swiftlang/sourcekit-lsp

Description Starting with Swift 6.4, a new parameter has been added to the initialization request: reportSyntacticHighlightInSemanticTokens. This new parameter is not part of the official lsp specification, and is placed in the initializationOptions section of the initialization request (user provided initialization options). When the reportSyntacticHighlightInSemanticTokens is false (the default), sourcekit-lsp does not include syntactic highlight in its response to the textDocument/semanticTokens/full request. It is assumed that the client already has its own syntactic highlighting, and sourcekit-lsp should only augment it with semantic tokens. The official lsp specification already has similar parameter: augmentsSyntaxTokens in semanic tokens client capabilities section. This parameter is currently not used by sourcekit-lsp. See these parameters placement in the params of the initialization request (note that when one parameter is true, the other is false for the same effect): {

Currently, many of the code actions in SwiftLanguageService/CodeActions directly conform to SyntaxCodeActionProvider. For those for which it makes sense, I think we could avoid some boilerplate (and add some more structure) by making the code action conform to SyntaxRefactoringCodeActionProvider as well as SyntaxRefactoringCodeActionProvider, removing the explicit implementation of codeActions(in:) and instead rely on the synthesized version. As this is a learning-oriented task, please don’t address this purely by using an AI tool, as described in https://github.com/swiftlang/project-operations/blob/main/contributing/ai-tools.md#learning-oriented-contributions.

Description Toggle a test between enabled and disabled states. Swift Testing: Before: @Test func testExample() { #expect(2 + 2 == 4) } After: @Test(.disabled()) func testExample() { #expect(2 + 2 == 4) } XCTest: Before: func testExample() { XCTAssertEqual(2 + 2, 4) } After: func testExample() throws { throw XCTSkip("Disabled") XCTAssertEqual(2 + 2, 4) }

Description Generate computed properties to extract associated values from enum cases. Before: enum Value { case text(String) case number(Int) } After: enum Value { case text(String) case number(Int) var asText: String? { if case let .text(v) = self { return v } return nil } var isText: Bool { if case .text = self { return true } return false } var asNumber: Int? { if case let .number(v) = self { return v } return nil } var isNumber: Bool { if case .number = self { return true } return false } } See also: rust-analyzer generate_enum_as_method

Description Replace a temporary variable with its value at all usage sites. Example Before: func example() { let basePrice = item.price let total = basePrice * quantity } After: func example() { let total = item.price * quantity }