mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-12 22:25:44 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

dont dedupe snapshots in sqlite on every run

Browse source
This commit is contained in:
Nick Sweeting 2020-07-13 11:25:43 -04:00
parent dfb83b4f27
commit 354a63ccd4
1 changed files with 1 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

17
archivebox/index/sql.py
View file

@ -26,23 +26,8 @@ def write_sql_main_index(links: List[Link], out_dir: str=OUTPUT_DIR) -> None:
from core.models import Snapshot
from django.db import transaction
all_urls = {link.url: link for link in links}
all_ts = {link.timestamp: link for link in links}
with transaction.atomic():
for snapshot in Snapshot.objects.all():
if snapshot.timestamp in all_ts:
info = {k: v for k, v in all_urls.pop(snapshot.url)._asdict().items() if k in Snapshot.keys}
snapshot.delete()
Snapshot.objects.create(**info)
elif snapshot.url in all_urls:
info = {k: v for k, v in all_urls.pop(snapshot.url)._asdict().items() if k in Snapshot.keys}
snapshot.delete()
Snapshot.objects.create(**info)
else:
snapshot.delete()
for url, link in all_urls.items():
for link in links:
info = {k: v for k, v in link._asdict().items() if k in Snapshot.keys}
Snapshot.objects.update_or_create(url=url, defaults=info)