Translate Server

What is Translate Server about?

Translate Server automates the localization workflow by reading .ts, .js, and .json files, detecting the language code embedded in the filename (e.g., messages.en-US.ts), and sending the text content to a large language model for translation. The tool preserves the original file structure and creates backups before modifying files.

How to use Translate Server?

1. Install – run npx -y translate-server (or install locally via npm/yarn).
2. Configure – add an MCP entry in Cursor (or your IDE) with the required API key, model name, and base URL.
3. Run – execute the server with the path to a single file or a directory. Use absolute paths; relative paths are resolved from the execution location.
4. Result – translated files are written alongside the originals, and a backup file original.backup.<timestamp>.<ext> is kept.

Key features of Translate Server

• Supports TypeScript, JavaScript, and JSON localization files.
• Automatic language code detection from filenames (xx-XX or xx-xx).
• Batch processing of entire directories with concurrent file handling.
• Automatic backup of original files before transformation.
• Chunked processing for large files (default 500 key‑value pairs per chunk).
• Error isolation – failure of one file does not stop the whole batch.
• Simple MCP configuration via JSON snippet for Windows/macOS.

Use cases of Translate Server

• Rapidly generate translated resource files for new locales in a software project.
• Maintain consistency across multiple language versions by using a single LLM backend.
• Integrate into CI/CD pipelines to keep translation assets up‑to‑date.
• Localize open‑source libraries or documentation without manual copy‑paste.

FAQ from the Translate Server

Q: Which language models are supported? A: Any model reachable through the provided baseUrl that accepts text translation prompts. The API key, model name, and base URL are supplied in the MCP configuration.

Q: Do I need to rename my files to match the naming convention? A: Yes. The tool only processes files whose names contain a valid language code (e.g., en-US, zh-CN). Files that do not match are skipped.

Q: What happens to the original files? A: Before any modification a backup is created with the pattern original.backup.<timestamp>.<ext>. You can restore from these backups if needed.

Q: Can I limit concurrency or chunk size? A: Advanced options are configurable in the source code or via environment variables; defaults are reasonable for most scenarios.

Q: Is there support for other file formats? A: Currently only .ts, .js, and .json are supported. Extending to additional formats would require custom parsing logic.

多语言本地化翻译工具

这是一个基于大语言模型的本地化翻译工具，可以自动处理并翻译各种本地化文件（.ts、.js、.json）中的文本内容。

功能特点

1. 支持多种文件格式

   • TypeScript (.ts)
   • JavaScript (.js)
   • JSON (.json)

2. 智能文件处理

   • 自动识别文件中的语言代码（如：en-US, zh-CN）
   • 自动备份原始文件
   • 保持源文件格式和结构

3. 高效批处理

   • 支持整个目录的批量处理
   • 自动创建输出目录结构
   • 并发处理多个文件
   • 支持单文件处理

4. 安全特性

   • 自动文件备份
   • 错误处理和日志记录
   • 格式验证和清理

使用说明

文件命名规范

翻译文件需要在文件名中包含语言代码，例如：

• messages.en-US.ts
• translations.zh-CN.json

支持的语言代码格式：

• xx-XX（如：zh-CN, en-US）
• xx-xx（如：zh-cn, en-us）

Cursor 配置说明

1. 打开 Cursor 设置
2. 找到 MCP 配置
3. 根据操作系统输入对应配置：

Mac 用户

{
  "mcpServers": {
    "transform": {
      "command": "node",
      "args": [
        "{\"apiKey\":\"你的API密钥\",\"modelName\":\"你使用的模型名称\",\"baseUrl\":\"API基础URL\"}"
      ]
    }
  }
}

Windows 用户

{
  "mcpServers": {
    "transform": {
      "command": "cmd",
      "args": [
        "/c",
        "node",
        "{\"apiKey\":\"你的API密钥\",\"modelName\":\"你使用的模型名称\",\"baseUrl\":\"API基础URL\"}"
      ]
    }
  }
}

注意事项

1. 文件处理

   • 处理前会自动备份原文件
   • 备份文件格式：原文件名.backup.时间戳.扩展名
   • 跳过不符合命名规范的文件

2. 错误处理

   • 单个文件处理失败不会影响其他文件
   • 控制台会输出处理状态和错误信息

3. 性能优化

   • 大文件会自动分块处理
   • 默认每块最大处理 500 个键值对
   • 并发处理多个文件提高效率

4. 路径问题

   • 请使用绝对路径，相对路径会查找代码执行位置