1 # SPDX-License-Identifier: GPL-3.0-or-later
3 # This file is part of Nominatim. (https://nominatim.org)
5 # Copyright (C) 2024 by the Nominatim developer community.
6 # For a full list of authors see the git log.
8 Implementation of the 'import' subcommand.
10 from typing import Optional
13 from pathlib import Path
18 from ..errors import UsageError
19 from ..config import Configuration
20 from ..db.connection import connect
21 from ..db import status, properties
22 from ..tokenizer.base import AbstractTokenizer
23 from ..version import NOMINATIM_VERSION
24 from .args import NominatimArgs
28 LOG = logging.getLogger()
33 Create a new Nominatim database from an OSM file.
35 This sub-command sets up a new Nominatim database from scratch starting
36 with creating a new database in Postgresql. The user running this command
37 needs superuser rights on the database.
40 def add_args(self, parser: argparse.ArgumentParser) -> None:
41 group1 = parser.add_argument_group('Required arguments')
42 group1.add_argument('--osm-file', metavar='FILE', action='append',
43 help='OSM file to be imported'
44 ' (repeat for importing multiple files)',
46 group1.add_argument('--continue', dest='continue_at',
47 choices=['import-from-file', 'load-data', 'indexing', 'db-postprocess'],
48 help='Continue an import that was interrupted',
50 group2 = parser.add_argument_group('Optional arguments')
51 group2.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
52 help='Size of cache to be used by osm2pgsql (in MB)')
53 group2.add_argument('--reverse-only', action='store_true',
54 help='Do not create tables and indexes for searching')
55 group2.add_argument('--no-partitions', action='store_true',
56 help="Do not partition search indices "
57 "(speeds up import of single country extracts)")
58 group2.add_argument('--no-updates', action='store_true',
59 help="Do not keep tables that are only needed for "
60 "updating the database later")
61 group2.add_argument('--offline', action='store_true',
62 help="Do not attempt to load any additional data from the internet")
63 group3 = parser.add_argument_group('Expert options')
64 group3.add_argument('--ignore-errors', action='store_true',
65 help='Continue import even when errors in SQL are present')
66 group3.add_argument('--index-noanalyse', action='store_true',
67 help='Do not perform analyse operations during index (expert only)')
68 group3.add_argument('--prepare-database', action='store_true',
69 help='Create the database but do not import any data')
71 def run(self, args: NominatimArgs) -> int:
72 if args.osm_file is None and args.continue_at is None and not args.prepare_database:
73 raise UsageError("No input files (use --osm-file).")
75 if args.osm_file is not None and args.continue_at not in ('import-from-file', None):
76 raise UsageError(f"Cannot use --continue {args.continue_at} and --osm-file together.")
78 if args.continue_at is not None and args.prepare_database:
80 "Cannot use --continue and --prepare-database together."
83 return asyncio.run(self.async_run(args))
85 async def async_run(self, args: NominatimArgs) -> int:
86 from ..data import country_info
87 from ..tools import database_import, postcodes, freeze
88 from ..indexer.indexer import Indexer
90 start_time = time.time()
92 num_threads = args.threads or psutil.cpu_count() or 1
93 country_info.setup_country_config(args.config)
95 if args.prepare_database or args.continue_at is None:
96 LOG.warning('Creating database')
97 database_import.setup_database_skeleton(args.config.get_libpq_dsn(),
98 rouser=args.config.DATABASE_WEBUSER)
99 if args.prepare_database:
102 if args.continue_at in (None, 'import-from-file'):
103 self._base_import(args)
105 if args.continue_at in ('import-from-file', 'load-data', None):
106 LOG.warning('Initialise tables')
107 with connect(args.config.get_libpq_dsn()) as conn:
108 database_import.truncate_data_tables(conn)
110 LOG.warning('Load data into placex table')
111 await database_import.load_data(args.config.get_libpq_dsn(), num_threads)
113 LOG.warning("Setting up tokenizer")
114 tokenizer = self._get_tokenizer(args.continue_at, args.config)
116 if args.continue_at in ('import-from-file', 'load-data', None):
117 LOG.warning('Calculate postcodes')
118 postcodes.update_postcodes(args.config.get_libpq_dsn(),
119 args.project_dir, tokenizer)
121 if args.continue_at in ('import-from-file', 'load-data', 'indexing', None):
122 LOG.warning('Indexing places')
123 indexer = Indexer(args.config.get_libpq_dsn(), tokenizer, num_threads)
124 await indexer.index_full(analyse=not args.index_noanalyse)
126 LOG.warning('Post-process tables')
127 with connect(args.config.get_libpq_dsn()) as conn:
128 conn.autocommit = True
129 await database_import.create_search_indices(conn, args.config,
130 drop=args.no_updates,
132 LOG.warning('Create search index for default country names.')
133 conn.autocommit = False
134 country_info.create_country_names(conn, tokenizer,
135 args.config.get_str_list('LANGUAGES'))
137 conn.autocommit = True
138 freeze.drop_update_tables(conn)
139 tokenizer.finalize_import(args.config)
141 LOG.warning('Recompute word counts')
142 tokenizer.update_statistics(args.config, threads=num_threads)
144 end_time = time.time()
145 elapsed = end_time - start_time
146 LOG.warning(f'Import completed successfully in {elapsed:.2f} seconds.')
148 self._finalize_database(args.config.get_libpq_dsn(), args.offline)
152 def _base_import(self, args: NominatimArgs) -> None:
153 from ..tools import database_import, refresh
154 from ..data import country_info
156 files = args.get_osm_file_list()
158 raise UsageError("No input files (use --osm-file).")
160 if args.continue_at in ('import-from-file', None):
161 # Check if the correct plugins are installed
162 database_import.check_existing_database_plugins(args.config.get_libpq_dsn())
163 LOG.warning('Setting up country tables')
164 country_info.setup_country_tables(args.config.get_libpq_dsn(),
165 args.config.lib_dir.data,
168 LOG.warning('Importing OSM data file')
169 database_import.import_osm_data(files,
170 args.osm2pgsql_options(0, 1),
171 drop=args.no_updates,
172 ignore_errors=args.ignore_errors)
174 LOG.warning('Importing wikipedia importance data')
175 data_path = Path(args.config.WIKIPEDIA_DATA_PATH or args.project_dir)
176 if refresh.import_wikipedia_articles(args.config.get_libpq_dsn(),
178 LOG.error('Wikipedia importance dump file not found. '
179 'Calculating importance values of locations will not '
180 'use Wikipedia importance data.')
182 LOG.warning('Importing secondary importance raster data')
183 if refresh.import_secondary_importance(args.config.get_libpq_dsn(),
184 args.project_dir) != 0:
185 LOG.error('Secondary importance file not imported. '
186 'Falling back to default ranking.')
188 self._setup_tables(args.config, args.reverse_only)
190 def _setup_tables(self, config: Configuration, reverse_only: bool) -> None:
191 """ Set up the basic database layout: tables, indexes and functions.
193 from ..tools import database_import, refresh
195 with connect(config.get_libpq_dsn()) as conn:
196 conn.autocommit = True
197 LOG.warning('Create functions (1st pass)')
198 refresh.create_functions(conn, config, False, False)
199 LOG.warning('Create tables')
200 database_import.create_tables(conn, config, reverse_only=reverse_only)
201 refresh.load_address_levels_from_config(conn, config)
202 LOG.warning('Create functions (2nd pass)')
203 refresh.create_functions(conn, config, False, False)
204 LOG.warning('Create table triggers')
205 database_import.create_table_triggers(conn, config)
206 LOG.warning('Create partition tables')
207 database_import.create_partition_tables(conn, config)
208 LOG.warning('Create functions (3rd pass)')
209 refresh.create_functions(conn, config, False, False)
211 def _get_tokenizer(self, continue_at: Optional[str],
212 config: Configuration) -> AbstractTokenizer:
213 """ Set up a new tokenizer or load an already initialised one.
215 from ..tokenizer import factory as tokenizer_factory
217 if continue_at in ('import-from-file', 'load-data', None):
218 # (re)initialise the tokenizer data
219 return tokenizer_factory.create_tokenizer(config)
221 # just load the tokenizer
222 return tokenizer_factory.get_tokenizer_for_db(config)
224 def _finalize_database(self, dsn: str, offline: bool) -> None:
225 """ Determine the database date and set the status accordingly.
227 with connect(dsn) as conn:
228 properties.set_property(conn, 'database_version', str(NOMINATIM_VERSION))
231 dbdate = status.compute_database_date(conn, offline)
232 status.set_status(conn, dbdate)
233 LOG.info('Database is at %s.', dbdate)
234 except Exception as exc:
235 LOG.error('Cannot determine date of database: %s', exc)