MRU and frecency ranking

[YodaEmbedding]

It would be nice to include MRU or frecency based rankings into the file sorting on a per-root-directory basis. This way, more frequently or recently accessed files are given higher priority, provided they meet some minimum threshold of fuzzy text matching score. This should have the effect of reducing characters typed to access common files, and also have them immediately visible in the list to remind the user which files they typically want to open. Perhaps the frecency filter could be disabled for inputs of >2 characters to ensure never-before-accessed files can also be easily accessed via the default search algorithms.

[zetashift]

I like how telescope did this by using a plugin: https://github.com/nvim-telescope/telescope-frecency.nvim

I'd imagine something similiar for snap would be the best.