Merge pull request #2472 from lonvia/word-count-computation

[nominatim.git] / nominatim / clicmd / refresh.py

diff --git a/nominatim/clicmd/refresh.py b/nominatim/clicmd/refresh.py
index f68e185ac2281ae98bf9053211c7b7efb3cd7d2b..e7d7d7ba1d33cc032b3c0260a5ffc3d05f772344 100644 (file)
--- a/nominatim/clicmd/refresh.py
+++ b/nominatim/clicmd/refresh.py
@@ -4,8 +4,7 @@ Implementation of 'refresh' subcommand.
 import logging
 from pathlib import Path
 
-from ..db.connection import connect
-from ..tools.exec_utils import run_legacy_script
+from nominatim.db.connection import connect
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111
@@ -18,8 +17,16 @@ class UpdateRefresh:
     """\
     Recompute auxiliary data used by the indexing process.
 
-    These functions must not be run in parallel with other update commands.
+    This sub-commands updates various static data and functions in the database.
+    It usually needs to be run after changing various aspects of the
+    configuration. The configuration documentation will mention the exact
+    command to use in such case.
+
+    Warning: the 'update' command must not be run in parallel with other update
+             commands like 'replication' or 'add-data'.
     """
+    def __init__(self):
+        self.tokenizer = None
 
     @staticmethod
     def add_args(parser):
@@ -33,7 +40,7 @@ class UpdateRefresh:
         group.add_argument('--functions', action='store_true',
                            help='Update the PL/pgSQL functions in the database')
         group.add_argument('--wiki-data', action='store_true',
-                           help='Update Wikipedia/data importance numbers.')
+                           help='Update Wikipedia/data importance numbers')
         group.add_argument('--importance', action='store_true',
                            help='Recompute place importances (expensive!)')
         group.add_argument('--website', action='store_true',
@@ -44,19 +51,28 @@ class UpdateRefresh:
         group.add_argument('--enable-debug-statements', action='store_true',
                            help='Enable debug warning statements in functions')
 
-    @staticmethod
-    def run(args):
-        from ..tools import refresh
+
+    def run(self, args):
+        from ..tools import refresh, postcodes
+        from ..indexer.indexer import Indexer
+
 
         if args.postcodes:
-            LOG.warning("Update postcodes centroid")
-            with connect(args.config.get_libpq_dsn()) as conn:
-                refresh.update_postcodes(conn, args.sqllib_dir)
+            if postcodes.can_compute(args.config.get_libpq_dsn()):
+                LOG.warning("Update postcodes centroid")
+                tokenizer = self._get_tokenizer(args.config)
+                postcodes.update_postcodes(args.config.get_libpq_dsn(),
+                                           args.project_dir, tokenizer)
+                indexer = Indexer(args.config.get_libpq_dsn(), tokenizer,
+                                  args.threads or 1)
+                indexer.index_postcodes()
+            else:
+                LOG.error("The place table doesn't exist. "
+                          "Postcode updates on a frozen database is not possible.")
 
         if args.word_counts:
-            LOG.warning('Recompute frequency of full-word search terms')
-            with connect(args.config.get_libpq_dsn()) as conn:
-                refresh.recompute_word_counts(conn, args.sqllib_dir)
+            LOG.warning('Recompute word statistics')
+            self._get_tokenizer(args.config).update_statistics()
 
         if args.address_levels:
             cfg = Path(args.config.ADDRESS_LEVEL_CONFIG)
@@ -67,19 +83,38 @@ class UpdateRefresh:
         if args.functions:
             LOG.warning('Create functions')
             with connect(args.config.get_libpq_dsn()) as conn:
-                refresh.create_functions(conn, args.config, args.sqllib_dir,
+                refresh.create_functions(conn, args.config,
                                          args.diffs, args.enable_debug_statements)
+                self._get_tokenizer(args.config).update_sql_functions(args.config)
 
         if args.wiki_data:
-            run_legacy_script('setup.php', '--import-wikipedia-articles',
-                              nominatim_env=args, throw_on_fail=True)
+            data_path = Path(args.config.WIKIPEDIA_DATA_PATH
+                             or args.project_dir)
+            LOG.warning('Import wikipdia article importance from %s', data_path)
+            if refresh.import_wikipedia_articles(args.config.get_libpq_dsn(),
+                                                 data_path) > 0:
+                LOG.fatal('FATAL: Wikipedia importance dump file not found')
+                return 1
+
         # Attention: importance MUST come after wiki data import.
         if args.importance:
-            run_legacy_script('update.php', '--recompute-importance',
-                              nominatim_env=args, throw_on_fail=True)
+            LOG.warning('Update importance values for database')
+            with connect(args.config.get_libpq_dsn()) as conn:
+                refresh.recompute_importance(conn)
+
         if args.website:
             webdir = args.project_dir / 'website'
             LOG.warning('Setting up website directory at %s', webdir)
-            refresh.setup_website(webdir, args.phplib_dir, args.config)
+            with connect(args.config.get_libpq_dsn()) as conn:
+                refresh.setup_website(webdir, args.config, conn)
 
         return 0
+
+
+    def _get_tokenizer(self, config):
+        if self.tokenizer is None:
+            from ..tokenizer import factory as tokenizer_factory
+
+            self.tokenizer = tokenizer_factory.get_tokenizer_for_db(config)
+
+        return self.tokenizer