Merge pull request #2629 from tareqpi/country-names-yaml-configuration

Move default country names into yaml configuration
This commit is contained in:
Sarah Hoffmann
2022-04-04 09:04:25 +02:00
committed by GitHub
7 changed files with 26887 additions and 295 deletions

View File

@@ -226,8 +226,7 @@ if (BUILD_IMPORTER)
PATTERN __pycache__ EXCLUDE)
install(DIRECTORY lib-sql DESTINATION ${NOMINATIM_LIBDIR})
install(FILES data/country_name.sql
${COUNTRY_GRID_FILE}
install(FILES ${COUNTRY_GRID_FILE}
data/words.sql
DESTINATION ${NOMINATIM_DATADIR})
endif()

File diff suppressed because one or more lines are too long

View File

@@ -43,8 +43,8 @@ BEGIN
END IF;
END LOOP;
-- anything will do as a fallback - just take the first name type thing there is
RETURN trim((avals(name))[1]);
-- as a fallback - take the last element since it is the default name
RETURN trim((avals(name))[array_length(avals(name), 1)]);
END;
$$
LANGUAGE plpgsql IMMUTABLE;

View File

@@ -12,13 +12,14 @@ import psycopg2.extras
from nominatim.db import utils as db_utils
from nominatim.db.connection import connect
class _CountryInfo:
""" Caches country-specific properties from the configuration file.
"""
def __init__(self):
self._info = {}
self._key_prefix = 'name'
def load(self, config):
""" Load the country properties from the configuration files,
@@ -33,16 +34,24 @@ class _CountryInfo:
elif not isinstance(prop['languages'], list):
prop['languages'] = [x.strip()
for x in prop['languages'].split(',')]
if 'names' not in prop or prop['names'] is None:
prop['names'] = {self._key_prefix: {}}
def items(self):
""" Return tuples of (country_code, property dict) as iterable.
"""
return self._info.items()
def key_prefix(self):
""" Return the prefix that will be attached to the keys of the country
names values when storing them in the database
"""
return self._key_prefix
_COUNTRY_INFO = _CountryInfo()
def setup_country_config(config):
""" Load country properties from the configuration file.
Needs to be called before using any other functions in this
@@ -61,9 +70,11 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False):
""" Create and populate the tables with basic static data that provides
the background for geocoding. Data is assumed to not yet exist.
"""
db_utils.execute_file(dsn, sql_dir / 'country_name.sql')
db_utils.execute_file(dsn, sql_dir / 'country_osm_grid.sql.gz')
def add_prefix_to_keys(names, prefix):
return {prefix+':'+k: v for k, v in names.items()}
params = []
for ccode, props in _COUNTRY_INFO.items():
if ccode is not None and props is not None:
@@ -71,16 +82,26 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False):
partition = 0
else:
partition = props.get('partition')
lang = props['languages'][0] if len(props['languages']) == 1 else None
params.append((ccode, partition, lang))
lang = props['languages'][0] if len(
props['languages']) == 1 else None
name = add_prefix_to_keys(props.get('names').get(
_COUNTRY_INFO.key_prefix()), _COUNTRY_INFO.key_prefix())
params.append((ccode, name, lang, partition))
with connect(dsn) as conn:
with conn.cursor() as cur:
psycopg2.extras.register_hstore(cur)
cur.execute(
""" CREATE TABLE public.country_name (
country_code character varying(2),
name public.hstore,
derived_name public.hstore,
country_default_language_code text,
partition integer
); """)
cur.execute_values(
""" UPDATE country_name
SET partition = part, country_default_language_code = lang
FROM (VALUES %s) AS v (cc, part, lang)
WHERE country_code = v.cc""", params)
""" INSERT INTO public.country_name
(country_code, name, country_default_language_code, partition) VALUES %s
""", params)
conn.commit()
@@ -94,8 +115,9 @@ def create_country_names(conn, tokenizer, languages=None):
languages = languages.split(',')
def _include_key(key):
return key == 'name' or \
(key.startswith('name:') and (not languages or key[5:] in languages))
return key == _COUNTRY_INFO.key_prefix() or \
(key.startswith(_COUNTRY_INFO.key_prefix()+':') and
(not languages or key[len(_COUNTRY_INFO.key_prefix())+1:] in languages))
with conn.cursor() as cur:
psycopg2.extras.register_hstore(cur)
@@ -112,7 +134,8 @@ def create_country_names(conn, tokenizer, languages=None):
# country names (only in languages as provided)
if name:
names.update(((k, v) for k, v in name.items() if _include_key(k)))
names.update(((k, v)
for k, v in name.items() if _include_key(k)))
analyzer.add_country_names(code, names)

File diff suppressed because it is too large Load Diff

View File

@@ -1,3 +1,3 @@
[pytest]
markers =
sanitizer_params
sanitizer_params

View File

@@ -12,20 +12,24 @@ import pytest
from nominatim.tools import country_info
@pytest.fixture(autouse=True)
def read_config(def_config):
country_info.setup_country_config(def_config)
@pytest.mark.parametrize("no_partitions", (True, False))
def test_setup_country_tables(src_dir, temp_db_with_extensions, dsn, temp_db_cursor,
def_config, no_partitions):
read_config(def_config)
country_info.setup_country_tables(dsn, src_dir / 'data', no_partitions)
assert temp_db_cursor.table_exists('country_name')
assert temp_db_cursor.table_rows('country_name') == \
temp_db_cursor.scalar('SELECT count(DISTINCT country_code) FROM country_name')
temp_db_cursor.scalar(
'SELECT count(DISTINCT country_code) FROM country_name')
partitions = temp_db_cursor.row_set("SELECT DISTINCT partition FROM country_name")
partitions = temp_db_cursor.row_set(
"SELECT DISTINCT partition FROM country_name")
if no_partitions:
assert partitions == {(0, )}
else:
@@ -37,7 +41,8 @@ def test_setup_country_tables(src_dir, temp_db_with_extensions, dsn, temp_db_cur
@pytest.mark.parametrize("languages", (None, ' fr,en'))
def test_create_country_names(temp_db_with_extensions, temp_db_conn, temp_db_cursor,
table_factory, tokenizer_mock, languages):
table_factory, tokenizer_mock, languages, def_config):
read_config(def_config)
table_factory('country_name', 'country_code varchar(2), name hstore',
content=(('us', '"name"=>"us1","name:af"=>"us2"'),
@@ -51,11 +56,66 @@ def test_create_country_names(temp_db_with_extensions, temp_db_conn, temp_db_cur
assert len(tokenizer.analyser_cache['countries']) == 2
result_set = {k: set(v.values()) for k, v in tokenizer.analyser_cache['countries']}
result_set = {k: set(v.values())
for k, v in tokenizer.analyser_cache['countries']}
if languages:
assert result_set == {'us' : set(('us', 'us1', 'United States')),
'fr' : set(('fr', 'Fra', 'Fren'))}
assert result_set == {'us': set(('us', 'us1', 'United States')),
'fr': set(('fr', 'Fra', 'Fren'))}
else:
assert result_set == {'us' : set(('us', 'us1', 'us2', 'United States')),
'fr' : set(('fr', 'Fra', 'Fren'))}
assert result_set == {'us': set(('us', 'us1', 'us2', 'United States')),
'fr': set(('fr', 'Fra', 'Fren'))}
def test_setup_country_config_languages_not_loaded(project_env):
(project_env.project_dir / 'country_settings.yaml').write_text("""
de:
partition: 3
names:
name:
default: Deutschland
""")
country_info._COUNTRY_INFO._info = None
country_info.setup_country_config(project_env)
assert country_info._COUNTRY_INFO._info == {'de': {'partition': 3,
'languages': [], 'names': {'name': {'default': 'Deutschland'}}}}
def test_setup_country_config_name_not_loaded(project_env):
(project_env.project_dir / 'country_settings.yaml').write_text("""
de:
partition: 3
languages: de
names:
""")
country_info._COUNTRY_INFO._info = None
country_info.setup_country_config(project_env)
assert country_info._COUNTRY_INFO._info == {'de': {'partition': 3,
'languages': ['de'], 'names': {'name': {}}}}
def test_setup_country_config_names_not_loaded(project_env):
(project_env.project_dir / 'country_settings.yaml').write_text("""
de:
partition: 3
languages: de
""")
country_info._COUNTRY_INFO._info = None
country_info.setup_country_config(project_env)
assert country_info._COUNTRY_INFO._info == {'de': {'partition': 3,
'languages': ['de'], 'names': {'name': {}}}}
def test_setup_country_config_special_character(project_env):
(project_env.project_dir / 'country_settings.yaml').write_text("""
bq:
partition: 250
languages: nl
names:
name:
default: "\\N"
""")
country_info._COUNTRY_INFO._info = None
country_info.setup_country_config(project_env)
assert country_info._COUNTRY_INFO._info == {'bq': {'partition': 250,
'languages': ['nl'], 'names': {'name': {'default': '\x85'}}}}