Integrated OSM views into importance computation

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

diff --git a/nominatim/clicmd/refresh.py b/nominatim/clicmd/refresh.py
index c741dcf63632fc0c01d8592a66f46d3be0c8bdbd..d03d682f89ba7fb6b850877b445c55b83e2300ed 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')
@@ -49,10 +63,21 @@ class UpdateRefresh:
                            help='Update the PL/pgSQL functions in the database')
         group.add_argument('--wiki-data', action='store_true',
                            help='Update Wikipedia/data importance numbers')
+        group.add_argument('--osm-views', action='store_true',
+                           help='Update OSM views/data importance numbers')
         group.add_argument('--importance', action='store_true',
                            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 +85,7 @@ class UpdateRefresh:
                            help='Enable debug warning statements in functions')
 
 
-    def run(self, args):
+    def run(self, args: NominatimArgs) -> int: #pylint: disable=too-many-branches, too-many-statements
         from ..tools import refresh, postcodes
         from ..indexer.indexer import Indexer
 
@@ -79,6 +104,7 @@ class UpdateRefresh:
                           "Postcode updates on a frozen database is not possible.")
 
         if args.word_tokens:
+            LOG.warning('Updating word tokens')
             tokenizer = self._get_tokenizer(args.config)
             tokenizer.update_word_tokens()
 
@@ -107,6 +133,17 @@ class UpdateRefresh:
                 LOG.fatal('FATAL: Wikipedia importance dump file not found')
                 return 1
 
+        if args.osm_views:
+            data_path = Path(args.project_dir)
+            LOG.warning('Import OSM views GeoTIFF data from %s', data_path)
+            num = refresh.import_osm_views_geotiff(args.config.get_libpq_dsn(), data_path)
+            if num == 1:
+                LOG.fatal('FATAL: OSM views GeoTIFF file not found')
+                return 1
+            if num == 2:
+                LOG.fatal('FATAL: PostGIS version number is less than 3')
+                return 1
+
         # Attention: importance MUST come after wiki data import.
         if args.importance:
             LOG.warning('Update importance values for database')
@@ -116,13 +153,25 @@ class UpdateRefresh:
         if args.website:
             webdir = args.project_dir / 'website'
             LOG.warning('Setting up website directory at %s', webdir)
+            # This is a little bit hacky: call the tokenizer setup, so that
+            # the tokenizer directory gets repopulated as well, in case it
+            # wasn't there yet.
+            self._get_tokenizer(args.config)
             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