MCP Translator Server

This project provides an MCP (Model Context Protocol) server designed to facilitate the translation of long-form text documents.

Features

• Accepts paths to multiple text files for concurrent translation jobs.
• Splits each text file into manageable chunks.
• Stores chunks and their translation status in a SQLite database (translation_progress.db), keeping track of which file each chunk belongs to.
• Provides tools to:
  • Initialize or restart the translation process for a specific file (start_translation file_path=...). Clears previous progress for that file if it existed.
  • List all ongoing translation jobs and their progress (list_jobs).
  • Fetch the next untranslated chunk for a specific file (get_next_chunk file_path=...).
  • Submit the translation for a specific chunk ID (submit_translation chunk_id=... translated_text=...).

Setup

1. Install dependencies:

   pip install .
   # Or pip install -e . for development mode
   

2. Download the spaCy model:

   python -m spacy download xx_sent_ud_sm
   

Running the Server

mcp-translator

This will start the MCP server, listening for client connections. The database file (translation_progress.db) will be created in the same directory as server.py.

Usage

Connect an MCP client (like a compatible AI assistant or script) to the running server. The client can then use the available tools, making sure to specify the file_path where required (start_translation, get_next_chunk), to manage the translation workflow for different files.