app.py

import io
import gradio as gr
import torch
from transformers import WhisperProcessor, WhisperForConditionalGeneration
import requests
from bs4 import BeautifulSoup
import tempfile
import os
import soundfile as sf
from spellchecker import SpellChecker
from pydub import AudioSegment
import librosa
import numpy as np

# Check if CUDA is available and set the device
device = "cuda" if torch.cuda.is_available() else "cpu"
print(f"Using device: {device}")

# Load the Whisper model and processor
model_name = "openai/whisper-small"
processor = WhisperProcessor.from_pretrained(model_name)
model = WhisperForConditionalGeneration.from_pretrained(model_name).to(device)

spell = SpellChecker()

def download_audio_from_url(url):
    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:
            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 correct_spelling(text):
    words = text.split()
    corrected_words = [spell.correction(word) or word for word in words]
    return ' '.join(corrected_words)

def format_transcript(transcript):
    sentences = transcript.split('.')
    formatted_transcript = []
    current_speaker = None
    for sentence in sentences:
        if ':' in sentence:
            speaker, content = sentence.split(':', 1)
            if speaker != current_speaker:
                formatted_transcript.append(f"\n\n{speaker.strip()}:{content.strip()}.")
                current_speaker = speaker
            else:
                formatted_transcript.append(f"{content.strip()}.")
        else:
            formatted_transcript.append(sentence.strip() + '.')
    return ' '.join(formatted_transcript)

def transcribe_audio(audio_file):
    try:
        # Load the entire audio file
        audio_input, sr = librosa.load(audio_file, sr=16000)
        
        # Convert to float32 numpy array
        audio_input = audio_input.astype(np.float32)

        # Process in chunks of 30 seconds with overlap
        chunk_length = 30 * sr
        overlap = 5 * sr  # 5 seconds overlap
        transcriptions = []

        for i in range(0, len(audio_input), chunk_length - overlap):
            chunk = audio_input[i:i+chunk_length]
            input_features = processor(chunk, sampling_rate=16000, return_tensors="pt").input_features.to(device)
            predicted_ids = model.generate(input_features)
            transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
            transcriptions.extend(transcription)

        # Join all transcriptions
        full_transcription = " ".join(transcriptions)

        print(f"Full transcription length: {len(full_transcription)} characters")
        return full_transcription
    except Exception as e:
        print(f"Error in transcribe_audio: {str(e)}")
        raise

def transcribe_video(url):
    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")
        
        # Convert audio bytes to AudioSegment
        audio = AudioSegment.from_file(io.BytesIO(audio_bytes))
        
        print(f"Audio duration: {len(audio) / 1000} seconds")
        
        # Save as WAV file
        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
            audio.export(temp_audio.name, format="wav")
            temp_audio_path = temp_audio.name

        print("Starting audio transcription...")
        transcript = transcribe_audio(temp_audio_path)
        print(f"Transcription completed. Transcript length: {len(transcript)} characters")
        
        # Clean up the temporary file
        os.unlink(temp_audio_path)

        # Apply spelling correction and formatting
        transcript = correct_spelling(transcript)
        transcript = format_transcript(transcript)

        return transcript
    except Exception as e:
        error_message = f"An error occurred: {str(e)}"
        print(error_message)
        return error_message

def download_transcript(transcript):
    with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt') as temp_file:
        temp_file.write(transcript)
        temp_file_path = temp_file.name
    return temp_file_path

# Create the Gradio interface
with gr.Blocks(title="Video Transcription") as demo:
    gr.Markdown("# Video Transcription")
    video_url = gr.Textbox(label="Video URL")
    transcribe_button = gr.Button("Transcribe")
    transcript_output = gr.Textbox(label="Transcript", lines=20)
    download_button = gr.Button("Download Transcript")
    download_link = gr.File(label="Download Transcript")
    
    transcribe_button.click(fn=transcribe_video, inputs=video_url, outputs=transcript_output)
    download_button.click(fn=download_transcript, inputs=transcript_output, outputs=download_link)

demo.launch()