mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-16 16:14:28 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: Overwrite(add command) was using snapshots without timestamps (and in memory), causing issues with some extractors

Browse source
This commit is contained in:
Cristian 2021-01-16 14:11:45 -05:00
parent a36f36866f
commit 30c886d4d4
3 changed files with 3 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

1
archivebox/core/models.py
View file

@ -106,6 +106,7 @@ class Snapshot(models.Model):
if "tags" in info: if "tags" in info:
# TODO: Handle tags # TODO: Handle tags
info.pop("tags") info.pop("tags")
info.pop("base_url", None)
return cls(**info) return cls(**info)
def get_history(self) -> dict: def get_history(self) -> dict:

2
archivebox/extractors/__init__.py
View file

@ -114,7 +114,7 @@ def archive_snapshot(snapshot: Model, overwrite: bool=False, methods: Optional[I
# print('{black} X {}{reset}'.format(method_name, **ANSI)) # print('{black} X {}{reset}'.format(method_name, **ANSI))
stats['skipped'] += 1 stats['skipped'] += 1
except Exception as e: except Exception as e:
raise Exception('Exception in archive_methods.save_{}(Link(url={}))'.format( raise Exception('Exception in archive_methods.save_{}(Snapshot(url={}))'.format(
method_name, method_name,
snapshot.url, snapshot.url,
)) from e )) from e

1
archivebox/main.py
View file

@ -597,6 +597,7 @@ def add(urls: Union[str, List[str]],
if update_all: if update_all:
archive_snapshots(all_snapshots, overwrite=overwrite, **archive_kwargs) archive_snapshots(all_snapshots, overwrite=overwrite, **archive_kwargs)
elif overwrite: elif overwrite:
imported_snapshots = Snapshot.objects.filter(url__in=[imported_snapshot.url for imported_snapshot in imported_snapshots])
archive_snapshots(imported_snapshots, overwrite=True, **archive_kwargs) archive_snapshots(imported_snapshots, overwrite=True, **archive_kwargs)
elif new_snapshots: elif new_snapshots:
archive_snapshots(new_snapshots, overwrite=False, **archive_kwargs) archive_snapshots(new_snapshots, overwrite=False, **archive_kwargs)