| 1 |
package uniseg |
| 2 |
|
| 3 |
import "unicode/utf8" |
| 4 |
|
| 5 |
// The states of the grapheme cluster parser. |
| 6 |
const ( |
| 7 |
grAny = iota |
| 8 |
grCR |
| 9 |
grControlLF |
| 10 |
grL |
| 11 |
grLVV |
| 12 |
grLVTT |
| 13 |
grPrepend |
| 14 |
grExtendedPictographic |
| 15 |
grExtendedPictographicZWJ |
| 16 |
grRIOdd |
| 17 |
grRIEven |
| 18 |
) |
| 19 |
|
| 20 |
// The grapheme cluster parser's breaking instructions. |
| 21 |
const ( |
| 22 |
grNoBoundary = iota |
| 23 |
grBoundary |
| 24 |
) |
| 25 |
|
| 26 |
// The grapheme cluster parser's state transitions. Maps (state, property) to |
| 27 |
// (new state, breaking instruction, rule number). The breaking instruction |
| 28 |
// always refers to the boundary between the last and next code point. |
| 29 |
// |
| 30 |
// This map is queried as follows: |
| 31 |
// |
| 32 |
// 1. Find specific state + specific property. Stop if found. |
| 33 |
// 2. Find specific state + any property. |
| 34 |
// 3. Find any state + specific property. |
| 35 |
// 4. If only (2) or (3) (but not both) was found, stop. |
| 36 |
// 5. If both (2) and (3) were found, use state and breaking instruction from |
| 37 |
// the transition with the lower rule number, prefer (3) if rule numbers |
| 38 |
// are equal. Stop. |
| 39 |
// 6. Assume grAny and grBoundary. |
| 40 |
var grTransitions = map[[2]int][3]int{ |
| 41 |
// GB5 |
| 42 |
{grAny, prCR}: {grCR, grBoundary, 50}, |
| 43 |
{grAny, prLF}: {grControlLF, grBoundary, 50}, |
| 44 |
{grAny, prControl}: {grControlLF, grBoundary, 50}, |
| 45 |
|
| 46 |
// GB4 |
| 47 |
{grCR, prAny}: {grAny, grBoundary, 40}, |
| 48 |
{grControlLF, prAny}: {grAny, grBoundary, 40}, |
| 49 |
|
| 50 |
// GB3. |
| 51 |
{grCR, prLF}: {grAny, grNoBoundary, 30}, |
| 52 |
|
| 53 |
// GB6. |
| 54 |
{grAny, prL}: {grL, grBoundary, 9990}, |
| 55 |
{grL, prL}: {grL, grNoBoundary, 60}, |
| 56 |
{grL, prV}: {grLVV, grNoBoundary, 60}, |
| 57 |
{grL, prLV}: {grLVV, grNoBoundary, 60}, |
| 58 |
{grL, prLVT}: {grLVTT, grNoBoundary, 60}, |
| 59 |
|
| 60 |
// GB7. |
| 61 |
{grAny, prLV}: {grLVV, grBoundary, 9990}, |
| 62 |
{grAny, prV}: {grLVV, grBoundary, 9990}, |
| 63 |
{grLVV, prV}: {grLVV, grNoBoundary, 70}, |
| 64 |
{grLVV, prT}: {grLVTT, grNoBoundary, 70}, |
| 65 |
|
| 66 |
// GB8. |
| 67 |
{grAny, prLVT}: {grLVTT, grBoundary, 9990}, |
| 68 |
{grAny, prT}: {grLVTT, grBoundary, 9990}, |
| 69 |
{grLVTT, prT}: {grLVTT, grNoBoundary, 80}, |
| 70 |
|
| 71 |
// GB9. |
| 72 |
{grAny, prExtend}: {grAny, grNoBoundary, 90}, |
| 73 |
{grAny, prZWJ}: {grAny, grNoBoundary, 90}, |
| 74 |
|
| 75 |
// GB9a. |
| 76 |
{grAny, prSpacingMark}: {grAny, grNoBoundary, 91}, |
| 77 |
|
| 78 |
// GB9b. |
| 79 |
{grAny, prPreprend}: {grPrepend, grBoundary, 9990}, |
| 80 |
{grPrepend, prAny}: {grAny, grNoBoundary, 92}, |
| 81 |
|
| 82 |
// GB11. |
| 83 |
{grAny, prExtendedPictographic}: {grExtendedPictographic, grBoundary, 9990}, |
| 84 |
{grExtendedPictographic, prExtend}: {grExtendedPictographic, grNoBoundary, 110}, |
| 85 |
{grExtendedPictographic, prZWJ}: {grExtendedPictographicZWJ, grNoBoundary, 110}, |
| 86 |
{grExtendedPictographicZWJ, prExtendedPictographic}: {grExtendedPictographic, grNoBoundary, 110}, |
| 87 |
|
| 88 |
// GB12 / GB13. |
| 89 |
{grAny, prRegionalIndicator}: {grRIOdd, grBoundary, 9990}, |
| 90 |
{grRIOdd, prRegionalIndicator}: {grRIEven, grNoBoundary, 120}, |
| 91 |
{grRIEven, prRegionalIndicator}: {grRIOdd, grBoundary, 120}, |
| 92 |
} |
| 93 |
|
| 94 |
// Graphemes implements an iterator over Unicode extended grapheme clusters, |
| 95 |
// specified in the Unicode Standard Annex #29. Grapheme clusters correspond to |
| 96 |
// "user-perceived characters". These characters often consist of multiple |
| 97 |
// code points (e.g. the "woman kissing woman" emoji consists of 8 code points: |
| 98 |
// woman + ZWJ + heavy black heart (2 code points) + ZWJ + kiss mark + ZWJ + |
| 99 |
// woman) and the rules described in Annex #29 must be applied to group those |
| 100 |
// code points into clusters perceived by the user as one character. |
| 101 |
type Graphemes struct { |
| 102 |
// The code points over which this class iterates. |
| 103 |
codePoints []rune |
| 104 |
|
| 105 |
// The (byte-based) indices of the code points into the original string plus |
| 106 |
// len(original string). Thus, len(indices) = len(codePoints) + 1. |
| 107 |
indices []int |
| 108 |
|
| 109 |
// The current grapheme cluster to be returned. These are indices into |
| 110 |
// codePoints/indices. If start == end, we either haven't started iterating |
| 111 |
// yet (0) or the iteration has already completed (1). |
| 112 |
start, end int |
| 113 |
|
| 114 |
// The index of the next code point to be parsed. |
| 115 |
pos int |
| 116 |
|
| 117 |
// The current state of the code point parser. |
| 118 |
state int |
| 119 |
} |
| 120 |
|
| 121 |
// NewGraphemes returns a new grapheme cluster iterator. |
| 122 |
func NewGraphemes(s string) *Graphemes { |
| 123 |
l := utf8.RuneCountInString(s) |
| 124 |
codePoints := make([]rune, l) |
| 125 |
indices := make([]int, l+1) |
| 126 |
i := 0 |
| 127 |
for pos, r := range s { |
| 128 |
codePoints[i] = r |
| 129 |
indices[i] = pos |
| 130 |
i++ |
| 131 |
} |
| 132 |
indices[l] = len(s) |
| 133 |
g := &Graphemes{ |
| 134 |
codePoints: codePoints, |
| 135 |
indices: indices, |
| 136 |
} |
| 137 |
g.Next() // Parse ahead. |
| 138 |
return g |
| 139 |
} |
| 140 |
|
| 141 |
// Next advances the iterator by one grapheme cluster and returns false if no |
| 142 |
// clusters are left. This function must be called before the first cluster is |
| 143 |
// accessed. |
| 144 |
func (g *Graphemes) Next() bool { |
| 145 |
g.start = g.end |
| 146 |
|
| 147 |
// The state transition gives us a boundary instruction BEFORE the next code |
| 148 |
// point so we always need to stay ahead by one code point. |
| 149 |
|
| 150 |
// Parse the next code point. |
| 151 |
for g.pos <= len(g.codePoints) { |
| 152 |
// GB2. |
| 153 |
if g.pos == len(g.codePoints) { |
| 154 |
g.end = g.pos |
| 155 |
g.pos++ |
| 156 |
break |
| 157 |
} |
| 158 |
|
| 159 |
// Determine the property of the next character. |
| 160 |
nextProperty := property(g.codePoints[g.pos]) |
| 161 |
g.pos++ |
| 162 |
|
| 163 |
// Find the applicable transition. |
| 164 |
var boundary bool |
| 165 |
transition, ok := grTransitions[[2]int{g.state, nextProperty}] |
| 166 |
if ok { |
| 167 |
// We have a specific transition. We'll use it. |
| 168 |
g.state = transition[0] |
| 169 |
boundary = transition[1] == grBoundary |
| 170 |
} else { |
| 171 |
// No specific transition found. Try the less specific ones. |
| 172 |
transAnyProp, okAnyProp := grTransitions[[2]int{g.state, prAny}] |
| 173 |
transAnyState, okAnyState := grTransitions[[2]int{grAny, nextProperty}] |
| 174 |
if okAnyProp && okAnyState { |
| 175 |
// Both apply. We'll use a mix (see comments for grTransitions). |
| 176 |
g.state = transAnyState[0] |
| 177 |
boundary = transAnyState[1] == grBoundary |
| 178 |
if transAnyProp[2] < transAnyState[2] { |
| 179 |
g.state = transAnyProp[0] |
| 180 |
boundary = transAnyProp[1] == grBoundary |
| 181 |
} |
| 182 |
} else if okAnyProp { |
| 183 |
// We only have a specific state. |
| 184 |
g.state = transAnyProp[0] |
| 185 |
boundary = transAnyProp[1] == grBoundary |
| 186 |
// This branch will probably never be reached because okAnyState will |
| 187 |
// always be true given the current transition map. But we keep it here |
| 188 |
// for future modifications to the transition map where this may not be |
| 189 |
// true anymore. |
| 190 |
} else if okAnyState { |
| 191 |
// We only have a specific property. |
| 192 |
g.state = transAnyState[0] |
| 193 |
boundary = transAnyState[1] == grBoundary |
| 194 |
} else { |
| 195 |
// No known transition. GB999: Any x Any. |
| 196 |
g.state = grAny |
| 197 |
boundary = true |
| 198 |
} |
| 199 |
} |
| 200 |
|
| 201 |
// If we found a cluster boundary, let's stop here. The current cluster will |
| 202 |
// be the one that just ended. |
| 203 |
if g.pos-1 == 0 /* GB1 */ || boundary { |
| 204 |
g.end = g.pos - 1 |
| 205 |
break |
| 206 |
} |
| 207 |
} |
| 208 |
|
| 209 |
return g.start != g.end |
| 210 |
} |
| 211 |
|
| 212 |
// Runes returns a slice of runes (code points) which corresponds to the current |
| 213 |
// grapheme cluster. If the iterator is already past the end or Next() has not |
| 214 |
// yet been called, nil is returned. |
| 215 |
func (g *Graphemes) Runes() []rune { |
| 216 |
if g.start == g.end { |
| 217 |
return nil |
| 218 |
} |
| 219 |
return g.codePoints[g.start:g.end] |
| 220 |
} |
| 221 |
|
| 222 |
// Str returns a substring of the original string which corresponds to the |
| 223 |
// current grapheme cluster. If the iterator is already past the end or Next() |
| 224 |
// has not yet been called, an empty string is returned. |
| 225 |
func (g *Graphemes) Str() string { |
| 226 |
if g.start == g.end { |
| 227 |
return "" |
| 228 |
} |
| 229 |
return string(g.codePoints[g.start:g.end]) |
| 230 |
} |
| 231 |
|
| 232 |
// Bytes returns a byte slice which corresponds to the current grapheme cluster. |
| 233 |
// If the iterator is already past the end or Next() has not yet been called, |
| 234 |
// nil is returned. |
| 235 |
func (g *Graphemes) Bytes() []byte { |
| 236 |
if g.start == g.end { |
| 237 |
return nil |
| 238 |
} |
| 239 |
return []byte(string(g.codePoints[g.start:g.end])) |
| 240 |
} |
| 241 |
|
| 242 |
// Positions returns the interval of the current grapheme cluster as byte |
| 243 |
// positions into the original string. The first returned value "from" indexes |
| 244 |
// the first byte and the second returned value "to" indexes the first byte that |
| 245 |
// is not included anymore, i.e. str[from:to] is the current grapheme cluster of |
| 246 |
// the original string "str". If Next() has not yet been called, both values are |
| 247 |
// 0. If the iterator is already past the end, both values are 1. |
| 248 |
func (g *Graphemes) Positions() (int, int) { |
| 249 |
return g.indices[g.start], g.indices[g.end] |
| 250 |
} |
| 251 |
|
| 252 |
// Reset puts the iterator into its initial state such that the next call to |
| 253 |
// Next() sets it to the first grapheme cluster again. |
| 254 |
func (g *Graphemes) Reset() { |
| 255 |
g.start, g.end, g.pos, g.state = 0, 0, 0, grAny |
| 256 |
g.Next() // Parse ahead again. |
| 257 |
} |
| 258 |
|
| 259 |
// GraphemeClusterCount returns the number of user-perceived characters |
| 260 |
// (grapheme clusters) for the given string. To calculate this number, it |
| 261 |
// iterates through the string using the Graphemes iterator. |
| 262 |
func GraphemeClusterCount(s string) (n int) { |
| 263 |
g := NewGraphemes(s) |
| 264 |
for g.Next() { |
| 265 |
n++ |
| 266 |
} |
| 267 |
return |
| 268 |
} |