github.com/twelsh-aw/go/src@v0.0.0-20230516233729-a56fe86a7c81/strings/strings_test.go (about) 1 // Copyright 2009 The Go Authors. All rights reserved. 2 // Use of this source code is governed by a BSD-style 3 // license that can be found in the LICENSE file. 4 5 package strings_test 6 7 import ( 8 "bytes" 9 "fmt" 10 "io" 11 "math" 12 "math/rand" 13 "reflect" 14 "strconv" 15 . "strings" 16 "testing" 17 "unicode" 18 "unicode/utf8" 19 "unsafe" 20 ) 21 22 func eq(a, b []string) bool { 23 if len(a) != len(b) { 24 return false 25 } 26 for i := 0; i < len(a); i++ { 27 if a[i] != b[i] { 28 return false 29 } 30 } 31 return true 32 } 33 34 var abcd = "abcd" 35 var faces = "☺☻☹" 36 var commas = "1,2,3,4" 37 var dots = "1....2....3....4" 38 39 type IndexTest struct { 40 s string 41 sep string 42 out int 43 } 44 45 var indexTests = []IndexTest{ 46 {"", "", 0}, 47 {"", "a", -1}, 48 {"", "foo", -1}, 49 {"fo", "foo", -1}, 50 {"foo", "foo", 0}, 51 {"oofofoofooo", "f", 2}, 52 {"oofofoofooo", "foo", 4}, 53 {"barfoobarfoo", "foo", 3}, 54 {"foo", "", 0}, 55 {"foo", "o", 1}, 56 {"abcABCabc", "A", 3}, 57 {"jrzm6jjhorimglljrea4w3rlgosts0w2gia17hno2td4qd1jz", "jz", 47}, 58 {"ekkuk5oft4eq0ocpacknhwouic1uua46unx12l37nioq9wbpnocqks6", "ks6", 52}, 59 {"999f2xmimunbuyew5vrkla9cpwhmxan8o98ec", "98ec", 33}, 60 {"9lpt9r98i04k8bz6c6dsrthb96bhi", "96bhi", 24}, 61 {"55u558eqfaod2r2gu42xxsu631xf0zobs5840vl", "5840vl", 33}, 62 // cases with one byte strings - test special case in Index() 63 {"", "a", -1}, 64 {"x", "a", -1}, 65 {"x", "x", 0}, 66 {"abc", "a", 0}, 67 {"abc", "b", 1}, 68 {"abc", "c", 2}, 69 {"abc", "x", -1}, 70 // test special cases in Index() for short strings 71 {"", "ab", -1}, 72 {"bc", "ab", -1}, 73 {"ab", "ab", 0}, 74 {"xab", "ab", 1}, 75 {"xab"[:2], "ab", -1}, 76 {"", "abc", -1}, 77 {"xbc", "abc", -1}, 78 {"abc", "abc", 0}, 79 {"xabc", "abc", 1}, 80 {"xabc"[:3], "abc", -1}, 81 {"xabxc", "abc", -1}, 82 {"", "abcd", -1}, 83 {"xbcd", "abcd", -1}, 84 {"abcd", "abcd", 0}, 85 {"xabcd", "abcd", 1}, 86 {"xyabcd"[:5], "abcd", -1}, 87 {"xbcqq", "abcqq", -1}, 88 {"abcqq", "abcqq", 0}, 89 {"xabcqq", "abcqq", 1}, 90 {"xyabcqq"[:6], "abcqq", -1}, 91 {"xabxcqq", "abcqq", -1}, 92 {"xabcqxq", "abcqq", -1}, 93 {"", "01234567", -1}, 94 {"32145678", "01234567", -1}, 95 {"01234567", "01234567", 0}, 96 {"x01234567", "01234567", 1}, 97 {"x0123456x01234567", "01234567", 9}, 98 {"xx01234567"[:9], "01234567", -1}, 99 {"", "0123456789", -1}, 100 {"3214567844", "0123456789", -1}, 101 {"0123456789", "0123456789", 0}, 102 {"x0123456789", "0123456789", 1}, 103 {"x012345678x0123456789", "0123456789", 11}, 104 {"xyz0123456789"[:12], "0123456789", -1}, 105 {"x01234567x89", "0123456789", -1}, 106 {"", "0123456789012345", -1}, 107 {"3214567889012345", "0123456789012345", -1}, 108 {"0123456789012345", "0123456789012345", 0}, 109 {"x0123456789012345", "0123456789012345", 1}, 110 {"x012345678901234x0123456789012345", "0123456789012345", 17}, 111 {"", "01234567890123456789", -1}, 112 {"32145678890123456789", "01234567890123456789", -1}, 113 {"01234567890123456789", "01234567890123456789", 0}, 114 {"x01234567890123456789", "01234567890123456789", 1}, 115 {"x0123456789012345678x01234567890123456789", "01234567890123456789", 21}, 116 {"xyz01234567890123456789"[:22], "01234567890123456789", -1}, 117 {"", "0123456789012345678901234567890", -1}, 118 {"321456788901234567890123456789012345678911", "0123456789012345678901234567890", -1}, 119 {"0123456789012345678901234567890", "0123456789012345678901234567890", 0}, 120 {"x0123456789012345678901234567890", "0123456789012345678901234567890", 1}, 121 {"x012345678901234567890123456789x0123456789012345678901234567890", "0123456789012345678901234567890", 32}, 122 {"xyz0123456789012345678901234567890"[:33], "0123456789012345678901234567890", -1}, 123 {"", "01234567890123456789012345678901", -1}, 124 {"32145678890123456789012345678901234567890211", "01234567890123456789012345678901", -1}, 125 {"01234567890123456789012345678901", "01234567890123456789012345678901", 0}, 126 {"x01234567890123456789012345678901", "01234567890123456789012345678901", 1}, 127 {"x0123456789012345678901234567890x01234567890123456789012345678901", "01234567890123456789012345678901", 33}, 128 {"xyz01234567890123456789012345678901"[:34], "01234567890123456789012345678901", -1}, 129 {"xxxxxx012345678901234567890123456789012345678901234567890123456789012", "012345678901234567890123456789012345678901234567890123456789012", 6}, 130 {"", "0123456789012345678901234567890123456789", -1}, 131 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456789", 2}, 132 {"xx012345678901234567890123456789012345678901234567890123456789012"[:41], "0123456789012345678901234567890123456789", -1}, 133 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456xxx", -1}, 134 {"xx0123456789012345678901234567890123456789012345678901234567890120123456789012345678901234567890123456xxx", "0123456789012345678901234567890123456xxx", 65}, 135 // test fallback to Rabin-Karp. 136 {"oxoxoxoxoxoxoxoxoxoxoxoy", "oy", 22}, 137 {"oxoxoxoxoxoxoxoxoxoxoxox", "oy", -1}, 138 } 139 140 var lastIndexTests = []IndexTest{ 141 {"", "", 0}, 142 {"", "a", -1}, 143 {"", "foo", -1}, 144 {"fo", "foo", -1}, 145 {"foo", "foo", 0}, 146 {"foo", "f", 0}, 147 {"oofofoofooo", "f", 7}, 148 {"oofofoofooo", "foo", 7}, 149 {"barfoobarfoo", "foo", 9}, 150 {"foo", "", 3}, 151 {"foo", "o", 2}, 152 {"abcABCabc", "A", 3}, 153 {"abcABCabc", "a", 6}, 154 } 155 156 var indexAnyTests = []IndexTest{ 157 {"", "", -1}, 158 {"", "a", -1}, 159 {"", "abc", -1}, 160 {"a", "", -1}, 161 {"a", "a", 0}, 162 {"\x80", "\xffb", 0}, 163 {"aaa", "a", 0}, 164 {"abc", "xyz", -1}, 165 {"abc", "xcz", 2}, 166 {"ab☺c", "x☺yz", 2}, 167 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 168 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 169 {"aRegExp*", ".(|)*+?^$[]", 7}, 170 {dots + dots + dots, " ", -1}, 171 {"012abcba210", "\xffb", 4}, 172 {"012\x80bcb\x80210", "\xffb", 3}, 173 {"0123456\xcf\x80abc", "\xcfb\x80", 10}, 174 } 175 176 var lastIndexAnyTests = []IndexTest{ 177 {"", "", -1}, 178 {"", "a", -1}, 179 {"", "abc", -1}, 180 {"a", "", -1}, 181 {"a", "a", 0}, 182 {"\x80", "\xffb", 0}, 183 {"aaa", "a", 2}, 184 {"abc", "xyz", -1}, 185 {"abc", "ab", 1}, 186 {"ab☺c", "x☺yz", 2}, 187 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 188 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 189 {"a.RegExp*", ".(|)*+?^$[]", 8}, 190 {dots + dots + dots, " ", -1}, 191 {"012abcba210", "\xffb", 6}, 192 {"012\x80bcb\x80210", "\xffb", 7}, 193 {"0123456\xcf\x80abc", "\xcfb\x80", 10}, 194 } 195 196 // Execute f on each test case. funcName should be the name of f; it's used 197 // in failure reports. 198 func runIndexTests(t *testing.T, f func(s, sep string) int, funcName string, testCases []IndexTest) { 199 for _, test := range testCases { 200 actual := f(test.s, test.sep) 201 if actual != test.out { 202 t.Errorf("%s(%q,%q) = %v; want %v", funcName, test.s, test.sep, actual, test.out) 203 } 204 } 205 } 206 207 func TestIndex(t *testing.T) { runIndexTests(t, Index, "Index", indexTests) } 208 func TestLastIndex(t *testing.T) { runIndexTests(t, LastIndex, "LastIndex", lastIndexTests) } 209 func TestIndexAny(t *testing.T) { runIndexTests(t, IndexAny, "IndexAny", indexAnyTests) } 210 func TestLastIndexAny(t *testing.T) { 211 runIndexTests(t, LastIndexAny, "LastIndexAny", lastIndexAnyTests) 212 } 213 214 func TestIndexByte(t *testing.T) { 215 for _, tt := range indexTests { 216 if len(tt.sep) != 1 { 217 continue 218 } 219 pos := IndexByte(tt.s, tt.sep[0]) 220 if pos != tt.out { 221 t.Errorf(`IndexByte(%q, %q) = %v; want %v`, tt.s, tt.sep[0], pos, tt.out) 222 } 223 } 224 } 225 226 func TestLastIndexByte(t *testing.T) { 227 testCases := []IndexTest{ 228 {"", "q", -1}, 229 {"abcdef", "q", -1}, 230 {"abcdefabcdef", "a", len("abcdef")}, // something in the middle 231 {"abcdefabcdef", "f", len("abcdefabcde")}, // last byte 232 {"zabcdefabcdef", "z", 0}, // first byte 233 {"a☺b☻c☹d", "b", len("a☺")}, // non-ascii 234 } 235 for _, test := range testCases { 236 actual := LastIndexByte(test.s, test.sep[0]) 237 if actual != test.out { 238 t.Errorf("LastIndexByte(%q,%c) = %v; want %v", test.s, test.sep[0], actual, test.out) 239 } 240 } 241 } 242 243 func simpleIndex(s, sep string) int { 244 n := len(sep) 245 for i := n; i <= len(s); i++ { 246 if s[i-n:i] == sep { 247 return i - n 248 } 249 } 250 return -1 251 } 252 253 func TestIndexRandom(t *testing.T) { 254 const chars = "abcdefghijklmnopqrstuvwxyz0123456789" 255 for times := 0; times < 10; times++ { 256 for strLen := 5 + rand.Intn(5); strLen < 140; strLen += 10 { // Arbitrary 257 s1 := make([]byte, strLen) 258 for i := range s1 { 259 s1[i] = chars[rand.Intn(len(chars))] 260 } 261 s := string(s1) 262 for i := 0; i < 50; i++ { 263 begin := rand.Intn(len(s) + 1) 264 end := begin + rand.Intn(len(s)+1-begin) 265 sep := s[begin:end] 266 if i%4 == 0 { 267 pos := rand.Intn(len(sep) + 1) 268 sep = sep[:pos] + "A" + sep[pos:] 269 } 270 want := simpleIndex(s, sep) 271 res := Index(s, sep) 272 if res != want { 273 t.Errorf("Index(%s,%s) = %d; want %d", s, sep, res, want) 274 } 275 } 276 } 277 } 278 } 279 280 func TestIndexRune(t *testing.T) { 281 tests := []struct { 282 in string 283 rune rune 284 want int 285 }{ 286 {"", 'a', -1}, 287 {"", '☺', -1}, 288 {"foo", '☹', -1}, 289 {"foo", 'o', 1}, 290 {"foo☺bar", '☺', 3}, 291 {"foo☺☻☹bar", '☹', 9}, 292 {"a A x", 'A', 2}, 293 {"some_text=some_value", '=', 9}, 294 {"☺a", 'a', 3}, 295 {"a☻☺b", '☺', 4}, 296 297 // RuneError should match any invalid UTF-8 byte sequence. 298 {"�", '�', 0}, 299 {"\xff", '�', 0}, 300 {"☻x�", '�', len("☻x")}, 301 {"☻x\xe2\x98", '�', len("☻x")}, 302 {"☻x\xe2\x98�", '�', len("☻x")}, 303 {"☻x\xe2\x98x", '�', len("☻x")}, 304 305 // Invalid rune values should never match. 306 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", -1, -1}, 307 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", 0xD800, -1}, // Surrogate pair 308 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", utf8.MaxRune + 1, -1}, 309 } 310 for _, tt := range tests { 311 if got := IndexRune(tt.in, tt.rune); got != tt.want { 312 t.Errorf("IndexRune(%q, %d) = %v; want %v", tt.in, tt.rune, got, tt.want) 313 } 314 } 315 316 haystack := "test世界" 317 allocs := testing.AllocsPerRun(1000, func() { 318 if i := IndexRune(haystack, 's'); i != 2 { 319 t.Fatalf("'s' at %d; want 2", i) 320 } 321 if i := IndexRune(haystack, '世'); i != 4 { 322 t.Fatalf("'世' at %d; want 4", i) 323 } 324 }) 325 if allocs != 0 && testing.CoverMode() == "" { 326 t.Errorf("expected no allocations, got %f", allocs) 327 } 328 } 329 330 const benchmarkString = "some_text=some☺value" 331 332 func BenchmarkIndexRune(b *testing.B) { 333 if got := IndexRune(benchmarkString, '☺'); got != 14 { 334 b.Fatalf("wrong index: expected 14, got=%d", got) 335 } 336 for i := 0; i < b.N; i++ { 337 IndexRune(benchmarkString, '☺') 338 } 339 } 340 341 var benchmarkLongString = Repeat(" ", 100) + benchmarkString 342 343 func BenchmarkIndexRuneLongString(b *testing.B) { 344 if got := IndexRune(benchmarkLongString, '☺'); got != 114 { 345 b.Fatalf("wrong index: expected 114, got=%d", got) 346 } 347 for i := 0; i < b.N; i++ { 348 IndexRune(benchmarkLongString, '☺') 349 } 350 } 351 352 func BenchmarkIndexRuneFastPath(b *testing.B) { 353 if got := IndexRune(benchmarkString, 'v'); got != 17 { 354 b.Fatalf("wrong index: expected 17, got=%d", got) 355 } 356 for i := 0; i < b.N; i++ { 357 IndexRune(benchmarkString, 'v') 358 } 359 } 360 361 func BenchmarkIndex(b *testing.B) { 362 if got := Index(benchmarkString, "v"); got != 17 { 363 b.Fatalf("wrong index: expected 17, got=%d", got) 364 } 365 for i := 0; i < b.N; i++ { 366 Index(benchmarkString, "v") 367 } 368 } 369 370 func BenchmarkLastIndex(b *testing.B) { 371 if got := Index(benchmarkString, "v"); got != 17 { 372 b.Fatalf("wrong index: expected 17, got=%d", got) 373 } 374 for i := 0; i < b.N; i++ { 375 LastIndex(benchmarkString, "v") 376 } 377 } 378 379 func BenchmarkIndexByte(b *testing.B) { 380 if got := IndexByte(benchmarkString, 'v'); got != 17 { 381 b.Fatalf("wrong index: expected 17, got=%d", got) 382 } 383 for i := 0; i < b.N; i++ { 384 IndexByte(benchmarkString, 'v') 385 } 386 } 387 388 type SplitTest struct { 389 s string 390 sep string 391 n int 392 a []string 393 } 394 395 var splittests = []SplitTest{ 396 {"", "", -1, []string{}}, 397 {abcd, "", 2, []string{"a", "bcd"}}, 398 {abcd, "", 4, []string{"a", "b", "c", "d"}}, 399 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 400 {faces, "", -1, []string{"☺", "☻", "☹"}}, 401 {faces, "", 3, []string{"☺", "☻", "☹"}}, 402 {faces, "", 17, []string{"☺", "☻", "☹"}}, 403 {"☺�☹", "", -1, []string{"☺", "�", "☹"}}, 404 {abcd, "a", 0, nil}, 405 {abcd, "a", -1, []string{"", "bcd"}}, 406 {abcd, "z", -1, []string{"abcd"}}, 407 {commas, ",", -1, []string{"1", "2", "3", "4"}}, 408 {dots, "...", -1, []string{"1", ".2", ".3", ".4"}}, 409 {faces, "☹", -1, []string{"☺☻", ""}}, 410 {faces, "~", -1, []string{faces}}, 411 {"1 2 3 4", " ", 3, []string{"1", "2", "3 4"}}, 412 {"1 2", " ", 3, []string{"1", "2"}}, 413 {"", "T", math.MaxInt / 4, []string{""}}, 414 {"\xff-\xff", "", -1, []string{"\xff", "-", "\xff"}}, 415 {"\xff-\xff", "-", -1, []string{"\xff", "\xff"}}, 416 } 417 418 func TestSplit(t *testing.T) { 419 for _, tt := range splittests { 420 a := SplitN(tt.s, tt.sep, tt.n) 421 if !eq(a, tt.a) { 422 t.Errorf("Split(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, a, tt.a) 423 continue 424 } 425 if tt.n == 0 { 426 continue 427 } 428 s := Join(a, tt.sep) 429 if s != tt.s { 430 t.Errorf("Join(Split(%q, %q, %d), %q) = %q", tt.s, tt.sep, tt.n, tt.sep, s) 431 } 432 if tt.n < 0 { 433 b := Split(tt.s, tt.sep) 434 if !reflect.DeepEqual(a, b) { 435 t.Errorf("Split disagrees with SplitN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 436 } 437 } 438 } 439 } 440 441 var splitaftertests = []SplitTest{ 442 {abcd, "a", -1, []string{"a", "bcd"}}, 443 {abcd, "z", -1, []string{"abcd"}}, 444 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 445 {commas, ",", -1, []string{"1,", "2,", "3,", "4"}}, 446 {dots, "...", -1, []string{"1...", ".2...", ".3...", ".4"}}, 447 {faces, "☹", -1, []string{"☺☻☹", ""}}, 448 {faces, "~", -1, []string{faces}}, 449 {faces, "", -1, []string{"☺", "☻", "☹"}}, 450 {"1 2 3 4", " ", 3, []string{"1 ", "2 ", "3 4"}}, 451 {"1 2 3", " ", 3, []string{"1 ", "2 ", "3"}}, 452 {"1 2", " ", 3, []string{"1 ", "2"}}, 453 {"123", "", 2, []string{"1", "23"}}, 454 {"123", "", 17, []string{"1", "2", "3"}}, 455 } 456 457 func TestSplitAfter(t *testing.T) { 458 for _, tt := range splitaftertests { 459 a := SplitAfterN(tt.s, tt.sep, tt.n) 460 if !eq(a, tt.a) { 461 t.Errorf(`Split(%q, %q, %d) = %v; want %v`, tt.s, tt.sep, tt.n, a, tt.a) 462 continue 463 } 464 s := Join(a, "") 465 if s != tt.s { 466 t.Errorf(`Join(Split(%q, %q, %d), %q) = %q`, tt.s, tt.sep, tt.n, tt.sep, s) 467 } 468 if tt.n < 0 { 469 b := SplitAfter(tt.s, tt.sep) 470 if !reflect.DeepEqual(a, b) { 471 t.Errorf("SplitAfter disagrees with SplitAfterN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 472 } 473 } 474 } 475 } 476 477 type FieldsTest struct { 478 s string 479 a []string 480 } 481 482 var fieldstests = []FieldsTest{ 483 {"", []string{}}, 484 {" ", []string{}}, 485 {" \t ", []string{}}, 486 {"\u2000", []string{}}, 487 {" abc ", []string{"abc"}}, 488 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 489 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 490 {"1\t\t2\t\t3\t4", []string{"1", "2", "3", "4"}}, 491 {"1\u20002\u20013\u20024", []string{"1", "2", "3", "4"}}, 492 {"\u2000\u2001\u2002", []string{}}, 493 {"\n™\t™\n", []string{"™", "™"}}, 494 {"\n\u20001™2\u2000 \u2001 ™", []string{"1™2", "™"}}, 495 {"\n1\uFFFD \uFFFD2\u20003\uFFFD4", []string{"1\uFFFD", "\uFFFD2", "3\uFFFD4"}}, 496 {"1\xFF\u2000\xFF2\xFF \xFF", []string{"1\xFF", "\xFF2\xFF", "\xFF"}}, 497 {faces, []string{faces}}, 498 } 499 500 func TestFields(t *testing.T) { 501 for _, tt := range fieldstests { 502 a := Fields(tt.s) 503 if !eq(a, tt.a) { 504 t.Errorf("Fields(%q) = %v; want %v", tt.s, a, tt.a) 505 continue 506 } 507 } 508 } 509 510 var FieldsFuncTests = []FieldsTest{ 511 {"", []string{}}, 512 {"XX", []string{}}, 513 {"XXhiXXX", []string{"hi"}}, 514 {"aXXbXXXcX", []string{"a", "b", "c"}}, 515 } 516 517 func TestFieldsFunc(t *testing.T) { 518 for _, tt := range fieldstests { 519 a := FieldsFunc(tt.s, unicode.IsSpace) 520 if !eq(a, tt.a) { 521 t.Errorf("FieldsFunc(%q, unicode.IsSpace) = %v; want %v", tt.s, a, tt.a) 522 continue 523 } 524 } 525 pred := func(c rune) bool { return c == 'X' } 526 for _, tt := range FieldsFuncTests { 527 a := FieldsFunc(tt.s, pred) 528 if !eq(a, tt.a) { 529 t.Errorf("FieldsFunc(%q) = %v, want %v", tt.s, a, tt.a) 530 } 531 } 532 } 533 534 // Test case for any function which accepts and returns a single string. 535 type StringTest struct { 536 in, out string 537 } 538 539 // Execute f on each test case. funcName should be the name of f; it's used 540 // in failure reports. 541 func runStringTests(t *testing.T, f func(string) string, funcName string, testCases []StringTest) { 542 for _, tc := range testCases { 543 actual := f(tc.in) 544 if actual != tc.out { 545 t.Errorf("%s(%q) = %q; want %q", funcName, tc.in, actual, tc.out) 546 } 547 } 548 } 549 550 var upperTests = []StringTest{ 551 {"", ""}, 552 {"ONLYUPPER", "ONLYUPPER"}, 553 {"abc", "ABC"}, 554 {"AbC123", "ABC123"}, 555 {"azAZ09_", "AZAZ09_"}, 556 {"longStrinGwitHmixofsmaLLandcAps", "LONGSTRINGWITHMIXOFSMALLANDCAPS"}, 557 {"RENAN BASTOS 93 AOSDAJDJAIDJAIDAJIaidsjjaidijadsjiadjiOOKKO", "RENAN BASTOS 93 AOSDAJDJAIDJAIDAJIAIDSJJAIDIJADSJIADJIOOKKO"}, 558 {"long\u0250string\u0250with\u0250nonascii\u2C6Fchars", "LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS"}, 559 {"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"}, // grows one byte per char 560 {"a\u0080\U0010FFFF", "A\u0080\U0010FFFF"}, // test utf8.RuneSelf and utf8.MaxRune 561 } 562 563 var lowerTests = []StringTest{ 564 {"", ""}, 565 {"abc", "abc"}, 566 {"AbC123", "abc123"}, 567 {"azAZ09_", "azaz09_"}, 568 {"longStrinGwitHmixofsmaLLandcAps", "longstringwithmixofsmallandcaps"}, 569 {"renan bastos 93 AOSDAJDJAIDJAIDAJIaidsjjaidijadsjiadjiOOKKO", "renan bastos 93 aosdajdjaidjaidajiaidsjjaidijadsjiadjiookko"}, 570 {"LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS", "long\u0250string\u0250with\u0250nonascii\u0250chars"}, 571 {"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"}, // shrinks one byte per char 572 {"A\u0080\U0010FFFF", "a\u0080\U0010FFFF"}, // test utf8.RuneSelf and utf8.MaxRune 573 } 574 575 const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000" 576 577 var trimSpaceTests = []StringTest{ 578 {"", ""}, 579 {"abc", "abc"}, 580 {space + "abc" + space, "abc"}, 581 {" ", ""}, 582 {" \t\r\n \t\t\r\r\n\n ", ""}, 583 {" \t\r\n x\t\t\r\r\n\n ", "x"}, 584 {" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, 585 {"1 \t\r\n2", "1 \t\r\n2"}, 586 {" x\x80", "x\x80"}, 587 {" x\xc0", "x\xc0"}, 588 {"x \xc0\xc0 ", "x \xc0\xc0"}, 589 {"x \xc0", "x \xc0"}, 590 {"x \xc0 ", "x \xc0"}, 591 {"x \xc0\xc0 ", "x \xc0\xc0"}, 592 {"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, 593 {"x ☺ ", "x ☺"}, 594 } 595 596 func tenRunes(ch rune) string { 597 r := make([]rune, 10) 598 for i := range r { 599 r[i] = ch 600 } 601 return string(r) 602 } 603 604 // User-defined self-inverse mapping function 605 func rot13(r rune) rune { 606 step := rune(13) 607 if r >= 'a' && r <= 'z' { 608 return ((r - 'a' + step) % 26) + 'a' 609 } 610 if r >= 'A' && r <= 'Z' { 611 return ((r - 'A' + step) % 26) + 'A' 612 } 613 return r 614 } 615 616 func TestMap(t *testing.T) { 617 // Run a couple of awful growth/shrinkage tests 618 a := tenRunes('a') 619 // 1. Grow. This triggers two reallocations in Map. 620 maxRune := func(rune) rune { return unicode.MaxRune } 621 m := Map(maxRune, a) 622 expect := tenRunes(unicode.MaxRune) 623 if m != expect { 624 t.Errorf("growing: expected %q got %q", expect, m) 625 } 626 627 // 2. Shrink 628 minRune := func(rune) rune { return 'a' } 629 m = Map(minRune, tenRunes(unicode.MaxRune)) 630 expect = a 631 if m != expect { 632 t.Errorf("shrinking: expected %q got %q", expect, m) 633 } 634 635 // 3. Rot13 636 m = Map(rot13, "a to zed") 637 expect = "n gb mrq" 638 if m != expect { 639 t.Errorf("rot13: expected %q got %q", expect, m) 640 } 641 642 // 4. Rot13^2 643 m = Map(rot13, Map(rot13, "a to zed")) 644 expect = "a to zed" 645 if m != expect { 646 t.Errorf("rot13: expected %q got %q", expect, m) 647 } 648 649 // 5. Drop 650 dropNotLatin := func(r rune) rune { 651 if unicode.Is(unicode.Latin, r) { 652 return r 653 } 654 return -1 655 } 656 m = Map(dropNotLatin, "Hello, 세계") 657 expect = "Hello" 658 if m != expect { 659 t.Errorf("drop: expected %q got %q", expect, m) 660 } 661 662 // 6. Identity 663 identity := func(r rune) rune { 664 return r 665 } 666 orig := "Input string that we expect not to be copied." 667 m = Map(identity, orig) 668 if unsafe.StringData(orig) != unsafe.StringData(m) { 669 t.Error("unexpected copy during identity map") 670 } 671 672 // 7. Handle invalid UTF-8 sequence 673 replaceNotLatin := func(r rune) rune { 674 if unicode.Is(unicode.Latin, r) { 675 return r 676 } 677 return utf8.RuneError 678 } 679 m = Map(replaceNotLatin, "Hello\255World") 680 expect = "Hello\uFFFDWorld" 681 if m != expect { 682 t.Errorf("replace invalid sequence: expected %q got %q", expect, m) 683 } 684 685 // 8. Check utf8.RuneSelf and utf8.MaxRune encoding 686 encode := func(r rune) rune { 687 switch r { 688 case utf8.RuneSelf: 689 return unicode.MaxRune 690 case unicode.MaxRune: 691 return utf8.RuneSelf 692 } 693 return r 694 } 695 s := string(rune(utf8.RuneSelf)) + string(utf8.MaxRune) 696 r := string(utf8.MaxRune) + string(rune(utf8.RuneSelf)) // reverse of s 697 m = Map(encode, s) 698 if m != r { 699 t.Errorf("encoding not handled correctly: expected %q got %q", r, m) 700 } 701 m = Map(encode, r) 702 if m != s { 703 t.Errorf("encoding not handled correctly: expected %q got %q", s, m) 704 } 705 706 // 9. Check mapping occurs in the front, middle and back 707 trimSpaces := func(r rune) rune { 708 if unicode.IsSpace(r) { 709 return -1 710 } 711 return r 712 } 713 m = Map(trimSpaces, " abc 123 ") 714 expect = "abc123" 715 if m != expect { 716 t.Errorf("trimSpaces: expected %q got %q", expect, m) 717 } 718 } 719 720 func TestToUpper(t *testing.T) { runStringTests(t, ToUpper, "ToUpper", upperTests) } 721 722 func TestToLower(t *testing.T) { runStringTests(t, ToLower, "ToLower", lowerTests) } 723 724 var toValidUTF8Tests = []struct { 725 in string 726 repl string 727 out string 728 }{ 729 {"", "\uFFFD", ""}, 730 {"abc", "\uFFFD", "abc"}, 731 {"\uFDDD", "\uFFFD", "\uFDDD"}, 732 {"a\xffb", "\uFFFD", "a\uFFFDb"}, 733 {"a\xffb\uFFFD", "X", "aXb\uFFFD"}, 734 {"a☺\xffb☺\xC0\xAFc☺\xff", "", "a☺b☺c☺"}, 735 {"a☺\xffb☺\xC0\xAFc☺\xff", "日本語", "a☺日本語b☺日本語c☺日本語"}, 736 {"\xC0\xAF", "\uFFFD", "\uFFFD"}, 737 {"\xE0\x80\xAF", "\uFFFD", "\uFFFD"}, 738 {"\xed\xa0\x80", "abc", "abc"}, 739 {"\xed\xbf\xbf", "\uFFFD", "\uFFFD"}, 740 {"\xF0\x80\x80\xaf", "☺", "☺"}, 741 {"\xF8\x80\x80\x80\xAF", "\uFFFD", "\uFFFD"}, 742 {"\xFC\x80\x80\x80\x80\xAF", "\uFFFD", "\uFFFD"}, 743 } 744 745 func TestToValidUTF8(t *testing.T) { 746 for _, tc := range toValidUTF8Tests { 747 got := ToValidUTF8(tc.in, tc.repl) 748 if got != tc.out { 749 t.Errorf("ToValidUTF8(%q, %q) = %q; want %q", tc.in, tc.repl, got, tc.out) 750 } 751 } 752 } 753 754 func BenchmarkToUpper(b *testing.B) { 755 for _, tc := range upperTests { 756 b.Run(tc.in, func(b *testing.B) { 757 for i := 0; i < b.N; i++ { 758 actual := ToUpper(tc.in) 759 if actual != tc.out { 760 b.Errorf("ToUpper(%q) = %q; want %q", tc.in, actual, tc.out) 761 } 762 } 763 }) 764 } 765 } 766 767 func BenchmarkToLower(b *testing.B) { 768 for _, tc := range lowerTests { 769 b.Run(tc.in, func(b *testing.B) { 770 for i := 0; i < b.N; i++ { 771 actual := ToLower(tc.in) 772 if actual != tc.out { 773 b.Errorf("ToLower(%q) = %q; want %q", tc.in, actual, tc.out) 774 } 775 } 776 }) 777 } 778 } 779 780 func BenchmarkMapNoChanges(b *testing.B) { 781 identity := func(r rune) rune { 782 return r 783 } 784 for i := 0; i < b.N; i++ { 785 Map(identity, "Some string that won't be modified.") 786 } 787 } 788 789 func TestSpecialCase(t *testing.T) { 790 lower := "abcçdefgğhıijklmnoöprsştuüvyz" 791 upper := "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ" 792 u := ToUpperSpecial(unicode.TurkishCase, upper) 793 if u != upper { 794 t.Errorf("Upper(upper) is %s not %s", u, upper) 795 } 796 u = ToUpperSpecial(unicode.TurkishCase, lower) 797 if u != upper { 798 t.Errorf("Upper(lower) is %s not %s", u, upper) 799 } 800 l := ToLowerSpecial(unicode.TurkishCase, lower) 801 if l != lower { 802 t.Errorf("Lower(lower) is %s not %s", l, lower) 803 } 804 l = ToLowerSpecial(unicode.TurkishCase, upper) 805 if l != lower { 806 t.Errorf("Lower(upper) is %s not %s", l, lower) 807 } 808 } 809 810 func TestTrimSpace(t *testing.T) { runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests) } 811 812 var trimTests = []struct { 813 f string 814 in, arg, out string 815 }{ 816 {"Trim", "abba", "a", "bb"}, 817 {"Trim", "abba", "ab", ""}, 818 {"TrimLeft", "abba", "ab", ""}, 819 {"TrimRight", "abba", "ab", ""}, 820 {"TrimLeft", "abba", "a", "bba"}, 821 {"TrimLeft", "abba", "b", "abba"}, 822 {"TrimRight", "abba", "a", "abb"}, 823 {"TrimRight", "abba", "b", "abba"}, 824 {"Trim", "<tag>", "<>", "tag"}, 825 {"Trim", "* listitem", " *", "listitem"}, 826 {"Trim", `"quote"`, `"`, "quote"}, 827 {"Trim", "\u2C6F\u2C6F\u0250\u0250\u2C6F\u2C6F", "\u2C6F", "\u0250\u0250"}, 828 {"Trim", "\x80test\xff", "\xff", "test"}, 829 {"Trim", " Ġ ", " ", "Ġ"}, 830 {"Trim", " Ġİ0", "0 ", "Ġİ"}, 831 //empty string tests 832 {"Trim", "abba", "", "abba"}, 833 {"Trim", "", "123", ""}, 834 {"Trim", "", "", ""}, 835 {"TrimLeft", "abba", "", "abba"}, 836 {"TrimLeft", "", "123", ""}, 837 {"TrimLeft", "", "", ""}, 838 {"TrimRight", "abba", "", "abba"}, 839 {"TrimRight", "", "123", ""}, 840 {"TrimRight", "", "", ""}, 841 {"TrimRight", "☺\xc0", "☺", "☺\xc0"}, 842 {"TrimPrefix", "aabb", "a", "abb"}, 843 {"TrimPrefix", "aabb", "b", "aabb"}, 844 {"TrimSuffix", "aabb", "a", "aabb"}, 845 {"TrimSuffix", "aabb", "b", "aab"}, 846 } 847 848 func TestTrim(t *testing.T) { 849 for _, tc := range trimTests { 850 name := tc.f 851 var f func(string, string) string 852 switch name { 853 case "Trim": 854 f = Trim 855 case "TrimLeft": 856 f = TrimLeft 857 case "TrimRight": 858 f = TrimRight 859 case "TrimPrefix": 860 f = TrimPrefix 861 case "TrimSuffix": 862 f = TrimSuffix 863 default: 864 t.Errorf("Undefined trim function %s", name) 865 } 866 actual := f(tc.in, tc.arg) 867 if actual != tc.out { 868 t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 869 } 870 } 871 } 872 873 func BenchmarkTrim(b *testing.B) { 874 b.ReportAllocs() 875 876 for i := 0; i < b.N; i++ { 877 for _, tc := range trimTests { 878 name := tc.f 879 var f func(string, string) string 880 switch name { 881 case "Trim": 882 f = Trim 883 case "TrimLeft": 884 f = TrimLeft 885 case "TrimRight": 886 f = TrimRight 887 case "TrimPrefix": 888 f = TrimPrefix 889 case "TrimSuffix": 890 f = TrimSuffix 891 default: 892 b.Errorf("Undefined trim function %s", name) 893 } 894 actual := f(tc.in, tc.arg) 895 if actual != tc.out { 896 b.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 897 } 898 } 899 } 900 } 901 902 func BenchmarkToValidUTF8(b *testing.B) { 903 tests := []struct { 904 name string 905 input string 906 }{ 907 {"Valid", "typical"}, 908 {"InvalidASCII", "foo\xffbar"}, 909 {"InvalidNonASCII", "日本語\xff日本語"}, 910 } 911 replacement := "\uFFFD" 912 b.ResetTimer() 913 for _, test := range tests { 914 b.Run(test.name, func(b *testing.B) { 915 for i := 0; i < b.N; i++ { 916 ToValidUTF8(test.input, replacement) 917 } 918 }) 919 } 920 } 921 922 type predicate struct { 923 f func(rune) bool 924 name string 925 } 926 927 var isSpace = predicate{unicode.IsSpace, "IsSpace"} 928 var isDigit = predicate{unicode.IsDigit, "IsDigit"} 929 var isUpper = predicate{unicode.IsUpper, "IsUpper"} 930 var isValidRune = predicate{ 931 func(r rune) bool { 932 return r != utf8.RuneError 933 }, 934 "IsValidRune", 935 } 936 937 func not(p predicate) predicate { 938 return predicate{ 939 func(r rune) bool { 940 return !p.f(r) 941 }, 942 "not " + p.name, 943 } 944 } 945 946 var trimFuncTests = []struct { 947 f predicate 948 in string 949 trimOut string 950 leftOut string 951 rightOut string 952 }{ 953 {isSpace, space + " hello " + space, 954 "hello", 955 "hello " + space, 956 space + " hello"}, 957 {isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", 958 "hello", 959 "hello34\u0e50\u0e51", 960 "\u0e50\u0e5212hello"}, 961 {isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", 962 "hello", 963 "helloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", 964 "\u2C6F\u2C6F\u2C6F\u2C6FABCDhello"}, 965 {not(isSpace), "hello" + space + "hello", 966 space, 967 space + "hello", 968 "hello" + space}, 969 {not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", 970 "\u0e50\u0e521234\u0e50\u0e51", 971 "\u0e50\u0e521234\u0e50\u0e51helo", 972 "hello\u0e50\u0e521234\u0e50\u0e51"}, 973 {isValidRune, "ab\xc0a\xc0cd", 974 "\xc0a\xc0", 975 "\xc0a\xc0cd", 976 "ab\xc0a\xc0"}, 977 {not(isValidRune), "\xc0a\xc0", 978 "a", 979 "a\xc0", 980 "\xc0a"}, 981 {isSpace, "", 982 "", 983 "", 984 ""}, 985 {isSpace, " ", 986 "", 987 "", 988 ""}, 989 } 990 991 func TestTrimFunc(t *testing.T) { 992 for _, tc := range trimFuncTests { 993 trimmers := []struct { 994 name string 995 trim func(s string, f func(r rune) bool) string 996 out string 997 }{ 998 {"TrimFunc", TrimFunc, tc.trimOut}, 999 {"TrimLeftFunc", TrimLeftFunc, tc.leftOut}, 1000 {"TrimRightFunc", TrimRightFunc, tc.rightOut}, 1001 } 1002 for _, trimmer := range trimmers { 1003 actual := trimmer.trim(tc.in, tc.f.f) 1004 if actual != trimmer.out { 1005 t.Errorf("%s(%q, %q) = %q; want %q", trimmer.name, tc.in, tc.f.name, actual, trimmer.out) 1006 } 1007 } 1008 } 1009 } 1010 1011 var indexFuncTests = []struct { 1012 in string 1013 f predicate 1014 first, last int 1015 }{ 1016 {"", isValidRune, -1, -1}, 1017 {"abc", isDigit, -1, -1}, 1018 {"0123", isDigit, 0, 3}, 1019 {"a1b", isDigit, 1, 1}, 1020 {space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes 1021 {"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, 1022 {"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, 1023 {"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, 1024 1025 // tests of invalid UTF-8 1026 {"\x801", isDigit, 1, 1}, 1027 {"\x80abc", isDigit, -1, -1}, 1028 {"\xc0a\xc0", isValidRune, 1, 1}, 1029 {"\xc0a\xc0", not(isValidRune), 0, 2}, 1030 {"\xc0☺\xc0", not(isValidRune), 0, 4}, 1031 {"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, 1032 {"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, 1033 {"a\xe0\x80cd", not(isValidRune), 1, 2}, 1034 {"\x80\x80\x80\x80", not(isValidRune), 0, 3}, 1035 } 1036 1037 func TestIndexFunc(t *testing.T) { 1038 for _, tc := range indexFuncTests { 1039 first := IndexFunc(tc.in, tc.f.f) 1040 if first != tc.first { 1041 t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) 1042 } 1043 last := LastIndexFunc(tc.in, tc.f.f) 1044 if last != tc.last { 1045 t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) 1046 } 1047 } 1048 } 1049 1050 func equal(m string, s1, s2 string, t *testing.T) bool { 1051 if s1 == s2 { 1052 return true 1053 } 1054 e1 := Split(s1, "") 1055 e2 := Split(s2, "") 1056 for i, c1 := range e1 { 1057 if i >= len(e2) { 1058 break 1059 } 1060 r1, _ := utf8.DecodeRuneInString(c1) 1061 r2, _ := utf8.DecodeRuneInString(e2[i]) 1062 if r1 != r2 { 1063 t.Errorf("%s diff at %d: U+%04X U+%04X", m, i, r1, r2) 1064 } 1065 } 1066 return false 1067 } 1068 1069 func TestCaseConsistency(t *testing.T) { 1070 // Make a string of all the runes. 1071 numRunes := int(unicode.MaxRune + 1) 1072 if testing.Short() { 1073 numRunes = 1000 1074 } 1075 a := make([]rune, numRunes) 1076 for i := range a { 1077 a[i] = rune(i) 1078 } 1079 s := string(a) 1080 // convert the cases. 1081 upper := ToUpper(s) 1082 lower := ToLower(s) 1083 1084 // Consistency checks 1085 if n := utf8.RuneCountInString(upper); n != numRunes { 1086 t.Error("rune count wrong in upper:", n) 1087 } 1088 if n := utf8.RuneCountInString(lower); n != numRunes { 1089 t.Error("rune count wrong in lower:", n) 1090 } 1091 if !equal("ToUpper(upper)", ToUpper(upper), upper, t) { 1092 t.Error("ToUpper(upper) consistency fail") 1093 } 1094 if !equal("ToLower(lower)", ToLower(lower), lower, t) { 1095 t.Error("ToLower(lower) consistency fail") 1096 } 1097 /* 1098 These fail because of non-one-to-oneness of the data, such as multiple 1099 upper case 'I' mapping to 'i'. We comment them out but keep them for 1100 interest. 1101 For instance: CAPITAL LETTER I WITH DOT ABOVE: 1102 unicode.ToUpper(unicode.ToLower('\u0130')) != '\u0130' 1103 1104 if !equal("ToUpper(lower)", ToUpper(lower), upper, t) { 1105 t.Error("ToUpper(lower) consistency fail"); 1106 } 1107 if !equal("ToLower(upper)", ToLower(upper), lower, t) { 1108 t.Error("ToLower(upper) consistency fail"); 1109 } 1110 */ 1111 } 1112 1113 var longString = "a" + string(make([]byte, 1<<16)) + "z" 1114 1115 var RepeatTests = []struct { 1116 in, out string 1117 count int 1118 }{ 1119 {"", "", 0}, 1120 {"", "", 1}, 1121 {"", "", 2}, 1122 {"-", "", 0}, 1123 {"-", "-", 1}, 1124 {"-", "----------", 10}, 1125 {"abc ", "abc abc abc ", 3}, 1126 // Tests for results over the chunkLimit 1127 {string(rune(0)), string(make([]byte, 1<<16)), 1 << 16}, 1128 {longString, longString + longString, 2}, 1129 } 1130 1131 func TestRepeat(t *testing.T) { 1132 for _, tt := range RepeatTests { 1133 a := Repeat(tt.in, tt.count) 1134 if !equal("Repeat(s)", a, tt.out, t) { 1135 t.Errorf("Repeat(%v, %d) = %v; want %v", tt.in, tt.count, a, tt.out) 1136 continue 1137 } 1138 } 1139 } 1140 1141 func repeat(s string, count int) (err error) { 1142 defer func() { 1143 if r := recover(); r != nil { 1144 switch v := r.(type) { 1145 case error: 1146 err = v 1147 default: 1148 err = fmt.Errorf("%s", v) 1149 } 1150 } 1151 }() 1152 1153 Repeat(s, count) 1154 1155 return 1156 } 1157 1158 // See Issue golang.org/issue/16237 1159 func TestRepeatCatchesOverflow(t *testing.T) { 1160 tests := [...]struct { 1161 s string 1162 count int 1163 errStr string 1164 }{ 1165 0: {"--", -2147483647, "negative"}, 1166 1: {"", int(^uint(0) >> 1), ""}, 1167 2: {"-", 10, ""}, 1168 3: {"gopher", 0, ""}, 1169 4: {"-", -1, "negative"}, 1170 5: {"--", -102, "negative"}, 1171 6: {string(make([]byte, 255)), int((^uint(0))/255 + 1), "overflow"}, 1172 } 1173 1174 for i, tt := range tests { 1175 err := repeat(tt.s, tt.count) 1176 if tt.errStr == "" { 1177 if err != nil { 1178 t.Errorf("#%d panicked %v", i, err) 1179 } 1180 continue 1181 } 1182 1183 if err == nil || !Contains(err.Error(), tt.errStr) { 1184 t.Errorf("#%d expected %q got %q", i, tt.errStr, err) 1185 } 1186 } 1187 } 1188 1189 func runesEqual(a, b []rune) bool { 1190 if len(a) != len(b) { 1191 return false 1192 } 1193 for i, r := range a { 1194 if r != b[i] { 1195 return false 1196 } 1197 } 1198 return true 1199 } 1200 1201 var RunesTests = []struct { 1202 in string 1203 out []rune 1204 lossy bool 1205 }{ 1206 {"", []rune{}, false}, 1207 {" ", []rune{32}, false}, 1208 {"ABC", []rune{65, 66, 67}, false}, 1209 {"abc", []rune{97, 98, 99}, false}, 1210 {"\u65e5\u672c\u8a9e", []rune{26085, 26412, 35486}, false}, 1211 {"ab\x80c", []rune{97, 98, 0xFFFD, 99}, true}, 1212 {"ab\xc0c", []rune{97, 98, 0xFFFD, 99}, true}, 1213 } 1214 1215 func TestRunes(t *testing.T) { 1216 for _, tt := range RunesTests { 1217 a := []rune(tt.in) 1218 if !runesEqual(a, tt.out) { 1219 t.Errorf("[]rune(%q) = %v; want %v", tt.in, a, tt.out) 1220 continue 1221 } 1222 if !tt.lossy { 1223 // can only test reassembly if we didn't lose information 1224 s := string(a) 1225 if s != tt.in { 1226 t.Errorf("string([]rune(%q)) = %x; want %x", tt.in, s, tt.in) 1227 } 1228 } 1229 } 1230 } 1231 1232 func TestReadByte(t *testing.T) { 1233 testStrings := []string{"", abcd, faces, commas} 1234 for _, s := range testStrings { 1235 reader := NewReader(s) 1236 if e := reader.UnreadByte(); e == nil { 1237 t.Errorf("Unreading %q at beginning: expected error", s) 1238 } 1239 var res bytes.Buffer 1240 for { 1241 b, e := reader.ReadByte() 1242 if e == io.EOF { 1243 break 1244 } 1245 if e != nil { 1246 t.Errorf("Reading %q: %s", s, e) 1247 break 1248 } 1249 res.WriteByte(b) 1250 // unread and read again 1251 e = reader.UnreadByte() 1252 if e != nil { 1253 t.Errorf("Unreading %q: %s", s, e) 1254 break 1255 } 1256 b1, e := reader.ReadByte() 1257 if e != nil { 1258 t.Errorf("Reading %q after unreading: %s", s, e) 1259 break 1260 } 1261 if b1 != b { 1262 t.Errorf("Reading %q after unreading: want byte %q, got %q", s, b, b1) 1263 break 1264 } 1265 } 1266 if res.String() != s { 1267 t.Errorf("Reader(%q).ReadByte() produced %q", s, res.String()) 1268 } 1269 } 1270 } 1271 1272 func TestReadRune(t *testing.T) { 1273 testStrings := []string{"", abcd, faces, commas} 1274 for _, s := range testStrings { 1275 reader := NewReader(s) 1276 if e := reader.UnreadRune(); e == nil { 1277 t.Errorf("Unreading %q at beginning: expected error", s) 1278 } 1279 res := "" 1280 for { 1281 r, z, e := reader.ReadRune() 1282 if e == io.EOF { 1283 break 1284 } 1285 if e != nil { 1286 t.Errorf("Reading %q: %s", s, e) 1287 break 1288 } 1289 res += string(r) 1290 // unread and read again 1291 e = reader.UnreadRune() 1292 if e != nil { 1293 t.Errorf("Unreading %q: %s", s, e) 1294 break 1295 } 1296 r1, z1, e := reader.ReadRune() 1297 if e != nil { 1298 t.Errorf("Reading %q after unreading: %s", s, e) 1299 break 1300 } 1301 if r1 != r { 1302 t.Errorf("Reading %q after unreading: want rune %q, got %q", s, r, r1) 1303 break 1304 } 1305 if z1 != z { 1306 t.Errorf("Reading %q after unreading: want size %d, got %d", s, z, z1) 1307 break 1308 } 1309 } 1310 if res != s { 1311 t.Errorf("Reader(%q).ReadRune() produced %q", s, res) 1312 } 1313 } 1314 } 1315 1316 var UnreadRuneErrorTests = []struct { 1317 name string 1318 f func(*Reader) 1319 }{ 1320 {"Read", func(r *Reader) { r.Read([]byte{0}) }}, 1321 {"ReadByte", func(r *Reader) { r.ReadByte() }}, 1322 {"UnreadRune", func(r *Reader) { r.UnreadRune() }}, 1323 {"Seek", func(r *Reader) { r.Seek(0, io.SeekCurrent) }}, 1324 {"WriteTo", func(r *Reader) { r.WriteTo(&bytes.Buffer{}) }}, 1325 } 1326 1327 func TestUnreadRuneError(t *testing.T) { 1328 for _, tt := range UnreadRuneErrorTests { 1329 reader := NewReader("0123456789") 1330 if _, _, err := reader.ReadRune(); err != nil { 1331 // should not happen 1332 t.Fatal(err) 1333 } 1334 tt.f(reader) 1335 err := reader.UnreadRune() 1336 if err == nil { 1337 t.Errorf("Unreading after %s: expected error", tt.name) 1338 } 1339 } 1340 } 1341 1342 var ReplaceTests = []struct { 1343 in string 1344 old, new string 1345 n int 1346 out string 1347 }{ 1348 {"hello", "l", "L", 0, "hello"}, 1349 {"hello", "l", "L", -1, "heLLo"}, 1350 {"hello", "x", "X", -1, "hello"}, 1351 {"", "x", "X", -1, ""}, 1352 {"radar", "r", "<r>", -1, "<r>ada<r>"}, 1353 {"", "", "<>", -1, "<>"}, 1354 {"banana", "a", "<>", -1, "b<>n<>n<>"}, 1355 {"banana", "a", "<>", 1, "b<>nana"}, 1356 {"banana", "a", "<>", 1000, "b<>n<>n<>"}, 1357 {"banana", "an", "<>", -1, "b<><>a"}, 1358 {"banana", "ana", "<>", -1, "b<>na"}, 1359 {"banana", "", "<>", -1, "<>b<>a<>n<>a<>n<>a<>"}, 1360 {"banana", "", "<>", 10, "<>b<>a<>n<>a<>n<>a<>"}, 1361 {"banana", "", "<>", 6, "<>b<>a<>n<>a<>n<>a"}, 1362 {"banana", "", "<>", 5, "<>b<>a<>n<>a<>na"}, 1363 {"banana", "", "<>", 1, "<>banana"}, 1364 {"banana", "a", "a", -1, "banana"}, 1365 {"banana", "a", "a", 1, "banana"}, 1366 {"☺☻☹", "", "<>", -1, "<>☺<>☻<>☹<>"}, 1367 } 1368 1369 func TestReplace(t *testing.T) { 1370 for _, tt := range ReplaceTests { 1371 if s := Replace(tt.in, tt.old, tt.new, tt.n); s != tt.out { 1372 t.Errorf("Replace(%q, %q, %q, %d) = %q, want %q", tt.in, tt.old, tt.new, tt.n, s, tt.out) 1373 } 1374 if tt.n == -1 { 1375 s := ReplaceAll(tt.in, tt.old, tt.new) 1376 if s != tt.out { 1377 t.Errorf("ReplaceAll(%q, %q, %q) = %q, want %q", tt.in, tt.old, tt.new, s, tt.out) 1378 } 1379 } 1380 } 1381 } 1382 1383 var TitleTests = []struct { 1384 in, out string 1385 }{ 1386 {"", ""}, 1387 {"a", "A"}, 1388 {" aaa aaa aaa ", " Aaa Aaa Aaa "}, 1389 {" Aaa Aaa Aaa ", " Aaa Aaa Aaa "}, 1390 {"123a456", "123a456"}, 1391 {"double-blind", "Double-Blind"}, 1392 {"ÿøû", "Ÿøû"}, 1393 {"with_underscore", "With_underscore"}, 1394 {"unicode \xe2\x80\xa8 line separator", "Unicode \xe2\x80\xa8 Line Separator"}, 1395 } 1396 1397 func TestTitle(t *testing.T) { 1398 for _, tt := range TitleTests { 1399 if s := Title(tt.in); s != tt.out { 1400 t.Errorf("Title(%q) = %q, want %q", tt.in, s, tt.out) 1401 } 1402 } 1403 } 1404 1405 var ContainsTests = []struct { 1406 str, substr string 1407 expected bool 1408 }{ 1409 {"abc", "bc", true}, 1410 {"abc", "bcd", false}, 1411 {"abc", "", true}, 1412 {"", "a", false}, 1413 1414 // cases to cover code in runtime/asm_amd64.s:indexShortStr 1415 // 2-byte needle 1416 {"xxxxxx", "01", false}, 1417 {"01xxxx", "01", true}, 1418 {"xx01xx", "01", true}, 1419 {"xxxx01", "01", true}, 1420 {"01xxxxx"[1:], "01", false}, 1421 {"xxxxx01"[:6], "01", false}, 1422 // 3-byte needle 1423 {"xxxxxxx", "012", false}, 1424 {"012xxxx", "012", true}, 1425 {"xx012xx", "012", true}, 1426 {"xxxx012", "012", true}, 1427 {"012xxxxx"[1:], "012", false}, 1428 {"xxxxx012"[:7], "012", false}, 1429 // 4-byte needle 1430 {"xxxxxxxx", "0123", false}, 1431 {"0123xxxx", "0123", true}, 1432 {"xx0123xx", "0123", true}, 1433 {"xxxx0123", "0123", true}, 1434 {"0123xxxxx"[1:], "0123", false}, 1435 {"xxxxx0123"[:8], "0123", false}, 1436 // 5-7-byte needle 1437 {"xxxxxxxxx", "01234", false}, 1438 {"01234xxxx", "01234", true}, 1439 {"xx01234xx", "01234", true}, 1440 {"xxxx01234", "01234", true}, 1441 {"01234xxxxx"[1:], "01234", false}, 1442 {"xxxxx01234"[:9], "01234", false}, 1443 // 8-byte needle 1444 {"xxxxxxxxxxxx", "01234567", false}, 1445 {"01234567xxxx", "01234567", true}, 1446 {"xx01234567xx", "01234567", true}, 1447 {"xxxx01234567", "01234567", true}, 1448 {"01234567xxxxx"[1:], "01234567", false}, 1449 {"xxxxx01234567"[:12], "01234567", false}, 1450 // 9-15-byte needle 1451 {"xxxxxxxxxxxxx", "012345678", false}, 1452 {"012345678xxxx", "012345678", true}, 1453 {"xx012345678xx", "012345678", true}, 1454 {"xxxx012345678", "012345678", true}, 1455 {"012345678xxxxx"[1:], "012345678", false}, 1456 {"xxxxx012345678"[:13], "012345678", false}, 1457 // 16-byte needle 1458 {"xxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEF", false}, 1459 {"0123456789ABCDEFxxxx", "0123456789ABCDEF", true}, 1460 {"xx0123456789ABCDEFxx", "0123456789ABCDEF", true}, 1461 {"xxxx0123456789ABCDEF", "0123456789ABCDEF", true}, 1462 {"0123456789ABCDEFxxxxx"[1:], "0123456789ABCDEF", false}, 1463 {"xxxxx0123456789ABCDEF"[:20], "0123456789ABCDEF", false}, 1464 // 17-31-byte needle 1465 {"xxxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEFG", false}, 1466 {"0123456789ABCDEFGxxxx", "0123456789ABCDEFG", true}, 1467 {"xx0123456789ABCDEFGxx", "0123456789ABCDEFG", true}, 1468 {"xxxx0123456789ABCDEFG", "0123456789ABCDEFG", true}, 1469 {"0123456789ABCDEFGxxxxx"[1:], "0123456789ABCDEFG", false}, 1470 {"xxxxx0123456789ABCDEFG"[:21], "0123456789ABCDEFG", false}, 1471 1472 // partial match cases 1473 {"xx01x", "012", false}, // 3 1474 {"xx0123x", "01234", false}, // 5-7 1475 {"xx01234567x", "012345678", false}, // 9-15 1476 {"xx0123456789ABCDEFx", "0123456789ABCDEFG", false}, // 17-31, issue 15679 1477 } 1478 1479 func TestContains(t *testing.T) { 1480 for _, ct := range ContainsTests { 1481 if Contains(ct.str, ct.substr) != ct.expected { 1482 t.Errorf("Contains(%s, %s) = %v, want %v", 1483 ct.str, ct.substr, !ct.expected, ct.expected) 1484 } 1485 } 1486 } 1487 1488 var ContainsAnyTests = []struct { 1489 str, substr string 1490 expected bool 1491 }{ 1492 {"", "", false}, 1493 {"", "a", false}, 1494 {"", "abc", false}, 1495 {"a", "", false}, 1496 {"a", "a", true}, 1497 {"aaa", "a", true}, 1498 {"abc", "xyz", false}, 1499 {"abc", "xcz", true}, 1500 {"a☺b☻c☹d", "uvw☻xyz", true}, 1501 {"aRegExp*", ".(|)*+?^$[]", true}, 1502 {dots + dots + dots, " ", false}, 1503 } 1504 1505 func TestContainsAny(t *testing.T) { 1506 for _, ct := range ContainsAnyTests { 1507 if ContainsAny(ct.str, ct.substr) != ct.expected { 1508 t.Errorf("ContainsAny(%s, %s) = %v, want %v", 1509 ct.str, ct.substr, !ct.expected, ct.expected) 1510 } 1511 } 1512 } 1513 1514 var ContainsRuneTests = []struct { 1515 str string 1516 r rune 1517 expected bool 1518 }{ 1519 {"", 'a', false}, 1520 {"a", 'a', true}, 1521 {"aaa", 'a', true}, 1522 {"abc", 'y', false}, 1523 {"abc", 'c', true}, 1524 {"a☺b☻c☹d", 'x', false}, 1525 {"a☺b☻c☹d", '☻', true}, 1526 {"aRegExp*", '*', true}, 1527 } 1528 1529 func TestContainsRune(t *testing.T) { 1530 for _, ct := range ContainsRuneTests { 1531 if ContainsRune(ct.str, ct.r) != ct.expected { 1532 t.Errorf("ContainsRune(%q, %q) = %v, want %v", 1533 ct.str, ct.r, !ct.expected, ct.expected) 1534 } 1535 } 1536 } 1537 1538 func TestContainsFunc(t *testing.T) { 1539 for _, ct := range ContainsRuneTests { 1540 if ContainsFunc(ct.str, func(r rune) bool { 1541 return ct.r == r 1542 }) != ct.expected { 1543 t.Errorf("ContainsFunc(%q, func(%q)) = %v, want %v", 1544 ct.str, ct.r, !ct.expected, ct.expected) 1545 } 1546 } 1547 } 1548 1549 var EqualFoldTests = []struct { 1550 s, t string 1551 out bool 1552 }{ 1553 {"abc", "abc", true}, 1554 {"ABcd", "ABcd", true}, 1555 {"123abc", "123ABC", true}, 1556 {"αβδ", "ΑΒΔ", true}, 1557 {"abc", "xyz", false}, 1558 {"abc", "XYZ", false}, 1559 {"abcdefghijk", "abcdefghijX", false}, 1560 {"abcdefghijk", "abcdefghij\u212A", true}, 1561 {"abcdefghijK", "abcdefghij\u212A", true}, 1562 {"abcdefghijkz", "abcdefghij\u212Ay", false}, 1563 {"abcdefghijKz", "abcdefghij\u212Ay", false}, 1564 {"1", "2", false}, 1565 {"utf-8", "US-ASCII", false}, 1566 } 1567 1568 func TestEqualFold(t *testing.T) { 1569 for _, tt := range EqualFoldTests { 1570 if out := EqualFold(tt.s, tt.t); out != tt.out { 1571 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.s, tt.t, out, tt.out) 1572 } 1573 if out := EqualFold(tt.t, tt.s); out != tt.out { 1574 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.t, tt.s, out, tt.out) 1575 } 1576 } 1577 } 1578 1579 func BenchmarkEqualFold(b *testing.B) { 1580 b.Run("Tests", func(b *testing.B) { 1581 for i := 0; i < b.N; i++ { 1582 for _, tt := range EqualFoldTests { 1583 if out := EqualFold(tt.s, tt.t); out != tt.out { 1584 b.Fatal("wrong result") 1585 } 1586 } 1587 } 1588 }) 1589 1590 const s1 = "abcdefghijKz" 1591 const s2 = "abcDefGhijKz" 1592 1593 b.Run("ASCII", func(b *testing.B) { 1594 for i := 0; i < b.N; i++ { 1595 EqualFold(s1, s2) 1596 } 1597 }) 1598 1599 b.Run("UnicodePrefix", func(b *testing.B) { 1600 for i := 0; i < b.N; i++ { 1601 EqualFold("αβδ"+s1, "ΑΒΔ"+s2) 1602 } 1603 }) 1604 1605 b.Run("UnicodeSuffix", func(b *testing.B) { 1606 for i := 0; i < b.N; i++ { 1607 EqualFold(s1+"αβδ", s2+"ΑΒΔ") 1608 } 1609 }) 1610 } 1611 1612 var CountTests = []struct { 1613 s, sep string 1614 num int 1615 }{ 1616 {"", "", 1}, 1617 {"", "notempty", 0}, 1618 {"notempty", "", 9}, 1619 {"smaller", "not smaller", 0}, 1620 {"12345678987654321", "6", 2}, 1621 {"611161116", "6", 3}, 1622 {"notequal", "NotEqual", 0}, 1623 {"equal", "equal", 1}, 1624 {"abc1231231123q", "123", 3}, 1625 {"11111", "11", 2}, 1626 } 1627 1628 func TestCount(t *testing.T) { 1629 for _, tt := range CountTests { 1630 if num := Count(tt.s, tt.sep); num != tt.num { 1631 t.Errorf("Count(%q, %q) = %d, want %d", tt.s, tt.sep, num, tt.num) 1632 } 1633 } 1634 } 1635 1636 var cutTests = []struct { 1637 s, sep string 1638 before, after string 1639 found bool 1640 }{ 1641 {"abc", "b", "a", "c", true}, 1642 {"abc", "a", "", "bc", true}, 1643 {"abc", "c", "ab", "", true}, 1644 {"abc", "abc", "", "", true}, 1645 {"abc", "", "", "abc", true}, 1646 {"abc", "d", "abc", "", false}, 1647 {"", "d", "", "", false}, 1648 {"", "", "", "", true}, 1649 } 1650 1651 func TestCut(t *testing.T) { 1652 for _, tt := range cutTests { 1653 if before, after, found := Cut(tt.s, tt.sep); before != tt.before || after != tt.after || found != tt.found { 1654 t.Errorf("Cut(%q, %q) = %q, %q, %v, want %q, %q, %v", tt.s, tt.sep, before, after, found, tt.before, tt.after, tt.found) 1655 } 1656 } 1657 } 1658 1659 var cutPrefixTests = []struct { 1660 s, sep string 1661 after string 1662 found bool 1663 }{ 1664 {"abc", "a", "bc", true}, 1665 {"abc", "abc", "", true}, 1666 {"abc", "", "abc", true}, 1667 {"abc", "d", "abc", false}, 1668 {"", "d", "", false}, 1669 {"", "", "", true}, 1670 } 1671 1672 func TestCutPrefix(t *testing.T) { 1673 for _, tt := range cutPrefixTests { 1674 if after, found := CutPrefix(tt.s, tt.sep); after != tt.after || found != tt.found { 1675 t.Errorf("CutPrefix(%q, %q) = %q, %v, want %q, %v", tt.s, tt.sep, after, found, tt.after, tt.found) 1676 } 1677 } 1678 } 1679 1680 var cutSuffixTests = []struct { 1681 s, sep string 1682 before string 1683 found bool 1684 }{ 1685 {"abc", "bc", "a", true}, 1686 {"abc", "abc", "", true}, 1687 {"abc", "", "abc", true}, 1688 {"abc", "d", "abc", false}, 1689 {"", "d", "", false}, 1690 {"", "", "", true}, 1691 } 1692 1693 func TestCutSuffix(t *testing.T) { 1694 for _, tt := range cutSuffixTests { 1695 if before, found := CutSuffix(tt.s, tt.sep); before != tt.before || found != tt.found { 1696 t.Errorf("CutSuffix(%q, %q) = %q, %v, want %q, %v", tt.s, tt.sep, before, found, tt.before, tt.found) 1697 } 1698 } 1699 } 1700 1701 func makeBenchInputHard() string { 1702 tokens := [...]string{ 1703 "<a>", "<p>", "<b>", "<strong>", 1704 "</a>", "</p>", "</b>", "</strong>", 1705 "hello", "world", 1706 } 1707 x := make([]byte, 0, 1<<20) 1708 for { 1709 i := rand.Intn(len(tokens)) 1710 if len(x)+len(tokens[i]) >= 1<<20 { 1711 break 1712 } 1713 x = append(x, tokens[i]...) 1714 } 1715 return string(x) 1716 } 1717 1718 var benchInputHard = makeBenchInputHard() 1719 1720 func benchmarkIndexHard(b *testing.B, sep string) { 1721 for i := 0; i < b.N; i++ { 1722 Index(benchInputHard, sep) 1723 } 1724 } 1725 1726 func benchmarkLastIndexHard(b *testing.B, sep string) { 1727 for i := 0; i < b.N; i++ { 1728 LastIndex(benchInputHard, sep) 1729 } 1730 } 1731 1732 func benchmarkCountHard(b *testing.B, sep string) { 1733 for i := 0; i < b.N; i++ { 1734 Count(benchInputHard, sep) 1735 } 1736 } 1737 1738 func BenchmarkIndexHard1(b *testing.B) { benchmarkIndexHard(b, "<>") } 1739 func BenchmarkIndexHard2(b *testing.B) { benchmarkIndexHard(b, "</pre>") } 1740 func BenchmarkIndexHard3(b *testing.B) { benchmarkIndexHard(b, "<b>hello world</b>") } 1741 func BenchmarkIndexHard4(b *testing.B) { 1742 benchmarkIndexHard(b, "<pre><b>hello</b><strong>world</strong></pre>") 1743 } 1744 1745 func BenchmarkLastIndexHard1(b *testing.B) { benchmarkLastIndexHard(b, "<>") } 1746 func BenchmarkLastIndexHard2(b *testing.B) { benchmarkLastIndexHard(b, "</pre>") } 1747 func BenchmarkLastIndexHard3(b *testing.B) { benchmarkLastIndexHard(b, "<b>hello world</b>") } 1748 1749 func BenchmarkCountHard1(b *testing.B) { benchmarkCountHard(b, "<>") } 1750 func BenchmarkCountHard2(b *testing.B) { benchmarkCountHard(b, "</pre>") } 1751 func BenchmarkCountHard3(b *testing.B) { benchmarkCountHard(b, "<b>hello world</b>") } 1752 1753 var benchInputTorture = Repeat("ABC", 1<<10) + "123" + Repeat("ABC", 1<<10) 1754 var benchNeedleTorture = Repeat("ABC", 1<<10+1) 1755 1756 func BenchmarkIndexTorture(b *testing.B) { 1757 for i := 0; i < b.N; i++ { 1758 Index(benchInputTorture, benchNeedleTorture) 1759 } 1760 } 1761 1762 func BenchmarkCountTorture(b *testing.B) { 1763 for i := 0; i < b.N; i++ { 1764 Count(benchInputTorture, benchNeedleTorture) 1765 } 1766 } 1767 1768 func BenchmarkCountTortureOverlapping(b *testing.B) { 1769 A := Repeat("ABC", 1<<20) 1770 B := Repeat("ABC", 1<<10) 1771 for i := 0; i < b.N; i++ { 1772 Count(A, B) 1773 } 1774 } 1775 1776 func BenchmarkCountByte(b *testing.B) { 1777 indexSizes := []int{10, 32, 4 << 10, 4 << 20, 64 << 20} 1778 benchStr := Repeat(benchmarkString, 1779 (indexSizes[len(indexSizes)-1]+len(benchmarkString)-1)/len(benchmarkString)) 1780 benchFunc := func(b *testing.B, benchStr string) { 1781 b.SetBytes(int64(len(benchStr))) 1782 for i := 0; i < b.N; i++ { 1783 Count(benchStr, "=") 1784 } 1785 } 1786 for _, size := range indexSizes { 1787 b.Run(fmt.Sprintf("%d", size), func(b *testing.B) { 1788 benchFunc(b, benchStr[:size]) 1789 }) 1790 } 1791 1792 } 1793 1794 var makeFieldsInput = func() string { 1795 x := make([]byte, 1<<20) 1796 // Input is ~10% space, ~10% 2-byte UTF-8, rest ASCII non-space. 1797 for i := range x { 1798 switch rand.Intn(10) { 1799 case 0: 1800 x[i] = ' ' 1801 case 1: 1802 if i > 0 && x[i-1] == 'x' { 1803 copy(x[i-1:], "χ") 1804 break 1805 } 1806 fallthrough 1807 default: 1808 x[i] = 'x' 1809 } 1810 } 1811 return string(x) 1812 } 1813 1814 var makeFieldsInputASCII = func() string { 1815 x := make([]byte, 1<<20) 1816 // Input is ~10% space, rest ASCII non-space. 1817 for i := range x { 1818 if rand.Intn(10) == 0 { 1819 x[i] = ' ' 1820 } else { 1821 x[i] = 'x' 1822 } 1823 } 1824 return string(x) 1825 } 1826 1827 var stringdata = []struct{ name, data string }{ 1828 {"ASCII", makeFieldsInputASCII()}, 1829 {"Mixed", makeFieldsInput()}, 1830 } 1831 1832 func BenchmarkFields(b *testing.B) { 1833 for _, sd := range stringdata { 1834 b.Run(sd.name, func(b *testing.B) { 1835 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1836 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1837 b.ReportAllocs() 1838 b.SetBytes(int64(j)) 1839 data := sd.data[:j] 1840 for i := 0; i < b.N; i++ { 1841 Fields(data) 1842 } 1843 }) 1844 } 1845 }) 1846 } 1847 } 1848 1849 func BenchmarkFieldsFunc(b *testing.B) { 1850 for _, sd := range stringdata { 1851 b.Run(sd.name, func(b *testing.B) { 1852 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1853 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1854 b.ReportAllocs() 1855 b.SetBytes(int64(j)) 1856 data := sd.data[:j] 1857 for i := 0; i < b.N; i++ { 1858 FieldsFunc(data, unicode.IsSpace) 1859 } 1860 }) 1861 } 1862 }) 1863 } 1864 } 1865 1866 func BenchmarkSplitEmptySeparator(b *testing.B) { 1867 for i := 0; i < b.N; i++ { 1868 Split(benchInputHard, "") 1869 } 1870 } 1871 1872 func BenchmarkSplitSingleByteSeparator(b *testing.B) { 1873 for i := 0; i < b.N; i++ { 1874 Split(benchInputHard, "/") 1875 } 1876 } 1877 1878 func BenchmarkSplitMultiByteSeparator(b *testing.B) { 1879 for i := 0; i < b.N; i++ { 1880 Split(benchInputHard, "hello") 1881 } 1882 } 1883 1884 func BenchmarkSplitNSingleByteSeparator(b *testing.B) { 1885 for i := 0; i < b.N; i++ { 1886 SplitN(benchInputHard, "/", 10) 1887 } 1888 } 1889 1890 func BenchmarkSplitNMultiByteSeparator(b *testing.B) { 1891 for i := 0; i < b.N; i++ { 1892 SplitN(benchInputHard, "hello", 10) 1893 } 1894 } 1895 1896 func BenchmarkRepeat(b *testing.B) { 1897 s := "0123456789" 1898 for _, n := range []int{5, 10} { 1899 for _, c := range []int{0, 1, 2, 6} { 1900 b.Run(fmt.Sprintf("%dx%d", n, c), func(b *testing.B) { 1901 for i := 0; i < b.N; i++ { 1902 Repeat(s[:n], c) 1903 } 1904 }) 1905 } 1906 } 1907 } 1908 1909 func BenchmarkRepeatLarge(b *testing.B) { 1910 s := Repeat("@", 8*1024) 1911 for j := 8; j <= 30; j++ { 1912 for _, k := range []int{1, 16, 4097} { 1913 s := s[:k] 1914 n := (1 << j) / k 1915 if n == 0 { 1916 continue 1917 } 1918 b.Run(fmt.Sprintf("%d/%d", 1<<j, k), func(b *testing.B) { 1919 for i := 0; i < b.N; i++ { 1920 Repeat(s, n) 1921 } 1922 b.SetBytes(int64(n * len(s))) 1923 }) 1924 } 1925 } 1926 } 1927 1928 func BenchmarkIndexAnyASCII(b *testing.B) { 1929 x := Repeat("#", 2048) // Never matches set 1930 cs := "0123456789abcdefghijklmnopqrstuvwxyz0123456789abcdefghijklmnopqrstuvwxyz" 1931 for k := 1; k <= 2048; k <<= 4 { 1932 for j := 1; j <= 64; j <<= 1 { 1933 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1934 for i := 0; i < b.N; i++ { 1935 IndexAny(x[:k], cs[:j]) 1936 } 1937 }) 1938 } 1939 } 1940 } 1941 1942 func BenchmarkIndexAnyUTF8(b *testing.B) { 1943 x := Repeat("#", 2048) // Never matches set 1944 cs := "你好世界, hello world. 你好世界, hello world. 你好世界, hello world." 1945 for k := 1; k <= 2048; k <<= 4 { 1946 for j := 1; j <= 64; j <<= 1 { 1947 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1948 for i := 0; i < b.N; i++ { 1949 IndexAny(x[:k], cs[:j]) 1950 } 1951 }) 1952 } 1953 } 1954 } 1955 1956 func BenchmarkLastIndexAnyASCII(b *testing.B) { 1957 x := Repeat("#", 2048) // Never matches set 1958 cs := "0123456789abcdefghijklmnopqrstuvwxyz0123456789abcdefghijklmnopqrstuvwxyz" 1959 for k := 1; k <= 2048; k <<= 4 { 1960 for j := 1; j <= 64; j <<= 1 { 1961 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1962 for i := 0; i < b.N; i++ { 1963 LastIndexAny(x[:k], cs[:j]) 1964 } 1965 }) 1966 } 1967 } 1968 } 1969 1970 func BenchmarkLastIndexAnyUTF8(b *testing.B) { 1971 x := Repeat("#", 2048) // Never matches set 1972 cs := "你好世界, hello world. 你好世界, hello world. 你好世界, hello world." 1973 for k := 1; k <= 2048; k <<= 4 { 1974 for j := 1; j <= 64; j <<= 1 { 1975 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1976 for i := 0; i < b.N; i++ { 1977 LastIndexAny(x[:k], cs[:j]) 1978 } 1979 }) 1980 } 1981 } 1982 } 1983 1984 func BenchmarkTrimASCII(b *testing.B) { 1985 cs := "0123456789abcdef" 1986 for k := 1; k <= 4096; k <<= 4 { 1987 for j := 1; j <= 16; j <<= 1 { 1988 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1989 x := Repeat(cs[:j], k) // Always matches set 1990 for i := 0; i < b.N; i++ { 1991 Trim(x[:k], cs[:j]) 1992 } 1993 }) 1994 } 1995 } 1996 } 1997 1998 func BenchmarkTrimByte(b *testing.B) { 1999 x := " the quick brown fox " 2000 for i := 0; i < b.N; i++ { 2001 Trim(x, " ") 2002 } 2003 } 2004 2005 func BenchmarkIndexPeriodic(b *testing.B) { 2006 key := "aa" 2007 for _, skip := range [...]int{2, 4, 8, 16, 32, 64} { 2008 b.Run(fmt.Sprintf("IndexPeriodic%d", skip), func(b *testing.B) { 2009 s := Repeat("a"+Repeat(" ", skip-1), 1<<16/skip) 2010 for i := 0; i < b.N; i++ { 2011 Index(s, key) 2012 } 2013 }) 2014 } 2015 } 2016 2017 func BenchmarkJoin(b *testing.B) { 2018 vals := []string{"red", "yellow", "pink", "green", "purple", "orange", "blue"} 2019 for l := 0; l <= len(vals); l++ { 2020 b.Run(strconv.Itoa(l), func(b *testing.B) { 2021 b.ReportAllocs() 2022 vals := vals[:l] 2023 for i := 0; i < b.N; i++ { 2024 Join(vals, " and ") 2025 } 2026 }) 2027 } 2028 } 2029 2030 func BenchmarkTrimSpace(b *testing.B) { 2031 tests := []struct{ name, input string }{ 2032 {"NoTrim", "typical"}, 2033 {"ASCII", " foo bar "}, 2034 {"SomeNonASCII", " \u2000\t\r\n x\t\t\r\r\ny\n \u3000 "}, 2035 {"JustNonASCII", "\u2000\u2000\u2000☺☺☺☺\u3000\u3000\u3000"}, 2036 } 2037 for _, test := range tests { 2038 b.Run(test.name, func(b *testing.B) { 2039 for i := 0; i < b.N; i++ { 2040 TrimSpace(test.input) 2041 } 2042 }) 2043 } 2044 } 2045 2046 var stringSink string 2047 2048 func BenchmarkReplaceAll(b *testing.B) { 2049 b.ReportAllocs() 2050 for i := 0; i < b.N; i++ { 2051 stringSink = ReplaceAll("banana", "a", "<>") 2052 } 2053 }