Is it possible to use a different markdown render for the chat?

[Chrispycode]

Hi I had some issues with render-markdown.nvim and didn't like the default config so I switched to markview.nvim.
I tried this config but obviously didn't work

{
    'olimorris/codecompanion.nvim',
    dependencies = {
      'nvim-treesitter/nvim-treesitter',
      { "nvim-lua/plenary.nvim",  branch = "master" },
      { 'OXY2DEV/markview.nvim',  ft = { "markdown", "codecompanion" } },
      { 'stevearc/dressing.nvim', opts = {} }, -- Optional: Improves `vim.ui.select`
    }
}

Any ideas?

[olimorris]

Assuming that's how you set up markview.nvim, it should work out of the box. Presume CodeCompanion is working for you?

[Chrispycode]

Sorry, the markview setup was incomplete but it still does not work here is my completion config

{
    'olimorris/codecompanion.nvim',
    dependencies = {
      'nvim-treesitter/nvim-treesitter',
      { "nvim-lua/plenary.nvim",  branch = "master" },
      {
        'OXY2DEV/markview.nvim',
        dependencies = {
          'nvim-treesitter/nvim-treesitter',
          'nvim-tree/nvim-web-devicons',
        },
        lazy = false,
        ft = { "markdown", "codecompanion" }
      },
      { 'stevearc/dressing.nvim', opts = {} }, -- Optional: Improves `vim.ui.select`
    },
    config = function()
      require('codecompanion').setup {
        strategies = {
          chat = {
            adapter = os.getenv 'LLM',
            keymaps = {
              debug = {
                modes = {
                  n = 'cd',
                },
                index = 14,
                callback = 'keymaps.debug',
                description = 'View debug info',
              },
            },
          },
          inline = {
            adapter = os.getenv 'LLM',
          },
        },
        adapters = {
          ollama = function()
            return require('codecompanion.adapters').extend('ollama', {
              schema = {
                num_ctx = {
                  default = 16384,
                },
                temperature = { default = 0.6 },
              },
            })
          end,
          copilot = function()
            return require('codecompanion.adapters').extend('copilot', {
              schema = {
                -- model = {
                --   default = 'gpt-4o-2024-05-13',
                -- },
                temperature = { default = 0.6 },
              },
            })
          end,
        },
      }
    end,
  }

And here is the completion result

[olimorris]

I'd advise working back from the minimal.lua file.

Maybe replace:

config = function()
  require('codecompanion').setup {
    -- Your config
  }
end,

with:

opts = {
    -- Your config
},

[Chrispycode]

thank you, I didn't see that but the result is the same I also removed some obsolete stuff

[olimorris]

Reading the config options, it seems to suggest you need to enable the plugin on specific filetypes. So something like:

filetypes = { "markdown", "quarto", "rmd", "codecompanion" },

[Chrispycode]

Also does not work so it must be a Issue with markview thank you for the help

[olimorris]

This should work:

{
  "OXY2DEV/markview.nvim",
  ft = { "markdown", "codecompanion" },
  opts = {
    filetypes = { "markdown", "codecompanion" },
    buf_ignore = {},
  },
},

[mathjiajia]

Thank you. Yes, it works.
BTW, is it possible to set a non-trivial buftype for codecompanion, eg `prompt'?

[olimorris]

It is possible but not something I'll implement in the plugin. The chat buffer is intended to be a temporary buffer and as such nofile is the correct buffer type.

You could use the CodeCompanionChatOpened event to change the buftype yourself.