"""Korean text utilities: chosung extraction and query helpers."""
from __future__ import annotations
# 19 Korean initial consonants (chosung) in Unicode order
_CHOSUNG = (
"ㄱ",
"ㄲ",
"ㄴ",
"ㄷ",
"ㄸ",
"ㄹ",
"ㅁ",
"ㅂ",
"ㅃ",
"ㅅ",
"ㅆ",
"ㅇ",
"ㅈ",
"ㅉ",
"ㅊ",
"ㅋ",
"ㅌ",
"ㅍ",
"ㅎ",
)
# Set of all chosung jamo for fast lookup
_CHOSUNG_SET = frozenset(_CHOSUNG)
_HANGUL_BASE = 0xAC00
_HANGUL_END = 0xD7A3
_JUNGSUNG_COUNT = 21
_JONGSUNG_COUNT = 28
def extract_chosung(text: str) -> str:
"""Extract initial consonants from Korean syllables; lowercase non-Hangul.
Examples:
>>> extract_chosung("삼성전자")
'ㅅㅅㅈㅈ'
>>> extract_chosung("삼성SDI")
'ㅅㅅsdi'
>>> extract_chosung("LG전자")
'lgㅈㅈ'
"""
result: list[str] = []
for ch in text:
code = ord(ch)
if _HANGUL_BASE <= code <= _HANGUL_END:
index = (code - _HANGUL_BASE) // (_JUNGSUNG_COUNT * _JONGSUNG_COUNT)
result.append(_CHOSUNG[index])
else:
result.append(ch.lower())
return "".join(result)
def has_chosung(query: str) -> bool:
"""Return True if *query* contains any chosung jamo (ㄱ-ㅎ)."""
return any(ch in _CHOSUNG_SET for ch in query)
def is_pure_chosung(query: str) -> bool:
"""Return True if *query* consists entirely of chosung jamo."""
return bool(query) and all(ch in _CHOSUNG_SET for ch in query)
def normalize_mixed_query(query: str) -> str:
"""Normalize a mixed chosung/non-chosung query for matching.
Chosung characters are kept as-is; non-chosung characters are lowered.
Examples:
>>> normalize_mixed_query("ㅅㅅSDI")
'ㅅㅅsdi'
"""
return "".join(ch if ch in _CHOSUNG_SET else ch.lower() for ch in query)
