tiny-code-only-tts / mini_tts /normalizer.py
abersbail's picture
Deploy tiny code-only TTS Space
0a88ee7 verified
raw
history blame contribute delete
895 Bytes
import re
DIGRAPH_MAP = {
"th": "T",
"sh": "S",
"ch": "C",
"ph": "F",
"oo": "U",
"ee": "I",
"ai": "A",
"ou": "W",
}
def normalize_text(text: str) -> str:
normalized = text.lower().strip()
normalized = re.sub(r"[^a-z0-9\s,.;:!?'-]", " ", normalized)
normalized = re.sub(r"\s+", " ", normalized)
return normalized
def text_to_symbols(text: str) -> list[str]:
normalized = normalize_text(text)
symbols: list[str] = []
i = 0
while i < len(normalized):
pair = normalized[i : i + 2]
if pair in DIGRAPH_MAP:
symbols.append(DIGRAPH_MAP[pair])
i += 2
continue
ch = normalized[i]
if ch in ",.;:!?":
symbols.append("|")
elif ch == " ":
symbols.append(" ")
else:
symbols.append(ch)
i += 1
return symbols