mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-15 15:44:26 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

Refactor should_save_extractor methods to accept overwrite parameter

Browse source
This commit is contained in:
Dan Arnfield 2021-01-21 15:45:11 -06:00
parent 553c3ca219
commit 5420903102
14 changed files with 56 additions and 50 deletions
Download patch file Download diff file Expand all files Collapse all files

9
archivebox/extractors/headers.py
View file

@ -22,11 +22,12 @@ from ..config import (
from ..logging_util import TimedProgress
@enforce_types
def should_save_headers(link: Link, out_dir: Optional[str]=None) -> bool:
out_dir = out_dir or link.link_dir
def should_save_headers(link: Link, out_dir: Optional[str]=None, overwrite: Optional[bool]=False) -> bool:
out_dir = out_dir or Path(link.link_dir)
if not overwrite and (out_dir / 'headers.json').exists():
return False
output = Path(out_dir or link.link_dir) / 'headers.json'
return not output.exists() and SAVE_HEADERS
return SAVE_HEADERS
@enforce_types