How to disable Diagnostics underlining ?

[julioyildo]

Hello !

What I am trying to do...

How I would like to learn how to do it...

I would like to disable the Diagnostics feature and its red and orange underlining... Is there a way to do it?

How the project might keep the docs accurate...

Thank you !

[krassowski]

Do you wish to disable diagnostics completely (both the underlines and the diagnostics manager) or only the underlines?

A simple approach for now would be to take advantage of regular expression pattern matching intended to provide diagnostics filtering. You could use a catch-all regexp to hide all diagnostics:

{
    "ignoreMessagesPatterns": [".*"]
}

This will cause all linter messages to be ignored.

[bollwyvl]

So we should probably figure out a way to auto-generate that insight into the docs, so that this returns something:

https://jupyterlab-lsp.readthedocs.io/en/latest/search.html?q=ignoreMessagesPatterns

Back in the day (#89), we mentioned @adobe/jsonschema2md. Here's a very rough cut of it (looks like I need to revist with myst-nb!) on the starters docs. I'll take a look...

[bollwyvl]

Sorry to keep hijaaking this for docs stuff. but I gave sphinx-jsonschema another go over here and like how it uses the page. I think we use some inline markdown in our schema, whereas it speaks rst, so maybe I need to figure out how to get that in there.

[krassowski]

I believe myst speaks both markdown and rst so it should work (after a minimal modification). And it does look nice!

[bollwyvl]

Yeah, sphinx-jsonschema operates down at the docutils layer (buh) so it's a matter of digging out some stuff. Not my favorite place to be. It also makes poorly-behaved duplicate html ids, which I need to unwrap. But if we get it working (here or on lite) it's probably worth there being a plugin, I guess, unless I can get something to take it as a PR... Maintaining docs packages is also not my favorite place to be.

[krassowski]

Yeah, sphinx-jsonschema operates down at the docutils layer (buh)

Oh, I thought it is a bit higher-level. So https://github.com/adobe/jsonschema2md seems better in that case, right?

[julioyildo]

Hi,

First of all, sorry for the response time,

I thank you for your feedback. @krassowski The setting you shared with me worked ! Thanks !

In the same way is it possible to disable auto completion?
This one is a bit slow, we would like to use native version of Jupyterlab with "Tab" or Shift + Tab".

(I have the impression that the auto-complete of the extension prevents the native version from working).

[krassowski]

There was a PR that implemented this #586 so it will be possible in the next release. you may want to use my pyls fork in the meantime to get faster completions

[krassowski]

Also to disable linting only for python-lsp-server:

{
    "language_servers": {
        "pylsp": {
            "serverSettings": {
                "pylsp.plugins.pycodestyle.enabled": false,
                "pylsp.plugins.pydocstyle.enabled": false,
                "pylsp.plugins.pyflakes.enabled": false,
                "pylsp.plugins.pylint.enabled": false,
                "pylsp.plugins.mccabe.enabled": false,
            }
        }
    }
}

or only for python-language-server:

{
    "language_servers": {
        "pyls": {
            "serverSettings": {
                "pyls.plugins.pycodestyle.enabled": false,
                "pyls.plugins.pydocstyle.enabled": false,
                "pyls.plugins.pyflakes.enabled": false,
                "pyls.plugins.pylint.enabled": false,
                "pyls.plugins.mccabe.enabled": false,
            }
        }
    }
}

[julioyildo]

Hi!

Thank you very much! That works well! :)