Eight Backwards Styles and Four Single Reversals
Compare eight outputs for one source: full reversal, token order, sentence order, alternating case and four explicitly decorated reversals. Below them a single-output panel offers four plain reversals: Reverse Characters, Reverse Word Order, Reverse Each Word and Reverse Lines. Mirror Text adds mirror symbols around a reversal; it does not reflect letters, and reversing text does not encrypt a message.
Six Choices With Their Exact Outputs
| Input | Choice | Output |
|---|---|---|
Hello | Character Reverse | olleH |
red blue | Word Reverse | blue red |
red blue | Reverse Each Word | der eulb |
One! Two? | Sentence Reverse | Two? One! |
a b c | Alternating Case | a B c |
A
B
| Reverse Lines |
B
A |
Paste Once, Compare, Then Copy
- Paste one original message or load Sample.
- Compare all eight labeled, selectable outputs and review punctuation and spacing in Sentence Reverse.
- For one plain operation, choose a mode in the single-reversal panel and inspect spaces, punctuation, blank lines and direction.
- Copy the chosen output with its own Copy button; changing the input clears feedback.
Characters, Tokens, Sentences and Lines
All reversed character outputs use whole grapheme clusters from Intl.Segmenter with the English locale. In Reverse Characters, CRLF remains one cluster but line breaks move with the reversal. Word Reverse and Sentence Reverse work independently inside CRLF, LF or CR physical lines and keep whitespace slots. Reverse Each Word reverses the clusters inside each token. Reverse Lines reverses line contents while retaining separators in their original slots, including the trailing empty line.
Sentence boundaries come from English Intl.Segmenter, including terminal punctuation; abbreviations and unfinished text can be ambiguous. Alternating Case changes successive clusters containing letters, beginning lowercase; spaces and symbols do not advance the case counter. Case conversion can expand letters and is not reversible.
Where Reversal Misleads
Word tokens are separated by whitespace-only grapheme clusters, not linguistic words, and punctuation stays with its token. Grapheme segmentation is not visual glyph reflection or language-aware reading order. A newly leading combining mark or regrouped regional indicators can change segmentation after reversal, so applying it twice is not a universal lossless round trip.
Decorated outputs add literal symbols to the copied text; they do not carry font styling. Sentence reversal is not sentence rewriting, and Unicode case mapping is not a guarantee of correct language-specific capitalization. Do not use reversal for confidential data or assume a decorated nickname is accepted by a destination. Text Mirror uses approximate character substitutions rather than decoration. Unicode text segmentation explains grapheme boundaries, and Unicode bidirectional text explains why stored order can differ from displayed order.