github.com/goproxy0/go@v0.0.0-20171111080102-49cc0c489d2c/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 {"ONLYUPPER", "ONLYUPPER"}, 522 {"abc", "ABC"}, 523 {"AbC123", "ABC123"}, 524 {"azAZ09_", "AZAZ09_"}, 525 {"longStrinGwitHmixofsmaLLandcAps", "LONGSTRINGWITHMIXOFSMALLANDCAPS"}, 526 {"long\u0250string\u0250with\u0250nonascii\u2C6Fchars", "LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS"}, 527 {"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"}, // grows one byte per char 528 } 529 530 var lowerTests = []StringTest{ 531 {"", ""}, 532 {"abc", "abc"}, 533 {"AbC123", "abc123"}, 534 {"azAZ09_", "azaz09_"}, 535 {"longStrinGwitHmixofsmaLLandcAps", "longstringwithmixofsmallandcaps"}, 536 {"LONG\u2C6FSTRING\u2C6FWITH\u2C6FNONASCII\u2C6FCHARS", "long\u0250string\u0250with\u0250nonascii\u0250chars"}, 537 {"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"}, // shrinks one byte per char 538 } 539 540 const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000" 541 542 var trimSpaceTests = []StringTest{ 543 {"", ""}, 544 {"abc", "abc"}, 545 {space + "abc" + space, "abc"}, 546 {" ", ""}, 547 {" \t\r\n \t\t\r\r\n\n ", ""}, 548 {" \t\r\n x\t\t\r\r\n\n ", "x"}, 549 {" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, 550 {"1 \t\r\n2", "1 \t\r\n2"}, 551 {" x\x80", "x\x80"}, 552 {" x\xc0", "x\xc0"}, 553 {"x \xc0\xc0 ", "x \xc0\xc0"}, 554 {"x \xc0", "x \xc0"}, 555 {"x \xc0 ", "x \xc0"}, 556 {"x \xc0\xc0 ", "x \xc0\xc0"}, 557 {"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, 558 {"x ☺ ", "x ☺"}, 559 } 560 561 func tenRunes(ch rune) string { 562 r := make([]rune, 10) 563 for i := range r { 564 r[i] = ch 565 } 566 return string(r) 567 } 568 569 // User-defined self-inverse mapping function 570 func rot13(r rune) rune { 571 step := rune(13) 572 if r >= 'a' && r <= 'z' { 573 return ((r - 'a' + step) % 26) + 'a' 574 } 575 if r >= 'A' && r <= 'Z' { 576 return ((r - 'A' + step) % 26) + 'A' 577 } 578 return r 579 } 580 581 func TestMap(t *testing.T) { 582 // Run a couple of awful growth/shrinkage tests 583 a := tenRunes('a') 584 // 1. Grow. This triggers two reallocations in Map. 585 maxRune := func(rune) rune { return unicode.MaxRune } 586 m := Map(maxRune, a) 587 expect := tenRunes(unicode.MaxRune) 588 if m != expect { 589 t.Errorf("growing: expected %q got %q", expect, m) 590 } 591 592 // 2. Shrink 593 minRune := func(rune) rune { return 'a' } 594 m = Map(minRune, tenRunes(unicode.MaxRune)) 595 expect = a 596 if m != expect { 597 t.Errorf("shrinking: expected %q got %q", expect, m) 598 } 599 600 // 3. Rot13 601 m = Map(rot13, "a to zed") 602 expect = "n gb mrq" 603 if m != expect { 604 t.Errorf("rot13: expected %q got %q", expect, m) 605 } 606 607 // 4. Rot13^2 608 m = Map(rot13, Map(rot13, "a to zed")) 609 expect = "a to zed" 610 if m != expect { 611 t.Errorf("rot13: expected %q got %q", expect, m) 612 } 613 614 // 5. Drop 615 dropNotLatin := func(r rune) rune { 616 if unicode.Is(unicode.Latin, r) { 617 return r 618 } 619 return -1 620 } 621 m = Map(dropNotLatin, "Hello, 세계") 622 expect = "Hello" 623 if m != expect { 624 t.Errorf("drop: expected %q got %q", expect, m) 625 } 626 627 // 6. Identity 628 identity := func(r rune) rune { 629 return r 630 } 631 orig := "Input string that we expect not to be copied." 632 m = Map(identity, orig) 633 if (*reflect.StringHeader)(unsafe.Pointer(&orig)).Data != 634 (*reflect.StringHeader)(unsafe.Pointer(&m)).Data { 635 t.Error("unexpected copy during identity map") 636 } 637 638 // 7. Handle invalid UTF-8 sequence 639 replaceNotLatin := func(r rune) rune { 640 if unicode.Is(unicode.Latin, r) { 641 return r 642 } 643 return '?' 644 } 645 m = Map(replaceNotLatin, "Hello\255World") 646 expect = "Hello?World" 647 if m != expect { 648 t.Errorf("replace invalid sequence: expected %q got %q", expect, m) 649 } 650 } 651 652 func TestToUpper(t *testing.T) { runStringTests(t, ToUpper, "ToUpper", upperTests) } 653 654 func TestToLower(t *testing.T) { runStringTests(t, ToLower, "ToLower", lowerTests) } 655 656 func BenchmarkToUpper(b *testing.B) { 657 for _, tc := range upperTests { 658 b.Run(tc.in, func(b *testing.B) { 659 for i := 0; i < b.N; i++ { 660 actual := ToUpper(tc.in) 661 if actual != tc.out { 662 b.Errorf("ToUpper(%q) = %q; want %q", tc.in, actual, tc.out) 663 } 664 } 665 }) 666 } 667 } 668 669 func BenchmarkToLower(b *testing.B) { 670 for _, tc := range lowerTests { 671 b.Run(tc.in, func(b *testing.B) { 672 for i := 0; i < b.N; i++ { 673 actual := ToLower(tc.in) 674 if actual != tc.out { 675 b.Errorf("ToLower(%q) = %q; want %q", tc.in, actual, tc.out) 676 } 677 } 678 }) 679 } 680 } 681 682 func BenchmarkMapNoChanges(b *testing.B) { 683 identity := func(r rune) rune { 684 return r 685 } 686 for i := 0; i < b.N; i++ { 687 Map(identity, "Some string that won't be modified.") 688 } 689 } 690 691 func TestSpecialCase(t *testing.T) { 692 lower := "abcçdefgğhıijklmnoöprsştuüvyz" 693 upper := "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ" 694 u := ToUpperSpecial(unicode.TurkishCase, upper) 695 if u != upper { 696 t.Errorf("Upper(upper) is %s not %s", u, upper) 697 } 698 u = ToUpperSpecial(unicode.TurkishCase, lower) 699 if u != upper { 700 t.Errorf("Upper(lower) is %s not %s", u, upper) 701 } 702 l := ToLowerSpecial(unicode.TurkishCase, lower) 703 if l != lower { 704 t.Errorf("Lower(lower) is %s not %s", l, lower) 705 } 706 l = ToLowerSpecial(unicode.TurkishCase, upper) 707 if l != lower { 708 t.Errorf("Lower(upper) is %s not %s", l, lower) 709 } 710 } 711 712 func TestTrimSpace(t *testing.T) { runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests) } 713 714 var trimTests = []struct { 715 f string 716 in, arg, out string 717 }{ 718 {"Trim", "abba", "a", "bb"}, 719 {"Trim", "abba", "ab", ""}, 720 {"TrimLeft", "abba", "ab", ""}, 721 {"TrimRight", "abba", "ab", ""}, 722 {"TrimLeft", "abba", "a", "bba"}, 723 {"TrimRight", "abba", "a", "abb"}, 724 {"Trim", "<tag>", "<>", "tag"}, 725 {"Trim", "* listitem", " *", "listitem"}, 726 {"Trim", `"quote"`, `"`, "quote"}, 727 {"Trim", "\u2C6F\u2C6F\u0250\u0250\u2C6F\u2C6F", "\u2C6F", "\u0250\u0250"}, 728 {"Trim", "\x80test\xff", "\xff", "test"}, 729 {"Trim", " Ġ ", " ", "Ġ"}, 730 {"Trim", " Ġİ0", "0 ", "Ġİ"}, 731 //empty string tests 732 {"Trim", "abba", "", "abba"}, 733 {"Trim", "", "123", ""}, 734 {"Trim", "", "", ""}, 735 {"TrimLeft", "abba", "", "abba"}, 736 {"TrimLeft", "", "123", ""}, 737 {"TrimLeft", "", "", ""}, 738 {"TrimRight", "abba", "", "abba"}, 739 {"TrimRight", "", "123", ""}, 740 {"TrimRight", "", "", ""}, 741 {"TrimRight", "☺\xc0", "☺", "☺\xc0"}, 742 {"TrimPrefix", "aabb", "a", "abb"}, 743 {"TrimPrefix", "aabb", "b", "aabb"}, 744 {"TrimSuffix", "aabb", "a", "aabb"}, 745 {"TrimSuffix", "aabb", "b", "aab"}, 746 } 747 748 func TestTrim(t *testing.T) { 749 for _, tc := range trimTests { 750 name := tc.f 751 var f func(string, string) string 752 switch name { 753 case "Trim": 754 f = Trim 755 case "TrimLeft": 756 f = TrimLeft 757 case "TrimRight": 758 f = TrimRight 759 case "TrimPrefix": 760 f = TrimPrefix 761 case "TrimSuffix": 762 f = TrimSuffix 763 default: 764 t.Errorf("Undefined trim function %s", name) 765 } 766 actual := f(tc.in, tc.arg) 767 if actual != tc.out { 768 t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 769 } 770 } 771 } 772 773 func BenchmarkTrim(b *testing.B) { 774 b.ReportAllocs() 775 776 for i := 0; i < b.N; i++ { 777 for _, tc := range trimTests { 778 name := tc.f 779 var f func(string, string) string 780 switch name { 781 case "Trim": 782 f = Trim 783 case "TrimLeft": 784 f = TrimLeft 785 case "TrimRight": 786 f = TrimRight 787 case "TrimPrefix": 788 f = TrimPrefix 789 case "TrimSuffix": 790 f = TrimSuffix 791 default: 792 b.Errorf("Undefined trim function %s", name) 793 } 794 actual := f(tc.in, tc.arg) 795 if actual != tc.out { 796 b.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 797 } 798 } 799 } 800 } 801 802 type predicate struct { 803 f func(rune) bool 804 name string 805 } 806 807 var isSpace = predicate{unicode.IsSpace, "IsSpace"} 808 var isDigit = predicate{unicode.IsDigit, "IsDigit"} 809 var isUpper = predicate{unicode.IsUpper, "IsUpper"} 810 var isValidRune = predicate{ 811 func(r rune) bool { 812 return r != utf8.RuneError 813 }, 814 "IsValidRune", 815 } 816 817 func not(p predicate) predicate { 818 return predicate{ 819 func(r rune) bool { 820 return !p.f(r) 821 }, 822 "not " + p.name, 823 } 824 } 825 826 var trimFuncTests = []struct { 827 f predicate 828 in, out string 829 }{ 830 {isSpace, space + " hello " + space, "hello"}, 831 {isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", "hello"}, 832 {isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", "hello"}, 833 {not(isSpace), "hello" + space + "hello", space}, 834 {not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", "\u0e50\u0e521234\u0e50\u0e51"}, 835 {isValidRune, "ab\xc0a\xc0cd", "\xc0a\xc0"}, 836 {not(isValidRune), "\xc0a\xc0", "a"}, 837 } 838 839 func TestTrimFunc(t *testing.T) { 840 for _, tc := range trimFuncTests { 841 actual := TrimFunc(tc.in, tc.f.f) 842 if actual != tc.out { 843 t.Errorf("TrimFunc(%q, %q) = %q; want %q", tc.in, tc.f.name, actual, tc.out) 844 } 845 } 846 } 847 848 var indexFuncTests = []struct { 849 in string 850 f predicate 851 first, last int 852 }{ 853 {"", isValidRune, -1, -1}, 854 {"abc", isDigit, -1, -1}, 855 {"0123", isDigit, 0, 3}, 856 {"a1b", isDigit, 1, 1}, 857 {space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes 858 {"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, 859 {"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, 860 {"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, 861 862 // tests of invalid UTF-8 863 {"\x801", isDigit, 1, 1}, 864 {"\x80abc", isDigit, -1, -1}, 865 {"\xc0a\xc0", isValidRune, 1, 1}, 866 {"\xc0a\xc0", not(isValidRune), 0, 2}, 867 {"\xc0☺\xc0", not(isValidRune), 0, 4}, 868 {"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, 869 {"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, 870 {"a\xe0\x80cd", not(isValidRune), 1, 2}, 871 {"\x80\x80\x80\x80", not(isValidRune), 0, 3}, 872 } 873 874 func TestIndexFunc(t *testing.T) { 875 for _, tc := range indexFuncTests { 876 first := IndexFunc(tc.in, tc.f.f) 877 if first != tc.first { 878 t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) 879 } 880 last := LastIndexFunc(tc.in, tc.f.f) 881 if last != tc.last { 882 t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) 883 } 884 } 885 } 886 887 func equal(m string, s1, s2 string, t *testing.T) bool { 888 if s1 == s2 { 889 return true 890 } 891 e1 := Split(s1, "") 892 e2 := Split(s2, "") 893 for i, c1 := range e1 { 894 if i >= len(e2) { 895 break 896 } 897 r1, _ := utf8.DecodeRuneInString(c1) 898 r2, _ := utf8.DecodeRuneInString(e2[i]) 899 if r1 != r2 { 900 t.Errorf("%s diff at %d: U+%04X U+%04X", m, i, r1, r2) 901 } 902 } 903 return false 904 } 905 906 func TestCaseConsistency(t *testing.T) { 907 // Make a string of all the runes. 908 numRunes := int(unicode.MaxRune + 1) 909 if testing.Short() { 910 numRunes = 1000 911 } 912 a := make([]rune, numRunes) 913 for i := range a { 914 a[i] = rune(i) 915 } 916 s := string(a) 917 // convert the cases. 918 upper := ToUpper(s) 919 lower := ToLower(s) 920 921 // Consistency checks 922 if n := utf8.RuneCountInString(upper); n != numRunes { 923 t.Error("rune count wrong in upper:", n) 924 } 925 if n := utf8.RuneCountInString(lower); n != numRunes { 926 t.Error("rune count wrong in lower:", n) 927 } 928 if !equal("ToUpper(upper)", ToUpper(upper), upper, t) { 929 t.Error("ToUpper(upper) consistency fail") 930 } 931 if !equal("ToLower(lower)", ToLower(lower), lower, t) { 932 t.Error("ToLower(lower) consistency fail") 933 } 934 /* 935 These fail because of non-one-to-oneness of the data, such as multiple 936 upper case 'I' mapping to 'i'. We comment them out but keep them for 937 interest. 938 For instance: CAPITAL LETTER I WITH DOT ABOVE: 939 unicode.ToUpper(unicode.ToLower('\u0130')) != '\u0130' 940 941 if !equal("ToUpper(lower)", ToUpper(lower), upper, t) { 942 t.Error("ToUpper(lower) consistency fail"); 943 } 944 if !equal("ToLower(upper)", ToLower(upper), lower, t) { 945 t.Error("ToLower(upper) consistency fail"); 946 } 947 */ 948 } 949 950 var RepeatTests = []struct { 951 in, out string 952 count int 953 }{ 954 {"", "", 0}, 955 {"", "", 1}, 956 {"", "", 2}, 957 {"-", "", 0}, 958 {"-", "-", 1}, 959 {"-", "----------", 10}, 960 {"abc ", "abc abc abc ", 3}, 961 } 962 963 func TestRepeat(t *testing.T) { 964 for _, tt := range RepeatTests { 965 a := Repeat(tt.in, tt.count) 966 if !equal("Repeat(s)", a, tt.out, t) { 967 t.Errorf("Repeat(%v, %d) = %v; want %v", tt.in, tt.count, a, tt.out) 968 continue 969 } 970 } 971 } 972 973 func repeat(s string, count int) (err error) { 974 defer func() { 975 if r := recover(); r != nil { 976 switch v := r.(type) { 977 case error: 978 err = v 979 default: 980 err = fmt.Errorf("%s", v) 981 } 982 } 983 }() 984 985 Repeat(s, count) 986 987 return 988 } 989 990 // See Issue golang.org/issue/16237 991 func TestRepeatCatchesOverflow(t *testing.T) { 992 tests := [...]struct { 993 s string 994 count int 995 errStr string 996 }{ 997 0: {"--", -2147483647, "negative"}, 998 1: {"", int(^uint(0) >> 1), ""}, 999 2: {"-", 10, ""}, 1000 3: {"gopher", 0, ""}, 1001 4: {"-", -1, "negative"}, 1002 5: {"--", -102, "negative"}, 1003 6: {string(make([]byte, 255)), int((^uint(0))/255 + 1), "overflow"}, 1004 } 1005 1006 for i, tt := range tests { 1007 err := repeat(tt.s, tt.count) 1008 if tt.errStr == "" { 1009 if err != nil { 1010 t.Errorf("#%d panicked %v", i, err) 1011 } 1012 continue 1013 } 1014 1015 if err == nil || !Contains(err.Error(), tt.errStr) { 1016 t.Errorf("#%d expected %q got %q", i, tt.errStr, err) 1017 } 1018 } 1019 } 1020 1021 func runesEqual(a, b []rune) bool { 1022 if len(a) != len(b) { 1023 return false 1024 } 1025 for i, r := range a { 1026 if r != b[i] { 1027 return false 1028 } 1029 } 1030 return true 1031 } 1032 1033 var RunesTests = []struct { 1034 in string 1035 out []rune 1036 lossy bool 1037 }{ 1038 {"", []rune{}, false}, 1039 {" ", []rune{32}, false}, 1040 {"ABC", []rune{65, 66, 67}, false}, 1041 {"abc", []rune{97, 98, 99}, false}, 1042 {"\u65e5\u672c\u8a9e", []rune{26085, 26412, 35486}, false}, 1043 {"ab\x80c", []rune{97, 98, 0xFFFD, 99}, true}, 1044 {"ab\xc0c", []rune{97, 98, 0xFFFD, 99}, true}, 1045 } 1046 1047 func TestRunes(t *testing.T) { 1048 for _, tt := range RunesTests { 1049 a := []rune(tt.in) 1050 if !runesEqual(a, tt.out) { 1051 t.Errorf("[]rune(%q) = %v; want %v", tt.in, a, tt.out) 1052 continue 1053 } 1054 if !tt.lossy { 1055 // can only test reassembly if we didn't lose information 1056 s := string(a) 1057 if s != tt.in { 1058 t.Errorf("string([]rune(%q)) = %x; want %x", tt.in, s, tt.in) 1059 } 1060 } 1061 } 1062 } 1063 1064 func TestReadByte(t *testing.T) { 1065 testStrings := []string{"", abcd, faces, commas} 1066 for _, s := range testStrings { 1067 reader := NewReader(s) 1068 if e := reader.UnreadByte(); e == nil { 1069 t.Errorf("Unreading %q at beginning: expected error", s) 1070 } 1071 var res bytes.Buffer 1072 for { 1073 b, e := reader.ReadByte() 1074 if e == io.EOF { 1075 break 1076 } 1077 if e != nil { 1078 t.Errorf("Reading %q: %s", s, e) 1079 break 1080 } 1081 res.WriteByte(b) 1082 // unread and read again 1083 e = reader.UnreadByte() 1084 if e != nil { 1085 t.Errorf("Unreading %q: %s", s, e) 1086 break 1087 } 1088 b1, e := reader.ReadByte() 1089 if e != nil { 1090 t.Errorf("Reading %q after unreading: %s", s, e) 1091 break 1092 } 1093 if b1 != b { 1094 t.Errorf("Reading %q after unreading: want byte %q, got %q", s, b, b1) 1095 break 1096 } 1097 } 1098 if res.String() != s { 1099 t.Errorf("Reader(%q).ReadByte() produced %q", s, res.String()) 1100 } 1101 } 1102 } 1103 1104 func TestReadRune(t *testing.T) { 1105 testStrings := []string{"", abcd, faces, commas} 1106 for _, s := range testStrings { 1107 reader := NewReader(s) 1108 if e := reader.UnreadRune(); e == nil { 1109 t.Errorf("Unreading %q at beginning: expected error", s) 1110 } 1111 res := "" 1112 for { 1113 r, z, e := reader.ReadRune() 1114 if e == io.EOF { 1115 break 1116 } 1117 if e != nil { 1118 t.Errorf("Reading %q: %s", s, e) 1119 break 1120 } 1121 res += string(r) 1122 // unread and read again 1123 e = reader.UnreadRune() 1124 if e != nil { 1125 t.Errorf("Unreading %q: %s", s, e) 1126 break 1127 } 1128 r1, z1, e := reader.ReadRune() 1129 if e != nil { 1130 t.Errorf("Reading %q after unreading: %s", s, e) 1131 break 1132 } 1133 if r1 != r { 1134 t.Errorf("Reading %q after unreading: want rune %q, got %q", s, r, r1) 1135 break 1136 } 1137 if z1 != z { 1138 t.Errorf("Reading %q after unreading: want size %d, got %d", s, z, z1) 1139 break 1140 } 1141 } 1142 if res != s { 1143 t.Errorf("Reader(%q).ReadRune() produced %q", s, res) 1144 } 1145 } 1146 } 1147 1148 var UnreadRuneErrorTests = []struct { 1149 name string 1150 f func(*Reader) 1151 }{ 1152 {"Read", func(r *Reader) { r.Read([]byte{0}) }}, 1153 {"ReadByte", func(r *Reader) { r.ReadByte() }}, 1154 {"UnreadRune", func(r *Reader) { r.UnreadRune() }}, 1155 {"Seek", func(r *Reader) { r.Seek(0, io.SeekCurrent) }}, 1156 {"WriteTo", func(r *Reader) { r.WriteTo(&bytes.Buffer{}) }}, 1157 } 1158 1159 func TestUnreadRuneError(t *testing.T) { 1160 for _, tt := range UnreadRuneErrorTests { 1161 reader := NewReader("0123456789") 1162 if _, _, err := reader.ReadRune(); err != nil { 1163 // should not happen 1164 t.Fatal(err) 1165 } 1166 tt.f(reader) 1167 err := reader.UnreadRune() 1168 if err == nil { 1169 t.Errorf("Unreading after %s: expected error", tt.name) 1170 } 1171 } 1172 } 1173 1174 var ReplaceTests = []struct { 1175 in string 1176 old, new string 1177 n int 1178 out string 1179 }{ 1180 {"hello", "l", "L", 0, "hello"}, 1181 {"hello", "l", "L", -1, "heLLo"}, 1182 {"hello", "x", "X", -1, "hello"}, 1183 {"", "x", "X", -1, ""}, 1184 {"radar", "r", "<r>", -1, "<r>ada<r>"}, 1185 {"", "", "<>", -1, "<>"}, 1186 {"banana", "a", "<>", -1, "b<>n<>n<>"}, 1187 {"banana", "a", "<>", 1, "b<>nana"}, 1188 {"banana", "a", "<>", 1000, "b<>n<>n<>"}, 1189 {"banana", "an", "<>", -1, "b<><>a"}, 1190 {"banana", "ana", "<>", -1, "b<>na"}, 1191 {"banana", "", "<>", -1, "<>b<>a<>n<>a<>n<>a<>"}, 1192 {"banana", "", "<>", 10, "<>b<>a<>n<>a<>n<>a<>"}, 1193 {"banana", "", "<>", 6, "<>b<>a<>n<>a<>n<>a"}, 1194 {"banana", "", "<>", 5, "<>b<>a<>n<>a<>na"}, 1195 {"banana", "", "<>", 1, "<>banana"}, 1196 {"banana", "a", "a", -1, "banana"}, 1197 {"banana", "a", "a", 1, "banana"}, 1198 {"☺☻☹", "", "<>", -1, "<>☺<>☻<>☹<>"}, 1199 } 1200 1201 func TestReplace(t *testing.T) { 1202 for _, tt := range ReplaceTests { 1203 if s := Replace(tt.in, tt.old, tt.new, tt.n); s != tt.out { 1204 t.Errorf("Replace(%q, %q, %q, %d) = %q, want %q", tt.in, tt.old, tt.new, tt.n, s, tt.out) 1205 } 1206 } 1207 } 1208 1209 var TitleTests = []struct { 1210 in, out string 1211 }{ 1212 {"", ""}, 1213 {"a", "A"}, 1214 {" aaa aaa aaa ", " Aaa Aaa Aaa "}, 1215 {" Aaa Aaa Aaa ", " Aaa Aaa Aaa "}, 1216 {"123a456", "123a456"}, 1217 {"double-blind", "Double-Blind"}, 1218 {"ÿøû", "Ÿøû"}, 1219 {"with_underscore", "With_underscore"}, 1220 {"unicode \xe2\x80\xa8 line separator", "Unicode \xe2\x80\xa8 Line Separator"}, 1221 } 1222 1223 func TestTitle(t *testing.T) { 1224 for _, tt := range TitleTests { 1225 if s := Title(tt.in); s != tt.out { 1226 t.Errorf("Title(%q) = %q, want %q", tt.in, s, tt.out) 1227 } 1228 } 1229 } 1230 1231 var ContainsTests = []struct { 1232 str, substr string 1233 expected bool 1234 }{ 1235 {"abc", "bc", true}, 1236 {"abc", "bcd", false}, 1237 {"abc", "", true}, 1238 {"", "a", false}, 1239 1240 // cases to cover code in runtime/asm_amd64.s:indexShortStr 1241 // 2-byte needle 1242 {"xxxxxx", "01", false}, 1243 {"01xxxx", "01", true}, 1244 {"xx01xx", "01", true}, 1245 {"xxxx01", "01", true}, 1246 {"01xxxxx"[1:], "01", false}, 1247 {"xxxxx01"[:6], "01", false}, 1248 // 3-byte needle 1249 {"xxxxxxx", "012", false}, 1250 {"012xxxx", "012", true}, 1251 {"xx012xx", "012", true}, 1252 {"xxxx012", "012", true}, 1253 {"012xxxxx"[1:], "012", false}, 1254 {"xxxxx012"[:7], "012", false}, 1255 // 4-byte needle 1256 {"xxxxxxxx", "0123", false}, 1257 {"0123xxxx", "0123", true}, 1258 {"xx0123xx", "0123", true}, 1259 {"xxxx0123", "0123", true}, 1260 {"0123xxxxx"[1:], "0123", false}, 1261 {"xxxxx0123"[:8], "0123", false}, 1262 // 5-7-byte needle 1263 {"xxxxxxxxx", "01234", false}, 1264 {"01234xxxx", "01234", true}, 1265 {"xx01234xx", "01234", true}, 1266 {"xxxx01234", "01234", true}, 1267 {"01234xxxxx"[1:], "01234", false}, 1268 {"xxxxx01234"[:9], "01234", false}, 1269 // 8-byte needle 1270 {"xxxxxxxxxxxx", "01234567", false}, 1271 {"01234567xxxx", "01234567", true}, 1272 {"xx01234567xx", "01234567", true}, 1273 {"xxxx01234567", "01234567", true}, 1274 {"01234567xxxxx"[1:], "01234567", false}, 1275 {"xxxxx01234567"[:12], "01234567", false}, 1276 // 9-15-byte needle 1277 {"xxxxxxxxxxxxx", "012345678", false}, 1278 {"012345678xxxx", "012345678", true}, 1279 {"xx012345678xx", "012345678", true}, 1280 {"xxxx012345678", "012345678", true}, 1281 {"012345678xxxxx"[1:], "012345678", false}, 1282 {"xxxxx012345678"[:13], "012345678", false}, 1283 // 16-byte needle 1284 {"xxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEF", false}, 1285 {"0123456789ABCDEFxxxx", "0123456789ABCDEF", true}, 1286 {"xx0123456789ABCDEFxx", "0123456789ABCDEF", true}, 1287 {"xxxx0123456789ABCDEF", "0123456789ABCDEF", true}, 1288 {"0123456789ABCDEFxxxxx"[1:], "0123456789ABCDEF", false}, 1289 {"xxxxx0123456789ABCDEF"[:20], "0123456789ABCDEF", false}, 1290 // 17-31-byte needle 1291 {"xxxxxxxxxxxxxxxxxxxxx", "0123456789ABCDEFG", false}, 1292 {"0123456789ABCDEFGxxxx", "0123456789ABCDEFG", true}, 1293 {"xx0123456789ABCDEFGxx", "0123456789ABCDEFG", true}, 1294 {"xxxx0123456789ABCDEFG", "0123456789ABCDEFG", true}, 1295 {"0123456789ABCDEFGxxxxx"[1:], "0123456789ABCDEFG", false}, 1296 {"xxxxx0123456789ABCDEFG"[:21], "0123456789ABCDEFG", false}, 1297 1298 // partial match cases 1299 {"xx01x", "012", false}, // 3 1300 {"xx0123x", "01234", false}, // 5-7 1301 {"xx01234567x", "012345678", false}, // 9-15 1302 {"xx0123456789ABCDEFx", "0123456789ABCDEFG", false}, // 17-31, issue 15679 1303 } 1304 1305 func TestContains(t *testing.T) { 1306 for _, ct := range ContainsTests { 1307 if Contains(ct.str, ct.substr) != ct.expected { 1308 t.Errorf("Contains(%s, %s) = %v, want %v", 1309 ct.str, ct.substr, !ct.expected, ct.expected) 1310 } 1311 } 1312 } 1313 1314 var ContainsAnyTests = []struct { 1315 str, substr string 1316 expected bool 1317 }{ 1318 {"", "", false}, 1319 {"", "a", false}, 1320 {"", "abc", false}, 1321 {"a", "", false}, 1322 {"a", "a", true}, 1323 {"aaa", "a", true}, 1324 {"abc", "xyz", false}, 1325 {"abc", "xcz", true}, 1326 {"a☺b☻c☹d", "uvw☻xyz", true}, 1327 {"aRegExp*", ".(|)*+?^$[]", true}, 1328 {dots + dots + dots, " ", false}, 1329 } 1330 1331 func TestContainsAny(t *testing.T) { 1332 for _, ct := range ContainsAnyTests { 1333 if ContainsAny(ct.str, ct.substr) != ct.expected { 1334 t.Errorf("ContainsAny(%s, %s) = %v, want %v", 1335 ct.str, ct.substr, !ct.expected, ct.expected) 1336 } 1337 } 1338 } 1339 1340 var ContainsRuneTests = []struct { 1341 str string 1342 r rune 1343 expected bool 1344 }{ 1345 {"", 'a', false}, 1346 {"a", 'a', true}, 1347 {"aaa", 'a', true}, 1348 {"abc", 'y', false}, 1349 {"abc", 'c', true}, 1350 {"a☺b☻c☹d", 'x', false}, 1351 {"a☺b☻c☹d", '☻', true}, 1352 {"aRegExp*", '*', true}, 1353 } 1354 1355 func TestContainsRune(t *testing.T) { 1356 for _, ct := range ContainsRuneTests { 1357 if ContainsRune(ct.str, ct.r) != ct.expected { 1358 t.Errorf("ContainsRune(%q, %q) = %v, want %v", 1359 ct.str, ct.r, !ct.expected, ct.expected) 1360 } 1361 } 1362 } 1363 1364 var EqualFoldTests = []struct { 1365 s, t string 1366 out bool 1367 }{ 1368 {"abc", "abc", true}, 1369 {"ABcd", "ABcd", true}, 1370 {"123abc", "123ABC", true}, 1371 {"αβδ", "ΑΒΔ", true}, 1372 {"abc", "xyz", false}, 1373 {"abc", "XYZ", false}, 1374 {"abcdefghijk", "abcdefghijX", false}, 1375 {"abcdefghijk", "abcdefghij\u212A", true}, 1376 {"abcdefghijK", "abcdefghij\u212A", true}, 1377 {"abcdefghijkz", "abcdefghij\u212Ay", false}, 1378 {"abcdefghijKz", "abcdefghij\u212Ay", false}, 1379 } 1380 1381 func TestEqualFold(t *testing.T) { 1382 for _, tt := range EqualFoldTests { 1383 if out := EqualFold(tt.s, tt.t); out != tt.out { 1384 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.s, tt.t, out, tt.out) 1385 } 1386 if out := EqualFold(tt.t, tt.s); out != tt.out { 1387 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.t, tt.s, out, tt.out) 1388 } 1389 } 1390 } 1391 1392 var CountTests = []struct { 1393 s, sep string 1394 num int 1395 }{ 1396 {"", "", 1}, 1397 {"", "notempty", 0}, 1398 {"notempty", "", 9}, 1399 {"smaller", "not smaller", 0}, 1400 {"12345678987654321", "6", 2}, 1401 {"611161116", "6", 3}, 1402 {"notequal", "NotEqual", 0}, 1403 {"equal", "equal", 1}, 1404 {"abc1231231123q", "123", 3}, 1405 {"11111", "11", 2}, 1406 } 1407 1408 func TestCount(t *testing.T) { 1409 for _, tt := range CountTests { 1410 if num := Count(tt.s, tt.sep); num != tt.num { 1411 t.Errorf("Count(\"%s\", \"%s\") = %d, want %d", tt.s, tt.sep, num, tt.num) 1412 } 1413 } 1414 } 1415 1416 func makeBenchInputHard() string { 1417 tokens := [...]string{ 1418 "<a>", "<p>", "<b>", "<strong>", 1419 "</a>", "</p>", "</b>", "</strong>", 1420 "hello", "world", 1421 } 1422 x := make([]byte, 0, 1<<20) 1423 for { 1424 i := rand.Intn(len(tokens)) 1425 if len(x)+len(tokens[i]) >= 1<<20 { 1426 break 1427 } 1428 x = append(x, tokens[i]...) 1429 } 1430 return string(x) 1431 } 1432 1433 var benchInputHard = makeBenchInputHard() 1434 1435 func benchmarkIndexHard(b *testing.B, sep string) { 1436 for i := 0; i < b.N; i++ { 1437 Index(benchInputHard, sep) 1438 } 1439 } 1440 1441 func benchmarkLastIndexHard(b *testing.B, sep string) { 1442 for i := 0; i < b.N; i++ { 1443 LastIndex(benchInputHard, sep) 1444 } 1445 } 1446 1447 func benchmarkCountHard(b *testing.B, sep string) { 1448 for i := 0; i < b.N; i++ { 1449 Count(benchInputHard, sep) 1450 } 1451 } 1452 1453 func BenchmarkIndexHard1(b *testing.B) { benchmarkIndexHard(b, "<>") } 1454 func BenchmarkIndexHard2(b *testing.B) { benchmarkIndexHard(b, "</pre>") } 1455 func BenchmarkIndexHard3(b *testing.B) { benchmarkIndexHard(b, "<b>hello world</b>") } 1456 func BenchmarkIndexHard4(b *testing.B) { 1457 benchmarkIndexHard(b, "<pre><b>hello</b><strong>world</strong></pre>") 1458 } 1459 1460 func BenchmarkLastIndexHard1(b *testing.B) { benchmarkLastIndexHard(b, "<>") } 1461 func BenchmarkLastIndexHard2(b *testing.B) { benchmarkLastIndexHard(b, "</pre>") } 1462 func BenchmarkLastIndexHard3(b *testing.B) { benchmarkLastIndexHard(b, "<b>hello world</b>") } 1463 1464 func BenchmarkCountHard1(b *testing.B) { benchmarkCountHard(b, "<>") } 1465 func BenchmarkCountHard2(b *testing.B) { benchmarkCountHard(b, "</pre>") } 1466 func BenchmarkCountHard3(b *testing.B) { benchmarkCountHard(b, "<b>hello world</b>") } 1467 1468 var benchInputTorture = Repeat("ABC", 1<<10) + "123" + Repeat("ABC", 1<<10) 1469 var benchNeedleTorture = Repeat("ABC", 1<<10+1) 1470 1471 func BenchmarkIndexTorture(b *testing.B) { 1472 for i := 0; i < b.N; i++ { 1473 Index(benchInputTorture, benchNeedleTorture) 1474 } 1475 } 1476 1477 func BenchmarkCountTorture(b *testing.B) { 1478 for i := 0; i < b.N; i++ { 1479 Count(benchInputTorture, benchNeedleTorture) 1480 } 1481 } 1482 1483 func BenchmarkCountTortureOverlapping(b *testing.B) { 1484 A := Repeat("ABC", 1<<20) 1485 B := Repeat("ABC", 1<<10) 1486 for i := 0; i < b.N; i++ { 1487 Count(A, B) 1488 } 1489 } 1490 1491 func BenchmarkCountByte(b *testing.B) { 1492 indexSizes := []int{10, 32, 4 << 10, 4 << 20, 64 << 20} 1493 benchStr := Repeat(benchmarkString, 1494 (indexSizes[len(indexSizes)-1]+len(benchmarkString)-1)/len(benchmarkString)) 1495 benchFunc := func(b *testing.B, benchStr string) { 1496 b.SetBytes(int64(len(benchStr))) 1497 for i := 0; i < b.N; i++ { 1498 Count(benchStr, "=") 1499 } 1500 } 1501 for _, size := range indexSizes { 1502 b.Run(fmt.Sprintf("%d", size), func(b *testing.B) { 1503 benchFunc(b, benchStr[:size]) 1504 }) 1505 } 1506 1507 } 1508 1509 var makeFieldsInput = func() string { 1510 x := make([]byte, 1<<20) 1511 // Input is ~10% space, ~10% 2-byte UTF-8, rest ASCII non-space. 1512 for i := range x { 1513 switch rand.Intn(10) { 1514 case 0: 1515 x[i] = ' ' 1516 case 1: 1517 if i > 0 && x[i-1] == 'x' { 1518 copy(x[i-1:], "χ") 1519 break 1520 } 1521 fallthrough 1522 default: 1523 x[i] = 'x' 1524 } 1525 } 1526 return string(x) 1527 } 1528 1529 var makeFieldsInputASCII = func() string { 1530 x := make([]byte, 1<<20) 1531 // Input is ~10% space, rest ASCII non-space. 1532 for i := range x { 1533 if rand.Intn(10) == 0 { 1534 x[i] = ' ' 1535 } else { 1536 x[i] = 'x' 1537 } 1538 } 1539 return string(x) 1540 } 1541 1542 var stringdata = []struct{ name, data string }{ 1543 {"ASCII", makeFieldsInputASCII()}, 1544 {"Mixed", makeFieldsInput()}, 1545 } 1546 1547 func BenchmarkFields(b *testing.B) { 1548 for _, sd := range stringdata { 1549 b.Run(sd.name, func(b *testing.B) { 1550 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1551 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1552 b.ReportAllocs() 1553 b.SetBytes(int64(j)) 1554 data := sd.data[:j] 1555 for i := 0; i < b.N; i++ { 1556 Fields(data) 1557 } 1558 }) 1559 } 1560 }) 1561 } 1562 } 1563 1564 func BenchmarkFieldsFunc(b *testing.B) { 1565 for _, sd := range stringdata { 1566 b.Run(sd.name, func(b *testing.B) { 1567 for j := 1 << 4; j <= 1<<20; j <<= 4 { 1568 b.Run(fmt.Sprintf("%d", j), func(b *testing.B) { 1569 b.ReportAllocs() 1570 b.SetBytes(int64(j)) 1571 data := sd.data[:j] 1572 for i := 0; i < b.N; i++ { 1573 FieldsFunc(data, unicode.IsSpace) 1574 } 1575 }) 1576 } 1577 }) 1578 } 1579 } 1580 1581 func BenchmarkSplitEmptySeparator(b *testing.B) { 1582 for i := 0; i < b.N; i++ { 1583 Split(benchInputHard, "") 1584 } 1585 } 1586 1587 func BenchmarkSplitSingleByteSeparator(b *testing.B) { 1588 for i := 0; i < b.N; i++ { 1589 Split(benchInputHard, "/") 1590 } 1591 } 1592 1593 func BenchmarkSplitMultiByteSeparator(b *testing.B) { 1594 for i := 0; i < b.N; i++ { 1595 Split(benchInputHard, "hello") 1596 } 1597 } 1598 1599 func BenchmarkSplitNSingleByteSeparator(b *testing.B) { 1600 for i := 0; i < b.N; i++ { 1601 SplitN(benchInputHard, "/", 10) 1602 } 1603 } 1604 1605 func BenchmarkSplitNMultiByteSeparator(b *testing.B) { 1606 for i := 0; i < b.N; i++ { 1607 SplitN(benchInputHard, "hello", 10) 1608 } 1609 } 1610 1611 func BenchmarkRepeat(b *testing.B) { 1612 for i := 0; i < b.N; i++ { 1613 Repeat("-", 80) 1614 } 1615 } 1616 1617 func BenchmarkIndexAnyASCII(b *testing.B) { 1618 x := Repeat("#", 4096) // Never matches set 1619 cs := "0123456789abcdef" 1620 for k := 1; k <= 4096; k <<= 4 { 1621 for j := 1; j <= 16; j <<= 1 { 1622 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1623 for i := 0; i < b.N; i++ { 1624 IndexAny(x[:k], cs[:j]) 1625 } 1626 }) 1627 } 1628 } 1629 } 1630 1631 func BenchmarkTrimASCII(b *testing.B) { 1632 cs := "0123456789abcdef" 1633 for k := 1; k <= 4096; k <<= 4 { 1634 for j := 1; j <= 16; j <<= 1 { 1635 b.Run(fmt.Sprintf("%d:%d", k, j), func(b *testing.B) { 1636 x := Repeat(cs[:j], k) // Always matches set 1637 for i := 0; i < b.N; i++ { 1638 Trim(x[:k], cs[:j]) 1639 } 1640 }) 1641 } 1642 } 1643 }