add format option handlers

[ZacLN]

Allows the passing of config options from client to DocumentFormat. Merging should be delayed until tagging of new DocFormat release + julia-vscode/julia-vscode#806 .

[codecov]

Codecov Report

Merging #374 into master will increase coverage by 0.19%.
The diff coverage is 30%.

@@            Coverage Diff             @@
##           master     #374      +/-   ##
==========================================
+ Coverage   33.14%   33.33%   +0.19%     
==========================================
  Files          22       22              
  Lines        1219     1230      +11     
==========================================
+ Hits          404      410       +6     
- Misses        815      820       +5

Impacted Files	Coverage Δ
src/requests/features.jl	0% <0%> (ø)	⬆️
src/requests/workspace.jl	0% <0%> (ø)	⬆️
src/requests/init.jl	54.76% <100%> (+1.1%)	⬆️
src/languageserverinstance.jl	88.09% <40%> (-6.5%)	⬇️
src/protocol/configuration.jl	100% <0%> (ø)	⬆️
src/jsonrpc.jl	72.72% <0%> (+3.03%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f31248d...8f69ebd. Read the comment docs.

[non-Jedi]

This broke master for use with eglot. The server sends:

[ Info: SENT: {"method":"workspace/configuration","id":-100,"params":{"items":[{"scopeUri":null,"section":"julia.format.indent"},{"scopeUri":null,"section":"julia.format.indents"},{"scopeUri":null,"section":"julia.format.ops"},{"scopeUri":null,"section":"julia.format.tuples"},{"scopeUri":null,"section":"julia.format.curly"},{"scopeUri":null,"section":"julia.format.calls"},{"scopeUri":null,"section":"julia.format.iterOps"},{"scopeUri":null,"section":"julia.format.comments"},{"scopeUri":null,"section":"julia.format.docs"},{"scopeUri":null,"section":"julia.format.lineends"}]},"jsonrpc":"2.0"}

and eglot responds with

[ Info: RECEIVED: {"jsonrpc":"2.0","id":-100,"result":null,"error":null}

And of course length is not defined for nothing.

Reading the specification, it seems like eglot may be in the wrong here since the type of result is supposed to be any[], but I could be misinterpreting it. I'm not very familiar with the language they're using to describe the interfaces.

[non-Jedi]

This PR doesn't follow the spec in that it sets scopeUri to nothing when it needs to be a valid DocumentUri (a string). Any chance of rolling this PR back and releasing a 1.0.1?

[davidanthoff]

Could you try whether #416 fixes the problem?

If I understand the @json_read macro correctly, then it will lead to a situation where serializing a struct with an optional field will actually just not put that field into the JSON representation if the value is nothing, which is probably what we want here, right?

[non-Jedi]

Ya, that would be what we want. I checked out 7330639 from #416, and I'm still seeing "scopeUri": null. Would you like me to open an issue for this, so it's tracked better than in the comments of an already merged PR?