Future of the project

[michaelb]

As the original project developer, this is how i envision sniprun's future:

The following is already a reality, and I'd like it to stay that way

• Simple in utilisation and understanding
• Allow simple contribution (addition of new languages)
• Documentation effort
• Sniprun only runs the code it needs to run; no unecessary imports etc..

This would be nice to have:

• Passing arguments from vimscript to interpreters
• Optional dependencies (features?) (for in-development interpreters)
• Auto-install dependencies (but high risk to collide with the user's personnal config)
• Popularity (more contributors would be welcome)
• Test automation (WIP)
  -̶ ̶l̶e̶t̶ ̶c̶o̶n̶f̶i̶g̶u̶r̶a̶t̶i̶o̶n̶=̶o̶p̶t̶i̶o̶n̶s̶ ̶i̶n̶ ̶v̶i̶m̶ done
• integrated parsing framework (to help implementation of higher support levels)

Seeing how there are few satisfying alternatives (comment if i am wrong), especially for compiled languages, and how quickly i could develop this by myself * , I'd like to see sniprun become a popular plugin. Maybe the all-language partial tester? ? I can dream alright.

[denstiny]

I think if sniprun can help me to test the practicability of a function, he will become the mainstream plug-in

[denstiny]

[denstiny]

But can you tell me how to do

[michaelb]

That's in the plans and was envisaged from the start !!

I've done some conclusive tests (auto function fetching), but I need a stable treesitter interface (due for neovim 0.5) before I can really work on this with confidence.

At "File" support level, Sniprun would be run from only the printf line. It will then understand it needs the variables i and b, and thus search for the definition of i and b. Sniprun would also include (intelligently) everything it needs between (because side effects may occur like presently) the printf and the definition and then understand it needs the function Swap, and fetch the definition.

Partial "File" support may be a bit impredictable / inconsistent if not done right, so I won't do a half-assed release.
Some more people are interested (with reason) to see this happen, and I'm quite hyped myself, though the technical challenge will be great - you won't see this tomorrow.

[denstiny]

I'm looking forward to, but I can't write a vim plug-in, or you really want to really help you

[michaelb]

Don't worry about it. Not many people (incl. me) are confident in writing a lua treesitter interface + communication with the Rust code. Programming is learning, give me enough time and i'll figure it out.