How Does Unicode Diacritics & Accent Removal Work?
The engine runs text.normalize("NFD").replace(/[\u0300-\u036f]/g, "").normalize("NFC"), separating accented letters into base Latin characters plus combining marks and deleting the combining marks.
Text Cleaner normalizes Unicode characters using canonical decomposition (NFD) and strips all combining diacritical marks (\u0300–\u036f). This transforms accented letters into their base Latin equivalents (e.g. é becomes e, ñ becomes n, ç becomes c, and ř becomes r).
What Unicode Code Points Are Stripped in a Worked Before / After Example?
In NFD form, precomposed characters like é (U+00E9) decompose into base e (U+0065) + Combining Acute Accent (U+0301), allowing exact removal of the accent while keeping the letter intact:
- Worked Before / After Example: Input
Příliš žluťoučký kůň úpěl ďábelské ódy.\nCafé, résumé, façade, naïve, señor, Über.(81 characters, 12 words, 2 lines) → OutputPrilis zlutoucky kun upel dabelske ody.\nCafe, resume, facade, naive, senor, Uber.(81 ASCII characters, 12 words, 2 lines; 23 diacritical marks stripped).
| Diacritic Type & Combining Code Point | Input Example (Before) | Output Result (After) | Marks Stripped (count) |
|---|---|---|---|
Acute & Grave Accents (U+0301, U+0300) |
Café résumé à bientôt (21 chars) |
Cafe resume a bientot (21 chars) |
5 accent marks stripped |
Caron / Háček & Ring Above (U+030C, U+030A) |
Příliš žluťoučký kůň (20 chars) |
Prilis zlutoucky kun (20 chars) |
9 Czech diacritics stripped |
Tilde & Cedilla (U+0303, U+0327) |
señor façade São Paulo (22 chars) |
senor facade Sao Paulo (22 chars) |
3 marks stripped |
Diaeresis / Umlaut (U+0308) |
naïve Über München (18 chars) |
naive Uber Munchen (18 chars) |
3 umlaut/diaeresis marks stripped |
When Should You Strip Accents from Text?
Strip accents when normalizing search indexes, generating ASCII-safe URL slugs and filenames, or fitting messages into 160-character GSM-7 SMS encoding.
- Database & Search Indexing: Normalize search queries so users can find names regardless of whether they typed accents or plain letters.
- URL Slugs & File Names: Web servers and file systems often choke on non-ASCII characters in paths.
- Legacy Systems & SMS: Sending text via standard GSM-7 SMS without paying double or triple character fees for Unicode encoding.
Is Diacritics Removal Completely Safe and On-Device?
Yes—all Unicode NFD/NFC normalization runs in local browser memory with zero network transmission.
Your sensitive names, strings, and document drafts are processed entirely in memory in your web browser. Nothing is ever sent to a remote server.
