Merge pull request #2709 from lonvia/less-strict-country-assignment

[nominatim.git] / nominatim / tools / check_database.py

diff --git a/nominatim/tools/check_database.py b/nominatim/tools/check_database.py
index d4f793b46334c27f304fbfe898268a4f23f564a9..7ac31271568382aedc011e6dc46d8c3b3e8bb57c 100644 (file)
--- a/nominatim/tools/check_database.py
+++ b/nominatim/tools/check_database.py
@@ -1,3 +1,9 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# This file is part of Nominatim. (https://nominatim.org)
+#
+# Copyright (C) 2022 by the Nominatim developer community.
+# For a full list of authors see the git log.
 """
 Collection of functions that check if the database is complete and functional.
 """
@@ -17,6 +23,7 @@ class CheckState(Enum):
     FAIL = 1
     FATAL = 2
     NOT_APPLICABLE = 3
+    WARN = 4
 
 def _check(hint=None):
     """ Decorator for checks. It adds the function to the list of
@@ -24,6 +31,7 @@ def _check(hint=None):
     """
     def decorator(func):
         title = func.__doc__.split('\n', 1)[0].strip()
+
         def run_check(conn, config):
             print(title, end=' ... ')
             ret = func(conn, config)
@@ -33,6 +41,11 @@ def _check(hint=None):
                 params = {}
             if ret == CheckState.OK:
                 print('\033[92mOK\033[0m')
+            elif ret == CheckState.WARN:
+                print('\033[93mWARNING\033[0m')
+                if hint:
+                    print('')
+                    print(dedent(hint.format(**params)))
             elif ret == CheckState.NOT_APPLICABLE:
                 print('not applicable')
             else:
@@ -98,13 +111,12 @@ def _get_indexes(conn):
     if conn.table_exists('place'):
         indexes.extend(('idx_placex_pendingsector',
                         'idx_location_area_country_place_id',
-                        'idx_place_osm_unique'
-                       ))
+                        'idx_place_osm_unique'))
 
     return indexes
 
 
-### CHECK FUNCTIONS
+# CHECK FUNCTIONS
 #
 # Functions are exectured in the order they appear here.
 
@@ -166,7 +178,7 @@ def check_tokenizer(_, config):
         return CheckState.FAIL, dict(msg="""\
             Cannot load tokenizer. Did the import finish sucessfully?""")
 
-    result = tokenizer.check_database()
+    result = tokenizer.check_database(config)
 
     if result is None:
         return CheckState.OK
@@ -174,6 +186,23 @@ def check_tokenizer(_, config):
     return CheckState.FAIL, dict(msg=result)
 
 
+@_check(hint="""\
+             Wikipedia/Wikidata importance tables missing.
+             Quality of search results may be degraded. Reverse geocoding is unaffected.
+             See https://nominatim.org/release-docs/latest/admin/Import/#wikipediawikidata-rankings
+             """)
+def check_existance_wikipedia(conn, _):
+    """ Checking for wikipedia/wikidata data
+    """
+    if not conn.table_exists('search_name'):
+        return CheckState.NOT_APPLICABLE
+
+    with conn.cursor() as cur:
+        cnt = cur.scalar('SELECT count(*) FROM wikipedia_article')
+
+        return CheckState.WARN if cnt == 0 else CheckState.OK
+
+
 @_check(hint="""\
              The indexing didn't finish. {count} entries are not yet indexed.