github.com/tidwall/go@v0.0.0-20170415222209-6694a6888b7d/src/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/rand" 12 "reflect" 13 . "strings" 14 "testing" 15 "unicode" 16 "unicode/utf8" 17 "unsafe" 18 ) 19 20 func eq(a, b []string) bool { 21 if len(a) != len(b) { 22 return false 23 } 24 for i := 0; i < len(a); i++ { 25 if a[i] != b[i] { 26 return false 27 } 28 } 29 return true 30 } 31 32 var abcd = "abcd" 33 var faces = "☺☻☹" 34 var commas = "1,2,3,4" 35 var dots = "1....2....3....4" 36 37 type IndexTest struct { 38 s string 39 sep string 40 out int 41 } 42 43 var indexTests = []IndexTest{ 44 {"", "", 0}, 45 {"", "a", -1}, 46 {"", "foo", -1}, 47 {"fo", "foo", -1}, 48 {"foo", "foo", 0}, 49 {"oofofoofooo", "f", 2}, 50 {"oofofoofooo", "foo", 4}, 51 {"barfoobarfoo", "foo", 3}, 52 {"foo", "", 0}, 53 {"foo", "o", 1}, 54 {"abcABCabc", "A", 3}, 55 // cases with one byte strings - test special case in Index() 56 {"", "a", -1}, 57 {"x", "a", -1}, 58 {"x", "x", 0}, 59 {"abc", "a", 0}, 60 {"abc", "b", 1}, 61 {"abc", "c", 2}, 62 {"abc", "x", -1}, 63 // test special cases in Index() for short strings 64 {"", "ab", -1}, 65 {"bc", "ab", -1}, 66 {"ab", "ab", 0}, 67 {"xab", "ab", 1}, 68 {"xab"[:2], "ab", -1}, 69 {"", "abc", -1}, 70 {"xbc", "abc", -1}, 71 {"abc", "abc", 0}, 72 {"xabc", "abc", 1}, 73 {"xabc"[:3], "abc", -1}, 74 {"xabxc", "abc", -1}, 75 {"", "abcd", -1}, 76 {"xbcd", "abcd", -1}, 77 {"abcd", "abcd", 0}, 78 {"xabcd", "abcd", 1}, 79 {"xyabcd"[:5], "abcd", -1}, 80 {"xbcqq", "abcqq", -1}, 81 {"abcqq", "abcqq", 0}, 82 {"xabcqq", "abcqq", 1}, 83 {"xyabcqq"[:6], "abcqq", -1}, 84 {"xabxcqq", "abcqq", -1}, 85 {"xabcqxq", "abcqq", -1}, 86 {"", "01234567", -1}, 87 {"32145678", "01234567", -1}, 88 {"01234567", "01234567", 0}, 89 {"x01234567", "01234567", 1}, 90 {"x0123456x01234567", "01234567", 9}, 91 {"xx01234567"[:9], "01234567", -1}, 92 {"", "0123456789", -1}, 93 {"3214567844", "0123456789", -1}, 94 {"0123456789", "0123456789", 0}, 95 {"x0123456789", "0123456789", 1}, 96 {"x012345678x0123456789", "0123456789", 11}, 97 {"xyz0123456789"[:12], "0123456789", -1}, 98 {"x01234567x89", "0123456789", -1}, 99 {"", "0123456789012345", -1}, 100 {"3214567889012345", "0123456789012345", -1}, 101 {"0123456789012345", "0123456789012345", 0}, 102 {"x0123456789012345", "0123456789012345", 1}, 103 {"x012345678901234x0123456789012345", "0123456789012345", 17}, 104 {"", "01234567890123456789", -1}, 105 {"32145678890123456789", "01234567890123456789", -1}, 106 {"01234567890123456789", "01234567890123456789", 0}, 107 {"x01234567890123456789", "01234567890123456789", 1}, 108 {"x0123456789012345678x01234567890123456789", "01234567890123456789", 21}, 109 {"xyz01234567890123456789"[:22], "01234567890123456789", -1}, 110 {"", "0123456789012345678901234567890", -1}, 111 {"321456788901234567890123456789012345678911", "0123456789012345678901234567890", -1}, 112 {"0123456789012345678901234567890", "0123456789012345678901234567890", 0}, 113 {"x0123456789012345678901234567890", "0123456789012345678901234567890", 1}, 114 {"x012345678901234567890123456789x0123456789012345678901234567890", "0123456789012345678901234567890", 32}, 115 {"xyz0123456789012345678901234567890"[:33], "0123456789012345678901234567890", -1}, 116 {"", "01234567890123456789012345678901", -1}, 117 {"32145678890123456789012345678901234567890211", "01234567890123456789012345678901", -1}, 118 {"01234567890123456789012345678901", "01234567890123456789012345678901", 0}, 119 {"x01234567890123456789012345678901", "01234567890123456789012345678901", 1}, 120 {"x0123456789012345678901234567890x01234567890123456789012345678901", "01234567890123456789012345678901", 33}, 121 {"xyz01234567890123456789012345678901"[:34], "01234567890123456789012345678901", -1}, 122 {"xxxxxx012345678901234567890123456789012345678901234567890123456789012", "012345678901234567890123456789012345678901234567890123456789012", 6}, 123 {"", "0123456789012345678901234567890123456789", -1}, 124 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456789", 2}, 125 {"xx012345678901234567890123456789012345678901234567890123456789012"[:41], "0123456789012345678901234567890123456789", -1}, 126 {"xx012345678901234567890123456789012345678901234567890123456789012", "0123456789012345678901234567890123456xxx", -1}, 127 {"xx0123456789012345678901234567890123456789012345678901234567890120123456789012345678901234567890123456xxx", "0123456789012345678901234567890123456xxx", 65}, 128 } 129 130 var lastIndexTests = []IndexTest{ 131 {"", "", 0}, 132 {"", "a", -1}, 133 {"", "foo", -1}, 134 {"fo", "foo", -1}, 135 {"foo", "foo", 0}, 136 {"foo", "f", 0}, 137 {"oofofoofooo", "f", 7}, 138 {"oofofoofooo", "foo", 7}, 139 {"barfoobarfoo", "foo", 9}, 140 {"foo", "", 3}, 141 {"foo", "o", 2}, 142 {"abcABCabc", "A", 3}, 143 {"abcABCabc", "a", 6}, 144 } 145 146 var indexAnyTests = []IndexTest{ 147 {"", "", -1}, 148 {"", "a", -1}, 149 {"", "abc", -1}, 150 {"a", "", -1}, 151 {"a", "a", 0}, 152 {"aaa", "a", 0}, 153 {"abc", "xyz", -1}, 154 {"abc", "xcz", 2}, 155 {"ab☺c", "x☺yz", 2}, 156 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 157 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 158 {"aRegExp*", ".(|)*+?^$[]", 7}, 159 {dots + dots + dots, " ", -1}, 160 {"012abcba210", "\xffb", 4}, 161 {"012\x80bcb\x80210", "\xffb", 3}, 162 } 163 164 var lastIndexAnyTests = []IndexTest{ 165 {"", "", -1}, 166 {"", "a", -1}, 167 {"", "abc", -1}, 168 {"a", "", -1}, 169 {"a", "a", 0}, 170 {"aaa", "a", 2}, 171 {"abc", "xyz", -1}, 172 {"abc", "ab", 1}, 173 {"ab☺c", "x☺yz", 2}, 174 {"a☺b☻c☹d", "cx", len("a☺b☻")}, 175 {"a☺b☻c☹d", "uvw☻xyz", len("a☺b")}, 176 {"a.RegExp*", ".(|)*+?^$[]", 8}, 177 {dots + dots + dots, " ", -1}, 178 {"012abcba210", "\xffb", 6}, 179 {"012\x80bcb\x80210", "\xffb", 7}, 180 } 181 182 // Execute f on each test case. funcName should be the name of f; it's used 183 // in failure reports. 184 func runIndexTests(t *testing.T, f func(s, sep string) int, funcName string, testCases []IndexTest) { 185 for _, test := range testCases { 186 actual := f(test.s, test.sep) 187 if actual != test.out { 188 t.Errorf("%s(%q,%q) = %v; want %v", funcName, test.s, test.sep, actual, test.out) 189 } 190 } 191 } 192 193 func TestIndex(t *testing.T) { runIndexTests(t, Index, "Index", indexTests) } 194 func TestLastIndex(t *testing.T) { runIndexTests(t, LastIndex, "LastIndex", lastIndexTests) } 195 func TestIndexAny(t *testing.T) { runIndexTests(t, IndexAny, "IndexAny", indexAnyTests) } 196 func TestLastIndexAny(t *testing.T) { runIndexTests(t, LastIndexAny, "LastIndexAny", lastIndexAnyTests) } 197 198 func TestLastIndexByte(t *testing.T) { 199 testCases := []IndexTest{ 200 {"", "q", -1}, 201 {"abcdef", "q", -1}, 202 {"abcdefabcdef", "a", len("abcdef")}, // something in the middle 203 {"abcdefabcdef", "f", len("abcdefabcde")}, // last byte 204 {"zabcdefabcdef", "z", 0}, // first byte 205 {"a☺b☻c☹d", "b", len("a☺")}, // non-ascii 206 } 207 for _, test := range testCases { 208 actual := LastIndexByte(test.s, test.sep[0]) 209 if actual != test.out { 210 t.Errorf("LastIndexByte(%q,%c) = %v; want %v", test.s, test.sep[0], actual, test.out) 211 } 212 } 213 } 214 215 func simpleIndex(s, sep string) int { 216 n := len(sep) 217 for i := n; i <= len(s); i++ { 218 if s[i-n:i] == sep { 219 return i - n 220 } 221 } 222 return -1 223 } 224 225 func TestIndexRandom(t *testing.T) { 226 const chars = "abcdefghijklmnopqrstuvwxyz0123456789" 227 for times := 0; times < 10; times++ { 228 for strLen := 5 + rand.Intn(5); strLen < 140; strLen += 10 { // Arbitrary 229 s1 := make([]byte, strLen) 230 for i := range s1 { 231 s1[i] = chars[rand.Intn(len(chars))] 232 } 233 s := string(s1) 234 for i := 0; i < 50; i++ { 235 begin := rand.Intn(len(s) + 1) 236 end := begin + rand.Intn(len(s)+1-begin) 237 sep := s[begin:end] 238 if i%4 == 0 { 239 pos := rand.Intn(len(sep) + 1) 240 sep = sep[:pos] + "A" + sep[pos:] 241 } 242 want := simpleIndex(s, sep) 243 res := Index(s, sep) 244 if res != want { 245 t.Errorf("Index(%s,%s) = %d; want %d", s, sep, res, want) 246 } 247 } 248 } 249 } 250 } 251 252 func TestIndexRune(t *testing.T) { 253 tests := []struct { 254 in string 255 rune rune 256 want int 257 }{ 258 {"", 'a', -1}, 259 {"", '☺', -1}, 260 {"foo", '☹', -1}, 261 {"foo", 'o', 1}, 262 {"foo☺bar", '☺', 3}, 263 {"foo☺☻☹bar", '☹', 9}, 264 {"a A x", 'A', 2}, 265 {"some_text=some_value", '=', 9}, 266 {"☺a", 'a', 3}, 267 {"a☻☺b", '☺', 4}, 268 269 // RuneError should match any invalid UTF-8 byte sequence. 270 {"�", '�', 0}, 271 {"\xff", '�', 0}, 272 {"☻x�", '�', len("☻x")}, 273 {"☻x\xe2\x98", '�', len("☻x")}, 274 {"☻x\xe2\x98�", '�', len("☻x")}, 275 {"☻x\xe2\x98x", '�', len("☻x")}, 276 277 // Invalid rune values should never match. 278 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", -1, -1}, 279 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", 0xD800, -1}, // Surrogate pair 280 {"a☺b☻c☹d\xe2\x98�\xff�\xed\xa0\x80", utf8.MaxRune + 1, -1}, 281 } 282 for _, tt := range tests { 283 if got := IndexRune(tt.in, tt.rune); got != tt.want { 284 t.Errorf("IndexRune(%q, %d) = %v; want %v", tt.in, tt.rune, got, tt.want) 285 } 286 } 287 288 haystack := "test世界" 289 allocs := testing.AllocsPerRun(1000, func() { 290 if i := IndexRune(haystack, 's'); i != 2 { 291 t.Fatalf("'s' at %d; want 2", i) 292 } 293 if i := IndexRune(haystack, '世'); i != 4 { 294 t.Fatalf("'世' at %d; want 4", i) 295 } 296 }) 297 if allocs != 0 && testing.CoverMode() == "" { 298 t.Errorf("expected no allocations, got %f", allocs) 299 } 300 } 301 302 const benchmarkString = "some_text=some☺value" 303 304 func BenchmarkIndexRune(b *testing.B) { 305 if got := IndexRune(benchmarkString, '☺'); got != 14 { 306 b.Fatalf("wrong index: expected 14, got=%d", got) 307 } 308 for i := 0; i < b.N; i++ { 309 IndexRune(benchmarkString, '☺') 310 } 311 } 312 313 var benchmarkLongString = Repeat(" ", 100) + benchmarkString 314 315 func BenchmarkIndexRuneLongString(b *testing.B) { 316 if got := IndexRune(benchmarkLongString, '☺'); got != 114 { 317 b.Fatalf("wrong index: expected 114, got=%d", got) 318 } 319 for i := 0; i < b.N; i++ { 320 IndexRune(benchmarkLongString, '☺') 321 } 322 } 323 324 func BenchmarkIndexRuneFastPath(b *testing.B) { 325 if got := IndexRune(benchmarkString, 'v'); got != 17 { 326 b.Fatalf("wrong index: expected 17, got=%d", got) 327 } 328 for i := 0; i < b.N; i++ { 329 IndexRune(benchmarkString, 'v') 330 } 331 } 332 333 func BenchmarkIndex(b *testing.B) { 334 if got := Index(benchmarkString, "v"); got != 17 { 335 b.Fatalf("wrong index: expected 17, got=%d", got) 336 } 337 for i := 0; i < b.N; i++ { 338 Index(benchmarkString, "v") 339 } 340 } 341 342 func BenchmarkLastIndex(b *testing.B) { 343 if got := Index(benchmarkString, "v"); got != 17 { 344 b.Fatalf("wrong index: expected 17, got=%d", got) 345 } 346 for i := 0; i < b.N; i++ { 347 LastIndex(benchmarkString, "v") 348 } 349 } 350 351 func BenchmarkIndexByte(b *testing.B) { 352 if got := IndexByte(benchmarkString, 'v'); got != 17 { 353 b.Fatalf("wrong index: expected 17, got=%d", got) 354 } 355 for i := 0; i < b.N; i++ { 356 IndexByte(benchmarkString, 'v') 357 } 358 } 359 360 type SplitTest struct { 361 s string 362 sep string 363 n int 364 a []string 365 } 366 367 var splittests = []SplitTest{ 368 {"", "", -1, []string{}}, 369 {abcd, "", 2, []string{"a", "bcd"}}, 370 {abcd, "", 4, []string{"a", "b", "c", "d"}}, 371 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 372 {faces, "", -1, []string{"☺", "☻", "☹"}}, 373 {faces, "", 3, []string{"☺", "☻", "☹"}}, 374 {faces, "", 17, []string{"☺", "☻", "☹"}}, 375 {"☺�☹", "", -1, []string{"☺", "�", "☹"}}, 376 {abcd, "a", 0, nil}, 377 {abcd, "a", -1, []string{"", "bcd"}}, 378 {abcd, "z", -1, []string{"abcd"}}, 379 {commas, ",", -1, []string{"1", "2", "3", "4"}}, 380 {dots, "...", -1, []string{"1", ".2", ".3", ".4"}}, 381 {faces, "☹", -1, []string{"☺☻", ""}}, 382 {faces, "~", -1, []string{faces}}, 383 {"1 2 3 4", " ", 3, []string{"1", "2", "3 4"}}, 384 {"1 2", " ", 3, []string{"1", "2"}}, 385 } 386 387 func TestSplit(t *testing.T) { 388 for _, tt := range splittests { 389 a := SplitN(tt.s, tt.sep, tt.n) 390 if !eq(a, tt.a) { 391 t.Errorf("Split(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, a, tt.a) 392 continue 393 } 394 if tt.n == 0 { 395 continue 396 } 397 s := Join(a, tt.sep) 398 if s != tt.s { 399 t.Errorf("Join(Split(%q, %q, %d), %q) = %q", tt.s, tt.sep, tt.n, tt.sep, s) 400 } 401 if tt.n < 0 { 402 b := Split(tt.s, tt.sep) 403 if !reflect.DeepEqual(a, b) { 404 t.Errorf("Split disagrees with SplitN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 405 } 406 } 407 } 408 } 409 410 var splitaftertests = []SplitTest{ 411 {abcd, "a", -1, []string{"a", "bcd"}}, 412 {abcd, "z", -1, []string{"abcd"}}, 413 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 414 {commas, ",", -1, []string{"1,", "2,", "3,", "4"}}, 415 {dots, "...", -1, []string{"1...", ".2...", ".3...", ".4"}}, 416 {faces, "☹", -1, []string{"☺☻☹", ""}}, 417 {faces, "~", -1, []string{faces}}, 418 {faces, "", -1, []string{"☺", "☻", "☹"}}, 419 {"1 2 3 4", " ", 3, []string{"1 ", "2 ", "3 4"}}, 420 {"1 2 3", " ", 3, []string{"1 ", "2 ", "3"}}, 421 {"1 2", " ", 3, []string{"1 ", "2"}}, 422 {"123", "", 2, []string{"1", "23"}}, 423 {"123", "", 17, []string{"1", "2", "3"}}, 424 } 425 426 func TestSplitAfter(t *testing.T) { 427 for _, tt := range splitaftertests { 428 a := SplitAfterN(tt.s, tt.sep, tt.n) 429 if !eq(a, tt.a) { 430 t.Errorf(`Split(%q, %q, %d) = %v; want %v`, tt.s, tt.sep, tt.n, a, tt.a) 431 continue 432 } 433 s := Join(a, "") 434 if s != tt.s { 435 t.Errorf(`Join(Split(%q, %q, %d), %q) = %q`, tt.s, tt.sep, tt.n, tt.sep, s) 436 } 437 if tt.n < 0 { 438 b := SplitAfter(tt.s, tt.sep) 439 if !reflect.DeepEqual(a, b) { 440 t.Errorf("SplitAfter disagrees with SplitAfterN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 441 } 442 } 443 } 444 } 445 446 type FieldsTest struct { 447 s string 448 a []string 449 } 450 451 var fieldstests = []FieldsTest{ 452 {"", []string{}}, 453 {" ", []string{}}, 454 {" \t ", []string{}}, 455 {"\u2000", []string{}}, 456 {" abc ", []string{"abc"}}, 457 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 458 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 459 {"1\t\t2\t\t3\t4", []string{"1", "2", "3", "4"}}, 460 {"1\u20002\u20013\u20024", []string{"1", "2", "3", "4"}}, 461 {"\u2000\u2001\u2002", []string{}}, 462 {"\n™\t™\n", []string{"™", "™"}}, 463 {"\n\u20001™2\u2000 \u2001 ™", []string{"1™2", "™"}}, 464 {"\n1\uFFFD \uFFFD2\u20003\uFFFD4", []string{"1\uFFFD", "\uFFFD2", "3\uFFFD4"}}, 465 {"1\xFF\u2000\xFF2\xFF \xFF", []string{"1\xFF", "\xFF2\xFF", "\xFF"}}, 466 {faces, []string{faces}}, 467 } 468 469 func TestFields(t *testing.T) { 470 for _, tt := range fieldstests { 471 a := Fields(tt.s) 472 if !eq(a, tt.a) { 473 t.Errorf("Fields(%q) = %v; want %v", tt.s, a, tt.a) 474 continue 475 } 476 } 477 } 478 479 var FieldsFuncTests = []FieldsTest{ 480 {"", []string{}}, 481 {"XX", []string{}}, 482 {"XXhiXXX", []string{"hi"}}, 483 {"aXXbXXXcX", []string{"a", "b", "c"}}, 484 } 485 486 func TestFieldsFunc(t *testing.T) { 487 for _, tt := range fieldstests { 488 a := FieldsFunc(tt.s, unicode.IsSpace) 489 if !eq(a, tt.a) { 490 t.Errorf("FieldsFunc(%q, unicode.IsSpace) = %v; want %v", tt.s, a, tt.a) 491 continue 492 } 493 } 494 pred := func(c rune) bool { return c == 'X' } 495 for _, tt := range FieldsFuncTests { 496 a := FieldsFunc(tt.s, pred) 497 if !eq(a, tt.a) { 498 t.Errorf("FieldsFunc(%q) = %v, want %v", tt.s, a, tt.a) 499 } 500 } 501 } 502 503 // Test case for any function which accepts and returns a single string. 504 type StringTest struct { 505 in, out string 506 } 507 508 // Execute f on each test case. funcName should be the name of f; it's used 509 // in failure reports. 510 func runStringTests(t *testing.T, f func(string) string, funcName string, testCases []StringTest) { 511 for _, tc := range testCases { 512 actual := f(tc.in) 513 if actual != tc.out { 514 t.Errorf("%s(%q) = %q; want %q", funcName, tc.in, actual, tc.out) 515 } 516 } 517 } 518 519 var upperTests = []StringTest{ 520 {"", ""}, 521 {"abc", "ABC"}, 522 {"AbC123", "ABC123"}, 523 {"azAZ09_", "AZAZ09_"}, 524 {"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"}, // grows one byte per char 525 } 526 527 var lowerTests = []StringTest{ 528 {"", ""}, 529 {"abc", "abc"}, 530 {"AbC123", "abc123"}, 531 {"azAZ09_", "azaz09_"}, 532 {"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"}, // shrinks one byte per char 533 } 534 535 const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000" 536 537 var trimSpaceTests = []StringTest{ 538 {"", ""}, 539 {"abc", "abc"}, 540 {space + "abc" + space, "abc"}, 541 {" ", ""}, 542 {" \t\r\n \t\t\r\r\n\n ", ""}, 543 {" \t\r\n x\t\t\r\r\n\n ", "x"}, 544 {" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, 545 {"1 \t\r\n2", "1 \t\r\n2"}, 546 {" x\x80", "x\x80"}, 547 {" x\xc0", "x\xc0"}, 548 {"x \xc0\xc0 ", "x \xc0\xc0"}, 549 {"x \xc0", "x \xc0"}, 550 {"x \xc0 ", "x \xc0"}, 551 {"x \xc0\xc0 ", "x \xc0\xc0"}, 552 {"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, 553 {"x ☺ ", "x ☺"}, 554 } 555 556 func tenRunes(ch rune) string { 557 r := make([]rune, 10) 558 for i := range r { 559 r[i] = ch 560 } 561 return string(r) 562 } 563 564 // User-defined self-inverse mapping function 565 func rot13(r rune) rune { 566 step := rune(13) 567 if r >= 'a' && r <= 'z' { 568 return ((r - 'a' + step) % 26) + 'a' 569 } 570 if r >= 'A' && r <= 'Z' { 571 return ((r - 'A' + step) % 26) + 'A' 572 } 573 return r 574 } 575 576 func TestMap(t *testing.T) { 577 // Run a couple of awful growth/shrinkage tests 578 a := tenRunes('a') 579 // 1. Grow. This triggers two reallocations in Map. 580 maxRune := func(rune) rune { return unicode.MaxRune } 581 m := Map(maxRune, a) 582 expect := tenRunes(unicode.MaxRune) 583 if m != expect { 584 t.Errorf("growing: expected %q got %q", expect, m) 585 } 586 587 // 2. Shrink 588 minRune := func(rune) rune { return 'a' } 589 m = Map(minRune, tenRunes(unicode.MaxRune)) 590 expect = a 591 if m != expect { 592 t.Errorf("shrinking: expected %q got %q", expect, m) 593 } 594 595 // 3. Rot13 596 m = Map(rot13, "a to zed") 597 expect = "n gb mrq" 598 if m != expect { 599 t.Errorf("rot13: expected %q got %q", expect, m) 600 } 601 602 // 4. Rot13^2 603 m = Map(rot13, Map(rot13, "a to zed")) 604 expect = "a to zed" 605 if m != expect { 606 t.Errorf("rot13: expected %q got %q", expect, m) 607 } 608 609 // 5. Drop 610 dropNotLatin := func(r rune) rune { 611 if unicode.Is(unicode.Latin, r) { 612 return r 613 } 614 return -1 615 } 616 m = Map(dropNotLatin, "Hello, 세계") 617 expect = "Hello" 618 if m != expect { 619 t.Errorf("drop: expected %q got %q", expect, m) 620 } 621 622 // 6. Identity 623 identity := func(r rune) rune { 624 return r 625 } 626 orig := "Input string that we expect not to be copied." 627 m = Map(identity, orig) 628 if (*reflect.StringHeader)(unsafe.Pointer(&orig)).Data != 629 (*reflect.StringHeader)(unsafe.Pointer(&m)).Data { 630 t.Error("unexpected copy during identity map") 631 } 632 633 // 7. Handle invalid UTF-8 sequence 634 replaceNotLatin := func(r rune) rune { 635 if unicode.Is(unicode.Latin, r) { 636 return r 637 } 638 return '?' 639 } 640 m = Map(replaceNotLatin, "Hello\255World") 641 expect = "Hello?World" 642 if m != expect { 643 t.Errorf("replace invalid sequence: expected %q got %q", expect, m) 644 } 645 } 646 647 func TestToUpper(t *testing.T) { runStringTests(t, ToUpper, "ToUpper", upperTests) } 648 649 func TestToLower(t *testing.T) { runStringTests(t, ToLower, "ToLower", lowerTests) } 650 651 func BenchmarkMapNoChanges(b *testing.B) { 652 identity := func(r rune) rune { 653 return r 654 } 655 for i := 0; i < b.N; i++ { 656 Map(identity, "Some string that won't be modified.") 657 } 658 } 659 660 func TestSpecialCase(t *testing.T) { 661 lower := "abcçdefgğhıijklmnoöprsştuüvyz" 662 upper := "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ" 663 u := ToUpperSpecial(unicode.TurkishCase, upper) 664 if u != upper { 665 t.Errorf("Upper(upper) is %s not %s", u, upper) 666 } 667 u = ToUpperSpecial(unicode.TurkishCase, lower) 668 if u != upper { 669 t.Errorf("Upper(lower) is %s not %s", u, upper) 670 } 671 l := ToLowerSpecial(unicode.TurkishCase, lower) 672 if l != lower { 673 t.Errorf("Lower(lower) is %s not %s", l, lower) 674 } 675 l = ToLowerSpecial(unicode.TurkishCase, upper) 676 if l != lower { 677 t.Errorf("Lower(upper) is %s not %s", l, lower) 678 } 679 } 680 681 func TestTrimSpace(t *testing.T) { runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests) } 682 683 var trimTests = []struct { 684 f string 685 in, arg, out string 686 }{ 687 {"Trim", "abba", "a", "bb"}, 688 {"Trim", "abba", "ab", ""}, 689 {"TrimLeft", "abba", "ab", ""}, 690 {"TrimRight", "abba", "ab", ""}, 691 {"TrimLeft", "abba", "a", "bba"}, 692 {"TrimRight", "abba", "a", "abb"}, 693 {"Trim", "<tag>", "<>", "tag"}, 694 {"Trim", "* listitem", " *", "listitem"}, 695 {"Trim", `"quote"`, `"`, "quote"}, 696 {"Trim", "\u2C6F\u2C6F\u0250\u0250\u2C6F\u2C6F", "\u2C6F", "\u0250\u0250"}, 697 {"Trim", "\x80test\xff", "\xff", "test"}, 698 {"Trim", " Ġ ", " ", "Ġ"}, 699 {"Trim", " Ġİ0", "0 ", "Ġİ"}, 700 //empty string tests 701 {"Trim", "abba", "", "abba"}, 702 {"Trim", "", "123", ""}, 703 {"Trim", "", "", ""}, 704 {"TrimLeft", "abba", "", "abba"}, 705 {"TrimLeft", "", "123", ""}, 706 {"TrimLeft", "", "", ""}, 707 {"TrimRight", "abba", "", "abba"}, 708 {"TrimRight", "", "123", ""}, 709 {"TrimRight", "", "", ""}, 710 {"TrimRight", "☺\xc0", "☺", "☺\xc0"}, 711 {"TrimPrefix", "aabb", "a", "abb"}, 712 {"TrimPrefix", "aabb", "b", "aabb"}, 713 {"TrimSuffix", "aabb", "a", "aabb"}, 714 {"TrimSuffix", "aabb", "b", "aab"}, 715 } 716 717 func TestTrim(t *testing.T) { 718 for _, tc := range trimTests { 719 name := tc.f 720 var f func(string, string) string 721 switch name { 722 case "Trim": 723 f = Trim 724 case "TrimLeft": 725 f = TrimLeft 726 case "TrimRight": 727 f = TrimRight 728 case "TrimPrefix": 729 f = TrimPrefix 730 case "TrimSuffix": 731 f = TrimSuffix 732 default: 733 t.Errorf("Undefined trim function %s", name) 734 } 735 actual := f(tc.in, tc.arg) 736 if actual != tc.out { 737 t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 738 } 739 } 740 } 741 742 func BenchmarkTrim(b *testing.B) { 743 b.ReportAllocs() 744 745 for i := 0; i < b.N; i++ { 746 for _, tc := range trimTests { 747 name := tc.f 748 var f func(string, string) string 749 switch name { 750 case "Trim": 751 f = Trim 752 case "TrimLeft": 753 f = TrimLeft 754 case "TrimRight": 755 f = TrimRight 756 case "TrimPrefix": 757 f = TrimPrefix 758 case "TrimSuffix": 759 f = TrimSuffix 760 default: 761 b.Errorf("Undefined trim function %s", name) 762 } 763 actual := f(tc.in, tc.arg) 764 if actual != tc.out { 765 b.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 766 } 767 } 768 } 769 } 770 771 type predicate struct { 772 f func(rune) bool 773 name string 774 } 775 776 var isSpace = predicate{unicode.IsSpace, "IsSpace"} 777 var isDigit = predicate{unicode.IsDigit, "IsDigit"} 778 var isUpper = predicate{unicode.IsUpper, "IsUpper"} 779 var isValidRune = predicate{ 780 func(r rune) bool { 781 return r != utf8.RuneError 782 }, 783 "IsValidRune", 784 } 785 786 func not(p predicate) predicate { 787 return predicate{ 788 func(r rune) bool { 789 return !p.f(r) 790 }, 791 "not " + p.name, 792 } 793 } 794 795 var trimFuncTests = []struct { 796 f predicate 797 in, out string 798 }{ 799 {isSpace, space + " hello " + space, "hello"}, 800 {isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", "hello"}, 801 {isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", "hello"}, 802 {not(isSpace), "hello" + space + "hello", space}, 803 {not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", "\u0e50\u0e521234\u0e50\u0e51"}, 804 {isValidRune, "ab\xc0a\xc0cd", "\xc0a\xc0"}, 805 {not(isValidRune), "\xc0a\xc0", "a"}, 806 } 807 808 func TestTrimFunc(t *testing.T) { 809 for _, tc := range trimFuncTests { 810 actual := TrimFunc(tc.in, tc.f.f) 811 if actual != tc.out { 812 t.Errorf("TrimFunc(%q, %q) = %q; want %q", tc.in, tc.f.name, actual, tc.out) 813 } 814 } 815 } 816 817 var indexFuncTests = []struct { 818 in string 819 f predicate 820 first, last int 821 }{ 822 {"", isValidRune, -1, -1}, 823 {"abc", isDigit, -1, -1}, 824 {"0123", isDigit, 0, 3}, 825 {"a1b", isDigit, 1, 1}, 826 {space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes 827 {"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, 828 {"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, 829 {"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, 830 831 // tests of invalid UTF-8 832 {"\x801", isDigit, 1, 1}, 833 {"\x80abc", isDigit, -1, -1}, 834 {"\xc0a\xc0", isValidRune, 1, 1}, 835 {"\xc0a\xc0", not(isValidRune), 0, 2}, 836 {"\xc0☺\xc0", not(isValidRune), 0, 4}, 837 {"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, 838 {"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, 839 {"a\xe0\x80cd", not(isValidRune), 1, 2}, 840 {"\x80\x80\x80\x80", not(isValidRune), 0, 3}, 841 } 842 843 func TestIndexFunc(t *testing.T) { 844 for _, tc := range indexFuncTests { 845 first := IndexFunc(tc.in, tc.f.f) 846 if first != tc.first { 847 t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) 848 } 849 last := LastIndexFunc(tc.in, tc.f.f) 850 if last != tc.last { 851 t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) 852 } 853 } 854 } 855 856 func equal(m string, s1, s2 string, t *testing.T) bool { 857 if s1 == s2 { 858 return true 859 } 860 e1 := Split(s1, "") 861 e2 := Split(s2, "") 862 for i, c1 := range e1 { 863 if i >= len(e2) { 864 break 865 } 866 r1, _ := utf8.DecodeRuneInString(c1) 867 r2, _ := utf8.DecodeRuneInString(e2[i]) 868 if r1 != r2 { 869 t.Errorf("%s diff at %d: U+%04X U+%04X", m, i, r1, r2) 870 } 871 } 872 return false 873 } 874 875 func TestCaseConsistency(t *testing.T) { 876 // Make a string of all the runes. 877 numRunes := int(unicode.MaxRune + 1) 878 if testing.Short() { 879 numRunes = 1000 880 } 881 a := make([]rune, numRunes) 882 for i := range a { 883 a[i] = rune(i) 884 } 885 s := string(a) 886 // convert the cases. 887 upper := ToUpper(s) 888 lower := ToLower(s) 889 890 // Consistency checks 891 if n := utf8.RuneCountInString(upper); n != numRunes { 892 t.Error("rune count wrong in upper:", n) 893 } 894 if n := utf8.RuneCountInString(lower); n != numRunes { 895 t.Error("rune count wrong in lower:", n) 896 } 897 if !equal("ToUpper(upper)", ToUpper(upper), upper, t) { 898 t.Error("ToUpper(upper) consistency fail") 899 } 900 if !equal("ToLower(lower)", ToLower(lower), lower, t) { 901 t.Error("ToLower(lower) consistency fail") 902 } 903 /* 904 These fail because of non-one-to-oneness of the data, such as multiple 905 upper case 'I' mapping to 'i'. We comment them out but keep them for 906 interest. 907 For instance: CAPITAL LETTER I WITH DOT ABOVE: 908 unicode.ToUpper(unicode.ToLower('\u0130')) != '\u0130' 909 910 if !equal("ToUpper(lower)", ToUpper(lower), upper, t) { 911 t.Error("ToUpper(lower) consistency fail"); 912 } 913 if !equal("ToLower(upper)", ToLower(upper), lower, t) { 914 t.Error("ToLower(upper) consistency fail"); 915 } 916 */ 917 } 918 919 var RepeatTests = []struct { 920 in, out string 921 count int 922 }{ 923 {"", "", 0}, 924 {"", "", 1}, 925 {"", "", 2}, 926 {"-", "", 0}, 927 {"-", "-", 1}, 928 {"-", "----------", 10}, 929 {"abc ", "abc abc abc ", 3}, 930 } 931 932 func TestRepeat(t *testing.T) { 933 for _, tt := range RepeatTests { 934 a := Repeat(tt.in, tt.count) 935 if !equal("Repeat(s)", a, tt.out, t) { 936 t.Errorf("Repeat(%v, %d) = %v; want %v", tt.in, tt.count, a, tt.out) 937 continue 938 } 939 } 940 } 941 942 func repeat(s string, count int) (err error) { 943 defer func() { 944 if r := recover(); r != nil { 945 switch v := r.(type) { 946 case error: 947 err = v 948 default: 949 err = fmt.Errorf("%s", v) 950 } 951 } 952 }() 953 954 Repeat(s, count) 955 956 return 957 } 958 959 // See Issue golang.org/issue/16237 960 func TestRepeatCatchesOverflow(t *testing.T) { 961 tests := [...]struct { 962 s string 963 count int 964 errStr string 965 }{ 966 0: {"--", -2147483647, "negative"}, 967 1: {"", int(^uint(0) >> 1), ""}, 968 2: {"-", 10, ""}, 969 3: {"gopher", 0, ""}, 970 4: {"-", -1, "negative"}, 971 5: {"--", -102, "negative"}, 972 6: {string(make([]byte, 255)), int((^uint(0))/255 + 1), "overflow"}, 973 } 974 975 for i, tt := range tests { 976 err := repeat(tt.s, tt.count) 977 if tt.errStr == "" { 978 if err != nil { 979 t.Errorf("#%d panicked %v", i, err) 980 } 981 continue 982 } 983 984 if err == nil || !Contains(err.Error(), tt.errStr) { 985 t.Errorf("#%d expected %q got %q", i, tt.errStr, err) 986 } 987 } 988 } 989 990 func runesEqual(a, b []rune) bool { 991 if len(a) != len(b) { 992 return false 993 } 994 for i, r := range a { 995 if r != b[i] { 996 return false 997 } 998 } 999 return true 1000 } 1001 1002 var RunesTests = []struct { 1003 in string 1004 out []rune 1005 lossy bool 1006 }{ 1007 {"", []rune{}, false}, 1008 {" ", []rune{32}, false}, 1009 {"ABC", []rune{65, 66, 67}, false}, 1010 {"abc", []rune{97, 98, 99}, false}, 1011 {"\u65e5\u672c\u8a9e", []rune{26085, 26412, 35486}, false}, 1012 {"ab\x80c", []rune{97, 98, 0xFFFD, 99}, true}, 1013 {"ab\xc0c", []rune{97, 98, 0xFFFD, 99}, true}, 1014 } 1015 1016 func TestRunes(t *testing.T) { 1017 for _, tt := range RunesTests { 1018 a := []rune(tt.in) 1019 if !runesEqual(a, tt.out) { 1020 t.Errorf("[]rune(%q) = %v; want %v", tt.in, a, tt.out) 1021 continue 1022 } 1023 if !tt.lossy { 1024 // can only test reassembly if we didn't lose information 1025 s := string(a) 1026 if s != tt.in { 1027 t.Errorf("string([]rune(%q)) = %x; want %x", tt.in, s, tt.in) 1028 } 1029 } 1030 } 1031 } 1032 1033 func TestReadByte(t *testing.T) { 1034 testStrings := []string{"", abcd, faces, commas} 1035 for _, s := range testStrings { 1036 reader := NewReader(s) 1037 if e := reader.UnreadByte(); e == nil { 1038 t.Errorf("Unreading %q at beginning: expected error", s) 1039 } 1040 var res bytes.Buffer 1041 for { 1042 b, e := reader.ReadByte() 1043 if e == io.EOF { 1044 break 1045 } 1046 if e != nil { 1047 t.Errorf("Reading %q: %s", s, e) 1048 break 1049 } 1050 res.WriteByte(b) 1051 // unread and read again 1052 e = reader.UnreadByte() 1053 if e != nil { 1054 t.Errorf("Unreading %q: %s", s, e) 1055 break 1056 } 1057 b1, e := reader.ReadByte() 1058 if e != nil { 1059 t.Errorf("Reading %q after unreading: %s", s, e) 1060 break 1061 } 1062 if b1 != b { 1063 t.Errorf("Reading %q after unreading: want byte %q, got %q", s, b, b1) 1064 break 1065 } 1066 } 1067 if res.String() != s { 1068 t.Errorf("Reader(%q).ReadByte() produced %q", s, res.String()) 1069 } 1070 } 1071 } 1072 1073 func TestReadRune(t *testing.T) { 1074 testStrings := []string{"", abcd, faces, commas} 1075 for _, s := range testStrings { 1076 reader := NewReader(s) 1077 if e := reader.UnreadRune(); e == nil { 1078 t.Errorf("Unreading %q at beginning: expected error", s) 1079 } 1080 res := "" 1081 for { 1082 r, z, e := reader.ReadRune() 1083 if e == io.EOF { 1084 break 1085 } 1086 if e != nil { 1087 t.Errorf("Reading %q: %s", s, e) 1088 break 1089 } 1090 res += string(r) 1091 // unread and read again 1092 e = reader.UnreadRune() 1093 if e != nil { 1094 t.Errorf("Unreading %q: %s", s, e) 1095 break 1096 } 1097 r1, z1, e := reader.ReadRune() 1098 if e != nil { 1099 t.Errorf("Reading %q after unreading: %s", s, e) 1100 break 1101 } 1102 if r1 != r { 1103 t.Errorf("Reading %q after unreading: want rune %q, got %q", s, r, r1) 1104 break 1105 } 1106 if z1 != z { 1107 t.Errorf("Reading %q after unreading: want size %d, got %d", s, z, z1) 1108 break 1109 } 1110 } 1111 if res != s { 1112 t.Errorf("Reader(%q).ReadRune() produced %q", s, res) 1113 } 1114 } 1115 } 1116 1117 var UnreadRuneErrorTests = []struct { 1118 name string 1119 f func(*Reader) 1120 }{ 1121 {"Read", func(r *Reader) { r.Read([]byte{0}) }}, 1122 {"ReadByte", func(r *Reader) { r.ReadByte() }}, 1123 {"UnreadRune", func(r *Reader) { r.UnreadRune() }}, 1124 {"Seek", func(r *Reader) { r.Seek(0, io.SeekCurrent) }}, 1125 {"WriteTo", func(r *Reader) { r.WriteTo(&bytes.Buffer{}) }}, 1126 } 1127 1128 func TestUnreadRuneError(t *testing.T) { 1129 for _, tt := range UnreadRuneErrorTests { 1130 reader := NewReader("0123456789") 1131 if _, _, err := reader.ReadRune(); err != nil { 1132 // should not happen 1133 t.Fatal(err) 1134 } 1135 tt.f(reader) 1136 err := reader.UnreadRune() 1137 if err == nil { 1138 t.Errorf("Unreading after %s: expected error", tt.name) 1139 } 1140 } 1141 } 1142 1143 var ReplaceTests = []struct { 1144 in string 1145 old, new string 1146 n int 1147 out string 1148 }{ 1149 {"hello", "l", "L", 0, "hello"}, 1150 {"hello", "l", "L", -1, "heLLo"}, 1151 {"hello", "x", "X", -1, "hello"}, 1152 {"", "x", "X", -1, ""}, 1153 {"radar", "r", "<r>", -1, "<r>ada<r>"}, 1154 {"", "", "<>", -1, "<>"}, 1155 {"banana", "a", "<>", -1, "b<>n<>n<>"}, 1156 {"banana", "a", "<>", 1, "b<>nana"}, 1157 {"banana", "a", "<>", 1000, "b<>n<>n<>"}, 1158 {"banana", "an", "<>", -1, "b<><>a"}, 1159 {"banana", "ana", "<>", -1, "b<>na"}, 1160 {"banana", "", "<>", -1, "<>b<>a<>n<>a<>n<>a<>"}, 1161 {"banana", "", "<>", 10, "<>b<>a<>n<>a<>n<>a<>"}, 1162 {"banana", "", "<>", 6, "<>b<>a<>n<>a<>n<>a"}, 1163 {"banana", "", "<>", 5, "<>b<>a<>n<>a<>na"}, 1164 {"banana", "", "<>", 1, "<>banana"}, 1165 {"banana", "a", "a", -1, "banana"}, 1166 {"banana", "a", "a", 1, "banana"}, 1167 {"☺☻☹", "", "<>", -1, "<>☺<>☻<>☹<>"}, 1168 } 1169 1170 func TestReplace(t *testing.T) { 1171 for _, tt := range ReplaceTests { 1172 if s := Replace(tt.in, tt.old, tt.new, tt.n); s != tt.out { 1173 t.Errorf("Replace(%q, %q, %q, %d) = %q, want %q", tt.in, tt.old, tt.new, tt.n, s, tt.out) 1174 } 1175 } 1176 } 1177 1178 var TitleTests = []struct { 1179 in, out string 1180 }{ 1181 {"", ""}, 1182 {"a", "A"}, 1183 {" aaa aaa aaa ", " Aaa Aaa Aaa "}, 1184 {" Aaa Aaa Aaa ", " Aaa Aaa Aaa "}, 1185 {"123a456", "123a456"}, 1186 {"double-blind", "Double-Blind"}, 1187 {"ÿøû", "Ÿøû"}, 1188 {"with_underscore", "With_underscore"}, 1189 {"unicode \xe2\x80\xa8 line separator", "Unicode \xe2\x80\xa8 Line Separator"}, 1190 } 1191 1192 func TestTitle(t *testing.T) { 1193 for _, tt := range TitleTests { 1194 if s := Title(tt.in); s != tt.out { 1195 t.Errorf("Title(%q) = %q, want %q", tt.in, s, tt.out) 1196 } 1197 } 1198 } 1199 1200 var ContainsTests = []struct { 1201 str, substr string 1202 expected bool 1203 }{ 1204 {"abc", "bc", true}, 1205 {"abc", "bcd", false}, 1206 {"abc", "", true}, 1207 {"", "a", false}, 1208 1209 // cases to cover code in runtime/asm_amd64.s:indexShortStr 1210 // 2-byte needle 1211 {"xxxxxx", "01", false}, 1212 {"01xxxx", "01", true}, 1213 {"xx01xx", "01", true}, 1214 {"xxxx01", "01", true}, 1215 {"01xxxxx"[1:], "01", false}, 1216 {"xxxxx01"[:6], "01", false}, 1217 // 3-byte needle 1218 {"xxxxxxx", "012", false}, 1219 {"012xxxx", "012", true}, 1220 {"xx012xx", "012", true}, 1221 {"xxxx012", "012", true}, 1222 {"012xxxxx"[1:], "012", false}, 1223 {"xxxxx012"[:7], "012", false}, 1224 // 4-byte needle 1225 {"xxxxxxxx", "0123", false}, 1226 {"0123xxxx", "0123", true}, 1227 {"xx0123xx", "0123", true}, 1228 {"xxxx0123", "0123", true}, 1229 {"0123xxxxx"[1:], "0123", false}, 1230 {"xxxxx0123"[:8], "0123", false}, 1231 // 5-7-byte needle 1232 {"xxxxxxxxx", "01234", false}, 1233 {"01234xxxx", "01234", true}, 1234 {"xx01234xx", "01234", true}, 1235 {"xxxx01234", "01234", true}, 1236 {"01234xxxxx"[1:], "01234", false}, 1237 {"xxxxx01234"[:9], "01234", false}, 1238 // 8-byte needle 1239 {"xxxxxxxxxxxx", "01234567", false}, 1240 {"01234567xxxx", "01234567", true}, 1241 {"xx01234567xx", "01234567", true}, 1242 {"xxxx01234567", "01234567", true}, 1243 {"01234567xxxxx"[1:], "01234567", false}, 1244 {"xxxxx01234567"[:12], "01234567", false}, 1245 // 9-15-byte needle 1246 {"xxxxxxxxxxxxx", "012345678", false}, 1247 {"012345678xxxx", "012345678", true}, 1248 {"xx012345678xx", "012345678", true}, 1249 {"xxxx012345678", "012345678", true}, 1250 {"012345678xxxxx"[1:], "012345678", false}, 1251 {"xxxxx012345678"[:13], "012345678", false}, 1252 // 16-byte needle 1253 {"xxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEF", false}, 1254 {"0123456789ABCDEFxxxx", "0123456789ABCDEF", true}, 1255 {"xx0123456789ABCDEFxx", "0123456789ABCDEF", true}, 1256 {"xxxx0123456789ABCDEF", "0123456789ABCDEF", true}, 1257 {"0123456789ABCDEFxxxxx"[1:], "0123456789ABCDEF", false}, 1258 {"xxxxx0123456789ABCDEF"[:20], "0123456789ABCDEF", false}, 1259 // 17-31-byte needle 1260 {"xxxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEFG", false}, 1261 {"0123456789ABCDEFGxxxx", "0123456789ABCDEFG", true}, 1262 {"xx0123456789ABCDEFGxx", "0123456789ABCDEFG", true}, 1263 {"xxxx0123456789ABCDEFG", "0123456789ABCDEFG", true}, 1264 {"0123456789ABCDEFGxxxxx"[1:], "0123456789ABCDEFG", false}, 1265 {"xxxxx0123456789ABCDEFG"[:21], "0123456789ABCDEFG", false}, 1266 1267 // partial match cases 1268 {"xx01x", "012", false}, // 3 1269 {"xx0123x", "01234", false}, // 5-7 1270 {"xx01234567x", "012345678", false}, // 9-15 1271 {"xx0123456789ABCDEFx", "0123456789ABCDEFG", false}, // 17-31, issue 15679 1272 } 1273 1274 func TestContains(t *testing.T) { 1275 for _, ct := range ContainsTests { 1276 if Contains(ct.str, ct.substr) != ct.expected { 1277 t.Errorf("Contains(%s, %s) = %v, want %v", 1278 ct.str, ct.substr, !ct.expected, ct.expected) 1279 } 1280 } 1281 } 1282 1283 var ContainsAnyTests = []struct { 1284 str, substr string 1285 expected bool 1286 }{ 1287 {"", "", false}, 1288 {"", "a", false}, 1289 {"", "abc", false}, 1290 {"a", "", false}, 1291 {"a", "a", true}, 1292 {"aaa", "a", true}, 1293 {"abc", "xyz", false}, 1294 {"abc", "xcz", true}, 1295 {"a☺b☻c☹d", "uvw☻xyz", true}, 1296 {"aRegExp*", ".(|)*+?^$[]", true}, 1297 {dots + dots + dots, " ", false}, 1298 } 1299 1300 func TestContainsAny(t *testing.T) { 1301 for _, ct := range ContainsAnyTests { 1302 if ContainsAny(ct.str, ct.substr) != ct.expected { 1303 t.Errorf("ContainsAny(%s, %s) = %v, want %v", 1304 ct.str, ct.substr, !ct.expected, ct.expected) 1305 } 1306 } 1307 } 1308 1309 var ContainsRuneTests = []struct { 1310 str string 1311 r rune 1312 expected bool 1313 }{ 1314 {"", 'a', false}, 1315 {"a", 'a', true}, 1316 {"aaa", 'a', true}, 1317 {"abc", 'y', false}, 1318 {"abc", 'c', true}, 1319 {"a☺b☻c☹d", 'x', false}, 1320 {"a☺b☻c☹d", '☻', true}, 1321 {"aRegExp*", '*', true}, 1322 } 1323 1324 func TestContainsRune(t *testing.T) { 1325 for _, ct := range ContainsRuneTests { 1326 if ContainsRune(ct.str, ct.r) != ct.expected { 1327 t.Errorf("ContainsRune(%q, %q) = %v, want %v", 1328 ct.str, ct.r, !ct.expected, ct.expected) 1329 } 1330 } 1331 } 1332 1333 var EqualFoldTests = []struct { 1334 s, t string 1335 out bool 1336 }{ 1337 {"abc", "abc", true}, 1338 {"ABcd", "ABcd", true}, 1339 {"123abc", "123ABC", true}, 1340 {"αβδ", "ΑΒΔ", true}, 1341 {"abc", "xyz", false}, 1342 {"abc", "XYZ", false}, 1343 {"abcdefghijk", "abcdefghijX", false}, 1344 {"abcdefghijk", "abcdefghij\u212A", true}, 1345 {"abcdefghijK", "abcdefghij\u212A", true}, 1346 {"abcdefghijkz", "abcdefghij\u212Ay", false}, 1347 {"abcdefghijKz", "abcdefghij\u212Ay", false}, 1348 } 1349 1350 func TestEqualFold(t *testing.T) { 1351 for _, tt := range EqualFoldTests { 1352 if out := EqualFold(tt.s, tt.t); out != tt.out { 1353 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.s, tt.t, out, tt.out) 1354 } 1355 if out := EqualFold(tt.t, tt.s); out != tt.out { 1356 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.t, tt.s, out, tt.out) 1357 } 1358 } 1359 } 1360 1361 var CountTests = []struct { 1362 s, sep string 1363 num int 1364 }{ 1365 {"", "", 1}, 1366 {"", "notempty", 0}, 1367 {"notempty", "", 9}, 1368 {"smaller", "not smaller", 0}, 1369 {"12345678987654321", "6", 2}, 1370 {"611161116", "6", 3}, 1371 {"notequal", "NotEqual", 0}, 1372 {"equal", "equal", 1}, 1373 {"abc1231231123q", "123", 3}, 1374 {"11111", "11", 2}, 1375 } 1376 1377 func TestCount(t *testing.T) { 1378 for _, tt := range CountTests { 1379 if num := Count(tt.s, tt.sep); num != tt.num { 1380 t.Errorf("Count(\"%s\", \"%s\") = %d, want %d", tt.s, tt.sep, num, tt.num) 1381 } 1382 } 1383 } 1384 1385 func makeBenchInputHard() string { 1386 tokens := [...]string{ 1387 "<a>", "<p>", "<b>", "<strong>", 1388 "</a>", "</p>", "</b>", "</strong>", 1389 "hello", "world", 1390 } 1391 x := make([]byte, 0, 1<<20) 1392 for { 1393 i := rand.Intn(len(tokens)) 1394 if len(x)+len(tokens[i]) >= 1<<20 { 1395 break 1396 } 1397 x = append(x, tokens[i]...) 1398 } 1399 return string(x) 1400 } 1401 1402 var benchInputHard = makeBenchInputHard() 1403 1404 func benchmarkIndexHard(b *testing.B, sep string) { 1405 for i := 0; i < b.N; i++ { 1406 Index(benchInputHard, sep) 1407 } 1408 } 1409 1410 func benchmarkLastIndexHard(b *testing.B, sep string) { 1411 for i := 0; i < b.N; i++ { 1412 LastIndex(benchInputHard, sep) 1413 } 1414 } 1415 1416 func benchmarkCountHard(b *testing.B, sep string) { 1417 for i := 0; i < b.N; i++ { 1418 Count(benchInputHard, sep) 1419 } 1420 } 1421 1422 func BenchmarkIndexHard1(b *testing.B) { benchmarkIndexHard(b, "<>") } 1423 func BenchmarkIndexHard2(b *testing.B) { benchmarkIndexHard(b, "</pre>") } 1424 func BenchmarkIndexHard3(b *testing.B) { benchmarkIndexHard(b, "<b>hello world</b>") } 1425 func BenchmarkIndexHard4(b *testing.B) { 1426 benchmarkIndexHard(b, "<pre><b>hello</b><strong>world</strong></pre>") 1427 } 1428 1429 func BenchmarkLastIndexHard1(b *testing.B) { benchmarkLastIndexHard(b, "<>") } 1430 func BenchmarkLastIndexHard2(b *testing.B) { benchmarkLastIndexHard(b, "</pre>") } 1431 func BenchmarkLastIndexHard3(b *testing.B) { benchmarkLastIndexHard(b, "<b>hello world</b>") } 1432 1433 func BenchmarkCountHard1(b *testing.B) { benchmarkCountHard(b, "<>") } 1434 func BenchmarkCountHard2(b *testing.B) { benchmarkCountHard(b, "</pre>") } 1435 func BenchmarkCountHard3(b *testing.B) { benchmarkCountHard(b, "<b>hello world</b>") } 1436 1437 var benchInputTorture = Repeat("ABC", 1<<10) + "123" + Repeat("ABC", 1<<10) 1438 var benchNeedleTorture = Repeat("ABC", 1<<10+1) 1439 1440 func BenchmarkIndexTorture(b *testing.B) { 1441 for i := 0; i < b.N; i++ { 1442 Index(benchInputTorture, benchNeedleTorture) 1443 } 1444 } 1445 1446 func BenchmarkCountTorture(b *testing.B) { 1447 for i := 0; i < b.N; i++ { 1448 Count(benchInputTorture, benchNeedleTorture) 1449 } 1450 } 1451 1452 func BenchmarkCountTortureOverlapping(b *testing.B) { 1453 A := Repeat("ABC", 1<<20) 1454 B := Repeat("ABC", 1<<10) 1455 for i := 0; i < b.N; i++ { 1456 Count(A, B) 1457 } 1458 } 1459 1460 func BenchmarkCountByte(b *testing.B) { 1461 indexSizes := []int{10, 32, 4 << 10, 4 << 20, 64 << 20} 1462 benchStr := Repeat(benchmarkString, 1463 (indexSizes[len(indexSizes)-1]+len(benchmarkString)-1)/len(benchmarkString)) 1464 benchFunc := func(b *testing.B, benchStr string) { 1465 b.SetBytes(int64(len(benchStr))) 1466 for i := 0; i < b.N; i++ { 1467 Count(benchStr, "=") 1468 } 1469 } 1470 for _, size := range indexSizes { 1471 b.Run(fmt.Sprintf("%d", size), func(b *testing.B) { 1472 benchFunc(b, benchStr[:size]) 1473 }) 1474 } 1475 1476 } 1477 1478 var makeFieldsInput = func() string { 1479 x := make([]byte, 1<<20) 1480 // Input is ~10% space, ~10% 2-byte UTF-8, rest ASCII non-space. 1481 for i := range x { 1482 switch rand.Intn(10) { 1483 case 0: 1484 x[i] = ' ' 1485 case 1: 1486 if i > 0 && x[i-1] == 'x' { 1487 copy(x[i-1:], "χ") 1488 break 1489 } 1490 fallthrough 1491 default: 1492 x[i] = 'x' 1493 } 1494 } 1495 return string(x) 1496 } 1497 1498 var makeFieldsInputASCII = func() string { 1499 x := make([]byte, 1<<20) 1500 // Input is ~10% space, rest ASCII non-space. 1501 for i := range x { 1502 if rand.Intn(10) == 0 { 1503 x[i] = ' ' 1504 } else { 1505 x[i] = 'x' 1506 } 1507 } 1508 return string(x) 1509 } 1510 1511 var stringdata = []struct{ name, data string }{ 1512 {"ASCII", makeFieldsInputASCII()}, 1513 {"Mixed", makeFieldsInput()}, 1514 } 1515 1516 func BenchmarkFields(b *testing.B) { 1517 for _, sd := range stringdata { 1518 b.Run(sd.name, func(b *testing.B) { 1519 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1520 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1521 b.ReportAllocs() 1522 b.SetBytes(int64(j)) 1523 data := sd.data[:j] 1524 for i := 0; i < b.N; i++ { 1525 Fields(data) 1526 } 1527 }) 1528 } 1529 }) 1530 } 1531 } 1532 1533 func BenchmarkFieldsFunc(b *testing.B) { 1534 for _, sd := range stringdata { 1535 b.Run(sd.name, func(b *testing.B) { 1536 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1537 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1538 b.ReportAllocs() 1539 b.SetBytes(int64(j)) 1540 data := sd.data[:j] 1541 for i := 0; i < b.N; i++ { 1542 FieldsFunc(data, unicode.IsSpace) 1543 } 1544 }) 1545 } 1546 }) 1547 } 1548 } 1549 1550 func BenchmarkSplitEmptySeparator(b *testing.B) { 1551 for i := 0; i < b.N; i++ { 1552 Split(benchInputHard, "") 1553 } 1554 } 1555 1556 func BenchmarkSplitSingleByteSeparator(b *testing.B) { 1557 for i := 0; i < b.N; i++ { 1558 Split(benchInputHard, "/") 1559 } 1560 } 1561 1562 func BenchmarkSplitMultiByteSeparator(b *testing.B) { 1563 for i := 0; i < b.N; i++ { 1564 Split(benchInputHard, "hello") 1565 } 1566 } 1567 1568 func BenchmarkSplitNSingleByteSeparator(b *testing.B) { 1569 for i := 0; i < b.N; i++ { 1570 SplitN(benchInputHard, "/", 10) 1571 } 1572 } 1573 1574 func BenchmarkSplitNMultiByteSeparator(b *testing.B) { 1575 for i := 0; i < b.N; i++ { 1576 SplitN(benchInputHard, "hello", 10) 1577 } 1578 } 1579 1580 func BenchmarkRepeat(b *testing.B) { 1581 for i := 0; i < b.N; i++ { 1582 Repeat("-", 80) 1583 } 1584 } 1585 1586 func BenchmarkIndexAnyASCII(b *testing.B) { 1587 x := Repeat("#", 4096) // Never matches set 1588 cs := "0123456789abcdef" 1589 for k := 1; k <= 4096; k <<= 4 { 1590 for j := 1; j <= 16; j <<= 1 { 1591 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1592 for i := 0; i < b.N; i++ { 1593 IndexAny(x[:k], cs[:j]) 1594 } 1595 }) 1596 } 1597 } 1598 } 1599 1600 func BenchmarkTrimASCII(b *testing.B) { 1601 cs := "0123456789abcdef" 1602 for k := 1; k <= 4096; k <<= 4 { 1603 for j := 1; j <= 16; j <<= 1 { 1604 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1605 x := Repeat(cs[:j], k) // Always matches set 1606 for i := 0; i < b.N; i++ { 1607 Trim(x[:k], cs[:j]) 1608 } 1609 }) 1610 } 1611 } 1612 }