commit 270aba89a096f9a26f8d9dac4b6ac8e43dbf9cd5 Author: TTS Service Date: Fri Mar 27 13:41:07 2026 +0800 first commit: TTS Book Service with MiMo TTS integration diff --git a/.env.example b/.env.example new file mode 100644 index 0000000..278ecb3 --- /dev/null +++ b/.env.example @@ -0,0 +1,7 @@ +# 小米 MiMo TTS API Key（必填） +MIMO_API_KEY=your_api_key_here + +# 以下为可选配置（有默认值） +# MIMO_API_ENDPOINT=https://api.xiaomimimo.com/v1/chat/completions +# MIMO_TTS_MODEL=mimo-v2-audio-tts +# MIMO_VOICE=mimo_default diff --git a/Dockerfile b/Dockerfile new file mode 100644 index 0000000..4cecacb --- /dev/null +++ b/Dockerfile @@ -0,0 +1,17 @@ +FROM python:3.11-alpine + +# 安装 ffmpeg（alpine 版很小） +RUN apk add --no-cache ffmpeg + +WORKDIR /app + +COPY requirements.txt . +RUN pip install --no-cache-dir -r requirements.txt + +COPY app/ . + +RUN mkdir -p /app/data /app/audio + +EXPOSE 17200 + +CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "17200"] diff --git a/README.md b/README.md new file mode 100644 index 0000000..993fe65 --- /dev/null +++ b/README.md @@ -0,0 +1,96 @@ +# 📚 TTS Book Service + +基于**小米 MiMo TTS**的听书音频转换服务，提供 Web 管理界面和听书 App 音频接入接口。 + +## 架构 + +``` +听书 App ──HTTP──▶ 本服务 ──API──▶ 小米 MiMo TTS + │ + ├── SQLite (书籍/章节管理) + └── MP3 文件缓存 +``` + +## 快速启动 + +```bash +# 1. 配置 API Key +cp .env.example .env +# 编辑 .env 填入你的 MIMO_API_KEY + +# 2. 启动 +docker compose up -d + +# 3. 访问管理界面 +# http://your-server:17200 +``` + +## 功能 + +### Web 管理界面 (`/`) +- 📖 书籍管理（添加/删除） +- 📑 章节管理（添加/编辑/删除） +- 🎙️ TTS 试听（支持风格设置） +- ⚡ 单章/批量音频生成 +- ⚙️ 配置查看 + +### 听书 App 接口 + +| 接口 | 方法 | 说明 | +|------|------|------| +| `/api/book/{book_id}` | GET | 获取书籍信息和章节列表 | +| `/api/book/{book_id}/chapter/{chapter_id}/audio` | GET | 下载章节 MP3 音频 | + +### 管理 API + +| 接口 | 方法 | 说明 | +|------|------|------| +| `/admin/api/books` | GET/POST | 书籍列表/创建 | +| `/admin/api/books/{book_id}` | DELETE | 删除书籍 | +| `/admin/api/books/{book_id}/chapters` | GET/POST | 章节列表/创建 | +| `/admin/api/books/{book_id}/chapters/{id}` | PUT/DELETE | 更新/删除章节 | +| `/admin/api/books/{book_id}/chapters/{id}/generate` | POST | 生成单章音频 | +| `/admin/api/books/{book_id}/generate-all` | POST | 批量生成 | +| `/admin/api/tts/preview` | POST | TTS 试听 | +| `/admin/api/config` | GET | 查看配置 | + +## 接入听书 App + +1. 在管理界面添加书籍和章节 +2. 为章节生成音频 +3. 在听书 App 配置： + - **接入方式**: 音频方式 + - **接入地址**: `http://your-server:17200` + - **音频类型**: mp3 + - **书籍地址**: `http://your-server:17200/api/book/{book_id}` + +## 环境变量 + +| 变量 | 必填 | 默认值 | 说明 | +|------|------|--------|------| +| `MIMO_API_KEY` | ✅ | - | 小米 MiMo TTS API Key | +| `MIMO_API_ENDPOINT` | ❌ | `https://api.xiaomimimo.com/v1/chat/completions` | API 地址 | +| `MIMO_TTS_MODEL` | ❌ | `mimo-v2-audio-tts` | 模型名称 | +| `MIMO_VOICE` | ❌ | `mimo_default` | 默认音色 | +| `SERVER_PORT` | ❌ | `17200` | 服务端口 | + +## MiMo TTS 风格参考 + +在「TTS 试听」中可填写风格，例如： + +| 类别 | 示例 | +|------|------| +| 情感 | 开心 / 悲伤 / 生气 / 平静 | +| 语速 | 语速慢 / 语速快 / 悄悄话 | +| 角色 | 像个大将军 / 像个小孩 / 孙悟空 | +| 方言 | 东北话 / 四川话 / 台湾腔 / 粤语 | + +## 不使用 Docker 运行 + +```bash +pip install -r requirements.txt +# 需要系统安装 ffmpeg +export MIMO_API_KEY=your_key +cd app +python main.py +``` diff --git a/app/__pycache__/main.cpython-312.pyc b/app/__pycache__/main.cpython-312.pyc new file mode 100644 index 0000000..ff643c4 Binary files /dev/null and b/app/__pycache__/main.cpython-312.pyc differ diff --git a/app/config.py b/app/config.py new file mode 100644 index 0000000..3e231a9 --- /dev/null +++ b/app/config.py @@ -0,0 +1,21 @@ +import os + +BASE_DIR = os.path.dirname(os.path.abspath(__file__)) + +# 小米 MiMo TTS API 配置 +MIMO_API_ENDPOINT = os.environ.get( + "MIMO_API_ENDPOINT", "https://api.xiaomimimo.com/v1/chat/completions" +) +MIMO_API_KEY = os.environ.get("MIMO_API_KEY", "") +MIMO_TTS_MODEL = os.environ.get("MIMO_TTS_MODEL", "mimo-v2-audio-tts") +MIMO_VOICE = os.environ.get("MIMO_VOICE", "mimo_default") + +# 服务器配置 +SERVER_HOST = os.environ.get("SERVER_HOST", "0.0.0.0") +SERVER_PORT = int(os.environ.get("SERVER_PORT", "17200")) + +# 音频存储目录 +AUDIO_DIR = os.environ.get("AUDIO_DIR", os.path.join(BASE_DIR, "audio")) + +# 数据库 +DATABASE_URL = os.environ.get("DATABASE_URL", "sqlite+aiosqlite:///./data/tts.db") diff --git a/app/main.py b/app/main.py new file mode 100644 index 0000000..13adce6 --- /dev/null +++ b/app/main.py @@ -0,0 +1,436 @@ +""" +TTS Book Service - 小米 MiMo TTS 转换服务 +为听书 App 提供音频接入接口 +""" + +import os +import json +import base64 +import subprocess +import uuid +import asyncio +from contextlib import asynccontextmanager +from pathlib import Path + +import httpx +from fastapi import FastAPI, HTTPException, Query, Request +from fastapi.responses import FileResponse, HTMLResponse, JSONResponse +from fastapi.staticfiles import StaticFiles +from sqlalchemy import Column, Integer, String, Text, DateTime, func, select, delete +from sqlalchemy.ext.asyncio import create_async_engine, AsyncSession, async_sessionmaker +from sqlalchemy.orm import DeclarativeBase + +import config + +# ── Database ────────────────────────────────────────────────────────────── + +engine = create_async_engine(config.DATABASE_URL, echo=False) +async_session = async_sessionmaker(engine, class_=AsyncSession, expire_on_commit=False) + + +class Base(DeclarativeBase): + pass + + +class Book(Base): + __tablename__ = "books" + id = Column(Integer, primary_key=True, autoincrement=True) + book_id = Column(String(100), unique=True, nullable=False, index=True) + title = Column(String(500), nullable=False) + author = Column(String(200), default="") + created_at = Column(DateTime, server_default=func.now()) + + +class Chapter(Base): + __tablename__ = "chapters" + id = Column(Integer, primary_key=True, autoincrement=True) + book_id = Column(String(100), nullable=False, index=True) + chapter_id = Column(String(100), nullable=False, index=True) + app_chapter_id = Column(String(100), default="") + title = Column(String(500), default="") + text_content = Column(Text, default="") + audio_file = Column(String(500), default="") + status = Column(String(20), default="pending") + error_msg = Column(Text, default="") + created_at = Column(DateTime, server_default=func.now()) + updated_at = Column(DateTime, server_default=func.now(), onupdate=func.now()) + + +# ── TTS Service ─────────────────────────────────────────────────────────── + +async def call_mimo_tts(text: str, style: str = "") -> bytes: + """调用小米 MiMo TTS API，返回 WAV 音频字节""" + if not config.MIMO_API_KEY: + raise HTTPException(500, "MIMO_API_KEY 未配置，请设置环境变量") + + content = f"{text}" if style else text + + payload = { + "model": config.MIMO_TTS_MODEL, + "audio": {"format": "wav", "voice": config.MIMO_VOICE}, + "messages": [{"role": "assistant", "content": content}], + } + + headers = { + "Content-Type": "application/json", + "api-key": config.MIMO_API_KEY, + } + + async with httpx.AsyncClient(timeout=120) as client: + resp = await client.post(config.MIMO_API_ENDPOINT, json=payload, headers=headers) + + if resp.status_code != 200: + raise HTTPException(502, f"MiMo TTS API 错误: HTTP {resp.status_code} - {resp.text[:300]}") + + data = resp.json() + if data.get("error"): + raise HTTPException(502, f"MiMo TTS 错误: {data['error']}") + + try: + audio_b64 = data["choices"][0]["message"]["audio"]["data"] + return base64.b64decode(audio_b64) + except (KeyError, IndexError, TypeError) as e: + raise HTTPException(502, f"MiMo TTS 响应解析失败: {e}") + + +def wav_to_mp3(wav_path: str, mp3_path: str): + """用 ffmpeg 将 WAV 转为 MP3""" + result = subprocess.run( + ["ffmpeg", "-y", "-i", wav_path, "-codec:a", "libmp3lame", "-qscale:a", "2", mp3_path], + capture_output=True, text=True, + ) + if result.returncode != 0: + raise RuntimeError(f"ffmpeg 转换失败: {result.stderr[:300]}") + + +async def generate_chapter_audio(chapter_id_str: str): + """为指定章节生成音频（WAV → MP3）""" + async with async_session() as db: + result = await db.execute(select(Chapter).where(Chapter.chapter_id == chapter_id_str)) + chapter = result.scalar_one_or_none() + if not chapter: + return + + if not chapter.text_content.strip(): + chapter.status = "error" + chapter.error_msg = "文本内容为空" + await db.commit() + return + + chapter.status = "generating" + await db.commit() + + try: + audio_dir = Path(config.AUDIO_DIR) / chapter.book_id + audio_dir.mkdir(parents=True, exist_ok=True) + + wav_path = str(audio_dir / f"{chapter.chapter_id}.wav") + mp3_path = str(audio_dir / f"{chapter.chapter_id}.mp3") + + # MiMo TTS 生成 WAV + wav_bytes = await call_mimo_tts(chapter.text_content) + with open(wav_path, "wb") as f: + f.write(wav_bytes) + + # WAV → MP3 + loop = asyncio.get_event_loop() + await loop.run_in_executor(None, wav_to_mp3, wav_path, mp3_path) + + # 删除 WAV 源文件，只保留 MP3 + os.remove(wav_path) + + chapter.audio_file = mp3_path + chapter.status = "ready" + chapter.error_msg = "" + except Exception as e: + chapter.status = "error" + chapter.error_msg = str(e)[:500] + + await db.commit() + + +# ── App Lifecycle ────────────────────────────────────────────────────────── + +@asynccontextmanager +async def lifespan(app: FastAPI): + os.makedirs(config.AUDIO_DIR, exist_ok=True) + os.makedirs(os.path.join(config.BASE_DIR, "data"), exist_ok=True) + async with engine.begin() as conn: + await conn.run_sync(Base.metadata.create_all) + yield + + +app = FastAPI(title="TTS Book Service", lifespan=lifespan) + + +# ── 听书 App 音频接入接口 ───────────────────────────────────────────────── + +@app.get("/api/book/{book_id}") +async def get_book_info(book_id: str): + """获取书籍信息及章节列表（听书 App 调用）""" + async with async_session() as db: + book_result = await db.execute(select(Book).where(Book.book_id == book_id)) + book = book_result.scalar_one_or_none() + if not book: + raise HTTPException(404, f"书籍 {book_id} 不存在") + + ch_result = await db.execute( + select(Chapter).where(Chapter.book_id == book_id).order_by(Chapter.id) + ) + chapters = ch_result.scalars().all() + + return { + "book_id": book.book_id, + "title": book.title, + "author": book.author, + "chapters": [ + { + "chapter_id": ch.chapter_id, + "app_chapter_id": ch.app_chapter_id, + "title": ch.title, + "status": ch.status, + "audio_url": f"/api/book/{book_id}/chapter/{ch.chapter_id}/audio" + if ch.status == "ready" else None, + } + for ch in chapters + ], + } + + +@app.get("/api/book/{book_id}/chapter/{chapter_id}/audio") +async def get_chapter_audio(book_id: str, chapter_id: str): + """获取章节音频文件（听书 App 调用）""" + async with async_session() as db: + result = await db.execute( + select(Chapter).where( + Chapter.book_id == book_id, Chapter.chapter_id == chapter_id + ) + ) + chapter = result.scalar_one_or_none() + + if not chapter: + raise HTTPException(404, "章节不存在") + if chapter.status != "ready" or not chapter.audio_file: + raise HTTPException(404, f"音频尚未生成，当前状态: {chapter.status}") + + if not os.path.exists(chapter.audio_file): + raise HTTPException(404, "音频文件丢失") + + return FileResponse(chapter.audio_file, media_type="audio/mpeg", filename=f"{chapter_id}.mp3") + + +# ── 管理 API ────────────────────────────────────────────────────────────── + +# --- Books --- + +@app.get("/admin/api/books") +async def list_books(): + async with async_session() as db: + result = await db.execute(select(Book).order_by(Book.id.desc())) + books = result.scalars().all() + return [{"book_id": b.book_id, "title": b.title, "author": b.author} for b in books] + + +@app.post("/admin/api/books") +async def create_book(request: Request): + data = await request.json() + book_id = data.get("book_id", "").strip() + title = data.get("title", "").strip() + author = data.get("author", "").strip() + + if not book_id or not title: + raise HTTPException(400, "book_id 和 title 不能为空") + + async with async_session() as db: + existing = await db.execute(select(Book).where(Book.book_id == book_id)) + if existing.scalar_one_or_none(): + raise HTTPException(409, f"书籍 {book_id} 已存在") + + book = Book(book_id=book_id, title=title, author=author) + db.add(book) + await db.commit() + return {"ok": True, "book_id": book_id} + + +@app.delete("/admin/api/books/{book_id}") +async def delete_book(book_id: str): + async with async_session() as db: + await db.execute(delete(Chapter).where(Chapter.book_id == book_id)) + await db.execute(delete(Book).where(Book.book_id == book_id)) + await db.commit() + return {"ok": True} + + +# --- Chapters --- + +@app.get("/admin/api/books/{book_id}/chapters") +async def list_chapters(book_id: str): + async with async_session() as db: + result = await db.execute( + select(Chapter).where(Chapter.book_id == book_id).order_by(Chapter.id) + ) + chapters = result.scalars().all() + return [ + { + "chapter_id": ch.chapter_id, + "app_chapter_id": ch.app_chapter_id, + "title": ch.title, + "text_content": ch.text_content[:200] + "..." if len(ch.text_content) > 200 else ch.text_content, + "text_length": len(ch.text_content), + "status": ch.status, + "error_msg": ch.error_msg, + "has_audio": ch.status == "ready", + } + for ch in chapters + ] + + +@app.post("/admin/api/books/{book_id}/chapters") +async def create_chapter(book_id: str, request: Request): + data = await request.json() + chapter_id = data.get("chapter_id", "").strip() + title = data.get("title", "").strip() + app_chapter_id = data.get("app_chapter_id", "").strip() + text_content = data.get("text_content", "").strip() + + if not chapter_id: + raise HTTPException(400, "chapter_id 不能为空") + + async with async_session() as db: + existing = await db.execute( + select(Chapter).where(Chapter.book_id == book_id, Chapter.chapter_id == chapter_id) + ) + if existing.scalar_one_or_none(): + raise HTTPException(409, f"章节 {chapter_id} 已存在") + + chapter = Chapter( + book_id=book_id, + chapter_id=chapter_id, + app_chapter_id=app_chapter_id or chapter_id, + title=title, + text_content=text_content, + ) + db.add(chapter) + await db.commit() + return {"ok": True, "chapter_id": chapter_id} + + +@app.put("/admin/api/books/{book_id}/chapters/{chapter_id}") +async def update_chapter(book_id: str, chapter_id: str, request: Request): + data = await request.json() + async with async_session() as db: + result = await db.execute( + select(Chapter).where(Chapter.book_id == book_id, Chapter.chapter_id == chapter_id) + ) + chapter = result.scalar_one_or_none() + if not chapter: + raise HTTPException(404, "章节不存在") + + if "text_content" in data: + chapter.text_content = data["text_content"] + if "title" in data: + chapter.title = data["title"] + if "app_chapter_id" in data: + chapter.app_chapter_id = data["app_chapter_id"] + + await db.commit() + return {"ok": True} + + +@app.delete("/admin/api/books/{book_id}/chapters/{chapter_id}") +async def delete_chapter(book_id: str, chapter_id: str): + async with async_session() as db: + await db.execute( + delete(Chapter).where(Chapter.book_id == book_id, Chapter.chapter_id == chapter_id) + ) + await db.commit() + return {"ok": True} + + +# --- TTS --- + +@app.post("/admin/api/books/{book_id}/chapters/{chapter_id}/generate") +async def generate_audio(book_id: str, chapter_id: str): + """手动生成单章音频""" + await generate_chapter_audio(chapter_id) + async with async_session() as db: + result = await db.execute( + select(Chapter).where(Chapter.book_id == book_id, Chapter.chapter_id == chapter_id) + ) + ch = result.scalar_one_or_none() + return {"ok": True, "status": ch.status, "error_msg": ch.error_msg} + + +@app.post("/admin/api/books/{book_id}/generate-all") +async def generate_all_chapters(book_id: str): + """批量生成书籍所有章节音频""" + async with async_session() as db: + result = await db.execute( + select(Chapter).where(Chapter.book_id == book_id, Chapter.status != "ready") + ) + chapters = result.scalars().all() + + chapter_ids = [ch.chapter_id for ch in chapters] + for cid in chapter_ids: + await generate_chapter_audio(cid) + + return {"ok": True, "total": len(chapter_ids), "chapter_ids": chapter_ids} + + +# --- TTS 试听 --- + +@app.post("/admin/api/tts/preview") +async def tts_preview(request: Request): + """试听 TTS 效果""" + data = await request.json() + text = data.get("text", "").strip() + style = data.get("style", "").strip() + + if not text: + raise HTTPException(400, "文本不能为空") + + wav_bytes = await call_mimo_tts(text, style) + audio_dir = Path(config.AUDIO_DIR) / "_preview" + audio_dir.mkdir(parents=True, exist_ok=True) + + filename = f"{uuid.uuid4().hex}.mp3" + wav_path = str(audio_dir / f"{uuid.uuid4().hex}.wav") + mp3_path = str(audio_dir / filename) + + with open(wav_path, "wb") as f: + f.write(wav_bytes) + + loop = asyncio.get_event_loop() + await loop.run_in_executor(None, wav_to_mp3, wav_path, mp3_path) + os.remove(wav_path) + + return {"ok": True, "url": f"/audio/_preview/{filename}"} + + +@app.get("/admin/api/config") +async def get_config(): + return { + "endpoint": config.MIMO_API_ENDPOINT, + "model": config.MIMO_TTS_MODEL, + "voice": config.MIMO_VOICE, + "api_key_masked": config.MIMO_API_KEY[:6] + "****" if config.MIMO_API_KEY else "未配置", + } + + +# ── 静态文件 & 前端 ────────────────────────────────────────────────────── + +app.mount("/audio", StaticFiles(directory=config.AUDIO_DIR), name="audio") + + +@app.get("/", response_class=HTMLResponse) +async def frontend(): + html_path = os.path.join(config.BASE_DIR, "static", "index.html") + with open(html_path, "r", encoding="utf-8") as f: + return HTMLResponse(f.read()) + + +# ── Main ────────────────────────────────────────────────────────────────── + +if __name__ == "__main__": + import uvicorn + uvicorn.run("main:app", host=config.SERVER_HOST, port=config.SERVER_PORT, reload=True) diff --git a/app/models.py b/app/models.py new file mode 100644 index 0000000..6bfd13d --- /dev/null +++ b/app/models.py @@ -0,0 +1,42 @@ +from sqlalchemy import Column, Integer, String, Text, DateTime, func +from sqlalchemy.ext.asyncio import create_async_engine, AsyncSession, async_sessionmaker +from sqlalchemy.orm import DeclarativeBase +import config + +engine = create_async_engine(config.DATABASE_URL, echo=False) +async_session = async_sessionmaker(engine, class_=AsyncSession, expire_on_commit=False) + + +class Base(DeclarativeBase): + pass + + +class Book(Base): + __tablename__ = "books" + + id = Column(Integer, primary_key=True, autoincrement=True) + book_id = Column(String(100), unique=True, nullable=False, index=True) # 平台书籍ID + title = Column(String(500), nullable=False) # 书名 + author = Column(String(200), default="") # 作者 + created_at = Column(DateTime, server_default=func.now()) + + +class Chapter(Base): + __tablename__ = "chapters" + + id = Column(Integer, primary_key=True, autoincrement=True) + book_id = Column(String(100), nullable=False, index=True) # 关联的书籍ID + chapter_id = Column(String(100), nullable=False, index=True) # 平台音频ID + app_chapter_id = Column(String(100), default="") # App章节ID + title = Column(String(500), default="") # 章节标题 + text_content = Column(Text, default="") # TTS 文本内容 + audio_file = Column(String(500), default="") # 生成的音频文件路径 + status = Column(String(20), default="pending") # pending / generating / ready / error + error_msg = Column(Text, default="") + created_at = Column(DateTime, server_default=func.now()) + updated_at = Column(DateTime, server_default=func.now(), onupdate=func.now()) + + +async def init_db(): + async with engine.begin() as conn: + await conn.run_sync(Base.metadata.create_all) diff --git a/app/static/index.html b/app/static/index.html new file mode 100644 index 0000000..c69cf89 --- /dev/null +++ b/app/static/index.html @@ -0,0 +1,480 @@ + + + + + +TTS Book Service + + + +
+

📚 TTS Book Service

+

小米 MiMo TTS 听书音频转换服务

+ +
+ + + +
+ + +
+
+
+
📖 书籍列表
+ +
+
+
+

加载中...

+
+
+ + +
+
+
🎙️ TTS 试听
+
+ + +
+
+ + +
+ + +
+
+ + +
+
+
⚙️ 当前配置
+ + + + + + +
配置项值
TTS API-
模型-
默认音色-
API Key-
+

通过环境变量配置：MIMO_API_KEY、MIMO_API_ENDPOINT、MIMO_TTS_MODEL、MIMO_VOICE

+
+
+
+ + + + + + + + + + + + + + + + diff --git a/docker-compose.yml b/docker-compose.yml new file mode 100644 index 0000000..e64efb2 --- /dev/null +++ b/docker-compose.yml @@ -0,0 +1,27 @@ +version: "3.8" + +services: + tts-book-service: + build: . + container_name: tts-book-service + ports: + - "17200:17200" + environment: + - MIMO_API_KEY=${MIMO_API_KEY} + - MIMO_API_ENDPOINT=${MIMO_API_ENDPOINT:-https://api.xiaomimimo.com/v1/chat/completions} + - MIMO_TTS_MODEL=${MIMO_TTS_MODEL:-mimo-v2-audio-tts} + - MIMO_VOICE=${MIMO_VOICE:-mimo_default} + - SERVER_PORT=17200 + volumes: + - tts-data:/app/data + - tts-audio:/app/audio + deploy: + resources: + limits: + memory: 512M + cpus: "1.0" + restart: unless-stopped + +volumes: + tts-data: + tts-audio: diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 0000000..c7b8ef3 --- /dev/null +++ b/requirements.txt @@ -0,0 +1,6 @@ +fastapi==0.115.0 +uvicorn[standard]==0.30.0 +sqlalchemy==2.0.35 +aiosqlite==0.20.0 +httpx==0.27.0 +python-multipart==0.0.12