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

#
# Copyright (C) 2016 and later: Unicode, Inc. and others.
# License & terms of use: http://www.unicode.org/copyright.html
# 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.
# Rules derived from Unicode Standard Annex #14 Revision 40 for Unicode 11.0
#
# 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)
#         It allows breaking before 201C and after 201D, for zh_Hans, zh_Hant, and ja.

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

AI = [:LineBreak =  Ambiguous:];
AL = [:LineBreak =  Alphabetic:];
BAX = [\u2010 \u2013];
BA = [[:LineBreak =  Break_After:] - BAX];
HH = [\u2010];      # \u2010 is HYPHEN, default line break is BA.
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:] \u201d];
CMS = [:LineBreak =  Combining_Mark:];
CP = [:LineBreak =  Close_Parenthesis:];
CR = [:LineBreak =  Carriage_Return:];
EB = [[:LineBreak =  EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
EM = [:LineBreak =  EM:];
EX = [:LineBreak =  Exclamation:];
GL = [:LineBreak =  Glue:];
HL = [:LineBreak =  Hebrew_Letter:];
HY = [:LineBreak =  Hyphen:];
H2 = [:LineBreak =  H2:];
H3 = [:LineBreak =  H3:];
ID = [[:LineBreak =  Ideographic:] CJ];  # CSS Normal tailoring: CJ resolves to ID
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:] \u201c];
PO = [:LineBreak =  Postfix_Numeric:];
PR = [:LineBreak =  Prefix_Numeric:];
QU = [[:LineBreak =  Quotation:] - [\u201c\u201d]];
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:];
ZWJ = [:LineBreak =  ZWJ:];

# LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes
AL = [AL AI SG XX ];
dictionary = SA;

# By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
#         list it in the numerous rules that use CM.
CM = [CMS ZWJ];

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);

# LB8   break after ZW SP*.
# Precedes LB7 because both rules will match the sequences like ZW SP,
# and LB8 must take precedence.

LB8:        ZW SP* ÷ [^ZW SP BK CR LF NL];

# Numbers. Equivalent to Tailoring example 8 from UAX 14.
# Moved up, before LB14, because it can match longer sequences which must take precedence.
LB25:        ((PR | PO)CM*)? ((OP | HY)CM*)? (IS CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PO))?;

# Rules LB14 - LB17.
# Moved up, 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, "OP CM SP" matches LB7.1
LB14:        OP CM* SP* .;

# LB 14a Break before an IS that begins a number and follows a space.
LB14a:       SP ÷ IS CM* NU;

# LB14b       × IS
LB14b.1:     [^SP] CM* IS;
LB14b.2:     SP IS;

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 Do not break before spaces or zero width space.

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

# LB8a
#      ZWJ x
#      Don't match a CM on the right - let other rules pick up CM sequences, where
#      the ZWJ behaves as just another generic CM.
LB8a:       ZWJ [^CM];


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

#LB11:       × WJ;
#            WJ ×

LB11.1:      [^SP] CM* WJ;
LB11.2:      SP WJ;
LB11.3:      WJ CM* [^CM];

LB12:        GL CM* [^CM];

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

# LB 13 Do not break before ‘]’ or ‘!’ or ‘/’, even after spaces.
LB13.1: [^SP] CM* [CL CP EX SY];
LB13.2: SP [CL CP EX 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:  ZWJ x . is tricky because CM includes ZWJ.
#                                 ZWJ acts like a CM to the left, combining with CB.
#                                 ZWJ acts independently to the right, no break after by LB8a.
LB20.1:      . CM* ZWJ CB;
LB20.2:      . CM* ÷ CB;

LB20.3:      CB CM* ZWJ [^CM];
LB20.4:      CB CM* ÷;

# LB 20.09    Don't break between Hyphens and Letters when there is a break preceding the hyphen.
LB20.09:     ^(HY | HH) CM* AL;

# 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]?" 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:      (AL | HL | CM) CM* NU;
LB23.2:      NU CM* (AL | HL);

LB23a.1:     PR CM* (ID | EB | EM);
LB23a.2:     (ID | EB | EM) CM* PO;

LB24.2:      (PR | PO) CM* (AL | HL);
LB24.3:      (AL | HL | CM) 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);

# LB30a  keep pairs of RI together.
LB30a.1:     RI CM* RI         ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
LB30a.2:     RI CM* RI CM* CMS ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
LB30a.3:     RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS ZWJ]?;

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

# LB31 Break Everywhere Else.
#      Include combining marks
LB31.1:        . CM* ZWJ [^CM];
LB31.2:        . CM* ÷;
Commit	Line	Data
f3c0d7a5 A	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.
2ca993e8 A	5	#
	6	# file: line_normal_cj.txt
	7	#
0f5d89e8 A	8	# Reference Line Break rules for intltest rbbi/RBBIMonkeyTest.
0f5d89e8 A	9	# Rules derived from Unicode Standard Annex #14 Revision 40 for Unicode 11.0
2ca993e8 A	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
f3c0d7a5	15	# Implement default line breaking as defined by
2ca993e8 A	16	# Unicode Standard Annex #14 Revision 34 for Unicode 8.0
2ca993e8 A	17	# http://www.unicode.org/reports/tr14/
f3c0d7a5	18	# tailored as noted in 2nd paragraph below.
2ca993e8 A	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 Chinese & Japanese.
	26	# It sets characters of class CJ to behave like ID.
	27	# In addition, it allows breaks:
	28	# * before hyphens 2010 & 2013 (both BA) and 301C, 30A0 (both NS)
3d1f044b	29	# It allows breaking before 201C and after 201D, for zh_Hans, zh_Hant, and ja.
2ca993e8 A	30
	31	type = line;
	32	locale = ja@lb=normal;
	33
f3c0d7a5 A	34	AI = [:LineBreak = Ambiguous:];
f3c0d7a5 A	35	AL = [:LineBreak = Alphabetic:];
2ca993e8 A	36	BAX = [\u2010 \u2013];
2ca993e8 A	37	BA = [[:LineBreak = Break_After:] - BAX];
3d1f044b	38	HH = [\u2010]; # \u2010 is HYPHEN, default line break is BA.
2ca993e8 A	39	BB = [:LineBreak = Break_Before:];
	40	BK = [:LineBreak = Mandatory_Break:];
	41	B2 = [:LineBreak = Break_Both:];
	42	CB = [:LineBreak = Contingent_Break:];
	43	CJ = [:LineBreak = Conditional_Japanese_Starter:];
3d1f044b A	44	CL = [[:LineBreak = Close_Punctuation:] \u201d];
3d1f044b A	45	CMS = [:LineBreak = Combining_Mark:];
2ca993e8 A	46	CP = [:LineBreak = Close_Parenthesis:];
2ca993e8 A	47	CR = [:LineBreak = Carriage_Return:];
f3c0d7a5 A	48	EB = [[:LineBreak = EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
f3c0d7a5 A	49	EM = [:LineBreak = EM:];
2ca993e8 A	50	EX = [:LineBreak = Exclamation:];
	51	GL = [:LineBreak = Glue:];
	52	HL = [:LineBreak = Hebrew_Letter:];
	53	HY = [:LineBreak = Hyphen:];
	54	H2 = [:LineBreak = H2:];
	55	H3 = [:LineBreak = H3:];
f3c0d7a5	56	ID = [[:LineBreak = Ideographic:] CJ]; # CSS Normal tailoring: CJ resolves to ID
2ca993e8 A	57	IN = [:LineBreak = Inseperable:];
	58	IS = [:LineBreak = Infix_Numeric:];
	59	JL = [:LineBreak = JL:];
	60	JV = [:LineBreak = JV:];
	61	JT = [:LineBreak = JT:];
	62	LF = [:LineBreak = Line_Feed:];
	63	NL = [:LineBreak = Next_Line:];
	64	NSX = [\u301C \u30A0];
	65	NS = [[:LineBreak = Nonstarter:] - NSX];
	66	NU = [:LineBreak = Numeric:];
3d1f044b	67	OP = [[:LineBreak = Open_Punctuation:] \u201c];
2ca993e8 A	68	PO = [:LineBreak = Postfix_Numeric:];
2ca993e8 A	69	PR = [:LineBreak = Prefix_Numeric:];
3d1f044b	70	QU = [[:LineBreak = Quotation:] - [\u201c\u201d]];
2ca993e8 A	71	RI = [:LineBreak = Regional_Indicator:];
	72	SA = [:LineBreak = Complex_Context:];
	73	SG = [:LineBreak = Surrogate:];
	74	SP = [:LineBreak = Space:];
	75	SY = [:LineBreak = Break_Symbols:];
	76	WJ = [:LineBreak = Word_Joiner:];
	77	XX = [:LineBreak = Unknown:];
	78	ZW = [:LineBreak = ZWSpace:];
f3c0d7a5	79	ZWJ = [:LineBreak = ZWJ:];
2ca993e8	80
f3c0d7a5 A	81	# LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes
	82	AL = [AL AI SG XX ];
	83	dictionary = SA;
2ca993e8	84
f3c0d7a5 A	85	# By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
f3c0d7a5 A	86	# list it in the numerous rules that use CM.
3d1f044b	87	CM = [CMS ZWJ];
2ca993e8 A	88
	89	LB4: BK ÷;
	90	LB5: CR LF;
	91	LB5.1: CR ÷;
	92	LB5.2: LF ÷;
	93	LB5.3: NL ÷;
	94
	95	LB6: . (BK \| CR \| LF \| NL);
	96	LB6.1: [^BK CR LF NL SP ZW] CM* (BK \| CR \| LF \| NL);
	97
3d1f044b A	98	# LB8 break after ZW SP*.
	99	# Precedes LB7 because both rules will match the sequences like ZW SP,
	100	# and LB8 must take precedence.
	101
	102	LB8: ZW SP* ÷ [^ZW SP BK CR LF NL];
	103
	104	# Numbers. Equivalent to Tailoring example 8 from UAX 14.
	105	# Moved up, before LB14, because it can match longer sequences which must take precedence.
	106	LB25: ((PR \| PO)CM)? ((OP \| HY)CM)? (IS CM)? NU (CM(NU \| SY \| IS))* (CM(CL \| CP))? (CM(PR \| PO))?;
	107
2ca993e8	108	# Rules LB14 - LB17.
3d1f044b A	109	# Moved up, before LB7, because they can match a longer sequence that would also match LB7.
	110	# For example, the sequence "OP CM SP AL" matches LB14
	111	# while the prefix of it, "OP CM SP" matches LB7.1
2ca993e8	112	LB14: OP CM* SP* .;
3d1f044b A	113
	114	# LB 14a Break before an IS that begins a number and follows a space.
	115	LB14a: SP ÷ IS CM* NU;
	116
	117	# LB14b × IS
	118	LB14b.1: [^SP] CM* IS;
	119	LB14b.2: SP IS;
	120
2ca993e8 A	121	LB15: QU CM* SP* OP;
	122
	123	# Do not break between closing punctuation and $NS, even with intervening spaces
	124	# But DO allow a break between closing punctuation and $NSX, don't include it here
	125	LB16: (CL \| CP)CM* SP* NS;
	126	LB17: B2 CM* SP* B2;
	127
3d1f044b A	128	# LB7 Do not break before spaces or zero width space.
3d1f044b A	129
2ca993e8 A	130	LB7.1: [^ZW SP] CM* [SP ZW];
	131	LB7.2: [ZW SP] [SP ZW];
	132
f3c0d7a5	133	# LB8a
0f5d89e8 A	134	# ZWJ x
	135	# Don't match a CM on the right - let other rules pick up CM sequences, where
	136	# the ZWJ behaves as just another generic CM.
	137	LB8a: ZWJ [^CM];
2ca993e8 A	138
	139
	140	# LB9: X CM -> X
	141	# LB10: Unattached CM -> AL
	142
	143	#LB11: × WJ;
	144	# WJ ×
	145
0f5d89e8	146	LB11.1: [^SP] CM* WJ;
2ca993e8 A	147	LB11.2: SP WJ;
	148	LB11.3: WJ CM* [^CM];
	149
	150	LB12: GL CM* [^CM];
	151
	152	LB12a: [^SP BA BAX HY] CM* GL;
	153
3d1f044b A	154	# LB 13 Do not break before ‘]’ or ‘!’ or ‘/’, even after spaces.
	155	LB13.1: [^SP] CM* [CL CP EX SY];
	156	LB13.2: SP [CL CP EX SY];
2ca993e8 A	157
	158	# LB 14-17 are moved above LB 7.
	159
	160	LB18: SP ÷;
	161
	162	LB19: . CM* QU;
	163	LB19.1: QU CM* [^CM];
	164
	165	# LB 20 Break before and after CB.
0f5d89e8	166	# Interaction with LB8a: ZWJ x . is tricky because CM includes ZWJ.
f3c0d7a5	167	# ZWJ acts like a CM to the left, combining with CB.
0f5d89e8 A	168	# ZWJ acts independently to the right, no break after by LB8a.
	169	LB20.1: . CM* ZWJ CB;
	170	LB20.2: . CM* ÷ CB;
	171
	172	LB20.3: CB CM* ZWJ [^CM];
	173	LB20.4: CB CM* ÷;
2ca993e8	174
3d1f044b A	175	# LB 20.09 Don't break between Hyphens and Letters when there is a break preceding the hyphen.
	176	LB20.09: ^(HY \| HH) CM* AL;
	177
2ca993e8 A	178	# Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
	179	# not picking up the continuing match after the BA from 21a.
	180	# TODO: For CJ tailorings (with BAX) does this rule want to include BAX? If so,
	181	# should "HL BAX" not break when followed by a CB? Thats what the current
f3c0d7a5	182	# rules do, which is why "[^CM CB]?" includes the ?.
2ca993e8 A	183	LB21a: HL CM* (HY \| BA \| BAX) CM* [^CM CB]?;
	184
	185	# DO allow breaks here before $BAXcm and $NSXcm, so don't include them
	186	LB21.1: . CM* [BA HY NS];
	187	LB21.2: BB CM* [^CM CB];
	188
	189	LB21b: SY CM* HL;
	190
	191	LB22.1: (AL \| HL \| CM) CM* IN; # The CM is from LB10, treat an unattached CM as AL.
	192	LB22.2: EX CM* IN;
	193	LB22.3: (ID \| EB \| EM) CM* IN;
	194	LB22.4: IN CM* IN;
	195	LB22.5: NU CM* IN;
	196
f3c0d7a5 A	197	LB23.1: (AL \| HL \| CM) CM* NU;
	198	LB23.2: NU CM* (AL \| HL);
	199
	200	LB23a.1: PR CM* (ID \| EB \| EM);
	201	LB23a.2: (ID \| EB \| EM) CM* PO;
2ca993e8	202
f3c0d7a5 A	203	LB24.2: (PR \| PO) CM* (AL \| HL);
f3c0d7a5 A	204	LB24.3: (AL \| HL \| CM) CM* (PR \| PO);
2ca993e8	205
2ca993e8 A	206	LB26.1: JL CM* (JL \| JV \| H2 \| H3);
	207	LB26.2: (JV \| H2) CM* (JV \| JT);
	208	LB26.3: (JT \| H3) CM* JT;
	209
	210	LB27.1: (JL \| JV \| JT \| H2 \| H3) CM* IN;
	211	LB27.2: (JL \| JV \| JT \| H2 \| H3) CM* PO;
	212	LB27.3: PR CM* (JL \| JV \| JT \| H2 \| H3);
	213
	214	# LB28 Do not break between Alphabetics.
	215	# Unattached (leading) CM treated as AL.
	216	LB28: (AL \| HL \| CM)CM* (AL \| HL);
	217
	218	LB29: IS CM* (AL \| HL);
	219
	220	# LB30 is adjusted for unattached leading CM being treated as AL.
	221	LB30.1: (AL \| CM \| HL \| NU) CM* OP;
	222	LB30.2: CP CM* (AL \| HL \| NU);
	223
0f5d89e8	224	# LB30a keep pairs of RI together.
3d1f044b A	225	LB30a.1: RI CM* RI ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
3d1f044b A	226	LB30a.2: RI CM* RI CM* CMS ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
0f5d89e8	227	LB30a.3: RI CM* RI CM* [BK CR LF NL SP ZW WJ GL CL CP EX IS SY QU BA HY NS ZWJ]?;
2ca993e8 A	228
	229	# LB30b Do not break between Emoji Base and Emoji Modifier
	230	LB30b: EB CM* EM;
	231
	232	# LB31 Break Everywhere Else.
	233	# Include combining marks
0f5d89e8	234	LB31.1: . CM* ZWJ [^CM];
2ca993e8	235	LB31.2: . CM* ÷;