[tomotakatakahashi]

Summary

This PR fixes a bug where relative paths are not resolved when configuration files are searched.

When a file is passed as a relative path (e.g. swiftlint lint Sources/App.swift), the nested configuration search does not recognize that it has reached the root directory, so it keeps walking up the file tree and merges .swiftlint.yml files from directories outside the project.

Context

According to Claude Code, another PR of mine ( #6825 ) would worsen this bug (although I haven't confirmed whether it is true or not). Maybe I should fix this bug before the other PR is merged.

Test

• The added test fails if the Source changes are not applied.
• Running the swiftlint scheme with lint --no-cache Sources/Core/Service.swift arguments and /Users/tomotaka/projects/SwiftLint2/Tests/IntegrationTests/Resources/_2_parent_child_different_dirs/project working directory:
  • Before this PR: Found 2 violations. (file_name and no_empty_block)
    • This repo's .swiftlint.yml (opt_in_rules: all) is used instead of the working dir's .swiftlint.yml.
  • After this PR: Found 0 violations.

Comment

• I wasn't sure which URL API to use for the comparison. standardizedFileURL, path(percentEncoded:) and absoluteString all resolve the relative URL:

import Playgrounds

#Playground {
    let baseURL = URL(string: "file:///Users/tomotaka/projects/SwiftLint2/Tests/IntegrationTests/Resources/_9_symlinked_paths/Real/")!
    let relURL = URL(fileURLWithPath: "../LinkToFile.swift", relativeTo: baseURL)

    // "/Users/tomotaka/projects/SwiftLint2/Tests/IntegrationTests/Resources/_9_symlinked_paths/LinkToFile.swift"
    relURL.path(percentEncoded: false)
    // file:///Users/tomotaka/projects/SwiftLint2/Tests/IntegrationTests/Resources/_9_symlinked_paths/LinkToFile.swift
    relURL.standardizedFileURL
    // "file:///Users/tomotaka/projects/SwiftLint2/Tests/IntegrationTests/Resources/_9_symlinked_paths/LinkToFile.swift"
    relURL.absoluteString
}

• I considered adding this test case to MultipleConfigurationsTests, but this test case does not necessarily catch a regression. If the Source changes are reverted, the test suite becomes flaky (due to the caching mechanism of Configuration according to Claude Code).

    @Test(.workingDirectory(Constants.Dir.level0))
    func level0WithRelativeFilePath() {
        let relativePath = URL(
            filePath: Constants.Swift._0.lastPathComponent,
            relativeTo: Constants.Swift._0.deletingLastPathComponent()
        )
        #expect(
            Constants.Config._0.configuration(for: SwiftLintFile(path: relativePath)!)
                == Constants.Config._0
        )
    }

[SwiftLintBot]

	20 Messages
📖	Building this branch resulted in the same binary size as when built on main.
📖	Linting Aerial with this PR took 0.68 s vs 0.68 s on main (0% slower).
📖	Linting Alamofire with this PR took 0.98 s vs 0.98 s on main (0% slower).
📖	Linting Brave with this PR took 5.53 s vs 5.51 s on main (0% slower).
📖	Linting Brigade with this PR took 18.34 s vs 18.19 s on main (0% slower).
📖	Linting DuckDuckGo with this PR took 31.16 s vs 31.04 s on main (0% slower).
📖	Linting Firefox with this PR took 11.75 s vs 11.75 s on main (0% slower).
📖	Linting Kickstarter with this PR took 7.41 s vs 7.45 s on main (0% faster).
📖	Linting Moya with this PR took 0.38 s vs 0.39 s on main (2% faster).
📖	Linting NetNewsWire with this PR took 2.59 s vs 2.59 s on main (0% slower).
📖	Linting Nimble with this PR took 0.56 s vs 0.57 s on main (1% faster).
📖	Linting PocketCasts with this PR took 7.45 s vs 7.42 s on main (0% slower).
📖	Linting Quick with this PR took 0.37 s vs 0.36 s on main (2% slower).
📖	Linting Realm with this PR took 2.98 s vs 2.98 s on main (0% slower).
📖	Linting Sourcery with this PR took 1.63 s vs 1.6 s on main (1% slower).
📖	Linting Swift with this PR took 4.58 s vs 4.57 s on main (0% slower).
📖	Linting SwiftLintPerformanceTests with this PR took 0.18 s vs 0.18 s on main (0% slower).
📖	Linting VLC with this PR took 1.41 s vs 1.4 s on main (0% slower).
📖	Linting Wire with this PR took 16.56 s vs 16.52 s on main (0% slower).
📖	Linting WordPress with this PR took 10.66 s vs 10.6 s on main (0% slower).

Generated by 🚫 Danger