utf16.go

     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 utf16 implements encoding and decoding of UTF-16 sequences.
     6  package utf16
     7  
     8  // The conditions replacementChar==unicode.ReplacementChar and
     9  // maxRune==unicode.MaxRune are verified in the tests.
    10  // Defining them locally avoids this package depending on package unicode.
    11  
    12  const (
    13  	replacementChar = '\uFFFD'     // Unicode replacement character
    14  	maxRune         = '\U0010FFFF' // Maximum valid Unicode code point.
    15  )
    16  
    17  const (
    18  	// 0xd800-0xdc00 encodes the high 10 bits of a pair.
    19  	// 0xdc00-0xe000 encodes the low 10 bits of a pair.
    20  	// the value is those 20 bits plus 0x10000.
    21  	surr1 = 0xd800
    22  	surr2 = 0xdc00
    23  	surr3 = 0xe000
    24  
    25  	surrSelf = 0x10000
    26  )
    27  
    28  // IsSurrogate reports whether the specified Unicode code point
    29  // can appear in a surrogate pair.
    30  func IsSurrogate(r rune) bool {
    31  	return surr1 <= r && r < surr3
    32  }
    33  
    34  // DecodeRune returns the UTF-16 decoding of a surrogate pair.
    35  // If the pair is not a valid UTF-16 surrogate pair, DecodeRune returns
    36  // the Unicode replacement code point U+FFFD.
    37  func DecodeRune(r1, r2 rune) rune {
    38  	if surr1 <= r1 && r1 < surr2 && surr2 <= r2 && r2 < surr3 {
    39  		return (r1-surr1)<<10 | (r2 - surr2) + surrSelf
    40  	}
    41  	return replacementChar
    42  }
    43  
    44  // EncodeRune returns the UTF-16 surrogate pair r1, r2 for the given rune.
    45  // If the rune is not a valid Unicode code point or does not need encoding,
    46  // EncodeRune returns U+FFFD, U+FFFD.
    47  func EncodeRune(r rune) (r1, r2 rune) {
    48  	if r < surrSelf || r > maxRune {
    49  		return replacementChar, replacementChar
    50  	}
    51  	r -= surrSelf
    52  	return surr1 + (r>>10)&0x3ff, surr2 + r&0x3ff
    53  }
    54  
    55  // RuneLen returns the number of 16-bit words in the UTF-16 encoding of the rune.
    56  // It returns -1 if the rune is not a valid value to encode in UTF-16.
    57  func RuneLen(r rune) int {
    58  	switch {
    59  	case 0 <= r && r < surr1, surr3 <= r && r < surrSelf:
    60  		return 1
    61  	case surrSelf <= r && r <= maxRune:
    62  		return 2
    63  	default:
    64  		return -1
    65  	}
    66  }
    67  
    68  // Encode returns the UTF-16 encoding of the Unicode code point sequence s.
    69  func Encode(s []rune) []uint16 {
    70  	n := len(s)
    71  	for _, v := range s {
    72  		if v >= surrSelf {
    73  			n++
    74  		}
    75  	}
    76  
    77  	a := make([]uint16, n)
    78  	n = 0
    79  	for _, v := range s {
    80  		switch RuneLen(v) {
    81  		case 1: // normal rune
    82  			a[n] = uint16(v)
    83  			n++
    84  		case 2: // needs surrogate sequence
    85  			r1, r2 := EncodeRune(v)
    86  			a[n] = uint16(r1)
    87  			a[n+1] = uint16(r2)
    88  			n += 2
    89  		default:
    90  			a[n] = uint16(replacementChar)
    91  			n++
    92  		}
    93  	}
    94  	return a[:n]
    95  }
    96  
    97  // AppendRune appends the UTF-16 encoding of the Unicode code point r
    98  // to the end of p and returns the extended buffer. If the rune is not
    99  // a valid Unicode code point, it appends the encoding of U+FFFD.
   100  func AppendRune(a []uint16, r rune) []uint16 {
   101  	// This function is inlineable for fast handling of ASCII.
   102  	switch {
   103  	case 0 <= r && r < surr1, surr3 <= r && r < surrSelf:
   104  		// normal rune
   105  		return append(a, uint16(r))
   106  	case surrSelf <= r && r <= maxRune:
   107  		// needs surrogate sequence
   108  		r1, r2 := EncodeRune(r)
   109  		return append(a, uint16(r1), uint16(r2))
   110  	}
   111  	return append(a, replacementChar)
   112  }
   113  
   114  // Decode returns the Unicode code point sequence represented
   115  // by the UTF-16 encoding s.
   116  func Decode(s []uint16) []rune {
   117  	// Preallocate capacity to hold up to 64 runes.
   118  	// Decode inlines, so the allocation can live on the stack.
   119  	buf := make([]rune, 0, 64)
   120  	return decode(s, buf)
   121  }
   122  
   123  // decode appends to buf the Unicode code point sequence represented
   124  // by the UTF-16 encoding s and return the extended buffer.
   125  func decode(s []uint16, buf []rune) []rune {
   126  	for i := 0; i < len(s); i++ {
   127  		var ar rune
   128  		switch r := s[i]; {
   129  		case r < surr1, surr3 <= r:
   130  			// normal rune
   131  			ar = rune(r)
   132  		case surr1 <= r && r < surr2 && i+1 < len(s) &&
   133  			surr2 <= s[i+1] && s[i+1] < surr3:
   134  			// valid surrogate sequence
   135  			ar = DecodeRune(rune(r), rune(s[i+1]))
   136  			i++
   137  		default:
   138  			// invalid surrogate sequence
   139  			ar = replacementChar
   140  		}
   141  		buf = append(buf, ar)
   142  	}
   143  	return buf
   144  }
   145
View as plain text