mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-13 06:34:25 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

feat: Add extract flag to add command

Browse source
This commit is contained in:
Cristian 2020-11-13 09:24:34 -05:00
parent 4372cb6eec
commit 44eede96e5
3 changed files with 28 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

14
archivebox/main.py
View file

@ -525,7 +525,8 @@ def add(urls: Union[str, List[str]],
index_only: bool=False,
overwrite: bool=False,
init: bool=False,
out_dir: Path=OUTPUT_DIR) -> List[Link]:
out_dir: Path=OUTPUT_DIR,
extractors: list=[]) -> List[Link]:
"""Add a new URL or list of URLs to your archive"""
assert depth in (0, 1), 'Depth must be 0 or 1 (depth >1 is not supported yet)'
@ -567,12 +568,17 @@ def add(urls: Union[str, List[str]],
return all_links
# Run the archive methods for each link
archive_kwargs = {
"out_dir": out_dir,
}
if extractors:
archive_kwargs["methods"] = extractors
if update_all:
archive_links(all_links, overwrite=overwrite, out_dir=out_dir)
archive_links(all_links, overwrite=overwrite, **archive_kwargs)
elif overwrite:
archive_links(imported_links, overwrite=True, out_dir=out_dir)
archive_links(imported_links, overwrite=True, **archive_kwargs)
elif new_links:
archive_links(new_links, overwrite=False, out_dir=out_dir)
archive_links(new_links, overwrite=False, **archive_kwargs)
return all_links