move warm script to python code

nominatim/tokenizer/base.py

@@ -13,6 +13,7 @@ from typing import List, Tuple, Dict, Any, Optional, Iterable
 from pathlib import Path
 
 from nominatim.config import Configuration
+from nominatim.db.connection import Connection
 from nominatim.data.place_info import PlaceInfo
 from nominatim.typing import Protocol
 
@@ -233,6 +234,13 @@ class AbstractTokenizer(ABC):
         """
 
 
+    @abstractmethod
+    def most_frequent_words(self, conn: Connection, num: int) -> List[str]:
+        """ Return a list of the `num` most frequent full words
+            in the database.
+        """
+
+
 class TokenizerModule(Protocol):
     """ Interface that must be exported by modules that implement their
         own tokenizer.

nominatim/tokenizer/icu_tokenizer.py

@@ -183,6 +183,18 @@ class ICUTokenizer(AbstractTokenizer):
                                self.loader.make_token_analysis())
 
 
+    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, sum((info->'count')::int) as count
+                             FROM word WHERE type = 'W'
+                             GROUP BY word
+                             ORDER BY count DESC LIMIT %s""", (num,))
+            return list(s[0].split('@')[0] for s in cur)
+
+
     def _install_php(self, phpdir: Path, overwrite: bool = True) -> None:
         """ Install the php script for the tokenizer.
         """

nominatim/tokenizer/legacy_tokenizer.py

@@ -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.
         """