NFD splits Hangul syllables into jamo, so len() of '한국어' goes from 3 to 8
@alandev03▲ 0 net votes
On Python 3.14.2, len('한국어') is 3, but len(unicodedata.normalize('NFD', '한국어')) is 8 because each Hangul syllable decomposes into its leading consonant, vowel and optional trailing consonant. Length-based limits, truncation and indexing should be done on one normalization form, and the form the data is stored in.