Source file src/bytes/bytes.go

Documentation: bytes

     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 bytes implements functions for the manipulation of byte slices.
     6  // It is analogous to the facilities of the strings package.
     7  package bytes
     8  
     9  import (
    10  	"internal/bytealg"
    11  	"unicode"
    12  	"unicode/utf8"
    13  )
    14  
    15  // Equal reports whether a and b
    16  // are the same length and contain the same bytes.
    17  // A nil argument is equivalent to an empty slice.
    18  func Equal(a, b []byte) bool {
    19  	// Neither cmd/compile nor gccgo allocates for these string conversions.
    20  	return string(a) == string(b)
    21  }
    22  
    23  // Compare returns an integer comparing two byte slices lexicographically.
    24  // The result will be 0 if a==b, -1 if a < b, and +1 if a > b.
    25  // A nil argument is equivalent to an empty slice.
    26  func Compare(a, b []byte) int {
    27  	return bytealg.Compare(a, b)
    28  }
    29  
    30  // explode splits s into a slice of UTF-8 sequences, one per Unicode code point (still slices of bytes),
    31  // up to a maximum of n byte slices. Invalid UTF-8 sequences are chopped into individual bytes.
    32  func explode(s []byte, n int) [][]byte {
    33  	if n <= 0 {
    34  		n = len(s)
    35  	}
    36  	a := make([][]byte, n)
    37  	var size int
    38  	na := 0
    39  	for len(s) > 0 {
    40  		if na+1 >= n {
    41  			a[na] = s
    42  			na++
    43  			break
    44  		}
    45  		_, size = utf8.DecodeRune(s)
    46  		a[na] = s[0:size:size]
    47  		s = s[size:]
    48  		na++
    49  	}
    50  	return a[0:na]
    51  }
    52  
    53  // Count counts the number of non-overlapping instances of sep in s.
    54  // If sep is an empty slice, Count returns 1 + the number of UTF-8-encoded code points in s.
    55  func Count(s, sep []byte) int {
    56  	// special case
    57  	if len(sep) == 0 {
    58  		return utf8.RuneCount(s) + 1
    59  	}
    60  	if len(sep) == 1 {
    61  		return bytealg.Count(s, sep[0])
    62  	}
    63  	n := 0
    64  	for {
    65  		i := Index(s, sep)
    66  		if i == -1 {
    67  			return n
    68  		}
    69  		n++
    70  		s = s[i+len(sep):]
    71  	}
    72  }
    73  
    74  // Contains reports whether subslice is within b.
    75  func Contains(b, subslice []byte) bool {
    76  	return Index(b, subslice) != -1
    77  }
    78  
    79  // ContainsAny reports whether any of the UTF-8-encoded code points in chars are within b.
    80  func ContainsAny(b []byte, chars string) bool {
    81  	return IndexAny(b, chars) >= 0
    82  }
    83  
    84  // ContainsRune reports whether the rune is contained in the UTF-8-encoded byte slice b.
    85  func ContainsRune(b []byte, r rune) bool {
    86  	return IndexRune(b, r) >= 0
    87  }
    88  
    89  // IndexByte returns the index of the first instance of c in b, or -1 if c is not present in b.
    90  func IndexByte(b []byte, c byte) int {
    91  	return bytealg.IndexByte(b, c)
    92  }
    93  
    94  func indexBytePortable(s []byte, c byte) int {
    95  	for i, b := range s {
    96  		if b == c {
    97  			return i
    98  		}
    99  	}
   100  	return -1
   101  }
   102  
   103  // LastIndex returns the index of the last instance of sep in s, or -1 if sep is not present in s.
   104  func LastIndex(s, sep []byte) int {
   105  	n := len(sep)
   106  	switch {
   107  	case n == 0:
   108  		return len(s)
   109  	case n == 1:
   110  		return LastIndexByte(s, sep[0])
   111  	case n == len(s):
   112  		if Equal(s, sep) {
   113  			return 0
   114  		}
   115  		return -1
   116  	case n > len(s):
   117  		return -1
   118  	}
   119  	// Rabin-Karp search from the end of the string
   120  	hashss, pow := hashStrRev(sep)
   121  	last := len(s) - n
   122  	var h uint32
   123  	for i := len(s) - 1; i >= last; i-- {
   124  		h = h*primeRK + uint32(s[i])
   125  	}
   126  	if h == hashss && Equal(s[last:], sep) {
   127  		return last
   128  	}
   129  	for i := last - 1; i >= 0; i-- {
   130  		h *= primeRK
   131  		h += uint32(s[i])
   132  		h -= pow * uint32(s[i+n])
   133  		if h == hashss && Equal(s[i:i+n], sep) {
   134  			return i
   135  		}
   136  	}
   137  	return -1
   138  }
   139  
   140  // LastIndexByte returns the index of the last instance of c in s, or -1 if c is not present in s.
   141  func LastIndexByte(s []byte, c byte) int {
   142  	for i := len(s) - 1; i >= 0; i-- {
   143  		if s[i] == c {
   144  			return i
   145  		}
   146  	}
   147  	return -1
   148  }
   149  
   150  // IndexRune interprets s as a sequence of UTF-8-encoded code points.
   151  // It returns the byte index of the first occurrence in s of the given rune.
   152  // It returns -1 if rune is not present in s.
   153  // If r is utf8.RuneError, it returns the first instance of any
   154  // invalid UTF-8 byte sequence.
   155  func IndexRune(s []byte, r rune) int {
   156  	switch {
   157  	case 0 <= r && r < utf8.RuneSelf:
   158  		return IndexByte(s, byte(r))
   159  	case r == utf8.RuneError:
   160  		for i := 0; i < len(s); {
   161  			r1, n := utf8.DecodeRune(s[i:])
   162  			if r1 == utf8.RuneError {
   163  				return i
   164  			}
   165  			i += n
   166  		}
   167  		return -1
   168  	case !utf8.ValidRune(r):
   169  		return -1
   170  	default:
   171  		var b [utf8.UTFMax]byte
   172  		n := utf8.EncodeRune(b[:], r)
   173  		return Index(s, b[:n])
   174  	}
   175  }
   176  
   177  // IndexAny interprets s as a sequence of UTF-8-encoded Unicode code points.
   178  // It returns the byte index of the first occurrence in s of any of the Unicode
   179  // code points in chars. It returns -1 if chars is empty or if there is no code
   180  // point in common.
   181  func IndexAny(s []byte, chars string) int {
   182  	if chars == "" {
   183  		// Avoid scanning all of s.
   184  		return -1
   185  	}
   186  	if len(s) > 8 {
   187  		if as, isASCII := makeASCIISet(chars); isASCII {
   188  			for i, c := range s {
   189  				if as.contains(c) {
   190  					return i
   191  				}
   192  			}
   193  			return -1
   194  		}
   195  	}
   196  	var width int
   197  	for i := 0; i < len(s); i += width {
   198  		r := rune(s[i])
   199  		if r < utf8.RuneSelf {
   200  			width = 1
   201  		} else {
   202  			r, width = utf8.DecodeRune(s[i:])
   203  		}
   204  		for _, ch := range chars {
   205  			if r == ch {
   206  				return i
   207  			}
   208  		}
   209  	}
   210  	return -1
   211  }
   212  
   213  // LastIndexAny interprets s as a sequence of UTF-8-encoded Unicode code
   214  // points. It returns the byte index of the last occurrence in s of any of
   215  // the Unicode code points in chars. It returns -1 if chars is empty or if
   216  // there is no code point in common.
   217  func LastIndexAny(s []byte, chars string) int {
   218  	if chars == "" {
   219  		// Avoid scanning all of s.
   220  		return -1
   221  	}
   222  	if len(s) > 8 {
   223  		if as, isASCII := makeASCIISet(chars); isASCII {
   224  			for i := len(s) - 1; i >= 0; i-- {
   225  				if as.contains(s[i]) {
   226  					return i
   227  				}
   228  			}
   229  			return -1
   230  		}
   231  	}
   232  	for i := len(s); i > 0; {
   233  		r, size := utf8.DecodeLastRune(s[:i])
   234  		i -= size
   235  		for _, c := range chars {
   236  			if r == c {
   237  				return i
   238  			}
   239  		}
   240  	}
   241  	return -1
   242  }
   243  
   244  // Generic split: splits after each instance of sep,
   245  // including sepSave bytes of sep in the subslices.
   246  func genSplit(s, sep []byte, sepSave, n int) [][]byte {
   247  	if n == 0 {
   248  		return nil
   249  	}
   250  	if len(sep) == 0 {
   251  		return explode(s, n)
   252  	}
   253  	if n < 0 {
   254  		n = Count(s, sep) + 1
   255  	}
   256  
   257  	a := make([][]byte, n)
   258  	n--
   259  	i := 0
   260  	for i < n {
   261  		m := Index(s, sep)
   262  		if m < 0 {
   263  			break
   264  		}
   265  		a[i] = s[: m+sepSave : m+sepSave]
   266  		s = s[m+len(sep):]
   267  		i++
   268  	}
   269  	a[i] = s
   270  	return a[:i+1]
   271  }
   272  
   273  // SplitN slices s into subslices separated by sep and returns a slice of
   274  // the subslices between those separators.
   275  // If sep is empty, SplitN splits after each UTF-8 sequence.
   276  // The count determines the number of subslices to return:
   277  //   n > 0: at most n subslices; the last subslice will be the unsplit remainder.
   278  //   n == 0: the result is nil (zero subslices)
   279  //   n < 0: all subslices
   280  func SplitN(s, sep []byte, n int) [][]byte { return genSplit(s, sep, 0, n) }
   281  
   282  // SplitAfterN slices s into subslices after each instance of sep and
   283  // returns a slice of those subslices.
   284  // If sep is empty, SplitAfterN splits after each UTF-8 sequence.
   285  // The count determines the number of subslices to return:
   286  //   n > 0: at most n subslices; the last subslice will be the unsplit remainder.
   287  //   n == 0: the result is nil (zero subslices)
   288  //   n < 0: all subslices
   289  func SplitAfterN(s, sep []byte, n int) [][]byte {
   290  	return genSplit(s, sep, len(sep), n)
   291  }
   292  
   293  // Split slices s into all subslices separated by sep and returns a slice of
   294  // the subslices between those separators.
   295  // If sep is empty, Split splits after each UTF-8 sequence.
   296  // It is equivalent to SplitN with a count of -1.
   297  func Split(s, sep []byte) [][]byte { return genSplit(s, sep, 0, -1) }
   298  
   299  // SplitAfter slices s into all subslices after each instance of sep and
   300  // returns a slice of those subslices.
   301  // If sep is empty, SplitAfter splits after each UTF-8 sequence.
   302  // It is equivalent to SplitAfterN with a count of -1.
   303  func SplitAfter(s, sep []byte) [][]byte {
   304  	return genSplit(s, sep, len(sep), -1)
   305  }
   306  
   307  var asciiSpace = [256]uint8{'\t': 1, '\n': 1, '\v': 1, '\f': 1, '\r': 1, ' ': 1}
   308  
   309  // Fields interprets s as a sequence of UTF-8-encoded code points.
   310  // It splits the slice s around each instance of one or more consecutive white space
   311  // characters, as defined by unicode.IsSpace, returning a slice of subslices of s or an
   312  // empty slice if s contains only white space.
   313  func Fields(s []byte) [][]byte {
   314  	// First count the fields.
   315  	// This is an exact count if s is ASCII, otherwise it is an approximation.
   316  	n := 0
   317  	wasSpace := 1
   318  	// setBits is used to track which bits are set in the bytes of s.
   319  	setBits := uint8(0)
   320  	for i := 0; i < len(s); i++ {
   321  		r := s[i]
   322  		setBits |= r
   323  		isSpace := int(asciiSpace[r])
   324  		n += wasSpace & ^isSpace
   325  		wasSpace = isSpace
   326  	}
   327  
   328  	if setBits >= utf8.RuneSelf {
   329  		// Some runes in the input slice are not ASCII.
   330  		return FieldsFunc(s, unicode.IsSpace)
   331  	}
   332  
   333  	// ASCII fast path
   334  	a := make([][]byte, n)
   335  	na := 0
   336  	fieldStart := 0
   337  	i := 0
   338  	// Skip spaces in the front of the input.
   339  	for i < len(s) && asciiSpace[s[i]] != 0 {
   340  		i++
   341  	}
   342  	fieldStart = i
   343  	for i < len(s) {
   344  		if asciiSpace[s[i]] == 0 {
   345  			i++
   346  			continue
   347  		}
   348  		a[na] = s[fieldStart:i:i]
   349  		na++
   350  		i++
   351  		// Skip spaces in between fields.
   352  		for i < len(s) && asciiSpace[s[i]] != 0 {
   353  			i++
   354  		}
   355  		fieldStart = i
   356  	}
   357  	if fieldStart < len(s) { // Last field might end at EOF.
   358  		a[na] = s[fieldStart:len(s):len(s)]
   359  	}
   360  	return a
   361  }
   362  
   363  // FieldsFunc interprets s as a sequence of UTF-8-encoded code points.
   364  // It splits the slice s at each run of code points c satisfying f(c) and
   365  // returns a slice of subslices of s. If all code points in s satisfy f(c), or
   366  // len(s) == 0, an empty slice is returned.
   367  // FieldsFunc makes no guarantees about the order in which it calls f(c).
   368  // If f does not return consistent results for a given c, FieldsFunc may crash.
   369  func FieldsFunc(s []byte, f func(rune) bool) [][]byte {
   370  	// A span is used to record a slice of s of the form s[start:end].
   371  	// The start index is inclusive and the end index is exclusive.
   372  	type span struct {
   373  		start int
   374  		end   int
   375  	}
   376  	spans := make([]span, 0, 32)
   377  
   378  	// Find the field start and end indices.
   379  	wasField := false
   380  	fromIndex := 0
   381  	for i := 0; i < len(s); {
   382  		size := 1
   383  		r := rune(s[i])
   384  		if r >= utf8.RuneSelf {
   385  			r, size = utf8.DecodeRune(s[i:])
   386  		}
   387  		if f(r) {
   388  			if wasField {
   389  				spans = append(spans, span{start: fromIndex, end: i})
   390  				wasField = false
   391  			}
   392  		} else {
   393  			if !wasField {
   394  				fromIndex = i
   395  				wasField = true
   396  			}
   397  		}
   398  		i += size
   399  	}
   400  
   401  	// Last field might end at EOF.
   402  	if wasField {
   403  		spans = append(spans, span{fromIndex, len(s)})
   404  	}
   405  
   406  	// Create subslices from recorded field indices.
   407  	a := make([][]byte, len(spans))
   408  	for i, span := range spans {
   409  		a[i] = s[span.start:span.end:span.end]
   410  	}
   411  
   412  	return a
   413  }
   414  
   415  // Join concatenates the elements of s to create a new byte slice. The separator
   416  // sep is placed between elements in the resulting slice.
   417  func Join(s [][]byte, sep []byte) []byte {
   418  	if len(s) == 0 {
   419  		return []byte{}
   420  	}
   421  	if len(s) == 1 {
   422  		// Just return a copy.
   423  		return append([]byte(nil), s[0]...)
   424  	}
   425  	n := len(sep) * (len(s) - 1)
   426  	for _, v := range s {
   427  		n += len(v)
   428  	}
   429  
   430  	b := make([]byte, n)
   431  	bp := copy(b, s[0])
   432  	for _, v := range s[1:] {
   433  		bp += copy(b[bp:], sep)
   434  		bp += copy(b[bp:], v)
   435  	}
   436  	return b
   437  }
   438  
   439  // HasPrefix tests whether the byte slice s begins with prefix.
   440  func HasPrefix(s, prefix []byte) bool {
   441  	return len(s) >= len(prefix) && Equal(s[0:len(prefix)], prefix)
   442  }
   443  
   444  // HasSuffix tests whether the byte slice s ends with suffix.
   445  func HasSuffix(s, suffix []byte) bool {
   446  	return len(s) >= len(suffix) && Equal(s[len(s)-len(suffix):], suffix)
   447  }
   448  
   449  // Map returns a copy of the byte slice s with all its characters modified
   450  // according to the mapping function. If mapping returns a negative value, the character is
   451  // dropped from the byte slice with no replacement. The characters in s and the
   452  // output are interpreted as UTF-8-encoded code points.
   453  func Map(mapping func(r rune) rune, s []byte) []byte {
   454  	// In the worst case, the slice can grow when mapped, making
   455  	// things unpleasant. But it's so rare we barge in assuming it's
   456  	// fine. It could also shrink but that falls out naturally.
   457  	maxbytes := len(s) // length of b
   458  	nbytes := 0        // number of bytes encoded in b
   459  	b := make([]byte, maxbytes)
   460  	for i := 0; i < len(s); {
   461  		wid := 1
   462  		r := rune(s[i])
   463  		if r >= utf8.RuneSelf {
   464  			r, wid = utf8.DecodeRune(s[i:])
   465  		}
   466  		r = mapping(r)
   467  		if r >= 0 {
   468  			rl := utf8.RuneLen(r)
   469  			if rl < 0 {
   470  				rl = len(string(utf8.RuneError))
   471  			}
   472  			if nbytes+rl > maxbytes {
   473  				// Grow the buffer.
   474  				maxbytes = maxbytes*2 + utf8.UTFMax
   475  				nb := make([]byte, maxbytes)
   476  				copy(nb, b[0:nbytes])
   477  				b = nb
   478  			}
   479  			nbytes += utf8.EncodeRune(b[nbytes:maxbytes], r)
   480  		}
   481  		i += wid
   482  	}
   483  	return b[0:nbytes]
   484  }
   485  
   486  // Repeat returns a new byte slice consisting of count copies of b.
   487  //
   488  // It panics if count is negative or if
   489  // the result of (len(b) * count) overflows.
   490  func Repeat(b []byte, count int) []byte {
   491  	if count == 0 {
   492  		return []byte{}
   493  	}
   494  	// Since we cannot return an error on overflow,
   495  	// we should panic if the repeat will generate
   496  	// an overflow.
   497  	// See Issue golang.org/issue/16237.
   498  	if count < 0 {
   499  		panic("bytes: negative Repeat count")
   500  	} else if len(b)*count/count != len(b) {
   501  		panic("bytes: Repeat count causes overflow")
   502  	}
   503  
   504  	nb := make([]byte, len(b)*count)
   505  	bp := copy(nb, b)
   506  	for bp < len(nb) {
   507  		copy(nb[bp:], nb[:bp])
   508  		bp *= 2
   509  	}
   510  	return nb
   511  }
   512  
   513  // ToUpper returns a copy of the byte slice s with all Unicode letters mapped to
   514  // their upper case.
   515  func ToUpper(s []byte) []byte {
   516  	isASCII, hasLower := true, false
   517  	for i := 0; i < len(s); i++ {
   518  		c := s[i]
   519  		if c >= utf8.RuneSelf {
   520  			isASCII = false
   521  			break
   522  		}
   523  		hasLower = hasLower || ('a' <= c && c <= 'z')
   524  	}
   525  
   526  	if isASCII { // optimize for ASCII-only byte slices.
   527  		if !hasLower {
   528  			// Just return a copy.
   529  			return append([]byte(""), s...)
   530  		}
   531  		b := make([]byte, len(s))
   532  		for i := 0; i < len(s); i++ {
   533  			c := s[i]
   534  			if 'a' <= c && c <= 'z' {
   535  				c -= 'a' - 'A'
   536  			}
   537  			b[i] = c
   538  		}
   539  		return b
   540  	}
   541  	return Map(unicode.ToUpper, s)
   542  }
   543  
   544  // ToLower returns a copy of the byte slice s with all Unicode letters mapped to
   545  // their lower case.
   546  func ToLower(s []byte) []byte {
   547  	isASCII, hasUpper := true, false
   548  	for i := 0; i < len(s); i++ {
   549  		c := s[i]
   550  		if c >= utf8.RuneSelf {
   551  			isASCII = false
   552  			break
   553  		}
   554  		hasUpper = hasUpper || ('A' <= c && c <= 'Z')
   555  	}
   556  
   557  	if isASCII { // optimize for ASCII-only byte slices.
   558  		if !hasUpper {
   559  			return append([]byte(""), s...)
   560  		}
   561  		b := make([]byte, len(s))
   562  		for i := 0; i < len(s); i++ {
   563  			c := s[i]
   564  			if 'A' <= c && c <= 'Z' {
   565  				c += 'a' - 'A'
   566  			}
   567  			b[i] = c
   568  		}
   569  		return b
   570  	}
   571  	return Map(unicode.ToLower, s)
   572  }
   573  
   574  // ToTitle treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their title case.
   575  func ToTitle(s []byte) []byte { return Map(unicode.ToTitle, s) }
   576  
   577  // ToUpperSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their
   578  // upper case, giving priority to the special casing rules.
   579  func ToUpperSpecial(c unicode.SpecialCase, s []byte) []byte {
   580  	return Map(c.ToUpper, s)
   581  }
   582  
   583  // ToLowerSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their
   584  // lower case, giving priority to the special casing rules.
   585  func ToLowerSpecial(c unicode.SpecialCase, s []byte) []byte {
   586  	return Map(c.ToLower, s)
   587  }
   588  
   589  // ToTitleSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their
   590  // title case, giving priority to the special casing rules.
   591  func ToTitleSpecial(c unicode.SpecialCase, s []byte) []byte {
   592  	return Map(c.ToTitle, s)
   593  }
   594  
   595  // ToValidUTF8 treats s as UTF-8-encoded bytes and returns a copy with each run of bytes
   596  // representing invalid UTF-8 replaced with the bytes in replacement, which may be empty.
   597  func ToValidUTF8(s, replacement []byte) []byte {
   598  	b := make([]byte, 0, len(s)+len(replacement))
   599  	invalid := false // previous byte was from an invalid UTF-8 sequence
   600  	for i := 0; i < len(s); {
   601  		c := s[i]
   602  		if c < utf8.RuneSelf {
   603  			i++
   604  			invalid = false
   605  			b = append(b, byte(c))
   606  			continue
   607  		}
   608  		_, wid := utf8.DecodeRune(s[i:])
   609  		if wid == 1 {
   610  			i++
   611  			if !invalid {
   612  				invalid = true
   613  				b = append(b, replacement...)
   614  			}
   615  			continue
   616  		}
   617  		invalid = false
   618  		b = append(b, s[i:i+wid]...)
   619  		i += wid
   620  	}
   621  	return b
   622  }
   623  
   624  // isSeparator reports whether the rune could mark a word boundary.
   625  // TODO: update when package unicode captures more of the properties.
   626  func isSeparator(r rune) bool {
   627  	// ASCII alphanumerics and underscore are not separators
   628  	if r <= 0x7F {
   629  		switch {
   630  		case '0' <= r && r <= '9':
   631  			return false
   632  		case 'a' <= r && r <= 'z':
   633  			return false
   634  		case 'A' <= r && r <= 'Z':
   635  			return false
   636  		case r == '_':
   637  			return false
   638  		}
   639  		return true
   640  	}
   641  	// Letters and digits are not separators
   642  	if unicode.IsLetter(r) || unicode.IsDigit(r) {
   643  		return false
   644  	}
   645  	// Otherwise, all we can do for now is treat spaces as separators.
   646  	return unicode.IsSpace(r)
   647  }
   648  
   649  // Title treats s as UTF-8-encoded bytes and returns a copy with all Unicode letters that begin
   650  // words mapped to their title case.
   651  //
   652  // BUG(rsc): The rule Title uses for word boundaries does not handle Unicode punctuation properly.
   653  func Title(s []byte) []byte {
   654  	// Use a closure here to remember state.
   655  	// Hackish but effective. Depends on Map scanning in order and calling
   656  	// the closure once per rune.
   657  	prev := ' '
   658  	return Map(
   659  		func(r rune) rune {
   660  			if isSeparator(prev) {
   661  				prev = r
   662  				return unicode.ToTitle(r)
   663  			}
   664  			prev = r
   665  			return r
   666  		},
   667  		s)
   668  }
   669  
   670  // TrimLeftFunc treats s as UTF-8-encoded bytes and returns a subslice of s by slicing off
   671  // all leading UTF-8-encoded code points c that satisfy f(c).
   672  func TrimLeftFunc(s []byte, f func(r rune) bool) []byte {
   673  	i := indexFunc(s, f, false)
   674  	if i == -1 {
   675  		return nil
   676  	}
   677  	return s[i:]
   678  }
   679  
   680  // TrimRightFunc returns a subslice of s by slicing off all trailing
   681  // UTF-8-encoded code points c that satisfy f(c).
   682  func TrimRightFunc(s []byte, f func(r rune) bool) []byte {
   683  	i := lastIndexFunc(s, f, false)
   684  	if i >= 0 && s[i] >= utf8.RuneSelf {
   685  		_, wid := utf8.DecodeRune(s[i:])
   686  		i += wid
   687  	} else {
   688  		i++
   689  	}
   690  	return s[0:i]
   691  }
   692  
   693  // TrimFunc returns a subslice of s by slicing off all leading and trailing
   694  // UTF-8-encoded code points c that satisfy f(c).
   695  func TrimFunc(s []byte, f func(r rune) bool) []byte {
   696  	return TrimRightFunc(TrimLeftFunc(s, f), f)
   697  }
   698  
   699  // TrimPrefix returns s without the provided leading prefix string.
   700  // If s doesn't start with prefix, s is returned unchanged.
   701  func TrimPrefix(s, prefix []byte) []byte {
   702  	if HasPrefix(s, prefix) {
   703  		return s[len(prefix):]
   704  	}
   705  	return s
   706  }
   707  
   708  // TrimSuffix returns s without the provided trailing suffix string.
   709  // If s doesn't end with suffix, s is returned unchanged.
   710  func TrimSuffix(s, suffix []byte) []byte {
   711  	if HasSuffix(s, suffix) {
   712  		return s[:len(s)-len(suffix)]
   713  	}
   714  	return s
   715  }
   716  
   717  // IndexFunc interprets s as a sequence of UTF-8-encoded code points.
   718  // It returns the byte index in s of the first Unicode
   719  // code point satisfying f(c), or -1 if none do.
   720  func IndexFunc(s []byte, f func(r rune) bool) int {
   721  	return indexFunc(s, f, true)
   722  }
   723  
   724  // LastIndexFunc interprets s as a sequence of UTF-8-encoded code points.
   725  // It returns the byte index in s of the last Unicode
   726  // code point satisfying f(c), or -1 if none do.
   727  func LastIndexFunc(s []byte, f func(r rune) bool) int {
   728  	return lastIndexFunc(s, f, true)
   729  }
   730  
   731  // indexFunc is the same as IndexFunc except that if
   732  // truth==false, the sense of the predicate function is
   733  // inverted.
   734  func indexFunc(s []byte, f func(r rune) bool, truth bool) int {
   735  	start := 0
   736  	for start < len(s) {
   737  		wid := 1
   738  		r := rune(s[start])
   739  		if r >= utf8.RuneSelf {
   740  			r, wid = utf8.DecodeRune(s[start:])
   741  		}
   742  		if f(r) == truth {
   743  			return start
   744  		}
   745  		start += wid
   746  	}
   747  	return -1
   748  }
   749  
   750  // lastIndexFunc is the same as LastIndexFunc except that if
   751  // truth==false, the sense of the predicate function is
   752  // inverted.
   753  func lastIndexFunc(s []byte, f func(r rune) bool, truth bool) int {
   754  	for i := len(s); i > 0; {
   755  		r, size := rune(s[i-1]), 1
   756  		if r >= utf8.RuneSelf {
   757  			r, size = utf8.DecodeLastRune(s[0:i])
   758  		}
   759  		i -= size
   760  		if f(r) == truth {
   761  			return i
   762  		}
   763  	}
   764  	return -1
   765  }
   766  
   767  // asciiSet is a 32-byte value, where each bit represents the presence of a
   768  // given ASCII character in the set. The 128-bits of the lower 16 bytes,
   769  // starting with the least-significant bit of the lowest word to the
   770  // most-significant bit of the highest word, map to the full range of all
   771  // 128 ASCII characters. The 128-bits of the upper 16 bytes will be zeroed,
   772  // ensuring that any non-ASCII character will be reported as not in the set.
   773  type asciiSet [8]uint32
   774  
   775  // makeASCIISet creates a set of ASCII characters and reports whether all
   776  // characters in chars are ASCII.
   777  func makeASCIISet(chars string) (as asciiSet, ok bool) {
   778  	for i := 0; i < len(chars); i++ {
   779  		c := chars[i]
   780  		if c >= utf8.RuneSelf {
   781  			return as, false
   782  		}
   783  		as[c>>5] |= 1 << uint(c&31)
   784  	}
   785  	return as, true
   786  }
   787  
   788  // contains reports whether c is inside the set.
   789  func (as *asciiSet) contains(c byte) bool {
   790  	return (as[c>>5] & (1 << uint(c&31))) != 0
   791  }
   792  
   793  func makeCutsetFunc(cutset string) func(r rune) bool {
   794  	if len(cutset) == 1 && cutset[0] < utf8.RuneSelf {
   795  		return func(r rune) bool {
   796  			return r == rune(cutset[0])
   797  		}
   798  	}
   799  	if as, isASCII := makeASCIISet(cutset); isASCII {
   800  		return func(r rune) bool {
   801  			return r < utf8.RuneSelf && as.contains(byte(r))
   802  		}
   803  	}
   804  	return func(r rune) bool {
   805  		for _, c := range cutset {
   806  			if c == r {
   807  				return true
   808  			}
   809  		}
   810  		return false
   811  	}
   812  }
   813  
   814  // Trim returns a subslice of s by slicing off all leading and
   815  // trailing UTF-8-encoded code points contained in cutset.
   816  func Trim(s []byte, cutset string) []byte {
   817  	return TrimFunc(s, makeCutsetFunc(cutset))
   818  }
   819  
   820  // TrimLeft returns a subslice of s by slicing off all leading
   821  // UTF-8-encoded code points contained in cutset.
   822  func TrimLeft(s []byte, cutset string) []byte {
   823  	return TrimLeftFunc(s, makeCutsetFunc(cutset))
   824  }
   825  
   826  // TrimRight returns a subslice of s by slicing off all trailing
   827  // UTF-8-encoded code points that are contained in cutset.
   828  func TrimRight(s []byte, cutset string) []byte {
   829  	return TrimRightFunc(s, makeCutsetFunc(cutset))
   830  }
   831  
   832  // TrimSpace returns a subslice of s by slicing off all leading and
   833  // trailing white space, as defined by Unicode.
   834  func TrimSpace(s []byte) []byte {
   835  	// Fast path for ASCII: look for the first ASCII non-space byte
   836  	start := 0
   837  	for ; start < len(s); start++ {
   838  		c := s[start]
   839  		if c >= utf8.RuneSelf {
   840  			// If we run into a non-ASCII byte, fall back to the
   841  			// slower unicode-aware method on the remaining bytes
   842  			return TrimFunc(s[start:], unicode.IsSpace)
   843  		}
   844  		if asciiSpace[c] == 0 {
   845  			break
   846  		}
   847  	}
   848  
   849  	// Now look for the first ASCII non-space byte from the end
   850  	stop := len(s)
   851  	for ; stop > start; stop-- {
   852  		c := s[stop-1]
   853  		if c >= utf8.RuneSelf {
   854  			return TrimFunc(s[start:stop], unicode.IsSpace)
   855  		}
   856  		if asciiSpace[c] == 0 {
   857  			break
   858  		}
   859  	}
   860  
   861  	// At this point s[start:stop] starts and ends with an ASCII
   862  	// non-space bytes, so we're done. Non-ASCII cases have already
   863  	// been handled above.
   864  	if start == stop {
   865  		// Special case to preserve previous TrimLeftFunc behavior,
   866  		// returning nil instead of empty slice if all spaces.
   867  		return nil
   868  	}
   869  	return s[start:stop]
   870  }
   871  
   872  // Runes interprets s as a sequence of UTF-8-encoded code points.
   873  // It returns a slice of runes (Unicode code points) equivalent to s.
   874  func Runes(s []byte) []rune {
   875  	t := make([]rune, utf8.RuneCount(s))
   876  	i := 0
   877  	for len(s) > 0 {
   878  		r, l := utf8.DecodeRune(s)
   879  		t[i] = r
   880  		i++
   881  		s = s[l:]
   882  	}
   883  	return t
   884  }
   885  
   886  // Replace returns a copy of the slice s with the first n
   887  // non-overlapping instances of old replaced by new.
   888  // If old is empty, it matches at the beginning of the slice
   889  // and after each UTF-8 sequence, yielding up to k+1 replacements
   890  // for a k-rune slice.
   891  // If n < 0, there is no limit on the number of replacements.
   892  func Replace(s, old, new []byte, n int) []byte {
   893  	m := 0
   894  	if n != 0 {
   895  		// Compute number of replacements.
   896  		m = Count(s, old)
   897  	}
   898  	if m == 0 {
   899  		// Just return a copy.
   900  		return append([]byte(nil), s...)
   901  	}
   902  	if n < 0 || m < n {
   903  		n = m
   904  	}
   905  
   906  	// Apply replacements to buffer.
   907  	t := make([]byte, len(s)+n*(len(new)-len(old)))
   908  	w := 0
   909  	start := 0
   910  	for i := 0; i < n; i++ {
   911  		j := start
   912  		if len(old) == 0 {
   913  			if i > 0 {
   914  				_, wid := utf8.DecodeRune(s[start:])
   915  				j += wid
   916  			}
   917  		} else {
   918  			j += Index(s[start:], old)
   919  		}
   920  		w += copy(t[w:], s[start:j])
   921  		w += copy(t[w:], new)
   922  		start = j + len(old)
   923  	}
   924  	w += copy(t[w:], s[start:])
   925  	return t[0:w]
   926  }
   927  
   928  // ReplaceAll returns a copy of the slice s with all
   929  // non-overlapping instances of old replaced by new.
   930  // If old is empty, it matches at the beginning of the slice
   931  // and after each UTF-8 sequence, yielding up to k+1 replacements
   932  // for a k-rune slice.
   933  func ReplaceAll(s, old, new []byte) []byte {
   934  	return Replace(s, old, new, -1)
   935  }
   936  
   937  // EqualFold reports whether s and t, interpreted as UTF-8 strings,
   938  // are equal under Unicode case-folding.
   939  func EqualFold(s, t []byte) bool {
   940  	for len(s) != 0 && len(t) != 0 {
   941  		// Extract first rune from each.
   942  		var sr, tr rune
   943  		if s[0] < utf8.RuneSelf {
   944  			sr, s = rune(s[0]), s[1:]
   945  		} else {
   946  			r, size := utf8.DecodeRune(s)
   947  			sr, s = r, s[size:]
   948  		}
   949  		if t[0] < utf8.RuneSelf {
   950  			tr, t = rune(t[0]), t[1:]
   951  		} else {
   952  			r, size := utf8.DecodeRune(t)
   953  			tr, t = r, t[size:]
   954  		}
   955  
   956  		// If they match, keep going; if not, return false.
   957  
   958  		// Easy case.
   959  		if tr == sr {
   960  			continue
   961  		}
   962  
   963  		// Make sr < tr to simplify what follows.
   964  		if tr < sr {
   965  			tr, sr = sr, tr
   966  		}
   967  		// Fast check for ASCII.
   968  		if tr < utf8.RuneSelf {
   969  			// ASCII only, sr/tr must be upper/lower case
   970  			if 'A' <= sr && sr <= 'Z' && tr == sr+'a'-'A' {
   971  				continue
   972  			}
   973  			return false
   974  		}
   975  
   976  		// General case. SimpleFold(x) returns the next equivalent rune > x
   977  		// or wraps around to smaller values.
   978  		r := unicode.SimpleFold(sr)
   979  		for r != sr && r < tr {
   980  			r = unicode.SimpleFold(r)
   981  		}
   982  		if r == tr {
   983  			continue
   984  		}
   985  		return false
   986  	}
   987  
   988  	// One string is empty. Are both?
   989  	return len(s) == len(t)
   990  }
   991  
   992  // Index returns the index of the first instance of sep in s, or -1 if sep is not present in s.
   993  func Index(s, sep []byte) int {
   994  	n := len(sep)
   995  	switch {
   996  	case n == 0:
   997  		return 0
   998  	case n == 1:
   999  		return IndexByte(s, sep[0])
  1000  	case n == len(s):
  1001  		if Equal(sep, s) {
  1002  			return 0
  1003  		}
  1004  		return -1
  1005  	case n > len(s):
  1006  		return -1
  1007  	case n <= bytealg.MaxLen:
  1008  		// Use brute force when s and sep both are small
  1009  		if len(s) <= bytealg.MaxBruteForce {
  1010  			return bytealg.Index(s, sep)
  1011  		}
  1012  		c0 := sep[0]
  1013  		c1 := sep[1]
  1014  		i := 0
  1015  		t := len(s) - n + 1
  1016  		fails := 0
  1017  		for i < t {
  1018  			if s[i] != c0 {
  1019  				// IndexByte is faster than bytealg.Index, so use it as long as
  1020  				// we're not getting lots of false positives.
  1021  				o := IndexByte(s[i:t], c0)
  1022  				if o < 0 {
  1023  					return -1
  1024  				}
  1025  				i += o
  1026  			}
  1027  			if s[i+1] == c1 && Equal(s[i:i+n], sep) {
  1028  				return i
  1029  			}
  1030  			fails++
  1031  			i++
  1032  			// Switch to bytealg.Index when IndexByte produces too many false positives.
  1033  			if fails > bytealg.Cutover(i) {
  1034  				r := bytealg.Index(s[i:], sep)
  1035  				if r >= 0 {
  1036  					return r + i
  1037  				}
  1038  				return -1
  1039  			}
  1040  		}
  1041  		return -1
  1042  	}
  1043  	c0 := sep[0]
  1044  	c1 := sep[1]
  1045  	i := 0
  1046  	fails := 0
  1047  	t := len(s) - n + 1
  1048  	for i < t {
  1049  		if s[i] != c0 {
  1050  			o := IndexByte(s[i:t], c0)
  1051  			if o < 0 {
  1052  				break
  1053  			}
  1054  			i += o
  1055  		}
  1056  		if s[i+1] == c1 && Equal(s[i:i+n], sep) {
  1057  			return i
  1058  		}
  1059  		i++
  1060  		fails++
  1061  		if fails >= 4+i>>4 && i < t {
  1062  			// Give up on IndexByte, it isn't skipping ahead
  1063  			// far enough to be better than Rabin-Karp.
  1064  			// Experiments (using IndexPeriodic) suggest
  1065  			// the cutover is about 16 byte skips.
  1066  			// TODO: if large prefixes of sep are matching
  1067  			// we should cutover at even larger average skips,
  1068  			// because Equal becomes that much more expensive.
  1069  			// This code does not take that effect into account.
  1070  			j := indexRabinKarp(s[i:], sep)
  1071  			if j < 0 {
  1072  				return -1
  1073  			}
  1074  			return i + j
  1075  		}
  1076  	}
  1077  	return -1
  1078  }
  1079  
  1080  func indexRabinKarp(s, sep []byte) int {
  1081  	// Rabin-Karp search
  1082  	hashsep, pow := hashStr(sep)
  1083  	n := len(sep)
  1084  	var h uint32
  1085  	for i := 0; i < n; i++ {
  1086  		h = h*primeRK + uint32(s[i])
  1087  	}
  1088  	if h == hashsep && Equal(s[:n], sep) {
  1089  		return 0
  1090  	}
  1091  	for i := n; i < len(s); {
  1092  		h *= primeRK
  1093  		h += uint32(s[i])
  1094  		h -= pow * uint32(s[i-n])
  1095  		i++
  1096  		if h == hashsep && Equal(s[i-n:i], sep) {
  1097  			return i - n
  1098  		}
  1099  	}
  1100  	return -1
  1101  }
  1102  
  1103  // primeRK is the prime base used in Rabin-Karp algorithm.
  1104  const primeRK = 16777619
  1105  
  1106  // hashStr returns the hash and the appropriate multiplicative
  1107  // factor for use in Rabin-Karp algorithm.
  1108  func hashStr(sep []byte) (uint32, uint32) {
  1109  	hash := uint32(0)
  1110  	for i := 0; i < len(sep); i++ {
  1111  		hash = hash*primeRK + uint32(sep[i])
  1112  	}
  1113  	var pow, sq uint32 = 1, primeRK
  1114  	for i := len(sep); i > 0; i >>= 1 {
  1115  		if i&1 != 0 {
  1116  			pow *= sq
  1117  		}
  1118  		sq *= sq
  1119  	}
  1120  	return hash, pow
  1121  }
  1122  
  1123  // hashStrRev returns the hash of the reverse of sep and the
  1124  // appropriate multiplicative factor for use in Rabin-Karp algorithm.
  1125  func hashStrRev(sep []byte) (uint32, uint32) {
  1126  	hash := uint32(0)
  1127  	for i := len(sep) - 1; i >= 0; i-- {
  1128  		hash = hash*primeRK + uint32(sep[i])
  1129  	}
  1130  	var pow, sq uint32 = 1, primeRK
  1131  	for i := len(sep); i > 0; i >>= 1 {
  1132  		if i&1 != 0 {
  1133  			pow *= sq
  1134  		}
  1135  		sq *= sq
  1136  	}
  1137  	return hash, pow
  1138  }
  1139  

View as plain text