Implement image culling capabilities.

[Ravennoir]

I see potential here to use the viewer as fast culling software to quickly find the best images from a image collection.

I have build some of the features needed for this, keeping the design philosophy of speed and simplicity here. If this would align with your goals I would like to contribute it.

Disclaimer: I am not a Rust developer and have build it with AI. Since I am aware of the discussion about code quality I want to ask upfront if you would be willing to accept contributions and state the requirements (eg. prompts used, tests, etc).

Features I have added:

• Keyboard first design philosophy: Shortcuts for all actions (exception: middle mouse does reset zoom)
• Multiimage display (Adding, removing images)
  • Re-ordering of images in multiimage view
  • Pinning reference image
• Zooming:
  • Zooming out is capped
  • Synchronised zoom across multiple images while in multi-image display mode
• Culling mode to filter by star rating
• Add star rating to file (currently writes to file, but I would like to consider sidecar files)
• Using embedded jpgs from Raw files (NEF) for faster display (not displaying the actual raw file)

So let me know if this would align with your goals

[dividebysandwich]

Hi,

Those look like some nice features! The only thing I'm a little worried about is the star ratings - sidecar files can be problematic and/or annoying, and at the moment lightningview doesn't store any additional information anywhere. My recommendation would be to use a crate that uses whatever user data storage the operating system uses (~/.local on linux for example).

I'll be happy to accept pull requests. I will review any pull request anyway.