line_phrase_cj.txt (14661B)
1 # Copyright (C) 2022 and later: Unicode, Inc. and others. 2 # License & terms of use: http://www.unicode.org/copyright.html 3 # 4 # file: line_phrase_cj.txt 5 # 6 # Line Breaking Rules 7 # Implement default line breaking as defined by 8 # Unicode Standard Annex #14 (https://www.unicode.org/reports/tr14/) 9 # for Unicode 14.0, with the following modification: 10 # 11 # Boundaries between hyphens and following letters are suppressed when 12 # there is a boundary preceding the hyphen. See rule 20.9 13 # 14 # This corresponds to CSS line-break-word-handling=phrase (BCP47 -u-lw-phrase). 15 # It sets characters of class CJ to behave like NS. 16 # It allows breaking before 201C and after 201D, for zh_Hans, zh_Hant, and ja. 17 # 18 # The content is the same as line_cj.txt except the following 19 # 1. Add CJK into dictionary. 20 # 2. Add East Asian Width with class F, W and H into $ALPlus. 21 # 22 # Character Classes defined by TR 14. 23 # 24 25 !!chain; 26 !!quoted_literals_only; 27 28 $AI = [:LineBreak = Ambiguous:]; 29 $AK = [:LineBreak = Aksara:]; 30 $AL = [:LineBreak = Alphabetic:]; 31 $AP = [:LineBreak = Aksara_Prebase:]; 32 $AS = [:LineBreak = Aksara_Start:]; 33 $BA = [:LineBreak = Break_After:]; 34 $HH = [:LineBreak = Unambiguous_Hyphen:]; 35 $BB = [:LineBreak = Break_Before:]; 36 $BK = [:LineBreak = Mandatory_Break:]; 37 $B2 = [:LineBreak = Break_Both:]; 38 $CB = [:LineBreak = Contingent_Break:]; 39 $CJ = [:LineBreak = Conditional_Japanese_Starter:]; 40 $CL = [[:LineBreak = Close_Punctuation:] \u201d]; 41 # $CM = [:LineBreak = Combining_Mark:]; 42 $CP = [:LineBreak = Close_Parenthesis:]; 43 $CR = [:LineBreak = Carriage_Return:]; 44 $EB = [:LineBreak = EB:]; 45 $EM = [:LineBreak = EM:]; 46 $EX = [:LineBreak = Exclamation:]; 47 $GL = [:LineBreak = Glue:]; 48 $HL = [:LineBreak = Hebrew_Letter:]; 49 $HY = [:LineBreak = Hyphen:]; 50 $H2 = [:LineBreak = H2:]; 51 $H3 = [:LineBreak = H3:]; 52 $ID = [:LineBreak = Ideographic:]; 53 $IN = [:LineBreak = Inseperable:]; 54 $IS = [:LineBreak = Infix_Numeric:]; 55 $JL = [:LineBreak = JL:]; 56 $JV = [:LineBreak = JV:]; 57 $JT = [:LineBreak = JT:]; 58 $LF = [:LineBreak = Line_Feed:]; 59 $NL = [:LineBreak = Next_Line:]; 60 # NS includes CJ for CSS strict line breaking. 61 $NS = [[:LineBreak = Nonstarter:] $CJ]; 62 $NU = [:LineBreak = Numeric:]; 63 $OP = [[:LineBreak = Open_Punctuation:] \u201c]; 64 $PO = [:LineBreak = Postfix_Numeric:]; 65 $PR = [:LineBreak = Prefix_Numeric:]; 66 $QU = [[:LineBreak = Quotation:] - [\u201c\u201d]]; 67 $RI = [:LineBreak = Regional_Indicator:]; 68 $SA = [:LineBreak = Complex_Context:]; 69 $SG = [:LineBreak = Surrogate:]; 70 $SP = [:LineBreak = Space:]; 71 $SY = [:LineBreak = Break_Symbols:]; 72 $VF = [:LineBreak = Virama_Final:]; 73 $VI = [:LineBreak = Virama:]; 74 $WJ = [:LineBreak = Word_Joiner:]; 75 $XX = [:LineBreak = Unknown:]; 76 $ZW = [:LineBreak = ZWSpace:]; 77 $ZWJ = [:LineBreak = ZWJ:]; 78 79 $EastAsian = [\p{ea=F}\p{ea=W}\p{ea=H}]; 80 81 $ExtPictUnassigned = [\p{Extended_Pictographic} & \p{Cn}]; 82 83 # By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly 84 # list it in the numerous rules that use CM. 85 # By LB1, SA characters with general categor of Mn or Mc also resolve to CM. 86 87 $CM = [[:LineBreak = Combining_Mark:] $ZWJ [$SA & [[:Mn:][:Mc:]]]]; 88 $CMX = [[$CM] - [$ZWJ]]; 89 90 # Dictionary character set, for triggering language-based break engines. Currently 91 # limited to LineBreak=Complex_Context (SA) and $dictionaryCJK. 92 93 # Add CJK dictionary 94 $Han = [:Han:]; 95 $Katakana = [:Katakana:]; 96 $Hiragana = [:Hiragana:]; 97 $HangulSyllable = [\uac00-\ud7a3]; 98 $ComplexContext = [:LineBreak = Complex_Context:]; 99 $KanaKanji = [$Han $Hiragana $Katakana \u30fc]; 100 $dictionaryCJK = [$KanaKanji $HangulSyllable]; 101 $dictionary = [$ComplexContext $dictionaryCJK]; 102 103 104 # 105 # Rule LB1. By default, treat AI (characters with ambiguous east Asian width), 106 # SA (Dictionary chars, excluding Mn and Mc) 107 # SG (Unpaired Surrogates) 108 # XX (Unknown, unassigned) 109 # as $AL (Alphabetic) 110 # 111 # Let fullwidth-ASCII digits and letters be part of words. 112 $FW_alphanum = [\uff10-\uff19\uff21-\uff3a\uff41-\uff5a]; 113 $ALPlus = [$AL $AI $SG $XX $FW_alphanum [$dictionary-[[:Mn:][:Mc:]]]]; 114 115 116 ## ------------------------------------------------- 117 118 # 119 # CAN_CM is the set of characters that may combine with CM combining chars. 120 # Note that Linebreak UAX 14's concept of a combining char and the rules 121 # for what they can combine with are _very_ different from the rest of Unicode. 122 # 123 # Note that $CM itself is left out of this set. If CM is needed as a base 124 # it must be listed separately in the rule. 125 # 126 $CAN_CM = [^$SP $BK $CR $LF $NL $ZW $CM]; # Bases that can take CMs 127 $CANT_CM = [ $SP $BK $CR $LF $NL $ZW $CM]; # Bases that can't take CMs 128 129 # 130 # AL_FOLLOW set of chars that can unconditionally follow an AL 131 # Needed in rules where stand-alone $CM s are treated as AL. 132 # 133 $AL_FOLLOW = [$BK $CR $LF $NL $ZW $SP $CL $CP $EX $HL $IS $SY $WJ $GL [$OP - $EastAsian] $QU $BA $HH $HY $NS $IN $NU $PR $PO $ALPlus]; 134 135 136 # 137 # Rule LB 4, 5 Mandatory (Hard) breaks. 138 # 139 $LB4Breaks = [$BK $CR $LF $NL]; 140 $LB4NonBreaks = [^$BK $CR $LF $NL $CM]; 141 $CR $LF {100}; 142 143 # 144 # LB 6 Do not break before hard line breaks. 145 # 146 $LB4NonBreaks? $LB4Breaks {100}; # LB 5 do not break before hard breaks. 147 $CAN_CM $CM* $LB4Breaks {100}; 148 ^$CM+ $LB4Breaks {100}; 149 150 # LB 7 x SP 151 # x ZW 152 $LB4NonBreaks [$SP $ZW]; 153 $CAN_CM $CM* [$SP $ZW]; 154 ^$CM+ [$SP $ZW]; 155 156 # 157 # LB 8 Break after zero width space 158 # ZW SP* ÷ 159 # 160 $LB8Breaks = [$LB4Breaks $ZW]; 161 $LB8NonBreaks = [[$LB4NonBreaks] - [$ZW]]; 162 $ZW $SP* / [^$SP $ZW $LB4Breaks]; 163 164 # LB 8a ZWJ x Do not break Emoji ZWJ sequences. 165 # 166 $ZWJ [^$CM]; 167 168 # LB 9 Combining marks. X $CM needs to behave like X, where X is not $SP, $BK $CR $LF $NL 169 # $CM not covered by the above needs to behave like $AL 170 # See definition of $CAN_CM. 171 172 $CAN_CM $CM+; # Stick together any combining sequences that don't match other rules. 173 ^$CM+; 174 175 # 176 # LB 11 Do not break before or after WORD JOINER & related characters. 177 # 178 $CAN_CM $CM* $WJ; 179 $LB8NonBreaks $WJ; 180 ^$CM+ $WJ; 181 182 $WJ $CM* .; 183 184 # 185 # LB 12 Do not break after NBSP and related characters. 186 # GL x 187 # 188 $GL $CM* .; 189 190 # 191 # LB 12a Do not break before NBSP and related characters ... 192 # [^SP BA HY HH] x GL 193 # 194 [[$LB8NonBreaks] - [$SP $BA $HY $HH]] $CM* $GL; 195 ^$CM+ $GL; 196 197 198 199 200 # LB 13 Don't break before ']' or '!' or '/', even after spaces. 201 # 202 $LB8NonBreaks $CL; 203 $CAN_CM $CM* $CL; 204 ^$CM+ $CL; # by rule 10, stand-alone CM behaves as AL 205 206 $LB8NonBreaks $CP; 207 $CAN_CM $CM* $CP; 208 ^$CM+ $CP; # by rule 10, stand-alone CM behaves as AL 209 210 $LB8NonBreaks $EX; 211 $CAN_CM $CM* $EX; 212 ^$CM+ $EX; # by rule 10, stand-alone CM behaves as AL 213 214 $LB8NonBreaks $SY; 215 $CAN_CM $CM* $SY; 216 ^$CM+ $SY; # by rule 10, stand-alone CM behaves as AL 217 218 219 # 220 # LB 14 Do not break after OP, even after spaces 221 # Note subtle interaction with "SP IS /" rules in LB14a. 222 # This rule consumes the SP, chaining happens on the IS, effectivley overriding the SP IS rules, 223 # which is the desired behavior. 224 # 225 $OP $CM* $SP* .; 226 227 $OP $CM* $SP+ $CM+ $AL_FOLLOW?; # by rule 10, stand-alone CM behaves as AL 228 # by rule 8, CM following a SP is stand-alone. 229 230 231 # LB 15a 232 ($OP $CM* $SP+ | [$OP $QU $GL] $CM*) ([\p{Pi} & $QU] $CM* $SP*)+ .; 233 ($OP $CM* $SP+ | [$OP $QU $GL] $CM*) ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?; 234 ^([\p{Pi} & $QU] $CM* $SP*)+ .; 235 ^([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?; 236 237 # LB 15b 238 $LB8NonBreaks [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}]; 239 $CAN_CM $CM* [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}]; 240 ^$CM+ [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}]; 241 242 # Messy interaction: manually chain between LB 15b and LB 15a on Pf Pi. 243 $LB8NonBreaks [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .; 244 $LB8NonBreaks [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?; 245 $CAN_CM $CM* [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .; 246 $CAN_CM $CM* [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?; 247 ^$CM+ [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .; 248 ^$CM+ [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?; 249 250 251 # LB 15c Force a break before start of a number with a leading decimal pt, e.g. " .23" 252 # Note: would be simpler to express as "$SP / $IS $CM* $NU;", but ICU rules have limitations. 253 # See issue ICU-20303 254 255 256 $CanFollowIS = [$BK $CR $LF $NL $SP $ZW $WJ $GL $CL $CP $EX $IS $SY $QU $BA $HH $HY $NS $ALPlus $HL $IN]; 257 $SP $IS / [^ $CanFollowIS $NU $CM]; 258 $SP $IS $CM* $CMX / [^ $CanFollowIS $NU $CM]; 259 260 # 261 # LB 15d Do not break before numeric separators (IS), even after spaces. 262 # SP IS QU is handled below as part of LB 19. 263 264 [$LB8NonBreaks - $SP] $IS; 265 $SP $IS $CM* [$CanFollowIS {eof}]; 266 $SP $IS $CM* $ZWJ [^$CM $NU]; 267 268 $CAN_CM $CM* $IS; 269 ^$CM+ $IS; # by rule 10, stand-alone CM behaves as AL 270 271 272 # LB 16 273 ($CL | $CP) $CM* $SP* $NS; 274 275 # LB 17 276 $B2 $CM* $SP* $B2; 277 278 # 279 # LB 18 Break after spaces. 280 # 281 $LB18NonBreaks = [$LB8NonBreaks - [$SP]]; 282 $LB18Breaks = [$LB8Breaks $SP]; 283 284 285 # LB 19 and LB 19a. 286 # Instead of implementing both as keep-together rules as in UAX #14, we have an 287 # East_Asian_Width and General_Category-insensitive keep-together rule 288 # equivalent to the old LB19 × QU and QU ×, and then we poke holes into it based 289 # on context. This avoids having to do manual chaining over multiple characters 290 # with many other rules over multiple characters, as a keep-together LB19a would 291 # overlap in context with at least LB14, LB15a, LB15a, LB15d, LB30a, and itself. 292 $LB18NonBreaks $CM* $QU; 293 ^$CM+ $QU; 294 295 [$LB18NonBreaks & $EastAsian - [$OP $GL]] / [\p{Pi} & $QU] $CM* [ $EastAsian - $CM]; 296 [$LB18NonBreaks & $EastAsian - [$OP $GL]] $CM* $CMX / [\p{Pi} & $QU] $CM* [ $EastAsian - $CM]; 297 298 $QU $CM* .; 299 [$LB18NonBreaks & $EastAsian] $CM* [\p{Pf} & $QU] / [ $EastAsian - [$NS $BA $EX $CL $IN $IS $GL $CM]]; 300 [$LB18NonBreaks & $EastAsian] $CM* [\p{Pf} & $QU] $CM* $CMX / [ $EastAsian - [$NS $BA $EX $CL $IN $IS $GL $CM]]; 301 302 # LB 20 303 # <break> $CB 304 # $CB <break> 305 # 306 $LB20NonBreaks = [$LB18NonBreaks - $CB]; 307 308 # LB 20a Don't break between Hyphens and Letters when there is a break preceding the hyphen. 309 # Originally added as a Finnish tailoring, promoted to default ICU behavior (ICU-8151), 310 # and then to default UAX #14 behaviour (UTC-179-C32). 311 # 312 ^($HY | $HH) $CM* ($ALPlus | $HL); 313 $GL $CM* ($HY | $HH) $CM* ($ALPlus | $HL ); 314 # Non-breaking CB from LB8a: 315 $CB $CM* $ZWJ ($HY | $HH) $CM* ($ALPlus | $HL ); 316 # Non-breaking SP from LB14: 317 $OP $CM* $SP+ ($HY | $HH) $CM* ($ALPlus | $HL ); 318 # Non-breaking SP from LB15a: 319 ($OP $CM* $SP+ | [$OP $QU $GL] $CM*) ([\p{Pi} & $QU] $CM* $SP*)+ $SP ($HY | $HH) $CM* ($ALPlus | $HL ); 320 ^([\p{Pi} & $QU] $CM* $SP*)+ $SP ($HY | $HH) $CM* ($ALPlus | $HL ); 321 # Non-breaking SP from LB15a following LB15b: 322 $LB8NonBreaks [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP ($HY | $HH) $CM* ($ALPlus | $HL ); 323 $CAN_CM $CM* [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP ($HY | $HH) $CM* ($ALPlus | $HL ); 324 ^$CM+ [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP ($HY | $HH) $CM* ($ALPlus | $HL ); 325 326 # LB 21 x (BA | HH | HY | NS) 327 # BB x 328 # 329 $LB20NonBreaks $CM* ($BA | $HH | $HY | $NS); 330 331 332 ^$CM+ ($BA | $HH | $HY | $NS); 333 334 $BB $CM* [^$CB]; # $BB x 335 $BB $CM* $LB20NonBreaks; 336 337 # LB 21a Do not break after the hyphen in Hebrew + Hyphen + non-Hebrew 338 # HL (HY | HH) x [^HL] 339 # 340 $HL $CM* ($HY | $HH) $CM* [^$CB $HL]?; 341 342 # LB 21b (forward) Don't break between SY and HL 343 # (break between HL and SY already disallowed by LB 13 above) 344 $SY $CM* $HL; 345 346 # LB 22 Do not break before ellipses 347 # 348 $LB20NonBreaks $CM* $IN; 349 ^$CM+ $IN; 350 351 352 # LB 23 353 # 354 ($ALPlus | $HL) $CM* $NU; 355 ^$CM+ $NU; # Rule 10, any otherwise unattached CM behaves as AL 356 $NU $CM* ($ALPlus | $HL); 357 358 # LB 23a 359 # 360 $PR $CM* ($ID | $EB | $EM); 361 ($ID | $EB | $EM) $CM* $PO; 362 363 364 # 365 # LB 24 366 # 367 ($PR | $PO) $CM* ($ALPlus | $HL); 368 ($ALPlus | $HL) $CM* ($PR | $PO); 369 ^$CM+ ($PR | $PO); # Rule 10, any otherwise unattached CM behaves as AL 370 371 # 372 # LB 25 Numbers. 373 # 374 (($PR | $PO) $CM*)? (($OP | $HY) $CM*)? ($IS $CM*)? $NU ($CM* ($NU | $SY | $IS))* 375 ($CM* ($CL | $CP))? ($CM* ($PR | $PO))?; 376 377 # LB 26 Do not break a Korean syllable 378 # 379 $JL $CM* ($JL | $JV | $H2 | $H3); 380 ($JV | $H2) $CM* ($JV | $JT); 381 ($JT | $H3) $CM* $JT; 382 383 # LB 27 Treat korean Syllable Block the same as ID (don't break it) 384 ($JL | $JV | $JT | $H2 | $H3) $CM* $PO; 385 $PR $CM* ($JL | $JV | $JT | $H2 | $H3); 386 387 388 # LB 28 Do not break between alphabetics 389 # 390 ($ALPlus | $HL) $CM* ($ALPlus | $HL); 391 ^$CM+ ($ALPlus | $HL); # The $CM+ is from rule 10, an unattached CM is treated as AL 392 393 #LB 28a Do not break Orthographic syllables 394 ($AP $CM*)? ($AS | $AK | [◌] ) ($CM* $VI $CM* ($AK | [◌] ))* ($CM* $VI | (($CM* ($AS | $AK | [◌] ) )? $CM* $VF))?; 395 396 # LB 29 397 $IS $CM* ($ALPlus | $HL); 398 399 # LB 30 400 ($ALPlus | $HL | $NU) $CM* [$OP - $EastAsian]; 401 ^$CM+ [$OP - $EastAsian]; # The $CM+ is from rule 10, an unattached CM is treated as AL. 402 [$CP - $EastAsian] $CM* ($ALPlus | $HL | $NU); 403 404 # LB 30a Do not break between regional indicators. Break after pairs of them. 405 # Tricky interaction with LB8a: ZWJ x . together with ZWJ acting like a CM. 406 $RI $CM* $RI / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HH $HY $NS $IN $CM]]; 407 $RI $CM* $RI $CM* [$CM-$ZWJ] / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HH $HY $NS $IN $CM]]; 408 $RI $CM* $RI $CM* [$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HH $HY $NS $IN $ZWJ {eof}]; 409 # note: the preceding rule includes {eof} rather than having the last [set] term qualified with '?' 410 # because of the chain-out behavior difference. The rule must chain out only from the [set characters], 411 # not from the preceding $RI or $CM, which it would be able to do if the set were optional. 412 413 # LB30b Do not break between an emoji base (or potential emoji) and an emoji modifier. 414 $EB $CM* $EM; 415 $ExtPictUnassigned $CM* $EM; 416 417 # LB 31 Break everywhere else. 418 # Match a single code point if no other rule applies. 419 .;