Unicode Escape Converter

Turns text into escape sequences and back. Encoding covers JavaScript and JSON \uXXXX, the U+XXXX notation used in specifications and bug reports, and HTML entities in both hex and decimal — keeping plain ASCII readable, or escaping every character if the destination is fussy. Decoding recognises all of those at once, plus \u{...}, \xXX and bare U+, so a log line with three different escape styles mixed together comes back as the text it was.

The reason \uXXXX needs company is that it is a UTF-16 escape, and UTF-16 only has room for four hex digits. Characters above U+FFFF have to be written as a surrogate pair: a grinning face is U+1F600, and in a JavaScript string it is \uD83D\uDE00, two escapes that mean one character. The newer \u{1F600} says the same thing in one piece, and the HTML forms are 😀 and 😀. All four are the same character; only the first exposes the machinery.

That machinery is worth understanding because it leaks. The values D800 through DFFF are reserved for the two halves of a pair and are not characters in their own right, so a lone surrogate — produced by slicing a string at the wrong offset, or by a truncation that counted UTF-16 units — is not valid text and cannot be encoded to UTF-8 at all. It survives inside a JavaScript string and then fails at the moment something tries to write it to a file or a network socket, which is a long way from where it was created.

The practical use is getting non-ASCII through a channel that cannot be trusted with it. A build pipeline that mangles UTF-8, a config format with an uncertain encoding, a system that accepts only ASCII bytes: escaping every non-ASCII character makes the data survive intact and stay diff-able, at the cost of being unreadable to a human. The reverse is more common day to day — pasting an escaped string out of a log or a stack trace to find out what it actually said. Everything runs in your browser.