github.com/ader1990/go@v0.0.0-20140630135419-8c24447fa791/src/pkg/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 "io" 10 "math/rand" 11 "reflect" 12 . "strings" 13 "testing" 14 "unicode" 15 "unicode/utf8" 16 "unsafe" 17 ) 18 19 func eq(a, b []string) bool { 20 if len(a) != len(b) { 21 return false 22 } 23 for i := 0; i < len(a); i++ { 24 if a[i] != b[i] { 25 return false 26 } 27 } 28 return true 29 } 30 31 var abcd = "abcd" 32 var faces = "☺☻☹" 33 var commas = "1,2,3,4" 34 var dots = "1....2....3....4" 35 36 type IndexTest struct { 37 s string 38 sep string 39 out int 40 } 41 42 var indexTests = []IndexTest{ 43 {"", "", 0}, 44 {"", "a", -1}, 45 {"", "foo", -1}, 46 {"fo", "foo", -1}, 47 {"foo", "foo", 0}, 48 {"oofofoofooo", "f", 2}, 49 {"oofofoofooo", "foo", 4}, 50 {"barfoobarfoo", "foo", 3}, 51 {"foo", "", 0}, 52 {"foo", "o", 1}, 53 {"abcABCabc", "A", 3}, 54 // cases with one byte strings - test special case in Index() 55 {"", "a", -1}, 56 {"x", "a", -1}, 57 {"x", "x", 0}, 58 {"abc", "a", 0}, 59 {"abc", "b", 1}, 60 {"abc", "c", 2}, 61 {"abc", "x", -1}, 62 } 63 64 var lastIndexTests = []IndexTest{ 65 {"", "", 0}, 66 {"", "a", -1}, 67 {"", "foo", -1}, 68 {"fo", "foo", -1}, 69 {"foo", "foo", 0}, 70 {"foo", "f", 0}, 71 {"oofofoofooo", "f", 7}, 72 {"oofofoofooo", "foo", 7}, 73 {"barfoobarfoo", "foo", 9}, 74 {"foo", "", 3}, 75 {"foo", "o", 2}, 76 {"abcABCabc", "A", 3}, 77 {"abcABCabc", "a", 6}, 78 } 79 80 var indexAnyTests = []IndexTest{ 81 {"", "", -1}, 82 {"", "a", -1}, 83 {"", "abc", -1}, 84 {"a", "", -1}, 85 {"a", "a", 0}, 86 {"aaa", "a", 0}, 87 {"abc", "xyz", -1}, 88 {"abc", "xcz", 2}, 89 {"a☺b☻c☹d", "uvw☻xyz", 2 + len("☺")}, 90 {"aRegExp*", ".(|)*+?^$[]", 7}, 91 {dots + dots + dots, " ", -1}, 92 } 93 var lastIndexAnyTests = []IndexTest{ 94 {"", "", -1}, 95 {"", "a", -1}, 96 {"", "abc", -1}, 97 {"a", "", -1}, 98 {"a", "a", 0}, 99 {"aaa", "a", 2}, 100 {"abc", "xyz", -1}, 101 {"abc", "ab", 1}, 102 {"a☺b☻c☹d", "uvw☻xyz", 2 + len("☺")}, 103 {"a.RegExp*", ".(|)*+?^$[]", 8}, 104 {dots + dots + dots, " ", -1}, 105 } 106 107 // Execute f on each test case. funcName should be the name of f; it's used 108 // in failure reports. 109 func runIndexTests(t *testing.T, f func(s, sep string) int, funcName string, testCases []IndexTest) { 110 for _, test := range testCases { 111 actual := f(test.s, test.sep) 112 if actual != test.out { 113 t.Errorf("%s(%q,%q) = %v; want %v", funcName, test.s, test.sep, actual, test.out) 114 } 115 } 116 } 117 118 func TestIndex(t *testing.T) { runIndexTests(t, Index, "Index", indexTests) } 119 func TestLastIndex(t *testing.T) { runIndexTests(t, LastIndex, "LastIndex", lastIndexTests) } 120 func TestIndexAny(t *testing.T) { runIndexTests(t, IndexAny, "IndexAny", indexAnyTests) } 121 func TestLastIndexAny(t *testing.T) { runIndexTests(t, LastIndexAny, "LastIndexAny", lastIndexAnyTests) } 122 123 var indexRuneTests = []struct { 124 s string 125 rune rune 126 out int 127 }{ 128 {"a A x", 'A', 2}, 129 {"some_text=some_value", '=', 9}, 130 {"☺a", 'a', 3}, 131 {"a☻☺b", '☺', 4}, 132 } 133 134 func TestIndexRune(t *testing.T) { 135 for _, test := range indexRuneTests { 136 if actual := IndexRune(test.s, test.rune); actual != test.out { 137 t.Errorf("IndexRune(%q,%d)= %v; want %v", test.s, test.rune, actual, test.out) 138 } 139 } 140 } 141 142 const benchmarkString = "some_text=some☺value" 143 144 func BenchmarkIndexRune(b *testing.B) { 145 if got := IndexRune(benchmarkString, '☺'); got != 14 { 146 b.Fatalf("wrong index: expected 14, got=%d", got) 147 } 148 for i := 0; i < b.N; i++ { 149 IndexRune(benchmarkString, '☺') 150 } 151 } 152 153 func BenchmarkIndexRuneFastPath(b *testing.B) { 154 if got := IndexRune(benchmarkString, 'v'); got != 17 { 155 b.Fatalf("wrong index: expected 17, got=%d", got) 156 } 157 for i := 0; i < b.N; i++ { 158 IndexRune(benchmarkString, 'v') 159 } 160 } 161 162 func BenchmarkIndex(b *testing.B) { 163 if got := Index(benchmarkString, "v"); got != 17 { 164 b.Fatalf("wrong index: expected 17, got=%d", got) 165 } 166 for i := 0; i < b.N; i++ { 167 Index(benchmarkString, "v") 168 } 169 } 170 171 func BenchmarkIndexByte(b *testing.B) { 172 if got := IndexByte(benchmarkString, 'v'); got != 17 { 173 b.Fatalf("wrong index: expected 17, got=%d", got) 174 } 175 for i := 0; i < b.N; i++ { 176 IndexByte(benchmarkString, 'v') 177 } 178 } 179 180 var explodetests = []struct { 181 s string 182 n int 183 a []string 184 }{ 185 {"", -1, []string{}}, 186 {abcd, 4, []string{"a", "b", "c", "d"}}, 187 {faces, 3, []string{"☺", "☻", "☹"}}, 188 {abcd, 2, []string{"a", "bcd"}}, 189 } 190 191 func TestExplode(t *testing.T) { 192 for _, tt := range explodetests { 193 a := SplitN(tt.s, "", tt.n) 194 if !eq(a, tt.a) { 195 t.Errorf("explode(%q, %d) = %v; want %v", tt.s, tt.n, a, tt.a) 196 continue 197 } 198 s := Join(a, "") 199 if s != tt.s { 200 t.Errorf(`Join(explode(%q, %d), "") = %q`, tt.s, tt.n, s) 201 } 202 } 203 } 204 205 type SplitTest struct { 206 s string 207 sep string 208 n int 209 a []string 210 } 211 212 var splittests = []SplitTest{ 213 {abcd, "a", 0, nil}, 214 {abcd, "a", -1, []string{"", "bcd"}}, 215 {abcd, "z", -1, []string{"abcd"}}, 216 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 217 {commas, ",", -1, []string{"1", "2", "3", "4"}}, 218 {dots, "...", -1, []string{"1", ".2", ".3", ".4"}}, 219 {faces, "☹", -1, []string{"☺☻", ""}}, 220 {faces, "~", -1, []string{faces}}, 221 {faces, "", -1, []string{"☺", "☻", "☹"}}, 222 {"1 2 3 4", " ", 3, []string{"1", "2", "3 4"}}, 223 {"1 2", " ", 3, []string{"1", "2"}}, 224 {"123", "", 2, []string{"1", "23"}}, 225 {"123", "", 17, []string{"1", "2", "3"}}, 226 } 227 228 func TestSplit(t *testing.T) { 229 for _, tt := range splittests { 230 a := SplitN(tt.s, tt.sep, tt.n) 231 if !eq(a, tt.a) { 232 t.Errorf("Split(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, a, tt.a) 233 continue 234 } 235 if tt.n == 0 { 236 continue 237 } 238 s := Join(a, tt.sep) 239 if s != tt.s { 240 t.Errorf("Join(Split(%q, %q, %d), %q) = %q", tt.s, tt.sep, tt.n, tt.sep, s) 241 } 242 if tt.n < 0 { 243 b := Split(tt.s, tt.sep) 244 if !reflect.DeepEqual(a, b) { 245 t.Errorf("Split disagrees with SplitN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 246 } 247 } 248 } 249 } 250 251 var splitaftertests = []SplitTest{ 252 {abcd, "a", -1, []string{"a", "bcd"}}, 253 {abcd, "z", -1, []string{"abcd"}}, 254 {abcd, "", -1, []string{"a", "b", "c", "d"}}, 255 {commas, ",", -1, []string{"1,", "2,", "3,", "4"}}, 256 {dots, "...", -1, []string{"1...", ".2...", ".3...", ".4"}}, 257 {faces, "☹", -1, []string{"☺☻☹", ""}}, 258 {faces, "~", -1, []string{faces}}, 259 {faces, "", -1, []string{"☺", "☻", "☹"}}, 260 {"1 2 3 4", " ", 3, []string{"1 ", "2 ", "3 4"}}, 261 {"1 2 3", " ", 3, []string{"1 ", "2 ", "3"}}, 262 {"1 2", " ", 3, []string{"1 ", "2"}}, 263 {"123", "", 2, []string{"1", "23"}}, 264 {"123", "", 17, []string{"1", "2", "3"}}, 265 } 266 267 func TestSplitAfter(t *testing.T) { 268 for _, tt := range splitaftertests { 269 a := SplitAfterN(tt.s, tt.sep, tt.n) 270 if !eq(a, tt.a) { 271 t.Errorf(`Split(%q, %q, %d) = %v; want %v`, tt.s, tt.sep, tt.n, a, tt.a) 272 continue 273 } 274 s := Join(a, "") 275 if s != tt.s { 276 t.Errorf(`Join(Split(%q, %q, %d), %q) = %q`, tt.s, tt.sep, tt.n, tt.sep, s) 277 } 278 if tt.n < 0 { 279 b := SplitAfter(tt.s, tt.sep) 280 if !reflect.DeepEqual(a, b) { 281 t.Errorf("SplitAfter disagrees with SplitAfterN(%q, %q, %d) = %v; want %v", tt.s, tt.sep, tt.n, b, a) 282 } 283 } 284 } 285 } 286 287 type FieldsTest struct { 288 s string 289 a []string 290 } 291 292 var fieldstests = []FieldsTest{ 293 {"", []string{}}, 294 {" ", []string{}}, 295 {" \t ", []string{}}, 296 {" abc ", []string{"abc"}}, 297 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 298 {"1 2 3 4", []string{"1", "2", "3", "4"}}, 299 {"1\t\t2\t\t3\t4", []string{"1", "2", "3", "4"}}, 300 {"1\u20002\u20013\u20024", []string{"1", "2", "3", "4"}}, 301 {"\u2000\u2001\u2002", []string{}}, 302 {"\n™\t™\n", []string{"™", "™"}}, 303 {faces, []string{faces}}, 304 } 305 306 func TestFields(t *testing.T) { 307 for _, tt := range fieldstests { 308 a := Fields(tt.s) 309 if !eq(a, tt.a) { 310 t.Errorf("Fields(%q) = %v; want %v", tt.s, a, tt.a) 311 continue 312 } 313 } 314 } 315 316 var FieldsFuncTests = []FieldsTest{ 317 {"", []string{}}, 318 {"XX", []string{}}, 319 {"XXhiXXX", []string{"hi"}}, 320 {"aXXbXXXcX", []string{"a", "b", "c"}}, 321 } 322 323 func TestFieldsFunc(t *testing.T) { 324 for _, tt := range fieldstests { 325 a := FieldsFunc(tt.s, unicode.IsSpace) 326 if !eq(a, tt.a) { 327 t.Errorf("FieldsFunc(%q, unicode.IsSpace) = %v; want %v", tt.s, a, tt.a) 328 continue 329 } 330 } 331 pred := func(c rune) bool { return c == 'X' } 332 for _, tt := range FieldsFuncTests { 333 a := FieldsFunc(tt.s, pred) 334 if !eq(a, tt.a) { 335 t.Errorf("FieldsFunc(%q) = %v, want %v", tt.s, a, tt.a) 336 } 337 } 338 } 339 340 // Test case for any function which accepts and returns a single string. 341 type StringTest struct { 342 in, out string 343 } 344 345 // Execute f on each test case. funcName should be the name of f; it's used 346 // in failure reports. 347 func runStringTests(t *testing.T, f func(string) string, funcName string, testCases []StringTest) { 348 for _, tc := range testCases { 349 actual := f(tc.in) 350 if actual != tc.out { 351 t.Errorf("%s(%q) = %q; want %q", funcName, tc.in, actual, tc.out) 352 } 353 } 354 } 355 356 var upperTests = []StringTest{ 357 {"", ""}, 358 {"abc", "ABC"}, 359 {"AbC123", "ABC123"}, 360 {"azAZ09_", "AZAZ09_"}, 361 {"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"}, // grows one byte per char 362 } 363 364 var lowerTests = []StringTest{ 365 {"", ""}, 366 {"abc", "abc"}, 367 {"AbC123", "abc123"}, 368 {"azAZ09_", "azaz09_"}, 369 {"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"}, // shrinks one byte per char 370 } 371 372 const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000" 373 374 var trimSpaceTests = []StringTest{ 375 {"", ""}, 376 {"abc", "abc"}, 377 {space + "abc" + space, "abc"}, 378 {" ", ""}, 379 {" \t\r\n \t\t\r\r\n\n ", ""}, 380 {" \t\r\n x\t\t\r\r\n\n ", "x"}, 381 {" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"}, 382 {"1 \t\r\n2", "1 \t\r\n2"}, 383 {" x\x80", "x\x80"}, 384 {" x\xc0", "x\xc0"}, 385 {"x \xc0\xc0 ", "x \xc0\xc0"}, 386 {"x \xc0", "x \xc0"}, 387 {"x \xc0 ", "x \xc0"}, 388 {"x \xc0\xc0 ", "x \xc0\xc0"}, 389 {"x ☺\xc0\xc0 ", "x ☺\xc0\xc0"}, 390 {"x ☺ ", "x ☺"}, 391 } 392 393 func tenRunes(ch rune) string { 394 r := make([]rune, 10) 395 for i := range r { 396 r[i] = ch 397 } 398 return string(r) 399 } 400 401 // User-defined self-inverse mapping function 402 func rot13(r rune) rune { 403 step := rune(13) 404 if r >= 'a' && r <= 'z' { 405 return ((r - 'a' + step) % 26) + 'a' 406 } 407 if r >= 'A' && r <= 'Z' { 408 return ((r - 'A' + step) % 26) + 'A' 409 } 410 return r 411 } 412 413 func TestMap(t *testing.T) { 414 // Run a couple of awful growth/shrinkage tests 415 a := tenRunes('a') 416 // 1. Grow. This triggers two reallocations in Map. 417 maxRune := func(rune) rune { return unicode.MaxRune } 418 m := Map(maxRune, a) 419 expect := tenRunes(unicode.MaxRune) 420 if m != expect { 421 t.Errorf("growing: expected %q got %q", expect, m) 422 } 423 424 // 2. Shrink 425 minRune := func(rune) rune { return 'a' } 426 m = Map(minRune, tenRunes(unicode.MaxRune)) 427 expect = a 428 if m != expect { 429 t.Errorf("shrinking: expected %q got %q", expect, m) 430 } 431 432 // 3. Rot13 433 m = Map(rot13, "a to zed") 434 expect = "n gb mrq" 435 if m != expect { 436 t.Errorf("rot13: expected %q got %q", expect, m) 437 } 438 439 // 4. Rot13^2 440 m = Map(rot13, Map(rot13, "a to zed")) 441 expect = "a to zed" 442 if m != expect { 443 t.Errorf("rot13: expected %q got %q", expect, m) 444 } 445 446 // 5. Drop 447 dropNotLatin := func(r rune) rune { 448 if unicode.Is(unicode.Latin, r) { 449 return r 450 } 451 return -1 452 } 453 m = Map(dropNotLatin, "Hello, 세계") 454 expect = "Hello" 455 if m != expect { 456 t.Errorf("drop: expected %q got %q", expect, m) 457 } 458 459 // 6. Identity 460 identity := func(r rune) rune { 461 return r 462 } 463 orig := "Input string that we expect not to be copied." 464 m = Map(identity, orig) 465 if (*reflect.StringHeader)(unsafe.Pointer(&orig)).Data != 466 (*reflect.StringHeader)(unsafe.Pointer(&m)).Data { 467 t.Error("unexpected copy during identity map") 468 } 469 } 470 471 func TestToUpper(t *testing.T) { runStringTests(t, ToUpper, "ToUpper", upperTests) } 472 473 func TestToLower(t *testing.T) { runStringTests(t, ToLower, "ToLower", lowerTests) } 474 475 func BenchmarkMapNoChanges(b *testing.B) { 476 identity := func(r rune) rune { 477 return r 478 } 479 for i := 0; i < b.N; i++ { 480 Map(identity, "Some string that won't be modified.") 481 } 482 } 483 484 func TestSpecialCase(t *testing.T) { 485 lower := "abcçdefgğhıijklmnoöprsştuüvyz" 486 upper := "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ" 487 u := ToUpperSpecial(unicode.TurkishCase, upper) 488 if u != upper { 489 t.Errorf("Upper(upper) is %s not %s", u, upper) 490 } 491 u = ToUpperSpecial(unicode.TurkishCase, lower) 492 if u != upper { 493 t.Errorf("Upper(lower) is %s not %s", u, upper) 494 } 495 l := ToLowerSpecial(unicode.TurkishCase, lower) 496 if l != lower { 497 t.Errorf("Lower(lower) is %s not %s", l, lower) 498 } 499 l = ToLowerSpecial(unicode.TurkishCase, upper) 500 if l != lower { 501 t.Errorf("Lower(upper) is %s not %s", l, lower) 502 } 503 } 504 505 func TestTrimSpace(t *testing.T) { runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests) } 506 507 var trimTests = []struct { 508 f string 509 in, arg, out string 510 }{ 511 {"Trim", "abba", "a", "bb"}, 512 {"Trim", "abba", "ab", ""}, 513 {"TrimLeft", "abba", "ab", ""}, 514 {"TrimRight", "abba", "ab", ""}, 515 {"TrimLeft", "abba", "a", "bba"}, 516 {"TrimRight", "abba", "a", "abb"}, 517 {"Trim", "<tag>", "<>", "tag"}, 518 {"Trim", "* listitem", " *", "listitem"}, 519 {"Trim", `"quote"`, `"`, "quote"}, 520 {"Trim", "\u2C6F\u2C6F\u0250\u0250\u2C6F\u2C6F", "\u2C6F", "\u0250\u0250"}, 521 //empty string tests 522 {"Trim", "abba", "", "abba"}, 523 {"Trim", "", "123", ""}, 524 {"Trim", "", "", ""}, 525 {"TrimLeft", "abba", "", "abba"}, 526 {"TrimLeft", "", "123", ""}, 527 {"TrimLeft", "", "", ""}, 528 {"TrimRight", "abba", "", "abba"}, 529 {"TrimRight", "", "123", ""}, 530 {"TrimRight", "", "", ""}, 531 {"TrimRight", "☺\xc0", "☺", "☺\xc0"}, 532 {"TrimPrefix", "aabb", "a", "abb"}, 533 {"TrimPrefix", "aabb", "b", "aabb"}, 534 {"TrimSuffix", "aabb", "a", "aabb"}, 535 {"TrimSuffix", "aabb", "b", "aab"}, 536 } 537 538 func TestTrim(t *testing.T) { 539 for _, tc := range trimTests { 540 name := tc.f 541 var f func(string, string) string 542 switch name { 543 case "Trim": 544 f = Trim 545 case "TrimLeft": 546 f = TrimLeft 547 case "TrimRight": 548 f = TrimRight 549 case "TrimPrefix": 550 f = TrimPrefix 551 case "TrimSuffix": 552 f = TrimSuffix 553 default: 554 t.Errorf("Undefined trim function %s", name) 555 } 556 actual := f(tc.in, tc.arg) 557 if actual != tc.out { 558 t.Errorf("%s(%q, %q) = %q; want %q", name, tc.in, tc.arg, actual, tc.out) 559 } 560 } 561 } 562 563 type predicate struct { 564 f func(rune) bool 565 name string 566 } 567 568 var isSpace = predicate{unicode.IsSpace, "IsSpace"} 569 var isDigit = predicate{unicode.IsDigit, "IsDigit"} 570 var isUpper = predicate{unicode.IsUpper, "IsUpper"} 571 var isValidRune = predicate{ 572 func(r rune) bool { 573 return r != utf8.RuneError 574 }, 575 "IsValidRune", 576 } 577 578 func not(p predicate) predicate { 579 return predicate{ 580 func(r rune) bool { 581 return !p.f(r) 582 }, 583 "not " + p.name, 584 } 585 } 586 587 var trimFuncTests = []struct { 588 f predicate 589 in, out string 590 }{ 591 {isSpace, space + " hello " + space, "hello"}, 592 {isDigit, "\u0e50\u0e5212hello34\u0e50\u0e51", "hello"}, 593 {isUpper, "\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", "hello"}, 594 {not(isSpace), "hello" + space + "hello", space}, 595 {not(isDigit), "hello\u0e50\u0e521234\u0e50\u0e51helo", "\u0e50\u0e521234\u0e50\u0e51"}, 596 {isValidRune, "ab\xc0a\xc0cd", "\xc0a\xc0"}, 597 {not(isValidRune), "\xc0a\xc0", "a"}, 598 } 599 600 func TestTrimFunc(t *testing.T) { 601 for _, tc := range trimFuncTests { 602 actual := TrimFunc(tc.in, tc.f.f) 603 if actual != tc.out { 604 t.Errorf("TrimFunc(%q, %q) = %q; want %q", tc.in, tc.f.name, actual, tc.out) 605 } 606 } 607 } 608 609 var indexFuncTests = []struct { 610 in string 611 f predicate 612 first, last int 613 }{ 614 {"", isValidRune, -1, -1}, 615 {"abc", isDigit, -1, -1}, 616 {"0123", isDigit, 0, 3}, 617 {"a1b", isDigit, 1, 1}, 618 {space, isSpace, 0, len(space) - 3}, // last rune in space is 3 bytes 619 {"\u0e50\u0e5212hello34\u0e50\u0e51", isDigit, 0, 18}, 620 {"\u2C6F\u2C6F\u2C6F\u2C6FABCDhelloEF\u2C6F\u2C6FGH\u2C6F\u2C6F", isUpper, 0, 34}, 621 {"12\u0e50\u0e52hello34\u0e50\u0e51", not(isDigit), 8, 12}, 622 623 // tests of invalid UTF-8 624 {"\x801", isDigit, 1, 1}, 625 {"\x80abc", isDigit, -1, -1}, 626 {"\xc0a\xc0", isValidRune, 1, 1}, 627 {"\xc0a\xc0", not(isValidRune), 0, 2}, 628 {"\xc0☺\xc0", not(isValidRune), 0, 4}, 629 {"\xc0☺\xc0\xc0", not(isValidRune), 0, 5}, 630 {"ab\xc0a\xc0cd", not(isValidRune), 2, 4}, 631 {"a\xe0\x80cd", not(isValidRune), 1, 2}, 632 {"\x80\x80\x80\x80", not(isValidRune), 0, 3}, 633 } 634 635 func TestIndexFunc(t *testing.T) { 636 for _, tc := range indexFuncTests { 637 first := IndexFunc(tc.in, tc.f.f) 638 if first != tc.first { 639 t.Errorf("IndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, first, tc.first) 640 } 641 last := LastIndexFunc(tc.in, tc.f.f) 642 if last != tc.last { 643 t.Errorf("LastIndexFunc(%q, %s) = %d; want %d", tc.in, tc.f.name, last, tc.last) 644 } 645 } 646 } 647 648 func equal(m string, s1, s2 string, t *testing.T) bool { 649 if s1 == s2 { 650 return true 651 } 652 e1 := Split(s1, "") 653 e2 := Split(s2, "") 654 for i, c1 := range e1 { 655 if i >= len(e2) { 656 break 657 } 658 r1, _ := utf8.DecodeRuneInString(c1) 659 r2, _ := utf8.DecodeRuneInString(e2[i]) 660 if r1 != r2 { 661 t.Errorf("%s diff at %d: U+%04X U+%04X", m, i, r1, r2) 662 } 663 } 664 return false 665 } 666 667 func TestCaseConsistency(t *testing.T) { 668 // Make a string of all the runes. 669 numRunes := int(unicode.MaxRune + 1) 670 if testing.Short() { 671 numRunes = 1000 672 } 673 a := make([]rune, numRunes) 674 for i := range a { 675 a[i] = rune(i) 676 } 677 s := string(a) 678 // convert the cases. 679 upper := ToUpper(s) 680 lower := ToLower(s) 681 682 // Consistency checks 683 if n := utf8.RuneCountInString(upper); n != numRunes { 684 t.Error("rune count wrong in upper:", n) 685 } 686 if n := utf8.RuneCountInString(lower); n != numRunes { 687 t.Error("rune count wrong in lower:", n) 688 } 689 if !equal("ToUpper(upper)", ToUpper(upper), upper, t) { 690 t.Error("ToUpper(upper) consistency fail") 691 } 692 if !equal("ToLower(lower)", ToLower(lower), lower, t) { 693 t.Error("ToLower(lower) consistency fail") 694 } 695 /* 696 These fail because of non-one-to-oneness of the data, such as multiple 697 upper case 'I' mapping to 'i'. We comment them out but keep them for 698 interest. 699 For instance: CAPITAL LETTER I WITH DOT ABOVE: 700 unicode.ToUpper(unicode.ToLower('\u0130')) != '\u0130' 701 702 if !equal("ToUpper(lower)", ToUpper(lower), upper, t) { 703 t.Error("ToUpper(lower) consistency fail"); 704 } 705 if !equal("ToLower(upper)", ToLower(upper), lower, t) { 706 t.Error("ToLower(upper) consistency fail"); 707 } 708 */ 709 } 710 711 var RepeatTests = []struct { 712 in, out string 713 count int 714 }{ 715 {"", "", 0}, 716 {"", "", 1}, 717 {"", "", 2}, 718 {"-", "", 0}, 719 {"-", "-", 1}, 720 {"-", "----------", 10}, 721 {"abc ", "abc abc abc ", 3}, 722 } 723 724 func TestRepeat(t *testing.T) { 725 for _, tt := range RepeatTests { 726 a := Repeat(tt.in, tt.count) 727 if !equal("Repeat(s)", a, tt.out, t) { 728 t.Errorf("Repeat(%v, %d) = %v; want %v", tt.in, tt.count, a, tt.out) 729 continue 730 } 731 } 732 } 733 734 func runesEqual(a, b []rune) bool { 735 if len(a) != len(b) { 736 return false 737 } 738 for i, r := range a { 739 if r != b[i] { 740 return false 741 } 742 } 743 return true 744 } 745 746 var RunesTests = []struct { 747 in string 748 out []rune 749 lossy bool 750 }{ 751 {"", []rune{}, false}, 752 {" ", []rune{32}, false}, 753 {"ABC", []rune{65, 66, 67}, false}, 754 {"abc", []rune{97, 98, 99}, false}, 755 {"\u65e5\u672c\u8a9e", []rune{26085, 26412, 35486}, false}, 756 {"ab\x80c", []rune{97, 98, 0xFFFD, 99}, true}, 757 {"ab\xc0c", []rune{97, 98, 0xFFFD, 99}, true}, 758 } 759 760 func TestRunes(t *testing.T) { 761 for _, tt := range RunesTests { 762 a := []rune(tt.in) 763 if !runesEqual(a, tt.out) { 764 t.Errorf("[]rune(%q) = %v; want %v", tt.in, a, tt.out) 765 continue 766 } 767 if !tt.lossy { 768 // can only test reassembly if we didn't lose information 769 s := string(a) 770 if s != tt.in { 771 t.Errorf("string([]rune(%q)) = %x; want %x", tt.in, s, tt.in) 772 } 773 } 774 } 775 } 776 777 func TestReadByte(t *testing.T) { 778 testStrings := []string{"", abcd, faces, commas} 779 for _, s := range testStrings { 780 reader := NewReader(s) 781 if e := reader.UnreadByte(); e == nil { 782 t.Errorf("Unreading %q at beginning: expected error", s) 783 } 784 var res bytes.Buffer 785 for { 786 b, e := reader.ReadByte() 787 if e == io.EOF { 788 break 789 } 790 if e != nil { 791 t.Errorf("Reading %q: %s", s, e) 792 break 793 } 794 res.WriteByte(b) 795 // unread and read again 796 e = reader.UnreadByte() 797 if e != nil { 798 t.Errorf("Unreading %q: %s", s, e) 799 break 800 } 801 b1, e := reader.ReadByte() 802 if e != nil { 803 t.Errorf("Reading %q after unreading: %s", s, e) 804 break 805 } 806 if b1 != b { 807 t.Errorf("Reading %q after unreading: want byte %q, got %q", s, b, b1) 808 break 809 } 810 } 811 if res.String() != s { 812 t.Errorf("Reader(%q).ReadByte() produced %q", s, res.String()) 813 } 814 } 815 } 816 817 func TestReadRune(t *testing.T) { 818 testStrings := []string{"", abcd, faces, commas} 819 for _, s := range testStrings { 820 reader := NewReader(s) 821 if e := reader.UnreadRune(); e == nil { 822 t.Errorf("Unreading %q at beginning: expected error", s) 823 } 824 res := "" 825 for { 826 r, z, e := reader.ReadRune() 827 if e == io.EOF { 828 break 829 } 830 if e != nil { 831 t.Errorf("Reading %q: %s", s, e) 832 break 833 } 834 res += string(r) 835 // unread and read again 836 e = reader.UnreadRune() 837 if e != nil { 838 t.Errorf("Unreading %q: %s", s, e) 839 break 840 } 841 r1, z1, e := reader.ReadRune() 842 if e != nil { 843 t.Errorf("Reading %q after unreading: %s", s, e) 844 break 845 } 846 if r1 != r { 847 t.Errorf("Reading %q after unreading: want rune %q, got %q", s, r, r1) 848 break 849 } 850 if z1 != z { 851 t.Errorf("Reading %q after unreading: want size %d, got %d", s, z, z1) 852 break 853 } 854 } 855 if res != s { 856 t.Errorf("Reader(%q).ReadRune() produced %q", s, res) 857 } 858 } 859 } 860 861 var UnreadRuneErrorTests = []struct { 862 name string 863 f func(*Reader) 864 }{ 865 {"Read", func(r *Reader) { r.Read([]byte{0}) }}, 866 {"ReadByte", func(r *Reader) { r.ReadByte() }}, 867 {"UnreadRune", func(r *Reader) { r.UnreadRune() }}, 868 {"Seek", func(r *Reader) { r.Seek(0, 1) }}, 869 {"WriteTo", func(r *Reader) { r.WriteTo(&bytes.Buffer{}) }}, 870 } 871 872 func TestUnreadRuneError(t *testing.T) { 873 for _, tt := range UnreadRuneErrorTests { 874 reader := NewReader("0123456789") 875 if _, _, err := reader.ReadRune(); err != nil { 876 // should not happen 877 t.Fatal(err) 878 } 879 tt.f(reader) 880 err := reader.UnreadRune() 881 if err == nil { 882 t.Errorf("Unreading after %s: expected error", tt.name) 883 } 884 } 885 } 886 887 var ReplaceTests = []struct { 888 in string 889 old, new string 890 n int 891 out string 892 }{ 893 {"hello", "l", "L", 0, "hello"}, 894 {"hello", "l", "L", -1, "heLLo"}, 895 {"hello", "x", "X", -1, "hello"}, 896 {"", "x", "X", -1, ""}, 897 {"radar", "r", "<r>", -1, "<r>ada<r>"}, 898 {"", "", "<>", -1, "<>"}, 899 {"banana", "a", "<>", -1, "b<>n<>n<>"}, 900 {"banana", "a", "<>", 1, "b<>nana"}, 901 {"banana", "a", "<>", 1000, "b<>n<>n<>"}, 902 {"banana", "an", "<>", -1, "b<><>a"}, 903 {"banana", "ana", "<>", -1, "b<>na"}, 904 {"banana", "", "<>", -1, "<>b<>a<>n<>a<>n<>a<>"}, 905 {"banana", "", "<>", 10, "<>b<>a<>n<>a<>n<>a<>"}, 906 {"banana", "", "<>", 6, "<>b<>a<>n<>a<>n<>a"}, 907 {"banana", "", "<>", 5, "<>b<>a<>n<>a<>na"}, 908 {"banana", "", "<>", 1, "<>banana"}, 909 {"banana", "a", "a", -1, "banana"}, 910 {"banana", "a", "a", 1, "banana"}, 911 {"☺☻☹", "", "<>", -1, "<>☺<>☻<>☹<>"}, 912 } 913 914 func TestReplace(t *testing.T) { 915 for _, tt := range ReplaceTests { 916 if s := Replace(tt.in, tt.old, tt.new, tt.n); s != tt.out { 917 t.Errorf("Replace(%q, %q, %q, %d) = %q, want %q", tt.in, tt.old, tt.new, tt.n, s, tt.out) 918 } 919 } 920 } 921 922 var TitleTests = []struct { 923 in, out string 924 }{ 925 {"", ""}, 926 {"a", "A"}, 927 {" aaa aaa aaa ", " Aaa Aaa Aaa "}, 928 {" Aaa Aaa Aaa ", " Aaa Aaa Aaa "}, 929 {"123a456", "123a456"}, 930 {"double-blind", "Double-Blind"}, 931 {"ÿøû", "Ÿøû"}, 932 {"with_underscore", "With_underscore"}, 933 {"unicode \xe2\x80\xa8 line separator", "Unicode \xe2\x80\xa8 Line Separator"}, 934 } 935 936 func TestTitle(t *testing.T) { 937 for _, tt := range TitleTests { 938 if s := Title(tt.in); s != tt.out { 939 t.Errorf("Title(%q) = %q, want %q", tt.in, s, tt.out) 940 } 941 } 942 } 943 944 var ContainsTests = []struct { 945 str, substr string 946 expected bool 947 }{ 948 {"abc", "bc", true}, 949 {"abc", "bcd", false}, 950 {"abc", "", true}, 951 {"", "a", false}, 952 } 953 954 func TestContains(t *testing.T) { 955 for _, ct := range ContainsTests { 956 if Contains(ct.str, ct.substr) != ct.expected { 957 t.Errorf("Contains(%s, %s) = %v, want %v", 958 ct.str, ct.substr, !ct.expected, ct.expected) 959 } 960 } 961 } 962 963 var ContainsAnyTests = []struct { 964 str, substr string 965 expected bool 966 }{ 967 {"", "", false}, 968 {"", "a", false}, 969 {"", "abc", false}, 970 {"a", "", false}, 971 {"a", "a", true}, 972 {"aaa", "a", true}, 973 {"abc", "xyz", false}, 974 {"abc", "xcz", true}, 975 {"a☺b☻c☹d", "uvw☻xyz", true}, 976 {"aRegExp*", ".(|)*+?^$[]", true}, 977 {dots + dots + dots, " ", false}, 978 } 979 980 func TestContainsAny(t *testing.T) { 981 for _, ct := range ContainsAnyTests { 982 if ContainsAny(ct.str, ct.substr) != ct.expected { 983 t.Errorf("ContainsAny(%s, %s) = %v, want %v", 984 ct.str, ct.substr, !ct.expected, ct.expected) 985 } 986 } 987 } 988 989 var ContainsRuneTests = []struct { 990 str string 991 r rune 992 expected bool 993 }{ 994 {"", 'a', false}, 995 {"a", 'a', true}, 996 {"aaa", 'a', true}, 997 {"abc", 'y', false}, 998 {"abc", 'c', true}, 999 {"a☺b☻c☹d", 'x', false}, 1000 {"a☺b☻c☹d", '☻', true}, 1001 {"aRegExp*", '*', true}, 1002 } 1003 1004 func TestContainsRune(t *testing.T) { 1005 for _, ct := range ContainsRuneTests { 1006 if ContainsRune(ct.str, ct.r) != ct.expected { 1007 t.Errorf("ContainsRune(%q, %q) = %v, want %v", 1008 ct.str, ct.r, !ct.expected, ct.expected) 1009 } 1010 } 1011 } 1012 1013 var EqualFoldTests = []struct { 1014 s, t string 1015 out bool 1016 }{ 1017 {"abc", "abc", true}, 1018 {"ABcd", "ABcd", true}, 1019 {"123abc", "123ABC", true}, 1020 {"αβδ", "ΑΒΔ", true}, 1021 {"abc", "xyz", false}, 1022 {"abc", "XYZ", false}, 1023 {"abcdefghijk", "abcdefghijX", false}, 1024 {"abcdefghijk", "abcdefghij\u212A", true}, 1025 {"abcdefghijK", "abcdefghij\u212A", true}, 1026 {"abcdefghijkz", "abcdefghij\u212Ay", false}, 1027 {"abcdefghijKz", "abcdefghij\u212Ay", false}, 1028 } 1029 1030 func TestEqualFold(t *testing.T) { 1031 for _, tt := range EqualFoldTests { 1032 if out := EqualFold(tt.s, tt.t); out != tt.out { 1033 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.s, tt.t, out, tt.out) 1034 } 1035 if out := EqualFold(tt.t, tt.s); out != tt.out { 1036 t.Errorf("EqualFold(%#q, %#q) = %v, want %v", tt.t, tt.s, out, tt.out) 1037 } 1038 } 1039 } 1040 1041 var CountTests = []struct { 1042 s, sep string 1043 num int 1044 }{ 1045 {"", "", 1}, 1046 {"", "notempty", 0}, 1047 {"notempty", "", 9}, 1048 {"smaller", "not smaller", 0}, 1049 {"12345678987654321", "6", 2}, 1050 {"611161116", "6", 3}, 1051 {"notequal", "NotEqual", 0}, 1052 {"equal", "equal", 1}, 1053 {"abc1231231123q", "123", 3}, 1054 {"11111", "11", 2}, 1055 } 1056 1057 func TestCount(t *testing.T) { 1058 for _, tt := range CountTests { 1059 if num := Count(tt.s, tt.sep); num != tt.num { 1060 t.Errorf("Count(\"%s\", \"%s\") = %d, want %d", tt.s, tt.sep, num, tt.num) 1061 } 1062 } 1063 } 1064 1065 func makeBenchInputHard() string { 1066 tokens := [...]string{ 1067 "<a>", "<p>", "<b>", "<strong>", 1068 "</a>", "</p>", "</b>", "</strong>", 1069 "hello", "world", 1070 } 1071 x := make([]byte, 0, 1<<20) 1072 for len(x) < 1<<20 { 1073 i := rand.Intn(len(tokens)) 1074 x = append(x, tokens[i]...) 1075 } 1076 return string(x) 1077 } 1078 1079 var benchInputHard = makeBenchInputHard() 1080 1081 func benchmarkIndexHard(b *testing.B, sep string) { 1082 for i := 0; i < b.N; i++ { 1083 Index(benchInputHard, sep) 1084 } 1085 } 1086 1087 func benchmarkCountHard(b *testing.B, sep string) { 1088 for i := 0; i < b.N; i++ { 1089 Count(benchInputHard, sep) 1090 } 1091 } 1092 1093 func BenchmarkIndexHard1(b *testing.B) { benchmarkIndexHard(b, "<>") } 1094 func BenchmarkIndexHard2(b *testing.B) { benchmarkIndexHard(b, "</pre>") } 1095 func BenchmarkIndexHard3(b *testing.B) { benchmarkIndexHard(b, "<b>hello world</b>") } 1096 1097 func BenchmarkCountHard1(b *testing.B) { benchmarkCountHard(b, "<>") } 1098 func BenchmarkCountHard2(b *testing.B) { benchmarkCountHard(b, "</pre>") } 1099 func BenchmarkCountHard3(b *testing.B) { benchmarkCountHard(b, "<b>hello world</b>") } 1100 1101 var benchInputTorture = Repeat("ABC", 1<<10) + "123" + Repeat("ABC", 1<<10) 1102 var benchNeedleTorture = Repeat("ABC", 1<<10+1) 1103 1104 func BenchmarkIndexTorture(b *testing.B) { 1105 for i := 0; i < b.N; i++ { 1106 Index(benchInputTorture, benchNeedleTorture) 1107 } 1108 } 1109 1110 func BenchmarkCountTorture(b *testing.B) { 1111 for i := 0; i < b.N; i++ { 1112 Count(benchInputTorture, benchNeedleTorture) 1113 } 1114 } 1115 1116 func BenchmarkCountTortureOverlapping(b *testing.B) { 1117 A := Repeat("ABC", 1<<20) 1118 B := Repeat("ABC", 1<<10) 1119 for i := 0; i < b.N; i++ { 1120 Count(A, B) 1121 } 1122 } 1123 1124 var makeFieldsInput = func() string { 1125 x := make([]byte, 1<<20) 1126 // Input is ~10% space, ~10% 2-byte UTF-8, rest ASCII non-space. 1127 for i := range x { 1128 switch rand.Intn(10) { 1129 case 0: 1130 x[i] = ' ' 1131 case 1: 1132 if i > 0 && x[i-1] == 'x' { 1133 copy(x[i-1:], "χ") 1134 break 1135 } 1136 fallthrough 1137 default: 1138 x[i] = 'x' 1139 } 1140 } 1141 return string(x) 1142 } 1143 1144 var fieldsInput = makeFieldsInput() 1145 1146 func BenchmarkFields(b *testing.B) { 1147 b.SetBytes(int64(len(fieldsInput))) 1148 for i := 0; i < b.N; i++ { 1149 Fields(fieldsInput) 1150 } 1151 } 1152 1153 func BenchmarkFieldsFunc(b *testing.B) { 1154 b.SetBytes(int64(len(fieldsInput))) 1155 for i := 0; i < b.N; i++ { 1156 FieldsFunc(fieldsInput, unicode.IsSpace) 1157 } 1158 } 1159 1160 func BenchmarkSplit1(b *testing.B) { 1161 for i := 0; i < b.N; i++ { 1162 Split(benchInputHard, "") 1163 } 1164 } 1165 1166 func BenchmarkSplit2(b *testing.B) { 1167 for i := 0; i < b.N; i++ { 1168 Split(benchInputHard, "/") 1169 } 1170 } 1171 1172 func BenchmarkSplit3(b *testing.B) { 1173 for i := 0; i < b.N; i++ { 1174 Split(benchInputHard, "hello") 1175 } 1176 }