mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-14 07:04:27 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat: Modify migration reverse function to restore index (WIP)

Browse source
This commit is contained in:
Cristian 2020-11-11 15:26:54 -05:00
parent 508a0bb06e
commit f7f0bebdcc
1 changed files with 21 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

22
archivebox/core/migrations/0007_archiveresult.py
View file

@ -33,9 +33,29 @@ def forwards_func(apps, schema_editor):
start_ts=result["start_ts"], end_ts=result["end_ts"], status=result["status"], pwd=result["pwd"], output=result["output"])
def verify_json_index_integrity(results):
results = snapshot.archiveresult_set.all()
out_dir = Path(CONFIG['ARCHIVE_DIR']) / snapshot.timestamp
with open(out_dir / "index.json", "r") as f:
index = json.load(f)
history = index["history"]
extractors = [extractor for extractor in history]
index_results = [(result, extractor) for result in history[extractor]]
flattened_results = [(result["start_ts"], extractor) for result, extractor in index_results]
missing = [result for result in results if result.start_ts not in flattened_results]
#process missing elements here. Re-add to the index.json
def reverse_func(apps, schema_editor):
ArchiveResult = apps.get_model("core", "ArchiveResult")
Snapshot = apps.get_model("core", "Snapshot")
for snapshot in Snapshot.objects.all():
verify_json_index_integrity(snapshot)
ArchiveResult.objects.all().delete()