Fixed rules violation in the linter code

[xperiandri]

• disable FL0046 for Identifiers.fs
• make constants instead of let bindings in SourceLength.fs

[xperiandri]

@Numpsy could you review?

[Copilot]

Pull Request Overview

This PR refactors linter rule code by disabling a specific lint rule, converting various test thresholds into compile-time constants, and modernizing string formatting in the build script.

• Add [<Literal>] to test threshold definitions in SourceLength.fs
• Disable FL0046 and replace sprintf with string interpolation in Identifiers.fs
• Update build.fsx to use interpolation and DotNet.nugetPush, remove Paket

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.

File	Description
tests/FSharpLint.Core.Tests/Rules/Conventions/SourceLength.fs	Convert test length thresholds to [<Literal>] constants
src/FSharpLint.Core/Rules/Identifiers.fs	Disable FL0046 and switch sprintf to string interpolation
src/FSharpLint.Core/Framework/Ast.fs	Adjust pattern match for SynExpr.Typar
build.fsx	Remove Paket ref, switch to interpolation, replace NuGet push

Comments suppressed due to low confidence (4)

tests/FSharpLint.Core.Tests/Rules/Conventions/SourceLength.fs:232

• The TestMaxLinesInMember fixture is declared but has no assertions; add concrete test cases to verify member-line limits.

    // TODO: Add tests.

tests/FSharpLint.Core.Tests/Rules/Conventions/SourceLength.fs:307

• The TestMaxLinesInUnion fixture is missing actual test assertions; include tests to validate union-line constraints.

    // TODO: Add tests.

tests/FSharpLint.Core.Tests/Rules/Conventions/SourceLength.fs:342

• The TestMaxLinesInEnum fixture has no implemented tests; consider adding assertions to cover enum-line length behavior.

    // TODO: Add tests.

src/FSharpLint.Core/Framework/Ast.fs:373

• If SynExpr.Typar actually has two parameters, collapsing to one may miss data; verify the AST shape change is intentional.

        | SynExpr.Typar(_) -> ()

[Numpsy]

Looks ok to me (I've been putting the format specifiers in interpolated strings for my own code, I don't know if copilot can be configured to not complain about that?)

[Numpsy]

I've been having a go at getting the Fornax doc generation working over in Numpsy@686d139, not quite working yet though

[knocte]

disable FL0046 for Identifiers.fs

You explained the what but not the why. What was the motivation for doing this?

[xperiandri]

Well, someone decided to name the identifiers in pascal case. Which is logical.
But F# does not allow compiling an interpolated string into a literal.
Probably I should have mentioned that

[knocte]

Probably I should have mentioned that

Or file that as a bug (with a testcase) so that we can take a closer look? I'm still not sure I understand the issue.

[knocte]

But F# does not allow compiling an interpolated string into a literal.

But what was the motivation for having constants (literals) in the first place?

[xperiandri]

I do not know. Someone did that before me and you in this project

[knocte]

I'm not referring to previous code, I'm referring to what you have put in this PR description: "make constants instead of let bindings in SourceLength.fs", what is the motivation of making them constants instead of let bindings?

[xperiandri]

Because constants are Pascal case and let bindings are Camel case. But in the code let bindings are Camel case

[knocte]

It is so hard to decipher what you exactly mean man. Let me see: are you referring to running FSharpLint on itself? Because I just tested removing that fsharp:disable line (see https://github.com/knocte/FSharpLint/commits/wip/removeRuleDisabling/ ) and CI still passes (SelfCheck still green) so I'm still wondering what this change is for.

[Numpsy]

The previously seen issue might have been a bug that was fixed by #758? (where FL0046 previously fired for let bindings, and now doesn't)