X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/d7f9d2bde90ccc12507598d87cc7fe57e225f5ea..18800a1a82f29d6d1dda177dcd60715b2c060fb2:/nominatim/tools/tiger_data.py?ds=sidebyside diff --git a/nominatim/tools/tiger_data.py b/nominatim/tools/tiger_data.py index fbcdb077..6e37df5e 100644 --- a/nominatim/tools/tiger_data.py +++ b/nominatim/tools/tiger_data.py @@ -1,6 +1,14 @@ +# 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 @@ -8,42 +16,82 @@ import tarfile from nominatim.db.connection import connect 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(pool, 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 # Using pool of database connections to execute sql statements - for sql_query in file: - pool.next_free_worker().perform(sql_query) + + 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: @@ -51,35 +99,31 @@ def handle_threaded_sql_statements(pool, file): 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 TigerInput(data_dir) as tar: + if not tar: + return - with connect(dsn) as conn: - sql = SQLPreprocessor(conn, config) - sql.run_sql_file(conn, 'tiger_import_start.sql') + 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 chunks. - place_threads = max(1, threads - 1) + # Reading files and then for each file line handling + # sql_query in chunks. + place_threads = max(1, threads - 1) - with WorkerPool(dsn, place_threads, ignore_sql_errors=True) as pool: - for sql_file in sql_files: - if not tar: - file = open(sql_file) - else: - file = tar.extractfile(sql_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) - handle_threaded_sql_statements(pool, file) + print('\n') - if tar: - tar.close() - print('\n') LOG.warning("Creating indexes on Tiger data") with connect(dsn) as conn: sql = SQLPreprocessor(conn, config)