Feature Requests: New apis and configuring the way output is displayed for python

[benzhou1]

I really like the plugin and would like to request the following possible features:

1. Api to cancel existing run.

• The use case is sometimes you write bad code which can do bad things especially in a scratch file.

2. Api to clear existing output in the buffer or maybe clear output before running

• I know we can just undo the output or just manually delete the output, but an api would be nice.

3. A way to configure how the output looks like. Specifically the equal signs that surround the output when running a python scratch file

• An idea I have is to put the output into a comment block instead. The reason for this is so we can avoid the lsp complaining about the syntax and if you forget to clear the output your next run will fail as well.

[YaroSpace]

Great to hear you like the plugin!
I will gladly look into your requests. Let me think about it and come back to you.

As for #3, you can configure the output with custom formatter: https://github.com/YaroSpace/lua-console.nvim#setting-up

opts = {
  external_evaluators = {
    python = {
      cmd = { 'python3', '-c' },
      env = { PYTHONPATH = '~/projects' },
      timeout = 100000,
      formatter = function(result) 
           do_something; 
         return result 
      end,
    },
  }
}

You can have a look at default formatter here:

lua-console.nvim/lua/lua-console/exev_config.lua

Line 4 in 2efa0de

local function generic_formatter(result)

local function generic_formatter(result)
  local width = vim.o.columns
  local sep_start = ('='):rep(width)
  local sep_end = ('='):rep(width)

  table.insert(result, 1, sep_start)
  table.insert(result, sep_end)

  return result
end

[benzhou1]

thanks I will check out the formatter.

[YaroSpace]

Regarding other points:

1. Is clear, I will add a key binding and an api to query/stop running execution process.
2. Please give more details about clearing the buffer. How do you see the use case?

[benzhou1]

For #2 I would first:

1. Run the entire scratch file and output will be appended to the buffer.
2. Lets says I see a bug in the output and make a fix to the scratch file.
3. I try to run the entire file again and realized I didin't clear the output from the previous run. Now the run will fail because the output that was appended is not valid code.
4. I have to now go to the bottom of the buffer and delete the output before I can run the entire scratch file again.

So if there was an api to clear the output from the buffer, I would clear the output with the api before running the scratch file again. I can can easily create a key map to do this.

For me, I would actually always want the output cleared before starting a new run because I would rarely need to see the output of a previous run. If I don't want to clear the output before a run then I can just create a separate key map for that.

[YaroSpace]

How do you usually run the code, key map to execute the whole buffer or visually select lines?

We need somehow to determine where to clear the buffer from, since you code is in the same buffer.

[benzhou1]

Usually whole buffer. I sometimes use visual but its annoying to keep having to do the selection and plus you have to make sure to select all the context you need to run the selection.

I believe right now you are using result prefix, maybe you can do something with that. Maybe a result prefix and suffix and then you can clear everything in between? You might need a more unique prefix/suffix to avoid removing actual code.

[YaroSpace]

Updated in #12. Try it out.

[benzhou1]

looks good thanks for this!