examples/download_real_samples.py

#!/usr/bin/env python3
"""
download_real_samples.py - Baixa arquivos de áudio reais (domínio público) para exemplos.
"""
import subprocess
from pathlib import Path
from urllib.request import urlretrieve

BASE_DIR = Path(__file__).parent
INPUTS_DIR = BASE_DIR / "sample_inputs"
OUTPUTS_DIR = BASE_DIR / "sample_outputs"

URLS = {
    "real_sample_1.wav": "https://www.voiptroubleshooter.com/open_speech/american/OSR_us_000_0010_8k.wav",
    "real_sample_2.wav": "https://www.voiptroubleshooter.com/open_speech/american/OSR_us_000_0011_8k.wav",
    "real_sample_3.ogg": "https://upload.wikimedia.org/wikipedia/commons/9/99/Short_melody_from_%22Pa%27s_Fiddle%22.ogg",
}

def download_file(url, dest_path):
    if dest_path.exists():
        print(f"⏭️  Already exists: {dest_path.name}")
        return True
    print(f"⬇️  Downloading {dest_path.name} ...")
    try:
        urlretrieve(url, dest_path)
        print(f"   ✅ Done: {dest_path.name}")
        return True
    except Exception as e:
        print(f"   ❌ Failed: {e}")
        return False

def main():
    INPUTS_DIR.mkdir(parents=True, exist_ok=True)
    for name, url in URLS.items():
        dest = INPUTS_DIR / name
        download_file(url, dest)
        if name.endswith(".ogg"):
            mp3_path = INPUTS_DIR / name.replace(".ogg", ".mp3")
            if not mp3_path.exists():
                subprocess.run([
                    "ffmpeg", "-y", "-i", str(dest),
                    "-acodec", "libmp3lame", "-b:a", "192k", str(mp3_path)
                ], check=True, capture_output=True)
                dest.unlink()
                print(f"   ✅ Converted to MP3: {mp3_path.name}")
    print("✅ Real samples downloaded.")

if __name__ == "__main__":
    main()