-# Using non-top-level imports to make pyosmium optional for replication only.
-# pylint: disable=E0012,C0415
-
-
-class SetupAll:
- """\
- Create a new Nominatim database from an OSM file.
- """
-
- @staticmethod
- def add_args(parser):
- group_name = parser.add_argument_group('Required arguments')
- group = group_name.add_mutually_exclusive_group(required=True)
- group.add_argument('--osm-file',
- help='OSM file to be imported.')
- group.add_argument('--continue', dest='continue_at',
- choices=['load-data', 'indexing', 'db-postprocess'],
- help='Continue an import that was interrupted')
- group = parser.add_argument_group('Optional arguments')
- group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
- help='Size of cache to be used by osm2pgsql (in MB)')
- group.add_argument('--reverse-only', action='store_true',
- help='Do not create tables and indexes for searching')
- group.add_argument('--enable-debug-statements', action='store_true',
- help='Include debug warning statements in SQL code')
- group.add_argument('--no-partitions', action='store_true',
- help="""Do not partition search indices
- (speeds up import of single country extracts)""")
- group.add_argument('--no-updates', action='store_true',
- help="""Do not keep tables that are only needed for
- updating the database later""")
- group = parser.add_argument_group('Expert options')
- group.add_argument('--ignore-errors', action='store_true',
- help='Continue import even when errors in SQL are present')
- group.add_argument('--index-noanalyse', action='store_true',
- help='Do not perform analyse operations during index')
-
-
- @staticmethod
- def run(args):
- params = ['setup.php']
- if args.osm_file:
- params.extend(('--all', '--osm-file', args.osm_file))
- else:
- if args.continue_at == 'load-data':
- params.append('--load-data')
- if args.continue_at in ('load-data', 'indexing'):
- params.append('--index')
- params.extend(('--create-search-indices', '--create-country-names',
- '--setup-website'))
- if args.osm2pgsql_cache:
- params.extend(('--osm2pgsql-cache', args.osm2pgsql_cache))
- if args.reverse_only:
- params.append('--reverse-only')
- if args.enable_debug_statements:
- params.append('--enable-debug-statements')
- if args.no_partitions:
- params.append('--no-partitions')
- if args.no_updates:
- params.append('--drop')
- if args.ignore_errors:
- params.append('--ignore-errors')
- if args.index_noanalyse:
- params.append('--index-noanalyse')
-
- return run_legacy_script(*params, nominatim_env=args)
-
-
-class SetupFreeze:
- """\
- Make database read-only.
-
- About half of data in the Nominatim database is kept only to be able to
- keep the data up-to-date with new changes made in OpenStreetMap. This
- command drops all this data and only keeps the part needed for geocoding
- itself.
-
- This command has the same effect as the `--no-updates` option for imports.
- """
-
- @staticmethod
- def add_args(parser):
- pass # No options
-
- @staticmethod
- def run(args):
- return run_legacy_script('setup.php', '--drop', nominatim_env=args)
-
-
-class SetupSpecialPhrases:
- """\
- Maintain special phrases.
- """
-
- @staticmethod
- def add_args(parser):
- group = parser.add_argument_group('Input arguments')
- group.add_argument('--from-wiki', action='store_true',
- help='Pull special phrases from the OSM wiki.')
- group = parser.add_argument_group('Output arguments')
- group.add_argument('-o', '--output', default='-',
- help="""File to write the preprocessed phrases to.
- If omitted, it will be written to stdout.""")
-
- @staticmethod
- def run(args):
- if args.output != '-':
- raise NotImplementedError('Only output to stdout is currently implemented.')
- return run_legacy_script('specialphrases.php', '--wiki-import', nominatim_env=args)
-
-
-class UpdateReplication:
- """\
- Update the database using an online replication service.
- """
-
- @staticmethod
- def add_args(parser):
- group = parser.add_argument_group('Arguments for initialisation')
- group.add_argument('--init', action='store_true',
- help='Initialise the update process')
- group.add_argument('--no-update-functions', dest='update_functions',
- action='store_false',
- help="""Do not update the trigger function to
- support differential updates.""")
- group = parser.add_argument_group('Arguments for updates')
- group.add_argument('--check-for-updates', action='store_true',
- help='Check if new updates are available and exit')
- group.add_argument('--once', action='store_true',
- help="""Download and apply updates only once. When
- not set, updates are continuously applied""")
- group.add_argument('--no-index', action='store_false', dest='do_index',
- help="""Do not index the new data. Only applicable
- together with --once""")
- group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
- help='Size of cache to be used by osm2pgsql (in MB)')
- group = parser.add_argument_group('Download parameters')
- group.add_argument('--socket-timeout', dest='socket_timeout', type=int, default=60,
- help='Set timeout for file downloads.')
-
- @staticmethod
- def _init_replication(args):
- from .tools import replication, refresh
-
- socket.setdefaulttimeout(args.socket_timeout)
-
- LOG.warning("Initialising replication updates")
- conn = connect(args.config.get_libpq_dsn())
- replication.init_replication(conn, base_url=args.config.REPLICATION_URL)
- if args.update_functions:
- LOG.warning("Create functions")
- refresh.create_functions(conn, args.config, args.data_dir,
- True, False)
- conn.close()
- return 0
-
-
- @staticmethod
- def _check_for_updates(args):
- from .tools import replication
-
- conn = connect(args.config.get_libpq_dsn())
- ret = replication.check_for_updates(conn, base_url=args.config.REPLICATION_URL)
- conn.close()
- return ret
-
- @staticmethod
- def _report_update(batchdate, start_import, start_index):
- def round_time(delta):
- return dt.timedelta(seconds=int(delta.total_seconds()))
-
- end = dt.datetime.now(dt.timezone.utc)
- LOG.warning("Update completed. Import: %s. %sTotal: %s. Remaining backlog: %s.",
- round_time((start_index or end) - start_import),
- "Indexing: {} ".format(round_time(end - start_index))
- if start_index else '',
- round_time(end - start_import),
- round_time(end - batchdate))
-
- @staticmethod
- def _update(args):
- from .tools import replication
- from .indexer.indexer import Indexer
-
- params = _osm2pgsql_options_from_args(args, 2000, 1)
- params.update(base_url=args.config.REPLICATION_URL,
- update_interval=args.config.get_int('REPLICATION_UPDATE_INTERVAL'),
- import_file=args.project_dir / 'osmosischange.osc',
- max_diff_size=args.config.get_int('REPLICATION_MAX_DIFF'),
- indexed_only=not args.once)
-
- # Sanity check to not overwhelm the Geofabrik servers.
- if 'download.geofabrik.de'in params['base_url']\
- and params['update_interval'] < 86400:
- LOG.fatal("Update interval too low for download.geofabrik.de.\n"
- "Please check install documentation "
- "(https://nominatim.org/release-docs/latest/admin/Import-and-Update#"
- "setting-up-the-update-process).")
- raise UsageError("Invalid replication update interval setting.")
-
- if not args.once:
- if not args.do_index:
- LOG.fatal("Indexing cannot be disabled when running updates continuously.")
- raise UsageError("Bad argument '--no-index'.")
- recheck_interval = args.config.get_int('REPLICATION_RECHECK_INTERVAL')
-
- while True:
- conn = connect(args.config.get_libpq_dsn())
- start = dt.datetime.now(dt.timezone.utc)
- state = replication.update(conn, params)
- if state is not replication.UpdateState.NO_CHANGES:
- status.log_status(conn, start, 'import')
- batchdate, _, _ = status.get_status(conn)
- conn.close()
-
- if state is not replication.UpdateState.NO_CHANGES and args.do_index:
- index_start = dt.datetime.now(dt.timezone.utc)
- indexer = Indexer(args.config.get_libpq_dsn(),
- args.threads or 1)
- indexer.index_boundaries(0, 30)
- indexer.index_by_rank(0, 30)
-
- conn = connect(args.config.get_libpq_dsn())
- status.set_indexed(conn, True)
- status.log_status(conn, index_start, 'index')
- conn.close()
- else:
- index_start = None
-
- if LOG.isEnabledFor(logging.WARNING):
- UpdateReplication._report_update(batchdate, start, index_start)
-
- if args.once:
- break
-
- if state is replication.UpdateState.NO_CHANGES:
- LOG.warning("No new changes. Sleeping for %d sec.", recheck_interval)
- time.sleep(recheck_interval)
-
- return state.value
-
- @staticmethod
- def run(args):
- try:
- import osmium # pylint: disable=W0611
- except ModuleNotFoundError:
- LOG.fatal("pyosmium not installed. Replication functions not available.\n"
- "To install pyosmium via pip: pip3 install osmium")
- return 1
-
- if args.init:
- return UpdateReplication._init_replication(args)
-
- if args.check_for_updates:
- return UpdateReplication._check_for_updates(args)
-
- return UpdateReplication._update(args)
-
-class UpdateAddData:
- """\
- Add additional data from a file or an online source.
-
- Data is only imported, not indexed. You need to call `nominatim-update index`
- to complete the process.
- """
-
- @staticmethod
- def add_args(parser):
- 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')
- group.add_argument('--diff', metavar='FILE',
- help='Import data from an OSM diff 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,
- help='Import a single way from the API')
- 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.')
- 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')
-
- @staticmethod
- def run(args):
- if args.tiger_data:
- os.environ['NOMINATIM_TIGER_DATA_PATH'] = args.tiger_data
- return run_legacy_script('setup.php', '--import-tiger-data', nominatim_env=args)
-
- 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)
-
-
-class UpdateIndex:
- """\
- Reindex all new and modified data.
- """
-
- @staticmethod
- def add_args(parser):
- group = parser.add_argument_group('Filter arguments')
- group.add_argument('--boundaries-only', action='store_true',
- help="""Index only administrative boundaries.""")
- group.add_argument('--no-boundaries', action='store_true',
- help="""Index everything except administrative boundaries.""")
- group.add_argument('--minrank', '-r', type=int, metavar='RANK', default=0,
- help='Minimum/starting rank')
- group.add_argument('--maxrank', '-R', type=int, metavar='RANK', default=30,
- help='Maximum/finishing rank')
-
- @staticmethod
- def run(args):
- from .indexer.indexer import Indexer
-
- indexer = Indexer(args.config.get_libpq_dsn(),
- args.threads or _num_system_cpus() 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)
-
- if not args.no_boundaries and not args.boundaries_only \
- and args.minrank == 0 and args.maxrank == 30:
- conn = connect(args.config.get_libpq_dsn())
- status.set_indexed(conn, True)
- conn.close()
-
- return 0
-
-
-class UpdateRefresh:
- """\
- Recompute auxiliary data used by the indexing process.
-
- These functions must not be run in parallel with other update commands.
- """
-
- @staticmethod
- def add_args(parser):
- group = parser.add_argument_group('Data arguments')
- group.add_argument('--postcodes', action='store_true',
- help='Update postcode centroid table')
- group.add_argument('--word-counts', action='store_true',
- help='Compute frequency of full-word search terms')
- group.add_argument('--address-levels', action='store_true',
- help='Reimport address level configuration')
- group.add_argument('--functions', action='store_true',
- 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('--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 = 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')
- group.add_argument('--enable-debug-statements', action='store_true',
- help='Enable debug warning statements in functions')
-
- @staticmethod
- def run(args):
- from .tools import refresh
-
- if args.postcodes:
- LOG.warning("Update postcodes centroid")
- conn = connect(args.config.get_libpq_dsn())
- refresh.update_postcodes(conn, args.data_dir)
- conn.close()
-
- if args.word_counts:
- LOG.warning('Recompute frequency of full-word search terms')
- conn = connect(args.config.get_libpq_dsn())
- refresh.recompute_word_counts(conn, args.data_dir)
- conn.close()
-
- if args.address_levels:
- cfg = Path(args.config.ADDRESS_LEVEL_CONFIG)
- LOG.warning('Updating address levels from %s', cfg)
- conn = connect(args.config.get_libpq_dsn())
- refresh.load_address_levels_from_file(conn, cfg)
- conn.close()
-
- if args.functions:
- LOG.warning('Create functions')
- conn = connect(args.config.get_libpq_dsn())
- refresh.create_functions(conn, args.config, args.data_dir,
- args.diffs, args.enable_debug_statements)
- conn.close()
-
- if args.wiki_data:
- run_legacy_script('setup.php', '--import-wikipedia-articles',
- nominatim_env=args, throw_on_fail=True)
- # Attention: importance MUST come after wiki data import.
- if args.importance:
- run_legacy_script('update.php', '--recompute-importance',
- nominatim_env=args, throw_on_fail=True)
- if args.website:
- run_legacy_script('setup.php', '--setup-website',
- nominatim_env=args, throw_on_fail=True)
-
- return 0
-
-
-class AdminCheckDatabase:
- """\
- Check that the database is complete and operational.
- """
-
- @staticmethod
- def add_args(parser):
- pass # No options
-
- @staticmethod
- def run(args):
- return run_legacy_script('check_import_finished.php', nominatim_env=args)
-
-
-class AdminWarm:
- """\
- Warm database caches for search and reverse queries.
- """
-
- @staticmethod
- def add_args(parser):
- group = parser.add_argument_group('Target arguments')
- group.add_argument('--search-only', action='store_const', dest='target',
- const='search',
- help="Only pre-warm tables for search queries")
- group.add_argument('--reverse-only', action='store_const', dest='target',
- const='reverse',
- help="Only pre-warm tables for reverse queries")
-
- @staticmethod
- def run(args):
- params = ['warm.php']
- if args.target == 'reverse':
- params.append('--reverse-only')
- if args.target == 'search':
- params.append('--search-only')
- return run_legacy_script(*params, nominatim_env=args)
-
-
projects / nominatim.git / blobdiff