Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-26 14:10:20 +00:00
Code Issues Projects Releases Packages Wiki Activity

add update flag support to archivebox schedule

Browse source
This commit is contained in:
Nick Sweeting 2022-05-09 20:18:43 -07:00
parent 8cfe6f4afb
commit 4b8b17e788
2 changed files with 8 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
archivebox/cli/archivebox_schedule.py
View file

@ -51,6 +51,11 @@ def main(args: Optional[List[str]]=None, stdin: Optional[IO]=None, pwd: Optional
action='store_true', action='store_true',
help='Re-archive any URLs that have been previously archived, overwriting existing Snapshots', help='Re-archive any URLs that have been previously archived, overwriting existing Snapshots',
) )
parser.add_argument(
'--update',
action='store_true',
help='Re-pull any URLs that have been previously added, as needed to fill missing ArchiveResults',
)
group.add_argument( group.add_argument(
'--clear', # '-c' '--clear', # '-c'
action='store_true', action='store_true',
@ -94,6 +99,7 @@ def main(args: Optional[List[str]]=None, stdin: Optional[IO]=None, pwd: Optional
every=command.every, every=command.every,
depth=command.depth, depth=command.depth,
overwrite=command.overwrite, overwrite=command.overwrite,
update=command.update,
import_path=command.import_path, import_path=command.import_path,
out_dir=pwd or OUTPUT_DIR, out_dir=pwd or OUTPUT_DIR,
) )

2
archivebox/main.py
View file

@ -1128,6 +1128,7 @@ def schedule(add: bool=False,
every: Optional[str]=None, every: Optional[str]=None,
depth: int=0, depth: int=0,
overwrite: bool=False, overwrite: bool=False,
update: bool=not ONLY_NEW,
import_path: Optional[str]=None, import_path: Optional[str]=None,
out_dir: Path=OUTPUT_DIR): out_dir: Path=OUTPUT_DIR):
"""Set ArchiveBox to regularly import URLs at specific times using cron""" """Set ArchiveBox to regularly import URLs at specific times using cron"""
@ -1157,6 +1158,7 @@ def schedule(add: bool=False,
*([ *([
'add', 'add',
*(['--overwrite'] if overwrite else []), *(['--overwrite'] if overwrite else []),
*(['--update'] if update else []),
f'--depth={depth}', f'--depth={depth}',
f'"{import_path}"', f'"{import_path}"',
] if import_path else ['update']), ] if import_path else ['update']),