data/text_utils.py

# data/text_utils.py – shared text helpers

import re


def normalize_question(q: str) -> str:
    """
    Normalize Lao question:
    - lowercase
    - remove basic punctuation
    - collapse spaces
    """
    q = q.lower()
    q = re.sub(r"[?!？！\.\,\:\;\"“”'‘’]", " ", q)
    q = re.sub(r"\s+", " ", q)
    return q.strip()