...
Run Format

Source file src/runtime/utf8.go

  // Copyright 2016 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 runtime
  
  // Numbers fundamental to the encoding.
  const (
  	runeError = '\uFFFD'     // the "error" Rune or "Unicode replacement character"
  	runeSelf  = 0x80         // characters below Runeself are represented as themselves in a single byte.
  	maxRune   = '\U0010FFFF' // Maximum valid Unicode code point.
  )
  
  // Code points in the surrogate range are not valid for UTF-8.
  const (
  	surrogateMin = 0xD800
  	surrogateMax = 0xDFFF
  )
  
  const (
  	t1 = 0x00 // 0000 0000
  	tx = 0x80 // 1000 0000
  	t2 = 0xC0 // 1100 0000
  	t3 = 0xE0 // 1110 0000
  	t4 = 0xF0 // 1111 0000
  	t5 = 0xF8 // 1111 1000
  
  	maskx = 0x3F // 0011 1111
  	mask2 = 0x1F // 0001 1111
  	mask3 = 0x0F // 0000 1111
  	mask4 = 0x07 // 0000 0111
  
  	rune1Max = 1<<7 - 1
  	rune2Max = 1<<11 - 1
  	rune3Max = 1<<16 - 1
  
  	// The default lowest and highest continuation byte.
  	locb = 0x80 // 1000 0000
  	hicb = 0xBF // 1011 1111
  )
  
  // decoderune returns the non-ASCII rune at the start of
  // s[k:] and the index after the rune in s.
  //
  // decoderune assumes that caller has checked that
  // the to be decoded rune is a non-ASCII rune.
  //
  // If the string appears to be incomplete or decoding problems
  // are encountered (runeerror, k + 1) is returned to ensure
  // progress when decoderune is used to iterate over a string.
  func decoderune(s string, k int) (r rune, pos int) {
  	pos = k
  
  	if k >= len(s) {
  		return runeError, k + 1
  	}
  
  	s = s[k:]
  
  	switch {
  	case t2 <= s[0] && s[0] < t3:
  		// 0080-07FF two byte sequence
  		if len(s) > 1 && (locb <= s[1] && s[1] <= hicb) {
  			r = rune(s[0]&mask2)<<6 | rune(s[1]&maskx)
  			pos += 2
  			if rune1Max < r {
  				return
  			}
  		}
  	case t3 <= s[0] && s[0] < t4:
  		// 0800-FFFF three byte sequence
  		if len(s) > 2 && (locb <= s[1] && s[1] <= hicb) && (locb <= s[2] && s[2] <= hicb) {
  			r = rune(s[0]&mask3)<<12 | rune(s[1]&maskx)<<6 | rune(s[2]&maskx)
  			pos += 3
  			if rune2Max < r && !(surrogateMin <= r && r <= surrogateMax) {
  				return
  			}
  		}
  	case t4 <= s[0] && s[0] < t5:
  		// 10000-1FFFFF four byte sequence
  		if len(s) > 3 && (locb <= s[1] && s[1] <= hicb) && (locb <= s[2] && s[2] <= hicb) && (locb <= s[3] && s[3] <= hicb) {
  			r = rune(s[0]&mask4)<<18 | rune(s[1]&maskx)<<12 | rune(s[2]&maskx)<<6 | rune(s[3]&maskx)
  			pos += 4
  			if rune3Max < r && r <= maxRune {
  				return
  			}
  		}
  	}
  
  	return runeError, k + 1
  }
  
  // encoderune writes into p (which must be large enough) the UTF-8 encoding of the rune.
  // It returns the number of bytes written.
  func encoderune(p []byte, r rune) int {
  	// Negative values are erroneous. Making it unsigned addresses the problem.
  	switch i := uint32(r); {
  	case i <= rune1Max:
  		p[0] = byte(r)
  		return 1
  	case i <= rune2Max:
  		_ = p[1] // eliminate bounds checks
  		p[0] = t2 | byte(r>>6)
  		p[1] = tx | byte(r)&maskx
  		return 2
  	case i > maxRune, surrogateMin <= i && i <= surrogateMax:
  		r = runeError
  		fallthrough
  	case i <= rune3Max:
  		_ = p[2] // eliminate bounds checks
  		p[0] = t3 | byte(r>>12)
  		p[1] = tx | byte(r>>6)&maskx
  		p[2] = tx | byte(r)&maskx
  		return 3
  	default:
  		_ = p[3] // eliminate bounds checks
  		p[0] = t4 | byte(r>>18)
  		p[1] = tx | byte(r>>12)&maskx
  		p[2] = tx | byte(r>>6)&maskx
  		p[3] = tx | byte(r)&maskx
  		return 4
  	}
  }
  

View as plain text