scuffed-code/icu4c/source/data/brkitr/word.txt

#
# Copyright (C) 2002-2003,
# 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 Version 4.0.0, dated 2003-04-17
#

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

!!chain;
!!LBCMNoChain;

$Katakana  = [[:Script = KATAKANA:]
			  [:name = KATAKANA-HIRAGANA PROLONGED SOUND MARK:]
			  [:name = HALFWIDTH KATAKANA-HIRAGANA PROLONGED SOUND MARK:]
			  [:name = HALFWIDTH KATAKANA VOICED SOUND MARK:]
			  [:name = HALFWIDTH KATAKANA SEMI-VOICED SOUND MARK:]];


$ALetter   = [[:Alphabetic:] [:name= HEBREW PUNCTUATION GERESH:]
						   - [:Ideographic:]
						   - $Katakana
						   - [:Script = Thai:]
						   - [:Script = Lao:]
						   - [:Script = Hiragana:]];

$ABaseLetter = [$ALetter - [:Grapheme_Extend = TRUE:]];
$ACMLetter   = [$ALetter & [:Grapheme_Extend = TRUE:]];

$MidLetter = [[:name = APOSTROPHE:] [:name = MIDDLE DOT:]
			  [:name = HEBREW PUNCTUATION GERSHAYIM:]
			  [:name = RIGHT SINGLE QUOTATION MARK:]
			  [:name = HYPHENATION POINT:]];

$MidNumLet = [[:name = FULL STOP:] [:name = COLON:]];

$MidNum    = [[:LineBreak = Infix_Numeric:] - $MidNumLet];
$Numeric   = [:LineBreak = Numeric:];

#
#  Character Class Definitions.
#    The names are those from TR29.
#

$CR      = \u000d;
$LF      = \u000a;
$Control = [[:Zl:] [:Zp:] [:Cc:] [:Cf:]];
$Extend  = [[:Grapheme_Extend = TRUE:]];
$Format  = [[:Cf:]];
$Hiragana = [:Hiragana:];
$Ideographic = [:IDEOGRAPHIC:];

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

!!forward;

$CR $LF;

# rule 3 and 4

$ALetterEx     = $ALetter     $Extend*;
$ABaseLetterEx = $ABaseLetter $Extend*;
$ACMLetterEx   = $ACMLetter   $Extend*;
$NumericEx     = $Numeric     $Extend*;
$MidNumEx      = $MidNum      $Extend*;
$MidNumLetEx   = $MidNumLet   $Extend*;
$MidLetterEx   = $MidLetter   $Extend*;
$KatakanaEx    = $Katakana    $Extend*;

# see character breaks

[^$Control] $Extend*;

# rule 5

$ALetterEx ($Format* $ALetterEx)* {200};

# rule 6 and 7

$MidALetterEx = ($ABaseLetterEx | $Format $ACMLetterEx);

$ALetterSeq =
$ALetterEx
(
    $Format* ($MidLetterEx | $MidNumLetEx) $Format* $MidALetterEx
)*;

$MidALetterSeq =
$MidALetterEx
(
    $Format* ($MidLetterEx | $MidNumLetEx) $Format* $MidALetterEx
)*;

# rule 8

$NumericEx ($Format* $NumericEx)* {100};

# rule 9

$ALetterSeq ($Format* ($NumericEx | $MidALetterSeq))* {200};

# rule 10

$NumericEx ($Format* $MidALetterSeq)+ ($Format* $NumericEx)* {200};

# rule 11 and 12 

$NumericEx ($Format* ($MidNumEx | $MidNumLetEx) $Format* $NumericEx)+ {100};

# rule 13

$KatakanaEx ($Format* $KatakanaEx)* {300};
$Hiragana $Extend* {300};
$Ideographic $Extend* {400};

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

!!reverse;

$BackALetterEx     = $Extend* $ALetter;
$BackABaseLetterEx = $Extend* $ABaseLetter;
$BackACMLetterEx   = $Extend* $ACMLetter;
$BackNumericEx     = $Extend* $Numeric;
$BackMidNumEx      = $Extend* $MidNum;
$BackMidNumLetEx   = $Extend* $MidNumLet;
$BackMidLetterEx   = $Extend* $MidLetter;
$BackKatakanaEx    = $Extend* $Katakana;

$LF $CR;

# see character breaks

$Extend* [^$Control];

# rule 5

($BackALetterEx $Format*)* $BackABaseLetterEx;
($BackALetterEx $Format*)* $BackACMLetterEx / $Control;

# rule 6 and 7

$BackMidALetterEx = ($BackABaseLetterEx | $BackACMLetterEx $Format);

$BackALetterSeq =
(
    $BackMidALetterEx $Format* ($BackMidLetterEx | $BackMidNumLetEx) $Format*
)*
$BackABaseLetterEx;

$BackMidALetterSeq =
(
    $BackMidALetterEx $Format* ($BackMidLetterEx | $BackMidNumLetEx) $Format*
)*
$BackMidALetterEx;

# rule 8

$BackNumericEx $Format* $BackNumericEx;

# rule 10

(($BackNumericEx | $BackMidALetterSeq) $Format*)* $BackALetterSeq;

# to handle letter sequences ending with a combining mark
(($BackNumericEx | $BackMidALetterSeq) $Format*)* 
(
    $BackMidALetterEx $Format* ($BackMidLetterEx | $BackMidNumLetEx) $Format*
)*
$BackACMLetterEx / $Control;

# rule 10

($BackNumericEx $Format*)* ($BackMidALetterSeq $Format*)* $BackNumericEx;

# rule 11 and 12

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

# rule 13

$BackKatakanaEx $Format* $BackKatakanaEx;

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

!!safe_reverse;

# rule 3
$Extend+ [^$Extend];

# rule 4
$Format+ $BackABaseLetterEx;
$Format+ $BackACMLetterEx / $Control;
$Format+ $BackNumericEx;
$Format+ $BackMidLetterEx;
$Format+ $BackMidNumLetEx;
$Format+ $BackMidNumEx;
$Format+ $BackKatakanaEx;

# rule 6
($MidLetter | $MidNumLet) $Format* $BackABaseLetterEx;
($MidLetter | $MidNumLet) $Format* $BackACMLetterEx / $Control;

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

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

!!safe_forward;

# rule 3
$Extend+;

# rule 4
$Format+ $ALetterEx;
$Format+ $NumericEx;
$Format+ $MidLetterEx;
$Format+ $MidNumLetEx;
$Format+ $MidNumEx;
$Format+ $KatakanaEx;

# rule 6
($MidLetter | $MidNumLet) $Format* $ALetterEx;

# rule 11
($MidNum | $MidNumLet) $Format* $NumericEx;