xiaoxia-saas/packages/domain/voice_presets.py

"""配音引擎音色预设.

与 CosyVoice 的 preset_voices 区分：
- preset_voices.py: CosyVoice 真实音色（阿里云）
- voice_presets.py: 配音引擎通用音色预设（含 mock/后续接入的真实 TTS）
"""

from __future__ import annotations

import sys
from dataclasses import dataclass

if sys.version_info >= (3, 11):
    from enum import StrEnum
else:
    from enum import Enum

    class StrEnum(str, Enum):
        pass


class VoiceGender(StrEnum):
    """音色性别."""

    MALE = "male"
    FEMALE = "female"
    CHILD = "child"


class VoiceStyle(StrEnum):
    """音色风格."""

    STABLE = "stable"  # 沉稳
    LIVELY = "lively"  # 活泼
    CUSTOMER_SERVICE = "customer_service"  # 客服
    NARRATION = "narration"  # 旁白
    NEWS = "news"  # 新闻
    STORY = "story"  # 故事


@dataclass(slots=True)
class VoicePreset:
    """音色预设.

    Attributes:
        voice_id: 音色唯一标识
        name: 音色名称
        gender: 性别
        style: 风格
        description: 描述
        provider: 供应商（mock/aliyun/xunfei）
        provider_voice_id: 供应商侧音色 ID
        default_speed: 默认语速
        default_pitch: 默认语调
        sample_rate: 采样率
        language: 语言
    """

    voice_id: str
    name: str
    gender: VoiceGender = VoiceGender.FEMALE
    style: VoiceStyle = VoiceStyle.NARRATION
    description: str = ""
    provider: str = "mock"
    provider_voice_id: str = ""
    default_speed: float = 1.0
    default_pitch: float = 0.0
    sample_rate: int = 22050
    language: str = "zh-CN"


# ─── Mock 音色预设列表 ──────────────────────────────────────

MOCK_VOICES: list[VoicePreset] = [
    VoicePreset(
        voice_id="female_warm",
        name="温暖女声",
        gender=VoiceGender.FEMALE,
        style=VoiceStyle.NARRATION,
        description="温柔温暖的女声，适合情感类、生活类视频",
        provider="mock",
        provider_voice_id="sine_220",
        default_speed=1.0,
        default_pitch=0.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="male_stable",
        name="沉稳男声",
        gender=VoiceGender.MALE,
        style=VoiceStyle.STABLE,
        description="沉稳厚重的男声，适合商务、知识类视频",
        provider="mock",
        provider_voice_id="sine_110",
        default_speed=0.9,
        default_pitch=0.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="female_lively",
        name="活泼女声",
        gender=VoiceGender.FEMALE,
        style=VoiceStyle.LIVELY,
        description="明亮活泼的女声，适合vlog、美食、旅行类视频",
        provider="mock",
        provider_voice_id="sine_280",
        default_speed=1.2,
        default_pitch=2.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="child_cute",
        name="可爱童声",
        gender=VoiceGender.CHILD,
        style=VoiceStyle.STORY,
        description="清脆可爱的童声，适合儿童教育、动画类视频",
        provider="mock",
        provider_voice_id="sine_380",
        default_speed=1.0,
        default_pitch=4.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="female_service",
        name="客服女声",
        gender=VoiceGender.FEMALE,
        style=VoiceStyle.CUSTOMER_SERVICE,
        description="专业清晰的客服女声，适合产品介绍、教程类视频",
        provider="mock",
        provider_voice_id="sine_250",
        default_speed=1.0,
        default_pitch=1.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="male_news",
        name="新闻男声",
        gender=VoiceGender.MALE,
        style=VoiceStyle.NEWS,
        description="字正腔圆的新闻播报声，适合资讯、时政类视频",
        provider="mock",
        provider_voice_id="sine_140",
        default_speed=1.0,
        default_pitch=0.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="female_soft",
        name="轻柔女声",
        gender=VoiceGender.FEMALE,
        style=VoiceStyle.STORY,
        description="轻柔舒缓的女声，适合睡前故事、冥想类视频",
        provider="mock",
        provider_voice_id="sine_180",
        default_speed=0.8,
        default_pitch=0.0,
        sample_rate=22050,
        language="zh-CN",
    ),
    VoicePreset(
        voice_id="male_magnetic",
        name="磁性男声",
        gender=VoiceGender.MALE,
        style=VoiceStyle.STORY,
        description="低沉磁性的男声，适合电影解说、读书类视频",
        provider="mock",
        provider_voice_id="sine_90",
        default_speed=0.85,
        default_pitch=-2.0,
        sample_rate=22050,
        language="zh-CN",
    ),
]


# voice_id → VoicePreset
_MOCK_VOICE_MAP: dict[str, VoicePreset] = {v.voice_id: v for v in MOCK_VOICES}


def get_voice(voice_id: str, *, provider: str = "mock") -> VoicePreset | None:
    """根据 voice_id 获取音色预设."""
    if provider == "mock":
        return _MOCK_VOICE_MAP.get(voice_id)
    return None


def list_voices(
    *,
    gender: str | None = None,
    style: str | None = None,
    provider: str | None = None,
    keyword: str | None = None,
) -> list[VoicePreset]:
    """按条件筛选音色列表."""
    # 目前只有 mock 音色
    result = list(MOCK_VOICES)

    if provider and provider != "mock":
        return []

    if gender:
        result = [v for v in result if v.gender.value == gender]

    if style:
        result = [v for v in result if v.style.value == style]

    if keyword:
        kw = keyword.lower()
        result = [v for v in result if kw in v.name.lower() or kw in v.description.lower() or kw in v.voice_id.lower()]

    return result


def get_default_voice() -> VoicePreset:
    """获取默认音色."""
    return MOCK_VOICES[0]