from nominatim.tools.special_phrases.importer_statistics import SpecialPhrasesImporterStatistics
LOG = logging.getLogger()
+
+def _classtype_table(phrase_class, phrase_type):
+ """ Return the name of the table for the given class and type.
+ """
+ return f'place_classtype_{phrase_class}_{phrase_type}'
+
class SPImporter():
# pylint: disable-msg=too-many-instance-attributes
"""
- Class handling the process of special phrases importations into the database.
+ Class handling the process of special phrases importation into the database.
- Take a SPLoader which load the phrases from an external source.
+ Take a sp loader which load the phrases from an external source.
"""
def __init__(self, config, phplib_dir, db_connection, sp_loader) -> None:
self.config = config
# This set will contain all existing phrases to be added.
# It contains tuples with the following format: (lable, class, type, operator)
self.word_phrases = set()
- #This set will contain all existing place_classtype tables which doesn't match any
- #special phrases class/type on the wiki.
+ # This set will contain all existing place_classtype tables which doesn't match any
+ # special phrases class/type on the wiki.
self.table_phrases_to_delete = set()
- def import_phrases(self, tokenizer):
+ def import_phrases(self, tokenizer, should_replace):
"""
- Iterate through all specified languages and
- extract corresponding special phrases from the wiki.
+ Iterate through all SpecialPhrases extracted from the
+ loader and import them into the database.
+
+ If should_replace is set to True only the loaded phrases
+ will be kept into the database. All other phrases already
+ in the database will be removed.
"""
LOG.warning('Special phrases importation starting')
self._fetch_existing_place_classtype_tables()
- #Store pairs of class/type for further processing
+ # Store pairs of class/type for further processing
class_type_pairs = set()
for loaded_phrases in self.sp_loader:
for phrase in loaded_phrases:
result = self._process_phrase(phrase)
if result:
- class_type_pairs.update(result)
+ class_type_pairs.add(result)
self._create_place_classtype_table_and_indexes(class_type_pairs)
- self._remove_non_existent_tables_from_db()
+ if should_replace:
+ self._remove_non_existent_tables_from_db()
self.db_connection.commit()
with tokenizer.name_analyzer() as analyzer:
- analyzer.update_special_phrases(self.word_phrases)
+ analyzer.update_special_phrases(self.word_phrases, should_replace)
LOG.warning('Import done.')
self.statistics_handler.notify_import_done()
Return the class/type pair corresponding to the phrase.
"""
- #blacklisting: disallow certain class/type combinations
- if (
- phrase.p_class in self.black_list.keys() and
- phrase.p_type in self.black_list[phrase.p_class]
- ): return None
+ # blacklisting: disallow certain class/type combinations
+ if phrase.p_class in self.black_list.keys() \
+ and phrase.p_type in self.black_list[phrase.p_class]:
+ return None
- #whitelisting: if class is in whitelist, allow only tags in the list
- if (
- phrase.p_class in self.white_list.keys() and
- phrase.p_type not in self.white_list[phrase.p_class]
- ): return None
+ # whitelisting: if class is in whitelist, allow only tags in the list
+ if phrase.p_class in self.white_list.keys() \
+ and phrase.p_type not in self.white_list[phrase.p_class]:
+ return None
- #sanity check, in case somebody added garbage in the wiki
+ # sanity check, in case somebody added garbage in the wiki
if not self._check_sanity(phrase):
self.statistics_handler.notify_one_phrase_invalid()
return None
self.word_phrases.add((phrase.p_label, phrase.p_class,
phrase.p_type, phrase.p_operator))
- return set({(phrase.p_class, phrase.p_type)})
+ return (phrase.p_class, phrase.p_type)
def _create_place_classtype_table_and_indexes(self, class_type_pairs):
sql_tablespace = self.config.TABLESPACE_AUX_DATA
if sql_tablespace:
- sql_tablespace = ' TABLESPACE '+sql_tablespace
+ sql_tablespace = ' TABLESPACE ' + sql_tablespace
with self.db_connection.cursor() as db_cursor:
db_cursor.execute("CREATE INDEX idx_placex_classtype ON placex (class, type)")
phrase_class = pair[0]
phrase_type = pair[1]
- table_name = 'place_classtype_{}_{}'.format(phrase_class, phrase_type)
+ table_name = _classtype_table(phrase_class, phrase_type)
if table_name in self.table_phrases_to_delete:
self.statistics_handler.notify_one_table_ignored()
- #Remove this table from the ones to delete as it match a class/type
- #still existing on the special phrases of the wiki.
+ # Remove this table from the ones to delete as it match a
+ # class/type still existing on the special phrases of the wiki.
self.table_phrases_to_delete.remove(table_name)
- #So dont need to create the table and indexes.
+ # So don't need to create the table and indexes.
continue
- #Table creation
+ # Table creation
self._create_place_classtype_table(sql_tablespace, phrase_class, phrase_type)
- #Indexes creation
+ # Indexes creation
self._create_place_classtype_indexes(sql_tablespace, phrase_class, phrase_type)
- #Grant access on read to the web user.
+ # Grant access on read to the web user.
self._grant_access_to_webuser(phrase_class, phrase_type)
self.statistics_handler.notify_one_table_created()
"""
Create table place_classtype of the given phrase_class/phrase_type if doesn't exit.
"""
- table_name = 'place_classtype_{}_{}'.format(phrase_class, phrase_type)
+ table_name = _classtype_table(phrase_class, phrase_type)
with self.db_connection.cursor() as db_cursor:
db_cursor.execute(SQL("""
- CREATE TABLE IF NOT EXISTS {{}} {}
- AS SELECT place_id AS place_id,st_centroid(geometry) AS centroid FROM placex
+ CREATE TABLE IF NOT EXISTS {{}} {}
+ AS SELECT place_id AS place_id,st_centroid(geometry) AS centroid FROM placex
WHERE class = {{}} AND type = {{}}""".format(sql_tablespace))
.format(Identifier(table_name), Literal(phrase_class),
Literal(phrase_type)))
Create indexes on centroid and place_id for the place_classtype table.
"""
index_prefix = 'idx_place_classtype_{}_{}_'.format(phrase_class, phrase_type)
- base_table = 'place_classtype_{}_{}'.format(phrase_class, phrase_type)
- #Index on centroid
+ base_table = _classtype_table(phrase_class, phrase_type)
+ # Index on centroid
if not self.db_connection.index_exists(index_prefix + 'centroid'):
with self.db_connection.cursor() as db_cursor:
db_cursor.execute(SQL("""
.format(Identifier(index_prefix + 'centroid'),
Identifier(base_table)), sql_tablespace)
- #Index on place_id
+ # Index on place_id
if not self.db_connection.index_exists(index_prefix + 'place_id'):
with self.db_connection.cursor() as db_cursor:
db_cursor.execute(SQL(
"""
Grant access on read to the table place_classtype for the webuser.
"""
- table_name = 'place_classtype_{}_{}'.format(phrase_class, phrase_type)
+ table_name = _classtype_table(phrase_class, phrase_type)
with self.db_connection.cursor() as db_cursor:
db_cursor.execute(SQL("""GRANT SELECT ON {} TO {}""")
.format(Identifier(table_name),
Delete the place_classtype tables.
"""
LOG.warning('Cleaning database...')
- #Array containing all queries to execute. Contain tuples of format (query, parameters)
+ # Array containing all queries to execute.
+ # Contains tuples of format (query, parameters)
queries_parameters = []
- #Delete place_classtype tables corresponding to class/type which are not on the wiki anymore
+ # Delete place_classtype tables corresponding to class/type which
+ # are not on the wiki anymore.
for table in self.table_phrases_to_delete:
self.statistics_handler.notify_one_table_deleted()
query = SQL('DROP TABLE IF EXISTS {}').format(Identifier(table))
file, extension = os.path.splitext(file_path)
json_file_path = Path(file + '.json').resolve()
- if extension not in('.php', '.json'):
+ if extension not in ('.php', '.json'):
raise UsageError('The custom NOMINATIM_PHRASE_CONFIG file has not a valid extension.')
if extension == '.php' and not isfile(json_file_path):
(self.phplib_dir / 'migration/PhraseSettingsToJson.php').resolve(),
file_path], check=True)
LOG.warning('special_phrase configuration file has been converted to json.')
- return json_file_path
except subprocess.CalledProcessError:
LOG.error('Error while converting %s to json.', file_path)
raise
- else:
- return json_file_path
+
+ return json_file_path
projects / nominatim.git / blobdiff