github.com/mtsmfm/go/src@v0.0.0-20221020090648-44bdcb9f8fde/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 // cases with one byte strings - test special case in Index() 58 {"", "a", -1}, 59 {"x", "a", -1}, 60 {"x", "x", 0}, 61 {"abc", "a", 0}, 62 {"abc", "b", 1}, 63 {"abc", "c", 2}, 64 {"abc", "x", -1}, 65 // test special cases in Index() for short strings 66 {"", "ab", -1}, 67 {"bc", "ab", -1}, 68 {"ab", "ab", 0}, 69 {"xab", "ab", 1}, 70 {"xab"[:2], "ab", -1}, 71 {"", "abc", -1}, 72 {"xbc", "abc", -1}, 73 {"abc", "abc", 0}, 74 {"xabc", "abc", 1}, 75 {"xabc"[:3], "abc", -1}, 76 {"xabxc", "abc", -1}, 77 {"", "abcd", -1}, 78 {"xbcd", "abcd", -1}, 79 {"abcd", "abcd", 0}, 80 {"xabcd", "abcd", 1}, 81 {"xyabcd"[:5], "abcd", -1}, 82 {"xbcqq", "abcqq", -1}, 83 {"abcqq", "abcqq", 0}, 84 {"xabcqq", "abcqq", 1}, 85 {"xyabcqq"[:6], "abcqq", -1}, 86 {"xabxcqq", "abcqq", -1}, 87 {"xabcqxq", "abcqq", -1}, 88 {"", "01234567", -1}, 89 {"32145678", "01234567", -1}, 90 {"01234567", "01234567", 0}, 91 {"x01234567", "01234567", 1}, 92 {"x0123456x01234567", "01234567", 9}, 93 {"xx01234567"[:9], "01234567", -1}, 94 {"", "0123456789", -1}, 95 {"3214567844", "0123456789", -1}, 96 {"0123456789", "0123456789", 0}, 97 {"x0123456789", "0123456789", 1}, 98 {"x012345678x0123456789", "0123456789", 11}, 99 {"xyz0123456789"[:12], "0123456789", -1}, 100 {"x01234567x89", "0123456789", -1}, 101 {"", "0123456789012345", -1}, 102 {"3214567889012345", "0123456789012345", -1}, 103 {"0123456789012345", "0123456789012345", 0}, 104 {"x0123456789012345", "0123456789012345", 1}, 105 {"x012345678901234x0123456789012345", "0123456789012345", 17}, 106 {"", "01234567890123456789", -1}, 107 {"32145678890123456789", "01234567890123456789", -1}, 108 {"01234567890123456789", "01234567890123456789", 0}, 109 {"x01234567890123456789", "01234567890123456789", 1}, 110 {"x0123456789012345678x01234567890123456789", "01234567890123456789", 21}, 111 {"xyz01234567890123456789"[:22], "01234567890123456789", -1}, 112 {"", "0123456789012345678901234567890", -1}, 113 {"321456788901234567890123456789012345678911", "0123456789012345678901234567890", -1}, 114 {"0123456789012345678901234567890", "0123456789012345678901234567890", 0}, 115 {"x0123456789012345678901234567890", "0123456789012345678901234567890", 1}, 116 {"x012345678901234567890123456789x0123456789012345678901234567890", "0123456789012345678901234567890", 32}, 117 {"xyz0123456789012345678901234567890"[:33], "0123456789012345678901234567890", -1}, 118 {"", "01234567890123456789012345678901", -1}, 119 {"32145678890123456789012345678901234567890211", "01234567890123456789012345678901", -1}, 120 {"01234567890123456789012345678901", "01234567890123456789012345678901", 0}, 121 {"x01234567890123456789012345678901", "01234567890123456789012345678901", 1}, 122 {"x0123456789012345678901234567890x01234567890123456789012345678901", "01234567890123456789012345678901", 33}, 123 {"xyz01234567890123456789012345678901"[:34], "01234567890123456789012345678901", -1}, 124 {"xxxxxx012345678901234567890123456789012345678901234567890123456789012", "012345678901234567890123456789012345678901234567890123456789012", 6}, 125 {"", "0123456789012345678901234567890123456789", -1}, 126 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456789", 2}, 127 {"xx012345678901234567890123456789012345678901234567890123456789012"[:41], "0123456789012345678901234567890123456789", -1}, 128 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456xxx", -1}, 129 {"xx0123456789012345678901234567890123456789012345678901234567890120123456789012345678901234567890123456xxx", "0123456789012345678901234567890123456xxx", 65}, 130 // test fallback to Rabin-Karp. 131 {"oxoxoxoxoxoxoxoxoxoxoxoy", "oy", 22}, 132 {"oxoxoxoxoxoxoxoxoxoxoxox", "oy", -1}, 133 } 134 135 var lastIndexTests = []IndexTest{ 136 {"", "", 0}, 137 {"", "a", -1}, 138 {"", "foo", -1}, 139 {"fo", "foo", -1}, 140 {"foo", "foo", 0}, 141 {"foo", "f", 0}, 142 {"oofofoofooo", "f", 7}, 143 {"oofofoofooo", "foo", 7}, 144 {"barfoobarfoo", "foo", 9}, 145 {"foo", "", 3}, 146 {"foo", "o", 2}, 147 {"abcABCabc", "A", 3}, 148 {"abcABCabc", "a", 6}, 149 } 150 151 var indexAnyTests = []IndexTest{ 152 {"", "", -1}, 153 {"", "a", -1}, 154 {"", "abc", -1}, 155 {"a", "", -1}, 156 {"a", "a", 0}, 157 {"\x80", "\xffb", 0}, 158 {"aaa", "a", 0}, 159 {"abc", "xyz", -1}, 160 {"abc", "xcz", 2}, 161 {"ab☺c", "x☺yz", 2}, 162 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 163 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 164 {"aRegExp*", ".(|)*+?^$[]", 7}, 165 {dots + dots + dots, " ", -1}, 166 {"012abcba210", "\xffb", 4}, 167 {"012\x80bcb\x80210", "\xffb", 3}, 168 {"0123456\xcf\x80abc", "\xcfb\x80", 10}, 169 } 170 171 var lastIndexAnyTests = []IndexTest{ 172 {"", "", -1}, 173 {"", "a", -1}, 174 {"", "abc", -1}, 175 {"a", "", -1}, 176 {"a", "a", 0}, 177 {"\x80", "\xffb", 0}, 178 {"aaa", "a", 2}, 179 {"abc", "xyz", -1}, 180 {"abc", "ab", 1}, 181 {"ab☺c", "x☺yz", 2}, 182 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 183 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 184 {"a.RegExp*", ".(|)*+?^$[]", 8}, 185 {dots + dots + dots, " ", -1}, 186 {"012abcba210", "\xffb", 6}, 187 {"012\x80bcb\x80210", "\xffb", 7}, 188 {"0123456\xcf\x80abc", "\xcfb\x80", 10}, 189 } 190 191 // Execute f on each test case. funcName should be the name of f; it's used 192 // in failure reports. 193 func runIndexTests(t *testing.T, f func(s, sep string) int, funcName string, testCases []IndexTest) { 194 for _, test := range testCases { 195 actual := f(test.s, test.sep) 196 if actual != test.out { 197 t.Errorf("%s(%q,%q) = %v; want %v", funcName, test.s, test.sep, actual, test.out) 198 } 199 } 200 } 201 202 func TestIndex(t *testing.T) { runIndexTests(t, Index, "Index", indexTests) } 203 func TestLastIndex(t *testing.T) { runIndexTests(t, LastIndex, "LastIndex", lastIndexTests) } 204 func TestIndexAny(t *testing.T) { runIndexTests(t, IndexAny, "IndexAny", indexAnyTests) } 205 func TestLastIndexAny(t *testing.T) { 206 runIndexTests(t, LastIndexAny, "LastIndexAny", lastIndexAnyTests) 207 } 208 209 func TestIndexByte(t *testing.T) { 210 for _, tt := range indexTests { 211 if len(tt.sep) != 1 { 212 continue 213 } 214 pos := IndexByte(tt.s, tt.sep[0]) 215 if pos != tt.out { 216 t.Errorf(`IndexByte(%q, %q) = %v; want %v`, tt.s, tt.sep[0], pos, tt.out) 217 } 218 } 219 } 220 221 func TestLastIndexByte(t *testing.T) { 222 testCases := []IndexTest{ 223 {"", "q", -1}, 224 {"abcdef", "q", -1}, 225 {"abcdefabcdef", "a", len("abcdef")}, // something in the middle 226 {"abcdefabcdef", "f", len("abcdefabcde")}, // last byte 227 {"zabcdefabcdef", "z", 0}, // first byte 228 {"a☺b☻c☹d", "b", len("a☺")}, // non-ascii 229 } 230 for _, test := range testCases { 231 actual := LastIndexByte(test.s, test.sep[0]) 232 if actual != test.out { 233 t.Errorf("LastIndexByte(%q,%c) = %v; want %v", test.s, test.sep[0], actual, test.out) 234 } 235 } 236 } 237 238 func simpleIndex(s, sep string) int { 239 n := len(sep) 240 for i := n; i <= len(s); i++ { 241 if s[i-n:i] == sep { 242 return i - n 243 } 244 } 245 return -1 246 } 247 248 func TestIndexRandom(t *testing.T) { 249 const chars = "abcdefghijklmnopqrstuvwxyz0123456789" 250 for times := 0; times < 10; times++ { 251 for strLen := 5 + rand.Intn(5); strLen < 140; strLen += 10 { // Arbitrary 252 s1 := make([]byte, strLen) 253 for i := range s1 { 254 s1[i] = chars[rand.Intn(len(chars))] 255 } 256 s := string(s1) 257 for i := 0; i < 50; i++ { 258 begin := rand.Intn(len(s) + 1) 259 end := begin + rand.Intn(len(s)+1-begin) 260 sep := s[begin:end] 261 if i%4 == 0 { 262 pos := rand.Intn(len(sep) + 1) 263 sep = sep[:pos] + "A" + sep[pos:] 264 } 265 want := simpleIndex(s, sep) 266 res := Index(s, sep) 267 if res != want { 268 t.Errorf("Index(%s,%s) = %d; want %d", s, sep, res, want) 269 } 270 } 271 } 272 } 273 } 274 275 func TestIndexRune(t *testing.T) { 276 tests := []struct { 277 in string 278 rune rune 279 want int 280 }{ 281 {"", 'a', -1}, 282 {"", '☺', -1}, 283 {"foo", '☹', -1}, 284 {"foo", 'o', 1}, 285 {"foo☺bar", '☺', 3}, 286 {"foo☺☻☹bar", '☹', 9}, 287 {"a A x", 'A', 2}, 288 {"some_text=some_value", '=', 9}, 289 {"☺a", 'a', 3}, 290 {"a☻☺b", '☺', 4}, 291 292 // RuneError should match any invalid UTF-8 byte sequence. 293 {"�", '�', 0}, 294 {"\xff", '�', 0}, 295 {"☻x�", '�', len("☻x")}, 296 {"☻x\xe2\x98", '�', len("☻x")}, 297 {"☻x\xe2\x98�", '�', len("☻x")}, 298 {"☻x\xe2\x98x", '�', len("☻x")}, 299 300 // Invalid rune values should never match. 301 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", -1, -1}, 302 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", 0xD800, -1}, // Surrogate pair 303 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", utf8.MaxRune + 1, -1}, 304 } 305 for _, tt := range tests { 306 if got := IndexRune(tt.in, tt.rune); got != tt.want { 307 t.Errorf("IndexRune(%q, %d) = %v; want %v", tt.in, tt.rune, got, tt.want) 308 } 309 } 310 311 haystack := "test世界" 312 allocs := testing.AllocsPerRun(1000, func() { 313 if i := IndexRune(haystack, 's'); i != 2 { 314 t.Fatalf("'s' at %d; want 2", i) 315 } 316 if i := IndexRune(haystack, '世'); i != 4 { 317 t.Fatalf("'世' at %d; want 4", i) 318 } 319 }) 320 if allocs != 0 && testing.CoverMode() == "" { 321 t.Errorf("expected no allocations, got %f", allocs) 322 } 323 } 324 325 const benchmarkString = "some_text=some☺value" 326 327 func BenchmarkIndexRune(b *testing.B) { 328 if got := IndexRune(benchmarkString, '☺'); got != 14 { 329 b.Fatalf("wrong index: expected 14, got=%d", got) 330 } 331 for i := 0; i < b.N; i++ { 332 IndexRune(benchmarkString, '☺') 333 } 334 } 335 336 var benchmarkLongString = Repeat(" ", 100) + benchmarkString 337 338 func BenchmarkIndexRuneLongString(b *testing.B) { 339 if got := IndexRune(benchmarkLongString, '☺'); got != 114 { 340 b.Fatalf("wrong index: expected 114, got=%d", got) 341 } 342 for i := 0; i < b.N; i++ { 343 IndexRune(benchmarkLongString, '☺') 344 } 345 } 346 347 func BenchmarkIndexRuneFastPath(b *testing.B) { 348 if got := IndexRune(benchmarkString, 'v'); got != 17 { 349 b.Fatalf("wrong index: expected 17, got=%d", got) 350 } 351 for i := 0; i < b.N; i++ { 352 IndexRune(benchmarkString, 'v') 353 } 354 } 355 356 func BenchmarkIndex(b *testing.B) { 357 if got := Index(benchmarkString, "v"); got != 17 { 358 b.Fatalf("wrong index: expected 17, got=%d", got) 359 } 360 for i := 0; i < b.N; i++ { 361 Index(benchmarkString, "v") 362 } 363 } 364 365 func BenchmarkLastIndex(b *testing.B) { 366 if got := Index(benchmarkString, "v"); got != 17 { 367 b.Fatalf("wrong index: expected 17, got=%d", got) 368 } 369 for i := 0; i < b.N; i++ { 370 LastIndex(benchmarkString, "v") 371 } 372 } 373 374 func BenchmarkIndexByte(b *testing.B) { 375 if got := IndexByte(benchmarkString, 'v'); got != 17 { 376 b.Fatalf("wrong index: expected 17, got=%d", got) 377 } 378 for i := 0; i < b.N; i++ { 379 IndexByte(benchmarkString, 'v') 380 } 381 } 382 383 type SplitTest struct { 384 s string 385 sep string 386 n int 387 a []string 388 } 389 390 var splittests = []SplitTest{ 391 {"", "", -1, []string{}}, 392 {abcd, "", 2, []string{"a", "bcd"}}, 393 {abcd, "", 4, []string{"a", "b", "c", "d"}}, 394 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 395 {faces, "", -1, []string{"☺", "☻", "☹"}}, 396 {faces, "", 3, []string{"☺", "☻", "☹"}}, 397 {faces, "", 17, []string{"☺", "☻", "☹"}}, 398 {"☺�☹", "", -1, []string{"☺", "�", "☹"}}, 399 {abcd, "a", 0, nil}, 400 {abcd, "a", -1, []string{"", "bcd"}}, 401 {abcd, "z", -1, []string{"abcd"}}, 402 {commas, ",", -1, []string{"1", "2", "3", "4"}}, 403 {dots, "...", -1, []string{"1", ".2", ".3", ".4"}}, 404 {faces, "☹", -1, []string{"☺☻", ""}}, 405 {faces, "~", -1, []string{faces}}, 406 {"1 2 3 4", " ", 3, []string{"1", "2", "3 4"}}, 407 {"1 2", " ", 3, []string{"1", "2"}}, 408 {"", "T", math.MaxInt / 4, []string{""}}, 409 {"\xff-\xff", "", -1, []string{"\xff", "-", "\xff"}}, 410 {"\xff-\xff", "-", -1, []string{"\xff", "\xff"}}, 411 } 412 413 func TestSplit(t *testing.T) { 414 for _, tt := range splittests { 415 a := SplitN(tt.s, tt.sep, tt.n) 416 if !eq(a, tt.a) { 417 t.Errorf("Split(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, a, tt.a) 418 continue 419 } 420 if tt.n == 0 { 421 continue 422 } 423 s := Join(a, tt.sep) 424 if s != tt.s { 425 t.Errorf("Join(Split(%q, %q, %d), %q) = %q", tt.s, tt.sep, tt.n, tt.sep, s) 426 } 427 if tt.n < 0 { 428 b := Split(tt.s, tt.sep) 429 if !reflect.DeepEqual(a, b) { 430 t.Errorf("Split disagrees with SplitN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 431 } 432 } 433 } 434 } 435 436 var splitaftertests = []SplitTest{ 437 {abcd, "a", -1, []string{"a", "bcd"}}, 438 {abcd, "z", -1, []string{"abcd"}}, 439 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 440 {commas, ",", -1, []string{"1,", "2,", "3,", "4"}}, 441 {dots, "...", -1, []string{"1...", ".2...", ".3...", ".4"}}, 442 {faces, "☹", -1, []string{"☺☻☹", ""}}, 443 {faces, "~", -1, []string{faces}}, 444 {faces, "", -1, []string{"☺", "☻", "☹"}}, 445 {"1 2 3 4", " ", 3, []string{"1 ", "2 ", "3 4"}}, 446 {"1 2 3", " ", 3, []string{"1 ", "2 ", "3"}}, 447 {"1 2", " ", 3, []string{"1 ", "2"}}, 448 {"123", "", 2, []string{"1", "23"}}, 449 {"123", "", 17, []string{"1", "2", "3"}}, 450 } 451 452 func TestSplitAfter(t *testing.T) { 453 for _, tt := range splitaftertests { 454 a := SplitAfterN(tt.s, tt.sep, tt.n) 455 if !eq(a, tt.a) { 456 t.Errorf(`Split(%q, %q, %d) = %v; want %v`, tt.s, tt.sep, tt.n, a, tt.a) 457 continue 458 } 459 s := Join(a, "") 460 if s != tt.s { 461 t.Errorf(`Join(Split(%q, %q, %d), %q) = %q`, tt.s, tt.sep, tt.n, tt.sep, s) 462 } 463 if tt.n < 0 { 464 b := SplitAfter(tt.s, tt.sep) 465 if !reflect.DeepEqual(a, b) { 466 t.Errorf("SplitAfter disagrees with SplitAfterN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 467 } 468 } 469 } 470 } 471 472 type FieldsTest struct { 473 s string 474 a []string 475 } 476 477 var fieldstests = []FieldsTest{ 478 {"", []string{}}, 479 {" ", []string{}}, 480 {" \t ", []string{}}, 481 {"\u2000", []string{}}, 482 {" abc ", []string{"abc"}}, 483 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 484 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 485 {"1\t\t2\t\t3\t4", []string{"1", "2", "3", "4"}}, 486 {"1\u20002\u20013\u20024", []string{"1", "2", "3", "4"}}, 487 {"\u2000\u2001\u2002", []string{}}, 488 {"\n™\t™\n", []string{"™", "™"}}, 489 {"\n\u20001™2\u2000 \u2001 ™", []string{"1™2", "™"}}, 490 {"\n1\uFFFD \uFFFD2\u20003\uFFFD4", []string{"1\uFFFD", "\uFFFD2", "3\uFFFD4"}}, 491 {"1\xFF\u2000\xFF2\xFF \xFF", []string{"1\xFF", "\xFF2\xFF", "\xFF"}}, 492 {faces, []string{faces}}, 493 } 494 495 func TestFields(t *testing.T) { 496 for _, tt := range fieldstests { 497 a := Fields(tt.s) 498 if !eq(a, tt.a) { 499 t.Errorf("Fields(%q) = %v; want %v", tt.s, a, tt.a) 500 continue 501 } 502 } 503 } 504 505 var FieldsFuncTests = []FieldsTest{ 506 {"", []string{}}, 507 {"XX", []string{}}, 508 {"XXhiXXX", []string{"hi"}}, 509 {"aXXbXXXcX", []string{"a", "b", "c"}}, 510 } 511 512 func TestFieldsFunc(t *testing.T) { 513 for _, tt := range fieldstests { 514 a := FieldsFunc(tt.s, unicode.IsSpace) 515 if !eq(a, tt.a) { 516 t.Errorf("FieldsFunc(%q, unicode.IsSpace) = %v; want %v", tt.s, a, tt.a) 517 continue 518 } 519 } 520 pred := func(c rune) bool { return c == 'X' } 521 for _, tt := range FieldsFuncTests { 522 a := FieldsFunc(tt.s, pred) 523 if !eq(a, tt.a) { 524 t.Errorf("FieldsFunc(%q) = %v, want %v", tt.s, a, tt.a) 525 } 526 } 527 } 528 529 // Test case for any function which accepts and returns a single string. 530 type StringTest struct { 531 in, out string 532 } 533 534 // Execute f on each test case. funcName should be the name of f; it's used 535 // in failure reports. 536 func runStringTests(t *testing.T, f func(string) string, funcName string, testCases []StringTest) { 537 for _, tc := range testCases { 538 actual := f(tc.in) 539 if actual != tc.out { 540 t.Errorf("%s(%q) = %q; want %q", funcName, tc.in, actual, tc.out) 541 } 542 } 543 } 544 545 var upperTests = []StringTest{ 546 {"", ""}, 547 {"ONLYUPPER", "ONLYUPPER"}, 548 {"abc", "ABC"}, 549 {"AbC123", "ABC123"}, 550 {"azAZ09_", "AZAZ09_"}, 551 {"longStrinGwitHmixofsmaLLandcAps", "LONGSTRINGWITHMIXOFSMALLANDCAPS"}, 552 {"RENAN BASTOS 93 AOSDAJDJAIDJAIDAJIaidsjjaidijadsjiadjiOOKKO", "RENAN BASTOS 93 AOSDAJDJAIDJAIDAJIAIDSJJAIDIJADSJIADJIOOKKO"}, 553 {"long\u0250string\u0250with\u0250nonascii\u2C6Fchars", "LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS"}, 554 {"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"}, // grows one byte per char 555 {"a\u0080\U0010FFFF", "A\u0080\U0010FFFF"}, // test utf8.RuneSelf and utf8.MaxRune 556 } 557 558 var lowerTests = []StringTest{ 559 {"", ""}, 560 {"abc", "abc"}, 561 {"AbC123", "abc123"}, 562 {"azAZ09_", "azaz09_"}, 563 {"longStrinGwitHmixofsmaLLandcAps", "longstringwithmixofsmallandcaps"}, 564 {"renan bastos 93 AOSDAJDJAIDJAIDAJIaidsjjaidijadsjiadjiOOKKO", "renan bastos 93 aosdajdjaidjaidajiaidsjjaidijadsjiadjiookko"}, 565 {"LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS", "long\u0250string\u0250with\u0250nonascii\u0250chars"}, 566 {"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"}, // shrinks one byte per char 567 {"A\u0080\U0010FFFF", "a\u0080\U0010FFFF"}, // test utf8.RuneSelf and utf8.MaxRune 568 } 569 570 const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000" 571 572 var trimSpaceTests = []StringTest{ 573 {"", ""}, 574 {"abc", "abc"}, 575 {space + "abc" + space, "abc"}, 576 {" ", ""}, 577 {" \t\r\n \t\t\r\r\n\n ", ""}, 578 {" \t\r\n x\t\t\r\r\n\n ", "x"}, 579 {" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, 580 {"1 \t\r\n2", "1 \t\r\n2"}, 581 {" x\x80", "x\x80"}, 582 {" x\xc0", "x\xc0"}, 583 {"x \xc0\xc0 ", "x \xc0\xc0"}, 584 {"x \xc0", "x \xc0"}, 585 {"x \xc0 ", "x \xc0"}, 586 {"x \xc0\xc0 ", "x \xc0\xc0"}, 587 {"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, 588 {"x ☺ ", "x ☺"}, 589 } 590 591 func tenRunes(ch rune) string { 592 r := make([]rune, 10) 593 for i := range r { 594 r[i] = ch 595 } 596 return string(r) 597 } 598 599 // User-defined self-inverse mapping function 600 func rot13(r rune) rune { 601 step := rune(13) 602 if r >= 'a' && r <= 'z' { 603 return ((r - 'a' + step) % 26) + 'a' 604 } 605 if r >= 'A' && r <= 'Z' { 606 return ((r - 'A' + step) % 26) + 'A' 607 } 608 return r 609 } 610 611 func TestMap(t *testing.T) { 612 // Run a couple of awful growth/shrinkage tests 613 a := tenRunes('a') 614 // 1. Grow. This triggers two reallocations in Map. 615 maxRune := func(rune) rune { return unicode.MaxRune } 616 m := Map(maxRune, a) 617 expect := tenRunes(unicode.MaxRune) 618 if m != expect { 619 t.Errorf("growing: expected %q got %q", expect, m) 620 } 621 622 // 2. Shrink 623 minRune := func(rune) rune { return 'a' } 624 m = Map(minRune, tenRunes(unicode.MaxRune)) 625 expect = a 626 if m != expect { 627 t.Errorf("shrinking: expected %q got %q", expect, m) 628 } 629 630 // 3. Rot13 631 m = Map(rot13, "a to zed") 632 expect = "n gb mrq" 633 if m != expect { 634 t.Errorf("rot13: expected %q got %q", expect, m) 635 } 636 637 // 4. Rot13^2 638 m = Map(rot13, Map(rot13, "a to zed")) 639 expect = "a to zed" 640 if m != expect { 641 t.Errorf("rot13: expected %q got %q", expect, m) 642 } 643 644 // 5. Drop 645 dropNotLatin := func(r rune) rune { 646 if unicode.Is(unicode.Latin, r) { 647 return r 648 } 649 return -1 650 } 651 m = Map(dropNotLatin, "Hello, 세계") 652 expect = "Hello" 653 if m != expect { 654 t.Errorf("drop: expected %q got %q", expect, m) 655 } 656 657 // 6. Identity 658 identity := func(r rune) rune { 659 return r 660 } 661 orig := "Input string that we expect not to be copied." 662 m = Map(identity, orig) 663 if unsafe.StringData(orig) != unsafe.StringData(m) { 664 t.Error("unexpected copy during identity map") 665 } 666 667 // 7. Handle invalid UTF-8 sequence 668 replaceNotLatin := func(r rune) rune { 669 if unicode.Is(unicode.Latin, r) { 670 return r 671 } 672 return utf8.RuneError 673 } 674 m = Map(replaceNotLatin, "Hello\255World") 675 expect = "Hello\uFFFDWorld" 676 if m != expect { 677 t.Errorf("replace invalid sequence: expected %q got %q", expect, m) 678 } 679 680 // 8. Check utf8.RuneSelf and utf8.MaxRune encoding 681 encode := func(r rune) rune { 682 switch r { 683 case utf8.RuneSelf: 684 return unicode.MaxRune 685 case unicode.MaxRune: 686 return utf8.RuneSelf 687 } 688 return r 689 } 690 s := string(rune(utf8.RuneSelf)) + string(utf8.MaxRune) 691 r := string(utf8.MaxRune) + string(rune(utf8.RuneSelf)) // reverse of s 692 m = Map(encode, s) 693 if m != r { 694 t.Errorf("encoding not handled correctly: expected %q got %q", r, m) 695 } 696 m = Map(encode, r) 697 if m != s { 698 t.Errorf("encoding not handled correctly: expected %q got %q", s, m) 699 } 700 701 // 9. Check mapping occurs in the front, middle and back 702 trimSpaces := func(r rune) rune { 703 if unicode.IsSpace(r) { 704 return -1 705 } 706 return r 707 } 708 m = Map(trimSpaces, " abc 123 ") 709 expect = "abc123" 710 if m != expect { 711 t.Errorf("trimSpaces: expected %q got %q", expect, m) 712 } 713 } 714 715 func TestToUpper(t *testing.T) { runStringTests(t, ToUpper, "ToUpper", upperTests) } 716 717 func TestToLower(t *testing.T) { runStringTests(t, ToLower, "ToLower", lowerTests) } 718 719 var toValidUTF8Tests = []struct { 720 in string 721 repl string 722 out string 723 }{ 724 {"", "\uFFFD", ""}, 725 {"abc", "\uFFFD", "abc"}, 726 {"\uFDDD", "\uFFFD", "\uFDDD"}, 727 {"a\xffb", "\uFFFD", "a\uFFFDb"}, 728 {"a\xffb\uFFFD", "X", "aXb\uFFFD"}, 729 {"a☺\xffb☺\xC0\xAFc☺\xff", "", "a☺b☺c☺"}, 730 {"a☺\xffb☺\xC0\xAFc☺\xff", "日本語", "a☺日本語b☺日本語c☺日本語"}, 731 {"\xC0\xAF", "\uFFFD", "\uFFFD"}, 732 {"\xE0\x80\xAF", "\uFFFD", "\uFFFD"}, 733 {"\xed\xa0\x80", "abc", "abc"}, 734 {"\xed\xbf\xbf", "\uFFFD", "\uFFFD"}, 735 {"\xF0\x80\x80\xaf", "☺", "☺"}, 736 {"\xF8\x80\x80\x80\xAF", "\uFFFD", "\uFFFD"}, 737 {"\xFC\x80\x80\x80\x80\xAF", "\uFFFD", "\uFFFD"}, 738 } 739 740 func TestToValidUTF8(t *testing.T) { 741 for _, tc := range toValidUTF8Tests { 742 got := ToValidUTF8(tc.in, tc.repl) 743 if got != tc.out { 744 t.Errorf("ToValidUTF8(%q, %q) = %q; want %q", tc.in, tc.repl, got, tc.out) 745 } 746 } 747 } 748 749 func BenchmarkToUpper(b *testing.B) { 750 for _, tc := range upperTests { 751 b.Run(tc.in, func(b *testing.B) { 752 for i := 0; i < b.N; i++ { 753 actual := ToUpper(tc.in) 754 if actual != tc.out { 755 b.Errorf("ToUpper(%q) = %q; want %q", tc.in, actual, tc.out) 756 } 757 } 758 }) 759 } 760 } 761 762 func BenchmarkToLower(b *testing.B) { 763 for _, tc := range lowerTests { 764 b.Run(tc.in, func(b *testing.B) { 765 for i := 0; i < b.N; i++ { 766 actual := ToLower(tc.in) 767 if actual != tc.out { 768 b.Errorf("ToLower(%q) = %q; want %q", tc.in, actual, tc.out) 769 } 770 } 771 }) 772 } 773 } 774 775 func BenchmarkMapNoChanges(b *testing.B) { 776 identity := func(r rune) rune { 777 return r 778 } 779 for i := 0; i < b.N; i++ { 780 Map(identity, "Some string that won't be modified.") 781 } 782 } 783 784 func TestSpecialCase(t *testing.T) { 785 lower := "abcçdefgğhıijklmnoöprsştuüvyz" 786 upper := "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ" 787 u := ToUpperSpecial(unicode.TurkishCase, upper) 788 if u != upper { 789 t.Errorf("Upper(upper) is %s not %s", u, upper) 790 } 791 u = ToUpperSpecial(unicode.TurkishCase, lower) 792 if u != upper { 793 t.Errorf("Upper(lower) is %s not %s", u, upper) 794 } 795 l := ToLowerSpecial(unicode.TurkishCase, lower) 796 if l != lower { 797 t.Errorf("Lower(lower) is %s not %s", l, lower) 798 } 799 l = ToLowerSpecial(unicode.TurkishCase, upper) 800 if l != lower { 801 t.Errorf("Lower(upper) is %s not %s", l, lower) 802 } 803 } 804 805 func TestTrimSpace(t *testing.T) { runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests) } 806 807 var trimTests = []struct { 808 f string 809 in, arg, out string 810 }{ 811 {"Trim", "abba", "a", "bb"}, 812 {"Trim", "abba", "ab", ""}, 813 {"TrimLeft", "abba", "ab", ""}, 814 {"TrimRight", "abba", "ab", ""}, 815 {"TrimLeft", "abba", "a", "bba"}, 816 {"TrimLeft", "abba", "b", "abba"}, 817 {"TrimRight", "abba", "a", "abb"}, 818 {"TrimRight", "abba", "b", "abba"}, 819 {"Trim", "<tag>", "<>", "tag"}, 820 {"Trim", "* listitem", " *", "listitem"}, 821 {"Trim", `"quote"`, `"`, "quote"}, 822 {"Trim", "\u2C6F\u2C6F\u0250\u0250\u2C6F\u2C6F", "\u2C6F", "\u0250\u0250"}, 823 {"Trim", "\x80test\xff", "\xff", "test"}, 824 {"Trim", " Ġ ", " ", "Ġ"}, 825 {"Trim", " Ġİ0", "0 ", "Ġİ"}, 826 //empty string tests 827 {"Trim", "abba", "", "abba"}, 828 {"Trim", "", "123", ""}, 829 {"Trim", "", "", ""}, 830 {"TrimLeft", "abba", "", "abba"}, 831 {"TrimLeft", "", "123", ""}, 832 {"TrimLeft", "", "", ""}, 833 {"TrimRight", "abba", "", "abba"}, 834 {"TrimRight", "", "123", ""}, 835 {"TrimRight", "", "", ""}, 836 {"TrimRight", "☺\xc0", "☺", "☺\xc0"}, 837 {"TrimPrefix", "aabb", "a", "abb"}, 838 {"TrimPrefix", "aabb", "b", "aabb"}, 839 {"TrimSuffix", "aabb", "a", "aabb"}, 840 {"TrimSuffix", "aabb", "b", "aab"}, 841 } 842 843 func TestTrim(t *testing.T) { 844 for _, tc := range trimTests { 845 name := tc.f 846 var f func(string, string) string 847 switch name { 848 case "Trim": 849 f = Trim 850 case "TrimLeft": 851 f = TrimLeft 852 case "TrimRight": 853 f = TrimRight 854 case "TrimPrefix": 855 f = TrimPrefix 856 case "TrimSuffix": 857 f = TrimSuffix 858 default: 859 t.Errorf("Undefined trim function %s", name) 860 } 861 actual := f(tc.in, tc.arg) 862 if actual != tc.out { 863 t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 864 } 865 } 866 } 867 868 func BenchmarkTrim(b *testing.B) { 869 b.ReportAllocs() 870 871 for i := 0; i < b.N; i++ { 872 for _, tc := range trimTests { 873 name := tc.f 874 var f func(string, string) string 875 switch name { 876 case "Trim": 877 f = Trim 878 case "TrimLeft": 879 f = TrimLeft 880 case "TrimRight": 881 f = TrimRight 882 case "TrimPrefix": 883 f = TrimPrefix 884 case "TrimSuffix": 885 f = TrimSuffix 886 default: 887 b.Errorf("Undefined trim function %s", name) 888 } 889 actual := f(tc.in, tc.arg) 890 if actual != tc.out { 891 b.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 892 } 893 } 894 } 895 } 896 897 func BenchmarkToValidUTF8(b *testing.B) { 898 tests := []struct { 899 name string 900 input string 901 }{ 902 {"Valid", "typical"}, 903 {"InvalidASCII", "foo\xffbar"}, 904 {"InvalidNonASCII", "日本語\xff日本語"}, 905 } 906 replacement := "\uFFFD" 907 b.ResetTimer() 908 for _, test := range tests { 909 b.Run(test.name, func(b *testing.B) { 910 for i := 0; i < b.N; i++ { 911 ToValidUTF8(test.input, replacement) 912 } 913 }) 914 } 915 } 916 917 type predicate struct { 918 f func(rune) bool 919 name string 920 } 921 922 var isSpace = predicate{unicode.IsSpace, "IsSpace"} 923 var isDigit = predicate{unicode.IsDigit, "IsDigit"} 924 var isUpper = predicate{unicode.IsUpper, "IsUpper"} 925 var isValidRune = predicate{ 926 func(r rune) bool { 927 return r != utf8.RuneError 928 }, 929 "IsValidRune", 930 } 931 932 func not(p predicate) predicate { 933 return predicate{ 934 func(r rune) bool { 935 return !p.f(r) 936 }, 937 "not " + p.name, 938 } 939 } 940 941 var trimFuncTests = []struct { 942 f predicate 943 in string 944 trimOut string 945 leftOut string 946 rightOut string 947 }{ 948 {isSpace, space + " hello " + space, 949 "hello", 950 "hello " + space, 951 space + " hello"}, 952 {isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", 953 "hello", 954 "hello34\u0e50\u0e51", 955 "\u0e50\u0e5212hello"}, 956 {isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", 957 "hello", 958 "helloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", 959 "\u2C6F\u2C6F\u2C6F\u2C6FABCDhello"}, 960 {not(isSpace), "hello" + space + "hello", 961 space, 962 space + "hello", 963 "hello" + space}, 964 {not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", 965 "\u0e50\u0e521234\u0e50\u0e51", 966 "\u0e50\u0e521234\u0e50\u0e51helo", 967 "hello\u0e50\u0e521234\u0e50\u0e51"}, 968 {isValidRune, "ab\xc0a\xc0cd", 969 "\xc0a\xc0", 970 "\xc0a\xc0cd", 971 "ab\xc0a\xc0"}, 972 {not(isValidRune), "\xc0a\xc0", 973 "a", 974 "a\xc0", 975 "\xc0a"}, 976 {isSpace, "", 977 "", 978 "", 979 ""}, 980 {isSpace, " ", 981 "", 982 "", 983 ""}, 984 } 985 986 func TestTrimFunc(t *testing.T) { 987 for _, tc := range trimFuncTests { 988 trimmers := []struct { 989 name string 990 trim func(s string, f func(r rune) bool) string 991 out string 992 }{ 993 {"TrimFunc", TrimFunc, tc.trimOut}, 994 {"TrimLeftFunc", TrimLeftFunc, tc.leftOut}, 995 {"TrimRightFunc", TrimRightFunc, tc.rightOut}, 996 } 997 for _, trimmer := range trimmers { 998 actual := trimmer.trim(tc.in, tc.f.f) 999 if actual != trimmer.out { 1000 t.Errorf("%s(%q, %q) = %q; want %q", trimmer.name, tc.in, tc.f.name, actual, trimmer.out) 1001 } 1002 } 1003 } 1004 } 1005 1006 var indexFuncTests = []struct { 1007 in string 1008 f predicate 1009 first, last int 1010 }{ 1011 {"", isValidRune, -1, -1}, 1012 {"abc", isDigit, -1, -1}, 1013 {"0123", isDigit, 0, 3}, 1014 {"a1b", isDigit, 1, 1}, 1015 {space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes 1016 {"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, 1017 {"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, 1018 {"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, 1019 1020 // tests of invalid UTF-8 1021 {"\x801", isDigit, 1, 1}, 1022 {"\x80abc", isDigit, -1, -1}, 1023 {"\xc0a\xc0", isValidRune, 1, 1}, 1024 {"\xc0a\xc0", not(isValidRune), 0, 2}, 1025 {"\xc0☺\xc0", not(isValidRune), 0, 4}, 1026 {"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, 1027 {"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, 1028 {"a\xe0\x80cd", not(isValidRune), 1, 2}, 1029 {"\x80\x80\x80\x80", not(isValidRune), 0, 3}, 1030 } 1031 1032 func TestIndexFunc(t *testing.T) { 1033 for _, tc := range indexFuncTests { 1034 first := IndexFunc(tc.in, tc.f.f) 1035 if first != tc.first { 1036 t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) 1037 } 1038 last := LastIndexFunc(tc.in, tc.f.f) 1039 if last != tc.last { 1040 t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) 1041 } 1042 } 1043 } 1044 1045 func equal(m string, s1, s2 string, t *testing.T) bool { 1046 if s1 == s2 { 1047 return true 1048 } 1049 e1 := Split(s1, "") 1050 e2 := Split(s2, "") 1051 for i, c1 := range e1 { 1052 if i >= len(e2) { 1053 break 1054 } 1055 r1, _ := utf8.DecodeRuneInString(c1) 1056 r2, _ := utf8.DecodeRuneInString(e2[i]) 1057 if r1 != r2 { 1058 t.Errorf("%s diff at %d: U+%04X U+%04X", m, i, r1, r2) 1059 } 1060 } 1061 return false 1062 } 1063 1064 func TestCaseConsistency(t *testing.T) { 1065 // Make a string of all the runes. 1066 numRunes := int(unicode.MaxRune + 1) 1067 if testing.Short() { 1068 numRunes = 1000 1069 } 1070 a := make([]rune, numRunes) 1071 for i := range a { 1072 a[i] = rune(i) 1073 } 1074 s := string(a) 1075 // convert the cases. 1076 upper := ToUpper(s) 1077 lower := ToLower(s) 1078 1079 // Consistency checks 1080 if n := utf8.RuneCountInString(upper); n != numRunes { 1081 t.Error("rune count wrong in upper:", n) 1082 } 1083 if n := utf8.RuneCountInString(lower); n != numRunes { 1084 t.Error("rune count wrong in lower:", n) 1085 } 1086 if !equal("ToUpper(upper)", ToUpper(upper), upper, t) { 1087 t.Error("ToUpper(upper) consistency fail") 1088 } 1089 if !equal("ToLower(lower)", ToLower(lower), lower, t) { 1090 t.Error("ToLower(lower) consistency fail") 1091 } 1092 /* 1093 These fail because of non-one-to-oneness of the data, such as multiple 1094 upper case 'I' mapping to 'i'. We comment them out but keep them for 1095 interest. 1096 For instance: CAPITAL LETTER I WITH DOT ABOVE: 1097 unicode.ToUpper(unicode.ToLower('\u0130')) != '\u0130' 1098 1099 if !equal("ToUpper(lower)", ToUpper(lower), upper, t) { 1100 t.Error("ToUpper(lower) consistency fail"); 1101 } 1102 if !equal("ToLower(upper)", ToLower(upper), lower, t) { 1103 t.Error("ToLower(upper) consistency fail"); 1104 } 1105 */ 1106 } 1107 1108 var longString = "a" + string(make([]byte, 1<<16)) + "z" 1109 1110 var RepeatTests = []struct { 1111 in, out string 1112 count int 1113 }{ 1114 {"", "", 0}, 1115 {"", "", 1}, 1116 {"", "", 2}, 1117 {"-", "", 0}, 1118 {"-", "-", 1}, 1119 {"-", "----------", 10}, 1120 {"abc ", "abc abc abc ", 3}, 1121 // Tests for results over the chunkLimit 1122 {string(rune(0)), string(make([]byte, 1<<16)), 1 << 16}, 1123 {longString, longString + longString, 2}, 1124 } 1125 1126 func TestRepeat(t *testing.T) { 1127 for _, tt := range RepeatTests { 1128 a := Repeat(tt.in, tt.count) 1129 if !equal("Repeat(s)", a, tt.out, t) { 1130 t.Errorf("Repeat(%v, %d) = %v; want %v", tt.in, tt.count, a, tt.out) 1131 continue 1132 } 1133 } 1134 } 1135 1136 func repeat(s string, count int) (err error) { 1137 defer func() { 1138 if r := recover(); r != nil { 1139 switch v := r.(type) { 1140 case error: 1141 err = v 1142 default: 1143 err = fmt.Errorf("%s", v) 1144 } 1145 } 1146 }() 1147 1148 Repeat(s, count) 1149 1150 return 1151 } 1152 1153 // See Issue golang.org/issue/16237 1154 func TestRepeatCatchesOverflow(t *testing.T) { 1155 tests := [...]struct { 1156 s string 1157 count int 1158 errStr string 1159 }{ 1160 0: {"--", -2147483647, "negative"}, 1161 1: {"", int(^uint(0) >> 1), ""}, 1162 2: {"-", 10, ""}, 1163 3: {"gopher", 0, ""}, 1164 4: {"-", -1, "negative"}, 1165 5: {"--", -102, "negative"}, 1166 6: {string(make([]byte, 255)), int((^uint(0))/255 + 1), "overflow"}, 1167 } 1168 1169 for i, tt := range tests { 1170 err := repeat(tt.s, tt.count) 1171 if tt.errStr == "" { 1172 if err != nil { 1173 t.Errorf("#%d panicked %v", i, err) 1174 } 1175 continue 1176 } 1177 1178 if err == nil || !Contains(err.Error(), tt.errStr) { 1179 t.Errorf("#%d expected %q got %q", i, tt.errStr, err) 1180 } 1181 } 1182 } 1183 1184 func runesEqual(a, b []rune) bool { 1185 if len(a) != len(b) { 1186 return false 1187 } 1188 for i, r := range a { 1189 if r != b[i] { 1190 return false 1191 } 1192 } 1193 return true 1194 } 1195 1196 var RunesTests = []struct { 1197 in string 1198 out []rune 1199 lossy bool 1200 }{ 1201 {"", []rune{}, false}, 1202 {" ", []rune{32}, false}, 1203 {"ABC", []rune{65, 66, 67}, false}, 1204 {"abc", []rune{97, 98, 99}, false}, 1205 {"\u65e5\u672c\u8a9e", []rune{26085, 26412, 35486}, false}, 1206 {"ab\x80c", []rune{97, 98, 0xFFFD, 99}, true}, 1207 {"ab\xc0c", []rune{97, 98, 0xFFFD, 99}, true}, 1208 } 1209 1210 func TestRunes(t *testing.T) { 1211 for _, tt := range RunesTests { 1212 a := []rune(tt.in) 1213 if !runesEqual(a, tt.out) { 1214 t.Errorf("[]rune(%q) = %v; want %v", tt.in, a, tt.out) 1215 continue 1216 } 1217 if !tt.lossy { 1218 // can only test reassembly if we didn't lose information 1219 s := string(a) 1220 if s != tt.in { 1221 t.Errorf("string([]rune(%q)) = %x; want %x", tt.in, s, tt.in) 1222 } 1223 } 1224 } 1225 } 1226 1227 func TestReadByte(t *testing.T) { 1228 testStrings := []string{"", abcd, faces, commas} 1229 for _, s := range testStrings { 1230 reader := NewReader(s) 1231 if e := reader.UnreadByte(); e == nil { 1232 t.Errorf("Unreading %q at beginning: expected error", s) 1233 } 1234 var res bytes.Buffer 1235 for { 1236 b, e := reader.ReadByte() 1237 if e == io.EOF { 1238 break 1239 } 1240 if e != nil { 1241 t.Errorf("Reading %q: %s", s, e) 1242 break 1243 } 1244 res.WriteByte(b) 1245 // unread and read again 1246 e = reader.UnreadByte() 1247 if e != nil { 1248 t.Errorf("Unreading %q: %s", s, e) 1249 break 1250 } 1251 b1, e := reader.ReadByte() 1252 if e != nil { 1253 t.Errorf("Reading %q after unreading: %s", s, e) 1254 break 1255 } 1256 if b1 != b { 1257 t.Errorf("Reading %q after unreading: want byte %q, got %q", s, b, b1) 1258 break 1259 } 1260 } 1261 if res.String() != s { 1262 t.Errorf("Reader(%q).ReadByte() produced %q", s, res.String()) 1263 } 1264 } 1265 } 1266 1267 func TestReadRune(t *testing.T) { 1268 testStrings := []string{"", abcd, faces, commas} 1269 for _, s := range testStrings { 1270 reader := NewReader(s) 1271 if e := reader.UnreadRune(); e == nil { 1272 t.Errorf("Unreading %q at beginning: expected error", s) 1273 } 1274 res := "" 1275 for { 1276 r, z, e := reader.ReadRune() 1277 if e == io.EOF { 1278 break 1279 } 1280 if e != nil { 1281 t.Errorf("Reading %q: %s", s, e) 1282 break 1283 } 1284 res += string(r) 1285 // unread and read again 1286 e = reader.UnreadRune() 1287 if e != nil { 1288 t.Errorf("Unreading %q: %s", s, e) 1289 break 1290 } 1291 r1, z1, e := reader.ReadRune() 1292 if e != nil { 1293 t.Errorf("Reading %q after unreading: %s", s, e) 1294 break 1295 } 1296 if r1 != r { 1297 t.Errorf("Reading %q after unreading: want rune %q, got %q", s, r, r1) 1298 break 1299 } 1300 if z1 != z { 1301 t.Errorf("Reading %q after unreading: want size %d, got %d", s, z, z1) 1302 break 1303 } 1304 } 1305 if res != s { 1306 t.Errorf("Reader(%q).ReadRune() produced %q", s, res) 1307 } 1308 } 1309 } 1310 1311 var UnreadRuneErrorTests = []struct { 1312 name string 1313 f func(*Reader) 1314 }{ 1315 {"Read", func(r *Reader) { r.Read([]byte{0}) }}, 1316 {"ReadByte", func(r *Reader) { r.ReadByte() }}, 1317 {"UnreadRune", func(r *Reader) { r.UnreadRune() }}, 1318 {"Seek", func(r *Reader) { r.Seek(0, io.SeekCurrent) }}, 1319 {"WriteTo", func(r *Reader) { r.WriteTo(&bytes.Buffer{}) }}, 1320 } 1321 1322 func TestUnreadRuneError(t *testing.T) { 1323 for _, tt := range UnreadRuneErrorTests { 1324 reader := NewReader("0123456789") 1325 if _, _, err := reader.ReadRune(); err != nil { 1326 // should not happen 1327 t.Fatal(err) 1328 } 1329 tt.f(reader) 1330 err := reader.UnreadRune() 1331 if err == nil { 1332 t.Errorf("Unreading after %s: expected error", tt.name) 1333 } 1334 } 1335 } 1336 1337 var ReplaceTests = []struct { 1338 in string 1339 old, new string 1340 n int 1341 out string 1342 }{ 1343 {"hello", "l", "L", 0, "hello"}, 1344 {"hello", "l", "L", -1, "heLLo"}, 1345 {"hello", "x", "X", -1, "hello"}, 1346 {"", "x", "X", -1, ""}, 1347 {"radar", "r", "<r>", -1, "<r>ada<r>"}, 1348 {"", "", "<>", -1, "<>"}, 1349 {"banana", "a", "<>", -1, "b<>n<>n<>"}, 1350 {"banana", "a", "<>", 1, "b<>nana"}, 1351 {"banana", "a", "<>", 1000, "b<>n<>n<>"}, 1352 {"banana", "an", "<>", -1, "b<><>a"}, 1353 {"banana", "ana", "<>", -1, "b<>na"}, 1354 {"banana", "", "<>", -1, "<>b<>a<>n<>a<>n<>a<>"}, 1355 {"banana", "", "<>", 10, "<>b<>a<>n<>a<>n<>a<>"}, 1356 {"banana", "", "<>", 6, "<>b<>a<>n<>a<>n<>a"}, 1357 {"banana", "", "<>", 5, "<>b<>a<>n<>a<>na"}, 1358 {"banana", "", "<>", 1, "<>banana"}, 1359 {"banana", "a", "a", -1, "banana"}, 1360 {"banana", "a", "a", 1, "banana"}, 1361 {"☺☻☹", "", "<>", -1, "<>☺<>☻<>☹<>"}, 1362 } 1363 1364 func TestReplace(t *testing.T) { 1365 for _, tt := range ReplaceTests { 1366 if s := Replace(tt.in, tt.old, tt.new, tt.n); s != tt.out { 1367 t.Errorf("Replace(%q, %q, %q, %d) = %q, want %q", tt.in, tt.old, tt.new, tt.n, s, tt.out) 1368 } 1369 if tt.n == -1 { 1370 s := ReplaceAll(tt.in, tt.old, tt.new) 1371 if s != tt.out { 1372 t.Errorf("ReplaceAll(%q, %q, %q) = %q, want %q", tt.in, tt.old, tt.new, s, tt.out) 1373 } 1374 } 1375 } 1376 } 1377 1378 var TitleTests = []struct { 1379 in, out string 1380 }{ 1381 {"", ""}, 1382 {"a", "A"}, 1383 {" aaa aaa aaa ", " Aaa Aaa Aaa "}, 1384 {" Aaa Aaa Aaa ", " Aaa Aaa Aaa "}, 1385 {"123a456", "123a456"}, 1386 {"double-blind", "Double-Blind"}, 1387 {"ÿøû", "Ÿøû"}, 1388 {"with_underscore", "With_underscore"}, 1389 {"unicode \xe2\x80\xa8 line separator", "Unicode \xe2\x80\xa8 Line Separator"}, 1390 } 1391 1392 func TestTitle(t *testing.T) { 1393 for _, tt := range TitleTests { 1394 if s := Title(tt.in); s != tt.out { 1395 t.Errorf("Title(%q) = %q, want %q", tt.in, s, tt.out) 1396 } 1397 } 1398 } 1399 1400 var ContainsTests = []struct { 1401 str, substr string 1402 expected bool 1403 }{ 1404 {"abc", "bc", true}, 1405 {"abc", "bcd", false}, 1406 {"abc", "", true}, 1407 {"", "a", false}, 1408 1409 // cases to cover code in runtime/asm_amd64.s:indexShortStr 1410 // 2-byte needle 1411 {"xxxxxx", "01", false}, 1412 {"01xxxx", "01", true}, 1413 {"xx01xx", "01", true}, 1414 {"xxxx01", "01", true}, 1415 {"01xxxxx"[1:], "01", false}, 1416 {"xxxxx01"[:6], "01", false}, 1417 // 3-byte needle 1418 {"xxxxxxx", "012", false}, 1419 {"012xxxx", "012", true}, 1420 {"xx012xx", "012", true}, 1421 {"xxxx012", "012", true}, 1422 {"012xxxxx"[1:], "012", false}, 1423 {"xxxxx012"[:7], "012", false}, 1424 // 4-byte needle 1425 {"xxxxxxxx", "0123", false}, 1426 {"0123xxxx", "0123", true}, 1427 {"xx0123xx", "0123", true}, 1428 {"xxxx0123", "0123", true}, 1429 {"0123xxxxx"[1:], "0123", false}, 1430 {"xxxxx0123"[:8], "0123", false}, 1431 // 5-7-byte needle 1432 {"xxxxxxxxx", "01234", false}, 1433 {"01234xxxx", "01234", true}, 1434 {"xx01234xx", "01234", true}, 1435 {"xxxx01234", "01234", true}, 1436 {"01234xxxxx"[1:], "01234", false}, 1437 {"xxxxx01234"[:9], "01234", false}, 1438 // 8-byte needle 1439 {"xxxxxxxxxxxx", "01234567", false}, 1440 {"01234567xxxx", "01234567", true}, 1441 {"xx01234567xx", "01234567", true}, 1442 {"xxxx01234567", "01234567", true}, 1443 {"01234567xxxxx"[1:], "01234567", false}, 1444 {"xxxxx01234567"[:12], "01234567", false}, 1445 // 9-15-byte needle 1446 {"xxxxxxxxxxxxx", "012345678", false}, 1447 {"012345678xxxx", "012345678", true}, 1448 {"xx012345678xx", "012345678", true}, 1449 {"xxxx012345678", "012345678", true}, 1450 {"012345678xxxxx"[1:], "012345678", false}, 1451 {"xxxxx012345678"[:13], "012345678", false}, 1452 // 16-byte needle 1453 {"xxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEF", false}, 1454 {"0123456789ABCDEFxxxx", "0123456789ABCDEF", true}, 1455 {"xx0123456789ABCDEFxx", "0123456789ABCDEF", true}, 1456 {"xxxx0123456789ABCDEF", "0123456789ABCDEF", true}, 1457 {"0123456789ABCDEFxxxxx"[1:], "0123456789ABCDEF", false}, 1458 {"xxxxx0123456789ABCDEF"[:20], "0123456789ABCDEF", false}, 1459 // 17-31-byte needle 1460 {"xxxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEFG", false}, 1461 {"0123456789ABCDEFGxxxx", "0123456789ABCDEFG", true}, 1462 {"xx0123456789ABCDEFGxx", "0123456789ABCDEFG", true}, 1463 {"xxxx0123456789ABCDEFG", "0123456789ABCDEFG", true}, 1464 {"0123456789ABCDEFGxxxxx"[1:], "0123456789ABCDEFG", false}, 1465 {"xxxxx0123456789ABCDEFG"[:21], "0123456789ABCDEFG", false}, 1466 1467 // partial match cases 1468 {"xx01x", "012", false}, // 3 1469 {"xx0123x", "01234", false}, // 5-7 1470 {"xx01234567x", "012345678", false}, // 9-15 1471 {"xx0123456789ABCDEFx", "0123456789ABCDEFG", false}, // 17-31, issue 15679 1472 } 1473 1474 func TestContains(t *testing.T) { 1475 for _, ct := range ContainsTests { 1476 if Contains(ct.str, ct.substr) != ct.expected { 1477 t.Errorf("Contains(%s, %s) = %v, want %v", 1478 ct.str, ct.substr, !ct.expected, ct.expected) 1479 } 1480 } 1481 } 1482 1483 var ContainsAnyTests = []struct { 1484 str, substr string 1485 expected bool 1486 }{ 1487 {"", "", false}, 1488 {"", "a", false}, 1489 {"", "abc", false}, 1490 {"a", "", false}, 1491 {"a", "a", true}, 1492 {"aaa", "a", true}, 1493 {"abc", "xyz", false}, 1494 {"abc", "xcz", true}, 1495 {"a☺b☻c☹d", "uvw☻xyz", true}, 1496 {"aRegExp*", ".(|)*+?^$[]", true}, 1497 {dots + dots + dots, " ", false}, 1498 } 1499 1500 func TestContainsAny(t *testing.T) { 1501 for _, ct := range ContainsAnyTests { 1502 if ContainsAny(ct.str, ct.substr) != ct.expected { 1503 t.Errorf("ContainsAny(%s, %s) = %v, want %v", 1504 ct.str, ct.substr, !ct.expected, ct.expected) 1505 } 1506 } 1507 } 1508 1509 var ContainsRuneTests = []struct { 1510 str string 1511 r rune 1512 expected bool 1513 }{ 1514 {"", 'a', false}, 1515 {"a", 'a', true}, 1516 {"aaa", 'a', true}, 1517 {"abc", 'y', false}, 1518 {"abc", 'c', true}, 1519 {"a☺b☻c☹d", 'x', false}, 1520 {"a☺b☻c☹d", '☻', true}, 1521 {"aRegExp*", '*', true}, 1522 } 1523 1524 func TestContainsRune(t *testing.T) { 1525 for _, ct := range ContainsRuneTests { 1526 if ContainsRune(ct.str, ct.r) != ct.expected { 1527 t.Errorf("ContainsRune(%q, %q) = %v, want %v", 1528 ct.str, ct.r, !ct.expected, ct.expected) 1529 } 1530 } 1531 } 1532 1533 var EqualFoldTests = []struct { 1534 s, t string 1535 out bool 1536 }{ 1537 {"abc", "abc", true}, 1538 {"ABcd", "ABcd", true}, 1539 {"123abc", "123ABC", true}, 1540 {"αβδ", "ΑΒΔ", true}, 1541 {"abc", "xyz", false}, 1542 {"abc", "XYZ", false}, 1543 {"abcdefghijk", "abcdefghijX", false}, 1544 {"abcdefghijk", "abcdefghij\u212A", true}, 1545 {"abcdefghijK", "abcdefghij\u212A", true}, 1546 {"abcdefghijkz", "abcdefghij\u212Ay", false}, 1547 {"abcdefghijKz", "abcdefghij\u212Ay", false}, 1548 {"1", "2", false}, 1549 {"utf-8", "US-ASCII", false}, 1550 } 1551 1552 func TestEqualFold(t *testing.T) { 1553 for _, tt := range EqualFoldTests { 1554 if out := EqualFold(tt.s, tt.t); out != tt.out { 1555 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.s, tt.t, out, tt.out) 1556 } 1557 if out := EqualFold(tt.t, tt.s); out != tt.out { 1558 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.t, tt.s, out, tt.out) 1559 } 1560 } 1561 } 1562 1563 func BenchmarkEqualFold(b *testing.B) { 1564 b.Run("Tests", func(b *testing.B) { 1565 for i := 0; i < b.N; i++ { 1566 for _, tt := range EqualFoldTests { 1567 if out := EqualFold(tt.s, tt.t); out != tt.out { 1568 b.Fatal("wrong result") 1569 } 1570 } 1571 } 1572 }) 1573 1574 const s1 = "abcdefghijKz" 1575 const s2 = "abcDefGhijKz" 1576 1577 b.Run("ASCII", func(b *testing.B) { 1578 for i := 0; i < b.N; i++ { 1579 EqualFold(s1, s2) 1580 } 1581 }) 1582 1583 b.Run("UnicodePrefix", func(b *testing.B) { 1584 for i := 0; i < b.N; i++ { 1585 EqualFold("αβδ"+s1, "ΑΒΔ"+s2) 1586 } 1587 }) 1588 1589 b.Run("UnicodeSuffix", func(b *testing.B) { 1590 for i := 0; i < b.N; i++ { 1591 EqualFold(s1+"αβδ", s2+"ΑΒΔ") 1592 } 1593 }) 1594 } 1595 1596 var CountTests = []struct { 1597 s, sep string 1598 num int 1599 }{ 1600 {"", "", 1}, 1601 {"", "notempty", 0}, 1602 {"notempty", "", 9}, 1603 {"smaller", "not smaller", 0}, 1604 {"12345678987654321", "6", 2}, 1605 {"611161116", "6", 3}, 1606 {"notequal", "NotEqual", 0}, 1607 {"equal", "equal", 1}, 1608 {"abc1231231123q", "123", 3}, 1609 {"11111", "11", 2}, 1610 } 1611 1612 func TestCount(t *testing.T) { 1613 for _, tt := range CountTests { 1614 if num := Count(tt.s, tt.sep); num != tt.num { 1615 t.Errorf("Count(%q, %q) = %d, want %d", tt.s, tt.sep, num, tt.num) 1616 } 1617 } 1618 } 1619 1620 var cutTests = []struct { 1621 s, sep string 1622 before, after string 1623 found bool 1624 }{ 1625 {"abc", "b", "a", "c", true}, 1626 {"abc", "a", "", "bc", true}, 1627 {"abc", "c", "ab", "", true}, 1628 {"abc", "abc", "", "", true}, 1629 {"abc", "", "", "abc", true}, 1630 {"abc", "d", "abc", "", false}, 1631 {"", "d", "", "", false}, 1632 {"", "", "", "", true}, 1633 } 1634 1635 func TestCut(t *testing.T) { 1636 for _, tt := range cutTests { 1637 if before, after, found := Cut(tt.s, tt.sep); before != tt.before || after != tt.after || found != tt.found { 1638 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) 1639 } 1640 } 1641 } 1642 1643 var cutPrefixTests = []struct { 1644 s, sep string 1645 after string 1646 found bool 1647 }{ 1648 {"abc", "a", "bc", true}, 1649 {"abc", "abc", "", true}, 1650 {"abc", "", "abc", true}, 1651 {"abc", "d", "abc", false}, 1652 {"", "d", "", false}, 1653 {"", "", "", true}, 1654 } 1655 1656 func TestCutPrefix(t *testing.T) { 1657 for _, tt := range cutPrefixTests { 1658 if after, found := CutPrefix(tt.s, tt.sep); after != tt.after || found != tt.found { 1659 t.Errorf("CutPrefix(%q, %q) = %q, %v, want %q, %v", tt.s, tt.sep, after, found, tt.after, tt.found) 1660 } 1661 } 1662 } 1663 1664 var cutSuffixTests = []struct { 1665 s, sep string 1666 after string 1667 found bool 1668 }{ 1669 {"abc", "bc", "a", true}, 1670 {"abc", "abc", "", true}, 1671 {"abc", "", "abc", true}, 1672 {"abc", "d", "abc", false}, 1673 {"", "d", "", false}, 1674 {"", "", "", true}, 1675 } 1676 1677 func TestCutSuffix(t *testing.T) { 1678 for _, tt := range cutSuffixTests { 1679 if after, found := CutSuffix(tt.s, tt.sep); after != tt.after || found != tt.found { 1680 t.Errorf("CutSuffix(%q, %q) = %q, %v, want %q, %v", tt.s, tt.sep, after, found, tt.after, tt.found) 1681 } 1682 } 1683 } 1684 1685 func makeBenchInputHard() string { 1686 tokens := [...]string{ 1687 "<a>", "<p>", "<b>", "<strong>", 1688 "</a>", "</p>", "</b>", "</strong>", 1689 "hello", "world", 1690 } 1691 x := make([]byte, 0, 1<<20) 1692 for { 1693 i := rand.Intn(len(tokens)) 1694 if len(x)+len(tokens[i]) >= 1<<20 { 1695 break 1696 } 1697 x = append(x, tokens[i]...) 1698 } 1699 return string(x) 1700 } 1701 1702 var benchInputHard = makeBenchInputHard() 1703 1704 func benchmarkIndexHard(b *testing.B, sep string) { 1705 for i := 0; i < b.N; i++ { 1706 Index(benchInputHard, sep) 1707 } 1708 } 1709 1710 func benchmarkLastIndexHard(b *testing.B, sep string) { 1711 for i := 0; i < b.N; i++ { 1712 LastIndex(benchInputHard, sep) 1713 } 1714 } 1715 1716 func benchmarkCountHard(b *testing.B, sep string) { 1717 for i := 0; i < b.N; i++ { 1718 Count(benchInputHard, sep) 1719 } 1720 } 1721 1722 func BenchmarkIndexHard1(b *testing.B) { benchmarkIndexHard(b, "<>") } 1723 func BenchmarkIndexHard2(b *testing.B) { benchmarkIndexHard(b, "</pre>") } 1724 func BenchmarkIndexHard3(b *testing.B) { benchmarkIndexHard(b, "<b>hello world</b>") } 1725 func BenchmarkIndexHard4(b *testing.B) { 1726 benchmarkIndexHard(b, "<pre><b>hello</b><strong>world</strong></pre>") 1727 } 1728 1729 func BenchmarkLastIndexHard1(b *testing.B) { benchmarkLastIndexHard(b, "<>") } 1730 func BenchmarkLastIndexHard2(b *testing.B) { benchmarkLastIndexHard(b, "</pre>") } 1731 func BenchmarkLastIndexHard3(b *testing.B) { benchmarkLastIndexHard(b, "<b>hello world</b>") } 1732 1733 func BenchmarkCountHard1(b *testing.B) { benchmarkCountHard(b, "<>") } 1734 func BenchmarkCountHard2(b *testing.B) { benchmarkCountHard(b, "</pre>") } 1735 func BenchmarkCountHard3(b *testing.B) { benchmarkCountHard(b, "<b>hello world</b>") } 1736 1737 var benchInputTorture = Repeat("ABC", 1<<10) + "123" + Repeat("ABC", 1<<10) 1738 var benchNeedleTorture = Repeat("ABC", 1<<10+1) 1739 1740 func BenchmarkIndexTorture(b *testing.B) { 1741 for i := 0; i < b.N; i++ { 1742 Index(benchInputTorture, benchNeedleTorture) 1743 } 1744 } 1745 1746 func BenchmarkCountTorture(b *testing.B) { 1747 for i := 0; i < b.N; i++ { 1748 Count(benchInputTorture, benchNeedleTorture) 1749 } 1750 } 1751 1752 func BenchmarkCountTortureOverlapping(b *testing.B) { 1753 A := Repeat("ABC", 1<<20) 1754 B := Repeat("ABC", 1<<10) 1755 for i := 0; i < b.N; i++ { 1756 Count(A, B) 1757 } 1758 } 1759 1760 func BenchmarkCountByte(b *testing.B) { 1761 indexSizes := []int{10, 32, 4 << 10, 4 << 20, 64 << 20} 1762 benchStr := Repeat(benchmarkString, 1763 (indexSizes[len(indexSizes)-1]+len(benchmarkString)-1)/len(benchmarkString)) 1764 benchFunc := func(b *testing.B, benchStr string) { 1765 b.SetBytes(int64(len(benchStr))) 1766 for i := 0; i < b.N; i++ { 1767 Count(benchStr, "=") 1768 } 1769 } 1770 for _, size := range indexSizes { 1771 b.Run(fmt.Sprintf("%d", size), func(b *testing.B) { 1772 benchFunc(b, benchStr[:size]) 1773 }) 1774 } 1775 1776 } 1777 1778 var makeFieldsInput = func() string { 1779 x := make([]byte, 1<<20) 1780 // Input is ~10% space, ~10% 2-byte UTF-8, rest ASCII non-space. 1781 for i := range x { 1782 switch rand.Intn(10) { 1783 case 0: 1784 x[i] = ' ' 1785 case 1: 1786 if i > 0 && x[i-1] == 'x' { 1787 copy(x[i-1:], "χ") 1788 break 1789 } 1790 fallthrough 1791 default: 1792 x[i] = 'x' 1793 } 1794 } 1795 return string(x) 1796 } 1797 1798 var makeFieldsInputASCII = func() string { 1799 x := make([]byte, 1<<20) 1800 // Input is ~10% space, rest ASCII non-space. 1801 for i := range x { 1802 if rand.Intn(10) == 0 { 1803 x[i] = ' ' 1804 } else { 1805 x[i] = 'x' 1806 } 1807 } 1808 return string(x) 1809 } 1810 1811 var stringdata = []struct{ name, data string }{ 1812 {"ASCII", makeFieldsInputASCII()}, 1813 {"Mixed", makeFieldsInput()}, 1814 } 1815 1816 func BenchmarkFields(b *testing.B) { 1817 for _, sd := range stringdata { 1818 b.Run(sd.name, func(b *testing.B) { 1819 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1820 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1821 b.ReportAllocs() 1822 b.SetBytes(int64(j)) 1823 data := sd.data[:j] 1824 for i := 0; i < b.N; i++ { 1825 Fields(data) 1826 } 1827 }) 1828 } 1829 }) 1830 } 1831 } 1832 1833 func BenchmarkFieldsFunc(b *testing.B) { 1834 for _, sd := range stringdata { 1835 b.Run(sd.name, func(b *testing.B) { 1836 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1837 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1838 b.ReportAllocs() 1839 b.SetBytes(int64(j)) 1840 data := sd.data[:j] 1841 for i := 0; i < b.N; i++ { 1842 FieldsFunc(data, unicode.IsSpace) 1843 } 1844 }) 1845 } 1846 }) 1847 } 1848 } 1849 1850 func BenchmarkSplitEmptySeparator(b *testing.B) { 1851 for i := 0; i < b.N; i++ { 1852 Split(benchInputHard, "") 1853 } 1854 } 1855 1856 func BenchmarkSplitSingleByteSeparator(b *testing.B) { 1857 for i := 0; i < b.N; i++ { 1858 Split(benchInputHard, "/") 1859 } 1860 } 1861 1862 func BenchmarkSplitMultiByteSeparator(b *testing.B) { 1863 for i := 0; i < b.N; i++ { 1864 Split(benchInputHard, "hello") 1865 } 1866 } 1867 1868 func BenchmarkSplitNSingleByteSeparator(b *testing.B) { 1869 for i := 0; i < b.N; i++ { 1870 SplitN(benchInputHard, "/", 10) 1871 } 1872 } 1873 1874 func BenchmarkSplitNMultiByteSeparator(b *testing.B) { 1875 for i := 0; i < b.N; i++ { 1876 SplitN(benchInputHard, "hello", 10) 1877 } 1878 } 1879 1880 func BenchmarkRepeat(b *testing.B) { 1881 s := "0123456789" 1882 for _, n := range []int{5, 10} { 1883 for _, c := range []int{0, 1, 2, 6} { 1884 b.Run(fmt.Sprintf("%dx%d", n, c), func(b *testing.B) { 1885 for i := 0; i < b.N; i++ { 1886 Repeat(s[:n], c) 1887 } 1888 }) 1889 } 1890 } 1891 } 1892 1893 func BenchmarkRepeatLarge(b *testing.B) { 1894 s := Repeat("@", 8*1024) 1895 for j := 8; j <= 30; j++ { 1896 for _, k := range []int{1, 16, 4097} { 1897 s := s[:k] 1898 n := (1 << j) / k 1899 if n == 0 { 1900 continue 1901 } 1902 b.Run(fmt.Sprintf("%d/%d", 1<<j, k), func(b *testing.B) { 1903 for i := 0; i < b.N; i++ { 1904 Repeat(s, n) 1905 } 1906 b.SetBytes(int64(n * len(s))) 1907 }) 1908 } 1909 } 1910 } 1911 1912 func BenchmarkIndexAnyASCII(b *testing.B) { 1913 x := Repeat("#", 2048) // Never matches set 1914 cs := "0123456789abcdefghijklmnopqrstuvwxyz0123456789abcdefghijklmnopqrstuvwxyz" 1915 for k := 1; k <= 2048; k <<= 4 { 1916 for j := 1; j <= 64; j <<= 1 { 1917 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1918 for i := 0; i < b.N; i++ { 1919 IndexAny(x[:k], cs[:j]) 1920 } 1921 }) 1922 } 1923 } 1924 } 1925 1926 func BenchmarkIndexAnyUTF8(b *testing.B) { 1927 x := Repeat("#", 2048) // Never matches set 1928 cs := "你好世界, hello world. 你好世界, hello world. 你好世界, hello world." 1929 for k := 1; k <= 2048; k <<= 4 { 1930 for j := 1; j <= 64; j <<= 1 { 1931 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1932 for i := 0; i < b.N; i++ { 1933 IndexAny(x[:k], cs[:j]) 1934 } 1935 }) 1936 } 1937 } 1938 } 1939 1940 func BenchmarkLastIndexAnyASCII(b *testing.B) { 1941 x := Repeat("#", 2048) // Never matches set 1942 cs := "0123456789abcdefghijklmnopqrstuvwxyz0123456789abcdefghijklmnopqrstuvwxyz" 1943 for k := 1; k <= 2048; k <<= 4 { 1944 for j := 1; j <= 64; j <<= 1 { 1945 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1946 for i := 0; i < b.N; i++ { 1947 LastIndexAny(x[:k], cs[:j]) 1948 } 1949 }) 1950 } 1951 } 1952 } 1953 1954 func BenchmarkLastIndexAnyUTF8(b *testing.B) { 1955 x := Repeat("#", 2048) // Never matches set 1956 cs := "你好世界, hello world. 你好世界, hello world. 你好世界, hello world." 1957 for k := 1; k <= 2048; k <<= 4 { 1958 for j := 1; j <= 64; j <<= 1 { 1959 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1960 for i := 0; i < b.N; i++ { 1961 LastIndexAny(x[:k], cs[:j]) 1962 } 1963 }) 1964 } 1965 } 1966 } 1967 1968 func BenchmarkTrimASCII(b *testing.B) { 1969 cs := "0123456789abcdef" 1970 for k := 1; k <= 4096; k <<= 4 { 1971 for j := 1; j <= 16; j <<= 1 { 1972 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1973 x := Repeat(cs[:j], k) // Always matches set 1974 for i := 0; i < b.N; i++ { 1975 Trim(x[:k], cs[:j]) 1976 } 1977 }) 1978 } 1979 } 1980 } 1981 1982 func BenchmarkTrimByte(b *testing.B) { 1983 x := " the quick brown fox " 1984 for i := 0; i < b.N; i++ { 1985 Trim(x, " ") 1986 } 1987 } 1988 1989 func BenchmarkIndexPeriodic(b *testing.B) { 1990 key := "aa" 1991 for _, skip := range [...]int{2, 4, 8, 16, 32, 64} { 1992 b.Run(fmt.Sprintf("IndexPeriodic%d", skip), func(b *testing.B) { 1993 s := Repeat("a"+Repeat(" ", skip-1), 1<<16/skip) 1994 for i := 0; i < b.N; i++ { 1995 Index(s, key) 1996 } 1997 }) 1998 } 1999 } 2000 2001 func BenchmarkJoin(b *testing.B) { 2002 vals := []string{"red", "yellow", "pink", "green", "purple", "orange", "blue"} 2003 for l := 0; l <= len(vals); l++ { 2004 b.Run(strconv.Itoa(l), func(b *testing.B) { 2005 b.ReportAllocs() 2006 vals := vals[:l] 2007 for i := 0; i < b.N; i++ { 2008 Join(vals, " and ") 2009 } 2010 }) 2011 } 2012 } 2013 2014 func BenchmarkTrimSpace(b *testing.B) { 2015 tests := []struct{ name, input string }{ 2016 {"NoTrim", "typical"}, 2017 {"ASCII", " foo bar "}, 2018 {"SomeNonASCII", " \u2000\t\r\n x\t\t\r\r\ny\n \u3000 "}, 2019 {"JustNonASCII", "\u2000\u2000\u2000☺☺☺☺\u3000\u3000\u3000"}, 2020 } 2021 for _, test := range tests { 2022 b.Run(test.name, func(b *testing.B) { 2023 for i := 0; i < b.N; i++ { 2024 TrimSpace(test.input) 2025 } 2026 }) 2027 } 2028 } 2029 2030 var stringSink string 2031 2032 func BenchmarkReplaceAll(b *testing.B) { 2033 b.ReportAllocs() 2034 for i := 0; i < b.N; i++ { 2035 stringSink = ReplaceAll("banana", "a", "<>") 2036 } 2037 }