[ASTGen] Add syntax tree based Fingerprint

Author: rintaro

lib/ASTGen/Sources/ASTGen/CMakeLists.txt

@@ -8,12 +8,14 @@ add_pure_swift_host_library(swiftASTGen STATIC CXX_INTEROP
   Diagnostics.swift
   DiagnosticsBridge.swift
   Exprs.swift
+  Fingerprint.swift
   Generics.swift
   Literals.swift
   ParameterClause.swift
   Patterns.swift
   Regex.swift
   SourceFile.swift
+  StableHasher.swift
   Stmts.swift
   TypeAttrs.swift
   Types.swift

lib/ASTGen/Sources/ASTGen/Fingerprint.swift

@@ -0,0 +1,79 @@
+//===--- Fingerprint.swift -------------------------------------------------===//
+//
+// This source file is part of the Swift.org open source project
+//
+// Copyright (c) 2024 Apple Inc. and the Swift project authors
+// Licensed under Apache License v2.0 with Runtime Library Exception
+//
+// See https://swift.org/LICENSE.txt for license information
+// See https://swift.org/CONTRIBUTORS.txt for the list of Swift project authors
+//
+//===----------------------------------------------------------------------===//
+
+@_spi(RawSyntax)
+import SwiftSyntax
+import SwiftIfConfig
+
+struct Fingerprint: Equatable {
+  typealias Core = (UInt64, UInt64)
+  var core: Core
+  init(core: Core) {
+    self.core = core
+  }
+
+  static func == (lhs: Self, rhs: Self) -> Bool {
+    return lhs.core == rhs.core
+  }
+}
+
+extension StableHasher {
+  @inline(__always)
+  mutating func combine(text: SyntaxText) {
+    let buffer = UnsafeRawBufferPointer(
+      start: UnsafeRawPointer(text.baseAddress),
+      count: text.count
+    )
+    self.combine(UInt(truncatingIfNeeded: buffer.count))
+    self.combine(bytes: buffer)
+  }
+}
+
+final class FingerprintVisitor: SyntaxVisitor {
+  let configuredRegions: ConfiguredRegions
+  var hasher: StableHasher
+
+  init(configuredRegions: ConfiguredRegions) {
+    self.configuredRegions = configuredRegions
+    self.hasher = StableHasher()
+    super.init(viewMode: .sourceAccurate)
+  }
+
+  func finalize() -> Fingerprint {
+    Fingerprint(core: hasher.finalize())
+  }
+
+  override func visit(_ token: TokenSyntax) -> SyntaxVisitorContinueKind {
+    // Collect all token texts.
+    hasher.combine(text: token.rawText)
+    return .skipChildren
+  }
+
+  override func visit(_ node: MemberBlockSyntax) -> SyntaxVisitorContinueKind {
+    // Skip nominal decl / extension member blocks.
+    return .skipChildren
+  }
+
+  override func visit(_ node: CodeBlockSyntax) -> SyntaxVisitorContinueKind {
+    // Skip function bodies.
+    // FIXME: Don't skip closure bodies or other control flow statement blocks.
+    //        E.g. 'var val = { if condition { "foo" } else { "bar" } }()'
+    return .skipChildren
+  }
+
+  override func visit(_ node: IfConfigDeclSyntax) -> SyntaxVisitorContinueKind {
+    if let active = configuredRegions.activeClause(for: node) {
+      self.walk(active)
+    }
+    return .skipChildren
+  }
+}