twitch-highlight-detector/detector_gameplay.py

#!/usr/bin/env python3
"""
Detector de GAMEPLAY ACTIVO:
Busca momentos donde está jugando, no solo hablando.
Filtra intros y momentos de solo hablar.
"""
import json
import logging
import re
import numpy as np

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


def detect_active_gameplay(chat_data, transcripcion_json, intro_skip=90, min_duration=20, max_duration=45, top=30):
    """
    Detecta momentos de gameplay activo (hablando + jugando).
    """
    logger.info("=== Buscando GAMEPLAY ACTIVO ===")

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

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

    # Palabras de GAMEPLAY (está jugando)
    gameplay_keywords = [
        r'\b(champion|campe[oó]n|ult[i|í]|habilidad|spell|q|w|e|r)\b',
        r'\b(kill|muert[e|é]|mate|muero|fui|mat[aá])\b',
        r'\b(fight|pelea|fight|team|equip|jungla|top|mid|adc|support)\b',
        r'\b(lane|linia|mina|drag[oó]n|baron|nashor|torre|inhib)\b',
        r'\b(ult|ultimate|flash|ignite|exhaust|teleport|heal)\b',
        r'\b(gank|roam|invade|invasi[oó])\b',
        r'\b(ward|vision|control|map|objetivo)\b',
        r'\b(damage|daño|dps|burst|poke)\b',
        r'\b(lethality|letalidad|crit|cr[i|í]tico)\b',
        r'\b(arma|item|build|objeto|poder|stats)\b',
        r'\b(level|nivel|exp|gold|oro|farm|cs)\b',
        r'\b(esquiv|evade|dodge|block|bloqueo)\b',
        r'\b(engage|pelear|inici|all[i|í]n|surrender|rindase)\b',
    ]

    # Palabras de SOLO HABLAR (excluir)
    talking_only_keywords = [
        r'\b(hola|buenas|buenas tardes|buenas noches|adi[oó]s)\b',
        r'\b(gracias|thank|agradezco)\b',
        r'\b(playlist|música|canci[oó]n|song)\b',
        r'\b(intro|presento|presentaci[oó]n|inicio)\b',
        r'\b(despedida|adi[oó]s|nos vemos|chao)\b',
        r'\b(merch|tienda|store|donar|donaci[oó]n)\b',
        r'\b(rrs|redes sociales|twitter|instagram|discord)\b',
        r'\b giveaway|sorteo|regalo\b',
    ]

    # Analizar segmentos
    gameplay_moments = []

    for i, seg in enumerate(segments):
        start = seg["start"]
        end = seg["end"]
        text = seg["text"]
        text_lower = text.lower()

        # Saltar intro
        if start < intro_skip:
            continue

        # Verificar que NO sea solo hablar
        is_talking_only = False
        for pattern in talking_only_keywords:
            if re.search(pattern, text_lower):
                is_talking_only = True
                break

        if is_talking_only:
            continue

        # Verificar que tenga palabras de gameplay
        gameplay_score = 0
        for pattern in gameplay_keywords:
            matches = len(re.findall(pattern, text_lower))
            gameplay_score += matches

        if gameplay_score > 0:
            gameplay_moments.append({
                "start": start,
                "end": end,
                "score": gameplay_score,
                "text": text.strip()[:80]
            })

    if not gameplay_moments:
        logger.warning("No se encontraron momentos de gameplay")
        return []

    # Ordenar por score
    gameplay_moments.sort(key=lambda x: -x["score"])

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

        # Duración del clip
        duration = max(min_duration, min(end - start, max_duration))

        # Extender el final para capturar la acción
        end = start + duration + 5  # +5s padding

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

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

    intervals.sort()

    logger.info(f"Moments de gameplay detectados: {len(intervals)}")

    return intervals, gameplay_moments


def main():
    import argparse
    parser = argparse.ArgumentParser()
    parser.add_argument("--chat", required=True)
    parser.add_argument("--transcripcion", required=True)
    parser.add_argument("--output", default="highlights_gameplay.json")
    parser.add_argument("--top", type=int, default=30)
    parser.add_argument("--intro-skip", type=int, default=90)
    parser.add_argument("--min-duration", type=int, default=20)
    parser.add_argument("--max-duration", type=int, default=45)
    args = parser.parse_args()

    with open(args.chat, 'r') as f:
        chat_data = json.load(f)

    intervals, moments = detect_active_gameplay(
        chat_data,
        args.transcripcion,
        args.intro_skip,
        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"GAMEPLAY ACTIVO - MOMENTOS JUGANDO".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"Intro excluida: primeros {args.intro_skip}s")
    print(f"{'-'*70}")

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

        for moment in moments:
            if abs(moment["start"] - start) < 10:
                text_preview = moment["text"][:60].replace('\n', ' ')
                print(f"{i:2d}. {h:02d}:{m:02d}:{sec:02d} - {duration}s - {text_preview}...")
                break

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


if __name__ == "__main__":
    main()