Improved performance of the postcodes query and some code cleaning

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

diff --git a/nominatim/clicmd/refresh.py b/nominatim/clicmd/refresh.py
index 8e69cacaf315fa8ab1c6c6a82bbfe0f2bea75efc..fbc23350c7212478f8c8b0c9fa484ec154f9f070 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
 
 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
 
 # Do not repeat documentation of subcommand classes.
 # pylint: disable=C0111


@@ -20,6 +19,8 @@ class UpdateRefresh:
 
     These functions must not be run in parallel with other update commands.
     """
 
     These functions must not be run in parallel with other update commands.
     """

+    def __init__(self):
+        self.tokenizer = None

 
     @staticmethod
     def add_args(parser):
 
     @staticmethod
     def add_args(parser):


@@ -44,45 +45,70 @@ class UpdateRefresh:
         group.add_argument('--enable-debug-statements', action='store_true',
                            help='Enable debug warning statements in functions')
 
         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:
 
         if args.postcodes:

-            LOG.warning("Update postcodes centroid")
-            conn = connect(args.config.get_libpq_dsn())
-            refresh.update_postcodes(conn, args.sqllib_dir)
-            conn.close()
+            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')
 
         if args.word_counts:
             LOG.warning('Recompute frequency of full-word search terms')

-            conn = connect(args.config.get_libpq_dsn())
-            refresh.recompute_word_counts(conn, args.sqllib_dir)
-            conn.close()
+            refresh.recompute_word_counts(args.config.get_libpq_dsn(), args.sqllib_dir)

 
         if args.address_levels:
             cfg = Path(args.config.ADDRESS_LEVEL_CONFIG)
             LOG.warning('Updating address levels from %s', cfg)
 
         if args.address_levels:
             cfg = Path(args.config.ADDRESS_LEVEL_CONFIG)
             LOG.warning('Updating address levels from %s', cfg)

-            conn = connect(args.config.get_libpq_dsn())
-            refresh.load_address_levels_from_file(conn, cfg)
-            conn.close()
+            with connect(args.config.get_libpq_dsn()) as conn:
+                refresh.load_address_levels_from_file(conn, cfg)

 
         if args.functions:
             LOG.warning('Create functions')
 
         if args.functions:
             LOG.warning('Create functions')

-            conn = connect(args.config.get_libpq_dsn())
-            refresh.create_functions(conn, args.config, args.sqllib_dir,
-                                     args.diffs, args.enable_debug_statements)
-            conn.close()
+            with connect(args.config.get_libpq_dsn()) as conn:
+                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:
 
         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:
         # 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:
         if args.website:

-            run_legacy_script('setup.php', '--setup-website',
-                              nominatim_env=args, throw_on_fail=True)
+            webdir = args.project_dir / 'website'
+            LOG.warning('Setting up website directory at %s', webdir)
+            with connect(args.config.get_libpq_dsn()) as conn:
+                refresh.setup_website(webdir, args.config, conn)

 
         return 0
 
         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