Spaces:
Sleeping
Sleeping
Update src/streamlit_app.py
Browse files- src/streamlit_app.py +258 -38
src/streamlit_app.py
CHANGED
|
@@ -1,40 +1,260 @@
|
|
| 1 |
-
import altair as alt
|
| 2 |
-
import numpy as np
|
| 3 |
-
import pandas as pd
|
| 4 |
import streamlit as st
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
Edit `/streamlit_app.py` to customize this app to your heart's desire :heart:.
|
| 10 |
-
If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
|
| 11 |
-
forums](https://discuss.streamlit.io).
|
| 12 |
-
|
| 13 |
-
In the meantime, below is an example of what you can do with just a few lines of code:
|
| 14 |
-
"""
|
| 15 |
-
|
| 16 |
-
num_points = st.slider("Number of points in spiral", 1, 10000, 1100)
|
| 17 |
-
num_turns = st.slider("Number of turns in spiral", 1, 300, 31)
|
| 18 |
-
|
| 19 |
-
indices = np.linspace(0, 1, num_points)
|
| 20 |
-
theta = 2 * np.pi * num_turns * indices
|
| 21 |
-
radius = indices
|
| 22 |
-
|
| 23 |
-
x = radius * np.cos(theta)
|
| 24 |
-
y = radius * np.sin(theta)
|
| 25 |
-
|
| 26 |
-
df = pd.DataFrame({
|
| 27 |
-
"x": x,
|
| 28 |
-
"y": y,
|
| 29 |
-
"idx": indices,
|
| 30 |
-
"rand": np.random.randn(num_points),
|
| 31 |
-
})
|
| 32 |
-
|
| 33 |
-
st.altair_chart(alt.Chart(df, height=700, width=700)
|
| 34 |
-
.mark_point(filled=True)
|
| 35 |
-
.encode(
|
| 36 |
-
x=alt.X("x", axis=None),
|
| 37 |
-
y=alt.Y("y", axis=None),
|
| 38 |
-
color=alt.Color("idx", legend=None, scale=alt.Scale()),
|
| 39 |
-
size=alt.Size("rand", legend=None, scale=alt.Scale(range=[1, 150])),
|
| 40 |
-
))
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
+
import PyPDF2
|
| 3 |
+
import pandas as pd
|
| 4 |
+
import os
|
| 5 |
+
import google.generativeai as genai
|
| 6 |
+
import csv
|
| 7 |
+
from datetime import datetime
|
| 8 |
+
import logging
|
| 9 |
+
import sys
|
| 10 |
+
import io
|
| 11 |
+
import tempfile
|
| 12 |
+
import base64
|
| 13 |
+
|
| 14 |
+
# 設定日誌
|
| 15 |
+
logging.basicConfig(
|
| 16 |
+
level=logging.INFO,
|
| 17 |
+
format='%(asctime)s - %(levelname)s - %(message)s',
|
| 18 |
+
handlers=[
|
| 19 |
+
logging.FileHandler("pdf_processing.log"),
|
| 20 |
+
logging.StreamHandler(sys.stdout)
|
| 21 |
+
]
|
| 22 |
+
)
|
| 23 |
+
logger = logging.getLogger(__name__)
|
| 24 |
+
|
| 25 |
+
# 設定頁面配置
|
| 26 |
+
st.set_page_config(
|
| 27 |
+
page_title="PDF處理與Gemini翻譯工具",
|
| 28 |
+
page_icon="📄",
|
| 29 |
+
layout="wide",
|
| 30 |
+
initial_sidebar_state="expanded"
|
| 31 |
+
)
|
| 32 |
+
|
| 33 |
+
# 應用程式標題和介紹
|
| 34 |
+
st.title("📄 PDF處理與Gemini翻譯工具")
|
| 35 |
+
st.markdown("上傳PDF檔案,選擇要處理的頁面,讓Gemini解釋內容並翻譯成繁體中文。")
|
| 36 |
+
|
| 37 |
+
# 側邊欄 - 設定區域
|
| 38 |
+
with st.sidebar:
|
| 39 |
+
st.header("設定")
|
| 40 |
+
|
| 41 |
+
# API金鑰輸入
|
| 42 |
+
api_key = st.text_input(
|
| 43 |
+
"Gemini API金鑰",
|
| 44 |
+
value="AIzaSyC3fbJVYQbINa8ztuOS5dFE4ud0I9jhy3o", # 預設值,實際使用應從設定檔讀取
|
| 45 |
+
type="password"
|
| 46 |
+
)
|
| 47 |
+
|
| 48 |
+
# 上傳PDF檔案
|
| 49 |
+
uploaded_file = st.file_uploader("上傳PDF檔案", type=["pdf"])
|
| 50 |
+
|
| 51 |
+
# 處理選項區塊
|
| 52 |
+
with st.expander("處理選項", expanded=True):
|
| 53 |
+
# 初始化session_state
|
| 54 |
+
if 'total_pages' not in st.session_state:
|
| 55 |
+
st.session_state.total_pages = 0
|
| 56 |
+
if 'page_content' not in st.session_state:
|
| 57 |
+
st.session_state.page_content = {}
|
| 58 |
+
|
| 59 |
+
# 頁面選擇(只在上傳檔案後顯示)
|
| 60 |
+
if uploaded_file is not None:
|
| 61 |
+
# 讀取PDF並獲取頁數
|
| 62 |
+
try:
|
| 63 |
+
pdf_reader = PyPDF2.PdfReader(uploaded_file)
|
| 64 |
+
st.session_state.total_pages = len(pdf_reader.pages)
|
| 65 |
+
|
| 66 |
+
# 加載PDF內容到session_state(如果尚未加載)
|
| 67 |
+
if len(st.session_state.page_content) == 0:
|
| 68 |
+
with st.spinner("正在加載PDF..."):
|
| 69 |
+
for i in range(st.session_state.total_pages):
|
| 70 |
+
st.session_state.page_content[i+1] = pdf_reader.pages[i].extract_text()
|
| 71 |
+
|
| 72 |
+
# 頁面選擇滑桿
|
| 73 |
+
page_to_process = st.slider(
|
| 74 |
+
"選擇要處理的頁面",
|
| 75 |
+
min_value=1,
|
| 76 |
+
max_value=st.session_state.total_pages,
|
| 77 |
+
value=1
|
| 78 |
+
)
|
| 79 |
+
|
| 80 |
+
st.info(f"PDF共有 {st.session_state.total_pages} 頁")
|
| 81 |
+
except Exception as e:
|
| 82 |
+
st.error(f"無法讀取PDF: {str(e)}")
|
| 83 |
+
else:
|
| 84 |
+
st.info("請先上傳PDF檔案")
|
| 85 |
+
page_to_process = 1
|
| 86 |
+
|
| 87 |
+
# 進階選項
|
| 88 |
+
with st.expander("進階選項"):
|
| 89 |
+
# 給Gemini的指示詞
|
| 90 |
+
instruction = st.text_area(
|
| 91 |
+
"給Gemini的指示詞",
|
| 92 |
+
value="請詳細解釋以下內容的主要要點和重要信息",
|
| 93 |
+
height=100
|
| 94 |
+
)
|
| 95 |
+
|
| 96 |
+
# 輸出檔名
|
| 97 |
+
output_filename = st.text_input(
|
| 98 |
+
"輸出CSV檔名",
|
| 99 |
+
value="gemini_translated_results.csv"
|
| 100 |
+
)
|
| 101 |
+
|
| 102 |
+
# 主要功能函數
|
| 103 |
+
def setup_gemini_api(api_key):
|
| 104 |
+
"""設置Gemini API"""
|
| 105 |
+
try:
|
| 106 |
+
os.environ["GOOGLE_API_KEY"] = api_key
|
| 107 |
+
genai.configure(api_key=api_key)
|
| 108 |
+
return genai.GenerativeModel("gemini-1.5-flash")
|
| 109 |
+
except Exception as e:
|
| 110 |
+
logger.error(f"Gemini API設置失敗: {e}")
|
| 111 |
+
st.error(f"API設置失敗: {str(e)}")
|
| 112 |
+
return None
|
| 113 |
+
|
| 114 |
+
def process_with_gemini(model, text, instruction):
|
| 115 |
+
"""使用Gemini處理文本"""
|
| 116 |
+
try:
|
| 117 |
+
prompt = f"{instruction}:\n\n{text}"
|
| 118 |
+
response = model.generate_content(prompt)
|
| 119 |
+
return response.text.strip()
|
| 120 |
+
except Exception as e:
|
| 121 |
+
logger.error(f"Gemini處理失敗: {e}")
|
| 122 |
+
return f"處理失敗: {str(e)}"
|
| 123 |
+
|
| 124 |
+
def translate_with_gemini(model, text):
|
| 125 |
+
"""使用Gemini將文本翻譯成繁體中文"""
|
| 126 |
+
try:
|
| 127 |
+
prompt = f"""
|
| 128 |
+
請將以下文本翻譯成繁體中文,保持專業和準確性:
|
| 129 |
+
|
| 130 |
+
{text}
|
| 131 |
+
|
| 132 |
+
只需要返回翻譯後的文本,不要加入其他解釋或備註。
|
| 133 |
+
"""
|
| 134 |
+
|
| 135 |
+
response = model.generate_content(prompt)
|
| 136 |
+
return response.text.strip()
|
| 137 |
+
except Exception as e:
|
| 138 |
+
logger.error(f"Gemini翻譯失敗: {e}")
|
| 139 |
+
return f"翻譯失敗: {str(e)}"
|
| 140 |
+
|
| 141 |
+
def get_csv_download_link(df, filename="data.csv"):
|
| 142 |
+
"""生成CSV檔案下載連結"""
|
| 143 |
+
csv = df.to_csv(index=False)
|
| 144 |
+
b64 = base64.b64encode(csv.encode()).decode()
|
| 145 |
+
href = f'<a href="data:file/csv;base64,{b64}" download="{filename}">下載 CSV 檔案</a>'
|
| 146 |
+
return href
|
| 147 |
+
|
| 148 |
+
# 主要內容區域
|
| 149 |
+
if uploaded_file is not None:
|
| 150 |
+
# 顯示頁面內容預覽
|
| 151 |
+
st.header("頁面內容預覽")
|
| 152 |
+
|
| 153 |
+
# 從session_state獲取選定頁面的內容
|
| 154 |
+
if page_to_process in st.session_state.page_content:
|
| 155 |
+
page_text = st.session_state.page_content[page_to_process]
|
| 156 |
+
st.text_area(
|
| 157 |
+
f"第 {page_to_process} 頁內容",
|
| 158 |
+
value=page_text,
|
| 159 |
+
height=150,
|
| 160 |
+
disabled=True
|
| 161 |
+
)
|
| 162 |
+
else:
|
| 163 |
+
st.warning("無法獲取選定頁面的內容")
|
| 164 |
+
|
| 165 |
+
# 處理按鈕
|
| 166 |
+
process_button = st.button("處理並翻譯", type="primary", use_container_width=True)
|
| 167 |
+
|
| 168 |
+
# 當按下處理按鈕
|
| 169 |
+
if process_button:
|
| 170 |
+
if not api_key:
|
| 171 |
+
st.error("請輸入Gemini API金鑰!")
|
| 172 |
+
else:
|
| 173 |
+
# 設置進度顯示
|
| 174 |
+
progress_placeholder = st.empty()
|
| 175 |
+
results_placeholder = st.empty()
|
| 176 |
+
|
| 177 |
+
with st.spinner("正在處理中..."):
|
| 178 |
+
progress_bar = progress_placeholder.progress(0)
|
| 179 |
+
|
| 180 |
+
# 設置API
|
| 181 |
+
model = setup_gemini_api(api_key)
|
| 182 |
+
if model:
|
| 183 |
+
progress_bar.progress(20)
|
| 184 |
+
|
| 185 |
+
# 獲取選定頁面的內容
|
| 186 |
+
page_text = st.session_state.page_content[page_to_process]
|
| 187 |
+
|
| 188 |
+
# 使用Gemini處理
|
| 189 |
+
progress_placeholder.text("正在使用Gemini解釋內容...")
|
| 190 |
+
explanation = process_with_gemini(model, page_text, instruction)
|
| 191 |
+
progress_bar.progress(60)
|
| 192 |
+
|
| 193 |
+
# 翻譯成繁體中文
|
| 194 |
+
progress_placeholder.text("正在翻譯成繁體中文...")
|
| 195 |
+
translation = translate_with_gemini(model, explanation)
|
| 196 |
+
progress_bar.progress(90)
|
| 197 |
+
|
| 198 |
+
# 創建結果DataFrame
|
| 199 |
+
results_data = {
|
| 200 |
+
"時間戳記": [datetime.now().isoformat()],
|
| 201 |
+
"原始內容": [page_text[:5000] + "..." if len(page_text) > 5000 else page_text],
|
| 202 |
+
"Gemini解釋": [explanation],
|
| 203 |
+
"繁體中文翻譯": [translation]
|
| 204 |
+
}
|
| 205 |
+
results_df = pd.DataFrame(results_data)
|
| 206 |
+
|
| 207 |
+
# 保存為CSV(臨時)
|
| 208 |
+
results_df.to_csv(output_filename, index=False, encoding="utf-8-sig")
|
| 209 |
+
|
| 210 |
+
# 完成
|
| 211 |
+
progress_bar.progress(100)
|
| 212 |
+
progress_placeholder.empty()
|
| 213 |
+
|
| 214 |
+
# 顯示結果
|
| 215 |
+
st.success("處理完成!")
|
| 216 |
+
|
| 217 |
+
# 創建選項卡顯示結果
|
| 218 |
+
tab1, tab2, tab3 = st.tabs(["Gemini解釋", "繁體中文翻譯", "CSV資料"])
|
| 219 |
+
|
| 220 |
+
with tab1:
|
| 221 |
+
st.subheader("Gemini解釋結果")
|
| 222 |
+
st.write(explanation)
|
| 223 |
+
|
| 224 |
+
with tab2:
|
| 225 |
+
st.subheader("繁體中文翻譯")
|
| 226 |
+
st.write(translation)
|
| 227 |
+
|
| 228 |
+
with tab3:
|
| 229 |
+
st.subheader("CSV資料預覽")
|
| 230 |
+
st.dataframe(results_df)
|
| 231 |
+
st.markdown(get_csv_download_link(results_df, output_filename), unsafe_allow_html=True)
|
| 232 |
+
st.info(f"CSV檔案已準備好下載。檔名: {output_filename}")
|
| 233 |
+
else:
|
| 234 |
+
# 未上傳檔案時顯示的內容
|
| 235 |
+
st.info("👈 請從側邊欄上傳PDF檔案開始")
|
| 236 |
+
|
| 237 |
+
# 顯示使用說明
|
| 238 |
+
with st.expander("使用說明", expanded=True):
|
| 239 |
+
st.markdown("""
|
| 240 |
+
### 如何使用這個工具:
|
| 241 |
+
|
| 242 |
+
1. **上傳PDF檔案** - 從側邊欄選擇並上傳PDF檔案
|
| 243 |
+
2. **選擇頁面** - 使用滑桿選擇要處理的頁面
|
| 244 |
+
3. **設定API金鑰** - 輸入您的Gemini API金鑰(預設已填入,可修改)
|
| 245 |
+
4. **自訂指示詞** - 可選擇修改給Gemini的指示詞
|
| 246 |
+
5. **處理與翻譯** - 點擊"處理並翻譯"按鈕
|
| 247 |
+
6. **查看結果** - 在選項卡中查看Gemini的解釋和繁體中文翻譯
|
| 248 |
+
7. **下載結果** - 下載CSV格式的結果檔案
|
| 249 |
+
|
| 250 |
+
### 功能特點:
|
| 251 |
+
|
| 252 |
+
- 逐頁預覽PDF內容
|
| 253 |
+
- 使用Gemini AI解釋文本
|
| 254 |
+
- 自動翻譯成繁體中文
|
| 255 |
+
- 結果以CSV格式儲存
|
| 256 |
+
""")
|
| 257 |
|
| 258 |
+
# 頁尾
|
| 259 |
+
st.markdown("---")
|
| 260 |
+
st.markdown("📄 PDF處理與Gemini翻譯工具 | 由Streamlit和Google Gemini AI提供技術支持")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|