]> git.openstreetmap.org Git - nominatim.git/blobdiff - nominatim/tools/tiger_data.py
remove postcodes from countries that don't have them
[nominatim.git] / nominatim / tools / tiger_data.py
index 07772c702a53f3a554cb8a42f54f9de388da56cd..6e37df5e9df7d2b6b3228c4ba79b8a5350865386 100644 (file)
+# 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 importing tiger data and handling tarbar and directory files
 """
+import csv
+import io
 import logging
 import os
 import tarfile
-import selectors
 
 from nominatim.db.connection import connect
-from nominatim.db.async_connection import DBConnection
+from nominatim.db.async_connection import WorkerPool
 from nominatim.db.sql_preprocessor import SQLPreprocessor
-
+from nominatim.errors import UsageError
+from nominatim.indexer.place_info import PlaceInfo
 
 LOG = logging.getLogger()
 
-
-def handle_tarfile_or_directory(data_dir):
-    """ Handles tarfile or directory for importing tiger data
+class TigerInput:
+    """ Context manager that goes through Tiger input files which may
+        either be in a directory or gzipped together in a tar file.
     """
 
-    tar = None
-    if data_dir.endswith('.tar.gz'):
-        tar = tarfile.open(data_dir)
-        sql_files = [i for i in tar.getmembers() if i.name.endswith('.sql')]
-        LOG.warning("Found %d SQL files in tarfile with path %s", len(sql_files), data_dir)
-        if not sql_files:
-            LOG.warning("Tiger data import selected but no files in tarfile's path %s", data_dir)
-            return None, None
-    else:
-        files = os.listdir(data_dir)
-        sql_files = [os.path.join(data_dir, i) for i in files if i.endswith('.sql')]
-        LOG.warning("Found %d SQL files in path %s", len(sql_files), data_dir)
-        if not sql_files:
-            LOG.warning("Tiger data import selected but no files found in path %s", data_dir)
-            return None, None
-
-    return sql_files, tar
-
-
-def handle_threaded_sql_statements(sel, file):
+    def __init__(self, data_dir):
+        self.tar_handle = None
+        self.files = []
+
+        if data_dir.endswith('.tar.gz'):
+            try:
+                self.tar_handle = tarfile.open(data_dir) # pylint: disable=consider-using-with
+            except tarfile.ReadError as err:
+                LOG.fatal("Cannot open '%s'. Is this a tar file?", data_dir)
+                raise UsageError("Cannot open Tiger data file.") from err
+
+            self.files = [i for i in self.tar_handle.getmembers() if i.name.endswith('.csv')]
+            LOG.warning("Found %d CSV files in tarfile with path %s", len(self.files), data_dir)
+        else:
+            files = os.listdir(data_dir)
+            self.files = [os.path.join(data_dir, i) for i in files if i.endswith('.csv')]
+            LOG.warning("Found %d CSV files in path %s", len(self.files), data_dir)
+
+        if not self.files:
+            LOG.warning("Tiger data import selected but no files found at %s", data_dir)
+
+
+    def __enter__(self):
+        return self
+
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        if self.tar_handle:
+            self.tar_handle.close()
+            self.tar_handle = None
+
+
+    def next_file(self):
+        """ Return a file handle to the next file to be processed.
+            Raises an IndexError if there is no file left.
+        """
+        fname = self.files.pop(0)
+
+        if self.tar_handle is not None:
+            return io.TextIOWrapper(self.tar_handle.extractfile(fname))
+
+        return open(fname, encoding='utf-8')
+
+
+    def __len__(self):
+        return len(self.files)
+
+
+def handle_threaded_sql_statements(pool, fd, analyzer):
     """ Handles sql statement with multiplexing
     """
-
     lines = 0
-    end_of_file = False
     # Using pool of database connections to execute sql statements
-    while not end_of_file:
-        for key, _ in sel.select(1):
-            conn = key.data
-            try:
-                if conn.is_done():
-                    sql_query = file.readline()
-                    lines += 1
-                    if not sql_query:
-                        end_of_file = True
-                        break
-                    conn.perform(sql_query)
-                    if lines == 1000:
-                        print('. ', end='', flush=True)
-                        lines = 0
-            except Exception as exc: # pylint: disable=broad-except
-                LOG.info('Wrong SQL statement: %s', exc)
-
-def handle_unregister_connection_pool(sel, place_threads):
-    """ Handles unregistering pool of connections
-    """
 
-    while place_threads > 0:
-        for key, _ in sel.select(1):
-            conn = key.data
-            sel.unregister(conn)
-            try:
-                conn.wait()
-            except Exception as exc: # pylint: disable=broad-except
-                LOG.info('Wrong SQL statement: %s', exc)
-            conn.close()
-            place_threads -= 1
+    sql = "SELECT tiger_line_import(%s, %s, %s, %s, %s, %s)"
+
+    for row in csv.DictReader(fd, delimiter=';'):
+        try:
+            address = dict(street=row['street'], postcode=row['postcode'])
+            args = ('SRID=4326;' + row['geometry'],
+                    int(row['from']), int(row['to']), row['interpolation'],
+                    PlaceInfo({'address': address}).analyze(analyzer),
+                    analyzer.normalize_postcode(row['postcode']))
+        except ValueError:
+            continue
+        pool.next_free_worker().perform(sql, args=args)
+
+        lines += 1
+        if lines == 1000:
+            print('.', end='', flush=True)
+            lines = 0
 
-def add_tiger_data(data_dir, config, threads):
+
+def add_tiger_data(data_dir, config, threads, tokenizer):
     """ Import tiger data from directory or tar file `data dir`.
     """
     dsn = config.get_libpq_dsn()
-    sql_files, tar = handle_tarfile_or_directory(data_dir)
-
-    if not sql_files:
-        return
-
-    with connect(dsn) as conn:
-        sql = SQLPreprocessor(conn, config)
-        sql.run_sql_file(conn, 'tiger_import_start.sql')
 
-    # Reading sql_files and then for each file line handling
-    # sql_query in <threads - 1> chunks.
-    sel = selectors.DefaultSelector()
-    place_threads = max(1, threads - 1)
+    with TigerInput(data_dir) as tar:
+        if not tar:
+            return
 
-    # Creates a pool of database connections
-    for _ in range(place_threads):
-        conn = DBConnection(dsn)
-        conn.connect()
-        sel.register(conn, selectors.EVENT_WRITE, conn)
+        with connect(dsn) as conn:
+            sql = SQLPreprocessor(conn, config)
+            sql.run_sql_file(conn, 'tiger_import_start.sql')
 
-    for sql_file in sql_files:
-        if not tar:
-            file = open(sql_file)
-        else:
-            file = tar.extractfile(sql_file)
+        # Reading files and then for each file line handling
+        # sql_query in <threads - 1> chunks.
+        place_threads = max(1, threads - 1)
 
-        handle_threaded_sql_statements(sel, file)
+        with WorkerPool(dsn, place_threads, ignore_sql_errors=True) as pool:
+            with tokenizer.name_analyzer() as analyzer:
+                while tar:
+                    with tar.next_file() as fd:
+                        handle_threaded_sql_statements(pool, fd, analyzer)
 
-    # Unregistering pool of database connections
-    handle_unregister_connection_pool(sel, place_threads)
+        print('\n')
 
-    if tar:
-        tar.close()
-    print('\n')
     LOG.warning("Creating indexes on Tiger data")
     with connect(dsn) as conn:
         sql = SQLPreprocessor(conn, config)