[apple/icu.git] / icuSources / test / testdata / break_rules / line_loose_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_loose_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..
#
#         This tailors the line break behavior to correspond to CSS
#         line-break=loose (BCP47 -u-lb-loose) 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)
#         * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS)
#         * between characters of LineBreak class IN such as 2026
#         * before some centered punct 203C, 2047, 2048, 2049, 30FB, FF1A, FF1B,
#           FF65 (all NS) and FF01, FF1F (both EX).
#         * before suffix characters with LineBreak class PO and EastAsianWidth A,F,W;
#           this includes: 00B0 2030 2032 2033 2035 2103 2109 FE6A FF05 FFE0
#         * after prefix characters with LineBreak class PR and EastAsianWidth A,F,W;
#           this includes: 00A4 00B1 20AC 2116 FE69 FF04 FFE1 FFE5 FFE6


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


AI = [:LineBreak =  Ambiguous:];
AL = [[:LineBreak =  Alphabetic:]];
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:];
CP = [:LineBreak =  Close_Parenthesis:];
CR = [:LineBreak =  Carriage_Return:];
EB = [[:LineBreak =  EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
EM = [:LineBreak =  EM:];
EXX = [\uFF01 \uFF1F];
EX = [[:LineBreak =  Exclamation:] - EXX];
GL = [:LineBreak =  Glue:];
HL = [:LineBreak =  Hebrew_Letter:];
HY = [:LineBreak =  Hyphen:];
H2 = [:LineBreak =  H2:];
H3 = [:LineBreak =  H3:];
ID = [[:LineBreak =  Ideographic:] CJ];  # CSS Loose 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 \u3005 \u303B \u309D \u309E \u30FD \u30FE \u203C \u2047 \u2048 \u2049 \u30FB \uFF1A \uFF1B \uFF65];
NS = [[:LineBreak =  Nonstarter:] - NSX];
NU = [:LineBreak =  Numeric:];
OP = [:LineBreak =  Open_Punctuation:];
POX = [\u00B0 \u2030 \u2032 \u2033 \u2035 \u2103 \u2109 \uFE6A \uFF05 \uFFE0];
PO = [[:LineBreak =  Postfix_Numeric:] - POX];
PRX = [\u00A4 \u00B1 \u20AC \u2116 \uFE69 \uFF04 \uFFE1 \uFFE5 \uFFE6];
PR = [[:LineBreak =  Prefix_Numeric:] - PRX];
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:];
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 = [CM 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);

# 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;
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
#      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 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:  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* ÷;

# 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.
# LB 21a Don't break after Hebrew + Hyphen
#   HL (HY | BA) x

LB21a:       HL CM* (HY | BA | BAX) CM* [^CM CB]?;

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;  # delete this rule for CSS loose.
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 | POX) CM* (AL | HL);
LB24.3:      (AL | HL | CM) CM* (PR | PO | POX);

# Numbers. Equivalent to Tailoring example 8 from UAx 14.
#          Loose_cj tailoring: do not include $PRX at the beginning or $POX at the end.
LB25:        ((PR | PO | POX)CM*)? ((OP | HY)CM*)? NU (CM*(NU | SY | IS))* (CM*(CL | CP))? (CM*(PR | PRX | 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* [CM-ZWJ] ÷ [^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
2ca993e8	1	#
f3c0d7a5 A	2	# Copyright (C) 2016 and later: Unicode, Inc. and others.
f3c0d7a5 A	3	# License & terms of use: http://www.unicode.org/copyright.html
2ca993e8 A	4	# Copyright (c) 2016, International Business Machines Corporation and others. All Rights Reserved.
	5	#
	6	# file: line_loose_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
	17	# http://www.unicode.org/reports/tr14/
	18	# tailored as noted in 2nd paragraph below..
	19	#
	20	# This tailors the line break behavior to correspond to CSS
	21	# line-break=loose (BCP47 -u-lb-loose) 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	# * before iteration marks 3005, 303B, 309D, 309E, 30FD, 30FE (all NS)
	26	# * between characters of LineBreak class IN such as 2026
	27	# * before some centered punct 203C, 2047, 2048, 2049, 30FB, FF1A, FF1B,
	28	# FF65 (all NS) and FF01, FF1F (both EX).
	29	# * before suffix characters with LineBreak class PO and EastAsianWidth A,F,W;
	30	# this includes: 00B0 2030 2032 2033 2035 2103 2109 FE6A FF05 FFE0
	31	# * after prefix characters with LineBreak class PR and EastAsianWidth A,F,W;
	32	# this includes: 00A4 00B1 20AC 2116 FE69 FF04 FFE1 FFE5 FFE6
	33
	34
	35	type = line;
	36	locale = ja@lb=loose;
	37
	38
	39	AI = [:LineBreak = Ambiguous:];
f3c0d7a5	40	AL = [[:LineBreak = Alphabetic:]];
2ca993e8 A	41	BAX = [\u2010 \u2013];
	42	BA = [[:LineBreak = Break_After:] - BAX];
	43	BB = [:LineBreak = Break_Before:];
	44	BK = [:LineBreak = Mandatory_Break:];
	45	B2 = [:LineBreak = Break_Both:];
	46	CB = [:LineBreak = Contingent_Break:];
	47	CJ = [:LineBreak = Conditional_Japanese_Starter:];
	48	CL = [:LineBreak = Close_Punctuation:];
	49	CM = [:LineBreak = Combining_Mark:];
	50	CP = [:LineBreak = Close_Parenthesis:];
	51	CR = [:LineBreak = Carriage_Return:];
f3c0d7a5 A	52	EB = [[:LineBreak = EB:] \U0001F46A-\U0001F46D\U0001F46F\U0001F91D\U0001F93C];
f3c0d7a5 A	53	EM = [:LineBreak = EM:];
2ca993e8 A	54	EXX = [\uFF01 \uFF1F];
	55	EX = [[:LineBreak = Exclamation:] - EXX];
	56	GL = [:LineBreak = Glue:];
	57	HL = [:LineBreak = Hebrew_Letter:];
	58	HY = [:LineBreak = Hyphen:];
	59	H2 = [:LineBreak = H2:];
	60	H3 = [:LineBreak = H3:];
f3c0d7a5	61	ID = [[:LineBreak = Ideographic:] CJ]; # CSS Loose tailoring: CJ resolves to ID
2ca993e8 A	62	IN = [:LineBreak = Inseperable:];
	63	IS = [:LineBreak = Infix_Numeric:];
	64	JL = [:LineBreak = JL:];
	65	JV = [:LineBreak = JV:];
	66	JT = [:LineBreak = JT:];
	67	LF = [:LineBreak = Line_Feed:];
	68	NL = [:LineBreak = Next_Line:];
	69	NSX = [\u301C \u30A0 \u3005 \u303B \u309D \u309E \u30FD \u30FE \u203C \u2047 \u2048 \u2049 \u30FB \uFF1A \uFF1B \uFF65];
	70	NS = [[:LineBreak = Nonstarter:] - NSX];
	71	NU = [:LineBreak = Numeric:];
	72	OP = [:LineBreak = Open_Punctuation:];
	73	POX = [\u00B0 \u2030 \u2032 \u2033 \u2035 \u2103 \u2109 \uFE6A \uFF05 \uFFE0];
	74	PO = [[:LineBreak = Postfix_Numeric:] - POX];
	75	PRX = [\u00A4 \u00B1 \u20AC \u2116 \uFE69 \uFF04 \uFFE1 \uFFE5 \uFFE6];
	76	PR = [[:LineBreak = Prefix_Numeric:] - PRX];
	77	QU = [:LineBreak = Quotation:];
	78	RI = [:LineBreak = Regional_Indicator:];
	79	SA = [:LineBreak = Complex_Context:];
	80	SG = [:LineBreak = Surrogate:];
	81	SP = [:LineBreak = Space:];
	82	SY = [:LineBreak = Break_Symbols:];
	83	WJ = [:LineBreak = Word_Joiner:];
	84	XX = [:LineBreak = Unknown:];
	85	ZW = [:LineBreak = ZWSpace:];
f3c0d7a5	86	ZWJ = [:LineBreak = ZWJ:];
2ca993e8	87
f3c0d7a5 A	88	# LB1 - Resolve AI, CB, CJ, SA, SG, and XX into other line breaking classes
	89	AL = [AL AI SG XX ];
	90	dictionary = SA;
2ca993e8	91
f3c0d7a5 A	92	# By LB9, a ZWJ also behaves as a CM. Including it in the definition of CM avoids having to explicitly
	93	# list it in the numerous rules that use CM.
	94	CM = [CM ZWJ];
2ca993e8 A	95
	96	LB4: BK ÷;
	97	LB5: CR LF;
	98	LB5.1: CR ÷;
	99	LB5.2: LF ÷;
	100	LB5.3: NL ÷;
	101
	102	LB6: . (BK \| CR \| LF \| NL);
	103	LB6.1: [^BK CR LF NL SP ZW] CM* (BK \| CR \| LF \| NL);
	104
	105	# Rules LB14 - LB17.
	106	# Moved before LB7, because they can match a longer sequence that would also match LB7,
	107	# for example, the sequence "OP CM SP AL" matches LB14 while the prefix of it,
	108	# "while only the prefix "OP CM SP" matches LB7.1
	109	LB14: OP CM* SP* .;
	110	LB15: QU CM* SP* OP;
	111	LB16: (CL \| CP)CM* SP* NS;
	112	LB17: B2 CM* SP* B2;
	113
	114	LB7.1: [^ZW SP] CM* [SP ZW];
	115	LB7.2: [ZW SP] [SP ZW];
	116
	117	# LB8, ICU differs from UAX-14,
	118	# ICU: ZW ÷;
	119	# UAX 14: ZW SP* ÷;
	120	LB8: ZW ÷;
	121
f3c0d7a5	122	# LB8a
0f5d89e8 A	123	# ZWJ x
	124	# Don't match a CM on the right - let other rules pick up CM sequences, where
	125	# the ZWJ behaves as just another generic CM.
	126	LB8a: ZWJ [^CM];
2ca993e8 A	127
	128
	129	# LB9: X CM -> X
	130	# LB10: Unattached CM -> AL
	131
	132	#LB11: × WJ;
	133	# WJ ×
	134
0f5d89e8	135	LB11.1: [^SP] CM* WJ;
2ca993e8 A	136	LB11.2: SP WJ;
	137	LB11.3: WJ CM* [^CM];
	138
	139	LB12: GL CM* [^CM];
	140
	141	LB12a: [^SP BA BAX HY] CM* GL;
	142
	143	# LB 13 ICU Tailoring, matches tailoring exmaple 8 from UAX 14.
	144	#
	145	# LB13.1 [^SP] CM* [CL CP EX IS SY] # original UAX 14 rule.
	146	# LB13.2 SP CM* [CL CP EX IS SY]
	147
	148	LB13.1: [^NU SP] CM* [CL CP IS SY];
	149	LB13.2: [^SP] CM* EX;
	150	LB13.2: SP [CL CP EX IS SY];
	151
	152
	153	# LB 14-17 are moved above LB 7.
	154
	155	LB18: SP ÷;
	156
	157	LB19: . CM* QU;
	158	LB19.1: QU CM* [^CM];
	159
	160	# LB 20 Break before and after CB.
0f5d89e8	161	# Interaction with LB8a: ZWJ x . is tricky because CM includes ZWJ.
f3c0d7a5	162	# ZWJ acts like a CM to the left, combining with CB.
0f5d89e8 A	163	# ZWJ acts independently to the right, no break after by LB8a.
	164	LB20.1: . CM* ZWJ CB;
	165	LB20.2: . CM* ÷ CB;
	166
	167	LB20.3: CB CM* ZWJ [^CM];
	168	LB20.4: CB CM* ÷;
2ca993e8 A	169
	170	# Note: Rule 21a must come before 21 to prevent 21.1 from matching HL BA, then
	171	# not picking up the continuing match after the BA from 21a.
	172	# LB 21a Don't break after Hebrew + Hyphen
	173	# HL (HY \| BA) x
	174
	175	LB21a: HL CM* (HY \| BA \| BAX) CM* [^CM CB]?;
	176
	177	LB21.1: . CM* [BA HY NS];
	178	LB21.2: BB CM* [^CM CB];
	179
	180	LB21b: SY CM* HL;
	181
	182	LB22.1: (AL \| HL \| CM) CM* IN; # The CM is from LB10, treat an unattached CM as AL.
	183	LB22.2: EX CM* IN;
	184	LB22.3: (ID \| EB \| EM) CM* IN;
	185	# LB22.4: IN CM* IN; # delete this rule for CSS loose.
	186	LB22.5: NU CM* IN;
	187
f3c0d7a5 A	188	LB23.1: (AL \| HL \| CM) CM* NU;
	189	LB23.2: NU CM* (AL \| HL);
	190
	191	LB23a.1: PR CM* (ID \| EB \| EM);
	192	LB23a.2: (ID \| EB \| EM) CM* PO;
2ca993e8	193
f3c0d7a5 A	194	LB24.2: (PR \| PO \| POX) CM* (AL \| HL);
f3c0d7a5 A	195	LB24.3: (AL \| HL \| CM) CM* (PR \| PO \| POX);
2ca993e8 A	196
	197	# Numbers. Equivalent to Tailoring example 8 from UAx 14.
	198	# Loose_cj tailoring: do not include $PRX at the beginning or $POX at the end.
	199	LB25: ((PR \| PO \| POX)CM)? ((OP \| HY)CM)? NU (CM(NU \| SY \| IS)) (CM(CL \| CP))? (CM(PR \| PRX \| PO))?;
	200
	201	LB26.1: JL CM* (JL \| JV \| H2 \| H3);
	202	LB26.2: (JV \| H2) CM* (JV \| JT);
	203	LB26.3: (JT \| H3) CM* JT;
	204
	205	LB27.1: (JL \| JV \| JT \| H2 \| H3) CM* IN;
	206	LB27.2: (JL \| JV \| JT \| H2 \| H3) CM* PO;
	207	LB27.3: PR CM* (JL \| JV \| JT \| H2 \| H3);
	208
	209	# LB28 Do not break between Alphabetics.
	210	# Unattached (leading) CM treated as AL.
	211	LB28: (AL \| HL \| CM)CM* (AL \| HL);
	212
	213	LB29: IS CM* (AL \| HL);
	214
	215	# LB30 is adjusted for unattached leading CM being treated as AL.
	216	LB30.1: (AL \| CM \| HL \| NU) CM* OP;
	217	LB30.2: CP CM* (AL \| HL \| NU);
	218
0f5d89e8 A	219	# LB30a keep pairs of RI together.
	220	LB30a.1: RI CM* RI ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
	221	LB30a.2: RI CM* RI CM* [CM-ZWJ] ÷ [^BK CR LF NL SP ZW WJ CL CP EX IS SY GL QU BA HY NS CM];
	222	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	223
	224	# LB30b Do not break between Emoji Base and Emoji Modifier
	225	LB30b: EB CM* EM;
	226
	227	# LB31 Break Everywhere Else.
	228	# Include combining marks
0f5d89e8	229	LB31.1: . CM* ZWJ [^CM];
2ca993e8	230	LB31.2: . CM* ÷;