feat(opts)!: restructure mappings and set default picker opts in setup

[al-ce]

BREAKING CHANGES: In this PR we would restructure the mappings table in in the setup configuration to follow Telescope's structure ( ["key"] = function) for action mappings.

All actions are named functions so they'll have names in the Telescope 'which key' popup.

Changes:

• The configuration for the Telescope picker is given some defaults that apply to the Hbac telescopecommand as well as a call to hbac.telescope(). Any opts added to the user's setup call or passed to the function call deep-extend into the config defaults
• Some require calls are moved around in hbac.command.actions, hbac.command.subcommands, and hbac.utils to prevent circular dependencies with the mappings in the setup configuration for the Telescope picker
• attach_mappings module is renamed to actions. It holds the definitions for the named functions that refresh the picker
• the attach_mappings function definition is moved to hbac.telescope and can accept an opts table so that the user can pass their own mappings configuration as they would to a builtin Telescope picker
• a minor fix to the ordinal value for the results. This ignores the absolute filepath above a project root for a cleaner fuzzy search (e.g. typing hu won't show /home/user/projectroot/main.c as a result)
• • README is updated with relevant instructions and examples

Example user config (extends default in hbac.config):

hbac.setup({
  telescope = {
    sort_lastused = false,  -- overrides default
    layout_strategy = "vertical",  -- additional opts
    prompt_prefix = "Hbac! ",
    mappings = {  -- extends default mappings
      i = {
        ["<CR>"] = telescope_actions.select_drop,
        ["<C-CR>"] = telescope_actions.select_default,
        ["<M-z>"] = function() print("hello") end,
      },
    },
    pin_picker = {
      -- etc.
    },
  },
})

example keymap for picker (extends user config):

vim.keymap.set("n", ",ht", function()
  require("hbac").telescope(
    {
      ignore_current_buffer = true,
      file_ignore_patterns = { ".json" },
      mappings = {
        i = {
          ["<CR>"] = telescope_actions.select_default,  -- overrides the mapping in user setup config
        },
      },
    }
  )
end, { desc = "Hbac: pin picker", noremap = true })

Addresses #11 and builds on #9 and #12 (partially addressed by f041ebabcd27aaadaebaccc9390edb56d06af731)

[axkirillov]

Looks good to me. I'll test it on my setup.

[axkirillov]

Did you decide to make the current buffer hidden by default after all?

[al-ce]

Did you decide to make the current buffer hidden by default after all?

Sorry I left that comment hanging! I think I'm on the same page with what you were saying, that we should not hide the current buffer (right?). Currently the defaults are

sort_mru = true,
sort_lastused = true,
selection_strategy = "row",

so the current buffer should be in the first result, selector/cursor should be on the second result which is the previously used buffer, and the rest are sorted by most recently used (if I've implemented the options correctly)

[axkirillov]

Did you decide to make the current buffer hidden by default after all?

Sorry I left that comment hanging! I think I'm on the same page with what you were saying, that we should not hide the current buffer (right?). Currently the defaults are

sort_mru = true,
sort_lastused = true,
selection_strategy = "row",

so the current buffer should be in the first result, selector/cursor should be on the second result which is the previously used buffer, and the rest are sorted by most recently used (if I've implemented the options correctly)

huh, weird, I think I have the current buffer hidden can you check? my setup function is empy, so I should be getting all default behaviours

[al-ce]

can you check? my setup function is empy, so I should be getting all default behaviours

I'm guilty of creating a million branches here, are you using the default_opts branch on my fork instead of the picker_opts branch from the last PR? (sorry I did that, I changed the merge message of the last PR and had a merge conflict so I created this new one) Otherwise I'll debug

[axkirillov]

can you check? my setup function is empy, so I should be getting all default behaviours

I'm guilty of creating a million branches here, are you using the default_opts branch on my fork instead of the picker_opts branch from the last PR? (sorry I did that, I changed the merge message of the last PR and had a merge conflict so I created this new one) Otherwise I'll debug

Yeah, I'm using this branch

    {
        'al-ce/hbac.nvim',
        config = function()
            require("hbac").setup()
        end,
        branch = "default_opts"
    },

[al-ce]

Does passing opts to the setup call or to the telescope function call change the behavior or does it throw any errors? Like

  {
    'al-ce/hbac.nvim',
    config = function()
      require("hbac").setup(
      {
        telescope = {
          sort_lastused = true,
        },
        }
    )

      vim.keymap.set("n", ",ht", function()
        require("hbac").telescope({
            sort_mru = true,
            mappings = {
              i = {
                ["<cr>"] = telescope_actions.select_drop,
              },
            },
          })
      end)

    end,
    branch = "default_opts"
  }

Just confirming whether it's just the default behavior or if any opts are registering

[axkirillov]

Does passing opts to the setup call or to the telescope function call change the behavior or does it throw any errors? Like

  {
    'al-ce/hbac.nvim',
    config = function()
      require("hbac").setup(
      {
        telescope = {
          sort_lastused = true,
        },
        }
    )

      vim.keymap.set("n", ",ht", function()
        require("hbac").telescope({
            sort_mru = true,
            mappings = {
              i = {
                ["<cr>"] = telescope_actions.select_drop,
              },
            },
          })
      end)

    end,
    branch = "default_opts"
  }

Just confirming whether it's just the default behavior or if any opts are registering

For me, even if I paste this configuration, the current buffer is hidden.

[al-ce]

Sorry for lack of updates, I've been traveling to visit family and I've been trying to figure this out in some downtime. I've tried a few different setups and I'm getting the behavior I'm expecting so I think it might be some mistake I made when I merged the last PR into this branch or I'm making some bad assumption about my machine. Sorry for the hold up and I'll keep trying, hope all is well

[axkirillov]

Sorry! I just realized I was passing my own opts 🤦🏻

vim.keymap.set("n", "<leader>b", function()
    local default_opts = {
        show_all_buffers = true,
        sort_mru = true,
        sort_lastused = true,
        ignore_current_buffer = true,
        cwd_only = false,
        cwd = nil,
    }
    require("hbac").telescope(default_opts)

[axkirillov]

I changed it to this and it works as expected

vim.keymap.set("n", "<leader>b", ":Hbac telescope<CR>", ns)

[al-ce]

I changed it to this and it works as expected

Ok that's good! I integrated your suggestions above and made a small refactor in the config module to not repeat the lines that assign the mappings, since they're the same for both. Is this ok?

local action_mappings = {
    ["<M-c>"] = actions.close_unpinned,
    ["<M-x>"] = actions.delete_buffer,
    ["<M-a>"] = actions.pin_all,
    ["<M-u>"] = actions.unpin_all,
    ["<M-y>"] = actions.toggle_pin,
}
local defaults = {
    telescope = {
        mappings = {
            n = action_mappings,
            i = action_mappings,
                }
        -- etc
}

[axkirillov]

Looks good and works for me.

[al-ce]

Hey, I'm back home from my travels. I'm good with the PR as it is if you're happy with it and ready to merge it, I think we've addressed the open issues it references. Will you do the merge into main when you're ready? I didn't want to push the new mapping config structure without knowing your procedures for announcements or versioning. Thanks again for the collaboration!