v5.0.0

This release brings support for JupyterLab 4.0. Known issues/limitations are listed in the changelog below. Please report any other regression on GitHub issue tracker and feel welcome to open a pull request if you would like to contribute an enhancement or a bug fix. For older JupyterLab versions please use jupyterlab-lsp 4.x line.

Extension authors: feature registration is now handled by upstream @jupyterlab/lsp package; migration overview notes are available in the documentation. Please note that better support for CodeMirror6 integration will be available in JupyterLab 4.1 release.

To upgrade to this release:

1. Make sure your JupyterLab version is 4.0.6 or newer and Python version is 3.8 or newer.
2. Update the extension:

   pip install jupyterlab-lsp==5.0.0 jupyter-lsp==2.2.0 'jupyterlab>=4.0.6'
   # or, for conda:
   # > Please note that conda packages may take a day or two since release before becoming available!
   # conda install -c conda-forge jupyterlab-lsp=5.0.0 jupyter-lsp=2.2.0 'jupyterlab>=4.0.6'

3. If installing this extension for the first time:
   • please read the full installation instructions as language servers of your choice need to be installed too, and additional configuration can make your experience smoother,
   • if you wish to use jedi-language-server or python-lsp-server, please make sure to upgrade IPython to 7.20 or newer.

Changelog

• enhancements:
  • uses toast notifications instead of messages on status bar
  • diagnostics panel will be re-opened on reload
• maintenance:
  • support JupyterLab 4
  • use upstream @jupyterlab/lsp package
  • use camelCase convention in TypeScript/JavaScript code
  • use @codemirror/linter to show diagnostics
    • this comes with a different style of underlines and custom tooltips
• known issues/limitations:
  • configuration of language servers via JSON Settings Editor may result in a spurious warning due to a transitive clash with settings from the UI editor when using nested pattern (e.g. {pylsp: {flake8: {enabled: true}}}); the dotted pattern (e.g. {"pylsp.flake8.enabled": true}) does not lead to such problem.
  • enabling auto-invoke of completer requires toggling checkbox in both native and LSP Code Completion settings
  • robot mode does not support JupyterLab 4.0, hence robot LSP will not work either
  • renaming in docker files may not work on certain variables due to upstream tokenizer issue

Full Changelog: v4.3.0...v5.0.0