adai/MemoryBear
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(rag): fix pdfplumber concurrency issue and add debug logging

Browse Source 
The pdfplumber parser now uses a global lock to prevent concurrent access issues during PDF image rendering. Additionally, added a warning log to trace knowledge retrieval results for debugging purposes. The syntax fix in knowledge node's match case ensures correct pattern matching behavior.

BREAKING CHANGE: The pdfplumber parser now requires LOCK_KEY_pdfplumber to be defined in sys.modules for thread safety.

Closes #841
This commit is contained in:
Timebomb2018
2026-04-09 17:48:16 +08:00
parent 0516822d42
commit a7b8ba0c66
3 changed files with 6 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
api/app/core/workflow/nodes/knowledge/node.py
View File

@@ -233,7 +233,7 @@ class KnowledgeRetrievalNode(BaseNode):
}
)
)
case RetrieveType.HYBRID | RetrieveType.Graph:
case (RetrieveType.HYBRID, RetrieveType.Graph):
rs1_task = asyncio.to_thread(
vector_service.search_by_vector, **{
"query": query,