diff --git a/trafilatura/cli.py b/trafilatura/cli.py index d69970b8..5c7a4337 100644 --- a/trafilatura/cli.py +++ b/trafilatura/cli.py @@ -16,7 +16,6 @@ url_processing_pipeline, write_result) from .settings import PARALLEL_CORES, SUPPORTED_FMT_CLI - # fix output encoding on some systems if sys.stdout.encoding != 'UTF-8': sys.stdout.reconfigure(encoding='utf-8') diff --git a/trafilatura/downloads.py b/trafilatura/downloads.py index 53333da8..2dc6e001 100644 --- a/trafilatura/downloads.py +++ b/trafilatura/downloads.py @@ -10,10 +10,11 @@ from concurrent.futures import ThreadPoolExecutor, as_completed from configparser import ConfigParser from functools import partial -from io import BytesIO from importlib.metadata import version +from io import BytesIO from time import sleep -from typing import Any, ByteString, Dict, Generator, List, Optional, Set, Tuple, Union +from typing import (Any, ByteString, Dict, Generator, List, Optional, Set, + Tuple, Union) import certifi import urllib3 @@ -22,8 +23,8 @@ from courlan.network import redirection_test from .settings import DEFAULT_CONFIG, Extractor -from .utils import URL_BLACKLIST_REGEX, decode_file, is_acceptable_length, make_chunks - +from .utils import (URL_BLACKLIST_REGEX, decode_file, is_acceptable_length, + make_chunks) try: from urllib3.contrib.socks import SOCKSProxyManager