[wip] add settings-based code extractors

[bollwyvl]

References

• fixes Define custom code extractors in settings #1079

Code changes

• adds cellTypes to RegExpForeignCodeExtractor.IOptions
• adds schema/transclusions.json to define a simplified extractor
• adds ILSPCustomTransclusionsManager token

User-facing changes

• users will be able to add new extractors from the Settings Editor

Backwards-incompatible changes

• n/a

Chores

• linted
• tested
• documented
• changelog entry

[github-actions]

👈 Launch a binder notebook on branch bollwyvl/jupyterlab-lsp/gh-1079-settings-code-extractors

[bollwyvl]

As of opening this PR, things kinda work, e.g. this naive extractor finding <script> tags inside .md files and markdown cells:

{
  "enabled": true,
  "codeExtractors": {
    "md-js-script": {
      "hostLanguage": "ipythongfm",
      "foreignLanguage": "javascript"
      "pattern": "<script>((.|\n)*?(?=</script>))</script>",
      "foreignCaptureGroups": [1],
      "fileExtension": "js",
      "cellTypes": ["code", "markdown"],
      "isStandalone": false,
    }
  }
}

Kinda, in that while diagnostics work, completion, go to definition, and others do not appear to work, but i haven't debugged further.

The internal state of the upstream manager (and the extractor itself) is kind of hard to reason about, and generally requires a full page reload after making any changes.

[bollwyvl]

[bollwyvl]

Apparently we aren't limited to integer-indexed groups:
https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Regular_expressions/Named_capturing_group#browser_compatibility

The syntax, alas, is not portable:

# python
r"(?P<name>.*)"

/* js */
/(?<name>.*)/ 

But still...

[bollwyvl]

Should likely consider allowing this to be an array of strings that get pre-concatenated.