blob: b615acdf0e54f2db8cac515ab41bf81d76cf520d [file] [log] [blame]
Bill Yi7fb3c4c2015-03-23 09:04:07 -07001// 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 regexp implements regular expression search.
6//
7// The syntax of the regular expressions accepted is the same
8// general syntax used by Perl, Python, and other languages.
9// More precisely, it is the syntax accepted by RE2 and described at
10// http://code.google.com/p/re2/wiki/Syntax, except for \C.
11// For an overview of the syntax, run
12// godoc regexp/syntax
13//
14// The regexp implementation provided by this package is
15// guaranteed to run in time linear in the size of the input.
16// (This is a property not guaranteed by most open source
17// implementations of regular expressions.) For more information
18// about this property, see
19// http://swtch.com/~rsc/regexp/regexp1.html
20// or any book about automata theory.
21//
22// All characters are UTF-8-encoded code points.
23//
24// There are 16 methods of Regexp that match a regular expression and identify
25// the matched text. Their names are matched by this regular expression:
26//
27// Find(All)?(String)?(Submatch)?(Index)?
28//
29// If 'All' is present, the routine matches successive non-overlapping
30// matches of the entire expression. Empty matches abutting a preceding
31// match are ignored. The return value is a slice containing the successive
32// return values of the corresponding non-'All' routine. These routines take
33// an extra integer argument, n; if n >= 0, the function returns at most n
34// matches/submatches.
35//
36// If 'String' is present, the argument is a string; otherwise it is a slice
37// of bytes; return values are adjusted as appropriate.
38//
39// If 'Submatch' is present, the return value is a slice identifying the
40// successive submatches of the expression. Submatches are matches of
41// parenthesized subexpressions (also known as capturing groups) within the
42// regular expression, numbered from left to right in order of opening
43// parenthesis. Submatch 0 is the match of the entire expression, submatch 1
44// the match of the first parenthesized subexpression, and so on.
45//
46// If 'Index' is present, matches and submatches are identified by byte index
47// pairs within the input string: result[2*n:2*n+1] identifies the indexes of
48// the nth submatch. The pair for n==0 identifies the match of the entire
49// expression. If 'Index' is not present, the match is identified by the
50// text of the match/submatch. If an index is negative, it means that
51// subexpression did not match any string in the input.
52//
53// There is also a subset of the methods that can be applied to text read
54// from a RuneReader:
55//
56// MatchReader, FindReaderIndex, FindReaderSubmatchIndex
57//
58// This set may grow. Note that regular expression matches may need to
59// examine text beyond the text returned by a match, so the methods that
60// match text from a RuneReader may read arbitrarily far into the input
61// before returning.
62//
63// (There are a few other methods that do not match this pattern.)
64//
65package regexp
66
67import (
68 "bytes"
69 "io"
70 "regexp/syntax"
71 "strconv"
72 "strings"
73 "sync"
74 "unicode"
75 "unicode/utf8"
76)
77
78var debug = false
79
80// Regexp is the representation of a compiled regular expression.
81// A Regexp is safe for concurrent use by multiple goroutines.
82type Regexp struct {
83 // read-only after Compile
84 expr string // as passed to Compile
85 prog *syntax.Prog // compiled program
86 onepass *onePassProg // onpass program or nil
87 prefix string // required prefix in unanchored matches
88 prefixBytes []byte // prefix, as a []byte
89 prefixComplete bool // prefix is the entire regexp
90 prefixRune rune // first rune in prefix
91 prefixEnd uint32 // pc for last rune in prefix
92 cond syntax.EmptyOp // empty-width conditions required at start of match
93 numSubexp int
94 subexpNames []string
95 longest bool
96
97 // cache of machines for running regexp
98 mu sync.Mutex
99 machine []*machine
100}
101
102// String returns the source text used to compile the regular expression.
103func (re *Regexp) String() string {
104 return re.expr
105}
106
107// Compile parses a regular expression and returns, if successful,
108// a Regexp object that can be used to match against text.
109//
110// When matching against text, the regexp returns a match that
111// begins as early as possible in the input (leftmost), and among those
112// it chooses the one that a backtracking search would have found first.
113// This so-called leftmost-first matching is the same semantics
114// that Perl, Python, and other implementations use, although this
115// package implements it without the expense of backtracking.
116// For POSIX leftmost-longest matching, see CompilePOSIX.
117func Compile(expr string) (*Regexp, error) {
118 return compile(expr, syntax.Perl, false)
119}
120
121// CompilePOSIX is like Compile but restricts the regular expression
122// to POSIX ERE (egrep) syntax and changes the match semantics to
123// leftmost-longest.
124//
125// That is, when matching against text, the regexp returns a match that
126// begins as early as possible in the input (leftmost), and among those
127// it chooses a match that is as long as possible.
128// This so-called leftmost-longest matching is the same semantics
129// that early regular expression implementations used and that POSIX
130// specifies.
131//
132// However, there can be multiple leftmost-longest matches, with different
133// submatch choices, and here this package diverges from POSIX.
134// Among the possible leftmost-longest matches, this package chooses
135// the one that a backtracking search would have found first, while POSIX
136// specifies that the match be chosen to maximize the length of the first
137// subexpression, then the second, and so on from left to right.
138// The POSIX rule is computationally prohibitive and not even well-defined.
139// See http://swtch.com/~rsc/regexp/regexp2.html#posix for details.
140func CompilePOSIX(expr string) (*Regexp, error) {
141 return compile(expr, syntax.POSIX, true)
142}
143
144// Longest makes future searches prefer the leftmost-longest match.
145// That is, when matching against text, the regexp returns a match that
146// begins as early as possible in the input (leftmost), and among those
147// it chooses a match that is as long as possible.
148func (re *Regexp) Longest() {
149 re.longest = true
150}
151
152func compile(expr string, mode syntax.Flags, longest bool) (*Regexp, error) {
153 re, err := syntax.Parse(expr, mode)
154 if err != nil {
155 return nil, err
156 }
157 maxCap := re.MaxCap()
158 capNames := re.CapNames()
159
160 re = re.Simplify()
161 prog, err := syntax.Compile(re)
162 if err != nil {
163 return nil, err
164 }
165 regexp := &Regexp{
166 expr: expr,
167 prog: prog,
168 onepass: compileOnePass(prog),
169 numSubexp: maxCap,
170 subexpNames: capNames,
171 cond: prog.StartCond(),
172 longest: longest,
173 }
174 if regexp.onepass == notOnePass {
175 regexp.prefix, regexp.prefixComplete = prog.Prefix()
176 } else {
177 regexp.prefix, regexp.prefixComplete, regexp.prefixEnd = onePassPrefix(prog)
178 }
179 if regexp.prefix != "" {
180 // TODO(rsc): Remove this allocation by adding
181 // IndexString to package bytes.
182 regexp.prefixBytes = []byte(regexp.prefix)
183 regexp.prefixRune, _ = utf8.DecodeRuneInString(regexp.prefix)
184 }
185 return regexp, nil
186}
187
188// get returns a machine to use for matching re.
189// It uses the re's machine cache if possible, to avoid
190// unnecessary allocation.
191func (re *Regexp) get() *machine {
192 re.mu.Lock()
193 if n := len(re.machine); n > 0 {
194 z := re.machine[n-1]
195 re.machine = re.machine[:n-1]
196 re.mu.Unlock()
197 return z
198 }
199 re.mu.Unlock()
200 z := progMachine(re.prog, re.onepass)
201 z.re = re
202 return z
203}
204
205// put returns a machine to the re's machine cache.
206// There is no attempt to limit the size of the cache, so it will
207// grow to the maximum number of simultaneous matches
208// run using re. (The cache empties when re gets garbage collected.)
209func (re *Regexp) put(z *machine) {
210 re.mu.Lock()
211 re.machine = append(re.machine, z)
212 re.mu.Unlock()
213}
214
215// MustCompile is like Compile but panics if the expression cannot be parsed.
216// It simplifies safe initialization of global variables holding compiled regular
217// expressions.
218func MustCompile(str string) *Regexp {
219 regexp, error := Compile(str)
220 if error != nil {
221 panic(`regexp: Compile(` + quote(str) + `): ` + error.Error())
222 }
223 return regexp
224}
225
226// MustCompilePOSIX is like CompilePOSIX but panics if the expression cannot be parsed.
227// It simplifies safe initialization of global variables holding compiled regular
228// expressions.
229func MustCompilePOSIX(str string) *Regexp {
230 regexp, error := CompilePOSIX(str)
231 if error != nil {
232 panic(`regexp: CompilePOSIX(` + quote(str) + `): ` + error.Error())
233 }
234 return regexp
235}
236
237func quote(s string) string {
238 if strconv.CanBackquote(s) {
239 return "`" + s + "`"
240 }
241 return strconv.Quote(s)
242}
243
244// NumSubexp returns the number of parenthesized subexpressions in this Regexp.
245func (re *Regexp) NumSubexp() int {
246 return re.numSubexp
247}
248
249// SubexpNames returns the names of the parenthesized subexpressions
250// in this Regexp. The name for the first sub-expression is names[1],
251// so that if m is a match slice, the name for m[i] is SubexpNames()[i].
252// Since the Regexp as a whole cannot be named, names[0] is always
253// the empty string. The slice should not be modified.
254func (re *Regexp) SubexpNames() []string {
255 return re.subexpNames
256}
257
258const endOfText rune = -1
259
260// input abstracts different representations of the input text. It provides
261// one-character lookahead.
262type input interface {
263 step(pos int) (r rune, width int) // advance one rune
264 canCheckPrefix() bool // can we look ahead without losing info?
265 hasPrefix(re *Regexp) bool
266 index(re *Regexp, pos int) int
267 context(pos int) syntax.EmptyOp
268}
269
270// inputString scans a string.
271type inputString struct {
272 str string
273}
274
275func (i *inputString) step(pos int) (rune, int) {
276 if pos < len(i.str) {
277 c := i.str[pos]
278 if c < utf8.RuneSelf {
279 return rune(c), 1
280 }
281 return utf8.DecodeRuneInString(i.str[pos:])
282 }
283 return endOfText, 0
284}
285
286func (i *inputString) canCheckPrefix() bool {
287 return true
288}
289
290func (i *inputString) hasPrefix(re *Regexp) bool {
291 return strings.HasPrefix(i.str, re.prefix)
292}
293
294func (i *inputString) index(re *Regexp, pos int) int {
295 return strings.Index(i.str[pos:], re.prefix)
296}
297
298func (i *inputString) context(pos int) syntax.EmptyOp {
299 r1, r2 := endOfText, endOfText
300 if pos > 0 && pos <= len(i.str) {
301 r1, _ = utf8.DecodeLastRuneInString(i.str[:pos])
302 }
303 if pos < len(i.str) {
304 r2, _ = utf8.DecodeRuneInString(i.str[pos:])
305 }
306 return syntax.EmptyOpContext(r1, r2)
307}
308
309// inputBytes scans a byte slice.
310type inputBytes struct {
311 str []byte
312}
313
314func (i *inputBytes) step(pos int) (rune, int) {
315 if pos < len(i.str) {
316 c := i.str[pos]
317 if c < utf8.RuneSelf {
318 return rune(c), 1
319 }
320 return utf8.DecodeRune(i.str[pos:])
321 }
322 return endOfText, 0
323}
324
325func (i *inputBytes) canCheckPrefix() bool {
326 return true
327}
328
329func (i *inputBytes) hasPrefix(re *Regexp) bool {
330 return bytes.HasPrefix(i.str, re.prefixBytes)
331}
332
333func (i *inputBytes) index(re *Regexp, pos int) int {
334 return bytes.Index(i.str[pos:], re.prefixBytes)
335}
336
337func (i *inputBytes) context(pos int) syntax.EmptyOp {
338 r1, r2 := endOfText, endOfText
339 if pos > 0 && pos <= len(i.str) {
340 r1, _ = utf8.DecodeLastRune(i.str[:pos])
341 }
342 if pos < len(i.str) {
343 r2, _ = utf8.DecodeRune(i.str[pos:])
344 }
345 return syntax.EmptyOpContext(r1, r2)
346}
347
348// inputReader scans a RuneReader.
349type inputReader struct {
350 r io.RuneReader
351 atEOT bool
352 pos int
353}
354
355func (i *inputReader) step(pos int) (rune, int) {
356 if !i.atEOT && pos != i.pos {
357 return endOfText, 0
358
359 }
360 r, w, err := i.r.ReadRune()
361 if err != nil {
362 i.atEOT = true
363 return endOfText, 0
364 }
365 i.pos += w
366 return r, w
367}
368
369func (i *inputReader) canCheckPrefix() bool {
370 return false
371}
372
373func (i *inputReader) hasPrefix(re *Regexp) bool {
374 return false
375}
376
377func (i *inputReader) index(re *Regexp, pos int) int {
378 return -1
379}
380
381func (i *inputReader) context(pos int) syntax.EmptyOp {
382 return 0
383}
384
385// LiteralPrefix returns a literal string that must begin any match
386// of the regular expression re. It returns the boolean true if the
387// literal string comprises the entire regular expression.
388func (re *Regexp) LiteralPrefix() (prefix string, complete bool) {
389 return re.prefix, re.prefixComplete
390}
391
392// MatchReader reports whether the Regexp matches the text read by the
393// RuneReader.
394func (re *Regexp) MatchReader(r io.RuneReader) bool {
395 return re.doExecute(r, nil, "", 0, 0) != nil
396}
397
398// MatchString reports whether the Regexp matches the string s.
399func (re *Regexp) MatchString(s string) bool {
400 return re.doExecute(nil, nil, s, 0, 0) != nil
401}
402
403// Match reports whether the Regexp matches the byte slice b.
404func (re *Regexp) Match(b []byte) bool {
405 return re.doExecute(nil, b, "", 0, 0) != nil
406}
407
408// MatchReader checks whether a textual regular expression matches the text
409// read by the RuneReader. More complicated queries need to use Compile and
410// the full Regexp interface.
411func MatchReader(pattern string, r io.RuneReader) (matched bool, err error) {
412 re, err := Compile(pattern)
413 if err != nil {
414 return false, err
415 }
416 return re.MatchReader(r), nil
417}
418
419// MatchString checks whether a textual regular expression
420// matches a string. More complicated queries need
421// to use Compile and the full Regexp interface.
422func MatchString(pattern string, s string) (matched bool, err error) {
423 re, err := Compile(pattern)
424 if err != nil {
425 return false, err
426 }
427 return re.MatchString(s), nil
428}
429
430// Match checks whether a textual regular expression
431// matches a byte slice. More complicated queries need
432// to use Compile and the full Regexp interface.
433func Match(pattern string, b []byte) (matched bool, err error) {
434 re, err := Compile(pattern)
435 if err != nil {
436 return false, err
437 }
438 return re.Match(b), nil
439}
440
441// ReplaceAllString returns a copy of src, replacing matches of the Regexp
442// with the replacement string repl. Inside repl, $ signs are interpreted as
443// in Expand, so for instance $1 represents the text of the first submatch.
444func (re *Regexp) ReplaceAllString(src, repl string) string {
445 n := 2
446 if strings.Index(repl, "$") >= 0 {
447 n = 2 * (re.numSubexp + 1)
448 }
449 b := re.replaceAll(nil, src, n, func(dst []byte, match []int) []byte {
450 return re.expand(dst, repl, nil, src, match)
451 })
452 return string(b)
453}
454
455// ReplaceAllLiteralString returns a copy of src, replacing matches of the Regexp
456// with the replacement string repl. The replacement repl is substituted directly,
457// without using Expand.
458func (re *Regexp) ReplaceAllLiteralString(src, repl string) string {
459 return string(re.replaceAll(nil, src, 2, func(dst []byte, match []int) []byte {
460 return append(dst, repl...)
461 }))
462}
463
464// ReplaceAllStringFunc returns a copy of src in which all matches of the
465// Regexp have been replaced by the return value of function repl applied
466// to the matched substring. The replacement returned by repl is substituted
467// directly, without using Expand.
468func (re *Regexp) ReplaceAllStringFunc(src string, repl func(string) string) string {
469 b := re.replaceAll(nil, src, 2, func(dst []byte, match []int) []byte {
470 return append(dst, repl(src[match[0]:match[1]])...)
471 })
472 return string(b)
473}
474
475func (re *Regexp) replaceAll(bsrc []byte, src string, nmatch int, repl func(dst []byte, m []int) []byte) []byte {
476 lastMatchEnd := 0 // end position of the most recent match
477 searchPos := 0 // position where we next look for a match
478 var buf []byte
479 var endPos int
480 if bsrc != nil {
481 endPos = len(bsrc)
482 } else {
483 endPos = len(src)
484 }
485 for searchPos <= endPos {
486 a := re.doExecute(nil, bsrc, src, searchPos, nmatch)
487 if len(a) == 0 {
488 break // no more matches
489 }
490
491 // Copy the unmatched characters before this match.
492 if bsrc != nil {
493 buf = append(buf, bsrc[lastMatchEnd:a[0]]...)
494 } else {
495 buf = append(buf, src[lastMatchEnd:a[0]]...)
496 }
497
498 // Now insert a copy of the replacement string, but not for a
499 // match of the empty string immediately after another match.
500 // (Otherwise, we get double replacement for patterns that
501 // match both empty and nonempty strings.)
502 if a[1] > lastMatchEnd || a[0] == 0 {
503 buf = repl(buf, a)
504 }
505 lastMatchEnd = a[1]
506
507 // Advance past this match; always advance at least one character.
508 var width int
509 if bsrc != nil {
510 _, width = utf8.DecodeRune(bsrc[searchPos:])
511 } else {
512 _, width = utf8.DecodeRuneInString(src[searchPos:])
513 }
514 if searchPos+width > a[1] {
515 searchPos += width
516 } else if searchPos+1 > a[1] {
517 // This clause is only needed at the end of the input
518 // string. In that case, DecodeRuneInString returns width=0.
519 searchPos++
520 } else {
521 searchPos = a[1]
522 }
523 }
524
525 // Copy the unmatched characters after the last match.
526 if bsrc != nil {
527 buf = append(buf, bsrc[lastMatchEnd:]...)
528 } else {
529 buf = append(buf, src[lastMatchEnd:]...)
530 }
531
532 return buf
533}
534
535// ReplaceAll returns a copy of src, replacing matches of the Regexp
536// with the replacement text repl. Inside repl, $ signs are interpreted as
537// in Expand, so for instance $1 represents the text of the first submatch.
538func (re *Regexp) ReplaceAll(src, repl []byte) []byte {
539 n := 2
540 if bytes.IndexByte(repl, '$') >= 0 {
541 n = 2 * (re.numSubexp + 1)
542 }
543 srepl := ""
544 b := re.replaceAll(src, "", n, func(dst []byte, match []int) []byte {
545 if len(srepl) != len(repl) {
546 srepl = string(repl)
547 }
548 return re.expand(dst, srepl, src, "", match)
549 })
550 return b
551}
552
553// ReplaceAllLiteral returns a copy of src, replacing matches of the Regexp
554// with the replacement bytes repl. The replacement repl is substituted directly,
555// without using Expand.
556func (re *Regexp) ReplaceAllLiteral(src, repl []byte) []byte {
557 return re.replaceAll(src, "", 2, func(dst []byte, match []int) []byte {
558 return append(dst, repl...)
559 })
560}
561
562// ReplaceAllFunc returns a copy of src in which all matches of the
563// Regexp have been replaced by the return value of function repl applied
564// to the matched byte slice. The replacement returned by repl is substituted
565// directly, without using Expand.
566func (re *Regexp) ReplaceAllFunc(src []byte, repl func([]byte) []byte) []byte {
567 return re.replaceAll(src, "", 2, func(dst []byte, match []int) []byte {
568 return append(dst, repl(src[match[0]:match[1]])...)
569 })
570}
571
572var specialBytes = []byte(`\.+*?()|[]{}^$`)
573
574func special(b byte) bool {
575 return bytes.IndexByte(specialBytes, b) >= 0
576}
577
578// QuoteMeta returns a string that quotes all regular expression metacharacters
579// inside the argument text; the returned string is a regular expression matching
580// the literal text. For example, QuoteMeta(`[foo]`) returns `\[foo\]`.
581func QuoteMeta(s string) string {
582 b := make([]byte, 2*len(s))
583
584 // A byte loop is correct because all metacharacters are ASCII.
585 j := 0
586 for i := 0; i < len(s); i++ {
587 if special(s[i]) {
588 b[j] = '\\'
589 j++
590 }
591 b[j] = s[i]
592 j++
593 }
594 return string(b[0:j])
595}
596
597// The number of capture values in the program may correspond
598// to fewer capturing expressions than are in the regexp.
599// For example, "(a){0}" turns into an empty program, so the
600// maximum capture in the program is 0 but we need to return
601// an expression for \1. Pad appends -1s to the slice a as needed.
602func (re *Regexp) pad(a []int) []int {
603 if a == nil {
604 // No match.
605 return nil
606 }
607 n := (1 + re.numSubexp) * 2
608 for len(a) < n {
609 a = append(a, -1)
610 }
611 return a
612}
613
614// Find matches in slice b if b is non-nil, otherwise find matches in string s.
615func (re *Regexp) allMatches(s string, b []byte, n int, deliver func([]int)) {
616 var end int
617 if b == nil {
618 end = len(s)
619 } else {
620 end = len(b)
621 }
622
623 for pos, i, prevMatchEnd := 0, 0, -1; i < n && pos <= end; {
624 matches := re.doExecute(nil, b, s, pos, re.prog.NumCap)
625 if len(matches) == 0 {
626 break
627 }
628
629 accept := true
630 if matches[1] == pos {
631 // We've found an empty match.
632 if matches[0] == prevMatchEnd {
633 // We don't allow an empty match right
634 // after a previous match, so ignore it.
635 accept = false
636 }
637 var width int
638 // TODO: use step()
639 if b == nil {
640 _, width = utf8.DecodeRuneInString(s[pos:end])
641 } else {
642 _, width = utf8.DecodeRune(b[pos:end])
643 }
644 if width > 0 {
645 pos += width
646 } else {
647 pos = end + 1
648 }
649 } else {
650 pos = matches[1]
651 }
652 prevMatchEnd = matches[1]
653
654 if accept {
655 deliver(re.pad(matches))
656 i++
657 }
658 }
659}
660
661// Find returns a slice holding the text of the leftmost match in b of the regular expression.
662// A return value of nil indicates no match.
663func (re *Regexp) Find(b []byte) []byte {
664 a := re.doExecute(nil, b, "", 0, 2)
665 if a == nil {
666 return nil
667 }
668 return b[a[0]:a[1]]
669}
670
671// FindIndex returns a two-element slice of integers defining the location of
672// the leftmost match in b of the regular expression. The match itself is at
673// b[loc[0]:loc[1]].
674// A return value of nil indicates no match.
675func (re *Regexp) FindIndex(b []byte) (loc []int) {
676 a := re.doExecute(nil, b, "", 0, 2)
677 if a == nil {
678 return nil
679 }
680 return a[0:2]
681}
682
683// FindString returns a string holding the text of the leftmost match in s of the regular
684// expression. If there is no match, the return value is an empty string,
685// but it will also be empty if the regular expression successfully matches
686// an empty string. Use FindStringIndex or FindStringSubmatch if it is
687// necessary to distinguish these cases.
688func (re *Regexp) FindString(s string) string {
689 a := re.doExecute(nil, nil, s, 0, 2)
690 if a == nil {
691 return ""
692 }
693 return s[a[0]:a[1]]
694}
695
696// FindStringIndex returns a two-element slice of integers defining the
697// location of the leftmost match in s of the regular expression. The match
698// itself is at s[loc[0]:loc[1]].
699// A return value of nil indicates no match.
700func (re *Regexp) FindStringIndex(s string) (loc []int) {
701 a := re.doExecute(nil, nil, s, 0, 2)
702 if a == nil {
703 return nil
704 }
705 return a[0:2]
706}
707
708// FindReaderIndex returns a two-element slice of integers defining the
709// location of the leftmost match of the regular expression in text read from
710// the RuneReader. The match text was found in the input stream at
711// byte offset loc[0] through loc[1]-1.
712// A return value of nil indicates no match.
713func (re *Regexp) FindReaderIndex(r io.RuneReader) (loc []int) {
714 a := re.doExecute(r, nil, "", 0, 2)
715 if a == nil {
716 return nil
717 }
718 return a[0:2]
719}
720
721// FindSubmatch returns a slice of slices holding the text of the leftmost
722// match of the regular expression in b and the matches, if any, of its
723// subexpressions, as defined by the 'Submatch' descriptions in the package
724// comment.
725// A return value of nil indicates no match.
726func (re *Regexp) FindSubmatch(b []byte) [][]byte {
727 a := re.doExecute(nil, b, "", 0, re.prog.NumCap)
728 if a == nil {
729 return nil
730 }
731 ret := make([][]byte, 1+re.numSubexp)
732 for i := range ret {
733 if 2*i < len(a) && a[2*i] >= 0 {
734 ret[i] = b[a[2*i]:a[2*i+1]]
735 }
736 }
737 return ret
738}
739
740// Expand appends template to dst and returns the result; during the
741// append, Expand replaces variables in the template with corresponding
742// matches drawn from src. The match slice should have been returned by
743// FindSubmatchIndex.
744//
745// In the template, a variable is denoted by a substring of the form
746// $name or ${name}, where name is a non-empty sequence of letters,
747// digits, and underscores. A purely numeric name like $1 refers to
748// the submatch with the corresponding index; other names refer to
749// capturing parentheses named with the (?P<name>...) syntax. A
750// reference to an out of range or unmatched index or a name that is not
751// present in the regular expression is replaced with an empty slice.
752//
753// In the $name form, name is taken to be as long as possible: $1x is
754// equivalent to ${1x}, not ${1}x, and, $10 is equivalent to ${10}, not ${1}0.
755//
756// To insert a literal $ in the output, use $$ in the template.
757func (re *Regexp) Expand(dst []byte, template []byte, src []byte, match []int) []byte {
758 return re.expand(dst, string(template), src, "", match)
759}
760
761// ExpandString is like Expand but the template and source are strings.
762// It appends to and returns a byte slice in order to give the calling
763// code control over allocation.
764func (re *Regexp) ExpandString(dst []byte, template string, src string, match []int) []byte {
765 return re.expand(dst, template, nil, src, match)
766}
767
768func (re *Regexp) expand(dst []byte, template string, bsrc []byte, src string, match []int) []byte {
769 for len(template) > 0 {
770 i := strings.Index(template, "$")
771 if i < 0 {
772 break
773 }
774 dst = append(dst, template[:i]...)
775 template = template[i:]
776 if len(template) > 1 && template[1] == '$' {
777 // Treat $$ as $.
778 dst = append(dst, '$')
779 template = template[2:]
780 continue
781 }
782 name, num, rest, ok := extract(template)
783 if !ok {
784 // Malformed; treat $ as raw text.
785 dst = append(dst, '$')
786 template = template[1:]
787 continue
788 }
789 template = rest
790 if num >= 0 {
791 if 2*num+1 < len(match) && match[2*num] >= 0 {
792 if bsrc != nil {
793 dst = append(dst, bsrc[match[2*num]:match[2*num+1]]...)
794 } else {
795 dst = append(dst, src[match[2*num]:match[2*num+1]]...)
796 }
797 }
798 } else {
799 for i, namei := range re.subexpNames {
800 if name == namei && 2*i+1 < len(match) && match[2*i] >= 0 {
801 if bsrc != nil {
802 dst = append(dst, bsrc[match[2*i]:match[2*i+1]]...)
803 } else {
804 dst = append(dst, src[match[2*i]:match[2*i+1]]...)
805 }
806 break
807 }
808 }
809 }
810 }
811 dst = append(dst, template...)
812 return dst
813}
814
815// extract returns the name from a leading "$name" or "${name}" in str.
816// If it is a number, extract returns num set to that number; otherwise num = -1.
817func extract(str string) (name string, num int, rest string, ok bool) {
818 if len(str) < 2 || str[0] != '$' {
819 return
820 }
821 brace := false
822 if str[1] == '{' {
823 brace = true
824 str = str[2:]
825 } else {
826 str = str[1:]
827 }
828 i := 0
829 for i < len(str) {
830 rune, size := utf8.DecodeRuneInString(str[i:])
831 if !unicode.IsLetter(rune) && !unicode.IsDigit(rune) && rune != '_' {
832 break
833 }
834 i += size
835 }
836 if i == 0 {
837 // empty name is not okay
838 return
839 }
840 name = str[:i]
841 if brace {
842 if i >= len(str) || str[i] != '}' {
843 // missing closing brace
844 return
845 }
846 i++
847 }
848
849 // Parse number.
850 num = 0
851 for i := 0; i < len(name); i++ {
852 if name[i] < '0' || '9' < name[i] || num >= 1e8 {
853 num = -1
854 break
855 }
856 num = num*10 + int(name[i]) - '0'
857 }
858 // Disallow leading zeros.
859 if name[0] == '0' && len(name) > 1 {
860 num = -1
861 }
862
863 rest = str[i:]
864 ok = true
865 return
866}
867
868// FindSubmatchIndex returns a slice holding the index pairs identifying the
869// leftmost match of the regular expression in b and the matches, if any, of
870// its subexpressions, as defined by the 'Submatch' and 'Index' descriptions
871// in the package comment.
872// A return value of nil indicates no match.
873func (re *Regexp) FindSubmatchIndex(b []byte) []int {
874 return re.pad(re.doExecute(nil, b, "", 0, re.prog.NumCap))
875}
876
877// FindStringSubmatch returns a slice of strings holding the text of the
878// leftmost match of the regular expression in s and the matches, if any, of
879// its subexpressions, as defined by the 'Submatch' description in the
880// package comment.
881// A return value of nil indicates no match.
882func (re *Regexp) FindStringSubmatch(s string) []string {
883 a := re.doExecute(nil, nil, s, 0, re.prog.NumCap)
884 if a == nil {
885 return nil
886 }
887 ret := make([]string, 1+re.numSubexp)
888 for i := range ret {
889 if 2*i < len(a) && a[2*i] >= 0 {
890 ret[i] = s[a[2*i]:a[2*i+1]]
891 }
892 }
893 return ret
894}
895
896// FindStringSubmatchIndex returns a slice holding the index pairs
897// identifying the leftmost match of the regular expression in s and the
898// matches, if any, of its subexpressions, as defined by the 'Submatch' and
899// 'Index' descriptions in the package comment.
900// A return value of nil indicates no match.
901func (re *Regexp) FindStringSubmatchIndex(s string) []int {
902 return re.pad(re.doExecute(nil, nil, s, 0, re.prog.NumCap))
903}
904
905// FindReaderSubmatchIndex returns a slice holding the index pairs
906// identifying the leftmost match of the regular expression of text read by
907// the RuneReader, and the matches, if any, of its subexpressions, as defined
908// by the 'Submatch' and 'Index' descriptions in the package comment. A
909// return value of nil indicates no match.
910func (re *Regexp) FindReaderSubmatchIndex(r io.RuneReader) []int {
911 return re.pad(re.doExecute(r, nil, "", 0, re.prog.NumCap))
912}
913
914const startSize = 10 // The size at which to start a slice in the 'All' routines.
915
916// FindAll is the 'All' version of Find; it returns a slice of all successive
917// matches of the expression, as defined by the 'All' description in the
918// package comment.
919// A return value of nil indicates no match.
920func (re *Regexp) FindAll(b []byte, n int) [][]byte {
921 if n < 0 {
922 n = len(b) + 1
923 }
924 result := make([][]byte, 0, startSize)
925 re.allMatches("", b, n, func(match []int) {
926 result = append(result, b[match[0]:match[1]])
927 })
928 if len(result) == 0 {
929 return nil
930 }
931 return result
932}
933
934// FindAllIndex is the 'All' version of FindIndex; it returns a slice of all
935// successive matches of the expression, as defined by the 'All' description
936// in the package comment.
937// A return value of nil indicates no match.
938func (re *Regexp) FindAllIndex(b []byte, n int) [][]int {
939 if n < 0 {
940 n = len(b) + 1
941 }
942 result := make([][]int, 0, startSize)
943 re.allMatches("", b, n, func(match []int) {
944 result = append(result, match[0:2])
945 })
946 if len(result) == 0 {
947 return nil
948 }
949 return result
950}
951
952// FindAllString is the 'All' version of FindString; it returns a slice of all
953// successive matches of the expression, as defined by the 'All' description
954// in the package comment.
955// A return value of nil indicates no match.
956func (re *Regexp) FindAllString(s string, n int) []string {
957 if n < 0 {
958 n = len(s) + 1
959 }
960 result := make([]string, 0, startSize)
961 re.allMatches(s, nil, n, func(match []int) {
962 result = append(result, s[match[0]:match[1]])
963 })
964 if len(result) == 0 {
965 return nil
966 }
967 return result
968}
969
970// FindAllStringIndex is the 'All' version of FindStringIndex; it returns a
971// slice of all successive matches of the expression, as defined by the 'All'
972// description in the package comment.
973// A return value of nil indicates no match.
974func (re *Regexp) FindAllStringIndex(s string, n int) [][]int {
975 if n < 0 {
976 n = len(s) + 1
977 }
978 result := make([][]int, 0, startSize)
979 re.allMatches(s, nil, n, func(match []int) {
980 result = append(result, match[0:2])
981 })
982 if len(result) == 0 {
983 return nil
984 }
985 return result
986}
987
988// FindAllSubmatch is the 'All' version of FindSubmatch; it returns a slice
989// of all successive matches of the expression, as defined by the 'All'
990// description in the package comment.
991// A return value of nil indicates no match.
992func (re *Regexp) FindAllSubmatch(b []byte, n int) [][][]byte {
993 if n < 0 {
994 n = len(b) + 1
995 }
996 result := make([][][]byte, 0, startSize)
997 re.allMatches("", b, n, func(match []int) {
998 slice := make([][]byte, len(match)/2)
999 for j := range slice {
1000 if match[2*j] >= 0 {
1001 slice[j] = b[match[2*j]:match[2*j+1]]
1002 }
1003 }
1004 result = append(result, slice)
1005 })
1006 if len(result) == 0 {
1007 return nil
1008 }
1009 return result
1010}
1011
1012// FindAllSubmatchIndex is the 'All' version of FindSubmatchIndex; it returns
1013// a slice of all successive matches of the expression, as defined by the
1014// 'All' description in the package comment.
1015// A return value of nil indicates no match.
1016func (re *Regexp) FindAllSubmatchIndex(b []byte, n int) [][]int {
1017 if n < 0 {
1018 n = len(b) + 1
1019 }
1020 result := make([][]int, 0, startSize)
1021 re.allMatches("", b, n, func(match []int) {
1022 result = append(result, match)
1023 })
1024 if len(result) == 0 {
1025 return nil
1026 }
1027 return result
1028}
1029
1030// FindAllStringSubmatch is the 'All' version of FindStringSubmatch; it
1031// returns a slice of all successive matches of the expression, as defined by
1032// the 'All' description in the package comment.
1033// A return value of nil indicates no match.
1034func (re *Regexp) FindAllStringSubmatch(s string, n int) [][]string {
1035 if n < 0 {
1036 n = len(s) + 1
1037 }
1038 result := make([][]string, 0, startSize)
1039 re.allMatches(s, nil, n, func(match []int) {
1040 slice := make([]string, len(match)/2)
1041 for j := range slice {
1042 if match[2*j] >= 0 {
1043 slice[j] = s[match[2*j]:match[2*j+1]]
1044 }
1045 }
1046 result = append(result, slice)
1047 })
1048 if len(result) == 0 {
1049 return nil
1050 }
1051 return result
1052}
1053
1054// FindAllStringSubmatchIndex is the 'All' version of
1055// FindStringSubmatchIndex; it returns a slice of all successive matches of
1056// the expression, as defined by the 'All' description in the package
1057// comment.
1058// A return value of nil indicates no match.
1059func (re *Regexp) FindAllStringSubmatchIndex(s string, n int) [][]int {
1060 if n < 0 {
1061 n = len(s) + 1
1062 }
1063 result := make([][]int, 0, startSize)
1064 re.allMatches(s, nil, n, func(match []int) {
1065 result = append(result, match)
1066 })
1067 if len(result) == 0 {
1068 return nil
1069 }
1070 return result
1071}
1072
1073// Split slices s into substrings separated by the expression and returns a slice of
1074// the substrings between those expression matches.
1075//
1076// The slice returned by this method consists of all the substrings of s
1077// not contained in the slice returned by FindAllString. When called on an expression
1078// that contains no metacharacters, it is equivalent to strings.SplitN.
1079//
1080// Example:
1081// s := regexp.MustCompile("a*").Split("abaabaccadaaae", 5)
1082// // s: ["", "b", "b", "c", "cadaaae"]
1083//
1084// The count determines the number of substrings to return:
1085// n > 0: at most n substrings; the last substring will be the unsplit remainder.
1086// n == 0: the result is nil (zero substrings)
1087// n < 0: all substrings
1088func (re *Regexp) Split(s string, n int) []string {
1089
1090 if n == 0 {
1091 return nil
1092 }
1093
1094 if len(re.expr) > 0 && len(s) == 0 {
1095 return []string{""}
1096 }
1097
1098 matches := re.FindAllStringIndex(s, n)
1099 strings := make([]string, 0, len(matches))
1100
1101 beg := 0
1102 end := 0
1103 for _, match := range matches {
1104 if n > 0 && len(strings) >= n-1 {
1105 break
1106 }
1107
1108 end = match[0]
1109 if match[1] != 0 {
1110 strings = append(strings, s[beg:end])
1111 }
1112 beg = match[1]
1113 }
1114
1115 if end != len(s) {
1116 strings = append(strings, s[beg:])
1117 }
1118
1119 return strings
1120}