文件内容
scripts/voice_clone.py
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Voice Clone - AI 声音复刻与语音合成工具
调用 AI Artist API 进行音色克隆和语音合成
"""
import os
import sys
import json
import time
import argparse
import requests
from pathlib import Path
# API 配置
BASE_URL = "https://ai.deepsop.com/prod-api"
FILE_UPLOAD_URL = f"{BASE_URL}/system/fileUpload/upload"
# 环境配置
API_KEY_ENV = "AI_ARTIST_TOKEN"
def get_api_key():
"""获取 API Key"""
api_key = os.environ.get(API_KEY_ENV)
if not api_key:
# Try loading from .env file (in script directory)
env_file = Path(__file__).parent / ".env"
if env_file.exists():
with open(env_file, "r", encoding="utf-8") as f:
for line in f:
line = line.strip()
if line and not line.startswith("#") and line.startswith(f"{API_KEY_ENV}="):
api_key = line.split("=", 1)[1].strip('"\'')
break
# Also check parent directory (skill root)
if not api_key:
env_file = Path(__file__).parent.parent / ".env"
if env_file.exists():
with open(env_file, "r", encoding="utf-8") as f:
for line in f:
line = line.strip()
if line and not line.startswith("#") and line.startswith(f"{API_KEY_ENV}="):
api_key = line.split("=", 1)[1].strip('"\'')
break
return api_key
def check_api_key():
"""检查 API Key 是否配置"""
api_key = get_api_key()
if not api_key:
print(f"[ERROR] 未配置 {API_KEY_ENV} 环境变量", file=sys.stderr)
print(f"\n请设置 API Key:")
print(f" Windows PowerShell: $env:{API_KEY_ENV}=\"sk-your_api_key_here\"")
print(f" Linux/macOS: export {API_KEY_ENV}=\"sk-your_api_key_here\"")
return None
return api_key
def get_headers():
"""获取请求头"""
api_key = get_api_key()
return {
"x-api-key": api_key,
"Content-Type": "application/json"
}
def upload_file(file_path):
"""上传本地文件到 OSS 获取 URL"""
if not os.path.exists(file_path):
print(f"[ERROR] 文件不存在：{file_path}", file=sys.stderr)
return None
try:
with open(file_path, "rb") as f:
files = {"file": (os.path.basename(file_path), f)}
response = requests.post(
FILE_UPLOAD_URL,
headers={"x-api-key": get_api_key()},
files=files,
timeout=60
)
response.raise_for_status()
result = response.json()
if result.get("code") == 200:
url = result.get("url")
print(f"[SUCCESS] 文件已上传：{file_path} -> {url}")
return url
else:
print(f"[ERROR] 文件上传失败：{result.get('msg', '未知错误')}", file=sys.stderr)
return None
except requests.exceptions.RequestException as e:
print(f"[ERROR] 文件上传错误：{e}", file=sys.stderr)
return None
def list_voices():
"""查询音色列表"""
url = f"{BASE_URL}/ai/voice/clone/list"
params = {"pageNum": 1, "pageSize": 10}
try:
response = requests.get(url, headers=get_headers(), params=params, timeout=30)
response.raise_for_status()
result = response.json()
if result.get("code") != 200:
print(f"[ERROR] 查询失败：{result.get('msg', '未知错误')}", file=sys.stderr)
return None
rows = result.get("rows", [])
total = result.get("total", 0)
print(f"[INFO] 共有 {total} 个音色")
print("\n可用音色列表:")
available = []
for voice in rows:
vid = voice.get("id")
name = voice.get("name")
status = voice.get("status")
model = voice.get("targetModel", "unknown")
status_mark = "[OK]" if status == "OK" else f"[{status}]"
print(f" [{vid}] {name} {status_mark} - {model}")
if status == "OK":
available.append(voice)
return available
except requests.exceptions.RequestException as e:
print(f"[ERROR] 查询音色列表出错：{e}", file=sys.stderr)
return None
def create_voice(name, audio_url=None, audio_path=None, prefix="DeepSop", remark=None):
"""创建新音色"""
# 如果提供本地文件路径，先上传获取 URL
if audio_path and not audio_url:
audio_url = upload_file(audio_path)
if not audio_url:
return None
if not audio_url:
print("[ERROR] 必须提供 audio_url 或 audio_path", file=sys.stderr)
return None
payload = {
"name": name,
"prefix": prefix,
"audioUrl": audio_url,
"remark": remark
}
url = f"{BASE_URL}/ai/voice/clone/sync/create"
try:
response = requests.post(url, headers=get_headers(), json=payload, timeout=60)
response.raise_for_status()
result = response.json()
if result.get("code") != 200:
print(f"[ERROR] 创建失败：{result.get('msg', '未知错误')}", file=sys.stderr)
return None
data = result.get("data", {})
vid = data.get("id")
voice_id = data.get("voiceId")
status = data.get("status")
print(f"[SUCCESS] 音色创建成功!")
print(f" ID: {vid}")
print(f" VoiceID: {voice_id}")
print(f" 名称：{name}")
print(f" 状态：{status}")
return data
except requests.exceptions.RequestException as e:
print(f"[ERROR] 创建音色出错：{e}", file=sys.stderr)
return None
def get_balance():
"""查询 K 币余额"""
url = f"{BASE_URL}/ai/vip/balance"
try:
response = requests.get(url, headers=get_headers(), timeout=30)
response.raise_for_status()
result = response.json()
if result.get("code") != 200:
print(f"[ERROR] 查询余额失败：{result.get('msg', '未知错误')}", file=sys.stderr)
return None
balance = result.get("data")
if balance is None:
print("[ERROR] 查询余额失败：返回数据缺少余额信息", file=sys.stderr)
return None
print(f"[INFO] 当前 K 币余额：{balance}")
return float(balance)
except (ValueError, TypeError):
print("[ERROR] 查询余额失败：余额数据格式异常", file=sys.stderr)
return None
except requests.exceptions.RequestException as e:
print(f"[ERROR] 查询余额出错：{e}", file=sys.stderr)
return None
def synthesize_voice(text, voice_id=None, voice_name=None):
"""语音合成"""
# 如果提供名称，先查询获取 ID
if voice_name and not voice_id:
voices = list_voices()
if voices:
for v in voices:
if v.get("name") == voice_name:
if v.get("status") != "OK":
print(f"[ERROR] 音色 '{voice_name}' 状态为 {v.get('status')}，不可用", file=sys.stderr)
return None
voice_id = v.get("id")
break
if not voice_id:
print(f"[ERROR] 未找到音色：{voice_name}", file=sys.stderr)
return None
if not voice_id:
print("[ERROR] 必须提供 voice_id 或 voice_name", file=sys.stderr)
return None
balance = get_balance()
if balance is None:
return None
if balance <= 0:
print("[ERROR] K 币余额不足，无法进行语音合成，请先充值 K 币", file=sys.stderr)
return None
payload = {
"text": text,
"id": voice_id
}
url = f"{BASE_URL}/ai/voice/clone/synthesize"
try:
response = requests.post(url, headers=get_headers(), json=payload, timeout=60)
response.raise_for_status()
result = response.json()
if result.get("code") != 200:
print(f"[ERROR] 合成失败：{result.get('msg', '未知错误')}", file=sys.stderr)
return None
audio_url = result.get("msg")
print(f"[SUCCESS] 语音合成成功!")
print(f" 音频链接：{audio_url}")
return {
"status": "SUCCESS",
"url": audio_url
}
except requests.exceptions.RequestException as e:
print(f"[ERROR] 语音合成出错：{e}", file=sys.stderr)
return None
def download_audio(url, output_dir=None):
"""下载音频文件"""
if not output_dir:
output_dir = os.path.join(os.path.expanduser("~"), ".openclaw", "workspace", "audio")
os.makedirs(output_dir, exist_ok=True)
try:
response = requests.get(url, timeout=60)
response.raise_for_status()
# Generate filename from URL
filename = url.split("/")[-1].split("?")[0]
if not filename.endswith(".mp3"):
filename = f"{int(time.time())}.mp3"
output_path = os.path.join(output_dir, filename)
with open(output_path, "wb") as f:
f.write(response.content)
print(f"[SAVE] 音频已保存：{output_path}")
return output_path
except requests.exceptions.RequestException as e:
print(f"[WARNING] 下载音频失败：{e}", file=sys.stderr)
return None
def main():
parser = argparse.ArgumentParser(
description="AI 声音复刻与语音合成工具",
formatter_class=argparse.RawDescriptionHelpFormatter
)
# 模式选择
mode_group = parser.add_mutually_exclusive_group(required=True)
mode_group.add_argument("--list", action="store_true", help="列出所有可用音色")
mode_group.add_argument("--synthesize", action="store_true", help="语音合成模式")
mode_group.add_argument("--create", action="store_true", help="创建新音色模式")
# 通用参数
parser.add_argument("--id", type=int, help="音色 ID")
parser.add_argument("--name", help="音色名称")
parser.add_argument("--text", help="要合成的文本内容")
# 创建音色参数
parser.add_argument("--audio", help="本地音频文件路径（创建音色时使用）")
parser.add_argument("--audio-url", help="在线音频 URL（创建音色时使用）")
parser.add_argument("--prefix", default="DeepSop", help="音色前缀（默认：DeepSop）")
# 下载参数
parser.add_argument("--download", action="store_true", help="下载合成的音频到本地")
parser.add_argument("--output-dir", help="音频保存目录")
args = parser.parse_args()
# 检查 API Key
if not check_api_key():
sys.exit(1)
# 列出音色
if args.list:
voices = list_voices()
sys.exit(0 if voices is not None else 1)
# 语音合成
elif args.synthesize:
if not args.text:
print("[ERROR] 合成模式必须提供 --text 参数", file=sys.stderr)
sys.exit(1)
if not args.id and not args.name:
print("[ERROR] 合成模式必须提供 --id 或 --name 参数", file=sys.stderr)
sys.exit(1)
result = synthesize_voice(args.text, voice_id=args.id, voice_name=args.name)
if result and result.get("status") == "SUCCESS":
if args.download:
download_audio(result["url"], args.output_dir)
print(result["url"])
sys.exit(0)
else:
sys.exit(1)
# 创建音色
elif args.create:
if not args.name:
print("[ERROR] 创建模式必须提供 --name 参数", file=sys.stderr)
sys.exit(1)
if not args.audio and not args.audio_url:
print("[ERROR] 创建模式必须提供 --audio 或 --audio-url 参数", file=sys.stderr)
sys.exit(1)
result = create_voice(
name=args.name,
audio_url=args.audio_url,
audio_path=args.audio,
prefix=args.prefix
)
if result:
sys.exit(0)
else:
sys.exit(1)
if __name__ == "__main__":
main()