Unicode Escape & Unescape

Because the classic \uXXXX form can only describe values up to FFFF, and emoji live far above that. Anything beyond that range is stored as a surrogate pair — two halves that together encode one character — so an emoji becomes something like \ud83d\ude00. The pair is correct and every JavaScript engine reassembles it into the single character. If you would rather see one sequence, switch to the ES6 \u{...} form, which has no such ceiling and writes the same emoji as \u{1f600}.

The braced form was added in ES6 and can hold any code point in one sequence, so it is both shorter and far easier to read for characters above FFFF. The unbraced form is limited to four hex digits but works absolutely everywhere, including old engines, JSON, Java and C#. JSON in particular only accepts the unbraced form — putting \u{1f600} in a JSON document produces a parse error — so pick based on where the escaped text is going, not on which looks nicer.

When the channel between you and the destination cannot be trusted to carry the bytes intact. That covers a surprising amount: build pipelines that mangle encodings, older APIs that only accept ASCII, database clients configured for the wrong charset, and configuration files edited by tools that rewrite them. Escaping turns every character into plain ASCII, which survives all of that. It is also invaluable for debugging — escaped text makes an invisible character visible, which is often the whole reason two identical-looking strings do not match.

That is one of its most useful jobs. Zero-width spaces, non-breaking spaces, byte-order marks and directional marks are all completely invisible on screen and all break string comparisons, URL slugs and search. Escaping makes each one appear as a distinct code, so a stray \u200b or \ufeff in text that looks perfectly ordinary becomes obvious. If two strings look identical but refuse to compare equal, escaping both is usually the fastest way to see why.

No. It is a purely reversible representation, exactly like hexadecimal — the same characters written a different way. Unescaping returns the original text byte for byte. That also means it offers no security or privacy whatsoever: escaped text is trivially readable by anyone, and treating it as obscured is a mistake. It is for surviving transport and for making the invisible visible, not for hiding anything.

Because escaping plain ASCII would balloon the output for no benefit — an English sentence would become several times longer and much harder to read. The default escapes only what is outside printable ASCII, which is precisely what is at risk in transport. You can switch to escaping everything when a destination is strict about the character set, or restrict it to non-printing characters only when you want to spot invisible ones without disturbing accented letters you can already read.