Upload 2 files

app.py

@@ -0,0 +1,234 @@
+# -*- coding: utf-8 -*-
+# 財政部財政資訊中心 江信宗
+
+import gradio as gr
+from googleapiclient.discovery import build
+import os
+import re
+import requests
+import io
+import json
+
+custom_css = """
+.center-aligned {
+    text-align: center !important;
+    color: #ff4081;
+    text-shadow: 2px 2px 4px rgba(0,0,0,0.1);
+    margin-bottom: 0 !important;
+}
+.input-background {
+    background-color: #B7E0FF !important;
+    padding: 15px !important;
+    border-radius: 10px !important;
+}
+.tabs-background {
+    background-color: #FFF5E4 !important;
+    padding: 0px !important;
+    border-radius: 10px !important;
+    margin: 0 !important;
+}
+.output-background {
+    background-color: #FFF4B5 !important;
+    padding: 15px !important;
+    border-radius: 10px !important;
+}
+.output-text-size textarea {
+    font-size: 18px !important;
+}
+.gen-button {
+    border-radius: 10px !important;
+    background-color: #ff4081 !important;
+    color: white !important;
+    font-weight: bold !important;
+    transition: all 0.3s ease !important;
+}
+.gen-button:hover {
+    background-color: #f50057 !important;
+    transform: scale(1.05);
+}
+.clear-button {
+    border-radius: 10px !important;
+    background-color: #333333 !important;
+    color: white !important;
+    font-weight: bold !important;
+    transition: all 0.3s ease !important;
+}
+.clear-button:hover {
+    background-color: #000000 !important;
+    transform: scale(1.05);
+}
+"""
+
+def check_description(video_id, api_key):
+    sponsored_keywords = ["折扣碼", "贊助廠商", "優惠碼", "專屬優惠", "預約表單", "賣場連結", "合作夥伴", "合作廠商"]
+    try:
+        youtube = build('youtube', 'v3', developerKey=api_key)
+        request = youtube.videos().list(
+            part="snippet",
+            id=video_id
+        )
+        response = request.execute()
+        if 'items' in response and len(response['items']) > 0:
+            description = response['items'][0]['snippet']['description'].lower().strip()
+            for keyword in sponsored_keywords:
+                if keyword.lower() in description:
+                    return True
+        return False
+    except Exception as e:
+        print(f"獲取 YouTube 說明欄資訊時發生錯誤：{str(e)}")
+        return False
+
+def extract_video_id(url):
+    match = re.search(r'(?:youtube\.com\/watch\?v=|youtu\.be\/)([^&\n?]+)', url)
+    if match:
+        return match.group(1)
+    return None
+
+def seconds_to_minutes_seconds(seconds):
+    minutes = int(seconds // 60)
+    remaining_seconds = int(seconds % 60)
+    return f"{minutes}分{remaining_seconds}秒"
+
+def format_segments(segments):
+    formatted_segments = []
+    for i, segment in enumerate(segments, 1):
+        start_time = seconds_to_minutes_seconds(segment[0])
+        end_time = seconds_to_minutes_seconds(segment[1])
+        formatted_segments.append(f"【廣告{i}時間軸】{start_time}至{end_time}")
+    return "\n".join(formatted_segments)
+
+def process_input(youtube_url, txt_file, api_key):
+    if not api_key:
+        api_key = os.getenv("YOUR_API_TOKEN")
+    if youtube_url:
+        video_id = extract_video_id(youtube_url)
+        if video_id:
+            api_url = f"https://sponsor.ajay.app/api/skipSegments?videoID={video_id}"
+            try:
+                response = requests.get(api_url)
+                if response.status_code == 404:
+                    # 使用 YouTube Data API 檢查說明欄
+                    if check_description(video_id, api_key):
+                        return "查核結果：影片內嵌贊助廣告，再請人工複查贊助事實。"
+                    else:
+                        return "查核結果：未找到贊助廣告片段。請使用地端影片分析程式查驗。"
+                response.raise_for_status()
+                data = response.json()
+                
+                if data:
+                    segments = [item['segment'] for item in data]
+                    formatted_output = format_segments(segments)
+                    return f"{formatted_output}\n查核結果：影片內嵌贊助廣告，再請人工複查贊助事實。"
+                else:
+                    # 使用 YouTube Data API 檢查說明欄
+                    if check_description(video_id):
+                        return "查核結果：影片內嵌贊助廣告，再請人工複查贊助事實。"
+                    else:
+                        return "查核結果：未找到贊助廣告片段。請人工複查是否有贊助內容。"
+            except requests.RequestException as e:
+                return f"請求發生錯誤：{str(e)}。請稍後再試或人工複查。"
+        else:
+            return "無效的 YouTube URL。請確保輸入正確的 YouTube 影片網址。"
+    elif txt_file is not None:
+        try:
+            if isinstance(txt_file, str):
+                with open(txt_file, 'r', encoding='utf-8') as file:
+                    content = file.read()
+            elif hasattr(txt_file, 'name'):
+                with open(txt_file.name, 'r', encoding='utf-8') as file:
+                    content = file.read()
+            else:
+                return "無法讀取上傳的文件。請確保上傳了有效的文本文件。"
+            gr.Info("檔案內容每列僅一個網址，以換行分隔下一個網址。")
+            urls = content.strip().split('\n')
+            results = []
+            for url in urls:
+                url = url.strip()
+                result = f"《URL》{url}\n"
+                video_id = extract_video_id(url)
+                if video_id:
+                    api_url = f"https://sponsor.ajay.app/api/skipSegments?videoID={video_id}"
+                    try:
+                        response = requests.get(api_url)
+                        if response.status_code == 404:
+                            result += "查核結果：未找到贊助廣告片段。請使用地端影片分析程式查驗。"
+                        else:
+                            response.raise_for_status()
+                            data = response.json()
+                            if data:
+                                segments = [item['segment'] for item in data]
+                                formatted_output = format_segments(segments)
+                                result += f"{formatted_output}\n查核結果：影片內嵌贊助廣告，再請人工複查贊助事實。"
+                            else:
+                                result += "查核結果：未找到贊助廣告片段。請人工複查是否有贊助內容。"
+                    except requests.RequestException as e:
+                        result += f"請求發生錯誤：{str(e)}。請稍後再試或人工複查。"
+                else:
+                    result += "無效的 YouTube URL。請確保輸入正確的 YouTube 影片網址。"
+                results.append(result)
+            return "\n\n".join(results)
+        except Exception as e:
+            return f"處理文件時發生錯誤：{str(e)}。請確保上傳了有效的文本文件。"
+    else:
+        gr.Warning("請輸入 YouTube 網址或上傳純文字檔！！")
+        return "請輸入 YouTube 網址或上傳純文字檔！！"
+
+with gr.Blocks(theme=gr.themes.Monochrome(), css=custom_css) as iface:
+    gr.Markdown("""
+    # 💰 YouTuber 營銷贊助稽查 - 財政部財政資訊中心 🏷️
+    > ### **※ 建構租稅公平，開拓更多可能性，自動化稽查 YouTuber 業配及銷售收益情形，系統布署：江信宗，適用稅目：綜合所得稅與營業稅。**
+    """, elem_classes="center-aligned")
+    
+    def clear_youtube_input():
+        return gr.update(value="")
+
+    with gr.Tabs(elem_classes="tabs-background") as tabs:
+        with gr.TabItem("YouTube 網址"):
+            youtube_input = gr.Textbox(label="請輸入 YouTube 網址：", autofocus=True, max_lines=2, elem_classes="input-background")
+        with gr.TabItem("整批網址稽查"):
+            file_input = gr.File(label="上傳純文字檔：", file_types=[".csv", ".txt"], elem_classes="input-background")
+        with gr.TabItem("API Key"):
+            api_key_input = gr.Textbox(type="password", label="請輸入您的 API Key：", placeholder="YouTube Data API v3 Key", elem_classes="input-background")
+
+    tabs.change(fn=clear_youtube_input, inputs=None, outputs=youtube_input)
+
+    with gr.Row():
+        submit_btn = gr.Button("傳送", variant="primary", scale=2, elem_classes="gen-button")
+        clear_btn = gr.Button("清除", variant="secondary", scale=1, elem_classes="clear-button")
+    output = gr.Textbox(label="稽查結果：", max_lines=20, elem_classes="output-background output-text-size")
+    gr.HTML(
+        """
+        <span style="font-size: 20px; color: black; font-weight:bold;">※ 影片或音訊檔建議使用</span><a href="https://huggingface.co/spaces/Jiangxz/Meeting_Minutes" style="font-size: 20px; color: red; font-weight:bold;">財資凊彩歐北寫</a><span style="font-size: 20px; color: black; font-weight:bold;">轉譯後，透過LLMs分析內容判斷營銷或贊助與否。</span>
+        """)
+    submit_btn.click(fn=process_input, inputs=[youtube_input, file_input, api_key_input], outputs=output)
+    def clear_inputs():
+        gr.Info("已成功清除所有內容。")
+        return "", None, ""
+    clear_btn.click(
+        fn=clear_inputs,
+        inputs=None,
+        outputs=[youtube_input, file_input, output]
+    )
+
+gr.HTML("""
+    <script>
+        document.addEventListener('click', function(e) {
+            if (e.target && e.target.textContent === '清除') {
+                var fileInputs = document.querySelectorAll('input[type="file"]');
+                fileInputs.forEach(function(input) {
+                    input.value = '';
+                });
+                var uploadTexts = document.querySelectorAll('.file-preview');
+                uploadTexts.forEach(function(text) {
+                    text.textContent = '';
+                });
+            }
+        });
+    </script>
+""")
+
+if __name__ == "__main__":
+    if "SPACE_ID" in os.environ:
+        iface.launch()
+    else:
+        iface.launch(share=True, show_api=False)

requirements.txt

@@ -0,0 +1,3 @@
+gradio
+requests
+google-api-python-client