dcs-jupyter 0.1.6

Jupyter kernel for live connection to DCS (Digital Combat Simulator) scripting environment

DCS Jupyter

A Jupyter kernel for live connection to DCS (Digital Combat Simulator) scripting environment. Execute Lua code in Jupyter notebooks that communicates directly with a running DCS mission via UDP socket connection.

Features

• Execute Lua code directly in DCS from Jupyter notebooks
• Jupyter Magic Commands (%dcs and %%dcs) for seamless integration with Python notebooks (demo notebook)
• Standalone DcsConnection class for use in Python scripts outside Jupyter
• Real-time communication with running DCS missions
• Support for both Jupyter Console and JupyterLab
• Easy installation and kernel management
• Cross-platform support (Windows, Linux, macOS)

Installation

Requirements

• Python 3.11 or higher
• DCS World (Digital Combat Simulator)

Install Dependencies (Windows)

If you don't have Python and pipx installed:

1. Install Python:

   • Download Python 3.11+ from python.org
   • During installation, check "Add Python to PATH"
   • Verify installation in Command Prompt: python --version

2. Install pipx: Open Command Prompt and run:

   python -m pip install --user pipx
   python -m pipx ensurepath
   

   • Restart Command Prompt after installation
   • Verify installation: pipx --version

Install Package

# Basic installation
pipx install dcs-jupyter

# With JupyterLab and extras (vim bindings, rich output, Dracula theme)
pipx install "dcs-jupyter[lab]"

Also available via pip or uv package managers.

Basic Setup (Windows)

1. Patch DCS

In dcs_install_dir/Scripts/MissionScripting.lua, find this section (around line 16):

do
  sanitizeModule('os')
  sanitizeModule('io')
  sanitizeModule('lfs')
  _G['require'] = nil
  _G['loadlib'] = nil
  _G['package'] = nil
end

Add this code BEFORE the above do block:

-- Initialize DCS Jupyter before security restrictions
if not _G['dcs_jupyter_udp'] then
    package.path = package.path..";.\\LuaSocket\\?.lua"
    package.cpath = package.cpath..";.\\LuaSocket\\?.dll"
    local socket = require("socket")
    local udp = assert(socket.udp())
    assert(udp:settimeout(0))
    assert(udp:setsockname("127.0.0.1", 8042))
    _G['dcs_jupyter_udp'] = udp
    
    -- Pre-store DCS user directory for script loading
    local lfs = require("lfs")
    _G['dcs_jupyter_user_dir'] = lfs.writedir()
end

Leave the original do block unchanged. This maintains DCS security while enabling the kernel connection.

2. Install DCS Plugin

Simple method (recommended for most users):

1. Download the jupyter_kernel_connection.lua file
2. Open Mission Editor → Triggers → New Trigger
3. Set Event: Mission Start, Action: Do Script
4. Copy the entire file contents into the text box

That's it! Your mission now has Jupyter support.

🔧 Alternative installation methods (click to expand)

Method B: Do Script File action

• Place jupyter_kernel_connection.lua anywhere
• Mission Editor → Triggers → New Trigger → Do Script File
• Select the script file
• Note: File is saved internally to mission

Method C: Dynamic loading (advanced)

• Place script in your DCS user directory (under 'Saved Games')
• Use: assert(loadfile(_G['dcs_jupyter_user_dir'] .. "/Scripts/jupyter_kernel_connection.lua"))()
• Advantage: Reloads script on every mission restart

3. Configure Network Settings

The kernel uses localhost (127.0.0.1:8042) by default, which works for native Windows installations.

Usage

Quick Start with Console

dcs-jupyter-console

Automatically installs the DCS kernel if needed and starts Jupyter Console with DCS Lua kernel.

Start with JupyterLab

dcs-jupyter-lab

Automatically installs the DCS kernel if needed and starts JupyterLab.

Examples

Download Example Notebooks

Get started with these example notebooks:

• Jupyter Magic Commands Demo - Using %dcs and %%dcs magic in Python notebooks
  • Download | View with outputs
• Airport Terrain Demo - Explore DCS terrain and airfield data
  • Download | View with outputs
• Vehicle Control - Control units and vehicles in DCS
  • Download | View with outputs

Right-click "Download" links and select "Save link as..." to save to your computer.

Basic Usage

Python Notebook Usage

Load the magic extension in your Python notebook:

%load_ext dcs_jupyter.magic
# or to reload: %reload_ext dcs_jupyter.magic

Then use magic commands to execute Lua code in DCS:

# Single line magic command
%dcs trigger.action.outText("Hello from magic command!", 5)

# Multi-line magic command
%%dcs
local currentTime = timer.getTime()
trigger.action.outText("Mission time: " .. math.floor(currentTime) .. " seconds", 5)

DCS Lua Kernel Usage

When using the DCS Lua kernel directly, execute Lua code:

-- Display message in DCS (visible in game)
trigger.action.outText("Hello from DCS Jupyter!", 10)

-- Get current mission time and display it
local currentTime = timer.getTime()
trigger.action.outText("Mission time: " .. math.floor(currentTime) .. " seconds", 5)

-- Create smoke at a specific location
local smokePos = {x = 0, y = 0, z = 0}  -- Adjust coordinates as needed
trigger.action.smoke(smokePos, trigger.smokeColor.Red)

-- Get all red coalition groups
local redGroups = coalition.getGroups(coalition.side.RED)
trigger.action.outText("Red coalition has " .. #redGroups .. " groups", 8)

Return current mission time:

return timer.getTime()

Get structured data for analysis:

-- Get all airfields with their properties
local airfields = {}
for coalition_id = 0, 2 do
    local airfield_list = coalition.getAirbases(coalition_id)
    for _, airfield in pairs(airfield_list) do
        airfields[#airfields + 1] = {
            name = airfield:getName(),
            coalition = coalition_id,
            position = airfield:getPosition().p,
            category = airfield:getDesc().category
        }
    end
end
return net.lua2json(airfields)

Configuration

Default settings: 127.0.0.1:8042, 10 second timeout. See Advanced Setup for customization.

Troubleshooting

Connection Issues

• Ensure DCS is running with the Lua plugin loaded
• Check firewall settings allow UDP traffic on port 8042
• Verify IP address is correct (127.0.0.1 for Windows, host IP for WSL2)

Kernel Not Found

If the kernel isn't recognized:

# Reinstall kernel
jupyter kernelspec uninstall dcs-lua
dcs-jupyter-console  # Will reinstall automatically

WSL2 Connection Problems

• WSL2 users need to modify IP addresses manually (see Advanced Setup)
• Use Windows host IP address instead of localhost
• Consider using native Windows installation for simpler setup

Advanced Setup

WSL2 Users

The default localhost configuration should work for most WSL2 setups, especially with mirrored networking mode (WSL2 default in newer versions). If you experience connection issues, you may need to use your WSL2 IP address:

1. Find your WSL2 IP address: ip addr - look for the inet address on your main interface (usually eth0)
   • Note: This is primarily for mirrored networking mode. In older NAT mode, the networking setup may differ.
2. Edit IP in the patching code: Replace 127.0.0.1 with your WSL2 IP address in the MissionScripting.lua patch above
3. Edit IP in kernel: src/dcs_jupyter/kernel.py: SOCKET_ADDR = 'YOUR_WSL2_IP'
4. Configure Windows Firewall for UDP port 8042 (might be needed)

Manual Kernel Installation

# Install kernel manually
jupyter kernelspec install kernel_spec/dcs-lua --user

# List installed kernels
jupyter kernelspec list

# Start Jupyter with DCS kernel
jupyter console --kernel=dcs-lua

Custom IP/Port Configuration

Edit SOCKET_ADDR and PORT constants in both src/dcs_jupyter/kernel.py and dcs_plugin/jupyter_kernel_connection.lua with identical values.

License

MIT License - see LICENSE file for details.

Support

For issues and questions:

• Check the troubleshooting section above
• Open an issue on GitHub with detailed information about your problem