Character conversion
Full/half-width · Roman letters · jamo · a warning for mixed characters
문자 변환
전각·반각, 한글·로마자, 한글·자모를 한 곳에서. 붙여 넣으면 섞인 글자를 먼저 알려 줍니다
전각 글자가 섞여 있습니다 — 검색·로그인·조회가 안 되던 이유가 대개 이것입니다.
Convert between full-width and half-width, Hangul and Roman letters, and Hangul and jamo on one screen. The moment you paste, any full-width characters mixed in are pointed out.
How to use it
- Choose which way to convert
- Paste the text and it converts straight away
- If full-width characters are mixed in, a notice appears above
- Copy it, or press “Result to input” and keep converting
- Copy the result, or press “result as input” to keep converting
Common questions
- What are full-width and half-width?
- ＡＢ and AB only look different in width on screen, but to a computer they are different characters. That is why a pasted value finds 0 results, a login fails, or a spreadsheet lookup misses. It is the hardest kind of cause to spot by eye.
- The romanisation differs from the official rules?
- It applies the character mapping only, without sound changes. 신라 comes out as sinra while the official form is Silla. Doing it properly needs morpheme boundaries, which needs a dictionary. So rather than presenting a wrong value as right, it tells you where a sound would change.
- Why is there no simplified/traditional Chinese?
- That needs a table, not a rule. A hand-written table goes stale and leaks, so it goes in when a proper data source can be brought in.
- What is splitting into jamo good for?
- Building an initial-consonant search index, or looking at why text got mangled. What you split can be joined back too.
- What if one simplified character maps to several traditional ones?
- It converts to the first candidate and notes on screen which characters were ambiguous. 发 can be 發 (to send) or 髮 (hair) — only the meaning decides. Picking one silently would ship the wrong text.
- Does it convert by word as well?
- No — by character. Catching vocabulary differences such as 打印機 ↔ 打印机 needs a word dictionary, and that table is far larger.
- Why does pinyin pause the first time I press it?
- The sound table is 20,000 characters (167KB), so it is not bundled with the tool. Someone who came only for full-width and half-width should not carry that weight, so it is fetched once, and only for those who choose pinyin. After that it is never fetched again.
- What about characters with several readings (多音字)?
- Where Unihan records several readings, it uses the first and shows the rest alongside. It cannot tell apart readings that depend on the word or the context (行 xíng / háng), and it says so in the result.
Other tools
Stays on your device (downloads data) What you enter is handled inside your browser. Only the data or model it needs is downloaded once.
This tool page states where what you enter goes — see the badge above. All tools · KarmoLab · Who made this · Built with AI. View source