A free online tool that normalizes Japanese text by unifying full-width/half-width characters, katakana/hiragana, and whitespace. Perfect for form input preprocessing, data cleansing, and text proofreading.
Includes 5 presets (Form Input, Alphanumerics Half-width, Kana Full-width, All Hiragana, All Katakana) plus advanced individual settings for alphanumerics, symbols, spaces, kana width, kana type, line endings, trailing whitespace, and tab conversion. A diff highlighting feature lets you visually review all changes.
All processing runs entirely in your browser — your data is never uploaded to any server. No registration, no installation. Just paste your text and start normalizing instantly.
Character Code Conversion
Full-width ↔ half-width conversion uses Unicode code point offset calculation. Full-width alphanumerics (U+FF10–U+FF5A) differ from half-width (U+0030–U+007A) by a fixed offset of 0xFEE0, enabling fast arithmetic conversion. Symbols use the same offset method.
Kana Conversion Algorithm
Half-width → full-width kana uses Unicode NFKC normalization. Voiced/semi-voiced half-width kana (e.g., ガ = 2 characters) are correctly merged into single full-width characters (ガ). Reverse conversion uses NFD decomposition to separate dakuten marks, then maps through a conversion table.
Diff Display (Myers Algorithm)
The diff view uses the Myers diff algorithm with O(ND) character-level comparison, highlighting deletions (red) and insertions (green). For performance, diff computation is skipped for texts exceeding 6,000 characters.