icuSources/test/testdata/break_rules/line_loose.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_loose.txt
   7 #
   8 # Reference Line Break rules for intltest rbbi/RBBIMonkeyTest
   9 #
  10 # Note: Rule syntax and the monkey test itself are still a work in progress.
  11 #       They are expected to change with review and the addition of support for rule tailoring.
  12 #
  13 #         This tailors the line break behavior to correspond to CSS
  14 #         line-break=loose (BCP47 -u-lb-loose) as defined for languages other than
  15 #         Chinese & Japanese.
  16 #         It sets characters of class CJ to behave like ID.
  17 #         In addition, it allows breaks:
  18 #         * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS)
  19 #         * between characters of LineBreak class IN
  20
  21 type = line;
  22 locale = en@lb=loose;
  23
  24
  25 AI = [:LineBreak =  Ambiguous:];
  26 AL = [:LineBreak =  Alphabetic:];
  27 BA = [:LineBreak =  Break_After:];
  28 BB = [:LineBreak =  Break_Before:];
  29 BK = [:LineBreak =  Mandatory_Break:];
  30 B2 = [:LineBreak =  Break_Both:];
  31 CB = [:LineBreak =  Contingent_Break:];
  32 CJ = [:LineBreak =  Conditional_Japanese_Starter:];
  33 CL = [:LineBreak =  Close_Punctuation:];
  34 CM = [:LineBreak =  Combining_Mark:];
  35 CP = [:LineBreak =  Close_Parenthesis:];
  36 CR = [:LineBreak =  Carriage_Return:];
  37 EB = [[:LineBreak =  EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
  38 EM = [:LineBreak =  EM:];
  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];  # CSS Normal tailoring: CJ resolves to ID
  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 ZWJ = [:LineBreak =  ZWJ:];
  69
  70 EmojiNRK = [[\p{Emoji}] - [[RI]\u002a\u00230-9©®™〰〽]];
  71 # Data for Extended Pictographic scraped from CLDR common/properties/ExtendedPictographic.txt, r13267
  72 Extended_Pict = [\U0001F774-\U0001F77F\U00002700-\U00002701\U00002703-\U00002704\U0000270E\U00002710-\U00002711\U00002765-\U00002767\U0001F030-\U0001F093\U0001F094-\U0001F09F\U0001F10D-\U0001F10F\U0001F12F\U0001F16C-\U0001F16F\U0001F1AD-\U0001F1E5\U0001F260-\U0001F265\U0001F203-\U0001F20F\U0001F23C-\U0001F23F\U0001F249-\U0001F24F\U0001F252-\U0001F25F\U0001F266-\U0001F2FF\U0001F7D5-\U0001F7FF\U0001F000-\U0001F003\U0001F005-\U0001F02B\U0001F02C-\U0001F02F\U0001F322-\U0001F323\U0001F394-\U0001F395\U0001F398\U0001F39C-\U0001F39D\U0001F3F1-\U0001F3F2\U0001F3F6\U0001F4FE\U0001F53E-\U0001F548\U0001F54F\U0001F568-\U0001F56E\U0001F571-\U0001F572\U0001F57B-\U0001F586\U0001F588-\U0001F589\U0001F58E-\U0001F58F\U0001F591-\U0001F594\U0001F597-\U0001F5A3\U0001F5A6-\U0001F5A7\U0001F5A9-\U0001F5B0\U0001F5B3-\U0001F5BB\U0001F5BD-\U0001F5C1\U0001F5C5-\U0001F5D0\U0001F5D4-\U0001F5DB\U0001F5DF-\U0001F5E0\U0001F5E2\U0001F5E4-\U0001F5E7\U0001F5E9-\U0001F5EE\U0001F5F0-\U0001F5F2\U0001F5F4-\U0001F5F9\U00002605\U00002607-\U0000260D\U0000260F-\U00002610\U00002612\U00002616-\U00002617\U00002619-\U0000261C\U0000261E-\U0000261F\U00002621\U00002624-\U00002625\U00002627-\U00002629\U0000262B-\U0000262D\U00002630-\U00002637\U0000263B-\U00002647\U00002654-\U0000265F\U00002661-\U00002662\U00002664\U00002667\U00002669-\U0000267A\U0000267C-\U0000267E\U00002680-\U00002691\U00002695\U00002698\U0000269A\U0000269D-\U0000269F\U000026A2-\U000026A9\U000026AC-\U000026AF\U000026B2-\U000026BC\U000026BF-\U000026C3\U000026C6-\U000026C7\U000026C9-\U000026CD\U000026D0\U000026D2\U000026D5-\U000026E8\U000026EB-\U000026EF\U000026F6\U000026FB-\U000026FC\U000026FE-\U000026FF\U00002388\U0001FA00-\U0001FFFD\U0001F0A0-\U0001F0AE\U0001F0B1-\U0001F0BF\U0001F0C1-\U0001F0CF\U0001F0D1-\U0001F0F5\U0001F0AF-\U0001F0B0\U0001F0C0\U0001F0D0\U0001F0F6-\U0001F0FF\U0001F80C-\U0001F80F\U0001F848-\U0001F84F\U0001F85A-\U0001F85F\U0001F888-\U0001F88F\U0001F8AE-\U0001F8FF\U0001F900-\U0001F90B\U0001F91F\U0001F928-\U0001F92F\U0001F931-\U0001F932\U0001F94C\U0001F95F-\U0001F96B\U0001F992-\U0001F997\U0001F9D0-\U0001F9E6\U0001F90C-\U0001F90F\U0001F93F\U0001F94D-\U0001F94F\U0001F96C-\U0001F97F\U0001F998-\U0001F9BF\U0001F9C1-\U0001F9CF\U0001F9E7-\U0001F9FF\U0001F6C6-\U0001F6CA\U0001F6D3-\U0001F6D4\U0001F6E6-\U0001F6E8\U0001F6EA\U0001F6F1-\U0001F6F2\U0001F6F7-\U0001F6F8\U0001F6D5-\U0001F6DF\U0001F6ED-\U0001F6EF\U0001F6F9-\U0001F6FF];
  73
  74 # LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes
  75 AL = [AL AI SG XX ];
  76 dictionary = SA;
  77
  78 # By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
  79 #         list it in the numerous rules that use CM.
  80 CM = [CM ZWJ];
  81
  82 LB4:        BK ÷;
  83 LB5:        CR LF;
  84 LB5.1:      CR ÷;
  85 LB5.2:      LF ÷;
  86 LB5.3:      NL ÷;
  87
  88 LB6:        . (BK | CR | LF | NL);
  89 LB6.1:      [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL);
  90
  91 # Rules LB14 - LB17.
  92 # Moved before LB7, because they can match a longer sequence that would also match LB7,
  93 # for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
  94 #                           "while only the prefix "OP CM SP" matches LB7.1
  95 LB14:        OP CM* SP* .;
  96 LB15:        QU CM* SP* OP;
  97 LB16:        (CL | CP)CM* SP* NS;
  98 LB17:        B2 CM* SP* B2;
  99
 100 LB7.1:      [^ZW SP] CM* [SP ZW];
 101 LB7.2:      [ZW SP] [SP ZW];
 102
 103 # LB8, ICU differs from UAX-14,
 104 #    ICU:    ZW ÷;
 105 #    UAX 14: ZW SP* ÷;
 106 LB8:        ZW ÷;
 107
 108 # LB8a
 109 #      ZWJ x (ID | Extended_Pict | EmojiNRK)
 110 LB8a:       ZWJ (ID | Extended_Pict | EmojiNRK);
 111
 112
 113 # LB9:  X CM -> X
 114 # LB10: Unattached CM -> AL
 115
 116 #LB11:       × WJ;
 117 #            WJ ×
 118
 119 LB11.1:      [^BK CR LF NL SP ZW] CM* WJ;
 120 LB11.2:      SP WJ;
 121 LB11.3:      WJ CM* [^CM];
 122
 123 LB12:        GL CM* [^CM];
 124
 125 LB12a:       [^SP BA HY] CM* GL;
 126
 127 # LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
 128 #
 129 #   LB13.1   [^SP] CM* [CL CP EX IS SY]    # original UAX 14 rule.
 130 #   LB13.2   SP    CM* [CL CP EX IS SY]
 131
 132 LB13.1: [^NU SP] CM* [CL CP IS SY];
 133 LB13.2: [^SP] CM* EX;
 134 LB13.2: SP [CL CP EX IS SY];
 135
 136
 137 # LB 14-17 are moved above LB 7.
 138
 139 LB18:        SP ÷;
 140
 141 LB19:        . CM* QU;
 142 LB19.1:      QU CM* [^CM];
 143
 144 # LB 20   Break before and after CB.
 145 #         Interaction with LB8a:  ZWJ x ID is tricky because CM includes ZWJ.
 146 #                                 ZWJ acts like a CM to the left, combining with CB.
 147 #                                 ZWJ acts independently to the right, no break from ID by LB8a.
 148 LB20:        . CM* ÷ CB;
 149 LB20.1a:     CB CM* ZWJ (ID | Extended_Pict | EmojiNRK);
 150 LB20.1b:      CB CM* ÷;
 151
 152 # Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
 153 #       not picking up the continuing match after the BA from 21a.
 154 LB21a:       HL CM* (HY | BA) CM* [^CM CB];
 155
 156 LB21.1:      . CM* [BA HY NS];
 157 LB21.2:      BB CM* [^CM CB];
 158
 159 LB21b:       SY CM* HL;
 160
 161 LB22.1:        (AL | HL | CM) CM* IN;   # The CM is from LB10, treat an unattached CM as AL.
 162 LB22.2:       EX CM* IN;
 163 LB22.3:       (ID | EB | EM) CM* IN;
 164 # LB22.4:       IN CM* IN;  # delete this rule for CSS loose.
 165 LB22.5:       NU CM* IN;
 166
 167 LB23.1:      (AL | HL | CM) CM* NU;
 168 LB23.2:      NU CM* (AL | HL);
 169
 170 LB23a.1:     PR CM* (ID | EB | EM);
 171 LB23a.2:     (ID | EB | EM) CM* PO;
 172
 173 LB24.2:      (PR | PO) CM* (AL | HL);
 174 LB24.3:      (AL | HL | CM) CM* (PR | PO);
 175
 176 # Numbers. Equivalent to Tailoring example 8 from UAx 14.
 177 LB25:        ((PR | PO)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?;
 178
 179 LB26.1:      JL CM* (JL | JV | H2 | H3);
 180 LB26.2:      (JV | H2) CM* (JV | JT);
 181 LB26.3:      (JT | H3) CM* JT;
 182
 183 LB27.1:      (JL | JV | JT | H2 | H3) CM* IN;
 184 LB27.2:      (JL | JV | JT | H2 | H3) CM* PO;
 185 LB27.3:      PR CM* (JL | JV | JT | H2 | H3);
 186
 187 # LB28 Do not break between Alphabetics.
 188 #      Unattached (leading) CM treated as AL.
 189 LB28:        (AL | HL | CM)CM* (AL | HL);
 190
 191 LB29:        IS CM* (AL | HL);
 192
 193 # LB30  is adjusted for unattached leading CM being treated as AL.
 194 LB30.1:      (AL | CM | HL | NU) CM* OP;
 195 LB30.2:      CP CM* (AL | HL | NU);
 196
 197 # LB31  keep pairs of RI together.
 198 LB30a.1:     RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS];
 199 LB30a.2:     RI CM* RI CM* ZWJ (ID | Extended_Pict | EmojiNRK);
 200 LB30a.3:     RI CM* RI CM* ÷;
 201
 202 # LB30b Do not break between Emoji Base and Emoji Modifier
 203 LB30b:       EB CM* EM;
 204
 205 # LB31 Break Everywhere Else.
 206 #      Include combining marks
 207 LB31.1:        . CM* ZWJ (ID | Extended_Pict | EmojiNRK);
 208 LB31.2:        . CM* ÷;