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