diff options
Diffstat (limited to 'intl/icu/source/data/translit/Han_Spacedhan.txt')
| -rw-r--r-- | intl/icu/source/data/translit/Han_Spacedhan.txt | 42 |
1 files changed, 42 insertions, 0 deletions
diff --git a/intl/icu/source/data/translit/Han_Spacedhan.txt b/intl/icu/source/data/translit/Han_Spacedhan.txt new file mode 100644 index 0000000000..b88c1dd14c --- /dev/null +++ b/intl/icu/source/data/translit/Han_Spacedhan.txt @@ -0,0 +1,42 @@ +# © 2016 and later: Unicode, Inc. and others. +# License & terms of use: http://www.unicode.org/copyright.html +# Generated using tools/cldr/cldr-to-icu/build-icu-data.xml +# +# File: Han_Spacedhan.txt +# Generated from CLDR +# + +# Only intended for internal use +# Make sure Han are normalized, including characters that contain them. +# The first set in the filter is computed with http://unicode.org/cldr/utility/list-unicodeset.jsp?a=[:tonfkd:/XXX/:]-[:ideographic:]-[:sc=han:] +# Where XXX is the resolved [:ideographic:][:sc=han:]. It needs updating with each Unicode release! +:: [[、。々《-』〜・㆒-㆟㈠-㉇㊀-㊰㋀-㋋ ㍘-㍰㍻-㍿㏠-㏾🈐-🈒🈔-🈺🉀-🉈🉐🉑][:ideographic:][:sc=han:]] nfkc; +:: fullwidth-halfwidth; +｡ → '.'; +。→ '.'; +、→ ','; +､→ ','; +《→ '«'; +》→ '»'; +〈 → '‹'; +〉→ '›'; +「→ '‘'; +」→ '’'; +｢→ '‘'; +｣→ '’'; +『→ '“'; +』→ '”'; +・→ '‧'; +･ → '‧'; +々→ '⓶'; +〜→ '~'; +$terminalPunct = [\.\,\:\;\?\!．，：？！｡、；[:Pe:][:Pf:]]; +$initialPunct = [:Ps:][:Pi:]; +# add space between any Han or terminal punctuation and letters, and +# between letters and Han or initial punct +[[:Ideographic:] $terminalPunct] {} [:Letter:] → ' ' ; +[:Letter:] [:Mark:]* {} [[:Ideographic:] $initialPunct] → ' ' ; +# remove spacing between ideographs and other letters +← [:Ideographic:] { ' ' } [:Letter:] ; +← [:Letter:] [:Mark:]* { ' ' } [:Ideographic:] ; + |