"""IPA-to-Kokoro phoneme mapping for Thai.
Maps FastThaiG2P's Wiktionary IPA output to Kokoro-82M's phoneme vocabulary.
Kokoro doesn't natively support Thai, but its phoneme set covers most Thai
segmental phonemes. Tones are approximated via Kokoro's intonation marks.
## License
Copyright 2026 Charin Polpanumas and Amazon Web Services
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
from __future__ import annotations
import re
# Thai tone → Kokoro intonation mark (one symbol per tone, no mergers)
# ↑ is an otherwise-unused Kokoro token (ID 170) repurposed as the high tone
# so that high (˦˥) and rising (˩˩˦) stay phonemically distinct.
_TONE_MAP = [
("˩˩˦", "↗"), # rising
("˥˩", "↘"), # falling
("˦˥", "↑"), # high
("˨˩", "↓"), # low
("˧", "→"), # mid
]
# Thai affricates → Kokoro precomposed symbols
_AFFRICATE_MAP = [
("t͡ɕʰ", "ʨʰ"),
("t͡ɕ", "ʨ"),
]
# Characters to strip (no Kokoro equivalent)
_STRIP_CHARS = ["̚", "̯", "͡"]
# Character replacements
_CHAR_MAP = {
"g": "ɡ", # ASCII g → IPA g
}
[docs]
def ipa_to_kokoro(ipa: str) -> str:
"""Convert FastThaiG2P IPA output to Kokoro-compatible phoneme string.
Strips word-boundary slashes and maps tones, affricates, and unsupported
characters to Kokoro equivalents.
Args:
ipa: Raw output from G2P.convert() e.g. '/sa˨˩.wat̚˨˩.diː˧/ /kʰrap̚˦˥/'
Returns:
Kokoro phoneme string e.g. 'sa↓.wat↓.diː→ kʰrap↗'
"""
# Strip slashes, join words with space
phonemes = " ".join(p.strip("/") for p in ipa.split() if p.startswith("/"))
if not phonemes:
phonemes = ipa.replace("/", "").strip()
# Affricates (before stripping tie bar)
for old, new in _AFFRICATE_MAP:
phonemes = phonemes.replace(old, new)
# Tones (longest match first)
for old, new in _TONE_MAP:
phonemes = phonemes.replace(old, new)
# Strip unsupported combining characters
for char in _STRIP_CHARS:
phonemes = phonemes.replace(char, "")
# Character replacements
for old, new in _CHAR_MAP.items():
phonemes = phonemes.replace(old, new)
return phonemes