bellingcat/auto-archiver
1
0
Fork 0
mirror of https://github.com/bellingcat/auto-archiver.git synced 2026-06-11 12:48:28 +03:00
Code Issues Packages Projects Releases Wiki Activity

chooses most complete result from api (#116)

Browse Source
This commit is contained in:
Miguel Sozinho Ramalho
2023-12-13 11:26:46 +00:00
committed by GitHub
parent 128d4136e3
commit a786d4bb0e
3 changed files with 18 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/auto_archiver/databases/api_db.py
View File

@@ -35,15 +35,15 @@ class AAApiDb(Database):
""" query the database for the existence of this item"""
if not self.allow_rearchive: return
params = {"url": item.get_url(), "limit": 1}
params = {"url": item.get_url(), "limit": 15}
headers = {"Authorization": f"Bearer {self.api_token}", "accept": "application/json"}
response = requests.get(os.path.join(self.api_endpoint, "tasks/search-url"), params=params, headers=headers)
if response.status_code == 200:
if len(response.json()):
logger.success(f"API returned a previously archived instance: {response.json()}")
# TODO: can we do better than just returning the most recent result?
return Metadata.from_dict(response.json()[0]["result"])
logger.success(f"API returned {len(response.json())} previously archived instance(s)")
fetched_metadata = [Metadata.from_dict(r["result"]) for r in response.json()]
return Metadata.choose_most_complete(fetched_metadata)
else:
logger.error(f"AA API FAIL ({response.status_code}): {response.json()}")
return False