2 Command-line interface to the Nominatim functions for import, update,
3 database administration and querying.
9 from pathlib import Path
11 from .config import Configuration
12 from .tools.exec_utils import run_legacy_script, run_api_script
13 from .db.connection import connect
15 LOG = logging.getLogger()
17 def _num_system_cpus():
19 cpus = len(os.sched_getaffinity(0))
20 except NotImplementedError:
23 return cpus or os.cpu_count()
26 class CommandlineParser:
27 """ Wraps some of the common functions for parsing the command line
28 and setting up subcommands.
30 def __init__(self, prog, description):
31 self.parser = argparse.ArgumentParser(
33 description=description,
34 formatter_class=argparse.RawDescriptionHelpFormatter)
36 self.subs = self.parser.add_subparsers(title='available commands',
39 # Arguments added to every sub-command
40 self.default_args = argparse.ArgumentParser(add_help=False)
41 group = self.default_args.add_argument_group('Default arguments')
42 group.add_argument('-h', '--help', action='help',
43 help='Show this help message and exit')
44 group.add_argument('-q', '--quiet', action='store_const', const=0,
45 dest='verbose', default=1,
46 help='Print only error messages')
47 group.add_argument('-v', '--verbose', action='count', default=1,
48 help='Increase verboseness of output')
49 group.add_argument('--project-dir', metavar='DIR', default='.',
50 help='Base directory of the Nominatim installation (default:.)')
51 group.add_argument('-j', '--threads', metavar='NUM', type=int,
52 help='Number of parallel threads to use')
55 def add_subcommand(self, name, cmd):
56 """ Add a subcommand to the parser. The subcommand must be a class
57 with a function add_args() that adds the parameters for the
58 subcommand and a run() function that executes the command.
60 parser = self.subs.add_parser(name, parents=[self.default_args],
61 help=cmd.__doc__.split('\n', 1)[0],
62 description=cmd.__doc__,
63 formatter_class=argparse.RawDescriptionHelpFormatter,
65 parser.set_defaults(command=cmd)
68 def run(self, **kwargs):
69 """ Parse the command line arguments of the program and execute the
70 appropriate subcommand.
72 args = self.parser.parse_args(args=kwargs.get('cli_args'))
74 if args.subcommand is None:
75 self.parser.print_help()
78 for arg in ('module_dir', 'osm2pgsql_path', 'phplib_dir', 'data_dir', 'phpcgi_path'):
79 setattr(args, arg, Path(kwargs[arg]))
80 args.project_dir = Path(args.project_dir)
82 logging.basicConfig(stream=sys.stderr,
83 format='%(asctime)s: %(message)s',
84 datefmt='%Y-%m-%d %H:%M:%S',
85 level=max(4 - args.verbose, 1) * 10)
87 args.config = Configuration(args.project_dir, args.data_dir / 'settings')
89 return args.command.run(args)
91 ##### Subcommand classes
93 # Each class needs to implement two functions: add_args() adds the CLI parameters
94 # for the subfunction, run() executes the subcommand.
96 # The class documentation doubles as the help text for the command. The
97 # first line is also used in the summary when calling the program without
100 # No need to document the functions each time.
101 # pylint: disable=C0111
106 Create a new Nominatim database from an OSM file.
110 def add_args(parser):
111 group_name = parser.add_argument_group('Required arguments')
112 group = group_name.add_mutually_exclusive_group(required=True)
113 group.add_argument('--osm-file',
114 help='OSM file to be imported.')
115 group.add_argument('--continue', dest='continue_at',
116 choices=['load-data', 'indexing', 'db-postprocess'],
117 help='Continue an import that was interrupted')
118 group = parser.add_argument_group('Optional arguments')
119 group.add_argument('--osm2pgsql-cache', metavar='SIZE', type=int,
120 help='Size of cache to be used by osm2pgsql (in MB)')
121 group.add_argument('--reverse-only', action='store_true',
122 help='Do not create tables and indexes for searching')
123 group.add_argument('--enable-debug-statements', action='store_true',
124 help='Include debug warning statements in SQL code')
125 group.add_argument('--no-partitions', action='store_true',
126 help="""Do not partition search indices
127 (speeds up import of single country extracts)""")
128 group.add_argument('--no-updates', action='store_true',
129 help="""Do not keep tables that are only needed for
130 updating the database later""")
131 group = parser.add_argument_group('Expert options')
132 group.add_argument('--ignore-errors', action='store_true',
133 help='Continue import even when errors in SQL are present')
134 group.add_argument('--index-noanalyse', action='store_true',
135 help='Do not perform analyse operations during index')
140 params = ['setup.php']
142 params.extend(('--all', '--osm-file', args.osm_file))
144 if args.continue_at == 'load-data':
145 params.append('--load-data')
146 if args.continue_at in ('load-data', 'indexing'):
147 params.append('--index')
148 params.extend(('--create-search-indices', '--create-country-names',
150 if args.osm2pgsql_cache:
151 params.extend(('--osm2pgsql-cache', args.osm2pgsql_cache))
152 if args.reverse_only:
153 params.append('--reverse-only')
154 if args.enable_debug_statements:
155 params.append('--enable-debug-statements')
156 if args.no_partitions:
157 params.append('--no-partitions')
159 params.append('--drop')
160 if args.ignore_errors:
161 params.append('--ignore-errors')
162 if args.index_noanalyse:
163 params.append('--index-noanalyse')
165 return run_legacy_script(*params, nominatim_env=args)
170 Make database read-only.
172 About half of data in the Nominatim database is kept only to be able to
173 keep the data up-to-date with new changes made in OpenStreetMap. This
174 command drops all this data and only keeps the part needed for geocoding
177 This command has the same effect as the `--no-updates` option for imports.
181 def add_args(parser):
186 return run_legacy_script('setup.php', '--drop', nominatim_env=args)
189 class SetupSpecialPhrases:
191 Maintain special phrases.
195 def add_args(parser):
196 group = parser.add_argument_group('Input arguments')
197 group.add_argument('--from-wiki', action='store_true',
198 help='Pull special phrases from the OSM wiki.')
199 group = parser.add_argument_group('Output arguments')
200 group.add_argument('-o', '--output', default='-',
201 help="""File to write the preprocessed phrases to.
202 If omitted, it will be written to stdout.""")
206 if args.output != '-':
207 raise NotImplementedError('Only output to stdout is currently implemented.')
208 return run_legacy_script('specialphrases.php', '--wiki-import', nominatim_env=args)
211 class UpdateReplication:
213 Update the database using an online replication service.
217 def add_args(parser):
218 group = parser.add_argument_group('Arguments for initialisation')
219 group.add_argument('--init', action='store_true',
220 help='Initialise the update process')
221 group.add_argument('--no-update-functions', dest='update_functions',
222 action='store_false',
223 help="""Do not update the trigger function to
224 support differential updates.""")
225 group = parser.add_argument_group('Arguments for updates')
226 group.add_argument('--check-for-updates', action='store_true',
227 help='Check if new updates are available and exit')
228 group.add_argument('--once', action='store_true',
229 help="""Download and apply updates only once. When
230 not set, updates are continuously applied""")
231 group.add_argument('--no-index', action='store_false', dest='do_index',
232 help="""Do not index the new data. Only applicable
233 together with --once""")
238 import osmium # pylint: disable=W0611
239 except ModuleNotFoundError:
240 LOG.fatal("pyosmium not installed. Replication functions not available.\n"
241 "To install pyosmium via pip: pip3 install osmium")
244 from .tools import replication, refresh
246 conn = connect(args.config.get_libpq_dsn())
248 params = ['update.php']
250 LOG.warning("Initialising replication updates")
251 replication.init_replication(conn, args.config.REPLICATION_URL)
252 if args.update_functions:
253 LOG.warning("Create functions")
254 refresh.create_functions(conn, args.config, args.data_dir,
259 if args.check_for_updates:
260 params.append('--check-for-updates')
263 params.append('--import-osmosis')
265 params.append('--import-osmosis-all')
266 if not args.do_index:
267 params.append('--no-index')
269 return run_legacy_script(*params, nominatim_env=args)
274 Add additional data from a file or an online source.
276 Data is only imported, not indexed. You need to call `nominatim-update index`
277 to complete the process.
281 def add_args(parser):
282 group_name = parser.add_argument_group('Source')
283 group = group_name.add_mutually_exclusive_group(required=True)
284 group.add_argument('--file', metavar='FILE',
285 help='Import data from an OSM file')
286 group.add_argument('--diff', metavar='FILE',
287 help='Import data from an OSM diff file')
288 group.add_argument('--node', metavar='ID', type=int,
289 help='Import a single node from the API')
290 group.add_argument('--way', metavar='ID', type=int,
291 help='Import a single way from the API')
292 group.add_argument('--relation', metavar='ID', type=int,
293 help='Import a single relation from the API')
294 group.add_argument('--tiger-data', metavar='DIR',
295 help='Add housenumbers from the US TIGER census database.')
296 group = parser.add_argument_group('Extra arguments')
297 group.add_argument('--use-main-api', action='store_true',
298 help='Use OSM API instead of Overpass to download objects')
303 os.environ['NOMINATIM_TIGER_DATA_PATH'] = args.tiger_data
304 return run_legacy_script('setup.php', '--import-tiger-data', nominatim_env=args)
306 params = ['update.php']
308 params.extend(('--import-file', args.file))
310 params.extend(('--import-diff', args.diff))
312 params.extend(('--import-node', args.node))
314 params.extend(('--import-way', args.way))
316 params.extend(('--import-relation', args.relation))
317 if args.use_main_api:
318 params.append('--use-main-api')
319 return run_legacy_script(*params, nominatim_env=args)
324 Reindex all new and modified data.
328 def add_args(parser):
329 group = parser.add_argument_group('Filter arguments')
330 group.add_argument('--boundaries-only', action='store_true',
331 help="""Index only administrative boundaries.""")
332 group.add_argument('--no-boundaries', action='store_true',
333 help="""Index everything except administrative boundaries.""")
334 group.add_argument('--minrank', '-r', type=int, metavar='RANK', default=0,
335 help='Minimum/starting rank')
336 group.add_argument('--maxrank', '-R', type=int, metavar='RANK', default=30,
337 help='Maximum/finishing rank')
341 from .indexer.indexer import Indexer
343 indexer = Indexer(args.config.get_libpq_dsn(),
344 args.threads or _num_system_cpus() or 1)
346 if not args.no_boundaries:
347 indexer.index_boundaries(args.minrank, args.maxrank)
348 if not args.boundaries_only:
349 indexer.index_by_rank(args.minrank, args.maxrank)
351 if not args.no_boundaries and not args.boundaries_only:
352 indexer.update_status_table()
359 Recompute auxiliary data used by the indexing process.
361 These functions must not be run in parallel with other update commands.
365 def add_args(parser):
366 group = parser.add_argument_group('Data arguments')
367 group.add_argument('--postcodes', action='store_true',
368 help='Update postcode centroid table')
369 group.add_argument('--word-counts', action='store_true',
370 help='Compute frequency of full-word search terms')
371 group.add_argument('--address-levels', action='store_true',
372 help='Reimport address level configuration')
373 group.add_argument('--functions', action='store_true',
374 help='Update the PL/pgSQL functions in the database')
375 group.add_argument('--wiki-data', action='store_true',
376 help='Update Wikipedia/data importance numbers.')
377 group.add_argument('--importance', action='store_true',
378 help='Recompute place importances (expensive!)')
379 group.add_argument('--website', action='store_true',
380 help='Refresh the directory that serves the scripts for the web API')
381 group = parser.add_argument_group('Arguments for function refresh')
382 group.add_argument('--no-diff-updates', action='store_false', dest='diffs',
383 help='Do not enable code for propagating updates')
384 group.add_argument('--enable-debug-statements', action='store_true',
385 help='Enable debug warning statements in functions')
389 from .tools import refresh
391 conn = connect(args.config.get_libpq_dsn())
394 LOG.warning("Update postcodes centroid")
395 refresh.update_postcodes(conn, args.data_dir)
398 LOG.warning('Recompute frequency of full-word search terms')
399 refresh.recompute_word_counts(conn, args.data_dir)
401 if args.address_levels:
402 cfg = Path(args.config.ADDRESS_LEVEL_CONFIG)
403 LOG.warning('Updating address levels from %s', cfg)
404 refresh.load_address_levels_from_file(conn, cfg)
407 LOG.warning('Create functions')
408 refresh.create_functions(conn, args.config, args.data_dir,
409 args.diffs, args.enable_debug_statements)
412 run_legacy_script('setup.php', '--import-wikipedia-articles',
413 nominatim_env=args, throw_on_fail=True)
414 # Attention: importance MUST come after wiki data import.
416 run_legacy_script('update.php', '--recompute-importance',
417 nominatim_env=args, throw_on_fail=True)
419 run_legacy_script('setup.php', '--setup-website',
420 nominatim_env=args, throw_on_fail=True)
427 class AdminCheckDatabase:
429 Check that the database is complete and operational.
433 def add_args(parser):
438 return run_legacy_script('check_import_finished.php', nominatim_env=args)
443 Warm database caches for search and reverse queries.
447 def add_args(parser):
448 group = parser.add_argument_group('Target arguments')
449 group.add_argument('--search-only', action='store_const', dest='target',
451 help="Only pre-warm tables for search queries")
452 group.add_argument('--reverse-only', action='store_const', dest='target',
454 help="Only pre-warm tables for reverse queries")
458 params = ['warm.php']
459 if args.target == 'reverse':
460 params.append('--reverse-only')
461 if args.target == 'search':
462 params.append('--search-only')
463 return run_legacy_script(*params, nominatim_env=args)
468 Export addresses as CSV file from the database.
472 def add_args(parser):
473 group = parser.add_argument_group('Output arguments')
474 group.add_argument('--output-type', default='street',
475 choices=('continent', 'country', 'state', 'county',
476 'city', 'suburb', 'street', 'path'),
477 help='Type of places to output (default: street)')
478 group.add_argument('--output-format',
479 default='street;suburb;city;county;state;country',
480 help="""Semicolon-separated list of address types
481 (see --output-type). Multiple ranks can be
482 merged into one column by simply using a
483 comma-separated list.""")
484 group.add_argument('--output-all-postcodes', action='store_true',
485 help="""List all postcodes for address instead of
486 just the most likely one""")
487 group.add_argument('--language',
488 help="""Preferred language for output
489 (use local name, if omitted)""")
490 group = parser.add_argument_group('Filter arguments')
491 group.add_argument('--restrict-to-country', metavar='COUNTRY_CODE',
492 help='Export only objects within country')
493 group.add_argument('--restrict-to-osm-node', metavar='ID', type=int,
494 help='Export only children of this OSM node')
495 group.add_argument('--restrict-to-osm-way', metavar='ID', type=int,
496 help='Export only children of this OSM way')
497 group.add_argument('--restrict-to-osm-relation', metavar='ID', type=int,
498 help='Export only children of this OSM relation')
503 params = ['export.php',
504 '--output-type', args.output_type,
505 '--output-format', args.output_format]
506 if args.output_all_postcodes:
507 params.append('--output-all-postcodes')
509 params.extend(('--language', args.language))
510 if args.restrict_to_country:
511 params.extend(('--restrict-to-country', args.restrict_to_country))
512 if args.restrict_to_osm_node:
513 params.extend(('--restrict-to-osm-node', args.restrict_to_osm_node))
514 if args.restrict_to_osm_way:
515 params.extend(('--restrict-to-osm-way', args.restrict_to_osm_way))
516 if args.restrict_to_osm_relation:
517 params.extend(('--restrict-to-osm-relation', args.restrict_to_osm_relation))
519 return run_legacy_script(*params, nominatim_env=args)
522 ('street', 'housenumber and street'),
523 ('city', 'city, town or village'),
524 ('county', 'county'),
526 ('country', 'country'),
527 ('postalcode', 'postcode')
531 ('addressdetails', 'Include a breakdown of the address into elements.'),
532 ('extratags', """Include additional information if available
533 (e.g. wikipedia link, opening hours)."""),
534 ('namedetails', 'Include a list of alternative names.')
538 ('addressdetails', 'Include a breakdown of the address into elements.'),
539 ('keywords', 'Include a list of name keywords and address keywords.'),
540 ('linkedplaces', 'Include a details of places that are linked with this one.'),
541 ('hierarchy', 'Include details of places lower in the address hierarchy.'),
542 ('group_hierarchy', 'Group the places by type.'),
543 ('polygon_geojson', 'Include geometry of result.')
546 def _add_api_output_arguments(parser):
547 group = parser.add_argument_group('Output arguments')
548 group.add_argument('--format', default='jsonv2',
549 choices=['xml', 'json', 'jsonv2', 'geojson', 'geocodejson'],
550 help='Format of result')
551 for name, desc in EXTRADATA_PARAMS:
552 group.add_argument('--' + name, action='store_true', help=desc)
554 group.add_argument('--lang', '--accept-language', metavar='LANGS',
555 help='Preferred language order for presenting search results')
556 group.add_argument('--polygon-output',
557 choices=['geojson', 'kml', 'svg', 'text'],
558 help='Output geometry of results as a GeoJSON, KML, SVG or WKT.')
559 group.add_argument('--polygon-threshold', type=float, metavar='TOLERANCE',
560 help="""Simplify output geometry.
561 Parameter is difference tolerance in degrees.""")
566 Execute API search query.
570 def add_args(parser):
571 group = parser.add_argument_group('Query arguments')
572 group.add_argument('--query',
573 help='Free-form query string')
574 for name, desc in STRUCTURED_QUERY:
575 group.add_argument('--' + name, help='Structured query: ' + desc)
577 _add_api_output_arguments(parser)
579 group = parser.add_argument_group('Result limitation')
580 group.add_argument('--countrycodes', metavar='CC,..',
581 help='Limit search results to one or more countries.')
582 group.add_argument('--exclude_place_ids', metavar='ID,..',
583 help='List of search object to be excluded')
584 group.add_argument('--limit', type=int,
585 help='Limit the number of returned results')
586 group.add_argument('--viewbox', metavar='X1,Y1,X2,Y2',
587 help='Preferred area to find search results')
588 group.add_argument('--bounded', action='store_true',
589 help='Strictly restrict results to viewbox area')
591 group = parser.add_argument_group('Other arguments')
592 group.add_argument('--no-dedupe', action='store_false', dest='dedupe',
593 help='Do not remove duplicates from the result list')
599 params = dict(q=args.query)
601 params = {k : getattr(args, k) for k, _ in STRUCTURED_QUERY if getattr(args, k)}
603 for param, _ in EXTRADATA_PARAMS:
604 if getattr(args, param):
606 for param in ('format', 'countrycodes', 'exclude_place_ids', 'limit', 'viewbox'):
607 if getattr(args, param):
608 params[param] = getattr(args, param)
610 params['accept-language'] = args.lang
611 if args.polygon_output:
612 params['polygon_' + args.polygon_output] = '1'
613 if args.polygon_threshold:
614 params['polygon_threshold'] = args.polygon_threshold
616 params['bounded'] = '1'
618 params['dedupe'] = '0'
620 return run_api_script('search', args.project_dir,
621 phpcgi_bin=args.phpcgi_path, params=params)
625 Execute API reverse query.
629 def add_args(parser):
630 group = parser.add_argument_group('Query arguments')
631 group.add_argument('--lat', type=float, required=True,
632 help='Latitude of coordinate to look up (in WGS84)')
633 group.add_argument('--lon', type=float, required=True,
634 help='Longitude of coordinate to look up (in WGS84)')
635 group.add_argument('--zoom', type=int,
636 help='Level of detail required for the address')
638 _add_api_output_arguments(parser)
643 params = dict(lat=args.lat, lon=args.lon)
644 if args.zoom is not None:
645 params['zoom'] = args.zoom
647 for param, _ in EXTRADATA_PARAMS:
648 if getattr(args, param):
651 params['format'] = args.format
653 params['accept-language'] = args.lang
654 if args.polygon_output:
655 params['polygon_' + args.polygon_output] = '1'
656 if args.polygon_threshold:
657 params['polygon_threshold'] = args.polygon_threshold
659 return run_api_script('reverse', args.project_dir,
660 phpcgi_bin=args.phpcgi_path, params=params)
665 Execute API reverse query.
669 def add_args(parser):
670 group = parser.add_argument_group('Query arguments')
671 group.add_argument('--id', metavar='OSMID',
672 action='append', required=True, dest='ids',
673 help='OSM id to lookup in format <NRW><id> (may be repeated)')
675 _add_api_output_arguments(parser)
680 params = dict(osm_ids=','.join(args.ids))
682 for param, _ in EXTRADATA_PARAMS:
683 if getattr(args, param):
686 params['format'] = args.format
688 params['accept-language'] = args.lang
689 if args.polygon_output:
690 params['polygon_' + args.polygon_output] = '1'
691 if args.polygon_threshold:
692 params['polygon_threshold'] = args.polygon_threshold
694 return run_api_script('lookup', args.project_dir,
695 phpcgi_bin=args.phpcgi_path, params=params)
700 Execute API lookup query.
704 def add_args(parser):
705 group = parser.add_argument_group('Query arguments')
706 objs = group.add_mutually_exclusive_group(required=True)
707 objs.add_argument('--node', '-n', type=int,
708 help="Look up the OSM node with the given ID.")
709 objs.add_argument('--way', '-w', type=int,
710 help="Look up the OSM way with the given ID.")
711 objs.add_argument('--relation', '-r', type=int,
712 help="Look up the OSM relation with the given ID.")
713 objs.add_argument('--place_id', '-p', type=int,
714 help='Database internal identifier of the OSM object to look up.')
715 group.add_argument('--class', dest='object_class',
716 help="""Class type to disambiguated multiple entries
717 of the same object.""")
719 group = parser.add_argument_group('Output arguments')
720 for name, desc in DETAILS_SWITCHES:
721 group.add_argument('--' + name, action='store_true', help=desc)
722 group.add_argument('--lang', '--accept-language', metavar='LANGS',
723 help='Preferred language order for presenting search results')
728 params = dict(osmtype='N', osmid=args.node)
730 params = dict(osmtype='W', osmid=args.node)
732 params = dict(osmtype='R', osmid=args.node)
734 params = dict(place_id=args.place_id)
735 if args.object_class:
736 params['class'] = args.object_class
737 for name, _ in DETAILS_SWITCHES:
738 params[name] = '1' if getattr(args, name) else '0'
740 return run_api_script('details', args.project_dir,
741 phpcgi_bin=args.phpcgi_path, params=params)
746 Execute API status query.
750 def add_args(parser):
751 group = parser.add_argument_group('API parameters')
752 group.add_argument('--format', default='text', choices=['text', 'json'],
753 help='Format of result')
757 return run_api_script('status', args.project_dir,
758 phpcgi_bin=args.phpcgi_path,
759 params=dict(format=args.format))
762 def nominatim(**kwargs):
764 Command-line tools for importing, updating, administrating and
765 querying the Nominatim database.
767 parser = CommandlineParser('nominatim', nominatim.__doc__)
769 parser.add_subcommand('import', SetupAll)
770 parser.add_subcommand('freeze', SetupFreeze)
771 parser.add_subcommand('replication', UpdateReplication)
773 parser.add_subcommand('check-database', AdminCheckDatabase)
774 parser.add_subcommand('warm', AdminWarm)
776 parser.add_subcommand('special-phrases', SetupSpecialPhrases)
778 parser.add_subcommand('add-data', UpdateAddData)
779 parser.add_subcommand('index', UpdateIndex)
780 parser.add_subcommand('refresh', UpdateRefresh)
782 parser.add_subcommand('export', QueryExport)
784 if kwargs.get('phpcgi_path'):
785 parser.add_subcommand('search', APISearch)
786 parser.add_subcommand('reverse', APIReverse)
787 parser.add_subcommand('lookup', APILookup)
788 parser.add_subcommand('details', APIDetails)
789 parser.add_subcommand('status', APIStatus)
791 parser.parser.epilog = 'php-cgi not found. Query commands not available.'
793 return parser.run(**kwargs)