Merge pull request '后端：稳定租户链路与 VLM 图片质量检测' (#9) from wren-dev into main

Reviewed-on: #9

fastapi_modules/fastapi_leaudit/leaudit_bridge/resilient_clients.py

@@ -239,7 +239,7 @@ class ResilientQwenVLMClient(QwenVLMClient):
         body = response.json()
         text = (body.get("choices") or [{}])[0].get("message", {}).get("content", "")
         parsed = _parse_json_loose(text)
-        return parsed if isinstance(parsed, dict) else {}
+        return parsed if isinstance(parsed, dict) else {"result": text, "reason": text}
 
 
 class ResilientChandraOCRClient(ChandraOCRClient):

fastapi_modules/fastapi_leaudit/services/impl/pageQualityServiceImpl.py

@@ -6,6 +6,7 @@ from typing import Any
 from pathlib import Path
 import tempfile
 import logging
+import json
 
 import fitz
 from leaudit.converters import doc2pdf
@@ -30,9 +31,11 @@ _PAGE_QUALITY_VLM_PROMPT = """
 你是文档扫描图片质量检测员。请判断这 1 页文档图片是否适合继续做 OCR 与合同/公文评查。
 
 判定标准：
-1. pass：文字主体清晰、方向正常、没有明显截断，能稳定阅读。
+1. 必须同时检查整页扫描质量，以及页面内所有内嵌照片、证据照片、现场照片、截图、印章和签名图片的清晰度。
-2. review：存在轻微模糊、倾斜、阴影、低对比度、局部遮挡、轻微截断，建议人工确认但仍可能可读。
+2. pass：文字主体清晰、方向正常、没有明显截断；页面内嵌照片/证据照片也能辨认关键视觉信息。
-3. reject：严重模糊、重影、过曝/过暗、页面大面积缺失、关键文字不可辨认、方向严重错误、空白页或非文档页，建议重拍。
+3. review：存在轻微模糊、倾斜、阴影、低对比度、局部遮挡、轻微截断；或内嵌照片/证据照片主体明显发虚、牌匾/场所/人物/关键物证不易辨认，建议人工确认但仍可能可用。
+4. reject：严重模糊、重影、过曝/过暗、页面大面积缺失、关键文字不可辨认、方向严重错误、空白页或非文档页；或内嵌证据照片主体无法辨认、关键证据信息不可用，建议重拍。
+5. 即使页面周边文字清楚，只要内嵌证据照片明显模糊，也不能判 pass，至少判 review，严重时判 reject。
 
 只输出 JSON，不要输出 Markdown，不要解释额外文本：
 {"status":"pass|review|reject","score":0.0到1.0,"reason":"20字以内中文原因"}
@@ -495,12 +498,28 @@ class PageQualityServiceImpl(IPageQualityService):
             logger.warning("VLM page quality detection failed: %s", exc)
             return "review", 0.5, "VLM图片质量检测失败，需人工确认"
 
-        status = str((result or {}).get("status") or "").strip().lower()
+        result_dict = self._coerce_vlm_result(result)
+        status = self._normalize_quality_status(
+            self._first_non_empty(
+                result_dict,
+                ("status", "quality_status", "qualityStatus", "result", "label", "decision", "conclusion"),
+            )
+        )
+        reason = self._normalize_quality_reason(
+            self._first_non_empty(
+                result_dict,
+                ("reason", "quality_reason", "qualityReason", "message", "msg", "detail", "explanation", "description"),
+            )
+        )
+        if status is None and reason:
+            status = self._normalize_quality_status(reason)
         if status not in {"pass", "review", "reject"}:
             return "review", 0.5, "VLM返回结果不可用，需人工确认"
 
-        score = self._normalize_quality_score((result or {}).get("score"), status)
+        score = self._normalize_quality_score(
-        reason = str((result or {}).get("reason") or "").strip() or None
+            self._first_non_empty(result_dict, ("score", "quality_score", "qualityScore", "confidence")),
+            status,
+        )
         if status != "pass" and not reason:
             reason = "页面图片质量需人工确认"
         return status, score, reason
@@ -526,6 +545,56 @@ class PageQualityServiceImpl(IPageQualityService):
             return defaults[status]
         return max(0.0, min(1.0, score))
 
+    def _coerce_vlm_result(self, result: Any) -> dict[str, Any]:
+        if isinstance(result, dict):
+            return result
+        if isinstance(result, str):
+            text_result = result.strip()
+            if not text_result:
+                return {}
+            try:
+                parsed = json.loads(text_result)
+            except json.JSONDecodeError:
+                return {"result": text_result, "reason": text_result}
+            return parsed if isinstance(parsed, dict) else {"result": text_result}
+        return {}
+
+    def _first_non_empty(self, payload: dict[str, Any], keys: tuple[str, ...]) -> Any:
+        for key in keys:
+            value = payload.get(key)
+            if value is not None and str(value).strip():
+                return value
+        return None
+
+    def _normalize_quality_status(self, raw_status: Any) -> str | None:
+        text_status = str(raw_status or "").strip().lower()
+        if not text_status:
+            return None
+        compact_status = text_status.replace(" ", "").replace("_", "").replace("-", "")
+        if compact_status in {"pass", "passed", "ok", "good", "clear", "readable"}:
+            return "pass"
+        if compact_status in {"review", "warn", "warning", "manual", "uncertain", "suspect", "suspicious"}:
+            return "review"
+        if compact_status in {"reject", "rejected", "fail", "failed", "bad", "unreadable", "retake"}:
+            return "reject"
+
+        reject_keywords = ("不通过", "拒绝", "重拍", "不可读", "无法辨认", "无法识别", "严重", "大面积缺失", "空白页")
+        review_keywords = ("复核", "人工", "疑似", "轻微", "建议确认", "建议人工", "模糊", "不清晰", "低对比", "发虚")
+        pass_keywords = ("通过", "合格", "清晰", "可读")
+        if any(keyword in text_status for keyword in reject_keywords):
+            return "reject"
+        if any(keyword in text_status for keyword in review_keywords):
+            return "review"
+        if any(keyword in text_status for keyword in pass_keywords):
+            return "pass"
+        return None
+
+    def _normalize_quality_reason(self, raw_reason: Any) -> str | None:
+        reason = str(raw_reason or "").strip()
+        if not reason:
+            return None
+        return reason[:80]
+
     def _document_service(self):
         if self.DocumentService is None:
             from fastapi_modules.fastapi_leaudit.services.impl.documentServiceImpl import DocumentServiceImpl

fastapi_modules/fastapi_leaudit/services/impl/ragChatServiceImpl.py

@@ -1,3 +1,5 @@
+"""RAG 聊天服务实现。"""
+
 from __future__ import annotations
 
 import asyncio
@@ -49,6 +51,8 @@ MANUAL_TITLE_SOURCE = "manual"
 
 
 class RagChatServiceImpl(IRagChatService):
+    """RAG 聊天服务实现。"""
+
     _message_tasks: dict[str, asyncio.Task] = {}
     _task_events: dict[str, list[dict]] = {}
     _task_done: dict[str, bool] = {}
@@ -1143,14 +1147,17 @@ class RagChatServiceImpl(IRagChatService):
                 except Exception:
                     followups = []
 
+                sources = self._build_sources(context_chunks, dataset_name)
                 if message_end_payload:
-                    message_end_payload.setdefault("metadata", {})["suggested_questions"] = followups
+                    message_end_metadata = message_end_payload.setdefault("metadata", {})
+                    message_end_metadata["suggested_questions"] = followups
+                    message_end_metadata["retriever_resources"] = sources
                     await self._append_task_event(task_id, message_end_payload)
                 await self._finalize_message_record(
                     conversation_id=conversation_id,
                     message_id=message_id,
                     content=collected_answer,
-                    sources=self._build_sources(context_chunks, dataset_name),
+                    sources=sources,
                     metadata={"suggested_questions": followups, "status": "completed", "task_id": task_id},
                 )
                 await self._maybe_schedule_auto_title(

fastapi_modules/fastapi_leaudit/services/impl/ragDatasetServiceImpl.py

@@ -1,3 +1,5 @@
+"""RAG 知识库服务实现。"""
+
 from __future__ import annotations
 
 import asyncio
@@ -42,6 +44,8 @@ from fastapi_modules.fastapi_leaudit.services.impl.tenantResolver import TenantR
 
 
 class RagDatasetServiceImpl(IRagDatasetService):
+    """RAG 知识库服务实现。"""
+
     _ACTIVE_INDEXING_STATUSES = {"waiting", "parsing", "cleaning", "splitting", "indexing"}
     _DELETABLE_DOCUMENT_STATUSES = {"completed", "error", "paused"}
     _APP_LINK_SQL = """
@@ -295,7 +299,32 @@ class RagDatasetServiceImpl(IRagDatasetService):
             if target_is_default:
                 await self._clear_default_flags(session, tenant_code=resolved_tenant_code)
             elif existing.get("is_default") and Body.get("is_default") is False:
-                raise LeauditException(StatusCodeEnum.HTTP_400_BAD_REQUEST, "默认知识库不能直接取消，请先将其他知识库设为默认")
+                normalized_tenant_code = str(resolved_tenant_code or "").strip()
+                default_filters = [
+                    "deleted_at IS NULL",
+                    "is_default = TRUE",
+                    "id <> :dataset_id",
+                ]
+                default_params = {"dataset_id": DatasetId}
+                if normalized_tenant_code:
+                    default_filters.append("tenant_code = :tenant_code")
+                    default_params["tenant_code"] = normalized_tenant_code
+                else:
+                    default_filters.append("(tenant_code IS NULL OR BTRIM(tenant_code) = '')")
+                other_default_count = (
+                    await session.execute(
+                        text(
+                            f"""
+                            SELECT COUNT(1)
+                            FROM rag_dataset
+                            WHERE {" AND ".join(default_filters)}
+                            """
+                        ),
+                        default_params,
+                    )
+                ).scalar_one()
+                if int(other_default_count or 0) <= 0:
+                    raise LeauditException(StatusCodeEnum.HTTP_400_BAD_REQUEST, "默认知识库不能直接取消，请先将其他知识库设为默认")
             await session.execute(
                 text(
                     """

tests/test_page_quality_vlm.py

@@ -1,5 +1,7 @@
 import pytest
+import httpx
 
+from fastapi_modules.fastapi_leaudit.leaudit_bridge.resilient_clients import ResilientQwenVLMClient
 from fastapi_modules.fastapi_leaudit.services.impl.pageQualityServiceImpl import PageQualityServiceImpl
 
 
@@ -32,6 +34,58 @@ async def test_vlm_page_quality_reject_result_is_used():
     assert score == 0.18
     assert "严重模糊" in reason
     assert "只输出 JSON" in service.VlmClient.prompts[0][0]
+    assert "内嵌照片" in service.VlmClient.prompts[0][0]
+    assert "即使页面周边文字清楚" in service.VlmClient.prompts[0][0]
+
+
+@pytest.mark.asyncio
+async def test_vlm_page_quality_embedded_evidence_blur_cannot_pass():
+    service = PageQualityServiceImpl()
+    service.VlmClient = _FakeVlmClient(
+        {
+            "quality_status": "疑似模糊",
+            "quality_score": "0.42",
+            "message": "内嵌证据照片主体发虚，门头文字不易辨认",
+        }
+    )
+
+    status, score, reason = await service._classify_page_image_by_vlm(b"image-bytes")
+
+    assert status == "review"
+    assert score == 0.42
+    assert "内嵌证据照片" in reason
+
+
+@pytest.mark.asyncio
+async def test_vlm_page_quality_chinese_reject_status_is_supported():
+    service = PageQualityServiceImpl()
+    service.VlmClient = _FakeVlmClient(
+        {
+            "result": "不通过",
+            "confidence": 0.1,
+            "detail": "证据照片严重模糊，关键场所无法辨认",
+        }
+    )
+
+    status, score, reason = await service._classify_page_image_by_vlm(b"image-bytes")
+
+    assert status == "reject"
+    assert score == 0.1
+    assert "严重模糊" in reason
+
+
+@pytest.mark.asyncio
+async def test_vlm_page_quality_json_string_result_is_supported():
+    service = PageQualityServiceImpl()
+    service.VlmClient = _FakeVlmClient(
+        '{"status":"review","score":0.33,"reason":"页面内照片模糊"}'
+    )
+
+    status, score, reason = await service._classify_page_image_by_vlm(b"image-bytes")
+
+    assert status == "review"
+    assert score == 0.33
+    assert reason == "页面内照片模糊"
 
 
 @pytest.mark.asyncio
@@ -56,3 +110,32 @@ async def test_vlm_page_quality_error_falls_back_to_review_not_pass():
     assert status == "review"
     assert score == 0.5
     assert "VLM图片质量检测失败" in reason
+
+
+@pytest.mark.asyncio
+async def test_resilient_vlm_extract_multifield_keeps_raw_text_when_json_parse_fails(monkeypatch):
+    client = ResilientQwenVLMClient(base_url="http://example.test", api_key="x", model="vlm-test")
+
+    async def fake_post_with_retry(payload):
+        return httpx.Response(
+            200,
+            json={
+                "choices": [
+                    {
+                        "message": {
+                            "content": "疑似模糊：内嵌证据照片主体发虚，建议人工复核",
+                        }
+                    }
+                ]
+            },
+        )
+
+    monkeypatch.setattr(client, "_post_with_retry", fake_post_with_retry)
+
+    result = await client.extract_multifield(
+        prompt="图片质量检测",
+        images_data_urls=["data:image/png;base64,xxx"],
+    )
+
+    assert result["result"].startswith("疑似模糊")
+    assert "内嵌证据照片" in result["reason"]

tests/test_rag_chat_streaming_sources.py

@@ -0,0 +1,91 @@
+"""RAG 聊天流式引用来源测试。"""
+
+import asyncio
+import json
+from unittest.mock import patch
+
+from fastapi_modules.fastapi_leaudit.services.impl.ragChatServiceImpl import RagChatServiceImpl
+
+
+async def _fake_generate_stream(**kwargs):
+    yield "data: " + json.dumps(
+        {
+            "event": "message",
+            "task_id": kwargs["task_id"],
+            "message_id": kwargs["message_id"],
+            "conversation_id": kwargs["conversation_id"],
+            "answer": "测试回答",
+        },
+        ensure_ascii=False,
+    ) + "\n\n"
+    yield "data: " + json.dumps(
+        {
+            "event": "message_end",
+            "task_id": kwargs["task_id"],
+            "message_id": kwargs["message_id"],
+            "conversation_id": kwargs["conversation_id"],
+            "metadata": {},
+        },
+        ensure_ascii=False,
+    ) + "\n\n"
+
+
+async def _run_streaming_task() -> list[dict]:
+    service = RagChatServiceImpl()
+    task_id = "task-test"
+    service._task_events[task_id] = []
+    service._task_done[task_id] = False
+    service._task_locks[task_id] = asyncio.Lock()
+
+    async def fake_retrieve_context(dataset_id, query):
+        return [
+            {
+                "dataset_id": dataset_id,
+                "document_id": "doc-1",
+                "document_name": "引用文档.pdf",
+                "source": "引用文档.pdf",
+                "id": "segment-1",
+                "score": 0.91,
+                "hit_count": 3,
+                "text": "引用片段内容",
+            }
+        ], "测试知识库"
+
+    async def noop_async(*args, **kwargs):
+        return None
+
+    service._retrieve_context = fake_retrieve_context
+    service._finalize_message_record = noop_async
+    service._maybe_schedule_auto_title = noop_async
+
+    with (
+        patch(
+            "fastapi_modules.fastapi_leaudit.services.impl.ragChatServiceImpl.generate_stream",
+            _fake_generate_stream,
+        ),
+        patch(
+            "fastapi_modules.fastapi_leaudit.services.impl.ragChatServiceImpl.generate_followups",
+            return_value=[],
+        ),
+    ):
+        await service._run_message_task(
+            task_id=task_id,
+            conversation_id="conversation-test",
+            message_id="message-test",
+            query="测试问题",
+            app={"dataset_id": 7},
+        )
+
+    return service._task_events[task_id]
+
+
+def test_streaming_message_end_includes_retriever_resources():
+    events = asyncio.run(_run_streaming_task())
+
+    message_end = next(event for event in events if event.get("event") == "message_end")
+
+    resources = message_end["metadata"].get("retriever_resources")
+    assert resources
+    assert resources[0]["dataset_id"] == "7"
+    assert resources[0]["dataset_name"] == "测试知识库"
+    assert resources[0]["document_name"] == "引用文档.pdf"

tests/test_rag_dataset_defaults.py

@@ -0,0 +1,127 @@
+"""RAG 知识库默认标记测试。"""
+
+import asyncio
+from unittest.mock import patch
+
+from fastapi_common.fastapi_common_web.domain.responses import StatusCodeEnum
+from fastapi_common.fastapi_common_web.exception.LeauditException import LeauditException
+
+from fastapi_modules.fastapi_leaudit.services.impl.ragDatasetServiceImpl import RagDatasetServiceImpl
+
+
+class _ScalarResult:
+    def __init__(self, value):
+        self.value = value
+
+    def scalar_one(self):
+        return self.value
+
+
+class _ExecuteResult:
+    def mappings(self):
+        return self
+
+    def first(self):
+        return {"id": 1}
+
+
+class _FakeSession:
+    def __init__(self, other_default_count: int):
+        self.other_default_count = other_default_count
+        self.executed_sql = []
+        self.executed_params = []
+
+    async def execute(self, sql, params=None):
+        sql_text = str(sql)
+        self.executed_sql.append(sql_text)
+        self.executed_params.append(params or {})
+        if "COUNT(1)" in sql_text and "is_default = TRUE" in sql_text:
+            return _ScalarResult(self.other_default_count)
+        return _ExecuteResult()
+
+
+class _FakeSessionContext:
+    def __init__(self, session):
+        self.session = session
+
+    async def __aenter__(self):
+        return self.session
+
+    async def __aexit__(self, exc_type, exc, tb):
+        return None
+
+
+async def _update_default_flag_with_other_count(other_default_count: int):
+    service = RagDatasetServiceImpl()
+    fake_session = _FakeSession(other_default_count=other_default_count)
+    existing_row = {
+        "id": 10,
+        "name": "旧默认知识库",
+        "description": "",
+        "area": "梅州",
+        "tenant_code": "MZ",
+        "tenant_name": "梅州",
+        "is_public": False,
+        "is_default": True,
+        "status": 1,
+        "document_count": 0,
+        "total_chunks": 0,
+        "chunk_max_size": 800,
+        "chunk_min_size": 20,
+        "sort_order": 0,
+        "retrieval_model": {},
+        "created_at": None,
+        "updated_at": None,
+    }
+
+    async def fake_get_dataset_row(dataset_id):
+        return existing_row
+
+    async def fake_resolve_tenant_context(**kwargs):
+        return {"tenant_code": "MZ", "tenant_name": "梅州", "tenant_type": "CUSTOM", "area": "梅州"}
+
+    async def fake_resolve_dataset_area_input(**kwargs):
+        return "梅州", "MZ", object()
+
+    async def noop_async(*args, **kwargs):
+        return None
+
+    service._get_dataset_row = fake_get_dataset_row
+    service._resolve_tenant_context = fake_resolve_tenant_context
+    service._resolve_dataset_area_input = fake_resolve_dataset_area_input
+    service._assert_manage_area_scope = noop_async
+    service._ensure_rag_tenant_schema = noop_async
+    service._ensure_linked_app = noop_async
+
+    with patch(
+        "fastapi_modules.fastapi_leaudit.services.impl.ragDatasetServiceImpl.GetAsyncSession",
+        return_value=_FakeSessionContext(fake_session),
+    ):
+        result = await service.UpdateAdminDataset(
+            CurrentUserId=1,
+            UserArea="梅州",
+            UserRole="super_admin",
+            TenantCode="MZ",
+            TenantName="梅州",
+            DatasetId=10,
+            Body={"is_default": False},
+        )
+
+    return result, fake_session
+
+
+def test_update_admin_dataset_allows_unsetting_default_when_same_tenant_has_another_default():
+    result, fake_session = asyncio.run(_update_default_flag_with_other_count(other_default_count=1))
+
+    assert result is not None
+    assert any("UPDATE rag_dataset" in sql_text for sql_text in fake_session.executed_sql)
+    assert any("tenant_code = :tenant_code" in sql_text for sql_text in fake_session.executed_sql)
+    assert any(params.get("tenant_code") == "MZ" for params in fake_session.executed_params)
+
+
+def test_update_admin_dataset_rejects_unsetting_only_default_in_tenant():
+    try:
+        asyncio.run(_update_default_flag_with_other_count(other_default_count=0))
+        assert False, "expected LeauditException"
+    except LeauditException as exc:
+        assert exc.status == StatusCodeEnum.HTTP_400_BAD_REQUEST