R2R
R2R copied to clipboard
when ingest document with collection_ids error occurs
trafficstars
Describe the bug when ingest document with collection_ids error occurs.
To Reproduce
ingest request like this, with collection_ids.
POST /v3/documents HTTP/1.1
Content-Length: 4372
Content-Type: multipart/form-data
Host: localhost:7272
User-Agent: HTTPie
--PieBoundary123456789012345678901234567
Content-Disposition: form-data; name="metadata"
{"catalog":"help"}
--PieBoundary123456789012345678901234567
Content-Disposition: form-data; name="file"; filename="10-dev-env.md"
Content-Type:
--PieBoundary123456789012345678901234567
Content-Disposition: form-data; name="collection_ids"
["2cf8ceb6-1106-4c12-b233-7ef15cf74530","7a027464-a07c-4a04-9cbe-5976756791a8"]
--PieBoundary123456789012345678901234567
Content-Disposition: form-data; name="file-name"
10-dev-env.md
--PieBoundary123456789012345678901234567--
In the server side an error log like this:
ERROR - ingestion_workflow.py 178 : Error during assigning document to collection: invalid input for query argument $2: '2cf8ceb6-1106-4c12-b233-7ef15cf74530' (a sized iterable container expected (got type 'str'))
Additional context Add any other context about the problem here.
in the source file documents_router.py:
documents_router.py:
workflow_input = {
"file_data": file_data,
"document_id": str(document_id),
"collection_ids": (
[str(cid) for cid in collection_ids]
if collection_ids
else None
),
and in the file documents.py:
ids = [id] if isinstance(id, UUID) else id
maybe should be
ids = id if isinstance(id, list) else [id]