1<?xml version="1.0" encoding="UTF-8" ?> 2<!DOCTYPE supplementalData SYSTEM "../../common/dtd/ldmlSupplemental.dtd"> 3<!-- 4Copyright © 1991-2016 Unicode, Inc. 5CLDR data files are interpreted according to the LDML specification (http://unicode.org/reports/tr35/) 6For terms of use, see http://www.unicode.org/copyright.html 7--> 8<supplementalData> 9 <version number="$Revision: 12355 $"/> 10 <transforms> 11 <transform source="es_419" target="fa" direction="forward" alias="fa-t-es-419"> 12 <tRule><![CDATA[ 13$Boundary = [^[:L:][:M:][:N:]]; 14$Vowel = [i e o u a]; 15 16::es-es_FONIPA; 17::es_FONIPA-es_419_FONIPA; 18 19# In Farsi transcription of Spanish, un-stressed [e] should be treated 20# like [ə] which gets stripped off. However, we currently do have not 21# have a good way of finding stress in Spanish words. In the long term, 22# it would be _much_ better to look at stress markers, but for now 23# we do this trivial heuristics to find unstressed [e] in the first 24# syllable. 25$Boundary [^Vowel] {e} [^$Vowel]* $Vowel → ə; 26 27::und_FONIPA-fa; 28 ]]></tRule> 29 </transform> 30 </transforms> 31</supplementalData> 32