# # Copyright (c) 2016, International Business Machines Corporation and others. All Rights Reserved. # # file: line_loose_cj.txt # # Reference Line Break rules for intltest rbbi/RBBIMonkeyTest # # Note: Rule syntax and the monkey test itself are still a work in progress. # They are expected to change with review and the addition of support for rule tailoring. # # Line Breaking Rules # Implement default line breaking as defined by # Unicode Standard Annex #14 Revision 34 for Unicode 8.0 # http://www.unicode.org/reports/tr14/ # tailored as noted in 2nd paragraph below.. # # This tailors the line break behavior to correspond to CSS # line-break=loose (BCP47 -u-lb-loose) as defined for Chinese & Japanese. # It sets characters of class CJ to behave like ID. # In addition, it allows breaks: # * before hyphens 2010 & 2013 (both BA) and 301C, 30A0 (both NS) # * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS) # * between characters of LineBreak class IN such as 2026 # * before some centered punct 203C, 2047, 2048, 2049, 30FB, FF1A, FF1B, # FF65 (all NS) and FF01, FF1F (both EX). # * before suffix characters with LineBreak class PO and EastAsianWidth A,F,W; # this includes: 00B0 2030 2032 2033 2035 2103 2109 FE6A FF05 FFE0 # * after prefix characters with LineBreak class PR and EastAsianWidth A,F,W; # this includes: 00A4 00B1 20AC 2116 FE69 FF04 FFE1 FFE5 FFE6 type = line; locale = ja@lb=loose; AI = [:LineBreak = Ambiguous:]; AL = [[:LineBreak = Alphabetic:]-[\u2764]]; BAX = [\u2010 \u2013]; BA = [[:LineBreak = Break_After:] - BAX]; BB = [:LineBreak = Break_Before:]; BK = [:LineBreak = Mandatory_Break:]; B2 = [:LineBreak = Break_Both:]; CB = [:LineBreak = Contingent_Break:]; CJ = [:LineBreak = Conditional_Japanese_Starter:]; CL = [:LineBreak = Close_Punctuation:]; CM = [:LineBreak = Combining_Mark:]; CP = [:LineBreak = Close_Parenthesis:]; CR = [:LineBreak = Carriage_Return:]; EB = [\u261D\u26F9\u270A-\u270D\U0001F385\U0001F3C3-\U0001F3C4\U0001F3CA-\U0001F3CB\U0001F442-\U0001F443\U0001F446-\U0001F450\U0001F466-\U0001F469\U0001F46E\U0001F470-\U0001F478\U0001F47C\U0001F481-\U0001F483\U0001F485-\U0001F487\U0001F4AA\U0001F575\U0001F590\U0001F595-\U0001F596\U0001F645-\U0001F647\U0001F64B-\U0001F64F\U0001F6A3\U0001F6B4-\U0001F6B6\U0001F6C0\U0001F918]; EM = [\U0001F3FB-\U0001F3FF]; EXX = [\uFF01 \uFF1F]; EX = [[:LineBreak = Exclamation:] - EXX]; GL = [:LineBreak = Glue:]; HL = [:LineBreak = Hebrew_Letter:]; HY = [:LineBreak = Hyphen:]; H2 = [:LineBreak = H2:]; H3 = [:LineBreak = H3:]; ID = [[:LineBreak = Ideographic:][\u2764]CJ]; IN = [:LineBreak = Inseperable:]; IS = [:LineBreak = Infix_Numeric:]; JL = [:LineBreak = JL:]; JV = [:LineBreak = JV:]; JT = [:LineBreak = JT:]; LF = [:LineBreak = Line_Feed:]; NL = [:LineBreak = Next_Line:]; NSX = [\u301C \u30A0 \u3005 \u303B \u309D \u309E \u30FD \u30FE \u203C \u2047 \u2048 \u2049 \u30FB \uFF1A \uFF1B \uFF65]; NS = [[:LineBreak = Nonstarter:] - NSX]; NU = [:LineBreak = Numeric:]; OP = [:LineBreak = Open_Punctuation:]; POX = [\u00B0 \u2030 \u2032 \u2033 \u2035 \u2103 \u2109 \uFE6A \uFF05 \uFFE0]; PO = [[:LineBreak = Postfix_Numeric:] - POX]; PRX = [\u00A4 \u00B1 \u20AC \u2116 \uFE69 \uFF04 \uFFE1 \uFFE5 \uFFE6]; PR = [[:LineBreak = Prefix_Numeric:] - PRX]; QU = [:LineBreak = Quotation:]; RI = [:LineBreak = Regional_Indicator:]; SA = [:LineBreak = Complex_Context:]; SG = [:LineBreak = Surrogate:]; SP = [:LineBreak = Space:]; SY = [:LineBreak = Break_Symbols:]; WJ = [:LineBreak = Word_Joiner:]; XX = [:LineBreak = Unknown:]; ZW = [:LineBreak = ZWSpace:]; ZJ = [\u200D]; # TODO: adjustment to sets needed only until Unicode properties are updated for Emoji. ID = [ID - EB]; AL = [AL - EM]; dictionary = [:LineBreak = Complex_Context:]; # Redfine AL. LB1. TODO: refine according to latest UAX. AL = [ AL AI SA SG XX ]; LB4: BK ÷; LB5: CR LF; LB5.1: CR ÷; LB5.2: LF ÷; LB5.3: NL ÷; LB6: . (BK | CR | LF | NL); LB6.1: [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL); # Rules LB14 - LB17. # Moved before LB7, because they can match a longer sequence that would also match LB7, # for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it, # "while only the prefix "OP CM SP" matches LB7.1 LB14: OP CM* SP* .; LB15: QU CM* SP* OP; LB16: (CL | CP)CM* SP* NS; LB17: B2 CM* SP* B2; LB7.1: [^ZW SP] CM* [SP ZW]; LB7.2: [ZW SP] [SP ZW]; # LB8, ICU differs from UAX-14, # ICU: ZW ÷; # UAX 14: ZW SP* ÷; LB8: ZW ÷; # LB8a, from Emoji proposal L2/16-011R3 # ZWJ x ID LB8a: ZJ (ID | EB | EM); # LB9: X CM -> X # LB10: Unattached CM -> AL #LB11: × WJ; # WJ × LB11.1: [^BK CR LF NL SP ZW] CM* WJ; LB11.2: SP WJ; LB11.3: WJ CM* [^CM]; LB12: GL CM* [^CM]; LB12a: [^SP BA BAX HY] CM* GL; # LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14. # # LB13.1 [^SP] CM* [CL CP EX IS SY] # original UAX 14 rule. # LB13.2 SP CM* [CL CP EX IS SY] LB13.1: [^NU SP] CM* [CL CP IS SY]; LB13.2: [^SP] CM* EX; LB13.2: SP [CL CP EX IS SY]; # LB 14-17 are moved above LB 7. LB18: SP ÷; LB19: . CM* QU; LB19.1: QU CM* [^CM]; # LB 20 Break before and after CB. # Interaction with LB8a: ZJ x ID is tricky because CM includes ZJ. # ZJ acts like a CM to the left, combining with CB. # ZJ acts independently to the right, no break from ID by LB8a. LB20: . CM* ÷ CB; LB20.1a: CB CM* ZJ (ID | EB | EM); LB20.1b: CB CM* ÷; # Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then # not picking up the continuing match after the BA from 21a. # LB 21a Don't break after Hebrew + Hyphen # HL (HY | BA) x LB21a: HL CM* (HY | BA | BAX) CM* [^CM CB]?; LB21.1: . CM* [BA HY NS]; LB21.2: BB CM* [^CM CB]; LB21b: SY CM* HL; LB22.1: (AL | HL | CM) CM* IN; # The CM is from LB10, treat an unattached CM as AL. LB22.2: EX CM* IN; LB22.3: (ID | EB | EM) CM* IN; # LB22.4: IN CM* IN; # delete this rule for CSS loose. LB22.5: NU CM* IN; LB23.1: (ID | EB | EM) CM* PO; LB23.2: (AL | HL | CM) CM* NU; LB23.3: NU CM* (AL | HL); LB24.1: PR CM* (ID | EB | EM); LB24.2: PR CM* (AL | HL); LB24.3: (PO | POX) CM* (AL | HL); LB24.4: (AL | HL | CM) CM* (PR | PO | POX); # Numbers. Equivalent to Tailoring example 8 from UAx 14. # Loose_cj tailoring: do not include $PRX at the beginning or $POX at the end. LB25: ((PR | PO | POX)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PRX | PO))?; LB26.1: JL CM* (JL | JV | H2 | H3); LB26.2: (JV | H2) CM* (JV | JT); LB26.3: (JT | H3) CM* JT; LB27.1: (JL | JV | JT | H2 | H3) CM* IN; LB27.2: (JL | JV | JT | H2 | H3) CM* PO; LB27.3: PR CM* (JL | JV | JT | H2 | H3); # LB28 Do not break between Alphabetics. # Unattached (leading) CM treated as AL. LB28: (AL | HL | CM)CM* (AL | HL); LB29: IS CM* (AL | HL); # LB30 is adjusted for unattached leading CM being treated as AL. LB30.1: (AL | CM | HL | NU) CM* OP; LB30.2: CP CM* (AL | HL | NU); # LB31 keep pairs of RI together. LB30a.1: RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS]; LB30a.2: RI CM* RI CM* ZJ (ID | EB | EM); LB30a.3: RI CM* RI CM* ÷; # LB30b Do not break between Emoji Base and Emoji Modifier LB30b: EB CM* EM; # LB31 Break Everywhere Else. # Include combining marks LB31.1: . CM* ZJ (ID | EB | EM); LB31.2: . CM* ÷;