adding prefix to keys is now more configurable

[nominatim.git] / nominatim / tools / country_info.py

diff --git a/nominatim/tools/country_info.py b/nominatim/tools/country_info.py
index 5b37c0b645dfffe301bfc649b38c5a77b758367b..2457e01b6b05178c7b7c41e55574953e8a4d9351 100644 (file)
--- a/nominatim/tools/country_info.py
+++ b/nominatim/tools/country_info.py
@@ -7,6 +7,8 @@
 """
 Functions for importing and managing static country information.
 """
 """
 Functions for importing and managing static country information.
 """

+import json
+from io import StringIO

 import psycopg2.extras
 
 from nominatim.db import utils as db_utils
 import psycopg2.extras
 
 from nominatim.db import utils as db_utils


@@ -33,7 +35,8 @@ class _CountryInfo:
                 elif not isinstance(prop['languages'], list):
                     prop['languages'] = [x.strip()
                                          for x in prop['languages'].split(',')]
                 elif not isinstance(prop['languages'], list):
                     prop['languages'] = [x.strip()
                                          for x in prop['languages'].split(',')]

-
+                if 'names' not in prop:
+                    prop['names']['name'] = {}

 
     def items(self):
         """ Return tuples of (country_code, property dict) as iterable.
 
     def items(self):
         """ Return tuples of (country_code, property dict) as iterable.


@@ -61,10 +64,12 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False):
     """ Create and populate the tables with basic static data that provides
         the background for geocoding. Data is assumed to not yet exist.
     """
     """ Create and populate the tables with basic static data that provides
         the background for geocoding. Data is assumed to not yet exist.
     """

-    db_utils.execute_file(dsn, sql_dir / 'country_name.sql')

     db_utils.execute_file(dsn, sql_dir / 'country_osm_grid.sql.gz')
 
     db_utils.execute_file(dsn, sql_dir / 'country_osm_grid.sql.gz')
 

-    params = []
+    def add_prefix_to_keys(name, prefix):
+        return {prefix+k: v for k, v in name.items()}
+
+    params, country_names_data = [], ''

     for ccode, props in _COUNTRY_INFO.items():
         if ccode is not None and props is not None:
             if ignore_partitions:
     for ccode, props in _COUNTRY_INFO.items():
         if ccode is not None and props is not None:
             if ignore_partitions:


@@ -74,8 +79,21 @@ def setup_country_tables(dsn, sql_dir, ignore_partitions=False):
             lang = props['languages'][0] if len(props['languages']) == 1 else None
             params.append((ccode, partition, lang))
 
             lang = props['languages'][0] if len(props['languages']) == 1 else None
             params.append((ccode, partition, lang))
 

+            name = add_prefix_to_keys(props.get('names').get('name'), 'name:')
+            name = json.dumps(name , ensure_ascii=False, separators=(', ', '=>'))
+            country_names_data += ccode + '\t' + name[1:-1] + '\n'

     with connect(dsn) as conn:
         with conn.cursor() as cur:
     with connect(dsn) as conn:
         with conn.cursor() as cur:

+            cur.execute(
+                """ CREATE TABLE public.country_name (
+                        country_code character varying(2),
+                        name public.hstore,
+                        derived_name public.hstore,
+                        country_default_language_code text,
+                        partition integer
+                    ); """)
+            data = StringIO(country_names_data)
+            cur.copy_from(data, 'country_name', columns=('country_code', 'name'))

             cur.execute_values(
                 """ UPDATE country_name
                     SET partition = part, country_default_language_code = lang
             cur.execute_values(
                 """ UPDATE country_name
                     SET partition = part, country_default_language_code = lang


@@ -94,8 +112,8 @@ def create_country_names(conn, tokenizer, languages=None):
         languages = languages.split(',')
 
     def _include_key(key):
         languages = languages.split(',')
 
     def _include_key(key):

-        return key == 'name' or \
-               (key.startswith('name:') and (not languages or key[5:] in languages))
+        return key.startswith('name:') and \
+            key[5:] in languages or key[5:] == 'default'

 
     with conn.cursor() as cur:
         psycopg2.extras.register_hstore(cur)
 
     with conn.cursor() as cur:
         psycopg2.extras.register_hstore(cur)