mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-27 21:24:16 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

switch .is_dir and .exists for os.access to avoid PermissionError on startup

Browse source
This commit is contained in:
Nick Sweeting 2024-10-08 03:02:34 -07:00
parent c3dd0f22e5
commit de2ab43f7f
No known key found for this signature in database
22 changed files with 119 additions and 97 deletions
Download patch file Download diff file Expand all files Collapse all files

28
archivebox/parsers/generic_txt.py
View file

@ -3,7 +3,6 @@ __description__ = 'Plain Text'
from typing import IO, Iterable
from datetime import datetime, timezone
from pathlib import Path
from ..index.schema import Link
from archivebox.misc.util import (
@ -22,19 +21,20 @@ def parse_generic_txt_export(text_file: IO[str], **_kwargs) -> Iterable[Link]:
if not line.strip():
continue
# if the line is a local file path that resolves, then we can archive it
try:
if Path(line).exists():
yield Link(
url=line,
timestamp=str(datetime.now(timezone.utc).timestamp()),
title=None,
tags=None,
sources=[text_file.name],
)
except (OSError, PermissionError):
# nvm, not a valid path...
pass
# # if the line is a local file path that resolves, then we can archive it
# if line.startswith('file://'):
# try:
# if Path(line).exists():
# yield Link(
# url=line,
# timestamp=str(datetime.now(timezone.utc).timestamp()),
# title=None,
# tags=None,
# sources=[text_file.name],
# )
# except (OSError, PermissionError):
# # nvm, not a valid path...
# pass
# otherwise look for anything that looks like a URL in the line
for url in find_all_urls(line):