Merge pull request #3328 from lonvia/word-count-into-new-table

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

diff --git a/nominatim/tokenizer/legacy_tokenizer.py b/nominatim/tokenizer/legacy_tokenizer.py
index a50dedb2da7f2ffb51ad836445d2282bd0408280..2d28a8b29891623e72e7f8c2f0f7b7c9cafbd160 100644 (file)
--- a/nominatim/tokenizer/legacy_tokenizer.py
+++ b/nominatim/tokenizer/legacy_tokenizer.py
@@ -210,7 +210,7 @@ class LegacyTokenizer(AbstractTokenizer):
             self._save_config(conn, config)
 
 
-    def update_statistics(self) -> None:
+    def update_statistics(self, _: Configuration) -> None:
         """ Recompute the frequency of full words.
         """
         with connect(self.dsn) as conn:
@@ -256,6 +256,16 @@ 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.
         """
@@ -564,14 +574,13 @@ 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: