Skip to content

Commit fa6c038

Browse files
valeriyvanvaleriyvan
committed
Adds benchmarks for UTF16 decoding
1 parent 2e8ae40 commit fa6c038

File tree

3 files changed

+213
-0
lines changed

3 files changed

+213
-0
lines changed

‎benchmark/CMakeLists.txt

+1
Original file line numberDiff line numberDiff line change
@@ -198,6 +198,7 @@ set(SWIFT_BENCH_MODULES
198198
single-source/TwoSum
199199
single-source/TypeFlood
200200
single-source/UTF8Decode
201+
single-source/UTF16Decode
201202
single-source/Walsh
202203
single-source/WordCount
203204
single-source/XorLoop

‎benchmark/single-source/UTF16Decode.swift

+210
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,210 @@
1+
//===--- UTF16Decode.swift -------------------------------------------------===//
2+
//
3+
// This source file is part of the Swift.org open source project
4+
//
5+
// Copyright (c) 2014 - 2017 Apple Inc. and the Swift project authors
6+
// Licensed under Apache License v2.0 with Runtime Library Exception
7+
//
8+
// See https://swift.org/LICENSE.txt for license information
9+
// See https://swift.org/CONTRIBUTORS.txt for the list of Swift project authors
10+
//
11+
//===----------------------------------------------------------------------===//
12+
13+
import TestsUtils
14+
import Foundation
15+
16+
public let UTF16Decode = [
17+
BenchmarkInfo(
18+
name: "UTF16Decode",
19+
runFunction: run_UTF16Decode,
20+
tags: [.validation, .api, .String]),
21+
BenchmarkInfo(
22+
name: "UTF16Decode_InitFromCustom_contiguous",
23+
runFunction: run_UTF16Decode_InitFromCustom_contiguous,
24+
tags: [.validation, .api, .String]),
25+
BenchmarkInfo(
26+
name: "UTF16Decode_InitFromCustom_contiguous_ascii",
27+
runFunction: run_UTF16Decode_InitFromCustom_contiguous_ascii,
28+
tags: [.validation, .api, .String]),
29+
BenchmarkInfo(
30+
name: "UTF16Decode_InitFromCustom_noncontiguous",
31+
runFunction: run_UTF16Decode_InitFromCustom_noncontiguous,
32+
tags: [.validation, .api, .String]),
33+
BenchmarkInfo(
34+
name: "UTF16Decode_InitFromCustom_noncontiguous_ascii",
35+
runFunction: run_UTF16Decode_InitFromCustom_noncontiguous_ascii,
36+
tags: [.validation, .api, .String]),
37+
BenchmarkInfo(
38+
name: "UTF16Decode_InitFromData",
39+
runFunction: run_UTF16Decode_InitFromData,
40+
tags: [.validation, .api, .String]),
41+
BenchmarkInfo(
42+
name: "UTF16Decode_InitDecoding",
43+
runFunction: run_UTF16Decode_InitDecoding,
44+
tags: [.validation, .api, .String]),
45+
BenchmarkInfo(
46+
name: "UTF16Decode_InitFromData_ascii",
47+
runFunction: run_UTF16Decode_InitFromData_ascii,
48+
tags: [.validation, .api, .String]),
49+
BenchmarkInfo(
50+
name: "UTF16Decode_InitDecoding_ascii",
51+
runFunction: run_UTF16Decode_InitDecoding_ascii,
52+
tags: [.validation, .api, .String]),
53+
BenchmarkInfo(
54+
name: "UTF16Decode_InitFromData_ascii_as_ascii",
55+
runFunction: run_UTF16Decode_InitFromData_ascii_as_ascii,
56+
tags: [.validation, .api, .String]),
57+
]
58+
59+
typealias CodeUnit = UInt16
60+
61+
// 1-byte sequences
62+
// This test case is the longest as it's the most performance sensitive.
63+
let ascii = "Swift is a multi-paradigm, compiled programming language created for iOS, OS X, watchOS, tvOS and Linux development by Apple Inc. Swift is designed to work with Apple's Cocoa and Cocoa Touch frameworks and the large body of existing Objective-C code written for Apple products. Swift is intended to be more resilient to erroneous code (\"safer\") than Objective-C and also more concise. It is built with the LLVM compiler framework included in Xcode 6 and later and uses the Objective-C runtime, which allows C, Objective-C, C++ and Swift code to run within a single program."
64+
let asciiCodeUnits: [CodeUnit] = Array(ascii.utf16)
65+
let asciiData: Data = asciiCodeUnits.withUnsafeBytes { Data($0) }
66+
67+
// 2-byte sequences
68+
let russian = "Ру́сский язы́к один из восточнославянских языков, национальный язык русского народа."
69+
// 3-byte sequences
70+
let japanese = "日本語（にほんご、にっぽんご）は、主に日本国内や日本人同士の間で使われている言語である。"
71+
// 4-byte sequences
72+
// Most commonly emoji, which are usually mixed with other text.
73+
let emoji = "Panda 🐼, Dog 🐶, Cat 🐱, Mouse 🐭."
74+
75+
let allStrings: [[CodeUnit]] = [ascii, russian, japanese, emoji].map { Array($0.utf16) }
76+
let allStringsCodeUnits: [CodeUnit] = Array(allStrings.joined())
77+
let allStringsData: Data = allStringsCodeUnits.withUnsafeBytes { Data($0) }
78+
79+
80+
@inline(never)
81+
public func run_UTF16Decode(_ N: Int) {
82+
let strings = allStrings
83+
84+
func isEmpty(_ result: UnicodeDecodingResult) -> Bool {
85+
switch result {
86+
case .emptyInput:
87+
return true
88+
default:
89+
return false
90+
}
91+
}
92+
93+
for _ in 1...200*N {
94+
for string in strings {
95+
var it = string.makeIterator()
96+
var utf16 = UTF16()
97+
while !isEmpty(utf16.decode(&it)) { }
98+
}
99+
}
100+
}
101+
102+
@inline(never)
103+
public func run_UTF16Decode_InitFromData(_ N: Int) {
104+
let input = allStringsData
105+
for _ in 0..<200*N {
106+
blackHole(String(data: input, encoding: .utf16))
107+
}
108+
}
109+
@inline(never)
110+
public func run_UTF16Decode_InitDecoding(_ N: Int) {
111+
let input: [CodeUnit] = allStringsCodeUnits
112+
for _ in 0..<200*N {
113+
blackHole(String(decoding: input, as: UTF16.self))
114+
}
115+
}
116+
117+
@inline(never)
118+
public func run_UTF16Decode_InitFromData_ascii(_ N: Int) {
119+
let input = asciiData
120+
for _ in 0..<1_000*N {
121+
blackHole(String(data: input, encoding: .utf16))
122+
}
123+
}
124+
@inline(never)
125+
public func run_UTF16Decode_InitDecoding_ascii(_ N: Int) {
126+
let input = asciiCodeUnits
127+
for _ in 0..<1_000*N {
128+
blackHole(String(decoding: input, as: UTF16.self))
129+
}
130+
}
131+
132+
@inline(never)
133+
public func run_UTF16Decode_InitFromData_ascii_as_ascii(_ N: Int) {
134+
let input = asciiData
135+
for _ in 0..<1_000*N {
136+
blackHole(String(data: input, encoding: .ascii))
137+
}
138+
}
139+
140+
struct CustomContiguousCollection: Collection {
141+
let storage: [CodeUnit]
142+
typealias Index = Int
143+
typealias Element = CodeUnit
144+
145+
init(_ codeUnits: [CodeUnit]) { self.storage = codeUnits }
146+
subscript(position: Int) -> Element { self.storage[position] }
147+
var startIndex: Index { 0 }
148+
var endIndex: Index { storage.count }
149+
func index(after i: Index) -> Index { i+1 }
150+
151+
@inline(never)
152+
func withContiguousStorageIfAvailable<R>(
153+
_ body: (UnsafeBufferPointer<CodeUnit>) throws -> R
154+
) rethrows -> R? {
155+
try storage.withContiguousStorageIfAvailable(body)
156+
}
157+
}
158+
struct CustomNoncontiguousCollection: Collection {
159+
let storage: [CodeUnit]
160+
typealias Index = Int
161+
typealias Element = CodeUnit
162+
163+
init(_ codeUnits: [CodeUnit]) { self.storage = codeUnits }
164+
subscript(position: Int) -> Element { self.storage[position] }
165+
var startIndex: Index { 0 }
166+
var endIndex: Index { storage.count }
167+
func index(after i: Index) -> Index { i+1 }
168+
169+
@inline(never)
170+
func withContiguousStorageIfAvailable<R>(
171+
_ body: (UnsafeBufferPointer<UInt8>) throws -> R
172+
) rethrows -> R? {
173+
nil
174+
}
175+
}
176+
let allStringsCustomContiguous = CustomContiguousCollection(allStringsCodeUnits)
177+
let asciiCustomContiguous = CustomContiguousCollection(Array(ascii.utf16))
178+
let allStringsCustomNoncontiguous = CustomNoncontiguousCollection(allStringsCodeUnits)
179+
let asciiCustomNoncontiguous = CustomNoncontiguousCollection(Array(ascii.utf16))
180+
181+
@inline(never)
182+
public func run_UTF16Decode_InitFromCustom_contiguous(_ N: Int) {
183+
let input = allStringsCustomContiguous
184+
for _ in 0..<200*N {
185+
blackHole(String(decoding: input, as: UTF16.self))
186+
}
187+
}
188+
@inline(never)
189+
public func run_UTF16Decode_InitFromCustom_contiguous_ascii(_ N: Int) {
190+
let input = asciiCustomContiguous
191+
for _ in 0..<1_000*N {
192+
blackHole(String(decoding: input, as: UTF16.self))
193+
}
194+
}
195+
196+
@inline(never)
197+
public func run_UTF16Decode_InitFromCustom_noncontiguous(_ N: Int) {
198+
let input = allStringsCustomNoncontiguous
199+
for _ in 0..<200*N {
200+
blackHole(String(decoding: input, as: UTF16.self))
201+
}
202+
}
203+
@inline(never)
204+
public func run_UTF16Decode_InitFromCustom_noncontiguous_ascii(_ N: Int) {
205+
let input = asciiCustomNoncontiguous
206+
for _ in 0..<1_000*N {
207+
blackHole(String(decoding: input, as: UTF16.self))
208+
}
209+
}
210+

‎benchmark/utils/main.swift

+2
Original file line numberDiff line numberDiff line change
@@ -199,6 +199,7 @@ import SuperChars
199199
import TwoSum
200200
import TypeFlood
201201
import UTF8Decode
202+
import UTF16Decode
202203
import Walsh
203204
import WordCount
204205
import XorLoop
@@ -393,6 +394,7 @@ register(SuperChars.benchmarks)
393394
register(TwoSum.benchmarks)
394395
register(TypeFlood.benchmarks)
395396
register(UTF8Decode.benchmarks)
397+
register(UTF16Decode.benchmarks)
396398
register(Walsh.benchmarks)
397399
register(WordCount.benchmarks)
398400
register(XorLoop.benchmarks)

0 commit comments

Comments
 (0)