TubeScript-API/test_transcript.py

#!/usr/bin/env python3
"""
Script de prueba rápida para verificar la función de transcripción
sin necesidad de Docker
"""
import sys
import os
import json
import subprocess
import requests

def test_transcript(video_id, lang='es'):
    """Probar la función de transcripción"""
    url = f"https://www.youtube.com/watch?v={video_id}"

    # Comando simplificado
    command = [
        "yt-dlp",
        "--skip-download",
        "--dump-json",
        "--no-warnings",
        "--extractor-args", "youtube:player_client=android",
        url
    ]

    print(f"🔍 Probando video: {video_id}")
    print(f"📝 Comando: {' '.join(command)}")
    print()

    try:
        result = subprocess.run(command, capture_output=True, text=True, timeout=60)

        if result.returncode != 0:
            print(f"❌ Error: {result.stderr[:500]}")
            return False

        if not result.stdout.strip():
            print("❌ No se obtuvieron datos")
            return False

        metadata = json.loads(result.stdout)

        # Buscar subtítulos
        print("🔍 Buscando subtítulos...")
        print()

        # Verificar requested_subtitles
        requested_subs = metadata.get('requested_subtitles', {})
        if requested_subs:
            print(f"✅ Encontrado en requested_subtitles:")
            for key, val in requested_subs.items():
                print(f"   - {key}: {val.get('ext', 'unknown')}")
        else:
            print("⚪ No hay requested_subtitles")

        # Verificar automatic_captions
        auto_captions = metadata.get('automatic_captions', {})
        if auto_captions:
            print(f"\n✅ Automatic captions disponibles:")
            for lang_key, formats in auto_captions.items():
                if lang in lang_key or lang_key.startswith(lang):
                    print(f"   - {lang_key}:")
                    for fmt in formats[:3]:  # Primeros 3 formatos
                        print(f"      • {fmt.get('ext', 'unknown')}: {fmt.get('url', 'N/A')[:80]}...")
        else:
            print("\n⚪ No hay automatic_captions")

        # Verificar subtitles
        subtitles = metadata.get('subtitles', {})
        if subtitles:
            print(f"\n✅ Subtitles manuales disponibles:")
            for lang_key, formats in subtitles.items():
                if lang in lang_key or lang_key.startswith(lang):
                    print(f"   - {lang_key}:")
                    for fmt in formats[:3]:
                        print(f"      • {fmt.get('ext', 'unknown')}: {fmt.get('url', 'N/A')[:80]}...")
        else:
            print("\n⚪ No hay subtitles manuales")

        # Probar obtener URL
        print("\n" + "="*60)
        print("🎯 Intentando obtener URL de subtítulos...")

        found_subs = requested_subs

        if not found_subs and auto_captions:
            for lang_key in auto_captions.keys():
                if lang in lang_key or lang_key.startswith(lang):
                    if auto_captions[lang_key]:
                        found_subs = {lang_key: auto_captions[lang_key][0]}
                        print(f"✅ Usando automatic_captions[{lang_key}]")
                        break

        if not found_subs and subtitles:
            for lang_key in subtitles.keys():
                if lang in lang_key or lang_key.startswith(lang):
                    if subtitles[lang_key]:
                        found_subs = {lang_key: subtitles[lang_key][0]}
                        print(f"✅ Usando subtitles[{lang_key}]")
                        break

        if not found_subs:
            print("❌ No se encontraron subtítulos para el idioma especificado")
            print(f"\n💡 Idiomas disponibles:")
            all_langs = set(list(auto_captions.keys()) + list(subtitles.keys()))
            for l in sorted(all_langs):
                print(f"   - {l}")
            return False

        # Intentar descargar
        lang_key = next(iter(found_subs))
        sub_url = found_subs[lang_key].get('url')
        sub_ext = found_subs[lang_key].get('ext', 'unknown')

        if not sub_url:
            print("❌ No se pudo obtener URL de subtítulos")
            return False

        print(f"✅ URL encontrada: {sub_url[:100]}...")
        print(f"📝 Formato: {sub_ext}")

        # Intentar descargar
        print("\n⬇️  Descargando subtítulos...")
        response = requests.get(sub_url, timeout=30)

        if response.status_code == 200:
            print(f"✅ Descarga exitosa ({len(response.content)} bytes)")

            # Mostrar muestra del contenido
            sample = response.text[:500] if hasattr(response, 'text') else str(response.content[:500])
            print(f"\n📄 Muestra del contenido:")
            print("─" * 60)
            print(sample)
            print("─" * 60)

            return True
        else:
            print(f"❌ Error HTTP {response.status_code}")
            return False

    except subprocess.TimeoutExpired:
        print("❌ Timeout al ejecutar yt-dlp")
        return False
    except Exception as e:
        print(f"❌ Error: {str(e)}")
        import traceback
        traceback.print_exc()
        return False

if __name__ == "__main__":
    video_id = sys.argv[1] if len(sys.argv) > 1 else "6hini9Xz_fc"
    lang = sys.argv[2] if len(sys.argv) > 2 else "es"

    print("="*60)
    print("🧪 TubeScript - Test de Transcripción")
    print("="*60)
    print()

    success = test_transcript(video_id, lang)

    print()
    print("="*60)
    if success:
        print("✅ TEST EXITOSO")
    else:
        print("❌ TEST FALLIDO")
    print("="*60)

    sys.exit(0 if success else 1)