diff --git a/app/api/routes.py b/app/api/routes.py index cc4a59d..43735b7 100644 --- a/app/api/routes.py +++ b/app/api/routes.py @@ -34,6 +34,7 @@ class QueryRequest(BaseModel): top_k: int = 5 stream: bool = False think: bool = False + only_rag: bool = False # 是否仅使用RAG检索结果，不进行LLM兜底 class IngestResponse(BaseModel): filename: str @@ -121,16 +122,41 @@ async def query_knowledge_base( enable_rerank=settings.RERANK_ENABLED ) - # 处理流式输出 (SSE 协议) + # 处理流式输出 (SSE 协议 - OpenAI 兼容格式) if request.stream: + import time async def stream_generator(): - # SSE 格式化辅助函数 - def sse_pack(event: str, text: str) -> str: - # 使用 JSON 包装 data 内容，确保换行符和特殊字符被正确转义 - data = json.dumps({"text": text}, ensure_ascii=False) - return f"event: {event}\ndata: {data}\n\n" + chat_id = f"chatcmpl-{secrets.token_hex(12)}" + created_time = int(time.time()) + model_name = settings.LLM_MODEL - yield sse_pack("thinking", "1. 上下文检索中...\n") + # 辅助函数：构造 OpenAI 兼容的 Chunk + def openai_chunk(content=None, reasoning_content=None, finish_reason=None, extra_delta=None): + delta = {} + if content: + delta["content"] = content + if reasoning_content: + delta["reasoning_content"] = reasoning_content + if extra_delta: + delta.update(extra_delta) + + chunk = { + "id": chat_id, + "object": "chat.completion.chunk", + "created": created_time, + "model": model_name, + "choices": [ + { + "index": 0, + "delta": delta, + "finish_reason": finish_reason + } + ] + } + return f"data: {json.dumps(chunk, ensure_ascii=False)}\n\n" + + # 1. 发送检索状态 (作为思考过程的一部分) + yield openai_chunk(reasoning_content="1. 上下文检索中...\n") context_param = QueryParam( mode=request.mode, @@ -139,9 +165,8 @@ async def query_knowledge_base( enable_rerank=settings.RERANK_ENABLED ) - # 获取上下文 (这步耗时较长，包含图遍历) + # 获取上下文 context_resp = await rag.aquery(request.query, param=context_param) - logging.info(f"Context Response: {context_resp}") # 判断检索状态 @@ -153,17 +178,28 @@ async def query_knowledge_base( think = request.think if has_context: - yield sse_pack("system", "retrieved") # 发送系统事件：已检索到信息 - yield sse_pack("thinking", f"2. 上下文已检索 (长度: {len(context_resp)} 字符).\n") + yield openai_chunk( + reasoning_content=f"2. 上下文已检索 (长度: {len(context_resp)} 字符).\n", + extra_delta={"x_rag_status": "hit"} + ) else: - yield sse_pack("system", "missed") # 发送系统事件：未检索到信息 - yield sse_pack("thinking", "2. 未找到相关上下文，将依赖 LLM 自身知识\n") + yield openai_chunk( + reasoning_content="2. 未找到相关上下文\n", + extra_delta={"x_rag_status": "miss"} + ) + + # 如果开启了仅RAG模式且未找到上下文，则直接结束 + if request.only_rag: + yield openai_chunk(content="未找到相关知识库内容。", finish_reason="stop") + yield "data: [DONE]\n\n" + return + + yield openai_chunk(reasoning_content=" (将依赖 LLM 自身知识)\n") think = False - yield sse_pack("thinking", "3. 答案生成中...\n") + yield openai_chunk(reasoning_content="3. 答案生成中...\n") # 2. 生成答案 - # 手动构建 System Prompt sys_prompt = CUSTOM_RAG_RESPONSE_PROMPT.format( context_data=context_resp, response_type="Multiple Paragraphs", @@ -179,23 +215,18 @@ async def query_knowledge_base( hashing_kv=rag.llm_response_cache ) - thinkState = 0 # think 状态 0: 未开始 1: 开始 2: 结束 async for chunk in stream_resp: if isinstance(chunk, dict): if chunk.get("type") == "thinking": - if thinkState == 0: - yield sse_pack("thinking", "\n思考:\n") - thinkState = 1 - - yield sse_pack("thinking", chunk["content"]) + yield openai_chunk(reasoning_content=chunk["content"]) elif chunk.get("type") == "content": - if thinkState == 1: - yield sse_pack("none", "\n\n\n") - thinkState = 2 - - yield sse_pack("answer", chunk["content"]) + yield openai_chunk(content=chunk["content"]) elif chunk: - yield sse_pack("answer", chunk) + yield openai_chunk(content=chunk) + + # 发送结束标记 + yield openai_chunk(finish_reason="stop") + yield "data: [DONE]\n\n" # 使用 text/event-stream Content-Type return StreamingResponse(stream_generator(), media_type="text/event-stream") diff --git a/app/static/admin.html b/app/static/admin.html index 9607935..fb17267 100644 --- a/app/static/admin.html +++ b/app/static/admin.html @@ -103,8 +103,9 @@
-
- +
+ + 发送
@@ -201,6 +202,7 @@ // 聊天 const queryInput = ref(''); + const onlyRag = ref(false); const chatHistory = ref([]); const chatLoading = ref(false); const chatBox = ref(null); @@ -375,7 +377,13 @@ 'Content-Type': 'application/json', 'X-Tenant-ID': currentTenantId.value }, - body: JSON.stringify({ query: q, stream: true, mode: 'mix', think: true }) + body: JSON.stringify({ + query: q, + stream: true, + mode: 'mix', + think: true, + only_rag: onlyRag.value + }) }); const reader = response.body.getReader(); @@ -397,32 +405,43 @@ buffer = blocks.pop(); // 保留最后一个可能不完整的块 for (const block of blocks) { + if (!block.trim() || block.trim() === 'data: [DONE]') continue; + const lines = block.split('\n'); - let eventType = 'answer'; - let dataText = ''; - for (const line of lines) { - if (line.startsWith('event: ')) { - eventType = line.slice(7).trim(); - } else if (line.startsWith('data: ')) { + if (line.startsWith('data: ')) { try { - const data = JSON.parse(line.slice(6)); - dataText = data.text; - } catch (e) {} + const jsonStr = line.slice(6); + const chunk = JSON.parse(jsonStr); + + // 解析 OpenAI 兼容格式 + if (chunk.choices && chunk.choices[0].delta) { + const delta = chunk.choices[0].delta; + + // 处理 x_rag_status + if (delta.x_rag_status) { + assistantMsg.retrievalStatus = delta.x_rag_status; + } + + // 处理思考过程 + if (delta.reasoning_content) { + assistantMsg.thinking += delta.reasoning_content; + } + + // 处理正文内容 + if (delta.content) { + assistantMsg.content += delta.content; + } + } + + // 滚动到底部 + if (chatBox.value) chatBox.value.scrollTop = chatBox.value.scrollHeight; + + } catch (e) { + console.error('JSON parse error:', e); + } } } - - if (eventType === 'system') { - assistantMsg.retrievalStatus = dataText; - } else if (dataText) { - if (eventType === 'thinking') { - assistantMsg.thinking += dataText; - } else if (eventType === 'answer') { - assistantMsg.content += dataText; - } - // 滚动到底部 - if (chatBox.value) chatBox.value.scrollTop = chatBox.value.scrollHeight; - } } } } catch (e) { @@ -436,9 +455,9 @@ let html = ''; if (retrievalStatus) { - const color = retrievalStatus === 'retrieved' ? '#67c23a' : '#e6a23c'; - const text = retrievalStatus === 'retrieved' ? '已检索到相关知识' : '未检索到相关知识，使用通用知识回答'; - const icon = retrievalStatus === 'retrieved' ? '✔️' : '⚠️'; + const color = retrievalStatus === 'hit' ? '#67c23a' : '#e6a23c'; + const text = retrievalStatus === 'hit' ? '已检索到相关知识' : '未检索到相关知识'; + const icon = retrievalStatus === 'hit' ? '✔️' : '⚠️'; html += `
${icon} ${text}
`; } @@ -468,7 +487,7 @@ goHome, refreshTenants, enterTenant, fetchDocuments, viewDocument, deleteDocument, deleteCurrentDoc, uploadFile, uploadText, addQA, removeQA, uploadQA, - sendQuery, renderMarkdown, formatDate, isAdmin + sendQuery, renderMarkdown, formatDate, isAdmin, onlyRag }; } });