Korean ZIP Filename Encoding Repair (CP949)

How it works

Korean Windows Explorer often stores ZIP entry names as CP949/EUC-KR bytes without the UTF-8 flag. On a Mac, Archive Utility or unzip reads those bytes as UTF-8 and you get mojibake (쳌쳌, 안녕). This is not a generic unzip site — it only repairs archive filename encoding.

1. Drop a ZIP in this tab. Nothing is uploaded.
2. Read raw filename bytes from the central directory, plus the UTF-8 flag and Info-ZIP Unicode Path extra field (0x7075).
3. Score Hangul density to auto-detect CP949 ZIP repair vs UTF-8. Override if auto is wrong.
4. Preview broken vs repaired names, then remux: file payloads stay compressed as-is, headers are rewritten as UTF-8.
5. If you already extracted on a Mac and re-zipped the garbled names, switch to Already extracted (mojibake). That path maps Latin-1 soup back to the original CP949 bytes (or undoes UTF-8-as-Latin-1 like 안녕 → 안녕) instead of re-reading header bytes.

If a Windows teammate sends 자료.zip and macOS shows garbage folder names, do not extract-and-rezip. Repair the original archive here and share the UTF-8 ZIP so both sides see the same Korean names. If the only copy left is the remuxed archive, use mojibake undo.

FAQ

Is my ZIP uploaded?

No. Decoding and header rewrite stay in the browser. Encrypted and ZIP64 archives are not supported.

Can it fix a folder I already extracted?

Prefer the ZIP created on Windows. If that is gone and you only have a remuxed archive of the garbled names, choose Already extracted (mojibake) (or leave Auto). The tool treats each stored name as Latin-1 soup, maps those bytes back to CP949, and also undoes UTF-8-as-Latin-1 (안녕). A loose folder on disk still cannot be repaired — zip it first.

What if auto-detect is wrong?

Pick CP949 / EUC-KR for a Windows Explorer archive, or Already extracted for a remuxed Mac zip. The last method and encoding are saved in this browser. ASCII-only names are left unchanged. Use Fixed names only when the archive is large.