Update app.py

app.py

@@ -15,21 +15,28 @@ import numpy as np
 transcription_pipeline = pipeline("automatic-speech-recognition", model="facebook/wav2vec2-base-960h")
 
 def download_audio_from_url(url):
-    if "share" in url:
-        # Extract the direct MP4 URL from the shareable link
-        response = requests.get(url)
-        soup = BeautifulSoup(response.content, 'html.parser')
-        video_tag = soup.find('video')
-        if video_tag and 'src' in video_tag.attrs:
-            video_url = video_tag['src']
+    try:
+        if "share" in url:
+            print("Processing shareable link...")
+            response = requests.get(url)
+            soup = BeautifulSoup(response.content, 'html.parser')
+            video_tag = soup.find('video')
+            if video_tag and 'src' in video_tag.attrs:
+                video_url = video_tag['src']
+                print(f"Extracted video URL: {video_url}")
+            else:
+                raise ValueError("Direct video URL not found in the shareable link.")
         else:
-            raise ValueError("Direct video URL not found in the shareable link.")
-    else:
-        video_url = url
-    
-    response = requests.get(video_url)
-    audio_bytes = response.content
-    return audio_bytes
+            video_url = url
+        
+        print(f"Downloading video from URL: {video_url}")
+        response = requests.get(video_url)
+        audio_bytes = response.content
+        print(f"Successfully downloaded {len(audio_bytes)} bytes of data")
+        return audio_bytes
+    except Exception as e:
+        print(f"Error in download_audio_from_url: {str(e)}")
+        raise
 
 def transcribe_audio(audio_bytes):
     audio = AudioSegment.from_file(io.BytesIO(audio_bytes))
@@ -77,10 +84,21 @@ def transcribe_audio(audio_bytes):
     return formatted_transcript
 
 def transcribe_video(url):
-    audio_bytes = download_audio_from_url(url)
-    transcript = transcribe_audio(audio_bytes)
-    return transcript
-
+    try:
+        print(f"Attempting to download audio from URL: {url}")
+        audio_bytes = download_audio_from_url(url)
+        print(f"Successfully downloaded {len(audio_bytes)} bytes of audio data")
+        
+        print("Starting audio transcription...")
+        transcript = transcribe_audio(audio_bytes)
+        print("Transcription completed successfully")
+        
+        return transcript
+    except Exception as e:
+        error_message = f"An error occurred: {str(e)}"
+        print(error_message)
+        return error_message
+        
 def download_transcript(transcript):
     return transcript, "transcript.txt"