add type annotations to ICU tokenizer helper modules

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

diff --git a/nominatim/tokenizer/factory.py b/nominatim/tokenizer/factory.py
index 1079c935985fd4ed92a24e6f30db63461b832906..67e221949911b19dcd10943d5833f5430194da1d 100644 (file)
--- a/nominatim/tokenizer/factory.py
+++ b/nominatim/tokenizer/factory.py
@@ -1,3 +1,9 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# This file is part of Nominatim. (https://nominatim.org)
+#
+# Copyright (C) 2022 by the Nominatim developer community.
+# For a full list of authors see the git log.
 """
 Functions for creating a tokenizer or initialising the right one for an
 existing database.
@@ -13,32 +19,41 @@ database.
 A tokenizer usually also includes PHP code for querying. The appropriate PHP
 normalizer module is installed, when the tokenizer is created.
 """
+from typing import Optional
 import logging
 import importlib
+from pathlib import Path
 
-from ..errors import UsageError
-from ..db import properties
-from ..db.connection import connect
+from nominatim.errors import UsageError
+from nominatim.db import properties
+from nominatim.db.connection import connect
+from nominatim.config import Configuration
+from nominatim.tokenizer.base import AbstractTokenizer, TokenizerModule
 
 LOG = logging.getLogger()
 
-def _import_tokenizer(name):
+def _import_tokenizer(name: str) -> TokenizerModule:
     """ Load the tokenizer.py module from project directory.
     """
-    try:
-        return importlib.import_module('nominatim.tokenizer.' + name + '_tokenizer')
-    except ModuleNotFoundError as exp:
+    src_file = Path(__file__).parent / (name + '_tokenizer.py')
+    if not src_file.is_file():
         LOG.fatal("No tokenizer named '%s' available. "
                   "Check the setting of NOMINATIM_TOKENIZER.", name)
-        raise UsageError('Tokenizer not found') from exp
+        raise UsageError('Tokenizer not found')
 
+    return importlib.import_module('nominatim.tokenizer.' + name + '_tokenizer')
 
-def create_tokenizer(config):
+
+def create_tokenizer(config: Configuration, init_db: bool = True,
+                     module_name: Optional[str] = None) -> AbstractTokenizer:
     """ Create a new tokenizer as defined by the given configuration.
 
         The tokenizer data and code is copied into the 'tokenizer' directory
         of the project directory and the tokenizer loaded from its new location.
     """
+    if module_name is None:
+        module_name = config.TOKENIZER
+
     # Create the directory for the tokenizer data
     basedir = config.project_dir / 'tokenizer'
     if not basedir.exists():
@@ -47,18 +62,19 @@ def create_tokenizer(config):
         LOG.fatal("Tokenizer directory '%s' cannot be created.", basedir)
         raise UsageError("Tokenizer setup failed.")
 
-    tokenizer_module = _import_tokenizer(config.TOKENIZER)
+    # Import and initialize the tokenizer.
+    tokenizer_module = _import_tokenizer(module_name)
 
     tokenizer = tokenizer_module.create(config.get_libpq_dsn(), basedir)
-    tokenizer.init_new_db(config)
+    tokenizer.init_new_db(config, init_db=init_db)
 
     with connect(config.get_libpq_dsn()) as conn:
-        properties.set_property(conn, 'tokenizer', config.TOKENIZER)
+        properties.set_property(conn, 'tokenizer', module_name)
 
     return tokenizer
 
 
-def get_tokenizer_for_db(config):
+def get_tokenizer_for_db(config: Configuration) -> AbstractTokenizer:
     """ Instantiate a tokenizer for an existing database.
 
         The function looks up the appropriate tokenizer in the database
@@ -66,8 +82,8 @@ def get_tokenizer_for_db(config):
     """
     basedir = config.project_dir / 'tokenizer'
     if not basedir.is_dir():
-        LOG.fatal("Cannot find tokenizer data in '%s'.", basedir)
-        raise UsageError('Cannot initialize tokenizer.')
+        # Directory will be repopulated by tokenizer below.
+        basedir.mkdir()
 
     with connect(config.get_libpq_dsn()) as conn:
         name = properties.get_property(conn, 'tokenizer')
@@ -79,6 +95,6 @@ def get_tokenizer_for_db(config):
     tokenizer_module = _import_tokenizer(name)
 
     tokenizer = tokenizer_module.create(config.get_libpq_dsn(), basedir)
-    tokenizer.init_from_project()
+    tokenizer.init_from_project(config)
 
     return tokenizer