Merge remote-tracking branch 'upstream/master'

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

diff --git a/nominatim/tools/refresh.py b/nominatim/tools/refresh.py
index c50493cc847f529377653e95d521f51bcdd55321..a200ee1348b9fdc717cd8db39c3e7bffd1438a64 100644 (file)
--- a/nominatim/tools/refresh.py
+++ b/nominatim/tools/refresh.py
@@ -8,6 +8,8 @@
 Functions for bringing auxiliary data in the database up-to-date.
 """
 from typing import MutableSequence, Tuple, Any, Type, Mapping, Sequence, List, cast
 Functions for bringing auxiliary data in the database up-to-date.
 """
 from typing import MutableSequence, Tuple, Any, Type, Mapping, Sequence, List, cast

+import csv
+import gzip

 import logging
 from textwrap import dedent
 from pathlib import Path
 import logging
 from textwrap import dedent
 from pathlib import Path


@@ -16,9 +18,9 @@ from psycopg2 import sql as pysql
 
 from nominatim.config import Configuration
 from nominatim.db.connection import Connection, connect
 
 from nominatim.config import Configuration
 from nominatim.db.connection import Connection, connect

-from nominatim.db.utils import execute_file
+from nominatim.db.utils import execute_file, CopyBuffer

 from nominatim.db.sql_preprocessor import SQLPreprocessor
 from nominatim.db.sql_preprocessor import SQLPreprocessor

-from nominatim.version import version_str
+from nominatim.version import NOMINATIM_VERSION

 
 LOG = logging.getLogger()
 
 
 LOG = logging.getLogger()
 


@@ -120,6 +122,7 @@ PHP_CONST_DEFS = (
     ('Search_NameOnlySearchFrequencyThreshold', 'SEARCH_NAME_ONLY_THRESHOLD', str),
     ('Use_US_Tiger_Data', 'USE_US_TIGER_DATA', bool),
     ('MapIcon_URL', 'MAPICON_URL', str),
     ('Search_NameOnlySearchFrequencyThreshold', 'SEARCH_NAME_ONLY_THRESHOLD', str),
     ('Use_US_Tiger_Data', 'USE_US_TIGER_DATA', bool),
     ('MapIcon_URL', 'MAPICON_URL', str),

+    ('Search_WithinCountries', 'SEARCH_WITHIN_COUNTRIES', bool),

 )
 
 
 )
 
 


@@ -131,21 +134,89 @@ def import_wikipedia_articles(dsn: str, data_path: Path, ignore_errors: bool = F
         Returns 0 if all was well and 1 if the importance file could not
         be found. Throws an exception if there was an error reading the file.
     """
         Returns 0 if all was well and 1 if the importance file could not
         be found. Throws an exception if there was an error reading the file.
     """

-    datafile = data_path / 'wikimedia-importance.sql.gz'
+    if import_importance_csv(dsn, data_path / 'wikimedia-importance.csv.gz') == 0 \
+       or import_importance_sql(dsn, data_path / 'wikimedia-importance.sql.gz',
+                                ignore_errors) == 0:
+        return 0

 
 

-    if not datafile.exists():
+    return 1
+
+
+def import_importance_csv(dsn: str, data_file: Path) -> int:
+    """ Replace wikipedia importance table with data from a
+        single CSV file.
+
+        The file must be a gzipped CSV and have the following columns:
+        language, title, importance, wikidata_id
+
+        Other columns may be present but will be ignored.
+    """
+    if not data_file.exists():
+        return 1
+
+    # Only import the first occurance of a wikidata ID.
+    # This keeps indexes and table small.
+    wd_done = set()
+
+    with connect(dsn) as conn:
+        with conn.cursor() as cur:
+            cur.drop_table('wikipedia_article')
+            cur.drop_table('wikipedia_redirect')
+            cur.drop_table('wikimedia_importance')
+            cur.execute("""CREATE TABLE wikimedia_importance (
+                             language TEXT NOT NULL,
+                             title TEXT NOT NULL,
+                             importance double precision NOT NULL,
+                             wikidata TEXT
+                           ) """)
+
+        with gzip.open(str(data_file), 'rt') as fd, CopyBuffer() as buf:
+            for row in csv.DictReader(fd, delimiter='\t', quotechar='|'):
+                wd_id = int(row['wikidata_id'][1:])
+                buf.add(row['language'], row['title'], row['importance'],
+                        None if wd_id in wd_done else row['wikidata_id'])
+                wd_done.add(wd_id)
+
+                if buf.size() > 10000000:
+                    with conn.cursor() as cur:
+                        buf.copy_out(cur, 'wikimedia_importance',
+                                     columns=['language', 'title', 'importance',
+                                              'wikidata'])
+
+            with conn.cursor() as cur:
+                buf.copy_out(cur, 'wikimedia_importance',
+                             columns=['language', 'title', 'importance', 'wikidata'])
+
+        with conn.cursor() as cur:
+            cur.execute("""CREATE INDEX IF NOT EXISTS idx_wikimedia_importance_title
+                           ON wikimedia_importance (title)""")
+            cur.execute("""CREATE INDEX IF NOT EXISTS idx_wikimedia_importance_wikidata
+                           ON wikimedia_importance (wikidata)
+                           WHERE wikidata is not null""")
+
+        conn.commit()
+
+    return 0
+
+
+def import_importance_sql(dsn: str, data_file: Path, ignore_errors: bool) -> int:
+    """ Replace wikipedia importance table with data from an SQL file.
+    """
+    if not data_file.exists():

         return 1
 
     pre_code = """BEGIN;
                   DROP TABLE IF EXISTS "wikipedia_article";
         return 1
 
     pre_code = """BEGIN;
                   DROP TABLE IF EXISTS "wikipedia_article";

-                  DROP TABLE IF EXISTS "wikipedia_redirect"
+                  DROP TABLE IF EXISTS "wikipedia_redirect";
+                  DROP TABLE IF EXISTS "wikipedia_importance";

                """
     post_code = "COMMIT"
                """
     post_code = "COMMIT"

-    execute_file(dsn, datafile, ignore_errors=ignore_errors,
+    execute_file(dsn, data_file, ignore_errors=ignore_errors,

                  pre_code=pre_code, post_code=post_code)
 
     return 0
 
                  pre_code=pre_code, post_code=post_code)
 
     return 0
 

+

 def import_secondary_importance(dsn: str, data_path: Path, ignore_errors: bool = False) -> int:
     """ Replaces the secondary importance raster data table with new data.
 
 def import_secondary_importance(dsn: str, data_path: Path, ignore_errors: bool = False) -> int:
     """ Replaces the secondary importance raster data table with new data.
 


@@ -176,7 +247,7 @@ def recompute_importance(conn: Connection) -> None:
         cur.execute("""
             UPDATE placex SET (wikipedia, importance) =
                (SELECT wikipedia, importance
         cur.execute("""
             UPDATE placex SET (wikipedia, importance) =
                (SELECT wikipedia, importance

-                FROM compute_importance(extratags, country_code, osm_type, osm_id, centroid))
+                FROM compute_importance(extratags, country_code, rank_search, centroid))

             """)
         cur.execute("""
             UPDATE placex s SET wikipedia = d.wikipedia, importance = d.importance
             """)
         cur.execute("""
             UPDATE placex s SET wikipedia = d.wikipedia, importance = d.importance


@@ -212,34 +283,41 @@ def _quote_php_variable(var_type: Type[Any], config: Configuration,
 def setup_website(basedir: Path, config: Configuration, conn: Connection) -> None:
     """ Create the website script stubs.
     """
 def setup_website(basedir: Path, config: Configuration, conn: Connection) -> None:
     """ Create the website script stubs.
     """

+    if config.lib_dir.php is None:
+        LOG.info("Python frontend does not require website setup. Skipping.")
+        return
+

     if not basedir.exists():
         LOG.info('Creating website directory.')
         basedir.mkdir()
 
     if not basedir.exists():
         LOG.info('Creating website directory.')
         basedir.mkdir()
 

-    template = dedent(f"""\
+    assert config.project_dir is not None
+    basedata = dedent(f"""\

                       <?php
 
                       @define('CONST_Debug', $_GET['debug'] ?? false);
                       @define('CONST_LibDir', '{config.lib_dir.php}');
                       @define('CONST_TokenizerDir', '{config.project_dir / 'tokenizer'}');
                       <?php
 
                       @define('CONST_Debug', $_GET['debug'] ?? false);
                       @define('CONST_LibDir', '{config.lib_dir.php}');
                       @define('CONST_TokenizerDir', '{config.project_dir / 'tokenizer'}');

-                      @define('CONST_NominatimVersion', '{version_str()}');
+                      @define('CONST_NominatimVersion', '{NOMINATIM_VERSION!s}');

 
                       """)
 
     for php_name, conf_name, var_type in PHP_CONST_DEFS:
         varout = _quote_php_variable(var_type, config, conf_name)
 
 
                       """)
 
     for php_name, conf_name, var_type in PHP_CONST_DEFS:
         varout = _quote_php_variable(var_type, config, conf_name)
 

-        template += f"@define('CONST_{php_name}', {varout});\n"
+        basedata += f"@define('CONST_{php_name}', {varout});\n"

 
 

-    template += f"\nrequire_once('{config.lib_dir.php}/website/{{}}');\n"
+    template = "\nrequire_once(CONST_LibDir.'/website/{}');\n"

 
     search_name_table_exists = bool(conn and conn.table_exists('search_name'))
 
     for script in WEBSITE_SCRIPTS:
         if not search_name_table_exists and script == 'search.php':
 
     search_name_table_exists = bool(conn and conn.table_exists('search_name'))
 
     for script in WEBSITE_SCRIPTS:
         if not search_name_table_exists and script == 'search.php':

-            (basedir / script).write_text(template.format('reverse-only-search.php'), 'utf-8')
+            out = template.format('reverse-only-search.php')

         else:
         else:

-            (basedir / script).write_text(template.format(script), 'utf-8')
+            out = template.format(script)
+
+        (basedir / script).write_text(basedata + out, 'utf-8')

 
 
 def invalidate_osm_object(osm_type: str, osm_id: int, conn: Connection,
 
 
 def invalidate_osm_object(osm_type: str, osm_id: int, conn: Connection,