Fix garbled text encoding (mojibake)

Mojibake (a Japanese word for garbled characters) happens when text saved in one character encoding is read as another. The most common case today: text saved as UTF-8 is opened as Windows-1252 (also called Latin-1 or "ANSI"). Each special character is stored as two or three bytes in UTF-8, and the wrong decoder shows each byte as a separate character:

You see this in CSV exports opened in Excel, old database dumps, email subjects, subtitle files and website content migrated between systems.

The tool reverses the mistake: it turns the garbled characters back into the bytes they came from, using the encoding they were wrongly read with, and decodes those bytes again as UTF-8. It tries Windows-1252, Windows-1250 and ISO-8859-2 and keeps the version that removes the garbage without creating new errors.

• Double encoding, where text was garbled twice and looks like é, is undone in up to three rounds.
• Mixed text, where some lines are fine and others garbled, is fixed line by line, and only the broken pieces are changed.
• Each step applied is listed under the result, so you can see what happened.

Everything runs in your browser; your text and files are not uploaded.

The replacement character � (U+FFFD) means a program already gave up: it met bytes it could not decode and replaced them. The original letter is gone and cannot be recovered from the text itself. The tool counts these and warns you.

In that case, go back to the original file, the export from the source system, and open it here with Open file. The tool reads the raw bytes: UTF-8 if valid, otherwise Windows-1252, and also UTF-16 files with a byte order mark.

Excel on Windows assumes a CSV without a marker is in the local legacy encoding, so a UTF-8 export with accented names often shows é and similar sequences. Three ways to avoid it:

1. Download the fixed text with Download as UTF-8 (with BOM, for Excel). The byte order mark (BOM) tells Excel the file is UTF-8, and double-clicking opens it correctly.
2. In Excel, use Data → From Text/CSV and choose 65001: Unicode (UTF-8) as the file origin.
3. Open the file in Google Sheets, which reads UTF-8 by default.

To browse and edit a CSV without Excel at all, try the CSV viewer.

• Save and exchange text as UTF-8 everywhere: files, databases, APIs and web pages.
• Declare it: <meta charset="utf-8"> in HTML, Content-Type: text/plain; charset=utf-8 in HTTP headers, utf8mb4 in MySQL.
• When importing data, pick the encoding explicitly instead of letting the program guess.
• If a file looks garbled, do not save it again in the program that shows it wrong; that can make the damage permanent. Fix it from the original.