]> git.openstreetmap.org Git - nominatim.git/blobdiff - src/nominatim_db/tools/refresh.py
look up all places at once
[nominatim.git] / src / nominatim_db / tools / refresh.py
index 2e2ffabd07c60ea959bf1cb9aeaf9efa4cfe2b21..dc98fe4140b1b552abce67609dda467690a9772d 100644 (file)
@@ -11,22 +11,21 @@ from typing import MutableSequence, Tuple, Any, Type, Mapping, Sequence, List, c
 import csv
 import gzip
 import logging
-from textwrap import dedent
 from pathlib import Path
 
-from psycopg2 import sql as pysql
+from psycopg import sql as pysql
 
 from ..config import Configuration
-from ..db.connection import Connection, connect, postgis_version_tuple,\
-                            drop_tables, table_exists
-from ..db.utils import execute_file, CopyBuffer
+from ..db.connection import Connection, connect, postgis_version_tuple, \
+                            drop_tables
+from ..db.utils import execute_file
 from ..db.sql_preprocessor import SQLPreprocessor
-from ..version import NOMINATIM_VERSION
 
 LOG = logging.getLogger()
 
 OSM_TYPE = {'N': 'node', 'W': 'way', 'R': 'relation'}
 
+
 def _add_address_level_rows_from_entry(rows: MutableSequence[Tuple[Any, ...]],
                                        entry: Mapping[str, Any]) -> None:
     """ Converts a single entry from the JSON format for address rank
@@ -53,7 +52,7 @@ def load_address_levels(conn: Connection, table: str, levels: Sequence[Mapping[s
         The table has the following columns:
             country, class, type, rank_search, rank_address
     """
-    rows: List[Tuple[Any, ...]]  = []
+    rows: List[Tuple[Any, ...]] = []
     for entry in levels:
         _add_address_level_rows_from_entry(rows, entry)
 
@@ -68,8 +67,8 @@ def load_address_levels(conn: Connection, table: str, levels: Sequence[Mapping[s
                                         rank_address SMALLINT)
                               """).format(pysql.Identifier(table)))
 
-        cur.execute_values(pysql.SQL("INSERT INTO {} VALUES %s")
-                           .format(pysql.Identifier(table)), rows)
+        cur.executemany(pysql.SQL("INSERT INTO {} VALUES (%s, %s, %s, %s, %s)")
+                             .format(pysql.Identifier(table)), rows)
 
         cur.execute(pysql.SQL('CREATE UNIQUE INDEX ON {} (country_code, class, type)')
                     .format(pysql.Identifier(table)))
@@ -99,34 +98,6 @@ def create_functions(conn: Connection, config: Configuration,
                      debug=enable_debug)
 
 
-
-WEBSITE_SCRIPTS = (
-    'deletable.php',
-    'details.php',
-    'lookup.php',
-    'polygons.php',
-    'reverse.php',
-    'search.php',
-    'status.php'
-)
-
-# constants needed by PHP scripts: PHP name, config name, type
-PHP_CONST_DEFS = (
-    ('Database_DSN', 'DATABASE_DSN', str),
-    ('Default_Language', 'DEFAULT_LANGUAGE', str),
-    ('Log_DB', 'LOG_DB', bool),
-    ('Log_File', 'LOG_FILE', Path),
-    ('NoAccessControl', 'CORS_NOACCESSCONTROL', bool),
-    ('Places_Max_ID_count', 'LOOKUP_MAX_COUNT', int),
-    ('PolygonOutput_MaximumTypes', 'POLYGON_OUTPUT_MAX_TYPES', int),
-    ('Search_BatchMode', 'SEARCH_BATCH_MODE', bool),
-    ('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),
-)
-
-
 def import_wikipedia_articles(dsn: str, data_path: Path, ignore_errors: bool = False) -> int:
     """ Replaces the wikipedia importance tables with new data.
         The import is run in a single transaction so that the new data
@@ -155,7 +126,7 @@ def import_importance_csv(dsn: str, data_file: Path) -> int:
     if not data_file.exists():
         return 1
 
-    # Only import the first occurence of a wikidata ID.
+    # Only import the first occurrence of a wikidata ID.
     # This keeps indexes and table small.
     wd_done = set()
 
@@ -169,24 +140,17 @@ def import_importance_csv(dsn: str, data_file: Path) -> int:
                              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'])
+            copy_cmd = """COPY wikimedia_importance(language, title, importance, wikidata)
+                          FROM STDIN"""
+            with gzip.open(str(data_file), 'rt') as fd, cur.copy(copy_cmd) as copy:
+                for row in csv.DictReader(fd, delimiter='\t', quotechar='|'):
+                    wd_id = int(row['wikidata_id'][1:])
+                    copy.write_row((row['language'],
+                                    row['title'],
+                                    row['importance'],
+                                    None if wd_id in wd_done else row['wikidata_id']))
+                    wd_done.add(wd_id)
 
-        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
@@ -236,6 +200,7 @@ def import_secondary_importance(dsn: str, data_path: Path, ignore_errors: bool =
 
     return 0
 
+
 def recompute_importance(conn: Connection) -> None:
     """ Recompute wikipedia links and importance for all entries in placex.
         This is a long-running operations that must not be executed in
@@ -279,46 +244,6 @@ def _quote_php_variable(var_type: Type[Any], config: Configuration,
     return f"'{quoted}'"
 
 
-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()
-
-    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'}');
-                      @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)
-
-        basedata += f"@define('CONST_{php_name}', {varout});\n"
-
-    template = "\nrequire_once(CONST_LibDir.'/website/{}');\n"
-
-    search_name_table_exists = bool(conn and table_exists(conn, 'search_name'))
-
-    for script in WEBSITE_SCRIPTS:
-        if not search_name_table_exists and script == 'search.php':
-            out = template.format('reverse-only-search.php')
-        else:
-            out = template.format(script)
-
-        (basedir / script).write_text(basedata + out, 'utf-8')
-
-
 def invalidate_osm_object(osm_type: str, osm_id: int, conn: Connection,
                           recursive: bool = True) -> None:
     """ Mark the given OSM object for reindexing. When 'recursive' is set