Merge pull request #3122 from miku0/sanitizer-final

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

diff --git a/nominatim/tokenizer/base.py b/nominatim/tokenizer/base.py
index 5a3d3b1276aa7c6bf3bb36787967d519fe70aa79..f0fd9dd0e178c5a1319ce6e79a560cb438567f12 100644 (file)
--- a/nominatim/tokenizer/base.py
+++ b/nominatim/tokenizer/base.py
@@ -5,16 +5,17 @@
 # Copyright (C) 2022 by the Nominatim developer community.
 # For a full list of authors see the git log.
 """
-Abstract class defintions for tokenizers. These base classes are here
+Abstract class definitions for tokenizers. These base classes are here
 mainly for documentation purposes.
 """
 from abc import ABC, abstractmethod
-from typing import List, Tuple, Dict, Any
+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
-
-# pylint: disable=unnecessary-pass
+from nominatim.typing import Protocol
 
 class AbstractAnalyzer(ABC):
     """ The analyzer provides the functions for analysing names and building
@@ -80,7 +81,8 @@ class AbstractAnalyzer(ABC):
 
 
     @abstractmethod
-    def update_special_phrases(self, phrases: List[Tuple[str, str, str, str]],
+    def update_special_phrases(self,
+                               phrases: Iterable[Tuple[str, str, str, str]],
                                should_replace: bool) -> None:
         """ Update the tokenizer's special phrase tokens from the given
             list of special phrases.
@@ -112,7 +114,7 @@ class AbstractAnalyzer(ABC):
             the search index.
 
             Arguments:
-                place: Place information retrived from the database.
+                place: Place information retrieved from the database.
 
             Returns:
                 A JSON-serialisable structure that will be handed into
@@ -140,7 +142,7 @@ class AbstractTokenizer(ABC):
 
               init_db: When set to False, then initialisation of database
                 tables should be skipped. This option is only required for
-                migration purposes and can be savely ignored by custom
+                migration purposes and can be safely ignored by custom
                 tokenizers.
 
             TODO: can we move the init_db parameter somewhere else?
@@ -186,7 +188,7 @@ class AbstractTokenizer(ABC):
 
 
     @abstractmethod
-    def check_database(self, config: Configuration) -> str:
+    def check_database(self, config: Configuration) -> Optional[str]:
         """ Check that the database is set up correctly and ready for being
             queried.
 
@@ -230,3 +232,20 @@ class AbstractTokenizer(ABC):
             When used outside the with construct, the caller must ensure to
             call the close() function before destructing the analyzer.
         """
+
+
+    @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.
+    """
+
+    def create(self, dsn: str, data_dir: Path) -> AbstractTokenizer:
+        """ Factory for new tokenizers.
+        """