PO Translation MCP Server
Enables AI-powered translation of Django gettext .po files while intelligently preserving variables, HTML tags, URLs, and JavaScript code with validation and automated updates.
README
PO Translation MCP Server
MCP server for translating Django gettext .po files with AI assistance, featuring intelligent preservation of variables, HTML tags, URLs, and JavaScript code.
✨ Features
- ✅ analyze_po_file: Analyze .po files with statistics and line numbers
- ✅ validate_and_update_po_file: Combined validation + update in one operation
- 🤖 Claude-Powered Translation: Leverages Claude's native translation capabilities
- 🔄 Variable Preservation: Python format strings, Django templates
- 🏷️ HTML Preservation: Keep all HTML tags intact
- 🌐 URL Preservation: URLs never get translated
- 💻 JavaScript Preservation: Smart code detection (not natural language patterns)
- 📍 Line Number Tracking: AI can locate entries precisely in source files
🚀 Installation
npm install
npm run build
📖 Usage with Claude Code
Add to your Claude Code MCP settings:
{
"mcpServers": {
"po-translation": {
"command": "node",
"args": ["/path/to/po-mcp/dist/index.js"]
}
}
}
🔧 MCP Tools
1. analyze_po_file
Analyze a .po file and get translation statistics with line numbers:
{
"po_file_path": "/absolute/path/to/locale/tr/LC_MESSAGES/django.po"
}
Output:
{
"statistics": {
"translated": 2876,
"untranslated": 0,
"fuzzy": 2,
"total": 2878
},
"untranslated_entries": [],
"fuzzy_entries": [
{
"msgid": "Departure Place ID",
"msgstr": "Kalkış Tarihi",
"reference": "transfer/models/booking/option.py:46",
"line_number": 12941,
"context": null,
"flags": ["fuzzy"]
}
],
"file_info": {
"path": "/absolute/path/to/django.po",
"locale": "tr",
"last_modified": "2025-10-20T15:18:10.591Z"
}
}
2. validate_and_update_po_file
Combined validation and update in one operation - validates translations and automatically updates the PO file if valid:
{
"po_file_path": "/absolute/path/to/locale/tr/LC_MESSAGES/django.po",
"translations": [
{
"msgid": "Optional Reservation (No Payment)",
"msgstr_original": "",
"msgstr_translated": "Opsiyonel Rezervasyon (Ödemesiz)",
"line_number": 2767,
"reference": "models.py:123",
"skipped": false,
"skip_reason": null
}
],
"strict_mode": true,
"check_variables": true,
"check_html": true,
"check_urls": true,
"check_javascript": true,
"check_length": false,
"dry_run": false,
"force_update": false
}
Output:
{
"validation": {
"validation_results": [
{
"msgid": "Optional Reservation (No Payment)",
"msgstr": "Opsiyonel Rezervasyon (Ödemesiz)",
"valid": true,
"issues": [],
"warnings": []
}
],
"summary": {
"total": 1,
"valid": 1,
"invalid": 0,
"has_warnings": 0
},
"overall_valid": true
},
"update": {
"success": true,
"updated_entries": 1,
"skipped_entries": 0,
"file_path": "/absolute/path/to/django.po",
"git_diff_preview": "1 entries updated",
"errors": []
},
"message": "✅ Successfully validated and updated 1 translations."
}
Options:
dry_run: true- Preview changes without writingforce_update: true- Update even if some translations are invalidstrict_mode: true- Treat warnings as errors
🔄 Complete Translation Workflow
With Claude Code
# 1. Analyze: Check what needs translation
analyze_po_file({
po_file_path: "/absolute/path/to/locale/tr/LC_MESSAGES/django.po"
})
# → Returns untranslated/fuzzy entries with line numbers
# 2. Translate: Claude translates the entries natively
# Claude receives the msgid entries and translates them directly
# (No separate translate_entries tool needed - Claude does this)
# Example conversation:
# User: "Translate these 3 untranslated entries to Turkish"
# Claude:
# 1. "Intermediate Stops" → "Ara Duraklar"
# 2. "No intermediate stops" → "Ara durak yok"
# 3. "Click 'Trip Details' to see stops" → "Durakları görmek için 'Sefer Detayları'na tıklayın"
# 3. Validate & Update: One operation validates and updates
validate_and_update_po_file({
po_file_path: "/absolute/path/to/locale/tr/LC_MESSAGES/django.po",
translations: [
{
"msgid": "Intermediate Stops",
"msgstr_original": "",
"msgstr_translated": "Ara Duraklar",
"line_number": 2767,
"reference": "models.py:123",
"skipped": false,
"skip_reason": null
}
],
dry_run: false // Set to true to preview first
})
# → Validates quality + updates PO file in one step
Why No translate_entries Tool?
This MCP server leverages Claude's native translation capabilities. Instead of implementing a separate translation API:
- analyze_po_file finds entries that need translation
- Claude translates them directly using its language understanding
- validate_and_update_po_file ensures quality (variables, HTML, URLs preserved) AND updates the file
This approach is:
- ✅ More accurate: Claude understands context better than machine translation
- ✅ Simpler: No API keys or external services needed
- ✅ Flexible: Claude can handle edge cases and ask for clarification
- ✅ Cost-effective: No separate translation API costs
- ✅ Efficient: Combined validation + update in one step
🧪 Testing
# Test analyzer with line numbers
node test-analyzer.js
# Test pattern protection (variables, HTML, URLs, JS)
node test-protector.js
Test Output:
📊 Statistics:
Total: 2878
Translated: 2876
Untranslated: 0
Fuzzy: 2
⚠️ Fuzzy Entries:
1. "Departure Place ID"
Current: "Kalkış Tarihi"
Line: 12941 | Reference: transfer/models/booking/option.py:46
✅ Test passed!
🛡️ Pattern Preservation
The server preserves 12+ pattern types:
Python Variables:
%(name)s- Python old-style format{0},{variable}- Python format strings{{ variable }}- Django template variables
Django Tags:
{% tag %}- Django template tags{# comment #}- Django comments
HTML Tags:
<tag>,</tag>- Opening/closing tags<tag attr="value">- Tags with attributes
URLs:
https://example.com- Full URLs with protocolwww.example.com- URLs without protocol/path/to/resource- Relative URLs
JavaScript Code (Smart Detection):
console.log("text")- String parametersmodule.method()- Dot notation callsgetData('value')- Quote-containing parameters- ❌ NOT
Reservation (No Payment)- Natural language (false positives avoided)
📁 Project Structure
po-mcp/
├── src/
│ ├── index.ts # MCP server entry point (2 tools)
│ ├── types.ts # TypeScript interfaces
│ ├── tools/
│ │ ├── analyze.ts # analyze_po_file tool
│ │ ├── validate-and-update.ts # validate_and_update_po_file tool
│ │ ├── validate.ts # Internal validation logic
│ │ └── update.ts # Internal update logic
│ └── parser/
│ ├── po-parser.ts # PO parsing with line numbers
│ └── gettext-types.d.ts # Custom type definitions
├── dist/ # Compiled JavaScript
├── locale/ # Test PO files
├── test-analyzer.js # Analyzer test script
├── package.json
├── tsconfig.json
└── README.md # This file
💻 Development
# Build
npm run build
# Watch mode
npm run watch
# Development mode
npm run dev
📦 Dependencies
{
"dependencies": {
"@modelcontextprotocol/sdk": "^0.5.0",
"gettext-parser": "^8.0.0",
"glob": "^10.3.0"
}
}
✅ Implementation Status
Completed Tools:
- ✅ analyze_po_file (with line number tracking)
- ✅ validate_and_update_po_file (combined validation + update)
Key Improvements:
- ✅ Fixed JavaScript pattern detection (no false positives on natural language)
- ✅ Simplified API: 2 tools instead of 4
- ✅ One-step workflow: validate + update in single operation
Test Status:
- ✅ Build: No errors
- ✅ analyze_po_file: Tested with 2878 entries
- ✅ Pattern detection: Smart JS detection working
- ✅ Line number tracking: Accurate line mapping
Ready for: Production use with Claude Code
📊 Statistics
- Total Files: 8 TypeScript source files
- MCP Tools: 2 (analyze + validate_and_update)
- Pattern Types: 12+ preserved (smart JS detection)
- Build Status: ✅ Success
🙏 License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.