+# 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
+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 = 'SELECT place_id FROM placex'
+ 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
@staticmethod
@functools.lru_cache(maxsize=1)
def _index_sql(num_places):
- return """ UPDATE placex
- SET indexed_status = 0, address = v.addr, token_info = v.ti
- FROM (VALUES {}) as v(id, addr, ti)
- WHERE place_id = v.id
- """.format(','.join(["(%s, %s::hstore, %s::jsonb)"] * num_places))
+ return pysql.SQL(
+ """ UPDATE placex
+ 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(AbstractPlacexRunner.UPDATE_LINE, num_places))
- def get_place_details(self, worker, ids):
- worker.perform("""SELECT place_id, (placex_prepare_update(placex)).*
- FROM placex WHERE place_id IN %s""",
+ @staticmethod
+ def get_place_details(worker, ids):
+ 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)
"""
def name(self):
- return "rank {}".format(self.rank)
+ return f"rank {self.rank}"
def sql_count_objects(self):
- return """SELECT count(*) FROM placex
- WHERE rank_address = {} and indexed_status > 0
- """.format(self.rank)
+ return pysql.SQL("""SELECT count(*) FROM placex
+ WHERE rank_address = {} and indexed_status > 0
+ """).format(pysql.Literal(self.rank))
def sql_get_objects(self):
- return """{} WHERE indexed_status > 0 and rank_address = {}
- ORDER BY geometry_sector
- """.format(self.SELECT_SQL, self.rank)
+ return self.SELECT_SQL + pysql.SQL(
+ """WHERE indexed_status > 0 and rank_address = {}
+ ORDER BY geometry_sector
+ """).format(pysql.Literal(self.rank))
class BoundaryRunner(AbstractPlacexRunner):
"""
def name(self):
- return "boundaries rank {}".format(self.rank)
+ return f"boundaries rank {self.rank}"
def sql_count_objects(self):
- return """SELECT count(*) FROM placex
- WHERE indexed_status > 0
- AND rank_search = {}
- AND class = 'boundary' and type = 'administrative'
- """.format(self.rank)
+ return pysql.SQL("""SELECT count(*) FROM placex
+ WHERE indexed_status > 0
+ AND rank_search = {}
+ AND class = 'boundary' and type = 'administrative'
+ """).format(pysql.Literal(self.rank))
def sql_get_objects(self):
- return """{} WHERE indexed_status > 0 and rank_search = {}
- and class = 'boundary' and type = 'administrative'
- ORDER BY partition, admin_level
- """.format(self.SELECT_SQL, self.rank)
+ return self.SELECT_SQL + pysql.SQL(
+ """WHERE indexed_status > 0 and rank_search = {}
+ and class = 'boundary' and type = 'administrative'
+ ORDER BY partition, admin_level
+ """).format(pysql.Literal(self.rank))
class InterpolationRunner:
@staticmethod
def sql_get_objects():
- return """SELECT place_id, get_interpolation_address(address, osm_id) as address
+ return """SELECT place_id
FROM location_property_osmline
WHERE indexed_status > 0
ORDER BY geometry_sector"""
+ @staticmethod
+ def get_place_details(worker, ids):
+ worker.perform("""SELECT place_id, get_interpolation_address(address, osm_id) as address
+ FROM location_property_osmline WHERE place_id IN %s""",
+ (tuple((p[0] for p in ids)), ))
+
+
@staticmethod
@functools.lru_cache(maxsize=1)
def _index_sql(num_places):
- return """ UPDATE location_property_osmline
- SET indexed_status = 0, address = v.addr, token_info = v.ti
- FROM (VALUES {}) as v(id, addr, ti)
- WHERE place_id = v.id
- """.format(','.join(["(%s, %s::hstore, %s::jsonb)"] * num_places))
+ return pysql.SQL("""UPDATE location_property_osmline
+ SET indexed_status = 0, address = v.addr, token_info = v.ti
+ FROM (VALUES {}) as v(id, addr, ti)
+ WHERE place_id = v.id
+ """).format(_mk_valuelist("(%s, %s::hstore, %s::jsonb)", num_places))
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)))
+ values.append(_analyze_place(place, self.analyzer))
worker.perform(self._index_sql(len(places)), values)
@staticmethod
def index_places(worker, ids):
- worker.perform(""" UPDATE location_postcode SET indexed_status = 0
- WHERE place_id IN ({})
- """.format(','.join((str(i[0]) for i in ids))))
+ worker.perform(pysql.SQL("""UPDATE location_postcode SET indexed_status = 0
+ WHERE place_id IN ({})""")
+ .format(pysql.SQL(',').join((pysql.Literal(i[0]) for i in ids))))
projects / nominatim.git / blobdiff