<?xml version="1.0" encoding="UTF-8" ?> <!DOCTYPE supplementalData SYSTEM "../../common/dtd/ldmlSupplemental.dtd"> <!-- Copyright © 1991-2013 Unicode, Inc. CLDR data files are interpreted according to the LDML specification (http://unicode.org/reports/tr35/) For terms of use, see http://www.unicode.org/copyright.html --> <supplementalData> <version number="$Revision: 12245 $"/> <transforms> <transform source="Latin" target="NumericPinyin" direction="both" alias="und-pinyin-t-d0-npinyin" backwardAlias="und-pinyin-t-s0-npinyin"> <tRule><![CDATA[ # According to the pinyin definitions I've been able to find: # 'a', 'e' are the preferred bases # otherwise 'o' # otherwise last vowel # The trailing form of syllables are the following: # "a", "ai", "ao", "an", "ang", # "o", "ou", "ong", # "e", "ei", "er", "en", "eng", # "i", "ia", "iao", "ie", "iu", "ian", "in", "iang", "ing", "iong", # "u", "ua", "uo", "uai", "ui", "uan", "un", "uang", "ueng", # "ü", "üe", "üan", "ün" # so the letters the tone will 'hop' are: ::NFD (NFC); $tone = [̄́̌̀̆] ; # Move the tone to the end of a syllable, and convert to number e {($tone) r} → r &Pinyin-NumericPinyin($1); ($tone) ( [i o n u {o n} {n g}]) → $2 &Pinyin-NumericPinyin($1); ($tone) → &Pinyin-NumericPinyin($1); # The following backs up until it finds the right vowel, then deposits the tone $vowel = [aAeEiIoOuU {ü} {Ü} vV]; $consonant = [[a-z A-Z] - [$vowel]]; $digit = [1-5]; $1 &NumericPinyin-Pinyin($3) $2 ← ([aAeE]) ($vowel* $consonant*) ($digit); $1 &NumericPinyin-Pinyin($3) $2 ← ([oO]) ([$vowel-[aeAE]]* $consonant*) ($digit); $1 &NumericPinyin-Pinyin($3) $2 ← ($vowel) ($consonant*) ($digit); &NumericPinyin-Pinyin($1) ← [:letter:] {($digit)}; ::NFC (NFD); ]]></tRule> </transform> </transforms> </supplementalData>