twitch-highlight-detector/detector_muertes.py

#!/usr/bin/env python3
"""
Detector de MUERTES Y AUTO-CRÍTICA:
Encuentra momentos donde el streamer muere o se critica por jugar mal.
"""
import json
import logging
import re

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)


def detect_death_and_failure_moments(transcripcion_json, min_duration=12, max_duration=35, top=30):
    """
    Detecta momentos de muerte o auto-crítica.
    """
    logger.info("=== Buscando MUERTES Y AUTO-CRÍTICA ===")

    with open(transcripcion_json, 'r', encoding='utf-8') as f:
        data = json.load(f)

    segments = data.get("segments", [])

    # Patrones de muerte
    death_patterns = [
        r'\bme han (matado|kill|limeado|pegado)\b',
        r'\b(me caigo|me muero|estoy muerto|muero|mor[íi])\b',
        r'\bme (matan|mate|kill|destruyen)\b',
        r'\bhaz (kill|muerte|limpieza)\b',
        r'\b(tf|trade)\b',  # trading deaths
        r'\bhe (muerto|matado)\b',
        r'\b(me llevan|me cargan|me comen)\b',
        r'\bfallec[íi]\b',
        r'\bdefunc[ií]on\b',

        # Sonidos de muerte/grito
        r'\burgh?\b',
        r'\baggh?\b',
        r'\bargh?\b',
        r'\b(ah|oh|ugh) (no|puta|mierda|dios)\b',
        r'\bno+[.,!]+\b',
        r'\bjoder\b',
        r'\bputa madre\b',
        r'\bputa\b',
        r'\bmierda\b',

        # Frases de muerte
        r'\bestoy (muerto|perdido|acabado)\b',
        r'\bno puedo\b',
        r'\bimposible\b',
        r'\bme (acaban|terminaron)\b',
    ]

    # Patrones de auto-crítica ("jugué muy mal")
    failure_patterns = [
        r'\b(la )?cagu[ée]\b',
        r'\b(jugu[ée]|he jugado) (mal|p[ée]simamente|horrible)\b',
        r'\b(qu[ée] (mal|p[ée]simo|terrible)|error|fail)\b',
        r'\b(lo hice|la hice) mal\b',
        r'\bputa (mala|terrible|fatal)\b',
        r'\bno (me sali[óo]|funcion[óo]|lo logr[ée])\b',
        r'\b(es)tupidez\b',
        r'\bimbecilidad\b',
        r'\bburrada\b',
        r'\bputada\b',
        r'\b(desastroso|catastr[óo]fico)\b',
        r'\b(qu[ée] pena|verg[üu]enza)\b',
        r'\b(he fallado|fall[ée])\b',
        r'\bperd[íi]\b',
        r'\bno deb[ií] (haber|hacer)\b',
        r'\bcagad[oa]\b',

        # Más patrones de fallo
        r'\bmal (jugu|he|estoy)\b',
        r'\bterrible\b',
        r'\bhorrible\b',
        r'\bfatal\b',
        r'\b(pesimo|p[ée]simo)\b',
        r'\bno (sé|pude|pude)\b',
        r'\b(dif[íi]cil|imposible)\b',
        r'\bperd[íi] (el tiempo|la oportunidad|el flash|la fight)\b',
        r'\berror (m[íi]o|grave)\b',
    ]

    # Analizar cada segmento
    moments = []
    for i, seg in enumerate(segments):
        text = seg["text"].lower()
        start = seg["start"]
        end = seg["end"]

        score = 0
        type_ = None

        # Buscar patrones de muerte
        for pattern in death_patterns:
            if re.search(pattern, text, re.IGNORECASE):
                score += 20
                type_ = "muerte"
                break

        # Buscar patrones de auto-crítica
        for pattern in failure_patterns:
            if re.search(pattern, text, re.IGNORECASE):
                score += 15
                if not type_:
                    type_ = "fallo"
                break

        if score > 0:
            moments.append({
                "start": start,
                "end": end,
                "score": score,
                "text": text.strip(),
                "type": type_
            })

    if not moments:
        logger.warning("No se encontraron momentos de muerte/fallo")
        return []

    # Ordenar por score y timestamp
    moments.sort(key=lambda x: (-x["score"], x["start"]))

    # Agrupar en intervalos sin solapamiento
    intervals = []
    for moment in moments:
        start = int(moment["start"])
        end = int(moment["end"])

        # Extender a duración mínima si es muy corto
        duration = max(min_duration, min(end - start, max_duration))
        end = start + duration

        # Verificar solapamiento con intervalos existentes
        overlaps = False
        for s, e in intervals:
            if not (end < s or start > e):  # Hay solapamiento
                overlaps = True
                break

        if not overlaps:
            intervals.append((start, end))
            if len(intervals) >= top:
                break

    # Ordenar por timestamp final
    intervals.sort()

    logger.info(f"Momentos detectados: {len(intervals)}")

    return intervals, moments


def main():
    import argparse
    parser = argparse.ArgumentParser()
    parser.add_argument("--transcripcion", required=True)
    parser.add_argument("--output", default="highlights_muertes.json")
    parser.add_argument("--top", type=int, default=30)
    parser.add_argument("--min-duration", type=int, default=12)
    parser.add_argument("--max-duration", type=int, default=35)
    args = parser.parse_args()

    intervals, moments = detect_death_and_failure_moments(
        args.transcripcion,
        args.min_duration,
        args.max_duration,
        args.top
    )

    # Guardar
    with open(args.output, 'w') as f:
        json.dump(intervals, f)

    logger.info(f"Guardado en {args.output}")

    # Imprimir resumen
    print(f"\n{'='*70}")
    print(f"MOMENTOS DE MUERTE Y AUTO-CRÍTICA".center(70))
    print(f"{'='*70}")
    print(f"Total: {len(intervals)} clips")
    print(f"Duración total: {sum(e-s for s,e in intervals)}s ({sum(e-s for s,e in intervals)/60:.1f} min)")
    print(f"{'-'*70}")

    for i, (start, end) in enumerate(intervals, 1):
        duration = end - start
        h = start // 3600
        m = (start % 3600) // 60
        sec = start % 60

        # Buscar el texto correspondiente
        for moment in moments:
            if abs(moment["start"] - start) < 5:
                type_icon = "💀" if moment["type"] == "muerte" else "❌"
                text_preview = moment["text"][:55].replace('\n', ' ')
                print(f"{i:2d}. {h:02d}:{m:02d}:{sec:02d} - {duration}s {type_icon} - {text_preview}...")
                break

    print(f"{'='*70}")


if __name__ == "__main__":
    main()