[benchmark] Add some distance(from:to:) benchmarks for String views

Author: lorentey

benchmark/CMakeLists.txt

@@ -179,6 +179,7 @@ set(SWIFT_BENCH_MODULES
     single-source/StrToInt
     single-source/StringBuilder
     single-source/StringComparison
+    single-source/StringDistance
     single-source/StringEdits
     single-source/StringEnum
     single-source/StringInterpolation

benchmark/single-source/StringDistance.swift

@@ -0,0 +1,128 @@
+//===--- StringEdits.swift ------------------------------------------------===//
+//
+// This source file is part of the Swift.org open source project
+//
+// Copyright (c) 2014 - 2022 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
+//
+//===----------------------------------------------------------------------===//
+
+import TestsUtils
+
+public let benchmarks: [BenchmarkInfo] = [
+  BenchmarkInfo(
+    name: "StringDistance.characters",
+    runFunction: { n in
+      run_characters(string: sampleString, ranges: sampleRanges, n: n)
+    },
+    tags: [.api, .String],
+    setUpFunction: { _ = sampleRanges }),
+  BenchmarkInfo(
+    name: "StringDistance.scalars",
+    runFunction: { n in
+      run_scalars(string: sampleString, ranges: sampleRanges, n: n)
+    },
+    tags: [.api, .String],
+    setUpFunction: { _ = sampleRanges }),
+  BenchmarkInfo(
+    name: "StringDistance.utf16",
+    runFunction: { n in
+      run_utf16(string: sampleString, ranges: sampleRanges, n: n)
+    },
+    tags: [.api, .String],
+    setUpFunction: { _ = sampleRanges }),
+  BenchmarkInfo(
+    name: "StringDistance.utf8",
+    runFunction: { n in
+      run_utf8(string: sampleString, ranges: sampleRanges, n: n)
+    },
+    tags: [.api, .String],
+    setUpFunction: { _ = sampleRanges }),
+]
+
+
+let sampleString =
+    #"""
+    The powerful programming language that is also easy to learn.
+    손쉽게 학습할 수 있는 강력한 프로그래밍 언어.
+    🪙 A 🥞 short 🍰 piece 🫘 of 🌰 text 👨‍👨‍👧‍👧 with 👨‍👩‍👦 some 🚶🏽 emoji 🇺🇸🇨🇦 characters 🧈
+    some🔩times 🛺 placed 🎣 in 🥌 the 🆘 mid🔀dle 🇦🇶or🏁 around 🏳️‍🌈 a 🍇 w🍑o🥒r🥨d
+    Unicode is such fun!
+    U̷n̷i̷c̷o̴d̴e̷ ̶i̸s̷ ̸s̵u̵c̸h̷ ̸f̵u̷n̴!̵
+    U̴̡̲͋̾n̵̻̳͌ì̶̠̕c̴̭̈͘ǫ̷̯͋̊d̸͖̩̈̈́ḛ̴́ ̴̟͎͐̈i̴̦̓s̴̜̱͘ ̶̲̮̚s̶̙̞͘u̵͕̯̎̽c̵̛͕̜̓h̶̘̍̽ ̸̜̞̿f̵̤̽ṷ̴͇̎͘ń̷͓̒!̷͍̾̚
+    U̷̢̢̧̨̼̬̰̪͓̞̠͔̗̼̙͕͕̭̻̗̮̮̥̣͉̫͉̬̲̺͍̺͊̂ͅn̶̨̢̨̯͓̹̝̲̣̖̞̼̺̬̤̝̊̌́̑̋̋͜͝ͅḭ̸̦̺̺͉̳͎́͑c̵̛̘̥̮̙̥̟̘̝͙̤̮͉͔̭̺̺̅̀̽̒̽̏̊̆͒͌̂͌̌̓̈́̐̔̿̂͑͠͝͝ͅö̶̱̠̱̤̙͚͖̳̜̰̹̖̣̻͎͉̞̫̬̯͕̝͔̝̟̘͔̙̪̭̲́̆̂͑̌͂̉̀̓́̏̎̋͗͛͆̌̽͌̄̎̚͝͝͝͝ͅd̶̨̨̡̡͙̟͉̱̗̝͙͍̮͍̘̮͔͑e̶̢͕̦̜͔̘̘̝͈̪̖̺̥̺̹͉͎͈̫̯̯̻͑͑̿̽͂̀̽͋́̎̈́̈̿͆̿̒̈́̽̔̇͐͛̀̓͆̏̾̀̌̈́̆̽̕ͅ ̷̢̳̫̣̼̙̯̤̬̥̱͓̹͇̽̄̄̋̿̐̇̌̒̾̑̆̈́̏͐̒̈̋̎͐̿̽̆̉͋͊̀̍͘̕̕̕͝͠͠͝ͅͅì̸̢̧̨̨̮͇̤͍̭̜̗̪̪͖̭͇͔̜̗͈̫̩͔̗͔̜̖̲̱͍̗̱̩͍̘̜̙̩͔̏̋̓̊́́̋̐̌͊͘̕͠s̶̨̢̧̥̲̖̝̩͖̱͋́͑͐̇̐̔̀̉͒͒́̐̉̔͘͠͠ ̵̧̛͕̦̭̣̝̩͕̠͎̮͓͉̟̠̘͎͋͗͆̋̌̓̃̏̊̔̾̒̿s̸̟͚̪̘̰̮͉̖̝̅̓͛̏̆ư̵͍̙̠͍̜͖͔̮̠̦̤̣̯̘̲͍͂͌̌̅̍͌̈́̆̋̎͋̓̍͆̃̑͌͘̕͜ͅç̸̟̗͉̟̤̙̹͓̖͇̳̈́̍̏͐̓̓̈̆̉̈͆̍ͅh̵̛̛̹̪͇͓̤̺̟͙̣̰͓̺̩̤̘̫͔̺͙͌́̑̓͗̏͆́͊̈́̋̿͒̐̀́̌͜͜͝ ̴̗͓͚͖̣̥͛́̓͐͂͛̐͑̈́͗̂̈͠f̶̡̩̟̤̭̩̱̥͈̼̥̳͕̣͓̱̰͎̖̦͎̦̻̫͉̝̗̝͚̎͌͑̾̿̊̉͆̉̏̅̔̓̈́̀͐̚͘ͅư̷̦̮͖͙̺̱̼̜̺̤͎̜͐͐̊̊̈͋̔̓̍͊̇̊̈́̈͑̐̎̿̑̋͋̀̅̓͛̚͜n̷̡̨͉̠̖̙͎̳̠̦̼̻̲̳̿̀̓̍͋̎͆̓̇̾̅͊̐͘͘̕!̷̡̨̧̢̡̡̼̹̭̝̝̭̫̫̥̰̤̪̦̤̼̖̖̳̰̲͙͕̖̬̳̪͖̹̮͐͐͊̈́̐͑͛̾̈͊̊͋͑̉͒̈̿̈̃̑͋͐́͊̀͝͠͝͠
+    """#
+
+let sampleRanges = (
+  generateRanges(for: sampleString, by: 1)
+  + generateRanges(for: sampleString, by: 2)
+  + generateRanges(for: sampleString, by: 4)
+  + generateRanges(for: sampleString, by: 8)
+  + generateRanges(for: sampleString, by: 16)
+  + generateRanges(for: sampleString, by: 32)
+  + generateRanges(for: sampleString, by: 64)
+  + generateRanges(for: sampleString, by: 128)
+  + generateRanges(for: sampleString, by: 256)
+  + generateRanges(for: sampleString, by: 512))
+
+func generateRanges(for string: String, by step: Int) -> [Range<String.Index>] {
+  var remaining = step
+  var i = string.startIndex
+  var last = i
+
+  var ranges: [Range<String.Index>] = []
+  while i < string.endIndex {
+    string.unicodeScalars.formIndex(after: &i)
+    remaining -= 1
+    if remaining == 0 {
+      ranges.append(last ..< i)
+      remaining = step
+      last = i
+    }
+  }
+  ranges.append(last ..< i)
+  return ranges
+}
+
+func run_characters(string: String, ranges: [Range<String.Index>], n: Int) {
+  var c = 0
+  for _ in 0 ..< n {
+    for r in ranges {
+      c += string.distance(from: r.lowerBound, to: r.upperBound)
+    }
+  }
+  blackHole(c)
+}
+
+func run_scalars(string: String, ranges: [Range<String.Index>], n: Int) {
+  var c = 0
+  for _ in 0 ..< n {
+    for r in ranges {
+      c += string.unicodeScalars.distance(from: r.lowerBound, to: r.upperBound)
+    }
+  }
+  blackHole(c)
+}
+
+func run_utf16(string: String, ranges: [Range<String.Index>], n: Int) {
+  var c = 0
+  for _ in 0 ..< n {
+    for r in ranges {
+      c += string.utf16.distance(from: r.lowerBound, to: r.upperBound)
+    }
+  }
+  blackHole(c)
+}
+
+func run_utf8(string: String, ranges: [Range<String.Index>], n: Int) {
+  var c = 0
+  for _ in 0 ..< n {
+    for r in ranges {
+      c += string.utf8.distance(from: r.lowerBound, to: r.upperBound)
+    }
+  }
+  blackHole(c)
+}

benchmark/utils/main.swift

@@ -178,6 +178,7 @@ import StrComplexWalk
 import StrToInt
 import StringBuilder
 import StringComparison
+import StringDistance
 import StringEdits
 import StringEnum
 import StringInterpolation
@@ -362,9 +363,10 @@ register(StaticArray.benchmarks)
 register(StrComplexWalk.benchmarks)
 register(StrToInt.benchmarks)
 register(StringBuilder.benchmarks)
+register(StringComparison.benchmarks)
+register(StringDistance.benchmarks)
 register(StringEdits.benchmarks)
 register(StringEnum.benchmarks)
-register(StringComparison.benchmarks)
 register(StringInterpolation.benchmarks)
 register(StringMatch.benchmarks)
 register(StringRemoveDupes.benchmarks)