add type annotations to ICU tokenizer helper modules

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

diff --git a/nominatim/tokenizer/base.py b/nominatim/tokenizer/base.py
index 5a3d3b1276aa7c6bf3bb36787967d519fe70aa79..6484ff6a06a798b206c9c8e54bf947cf2da4b1d4 100644 (file)
--- a/nominatim/tokenizer/base.py
+++ b/nominatim/tokenizer/base.py
@@ -10,12 +10,13 @@ mainly for documentation purposes.
 """
 from abc import ABC, abstractmethod
 from typing import List, Tuple, Dict, Any
+from pathlib import Path
+
+from typing_extensions import Protocol
 
 from nominatim.config import Configuration
 from nominatim.data.place_info import PlaceInfo
 
-# pylint: disable=unnecessary-pass
-
 class AbstractAnalyzer(ABC):
     """ The analyzer provides the functions for analysing names and building
         the token database.
@@ -230,3 +231,13 @@ class AbstractTokenizer(ABC):
             When used outside the with construct, the caller must ensure to
             call the close() function before destructing the analyzer.
         """
+
+
+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.
+        """