1# Copyright (C) 2016 and later: Unicode, Inc. and others.
2# License & terms of use: http://www.unicode.org/copyright.html
3# Copyright (c) 2002-2016  International Business Machines Corporation and
4# others. All Rights Reserved.
5#
6#  file:  line_loose_cj.txt
7#
8#         Line Breaking Rules
9#         Implement default line breaking as defined by
10#         Unicode Standard Annex #14 Revision 37 for Unicode 9.0
11#         http://www.unicode.org/reports/tr14/
12#
13#         Includes extensions to the handling of emoji ZWJ sequences from
14#         https://goo.gl/cluFCn
15#
16#         tailored as noted in 2nd paragraph below.
17#
18#         TODO:  Rule LB 8 remains as it was in Unicode 5.2
19#         This is only because of a limitation of ICU break engine implementation,
20#         not because the older behavior is desirable.
21#
22#         This tailors the line break behavior to correspond to CSS
23#         line-break=loose (BCP47 -u-lb-loose) as defined for Chinese & Japanese.
24#         It sets characters of class CJ to behave like ID.
25#         In addition, it allows breaks:
26#         * before hyphens 2010 & 2013 (both BA) and 301C, 30A0 (both NS)
27#         * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS)
28#         * between characters of LineBreak class IN such as 2026
29#         * before some centered punct 203C, 2047, 2048, 2049, 30FB, FF1A, FF1B,
30#           FF65 (all NS) and FF01, FF1F (both EX).
31#         * before suffix characters with LineBreak class PO and EastAsianWidth A,F,W;
32#           this includes: 00B0 2030 2032 2033 2035 2103 2109 FE6A FF05 FFE0
33#         * after prefix characters with LineBreak class PR and EastAsianWidth A,F,W;
34#           this includes: 00A4 00B1 20AC 2116 FE69 FF04 FFE1 FFE5 FFE6
35
36
37#
38#  Character Classes defined by TR 14.
39#
40
41!!chain;
42
43$AI = [:LineBreak =  Ambiguous:];
44$AL = [:LineBreak =  Alphabetic:];
45$BAX = [\u2010 \u2013];
46$BA = [[:LineBreak =  Break_After:] - $BAX];
47$BB = [:LineBreak =  Break_Before:];
48$BK = [:LineBreak =  Mandatory_Break:];
49$B2 = [:LineBreak =  Break_Both:];
50$CB = [:LineBreak =  Contingent_Break:];
51$CJ = [:LineBreak =  Conditional_Japanese_Starter:];
52$CL = [:LineBreak =  Close_Punctuation:];
53# $CM = [:LineBreak =  Combining_Mark:];
54$CP = [:LineBreak =  Close_Parenthesis:];
55$CR = [:LineBreak =  Carriage_Return:];
56$EB = [[:LineBreak =  EB:]\U0001F3C2\U0001F3C7\U0001F3CC\U0001F46A-\U0001F46D\U0001F46F\U0001F574\U0001F6CC];
57$EM = [:LineBreak =  EM:];
58$EXX = [\uFF01 \uFF1F];
59$EX = [[:LineBreak =  Exclamation:] - $EXX];
60$GL = [:LineBreak =  Glue:];
61$HL = [:LineBreak =  Hebrew_Letter:];
62$HY = [:LineBreak =  Hyphen:];
63$H2 = [:LineBreak =  H2:];
64$H3 = [:LineBreak =  H3:];
65# CSS Loose tailoring: CJ resolves to ID
66$ID = [[:LineBreak =  Ideographic:] $CJ];
67$IN = [:LineBreak =  Inseperable:];
68$IS = [:LineBreak =  Infix_Numeric:];
69$JL = [:LineBreak =  JL:];
70$JV = [:LineBreak =  JV:];
71$JT = [:LineBreak =  JT:];
72$LF = [:LineBreak =  Line_Feed:];
73$NL = [:LineBreak =  Next_Line:];
74$NSX = [\u301C \u30A0 \u3005 \u303B \u309D \u309E \u30FD \u30FE \u203C \u2047 \u2048 \u2049 \u30FB \uFF1A \uFF1B \uFF65];
75$NS = [[:LineBreak =  Nonstarter:] - $NSX];
76$NU = [:LineBreak =  Numeric:];
77$OP = [:LineBreak =  Open_Punctuation:];
78$POX = [\u00B0 \u2030 \u2032 \u2033 \u2035 \u2103 \u2109 \uFE6A \uFF05 \uFFE0];
79$PO = [[:LineBreak =  Postfix_Numeric:] - $POX];
80$PRX = [\u00A4 \u00B1 \u20AC \u2116 \uFE69 \uFF04 \uFFE1 \uFFE5 \uFFE6];
81$PR = [[:LineBreak =  Prefix_Numeric:] - $PRX];
82$QU = [:LineBreak =  Quotation:];
83$RI = [:LineBreak =  Regional_Indicator:];
84$SA = [:LineBreak =  Complex_Context:];
85$SG = [:LineBreak =  Surrogate:];
86$SP = [:LineBreak =  Space:];
87$SY = [:LineBreak =  Break_Symbols:];
88$WJ = [:LineBreak =  Word_Joiner:];
89$XX = [:LineBreak =  Unknown:];
90$ZW = [:LineBreak =  ZWSpace:];
91$ZWJ = [:LineBreak = ZWJ:];
92
93$EmojiNRK = [[\p{Emoji}] - [$RI \u002a\u00230-9©®™〰〽]];
94# Data for Extended Pictographic scraped from CLDR common/properties/ExtendedPictographic.txt, r12773
95$Extended_Pict = [\U0001F774-\U0001F77F\u2700-\u2701\u2703-\u2704\u270E\u2710-\u2711\u2765-\u2767\U0001F030-\U0001F093\U0001F094-\U0001F09F\U0001F10D-\U0001F10F\U0001F12F\U0001F16C-\U0001F16F\U0001F1AD-\U0001F1E5\U0001F203-\U0001F20F\U0001F23C-\U0001F23F\U0001F249-\U0001F24F\U0001F252-\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\u2605\u2607-\u260D\u260F-\u2610\u2612\u2616-\u2617\u2619-\u261C\u261E-\u261F\u2621\u2624-\u2625\u2627-\u2629\u262B-\u262D\u2630-\u2637\u263B-\u2647\u2654-\u265F\u2661-\u2662\u2664\u2667\u2669-\u267A\u267C-\u267E\u2680-\u2691\u2695\u2698\u269A\u269D-\u269F\u26A2-\u26A9\u26AC-\u26AF\u26B2-\u26BC\u26BF-\u26C3\u26C6-\u26C7\u26C9-\u26CD\u26D0\u26D2\u26D5-\u26E8\u26EB-\u26EF\u26F6\u26FB-\u26FC\u26FE-\u26FF\u2388\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-\U0001F90F\U0001F91F\U0001F928-\U0001F92F\U0001F931-\U0001F932\U0001F93F\U0001F94C-\U0001F94F\U0001F95F-\U0001F97F\U0001F992-\U0001F9BF\U0001F9C1-\U0001F9FF\U0001F6C6-\U0001F6CA\U0001F6E6-\U0001F6E8\U0001F6EA\U0001F6F1-\U0001F6F2\U0001F6D3-\U0001F6DF\U0001F6ED-\U0001F6EF\U0001F6F7-\U0001F6FF];
96
97# By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
98#         list it in the numerous rules that use CM.
99# By LB1, SA characters with general categor of Mn or Mc also resolve to CM.
100
101$CM = [[:LineBreak = Combining_Mark:] $ZWJ [$SA & [[:Mn:][:Mc:]]]];
102
103#   Dictionary character set, for triggering language-based break engines. Currently
104#   limited to LineBreak=Complex_Context (SA).
105
106$dictionary = [$SA];
107
108#
109#  Rule LB1.  By default, treat AI  (characters with ambiguous east Asian width),
110#                               SA  (Dictionary chars, excluding Mn and Mc)
111#                               SG  (Unpaired Surrogates)
112#                               XX  (Unknown, unassigned)
113#                         as $AL  (Alphabetic)
114#
115$ALPlus = [$AL $AI $SG $XX [$SA-[[:Mn:][:Mc:]]]];
116
117
118## -------------------------------------------------
119
120!!forward;
121
122#
123# CAN_CM  is the set of characters that may combine with CM combining chars.
124#         Note that Linebreak UAX 14's concept of a combining char and the rules
125#         for what they can combine with are _very_ different from the rest of Unicode.
126#
127#         Note that $CM itself is left out of this set.  If CM is needed as a base
128#         it must be listed separately in the rule.
129#
130$CAN_CM  = [^$SP $BK $CR $LF $NL $ZW $CM];       # Bases that can   take CMs
131$CANT_CM = [ $SP $BK $CR $LF $NL $ZW $CM];       # Bases that can't take CMs
132
133#
134# AL_FOLLOW  set of chars that can unconditionally follow an AL
135#            Needed in rules where stand-alone $CM s are treated as AL.
136#
137$AL_FOLLOW      = [$BK $CR $LF $NL $ZW $SP $CL $CP $EX $HL $IS $SY $WJ $GL $OP $QU $BA $HY $NS $IN $NU $PR $PO $POX $ALPlus];
138
139
140#
141#  Rule LB 4, 5    Mandatory (Hard) breaks.
142#
143$LB4Breaks    = [$BK $CR $LF $NL];
144$LB4NonBreaks = [^$BK $CR $LF $NL $CM];
145$CR $LF {100};
146
147#
148#  LB 6    Do not break before hard line breaks.
149#
150$LB4NonBreaks?  $LB4Breaks {100};    # LB 5  do not break before hard breaks.
151$CAN_CM $CM*    $LB4Breaks {100};
152^$CM+           $LB4Breaks {100};
153
154# LB 7         x SP
155#              x ZW
156$LB4NonBreaks [$SP $ZW];
157$CAN_CM $CM*  [$SP $ZW];
158^$CM+         [$SP $ZW];
159
160#
161# LB 8         Break after zero width space
162#              TODO:  ZW SP* <break>
163#              An engine change is required to write the reverse rule for this.
164#              For now, leave the Unicode 5.2 rule, ZW <break>
165#
166$LB8Breaks    = [$LB4Breaks $ZW];
167$LB8NonBreaks = [[$LB4NonBreaks] - [$ZW]];
168
169# LB 8a        ZWJ x (ID | Extended_Pict | EmojiNRK)   Emoji ZWJ sequences.
170#
171$ZWJ ($ID | $Extended_Pict | $EmojiNRK);
172
173# LB 9     Combining marks.      X   $CM needs to behave like X, where X is not $SP, $BK $CR $LF $NL
174#                                $CM not covered by the above needs to behave like $AL
175#                                See definition of $CAN_CM.
176
177$CAN_CM $CM+;                   #  Stick together any combining sequences that don't match other rules.
178^$CM+;
179
180#
181# LB 11  Do not break before or after WORD JOINER & related characters.
182#
183$CAN_CM $CM*  $WJ;
184$LB8NonBreaks $WJ;
185^$CM+         $WJ;
186
187$WJ $CM* .;
188
189#
190# LB 12  Do not break after NBSP and related characters.
191#         GL  x
192#
193$GL $CM* .;
194
195#
196# LB 12a  Do not break before NBSP and related characters ...
197#            [^SP BA HY] x GL
198#
199[[$LB8NonBreaks] - [$SP $BA $BAX $HY]] $CM* $GL;
200^$CM+ $GL;
201
202
203#
204# LB 13   Don't break before ']' or '!' or ';' or '/', even after spaces.
205#
206# Do not include $EXX here
207$LB8NonBreaks $CL;
208$CAN_CM $CM*  $CL;
209^$CM+         $CL;              # by rule 10, stand-alone CM behaves as AL
210
211$LB8NonBreaks $CP;
212$CAN_CM $CM*  $CP;
213^$CM+         $CP;              # by rule 10, stand-alone CM behaves as AL
214
215$LB8NonBreaks $EX;
216$CAN_CM $CM*  $EX;
217^$CM+         $EX;              # by rule 10, stand-alone CM behaves as AL
218
219$LB8NonBreaks $IS;
220$CAN_CM $CM*  $IS;
221^$CM+         $IS;              # by rule 10, stand-alone CM behaves as AL
222
223$LB8NonBreaks $SY;
224$CAN_CM $CM*  $SY;
225^$CM+         $SY;              # by rule 10, stand-alone CM behaves as AL
226
227
228#
229# LB 14  Do not break after OP, even after spaces
230#
231$OP $CM* $SP* .;
232
233$OP $CM* $SP+ $CM+ $AL_FOLLOW?;    # by rule 10, stand-alone CM behaves as AL
234                                   # by rule 8, CM following a SP is stand-alone.
235
236# LB 15
237$QU $CM* $SP* $OP;
238
239# LB 16
240# Do not break between closing punctuation and $NS, even with intervening spaces
241# But DO allow a break between closing punctuation and $NSX, don't include it here
242($CL | $CP) $CM* $SP* $NS;
243
244# LB 17
245$B2 $CM* $SP* $B2;
246
247#
248# LB 18  Break after spaces.
249#
250$LB18NonBreaks = [$LB8NonBreaks - [$SP]];
251$LB18Breaks    = [$LB8Breaks $SP];
252
253
254# LB 19
255#         x QU
256$LB18NonBreaks $CM* $QU;
257^$CM+               $QU;
258
259#         QU  x
260$QU $CM* .;
261
262
263# LB 20
264#        <break>  $CB
265#        $CB   <break>
266
267$LB20NonBreaks = [$LB18NonBreaks - $CB];
268
269# LB 21        x   (BA | HY | NS)
270#           BB x
271#
272# DO allow breaks here before $BAX and $NSX, so don't include them
273$LB20NonBreaks $CM* ($BA | $HY | $NS);
274^$CM+ ($BA | $HY | $NS);
275
276$BB $CM* [^$CB];                                  #  $BB  x
277$BB $CM* $LB20NonBreaks;
278
279# LB 21a Don't break after Hebrew + Hyphen
280#   HL (HY | BA) x
281#
282$HL $CM* ($HY | $BA | $BAX) $CM* [^$CB]?;
283
284# LB 21b (forward) Don't break between SY and HL
285# (break between HL and SY already disallowed by LB 13 above)
286$SY $CM* $HL;
287
288# LB 22
289($ALPlus | $HL) $CM* $IN;
290^$CM+    $IN;     #  by rule 10, any otherwise unattached CM behaves as AL
291$EX $CM*    $IN;
292($ID | $EB | $EM) $CM*  $IN;
293# $IN $CM*    $IN;  # delete this rule for CSS loose
294$NU $CM*    $IN;
295
296
297# $LB 23
298#
299($ALPlus | $HL) $CM* $NU;
300^$CM+  $NU;       # Rule 10, any otherwise unattached CM behaves as AL
301$NU $CM* ($ALPlus | $HL);
302
303# LB 23a
304# Do not include $POX here
305#
306$PR $CM* ($ID | $EB | $EM);
307($ID | $EB | $EM) $CM*  $PO;
308
309
310#
311# LB 24
312#
313# Do not include $PRX here
314($PR | $PO | $POX) $CM* ($ALPlus | $HL);
315($ALPlus | $HL) $CM* ($PR | $PO | $POX);     # TODO: should this be ($PR | $PRX | $PO)
316^$CM+ ($PR | $PO | $POX);       # Rule 10, any otherwise unattached CM behaves as AL
317
318#
319# LB 25   Numbers.
320#
321# Here do not include $PRX at the beginning or $POX at the end
322(($PR | $PO | $POX) $CM*)? (($OP | $HY) $CM*)? $NU ($CM* ($NU | $SY | $IS))*
323    ($CM* ($CL | $CP))? ($CM* ($PR | $PRX | $PO))?;
324
325# LB 26  Do not break a Korean syllable
326#
327$JL $CM* ($JL | $JV | $H2 | $H3);
328($JV | $H2) $CM* ($JV | $JT);
329($JT | $H3) $CM* $JT;
330
331# LB 27  Treat korean Syllable Block the same as ID  (don't break it)
332# Do not include $POX or $PRX here
333($JL | $JV | $JT | $H2 | $H3) $CM* $IN;
334($JL | $JV | $JT | $H2 | $H3) $CM* $PO;
335$PR $CM* ($JL | $JV | $JT | $H2 | $H3);
336
337
338# LB 28   Do not break between alphabetics
339#
340($ALPlus | $HL) $CM* ($ALPlus | $HL);
341^$CM+ ($ALPlus | $HL);      # The $CM+ is from rule 10, an unattached CM is treated as AL
342
343# LB 29
344$IS $CM* ($ALPlus | $HL);
345
346# LB 30
347($ALPlus | $HL | $NU) $CM* $OP;
348^$CM+ $OP;         # The $CM+ is from rule 10, an unattached CM is treated as AL.
349$CP $CM* ($ALPlus | $HL | $NU);
350
351# LB 30a  Do not break between regional indicators. Break after pairs of them.
352#         Tricky interaction with LB8a: ZWJ x ID
353$RI $CM* $RI                / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $CM] {eof}];
354$RI $CM* $RI $CM*  $ZWJ     / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $CM $ID $Extended_Pict $EmojiNRK] {eof}];
355$RI $CM* $RI $CM* [$CM-$ZWJ] / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $CM] {eof}];
356
357$RI $CM* $RI $CM* [$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS {eof}];
358$RI $CM* $RI $CM* $ZWJ ($ID | $Extended_Pict | $EmojiNRK);
359
360# LB 30b Do not break between an Emoji Base and an Emoji Modifier
361$EB $CM* $EM;
362
363#
364#  Reverse Rules.
365#
366## -------------------------------------------------
367
368!!reverse;
369
370#  LB 9 Combining Marks.
371#  Stick together any combining sequences that don't match other rules.
372
373^$CM+ $CAN_CM?;
374
375#
376#  Sequences of the form  (shown forwards)
377#      [CANT_CM]  <break>  [CM]  [whatever]
378#  The CM needs to behave as an AL
379#
380$AL_FOLLOW $CM+ / (
381          [$BK $CR $LF $NL $ZW {eof}] |
382          $SP+ $CM+ $SP |
383          $SP+ $CM* ([^$OP $CM $SP] | [$AL {eof}]));   # if LB 14 will match, need to surpress this break.
384                                               #  LB14 says    OP SP* x .
385                                               #    becomes    OP SP* x AL
386                                               #    becomes    OP SP* x CM+ AL_FOLLOW
387                                               #
388                                               # Further note:  the $AL in [$AL {eof}] is only to work around
389                                               #                a rule compiler bug which complains about
390                                               #                empty sets otherwise.
391
392
393# LB 4, 5, 6
394
395$LB4Breaks [$LB4NonBreaks-$CM];
396$LB4Breaks $CM+ $CAN_CM;
397$LF $CR;
398
399
400# LB 7         x SP
401#              x ZW
402[$SP $ZW] [$LB4NonBreaks-$CM];
403[$SP $ZW] $CM+ $CAN_CM;
404
405# LB 8 ZW SP* <break>
406#     TODO: to implement this, we need more than one look-ahead hard break in play at a time.
407#           Requires an engine enhancement.
408#   / $SP* $ZW
409
410# LB 8a        ZWJ x (ID | Extended_Pict | EmojiNRK)
411#
412($ID | $Extended_Pict | $EmojiNRK) $ZWJ $CM* $CAN_CM?;
413
414
415# LB 9,10  Combining marks.
416#    X   $CM needs to behave like X, where X is not $SP or controls.
417#    $CM not covered by the above needs to behave like $AL
418# Stick together any combining sequences that don't match other rules.
419^$CM+ $CAN_CM;
420
421
422# LB 11
423#
424$WJ $CM* $CAN_CM;
425$WJ      [$LB8NonBreaks-$CM];
426
427     $CANT_CM $CM* $WJ;
428$CAN_CM  $CM* $WJ;
429
430# LB 12a
431#      [^SP BA HY] x GL
432#
433$GL $CM* [$LB8NonBreaks-[$CM $SP $BA $BAX $HY]];
434
435# LB 12
436#     GL  x
437#
438$CANT_CM $CM* $GL;
439$CAN_CM $CM* $GL;
440
441
442# LB 13
443# Do not include $EXX here
444$CL $CM+ $CAN_CM;
445$CP $CM+ $CAN_CM;
446$EX $CM+ $CAN_CM;
447$IS $CM+ $CAN_CM;
448$SY $CM+ $CAN_CM;
449
450$CL [$LB8NonBreaks-$CM];
451$CP [$LB8NonBreaks-$CM];
452$EX [$LB8NonBreaks-$CM];
453$IS [$LB8NonBreaks-$CM];
454$SY [$LB8NonBreaks-$CM];
455
456
457# LB 14    OP SP* x
458#
459.   $SP* $CM* $OP;
460$AL_FOLLOW? $CM+ $SP+ $CM* $OP;     #  by LB 10, behaves like $AL_FOLLOW? $AL $SP+ $CM* $OP
461
462
463# LB 15
464$OP $SP* $CM* $QU;
465
466# LB 16
467# Don't include $NSX here
468$NS $SP* $CM* ($CL | $CP);
469
470# LB 17
471$B2 $SP* $CM* $B2;
472
473# LB 18  break after spaces
474#        Nothing explicit needed here.
475
476
477#
478# LB 19
479#
480$QU $CM* $CAN_CM;                                #   . x QU
481$QU      $LB18NonBreaks;
482
483
484$CAN_CM  $CM* $QU;                               #   QU x .
485     $CANT_CM $CM* $QU;
486
487#
488#  LB 20  Break before and after CB.
489#         nothing needed here.
490#
491
492# LB 21
493# Don't include $BAX or $NSX here
494($BA | $HY | $NS) $CM* [$LB20NonBreaks-$CM];     #  . x (BA | HY | NS)
495
496[$LB20NonBreaks-$CM] $CM* $BB;                   #  BB x .
497[^$CB] $CM* $BB;                                      #
498
499# LB21a
500[^$CM $CB]? $CM* ($HY | $BA | $BAX) $CM* $HL;
501
502# LB21b (reverse)
503$HL $CM* $SY;
504
505# LB 22
506$IN $CM* ($ALPlus | $HL);
507$IN $CM* $EX;
508$IN $CM* ($ID | $EB | $EM);
509# $IN $CM* $IN; # delete this rule for CSS loose
510$IN $CM* $NU;
511
512# LB 23
513# Do not include $POX here
514$NU $CM* ($ALPlus | $HL);
515($ALPlus | $HL) $CM* $NU;
516
517# LB23a
518# Do not include $PRX here
519($ID | $EB | $EM) $CM* $PR;
520$PO $CM* ($ID | $EB | $EM);
521
522# LB 24
523# Do not include $PRX here
524($ALPlus | $HL) $CM* ($PR | $PO | $POX);
525($PR | $PO | $POX) $CM* ($ALPlus | $HL);
526
527
528# LB 25
529# Here do not include $POX at the beginning or $PRX at the end
530($CM* ($PR | $PRX | $PO))? ($CM* ($CL | $CP))? ($CM* ($NU | $IS | $SY))* $CM* $NU ($CM* ($OP | $HY))? ($CM* ($PR | $PO | $POX))?;
531
532# LB 26
533($H3 | $H2 | $JV | $JL) $CM* $JL;
534($JT | $JV) $CM* ($H2 | $JV);
535$JT $CM* ($H3 | $JT);
536
537# LB 27
538# Do not include $POX or $PRX here
539$IN $CM* ($H3 | $H2 | $JT | $JV | $JL);
540$PO $CM* ($H3 | $H2 | $JT | $JV | $JL);
541 ($H3 | $H2 | $JT | $JV | $JL) $CM* $PR;
542
543# LB 28
544($ALPlus | $HL) $CM* ($ALPlus | $HL);
545
546
547# LB 29
548($ALPlus | $HL) $CM* $IS;
549
550# LB 30
551$OP $CM* ($ALPlus | $HL | $NU);
552($ALPlus | $HL | $NU) $CM* $CP;
553
554# LB 30a
555#    Pairs of Regional Indicators.
556#    The following two rules are nearly identical. The first matches only sequences with an odd number of adjacent RIs,
557#    the second with an even number. Stripping away the cruft they look like
558#         [^RI] RI / (RI RI)+ ^RI;
559#         [^RI] RI RI / (RI RI)+ ^RI;
560#
561# Line Loose tailoring: Don't include NSX here.
562[{bof} $NS $HY $BA $QU $CL $CP $EX $IS $SY $WJ $GL $ZW $SP $BK $CR $LF $NL $ZWJ] $CM* $RI / ($CM* $RI $CM* $RI)+ $CM* [{eof}[^$RI $CM]];
563[{bof} $NS $HY $BA $QU $CL $CP $EX $IS $SY $WJ $GL $ZW $SP $BK $CR $LF $NL $ZWJ] $CM* $RI $CM* $RI / ($CM* $RI $CM* $RI)+ $CM* [{eof}[^$RI $CM]];
564
565# In general, adjacent RIs stay together. The hard-break rules, above, overide this, forcing in the boundaries between pairs.
566$RI $CM* $RI;
567
568#    WJ, GL, QU, etc. are classes with rules like "WJ x "   which includes "WJ x RI".
569$RI $CM* ([$WJ $GL $QU $BB] |  (($HY | $BA)$CM* $HL));
570
571
572# LB 30b Do not break between an Emoji Base and an Emoji Modifier
573$EM $CM* $EB;
574
575
576## -------------------------------------------------
577
578!!safe_reverse;
579
580# LB 9
581^$CM+ [^$CM $BK $CR $LF $NL $ZW $SP];
582^$CM+ $SP / .;
583
584# LB 14
585$SP+ $CM* $OP;
586
587# LB 15
588$SP+ $CM* $QU;
589
590# LB 16
591$SP+ $CM* ($CL | $CP);
592
593# LB 17
594$SP+ $CM* $B2;
595
596# LB 21
597$CM* ($HY | $BA | $BAX) $CM* $HL;
598
599# LB 25
600($CM* ($IS | $SY))+ $CM* $NU;
601($CL | $CP) $CM* ($NU | $IS | $SY);
602
603#  LB 30
604($CM* $RI)+;
605
606# For dictionary-based break
607$dictionary $dictionary;
608
609## -------------------------------------------------
610
611!!safe_forward;
612
613# Skip forward over all character classes that are involved in
614#   rules containing patterns with possibly more than one char
615#   of context.
616#
617#  It might be slightly more efficient to have specific rules
618#  instead of one generic one, but only if we could
619#  turn off rule chaining.  We don't want to move more
620#  than necessary.
621#
622^[$CM $OP $QU $CL $CP $B2 $PR $PRX $HY $BA $BAX $SP $RI $ZWJ $dictionary]+ [^$CM $OP $QU $CL $CP $B2 $PR $PRX $HY $BA $BAX $RI $ZWJ $dictionary];
623$dictionary $dictionary;
624
625