| diff --git a/source/data/brkitr/rules/word.txt b/source/data/brkitr/rules/word.txt |
| index aa7c47c..9c93dd5 100644 |
| --- a/source/data/brkitr/rules/word.txt |
| +++ b/source/data/brkitr/rules/word.txt |
| @@ -39,10 +39,16 @@ $Hebrew_Letter = [\p{Word_Break = Hebrew_Letter}]; |
| $ALetter = [\p{Word_Break = ALetter}]; |
| $Single_Quote = [\p{Word_Break = Single_Quote}]; |
| $Double_Quote = [\p{Word_Break = Double_Quote}]; |
| -$MidNumLet = [\p{Word_Break = MidNumLet}]; |
| +# Remove two full stop characters from $MidNumLet and add them to $MidNum |
| +# to break a hostname into its components at the cost of breaking |
| +# 'e.g.' and 'i.e.' as well. |
| +# $MidNumLet is used in rules 6/7 (rules of our interest) and rules 11/12. |
| +# Because it's OR'd with $MidNum in rules 11/12, rules 11/12 are not affected |
| +# while rules 6/7 are reverted to the old behavior we want. |
| +$MidNumLet = [[\p{Word_Break = MidNumLet}] - [\u002E \uFF0E]]; |
| $MidLetter = [\p{Word_Break = MidLetter}]; |
| -$MidNum = [\p{Word_Break = MidNum}]; |
| -$Numeric = [\p{Word_Break = Numeric}]; |
| +$MidNum = [\p{Word_Break = MidNum}[\u002E \uFF0E]]; |
| +$Numeric = [\p{Word_Break = Numeric}[\uff10-\uff19]]; #includes fullwidth digits |
| $ExtendNumLet = [\p{Word_Break = ExtendNumLet}]; |
| $E_Base = [\p{Word_Break = EB}\U0001F3C2\U0001F3C7\U0001F3CC\U0001F46A-\U0001F46D\U0001F46F\U0001F574\U0001F6CC]; |
| $E_Modifier = [\p{Word_Break = EM}]; |