split search SQL in windowed search_name lookup and constraint search

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

diff --git a/nominatim/clicmd/replication.py b/nominatim/clicmd/replication.py
index 9d94630488cab8c9ad8f6f6d8a4fddb0e2a54304..b795650694d207ed73b32db5f566ae8289ffad58 100644 (file)
--- a/nominatim/clicmd/replication.py
+++ b/nominatim/clicmd/replication.py
@@ -7,6 +7,8 @@
 """
 Implementation of the 'replication' sub-command.
 """
 """
 Implementation of the 'replication' sub-command.
 """

+from typing import Optional
+import argparse

 import datetime as dt
 import logging
 import socket
 import datetime as dt
 import logging
 import socket


@@ -15,6 +17,7 @@ import time
 from nominatim.db import status
 from nominatim.db.connection import connect
 from nominatim.errors import UsageError
 from nominatim.db import status
 from nominatim.db.connection import connect
 from nominatim.errors import UsageError

+from nominatim.clicmd.args import NominatimArgs

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


@@ -41,8 +44,7 @@ class UpdateReplication:
     downloads and imports the next batch of updates.
     """
 
     downloads and imports the next batch of updates.
     """
 

-    @staticmethod
-    def add_args(parser):
+    def add_args(self, parser: argparse.ArgumentParser) -> None:

         group = parser.add_argument_group('Arguments for initialisation')
         group.add_argument('--init', action='store_true',
                            help='Initialise the update process')
         group = parser.add_argument_group('Arguments for initialisation')
         group.add_argument('--init', action='store_true',
                            help='Initialise the update process')


@@ -68,29 +70,32 @@ class UpdateReplication:
         group.add_argument('--socket-timeout', dest='socket_timeout', type=int, default=60,
                            help='Set timeout for file downloads')
 
         group.add_argument('--socket-timeout', dest='socket_timeout', type=int, default=60,
                            help='Set timeout for file downloads')
 

-    @staticmethod
-    def _init_replication(args):
+
+    def _init_replication(self, args: NominatimArgs) -> int:

         from ..tools import replication, refresh
 
         LOG.warning("Initialising replication updates")
         with connect(args.config.get_libpq_dsn()) as conn:
         from ..tools import replication, refresh
 
         LOG.warning("Initialising replication updates")
         with connect(args.config.get_libpq_dsn()) as conn:

-            replication.init_replication(conn, base_url=args.config.REPLICATION_URL)
+            replication.init_replication(conn, base_url=args.config.REPLICATION_URL,
+                                         socket_timeout=args.socket_timeout)

             if args.update_functions:
                 LOG.warning("Create functions")
                 refresh.create_functions(conn, args.config, True, False)
         return 0
 
 
             if args.update_functions:
                 LOG.warning("Create functions")
                 refresh.create_functions(conn, args.config, True, False)
         return 0
 
 

-    @staticmethod
-    def _check_for_updates(args):
+    def _check_for_updates(self, args: NominatimArgs) -> int:

         from ..tools import replication
 
         with connect(args.config.get_libpq_dsn()) as conn:
         from ..tools import replication
 
         with connect(args.config.get_libpq_dsn()) as conn:

-            return replication.check_for_updates(conn, base_url=args.config.REPLICATION_URL)
+            return replication.check_for_updates(conn, base_url=args.config.REPLICATION_URL,
+                                                 socket_timeout=args.socket_timeout)
+

 
 

-    @staticmethod
-    def _report_update(batchdate, start_import, start_index):
-        def round_time(delta):
+    def _report_update(self, batchdate: dt.datetime,
+                       start_import: dt.datetime,
+                       start_index: Optional[dt.datetime]) -> None:
+        def round_time(delta: dt.timedelta) -> dt.timedelta:

             return dt.timedelta(seconds=int(delta.total_seconds()))
 
         end = dt.datetime.now(dt.timezone.utc)
             return dt.timedelta(seconds=int(delta.total_seconds()))
 
         end = dt.datetime.now(dt.timezone.utc)


@@ -101,8 +106,7 @@ class UpdateReplication:
                     round_time(end - batchdate))
 
 
                     round_time(end - batchdate))
 
 

-    @staticmethod
-    def _compute_update_interval(args):
+    def _compute_update_interval(self, args: NominatimArgs) -> int:

         if args.catch_up:
             return 0
 
         if args.catch_up:
             return 0
 


@@ -119,13 +123,13 @@ class UpdateReplication:
         return update_interval
 
 
         return update_interval
 
 

-    @staticmethod
-    def _update(args):
+    def _update(self, args: NominatimArgs) -> None:
+        # pylint: disable=too-many-locals

         from ..tools import replication
         from ..indexer.indexer import Indexer
         from ..tokenizer import factory as tokenizer_factory
 
         from ..tools import replication
         from ..indexer.indexer import Indexer
         from ..tokenizer import factory as tokenizer_factory
 

-        update_interval = UpdateReplication._compute_update_interval(args)
+        update_interval = self._compute_update_interval(args)

 
         params = args.osm2pgsql_options(default_cache=2000, default_threads=1)
         params.update(base_url=args.config.REPLICATION_URL,
 
         params = args.osm2pgsql_options(default_cache=2000, default_threads=1)
         params.update(base_url=args.config.REPLICATION_URL,


@@ -143,10 +147,13 @@ class UpdateReplication:
         tokenizer = tokenizer_factory.get_tokenizer_for_db(args.config)
         indexer = Indexer(args.config.get_libpq_dsn(), tokenizer, args.threads or 1)
 
         tokenizer = tokenizer_factory.get_tokenizer_for_db(args.config)
         indexer = Indexer(args.config.get_libpq_dsn(), tokenizer, args.threads or 1)
 

+        dsn = args.config.get_libpq_dsn()
+

         while True:
         while True:

-            with connect(args.config.get_libpq_dsn()) as conn:
-                start = dt.datetime.now(dt.timezone.utc)
-                state = replication.update(conn, params)
+            start = dt.datetime.now(dt.timezone.utc)
+            state = replication.update(dsn, params, socket_timeout=args.socket_timeout)
+
+            with connect(dsn) as conn:

                 if state is not replication.UpdateState.NO_CHANGES:
                     status.log_status(conn, start, 'import')
                 batchdate, _, _ = status.get_status(conn)
                 if state is not replication.UpdateState.NO_CHANGES:
                     status.log_status(conn, start, 'import')
                 batchdate, _, _ = status.get_status(conn)


@@ -156,7 +163,7 @@ class UpdateReplication:
                 index_start = dt.datetime.now(dt.timezone.utc)
                 indexer.index_full(analyse=False)
 
                 index_start = dt.datetime.now(dt.timezone.utc)
                 indexer.index_full(analyse=False)
 

-                with connect(args.config.get_libpq_dsn()) as conn:
+                with connect(dsn) as conn:

                     status.set_indexed(conn, True)
                     status.log_status(conn, index_start, 'index')
                     conn.commit()
                     status.set_indexed(conn, True)
                     status.log_status(conn, index_start, 'index')
                     conn.commit()


@@ -169,7 +176,8 @@ class UpdateReplication:
                     indexer.index_full(analyse=False)
 
             if LOG.isEnabledFor(logging.WARNING):
                     indexer.index_full(analyse=False)
 
             if LOG.isEnabledFor(logging.WARNING):

-                UpdateReplication._report_update(batchdate, start, index_start)
+                assert batchdate is not None
+                self._report_update(batchdate, start, index_start)

 
             if args.once or (args.catch_up and state is replication.UpdateState.NO_CHANGES):
                 break
 
             if args.once or (args.catch_up and state is replication.UpdateState.NO_CHANGES):
                 break


@@ -179,15 +187,14 @@ class UpdateReplication:
                 time.sleep(recheck_interval)
 
 
                 time.sleep(recheck_interval)
 
 

-    @staticmethod
-    def run(args):
+    def run(self, args: NominatimArgs) -> int:

         socket.setdefaulttimeout(args.socket_timeout)
 
         if args.init:
         socket.setdefaulttimeout(args.socket_timeout)
 
         if args.init:

-            return UpdateReplication._init_replication(args)
+            return self._init_replication(args)

 
         if args.check_for_updates:
 
         if args.check_for_updates:

-            return UpdateReplication._check_for_updates(args)
+            return self._check_for_updates(args)

 
 

-        UpdateReplication._update(args)
+        self._update(args)

         return 0
         return 0