hi

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.db filter=lfs diff=lfs merge=lfs -text

Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.9-slim
+
+WORKDIR /app
+
+COPY . /app
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+EXPOSE 7860
+
+CMD ["streamlit", "run", "app.py", "--server.port=7860", "--server.address=0.0.0.0"]

app.py

@@ -0,0 +1,103 @@
+import streamlit as st
+import pandas as pd
+import sqlite3
+import requests
+import numpy as np  # 新增：用於處理 NaN
+from io import BytesIO
+from PIL import Image
+from database import recommend_books
+import time  # 新增：用於進度條
+
+# 標題
+st.title("圖書搜尋")
+
+@st.cache_data
+def load_books():
+    conn = sqlite3.connect('books.db')
+    df = pd.read_sql_query("SELECT * FROM books", conn)
+    conn.close()
+    # 處理必要欄位型態
+    for col in ['discount_price', 'list_price', 'stock']:
+        if col in df.columns:
+            df[col] = pd.to_numeric(df[col].replace('N/A', np.nan), errors='coerce').fillna(0)
+    return df
+
+df = load_books()
+
+def run_recommendation():
+    with st.spinner("正在計算推薦..."):
+        time.sleep(0.5)
+        st.session_state.recs = recommend_books(st.session_state.search_query)
+        st.session_state.selected_display = None
+    if st.session_state.recs.empty:
+        st.warning("無推薦結果，請調整關鍵字。")
+    else:
+        st.success(f"找到 {len(st.session_state.recs)} 本推薦書籍！")
+
+if "search_query" not in st.session_state:
+    st.session_state.search_query = ""
+
+search_query = st.text_input(
+    "輸入關鍵字（如：靈修、耶穌）",
+    value=st.session_state.search_query,
+    key="search_query",
+    on_change=run_recommendation
+)
+if "recs" not in st.session_state:
+    st.session_state.recs = pd.DataFrame()
+if st.session_state.recs is None or st.session_state.recs.empty:
+    st.session_state.recs = pd.DataFrame()
+if "selected_display" not in st.session_state:
+    st.session_state.selected_display = None
+
+if not st.session_state.recs.empty:
+    st.subheader("推薦書籍列表（相似度排序）")
+    recs_display = st.session_state.recs[['product_id', 'title', 'author', 'similarity']].copy()
+    import re
+    recs_display['顯示'] = recs_display['title'].apply(lambda x: re.sub(r'<[^>]+>', '', str(x)))
+    recs_display['顯示'] = recs_display['顯示'].apply(lambda x: re.sub(r'~', '～', str(x)))
+    for idx, row in recs_display.iterrows():
+        if st.button(row['顯示'], key=f"book_{row['product_id']}"):
+            st.session_state.selected_display = row['顯示']
+
+if st.session_state.selected_display:
+    selected_row = None
+    if not st.session_state.recs.empty:
+        recs_display = st.session_state.recs[['product_id', 'title', 'author', 'similarity']].copy()
+        import re
+        recs_display['顯示'] = recs_display['title'].apply(lambda x: re.sub(r'<[^>]+>', '', str(x)))
+        recs_display['顯示'] = recs_display['顯示'].apply(lambda x: re.sub(r'~', '～', str(x)))
+        selected_rows = recs_display[recs_display['顯示'] == st.session_state.selected_display]
+        if not selected_rows.empty:
+            selected_row = selected_rows.iloc[0]
+    if selected_row is not None:
+        book_data = df[df['product_id'] == selected_row['product_id']].iloc[0]
+        def clean_text(val):
+            import re
+            return re.sub(r'~', '～', re.sub(r'<[^>]+>', '', str(val)))
+        col1, col2 = st.columns([1, 2])
+        with col1:
+            if 'image_url' in book_data and book_data['image_url']:
+                try:
+                    response = requests.get(book_data['image_url'], verify=False, timeout=10)
+                    response.raise_for_status()
+                    img = Image.open(BytesIO(response.content))
+                    img = img.resize((180, 250))
+                    st.image(img, caption=f"{clean_text(book_data['title'])} - 作者：{clean_text(book_data['author'])}", width=180)
+                except Exception as e:
+                    st.error(f"圖片載入失敗：{e}（已跳過 SSL 驗證）")
+        with col2:
+            st.header("書籍詳情")
+            product_id = book_data.get('product_id', '')
+            st.markdown(f"""
+            **書名**：{clean_text(book_data.get('title', 'N/A'))}
+            **作者**：{clean_text(book_data.get('author', 'N/A'))}
+            **出版社**：{clean_text(book_data.get('publisher', 'N/A'))}
+            **優惠價**：{book_data.get('discount_price', 'N/A')}
+            **定價**：{book_data.get('list_price', 'N/A')}
+            **購買連結**：<a href="https://shop.campus.org.tw/ProductDetails.aspx?productID={product_id}" target="_blank" rel="noopener noreferrer">前往購買</a>
+            """, unsafe_allow_html=True)
+            content_intro = book_data.get('content_intro', 'N/A')
+            content_intro = clean_text(content_intro)
+            st.markdown("**內容簡介**：")
+            st.write(content_intro[:500] + "..." if len(content_intro) > 500 else content_intro)

books.db

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edd29142b8ff8644cd700ce8fa7fbedc8eaa5441e417aa3d7fe282aa49504682
+size 1269760

database.py

@@ -0,0 +1,60 @@
+import json
+import sqlite3
+import pandas as pd
+import re
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+
+# 載入JSON數據
+def load_data(json_file):
+    with open(json_file, 'r', encoding='utf-8') as f:
+        data = json.load(f)
+    return pd.DataFrame(data)
+
+# 初始化資料庫
+def init_db(db_name='books.db'):
+    conn = sqlite3.connect(db_name)
+    df = load_data('data.json')  # 假設JSON命名為data.json
+    df.to_sql('books', conn, if_exists='replace', index=False)
+    conn.close()
+    print("資料庫初始化完成！")
+
+# 簡單推薦函數（基於內容簡介的TF-IDF，優化中文處理）
+def recommend_books(query, top_n=5):
+    conn = sqlite3.connect('books.db')
+    df = pd.read_sql_query("SELECT * FROM books", conn)
+    conn.close()
+    
+    if df.empty or 'content_intro' not in df.columns:
+        return pd.DataFrame()
+    
+    # 預處理：合併 title、author 與 content_intro，提升相關性
+    df['text'] = (df['title'].fillna('') + ' ' + df['author'].fillna('') + ' ' + df['content_intro'].fillna('')).astype(str)
+    query = re.sub(r'[^\w\s]', ' ', query.lower())  # 移除標點，簡化查詢
+    
+    # TF-IDF向量化（中文優化：使用 ngrams 捕捉詞組，無需分詞）
+    vectorizer = TfidfVectorizer(
+        max_features=2000,
+        ngram_range=(1, 3),  # 單詞、雙詞、三詞組合，適合中文
+        min_df=1,            # 最小文件頻率
+        max_df=0.95,         # 最大文件頻率（過濾過常見詞）
+        lowercase=True
+    )
+    tfidf_matrix = vectorizer.fit_transform(df['text'].values)
+    
+    query_vec = vectorizer.transform([query])
+    similarities = cosine_similarity(query_vec, tfidf_matrix).flatten()
+    
+    df['similarity'] = similarities
+    
+    # 若所有相似度為0，使用關鍵字匹配 fallback
+    if similarities.max() == 0:
+        df['keyword_score'] = df['text'].str.contains(query, case=False, na=False).astype(int)
+        df['similarity'] = df['keyword_score']  # 簡單計數匹配
+    
+    # 回傳推薦結果包含 product_id，並依相似度排序
+    recommendations = df.nlargest(top_n, 'similarity')[['product_id', 'title', 'author', 'discount_price', 'similarity']]
+    return recommendations
+
+if __name__ == "__main__":
+    init_db()

requirements.txt

@@ -0,0 +1,6 @@
+streamlit
+pandas
+requests
+numpy
+Pillow
+scikit-learn