feat(gear_sonic_deploy): Add Python network setup script for Unitree G1 robot

[klakhi]

Summary

Add gear_sonic_deploy/scripts/setup_network.py — a Python script for configuring a static IP on the workstation NIC to communicate with the Unitree G1 robot over the 192.168.123.x subnet.

Motivation

The existing setup scripts are shell-based and lack:

• Interactive interface discovery (lists all physical NICs with their state and current IP)
• Scoped firewall handling — adds a targeted ufw rule per interface instead of disabling the firewall globally
• Pre-change config backup to /tmp/network_backup_<timestamp>.txt
• Post-setup connectivity ping to verify the robot is reachable
• A --interface flag for non-interactive/scripted use

What this script does

1. Robot selection — interactive menu (currently supports Unitree G1; extensible via ROBOT_CONFIGURATIONS dict)
2. Interface discovery — enumerates physical NICs (skips loopback, Docker, veth, bridge interfaces), shows UP/DOWN state and current IP
3. Config backup — saves existing IP/route state before making changes
4. Static IP assignment — flushes existing addresses, brings the interface up, assigns 192.168.123.222/24 via iproute2
5. Firewall — if ufw is active, adds scoped allow in/out on <interface> rules (does not disable ufw globally)
6. Connectivity check — pings the robot at 192.168.123.164 to confirm end-to-end reachability

Usage

# Interactive (prompts for robot type + interface)
python3 gear_sonic_deploy/scripts/setup_network.py

# Skip interface selection
python3 gear_sonic_deploy/scripts/setup_network.py -i enp5s0

Requirements

• Linux (Ubuntu 20.04 / 22.04 / 24.04)
• iproute2, sudo privileges
• ufw optional — skipped gracefully if not present

Files changed

• gear_sonic_deploy/scripts/setup_network.py (new)

Branch

klakhi/setup_network_script on klakhi/GR00T-WholeBodyControl

Fixed by #238

[KevinMi2023p]

The open-source version, here, rely on manual/static network setup. Feel free to ping more reviewers if you want this to get merged.

[klakhi]

@jsepulveda-nvidia please help merge, thanks