lib/media.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

import subprocess
from pathlib import Path
from huggingface_hub import hf_hub_url as hf_link

from .config import (
    AUDIO_DIR,
    TRANSCRIPTS_DIR,
    AUDIO_EXT,
    TEXT_EXT,
)

# -->> Tunables <<---------------------


# -->> Definitions <<------------------


# -->> API <<--------------------------


def extract_audio_from_video(video_file, audio_folder_path):
    """Converts video to audio directly using `ffmpeg` command
    with the help of subprocess module"""
    # Create the audio folder if it doesn't exist
    Path(audio_folder_path).mkdir(parents=True, exist_ok=True)

    # Extract the filename from the video_file path
    video_filename = Path(video_file).name
    video_name = Path(video_filename).stem
    output_audio_path = Path(audio_folder_path) / (video_name + AUDIO_EXT)
    output_audio_pathstr = str(output_audio_path)

    # Check if the audio file already exists and remove it
    if output_audio_path.exists():
        output_audio_path.unlink()

    # Command to extract audio using ffmpeg
    command = [
        "ffmpeg",
        "-i",
        video_file,
        "-ab",
        "160k",
        "-ac",
        "2",
        "-ar",
        "44100",
        "-vn",
        output_audio_pathstr,
    ]

    # Run the command
    print("Extracting audio...")
    subprocess.run(command, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    print("Audio extracted!")

    # If save_audio is True, provide a link to save the file
    return output_audio_pathstr, output_audio_path


def save_file(audio_path, transcript_folder_path, text):
    """Saves the text to a file"""

    # Create the transcript folder if it doesn't exist
    transcript_folder_path = Path(transcript_folder_path)
    transcript_folder_path.mkdir(parents=True, exist_ok=True)

    # Extract the filename from the audio_file path
    audio_filename = Path(audio_path).stem
    output_transcript_path = transcript_folder_path / (audio_filename + TEXT_EXT)
    output_transcript_pathstr = str(output_transcript_path)

    # Check if the transcript file already exists and remove it
    if output_transcript_path.exists():
        output_transcript_path.unlink(missing_ok=True)

    # Save the transcript to a file
    with open(output_transcript_pathstr, "w") as f:
        f.write(text)

    # If save_transcript is True, provide a link to save the file
    link = hf_link(
        repo_id="Madhuslista/whisper_wrapper", 
        filename=output_transcript_pathstr,
        subfolder=str(output_transcript_path.parent)
    )
    

    return output_transcript_pathstr, link

# -->> Execute <<----------------------


# -->> Export <<-----------------------