hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-02-14 01:47:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #3986 from lonvia/rework-tiger-tests

Browse Source 
Rework unit tests for import of tiger data
This commit is contained in:
Sarah Hoffmann
2026-02-13 14:11:04 +01:00
committed by GitHub
parent 3f14f89bdf 8ed7a3875a
commit 2ddb19c0b0
3 changed files with 88 additions and 135 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
lib-sql/tables/search_name.sql
View File

@@ -7,7 +7,7 @@
DROP TABLE IF EXISTS search_name;
{% if not db.reverse_only %}
{% if not create_reverse_only %}
CREATE TABLE search_name (
place_id BIGINT NOT NULL,

9
src/nominatim_db/tools/database_import.py
View File

@@ -152,12 +152,11 @@ def create_tables(conn: Connection, config: Configuration, reverse_only: bool =
When `reverse_only` is True, then the main table for searching will
be skipped and only reverse search is possible.
"""
sql = SQLPreprocessor(conn, config)
sql.env.globals['db']['reverse_only'] = reverse_only
SQLPreprocessor(conn, config).run_sql_file(conn, 'tables.sql',
create_reverse_only=reverse_only)
sql.run_sql_file(conn, 'tables.sql')
sql.run_sql_file(conn, 'grants.sql')
# reinitiate the preprocessor to get all the newly created tables
SQLPreprocessor(conn, config).run_sql_file(conn, 'grants.sql')
def create_table_triggers(conn: Connection, config: Configuration) -> None:

212
test/python/tools/test_tiger_data.py
View File

@@ -2,7 +2,7 @@
#
# This file is part of Nominatim. (https://nominatim.org)
#
# Copyright (C) 2025 by the Nominatim developer community.
# Copyright (C) 2026 by the Nominatim developer community.
# For a full list of authors see the git log.
"""
Test for tiger data function
@@ -13,59 +13,10 @@ from textwrap import dedent
import pytest
import pytest_asyncio # noqa: F401
from nominatim_db.db.connection import execute_scalar
from nominatim_db.tools import tiger_data, freeze
from nominatim_db.tools import tiger_data
from nominatim_db.errors import UsageError
class MockTigerTable:
def __init__(self, conn):
self.conn = conn
with conn.cursor() as cur:
cur.execute("""CREATE TABLE tiger (linegeo GEOMETRY,
start INTEGER,
stop INTEGER,
interpol TEXT,
token_info JSONB,
postcode TEXT)""")
# We need this table to determine if the database is frozen or not
cur.execute("CREATE TABLE place (number INTEGER)")
# We need this table to determine if the database is in reverse-only mode
cur.execute("CREATE TABLE search_name (place_id BIGINT)")
def count(self):
return execute_scalar(self.conn, "SELECT count(*) FROM tiger")
def row(self):
with self.conn.cursor() as cur:
cur.execute("SELECT * FROM tiger LIMIT 1")
return cur.fetchone()
@pytest.fixture
def tiger_table(def_config, temp_db_conn, sql_preprocessor,
temp_db_with_extensions, tmp_path):
def_config.lib_dir.sql = tmp_path / 'sql'
def_config.lib_dir.sql.mkdir()
(def_config.lib_dir.sql / 'tiger_import_start.sql').write_text(
"""CREATE OR REPLACE FUNCTION tiger_line_import(linegeo GEOMETRY, start INTEGER,
stop INTEGER, interpol TEXT,
token_info JSONB, postcode TEXT)
RETURNS INTEGER AS $$
INSERT INTO tiger VALUES(linegeo, start, stop, interpol, token_info, postcode)
RETURNING 1
$$ LANGUAGE SQL;""", encoding='utf-8')
(def_config.lib_dir.sql / 'tiger_import_finish.sql').write_text(
"""DROP FUNCTION tiger_line_import (linegeo GEOMETRY, in_startnumber INTEGER,
in_endnumber INTEGER, interpolationtype TEXT,
token_info JSONB, in_postcode TEXT);""", encoding='utf-8')
return MockTigerTable(temp_db_conn)
@pytest.fixture
def csv_factory(tmp_path):
def _mk_file(fname, hnr_from=1, hnr_to=9, interpol='odd', street='Main St',
@@ -80,107 +31,110 @@ def csv_factory(tmp_path):
return _mk_file
@pytest.mark.parametrize("threads", (1, 5))
@pytest.mark.asyncio
async def test_add_tiger_data(def_config, src_dir, tiger_table, tokenizer_mock, threads):
await tiger_data.add_tiger_data(str(src_dir / 'test' / 'testdb' / 'tiger'),
def_config, threads, tokenizer_mock())
class TestTiger:
assert tiger_table.count() == 6213
@pytest.fixture(autouse=True)
def setup(self, temp_db_conn, placex_row, load_sql):
load_sql('tables/search_name.sql', create_reverse_only=False)
load_sql('tables/tiger.sql')
# fake parent roads
for x in range(-870, -863):
for y in range(323, 328):
placex_row(rank_search=26, rank_address=26,
geom=f"LINESTRING({x/10 - 0.1} {y/10}, {x/10 + 0.1} {y/10})")
@pytest.mark.parametrize("threads", (1, 5))
@pytest.mark.asyncio
async def test_add_tiger_data_database_frozen(def_config, src_dir, temp_db_conn, tiger_table,
tokenizer_mock, threads):
freeze.drop_update_tables(temp_db_conn)
temp_db_conn.execute("""
CREATE OR REPLACE FUNCTION get_partition(cc VARCHAR(10)) RETURNS INTEGER AS $$
SELECT 0;
$$ LANGUAGE sql;
CREATE OR REPLACE FUNCTION token_matches_street(i JSONB, s INT[]) RETURNS BOOLEAN AS $$
SELECT false
$$ LANGUAGE SQL IMMUTABLE STRICT PARALLEL SAFE;
""")
await tiger_data.add_tiger_data(str(src_dir / 'test' / 'testdb' / 'tiger'),
def_config, threads, tokenizer_mock())
assert tiger_table.count() == 6213
@pytest.mark.asyncio
async def test_add_tiger_data_reverse_only(def_config, src_dir, temp_db_conn, tiger_table,
tokenizer_mock):
with temp_db_conn.cursor() as cur:
cur.execute("DROP TABLE search_name")
temp_db_conn.commit()
with pytest.raises(UsageError,
match="Cannot perform tiger import: required tables are missing. "
"See https://github.com/osm-search/Nominatim/issues/2463 for details."):
@pytest.mark.parametrize("threads", (1, 5))
@pytest.mark.asyncio
async def test_add_tiger_data_database_frozen(self, def_config, src_dir, temp_db_cursor,
tokenizer_mock, threads):
await tiger_data.add_tiger_data(str(src_dir / 'test' / 'testdb' / 'tiger'),
def_config, 1, tokenizer_mock())
def_config, threads, tokenizer_mock())
assert tiger_table.count() == 0
assert temp_db_cursor.table_rows('location_property_tiger') == 6209
@pytest.mark.asyncio
async def test_add_tiger_data_reverse_only(self, def_config, src_dir, temp_db_cursor,
tokenizer_mock):
temp_db_cursor.execute("DROP TABLE search_name")
@pytest.mark.asyncio
async def test_add_tiger_data_no_files(def_config, tiger_table, tokenizer_mock,
tmp_path):
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
with pytest.raises(UsageError,
match="Cannot perform tiger import: required tables are missing. "
"See https://github.com/osm-search/Nominatim/issues/2463 for details."):
await tiger_data.add_tiger_data(str(src_dir / 'test' / 'testdb' / 'tiger'),
def_config, 1, tokenizer_mock())
assert tiger_table.count() == 0
assert temp_db_cursor.table_rows('location_property_tiger') == 0
@pytest.mark.asyncio
async def test_add_tiger_data_no_files(self, def_config, temp_db_cursor, tokenizer_mock,
tmp_path):
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
@pytest.mark.asyncio
async def test_add_tiger_data_bad_file(def_config, tiger_table, tokenizer_mock,
tmp_path):
sqlfile = tmp_path / '1010.csv'
sqlfile.write_text("""Random text""", encoding='utf-8')
assert temp_db_cursor.table_rows('location_property_tiger') == 0
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
@pytest.mark.asyncio
async def test_add_tiger_data_bad_file(self, def_config, temp_db_cursor, tokenizer_mock,
tmp_path):
sqlfile = tmp_path / '1010.csv'
sqlfile.write_text('Random text', encoding='utf-8')
assert tiger_table.count() == 0
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
assert temp_db_cursor.table_rows('location_property_tiger') == 0
@pytest.mark.asyncio
async def test_add_tiger_data_hnr_nan(def_config, tiger_table, tokenizer_mock,
csv_factory, tmp_path):
csv_factory('file1', hnr_from=99)
csv_factory('file2', hnr_from='L12')
csv_factory('file3', hnr_to='12.4')
@pytest.mark.asyncio
async def test_add_tiger_data_hnr_nan(self, def_config, temp_db_cursor, tokenizer_mock,
csv_factory, tmp_path):
csv_factory('file1', hnr_to=99)
csv_factory('file2', hnr_from='L12')
csv_factory('file3', hnr_to='12.4')
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
await tiger_data.add_tiger_data(str(tmp_path), def_config, 1, tokenizer_mock())
assert tiger_table.count() == 1
assert tiger_table.row().start == 99
rows = temp_db_cursor.row_set("""
SELECT startnumber, endnumber FROM location_property_tiger""")
assert rows == {(1, 99)}
@pytest.mark.parametrize("threads", (1, 5))
@pytest.mark.asyncio
async def test_add_tiger_data_tarfile(def_config, tiger_table, tokenizer_mock,
tmp_path, src_dir, threads):
tar = tarfile.open(str(tmp_path / 'sample.tar.gz'), "w:gz")
tar.add(str(src_dir / 'test' / 'testdb' / 'tiger' / '01001.csv'))
tar.close()
@pytest.mark.parametrize("threads", (1, 5))
@pytest.mark.asyncio
async def test_add_tiger_data_tarfile(self, def_config, temp_db_cursor, tokenizer_mock,
tmp_path, src_dir, threads):
tar = tarfile.open(str(tmp_path / 'sample.tar.gz'), "w:gz")
tar.add(str(src_dir / 'test' / 'testdb' / 'tiger' / '01001.csv'))
tar.close()
await tiger_data.add_tiger_data(str(tmp_path / 'sample.tar.gz'), def_config, threads,
tokenizer_mock())
await tiger_data.add_tiger_data(str(tmp_path / 'sample.tar.gz'), def_config, threads,
tokenizer_mock())
assert tiger_table.count() == 6213
assert temp_db_cursor.table_rows('location_property_tiger') == 6209
@pytest.mark.asyncio
async def test_add_tiger_data_bad_tarfile(self, def_config, tokenizer_mock, tmp_path):
tarfile = tmp_path / 'sample.tar.gz'
tarfile.write_text("""Random text""", encoding='utf-8')
@pytest.mark.asyncio
async def test_add_tiger_data_bad_tarfile(def_config, tiger_table, tokenizer_mock,
tmp_path):
tarfile = tmp_path / 'sample.tar.gz'
tarfile.write_text("""Random text""", encoding='utf-8')
with pytest.raises(UsageError):
await tiger_data.add_tiger_data(str(tarfile), def_config, 1, tokenizer_mock())
with pytest.raises(UsageError):
await tiger_data.add_tiger_data(str(tarfile), def_config, 1, tokenizer_mock())
@pytest.mark.asyncio
async def test_add_tiger_data_empty_tarfile(self, def_config, temp_db_cursor, tokenizer_mock,
tmp_path):
tar = tarfile.open(str(tmp_path / 'sample.tar.gz'), "w:gz")
tar.add(__file__)
tar.close()
await tiger_data.add_tiger_data(str(tmp_path / 'sample.tar.gz'), def_config, 1,
tokenizer_mock())
@pytest.mark.asyncio
async def test_add_tiger_data_empty_tarfile(def_config, tiger_table, tokenizer_mock,
tmp_path):
tar = tarfile.open(str(tmp_path / 'sample.tar.gz'), "w:gz")
tar.add(__file__)
tar.close()
await tiger_data.add_tiger_data(str(tmp_path / 'sample.tar.gz'), def_config, 1,
tokenizer_mock())
assert tiger_table.count() == 0
assert temp_db_cursor.table_rows('location_property_tiger') == 0