reintroduce cutoffs when searching for very frequent words

[nominatim.git] / nominatim / tokenizer / legacy_tokenizer.py

diff --git a/nominatim/tokenizer/legacy_tokenizer.py b/nominatim/tokenizer/legacy_tokenizer.py
index 1b0b2980f9048f0fb42ac82e9b5bc8c8fe4f6298..93808cc39f3407458bb2d570d2a8740128f2c168 100644 (file)
--- a/nominatim/tokenizer/legacy_tokenizer.py
+++ b/nominatim/tokenizer/legacy_tokenizer.py
@@ -106,6 +106,7 @@ class LegacyTokenizer(AbstractTokenizer):
             This copies all necessary data in the project directory to make
             sure the tokenizer remains stable even over updates.
         """
+        assert config.project_dir is not None
         module_dir = _install_module(config.DATABASE_MODULE_PATH,
                                      config.lib_dir.module,
                                      config.project_dir / 'module')
@@ -127,6 +128,8 @@ class LegacyTokenizer(AbstractTokenizer):
     def init_from_project(self, config: Configuration) -> None:
         """ Initialise the tokenizer from the project directory.
         """
+        assert config.project_dir is not None
+
         with connect(self.dsn) as conn:
             self.normalization = properties.get_property(conn, DBCFG_NORMALIZATION)
 
@@ -149,6 +152,8 @@ class LegacyTokenizer(AbstractTokenizer):
     def update_sql_functions(self, config: Configuration) -> None:
         """ Reimport the SQL functions for this tokenizer.
         """
+        assert config.project_dir is not None
+
         with connect(self.dsn) as conn:
             max_word_freq = properties.get_property(conn, DBCFG_MAXWORDFREQ)
             modulepath = config.DATABASE_MODULE_PATH or \
@@ -193,6 +198,8 @@ class LegacyTokenizer(AbstractTokenizer):
             This is a special migration function for updating existing databases
             to new software versions.
         """
+        assert config.project_dir is not None
+
         self.normalization = config.TERM_NORMALIZATION
         module_dir = _install_module(config.DATABASE_MODULE_PATH,
                                      config.lib_dir.module,
@@ -203,7 +210,7 @@ class LegacyTokenizer(AbstractTokenizer):
             self._save_config(conn, config)
 
 
-    def update_statistics(self) -> None:
+    def update_statistics(self, config: Configuration, threads: int = 1) -> None:
         """ Recompute the frequency of full words.
         """
         with connect(self.dsn) as conn:
@@ -249,18 +256,29 @@ class LegacyTokenizer(AbstractTokenizer):
         return LegacyNameAnalyzer(self.dsn, normalizer)
 
 
+    def most_frequent_words(self, conn: Connection, num: int) -> List[str]:
+        """ Return a list of the `num` most frequent full words
+            in the database.
+        """
+        with conn.cursor() as cur:
+            cur.execute(""" SELECT word FROM word WHERE word is not null
+                              ORDER BY search_name_count DESC LIMIT %s""", (num,))
+            return list(s[0] for s in cur)
+
+
     def _install_php(self, config: Configuration, overwrite: bool = True) -> None:
         """ Install the php script for the tokenizer.
         """
-        php_file = self.data_dir / "tokenizer.php"
+        if config.lib_dir.php is not None:
+            php_file = self.data_dir / "tokenizer.php"
 
-        if not php_file.exists() or overwrite:
-            php_file.write_text(dedent(f"""\
-                <?php
-                @define('CONST_Max_Word_Frequency', {config.MAX_WORD_FREQUENCY});
-                @define('CONST_Term_Normalization_Rules', "{config.TERM_NORMALIZATION}");
-                require_once('{config.lib_dir.php}/tokenizer/legacy_tokenizer.php');
-                """), encoding='utf-8')
+            if not php_file.exists() or overwrite:
+                php_file.write_text(dedent(f"""\
+                    <?php
+                    @define('CONST_Max_Word_Frequency', {config.MAX_WORD_FREQUENCY});
+                    @define('CONST_Term_Normalization_Rules', "{config.TERM_NORMALIZATION}");
+                    require_once('{config.lib_dir.php}/tokenizer/legacy_tokenizer.php');
+                    """), encoding='utf-8')
 
 
     def _init_db_tables(self, config: Configuration) -> None:
@@ -557,14 +575,13 @@ class _TokenInfo:
     def add_street(self, conn: Connection, street: str) -> None:
         """ Add addr:street match terms.
         """
-        def _get_street(name: str) -> List[int]:
+        def _get_street(name: str) -> Optional[str]:
             with conn.cursor() as cur:
-                return cast(List[int],
+                return cast(Optional[str],
                             cur.scalar("SELECT word_ids_from_name(%s)::text", (name, )))
 
         tokens = self.cache.streets.get(street, _get_street)
-        if tokens:
-            self.data['street'] = tokens
+        self.data['street'] = tokens or '{}'
 
 
     def add_place(self, conn: Connection, place: str) -> None: