Home | History | Annotate | Download | only in regexp
      1 // Copyright 2010 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 regexp
      6 
      7 import (
      8 	"bufio"
      9 	"compress/bzip2"
     10 	"fmt"
     11 	"internal/testenv"
     12 	"io"
     13 	"os"
     14 	"path/filepath"
     15 	"regexp/syntax"
     16 	"strconv"
     17 	"strings"
     18 	"testing"
     19 	"unicode/utf8"
     20 )
     21 
     22 // TestRE2 tests this package's regexp API against test cases
     23 // considered during RE2's exhaustive tests, which run all possible
     24 // regexps over a given set of atoms and operators, up to a given
     25 // complexity, over all possible strings over a given alphabet,
     26 // up to a given size. Rather than try to link with RE2, we read a
     27 // log file containing the test cases and the expected matches.
     28 // The log file, re2-exhaustive.txt, is generated by running 'make log'
     29 // in the open source RE2 distribution https://github.com/google/re2/.
     30 //
     31 // The test file format is a sequence of stanzas like:
     32 //
     33 //	strings
     34 //	"abc"
     35 //	"123x"
     36 //	regexps
     37 //	"[a-z]+"
     38 //	0-3;0-3
     39 //	-;-
     40 //	"([0-9])([0-9])([0-9])"
     41 //	-;-
     42 //	-;0-3 0-1 1-2 2-3
     43 //
     44 // The stanza begins by defining a set of strings, quoted
     45 // using Go double-quote syntax, one per line. Then the
     46 // regexps section gives a sequence of regexps to run on
     47 // the strings. In the block that follows a regexp, each line
     48 // gives the semicolon-separated match results of running
     49 // the regexp on the corresponding string.
     50 // Each match result is either a single -, meaning no match, or a
     51 // space-separated sequence of pairs giving the match and
     52 // submatch indices. An unmatched subexpression formats
     53 // its pair as a single - (not illustrated above).  For now
     54 // each regexp run produces two match results, one for a
     55 // ``full match'' that restricts the regexp to matching the entire
     56 // string or nothing, and one for a ``partial match'' that gives
     57 // the leftmost first match found in the string.
     58 //
     59 // Lines beginning with # are comments. Lines beginning with
     60 // a capital letter are test names printed during RE2's test suite
     61 // and are echoed into t but otherwise ignored.
     62 //
     63 // At time of writing, re2-exhaustive.txt is 59 MB but compresses to 385 kB,
     64 // so we store re2-exhaustive.txt.bz2 in the repository and decompress it on the fly.
     65 //
     66 func TestRE2Search(t *testing.T) {
     67 	testRE2(t, "testdata/re2-search.txt")
     68 }
     69 
     70 func testRE2(t *testing.T, file string) {
     71 	f, err := os.Open(file)
     72 	if err != nil {
     73 		t.Fatal(err)
     74 	}
     75 	defer f.Close()
     76 	var txt io.Reader
     77 	if strings.HasSuffix(file, ".bz2") {
     78 		z := bzip2.NewReader(f)
     79 		txt = z
     80 		file = file[:len(file)-len(".bz2")] // for error messages
     81 	} else {
     82 		txt = f
     83 	}
     84 	lineno := 0
     85 	scanner := bufio.NewScanner(txt)
     86 	var (
     87 		str       []string
     88 		input     []string
     89 		inStrings bool
     90 		re        *Regexp
     91 		refull    *Regexp
     92 		nfail     int
     93 		ncase     int
     94 	)
     95 	for lineno := 1; scanner.Scan(); lineno++ {
     96 		line := scanner.Text()
     97 		switch {
     98 		case line == "":
     99 			t.Fatalf("%s:%d: unexpected blank line", file, lineno)
    100 		case line[0] == '#':
    101 			continue
    102 		case 'A' <= line[0] && line[0] <= 'Z':
    103 			// Test name.
    104 			t.Logf("%s\n", line)
    105 			continue
    106 		case line == "strings":
    107 			str = str[:0]
    108 			inStrings = true
    109 		case line == "regexps":
    110 			inStrings = false
    111 		case line[0] == '"':
    112 			q, err := strconv.Unquote(line)
    113 			if err != nil {
    114 				// Fatal because we'll get out of sync.
    115 				t.Fatalf("%s:%d: unquote %s: %v", file, lineno, line, err)
    116 			}
    117 			if inStrings {
    118 				str = append(str, q)
    119 				continue
    120 			}
    121 			// Is a regexp.
    122 			if len(input) != 0 {
    123 				t.Fatalf("%s:%d: out of sync: have %d strings left before %#q", file, lineno, len(input), q)
    124 			}
    125 			re, err = tryCompile(q)
    126 			if err != nil {
    127 				if err.Error() == "error parsing regexp: invalid escape sequence: `\\C`" {
    128 					// We don't and likely never will support \C; keep going.
    129 					continue
    130 				}
    131 				t.Errorf("%s:%d: compile %#q: %v", file, lineno, q, err)
    132 				if nfail++; nfail >= 100 {
    133 					t.Fatalf("stopping after %d errors", nfail)
    134 				}
    135 				continue
    136 			}
    137 			full := `\A(?:` + q + `)\z`
    138 			refull, err = tryCompile(full)
    139 			if err != nil {
    140 				// Fatal because q worked, so this should always work.
    141 				t.Fatalf("%s:%d: compile full %#q: %v", file, lineno, full, err)
    142 			}
    143 			input = str
    144 		case line[0] == '-' || '0' <= line[0] && line[0] <= '9':
    145 			// A sequence of match results.
    146 			ncase++
    147 			if re == nil {
    148 				// Failed to compile: skip results.
    149 				continue
    150 			}
    151 			if len(input) == 0 {
    152 				t.Fatalf("%s:%d: out of sync: no input remaining", file, lineno)
    153 			}
    154 			var text string
    155 			text, input = input[0], input[1:]
    156 			if !isSingleBytes(text) && strings.Contains(re.String(), `\B`) {
    157 				// RE2's \B considers every byte position,
    158 				// so it sees 'not word boundary' in the
    159 				// middle of UTF-8 sequences. This package
    160 				// only considers the positions between runes,
    161 				// so it disagrees. Skip those cases.
    162 				continue
    163 			}
    164 			res := strings.Split(line, ";")
    165 			if len(res) != len(run) {
    166 				t.Fatalf("%s:%d: have %d test results, want %d", file, lineno, len(res), len(run))
    167 			}
    168 			for i := range res {
    169 				have, suffix := run[i](re, refull, text)
    170 				want := parseResult(t, file, lineno, res[i])
    171 				if !same(have, want) {
    172 					t.Errorf("%s:%d: %#q%s.FindSubmatchIndex(%#q) = %v, want %v", file, lineno, re, suffix, text, have, want)
    173 					if nfail++; nfail >= 100 {
    174 						t.Fatalf("stopping after %d errors", nfail)
    175 					}
    176 					continue
    177 				}
    178 				b, suffix := match[i](re, refull, text)
    179 				if b != (want != nil) {
    180 					t.Errorf("%s:%d: %#q%s.MatchString(%#q) = %v, want %v", file, lineno, re, suffix, text, b, !b)
    181 					if nfail++; nfail >= 100 {
    182 						t.Fatalf("stopping after %d errors", nfail)
    183 					}
    184 					continue
    185 				}
    186 			}
    187 
    188 		default:
    189 			t.Fatalf("%s:%d: out of sync: %s\n", file, lineno, line)
    190 		}
    191 	}
    192 	if err := scanner.Err(); err != nil {
    193 		t.Fatalf("%s:%d: %v", file, lineno, err)
    194 	}
    195 	if len(input) != 0 {
    196 		t.Fatalf("%s:%d: out of sync: have %d strings left at EOF", file, lineno, len(input))
    197 	}
    198 	t.Logf("%d cases tested", ncase)
    199 }
    200 
    201 var run = []func(*Regexp, *Regexp, string) ([]int, string){
    202 	runFull,
    203 	runPartial,
    204 	runFullLongest,
    205 	runPartialLongest,
    206 }
    207 
    208 func runFull(re, refull *Regexp, text string) ([]int, string) {
    209 	refull.longest = false
    210 	return refull.FindStringSubmatchIndex(text), "[full]"
    211 }
    212 
    213 func runPartial(re, refull *Regexp, text string) ([]int, string) {
    214 	re.longest = false
    215 	return re.FindStringSubmatchIndex(text), ""
    216 }
    217 
    218 func runFullLongest(re, refull *Regexp, text string) ([]int, string) {
    219 	refull.longest = true
    220 	return refull.FindStringSubmatchIndex(text), "[full,longest]"
    221 }
    222 
    223 func runPartialLongest(re, refull *Regexp, text string) ([]int, string) {
    224 	re.longest = true
    225 	return re.FindStringSubmatchIndex(text), "[longest]"
    226 }
    227 
    228 var match = []func(*Regexp, *Regexp, string) (bool, string){
    229 	matchFull,
    230 	matchPartial,
    231 	matchFullLongest,
    232 	matchPartialLongest,
    233 }
    234 
    235 func matchFull(re, refull *Regexp, text string) (bool, string) {
    236 	refull.longest = false
    237 	return refull.MatchString(text), "[full]"
    238 }
    239 
    240 func matchPartial(re, refull *Regexp, text string) (bool, string) {
    241 	re.longest = false
    242 	return re.MatchString(text), ""
    243 }
    244 
    245 func matchFullLongest(re, refull *Regexp, text string) (bool, string) {
    246 	refull.longest = true
    247 	return refull.MatchString(text), "[full,longest]"
    248 }
    249 
    250 func matchPartialLongest(re, refull *Regexp, text string) (bool, string) {
    251 	re.longest = true
    252 	return re.MatchString(text), "[longest]"
    253 }
    254 
    255 func isSingleBytes(s string) bool {
    256 	for _, c := range s {
    257 		if c >= utf8.RuneSelf {
    258 			return false
    259 		}
    260 	}
    261 	return true
    262 }
    263 
    264 func tryCompile(s string) (re *Regexp, err error) {
    265 	// Protect against panic during Compile.
    266 	defer func() {
    267 		if r := recover(); r != nil {
    268 			err = fmt.Errorf("panic: %v", r)
    269 		}
    270 	}()
    271 	return Compile(s)
    272 }
    273 
    274 func parseResult(t *testing.T, file string, lineno int, res string) []int {
    275 	// A single - indicates no match.
    276 	if res == "-" {
    277 		return nil
    278 	}
    279 	// Otherwise, a space-separated list of pairs.
    280 	n := 1
    281 	for j := 0; j < len(res); j++ {
    282 		if res[j] == ' ' {
    283 			n++
    284 		}
    285 	}
    286 	out := make([]int, 2*n)
    287 	i := 0
    288 	n = 0
    289 	for j := 0; j <= len(res); j++ {
    290 		if j == len(res) || res[j] == ' ' {
    291 			// Process a single pair.  - means no submatch.
    292 			pair := res[i:j]
    293 			if pair == "-" {
    294 				out[n] = -1
    295 				out[n+1] = -1
    296 			} else {
    297 				k := strings.Index(pair, "-")
    298 				if k < 0 {
    299 					t.Fatalf("%s:%d: invalid pair %s", file, lineno, pair)
    300 				}
    301 				lo, err1 := strconv.Atoi(pair[:k])
    302 				hi, err2 := strconv.Atoi(pair[k+1:])
    303 				if err1 != nil || err2 != nil || lo > hi {
    304 					t.Fatalf("%s:%d: invalid pair %s", file, lineno, pair)
    305 				}
    306 				out[n] = lo
    307 				out[n+1] = hi
    308 			}
    309 			n += 2
    310 			i = j + 1
    311 		}
    312 	}
    313 	return out
    314 }
    315 
    316 func same(x, y []int) bool {
    317 	if len(x) != len(y) {
    318 		return false
    319 	}
    320 	for i, xi := range x {
    321 		if xi != y[i] {
    322 			return false
    323 		}
    324 	}
    325 	return true
    326 }
    327 
    328 // TestFowler runs this package's regexp API against the
    329 // POSIX regular expression tests collected by Glenn Fowler
    330 // at http://www2.research.att.com/~astopen/testregex/testregex.html.
    331 func TestFowler(t *testing.T) {
    332 	files, err := filepath.Glob("testdata/*.dat")
    333 	if err != nil {
    334 		t.Fatal(err)
    335 	}
    336 	for _, file := range files {
    337 		t.Log(file)
    338 		testFowler(t, file)
    339 	}
    340 }
    341 
    342 var notab = MustCompilePOSIX(`[^\t]+`)
    343 
    344 func testFowler(t *testing.T, file string) {
    345 	f, err := os.Open(file)
    346 	if err != nil {
    347 		t.Error(err)
    348 		return
    349 	}
    350 	defer f.Close()
    351 	b := bufio.NewReader(f)
    352 	lineno := 0
    353 	lastRegexp := ""
    354 Reading:
    355 	for {
    356 		lineno++
    357 		line, err := b.ReadString('\n')
    358 		if err != nil {
    359 			if err != io.EOF {
    360 				t.Errorf("%s:%d: %v", file, lineno, err)
    361 			}
    362 			break Reading
    363 		}
    364 
    365 		// http://www2.research.att.com/~astopen/man/man1/testregex.html
    366 		//
    367 		// INPUT FORMAT
    368 		//   Input lines may be blank, a comment beginning with #, or a test
    369 		//   specification. A specification is five fields separated by one
    370 		//   or more tabs. NULL denotes the empty string and NIL denotes the
    371 		//   0 pointer.
    372 		if line[0] == '#' || line[0] == '\n' {
    373 			continue Reading
    374 		}
    375 		line = line[:len(line)-1]
    376 		field := notab.FindAllString(line, -1)
    377 		for i, f := range field {
    378 			if f == "NULL" {
    379 				field[i] = ""
    380 			}
    381 			if f == "NIL" {
    382 				t.Logf("%s:%d: skip: %s", file, lineno, line)
    383 				continue Reading
    384 			}
    385 		}
    386 		if len(field) == 0 {
    387 			continue Reading
    388 		}
    389 
    390 		//   Field 1: the regex(3) flags to apply, one character per REG_feature
    391 		//   flag. The test is skipped if REG_feature is not supported by the
    392 		//   implementation. If the first character is not [BEASKLP] then the
    393 		//   specification is a global control line. One or more of [BEASKLP] may be
    394 		//   specified; the test will be repeated for each mode.
    395 		//
    396 		//     B 	basic			BRE	(grep, ed, sed)
    397 		//     E 	REG_EXTENDED		ERE	(egrep)
    398 		//     A	REG_AUGMENTED		ARE	(egrep with negation)
    399 		//     S	REG_SHELL		SRE	(sh glob)
    400 		//     K	REG_SHELL|REG_AUGMENTED	KRE	(ksh glob)
    401 		//     L	REG_LITERAL		LRE	(fgrep)
    402 		//
    403 		//     a	REG_LEFT|REG_RIGHT	implicit ^...$
    404 		//     b	REG_NOTBOL		lhs does not match ^
    405 		//     c	REG_COMMENT		ignore space and #...\n
    406 		//     d	REG_SHELL_DOT		explicit leading . match
    407 		//     e	REG_NOTEOL		rhs does not match $
    408 		//     f	REG_MULTIPLE		multiple \n separated patterns
    409 		//     g	FNM_LEADING_DIR		testfnmatch only -- match until /
    410 		//     h	REG_MULTIREF		multiple digit backref
    411 		//     i	REG_ICASE		ignore case
    412 		//     j	REG_SPAN		. matches \n
    413 		//     k	REG_ESCAPE		\ to escape [...] delimiter
    414 		//     l	REG_LEFT		implicit ^...
    415 		//     m	REG_MINIMAL		minimal match
    416 		//     n	REG_NEWLINE		explicit \n match
    417 		//     o	REG_ENCLOSED		(|&) magic inside [@|&](...)
    418 		//     p	REG_SHELL_PATH		explicit / match
    419 		//     q	REG_DELIMITED		delimited pattern
    420 		//     r	REG_RIGHT		implicit ...$
    421 		//     s	REG_SHELL_ESCAPED	\ not special
    422 		//     t	REG_MUSTDELIM		all delimiters must be specified
    423 		//     u	standard unspecified behavior -- errors not counted
    424 		//     v	REG_CLASS_ESCAPE	\ special inside [...]
    425 		//     w	REG_NOSUB		no subexpression match array
    426 		//     x	REG_LENIENT		let some errors slide
    427 		//     y	REG_LEFT		regexec() implicit ^...
    428 		//     z	REG_NULL		NULL subexpressions ok
    429 		//     $	                        expand C \c escapes in fields 2 and 3
    430 		//     /	                        field 2 is a regsubcomp() expression
    431 		//     =	                        field 3 is a regdecomp() expression
    432 		//
    433 		//   Field 1 control lines:
    434 		//
    435 		//     C		set LC_COLLATE and LC_CTYPE to locale in field 2
    436 		//
    437 		//     ?test ...	output field 5 if passed and != EXPECTED, silent otherwise
    438 		//     &test ...	output field 5 if current and previous passed
    439 		//     |test ...	output field 5 if current passed and previous failed
    440 		//     ; ...	output field 2 if previous failed
    441 		//     {test ...	skip if failed until }
    442 		//     }		end of skip
    443 		//
    444 		//     : comment		comment copied as output NOTE
    445 		//     :comment:test	:comment: ignored
    446 		//     N[OTE] comment	comment copied as output NOTE
    447 		//     T[EST] comment	comment
    448 		//
    449 		//     number		use number for nmatch (20 by default)
    450 		flag := field[0]
    451 		switch flag[0] {
    452 		case '?', '&', '|', ';', '{', '}':
    453 			// Ignore all the control operators.
    454 			// Just run everything.
    455 			flag = flag[1:]
    456 			if flag == "" {
    457 				continue Reading
    458 			}
    459 		case ':':
    460 			i := strings.Index(flag[1:], ":")
    461 			if i < 0 {
    462 				t.Logf("skip: %s", line)
    463 				continue Reading
    464 			}
    465 			flag = flag[1+i+1:]
    466 		case 'C', 'N', 'T', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9':
    467 			t.Logf("skip: %s", line)
    468 			continue Reading
    469 		}
    470 
    471 		// Can check field count now that we've handled the myriad comment formats.
    472 		if len(field) < 4 {
    473 			t.Errorf("%s:%d: too few fields: %s", file, lineno, line)
    474 			continue Reading
    475 		}
    476 
    477 		// Expand C escapes (a.k.a. Go escapes).
    478 		if strings.Contains(flag, "$") {
    479 			f := `"` + field[1] + `"`
    480 			if field[1], err = strconv.Unquote(f); err != nil {
    481 				t.Errorf("%s:%d: cannot unquote %s", file, lineno, f)
    482 			}
    483 			f = `"` + field[2] + `"`
    484 			if field[2], err = strconv.Unquote(f); err != nil {
    485 				t.Errorf("%s:%d: cannot unquote %s", file, lineno, f)
    486 			}
    487 		}
    488 
    489 		//   Field 2: the regular expression pattern; SAME uses the pattern from
    490 		//     the previous specification.
    491 		//
    492 		if field[1] == "SAME" {
    493 			field[1] = lastRegexp
    494 		}
    495 		lastRegexp = field[1]
    496 
    497 		//   Field 3: the string to match.
    498 		text := field[2]
    499 
    500 		//   Field 4: the test outcome...
    501 		ok, shouldCompile, shouldMatch, pos := parseFowlerResult(field[3])
    502 		if !ok {
    503 			t.Errorf("%s:%d: cannot parse result %#q", file, lineno, field[3])
    504 			continue Reading
    505 		}
    506 
    507 		//   Field 5: optional comment appended to the report.
    508 
    509 	Testing:
    510 		// Run test once for each specified capital letter mode that we support.
    511 		for _, c := range flag {
    512 			pattern := field[1]
    513 			syn := syntax.POSIX | syntax.ClassNL
    514 			switch c {
    515 			default:
    516 				continue Testing
    517 			case 'E':
    518 				// extended regexp (what we support)
    519 			case 'L':
    520 				// literal
    521 				pattern = QuoteMeta(pattern)
    522 			}
    523 
    524 			for _, c := range flag {
    525 				switch c {
    526 				case 'i':
    527 					syn |= syntax.FoldCase
    528 				}
    529 			}
    530 
    531 			re, err := compile(pattern, syn, true)
    532 			if err != nil {
    533 				if shouldCompile {
    534 					t.Errorf("%s:%d: %#q did not compile", file, lineno, pattern)
    535 				}
    536 				continue Testing
    537 			}
    538 			if !shouldCompile {
    539 				t.Errorf("%s:%d: %#q should not compile", file, lineno, pattern)
    540 				continue Testing
    541 			}
    542 			match := re.MatchString(text)
    543 			if match != shouldMatch {
    544 				t.Errorf("%s:%d: %#q.Match(%#q) = %v, want %v", file, lineno, pattern, text, match, shouldMatch)
    545 				continue Testing
    546 			}
    547 			have := re.FindStringSubmatchIndex(text)
    548 			if (len(have) > 0) != match {
    549 				t.Errorf("%s:%d: %#q.Match(%#q) = %v, but %#q.FindSubmatchIndex(%#q) = %v", file, lineno, pattern, text, match, pattern, text, have)
    550 				continue Testing
    551 			}
    552 			if len(have) > len(pos) {
    553 				have = have[:len(pos)]
    554 			}
    555 			if !same(have, pos) {
    556 				t.Errorf("%s:%d: %#q.FindSubmatchIndex(%#q) = %v, want %v", file, lineno, pattern, text, have, pos)
    557 			}
    558 		}
    559 	}
    560 }
    561 
    562 func parseFowlerResult(s string) (ok, compiled, matched bool, pos []int) {
    563 	//   Field 4: the test outcome. This is either one of the posix error
    564 	//     codes (with REG_ omitted) or the match array, a list of (m,n)
    565 	//     entries with m and n being first and last+1 positions in the
    566 	//     field 3 string, or NULL if REG_NOSUB is in effect and success
    567 	//     is expected. BADPAT is acceptable in place of any regcomp(3)
    568 	//     error code. The match[] array is initialized to (-2,-2) before
    569 	//     each test. All array elements from 0 to nmatch-1 must be specified
    570 	//     in the outcome. Unspecified endpoints (offset -1) are denoted by ?.
    571 	//     Unset endpoints (offset -2) are denoted by X. {x}(o:n) denotes a
    572 	//     matched (?{...}) expression, where x is the text enclosed by {...},
    573 	//     o is the expression ordinal counting from 1, and n is the length of
    574 	//     the unmatched portion of the subject string. If x starts with a
    575 	//     number then that is the return value of re_execf(), otherwise 0 is
    576 	//     returned.
    577 	switch {
    578 	case s == "":
    579 		// Match with no position information.
    580 		ok = true
    581 		compiled = true
    582 		matched = true
    583 		return
    584 	case s == "NOMATCH":
    585 		// Match failure.
    586 		ok = true
    587 		compiled = true
    588 		matched = false
    589 		return
    590 	case 'A' <= s[0] && s[0] <= 'Z':
    591 		// All the other error codes are compile errors.
    592 		ok = true
    593 		compiled = false
    594 		return
    595 	}
    596 	compiled = true
    597 
    598 	var x []int
    599 	for s != "" {
    600 		var end byte = ')'
    601 		if len(x)%2 == 0 {
    602 			if s[0] != '(' {
    603 				ok = false
    604 				return
    605 			}
    606 			s = s[1:]
    607 			end = ','
    608 		}
    609 		i := 0
    610 		for i < len(s) && s[i] != end {
    611 			i++
    612 		}
    613 		if i == 0 || i == len(s) {
    614 			ok = false
    615 			return
    616 		}
    617 		var v = -1
    618 		var err error
    619 		if s[:i] != "?" {
    620 			v, err = strconv.Atoi(s[:i])
    621 			if err != nil {
    622 				ok = false
    623 				return
    624 			}
    625 		}
    626 		x = append(x, v)
    627 		s = s[i+1:]
    628 	}
    629 	if len(x)%2 != 0 {
    630 		ok = false
    631 		return
    632 	}
    633 	ok = true
    634 	matched = true
    635 	pos = x
    636 	return
    637 }
    638 
    639 var text []byte
    640 
    641 func makeText(n int) []byte {
    642 	if len(text) >= n {
    643 		return text[:n]
    644 	}
    645 	text = make([]byte, n)
    646 	x := ^uint32(0)
    647 	for i := range text {
    648 		x += x
    649 		x ^= 1
    650 		if int32(x) < 0 {
    651 			x ^= 0x88888eef
    652 		}
    653 		if x%31 == 0 {
    654 			text[i] = '\n'
    655 		} else {
    656 			text[i] = byte(x%(0x7E+1-0x20) + 0x20)
    657 		}
    658 	}
    659 	return text
    660 }
    661 
    662 func BenchmarkMatch(b *testing.B) {
    663 	isRaceBuilder := strings.HasSuffix(testenv.Builder(), "-race")
    664 
    665 	for _, data := range benchData {
    666 		r := MustCompile(data.re)
    667 		for _, size := range benchSizes {
    668 			if isRaceBuilder && size.n > 1<<10 {
    669 				continue
    670 			}
    671 			t := makeText(size.n)
    672 			b.Run(data.name+"/"+size.name, func(b *testing.B) {
    673 				b.SetBytes(int64(size.n))
    674 				for i := 0; i < b.N; i++ {
    675 					if r.Match(t) {
    676 						b.Fatal("match!")
    677 					}
    678 				}
    679 			})
    680 		}
    681 	}
    682 }
    683 
    684 func BenchmarkMatch_onepass_regex(b *testing.B) {
    685 	isRaceBuilder := strings.HasSuffix(testenv.Builder(), "-race")
    686 	r := MustCompile(`(?s)\A.*\z`)
    687 	if r.get().op == notOnePass {
    688 		b.Fatalf("want onepass regex, but %q is not onepass", r)
    689 	}
    690 	for _, size := range benchSizes {
    691 		if isRaceBuilder && size.n > 1<<10 {
    692 			continue
    693 		}
    694 		t := makeText(size.n)
    695 		bs := make([][]byte, len(t))
    696 		for i, s := range t {
    697 			bs[i] = []byte{s}
    698 		}
    699 		b.Run(size.name, func(b *testing.B) {
    700 			b.SetBytes(int64(size.n))
    701 			b.ReportAllocs()
    702 			for i := 0; i < b.N; i++ {
    703 				for _, byts := range bs {
    704 					if !r.Match(byts) {
    705 						b.Fatal("not match!")
    706 					}
    707 				}
    708 			}
    709 		})
    710 	}
    711 }
    712 
    713 var benchData = []struct{ name, re string }{
    714 	{"Easy0", "ABCDEFGHIJKLMNOPQRSTUVWXYZ$"},
    715 	{"Easy0i", "(?i)ABCDEFGHIJklmnopqrstuvwxyz$"},
    716 	{"Easy1", "A[AB]B[BC]C[CD]D[DE]E[EF]F[FG]G[GH]H[HI]I[IJ]J$"},
    717 	{"Medium", "[XYZ]ABCDEFGHIJKLMNOPQRSTUVWXYZ$"},
    718 	{"Hard", "[ -~]*ABCDEFGHIJKLMNOPQRSTUVWXYZ$"},
    719 	{"Hard1", "ABCD|CDEF|EFGH|GHIJ|IJKL|KLMN|MNOP|OPQR|QRST|STUV|UVWX|WXYZ"},
    720 }
    721 
    722 var benchSizes = []struct {
    723 	name string
    724 	n    int
    725 }{
    726 	{"32", 32},
    727 	{"1K", 1 << 10},
    728 	{"32K", 32 << 10},
    729 	{"1M", 1 << 20},
    730 	{"32M", 32 << 20},
    731 }
    732 
    733 func TestLongest(t *testing.T) {
    734 	re, err := Compile(`a(|b)`)
    735 	if err != nil {
    736 		t.Fatal(err)
    737 	}
    738 	if g, w := re.FindString("ab"), "a"; g != w {
    739 		t.Errorf("first match was %q, want %q", g, w)
    740 	}
    741 	re.Longest()
    742 	if g, w := re.FindString("ab"), "ab"; g != w {
    743 		t.Errorf("longest match was %q, want %q", g, w)
    744 	}
    745 }
    746 
    747 // TestProgramTooLongForBacktrack tests that a regex which is too long
    748 // for the backtracker still executes properly.
    749 func TestProgramTooLongForBacktrack(t *testing.T) {
    750 	longRegex := MustCompile(`(one|two|three|four|five|six|seven|eight|nine|ten|eleven|twelve|thirteen|fourteen|fifteen|sixteen|seventeen|eighteen|nineteen|twenty|twentyone|twentytwo|twentythree|twentyfour|twentyfive|twentysix|twentyseven|twentyeight|twentynine|thirty|thirtyone|thirtytwo|thirtythree|thirtyfour|thirtyfive|thirtysix|thirtyseven|thirtyeight|thirtynine|forty|fortyone|fortytwo|fortythree|fortyfour|fortyfive|fortysix|fortyseven|fortyeight|fortynine|fifty|fiftyone|fiftytwo|fiftythree|fiftyfour|fiftyfive|fiftysix|fiftyseven|fiftyeight|fiftynine|sixty|sixtyone|sixtytwo|sixtythree|sixtyfour|sixtyfive|sixtysix|sixtyseven|sixtyeight|sixtynine|seventy|seventyone|seventytwo|seventythree|seventyfour|seventyfive|seventysix|seventyseven|seventyeight|seventynine|eighty|eightyone|eightytwo|eightythree|eightyfour|eightyfive|eightysix|eightyseven|eightyeight|eightynine|ninety|ninetyone|ninetytwo|ninetythree|ninetyfour|ninetyfive|ninetysix|ninetyseven|ninetyeight|ninetynine|onehundred)`)
    751 	if !longRegex.MatchString("two") {
    752 		t.Errorf("longRegex.MatchString(\"two\") was false, want true")
    753 	}
    754 	if longRegex.MatchString("xxx") {
    755 		t.Errorf("longRegex.MatchString(\"xxx\") was true, want false")
    756 	}
    757 }
    758