run Tiger import with parallel threads per default

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

diff --git a/nominatim/clicmd/add_data.py b/nominatim/clicmd/add_data.py
index 18d67ce08055c7ddf26f9337304b0f4329b96c17..9f0d311898dbadbf54a6f77b1900735a3b5d532a 100644 (file)
--- a/nominatim/clicmd/add_data.py
+++ b/nominatim/clicmd/add_data.py
@@ -3,7 +3,7 @@ Implementation of the 'add-data' subcommand.
 """
 import logging
 
-from nominatim.tools.exec_utils import run_legacy_script
+import psutil
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111
@@ -16,8 +16,17 @@ class UpdateAddData:
     """\
     Add additional data from a file or an online source.
 
-    Data is only imported, not indexed. You need to call `nominatim index`
-    to complete the process.
+    This command allows to add or update the search data in the database.
+    The data can come either from an OSM file or single OSM objects can
+    directly be downloaded from the OSM API. This function only loads the
+    data into the database. Afterwards it still needs to be integrated
+    in the search index. Use the `nominatim index` command for that.
+
+    The command can also be used to add external non-OSM data to the
+    database. At the moment the only supported format is TIGER housenumber
+    data. See the online documentation at
+    https://nominatim.org/release-docs/latest/admin/Import/#installing-tiger-housenumber-data-for-the-us
+    for more information.
     """
 
     @staticmethod
@@ -25,9 +34,9 @@ class UpdateAddData:
         group_name = parser.add_argument_group('Source')
         group = group_name.add_mutually_exclusive_group(required=True)
         group.add_argument('--file', metavar='FILE',
-                           help='Import data from an OSM file')
+                           help='Import data from an OSM file or diff file')
         group.add_argument('--diff', metavar='FILE',
-                           help='Import data from an OSM diff file')
+                           help='Import data from an OSM diff file (deprecated: use --file)')
         group.add_argument('--node', metavar='ID', type=int,
                            help='Import a single node from the API')
         group.add_argument('--way', metavar='ID', type=int,
@@ -35,33 +44,45 @@ class UpdateAddData:
         group.add_argument('--relation', metavar='ID', type=int,
                            help='Import a single relation from the API')
         group.add_argument('--tiger-data', metavar='DIR',
-                           help='Add housenumbers from the US TIGER census database.')
+                           help='Add housenumbers from the US TIGER census database')
         group = parser.add_argument_group('Extra arguments')
         group.add_argument('--use-main-api', action='store_true',
                            help='Use OSM API instead of Overpass to download objects')
+        group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
+                           help='Size of cache to be used by osm2pgsql (in MB)')
+        group.add_argument('--socket-timeout', dest='socket_timeout', type=int, default=60,
+                           help='Set timeout for file downloads')
 
     @staticmethod
     def run(args):
         from nominatim.tokenizer import factory as tokenizer_factory
-        from nominatim.tools import tiger_data
+        from nominatim.tools import tiger_data, add_osm_data
 
         if args.tiger_data:
             tokenizer = tokenizer_factory.get_tokenizer_for_db(args.config)
             return tiger_data.add_tiger_data(args.tiger_data,
-                                             args.config, args.threads or 1,
+                                             args.config,
+                                             args.threads or psutil.cpu_count()  or 1,
                                              tokenizer)
 
-        params = ['update.php']
-        if args.file:
-            params.extend(('--import-file', args.file))
-        elif args.diff:
-            params.extend(('--import-diff', args.diff))
-        elif args.node:
-            params.extend(('--import-node', args.node))
-        elif args.way:
-            params.extend(('--import-way', args.way))
-        elif args.relation:
-            params.extend(('--import-relation', args.relation))
-        if args.use_main_api:
-            params.append('--use-main-api')
-        return run_legacy_script(*params, nominatim_env=args)
+        osm2pgsql_params = args.osm2pgsql_options(default_cache=1000, default_threads=1)
+        if args.file or args.diff:
+            return add_osm_data.add_data_from_file(args.file or args.diff,
+                                                   osm2pgsql_params)
+
+        if args.node:
+            return add_osm_data.add_osm_object('node', args.node,
+                                               args.use_main_api,
+                                               osm2pgsql_params)
+
+        if args.way:
+            return add_osm_data.add_osm_object('way', args.way,
+                                               args.use_main_api,
+                                               osm2pgsql_params)
+
+        if args.relation:
+            return add_osm_data.add_osm_object('relation', args.relation,
+                                               args.use_main_api,
+                                               osm2pgsql_params)
+
+        return 0