Newer
Older
// Copyright 2011 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package syntax
import (
"fmt"
"strings"
"testing"
"unicode"
)
type parseTest struct {
Regexp string
Dump string
}
var parseTests = []parseTest{
{`a`, `lit{a}`},
{`a.`, `cat{lit{a}dot{}}`},
{`a.b`, `cat{lit{a}dot{}lit{b}}`},
{`ab`, `str{ab}`},
{`a.b.c`, `cat{lit{a}dot{}lit{b}dot{}lit{c}}`},
{`abc`, `str{abc}`},
{`a|^`, `alt{lit{a}bol{}}`},
{`a|b`, `cc{0x61-0x62}`},
{`(a)`, `cap{lit{a}}`},
{`(a)|b`, `alt{cap{lit{a}}lit{b}}`},
{`a*`, `star{lit{a}}`},
{`a+`, `plus{lit{a}}`},
{`a?`, `que{lit{a}}`},
{`a{2}`, `rep{2,2 lit{a}}`},
{`a{2,3}`, `rep{2,3 lit{a}}`},
{`a{2,}`, `rep{2,-1 lit{a}}`},
{`a*?`, `nstar{lit{a}}`},
{`a+?`, `nplus{lit{a}}`},
{`a??`, `nque{lit{a}}`},
{`a{2}?`, `nrep{2,2 lit{a}}`},
{`a{2,3}?`, `nrep{2,3 lit{a}}`},
{`a{2,}?`, `nrep{2,-1 lit{a}}`},
// Malformed { } are treated as literals.
{`x{1001`, `str{x{1001}`},
{`x{9876543210`, `str{x{9876543210}`},
{`x{9876543210,`, `str{x{9876543210,}`},
{`x{2,1`, `str{x{2,1}`},
{`x{1,9876543210`, `str{x{1,9876543210}`},
{`|`, `emp{}`}, // alt{emp{}emp{}} but got factored
{`|x|`, `alt{emp{}lit{x}emp{}}`},
{`.`, `dot{}`},
{`^`, `bol{}`},
{`$`, `eol{}`},
{`\|`, `lit{|}`},
{`\(`, `lit{(}`},
{`\)`, `lit{)}`},
{`\*`, `lit{*}`},
{`\+`, `lit{+}`},
{`\?`, `lit{?}`},
{`{`, `lit{{}`},
{`}`, `lit{}}`},
{`\.`, `lit{.}`},
{`\^`, `lit{^}`},
{`\$`, `lit{$}`},
{`\\`, `lit{\}`},
{`[ace]`, `cc{0x61 0x63 0x65}`},
{`[abc]`, `cc{0x61-0x63}`},
{`[a-z]`, `cc{0x61-0x7a}`},
{`[a]`, `lit{a}`},
{`\-`, `lit{-}`},
{`-`, `lit{-}`},
{`\_`, `lit{_}`},
{`abc`, `str{abc}`},
{`abc|def`, `alt{str{abc}str{def}}`},
{`abc|def|ghi`, `alt{str{abc}str{def}str{ghi}}`},
// Posix and Perl extensions
{`[[:lower:]]`, `cc{0x61-0x7a}`},
{`[a-z]`, `cc{0x61-0x7a}`},
{`[^[:lower:]]`, `cc{0x0-0x60 0x7b-0x10ffff}`},
{`[[:^lower:]]`, `cc{0x0-0x60 0x7b-0x10ffff}`},
{`(?i)[[:lower:]]`, `cc{0x41-0x5a 0x61-0x7a 0x17f 0x212a}`},
{`(?i)[a-z]`, `cc{0x41-0x5a 0x61-0x7a 0x17f 0x212a}`},
{`(?i)[^[:lower:]]`, `cc{0x0-0x40 0x5b-0x60 0x7b-0x17e 0x180-0x2129 0x212b-0x10ffff}`},
{`(?i)[[:^lower:]]`, `cc{0x0-0x40 0x5b-0x60 0x7b-0x17e 0x180-0x2129 0x212b-0x10ffff}`},
{`\d`, `cc{0x30-0x39}`},
{`\D`, `cc{0x0-0x2f 0x3a-0x10ffff}`},
{`\s`, `cc{0x9-0xa 0xc-0xd 0x20}`},
{`\S`, `cc{0x0-0x8 0xb 0xe-0x1f 0x21-0x10ffff}`},
{`\w`, `cc{0x30-0x39 0x41-0x5a 0x5f 0x61-0x7a}`},
{`\W`, `cc{0x0-0x2f 0x3a-0x40 0x5b-0x5e 0x60 0x7b-0x10ffff}`},
{`(?i)\w`, `cc{0x30-0x39 0x41-0x5a 0x5f 0x61-0x7a 0x17f 0x212a}`},
{`(?i)\W`, `cc{0x0-0x2f 0x3a-0x40 0x5b-0x5e 0x60 0x7b-0x17e 0x180-0x2129 0x212b-0x10ffff}`},
{`[^\\]`, `cc{0x0-0x5b 0x5d-0x10ffff}`},
// { `\C`, `byte{}` }, // probably never
// Unicode, negatives, and a double negative.
{`\p{Braille}`, `cc{0x2800-0x28ff}`},
{`\P{Braille}`, `cc{0x0-0x27ff 0x2900-0x10ffff}`},
{`\p{^Braille}`, `cc{0x0-0x27ff 0x2900-0x10ffff}`},
{`\P{^Braille}`, `cc{0x2800-0x28ff}`},
{`\pZ`, `cc{0x20 0xa0 0x1680 0x2000-0x200a 0x2028-0x2029 0x202f 0x205f 0x3000}`},
{`[\p{Braille}]`, `cc{0x2800-0x28ff}`},
{`[\P{Braille}]`, `cc{0x0-0x27ff 0x2900-0x10ffff}`},
{`[\p{^Braille}]`, `cc{0x0-0x27ff 0x2900-0x10ffff}`},
{`[\P{^Braille}]`, `cc{0x2800-0x28ff}`},
{`[\pZ]`, `cc{0x20 0xa0 0x1680 0x2000-0x200a 0x2028-0x2029 0x202f 0x205f 0x3000}`},
{`\p{Lu}`, mkCharClass(unicode.IsUpper)},
{`\p{Uppercase_Letter}`, mkCharClass(unicode.IsUpper)},
{`\p{upper case-let ter}`, mkCharClass(unicode.IsUpper)},
{`\p{__upper case-let ter}`, mkCharClass(unicode.IsUpper)},
{`[\p{Lu}]`, mkCharClass(unicode.IsUpper)},
{`(?i)[\p{Lu}]`, mkCharClass(isUpperFold)},
{`\p{Any}`, `dot{}`},
{`\p{^Any}`, `cc{}`},
{`(?i)\p{ascii}`, `cc{0x0-0x7f 0x17f 0x212a}`},
{`\p{Assigned}`, mkCharClass(func(r rune) bool { return !unicode.In(r, unicode.Cn) })},
{`\p{^Assigned}`, mkCharClass(func(r rune) bool { return unicode.In(r, unicode.Cn) })},
{`[\012-\234]\141`, `cat{cc{0xa-0x9c}lit{a}}`},
{`[\x{41}-\x7a]\x61`, `cat{cc{0x41-0x7a}lit{a}}`},
// More interesting regular expressions.
{`a{,2}`, `str{a{,2}}`},
{`\.\^\$\\`, `str{.^$\}`},
{`[a-zABC]`, `cc{0x41-0x43 0x61-0x7a}`},
{`[^a]`, `cc{0x0-0x60 0x62-0x10ffff}`},
{`[α-ε☺]`, `cc{0x3b1-0x3b5 0x263a}`}, // utf-8
{`a*{`, `cat{star{lit{a}}lit{{}}`},
{`(?:ab)*`, `star{str{ab}}`},
{`(ab)*`, `star{cap{str{ab}}}`},
{`ab|cd`, `alt{str{ab}str{cd}}`},
{`a(b|c)d`, `cat{lit{a}cap{cc{0x62-0x63}}lit{d}}`},
{`(?:a)`, `lit{a}`},
{`(?:ab)(?:cd)`, `str{abcd}`},
{`(?:a+b+)(?:c+d+)`, `cat{plus{lit{a}}plus{lit{b}}plus{lit{c}}plus{lit{d}}}`},
{`(?:a+|b+)|(?:c+|d+)`, `alt{plus{lit{a}}plus{lit{b}}plus{lit{c}}plus{lit{d}}}`},
{`(?:a|b)|(?:c|d)`, `cc{0x61-0x64}`},
{`a|.`, `dot{}`},
{`.|a`, `dot{}`},
{`(?:[abc]|A|Z|hello|world)`, `alt{cc{0x41 0x5a 0x61-0x63}str{hello}str{world}}`},
{`(?:[abc]|A|Z)`, `cc{0x41 0x5a 0x61-0x63}`},
// Test Perl quoted literals
{`\Q+|*?{[\E`, `str{+|*?{[}`},
{`\Q+\E+`, `plus{lit{+}}`},
{`\Qab\E+`, `cat{lit{a}plus{lit{b}}}`},
{`\Q\\E`, `lit{\}`},
{`\Q\\\E`, `str{\\}`},
{`(?m)^`, `bol{}`},
{`(?m)$`, `eol{}`},
{`(?-m)^`, `bot{}`},
{`(?-m)$`, `eot{}`},
{`(?m)\A`, `bot{}`},
{`(?m)\z`, `eot{\z}`},
{`(?-m)\A`, `bot{}`},
{`(?-m)\z`, `eot{\z}`},
{`(?P<name>a)`, `cap{name:lit{a}}`},
{`(?<name>a)`, `cap{name:lit{a}}`},
// Case-folded literals
{`[Aa]`, `litfold{A}`},
{`[\x{100}\x{101}]`, `litfold{Ā}`},
{`[Δδ]`, `litfold{Δ}`},
{`abcde`, `str{abcde}`},
{`[Aa][Bb]cd`, `cat{strfold{AB}str{cd}}`},
// Factoring.
{`abc|abd|aef|bcx|bcy`, `alt{cat{lit{a}alt{cat{lit{b}cc{0x63-0x64}}str{ef}}}cat{str{bc}cc{0x78-0x79}}}`},
{`ax+y|ax+z|ay+w`, `cat{lit{a}alt{cat{plus{lit{x}}lit{y}}cat{plus{lit{x}}lit{z}}cat{plus{lit{y}}lit{w}}}}`},
// Bug fixes.
{`(?:.)`, `dot{}`},
{`(?:x|(?:xa))`, `cat{lit{x}alt{emp{}lit{a}}}`},
{`(?:.|(?:.a))`, `cat{dot{}alt{emp{}lit{a}}}`},
{`(?:A(?:A|a))`, `cat{lit{A}litfold{A}}`},
{`(?:A|a)`, `litfold{A}`},
{`A|(?:A|a)`, `litfold{A}`},
{`(?s).`, `dot{}`},
{`(?-s).`, `dnl{}`},
{`(?:(?:^).)`, `cat{bol{}dot{}}`},
{`(?-s)(?:(?:^).)`, `cat{bol{}dnl{}}`},
{`[\s\S]a`, `cat{cc{0x0-0x10ffff}lit{a}}`},
// RE2 prefix_tests
{`abc|abd`, `cat{str{ab}cc{0x63-0x64}}`},
{`a(?:b)c|abd`, `cat{str{ab}cc{0x63-0x64}}`},
{`abc|abd|aef|bcx|bcy`,
`alt{cat{lit{a}alt{cat{lit{b}cc{0x63-0x64}}str{ef}}}` +
`cat{str{bc}cc{0x78-0x79}}}`},
{`abc|x|abd`, `alt{str{abc}lit{x}str{abd}}`},
{`(?i)abc|ABD`, `cat{strfold{AB}cc{0x43-0x44 0x63-0x64}}`},
{`[ab]c|[ab]d`, `cat{cc{0x61-0x62}cc{0x63-0x64}}`},
{`.c|.d`, `cat{dot{}cc{0x63-0x64}}`},
{`x{2}|x{2}[0-9]`,
`cat{rep{2,2 lit{x}}alt{emp{}cc{0x30-0x39}}}`},
{`x{2}y|x{2}[0-9]y`,
`cat{rep{2,2 lit{x}}alt{lit{y}cat{cc{0x30-0x39}lit{y}}}}`},
{`a.*?c|a.*?b`,
`cat{lit{a}alt{cat{nstar{dot{}}lit{c}}cat{nstar{dot{}}lit{b}}}}`},
// Valid repetitions.
{`((((((((((x{2}){2}){2}){2}){2}){2}){2}){2}){2}))`, ``},
{`((((((((((x{1}){2}){2}){2}){2}){2}){2}){2}){2}){2})`, ``},
// Valid nesting.
{strings.Repeat("(", 999) + strings.Repeat(")", 999), ``},
{strings.Repeat("(?:", 999) + strings.Repeat(")*", 999), ``},
{"(" + strings.Repeat("|", 12345) + ")", ``}, // not nested at all
}
const testFlags = MatchNL | PerlX | UnicodeGroups
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
func TestParseSimple(t *testing.T) {
testParseDump(t, parseTests, testFlags)
}
var foldcaseTests = []parseTest{
{`AbCdE`, `strfold{ABCDE}`},
{`[Aa]`, `litfold{A}`},
{`a`, `litfold{A}`},
// 0x17F is an old English long s (looks like an f) and folds to s.
// 0x212A is the Kelvin symbol and folds to k.
{`A[F-g]`, `cat{litfold{A}cc{0x41-0x7a 0x17f 0x212a}}`}, // [Aa][A-z...]
{`[[:upper:]]`, `cc{0x41-0x5a 0x61-0x7a 0x17f 0x212a}`},
{`[[:lower:]]`, `cc{0x41-0x5a 0x61-0x7a 0x17f 0x212a}`},
}
func TestParseFoldCase(t *testing.T) {
testParseDump(t, foldcaseTests, FoldCase)
}
var literalTests = []parseTest{
{"(|)^$.[*+?]{5,10},\\", "str{(|)^$.[*+?]{5,10},\\}"},
}
func TestParseLiteral(t *testing.T) {
testParseDump(t, literalTests, Literal)
}
var matchnlTests = []parseTest{
{`.`, `dot{}`},
{"\n", "lit{\n}"},
{`[^a]`, `cc{0x0-0x60 0x62-0x10ffff}`},
{`[a\n]`, `cc{0xa 0x61}`},
}
func TestParseMatchNL(t *testing.T) {
testParseDump(t, matchnlTests, MatchNL)
}
var nomatchnlTests = []parseTest{
{`.`, `dnl{}`},
{"\n", "lit{\n}"},
{`[^a]`, `cc{0x0-0x9 0xb-0x60 0x62-0x10ffff}`},
{`[a\n]`, `cc{0xa 0x61}`},
}
func TestParseNoMatchNL(t *testing.T) {
testParseDump(t, nomatchnlTests, 0)
}
func testParseDump(t *testing.T, tests []parseTest, flags Flags) {
for _, tt := range tests {
re, err := Parse(tt.Regexp, flags)
if err != nil {
t.Errorf("Parse(%#q): %v", tt.Regexp, err)
continue
}
if tt.Dump == "" {
// It parsed. That's all we care about.
continue
}
d := dump(re)
if d != tt.Dump {
t.Errorf("Parse(%#q).Dump() = %#q want %#q", tt.Regexp, d, tt.Dump)
}
}
}
// dump prints a string representation of the regexp showing
// the structure explicitly.
func dump(re *Regexp) string {
var b strings.Builder
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
dumpRegexp(&b, re)
return b.String()
}
var opNames = []string{
OpNoMatch: "no",
OpEmptyMatch: "emp",
OpLiteral: "lit",
OpCharClass: "cc",
OpAnyCharNotNL: "dnl",
OpAnyChar: "dot",
OpBeginLine: "bol",
OpEndLine: "eol",
OpBeginText: "bot",
OpEndText: "eot",
OpWordBoundary: "wb",
OpNoWordBoundary: "nwb",
OpCapture: "cap",
OpStar: "star",
OpPlus: "plus",
OpQuest: "que",
OpRepeat: "rep",
OpConcat: "cat",
OpAlternate: "alt",
}
// dumpRegexp writes an encoding of the syntax tree for the regexp re to b.
// It is used during testing to distinguish between parses that might print
// the same using re's String method.
func dumpRegexp(b *strings.Builder, re *Regexp) {
if int(re.Op) >= len(opNames) || opNames[re.Op] == "" {
fmt.Fprintf(b, "op%d", re.Op)
} else {
switch re.Op {
default:
b.WriteString(opNames[re.Op])
case OpStar, OpPlus, OpQuest, OpRepeat:
if re.Flags&NonGreedy != 0 {
b.WriteByte('n')
}
b.WriteString(opNames[re.Op])
case OpLiteral:
if len(re.Rune) > 1 {
b.WriteString("str")
} else {
b.WriteString("lit")
}
if re.Flags&FoldCase != 0 {
for _, r := range re.Rune {
if unicode.SimpleFold(r) != r {
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
}
}
}
}
}
b.WriteByte('{')
switch re.Op {
case OpEndText:
if re.Flags&WasDollar == 0 {
b.WriteString(`\z`)
}
case OpLiteral:
for _, r := range re.Rune {
b.WriteRune(r)
}
case OpConcat, OpAlternate:
for _, sub := range re.Sub {
dumpRegexp(b, sub)
}
case OpStar, OpPlus, OpQuest:
dumpRegexp(b, re.Sub[0])
case OpRepeat:
fmt.Fprintf(b, "%d,%d ", re.Min, re.Max)
dumpRegexp(b, re.Sub[0])
case OpCapture:
if re.Name != "" {
b.WriteString(re.Name)
b.WriteByte(':')
}
dumpRegexp(b, re.Sub[0])
case OpCharClass:
sep := ""
for i := 0; i < len(re.Rune); i += 2 {
b.WriteString(sep)
sep = " "
lo, hi := re.Rune[i], re.Rune[i+1]
if lo == hi {
fmt.Fprintf(b, "%#x", lo)
} else {
fmt.Fprintf(b, "%#x-%#x", lo, hi)
}
}
}
b.WriteByte('}')
}
re := &Regexp{Op: OpCharClass}
lo := rune(-1)
for i := rune(0); i <= unicode.MaxRune; i++ {
if f(i) {
if lo < 0 {
lo = i
}
} else {
if lo >= 0 {
re.Rune = append(re.Rune, lo, i-1)
lo = -1
}
}
}
if lo >= 0 {
re.Rune = append(re.Rune, lo, unicode.MaxRune)
}
return dump(re)
}
if unicode.IsUpper(c) {
return true
}
c = unicode.SimpleFold(c)
}
return false
}
func TestFoldConstants(t *testing.T) {
last := rune(-1)
for i := rune(0); i <= unicode.MaxRune; i++ {
if unicode.SimpleFold(i) == i {
continue
}
if last == -1 && minFold != i {
t.Errorf("minFold=%#U should be %#U", minFold, i)
}
last = i
}
if maxFold != last {
t.Errorf("maxFold=%#U should be %#U", maxFold, last)
}
}
func TestAppendRangeCollapse(t *testing.T) {
// AppendRange should collapse each of the new ranges
// into the earlier ones (it looks back two ranges), so that
// the slice never grows very large.
// Note that we are not calling cleanClass.
r = appendRange(r, i, i)
r = appendRange(r, i+'a'-'A', i+'a'-'A')
}
if string(r) != "AZaz" {
t.Errorf("appendRange interlaced A-Z a-z = %s, want AZaz", string(r))
}
}
var invalidRegexps = []string{
`(`,
`)`,
`(a`,
`a)`,
`(a))`,
`a|b|)`,
`(a|b|))`,
`a|b)`,
`(a|b))`,
`[a-z`,
`([a-z)`,
`[a-z)`,
`([a-z]))`,
`x{1001}`,
`x{9876543210}`,
`x{2,1}`,
`x{1,9876543210}`,
"\xff", // Invalid UTF-8
"[\xff]",
"[\\\xff]",
"\\\xff",
`(?P<name>a`,
`(?P<name>`,
`(?P<name`,
`(?P<x y>a)`,
`(?P<>a)`,
`(?<name>a`,
`(?<name>`,
`(?<name`,
`(?<x y>a)`,
`(?<>a)`,
`[a-Z]`,
`(?i)[a-Z]`,
`a{100000}`, // too much repetition
`a{100000,}`, // too much repetition
"((((((((((x{2}){2}){2}){2}){2}){2}){2}){2}){2}){2})", // too much repetition
strings.Repeat("(", 1000) + strings.Repeat(")", 1000), // too deep
strings.Repeat("(?:", 1000) + strings.Repeat(")*", 1000), // too deep
"(" + strings.Repeat("(xx?)", 1000) + "){1000}", // too long
strings.Repeat("(xx?){1000}", 1000), // too long
strings.Repeat(`\pL`, 27000), // too many runes
}
var onlyPerl = []string{
`[a-b-c]`,
`\Qabc\E`,
`\Q*+?{[\E`,
`\Q\\E`,
`\Q\\\E`,
`\Q\\\\E`,
`\Q\\\\\E`,
`(?:a)`,
`(?P<name>a)`,
}
var onlyPOSIX = []string{
"a++",
"a**",
"a?*",
"a+*",
"a{1}*",
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
}
func TestParseInvalidRegexps(t *testing.T) {
for _, regexp := range invalidRegexps {
if re, err := Parse(regexp, Perl); err == nil {
t.Errorf("Parse(%#q, Perl) = %s, should have failed", regexp, dump(re))
}
if re, err := Parse(regexp, POSIX); err == nil {
t.Errorf("Parse(%#q, POSIX) = %s, should have failed", regexp, dump(re))
}
}
for _, regexp := range onlyPerl {
if _, err := Parse(regexp, Perl); err != nil {
t.Errorf("Parse(%#q, Perl): %v", regexp, err)
}
if re, err := Parse(regexp, POSIX); err == nil {
t.Errorf("Parse(%#q, POSIX) = %s, should have failed", regexp, dump(re))
}
}
for _, regexp := range onlyPOSIX {
if re, err := Parse(regexp, Perl); err == nil {
t.Errorf("Parse(%#q, Perl) = %s, should have failed", regexp, dump(re))
}
if _, err := Parse(regexp, POSIX); err != nil {
t.Errorf("Parse(%#q, POSIX): %v", regexp, err)
}
}
}
func TestToStringEquivalentParse(t *testing.T) {
for _, tt := range parseTests {
re, err := Parse(tt.Regexp, testFlags)
if err != nil {
t.Errorf("Parse(%#q): %v", tt.Regexp, err)
continue
}
if tt.Dump == "" {
// It parsed. That's all we care about.
continue
}
d := dump(re)
if d != tt.Dump {
t.Errorf("Parse(%#q).Dump() = %#q want %#q", tt.Regexp, d, tt.Dump)
continue
}
s := re.String()
if s != tt.Regexp {
// If ToString didn't return the original regexp,
// it must have found one with fewer parens.
// Unfortunately we can't check the length here, because
// ToString produces "\\{" for a literal brace,
// but "{" is a shorter equivalent in some contexts.
nre, err := Parse(s, testFlags)
if err != nil {
t.Errorf("Parse(%#q.String() = %#q): %v", tt.Regexp, s, err)
continue
}
nd := dump(nre)
if d != nd {
t.Errorf("Parse(%#q) -> %#q; %#q vs %#q", tt.Regexp, s, d, nd)
}
ns := nre.String()
if s != ns {
t.Errorf("Parse(%#q) -> %#q -> %#q", tt.Regexp, s, ns)
}
}
}
}
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
var stringTests = []struct {
re string
out string
}{
{`x(?i:ab*c|d?e)1`, `x(?i:AB*C|D?E)1`},
{`x(?i:ab*cd?e)1`, `x(?i:AB*CD?E)1`},
{`0(?i:ab*c|d?e)1`, `(?i:0(?:AB*C|D?E)1)`},
{`0(?i:ab*cd?e)1`, `(?i:0AB*CD?E1)`},
{`x(?i:ab*c|d?e)`, `x(?i:AB*C|D?E)`},
{`x(?i:ab*cd?e)`, `x(?i:AB*CD?E)`},
{`0(?i:ab*c|d?e)`, `(?i:0(?:AB*C|D?E))`},
{`0(?i:ab*cd?e)`, `(?i:0AB*CD?E)`},
{`(?i:ab*c|d?e)1`, `(?i:(?:AB*C|D?E)1)`},
{`(?i:ab*cd?e)1`, `(?i:AB*CD?E1)`},
{`(?i:ab)[123](?i:cd)`, `(?i:AB[1-3]CD)`},
{`(?i:ab*c|d?e)`, `(?i:AB*C|D?E)`},
{`[Aa][Bb]`, `(?i:AB)`},
{`[Aa][Bb]*[Cc]`, `(?i:AB*C)`},
{`A(?:[Bb][Cc]|[Dd])[Zz]`, `A(?i:(?:BC|D)Z)`},
{`[Aa](?:[Bb][Cc]|[Dd])Z`, `(?i:A(?:BC|D))Z`},
}
func TestString(t *testing.T) {
for _, tt := range stringTests {
re, err := Parse(tt.re, Perl)
if err != nil {
t.Errorf("Parse(%#q): %v", tt.re, err)
continue
}
out := re.String()
if out != tt.out {
t.Errorf("Parse(%#q).String() = %#q, want %#q", tt.re, out, tt.out)
}
}
}