import io
import json

import requests
from elevenlabs import voices, generate, set_api_key, VoiceSettings, Voice
from pydub import AudioSegment

from .tts_utils import mix_background_music


def merge_audio(token, text, 背景音乐, 音色选择, stability, similarity_boost, style, use_speaker_boost, TTS_up, bg_up):
    try:
        set_api_key(token)

        original_audio_bytes = generate(
            text=text,
            voice=Voice(voice_id=voices_eleven[speakers_eleven[音色选择]].voice_id,
                        settings=VoiceSettings(stability=stability, similarity_boost=similarity_boost, style=style,
                                               use_speaker_boost=use_speaker_boost)),
            model="eleven_multilingual_v2",

        )

        # 将字节串转换为AudioSegment对象
        original_audio = AudioSegment.from_file(io.BytesIO(original_audio_bytes), format="mp3")

        return None, *mix_background_music(original_audio, 背景音乐, TTS_up, bg_up)
    except Exception as e:
        return str(e), None, None


def get_eleven_spk():
    try:
        voices_eleven = voices()
        speakers_eleven = {}
        for i, v in enumerate(voices_eleven):
            label = {}
            label['口音'] = v.labels.get('accent')

            if 'description ' in v.labels:
                label['描述'] = v.labels['description ']
            elif 'description' in v.labels:
                label['描述'] = v.labels['description']

            label['年龄'] = v.labels['age']
            label['性别'] = v.labels['gender']

            if 'use case' in v.labels:
                label['用例'] = v.labels['use case']
            elif 'usecase' in v.labels:
                label['用例'] = v.labels['usecase']
            else:
                raise Exception('use case not found')

            key = v.name + '-' + str(label)
            speakers_eleven[key] = i
        select_key = list(speakers_eleven.keys())

        return voices_eleven, speakers_eleven, select_key
    except Exception as e:
        print(e)
        return [], {'获取音色失败': '获取音色失败'}, ['获取音色失败']



voices_eleven, speakers_eleven, select_key = get_eleven_spk()