# 啟用 Agentic Vision 的 API 呼叫
response = client.models.generate_content(
    model="gemini-3-flash-preview",
    contents=[image_part, "幫我把咖啡標記出來"],
    config=types.GenerateContentConfig(
        tools=[types.Tool(code_execution=types.ToolCodeExecution)],
        thinking_config=types.ThinkingConfig(thinkingBudget=2048),
    )
)

# Response 包含多種 part：文字、程式碼、執行結果、標註圖片
for part in response.candidates[0].content.parts:
    if part.text:           # 文字分析
    if part.executable_code: # 模型寫的 Python 程式碼
    if part.code_execution_result:  # 程式碼執行結果
    if part.as_image():     # 生成的標註圖片！

# main.py
image_temp_store: Dict[str, bytes] = {}        # 暫存圖片（user_id → bytes）
pending_agentic_vision: Dict[str, bool] = {}    # 等待使用者輸入指令

quick_reply_buttons = QuickReply(
    items=[
        QuickReplyButton(
            action=PostbackAction(
                label="識別圖片",
                data=json.dumps({"action": "image_analyze", "mode": "recognize"}),
                display_text="識別圖片"
            )
        ),
        QuickReplyButton(
            action=PostbackAction(
                label="Agentic Vision",
                data=json.dumps({"action": "image_analyze", "mode": "agentic_vision"}),
                display_text="Agentic Vision"
            )
        ),
    ]
)

# tools/summarizer.py
def analyze_image_agentic(image_data: bytes, prompt: str) -> dict:
    client = _get_vertex_client()

    contents = [
        types.Part.from_text(text=prompt),
        types.Part.from_bytes(data=image_data, mime_type="image/png")
    ]

    response = client.models.generate_content(
        model="gemini-3-flash-preview",
        contents=contents,
        config=types.GenerateContentConfig(
            temperature=0.5,
            max_output_tokens=4096,
            tools=[types.Tool(code_execution=types.ToolCodeExecution)],
            thinking_config=types.ThinkingConfig(thinkingBudget=2048),
        )
    )

    result_parts = []
    generated_images = []

    for part in response.candidates[0].content.parts:
        if hasattr(part, 'thought') and part.thought:
            continue  # 跳過 thinking parts
        if part.text is not None:
            result_parts.append(part.text)
        if part.code_execution_result is not None:
            result_parts.append(f"[Code Output]: {part.code_execution_result.output}")
        # 提取模型生成的標註圖片
        img = part.as_image()
        if img is not None:
            generated_images.append(img.image_bytes)

    return {
        "status": "success",
        "analysis": "\n".join(result_parts),
        "images": generated_images  # 標註後的圖片 bytes
    }

# 暫存標註圖片（UUID → bytes，5 分鐘 TTL）
annotated_image_store: Dict[str, dict] = {}

@app.get("/images/{image_id}")
def serve_annotated_image(image_id: str):
    """提供暫存的標註圖片給 LINE 下載"""
    entry = annotated_image_store.get(image_id)
    if not entry:
        raise HTTPException(status_code=404)
    if time.time() - entry["created_at"] > 300:  # 5 分鐘過期
        annotated_image_store.pop(image_id, None)
        raise HTTPException(status_code=404)
    return Response(content=entry["data"], media_type="image/png")

@app.post("/")
async def handle_webhook_callback(request: Request):
    global app_base_url
    if not app_base_url:
        forwarded_proto = request.headers.get('x-forwarded-proto', 'https')
        host = request.headers.get('x-forwarded-host') or request.headers.get('host', '')
        if host:
            app_base_url = f"{forwarded_proto}://{host}"

def _create_image_send_message(image_bytes: bytes):
    image_id = store_annotated_image(image_bytes)
    image_url = f"{app_base_url}/images/{image_id}"
    return ImageSendMessage(
        original_content_url=image_url,
        preview_image_url=image_url,
    )

# ❌ 錯誤
types.Part.from_image_bytes(data=image_data, mime_type="image/png")

# ✅ 正確
types.Part.from_bytes(data=image_data, mime_type="image/png")

# ❌ 在 v1.49.0 不存在
types.ThinkingConfig(thinking_level=types.ThinkingLevel.MEDIUM)

# ✅ v1.49.0 支援的寫法
types.ThinkingConfig(thinkingBudget=2048)

# ❌ 改前：thinking 消耗了大部分 token 額度
config=types.GenerateContentConfig(
    max_output_tokens=2048,
)

# ✅ 改後：關閉 thinking + 加大 token 額度
config=types.GenerateContentConfig(
    max_output_tokens=8192,
    thinking_config=types.ThinkingConfig(thinkingBudget=0),  # 關閉 thinking
)

image_temp_store: Dict[str, bytes] = {}       # 圖片暫存
pending_agentic_vision: Dict[str, bool] = {}   # 等待指令