[apple/icu.git] / icuSources / test / testdata / break_rules / line_normal_cj.txt

# Copyright (c) 2016  International Business Machines Corporation and # others. All Rights Reserved.
#
#  file:  line_normal_cj.txt
#
# Reference Line Break rules for intltest rbbi/RBBIMonkeyTest
#
# Note: Rule syntax and the monkey test itself are still a work in progress.
#       They are expected to change with review and the addition of support for rule tailoring.
#
#         Line Breaking Rules
#         Implement default line breaking as defined by 
#         Unicode Standard Annex #14 Revision 34 for Unicode 8.0
#         http://www.unicode.org/reports/tr14/
#         tailored as noted in 2nd paragraph below..
#
#         TODO:  Rule LB 8 remains as it was in Unicode 5.2
#         This is only because of a limitation of ICU break engine implementation,
#         not because the older behavior is desirable.
#
#         This tailors the line break behavior to correspond to CSS
#         line-break=normal (BCP47 -u-lb-normal) as defined for Chinese & Japanese.
#         It sets characters of class CJ to behave like ID.
#         In addition, it allows breaks:
#         * before hyphens 2010 & 2013 (both BA) and 301C, 30A0 (both NS)

type = line;
locale = ja@lb=normal;

AI = [[:LineBreak =  Ambiguous:] - [\u2640\u2642]];
AL = [[:LineBreak =  Alphabetic:]-[\u2695\u2696\u2764]];
BAX = [\u2010 \u2013];
BA = [[:LineBreak =  Break_After:] - BAX];
BB = [:LineBreak =  Break_Before:];
BK = [:LineBreak =  Mandatory_Break:];
B2 = [:LineBreak =  Break_Both:];
CB = [:LineBreak =  Contingent_Break:];
CJ = [:LineBreak =  Conditional_Japanese_Starter:];
CL = [:LineBreak =  Close_Punctuation:];
CM = [[:LineBreak =  Combining_Mark:] \u200D];
CP = [:LineBreak =  Close_Parenthesis:];
CR = [:LineBreak =  Carriage_Return:];

EB = [\u261D\u26F9\u270A-\u270D\U0001F385\U0001F3C2-\U0001F3C4\U0001F3C7\U0001F3CA-\U0001F3CC\U0001F442-\U0001F443\U0001F446-\U0001F450\U0001F466-\U0001F478\U0001F47C\U0001F481-\U0001F483\U0001F485-\U0001F487\U0001F4AA\U0001F574-\U0001F575\U0001F57A\U0001F590\U0001F595-\U0001F596\U0001F645-\U0001F647\U0001F64B-\U0001F64F\U0001F6A3\U0001F6B4-\U0001F6B6\U0001F6C0\U0001F6CC\U0001F918-\U0001F91E\U0001F926\U0001F930\U0001F933-\U0001F939\U0001F93C-\U0001F93E];
EM = [\U0001F3FB-\U0001F3FF];

EX = [:LineBreak =  Exclamation:];
GL = [:LineBreak =  Glue:];
HL = [:LineBreak =  Hebrew_Letter:];
HY = [:LineBreak =  Hyphen:];
H2 = [:LineBreak =  H2:];
H3 = [:LineBreak =  H3:];
ID = [[:LineBreak =  Ideographic:] CJ [\u2640\u2642\u2695\u2696\u2764]];
IN = [:LineBreak =  Inseperable:];
IS = [:LineBreak =  Infix_Numeric:];
JL = [:LineBreak =  JL:];
JV = [:LineBreak =  JV:];
JT = [:LineBreak =  JT:];
LF = [:LineBreak =  Line_Feed:];
NL = [:LineBreak =  Next_Line:];
NSX = [\u301C \u30A0];
NS = [[:LineBreak =  Nonstarter:] - NSX];
NU = [:LineBreak =  Numeric:];
OP = [:LineBreak =  Open_Punctuation:];
PO = [:LineBreak =  Postfix_Numeric:];
PR = [:LineBreak =  Prefix_Numeric:];
QU = [:LineBreak =  Quotation:];
RI = [:LineBreak =  Regional_Indicator:];
SA = [:LineBreak =  Complex_Context:];
SG = [:LineBreak =  Surrogate:];
SP = [:LineBreak =  Space:];
SY = [:LineBreak =  Break_Symbols:];
WJ = [:LineBreak =  Word_Joiner:];
XX = [:LineBreak =  Unknown:];
ZW = [:LineBreak =  ZWSpace:];
ZJ = [\u200D];

# TODO: adjustment to sets needed only until Unicode properties are updated for Emoji.
ID   = [ID - EB];
AL   = [AL - EM];

dictionary = [:LineBreak = Complex_Context:];

# Redfine AL. LB1.  TODO: refine according to latest UAX.
AL = [ AL AI SA SG XX ];

LB4:        BK ÷;
LB5:        CR LF;
LB5.1:      CR ÷;
LB5.2:      LF ÷;
LB5.3:      NL ÷;

LB6:        . (BK | CR | LF | NL);
LB6.1:      [^BK CR LF NL SP ZW] CM* (BK | CR | LF | NL);

# Rules LB14 - LB17.
# Moved before LB7, because they can match a longer sequence that would also match LB7,
# for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
#                           "while only the prefix "OP CM SP" matches LB7.1
LB14:        OP CM* SP* .;
LB15:        QU CM* SP* OP;

# Do not break between closing punctuation and $NS, even with intervening spaces
# But DO allow a break between closing punctuation and $NSX, don't include it here
LB16:        (CL | CP)CM* SP* NS;
LB17:        B2 CM* SP* B2;

LB7.1:      [^ZW SP] CM* [SP ZW];
LB7.2:      [ZW SP] [SP ZW];

# LB8, ICU differs from UAX-14,
#    ICU:    ZW ÷;
#    UAX 14: ZW SP* ÷;
LB8:        ZW ÷;

# LB8a, from Emoji proposal L2/16-011R3
#      ZWJ x ID
LB8a:       ZJ (ID | EB | EM);


# LB9:  X CM -> X
# LB10: Unattached CM -> AL

#LB11:       × WJ;
#            WJ ×

LB11.1:      [^BK CR LF NL SP ZW] CM* WJ;
LB11.2:      SP WJ;
LB11.3:      WJ CM* [^CM];

LB12:        GL CM* [^CM];

LB12a:       [^SP BA BAX HY] CM* GL;

# LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
#
#   LB13.1   [^SP] CM* [CL CP EX IS SY]    # original UAX 14 rule.
#   LB13.2   SP    CM* [CL CP EX IS SY]

LB13.1: [^NU SP] CM* [CL CP IS SY];
LB13.2: [^SP] CM* EX;
LB13.2: SP [CL CP EX IS SY];


# LB 14-17 are moved above LB 7.

LB18:        SP ÷;

LB19:        . CM* QU;
LB19.1:      QU CM* [^CM];

# LB 20   Break before and after CB.
#         Interaction with LB8a:  ZJ x ID is tricky because CM includes ZJ.
#                                 ZJ acts like a CM to the left, combining with CB.
#                                 ZJ acts independently to the right, no break from ID by LB8a.
LB20:        . CM* ÷ CB;
LB20.1a:     CB CM* ZJ (ID | EB | EM);
LB20.1b:      CB CM* ÷;

# Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
#       not picking up the continuing match after the BA from 21a.
# TODO: For CJ tailorings (with BAX) does this rule want to include BAX? If so,
#       should "HL BAX" not break when followed by a CB? Thats what the current
#       rules do, which is why "[^CM CB]?"\13 includes the ?.
LB21a:       HL CM* (HY | BA | BAX) CM* [^CM CB]?;

# DO allow breaks here before $BAXcm and $NSXcm, so don't include them
LB21.1:      . CM* [BA HY NS];
LB21.2:      BB CM* [^CM CB];

LB21b:       SY CM* HL;

LB22.1:        (AL | HL | CM) CM* IN;   # The CM is from LB10, treat an unattached CM as AL.
LB22.2:       EX CM* IN;
LB22.3:       (ID | EB | EM) CM* IN;
LB22.4:       IN CM* IN;
LB22.5:       NU CM* IN;

LB23.1:      (ID | EB | EM) CM* PO;
LB23.2:      (AL | HL | CM) CM* NU;
LB23.3:      NU CM* (AL | HL);

LB24.1:      PR CM* (ID | EB | EM);
LB24.2:      PR CM* (AL | HL);
LB24.3:      PO CM* (AL | HL);
LB24.4:      (AL | HL | CM) CM* (PR | PO);

# Numbers. Equivalent to Tailoring example 8 from UAx 14.
LB25:        ((PR | PO)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?;

LB26.1:      JL CM* (JL | JV | H2 | H3);
LB26.2:      (JV | H2) CM* (JV | JT);
LB26.3:      (JT | H3) CM* JT;

LB27.1:      (JL | JV | JT | H2 | H3) CM* IN;
LB27.2:      (JL | JV | JT | H2 | H3) CM* PO;
LB27.3:      PR CM* (JL | JV | JT | H2 | H3);

# LB28 Do not break between Alphabetics.
#      Unattached (leading) CM treated as AL.
LB28:        (AL | HL | CM)CM* (AL | HL);

LB29:        IS CM* (AL | HL);

# LB30  is adjusted for unattached leading CM being treated as AL.
LB30.1:      (AL | CM | HL | NU) CM* OP;
LB30.2:      CP CM* (AL | HL | NU);

# LB31  keep pairs of RI together.
LB30a.1:     RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS];
LB30a.2:     RI CM* RI CM* ZJ (ID | EB | EM);
LB30a.3:     RI CM* RI CM* ÷;

# LB30b Do not break between Emoji Base and Emoji Modifier
LB30b:       EB CM* EM;

# LB31 Break Everywhere Else.
#      Include combining marks
LB31.1:        . CM* ZJ (ID | EB | EM);
LB31.2:        . CM* ÷;
Commit	Line	Data
2ca993e8 A	1	# Copyright (c) 2016 International Business Machines Corporation and # others. All Rights Reserved.
	2	#
	3	# file: line_normal_cj.txt
	4	#
	5	# Reference Line Break rules for intltest rbbi/RBBIMonkeyTest
	6	#
	7	# Note: Rule syntax and the monkey test itself are still a work in progress.
	8	# They are expected to change with review and the addition of support for rule tailoring.
	9	#
	10	# Line Breaking Rules
	11	# Implement default line breaking as defined by
	12	# Unicode Standard Annex #14 Revision 34 for Unicode 8.0
	13	# http://www.unicode.org/reports/tr14/
	14	# tailored as noted in 2nd paragraph below..
	15	#
	16	# TODO: Rule LB 8 remains as it was in Unicode 5.2
	17	# This is only because of a limitation of ICU break engine implementation,
	18	# not because the older behavior is desirable.
	19	#
	20	# This tailors the line break behavior to correspond to CSS
	21	# line-break=normal (BCP47 -u-lb-normal) as defined for Chinese & Japanese.
	22	# It sets characters of class CJ to behave like ID.
	23	# In addition, it allows breaks:
	24	# * before hyphens 2010 & 2013 (both BA) and 301C, 30A0 (both NS)
	25
	26	type = line;
	27	locale = ja@lb=normal;
	28
	29	AI = [[:LineBreak = Ambiguous:] - [\u2640\u2642]];
a62d09fc	30	AL = [[:LineBreak = Alphabetic:]-[\u2695\u2696\u2764]];
2ca993e8 A	31	BAX = [\u2010 \u2013];
	32	BA = [[:LineBreak = Break_After:] - BAX];
	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:] \u200D];
	40	CP = [:LineBreak = Close_Parenthesis:];
	41	CR = [:LineBreak = Carriage_Return:];
	42
a62d09fc	43	EB = [\u261D\u26F9\u270A-\u270D\U0001F385\U0001F3C2-\U0001F3C4\U0001F3C7\U0001F3CA-\U0001F3CC\U0001F442-\U0001F443\U0001F446-\U0001F450\U0001F466-\U0001F478\U0001F47C\U0001F481-\U0001F483\U0001F485-\U0001F487\U0001F4AA\U0001F574-\U0001F575\U0001F57A\U0001F590\U0001F595-\U0001F596\U0001F645-\U0001F647\U0001F64B-\U0001F64F\U0001F6A3\U0001F6B4-\U0001F6B6\U0001F6C0\U0001F6CC\U0001F918-\U0001F91E\U0001F926\U0001F930\U0001F933-\U0001F939\U0001F93C-\U0001F93E];
2ca993e8 A	44	EM = [\U0001F3FB-\U0001F3FF];
	45
	46	EX = [:LineBreak = Exclamation:];
	47	GL = [:LineBreak = Glue:];
	48	HL = [:LineBreak = Hebrew_Letter:];
	49	HY = [:LineBreak = Hyphen:];
	50	H2 = [:LineBreak = H2:];
	51	H3 = [:LineBreak = H3:];
a62d09fc	52	ID = [[:LineBreak = Ideographic:] CJ [\u2640\u2642\u2695\u2696\u2764]];
2ca993e8 A	53	IN = [:LineBreak = Inseperable:];
	54	IS = [:LineBreak = Infix_Numeric:];
	55	JL = [:LineBreak = JL:];
	56	JV = [:LineBreak = JV:];
	57	JT = [:LineBreak = JT:];
	58	LF = [:LineBreak = Line_Feed:];
	59	NL = [:LineBreak = Next_Line:];
	60	NSX = [\u301C \u30A0];
	61	NS = [[:LineBreak = Nonstarter:] - NSX];
	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	ZJ = [\u200D];
	76
	77	# TODO: adjustment to sets needed only until Unicode properties are updated for Emoji.
	78	ID = [ID - EB];
	79	AL = [AL - EM];
	80
	81	dictionary = [:LineBreak = Complex_Context:];
	82
	83	# Redfine AL. LB1. TODO: refine according to latest UAX.
	84	AL = [ AL AI SA SG XX ];
	85
	86	LB4: BK ÷;
	87	LB5: CR LF;
	88	LB5.1: CR ÷;
	89	LB5.2: LF ÷;
	90	LB5.3: NL ÷;
	91
	92	LB6: . (BK \| CR \| LF \| NL);
	93	LB6.1: [^BK CR LF NL SP ZW] CM* (BK \| CR \| LF \| NL);
	94
	95	# Rules LB14 - LB17.
	96	# Moved before LB7, because they can match a longer sequence that would also match LB7,
	97	# for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
	98	# "while only the prefix "OP CM SP" matches LB7.1
	99	LB14: OP CM* SP* .;
	100	LB15: QU CM* SP* OP;
	101
	102	# Do not break between closing punctuation and $NS, even with intervening spaces
	103	# But DO allow a break between closing punctuation and $NSX, don't include it here
	104	LB16: (CL \| CP)CM* SP* NS;
	105	LB17: B2 CM* SP* B2;
	106
	107	LB7.1: [^ZW SP] CM* [SP ZW];
	108	LB7.2: [ZW SP] [SP ZW];
	109
	110	# LB8, ICU differs from UAX-14,
	111	# ICU: ZW ÷;
	112	# UAX 14: ZW SP* ÷;
	113	LB8: ZW ÷;
	114
	115	# LB8a, from Emoji proposal L2/16-011R3
	116	# ZWJ x ID
117	LB8a: ZJ (ID \| EB \| EM);
118
119
120	# LB9: X CM -> X
121	# LB10: Unattached CM -> AL
122
123	#LB11: × WJ;
124	# WJ ×
125
126	LB11.1: [^BK CR LF NL SP ZW] CM* WJ;
127	LB11.2: SP WJ;
128	LB11.3: WJ CM* [^CM];
129
130	LB12: GL CM* [^CM];
131
132	LB12a: [^SP BA BAX HY] CM* GL;
133
134	# LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
135	#
136	# LB13.1 [^SP] CM* [CL CP EX IS SY] # original UAX 14 rule.
137	# LB13.2 SP CM* [CL CP EX IS SY]
138
139	LB13.1: [^NU SP] CM* [CL CP IS SY];
140	LB13.2: [^SP] CM* EX;
141	LB13.2: SP [CL CP EX IS SY];
142
143
144	# LB 14-17 are moved above LB 7.
145
146	LB18: SP ÷;
147
148	LB19: . CM* QU;
149	LB19.1: QU CM* [^CM];
150
151	# LB 20 Break before and after CB.
152	# Interaction with LB8a: ZJ x ID is tricky because CM includes ZJ.
153	# ZJ acts like a CM to the left, combining with CB.
154	# ZJ acts independently to the right, no break from ID by LB8a.
155	LB20: . CM* ÷ CB;
156	LB20.1a: CB CM* ZJ (ID \| EB \| EM);
157	LB20.1b: 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	# TODO: For CJ tailorings (with BAX) does this rule want to include BAX? If so,
162	# should "HL BAX" not break when followed by a CB? Thats what the current
163	# rules do, which is why "[^CM CB]?"\13 includes the ?.
164	LB21a: HL CM* (HY \| BA \| BAX) CM* [^CM CB]?;
165
166	# DO allow breaks here before $BAXcm and $NSXcm, so don't include them
167	LB21.1: . CM* [BA HY NS];
168	LB21.2: BB CM* [^CM CB];
169
170	LB21b: SY CM* HL;
171
172	LB22.1: (AL \| HL \| CM) CM* IN; # The CM is from LB10, treat an unattached CM as AL.
173	LB22.2: EX CM* IN;
174	LB22.3: (ID \| EB \| EM) CM* IN;
175	LB22.4: IN CM* IN;
176	LB22.5: NU CM* IN;
177
178	LB23.1: (ID \| EB \| EM) CM* PO;
179	LB23.2: (AL \| HL \| CM) CM* NU;
180	LB23.3: NU CM* (AL \| HL);
181
182	LB24.1: PR CM* (ID \| EB \| EM);
183	LB24.2: PR CM* (AL \| HL);
184	LB24.3: PO CM* (AL \| HL);
185	LB24.4: (AL \| HL \| CM) CM* (PR \| PO);
186
187	# Numbers. Equivalent to Tailoring example 8 from UAx 14.
188	LB25: ((PR \| PO)CM)? ((OP \| HY)CM)? NU (CM(NU \| SY \| IS)) (CM(CL \| CP))? (CM(PR \| PO))?;
189
190	LB26.1: JL CM* (JL \| JV \| H2 \| H3);
191	LB26.2: (JV \| H2) CM* (JV \| JT);
192	LB26.3: (JT \| H3) CM* JT;
193
194	LB27.1: (JL \| JV \| JT \| H2 \| H3) CM* IN;
195	LB27.2: (JL \| JV \| JT \| H2 \| H3) CM* PO;
196	LB27.3: PR CM* (JL \| JV \| JT \| H2 \| H3);
197
198	# LB28 Do not break between Alphabetics.
199	# Unattached (leading) CM treated as AL.
200	LB28: (AL \| HL \| CM)CM* (AL \| HL);
201
202	LB29: IS CM* (AL \| HL);
203
204	# LB30 is adjusted for unattached leading CM being treated as AL.
205	LB30.1: (AL \| CM \| HL \| NU) CM* OP;
206	LB30.2: CP CM* (AL \| HL \| NU);
207
208	# LB31 keep pairs of RI together.
209	LB30a.1: RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS];
210	LB30a.2: RI CM* RI CM* ZJ (ID \| EB \| EM);
211	LB30a.3: RI CM* RI CM* ÷;
212
213	# LB30b Do not break between Emoji Base and Emoji Modifier
214	LB30b: EB CM* EM;
215
216	# LB31 Break Everywhere Else.
217	# Include combining marks
218	LB31.1: . CM* ZJ (ID \| EB \| EM);
219	LB31.2: . CM* ÷;