[apple/icu.git] / icuSources / data / brkitr / word.txt

#
# Copyright (C) 2002-2013, International Business Machines Corporation 
# and others. All Rights Reserved.
#
# file:  word.txt
#
# ICU Word Break Rules
#      See Unicode Standard Annex #29.
#      These rules are based on UAX #29 Revision 20 for Unicode Version 6.2
#
# Note:  Updates to word.txt will usually need to be merged into
#        word_POSIX.txt also.

##############################################################################
#
#  Character class definitions from TR 29
#
##############################################################################

!!chain;


#
#  Character Class Definitions.
#

$CR           = [\p{Word_Break = CR}];
$LF           = [\p{Word_Break = LF}];
$Newline      = [\p{Word_Break = Newline}];
$Extend       = [\p{Word_Break = Extend}];
$Format       = [\p{Word_Break = Format}];
$Hiragana     = [:Hiragana:];
$Katakana     = [\p{Word_Break = Katakana}];
$Han          = [:Han:];
$ALetter      = [\p{Word_Break = ALetter}];
$MidNumLet    = [\p{Word_Break = MidNumLet}];
$MidLetter    = [\p{Word_Break = MidLetter}];
$MidNum       = [\p{Word_Break = MidNum}];
$Numeric      = [\p{Word_Break = Numeric}];
$ExtendNumLet = [\p{Word_Break = ExtendNumLet}];

$RI_A        = \U0001F1E6;  # Trail ERTU
$RI_B        = \U0001F1E7;  # Trail EGR
$RI_C        = \U0001F1E8;  # Trail AHLNZ
$RI_D        = \U0001F1E9;  # Trail EK
$RI_E        = \U0001F1EA;  # Trail GS
$RI_F        = \U0001F1EB;  # Trail IR
$RI_G        = \U0001F1EC;  # Trail BR
$RI_H        = \U0001F1ED;  # Trail KU
$RI_I        = \U0001F1EE;  # Trail DLNT
$RI_J        = \U0001F1EF;  # Trail OP
$RI_K        = \U0001F1F0;  # Trail R
$RI_L        = \U0001F1F1;  # Trail B
$RI_M        = \U0001F1F2;  # Trail OXY
$RI_N        = \U0001F1F3;  # Trail LO
$RI_P        = \U0001F1F5;  # Trail LT
$RI_R        = \U0001F1F7;  # Trail OU
$RI_S        = \U0001F1F8;  # Trail AEGK
$RI_T        = \U0001F1F9;  # Trail HRW
$RI_U        = \U0001F1FA;  # Trail AS
$RI_V        = \U0001F1FB;	# Trail N

$RI_A_End    = [\U0001F1EA \U0001F1F7 \U0001F1F9 \U0001F1FA];  	# ERTU
$RI_B_End    = [\U0001F1EA \U0001F1EC \U0001F1F7];              # EGR
$RI_C_End    = [\U0001F1E6 \U0001F1ED \U0001F1F1 \U0001F1F3 \U0001F1FF]; # AHLNZ
$RI_D_End    = [\U0001F1EA \U0001F1F0];                         # EK
$RI_E_End    = [\U0001F1EC \U0001F1F8];                         # GS
$RI_F_End    = [\U0001F1EE \U0001F1F7];                         # IR
$RI_G_End    = [\U0001F1E7 \U0001F1F7];                         # BR
$RI_H_End    = [\U0001F1F0 \U0001F1FA];                         # KU
$RI_I_End    = [\U0001F1E9 \U0001F1F1 \U0001F1F3 \U0001F1F9];	# DLNT
$RI_J_End    = [\U0001F1F4 \U0001F1F5];                         # OP
$RI_K_End    = \U0001F1F7;                                      # R
$RI_L_End    = \U0001F1E7;                                      # B
$RI_M_End    = [\U0001F1F4 \U0001F1FD \U0001F1FE];              # OXY
$RI_N_End    = [\U0001F1F1 \U0001F1F4];                         # LO
$RI_P_End    = [\U0001F1F1 \U0001F1F9];                         # LT
$RI_R_End    = [\U0001F1F4 \U0001F1FA];                         # OU
$RI_S_End    = [\U0001F1E6 \U0001F1EA \U0001F1EC \U0001F1F0];   # AEGK
$RI_T_End    = [\U0001F1ED \U0001F1F7 \U0001F1FC];              # HRW
$RI_U_End    = [\U0001F1E6 \U0001F1F8];                         # AS
$RI_V_End    = \U0001F1F3;                                      # N


#   Dictionary character set, for triggering language-based break engines. Currently
#   limited to LineBreak=Complex_Context. Note that this set only works in Unicode
#   5.0 or later as the definition of Complex_Context was corrected to include all
#   characters requiring dictionary break.

$Control        = [\p{Grapheme_Cluster_Break = Control}]; 
$HangulSyllable = [\uac00-\ud7a3];
$ComplexContext = [:LineBreak = Complex_Context:];
$KanaKanji      = [$Han $Hiragana $Katakana];
$dictionaryCJK  = [$KanaKanji $HangulSyllable];
$dictionary     = [$ComplexContext $dictionaryCJK];

# leave CJK scripts out of ALetterPlus
$ALetterPlus  = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]];


#
#  Rules 4    Ignore Format and Extend characters, 
#             except when they appear at the beginning of a region of text.
#
# TODO: check if handling of katakana in dictionary makes rules incorrect/void
$KatakanaEx     = $Katakana     ($Extend |  $Format)*;
$ALetterEx      = $ALetterPlus  ($Extend |  $Format)*;
$MidNumLetEx    = $MidNumLet    ($Extend |  $Format)*;
$MidLetterEx    = $MidLetter    ($Extend |  $Format)*;
$MidNumEx       = $MidNum       ($Extend |  $Format)*;
$NumericEx      = $Numeric      ($Extend |  $Format)*;
$ExtendNumLetEx = $ExtendNumLet ($Extend |  $Format)*;

$Ideographic    = [\p{Ideographic}];
$HiraganaEx     = $Hiragana     ($Extend |  $Format)*;
$IdeographicEx  = $Ideographic  ($Extend |  $Format)*;

## -------------------------------------------------

!!forward;


# Rule 3 - CR x LF
#
$CR $LF;

# Rule 4 - ignore Format and Extend characters, except when they appear at the beginning
#          of a region of Text.   The rule here comes into play when the start of text
#          begins with a group of Format chars, or with a "word" consisting of a single
#          char that is not in any of the listed word break categories followed by
#          format char(s), or is not a CJK dictionary character.
[^$CR $LF $Newline]? ($Extend |  $Format)+;

$NumericEx {100};
$ALetterEx {200};
$HangulSyllable {200};
$KatakanaEx {400};       # note:  these status values override those from rule 5
$HiraganaEx {400};       #        by virtue of being numerically larger.
$IdeographicEx {400};    #

#
# rule 5
#    Do not break between most letters.
#
$ALetterEx $ALetterEx {200};

# rule 6 and 7
$ALetterEx ($MidLetterEx | $MidNumLetEx) $ALetterEx {200};

# rule 8

$NumericEx $NumericEx {100};

# rule 9

$ALetterEx $NumericEx {200};

# rule 10

$NumericEx $ALetterEx {200};

# rule 11 and 12 

$NumericEx ($MidNumEx | $MidNumLetEx) $NumericEx {100};

# rule 13
# to be consistent with $KanaKanji $KanaKanhi, changed
# from 300 to 400.
# See also TestRuleStatus in intltest/rbbiapts.cpp
$KatakanaEx  $KatakanaEx {400};

# rule 13a/b

$ALetterEx      $ExtendNumLetEx {200};    #  (13a)
$NumericEx      $ExtendNumLetEx {100};    #  (13a)
$KatakanaEx     $ExtendNumLetEx {400};    #  (13a)
$ExtendNumLetEx $ExtendNumLetEx {200};    #  (13a)

$ExtendNumLetEx $ALetterEx  {200};    #  (13b)
$ExtendNumLetEx $NumericEx  {100};    #  (13b)
$ExtendNumLetEx $KatakanaEx {400};    #  (13b)

# rule 13c

$RI_A ($Extend|$Format)* $RI_A_End ($Extend|$Format)*;
$RI_B ($Extend|$Format)* $RI_B_End ($Extend|$Format)*;
$RI_C ($Extend|$Format)* $RI_C_End ($Extend|$Format)*;
$RI_D ($Extend|$Format)* $RI_D_End ($Extend|$Format)*;
$RI_E ($Extend|$Format)* $RI_E_End ($Extend|$Format)*;
$RI_F ($Extend|$Format)* $RI_F_End ($Extend|$Format)*;
$RI_G ($Extend|$Format)* $RI_G_End ($Extend|$Format)*;
$RI_H ($Extend|$Format)* $RI_H_End ($Extend|$Format)*;
$RI_I ($Extend|$Format)* $RI_I_End ($Extend|$Format)*;
$RI_J ($Extend|$Format)* $RI_J_End ($Extend|$Format)*;
$RI_K ($Extend|$Format)* $RI_K_End ($Extend|$Format)*;
$RI_L ($Extend|$Format)* $RI_L_End ($Extend|$Format)*;
$RI_M ($Extend|$Format)* $RI_M_End ($Extend|$Format)*;
$RI_N ($Extend|$Format)* $RI_N_End ($Extend|$Format)*;
$RI_P ($Extend|$Format)* $RI_P_End ($Extend|$Format)*;
$RI_R ($Extend|$Format)* $RI_R_End ($Extend|$Format)*;
$RI_S ($Extend|$Format)* $RI_S_End ($Extend|$Format)*;
$RI_T ($Extend|$Format)* $RI_T_End ($Extend|$Format)*;
$RI_U ($Extend|$Format)* $RI_U_End ($Extend|$Format)*;
$RI_V ($Extend|$Format)* $RI_V_End ($Extend|$Format)*;

# special handling for CJK characters: chain for later dictionary segmentation
$HangulSyllable $HangulSyllable {200};
$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found 


## -------------------------------------------------

!!reverse;

$BackALetterEx            = ($Format | $Extend)* $ALetterPlus;
$BackMidNumLetEx          = ($Format | $Extend)* $MidNumLet;
$BackNumericEx            = ($Format | $Extend)* $Numeric;
$BackMidNumEx             = ($Format | $Extend)* $MidNum;
$BackMidLetterEx          = ($Format | $Extend)* $MidLetter;
$BackKatakanaEx           = ($Format | $Extend)* $Katakana;
$BackHiraganaEx           = ($Format | $Extend)* $Hiragana;
$BackExtendNumLetEx       = ($Format | $Extend)* $ExtendNumLet;

# rule 3
$LF $CR;

# rule 4
($Format | $Extend)*  [^$CR $LF $Newline]?;

# rule 5

$BackALetterEx $BackALetterEx;

# rule 6 and 7

$BackALetterEx ($BackMidLetterEx | $BackMidNumLetEx) $BackALetterEx;


# rule 8

$BackNumericEx $BackNumericEx;

# rule 9

$BackNumericEx $BackALetterEx;

# rule 10

$BackALetterEx $BackNumericEx;

# rule 11 and 12

$BackNumericEx ($BackMidNumEx | $BackMidNumLetEx) $BackNumericEx;

# rule 13

$BackKatakanaEx $BackKatakanaEx;

# rules 13 a/b
#
$BackExtendNumLetEx ($BackALetterEx | $BackNumericEx | $BackKatakanaEx | $BackExtendNumLetEx);
($BackALetterEx | $BackNumericEx | $BackKatakanaEx) $BackExtendNumLetEx; 

# rule 13c

($Format|$Extend)* $RI_A_End ($Format|$Extend)* $RI_A;
($Format|$Extend)* $RI_B_End ($Format|$Extend)* $RI_B;
($Format|$Extend)* $RI_C_End ($Format|$Extend)* $RI_C;
($Format|$Extend)* $RI_D_End ($Format|$Extend)* $RI_D;
($Format|$Extend)* $RI_E_End ($Format|$Extend)* $RI_E;
($Format|$Extend)* $RI_F_End ($Format|$Extend)* $RI_F;
($Format|$Extend)* $RI_G_End ($Format|$Extend)* $RI_G;
($Format|$Extend)* $RI_H_End ($Format|$Extend)* $RI_H;
($Format|$Extend)* $RI_I_End ($Format|$Extend)* $RI_I;
($Format|$Extend)* $RI_J_End ($Format|$Extend)* $RI_J;
($Format|$Extend)* $RI_K_End ($Format|$Extend)* $RI_K;
($Format|$Extend)* $RI_L_End ($Format|$Extend)* $RI_L;
($Format|$Extend)* $RI_M_End ($Format|$Extend)* $RI_M;
($Format|$Extend)* $RI_N_End ($Format|$Extend)* $RI_N;
($Format|$Extend)* $RI_P_End ($Format|$Extend)* $RI_P;
($Format|$Extend)* $RI_R_End ($Format|$Extend)* $RI_R;
($Format|$Extend)* $RI_S_End ($Format|$Extend)* $RI_S;
($Format|$Extend)* $RI_T_End ($Format|$Extend)* $RI_T;
($Format|$Extend)* $RI_U_End ($Format|$Extend)* $RI_U;
($Format|$Extend)* $RI_V_End ($Format|$Extend)* $RI_V;

# special handling for CJK characters: chain for later dictionary segmentation
$HangulSyllable $HangulSyllable;
$KanaKanji $KanaKanji; #different rule status if both kanji and kana found

## -------------------------------------------------

!!safe_reverse;

# rule 3
($Extend | $Format)+ .?;

# rule 6
($MidLetter | $MidNumLet) $BackALetterEx;

# rule 11
($MidNum | $MidNumLet) $BackNumericEx;

# For dictionary-based break
$dictionary $dictionary;

## -------------------------------------------------

!!safe_forward;

# rule 4
($Extend | $Format)+ .?;

# rule 6
($MidLetterEx | $MidNumLetEx) $ALetterEx;

# rule 11
($MidNumEx | $MidNumLetEx) $NumericEx;

# For dictionary-based break
$dictionary $dictionary;
Commit	Line	Data
b75a7d8f	1	#
51004dcb	2	# Copyright (C) 2002-2013, International Business Machines Corporation
374ca955	3	# and others. All Rights Reserved.
b75a7d8f	4	#
374ca955	5	# file: word.txt
b75a7d8f	6	#
374ca955	7	# ICU Word Break Rules
b75a7d8f	8	# See Unicode Standard Annex #29.
51004dcb	9	# These rules are based on UAX #29 Revision 20 for Unicode Version 6.2
b75a7d8f	10	#
73c04bcf	11	# Note: Updates to word.txt will usually need to be merged into
51004dcb	12	# word_POSIX.txt also.
b75a7d8f	13
374ca955	14	##############################################################################
b75a7d8f A	15	#
	16	# Character class definitions from TR 29
	17	#
374ca955 A	18	##############################################################################
	19
	20	!!chain;
	21
b75a7d8f A	22
	23	#
	24	# Character Class Definitions.
b75a7d8f	25	#
b75a7d8f	26
46f4442e A	27	$CR = [\p{Word_Break = CR}];
	28	$LF = [\p{Word_Break = LF}];
	29	$Newline = [\p{Word_Break = Newline}];
	30	$Extend = [\p{Word_Break = Extend}];
73c04bcf	31	$Format = [\p{Word_Break = Format}];
51004dcb	32	$Hiragana = [:Hiragana:];
46f4442e	33	$Katakana = [\p{Word_Break = Katakana}];
51004dcb	34	$Han = [:Han:];
73c04bcf	35	$ALetter = [\p{Word_Break = ALetter}];
46f4442e	36	$MidNumLet = [\p{Word_Break = MidNumLet}];
73c04bcf A	37	$MidLetter = [\p{Word_Break = MidLetter}];
	38	$MidNum = [\p{Word_Break = MidNum}];
	39	$Numeric = [\p{Word_Break = Numeric}];
	40	$ExtendNumLet = [\p{Word_Break = ExtendNumLet}];
	41
51004dcb A	42	$RI_A = \U0001F1E6; # Trail ERTU
	43	$RI_B = \U0001F1E7; # Trail EGR
	44	$RI_C = \U0001F1E8; # Trail AHLNZ
	45	$RI_D = \U0001F1E9; # Trail EK
	46	$RI_E = \U0001F1EA; # Trail GS
	47	$RI_F = \U0001F1EB; # Trail IR
	48	$RI_G = \U0001F1EC; # Trail BR
	49	$RI_H = \U0001F1ED; # Trail KU
	50	$RI_I = \U0001F1EE; # Trail DLNT
	51	$RI_J = \U0001F1EF; # Trail OP
	52	$RI_K = \U0001F1F0; # Trail R
	53	$RI_L = \U0001F1F1; # Trail B
	54	$RI_M = \U0001F1F2; # Trail OXY
	55	$RI_N = \U0001F1F3; # Trail LO
	56	$RI_P = \U0001F1F5; # Trail LT
	57	$RI_R = \U0001F1F7; # Trail OU
	58	$RI_S = \U0001F1F8; # Trail AEGK
	59	$RI_T = \U0001F1F9; # Trail HRW
	60	$RI_U = \U0001F1FA; # Trail AS
	61	$RI_V = \U0001F1FB; # Trail N
	62
	63	$RI_A_End = [\U0001F1EA \U0001F1F7 \U0001F1F9 \U0001F1FA]; # ERTU
	64	$RI_B_End = [\U0001F1EA \U0001F1EC \U0001F1F7]; # EGR
	65	$RI_C_End = [\U0001F1E6 \U0001F1ED \U0001F1F1 \U0001F1F3 \U0001F1FF]; # AHLNZ
	66	$RI_D_End = [\U0001F1EA \U0001F1F0]; # EK
	67	$RI_E_End = [\U0001F1EC \U0001F1F8]; # GS
	68	$RI_F_End = [\U0001F1EE \U0001F1F7]; # IR
	69	$RI_G_End = [\U0001F1E7 \U0001F1F7]; # BR
	70	$RI_H_End = [\U0001F1F0 \U0001F1FA]; # KU
	71	$RI_I_End = [\U0001F1E9 \U0001F1F1 \U0001F1F3 \U0001F1F9]; # DLNT
	72	$RI_J_End = [\U0001F1F4 \U0001F1F5]; # OP
	73	$RI_K_End = \U0001F1F7; # R
	74	$RI_L_End = \U0001F1E7; # B
	75	$RI_M_End = [\U0001F1F4 \U0001F1FD \U0001F1FE]; # OXY
	76	$RI_N_End = [\U0001F1F1 \U0001F1F4]; # LO
	77	$RI_P_End = [\U0001F1F1 \U0001F1F9]; # LT
	78	$RI_R_End = [\U0001F1F4 \U0001F1FA]; # OU
	79	$RI_S_End = [\U0001F1E6 \U0001F1EA \U0001F1EC \U0001F1F0]; # AEGK
	80	$RI_T_End = [\U0001F1ED \U0001F1F7 \U0001F1FC]; # HRW
	81	$RI_U_End = [\U0001F1E6 \U0001F1F8]; # AS
	82	$RI_V_End = \U0001F1F3; # N
	83
73c04bcf	84
73c04bcf A	85	# Dictionary character set, for triggering language-based break engines. Currently
	86	# limited to LineBreak=Complex_Context. Note that this set only works in Unicode
	87	# 5.0 or later as the definition of Complex_Context was corrected to include all
	88	# characters requiring dictionary break.
	89
46f4442e	90	$Control = [\p{Grapheme_Cluster_Break = Control}];
51004dcb A	91	$HangulSyllable = [\uac00-\ud7a3];
	92	$ComplexContext = [:LineBreak = Complex_Context:];
	93	$KanaKanji = [$Han $Hiragana $Katakana];
	94	$dictionaryCJK = [$KanaKanji $HangulSyllable];
	95	$dictionary = [$ComplexContext $dictionaryCJK];
	96
	97	# leave CJK scripts out of ALetterPlus
	98	$ALetterPlus = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]];
	99
73c04bcf A	100
73c04bcf A	101	#
46f4442e A	102	# Rules 4 Ignore Format and Extend characters,
46f4442e A	103	# except when they appear at the beginning of a region of text.
73c04bcf	104	#
51004dcb	105	# TODO: check if handling of katakana in dictionary makes rules incorrect/void
73c04bcf A	106	$KatakanaEx = $Katakana ($Extend \| $Format)*;
73c04bcf A	107	$ALetterEx = $ALetterPlus ($Extend \| $Format)*;
46f4442e	108	$MidNumLetEx = $MidNumLet ($Extend \| $Format)*;
73c04bcf A	109	$MidLetterEx = $MidLetter ($Extend \| $Format)*;
	110	$MidNumEx = $MidNum ($Extend \| $Format)*;
	111	$NumericEx = $Numeric ($Extend \| $Format)*;
	112	$ExtendNumLetEx = $ExtendNumLet ($Extend \| $Format)*;
	113
46f4442e	114	$Ideographic = [\p{Ideographic}];
73c04bcf A	115	$HiraganaEx = $Hiragana ($Extend \| $Format)*;
73c04bcf A	116	$IdeographicEx = $Ideographic ($Extend \| $Format)*;
b75a7d8f	117
374ca955	118	## -------------------------------------------------
b75a7d8f	119
374ca955	120	!!forward;
b75a7d8f	121
b75a7d8f	122
73c04bcf	123	# Rule 3 - CR x LF
46f4442e A	124	#
46f4442e A	125	$CR $LF;
73c04bcf A	126
	127	# Rule 4 - ignore Format and Extend characters, except when they appear at the beginning
	128	# of a region of Text. The rule here comes into play when the start of text
	129	# begins with a group of Format chars, or with a "word" consisting of a single
	130	# char that is not in any of the listed word break categories followed by
51004dcb	131	# format char(s), or is not a CJK dictionary character.
46f4442e	132	[^$CR $LF $Newline]? ($Extend \| $Format)+;
b75a7d8f	133
374ca955 A	134	$NumericEx {100};
374ca955 A	135	$ALetterEx {200};
51004dcb A	136	$HangulSyllable {200};
	137	$KatakanaEx {400}; # note: these status values override those from rule 5
	138	$HiraganaEx {400}; # by virtue of being numerically larger.
46f4442e	139	$IdeographicEx {400}; #
b75a7d8f	140
46f4442e	141	#
374ca955	142	# rule 5
46f4442e A	143	# Do not break between most letters.
46f4442e A	144	#
73c04bcf	145	$ALetterEx $ALetterEx {200};
b75a7d8f	146
374ca955	147	# rule 6 and 7
46f4442e	148	$ALetterEx ($MidLetterEx \| $MidNumLetEx) $ALetterEx {200};
b75a7d8f	149
374ca955	150	# rule 8
b75a7d8f	151
73c04bcf	152	$NumericEx $NumericEx {100};
b75a7d8f	153
374ca955	154	# rule 9
b75a7d8f	155
46f4442e	156	$ALetterEx $NumericEx {200};
b75a7d8f	157
374ca955	158	# rule 10
b75a7d8f	159
73c04bcf	160	$NumericEx $ALetterEx {200};
b75a7d8f	161
374ca955 A	162	# rule 11 and 12
374ca955 A	163
46f4442e	164	$NumericEx ($MidNumEx \| $MidNumLetEx) $NumericEx {100};
374ca955 A	165
374ca955 A	166	# rule 13
51004dcb A	167	# to be consistent with $KanaKanji $KanaKanhi, changed
	168	# from 300 to 400.
	169	# See also TestRuleStatus in intltest/rbbiapts.cpp
	170	$KatakanaEx $KatakanaEx {400};
374ca955 A	171
	172	# rule 13a/b
	173
73c04bcf A	174	$ALetterEx $ExtendNumLetEx {200}; # (13a)
73c04bcf A	175	$NumericEx $ExtendNumLetEx {100}; # (13a)
51004dcb	176	$KatakanaEx $ExtendNumLetEx {400}; # (13a)
46f4442e	177	$ExtendNumLetEx $ExtendNumLetEx {200}; # (13a)
374ca955	178
73c04bcf A	179	$ExtendNumLetEx $ALetterEx {200}; # (13b)
73c04bcf A	180	$ExtendNumLetEx $NumericEx {100}; # (13b)
51004dcb A	181	$ExtendNumLetEx $KatakanaEx {400}; # (13b)
	182
	183	# rule 13c
	184
	185	$RI_A ($Extend\|$Format)* $RI_A_End ($Extend\|$Format)*;
	186	$RI_B ($Extend\|$Format)* $RI_B_End ($Extend\|$Format)*;
	187	$RI_C ($Extend\|$Format)* $RI_C_End ($Extend\|$Format)*;
	188	$RI_D ($Extend\|$Format)* $RI_D_End ($Extend\|$Format)*;
	189	$RI_E ($Extend\|$Format)* $RI_E_End ($Extend\|$Format)*;
	190	$RI_F ($Extend\|$Format)* $RI_F_End ($Extend\|$Format)*;
	191	$RI_G ($Extend\|$Format)* $RI_G_End ($Extend\|$Format)*;
	192	$RI_H ($Extend\|$Format)* $RI_H_End ($Extend\|$Format)*;
	193	$RI_I ($Extend\|$Format)* $RI_I_End ($Extend\|$Format)*;
	194	$RI_J ($Extend\|$Format)* $RI_J_End ($Extend\|$Format)*;
	195	$RI_K ($Extend\|$Format)* $RI_K_End ($Extend\|$Format)*;
	196	$RI_L ($Extend\|$Format)* $RI_L_End ($Extend\|$Format)*;
	197	$RI_M ($Extend\|$Format)* $RI_M_End ($Extend\|$Format)*;
	198	$RI_N ($Extend\|$Format)* $RI_N_End ($Extend\|$Format)*;
	199	$RI_P ($Extend\|$Format)* $RI_P_End ($Extend\|$Format)*;
	200	$RI_R ($Extend\|$Format)* $RI_R_End ($Extend\|$Format)*;
	201	$RI_S ($Extend\|$Format)* $RI_S_End ($Extend\|$Format)*;
	202	$RI_T ($Extend\|$Format)* $RI_T_End ($Extend\|$Format)*;
	203	$RI_U ($Extend\|$Format)* $RI_U_End ($Extend\|$Format)*;
	204	$RI_V ($Extend\|$Format)* $RI_V_End ($Extend\|$Format)*;
	205
	206	# special handling for CJK characters: chain for later dictionary segmentation
	207	$HangulSyllable $HangulSyllable {200};
	208	$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found
374ca955 A	209
	210
	211	## -------------------------------------------------
b75a7d8f	212
374ca955 A	213	!!reverse;
374ca955 A	214
51004dcb A	215	$BackALetterEx = ($Format \| $Extend)* $ALetterPlus;
	216	$BackMidNumLetEx = ($Format \| $Extend)* $MidNumLet;
	217	$BackNumericEx = ($Format \| $Extend)* $Numeric;
	218	$BackMidNumEx = ($Format \| $Extend)* $MidNum;
	219	$BackMidLetterEx = ($Format \| $Extend)* $MidLetter;
	220	$BackKatakanaEx = ($Format \| $Extend)* $Katakana;
	221	$BackHiraganaEx = ($Format \| $Extend)* $Hiragana;
	222	$BackExtendNumLetEx = ($Format \| $Extend)* $ExtendNumLet;
374ca955	223
73c04bcf	224	# rule 3
46f4442e	225	$LF $CR;
374ca955	226
73c04bcf	227	# rule 4
46f4442e	228	($Format \| $Extend)* [^$CR $LF $Newline]?;
374ca955 A	229
	230	# rule 5
	231
73c04bcf	232	$BackALetterEx $BackALetterEx;
374ca955 A	233
	234	# rule 6 and 7
	235
46f4442e	236	$BackALetterEx ($BackMidLetterEx \| $BackMidNumLetEx) $BackALetterEx;
374ca955 A	237
	238
	239	# rule 8
	240
73c04bcf	241	$BackNumericEx $BackNumericEx;
374ca955 A	242
	243	# rule 9
	244
73c04bcf	245	$BackNumericEx $BackALetterEx;
374ca955 A	246
	247	# rule 10
	248
73c04bcf	249	$BackALetterEx $BackNumericEx;
374ca955 A	250
	251	# rule 11 and 12
	252
46f4442e	253	$BackNumericEx ($BackMidNumEx \| $BackMidNumLetEx) $BackNumericEx;
374ca955 A	254
	255	# rule 13
	256
73c04bcf	257	$BackKatakanaEx $BackKatakanaEx;
374ca955 A	258
374ca955 A	259	# rules 13 a/b
b75a7d8f	260	#
46f4442e A	261	$BackExtendNumLetEx ($BackALetterEx \| $BackNumericEx \| $BackKatakanaEx \| $BackExtendNumLetEx);
46f4442e A	262	($BackALetterEx \| $BackNumericEx \| $BackKatakanaEx) $BackExtendNumLetEx;
374ca955	263
51004dcb A	264	# rule 13c
	265
	266	($Format\|$Extend)* $RI_A_End ($Format\|$Extend)* $RI_A;
	267	($Format\|$Extend)* $RI_B_End ($Format\|$Extend)* $RI_B;
	268	($Format\|$Extend)* $RI_C_End ($Format\|$Extend)* $RI_C;
	269	($Format\|$Extend)* $RI_D_End ($Format\|$Extend)* $RI_D;
	270	($Format\|$Extend)* $RI_E_End ($Format\|$Extend)* $RI_E;
	271	($Format\|$Extend)* $RI_F_End ($Format\|$Extend)* $RI_F;
	272	($Format\|$Extend)* $RI_G_End ($Format\|$Extend)* $RI_G;
	273	($Format\|$Extend)* $RI_H_End ($Format\|$Extend)* $RI_H;
	274	($Format\|$Extend)* $RI_I_End ($Format\|$Extend)* $RI_I;
	275	($Format\|$Extend)* $RI_J_End ($Format\|$Extend)* $RI_J;
	276	($Format\|$Extend)* $RI_K_End ($Format\|$Extend)* $RI_K;
	277	($Format\|$Extend)* $RI_L_End ($Format\|$Extend)* $RI_L;
	278	($Format\|$Extend)* $RI_M_End ($Format\|$Extend)* $RI_M;
	279	($Format\|$Extend)* $RI_N_End ($Format\|$Extend)* $RI_N;
	280	($Format\|$Extend)* $RI_P_End ($Format\|$Extend)* $RI_P;
	281	($Format\|$Extend)* $RI_R_End ($Format\|$Extend)* $RI_R;
	282	($Format\|$Extend)* $RI_S_End ($Format\|$Extend)* $RI_S;
	283	($Format\|$Extend)* $RI_T_End ($Format\|$Extend)* $RI_T;
	284	($Format\|$Extend)* $RI_U_End ($Format\|$Extend)* $RI_U;
	285	($Format\|$Extend)* $RI_V_End ($Format\|$Extend)* $RI_V;
	286
	287	# special handling for CJK characters: chain for later dictionary segmentation
	288	$HangulSyllable $HangulSyllable;
	289	$KanaKanji $KanaKanji; #different rule status if both kanji and kana found
	290
374ca955 A	291	## -------------------------------------------------
	292
	293	!!safe_reverse;
	294
	295	# rule 3
73c04bcf	296	($Extend \| $Format)+ .?;
374ca955 A	297
374ca955 A	298	# rule 6
46f4442e	299	($MidLetter \| $MidNumLet) $BackALetterEx;
374ca955 A	300
374ca955 A	301	# rule 11
46f4442e	302	($MidNum \| $MidNumLet) $BackNumericEx;
73c04bcf A	303
	304	# For dictionary-based break
	305	$dictionary $dictionary;
374ca955 A	306
	307	## -------------------------------------------------
	308
	309	!!safe_forward;
	310
374ca955	311	# rule 4
73c04bcf	312	($Extend \| $Format)+ .?;
374ca955 A	313
374ca955 A	314	# rule 6
46f4442e	315	($MidLetterEx \| $MidNumLetEx) $ALetterEx;
b75a7d8f	316
374ca955	317	# rule 11
46f4442e	318	($MidNumEx \| $MidNumLetEx) $NumericEx;
73c04bcf A	319
	320	# For dictionary-based break
	321	$dictionary $dictionary;