feat: Support syntax coloration for code block in LSP hover

[angelozerr]

Support syntax coloration for code block in LSP hover

Fixes #297

• Here the result with Lua (the coloration comes from a custom lexer):

Do you think it is correct CppCXY?

I copied / pasted some piece of code from QuickDocHighlightingHelper and it seems it is working with Language. For TextMate Language is not enough (because Language is equals to "textmate). I have implemented TextMate support too.

• Here a screenshot with Python TextMate (//cc @InSyncWithFoo):

• Here a screenshot with TypeScript TextMate:

• Here a demo with Rust TextMate:

• Here a demo with indented blockquote:

[fbricon]

the code block content is duplicated in your screenshot, or you can compare the following

with the JetBrains version:

[angelozerr]

the code block content is duplicated in your screenshot, or you can compare the following

Good catch, my translate of kotlin to Java code was bad, it should be fixed now (I have updated screenshot).

[codecov]

Codecov Report

Attention: Patch coverage is 55.30303% with 118 lines in your changes missing coverage. Please review.

Project coverage is 24.44%. Comparing base (d4d186d) to head (5c4fef8).
Report is 2 commits behind head on main.

❗ Current head 5c4fef8 differs from pull request most recent head 9a8da99

Please upload reports for the commit 9a8da99 to get more accurate results.

Files	Patch %	Lines
...features/documentation/LSPDocumentationHelper.java	40.00%	29 Missing and 10 partials ⚠️
...documentation/LightQuickDocHighlightingHelper.java	41.46%	18 Missing and 6 partials ⚠️
...cumentation/SyntaxColorationCodeBlockRenderer.java	70.49%	11 Missing and 7 partials ⚠️
...tures/documentation/TextMateHighlighterHelper.java	61.11%	9 Missing and 5 partials ⚠️
...4ij/features/completion/LSPCompletionProposal.java	0.00%	12 Missing ⚠️
...edhat/devtools/lsp4ij/LanguageServersRegistry.java	80.00%	1 Missing and 2 partials ⚠️
...dhat/devtools/lsp4ij/commands/CommandExecutor.java	0.00%	3 Missing ⚠️
...m/redhat/devtools/lsp4ij/ServerMessageHandler.java	0.00%	2 Missing ⚠️
...ols/lsp4ij/launching/ui/ShowInstructionDialog.java	0.00%	2 Missing ⚠️
...hat/devtools/lsp4ij/client/LanguageClientImpl.java	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #360      +/-   ##
==========================================
+ Coverage   22.99%   24.44%   +1.45%     
==========================================
  Files         267      270       +3     
  Lines        9369     9410      +41     
  Branches     1728     1734       +6     
==========================================
+ Hits         2154     2300     +146     
+ Misses       6809     6685     -124     
- Partials      406      425      +19     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[InSyncWithFoo]

Thanks. Glad to know that I'll be able to customize hover support using something more than just a boolean. For comparison, here's how the native API renders the same signature:

Notable differences:

• None and False have the same style as def, as they are all keywords.
• Monospace font is used (probably the editor font, but definitely not the UI font), since the entire block is a Markdown code block.
• Everything is rendered with respect to the current color scheme.

[fbricon]

Testing this PR's build on Pycharm, with based-pyright enabled:

VS stock Pycharm:

There's a chance the (function) prefix based-pyright adds is disrupting the built-in textmate grammar, causing the style discrepancy with stock pycharm.

As for the missing monospace font, is it a big deal?

[InSyncWithFoo]

As for the missing monospace font, is it a big deal?

It is. A Python function may have many parameters, with type hints for each.

Take json.dump() as example:

def dump(
    obj: Any,
    fp: SupportsWrite[str],
    *,
    skipkeys: bool = False,
    ensure_ascii: bool = True,
    check_circular: bool = True,
    allow_nan: bool = True,
    cls: type[JSONEncoder] | None = None,
    indent: None | int | str = None,
    separators: tuple[str, str] | None = None,
    default: Callable[[Any], Any] | None = None,
    sort_keys: bool = False,
    **kwds: Any,
) -> None: ...

Not crazy enough? Check out rich.console.Console.__init__():

def __init__(
    self,
    *,
    color_system: Optional[
        Literal["auto", "standard", "256", "truecolor", "windows"]
    ] = "auto",
    force_terminal: Optional[bool] = None,
    force_jupyter: Optional[bool] = None,
    force_interactive: Optional[bool] = None,
    soft_wrap: bool = False,
    theme: Optional[Theme] = None,
    stderr: bool = False,
    file: Optional[IO[str]] = None,
    quiet: bool = False,
    width: Optional[int] = None,
    height: Optional[int] = None,
    style: Optional[StyleType] = None,
    no_color: Optional[bool] = None,
    tab_size: int = 8,
    record: bool = False,
    markup: bool = True,
    emoji: bool = True,
    emoji_variant: Optional[EmojiVariant] = None,
    highlight: bool = True,
    log_time: bool = True,
    log_path: bool = True,
    log_time_format: Union[str, FormatTimeCallable] = "[%X]",
    highlighter: Optional["HighlighterType"] = ReprHighlighter(),
    legacy_windows: Optional[bool] = None,
    safe_box: bool = True,
    get_datetime: Optional[Callable[[], datetime]] = None,
    get_time: Optional[Callable[[], float]] = None,
    _environ: Optional[Mapping[str, str]] = None,
):

For the main part, these signatures are code, and (unless you are into that kind of thing) code are meant to be displayed in monospace font.

[fbricon]

it's still readable

[InSyncWithFoo]

I, for one, largely prefer it this way, and I think dyslexic people would agree:

Readability notwithstanding, here's how the block is sent via LSP (pseudo-JSON):

{
  "result": {
    "contents": {
      "kind": "markdown",
      "value": """
        ```python
        (function) def dumps(
            obj: Any,
            *,
            skipkeys: bool = False,
            ...
        ) -> str
        ```
        ---
        Serialize `obj` to a JSON formatted `str`.
        
        If `skipkeys` is true then `dict` keys that are not basic types [...]
      """
    },
    "range": {/* ... */}
}

That message should, obviously, be rendered as one code block followed by a horizontal ruler and multiple paragraphs:

(function) def dumps(
    obj: Any,
    *,
    skipkeys: bool = False,
    ...
) -> str

---

Serialize obj to a JSON formatted str.

If skipkeys is true then dict keys that are not basic types [...]

[fbricon]

There you go:

just needed to wrap the code rendering with <pre>

[angelozerr]

@fbricon if you want to test quickly this PR, you can create a ts file with documentation like:

/**

ts: const s= '';

```xml
xml: <foo></foo>

 java: String s = new String("abed");

  > const s = "";

def dump(
    obj: Any,
    fp: SupportsWrite[str],
    *,
    skipkeys: bool = False,
    ensure_ascii: bool = True,
    check_circular: bool = True,
    allow_nan: bool = True,
    cls: type[JSONEncoder] | None = None,
    indent: None | int | str = None,
    separators: tuple[str, str] | None = None,
    default: Callable[[Any], Any] | None = None,
    sort_keys: bool = False,
    **kwds: Any,
) -> None: ...

**/

function foo() {
}

foo();

and hover the foo:

[fbricon]

return should be @NotNull

[angelozerr]

It should, because it is filtered after.

[fbricon]

Thanks @angelozerr!