/
sentencerules.go
205 lines (189 loc) · 7.53 KB
/
sentencerules.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
package uniseg
import "unicode/utf8"
// The states of the sentence break parser.
const (
sbAny = iota
sbCR
sbParaSep
sbATerm
sbUpper
sbLower
sbSB7
sbSB8Close
sbSB8Sp
sbSTerm
sbSB8aClose
sbSB8aSp
)
// The sentence break parser's breaking instructions.
const (
sbDontBreak = iota
sbBreak
)
// The sentence break parser's state transitions. It's anologous to
// grTransitions, see comments there for details. Unicode version 14.0.0.
var sbTransitions = map[[2]int][3]int{
// SB3.
{sbAny, prCR}: {sbCR, sbDontBreak, 9990},
{sbCR, prLF}: {sbParaSep, sbDontBreak, 30},
// SB4.
{sbAny, prSep}: {sbParaSep, sbDontBreak, 9990},
{sbAny, prLF}: {sbParaSep, sbDontBreak, 9990},
{sbParaSep, prAny}: {sbAny, sbBreak, 40},
{sbCR, prAny}: {sbAny, sbBreak, 40},
// SB6.
{sbAny, prATerm}: {sbATerm, sbDontBreak, 9990},
{sbATerm, prNumeric}: {sbAny, sbDontBreak, 60},
{sbSB7, prNumeric}: {sbAny, sbDontBreak, 60}, // Because ATerm also appears in SB7.
// SB7.
{sbAny, prUpper}: {sbUpper, sbDontBreak, 9990},
{sbAny, prLower}: {sbLower, sbDontBreak, 9990},
{sbUpper, prATerm}: {sbSB7, sbDontBreak, 70},
{sbLower, prATerm}: {sbSB7, sbDontBreak, 70},
{sbSB7, prUpper}: {sbUpper, sbDontBreak, 70},
// SB8a.
{sbAny, prSTerm}: {sbSTerm, sbDontBreak, 9990},
{sbATerm, prSContinue}: {sbAny, sbDontBreak, 81},
{sbATerm, prATerm}: {sbATerm, sbDontBreak, 81},
{sbATerm, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSB7, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSB7, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSB7, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSB8Close, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSB8Close, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSB8Close, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSB8Sp, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSB8Sp, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSB8Sp, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSTerm, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSTerm, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSTerm, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSB8aClose, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSB8aClose, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSB8aClose, prSTerm}: {sbSTerm, sbDontBreak, 81},
{sbSB8aSp, prSContinue}: {sbAny, sbDontBreak, 81},
{sbSB8aSp, prATerm}: {sbATerm, sbDontBreak, 81},
{sbSB8aSp, prSTerm}: {sbSTerm, sbDontBreak, 81},
// SB9.
{sbATerm, prClose}: {sbSB8Close, sbDontBreak, 90},
{sbSB7, prClose}: {sbSB8Close, sbDontBreak, 90},
{sbSB8Close, prClose}: {sbSB8Close, sbDontBreak, 90},
{sbATerm, prSp}: {sbSB8Sp, sbDontBreak, 90},
{sbSB7, prSp}: {sbSB8Sp, sbDontBreak, 90},
{sbSB8Close, prSp}: {sbSB8Sp, sbDontBreak, 90},
{sbSTerm, prClose}: {sbSB8aClose, sbDontBreak, 90},
{sbSB8aClose, prClose}: {sbSB8aClose, sbDontBreak, 90},
{sbSTerm, prSp}: {sbSB8aSp, sbDontBreak, 90},
{sbSB8aClose, prSp}: {sbSB8aSp, sbDontBreak, 90},
{sbATerm, prSep}: {sbParaSep, sbDontBreak, 90},
{sbATerm, prCR}: {sbParaSep, sbDontBreak, 90},
{sbATerm, prLF}: {sbParaSep, sbDontBreak, 90},
{sbSB7, prSep}: {sbParaSep, sbDontBreak, 90},
{sbSB7, prCR}: {sbParaSep, sbDontBreak, 90},
{sbSB7, prLF}: {sbParaSep, sbDontBreak, 90},
{sbSB8Close, prSep}: {sbParaSep, sbDontBreak, 90},
{sbSB8Close, prCR}: {sbParaSep, sbDontBreak, 90},
{sbSB8Close, prLF}: {sbParaSep, sbDontBreak, 90},
{sbSTerm, prSep}: {sbParaSep, sbDontBreak, 90},
{sbSTerm, prCR}: {sbParaSep, sbDontBreak, 90},
{sbSTerm, prLF}: {sbParaSep, sbDontBreak, 90},
{sbSB8aClose, prSep}: {sbParaSep, sbDontBreak, 90},
{sbSB8aClose, prCR}: {sbParaSep, sbDontBreak, 90},
{sbSB8aClose, prLF}: {sbParaSep, sbDontBreak, 90},
// SB10.
{sbSB8Sp, prSp}: {sbSB8Sp, sbDontBreak, 100},
{sbSB8aSp, prSp}: {sbSB8aSp, sbDontBreak, 100},
{sbSB8Sp, prSep}: {sbParaSep, sbDontBreak, 100},
{sbSB8Sp, prCR}: {sbParaSep, sbDontBreak, 100},
{sbSB8Sp, prLF}: {sbParaSep, sbDontBreak, 100},
// SB11.
{sbATerm, prAny}: {sbAny, sbBreak, 110},
{sbSB7, prAny}: {sbAny, sbBreak, 110},
{sbSB8Close, prAny}: {sbAny, sbBreak, 110},
{sbSB8Sp, prAny}: {sbAny, sbBreak, 110},
{sbSTerm, prAny}: {sbAny, sbBreak, 110},
{sbSB8aClose, prAny}: {sbAny, sbBreak, 110},
{sbSB8aSp, prAny}: {sbAny, sbBreak, 110},
// We'll always break after ParaSep due to SB4.
}
// transitionSentenceBreakState determines the new state of the sentence break
// parser given the current state and the next code point. It also returns
// whether a sentence boundary was detected. If more than one code point is
// needed to determine the new state, the byte slice or the string starting
// after rune "r" can be used (whichever is not nil or empty) for further
// lookups.
func transitionSentenceBreakState(state int, r rune, b []byte, str string) (newState int, sentenceBreak bool) {
// Determine the property of the next character.
nextProperty := property(sentenceBreakCodePoints, r)
// SB5 (Replacing Ignore Rules).
if nextProperty == prExtend || nextProperty == prFormat {
if state == sbParaSep || state == sbCR {
return sbAny, true // Make sure we don't apply SB5 to SB3 or SB4.
}
if state < 0 {
return sbAny, true // SB1.
}
return state, false
}
// Find the applicable transition in the table.
var rule int
transition, ok := sbTransitions[[2]int{state, nextProperty}]
if ok {
// We have a specific transition. We'll use it.
newState, sentenceBreak, rule = transition[0], transition[1] == sbBreak, transition[2]
} else {
// No specific transition found. Try the less specific ones.
transAnyProp, okAnyProp := sbTransitions[[2]int{state, prAny}]
transAnyState, okAnyState := sbTransitions[[2]int{sbAny, nextProperty}]
if okAnyProp && okAnyState {
// Both apply. We'll use a mix (see comments for grTransitions).
newState, sentenceBreak, rule = transAnyState[0], transAnyState[1] == sbBreak, transAnyState[2]
if transAnyProp[2] < transAnyState[2] {
sentenceBreak, rule = transAnyProp[1] == sbBreak, transAnyProp[2]
}
} else if okAnyProp {
// We only have a specific state.
newState, sentenceBreak, rule = transAnyProp[0], transAnyProp[1] == sbBreak, transAnyProp[2]
// This branch will probably never be reached because okAnyState will
// always be true given the current transition map. But we keep it here
// for future modifications to the transition map where this may not be
// true anymore.
} else if okAnyState {
// We only have a specific property.
newState, sentenceBreak, rule = transAnyState[0], transAnyState[1] == sbBreak, transAnyState[2]
} else {
// No known transition. SB999: Any × Any.
newState, sentenceBreak, rule = sbAny, false, 9990
}
}
// SB8.
if rule > 80 && (state == sbATerm || state == sbSB8Close || state == sbSB8Sp || state == sbSB7) {
// Check the right side of the rule.
var length int
for nextProperty != prOLetter &&
nextProperty != prUpper &&
nextProperty != prLower &&
nextProperty != prSep &&
nextProperty != prCR &&
nextProperty != prLF &&
nextProperty != prATerm &&
nextProperty != prSTerm {
// Move on to the next rune.
if b != nil { // Byte slice version.
r, length = utf8.DecodeRune(b)
b = b[length:]
} else { // String version.
r, length = utf8.DecodeRuneInString(str)
str = str[length:]
}
if r == utf8.RuneError {
break
}
nextProperty = property(sentenceBreakCodePoints, r)
}
if nextProperty == prLower {
return sbLower, false
}
}
return
}