Adding a clue vs mapping to ’<Nop>’

[dhemery]

Contributing guidelines

• I have read CODE_OF_CONDUCT.md

Module(s)

mini.clue

Question

Suppose I’ve added ‘<Leader>’ as a trigger for mini.clue, and I want a sub-menu for ‘<Leader>x’.

Up until now, I've been adding that menu by adding a ‘<Leader>x’ clue in my mini.clue config.

As far as I can tell, I can get the same effect by mapping ‘<Leader>x’ to ‘<Nop>’.

Either way, when I type ‘<Leader>’ the x menu appears in the clue window.

Is that correct? Is there some additional benefit to adding a clue vs mapping the keys to '<Nop>'?

[echasnovski]

The main goal of 'mini.clue' is to show all available sequences of keys that actually do something, usually via available mappings.

To be more useful for nested mappings, after only small common part is typed 'mini.clue' also shows something to indicate that there are more keys under particular keys. For example, if there are <Leader>ff and <Leader>fh mappings, after pressing <Leader> there should be an entry for key f indicating that those two mappings are available. Those entries are generally refered to as "group next key".

There doesn't have to be anything extra done for it to be shown, especially no need to create a <Nop> mapping. But if you want to adjust the description of this group to be more helpful, adding a dedicated clue is the recommended approach.

Does this make sense?

[dhemery]

It makes sense. And it shows that I omitted an important part of my question: I know that the clue adds a description to the mapping group. I'm wondering whether there are any important differences between using a clue to add the description and using a <Nop> mapping to add the description.

As far as I can tell, once I’ve set up <Leader> as a trigger, the only reason to add a clue is to give a description to a mapping group, such as <Leader>f, like this:

clues = {
	{ mode = 'n', keys = '<Leader>f', desc = '+find'  },
}

I can get the same effect by creating a mapping to name the group. That is, mapping from <Leader>f to <Nop> and including a description in the mapping, like this:

vim.keymap.set('n', '<Leader>f', '<Nop>', { desc = '+find' } )

Adding the group descriptions via mappings has a nice benefit that I can name the groups in the same place where I define all of the mappings that have actions. Like this:

vim.keymap.set('n', '<Leader>f', '<Nop>', { desc = '+find' } )
vim.keymap.set('n', '<Leader>ff', '<Cmd>Pick files<CR>', { desc = 'find files' } )

If I use a clue to add the description, I define the mappings in one place, and the clues in another (the mini.clue config).

I can use either a clue or a <Nop> mapping to name the mapping group. As far as I can tell, the effect is the same either way.

What I don’t know: Are there important differences between the two mechanisms?

[echasnovski]

What I don’t know: Are there important differences between the two mechanisms?

The 'mini.clue' was not initially designed to handle the case of "create a <Nop> mapping to set a group's name". If it works right now, then it is totally by coincidence. It also will probably keep working, unless some major and/or tricky rewrite will be done, which is not planned.

It is possible to add test for this case, but I think suggesting a dedicated clue is good enough here.