icuSources/test/testdata/break_rules/line_normal.txt

   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_normal.txt
   7 #
   8 # Reference Line Break rules for intltest rbbi/RBBIMonkeyTest.
   9 # Rules derived from Unicode Standard Annex #14 Revision 40 for Unicode 11.0
  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 #         Line Breaking Rules
  15 #         Implement default line breaking as defined by
  16 #         Unicode Standard Annex #14 Revision 34 for Unicode 8.0
  17 #         http://www.unicode.org/reports/tr14/
  18 #         tailored as noted in 2nd paragraph below.
  19 #
  20 #         TODO:  Rule LB 8 remains as it was in Unicode 5.2
  21 #         This is only because of a limitation of ICU break engine implementation,
  22 #         not because the older behavior is desirable.
  23 #
  24 #         This tailors the line break behavior to correspond to CSS
  25 #         line-break=normal (BCP47 -u-lb-normal) as defined for languages other than
  26 #         Chinese & Japanese.
  27 #         It sets characters of class CJ to behave like ID.
  28
  29
  30 type = line;
  31 locale = en@lb=normal;
  32
  33 AI = [:LineBreak =  Ambiguous:];
  34 AL = [:LineBreak =  Alphabetic:];
  35 BA = [:LineBreak =  Break_After:];
  36 BB = [:LineBreak =  Break_Before:];
  37 BK = [:LineBreak =  Mandatory_Break:];
  38 B2 = [:LineBreak =  Break_Both:];
  39 CB = [:LineBreak =  Contingent_Break:];
  40 CJ = [:LineBreak =  Conditional_Japanese_Starter:];
  41 CL = [:LineBreak =  Close_Punctuation:];
  42 CM = [:LineBreak =  Combining_Mark:];
  43 CP = [:LineBreak =  Close_Parenthesis:];
  44 CR = [:LineBreak =  Carriage_Return:];
  45 EB = [[:LineBreak =  EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
  46 EM = [:LineBreak =  EM:];
  47 EX = [:LineBreak =  Exclamation:];
  48 GL = [:LineBreak =  Glue:];
  49 HL = [:LineBreak =  Hebrew_Letter:];
  50 HY = [:LineBreak =  Hyphen:];
  51 H2 = [:LineBreak =  H2:];
  52 H3 = [:LineBreak =  H3:];
  53 ID = [[:LineBreak =  Ideographic:] CJ];  # CSS Normal tailoring: CJ resolves to ID
  54 IN = [:LineBreak =  Inseperable:];
  55 IS = [:LineBreak =  Infix_Numeric:];
  56 JL = [:LineBreak =  JL:];
  57 JV = [:LineBreak =  JV:];
  58 JT = [:LineBreak =  JT:];
  59 LF = [:LineBreak =  Line_Feed:];
  60 NL = [:LineBreak =  Next_Line:];
  61 NS = [:LineBreak =  Nonstarter:];
  62 NU = [:LineBreak =  Numeric:];
  63 OP = [:LineBreak =  Open_Punctuation:];
  64 PO = [:LineBreak =  Postfix_Numeric:];
  65 PR = [:LineBreak =  Prefix_Numeric:];
  66 QU = [:LineBreak =  Quotation:];
  67 RI = [:LineBreak =  Regional_Indicator:];
  68 SA = [:LineBreak =  Complex_Context:];
  69 SG = [:LineBreak =  Surrogate:];
  70 SP = [:LineBreak =  Space:];
  71 SY = [:LineBreak =  Break_Symbols:];
  72 WJ = [:LineBreak =  Word_Joiner:];
  73 XX = [:LineBreak =  Unknown:];
  74 ZW = [:LineBreak =  ZWSpace:];
  75 ZWJ = [:LineBreak =  ZWJ:];
  76
  77 # LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes
  78 AL = [AL AI SG XX ];
  79 dictionary = SA;
  80
  81 # By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
  82 #         list it in the numerous rules that use CM.
  83 CM = [CM ZWJ];
  84
  85 LB4:        BK ÷;
  86 LB5:        CR LF;
  87 LB5.1:      CR ÷;
  88 LB5.2:      LF ÷;
  89 LB5.3:      NL ÷;
  90
  91 LB6:        . (BK | CR | LF | NL);
  92 LB6.1:      [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL);
  93
  94 # Rules LB14 - LB17.
  95 # Moved before LB7, because they can match a longer sequence that would also match LB7,
  96 # for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
  97 #                           "while only the prefix "OP CM SP" matches LB7.1
  98 LB14:        OP CM* SP* .;
  99 LB15:        QU CM* SP* OP;
 100 LB16:        (CL | CP)CM* SP* NS;
 101 LB17:        B2 CM* SP* B2;
 102
 103 LB7.1:      [^ZW SP] CM* [SP ZW];
 104 LB7.2:      [ZW SP] [SP ZW];
 105
 106 # LB8, ICU differs from UAX-14,
 107 #    ICU:    ZW ÷;
 108 #    UAX 14: ZW SP* ÷;
 109 LB8:        ZW ÷;
 110
 111 # LB8a
 112 #      ZWJ x
 113 #      Don't match a CM on the right - let other rules pick up CM sequences, where
 114 #      the ZWJ behaves as just another generic CM.
 115 LB8a:       ZWJ [^CM];
 116
 117
 118 # LB9:  X CM -> X
 119 # LB10: Unattached CM -> AL
 120
 121 #LB11:       × WJ;
 122 #            WJ ×
 123
 124 LB11.1:      [^SP] CM* WJ;
 125 LB11.2:      SP WJ;
 126 LB11.3:      WJ CM* [^CM];
 127
 128 LB12:        GL CM* [^CM];
 129
 130 LB12a:       [^SP BA HY] CM* GL;
 131
 132 # LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
 133 #
 134 #   LB13.1   [^SP] CM* [CL CP EX IS SY]    # original UAX 14 rule.
 135 #   LB13.2   SP    CM* [CL CP EX IS SY]
 136
 137 LB13.1: [^NU SP] CM* [CL CP IS SY];
 138 LB13.2: [^SP] CM* EX;
 139 LB13.2: SP [CL CP EX IS SY];
 140
 141
 142 # LB 14-17 are moved above LB 7.
 143
 144 LB18:        SP ÷;
 145
 146 LB19:        . CM* QU;
 147 LB19.1:      QU CM* [^CM];
 148
 149 # LB 20   Break before and after CB.
 150 #         Interaction with LB8a:  ZWJ x . is tricky because CM includes ZWJ.
 151 #                                 ZWJ acts like a CM to the left, combining with CB.
 152 #                                 ZWJ acts independently to the right, no break after by LB8a.
 153 LB20.1:      . CM* ZWJ CB;
 154 LB20.2:      . CM* ÷ CB;
 155
 156 LB20.3:      CB CM* ZWJ [^CM];
 157 LB20.4:      CB CM* ÷;
 158
 159 # Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
 160 #       not picking up the continuing match after the BA from 21a.
 161 LB21a:       HL CM* (HY | BA) CM* [^CM CB];
 162
 163 LB21.1:      . CM* [BA HY NS];
 164 LB21.2:      BB CM* [^CM CB];
 165
 166 LB21b:       SY CM* HL;
 167
 168 LB22.1:        (AL | HL | CM) CM* IN;   # The CM is from LB10, treat an unattached CM as AL.
 169 LB22.2:       EX CM* IN;
 170 LB22.3:       (ID | EB | EM) CM* IN;
 171 LB22.4:       IN CM* IN;
 172 LB22.5:       NU CM* IN;
 173
 174 LB23.1:      (AL | HL | CM) CM* NU;
 175 LB23.2:      NU CM* (AL | HL);
 176
 177 LB23a.1:     PR CM* (ID | EB | EM);
 178 LB23a.2:     (ID | EB | EM) CM* PO;
 179
 180 LB24.2:      (PR | PO) CM* (AL | HL);
 181 LB24.3:      (AL | HL | CM) CM* (PR | PO);
 182
 183 # Numbers. Equivalent to Tailoring example 8 from UAx 14.
 184 LB25:        ((PR | PO)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?;
 185
 186 LB26.1:      JL CM* (JL | JV | H2 | H3);
 187 LB26.2:      (JV | H2) CM* (JV | JT);
 188 LB26.3:      (JT | H3) CM* JT;
 189
 190 LB27.1:      (JL | JV | JT | H2 | H3) CM* IN;
 191 LB27.2:      (JL | JV | JT | H2 | H3) CM* PO;
 192 LB27.3:      PR CM* (JL | JV | JT | H2 | H3);
 193
 194 # LB28 Do not break between Alphabetics.
 195 #      Unattached (leading) CM treated as AL.
 196 LB28:        (AL | HL | CM)CM* (AL | HL);
 197
 198 LB29:        IS CM* (AL | HL);
 199
 200 # LB30  is adjusted for unattached leading CM being treated as AL.
 201 LB30.1:      (AL | CM | HL | NU) CM* OP;
 202 LB30.2:      CP CM* (AL | HL | NU);
 203
 204 # LB30a  keep pairs of RI together.
 205 LB30a.1:     RI CM* RI              ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
 206 LB30a.2:     RI CM* RI CM* [CM-ZWJ] ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
 207 LB30a.3:     RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS ZWJ]?;
 208
 209 # LB30b Do not break between Emoji Base and Emoji Modifier
 210 LB30b:       EB CM* EM;
 211
 212 # LB31 Break Everywhere Else.
 213 #      Include combining marks
 214 LB31.1:        . CM* ZWJ [^CM];
 215 LB31.2:        . CM* ÷;