Swiftlint源码学习
最后更新于
最后更新于
#swiftlint #开源库
Swiftlint 是统一Swift代码风格的工具，本身就是使用Swift语言编写。Swiftlint 使用 Swift_Package_Manager来组织源代码，可以直接用 Xcode打开。
产物
swiftlint 可执行文件
SwiftLintFramework 库
依赖
swift-argument-parser swift命令行的脚手架工具，用于使用swift快速创建命令行工具
SwiftSyntax 允许 Swift 工具解析、检查、生成和转换 Swift 源代码。
SourceKitten 用于和SourceKit进行交互的工具
Yams 解析yaml格式的工具
SwiftyTextTable 生成文本格式表格的工具
关于Swift Package Manager 请参考
- swiftlint
- main
- Commands // 命令的定义
- Extensions
- Helpers
- LintOrAnalyzeCommand // 对主流程的封装(查找swift文件，读取规则的配置，调用rule的检查方法)
- SwiftLintFramework
- Rules // rule的定义
- Documentation
- Reporters
- Protocols
- Models
- Helpers
- Extensions// 1. 命令的解析
// 2. 查询出所有的Swift文件，及文件对应的Swiftlint规格及配置
// 3. 利用规则扫描swift文件
// 4. 输出表格
swiflint lint
命令的解析
// ParsableCommand.run()
Lint.Run()struct LintOrAnalyzeOptions {
let mode: LintOrAnalyzeMode
let paths: [String]
let useSTDIN: Bool
let configurationFiles: [String]
let strict: Bool
let lenient: Bool
let forceExclude: Bool
let useExcludingByPrefix: Bool
let useScriptInputFiles: Bool
let benchmark: Bool
let reporter: String?
let quiet: Bool
let cachePath: String?
let ignoreCache: Bool
let enableAllRules: Bool
let autocorrect: Bool
let format: Bool
let compilerLogPath: String?
let compileCommands: String?
let inProcessSourcekit: Bool
} 查询出所有的Swift文件，及文件对应的Swiftlint配置及规则， 返回 [CollectedLinter]
LintOrAnalyzeCommand.run(_ options: LintOrAnalyzeOptions)public struct CollectedLinter {
/// The file to lint with this linter.
public let file: SwiftLintFile
private let rules: [Rule]
private let cache: LinterCache?
private let configuration: Configuration
private let compilerArguments: [String]
}利用规则扫描swift文件, 返回 [StyleViolation]
Rules.validate(file: SwiftLintFile) -> [StyleViolation]public struct StyleViolation: CustomStringConvertible, Equatable, Codable {
/// The identifier of the rule that generated this violation.
public let ruleIdentifier: String
/// The description of the rule that generated this violation.
public let ruleDescription: String
/// The name of the rule that generated this violation.
public let ruleName: String
/// The severity of this violation.
public private(set) var severity: ViolationSeverity
/// The location of this violation.
public private(set) var location: Location
/// The justification for this violation.
public let reason: String
/// A printable description for this violation.
public var description: String {
return XcodeReporter.generateForSingleViolation(self)
}按照指定格式输出表格
Reporter.report(violations: [StyleViolation], realtimeCondition: Bool)/// An executable value that can identify issues (violations) in Swift source code.
public protocol Rule {
static var description: RuleDescription { get }
var configurationDescription: String { get }
init()
init(configuration: Any) throws
func validate(file: SwiftLintFile, compilerArguments: [String]) -> [StyleViolation]
func validate(file: SwiftLintFile) -> [StyleViolation]
func isEqualTo(_ rule: Rule) -> Bool
func collectInfo(for file: SwiftLintFile, into storage: RuleStorage, compilerArguments: [String])
/// Executes the rule on a file after collecting file info for all files and returns any violations to the rule's
/// expectations.
///
/// - note: This function is called by the linter and is always implemented in extensions.
///
/// - parameter file: The file for which to execute the rule.
/// - parameter storage: The storage object containing all collected info.
/// - parameter compilerArguments: The compiler arguments needed to compile this file.
///
/// - returns: All style violations to the rule's expectations.
func validate(file: SwiftLintFile, using storage: RuleStorage, compilerArguments: [String]) -> [StyleViolation]
}不需要AST抽象语法树，直接通过文本查询或正则匹配实现的Rule
以 LeadingWhitespaceRule 为例：
public struct LeadingWhitespaceRule: CorrectableRule, ConfigurationProviderRule, SourceKitFreeRule {
public func validate(file: SwiftLintFile) -> [StyleViolation] {
let countOfLeadingWhitespace = file.contents.countOfLeadingCharacters(in: .whitespacesAndNewlines)
if countOfLeadingWhitespace == 0 {
return []
}
let reason = "File shouldn't start with whitespace: " +
"currently starts with \(countOfLeadingWhitespace) whitespace characters"
return [StyleViolation(ruleDescription: Self.description,
severity: configuration.severity,
location: Location(file: file.path, line: 1),
reason: reason)]
}
}
// 检查文件开头的空格数量
public func countOfLeadingCharacters(in characterSet: CharacterSet) -> Int {
let characterSet = characterSet.bridge()
var count = 0
for char in utf16 {
if !characterSet.characterIsMember(char) {
break
}
count += 1
}
return count
}需要抽象语法树配合的Rule, 抽象语法树通过sourcekit来获取
以 ArrayInitRule 为例：
import Foundation
let array = [1,2,3]
let array2 = array.map { return $0 }通过 sourcekitten structure --file main.swift获取AST
{
"key.diagnostic_stage" : "source.diagnostic.stage.swift.parse",
"key.length" : 74,
"key.offset" : 0,
"key.substructure" : [
{
"key.accessibility" : "source.lang.swift.accessibility.internal",
"key.kind" : "source.lang.swift.decl.var.global",
"key.length" : 19,
"key.name" : "array",
"key.namelength" : 5,
"key.nameoffset" : 23,
"key.offset" : 19
},
{
"key.bodylength" : 5,
"key.bodyoffset" : 32,
"key.elements" : [
{
"key.kind" : "source.lang.swift.structure.elem.expr",
"key.length" : 1,
"key.offset" : 32
},
{
"key.kind" : "source.lang.swift.structure.elem.expr",
"key.length" : 1,
"key.offset" : 34
},
{
"key.kind" : "source.lang.swift.structure.elem.expr",
"key.length" : 1,
"key.offset" : 36
}
],
"key.kind" : "source.lang.swift.expr.array",
"key.length" : 7,
"key.namelength" : 0,
"key.nameoffset" : 0,
"key.offset" : 31
},
{
"key.accessibility" : "source.lang.swift.accessibility.internal",
"key.kind" : "source.lang.swift.decl.var.global",
"key.length" : 29,
"key.name" : "array2",
"key.namelength" : 6,
"key.nameoffset" : 44,
"key.offset" : 40
},
{
"key.bodylength" : 4,
"key.bodyoffset" : 64,
"key.kind" : "source.lang.swift.expr.call",
"key.length" : 16,
"key.name" : "array.map",
"key.namelength" : 9,
"key.nameoffset" : 53,
"key.offset" : 53,
"key.substructure" : [
{
"key.bodylength" : 4,
"key.bodyoffset" : 64,
"key.kind" : "source.lang.swift.expr.closure",
"key.length" : 6,
"key.namelength" : 0,
"key.nameoffset" : 0,
"key.offset" : 63,
"key.substructure" : [
{
"key.bodylength" : 4,
"key.bodyoffset" : 64,
"key.kind" : "source.lang.swift.stmt.brace",
"key.length" : 6,
"key.namelength" : 0,
"key.nameoffset" : 0,
"key.offset" : 63
}
]
}
]
}
}通过 sourcekitten syntax --file main.swift获取Token list
[
{
"length" : 6,
"offset" : 0,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 10,
"offset" : 7,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 3,
"offset" : 19,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 5,
"offset" : 23,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 1,
"offset" : 32,
"type" : "source.lang.swift.syntaxtype.number"
},
{
"length" : 1,
"offset" : 34,
"type" : "source.lang.swift.syntaxtype.number"
},
{
"length" : 1,
"offset" : 36,
"type" : "source.lang.swift.syntaxtype.number"
},
{
"length" : 3,
"offset" : 40,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 6,
"offset" : 44,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 5,
"offset" : 53,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 3,
"offset" : 59,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 6,
"offset" : 65,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 2,
"offset" : 72,
"type" : "source.lang.swift.syntaxtype.identifier"
}
]
// 结合抽象语法树进行分析
public func validate(file: SwiftLintFile, kind: SwiftExpressionKind,
dictionary: SourceKittenDictionary) -> [StyleViolation] {
guard kind == .call, let name = dictionary.name, name.hasSuffix(".map"),
let bodyOffset = dictionary.bodyOffset,
let bodyLength = dictionary.bodyLength,
let bodyRange = dictionary.bodyByteRange,
let nameOffset = dictionary.nameOffset,
let nameLength = dictionary.nameLength,
let offset = dictionary.offset else {
return []
}
let tokens = file.syntaxMap.tokens(inByteRange: bodyRange).filter { token in
guard let kind = token.kind else {
return false
}
return !SyntaxKind.commentKinds.contains(kind)
}
guard let firstToken = tokens.first,
case let nameEndPosition = nameOffset + nameLength,
isClosureParameter(firstToken: firstToken, nameEndPosition: nameEndPosition, file: file),
isShortParameterStyleViolation(file: file, tokens: tokens) ||
isParameterStyleViolation(file: file, dictionary: dictionary, tokens: tokens),
let lastToken = tokens.last,
case let bodyEndPosition = bodyOffset + bodyLength,
!containsTrailingContent(lastToken: lastToken, bodyEndPosition: bodyEndPosition, file: file),
!containsLeadingContent(tokens: tokens, bodyStartPosition: bodyOffset, file: file) else {
return []
}
return [
StyleViolation(ruleDescription: Self.description,
severity: configuration.severity,
location: Location(file: file, byteOffset: offset))
]
}需要抽象语法树和编译日志配合的Rule, 抽象语法树通过sourcekit来获取
以 UnusedImportRule 为例：
import UIKit
import WebKit
class TestObject: NSObject {
var a = 11
}[
{
"length" : 6,
"offset" : 0,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 5,
"offset" : 7,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 6,
"offset" : 13,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 6,
"offset" : 20,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 5,
"offset" : 28,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 10,
"offset" : 34,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 8,
"offset" : 46,
"type" : "source.lang.swift.syntaxtype.typeidentifier"
},
{
"length" : 3,
"offset" : 61,
"type" : "source.lang.swift.syntaxtype.keyword"
},
{
"length" : 1,
"offset" : 65,
"type" : "source.lang.swift.syntaxtype.identifier"
},
{
"length" : 2,
"offset" : 69,
"type" : "source.lang.swift.syntaxtype.number"
}
] func getImportsAndUSRFragments(compilerArguments: [String]) -> (imports: Set<String>, usrFragments: Set<String>) {
var imports = Set<String>()
var usrFragments = Set<String>()
var nextIsModuleImport = false
for token in syntaxMap.tokens {
guard let tokenKind = token.kind else {
continue
}
if tokenKind == .keyword, contents(for: token) == "import" {
nextIsModuleImport = true
continue
}
// 过滤掉没有模块信息的token
if SyntaxKind.kindsWithoutModuleInfo.contains(tokenKind) {
continue
}
// 获取当前token对应的抽象语法树
let cursorInfoRequest = Request.cursorInfo(file: path!, offset: token.offset,
arguments: compilerArguments)
guard let cursorInfo = (try? cursorInfoRequest.sendIfNotDisabled()).map(SourceKittenDictionary.init) else {
queuedPrintError("Could not get cursor info")
continue
}
if nextIsModuleImport {
// 通过语法树并获取当前文件的所有import
if let importedModule = cursorInfo.moduleName,
cursorInfo.kind == "source.lang.swift.ref.module" {
imports.insert(importedModule)
nextIsModuleImport = false
continue
} else {
nextIsModuleImport = false
}
}
// 获取使用的import
appendUsedImports(cursorInfo: cursorInfo, usrFragments: &usrFragments)
}
return (imports: imports, usrFragments: usrFragments)
}
func appendUsedImports(cursorInfo: SourceKittenDictionary, usrFragments: inout Set<String>) {
if let rootModuleName = cursorInfo.moduleName?.split(separator: ".").first.map(String.init) {
usrFragments.insert(rootModuleName)
if rootModuleName == moduleToLog, let filePath = path, let usr = cursorInfo.value["key.usr"] as? String {
queuedPrintError(
"[SWIFTLINT_LOG_MODULE_USAGE] \(rootModuleName) referenced by USR '\(usr)' in file '\(filePath)'"
)
}
}
}SourceKit 是一套工具集，使得大多数 Swift 源代码层面的操作特性得以支持，例如Xcode中的源代码解析、语法高亮、排版、自动补全、跨语言头文件生成等等。 SourceKitten 是与 SourceKit 后台进程进行交互的工具 ，基于 SourceKitten 还有很多开源工具
Jazzy: 为Swift/OC生成文档
Sourcery: 生成模版代码
SwiftyMocky: 生成Mock代码的框架
SourceKittenDaemon: 为任何文本编辑器提供swift代码补全
SourceDocs: 从源码注释中生成Markdown文档的命令行工具
Cuckoo: Swift Mock 框架
IBAnalyzer: 无需运行代码或单元测试发现xib或storyboard相关的问题