Merge pull request #2391 from lonvia/fix-sonar-issues

[nominatim.git] / nominatim / tools / special_phrases / sp_importer.py

diff --git a/nominatim/tools/special_phrases/sp_importer.py b/nominatim/tools/special_phrases/sp_importer.py
index e82f721e6aca52e32ff0f8d1f8574a6fb8c07d0a..681990fa6175be2935f6b59e763fe4dab6239792 100644 (file)
--- a/nominatim/tools/special_phrases/sp_importer.py
+++ b/nominatim/tools/special_phrases/sp_importer.py
@@ -20,10 +20,16 @@ from nominatim.errors import UsageError
 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 sp loader which load the phrases from an external source.
     """
@@ -42,10 +48,14 @@ class SPImporter():
         #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()
@@ -57,14 +67,15 @@ class SPImporter():
             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()
@@ -138,7 +149,7 @@ class SPImporter():
         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):
@@ -159,7 +170,7 @@ class SPImporter():
             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()
@@ -188,7 +199,7 @@ class SPImporter():
         """
             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 {{}} {} 
@@ -203,7 +214,7 @@ class SPImporter():
             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)
+        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:
@@ -225,7 +236,7 @@ class SPImporter():
         """
             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),
@@ -269,9 +280,8 @@ class SPImporter():
                                 (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