[apple/icu.git] / icuSources / test / testdata / wordNLLTu8.txt

## wordNLLTu8.txt\r
## The following corresponds to source file from CoreLP of 2018-Jan-17:\r
##   CoreNLP/Tagger.subproj/Tagger.subproj/Source/Tagging/Data/word.txt\r
## and binary files in CoreLP of 2018-Feb-03:\r
##   CoreNLP.framework/Resources/word.dat       [for ICU before 60, including ICU 57]\r
##   CoreNLP.framework/Resources/word_icu60.dat\r
##############################################################################\r
#\r
# Copyright (C) 2002-2016, International Business Machines Corporation\r
# and others. All Rights Reserved.\r
#\r
# file:  word.txt\r
#\r
# ICU Word Break Rules\r
#      See Unicode Standard Annex #29.\r
#      These rules are based on UAX #29 Revision 27 for Unicode Version 8.0\r
#      with additions from L2/16-011R3 for Emoji sequences.\r
#\r
# Note:  Updates to word.txt will usually need to be merged into\r
#        word_POSIX.txt also.\r
\r
##############################################################################\r
#\r
#  Character class definitions from TR 29\r
#\r
##############################################################################\r
\r
!!chain;\r
\r
\r
# Smart Tokenizer customization\r
$Space		    = [\u0020];\r
$Fullstop       = [\u002E\uFE52\uFF0E];\r
$Dash           = [\u002D\uFE63\uFF0D];\r
\r
$Terminators    = [\u0021\u00A1\u00A1\u01C3\u01C3\u01C3\u203C\u2048\uFE15\uFF01\u003F];\r
$Parenthesis    = [\u0028\u005B\u007B\u0029\u005D\u007D];\r
$Punctuation    = [\u007E\u003D\u002A\u003C\u003E];\r
$OtherSingle    = [\u0060\u0300\u030F];\r
\r
$Hash           = [\u0023 \uFE5F \uFF03];\r
$At             = [\u0040 \uFE6B \uFF20];\r
$Dollar         = [\u0024];\r
\r
$Semicolon      = [\u003B];\r
$Colon          = [\u003A];\r
\r
#\r
#  Character Class Definitions.\r
#\r
\r
$E_Base = [\u261D\u26F9\u270A-\u270D\U0001F385\U0001F3C3-\U0001F3C4\U0001F3CA-\U0001F3CB\U0001F442-\U0001F443\U0001F446-\U0001F450\U0001F466-\U0001F469\U0001F46E\U0001F470-\U0001F478\U0001F47C\U0001F481-\U0001F483\U0001F485-\U0001F487\U0001F4AA\U0001F575\U0001F590\U0001F595-\U0001F596\U0001F645-\U0001F647\U0001F64B-\U0001F64F\U0001F6A3\U0001F6B4-\U0001F6B6\U0001F6C0\U0001F918];\r
$E_Modifier = [\U0001F3FB-\U0001F3FF];\r
$ZWJ = [\u200D];\r
$GAZ = [\U0001F466-\U0001F469\U0001F48B\U0001F5E8\u2764];\r
\r
$CR                 = [\p{Word_Break = CR}];\r
$LF                 = [\p{Word_Break = LF}];\r
$Newline            = [\p{Word_Break = Newline} ];\r
$Extend             = [[\p{Word_Break = Extend}][:Block=Tags:]];\r
$Regional_Indicator = [\p{Word_Break = Regional_Indicator}];\r
$Format             = [[\p{Word_Break = Format}] - [:Block=Tags:]];\r
$Katakana           = [\p{Word_Break = Katakana}];\r
$Hebrew_Letter      = [\p{Word_Break = Hebrew_Letter}];\r
$ALetter            = [\p{Word_Break = ALetter}];\r
$Single_Quote       = [\p{Word_Break = Single_Quote}\u2018\u2019];\r
$Double_Quote       = [\p{Word_Break = Double_Quote}];\r
$MidNumLet          = [\p{Word_Break = MidNumLet}-[$Fullstop $Single_Quote]];\r
$MidLetter          = [\p{Word_Break = MidLetter}-[$Single_Quote]];\r
$MidNum             = [\p{Word_Break = MidNum}];\r
$Numeric            = [\p{Word_Break = Numeric}];\r
$ExtendNumLet       = [\p{Word_Break = ExtendNumLet}];\r
\r
$Han                = [:Han:];\r
$Hiragana           = [:Hiragana:];\r
\r
\r
#   Dictionary character set, for triggering language-based break engines. Currently\r
#   limited to LineBreak=Complex_Context. Note that this set only works in Unicode\r
#   5.0 or later as the definition of Complex_Context was corrected to include all\r
#   characters requiring dictionary break.\r
\r
$Control        = [\p{Grapheme_Cluster_Break = Control}];\r
$HangulSyllable = [\uac00-\ud7a3];\r
$ComplexContext = [:LineBreak = Complex_Context:];\r
$KanaKanji      = [$Han $Hiragana $Katakana];\r
$dictionaryCJK  = [$KanaKanji $HangulSyllable];\r
$dictionary     = [$ComplexContext $dictionaryCJK];\r
\r
# leave CJK scripts out of ALetterPlus\r
$ALetterPlus  = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]];\r
\r
#\r
#  Rules 4    Ignore Format and Extend characters,\r
#             except when they appear at the beginning of a region of text.\r
#\r
# TODO: check if handling of katakana in dictionary makes rules incorrect/void\r
$KatakanaEx           = $Katakana           ($Extend |  $Format | $ZWJ)*;\r
$Hebrew_LetterEx      = $Hebrew_Letter      ($Extend |  $Format | $ZWJ)*;\r
$ALetterEx            = $ALetterPlus        ($Extend |  $Format | $ZWJ)*;\r
$Single_QuoteEx       = $Single_Quote       ($Extend |  $Format | $ZWJ)*;\r
$Double_QuoteEx       = $Double_Quote       ($Extend |  $Format | $ZWJ)*;\r
$MidNumLetEx          = $MidNumLet          ($Extend |  $Format | $ZWJ)*;\r
$MidLetterEx          = $MidLetter          ($Extend |  $Format | $ZWJ)*;\r
$MidNumEx             = $MidNum             ($Extend |  $Format | $ZWJ)*;\r
$NumericEx            = $Numeric            ($Extend |  $Format | $ZWJ)*;\r
$ExtendNumLetEx       = $ExtendNumLet       ($Extend |  $Format | $ZWJ)*;\r
$Regional_IndicatorEx = $Regional_Indicator ($Extend |  $Format | $ZWJ)*;\r
\r
$Ideographic    = [\p{Ideographic}];\r
$HiraganaEx     = $Hiragana     ($Extend |  $Format | $ZWJ)*;\r
$IdeographicEx  = $Ideographic  ($Extend |  $Format | $ZWJ)*;\r
\r
$Letter = [A-Za-z];\r
\r
$Contraction = (N | n) $Single_Quote (T | t);\r
$Yall = (Y | y) $Single_Quote (A | a) (L | l)*;\r
\r
$Time = (0 | 1 | 2)? $Numeric $Colon (0 | 1 | 2 | 3 | 4 | 5) $Numeric;\r
#$IPAddress = $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric;\r
$IPAddress = $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric;\r
\r
## -------------------------------------------------\r
\r
!!forward;\r
\r
$Space* {0};\r
\r
$Contraction {202};\r
$Yall {202};\r
\r
$Fullstop {20};\r
$Dash {21};\r
$Single_Quote {22};\r
\r
$Fullstop $Fullstop ($Fullstop*) {40};\r
$Dash $Dash ($Dash*) {40};\r
$Terminators $Terminators ($Terminators*) {40};\r
$Punctuation $Punctuation ($Punctuation*) {40};\r
$OtherSingle $OtherSingle ($OtherSingle*) {40};\r
\r
$Single_Quote $Single_Quote {40};\r
$Single_Quote $Single_Quote $Single_Quote* {40};\r
\r
$Hash ($ALetter | $Numeric)* {30};\r
$At ($ALetter | $Numeric)* {31};\r
$Dollar $ALetter ($ALetter | $Numeric)* {31};\r
\r
$Time {50};\r
$IPAddress {51};\r
\r
$ALetter / $IPAddress;\r
$ALetter / $Time;\r
\r
$IPAddress / $ALetter;\r
$Time / $ALetter;\r
\r
($Colon | $Semicolon) $Parenthesis {32};\r
\r
$Fullstop $Numeric $Numeric;\r
\r
# Rule 3 - CR x LF\r
#\r
$CR $LF;\r
\r
# Rule 3c   ZWJ x GAZ.  Preceeds WB4, so no intervening Extend chars allowed.\r
#\r
$ZWJ $GAZ;\r
\r
\r
# Rule 4 - ignore Format and Extend characters, except when they appear at the beginning\r
#          of a region of Text.   The rule here comes into play when the start of text\r
#          begins with a group of Format chars, or with a "word" consisting of a single\r
#          char that is not in any of the listed word break categories followed by\r
#          format char(s), or is not a CJK dictionary character.\r
[^$CR $LF $Newline]? ($Extend |  $Format | $ZWJ)+;\r
\r
$NumericEx {100};\r
$ALetterEx {200};\r
$HangulSyllable {200};\r
$Hebrew_LetterEx{200};\r
$KatakanaEx {400};       # note:  these status values override those from rule 5\r
$HiraganaEx {400};       #        by virtue of being numerically larger.\r
$IdeographicEx {400};    #\r
\r
$E_Base ($Extend | $Format | $ZWJ)*;\r
$E_Modifier ($Extend | $Format | $ZWJ)*;\r
$GAZ ($Extend | $Format | $ZWJ)*;\r
\r
#\r
# rule 5\r
#    Do not break between most letters.\r
#\r
($ALetterEx | $Hebrew_LetterEx)  ($ALetterEx | $Hebrew_LetterEx) {200};\r
\r
# rule 6 and 7\r
#($ALetterEx | $Hebrew_LetterEx) ($MidLetterEx | $MidNumLetEx | $Single_QuoteEx) ($ALetterEx | $Hebrew_LetterEx) {200};\r
($ALetterEx | $Hebrew_LetterEx) ($MidLetterEx | $MidNumLetEx) ($ALetterEx | $Hebrew_LetterEx) {200};\r
\r
# rule 7a\r
#$Hebrew_LetterEx $Single_QuoteEx {200};\r
\r
# rule 7b and 7c\r
$Hebrew_LetterEx $Double_QuoteEx $Hebrew_LetterEx {200};\r
\r
# rule 8\r
\r
$NumericEx $NumericEx {100};\r
\r
# rule 9\r
\r
($ALetterEx | $Hebrew_LetterEx) $NumericEx {200};\r
\r
# rule 10\r
\r
$NumericEx ($ALetterEx | $Hebrew_LetterEx) {200};\r
\r
# rule 11 and 12\r
\r
$NumericEx ($MidNumEx | $MidNumLetEx | $Single_QuoteEx) $NumericEx {100};\r
\r
# rule 13\r
# to be consistent with $KanaKanji $KanaKanhi, changed\r
# from 300 to 400.\r
# See also TestRuleStatus in intltest/rbbiapts.cpp\r
$KatakanaEx  $KatakanaEx {400};\r
\r
# rule 13a/b\r
\r
$ALetterEx       $ExtendNumLetEx {200};    #  (13a)\r
$Hebrew_LetterEx $ExtendNumLetEx {200};    #  (13a)\r
$NumericEx       $ExtendNumLetEx {100};    #  (13a)\r
$KatakanaEx      $ExtendNumLetEx {400};    #  (13a)\r
$ExtendNumLetEx  $ExtendNumLetEx {200};    #  (13a)\r
\r
$ExtendNumLetEx  $ALetterEx      {200};    #  (13b)\r
$ExtendNumLetEx  $Hebrew_Letter  {200};    #  (13b)\r
$ExtendNumLetEx  $NumericEx      {100};    #  (13b)\r
$ExtendNumLetEx  $KatakanaEx     {400};    #  (13b)\r
\r
# rule 13c\r
#    Pairs of Regional Indicators stay together.\r
#    With rule chaining disabled by ^, this rule will match exactly two of them.\r
#    No other rule begins with a Regional_Indicator, so chaining cannot extend the match.\r
#\r
^$Regional_IndicatorEx $Regional_IndicatorEx;\r
\r
# special handling for CJK characters: chain for later dictionary segmentation\r
$HangulSyllable $HangulSyllable {200};\r
$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found\r
\r
# rule 13d\r
#    E_Base x E_Modifier\r
#\r
($E_Base | $GAZ) ($Format | $Extend | $ZWJ)* $E_Modifier;\r
\r
## -------------------------------------------------\r
\r
!!reverse;\r
\r
$BackHebrew_LetterEx      = ($Format | $Extend | $ZWJ)* $Hebrew_Letter;\r
$BackALetterEx            = ($Format | $Extend | $ZWJ)* $ALetterPlus;\r
$BackSingle_QuoteEx       = ($Format | $Extend | $ZWJ)* $Single_Quote;\r
$BackDouble_QuoteEx       = ($Format | $Extend | $ZWJ)* $Double_Quote;\r
$BackMidNumLetEx          = ($Format | $Extend | $ZWJ)* $MidNumLet;\r
$BackNumericEx            = ($Format | $Extend | $ZWJ)* $Numeric;\r
$BackMidNumEx             = ($Format | $Extend | $ZWJ)* $MidNum;\r
$BackMidLetterEx          = ($Format | $Extend | $ZWJ)* $MidLetter;\r
$BackKatakanaEx           = ($Format | $Extend | $ZWJ)* $Katakana;\r
$BackHiraganaEx           = ($Format | $Extend | $ZWJ)* $Hiragana;\r
$BackExtendNumLetEx       = ($Format | $Extend | $ZWJ)* $ExtendNumLet;\r
$BackRegional_IndicatorEx = ($Format | $Extend | $ZWJ)* $Regional_Indicator;\r
\r
# rule 3\r
$LF $CR;\r
\r
# Rule 3c   ZWJ x GAZ.  Preceeds WB4, so no intervening Extend chars allowed.\r
#\r
$GAZ $ZWJ;\r
\r
# rule 4\r
($Format | $Extend | $ZWJ)*  [^$CR $LF $Newline]?;\r
\r
# rule 5\r
\r
($BackALetterEx | $BackHebrew_LetterEx) ($BackALetterEx | $BackHebrew_LetterEx);\r
\r
# rule 6 and 7\r
\r
#($BackALetterEx | $BackHebrew_LetterEx) ($BackMidLetterEx | $BackMidNumLetEx | $BackSingle_QuoteEx) ($BackALetterEx | $BackHebrew_LetterEx);\r
($BackALetterEx | $BackHebrew_LetterEx) ($BackMidLetterEx | $BackMidNumLetEx) ($BackALetterEx | $BackHebrew_LetterEx);\r
\r
# rule 7a\r
#$BackSingle_QuoteEx $BackHebrew_LetterEx;\r
\r
# Rule 7b and 7c\r
$BackHebrew_LetterEx $BackDouble_QuoteEx $BackHebrew_LetterEx;\r
\r
# rule 8\r
\r
$BackNumericEx $BackNumericEx;\r
\r
# rule 9\r
\r
$BackNumericEx ($BackALetterEx | $BackHebrew_LetterEx);\r
\r
# rule 10\r
\r
($BackALetterEx | $BackHebrew_LetterEx) $BackNumericEx;\r
\r
# rule 11 and 12\r
\r
$BackNumericEx ($BackMidNumEx | $BackMidNumLetEx | $BackSingle_QuoteEx) $BackNumericEx;\r
\r
# rule 13\r
\r
$BackKatakanaEx $BackKatakanaEx;\r
\r
# rules 13 a/b\r
#\r
$BackExtendNumLetEx ($BackALetterEx | $BackHebrew_LetterEx | $BackNumericEx | $BackKatakanaEx | $BackExtendNumLetEx);\r
($BackALetterEx | $BackHebrew_LetterEx | $BackNumericEx | $BackKatakanaEx) $BackExtendNumLetEx;\r
\r
# rule 13c\r
\r
^$BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
($Format | $Extend | $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
^$BackRegional_IndicatorEx $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
($Format | $Extend | $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
\r
$GAZ $ZWJ $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
($Format | $Extend | $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
$GAZ $ZWJ $BackRegional_IndicatorEx $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
($Format | $Extend | $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
\r
# special handling for CJK characters: chain for later dictionary segmentation\r
$HangulSyllable $HangulSyllable;\r
$KanaKanji $KanaKanji; #different rule status if both kanji and kana found\r
\r
# rule 13d\r
\r
$E_Modifier ($Format | $Extend | $ZWJ)* ($E_Base | $GAZ);\r
\r
\r
\r
## -------------------------------------------------\r
\r
!!safe_reverse;\r
\r
\r
# rule 3\r
($Extend | $Format | $ZWJ)+ .?;\r
\r
# rule 6\r
#($MidLetter | $MidNumLet | $Single_Quote) ($BackALetterEx | $BackHebrew_LetterEx);\r
($MidLetter | $MidNumLet) ($BackALetterEx | $BackHebrew_LetterEx);\r
\r
# rule 7b\r
$Double_Quote $BackHebrew_LetterEx;\r
\r
\r
# rule 11\r
#($MidNum | $MidNumLet | $Single_Quote) $BackNumericEx;\r
($MidNum | $MidNumLet) $BackNumericEx;\r
\r
# rule 13c\r
$BackRegional_IndicatorEx*;\r
\r
# For dictionary-based break\r
$dictionary $dictionary;\r
\r
## -------------------------------------------------\r
\r
!!safe_forward;\r
\r
\r
# rule 4\r
($Extend | $Format | $ZWJ)+ .?;\r
\r
# rule 6\r
#($MidLetterEx | $MidNumLetEx | $Single_QuoteEx) ($ALetterEx | $Hebrew_LetterEx);\r
($MidLetterEx | $MidNumLetEx) ($ALetterEx | $Hebrew_LetterEx);\r
\r
# rule 7b\r
$Double_QuoteEx $Hebrew_LetterEx;\r
\r
# rule 11\r
#($MidNumEx | $MidNumLetEx | $Single_QuoteEx) $NumericEx;\r
($MidNumEx | $MidNumLetEx) $NumericEx;\r
\r
# rule 13c\r
$Regional_IndicatorEx*;\r
\r
# For dictionary-based break\r
$dictionary $dictionary;\r
Commit	Line	Data
0f5d89e8 A	1	## wordNLLTu8.txt\r
	2	## The following corresponds to source file from CoreLP of 2018-Jan-17:\r
	3	## CoreNLP/Tagger.subproj/Tagger.subproj/Source/Tagging/Data/word.txt\r
	4	## and binary files in CoreLP of 2018-Feb-03:\r
	5	## CoreNLP.framework/Resources/word.dat [for ICU before 60, including ICU 57]\r
	6	## CoreNLP.framework/Resources/word_icu60.dat\r
	7	##############################################################################\r
	8	#\r
	9	# Copyright (C) 2002-2016, International Business Machines Corporation\r
	10	# and others. All Rights Reserved.\r
	11	#\r
	12	# file: word.txt\r
	13	#\r
	14	# ICU Word Break Rules\r
	15	# See Unicode Standard Annex #29.\r
	16	# These rules are based on UAX #29 Revision 27 for Unicode Version 8.0\r
	17	# with additions from L2/16-011R3 for Emoji sequences.\r
	18	#\r
	19	# Note: Updates to word.txt will usually need to be merged into\r
	20	# word_POSIX.txt also.\r
	21	\r
	22	##############################################################################\r
	23	#\r
	24	# Character class definitions from TR 29\r
	25	#\r
	26	##############################################################################\r
	27	\r
	28	!!chain;\r
	29	\r
	30	\r
	31	# Smart Tokenizer customization\r
	32	$Space = [\u0020];\r
	33	$Fullstop = [\u002E\uFE52\uFF0E];\r
	34	$Dash = [\u002D\uFE63\uFF0D];\r
	35	\r
	36	$Terminators = [\u0021\u00A1\u00A1\u01C3\u01C3\u01C3\u203C\u2048\uFE15\uFF01\u003F];\r
	37	$Parenthesis = [\u0028\u005B\u007B\u0029\u005D\u007D];\r
	38	$Punctuation = [\u007E\u003D\u002A\u003C\u003E];\r
	39	$OtherSingle = [\u0060\u0300\u030F];\r
	40	\r
	41	$Hash = [\u0023 \uFE5F \uFF03];\r
	42	$At = [\u0040 \uFE6B \uFF20];\r
	43	$Dollar = [\u0024];\r
	44	\r
	45	$Semicolon = [\u003B];\r
	46	$Colon = [\u003A];\r
	47	\r
	48	#\r
	49	# Character Class Definitions.\r
	50	#\r
	51	\r
	52	$E_Base = [\u261D\u26F9\u270A-\u270D\U0001F385\U0001F3C3-\U0001F3C4\U0001F3CA-\U0001F3CB\U0001F442-\U0001F443\U0001F446-\U0001F450\U0001F466-\U0001F469\U0001F46E\U0001F470-\U0001F478\U0001F47C\U0001F481-\U0001F483\U0001F485-\U0001F487\U0001F4AA\U0001F575\U0001F590\U0001F595-\U0001F596\U0001F645-\U0001F647\U0001F64B-\U0001F64F\U0001F6A3\U0001F6B4-\U0001F6B6\U0001F6C0\U0001F918];\r
	53	$E_Modifier = [\U0001F3FB-\U0001F3FF];\r
	54	$ZWJ = [\u200D];\r
	55	$GAZ = [\U0001F466-\U0001F469\U0001F48B\U0001F5E8\u2764];\r
	56	\r
	57	$CR = [\p{Word_Break = CR}];\r
	58	$LF = [\p{Word_Break = LF}];\r
	59	$Newline = [\p{Word_Break = Newline} ];\r
	60	$Extend = [[\p{Word_Break = Extend}][:Block=Tags:]];\r
	61	$Regional_Indicator = [\p{Word_Break = Regional_Indicator}];\r
	62	$Format = [[\p{Word_Break = Format}] - [:Block=Tags:]];\r
	63	$Katakana = [\p{Word_Break = Katakana}];\r
	64	$Hebrew_Letter = [\p{Word_Break = Hebrew_Letter}];\r
65	$ALetter = [\p{Word_Break = ALetter}];\r
66	$Single_Quote = [\p{Word_Break = Single_Quote}\u2018\u2019];\r
67	$Double_Quote = [\p{Word_Break = Double_Quote}];\r
68	$MidNumLet = [\p{Word_Break = MidNumLet}-[$Fullstop $Single_Quote]];\r
69	$MidLetter = [\p{Word_Break = MidLetter}-[$Single_Quote]];\r
70	$MidNum = [\p{Word_Break = MidNum}];\r
71	$Numeric = [\p{Word_Break = Numeric}];\r
72	$ExtendNumLet = [\p{Word_Break = ExtendNumLet}];\r
73	\r
74	$Han = [:Han:];\r
75	$Hiragana = [:Hiragana:];\r
76	\r
77	\r
78	# Dictionary character set, for triggering language-based break engines. Currently\r
79	# limited to LineBreak=Complex_Context. Note that this set only works in Unicode\r
80	# 5.0 or later as the definition of Complex_Context was corrected to include all\r
81	# characters requiring dictionary break.\r
82	\r
83	$Control = [\p{Grapheme_Cluster_Break = Control}];\r
84	$HangulSyllable = [\uac00-\ud7a3];\r
85	$ComplexContext = [:LineBreak = Complex_Context:];\r
86	$KanaKanji = [$Han $Hiragana $Katakana];\r
87	$dictionaryCJK = [$KanaKanji $HangulSyllable];\r
88	$dictionary = [$ComplexContext $dictionaryCJK];\r
89	\r
90	# leave CJK scripts out of ALetterPlus\r
91	$ALetterPlus = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]];\r
92	\r
93	#\r
94	# Rules 4 Ignore Format and Extend characters,\r
95	# except when they appear at the beginning of a region of text.\r
96	#\r
97	# TODO: check if handling of katakana in dictionary makes rules incorrect/void\r
98	$KatakanaEx = $Katakana ($Extend \| $Format \| $ZWJ)*;\r
99	$Hebrew_LetterEx = $Hebrew_Letter ($Extend \| $Format \| $ZWJ)*;\r
100	$ALetterEx = $ALetterPlus ($Extend \| $Format \| $ZWJ)*;\r
101	$Single_QuoteEx = $Single_Quote ($Extend \| $Format \| $ZWJ)*;\r
102	$Double_QuoteEx = $Double_Quote ($Extend \| $Format \| $ZWJ)*;\r
103	$MidNumLetEx = $MidNumLet ($Extend \| $Format \| $ZWJ)*;\r
104	$MidLetterEx = $MidLetter ($Extend \| $Format \| $ZWJ)*;\r
105	$MidNumEx = $MidNum ($Extend \| $Format \| $ZWJ)*;\r
106	$NumericEx = $Numeric ($Extend \| $Format \| $ZWJ)*;\r
107	$ExtendNumLetEx = $ExtendNumLet ($Extend \| $Format \| $ZWJ)*;\r
108	$Regional_IndicatorEx = $Regional_Indicator ($Extend \| $Format \| $ZWJ)*;\r
109	\r
110	$Ideographic = [\p{Ideographic}];\r
111	$HiraganaEx = $Hiragana ($Extend \| $Format \| $ZWJ)*;\r
112	$IdeographicEx = $Ideographic ($Extend \| $Format \| $ZWJ)*;\r
113	\r
114	$Letter = [A-Za-z];\r
115	\r
116	$Contraction = (N \| n) $Single_Quote (T \| t);\r
117	$Yall = (Y \| y) $Single_Quote (A \| a) (L \| l)*;\r
118	\r
119	$Time = (0 \| 1 \| 2)? $Numeric $Colon (0 \| 1 \| 2 \| 3 \| 4 \| 5) $Numeric;\r
120	#$IPAddress = $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric;\r
121	$IPAddress = $Numeric? $Numeric? $Numeric $Fullstop $Numeric? $Numeric? $Numeric;\r
122	\r
123	## -------------------------------------------------\r
124	\r
125	!!forward;\r
126	\r
127	$Space* {0};\r
128	\r
129	$Contraction {202};\r
130	$Yall {202};\r
131	\r
132	$Fullstop {20};\r
133	$Dash {21};\r
134	$Single_Quote {22};\r
135	\r
136	$Fullstop $Fullstop ($Fullstop*) {40};\r
137	$Dash $Dash ($Dash*) {40};\r
138	$Terminators $Terminators ($Terminators*) {40};\r
139	$Punctuation $Punctuation ($Punctuation*) {40};\r
140	$OtherSingle $OtherSingle ($OtherSingle*) {40};\r
141	\r
142	$Single_Quote $Single_Quote {40};\r
143	$Single_Quote $Single_Quote $Single_Quote* {40};\r
144	\r
145	$Hash ($ALetter \| $Numeric)* {30};\r
146	$At ($ALetter \| $Numeric)* {31};\r
147	$Dollar $ALetter ($ALetter \| $Numeric)* {31};\r
148	\r
149	$Time {50};\r
150	$IPAddress {51};\r
151	\r
152	$ALetter / $IPAddress;\r
153	$ALetter / $Time;\r
154	\r
155	$IPAddress / $ALetter;\r
156	$Time / $ALetter;\r
157	\r
158	($Colon \| $Semicolon) $Parenthesis {32};\r
159	\r
160	$Fullstop $Numeric $Numeric;\r
161	\r
162	# Rule 3 - CR x LF\r
163	#\r
164	$CR $LF;\r
165	\r
166	# Rule 3c ZWJ x GAZ. Preceeds WB4, so no intervening Extend chars allowed.\r
167	#\r
168	$ZWJ $GAZ;\r
169	\r
170	\r
171	# Rule 4 - ignore Format and Extend characters, except when they appear at the beginning\r
172	# of a region of Text. The rule here comes into play when the start of text\r
173	# begins with a group of Format chars, or with a "word" consisting of a single\r
174	# char that is not in any of the listed word break categories followed by\r
175	# format char(s), or is not a CJK dictionary character.\r
176	[^$CR $LF $Newline]? ($Extend \| $Format \| $ZWJ)+;\r
177	\r
178	$NumericEx {100};\r
179	$ALetterEx {200};\r
180	$HangulSyllable {200};\r
181	$Hebrew_LetterEx{200};\r
182	$KatakanaEx {400}; # note: these status values override those from rule 5\r
183	$HiraganaEx {400}; # by virtue of being numerically larger.\r
184	$IdeographicEx {400}; #\r
185	\r
186	$E_Base ($Extend \| $Format \| $ZWJ)*;\r
187	$E_Modifier ($Extend \| $Format \| $ZWJ)*;\r
188	$GAZ ($Extend \| $Format \| $ZWJ)*;\r
189	\r
190	#\r
191	# rule 5\r
192	# Do not break between most letters.\r
193	#\r
194	($ALetterEx \| $Hebrew_LetterEx) ($ALetterEx \| $Hebrew_LetterEx) {200};\r
195	\r
196	# rule 6 and 7\r
197	#($ALetterEx \| $Hebrew_LetterEx) ($MidLetterEx \| $MidNumLetEx \| $Single_QuoteEx) ($ALetterEx \| $Hebrew_LetterEx) {200};\r
198	($ALetterEx \| $Hebrew_LetterEx) ($MidLetterEx \| $MidNumLetEx) ($ALetterEx \| $Hebrew_LetterEx) {200};\r
199	\r
200	# rule 7a\r
201	#$Hebrew_LetterEx $Single_QuoteEx {200};\r
202	\r
203	# rule 7b and 7c\r
204	$Hebrew_LetterEx $Double_QuoteEx $Hebrew_LetterEx {200};\r
205	\r
206	# rule 8\r
207	\r
208	$NumericEx $NumericEx {100};\r
209	\r
210	# rule 9\r
211	\r
212	($ALetterEx \| $Hebrew_LetterEx) $NumericEx {200};\r
213	\r
214	# rule 10\r
215	\r
216	$NumericEx ($ALetterEx \| $Hebrew_LetterEx) {200};\r
217	\r
218	# rule 11 and 12\r
219	\r
220	$NumericEx ($MidNumEx \| $MidNumLetEx \| $Single_QuoteEx) $NumericEx {100};\r
221	\r
222	# rule 13\r
223	# to be consistent with $KanaKanji $KanaKanhi, changed\r
224	# from 300 to 400.\r
225	# See also TestRuleStatus in intltest/rbbiapts.cpp\r
226	$KatakanaEx $KatakanaEx {400};\r
227	\r
228	# rule 13a/b\r
229	\r
230	$ALetterEx $ExtendNumLetEx {200}; # (13a)\r
231	$Hebrew_LetterEx $ExtendNumLetEx {200}; # (13a)\r
232	$NumericEx $ExtendNumLetEx {100}; # (13a)\r
233	$KatakanaEx $ExtendNumLetEx {400}; # (13a)\r
234	$ExtendNumLetEx $ExtendNumLetEx {200}; # (13a)\r
235	\r
236	$ExtendNumLetEx $ALetterEx {200}; # (13b)\r
237	$ExtendNumLetEx $Hebrew_Letter {200}; # (13b)\r
238	$ExtendNumLetEx $NumericEx {100}; # (13b)\r
239	$ExtendNumLetEx $KatakanaEx {400}; # (13b)\r
240	\r
241	# rule 13c\r
242	# Pairs of Regional Indicators stay together.\r
243	# With rule chaining disabled by ^, this rule will match exactly two of them.\r
244	# No other rule begins with a Regional_Indicator, so chaining cannot extend the match.\r
245	#\r
246	^$Regional_IndicatorEx $Regional_IndicatorEx;\r
247	\r
248	# special handling for CJK characters: chain for later dictionary segmentation\r
249	$HangulSyllable $HangulSyllable {200};\r
250	$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found\r
251	\r
252	# rule 13d\r
253	# E_Base x E_Modifier\r
254	#\r
255	($E_Base \| $GAZ) ($Format \| $Extend \| $ZWJ)* $E_Modifier;\r
256	\r
257	## -------------------------------------------------\r
258	\r
259	!!reverse;\r
260	\r
261	$BackHebrew_LetterEx = ($Format \| $Extend \| $ZWJ)* $Hebrew_Letter;\r
262	$BackALetterEx = ($Format \| $Extend \| $ZWJ)* $ALetterPlus;\r
263	$BackSingle_QuoteEx = ($Format \| $Extend \| $ZWJ)* $Single_Quote;\r
264	$BackDouble_QuoteEx = ($Format \| $Extend \| $ZWJ)* $Double_Quote;\r
265	$BackMidNumLetEx = ($Format \| $Extend \| $ZWJ)* $MidNumLet;\r
266	$BackNumericEx = ($Format \| $Extend \| $ZWJ)* $Numeric;\r
267	$BackMidNumEx = ($Format \| $Extend \| $ZWJ)* $MidNum;\r
268	$BackMidLetterEx = ($Format \| $Extend \| $ZWJ)* $MidLetter;\r
269	$BackKatakanaEx = ($Format \| $Extend \| $ZWJ)* $Katakana;\r
270	$BackHiraganaEx = ($Format \| $Extend \| $ZWJ)* $Hiragana;\r
271	$BackExtendNumLetEx = ($Format \| $Extend \| $ZWJ)* $ExtendNumLet;\r
272	$BackRegional_IndicatorEx = ($Format \| $Extend \| $ZWJ)* $Regional_Indicator;\r
273	\r
274	# rule 3\r
275	$LF $CR;\r
276	\r
277	# Rule 3c ZWJ x GAZ. Preceeds WB4, so no intervening Extend chars allowed.\r
278	#\r
279	$GAZ $ZWJ;\r
280	\r
281	# rule 4\r
282	($Format \| $Extend \| $ZWJ)* [^$CR $LF $Newline]?;\r
283	\r
284	# rule 5\r
285	\r
286	($BackALetterEx \| $BackHebrew_LetterEx) ($BackALetterEx \| $BackHebrew_LetterEx);\r
287	\r
288	# rule 6 and 7\r
289	\r
290	#($BackALetterEx \| $BackHebrew_LetterEx) ($BackMidLetterEx \| $BackMidNumLetEx \| $BackSingle_QuoteEx) ($BackALetterEx \| $BackHebrew_LetterEx);\r
291	($BackALetterEx \| $BackHebrew_LetterEx) ($BackMidLetterEx \| $BackMidNumLetEx) ($BackALetterEx \| $BackHebrew_LetterEx);\r
292	\r
293	# rule 7a\r
294	#$BackSingle_QuoteEx $BackHebrew_LetterEx;\r
295	\r
296	# Rule 7b and 7c\r
297	$BackHebrew_LetterEx $BackDouble_QuoteEx $BackHebrew_LetterEx;\r
298	\r
299	# rule 8\r
300	\r
301	$BackNumericEx $BackNumericEx;\r
302	\r
303	# rule 9\r
304	\r
305	$BackNumericEx ($BackALetterEx \| $BackHebrew_LetterEx);\r
306	\r
307	# rule 10\r
308	\r
309	($BackALetterEx \| $BackHebrew_LetterEx) $BackNumericEx;\r
310	\r
311	# rule 11 and 12\r
312	\r
313	$BackNumericEx ($BackMidNumEx \| $BackMidNumLetEx \| $BackSingle_QuoteEx) $BackNumericEx;\r
314	\r
315	# rule 13\r
316	\r
317	$BackKatakanaEx $BackKatakanaEx;\r
318	\r
319	# rules 13 a/b\r
320	#\r
321	$BackExtendNumLetEx ($BackALetterEx \| $BackHebrew_LetterEx \| $BackNumericEx \| $BackKatakanaEx \| $BackExtendNumLetEx);\r
322	($BackALetterEx \| $BackHebrew_LetterEx \| $BackNumericEx \| $BackKatakanaEx) $BackExtendNumLetEx;\r
323	\r
324	# rule 13c\r
325	\r
326	^$BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
327	($Format \| $Extend \| $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
328	^$BackRegional_IndicatorEx $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
329	($Format \| $Extend \| $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
330	\r
331	$GAZ $ZWJ $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
332	($Format \| $Extend \| $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
333	$GAZ $ZWJ $BackRegional_IndicatorEx $BackRegional_IndicatorEx / ($BackRegional_IndicatorEx $BackRegional_IndicatorEx)*\r
334	($Format \| $Extend \| $ZWJ)* [[^$Regional_Indicator $Format $Extend $ZWJ] {eof}];\r
335	\r
336	# special handling for CJK characters: chain for later dictionary segmentation\r
337	$HangulSyllable $HangulSyllable;\r
338	$KanaKanji $KanaKanji; #different rule status if both kanji and kana found\r
339	\r
340	# rule 13d\r
341	\r
342	$E_Modifier ($Format \| $Extend \| $ZWJ)* ($E_Base \| $GAZ);\r
343	\r
344	\r
345	\r
346	## -------------------------------------------------\r
347	\r
348	!!safe_reverse;\r
349	\r
350	\r
351	# rule 3\r
352	($Extend \| $Format \| $ZWJ)+ .?;\r
353	\r
354	# rule 6\r
355	#($MidLetter \| $MidNumLet \| $Single_Quote) ($BackALetterEx \| $BackHebrew_LetterEx);\r
356	($MidLetter \| $MidNumLet) ($BackALetterEx \| $BackHebrew_LetterEx);\r
357	\r
358	# rule 7b\r
359	$Double_Quote $BackHebrew_LetterEx;\r
360	\r
361	\r
362	# rule 11\r
363	#($MidNum \| $MidNumLet \| $Single_Quote) $BackNumericEx;\r
364	($MidNum \| $MidNumLet) $BackNumericEx;\r
365	\r
366	# rule 13c\r
367	$BackRegional_IndicatorEx*;\r
368	\r
369	# For dictionary-based break\r
370	$dictionary $dictionary;\r
371	\r
372	## -------------------------------------------------\r
373	\r
374	!!safe_forward;\r
375	\r
376	\r
377	# rule 4\r
378	($Extend \| $Format \| $ZWJ)+ .?;\r
379	\r
380	# rule 6\r
381	#($MidLetterEx \| $MidNumLetEx \| $Single_QuoteEx) ($ALetterEx \| $Hebrew_LetterEx);\r
382	($MidLetterEx \| $MidNumLetEx) ($ALetterEx \| $Hebrew_LetterEx);\r
383	\r
384	# rule 7b\r
385	$Double_QuoteEx $Hebrew_LetterEx;\r
386	\r
387	# rule 11\r
388	#($MidNumEx \| $MidNumLetEx \| $Single_QuoteEx) $NumericEx;\r
389	($MidNumEx \| $MidNumLetEx) $NumericEx;\r
390	\r
391	# rule 13c\r
392	$Regional_IndicatorEx*;\r
393	\r
394	# For dictionary-based break\r
395	$dictionary $dictionary;\r