Spaces:
Sleeping
Sleeping
TRASCRIPTS = []
Browse files
app.py
CHANGED
|
@@ -49,13 +49,11 @@ from urllib.parse import urlparse, parse_qs
|
|
| 49 |
|
| 50 |
|
| 51 |
OUTPUT_PATH = 'videos'
|
| 52 |
-
|
| 53 |
|
| 54 |
OPEN_AI_KEY = os.getenv("OPEN_AI_KEY")
|
| 55 |
client = OpenAI(api_key=OPEN_AI_KEY)
|
| 56 |
|
| 57 |
-
JS = ''
|
| 58 |
-
CSS = ''
|
| 59 |
|
| 60 |
# # ====drive====初始化Google Drive服务
|
| 61 |
def init_drive_service():
|
|
@@ -311,92 +309,28 @@ def process_youtube_link(link):
|
|
| 311 |
df_string_output = json.dumps(transcript, ensure_ascii=False, indent=2)
|
| 312 |
df_summarise = generate_df_summarise(transcript)
|
| 313 |
|
|
|
|
|
|
|
|
|
|
| 314 |
# 确保返回与 UI 组件预期匹配的输出
|
| 315 |
return questions[0] if len(questions) > 0 else "", \
|
| 316 |
questions[1] if len(questions) > 1 else "", \
|
| 317 |
questions[2] if len(questions) > 2 else "", \
|
| 318 |
df_string_output, \
|
| 319 |
df_summarise, \
|
| 320 |
-
html_content
|
|
|
|
| 321 |
|
| 322 |
def format_transcript_to_html(formatted_transcript):
|
| 323 |
-
|
| 324 |
-
|
| 325 |
-
|
| 326 |
-
|
| 327 |
-
|
| 328 |
-
# return html_content
|
| 329 |
-
|
| 330 |
-
CSS = """
|
| 331 |
-
#contentDisplay {{
|
| 332 |
-
width: 500px;
|
| 333 |
-
margin: auto;
|
| 334 |
-
text-align: center;
|
| 335 |
-
}}
|
| 336 |
-
#imageDisplay {{
|
| 337 |
-
width: 500px;
|
| 338 |
-
height: 300px;
|
| 339 |
-
background-color: #f0f0f0;
|
| 340 |
-
background-size: contain;
|
| 341 |
-
background-position: center center;
|
| 342 |
-
background-repeat: no-repeat;
|
| 343 |
-
margin-bottom: 20px;
|
| 344 |
-
}}
|
| 345 |
-
"""
|
| 346 |
-
|
| 347 |
-
# 将逐字稿数据转换为 JavaScript 对象数组
|
| 348 |
-
entries_js_array = ',\n '.join([
|
| 349 |
-
json.dumps({
|
| 350 |
-
"start_time": entry['start_time'],
|
| 351 |
-
"end_time": entry['end_time'],
|
| 352 |
-
"text": entry['text'],
|
| 353 |
-
"screenshot_path": entry['screenshot_path']
|
| 354 |
-
}, ensure_ascii=False) for entry in formatted_transcript
|
| 355 |
-
])
|
| 356 |
-
|
| 357 |
-
JS = f"""
|
| 358 |
-
var entries = [
|
| 359 |
-
{entries_js_array}
|
| 360 |
-
];
|
| 361 |
-
var currentIndex = 0;
|
| 362 |
-
|
| 363 |
-
function updateContentDisplay(index) {{
|
| 364 |
-
var entry = entries[index];
|
| 365 |
-
document.getElementById('imageDisplay').style.backgroundImage = 'url(' + entry.screenshot_path + ')';
|
| 366 |
-
document.getElementById('timeDisplay').innerText = entry.start_time + " - " + entry.end_time;
|
| 367 |
-
document.getElementById('textDisplay').innerText = entry.text;
|
| 368 |
-
}}
|
| 369 |
-
|
| 370 |
-
updateContentDisplay(currentIndex);
|
| 371 |
-
|
| 372 |
-
document.getElementById('prevButton').addEventListener('click', function() {{
|
| 373 |
-
if (currentIndex > 0) {{
|
| 374 |
-
currentIndex -= 1;
|
| 375 |
-
updateContentDisplay(currentIndex);
|
| 376 |
-
}}
|
| 377 |
-
}});
|
| 378 |
-
|
| 379 |
-
document.getElementById('nextButton').addEventListener('click', function() {{
|
| 380 |
-
if (currentIndex < entries.length - 1) {{
|
| 381 |
-
currentIndex += 1;
|
| 382 |
-
updateContentDisplay(currentIndex);
|
| 383 |
-
}}
|
| 384 |
-
}});
|
| 385 |
-
"""
|
| 386 |
-
|
| 387 |
-
html_content = """
|
| 388 |
-
<div id="contentDisplay">
|
| 389 |
-
<div id="imageDisplay"></div>
|
| 390 |
-
<p id="timeDisplay"></p>
|
| 391 |
-
<p id="textDisplay"></p>
|
| 392 |
-
<button id="prevButton">上一個</button>
|
| 393 |
-
<button id="nextButton">下一個</button>
|
| 394 |
-
</div>
|
| 395 |
-
"""
|
| 396 |
-
|
| 397 |
return html_content
|
| 398 |
|
| 399 |
|
|
|
|
| 400 |
def get_embedded_youtube_link(video_id, start_time):
|
| 401 |
embed_url = f"https://www.youtube.com/embed/{video_id}?start={start_time}&autoplay=1"
|
| 402 |
return embed_url
|
|
@@ -544,7 +478,21 @@ def respond(user_message, df_string_output, chat_history):
|
|
| 544 |
# 返回聊天历史和空字符串清空输入框
|
| 545 |
return "", chat_history
|
| 546 |
|
| 547 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 548 |
with gr.Row():
|
| 549 |
with gr.Column():
|
| 550 |
file_upload = gr.File(label="Upload your CSV or Word file")
|
|
@@ -555,6 +503,18 @@ with gr.Blocks(css=CSS, js=JS) as demo:
|
|
| 555 |
send_button = gr.Button("Send")
|
| 556 |
|
| 557 |
with gr.Column():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 558 |
with gr.Tab("YouTube Transcript and Video"):
|
| 559 |
transcript_html = gr.HTML(label="YouTube Transcript and Video")
|
| 560 |
with gr.Tab("資料本文"):
|
|
|
|
| 49 |
|
| 50 |
|
| 51 |
OUTPUT_PATH = 'videos'
|
| 52 |
+
TRASCRIPTS = []
|
| 53 |
|
| 54 |
OPEN_AI_KEY = os.getenv("OPEN_AI_KEY")
|
| 55 |
client = OpenAI(api_key=OPEN_AI_KEY)
|
| 56 |
|
|
|
|
|
|
|
| 57 |
|
| 58 |
# # ====drive====初始化Google Drive服务
|
| 59 |
def init_drive_service():
|
|
|
|
| 309 |
df_string_output = json.dumps(transcript, ensure_ascii=False, indent=2)
|
| 310 |
df_summarise = generate_df_summarise(transcript)
|
| 311 |
|
| 312 |
+
global TRASCRIPTS
|
| 313 |
+
TRASCRIPTS = formatted_transcript
|
| 314 |
+
|
| 315 |
# 确保返回与 UI 组件预期匹配的输出
|
| 316 |
return questions[0] if len(questions) > 0 else "", \
|
| 317 |
questions[1] if len(questions) > 1 else "", \
|
| 318 |
questions[2] if len(questions) > 2 else "", \
|
| 319 |
df_string_output, \
|
| 320 |
df_summarise, \
|
| 321 |
+
html_content
|
| 322 |
+
|
| 323 |
|
| 324 |
def format_transcript_to_html(formatted_transcript):
|
| 325 |
+
html_content = ""
|
| 326 |
+
for entry in formatted_transcript:
|
| 327 |
+
html_content += f"<h3>{entry['start_time']} - {entry['end_time']}</h3>"
|
| 328 |
+
html_content += f"<p>{entry['text']}</p>"
|
| 329 |
+
html_content += f"<img src='{entry['screenshot_path']}' width='500px' />"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 330 |
return html_content
|
| 331 |
|
| 332 |
|
| 333 |
+
|
| 334 |
def get_embedded_youtube_link(video_id, start_time):
|
| 335 |
embed_url = f"https://www.youtube.com/embed/{video_id}?start={start_time}&autoplay=1"
|
| 336 |
return embed_url
|
|
|
|
| 478 |
# 返回聊天历史和空字符串清空输入框
|
| 479 |
return "", chat_history
|
| 480 |
|
| 481 |
+
def update_slide(direction):
|
| 482 |
+
current_index[0] += direction
|
| 483 |
+
if current_index[0] < 0:
|
| 484 |
+
current_index[0] = 0 # 防止索引小于0
|
| 485 |
+
elif current_index[0] >= len(TRASCRIPTS):
|
| 486 |
+
current_index[0] = len(TRASCRIPTS) - 1 # 防止索引超出范围
|
| 487 |
+
|
| 488 |
+
# 获取当前条目的文本和截图 URL
|
| 489 |
+
current_transcript = TRASCRIPTS[current_index[0]]
|
| 490 |
+
return current_transcript["screenshot_url"], current_transcript["text"]
|
| 491 |
+
|
| 492 |
+
|
| 493 |
+
current_index = [0]
|
| 494 |
+
|
| 495 |
+
with gr.Blocks() as demo:
|
| 496 |
with gr.Row():
|
| 497 |
with gr.Column():
|
| 498 |
file_upload = gr.File(label="Upload your CSV or Word file")
|
|
|
|
| 503 |
send_button = gr.Button("Send")
|
| 504 |
|
| 505 |
with gr.Column():
|
| 506 |
+
with gr.Tab("投影片"):
|
| 507 |
+
|
| 508 |
+
image = gr.Image()
|
| 509 |
+
text = gr.Textbox()
|
| 510 |
+
with gr.Row():
|
| 511 |
+
prev_button = gr.Button("上一个")
|
| 512 |
+
next_button = gr.Button("下一个")
|
| 513 |
+
|
| 514 |
+
# 设置按钮的动作
|
| 515 |
+
prev_button.click(fn=update_slide, inputs=-1, outputs=[image, text])
|
| 516 |
+
next_button.click(fn=update_slide, inputs=1, outputs=[image, text])
|
| 517 |
+
update_slide(0)
|
| 518 |
with gr.Tab("YouTube Transcript and Video"):
|
| 519 |
transcript_html = gr.HTML(label="YouTube Transcript and Video")
|
| 520 |
with gr.Tab("資料本文"):
|