Voice configuration

VoiceConfig.from_json reads a Piper .onnx.json file. Strict parsing is the default and requires num_symbols, num_speakers, audio.sample_rate, and a non-empty phoneme_id_map.

from piperg2p import VoiceConfig

config = VoiceConfig.from_json("voice.onnx.json")
config = VoiceConfig.from_dict(config.to_dict())

phoneme_type is a string-compatible PhonemeType enum with values text, espeak, pinyin, hebrew, japanese, and thai. This Phase 1 release implements text and espeak. strict=False is available for legacy configs and emits a compatibility warning when values are inferred.

ID values are normalized to immutable integer tuples. IDs must be non-negative and below num_symbols. Voice maps are never replaced by a universal vocabulary. Vowel clusters must have at least two elements and their merged token must be in the map.