common/transforms/my-my_FONIPA.xml

<?xml version="1.0" encoding="UTF-8" ?>
<!DOCTYPE supplementalData SYSTEM "../../common/dtd/ldmlSupplemental.dtd">
<!-- Copyright  1991-2015 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: 12347 $" />
	<transforms>
		<transform source="my" target="my_FONIPA" direction="forward" alias="my-fonipa-t-my">
			<tRule><![CDATA[

# Pronunciation rules for Burmese.
#
# The following rules are lexical and heuristic: lexical in the sense
# that they generate phoneme strings which may further undergo
# post-lexical phonological processes, in particular voicing, to
# result in actual surface forms; heuristic in the sense that they try
# to resolve ambiguities, especially around reduced vowels, in a
# systematic way that may be incorrect in many situations. Vowel
# reduction depends on many factors, such as morphemic structure,
# which are not available here.

#
# Definitions
#

# Dependent vowel signs
$vs_AA = \u102B;
$vs_aa = \u102C;
$vs_i = \u102D;
$vs_ii = \u102E;
$vs_u = \u102F;
$vs_uu = \u1030;
$vs_e = \u1031;
$vs_ai = \u1032;

# Various signs
$anusvara = \u1036;
$visarga = \u1038;
$virama = \u1039;
$asat = \u103A;

# Dependent (medial) consonant signs
$med_y = \u103B;
$med_r = \u103C;
$med_w = \u103D;
$med_h = \u103E;

# Independent letters and letter-like punctuation symbols
$independent = [\u1000-\u102A \u103F \u104C-\u104F \u1050-\u1055];

$creaky = \u0330;
$high = \u0301;
$low = \u0300;
$coda = [$creaky $high $low   ];  # TODO: remove if unused

#
# Preprocessing
#

::NFC;

# Replace U+102B TALL AA with U+102C AA. Their pronunciation is identical.
$vs_AA  $vs_aa;

# Unstack kinzi ( plus U+1039 VIRAMA) into plain .
# Hmm, what would happen if the syllable ending in kinzi had non-low tone?
 $virama  ;

# Unstack everything else, i.e. replace U+1039 VIRAMA with U+103A ASAT.
$virama  $asat;

# Unstack U+103F GREAT SA.
  ;

# Insert a syllable boundary marker /./ before every independent letter.
::Null;
[^.$] { } $independent ([\u1037\u103B-\u103E])* [^\u103A]  \.;

# Insert default inherent vowel: /a/ at the end, // everywhere else.
::Null;
([\u1000-\u1021\u103F] [\u103B-\u103E]*) } [$]  $1 a $creaky;
([\u1000-\u1021\u103F] [\u103B-\u103E]*) } \.   $1 ;

# Allow for additional coda consonants.
#
# This only covers a few of the cases in which full coda consonants
# can appear in loanwords. The general situation is somewhat rare and
# is more easily dealt with in a formalism that can impose structural
# constraints on syllables more easily.
::Null;
$asat ($visarga)? [\u1000-\u102A] { $asat  ;

# Deal with  early.
  l\.a $high ;

#
# Rhymes
#

::Null;

  ;

  ;  # in  ~  /m..z/

   $creaky ;
   $high ;
   $low ;

  ;  # maybe sometimes /e/

   $creaky ;
   $high ;
   $low ;

   $creaky;
   $high;
   $low;

  a $creaky ;
  a $high ;
  a $low ;

  a;

  a $creaky ;
  a $high ;
  a $low ;

  a;

  a $creaky ;
  a $high ;
  a $low ;

   $creaky;
   $high;
   $low;

  a;

$vs_aa    $creaky ;
$vs_aa    $high ;
$vs_aa    $low ;
$vs_aa   a;
$vs_aa   a $creaky ;
$vs_aa   a $high ;
$vs_aa   a $low ;
$vs_aa   a $creaky ;
$vs_aa   a $high ;
$vs_aa   a $low ;
$vs_aa   a;  # in  /k.la.s/ (club cell)
$vs_aa    $creaky;
$vs_aa    $high;
$vs_aa    $low;
$vs_aa   a $creaky;  # redundant creaky tone
$vs_aa   a $high;
$vs_aa  a $low;

$vs_i   e;
$vs_i   e;
$vs_i   e;
$vs_i   e $creaky ;
$vs_i   e $high ;
$vs_i   e $low ;
$vs_i   e;
$vs_i   e $creaky ;
$vs_i   e $high ;
$vs_i   e $low ;
$vs_i $vs_u   a;
$vs_i $vs_u   a $creaky ;
$vs_i $vs_u   a $high ;
$vs_i $vs_u   a $low ;
$vs_i $vs_u   a $creaky ;
$vs_i $vs_u   a $high ;
$vs_i $vs_u   a $low ;
$vs_i $vs_u   o $creaky;
$vs_i $vs_u   o $high;
$vs_i $vs_u   o $low;  # in  /k/
$vs_i $vs_u   o $creaky;
$vs_i $vs_u   o $high;
$vs_i $vs_u  o $low;
$vs_i $anusvara   e $creaky ;
$vs_i $anusvara   e $high ;
$vs_i $anusvara  e $low ;
$vs_i  i $creaky;

$vs_ii   i $creaky;  # this does not usually occur
$vs_ii   i $high;
$vs_ii  i $low;

$vs_u   o;
$vs_u   o;
$vs_u   o $creaky ;
$vs_u   o $high ;
$vs_u   o $low ;
$vs_u   o;
$vs_u   o $creaky ;
$vs_u   o $high ;
$vs_u   o $low ;
$vs_u   o;
$vs_u   o $creaky ;
$vs_u   o $high ;
$vs_u   o $low ;
$vs_u $anusvara   o $creaky ;
$vs_u $anusvara   o $high ;
$vs_u $anusvara  o $low ;
$vs_u  u $creaky;

$vs_uu   u $creaky;  # this does not usually occur
$vs_uu   u $high;
$vs_uu  u $low;

$vs_e   ;
$vs_e $vs_aa   a;
$vs_e $vs_aa   a $creaky ;
$vs_e $vs_aa   a $high ;
$vs_e $vs_aa   a $low ;
$vs_e $vs_aa    $creaky;
$vs_e $vs_aa    $high;  # redundant high tone; this does not usually occur
$vs_e $vs_aa    $low;
$vs_e $vs_aa   $high;
$vs_e   e $creaky;
$vs_e   e $high;
$vs_e  e $low;

$vs_ai    $creaky;
$vs_ai    $high;  # redundant high tone; this does not usually occur
$vs_ai   $high;

$anusvara   a $creaky ;
$anusvara   a $high ;
$anusvara  a $low ;

$med_w   ;
$med_w    $creaky ;
$med_w    $high ;
$med_w    $low ;
$med_w   ;
$med_w    $creaky ;
$med_w    $high ;
$med_w    $low ;

#
# Medials
#

::Null;

# Palatalization of the velar stops before MEDIAL YA and MEDIAL RA:
# velar + /j/ ==> modern palatals.

  t;
  t;
  d;
  d;

  t;
  t;
  d;
  d;

# Remove redundant MEDIAL YA and MEDIAL RA after initial YA.
 { [$med_y $med_r]  ;

# Reorder the medials so that U+103E SIGN MEDIAL HA comes before any
# other medials.

# First, push U+103E MEDIAL HA before U+103D MEDIAL WA.
\u103D \u103E  \u103E \u103D;
::Null;
# Now MEDIAL WA comes last.

# Produce the palatal  from (SA|LA)+YA+HA.
  ;
  ;

# Second, push U+103E MEDIAL HA before U+103C MEDIAL RA.
\u103C \u103E  \u103E \u103C;
::Null;

# Finally, push U+103E MEDIAL HA before U+103B MEDIAL YA.
\u103B \u103E  \u103E \u103B;
::Null;

# Consume MEDIAL HA and apply devoicing.

  ;
  ;
  ;
  n;
  n;
  m;
  ;
  ;
  l;
  w;
  l;

# Drop any remaining U+103E MEDIAL HA.
\u103E  ;

# Simplify medial cluster /jw/ to /w/, i.e. drop U+103B MEDIAL YA and
# U+103C MEDIAL RA before U+103D MEDIAL WA.  # TODO: revisit this
\u103B } \u103D  ;
\u103C } \u103D  ;

\u103B  j;
\u103C  j;
\u103D  w;

#
# Initials
#

# Velars
  k;
  k;
  ;
  ;
  ;

# Historic palatals
  s;
  s;
  z;
  z;
  ;
  ;

# Alveolars
  t;
  t;
  d;
  d;
  n;

# Historic dentals ==> alveolars
  t;
  t;
  d;
  d;
  n;

# Labials
  p;
  p;
  b;
  b;
  m;

# Other letters
  j;
  j;  # historic /r/
  ;  # final, typically not pronounced in native words
  l;
  w;
  ;  # historic /s/ ==> modern dental
  h;
  l;
  ;

# Independent vowels

  ;  # redundant creaky tone; this does not usually occur
  ;  # this does not usually occur
  ;

  ;  # this does not usually occur
  ;  # this does not usually occur
  ;

  ;  # redundant creaky tone; this does not usually occur
  ;  # this does not usually occur
  ;

  ;  # this does not usually occur
  ;
  ;

  ;  # this does not usually occur
  ;
  ;

  ;  # this does not usually occur
  ;  # redundant high tone; this does not usually occur
  ;

  ;  # this does not usually occur
  ;  # this does not usually occur
  ;

# Various signs

  na;
  jw;
#  was handled earlier.
  ;

#
# Postprocessing
#

# Delete any remaining U+103A ASAT.
$asat  ;

# Delete zero-width space, non-joiner, joiner.
[\u200B-\u200D]  ;

::NFC;

			]]></tRule>
		</transform>
	</transforms>
</supplementalData>