Merge remote-tracking branch 'upstream/master'

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

diff --git a/nominatim/clicmd/refresh.py b/nominatim/clicmd/refresh.py
index 3c245cd46e84b385a39288f537bb50813f8f284f..dce28d983a1365656c43b3e57bc79eb20c8befc8 100644 (file)
--- a/nominatim/clicmd/refresh.py
+++ b/nominatim/clicmd/refresh.py
@@ -7,10 +7,15 @@
 """
 Implementation of 'refresh' subcommand.
 """
+from typing import Tuple, Optional
+import argparse
 import logging
 from pathlib import Path
 
+from nominatim.config import Configuration
 from nominatim.db.connection import connect
+from nominatim.tokenizer.base import AbstractTokenizer
+from nominatim.clicmd.args import NominatimArgs
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111
@@ -19,6 +24,16 @@ from nominatim.db.connection import connect
 
 LOG = logging.getLogger()
 
+def _parse_osm_object(obj: str) -> Tuple[str, int]:
+    """ Parse the given argument into a tuple of OSM type and ID.
+        Raises an ArgumentError if the format is not recognized.
+    """
+    if len(obj) < 2 or obj[0].lower() not in 'nrw' or not obj[1:].isdigit():
+        raise argparse.ArgumentTypeError("Cannot parse OSM ID. Expect format: [N|W|R]<id>.")
+
+    return (obj[0].upper(), int(obj[1:]))
+
+
 class UpdateRefresh:
     """\
     Recompute auxiliary data used by the indexing process.
@@ -31,11 +46,10 @@ class UpdateRefresh:
     Warning: the 'update' command must not be run in parallel with other update
              commands like 'replication' or 'add-data'.
     """
-    def __init__(self):
-        self.tokenizer = None
+    def __init__(self) -> None:
+        self.tokenizer: Optional[AbstractTokenizer] = None
 
-    @staticmethod
-    def add_args(parser):
+    def add_args(self, parser: argparse.ArgumentParser) -> None:
         group = parser.add_argument_group('Data arguments')
         group.add_argument('--postcodes', action='store_true',
                            help='Update postcode centroid table')
@@ -53,6 +67,15 @@ class UpdateRefresh:
                            help='Recompute place importances (expensive!)')
         group.add_argument('--website', action='store_true',
                            help='Refresh the directory that serves the scripts for the web API')
+        group.add_argument('--data-object', action='append',
+                           type=_parse_osm_object, metavar='OBJECT',
+                           help='Mark the given OSM object as requiring an update'
+                                ' (format: [NWR]<id>)')
+        group.add_argument('--data-area', action='append',
+                           type=_parse_osm_object, metavar='OBJECT',
+                           help='Mark the area around the given OSM object as requiring an update'
+                                ' (format: [NWR]<id>)')
+
         group = parser.add_argument_group('Arguments for function refresh')
         group.add_argument('--no-diff-updates', action='store_false', dest='diffs',
                            help='Do not enable code for propagating updates')
@@ -60,7 +83,7 @@ class UpdateRefresh:
                            help='Enable debug warning statements in functions')
 
 
-    def run(self, args):
+    def run(self, args: NominatimArgs) -> int: #pylint: disable=too-many-branches
         from ..tools import refresh, postcodes
         from ..indexer.indexer import Indexer
 
@@ -124,10 +147,18 @@ class UpdateRefresh:
             with connect(args.config.get_libpq_dsn()) as conn:
                 refresh.setup_website(webdir, args.config, conn)
 
+        if args.data_object or args.data_area:
+            with connect(args.config.get_libpq_dsn()) as conn:
+                for obj in args.data_object or []:
+                    refresh.invalidate_osm_object(*obj, conn, recursive=False)
+                for obj in args.data_area or []:
+                    refresh.invalidate_osm_object(*obj, conn, recursive=True)
+                conn.commit()
+
         return 0
 
 
-    def _get_tokenizer(self, config):
+    def _get_tokenizer(self, config: Configuration) -> AbstractTokenizer:
         if self.tokenizer is None:
             from ..tokenizer import factory as tokenizer_factory