Merge pull request #2437 from lonvia/tweak-ranking-searches

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

diff --git a/nominatim/tokenizer/factory.py b/nominatim/tokenizer/factory.py
index 1079c935985fd4ed92a24e6f30db63461b832906..069672d4a1fd4d9b874943b5d44a367d4f2ef9e8 100644 (file)
--- a/nominatim/tokenizer/factory.py
+++ b/nominatim/tokenizer/factory.py
@@ -15,6 +15,7 @@ normalizer module is installed, when the tokenizer is created.
 """
 import logging
 import importlib
 """
 import logging
 import importlib

+from pathlib import Path

 
 from ..errors import UsageError
 from ..db import properties
 
 from ..errors import UsageError
 from ..db import properties


@@ -25,20 +26,24 @@ LOG = logging.getLogger()
 def _import_tokenizer(name):
     """ Load the tokenizer.py module from project directory.
     """
 def _import_tokenizer(name):
     """ 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)
         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, init_db=True, module_name=None):

     """ 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.
     """
     """ 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():
     # Create the directory for the tokenizer data
     basedir = config.project_dir / 'tokenizer'
     if not basedir.exists():


@@ -47,13 +52,14 @@ def create_tokenizer(config):
         LOG.fatal("Tokenizer directory '%s' cannot be created.", basedir)
         raise UsageError("Tokenizer setup failed.")
 
         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 = 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:
 
     with connect(config.get_libpq_dsn()) as conn:

-        properties.set_property(conn, 'tokenizer', config.TOKENIZER)
+        properties.set_property(conn, 'tokenizer', module_name)

 
     return tokenizer
 
 
     return tokenizer