Reverse Text

About text reversal

A reverse-text tool flips a string so that the last character becomes the first, the second-to-last becomes the second, and so on. It is the most basic non-identity string operation and shows up in interview questions, palindrome checks, brain-teaser puzzles, escape-room clues, and right-to-left display experiments. The naïve implementation in JavaScript, str.split('').reverse().join(''), dates from the language's first standard library and has appeared in beginner tutorials since 1996, but it has a subtle bug: it splits on UTF-16 code units rather than Unicode code points, so emoji and other astral characters get torn in half.

This tool uses the modern, Unicode-aware approach: it converts the string to an array with Array.from(str) or the spread iterator, which walks proper Unicode code points and keeps surrogate pairs together. Where the browser exposes Intl.Segmenter('en', {granularity: 'grapheme'}), we also segment grapheme clusters so multi-codepoint emoji (family, profession, regional flags) reverse as one unit instead of scrambling.

How the reverser works

// Code-point safe one-liner
output = [...input].reverse().join('')

// Grapheme-cluster safe (handles ZWJ emoji sequences)
const seg = new Intl.Segmenter('en', {granularity: 'grapheme'})
output = [...seg.segment(input)].map(s => s.segment).reverse().join('')

• Code-point segmentation. The spread operator iterates code points (using the StringIterator from ECMAScript 2015) rather than UTF-16 16-bit code units, so a rocket emoji at U+1F680 stays as one unit instead of becoming two unpaired surrogates.
• Reverse the array. Array.prototype.reverse() mutates in place in O(n) time, walking from both ends toward the middle.
• Rejoin. join('') concatenates with no separator, producing a single output string of the same code-point length as the input.
• Involutive. Reversing the output again returns the original, with one caveat: NFD decomposed accents end up before their base letter and stay that way unless you renormalise with str.normalize('NFC').

Worked example

Input: a mixed string with ASCII, an accented letter, and an emoji.

Café 2026 🚀

1. Spread to code points. ['C', 'a', 'f', 'é', ' ', '2', '0', '2', '6', ' ', '🚀']. The rocket is one code point at U+1F680; in UTF-16 storage it is a surrogate pair but the spread iterator yields it as one element.
2. Reverse the array. ['🚀', ' ', '6', '2', '0', '2', ' ', 'é', 'f', 'a', 'C'].
3. Join. Concatenate with empty separator.

Contrast with the naive split: 'Café 2026 🚀'.split('').reverse().join('') on Chrome 119 yields the same visible result for this string because Café uses precomposed é, but a string built with NFD decomposed accents or a flag emoji (which uses two regional indicator code points) would scramble.

Reversal under different segmentations

Pitfalls to watch for

• Decomposed accents jump positions. a + U+0301 combining acute renders as á. After reversal the combining mark sits before the base letter and attaches to whatever was previously to its right. Normalise to NFC before reversing if you handle mixed input.
• Flag emoji split in two. The Indian flag is two regional indicator letters (IN). Each is a separate code point, so without grapheme-level segmentation the reversal produces NI, the Nicaraguan flag. Use the Intl.Segmenter path.
• Bidi control characters strand. U+202E right-to-left override and U+202D left-to-right override change the display order of everything after them. Reversal moves them mid-string and can produce surprising visual results.
• URLs become invalid. Reversing https://example.com/path produces htap/moc.elpmaxe//:sptth, which is not a parseable URL.
• It is not encryption. Reversed text is decoded instantly by any tool, including this one. Use proper cryptography for anything sensitive.
• Word-reversal needs a different tool. If you want sentence-with-reversed-words rather than reversed characters, split on whitespace, reverse the array, and rejoin with a space.

Related text utilities

Frequently asked questions