X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/bc8b2d4ae0dbaef64448ddcb530de9626da9d82d..a6dab5e300de9c5664f714eb2c9290d18f01067f:/nominatim/indexer/runners.py?ds=sidebyside diff --git a/nominatim/indexer/runners.py b/nominatim/indexer/runners.py index 068d7d0f..c8495ee4 100644 --- a/nominatim/indexer/runners.py +++ b/nominatim/indexer/runners.py @@ -1,21 +1,33 @@ +# SPDX-License-Identifier: GPL-2.0-only +# +# This file is part of Nominatim. (https://nominatim.org) +# +# Copyright (C) 2022 by the Nominatim developer community. +# For a full list of authors see the git log. """ Mix-ins that provide the actual commands for the indexer for various indexing tasks. """ import functools -import psycopg2.extras from psycopg2 import sql as pysql +import psycopg2.extras + +from nominatim.data.place_info import PlaceInfo # pylint: disable=C0111 def _mk_valuelist(template, num): return pysql.SQL(',').join([pysql.SQL(template)] * num) +def _analyze_place(place, analyzer): + return psycopg2.extras.Json(analyzer.process_place(PlaceInfo(place))) + class AbstractPlacexRunner: """ Returns SQL commands for indexing of the placex table. """ SELECT_SQL = pysql.SQL('SELECT place_id FROM placex ') + UPDATE_LINE = "(%s, %s::hstore, %s::hstore, %s::int, %s::jsonb)" def __init__(self, rank, analyzer): self.rank = rank @@ -27,24 +39,27 @@ class AbstractPlacexRunner: def _index_sql(num_places): return pysql.SQL( """ UPDATE placex - SET indexed_status = 0, address = v.addr, token_info = v.ti - FROM (VALUES {}) as v(id, addr, ti) + SET indexed_status = 0, address = v.addr, token_info = v.ti, + name = v.name, linked_place_id = v.linked_place_id + FROM (VALUES {}) as v(id, name, addr, linked_place_id, ti) WHERE place_id = v.id - """).format(_mk_valuelist("(%s, %s::hstore, %s::jsonb)", num_places)) + """).format(_mk_valuelist(AbstractPlacexRunner.UPDATE_LINE, num_places)) @staticmethod def get_place_details(worker, ids): - worker.perform("""SELECT place_id, (placex_prepare_update(placex)).* - FROM placex WHERE place_id IN %s""", + worker.perform("""SELECT place_id, extra.* + FROM placex, LATERAL placex_indexing_prepare(placex) as extra + WHERE place_id IN %s""", (tuple((p[0] for p in ids)), )) def index_places(self, worker, places): values = [] for place in places: - values.extend((place[x] for x in ('place_id', 'address'))) - values.append(psycopg2.extras.Json(self.analyzer.process_place(place))) + for field in ('place_id', 'name', 'address', 'linked_place_id'): + values.append(place[field]) + values.append(_analyze_place(place, self.analyzer)) worker.perform(self._index_sql(len(places)), values) @@ -54,7 +69,7 @@ class RankRunner(AbstractPlacexRunner): """ def name(self): - return "rank {}".format(self.rank) + return f"rank {self.rank}" def sql_count_objects(self): return pysql.SQL("""SELECT count(*) FROM placex @@ -74,7 +89,7 @@ class BoundaryRunner(AbstractPlacexRunner): """ def name(self): - return "boundaries rank {}".format(self.rank) + return f"boundaries rank {self.rank}" def sql_count_objects(self): return pysql.SQL("""SELECT count(*) FROM placex @@ -138,7 +153,7 @@ class InterpolationRunner: values = [] for place in places: values.extend((place[x] for x in ('place_id', 'address'))) - values.append(psycopg2.extras.Json(self.analyzer.process_place(place))) + values.append(_analyze_place(place, self.analyzer)) worker.perform(self._index_sql(len(places)), values)