mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-21 10:25:11 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

deduplicate method history when merging links

Browse source
This commit is contained in:
Nick Sweeting 2019-04-11 03:41:25 -04:00
parent 0d2f7eb58e
commit 0272c9b8c0
1 changed files with 11 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

11
archivebox/legacy/index.py
View file

@ -28,6 +28,7 @@ from .util import (
TimedProgress,
copy_and_overwrite,
atomic_write,
ExtendedEncoder,
)
from .parse import parse_links
from .logs import (
@ -93,6 +94,16 @@ def merge_links(a: Link, b: Link) -> Link:
method: (a.history.get(method) or []) + (b.history.get(method) or [])
for method in all_methods
}
for method in all_methods:
deduped_jsons = {
json.dumps(result, sort_keys=True, cls=ExtendedEncoder)
for result in history[method]
}
history[method] = list(reversed(sorted(
(ArchiveResult.from_json(json.loads(result)) for result in deduped_jsons),
key=lambda result: result.start_ts,
)))
return Link(
url=url,