blob: 4436e7c1c35366981727b6d329adb024f6f74541 [file] [log] [blame]
# © 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: my_my_FONIPA.txt
# Generated from CLDR
#
# 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 ɴ ʔ ə]; # 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 (င\u103A plus U+1039 VIRAMA) into plain င\u103A.
# Hmm, what would happen if the syllable ending in kinzi had non-low tone?
င\u103A $virama → င\u103A;
# Unstack everything else, i.e. replace U+1039 VIRAMA with U+103A ASAT.
$virama → $asat;
# Unstack U+103F GREAT SA.
ဿ → သ\u103Aသ;
# Insert a syllable boundary marker /./ before every independent letter.
::Null;
[^.$] { } $independent ([\u1037\u103B-\u103E])* [^\u103A] → \.;
# Insert default inherent vowel: /a\u0330/ at the end, /ə/ everywhere else.
::Null;
([\u1000-\u1021\u103F] [\u103B-\u103E]*) } [$] → $1 a $creaky;
([\u1000-\u1021\u103F] [\u103B-\u103E]*) } \. → $1 ə;
# 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 ၎င\u103Aး early.
၎င\u103Aး → lə\.ɡa $high ʊ\u032Fɴ;
#
# Rhymes
#
::Null;
က\u103A → ɛʔ;
ဂ\u103A → ɛʔ; # in မဂ\u1039ဂဇင\u103Aး ~ မဂ\u103Aဂဇင\u103Aး /mɛʔ.ɡə.zɪ\u0301ɴ/
င\u1037\u103A → ɪ $creaky ɴ;
င\u103Aး → ɪ $high ɴ;
င\u103A → ɪ $low ɴ;
စ\u103A → ɪʔ; # maybe sometimes /eɪ\u032Fʔ/
ဉ\u1037\u103A → ɪ $creaky ɴ;
ဉ\u103Aး → ɪ $high ɴ;
ဉ\u103A → ɪ $low ɴ;
ည\u1037\u103A → ɛ $creaky;
ည\u103Aး → ɛ $high;
ည\u103A → ɛ $low;
ဏ\u1037\u103A → a $creaky ɴ;
ဏ\u103Aး → a $high ɴ;
ဏ\u103A → a $low ɴ;
တ\u103A → aʔ;
န\u1037\u103A → a $creaky ɴ;
န\u103Aး → a $high ɴ;
န\u103A → a $low ɴ;
ပ\u103A → aʔ;
မ\u1037\u103A → a $creaky ɴ;
မ\u103Aး → a $high ɴ;
မ\u103A → a $low ɴ;
ယ\u1037\u103A → ɛ $creaky;
ယ\u103Aး → ɛ $high;
ယ\u103A → ɛ $low;
သ\u103A → aʔ;
$vs_aa ဉ\u1037\u103A → ɪ $creaky ɴ;
$vs_aa ဉ\u103Aး → ɪ $high ɴ;
$vs_aa ဉ\u103A → ɪ $low ɴ;
$vs_aa တ\u103A → aʔ;
$vs_aa ဏ\u1037\u103A → a $creaky ɴ;
$vs_aa ဏ\u103Aး → a $high ɴ;
$vs_aa ဏ\u103A → a $low ɴ;
$vs_aa န\u1037\u103A → a $creaky ɴ;
$vs_aa န\u103Aး → a $high ɴ;
$vs_aa န\u103A → a $low ɴ;
$vs_aa ပ\u103A → aʔ; # in ကလာပ\u103Aစည\u103Aး /kə.laʔ.sɛ\u0301/ (club cell)
$vs_aa ယ\u1037\u103A → ɛ $creaky;
$vs_aa ယ\u103Aး → ɛ $high;
$vs_aa ယ\u103A → ɛ $low;
$vs_aa \u1037 → a $creaky; # redundant creaky tone
$vs_aa း → a $high;
$vs_aa → a $low;
$vs_i က\u103A → eɪ\u032Fʔ;
$vs_i စ\u103A → eɪ\u032Fʔ;
$vs_i တ\u103A → eɪ\u032Fʔ;
$vs_i န\u1037\u103A → e $creaky ɪ\u032Fɴ;
$vs_i န\u103Aး → e $high ɪ\u032Fɴ;
$vs_i န\u103A → e $low ɪ\u032Fɴ;
$vs_i ပ\u103A → eɪ\u032Fʔ;
$vs_i မ\u1037\u103A → e $creaky ɪ\u032Fɴ;
$vs_i မ\u103Aး → e $high ɪ\u032Fɴ;
$vs_i မ\u103A → e $low ɪ\u032Fɴ;
$vs_i $vs_u က\u103A → aɪ\u032Fʔ;
$vs_i $vs_u င\u1037\u103A → a $creaky ɪ\u032Fɴ;
$vs_i $vs_u င\u103Aး → a $high ɪ\u032Fɴ;
$vs_i $vs_u င\u103A → a $low ɪ\u032Fɴ;
$vs_i $vs_u ဏ\u1037\u103A → a $creaky ɪ\u032Fɴ;
$vs_i $vs_u ဏ\u103Aး → a $high ɪ\u032Fɴ;
$vs_i $vs_u ဏ\u103A → a $low ɪ\u032Fɴ;
$vs_i $vs_u ယ\u1037\u103A → o $creaky;
$vs_i $vs_u ယ\u103Aး → o $high;
$vs_i $vs_u ယ\u103A → o $low; # in က\u102D\u102Fယ\u103A /kò/
$vs_i $vs_u \u1037 → o $creaky;
$vs_i $vs_u း → o $high;
$vs_i $vs_u → o $low;
$vs_i $anusvara \u1037 → e $creaky ɪ\u032Fɴ;
$vs_i $anusvara း → e $high ɪ\u032Fɴ;
$vs_i $anusvara → e $low ɪ\u032Fɴ;
$vs_i → i $creaky;
$vs_ii \u1037 → i $creaky; # this does not usually occur
$vs_ii း → i $high;
$vs_ii → i $low;
$vs_u က\u103A → oʊ\u032Fʔ;
$vs_u ဂ\u103A → oʊ\u032Fʔ;
$vs_u ဏ\u1037\u103A → o $creaky ʊ\u032Fɴ;
$vs_u ဏ\u103Aး → o $high ʊ\u032Fɴ;
$vs_u ဏ\u103A → o $low ʊ\u032Fɴ;
$vs_u တ\u103A → oʊ\u032Fʔ;
$vs_u န\u1037\u103A → o $creaky ʊ\u032Fɴ;
$vs_u န\u103Aး → o $high ʊ\u032Fɴ;
$vs_u န\u103A → o $low ʊ\u032Fɴ;
$vs_u ပ\u103A → oʊ\u032Fʔ;
$vs_u မ\u1037\u103A → o $creaky ʊ\u032Fɴ;
$vs_u မ\u103Aး → o $high ʊ\u032Fɴ;
$vs_u မ\u103A → o $low ʊ\u032Fɴ;
$vs_u $anusvara \u1037 → o $creaky ʊ\u032Fɴ;
$vs_u $anusvara း → o $high ʊ\u032Fɴ;
$vs_u $anusvara → o $low ʊ\u032Fɴ;
$vs_u → u $creaky;
$vs_uu \u1037 → u $creaky; # this does not usually occur
$vs_uu း → u $high;
$vs_uu → u $low;
$vs_e တ\u103A → ɪʔ;
$vs_e $vs_aa က\u103A → aʊ\u032Fʔ;
$vs_e $vs_aa င\u1037\u103A → a $creaky ʊ\u032Fɴ;
$vs_e $vs_aa င\u103Aး → a $high ʊ\u032Fɴ;
$vs_e $vs_aa င\u103A → a $low ʊ\u032Fɴ;
$vs_e $vs_aa \u1037 → ɔ $creaky;
$vs_e $vs_aa း → ɔ $high; # redundant high tone; this does not usually occur
$vs_e $vs_aa \u103A → ɔ $low;
$vs_e $vs_aa → ɔ $high;
$vs_e \u1037 → e $creaky;
$vs_e း → e $high;
$vs_e → e $low;
$vs_ai \u1037 → ɛ $creaky;
$vs_ai း → ɛ $high; # redundant high tone; this does not usually occur
$vs_ai → ɛ $high;
$anusvara \u1037 → a $creaky ɴ;
$anusvara း → a $high ɴ;
$anusvara → a $low ɴ;
$med_w တ\u103A → ʊʔ;
$med_w န\u1037\u103A → ʊ $creaky ɴ;
$med_w န\u103Aး → ʊ $high ɴ;
$med_w န\u103A → ʊ $low ɴ;
$med_w ပ\u103A → ʊʔ;
$med_w မ\u1037\u103A → ʊ $creaky ɴ;
$med_w မ\u103Aး → ʊ $high ɴ;
$med_w မ\u103A → ʊ $low ɴ;
#
# Medials
#
::Null;
# Palatalization of the velar stops before MEDIAL YA and MEDIAL RA:
# velar + /j/ ==> modern palatals.
ကျ → t\u0361ɕ;
ချ → t\u0361ɕʰ;
ဂျ → d\u0361ʑ;
ဃျ → d\u0361ʑ;
ကြ → t\u0361ɕ;
ခြ → t\u0361ɕʰ;
ဂြ → d\u0361ʑ;
ဃြ → d\u0361ʑ;
# Remove redundant MEDIAL YA and MEDIAL RA after initial YA.
ယ { [$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 ʃ from (SA|LA)+YA+HA.
သျ\u103E → ʃ;
လျ\u103E → ʃ;
# 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.
င\u103E → ŋ\u030A;
ဉ\u103E → ɲ\u0325;
ည\u103E → ɲ\u0325;
ဏ\u103E → n\u0325;
န\u103E → n\u0325;
မ\u103E → m\u0325;
ယ\u103E → ʃ;
ရ\u103E → ʃ;
လ\u103E → l\u0325;
ဝ\u103E → w\u0325;
ဠ\u103E → l\u0325;
# 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
က → k;
ခ → kʰ;
ဂ → ɡ;
ဃ → ɡ;
င → ŋ;
# Historic palatals
စ → s;
ဆ → sʰ;
ဇ → z;
ဈ → z;
ဉ → ɲ;
ည → ɲ;
# Alveolars
ဋ → t;
ဌ → tʰ;
ဍ → d;
ဎ → d;
ဏ → n;
# Historic dentals ==> alveolars
တ → t;
ထ → tʰ;
ဒ → d;
ဓ → d;
န → n;
# Labials
ပ → p;
ဖ → pʰ;
ဗ → b;
ဘ → b;
မ → m;
# Other letters
ယ → j;
ရ → j; # historic /r/
လ\u103A → ; # final, typically not pronounced in native words
လ → l;
ဝ → w;
သ → θ; # historic /s/ ==> modern dental
ဟ → h;
ဠ → l;
အ → ʔ;
# Independent vowels
ဣ\u1037 → ʔḭ; # redundant creaky tone; this does not usually occur
ဣး → ʔí; # this does not usually occur
ဣ → ʔḭ;
ဤ\u1037 → ʔḭ; # this does not usually occur
ဤး → ʔí; # this does not usually occur
ဤ → ʔì;
ဥ\u1037 → ʔṵ; # redundant creaky tone; this does not usually occur
ဥး → ʔú; # this does not usually occur
ဥ → ʔṵ;
ဦ\u1037 → ʔṵ; # this does not usually occur
ဦး → ʔú;
ဦ → ʔù;
ဧ\u1037 → ʔḛ; # this does not usually occur
ဧး → ʔé;
ဧ → ʔè;
ဩ\u1037 → ʔɔ\u0330; # this does not usually occur
ဩး → ʔɔ\u0301; # redundant high tone; this does not usually occur
ဩ → ʔɔ\u0301;
ဪ\u1037 → ʔɔ\u0330; # this does not usually occur
ဪး → ʔɔ\u0301; # this does not usually occur
ဪ → ʔɔ\u0300;
# Various signs
၌ → n\u0325aɪ\u032Fʔ;
၍ → jwḛ;
# ၎င\u103Aး was handled earlier.
၏ → ʔḭ;
#
# Postprocessing
#
# Delete any remaining U+103A ASAT.
$asat → ;
# Delete zero-width space, non-joiner, joiner.
[\u200B-\u200D] → ;
::NFC;