mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-31 14:58:25 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

enforce utf8 on literally all file operations because windows sucks

Browse source
This commit is contained in:
Nick Sweeting 2021-03-27 01:01:29 -04:00
parent 185d2f9f9b
commit bd6d9c165b
9 changed files with 29 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

2
archivebox/extractors/archive_org.py
View file

@ -31,7 +31,7 @@ def should_save_archive_dot_org(link: Link, out_dir: Optional[Path]=None, overwr
out_dir = out_dir or Path(link.link_dir)
if not overwrite and (out_dir / 'archive.org.txt').exists():
# if open(path, 'r').read().strip() != 'None':
# if open(path, 'r', encoding='utf-8').read().strip() != 'None':
return False
return SAVE_ARCHIVE_DOT_ORG

2
archivebox/extractors/readability.py
View file

@ -35,7 +35,7 @@ def get_html(link: Link, path: Path) -> str:
document = None
for source in sources:
try:
with open(abs_path / source, "r") as f:
with open(abs_path / source, "r", encoding="utf-8") as f:
document = f.read()
break
except (FileNotFoundError, TypeError):