# -*- coding: utf-8 -*-
import logging
import os
import subprocess
import tempfile
from typing import List, Optional, Tuple

from app.schemas import AsrHit, KeywordConfig
from app.video_io import get_ffmpeg_cmd

logger = logging.getLogger(__name__)

_whisper_model = None


def _get_whisper():
    global _whisper_model
    if _whisper_model is None:
        from faster_whisper import WhisperModel
        model_size = os.environ.get("WHISPER_MODEL", "tiny")
        _whisper_model = WhisperModel(model_size, device="cpu", compute_type="int8")
        logger.info("加载 Whisper 模型: %s", model_size)
    return _whisper_model


def extract_audio_wav(video_path: str) -> str:
    out = tempfile.NamedTemporaryFile(suffix=".wav", delete=False)
    out.close()
    cmd = [
        get_ffmpeg_cmd("ffmpeg"), "-y", "-i", video_path,
        "-vn", "-acodec", "pcm_s16le", "-ar", "16000", "-ac", "1",
        out.name,
    ]
    subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
    return out.name


def transcribe(video_path: str) -> List[Tuple[str, float, float]]:
    wav = extract_audio_wav(video_path)
    try:
        model = _get_whisper()
        segments, _ = model.transcribe(wav, language="zh", vad_filter=True)
        return [(seg.text.strip(), seg.start, seg.end) for seg in segments if seg.text.strip()]
    except Exception as e:
        logger.warning("ASR 失败: %s", e)
        return []
    finally:
        if os.path.isfile(wav):
            os.remove(wav)


def match_keywords(
    segments: List[Tuple[str, float, float]],
    keywords: KeywordConfig,
) -> List[AsrHit]:
    hits: List[AsrHit] = []
    for text, start, _end in segments:
        for kw in keywords.storefront:
            if kw in text:
                hits.append(AsrHit(keyword=kw, time_sec=round(start, 2)))
                break
        for kw in keywords.handover:
            if kw in text:
                hits.append(AsrHit(keyword=kw, time_sec=round(start, 2)))
                break
    return hits


def best_asr_time(hits: List[AsrHit], keywords: List[str]) -> Optional[float]:
    for hit in hits:
        if hit.keyword in keywords:
            return hit.time_sec
    return None


def asr_available() -> bool:
    try:
        import faster_whisper  # noqa: F401
        return True
    except ImportError:
        return False