v0.2.0 — Native matplotlib viewer and campt coordinate conversion

What's New

Native Matplotlib Footprint Viewer (--win)

A lightweight alternative to the browser-based Panel/Bokeh viewer. Run:

isistools footprints image_list.lis --win
isistools footprints image_list.lis --cnet network.net --win

Opens a native window instantly — no browser, no server. Features:

• Per-filename footprint colors with hover tooltips showing CTX product ID
• Right-side legend with 18-char product IDs
• Control point overlay: registered (black ×), unregistered (red ●), ignored (gray ●)
• Built-in zoom/pan/home toolbar
• Window centered on screen (Qt and Tk backends)

Precise campt Coordinate Conversion

Pre-jigsaw control networks without ground coordinates are now converted from pixel (sample/line) to lon/lat using ISIS campt in batch mode via kalasiris. This provides camera-model-accurate positions instead of approximate footprint-based estimation.

Faster Startup

Holoviews/hvplot/panel imports are now lazy in the plotting layer. The --win path loads only matplotlib and lightweight dependencies, avoiding the heavy browser-plotting stack entirely.

Other Changes

• Serial lookup utilities moved from apps.tiepoint_review to io.cubes for clean reuse
• Fixed latent NameError in mosaic_review.py
• Fixed FileNotFoundError caused by transitive holoviews import on the matplotlib path
• New dependencies: matplotlib, mplcursors, kalasiris, PyQt6

Full Changelog: v0.1.1...v0.2.0