1// Copyright 2011 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 5package unicode 6 7// Bit masks for each code point under U+0100, for fast lookup. 8const ( 9 pC = 1 << iota // a control character. 10 pP // a punctuation character. 11 pN // a numeral. 12 pS // a symbolic character. 13 pZ // a spacing character. 14 pLu // an upper-case letter. 15 pLl // a lower-case letter. 16 pp // a printable character according to Go's definition. 17 pg = pp | pZ // a graphical character according to the Unicode definition. 18 pLo = pLl | pLu // a letter that is neither upper nor lower case. 19 pLmask = pLo 20) 21 22// GraphicRanges defines the set of graphic characters according to Unicode. 23var GraphicRanges = []*RangeTable{ 24 L, M, N, P, S, Zs, 25} 26 27// PrintRanges defines the set of printable characters according to Go. 28// ASCII space, U+0020, is handled separately. 29var PrintRanges = []*RangeTable{ 30 L, M, N, P, S, 31} 32 33// IsGraphic reports whether the rune is defined as a Graphic by Unicode. 34// Such characters include letters, marks, numbers, punctuation, symbols, and 35// spaces, from categories L, M, N, P, S, Zs. 36func IsGraphic(r rune) bool { 37 // We convert to uint32 to avoid the extra test for negative, 38 // and in the index we convert to uint8 to avoid the range check. 39 if uint32(r) <= MaxLatin1 { 40 return properties[uint8(r)]&pg != 0 41 } 42 return In(r, GraphicRanges...) 43} 44 45// IsPrint reports whether the rune is defined as printable by Go. Such 46// characters include letters, marks, numbers, punctuation, symbols, and the 47// ASCII space character, from categories L, M, N, P, S and the ASCII space 48// character. This categorization is the same as IsGraphic except that the 49// only spacing character is ASCII space, U+0020. 50func IsPrint(r rune) bool { 51 if uint32(r) <= MaxLatin1 { 52 return properties[uint8(r)]&pp != 0 53 } 54 return In(r, PrintRanges...) 55} 56 57// IsOneOf reports whether the rune is a member of one of the ranges. 58// The function "In" provides a nicer signature and should be used in preference to IsOneOf. 59func IsOneOf(ranges []*RangeTable, r rune) bool { 60 for _, inside := range ranges { 61 if Is(inside, r) { 62 return true 63 } 64 } 65 return false 66} 67 68// In reports whether the rune is a member of one of the ranges. 69func In(r rune, ranges ...*RangeTable) bool { 70 for _, inside := range ranges { 71 if Is(inside, r) { 72 return true 73 } 74 } 75 return false 76} 77 78// IsControl reports whether the rune is a control character. 79// The C (Other) Unicode category includes more code points 80// such as surrogates; use Is(C, r) to test for them. 81func IsControl(r rune) bool { 82 if uint32(r) <= MaxLatin1 { 83 return properties[uint8(r)]&pC != 0 84 } 85 // All control characters are < MaxLatin1. 86 return false 87} 88 89// IsLetter reports whether the rune is a letter (category L). 90func IsLetter(r rune) bool { 91 if uint32(r) <= MaxLatin1 { 92 return properties[uint8(r)]&(pLmask) != 0 93 } 94 return isExcludingLatin(Letter, r) 95} 96 97// IsMark reports whether the rune is a mark character (category M). 98func IsMark(r rune) bool { 99 // There are no mark characters in Latin-1. 100 return isExcludingLatin(Mark, r) 101} 102 103// IsNumber reports whether the rune is a number (category N). 104func IsNumber(r rune) bool { 105 if uint32(r) <= MaxLatin1 { 106 return properties[uint8(r)]&pN != 0 107 } 108 return isExcludingLatin(Number, r) 109} 110 111// IsPunct reports whether the rune is a Unicode punctuation character 112// (category P). 113func IsPunct(r rune) bool { 114 if uint32(r) <= MaxLatin1 { 115 return properties[uint8(r)]&pP != 0 116 } 117 return Is(Punct, r) 118} 119 120// IsSpace reports whether the rune is a space character as defined 121// by Unicode's White Space property; in the Latin-1 space 122// this is 123// '\t', '\n', '\v', '\f', '\r', ' ', U+0085 (NEL), U+00A0 (NBSP). 124// Other definitions of spacing characters are set by category 125// Z and property Pattern_White_Space. 126func IsSpace(r rune) bool { 127 // This property isn't the same as Z; special-case it. 128 if uint32(r) <= MaxLatin1 { 129 switch r { 130 case '\t', '\n', '\v', '\f', '\r', ' ', 0x85, 0xA0: 131 return true 132 } 133 return false 134 } 135 return isExcludingLatin(White_Space, r) 136} 137 138// IsSymbol reports whether the rune is a symbolic character. 139func IsSymbol(r rune) bool { 140 if uint32(r) <= MaxLatin1 { 141 return properties[uint8(r)]&pS != 0 142 } 143 return isExcludingLatin(Symbol, r) 144} 145