def check_database_integrity(context):
""" Check some generic constraints on the tables.
"""
- # place_addressline should not have duplicate (place_id, address_place_id)
- cur = context.db.cursor()
- cur.execute("""SELECT count(*) FROM
- (SELECT place_id, address_place_id, count(*) as c
- FROM place_addressline GROUP BY place_id, address_place_id) x
- WHERE c > 1""")
- assert cur.fetchone()[0] == 0, "Duplicates found in place_addressline"
+ with context.db.cursor() as cur:
+ # place_addressline should not have duplicate (place_id, address_place_id)
+ cur.execute("""SELECT count(*) FROM
+ (SELECT place_id, address_place_id, count(*) as c
+ FROM place_addressline GROUP BY place_id, address_place_id) x
+ WHERE c > 1""")
+ assert cur.fetchone()[0] == 0, "Duplicates found in place_addressline"
+
+ # word table must not have empty word_tokens
+ if context.nominatim.tokenizer != 'legacy':
+ cur.execute("SELECT count(*) FROM word WHERE word_token = ''")
+ assert cur.fetchone()[0] == 0, "Empty word tokens found in word table"
+
################################ GIVEN ##################################
for tests on data that looks up members.
"""
with context.db.cursor() as cur:
- for r in context.table:
- last_node = 0
- last_way = 0
- parts = []
- if r['members']:
- members = []
- for m in r['members'].split(','):
- mid = NominatimID(m)
- if mid.typ == 'N':
- parts.insert(last_node, int(mid.oid))
- last_node += 1
- last_way += 1
- elif mid.typ == 'W':
- parts.insert(last_way, int(mid.oid))
- last_way += 1
- else:
- parts.append(int(mid.oid))
-
- members.extend((mid.typ.lower() + mid.oid, mid.cls or ''))
- else:
- members = None
-
- tags = chain.from_iterable([(h[5:], r[h]) for h in r.headings if h.startswith("tags+")])
-
- cur.execute("""INSERT INTO planet_osm_rels (id, way_off, rel_off, parts, members, tags)
- VALUES (%s, %s, %s, %s, %s, %s)""",
- (r['id'], last_node, last_way, parts, members, list(tags)))
+ cur.execute("SELECT value FROM osm2pgsql_properties WHERE property = 'db_format'")
+ row = cur.fetchone()
+ if row is None or row[0] == '1':
+ for r in context.table:
+ last_node = 0
+ last_way = 0
+ parts = []
+ if r['members']:
+ members = []
+ for m in r['members'].split(','):
+ mid = NominatimID(m)
+ if mid.typ == 'N':
+ parts.insert(last_node, int(mid.oid))
+ last_node += 1
+ last_way += 1
+ elif mid.typ == 'W':
+ parts.insert(last_way, int(mid.oid))
+ last_way += 1
+ else:
+ parts.append(int(mid.oid))
+
+ members.extend((mid.typ.lower() + mid.oid, mid.cls or ''))
+ else:
+ members = None
+
+ tags = chain.from_iterable([(h[5:], r[h]) for h in r.headings if h.startswith("tags+")])
+
+ cur.execute("""INSERT INTO planet_osm_rels (id, way_off, rel_off, parts, members, tags)
+ VALUES (%s, %s, %s, %s, %s, %s)""",
+ (r['id'], last_node, last_way, parts, members, list(tags)))
+ else:
+ for r in context.table:
+ if r['members']:
+ members = []
+ for m in r['members'].split(','):
+ mid = NominatimID(m)
+ members.append({'ref': mid.oid, 'role': mid.cls or '', 'type': mid.typ})
+ else:
+ members = []
+
+ tags = {h[5:]: r[h] for h in r.headings if h.startswith("tags+")}
+
+ cur.execute("""INSERT INTO planet_osm_rels (id, tags, members)
+ VALUES (%s, %s, %s)""",
+ (r['id'], psycopg2.extras.Json(tags),
+ psycopg2.extras.Json(members)))
@given("the ways")
def add_data_to_planet_ways(context):
tests on that that looks up node ids in this table.
"""
with context.db.cursor() as cur:
+ cur.execute("SELECT value FROM osm2pgsql_properties WHERE property = 'db_format'")
+ row = cur.fetchone()
+ json_tags = row is not None and row[0] != '1'
for r in context.table:
- tags = chain.from_iterable([(h[5:], r[h]) for h in r.headings if h.startswith("tags+")])
+ if json_tags:
+ tags = psycopg2.extras.Json({h[5:]: r[h] for h in r.headings if h.startswith("tags+")})
+ else:
+ tags = list(chain.from_iterable([(h[5:], r[h])
+ for h in r.headings if h.startswith("tags+")]))
nodes = [ int(x.strip()) for x in r['nodes'].split(',') ]
cur.execute("INSERT INTO planet_osm_ways (id, nodes, tags) VALUES (%s, %s, %s)",
- (r['id'], nodes, list(tags)))
+ (r['id'], nodes, tags))
################################ WHEN ##################################
context.nominatim.run_nominatim('refresh', '--functions')
with context.db.cursor() as cur:
for row in context.table:
- PlaceColumn(context).add_row(row, False).db_insert(cur)
+ col = PlaceColumn(context).add_row(row, False)
+ col.db_delete(cur)
+ col.db_insert(cur)
+ cur.execute('SELECT flush_deleted_places()')
context.nominatim.reindex_placex(context.db)
check_database_integrity(context)
"""
context.nominatim.run_nominatim('refresh', '--functions')
with context.db.cursor() as cur:
+ cur.execute('TRUNCATE place_to_be_deleted')
for oid in oids.split(','):
NominatimID(oid).query_osm_id(cur, 'DELETE FROM place WHERE {}')
+ cur.execute('SELECT flush_deleted_places()')
context.nominatim.reindex_placex(context.db)
def check_place_contents(context, table, exact):
""" Check contents of place/placex tables. Each row represents a table row
and all data must match. Data not present in the expected table, may
- be arbitry. The rows are identified via the 'object' column which must
+ be arbitrary. The rows are identified via the 'object' column which must
have an identifier of the form '<NRW><osm id>[:<class>]'. When multiple
rows match (for example because 'class' was left out and there are
multiple entries for the given OSM object) then all must match. All
if exact:
cur.execute('SELECT osm_type, osm_id, class from {}'.format(table))
- assert expected_content == set([(r[0], r[1], r[2]) for r in cur])
+ actual = set([(r[0], r[1], r[2]) for r in cur])
+ assert expected_content == actual, \
+ f"Missing entries: {expected_content - actual}\n" \
+ f"Not expected in table: {actual - expected_content}"
@then("(?P<table>placex|place) has no entry for (?P<oid>.*)")
def check_search_name_contents(context, exclude):
""" Check contents of place/placex tables. Each row represents a table row
and all data must match. Data not present in the expected table, may
- be arbitry. The rows are identified via the 'object' column which must
+ be arbitrary. The rows are identified via the 'object' column which must
have an identifier of the form '<NRW><osm id>[:<class>]'. All
expected rows are expected to be present with at least one database row.
"""
def check_location_postcode(context):
""" Check full contents for location_postcode table. Each row represents a table row
and all data must match. Data not present in the expected table, may
- be arbitry. The rows are identified via 'country' and 'postcode' columns.
+ be arbitrary. The rows are identified via 'country' and 'postcode' columns.
All rows must be present as excepted and there must not be additional
rows.
"""
def check_place_addressline(context):
""" Check the contents of the place_addressline table. Each row represents
a table row and all data must match. Data not present in the expected
- table, may be arbitry. The rows are identified via the 'object' column,
+ table, may be arbitrary. The rows are identified via the 'object' column,
representing the addressee and the 'address' column, representing the
address item.
"""
DBRow(oid, res, context).assert_row(row, ('start', 'end'))
- assert not todo
+ assert not todo, f"Unmatched lines in table: {list(context.table[i] for i in todo)}"
+
+@then("location_property_osmline contains(?P<exact> exactly)?")
+def check_place_contents(context, exact):
+ """ Check contents of the interpolation table. Each row represents a table row
+ and all data must match. Data not present in the expected table, may
+ be arbitrary. The rows are identified via the 'object' column which must
+ have an identifier of the form '<osm id>[:<startnumber>]'. When multiple
+ rows match (for example because 'startnumber' was left out and there are
+ multiple entries for the given OSM object) then all must match. All
+ expected rows are expected to be present with at least one database row.
+ When 'exactly' is given, there must not be additional rows in the database.
+ """
+ with context.db.cursor(cursor_factory=psycopg2.extras.DictCursor) as cur:
+ expected_content = set()
+ for row in context.table:
+ if ':' in row['object']:
+ nid, start = row['object'].split(':', 2)
+ start = int(start)
+ else:
+ nid, start = row['object'], None
+
+ query = """SELECT *, ST_AsText(linegeo) as geomtxt,
+ ST_GeometryType(linegeo) as geometrytype
+ FROM location_property_osmline WHERE osm_id=%s"""
+
+ if ':' in row['object']:
+ query += ' and startnumber = %s'
+ params = [int(val) for val in row['object'].split(':', 2)]
+ else:
+ params = (int(row['object']), )
+ cur.execute(query, params)
+ assert cur.rowcount > 0, "No rows found for " + row['object']
+
+ for res in cur:
+ if exact:
+ expected_content.add((res['osm_id'], res['startnumber']))
+
+ DBRow(nid, res, context).assert_row(row, ['object'])
+
+ if exact:
+ cur.execute('SELECT osm_id, startnumber from location_property_osmline')
+ actual = set([(r[0], r[1]) for r in cur])
+ assert expected_content == actual, \
+ f"Missing entries: {expected_content - actual}\n" \
+ f"Not expected in table: {actual - expected_content}"
projects / nominatim.git / blobdiff