Add native Celigo image cytometer support

[rickwierenga]

Summary

Adds native PyLabRobot support for the Revvity Celigo image cytometer (formerly sold by Nexcelom/Cyntellect) without requiring the vendor application.

• Organizes the device under pylabrobot.revvity.celigo, with Celigo and CeligoConfig available from pylabrobot.revvity
• Loads complete per-instrument configuration through CeligoConfig.from_install()
• Communicates directly with the controller over FTDI USB
• Controls the Lumenera camera through its SDK
• Exposes axes, filter wheels, galvo, and laser as dedicated components
• Uses millimeters and seconds in public APIs, with controller-unit conversion kept internal
• Supports PyLabRobot plate resources through explicit celigo.set_plate()
• Provides high-level acquisition, image autofocus, auto-exposure, multichannel imaging, and galvo FOV scans
• Adds a Revvity manufacturer guide plus Celigo Hello World, Advanced Imaging, and Components/Diagnostics notebooks

Safety and reliability

• Encoder-verified movement and homing
• Configured motion limits, currents, and timeouts
• Typed controller and galvo status results
• Acquisition-scoped illumination with safe cleanup on success or failure
• Correct handling of inverted digital and analog outputs
• Serialized, time-bounded camera and controller communication
• Explicit laser opt-in with controller interlock checks
• USB-topology FTDI selection with kernel-driver detach/reattach support

Validation

• 142 Celigo tests pass
• Ruff passes
• mypy passes
• All Celigo notebooks parse and compile
• Celigo API autosummary builds with warnings treated as errors

Live hardware verification covers controller setup/status/self-test, XYZ and filter homing, drawer motion, galvo calibration and centering, native and calibrated camera capture, image autofocus, machine auto-exposure, all five configured illumination channels, camera-trigger diagnostics, and a 16-FOV galvo scan.

Hardware-displacement autofocus is not implemented. Laser firing and externally triggered frame acquisition have not been exercised.