串流回應

串流回應

透過 Server-Sent Events（SSE）串流，即時接收 Lawbot AI 的逐字生成回應， 提供更流暢的使用者體驗。

串流運作方式

在請求中設定 stream: true， API 將以 Server-Sent Events 格式回傳一系列 data: 事件， 每個事件包含一個 JSON chunk。串流結束時會傳送 data: [DONE]。

cURL 串流範例

terminal
curl https://developer-api.lawbot.tw/api/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "lawbot-flash",
    "stream": true,
    "reasoning_effort": "medium",
    "temperature": 0.3,
    "messages": [
      { "role": "user", "content": "請問股東會決議瑕疵的法律效果為何？" }
    ]
  }'

SSE 回應格式：

sse_response.txt
data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"delta":{"role":"assistant"},"index":0}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"delta":{"content":"股東會"},"index":0}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"delta":{"content":"決議瑕疵"},"index":0}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"delta":{"content":"依公司法"},"index":0}]}

data: {"id":"chatcmpl-abc","object":"chat.completion.chunk","choices":[{"delta":{},"finish_reason":"stop","index":0}],"x_lawbot_references":[{"doc_id":"law-corp-189","title":"公司法第 189 條","doc_type":"法規命令"}]}

data: [DONE]

Python 串流

使用 OpenAI Python SDK，SDK 已自動處理 SSE 解析：

streaming.py
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://developer-api.lawbot.tw/api/v1",
)

# 傳入 stream=True 啟用串流
stream = client.chat.completions.create(
    model="lawbot-flash",
    stream=True,
    temperature=0.3,
    messages=[
        {
            "role": "user",
            "content": "請問股東會決議瑕疵的法律效果為何？",
        }
    ],
    extra_body={"reasoning_effort": "medium"},
)

# 逐 chunk 處理回應
for chunk in stream:
    content = chunk.choices[0].delta.content
    if content is not None:
        print(content, end="", flush=True)

print()  # 最後換行

Node.js / TypeScript 串流

使用 OpenAI Node.js SDK，透過 for await...of 非同步迭代：

streaming.ts
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.LAWBOT_API_KEY,
  baseURL: "https://developer-api.lawbot.tw/api/v1",
});

const stream = await client.chat.completions.create({
  model: "lawbot-flash",
  stream: true,
  messages: [
    {
      role: "user",
      content: "請問股東會決議瑕疵的法律效果為何？",
    },
  ],
});

// 逐 chunk 處理回應
for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content;
  if (content) process.stdout.write(content);
}

原生 Fetch API 處理 SSE

若不使用 SDK，可透過原生 Fetch API 手動解析 SSE 串流：

raw_stream.ts
const response = await fetch(
  "https://developer-api.lawbot.tw/api/v1/chat/completions",
  {
    method: "POST",
    headers: {
      "Authorization": `Bearer ${apiKey}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      model: "lawbot-flash",
      stream: true,
      messages: [{ role: "user", content: "..." }],
    }),
  }
);

const reader = response.body!.getReader();
const decoder = new TextDecoder();

while (true) {
  const { done, value } = await reader.read();
  if (done) break;

  const text = decoder.decode(value);
  const lines = text.split("\n").filter((l) => l.startsWith("data: "));

  for (const line of lines) {
    const data = line.slice(6); // remove "data: "
    if (data === "[DONE]") break;

    const parsed = JSON.parse(data);
    const content = parsed.choices[0]?.delta?.content;
    if (content) console.log(content);
  }
}