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 0225c5ed41561b35ca257eb622f6804204ac56f6..86a504e71f15b56e53d9886cabcde14fee849cb5 100644 (file)
--- a/nominatim/clicmd/index.py
+++ b/nominatim/clicmd/index.py
@@ -1,10 +1,19 @@
+# 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 argparse
+

 import psutil
 
 import psutil
 

-from ..db import status
-from ..db.connection import connect
+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
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111


@@ -15,10 +24,14 @@ from ..db.connection import connect
 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.""")


@@ -29,17 +42,21 @@ 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(),
+        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.boundaries_only:
             indexer.index_by_rank(args.minrank, args.maxrank)
                           args.threads or psutil.cpu_count() or 1)
 
         if not args.no_boundaries:
             indexer.index_boundaries(args.minrank, args.maxrank)
         if not args.boundaries_only:
             indexer.index_by_rank(args.minrank, args.maxrank)

+            indexer.index_postcodes()

 
         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: