageError("Bad argument '--no-index'.")
             recheck_interval = args.config.get_int('REPLICATION_RECHECK_INTERVAL')
 

+        tokenizer = tokenizer_factory.get_tokenizer_for_db(args.config)
+

         while True:
         while True:

-            conn = connect(args.config.get_libpq_dsn())
-            start = dt.datetime.now(dt.timezone.utc)
-            state = replication.update(conn, params)
-            if state is not replication.UpdateState.NO_CHANGES:
-                status.log_status(conn, start, 'import')
-            batchdate, _, _ = status.get_status(conn)
-            conn.close()
+            with connect(args.config.get_libpq_dsn()) as conn:
+                start = dt.datetime.now(dt.timezone.utc)
+                state = replication.update(conn, params)
+                if state is not replication.UpdateState.NO_CHANGES:
+                    status.log_status(conn, start, 'import')
+                batchdate, _, _ = status.get_status(conn)
+                conn.commit()

 
             if state is not replication.UpdateState.NO_CHANGES and args.do_index:
                 index_start = dt.datetime.now(dt.timezone.utc)
 
             if state is not replication.UpdateState.NO_CHANGES and args.do_index:
                 index_start = dt.datetime.now(dt.timezone.utc)

-                indexer = Indexer(args.config.get_libpq_dsn(),
+                indexer = Indexer(args.config.get_libpq_dsn(), tokenizer,

                                   args.threads or 1)
                 indexer.index_boundaries(0, 30)
                 indexer.index_by_rank(0, 30)
 
                                   args.threads or 1)
                 indexer.index_boundaries(0, 30)
                 indexer.index_by_rank(0, 30)
 

-                conn = connect(args.config.get_libpq_dsn())
-                status.set_indexed(conn, True)
-                status.log_status(conn, index_start, 'index')
-                conn.close()
+                with connect(args.config.get_libpq_dsn()) as conn:
+                    status.set_indexed(conn, True)
+                    status.log_status(conn, index_start, 'index')
+                    conn.commit()

             else:
                 index_start = None
 
             else:
                 index_start = None
 


@@ -154,8 +142,6 @@ class UpdateReplication:
                 LOG.warning("No new changes. Sleeping for %d sec.", recheck_interval)
                 time.sleep(recheck_interval)
 
                 LOG.warning("No new changes. Sleeping for %d sec.", recheck_interval)
                 time.sleep(recheck_interval)
 

-        return state.value
-

 
     @staticmethod
     def run(args):
 
     @staticmethod
     def run(args):


@@ -167,4 +153,5 @@ class UpdateReplication:
         if args.check_for_updates:
             return UpdateReplication._check_for_updates(args)
 
         if args.check_for_updates:
             return UpdateReplication._check_for_updates(args)
 

-        return UpdateReplication._update(args)
+        UpdateReplication._update(args)
+        return 0