icuSources/test/testdata/break_rules/line_loose.txt

   1 #
   2 # Copyright (c) 2016, International Business Machines Corporation and others. All Rights Reserved.
   3 #
   4 #  file:  line_loose.txt
   5 #
   6 # Reference Line Break rules for intltest rbbi/RBBIMonkeyTest
   7 #
   8 # Note: Rule syntax and the monkey test itself are still a work in progress.
   9 #       They are expected to change with review and the addition of support for rule tailoring.
  10 #
  11 #         This tailors the line break behavior to correspond to CSS
  12 #         line-break=loose (BCP47 -u-lb-loose) as defined for languages other than
  13 #         Chinese & Japanese.
  14 #         It sets characters of class CJ to behave like ID.
  15 #         In addition, it allows breaks:
  16 #         * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS)
  17 #         * between characters of LineBreak class IN
  18
  19 type = line;
  20 locale = en@lb=loose;
  21
  22
  23 AI = [[:LineBreak =  Ambiguous:] - [\u2640\u2642]];
  24 AL = [[:LineBreak =  Alphabetic:]-[\u2764]];
  25 BA = [:LineBreak =  Break_After:];
  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:];
  32 CM = [[:LineBreak =  Combining_Mark:] \u200D];
  33 CP = [:LineBreak =  Close_Parenthesis:];
  34 CR = [:LineBreak =  Carriage_Return:];
  35
  36 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];
  37 EM = [\U0001F3FB-\U0001F3FF];
  38
  39 EX = [:LineBreak =  Exclamation:];
  40 GL = [:LineBreak =  Glue:];
  41 HL = [:LineBreak =  Hebrew_Letter:];
  42 HY = [:LineBreak =  Hyphen:];
  43 H2 = [:LineBreak =  H2:];
  44 H3 = [:LineBreak =  H3:];
  45 ID = [[:LineBreak =  Ideographic:] CJ [\u2640\u2642\u2764]];
  46 IN = [:LineBreak =  Inseperable:];
  47 IS = [:LineBreak =  Infix_Numeric:];
  48 JL = [:LineBreak =  JL:];
  49 JV = [:LineBreak =  JV:];
  50 JT = [:LineBreak =  JT:];
  51 LF = [:LineBreak =  Line_Feed:];
  52 NL = [:LineBreak =  Next_Line:];
  53 NSX = [\u3005 \u303B \u309D \u309E \u30FD \u30FE];
  54 NS = [[:LineBreak =  Nonstarter:] - NSX];
  55 NU = [:LineBreak =  Numeric:];
  56 OP = [:LineBreak =  Open_Punctuation:];
  57 PO = [:LineBreak =  Postfix_Numeric:];
  58 PR = [:LineBreak =  Prefix_Numeric:];
  59 QU = [:LineBreak =  Quotation:];
  60 RI = [:LineBreak =  Regional_Indicator:];
  61 SA = [:LineBreak =  Complex_Context:];
  62 SG = [:LineBreak =  Surrogate:];
  63 SP = [:LineBreak =  Space:];
  64 SY = [:LineBreak =  Break_Symbols:];
  65 WJ = [:LineBreak =  Word_Joiner:];
  66 XX = [:LineBreak =  Unknown:];
  67 ZW = [:LineBreak =  ZWSpace:];
  68 ZJ = [\u200D];
  69
  70 # TODO: adjustment to sets needed only until Unicode properties are updated for Emoji.
  71 ID   = [ID - EB];
  72 AL   = [AL - EM];
  73
  74 dictionary = [:LineBreak = Complex_Context:];
  75
  76 # Redfine AL. LB1.  TODO: refine according to latest UAX.
  77 AL = [ AL AI SA SG XX ];
  78
  79 LB4:        BK ÷;
  80 LB5:        CR LF;
  81 LB5.1:      CR ÷;
  82 LB5.2:      LF ÷;
  83 LB5.3:      NL ÷;
  84
  85 LB6:        . (BK | CR | LF | NL);
  86 LB6.1:      [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL);
  87
  88 # Rules LB14 - LB17.
  89 # Moved before LB7, because they can match a longer sequence that would also match LB7,
  90 # for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
  91 #                           "while only the prefix "OP CM SP" matches LB7.1
  92 LB14:        OP CM* SP* .;
  93 LB15:        QU CM* SP* OP;
  94 LB16:        (CL | CP)CM* SP* NS;
  95 LB17:        B2 CM* SP* B2;
  96
  97 LB7.1:      [^ZW SP] CM* [SP ZW];
  98 LB7.2:      [ZW SP] [SP ZW];
  99
 100 # LB8, ICU differs from UAX-14,
 101 #    ICU:    ZW ÷;
 102 #    UAX 14: ZW SP* ÷;
 103 LB8:        ZW ÷;
 104
 105 # LB8a, from Emoji proposal L2/16-011R3
 106 #      ZWJ x ID
 107 LB8a:       ZJ (ID | EB | EM);
 108
 109
 110 # LB9:  X CM -> X
 111 # LB10: Unattached CM -> AL
 112
 113 #LB11:       × WJ;
 114 #            WJ ×
 115
 116 LB11.1:      [^BK CR LF NL SP ZW] CM* WJ;
 117 LB11.2:      SP WJ;
 118 LB11.3:      WJ CM* [^CM];
 119
 120 LB12:        GL CM* [^CM];
 121
 122 LB12a:       [^SP BA HY] CM* GL;
 123
 124 # LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
 125 #
 126 #   LB13.1   [^SP] CM* [CL CP EX IS SY]    # original UAX 14 rule.
 127 #   LB13.2   SP    CM* [CL CP EX IS SY]
 128
 129 LB13.1: [^NU SP] CM* [CL CP IS SY];
 130 LB13.2: [^SP] CM* EX;
 131 LB13.2: SP [CL CP EX IS SY];
 132
 133
 134 # LB 14-17 are moved above LB 7.
 135
 136 LB18:        SP ÷;
 137
 138 LB19:        . CM* QU;
 139 LB19.1:      QU CM* [^CM];
 140
 141 # LB 20   Break before and after CB.
 142 #         Interaction with LB8a:  ZJ x ID is tricky because CM includes ZJ.
 143 #                                 ZJ acts like a CM to the left, combining with CB.
 144 #                                 ZJ acts independently to the right, no break from ID by LB8a.
 145 LB20:        . CM* ÷ CB;
 146 LB20.1a:     CB CM* ZJ (ID | EB | EM);
 147 LB20.1b:      CB CM* ÷;
 148
 149 # Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
 150 #       not picking up the continuing match after the BA from 21a.
 151 LB21a:       HL CM* (HY | BA) CM* [^CM CB];
 152
 153 LB21.1:      . CM* [BA HY NS];
 154 LB21.2:      BB CM* [^CM CB];
 155
 156 LB21b:       SY CM* HL;
 157
 158 LB22.1:        (AL | HL | CM) CM* IN;   # The CM is from LB10, treat an unattached CM as AL.
 159 LB22.2:       EX CM* IN;
 160 LB22.3:       (ID | EB | EM) CM* IN;
 161 # LB22.4:       IN CM* IN;  # delete this rule for CSS loose.
 162 LB22.5:       NU CM* IN;
 163
 164 LB23.1:      (ID | EB | EM) CM* PO;
 165 LB23.2:      (AL | HL | CM) CM* NU;
 166 LB23.3:      NU CM* (AL | HL);
 167
 168 LB24.1:      PR CM* (ID | EB | EM);
 169 LB24.2:      PR CM* (AL | HL);
 170 LB24.3:      PO CM* (AL | HL);
 171 LB24.4:      (AL | HL | CM) CM* (PR | PO);
 172
 173 # Numbers. Equivalent to Tailoring example 8 from UAx 14.
 174 LB25:        ((PR | PO)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?;
 175
 176 LB26.1:      JL CM* (JL | JV | H2 | H3);
 177 LB26.2:      (JV | H2) CM* (JV | JT);
 178 LB26.3:      (JT | H3) CM* JT;
 179
 180 LB27.1:      (JL | JV | JT | H2 | H3) CM* IN;
 181 LB27.2:      (JL | JV | JT | H2 | H3) CM* PO;
 182 LB27.3:      PR CM* (JL | JV | JT | H2 | H3);
 183
 184 # LB28 Do not break between Alphabetics.
 185 #      Unattached (leading) CM treated as AL.
 186 LB28:        (AL | HL | CM)CM* (AL | HL);
 187
 188 LB29:        IS CM* (AL | HL);
 189
 190 # LB30  is adjusted for unattached leading CM being treated as AL.
 191 LB30.1:      (AL | CM | HL | NU) CM* OP;
 192 LB30.2:      CP CM* (AL | HL | NU);
 193
 194 # LB31  keep pairs of RI together.
 195 LB30a.1:     RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS];
 196 LB30a.2:     RI CM* RI CM* ZJ (ID | EB | EM);
 197 LB30a.3:     RI CM* RI CM* ÷;
 198
 199 # LB30b Do not break between Emoji Base and Emoji Modifier
 200 LB30b:       EB CM* EM;
 201
 202 # LB31 Break Everywhere Else.
 203 #      Include combining marks
 204 LB31.1:        . CM* ZJ (ID | EB | EM);
 205 LB31.2:        . CM* ÷;