zlei6/ragflow
1
0
Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2026-06-29 23:41:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(api): use dataset-owner tenant for legacy /chunks docstore cleanup (#15961)

Browse Source
This commit is contained in:
kpdev
2026-06-23 23:24:40 -07:00
committed by GitHub
parent ede46e0bb8
commit 68d2ca0ff1
2 changed files with 83 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

69
test/testcases/test_http_api/test_file_management_within_dataset/test_doc_sdk_routes_unit.py
View File

@@ -731,6 +731,75 @@ class TestDocRoutesUnit:
res = _run(module.stop_parsing.__wrapped__("tenant-1", "ds-1"))
assert res["code"] == 0
def test_legacy_chunks_parse_uses_dataset_owner_tenant_for_delete(self, monkeypatch):
module = _load_doc_module(monkeypatch)
deleted = []
requester_tenant = "team-member"
owner_tenant = "dataset-owner"
monkeypatch.setattr(module.KnowledgebaseService, "accessible", lambda **_kwargs: True)
monkeypatch.setattr(
module.KnowledgebaseService,
"get_by_id",
lambda _id: (True, SimpleNamespace(tenant_id=owner_tenant)),
)
monkeypatch.setattr(module, "get_request_json", lambda: _AwaitableValue({"document_ids": ["doc-1"]}))
monkeypatch.setattr(module, "check_duplicate_ids", lambda ids, _kind: (ids, []))
monkeypatch.setattr(
module.DocumentService,
"query",
lambda **_kwargs: [_DummyDoc(doc_id="doc-1", run=module.TaskStatus.UNSTART.value)],
)
monkeypatch.setattr(module.DocumentService, "filter_update", lambda *_args, **_kwargs: 1)
monkeypatch.setattr(module.DocumentService, "get_by_id", lambda _id: (True, _DummyDoc(doc_id="doc-1")))
monkeypatch.setattr(module.File2DocumentService, "get_storage_address", lambda **_kwargs: ("b", "n"))
monkeypatch.setattr(module.TaskService, "filter_delete", lambda *_args, **_kwargs: None)
monkeypatch.setattr(module, "queue_tasks", lambda *_args, **_kwargs: None)
_patch_docstore(
monkeypatch,
module,
index_exist=lambda *_args, **_kwargs: True,
delete=lambda condition, index, kb_id: deleted.append((condition, index, kb_id)),
)
res = _run(module.parse.__wrapped__(requester_tenant, "ds-1"))
assert res["code"] == 0
assert deleted == [({"doc_id": "doc-1"}, module.search.index_name(owner_tenant), "kb-1")]
def test_legacy_chunks_stop_uses_dataset_owner_tenant_for_delete(self, monkeypatch):
module = _load_doc_module(monkeypatch)
deleted = []
requester_tenant = "team-member"
owner_tenant = "dataset-owner"
monkeypatch.setattr(module.KnowledgebaseService, "accessible", lambda **_kwargs: True)
monkeypatch.setattr(
module.KnowledgebaseService,
"get_by_id",
lambda _id: (True, SimpleNamespace(tenant_id=owner_tenant)),
)
monkeypatch.setattr(module, "get_request_json", lambda: _AwaitableValue({"document_ids": ["doc-1"]}))
monkeypatch.setattr(module, "check_duplicate_ids", lambda ids, _kind: (ids, []))
monkeypatch.setattr(
module.DocumentService,
"query",
lambda **_kwargs: [_DummyDoc(doc_id="doc-1", run=module.TaskStatus.RUNNING.value)],
)
monkeypatch.setattr(module, "cancel_all_task_of", lambda *_args, **_kwargs: None)
monkeypatch.setattr(module.DocumentService, "update_by_id", lambda *_args, **_kwargs: True)
_patch_docstore(
monkeypatch,
module,
index_exist=lambda *_args, **_kwargs: True,
delete=lambda condition, index, kb_id: deleted.append((condition, index, kb_id)),
)
res = _run(module.stop_parsing.__wrapped__(requester_tenant, "ds-1"))
assert res["code"] == 0
assert deleted == [({"doc_id": "doc-1"}, module.search.index_name(owner_tenant), "kb-1")]
def test_stop_parse_documents_cleans_partial_chunks(self, monkeypatch):
module = _load_doc_module(monkeypatch, module_basename="document_api")
updated = []