git-blame.nvim VS treesitter-unit

Sure, for the blame message I'm using git-blame.nvim

Two years ago this day, I made the initial commit to git-blame.nvim. And it turned out to be an important thing in my life. The plugin grew from a little thing I needed to something lots of people use (and love ^_^).

I use git-blame.nvim and I think it performs very well. Would you mind comparing blame_line.nvim with git-blame.nvim?

Plugin author: https://github.com/f-person/git-blame.nvim

I'm the author of git-blame.nvim, it's a simple plugin that displays blame info in virtual text. There's an open issue for displaying the blame info in the status bar instead of virtual text.

Thanks for the reply. I need it for git-blame.nvim. In fact, what you said is exactly what I do in git-blame.nvim – I get all lines of current buffer, concatinate them with `\n` and pass to `git blame`. The problem is, the plugin gives incorrect output for last lines of files that have no ending. You can see this issue for more details.

It's also relatively easy to write plugins based on the parsed AST. I wrote treesitter-unit[1] making it easy to select/modify the subtree of the selection.

[1]: https://github.com/David-Kunz/treesitter-unit

I built a plugin to select units of tree-sitter objects https://github.com/David-Kunz/treesitter-unit

To delete/yank the node, the combination would be `diu`, using this plugin: https://github.com/David-Kunz/treesitter-unit

Hi, I also think treesitter-textobjects are a bit overspecific, that's why I created https://github.com/David-Kunz/treesitter-unit/ .

In this video I create a Neovim plugin called 'treesitter-unit' using Lua. You can find the final plugin here: https://github.com/David-Kunz/treesitter-unit/

I tried it with operator mode maps and it works perfectly! I updated the README to include: vim.api.nvim_set_keymap('o', 'x', ':lua require"treesitter-unit".select()', {noremap=true})