src/streamlit_app.py

import streamlit as st
import google.generativeai as genai
import requests
import os

st.set_page_config(page_title="AI 新知小助手", page_icon="📚", layout="wide")

# 從環境變數中取得 API Key
api_key = os.environ.get("GEMINI_API_KEY")

if not api_key:
    st.error("請確認已經在 Space 的 Settings -> Variables and secrets 中設定了 GEMINI_API_KEY")
    st.stop()

genai.configure(api_key=api_key)

# 1. 基礎設定與連結
BASE_URL = "https://raw.githubusercontent.com/Deep-Learning-101/deep-learning-101.github.io/main/"
LOGO_URL = f"{BASE_URL}images/DeepLearning101-LOGO.png"
HOME_URL = "https://deep-learning-101.github.io"

# 定義知識庫檔案與對應的社群連結
KNOWLEDGE_MAP = {
    "大型語言模型 (LLM)": {
        "raw_url": f"{BASE_URL}Large-Language-Model.md",
        "page_url": "https://deep-learning-101.github.io/Large-Language-Model",
        "repo_url": "https://github.com/Deep-Learning-101/Natural-Language-Processing-Paper/blob/main/Large-Language-Model.md"
    },
    "自然語言處理 (NLP)": {
        "raw_url": f"{BASE_URL}Natural-Language-Processing.md",
        "page_url": "https://deep-learning-101.github.io/Natural-Language-Processing",
        "repo_url": "https://github.com/Deep-Learning-101/Natural-Language-Processing-Paper"
    },
    "語音處理 (Speech)": {
        "raw_url": f"{BASE_URL}Speech-Processing.md",
        "page_url": "https://deep-learning-101.github.io/Speech-Processing",
        "repo_url": "https://github.com/Deep-Learning-101/Speech-Processing-Paper"
    },
    "電腦視覺 (CV)": {
        "raw_url": f"{BASE_URL}Computer-Vision.md",
        "page_url": "https://deep-learning-101.github.io/Computer-Vision",
        "repo_url": "https://github.com/Deep-Learning-101/Computer-Vision-Paper"
    }
}

# 2. 批量抓取內容
def fetch_all_knowledge():
    combined_knowledge = ""
    with st.spinner("正在同步 GitHub 最新資訊..."):
        for category, info in KNOWLEDGE_MAP.items():
            try:
                response = requests.get(info["raw_url"])
                response.raise_for_status()
                combined_knowledge += f"\n\n## 【領域：{category}】\n"
                combined_knowledge += response.text
            except Exception as e:
                st.warning(f"無法同步 {category} 的資料：{e}")
    return combined_knowledge

if "knowledge" not in st.session_state:
    st.session_state.knowledge = fetch_all_knowledge()

if "messages" not in st.session_state:
    st.session_state.messages = []

# 3. 側邊欄設計
with st.sidebar:
    # 放置可點擊的 Logo
    st.markdown(
        f"""
        <a href="{HOME_URL}" target="_blank">
            <img src="{LOGO_URL}" width="85%">
        </a>
        """,
        unsafe_allow_html=True
    )
    
    st.title("⚙️ 知識庫狀態")
    st.write("目前收錄領域與連結：")
    
    # 迴圈產生各領域連結
    for category, info in KNOWLEDGE_MAP.items():
        with st.expander(category):
            st.markdown(f"🔗 [瀏覽網頁]({info['page_url']})")
            st.markdown(f"📂 [GitHub 原始碼]({info['repo_url']})")
        
    st.markdown("---")
    if st.button("🔄 手動更新知識庫"):
        st.session_state.knowledge = fetch_all_knowledge()
        st.success("資料已重新抓取！")

# 4. 主介面
st.title("📚 AI 演算法與論文社群助手")
st.caption("知識庫涵蓋 LLM、NLP、Speech、CV。歡迎直接提問！")

def get_gemini_response(user_input):
    system_instruction = f"""
    你是一位專業的 AI 技術分析專家。
    以下是從 GitHub 同步的技術資訊：
    ---
    {st.session_state.knowledge}
    ---
    請嚴格基於上述提供的資訊來回答問題。
    如果資訊中未收錄，請告知：「目前懶人包中尚未收錄此細節」。
    """
    model = genai.GenerativeModel(
        model_name="gemini-flash-latest",
        system_instruction=system_instruction
    )
    
    chat = model.start_chat(history=[])
    response = chat.send_message(user_input)
    return response.text

# 顯示對話紀錄
for message in st.session_state.messages:
    with st.chat_message(message["role"]):
        st.markdown(message["content"])

# 接收輸入
if prompt := st.chat_input("想瞭解哪方面的技術？"):
    st.session_state.messages.append({"role": "user", "content": prompt})
    with st.chat_message("user"):
        st.markdown(prompt)

    with st.chat_message("assistant"):
        response_text = get_gemini_response(prompt)
        st.markdown(response_text)
    
    st.session_state.messages.append({"role": "assistant", "content": response_text})