Update app.py

app.py

@@ -3,14 +3,21 @@ from datetime import datetime
 import random
 from transformers import pipeline
 from transformers.pipelines.audio_utils import ffmpeg_read
-import moviepy.editor as mp
+from moviepy import (
+    ImageClip, 
+    VideoFileClip,
+    TextClip, 
+    CompositeVideoClip, 
+    AudioFileClip,
+    concatenate_videoclips
+)
 import speech_recognition as sr
 import json
 from nltk.tokenize import sent_tokenize
 
 def transcribe_video(video_path):
     # Load the video file and extract audio
-    video = mp.VideoFileClip(video_path)
+    video = VideoFileClip(video_path)
     audio_path = "audio.wav"
     video.audio.write_audiofile(audio_path)
 
@@ -79,14 +86,14 @@ def translate_text(timestamps_json, target_language):
 
 def add_transcript_to_video(video_path, timestamps, output_path):
     # Load the video file
-    video = mp.VideoFileClip(video_path)
+    video = VideoFileClip(video_path)
 
     # Create text clips based on timestamps
     text_clips = []
     
     for entry in timestamps:
         # Create a text clip for each sentence
-        txt_clip = mp.TextClip(entry["text"], fontsize=24, color='white', bg_color='black', size=video.size)
+        txt_clip = TextClip(entry["text"], fontsize=24, color='white', bg_color='black', size=video.size)
         
         # Set the start time and duration for each text clip
         txt_clip = txt_clip.set_start(entry["start"]).set_duration(3).set_position(('bottom')).set_opacity(0.7)  # Display each sentence for 3 seconds
@@ -95,7 +102,7 @@ def add_transcript_to_video(video_path, timestamps, output_path):
         text_clips.append(txt_clip)
 
     # Overlay all text clips on the original video
-    final_video = mp.CompositeVideoClip([video] + text_clips)
+    final_video = CompositeVideoClip([video] + text_clips)
 
     # Write the result to a file
     final_video.write_videofile(output_path, codec='libx264', audio_codec='aac')