zlei6/ragflow
1
0
Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2026-07-14 00:48:26 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refactor: Consolidation WEB API & HTTP API for document get_filter (#14248)

Browse Source 
### What problem does this PR solve?

Before consolidation
Web API: POST /v1/document/filter
Http API - GET /api/v1/datasets/<dataset_id>/documents

After consolidation, Restful API -- GET
/api/v1/datasets/<dataset_id>/documents?type=filter
### Type of change

- [x] Refactoring
This commit is contained in:
Jack
2026-04-21 18:55:30 +08:00
committed by GitHub
parent a33d0737cd
commit 009e538a4e
8 changed files with 101 additions and 149 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
test/testcases/test_web_api/test_common.py
View File

@@ -375,7 +375,7 @@ def create_document(auth, payload=None, *, headers=HEADERS, data=None):
def list_documents(auth, params=None, payload=None, *, headers=HEADERS, data=None):
kb_id = params.get("kb_id") if params else None
url = f"{HOST_ADDRESS}/api/{VERSION}/datasets/{kb_id}/documents"
url = f"{HOST_ADDRESS}{DATASETS_URL}/{kb_id}/documents"
if payload is None:
payload = {}
res = requests.get(url=url, headers=headers, auth=auth, params=params, json=payload, data=data)
@@ -392,8 +392,8 @@ def parse_documents(auth, payload=None, *, headers=HEADERS, data=None):
return res.json()
def document_filter(auth, payload=None, *, headers=HEADERS, data=None):
res = requests.post(url=f"{HOST_ADDRESS}{DOCUMENT_APP_URL}/filter", headers=headers, auth=auth, json=payload, data=data)
def document_filter(auth, dataset_id, payload=None, *, headers=HEADERS, data=None):
res = requests.get(url=f"{HOST_ADDRESS}{DATASETS_URL}/{dataset_id}/documents?type=filter", params=payload, headers=headers, auth=auth, data=data)
return res.json()

85
test/testcases/test_web_api/test_document_app/test_document_metadata.py
View File

@@ -37,7 +37,7 @@ class TestAuthorization:
@pytest.mark.p2
@pytest.mark.parametrize("invalid_auth, expected_code, expected_fragment", INVALID_AUTH_CASES)
def test_filter_auth_invalid(self, invalid_auth, expected_code, expected_fragment):
res = document_filter(invalid_auth, {"kb_id": "kb_id"})
res = document_filter(invalid_auth, "kb_id", {})
assert res["code"] == expected_code, res
assert expected_fragment in res["message"], res
@@ -84,7 +84,7 @@ class TestDocumentMetadata:
@pytest.mark.p2
def test_filter(self, WebApiAuth, add_dataset_func):
kb_id = add_dataset_func
res = document_filter(WebApiAuth, {"kb_id": kb_id})
res = document_filter(WebApiAuth, kb_id, {})
assert res["code"] == 0, res
assert "filter" in res["data"], res
assert "total" in res["data"], res
@@ -148,12 +148,12 @@ class TestDocumentMetadata:
class TestDocumentMetadataNegative:
@pytest.mark.p3
@pytest.mark.p2
def test_filter_missing_kb_id(self, WebApiAuth, add_document_func):
_, doc_id = add_document_func
res = document_filter(WebApiAuth, {"doc_ids": [doc_id]})
assert res["code"] == 101, res
assert "KB ID" in res["message"], res
kb_id, doc_id = add_document_func
res = document_filter(WebApiAuth, "", {"doc_ids": [doc_id]})
assert res["code"] == 100, res
assert "<MethodNotAllowed '405: Method Not Allowed'>" == res["message"], res
@pytest.mark.p3
def test_metadata_summary_missing_kb_id(self, WebApiAuth, add_document_func):
@@ -228,77 +228,6 @@ class TestDocumentMetadataUnit:
monkeypatch.setattr(module.UserTenantService, "query", lambda **_kwargs: [SimpleNamespace(tenant_id=tenant_id)])
monkeypatch.setattr(module.KnowledgebaseService, "query", lambda **_kwargs: True if _kwargs.get("id") == kb_id else False)
def test_filter_missing_kb_id(self, document_app_module, monkeypatch):
module = document_app_module
async def fake_request_json():
return {}
monkeypatch.setattr(module, "get_request_json", fake_request_json)
res = _run(module.get_filter())
assert res["code"] == 101
assert "KB ID" in res["message"]
def test_filter_unauthorized(self, document_app_module, monkeypatch):
module = document_app_module
monkeypatch.setattr(module.UserTenantService, "query", lambda **_kwargs: [SimpleNamespace(tenant_id="tenant1")])
monkeypatch.setattr(module.KnowledgebaseService, "query", lambda **_kwargs: False)
async def fake_request_json():
return {"kb_id": "kb1"}
monkeypatch.setattr(module, "get_request_json", fake_request_json)
res = _run(module.get_filter())
assert res["code"] == 103
def test_filter_invalid_filters(self, document_app_module, monkeypatch):
module = document_app_module
self._allow_kb(module, monkeypatch)
async def fake_request_json():
return {"kb_id": "kb1", "run_status": ["INVALID"]}
monkeypatch.setattr(module, "get_request_json", fake_request_json)
res = _run(module.get_filter())
assert res["code"] == 102
assert "Invalid filter run status" in res["message"]
async def fake_request_json_types():
return {"kb_id": "kb1", "types": ["INVALID"]}
monkeypatch.setattr(module, "get_request_json", fake_request_json_types)
res = _run(module.get_filter())
assert res["code"] == 102
assert "Invalid filter conditions" in res["message"]
def test_filter_keywords_suffix(self, document_app_module, monkeypatch):
module = document_app_module
self._allow_kb(module, monkeypatch)
monkeypatch.setattr(module.DocumentService, "get_filter_by_kb_id", lambda *_args, **_kwargs: ({"run": {}}, 1))
async def fake_request_json():
return {"kb_id": "kb1", "keywords": "ragflow", "suffix": ["txt"]}
monkeypatch.setattr(module, "get_request_json", fake_request_json)
res = _run(module.get_filter())
assert res["code"] == 0
assert "filter" in res["data"]
def test_filter_exception(self, document_app_module, monkeypatch):
module = document_app_module
self._allow_kb(module, monkeypatch)
def raise_error(*_args, **_kwargs):
raise RuntimeError("boom")
monkeypatch.setattr(module.DocumentService, "get_filter_by_kb_id", raise_error)
async def fake_request_json():
return {"kb_id": "kb1"}
monkeypatch.setattr(module, "get_request_json", fake_request_json)
res = _run(module.get_filter())
assert res["code"] == 100
def test_infos_meta_fields(self, document_app_module, monkeypatch):
module = document_app_module