mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-28 05:34:14 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix statemachine create_root_snapshot and retry timing

Browse source
This commit is contained in:
Nick Sweeting 2024-11-17 22:57:09 -08:00
parent 67c22b2df0
commit 1ec2753664
No known key found for this signature in database
8 changed files with 48 additions and 40 deletions
Download patch file Download diff file Expand all files Collapse all files

10
archivebox/crawls/models.py
View file

@ -190,22 +190,28 @@ class Crawl(ABIDModel, ModelWithHealthStats, ModelWithStateMachine):
from core.models import ArchiveResult
snapshot_ids = self.snapshot_set.values_list('id', flat=True)
pending_archiveresults = ArchiveResult.objects.filter(snapshot_id__in=snapshot_ids, retry_at__isnull=True)
pending_archiveresults = ArchiveResult.objects.filter(snapshot_id__in=snapshot_ids, retry_at__isnull=False)
return pending_archiveresults
def create_root_snapshot(self) -> 'Snapshot':
from core.models import Snapshot
try:
return Snapshot.objects.get(crawl=self, url=self.seed.uri)
except Snapshot.DoesNotExist:
pass
root_snapshot, _ = Snapshot.objects.update_or_create(
crawl=self,
url=self.seed.uri,
defaults={
'crawl': self,
'status': Snapshot.INITIAL_STATE,
'retry_at': timezone.now(),
'timestamp': str(timezone.now().timestamp()),
# 'config': self.seed.config,
},
)
root_snapshot.save()
return root_snapshot