Basic implementation of multi buffer editor

[dunkyp]

This is more of an RFC, it implements multiple buffers using a std::vector. There are a few changes I'd be interested in making but wanted to see what your preferred approach would be. For instance I'm assigning a numerical id to buffers, it'd probably be better to use a unique id (perhaps based on a uuid) and place the buffers into a map. I've also added a default scratch buffer, which I thought would be useful if the intent is to include a scheme interpreter in the editor.

[arximboldi]

Wow, that's cool! I wanted to add this eventually so it's nice that someone is thinking about it too. I am still on a road-trip so I can't do a very thorough review, here are a couple of things I just saw:

1. Instead of an std::vector I'd use an immer::vector (or maybe even an immer::flex_vector) but...
2. ...I agree that a map is better than a vector. Immer now has an immer::map type. I'd suggest using a std::filesystem::path as the key (probably wrapped in an immer::box) using the std::filesystem::canonical(...) form of the path (unless the file does not exist). Then the order could be dealt with explicitly by using a float member (when placing a buffer between two other buffers you just need to set the order to some value between the order of those).
3. From a style point of view, I am trying to use only free functions in the "data model" to keep things simple, so I'd change the current() method for a current_buffer free function.
4. The tricky part down this road is to add commands similar to find-file or switch-to-buffer which open an interactive prompt in the mini-buffer. I have some ideas on how to do this but I do not have time to elaborate now...