hi

Dockerfile

@@ -0,0 +1,33 @@
+FROM python:3.11-slim
+
+# ---- System deps ----
+RUN apt-get update && apt-get upgrade -y && apt-get install -y --no-install-recommends \
+    build-essential \
+    git \
+    curl \
+ && rm -rf /var/lib/apt/lists/*
+
+# ---- Workdir ----
+WORKDIR /app
+
+# ---- Copy requirement & install ----
+COPY requirements.txt /app/requirements.txt
+RUN pip install --no-cache-dir -U pip \
+ && pip install --no-cache-dir -r /app/requirements.txt
+
+# ---- Runtime cache to /tmp (writeable) ----
+ENV XDG_CACHE_HOME=/tmp/.cache
+
+# ---- Copy app ----
+COPY . /app
+
+# ---- Healthcheck ----
+HEALTHCHECK --interval=30s --timeout=10s --retries=3 \
+    CMD curl -f http://localhost:${PORT:-7860}/health || exit 1
+
+# ---- Port & CMD ----
+EXPOSE 7860
+ENV PORT=7860 \
+    PYTHONUNBUFFERED=1
+
+CMD ["sh", "-c", "uvicorn app:app --host 0.0.0.0 --port ${PORT:-7860} --log-level info"]

app.py

@@ -0,0 +1,161 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# ---------- 環境與快取設定 (應置於最前) ----------
+import os
+import json
+import time
+import logging
+from typing import List, Dict, Any
+from contextlib import asynccontextmanager
+from fastapi import FastAPI, Request, HTTPException, status
+import uvicorn
+from linebot import LineBotApi
+from linebot.models import TextSendMessage
+from openai import OpenAI
+import hmac
+import base64
+import hashlib
+
+# ==== CONFIG (從環境變數載入，或使用預設值) ====
+def _require_env(var: str) -> str:
+    v = os.getenv(var)
+    if not v:
+        raise RuntimeError(f"FATAL: Missing required environment variable: {var}")
+    return v
+
+CHANNEL_SECRET = _require_env("CHANNEL_SECRET")
+CHANNEL_ACCESS_TOKEN = _require_env("CHANNEL_ACCESS_TOKEN")
+
+LLM_API_CONFIG = {
+    "base_url": "https://litellm-ekkks8gsocw.dgx-coolify.apmic.ai/",
+    "api_key": "sk-eT_04m428oAPUD5kUmIhVA",
+}
+LLM_MODEL_CONFIG = {
+    "model": "gpt-oss-120b",
+    "max_tokens": int(os.getenv("MAX_TOKENS", 1024)),
+    "temperature": float(os.getenv("TEMPERATURE", 0.3)),
+    "seed": int(os.getenv("LLM_SEED", 42)),
+}
+
+SYSTEM_PROMPT = "你是一個友好的AI助手，請用簡單、親切的中文回覆用戶的問題。"
+
+# ---------- 日誌設定 ----------
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+log = logging.getLogger(__name__)
+
+# ---------- 記憶體儲存對話歷史 ----------
+conversations: Dict[str, List[Dict[str, str]]] = {}  # {user_id: messages_list}
+
+# ---------- Helper to split long message for LINE API ----------
+def split_text_for_line(text: str, max_length: int = 4800) -> List[str]:
+    if len(text) <= max_length:
+        return [text]
+    chunks = []
+    while text:
+        if len(text) <= max_length:
+            chunks.append(text)
+            break
+        split_pos = text.rfind('\n', 0, max_length)
+        if split_pos == -1:
+            split_pos = max_length
+        chunks.append(text[:split_pos])
+        text = text[split_pos:].lstrip()
+    return chunks
+
+# ---------- Chat Pipeline ----------
+class ChatPipeline:
+    def __init__(self):
+        if not LLM_API_CONFIG["api_key"] or not LLM_API_CONFIG["base_url"]:
+            raise ValueError("LLM API Key or Base URL is not configured.")
+        self.llm_client = OpenAI(api_key=LLM_API_CONFIG["api_key"], base_url=LLM_API_CONFIG["base_url"])
+
+    def _llm_call(self, messages: List[Dict[str, str]]) -> str:
+        log.info("LLM 呼叫開始.")
+        start_time = time.time()
+        response = self.llm_client.chat.completions.create(
+            model=LLM_MODEL_CONFIG["model"],
+            messages=messages,
+            max_tokens=LLM_MODEL_CONFIG["max_tokens"],
+            temperature=LLM_MODEL_CONFIG["temperature"],
+            seed=LLM_MODEL_CONFIG["seed"],
+        )
+        content = response.choices[0].message.content or ""
+        log.info(f"LLM 呼叫完成，耗時: {time.time() - start_time:.2f} 秒。內容長度: {len(content)}.")
+        return content
+
+    def get_conversation_history(self, user_id: str) -> List[Dict[str, str]]:
+        return conversations.get(user_id, [])
+
+    def update_conversation_history(self, user_id: str, messages: List[Dict[str, str]]):
+        # 保留最近10條訊息
+        if len(messages) > 10:
+            messages = messages[-10:]
+        conversations[user_id] = messages
+
+    def clear_conversation_history(self, user_id: str):
+        if user_id in conversations:
+            del conversations[user_id]
+
+    def answer_question(self, user_id: str, user_text: str) -> str:
+        if user_text.strip() == "/clear":
+            self.clear_conversation_history(user_id)
+            return "對話紀錄已清除！現在開始新的對話。"
+        
+        history = self.get_conversation_history(user_id)
+        messages = [{"role": "system", "content": SYSTEM_PROMPT}]
+        messages.extend(history)
+        messages.append({"role": "user", "content": user_text})
+        response = self._llm_call(messages)
+        # 更新歷史
+        history.append({"role": "user", "content": user_text})
+        history.append({"role": "assistant", "content": response})
+        self.update_conversation_history(user_id, history)
+        return response
+
+# ---------- FastAPI 應用程式設置 ----------
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    log.info("應用程式啟動...")
+    global chat_pipeline
+    chat_pipeline = ChatPipeline()
+    yield
+    log.info("應用程式關閉...")
+
+app = FastAPI(lifespan=lifespan)
+chat_pipeline = None
+
+# LINE Bot API
+line_bot_api = LineBotApi(CHANNEL_ACCESS_TOKEN)
+
+# ---------- LINE Webhook 處理 ----------
+@app.post("/webhook")
+async def line_webhook(request: Request):
+    signature = request.headers.get('X-Line-Signature', '')
+    body = await request.body()
+    hash_ = hmac.new(CHANNEL_SECRET.encode('utf-8'), body, hashlib.sha256).digest()
+    expected_signature = base64.b64encode(hash_).decode()
+    if not hmac.compare_digest(signature, expected_signature):
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Invalid signature")
+    
+    body_str = body.decode('utf-8')
+    payload = json.loads(body_str)
+    for event in payload.get('events', []):
+        user_id = event['source']['userId']
+        reply_token = event.get('replyToken')
+        if event['type'] == 'message' and event['message']['type'] == 'text':
+            user_text = event['message']['text'].strip()
+            if not user_text:
+                continue
+            try:
+                ai_response = chat_pipeline.answer_question(user_id, user_text)
+                chunks = split_text_for_line(ai_response)
+                messages_to_send = [TextSendMessage(text=chunk) for chunk in chunks]
+                line_bot_api.reply_message(reply_token, messages_to_send[:5])  # LINE限5條
+            except Exception as e:
+                log.error(f"Error processing request: {e}", exc_info=True)
+                error_message = "抱歉，系統發生錯誤，請稍後再試。"
+                line_bot_api.reply_message(reply_token, [TextSendMessage(text=error_message)])
+    return {"status": "ok"}
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -0,0 +1,4 @@
+fastapi
+uvicorn
+line-bot-sdk
+openai