Merge remote-tracking branch 'upstream/master'

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

diff --git a/nominatim/clicmd/index.py b/nominatim/clicmd/index.py
index ca3f9deedf4e0546e211679336474a1615ef524d..16b5311c79f5764c2f9cde048b5635e0a1619e03 100644 (file)
--- a/nominatim/clicmd/index.py
+++ b/nominatim/clicmd/index.py
@@ -1,32 +1,37 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# This file is part of Nominatim. (https://nominatim.org)
+#
+# Copyright (C) 2022 by the Nominatim developer community.
+# For a full list of authors see the git log.

 """
 Implementation of the 'index' subcommand.
 """
 """
 Implementation of the 'index' subcommand.
 """

-import os
+import argparse

 
 

-from ..db import status
-from ..db.connection import connect
+import psutil
+
+from nominatim.db import status
+from nominatim.db.connection import connect
+from nominatim.clicmd.args import NominatimArgs

 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111
 # Using non-top-level imports to avoid eventually unused imports.
 # pylint: disable=E0012,C0415
 
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111
 # Using non-top-level imports to avoid eventually unused imports.
 # pylint: disable=E0012,C0415
 

-def _num_system_cpus():
-    try:
-        cpus = len(os.sched_getaffinity(0))
-    except NotImplementedError:
-        cpus = None
-
-    return cpus or os.cpu_count()
-

 
 class UpdateIndex:
     """\
     Reindex all new and modified data.
 
 class UpdateIndex:
     """\
     Reindex all new and modified data.

+
+    Indexing is the process of computing the address and search terms for
+    the places in the database. Every time data is added or changed, indexing
+    needs to be run. Imports and replication updates automatically take care
+    of indexing. For other cases, this function allows to run indexing manually.

     """
 
     """
 

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

         group = parser.add_argument_group('Filter arguments')
         group.add_argument('--boundaries-only', action='store_true',
                            help="""Index only administrative boundaries.""")
         group = parser.add_argument_group('Filter arguments')
         group.add_argument('--boundaries-only', action='store_true',
                            help="""Index only administrative boundaries.""")


@@ -37,12 +42,15 @@ class UpdateIndex:
         group.add_argument('--maxrank', '-R', type=int, metavar='RANK', default=30,
                            help='Maximum/finishing rank')
 
         group.add_argument('--maxrank', '-R', type=int, metavar='RANK', default=30,
                            help='Maximum/finishing rank')
 

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

         from ..indexer.indexer import Indexer
         from ..indexer.indexer import Indexer

+        from ..tokenizer import factory as tokenizer_factory
+
+        tokenizer = tokenizer_factory.get_tokenizer_for_db(args.config)

 
 

-        indexer = Indexer(args.config.get_libpq_dsn(),
-                          args.threads or _num_system_cpus() or 1)
+        indexer = Indexer(args.config.get_libpq_dsn(), tokenizer,
+                          args.threads or psutil.cpu_count() or 1)

 
         if not args.no_boundaries:
             indexer.index_boundaries(args.minrank, args.maxrank)
 
         if not args.no_boundaries:
             indexer.index_boundaries(args.minrank, args.maxrank)


@@ -51,8 +59,7 @@ class UpdateIndex:
 
         if not args.no_boundaries and not args.boundaries_only \
            and args.minrank == 0 and args.maxrank == 30:
 
         if not args.no_boundaries and not args.boundaries_only \
            and args.minrank == 0 and args.maxrank == 30:

-            conn = connect(args.config.get_libpq_dsn())
-            status.set_indexed(conn, True)
-            conn.close()
+            with connect(args.config.get_libpq_dsn()) as conn:
+                status.set_indexed(conn, True)

 
         return 0
 
         return 0