• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
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.txt
7#
8#         Line Breaking Rules
9#         Implement default line breaking as defined by
10#         Unicode Standard Annex #14 (https://www.unicode.org/reports/tr14/)
11#         for Unicode 14.0, with the following modification:
12#
13#         Boundaries between hyphens and following letters are suppressed when
14#         there is a boundary preceding the hyphen. See rule 20.9
15#
16#         This corresponds to CSS line-break=strict (BCP47 -u-lb-strict).
17#         It sets characters of class CJ to behave like NS.
18
19#
20#  Character Classes defined by TR 14.
21#
22
23!!chain;
24!!quoted_literals_only;
25
26$AI = [:LineBreak =  Ambiguous:];
27$AK = [:LineBreak =  Aksara:];
28$AL = [:LineBreak =  Alphabetic:];
29$AP = [:LineBreak =  Aksara_Prebase:];
30$AS = [:LineBreak =  Aksara_Start:];
31$BA = [:LineBreak =  Break_After:];
32$HH = [\u2010];     # \u2010 is HYPHEN, default line break is BA.
33$BB = [:LineBreak =  Break_Before:];
34$BK = [:LineBreak =  Mandatory_Break:];
35$B2 = [:LineBreak =  Break_Both:];
36$CB = [:LineBreak =  Contingent_Break:];
37$CJ = [:LineBreak =  Conditional_Japanese_Starter:];
38$CL = [:LineBreak =  Close_Punctuation:];
39# $CM = [:LineBreak =  Combining_Mark:];
40$CP = [:LineBreak =  Close_Parenthesis:];
41$CR = [:LineBreak =  Carriage_Return:];
42$EB = [:LineBreak =  EB:];
43$EM = [:LineBreak =  EM:];
44$EX = [:LineBreak =  Exclamation:];
45$GL = [:LineBreak =  Glue:];
46$HL = [:LineBreak =  Hebrew_Letter:];
47$HY = [:LineBreak =  Hyphen:];
48$H2 = [:LineBreak =  H2:];
49$H3 = [:LineBreak =  H3:];
50$ID = [:LineBreak =  Ideographic:];
51$IN = [:LineBreak =  Inseperable:];
52$IS = [:LineBreak =  Infix_Numeric:];
53$JL = [:LineBreak =  JL:];
54$JV = [:LineBreak =  JV:];
55$JT = [:LineBreak =  JT:];
56$LF = [:LineBreak =  Line_Feed:];
57$NL = [:LineBreak =  Next_Line:];
58# NS includes CJ for CSS strict line breaking.
59$NS = [[:LineBreak =  Nonstarter:] $CJ];
60$NU = [:LineBreak =  Numeric:];
61$OP = [:LineBreak =  Open_Punctuation:];
62$PO = [:LineBreak =  Postfix_Numeric:];
63$PR = [:LineBreak =  Prefix_Numeric:];
64$QU = [:LineBreak =  Quotation:];
65$RI = [:LineBreak =  Regional_Indicator:];
66$SA = [:LineBreak =  Complex_Context:];
67$SG = [:LineBreak =  Surrogate:];
68$SP = [:LineBreak =  Space:];
69$SY = [:LineBreak =  Break_Symbols:];
70$VF = [:LineBreak =  Virama_Final:];
71$VI = [:LineBreak =  Virama:];
72$WJ = [:LineBreak =  Word_Joiner:];
73$XX = [:LineBreak =  Unknown:];
74$ZW = [:LineBreak =  ZWSpace:];
75$ZWJ = [:LineBreak = ZWJ:];
76
77# OP30 and CP30 are variants of OP and CP that appear in-line in rule LB30 from UAX 14,
78# without a formal name. Because ICU rules require multiple uses of the expressions,
79# give them a single definition with a name
80
81$OP30 = [$OP - [\p{ea=F}\p{ea=W}\p{ea=H}]];
82$CP30 = [$CP - [\p{ea=F}\p{ea=W}\p{ea=H}]];
83
84$ExtPictUnassigned = [\p{Extended_Pictographic} & \p{Cn}];
85
86# By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
87#         list it in the numerous rules that use CM.
88# By LB1, SA characters with general categor of Mn or Mc also resolve to CM.
89
90$CM = [[:LineBreak = Combining_Mark:] $ZWJ [$SA & [[:Mn:][:Mc:]]]];
91$CMX = [[$CM] - [$ZWJ]];
92
93#   Dictionary character set, for triggering language-based break engines. Currently
94#   limited to LineBreak=Complex_Context (SA).
95
96$dictionary = [$SA];
97
98#
99#  Rule LB1.  By default, treat AI  (characters with ambiguous east Asian width),
100#                               SA  (Dictionary chars, excluding Mn and Mc)
101#                               SG  (Unpaired Surrogates)
102#                               XX  (Unknown, unassigned)
103#                         as $AL  (Alphabetic)
104#
105$ALPlus = [$AL $AI $SG $XX [$SA-[[:Mn:][:Mc:]]]];
106
107
108## -------------------------------------------------
109
110#
111# CAN_CM  is the set of characters that may combine with CM combining chars.
112#         Note that Linebreak UAX 14's concept of a combining char and the rules
113#         for what they can combine with are _very_ different from the rest of Unicode.
114#
115#         Note that $CM itself is left out of this set.  If CM is needed as a base
116#         it must be listed separately in the rule.
117#
118$CAN_CM  = [^$SP $BK $CR $LF $NL $ZW $CM];       # Bases that can   take CMs
119$CANT_CM = [ $SP $BK $CR $LF $NL $ZW $CM];       # Bases that can't take CMs
120
121#
122# AL_FOLLOW  set of chars that can unconditionally follow an AL
123#            Needed in rules where stand-alone $CM s are treated as AL.
124#
125$AL_FOLLOW      = [$BK $CR $LF $NL $ZW $SP $CL $CP $EX $HL $IS $SY $WJ $GL $OP30 $QU $BA $HY $NS $IN $NU $PR $PO $ALPlus];
126
127
128#
129#  Rule LB 4, 5    Mandatory (Hard) breaks.
130#
131$LB4Breaks    = [$BK $CR $LF $NL];
132$LB4NonBreaks = [^$BK $CR $LF $NL $CM];
133$CR $LF {100};
134
135#
136#  LB 6    Do not break before hard line breaks.
137#
138$LB4NonBreaks?  $LB4Breaks {100};    # LB 5  do not break before hard breaks.
139$CAN_CM $CM*    $LB4Breaks {100};
140^$CM+           $LB4Breaks {100};
141
142# LB 7         x SP
143#              x ZW
144$LB4NonBreaks [$SP $ZW];
145$CAN_CM $CM*  [$SP $ZW];
146^$CM+         [$SP $ZW];
147
148#
149# LB 8         Break after zero width space
150#              ZW SP* ÷
151#
152$LB8Breaks    = [$LB4Breaks $ZW];
153$LB8NonBreaks = [[$LB4NonBreaks] - [$ZW]];
154$ZW $SP* / [^$SP $ZW $LB4Breaks];
155
156# LB 8a        ZWJ x            Do not break Emoji ZWJ sequences.
157#
158$ZWJ [^$CM];
159
160# LB 9     Combining marks.      X   $CM needs to behave like X, where X is not $SP, $BK $CR $LF $NL
161#                                $CM not covered by the above needs to behave like $AL
162#                                See definition of $CAN_CM.
163
164$CAN_CM $CM+;                   #  Stick together any combining sequences that don't match other rules.
165^$CM+;
166
167#
168# LB 11  Do not break before or after WORD JOINER & related characters.
169#
170$CAN_CM $CM*  $WJ;
171$LB8NonBreaks $WJ;
172^$CM+         $WJ;
173
174$WJ $CM* .;
175
176#
177# LB 12  Do not break after NBSP and related characters.
178#         GL  x
179#
180$GL $CM* .;
181
182#
183# LB 12a  Do not break before NBSP and related characters ...
184#            [^SP BA HY] x GL
185#
186[[$LB8NonBreaks] - [$SP $BA $HY]] $CM* $GL;
187^$CM+ $GL;
188
189
190
191
192# LB 13   Don't break before ']' or '!' or '/', even after spaces.
193#
194$LB8NonBreaks $CL;
195$CAN_CM $CM*  $CL;
196^$CM+         $CL;              # by rule 10, stand-alone CM behaves as AL
197
198$LB8NonBreaks $CP;
199$CAN_CM $CM*  $CP;
200^$CM+         $CP;              # by rule 10, stand-alone CM behaves as AL
201
202$LB8NonBreaks $EX;
203$CAN_CM $CM*  $EX;
204^$CM+         $EX;              # by rule 10, stand-alone CM behaves as AL
205
206$LB8NonBreaks $SY;
207$CAN_CM $CM*  $SY;
208^$CM+         $SY;              # by rule 10, stand-alone CM behaves as AL
209
210
211#
212# LB 14  Do not break after OP, even after spaces
213#        Note subtle interaction with "SP IS /" rules in LB14a.
214#        This rule consumes the SP, chaining happens on the IS, effectivley overriding the  SP IS rules,
215#        which is the desired behavior.
216#
217$OP $CM* $SP* .;
218
219$OP $CM* $SP+ $CM+ $AL_FOLLOW?;    # by rule 10, stand-alone CM behaves as AL
220                                   # by rule 8, CM following a SP is stand-alone.
221
222
223# LB 15a
224($OP $CM* $SP+ | [$OP $QU $GL] $CM*) ([\p{Pi} & $QU] $CM* $SP*)+ .;
225($OP $CM* $SP+ | [$OP $QU $GL] $CM*) ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?;
226^([\p{Pi} & $QU] $CM* $SP*)+ .;
227^([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?;
228
229# LB 15b
230$LB8NonBreaks [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}];
231$CAN_CM $CM*  [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}];
232^$CM+  [\p{Pf} & $QU] $CM* [$SP $GL $WJ $CL $QU $CP $EX $IS $SY $BK $CR $LF $NL $ZW {eof}];
233
234# Messy interaction: manually chain between LB 15b and LB 15a on Pf Pi.
235$LB8NonBreaks [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .;
236$LB8NonBreaks [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?;
237$CAN_CM $CM*  [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .;
238$CAN_CM $CM*  [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?;
239^$CM+  [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ .;
240^$CM+  [\p{Pf} & $QU] $CM* ([\p{Pi} & $QU] $CM* $SP*)+ $SP $CM+ $AL_FOLLOW?;
241
242
243# LB 15c Force a break before start of a number with a leading decimal pt, e.g. " .23"
244#        Note: would be simpler to express as "$SP / $IS $CM* $NU;", but ICU rules have limitations.
245#        See issue ICU-20303
246
247
248$CanFollowIS = [$BK $CR $LF $NL $SP $ZW $WJ $GL $CL $CP $EX $IS $SY $QU $BA $HY $NS $ALPlus $HL $IN];
249$SP $IS           / [^ $CanFollowIS $NU $CM];
250$SP $IS $CM* $CMX / [^ $CanFollowIS $NU $CM];
251
252#
253# LB 15d Do not break before numeric separators (IS), even after spaces.
254
255[$LB8NonBreaks - $SP] $IS;
256$SP $IS $CM* [$CanFollowIS {eof}];
257$SP $IS $CM* $ZWJ [^$CM $NU];
258
259$CAN_CM $CM*  $IS;
260^$CM+         $IS;              # by rule 10, stand-alone CM behaves as AL
261
262
263# LB 16
264($CL | $CP) $CM* $SP* $NS;
265
266# LB 17
267$B2 $CM* $SP* $B2;
268
269#
270# LB 18  Break after spaces.
271#
272$LB18NonBreaks = [$LB8NonBreaks - [$SP]];
273$LB18Breaks    = [$LB8Breaks $SP];
274
275
276# LB 19
277#         x QU
278$LB18NonBreaks $CM* $QU;
279^$CM+               $QU;
280
281#         QU  x
282$QU $CM* .;
283
284# LB 20
285#        <break>  $CB
286#        $CB   <break>
287#
288$LB20NonBreaks = [$LB18NonBreaks - $CB];
289
290# LB 20.09    Don't break between Hyphens and Letters when there is a break preceding the hyphen.
291#             Originally added as a Finnish tailoring, now promoted to default ICU behavior.
292#             Note: this is not default UAX-14 behaviour. See issue ICU-8151.
293#
294^($HY | $HH) $CM* $ALPlus;
295
296# LB 21        x   (BA | HY | NS)
297#           BB x
298#
299$LB20NonBreaks $CM* ($BA | $HY | $NS);
300
301
302^$CM+ ($BA | $HY | $NS);
303
304$BB $CM* [^$CB];                                  #  $BB  x
305$BB $CM* $LB20NonBreaks;
306
307# LB 21a Don't break after Hebrew + Hyphen
308#   HL (HY | BA) x
309#
310$HL $CM* ($HY | $BA) $CM* [^$CB]?;
311
312# LB 21b (forward) Don't break between SY and HL
313# (break between HL and SY already disallowed by LB 13 above)
314$SY $CM* $HL;
315
316# LB 22  Do not break before ellipses
317#
318$LB20NonBreaks $CM*    $IN;
319^$CM+ $IN;
320
321
322# LB 23
323#
324($ALPlus | $HL) $CM* $NU;
325^$CM+  $NU;       # Rule 10, any otherwise unattached CM behaves as AL
326$NU $CM* ($ALPlus | $HL);
327
328# LB 23a
329#
330$PR $CM* ($ID | $EB | $EM);
331($ID | $EB | $EM) $CM*  $PO;
332
333
334#
335# LB 24
336#
337($PR | $PO) $CM* ($ALPlus | $HL);
338($ALPlus | $HL) $CM* ($PR | $PO);
339^$CM+ ($PR | $PO);       # Rule 10, any otherwise unattached CM behaves as AL
340
341#
342# LB 25   Numbers.
343#
344(($PR | $PO) $CM*)? (($OP | $HY) $CM*)? ($IS $CM*)? $NU ($CM* ($NU | $SY | $IS))*
345    ($CM* ($CL | $CP))? ($CM* ($PR | $PO))?;
346
347# LB 26  Do not break a Korean syllable
348#
349$JL $CM* ($JL | $JV | $H2 | $H3);
350($JV | $H2) $CM* ($JV | $JT);
351($JT | $H3) $CM* $JT;
352
353# LB 27  Treat korean Syllable Block the same as ID  (don't break it)
354($JL | $JV | $JT | $H2 | $H3) $CM* $PO;
355$PR $CM* ($JL | $JV | $JT | $H2 | $H3);
356
357
358# LB 28   Do not break between alphabetics
359#
360($ALPlus | $HL) $CM* ($ALPlus | $HL);
361^$CM+ ($ALPlus | $HL);      # The $CM+ is from rule 10, an unattached CM is treated as AL
362
363#LB 28a  Do not break Orthographic syllables
364($AP $CM*)? ($AS | $AK | [◌] ) ($CM* $VI $CM* ($AK | [◌] ))* ($CM* $VI | (($CM* ($AS | $AK | [◌] ) )? $CM* $VF))?;
365
366# LB 29
367$IS $CM* ($ALPlus | $HL);
368
369# LB 30
370($ALPlus | $HL | $NU) $CM* $OP30;
371^$CM+ $OP30;         # The $CM+ is from rule 10, an unattached CM is treated as AL.
372$CP30 $CM* ($ALPlus | $HL | $NU);
373
374# LB 30a  Do not break between regional indicators. Break after pairs of them.
375#         Tricky interaction with LB8a: ZWJ x .   together with ZWJ acting like a CM.
376$RI $CM* $RI                 / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $IN $CM]];
377$RI $CM* $RI $CM* [$CM-$ZWJ] / [[^$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $IN $CM]];
378$RI $CM* $RI $CM* [$BK $CR $LF $NL $SP $ZW $WJ $CL $CP $EX $IS $SY $GL $QU $BA $HY $NS $IN $ZWJ {eof}];
379# note: the preceding rule includes {eof} rather than having the last [set] term qualified with '?'
380#       because of the chain-out behavior difference. The rule must chain out only from the [set characters],
381#       not from the preceding $RI or $CM, which it would be able to do if the set were optional.
382
383# LB30b Do not break between an emoji base (or potential emoji) and an emoji modifier.
384$EB $CM* $EM;
385$ExtPictUnassigned $CM* $EM;
386
387# LB 31 Break everywhere else.
388#       Match a single code point if no other rule applies.
389.;
390