test/python/tools/test_refresh_wiki_data.py

   1 # SPDX-License-Identifier: GPL-2.0-only
   2 #
   3 # This file is part of Nominatim. (https://nominatim.org)
   4 #
   5 # Copyright (C) 2025 by the Nominatim developer community.
   6 # For a full list of authors see the git log.
   7 """
   8 Tests for correctly assigning wikipedia pages to places.
   9 """
  10 import gzip
  11 import csv
  12
  13 import pytest
  14
  15 from nominatim_db.tools.refresh import (import_wikipedia_articles,
  16                                         recompute_importance,
  17                                         create_functions)
  18
  19
  20 @pytest.fixture
  21 def wiki_csv(tmp_path, sql_preprocessor):
  22     def _import(data):
  23         with gzip.open(tmp_path / 'wikimedia-importance.csv.gz', mode='wt') as fd:
  24             writer = csv.DictWriter(fd, fieldnames=['language', 'type', 'title',
  25                                                     'importance', 'wikidata_id'],
  26                                     delimiter='\t', quotechar='|')
  27             writer.writeheader()
  28             for lang, title, importance, wd in data:
  29                 writer.writerow({'language': lang, 'type': 'a',
  30                                  'title': title, 'importance': str(importance),
  31                                  'wikidata_id': wd})
  32         return tmp_path
  33
  34     return _import
  35
  36
  37 @pytest.mark.parametrize('extra', [{'wikipedia:en': 'Test'},
  38                                    {'wikipedia': 'en:Test'},
  39                                    {'wikidata': 'Q123'}])
  40 def test_wikipedia(dsn, temp_db_conn, temp_db_cursor, table_factory,
  41                    def_config, wiki_csv, placex_table, extra):
  42     import_wikipedia_articles(dsn, wiki_csv([('en', 'Test', 0.3, 'Q123')]))
  43     create_functions(temp_db_conn, def_config)
  44
  45     content = temp_db_cursor.row_set(
  46         'SELECT language, title, importance, wikidata FROM wikimedia_importance')
  47     assert content == set([('en', 'Test', 0.3, 'Q123')])
  48
  49     place_id = placex_table.add(osm_id=12, extratags=extra)
  50     table_factory('search_name',
  51                   'place_id BIGINT, importance FLOAT',
  52                   [(place_id, 0.2)])
  53
  54     recompute_importance(temp_db_conn)
  55
  56     content = temp_db_cursor.row_set('SELECT wikipedia, importance FROM placex')
  57     assert content == set([('en:Test', 0.3)])
  58     simp = temp_db_cursor.scalar('SELECT importance FROM search_name WHERE place_id = %s',
  59                                  (place_id,))
  60     assert simp == 0.3
  61
  62
  63 def test_wikipedia_no_match(dsn, temp_db_conn, temp_db_cursor, def_config, wiki_csv,
  64                             placex_table, table_factory):
  65     import_wikipedia_articles(dsn, wiki_csv([('de', 'Test', 0.3, 'Q123')]))
  66     create_functions(temp_db_conn, def_config)
  67
  68     place_id = placex_table.add(osm_id=12, extratags={'wikipedia': 'en:Test'}, rank_search=10)
  69     table_factory('search_name',
  70                   'place_id BIGINT, importance FLOAT',
  71                   [(place_id, 0.2)])
  72
  73     recompute_importance(temp_db_conn)
  74
  75     content = temp_db_cursor.row_set('SELECT wikipedia, importance FROM placex')
  76     assert list(content) == [(None, pytest.approx(0.26667666))]
  77     simp = temp_db_cursor.scalar('SELECT importance FROM search_name WHERE place_id = %s',
  78                                  (place_id,))
  79     assert simp == pytest.approx(0.26667666)