line_cj.txt (9006B)
1 # 2 # Copyright (C) 2016 and later: Unicode, Inc. and others. 3 # License & terms of use: http://www.unicode.org/copyright.html 4 # Copyright (c) 2016, International Business Machines Corporation and others. All Rights Reserved. 5 6 # file: line.txt 7 # 8 # Reference Line Break rules for intltest rbbi/RBBIMonkeyTest. 9 # Rules derived from Unicode Standard Annex #14 for Unicode 16.0. 10 # 11 # Note: Rule syntax and the monkey test itself are still a work in progress. 12 # They are expected to change with review and the addition of support for rule tailoring. 13 14 15 type = line; 16 locale = zh; 17 18 19 AI = [:LineBreak = Ambiguous:]; 20 AK = [:LineBreak = Aksara:]; 21 AL = [:LineBreak = Alphabetic:]; 22 AP = [:LineBreak = Aksara_Prebase:]; 23 AS = [:LineBreak = Aksara_Start:]; 24 BA = [:LineBreak = Break_After:]; 25 HH = [:LineBreak = Unambiguous_Hyphen:]; 26 BB = [:LineBreak = Break_Before:]; 27 BK = [:LineBreak = Mandatory_Break:]; 28 B2 = [:LineBreak = Break_Both:]; 29 CB = [:LineBreak = Contingent_Break:]; 30 CJ = [:LineBreak = Conditional_Japanese_Starter:]; 31 CL = [[:LineBreak = Close_Punctuation:] \u201d]; 32 CMS = [:LineBreak = Combining_Mark:]; 33 CP = [:LineBreak = Close_Parenthesis:]; 34 CR = [:LineBreak = Carriage_Return:]; 35 EB = [:LineBreak = EB:]; 36 EM = [:LineBreak = EM:]; 37 EX = [:LineBreak = Exclamation:]; 38 GL = [:LineBreak = Glue:]; 39 HL = [:LineBreak = Hebrew_Letter:]; 40 HY = [:LineBreak = Hyphen:]; 41 H2 = [:LineBreak = H2:]; 42 H3 = [:LineBreak = H3:]; 43 ID = [:LineBreak = Ideographic:]; 44 IN = [:LineBreak = Inseperable:]; 45 IS = [:LineBreak = Infix_Numeric:]; 46 JL = [:LineBreak = JL:]; 47 JV = [:LineBreak = JV:]; 48 JT = [:LineBreak = JT:]; 49 LF = [:LineBreak = Line_Feed:]; 50 NL = [:LineBreak = Next_Line:]; 51 NS = [[:LineBreak = Nonstarter:] CJ]; # CSS Strict tailoring: CJ resolves to NS. 52 NU = [:LineBreak = Numeric:]; 53 OP = [[:LineBreak = Open_Punctuation:] \u201c]; 54 PO = [:LineBreak = Postfix_Numeric:]; 55 PR = [:LineBreak = Prefix_Numeric:]; 56 QU = [[:LineBreak = Quotation:] - [\u201c\u201d]]; 57 RI = [:LineBreak = Regional_Indicator:]; 58 SA = [:LineBreak = Complex_Context:]; 59 SG = [:LineBreak = Surrogate:]; 60 SP = [:LineBreak = Space:]; 61 SY = [:LineBreak = Break_Symbols:]; 62 VF = [:LineBreak = Virama_Final:]; 63 VI = [:LineBreak = Virama:]; 64 WJ = [:LineBreak = Word_Joiner:]; 65 XX = [:LineBreak = Unknown:]; 66 ZW = [:LineBreak = ZWSpace:]; 67 ZWJ = [:LineBreak = ZWJ:]; 68 69 # OP30 and CP30 are variants of OP and CP that appear in rule LB30 from UAX 14. 70 # Limitations of this monkey test rule parser require that these definitions be pulled out 71 # rather than appearing in-line in LB 30. 72 73 OP30 = [OP - [\p{ea=F}\p{ea=W}\p{ea=H}]]; 74 CP30 = [CP - [\p{ea=F}\p{ea=W}\p{ea=H}]]; 75 76 # OP30 and CP30 are variants of OP and CP that appear in rule LB30 from UAX 14. 77 # Limitations of this monkey test rule parser require that these definitions be pulled out 78 # rather than appearing in-line in LB 30. 79 80 OP30 = [OP - [\p{ea=F}\p{ea=W}\p{ea=H}]]; 81 CP30 = [CP - [\p{ea=F}\p{ea=W}\p{ea=H}]]; 82 83 eaFWH = [\p{ea=F}\p{ea=W}\p{ea=H}]; 84 eaFWHminusCMOPGL = [ eaFWH - [CMS OP GL] ]; 85 eaFWHminusCM = [ eaFWH - CMS ]; 86 eaFWHBreakableAtLB19 = [ eaFWH - [NS BA EX CL IN IS GL CMS] ]; 87 BAminuseaFWH = [BA - eaFWH ]; 88 89 PiQU = [\p{Pi}&QU]; 90 PfQU = [\p{Pf}&QU]; 91 92 # The redundant-looking inner brackets are required for the current parser in the test code. 93 ExtPictUnassigned = [[\p{Extended_Pictographic}]&[\p{Cn}]]; 94 95 # LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes 96 AL = [AL AI SG XX ]; 97 dictionary = SA; 98 99 # By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly 100 # list it in the numerous rules that use CM. 101 CM = [CMS ZWJ]; 102 103 LB4: BK ÷; 104 LB5: CR LF; 105 LB5.1: CR ÷; 106 LB5.2: LF ÷; 107 LB5.3: NL ÷; 108 109 LB6: . (BK | CR | LF | NL); 110 LB6.1: [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL); 111 112 # LB8 break after ZW SP*. 113 # Precedes LB7 because both rules will match the sequences like ZW SP, 114 # and LB8 must take precedence. 115 116 LB8: ZW SP* ÷ [^ZW SP BK CR LF NL]; 117 118 # Numbers. Equivalent to Tailoring example 8 from UAX 14. 119 # Moved up, before LB14, because it can match longer sequences which must take precedence. 120 LB25: ((PR | PO)CM*)? ((OP | HY)CM*)? (IS CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?; 121 122 # Rules LB14 - LB17. 123 124 # Moved before LB14, because it matches a supersequence. 125 LB20a.3: ( OP CM* SP* | QU CM* | GL CM* ) (PiQU CM* SP*)+ SP (HY | HH) CM* (AL | HL); 126 LB15a.1: ( OP CM* SP* | QU CM* | GL CM* ) (PiQU CM* SP*)+ .; 127 128 # Moved before LB14. These are really the cases where LB19a does not apply, but 129 # the old LB19 would. This is to avoid many instances of chaining over two code 130 # points. 131 LB19a.1: eaFWHminusCMOPGL ÷ PiQU CM* eaFWHminusCM; 132 LB19a.2: eaFWHminusCMOPGL CM* CMS ÷ PiQU CM* eaFWHminusCM; 133 LB19a.5: eaFWHminusCM CM* PfQU ÷ eaFWHBreakableAtLB19; 134 LB19a.6: eaFWHminusCM CM* PfQU CM* CMS ÷ eaFWHBreakableAtLB19; 135 136 # Moved up, before LB7, because they can match a longer sequence that would also match LB7. 137 # For example, the sequence "OP CM SP AL" matches LB14 138 # while the prefix of it, "OP CM SP" matches LB7.1 139 LB20a.7: OP CM* SP+ (HY | HH) CM* (AL | HL); 140 LB14: OP CM* SP* .; 141 142 LB20a.4: ^ (PiQU CM* SP*)+ SP (HY | HH) CM* (AL | HL); 143 LB15a.2: ^ (PiQU CM* SP*)+ .; 144 # LB15b/LB15a chaining. 145 LB20a.5: ([^SP] CM* | SP) PfQU CM* (PiQU CM* SP*)+ SP (HY | HH) CM* (AL | HL); 146 LB15b.1: ([^SP] CM* | SP) PfQU CM* (PiQU CM* SP*)+ .; 147 LB15b.2: ([^SP] CM* | SP) PfQU CM* ( SP | GL | WJ | CL | QU | CP | EX | IS | SY | BK | CR | LF | NL | ZW | $ ); 148 149 # LB 15c Break before an IS that begins a number and follows a space. 150 LB15c: SP ÷ IS CM* NU; 151 152 # LB15d × IS 153 LB15d.1: [^SP] CM* IS; 154 LB15d.2: SP IS; 155 156 LB16: (CL | CP)CM* SP* NS; 157 LB17: B2 CM* SP* B2; 158 159 160 # LB7 Do not break before spaces or zero width space. 161 162 LB7.1: [^ZW SP] CM* [SP ZW]; 163 LB7.2: [ZW SP] [SP ZW]; 164 165 # LB8a 166 # ZWJ x 167 # Don't match a CM on the right - let other rules pick up CM sequences, where 168 # the ZWJ behaves as just another generic CM. 169 LB8a: ZWJ [^CM]; 170 171 172 # LB9: X CM -> X 173 # LB10: Unattached CM -> AL 174 175 #LB11: × WJ; 176 # WJ × 177 178 LB11.1: [^SP] CM* WJ; 179 LB11.2: SP WJ; 180 LB11.3: WJ CM* [^CM]; 181 182 # Needs to apply before LB12, because the new monkeys are not greedy. 183 LB20a.2: GL CM* (HY | HH) CM* (AL | HL); 184 LB12: GL CM* [^CM]; 185 186 LB12a: [^SP BA HY HH] CM* GL; 187 188 # LB 13 Do not break before ‘]’ or ‘!’ or ‘/’, even after spaces. 189 LB13.1: [^SP] CM* [CL CP EX SY]; 190 LB13.2: SP [CL CP EX SY]; 191 192 # LB 14-17 are moved above LB 7. 193 194 LB18: SP ÷; 195 196 LB19: . CM* QU; 197 LB19.1: QU CM* [^CM]; 198 199 # LB 20 Break before and after CB. 200 # Interaction with LB8a: ZWJ x . is tricky because CM includes ZWJ. 201 # ZWJ acts like a CM to the left, combining with CB. 202 # ZWJ acts independently to the right, no break after by LB8a. 203 LB20.1: . CM* ZWJ CB; 204 LB20.2: . CM* ÷ CB; 205 206 LB20a.6: CB CM* ZWJ (HY | HH) CM* (AL | HL); 207 LB20.3: CB CM* ZWJ [^CM]; 208 LB20.4: CB CM* ÷; 209 210 # LB 20a Do not break after a word-initial hyphen. 211 LB20a.1: ^(HY | HH) CM* (AL | HL); 212 213 # Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then 214 # not picking up the continuing match after the BA from 21a. 215 LB21a: HL CM* (HY | HH) CM* [^CM CB HL]; 216 217 LB21.1: . CM* [BA HY HH NS]; 218 LB21.2: BB CM* [^CM CB]; 219 220 LB21b: SY CM* HL; 221 222 LB22: . CM* IN; 223 224 LB23.1: (AL | HL | CM) CM* NU; 225 LB23.2: NU CM* (AL | HL); 226 227 LB23a.1: PR CM* (ID | EB | EM); 228 LB23a.2: (ID | EB | EM) CM* PO; 229 230 LB24.2: (PR | PO) CM* (AL | HL); 231 LB24.3: (AL | HL | CM) CM* (PR | PO); 232 233 234 LB26.1: JL CM* (JL | JV | H2 | H3); 235 LB26.2: (JV | H2) CM* (JV | JT); 236 LB26.3: (JT | H3) CM* JT; 237 238 LB27.1: (JL | JV | JT | H2 | H3) CM* PO; 239 LB27.2: PR CM* (JL | JV | JT | H2 | H3); 240 241 # LB28 Do not break between Alphabetics. 242 # Unattached (leading) CM treated as AL. 243 LB28: (AL | HL | CM)CM* (AL | HL); 244 245 LB28: (AP CM*)? (AS | AK | [◌] ) (CM* VI CM* (AK | [◌] ))* (CM* VI | ((CM* (AS | AK | [◌] ) )? CM* VF))?; 246 247 LB29: IS CM* (AL | HL); 248 249 # LB30 is adjusted for unattached leading CM being treated as AL. 250 LB30.1: (AL | CM | HL | NU) CM* OP30; 251 LB30.2: CP30 CM* (AL | HL | NU); 252 253 # LB30a keep pairs of RI together. 254 LB30a.1: RI CM* RI ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HH HY NS IN CM]; 255 LB30a.2: RI CM* RI CM* CMS ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HH HY NS IN CM]; 256 LB30a.3: RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HH HY NS IN ZWJ]?; 257 258 # LB30b Do not break between Emoji Base (or potential emoji) and Emoji Modifier 259 LB30b.1: EB CM* EM; 260 LB30b.2: ExtPictUnassigned CM* EM; 261 262 # LB31 Break Everywhere Else. 263 # Include combining marks 264 LB31.1: . CM* ZWJ [^CM]; 265 LB31.2: . CM* ÷;