LSP (otter.nvim) issues when opening ipynb files

[life00]

Hi. I am using LazyVim with the following configuration:

Details

return {
  -- quarto-nvim
  -- for preview, LSP & suggestions integration, and code running
  {
    "quarto-dev/quarto-nvim",
    dependencies = {
      "jmbuhr/otter.nvim",
      "nvim-treesitter/nvim-treesitter",
    },
    ft = { "quarto" },
    keys = {
      {
        "<leader>cp",
        function()
          require("quarto").quartoPreview()
        end,
        desc = "Quarto preview",
      },
    },
    opts = {
      closePreviewOnExit = false,
      lsp_features = {
        languages = { "python", "r", "julia", "bash", "sql" },
        chunks = "all",
        diagnostics = {
          enabled = true,
          triggers = { "BufWritePost" },
        },
        completion = { enabled = true },
      },
      codeRunner = {
        enabled = true,
        default_method = "molten",
      },
    },
  },
  -- jupytext
  -- for converting ipynb files to quarto
  {
    "GCBallesteros/jupytext.nvim",
    event = "BufReadCmd *.ipynb",
    opts = {
      custom_language_formatting = {
        python = {
          extension = "qmd",
          style = "quarto",
          force_ft = "quarto",
        },
      },
    },
  },
  -- molten
  -- for running the code in jupyter environment
  {
    "benlubas/molten-nvim",
    ft = { "quarto" },
    version = "^1.0.0",
    build = ":UpdateRemotePlugins",
    init = function()
      vim.g.molten_auto_image_popup = true
      vim.g.molten_cover_empty_lines = true
      vim.g.molten_enter_output_behavior = "open_and_enter"
      vim.g.molten_image_provider = "none"
      vim.g.molten_wrap_output = true
      vim.g.molten_output_win_max_height = 32
      vim.g.molten_virt_text_max_lines = 8
      vim.g.molten_virt_text_output = true
    end,
    keys = {
      { "<leader>cji", ":MoltenInit<CR>", desc = "Molten initialize" },
      { "<leader>cjI", ":MoltenInfo<CR>", desc = "Molten info" },
      {
        "<leader>cjX",
        function()
          require("quarto.runner").run_all()
        end,
        desc = "Run all",
      },
      {
        "<leader>cjxk",
        function()
          require("quarto.runner").run_above()
        end,
        desc = "Run all above",
      },
      {
        "<leader>cjxj",
        function()
          require("quarto.runner").run_below()
        end,
        desc = "Run all below",
      },
      { "<leader>cjc", ":MoltenDelete<CR>", desc = "Clear active cell" },
      { "<leader>cjr", ":MoltenRestart<CR>", desc = "Restart Molten" },
      { "<leader>cjC", ":MoltenInterrupt<CR>", desc = "Cancel execution" },
      { "<leader>cjj", ":MoltenNext<CR>", desc = "Go to next cell" },
      { "<leader>cjk", ":MoltenPrev<CR>", desc = "Go to prev cell" },
      { "<leader>cjo", ":noautocmd MoltenEnterOutput<CR>", desc = "Enter output" },
      {
        "<C-cr>",
        function()
          if vim.bo.filetype == "quarto" then
            require("quarto.runner").run_cell()
          end
        end,
        desc = "Run cell",
      },
      { "<C-cr>", ":<C-u>MoltenEvaluateVisual<CR>gv", mode = "v", desc = "Run visual" },
      { "<leader>cjp", ":MoltenImagePopup<CR>", desc = "Preview image" },
      { "<leader>cjni", ":MoltenImportOutput<CR>", desc = "Import notebook output" },
      { "<leader>cjne", ":MoltenExportOutput<CR>", desc = "Export notebook output" },
    },
  },
  -- venv selector
  -- configure keybinding to also appear in quarto
  {
    "linux-cultist/venv-selector.nvim",
    cmd = "VenvSelect",
    ft = { "python", "quarto" },
    keys = { { "<leader>cv", "<cmd>:VenvSelect<cr>", desc = "Select VirtualEnv", ft = { "python", "quarto" } } },
  },
}

I managed to configure everything perfectly, however there is one remaining issue. When I open a raw ipynb file which is then automatically converted to qmd by jupytext.nvim I do not get LSP diagnostics or suggestions from otter.nvim. However, everything works when I open a raw qmd file.

I found a related discussion about this issue: jmbuhr/otter.nvim#154

I followed the same configuration as suggested in jmbuhr/otter.nvim#154 (comment) however it doesn't seem to work (see above config). I also tried manually activating it after I open the converted buffer using :lua require'otter'.activate() and it doesn't work for some reason as well.

If you have any ideas why it doesn't work for me, let me know. Thanks.