HTML Entity Encoder / Decoder

Paste a line of user-typed text straight into a page and the first stray < swallows the rest of your paragraph into a broken tag; a lone & can mangle the character that follows it. Escaping rewrites those characters as references the browser draws as literal symbols, so the text shows up exactly as written instead of being parsed as markup.

How to use

1. Leave Direction on Encode to turn text into entities, or switch it to Decode to read entities back as plain characters.
2. Under Characters to encode, keep Special characters only for ordinary page text, or pick Everything non-ASCII too when the result has to be pure ASCII (accents, dashes and emoji become references as well).
3. Tick or clear Prefer named references to choose between readable names like é and numeric forms like é.
4. Type or paste into the input box — the output updates as you type. Swap ⇅ moves the result into the input and flips the direction in one press.
5. Copy the output, or Download it as a .txt file. When decoding, the scope and named-reference controls switch off, since decoding accepts every form at once.

How it works

The five characters HTML treats specially — &, <, >, " and ' — never need the full entity tables, so encoding them runs off a small built-in lookup: each match is swapped for its named reference (&, <, and so on) or its decimal reference (&, <) depending on the checkbox. The heavier work — encoding every non-ASCII character, and all decoding — is handed to the he library, whose reference tables load once on first use rather than with the page. Very large pastes are processed in 256k-character slices so a big block never freezes the tab.

Take Beyoncé & Jay — 2 < 3. In special characters only mode the output is Beyoncé & Jay — 2 < 3 — only the ampersand and the less-than sign are touched; the accent and em dash stay as raw characters, which is correct on any UTF-8 page. Switch to everything non-ASCII too with named references on and it becomes Beyoncé & Jay — 2 < 3, now valid plain ASCII. Turn named references off and the same input yields Beyoncé & Jay — 2 < 3. Decoding any of those three reverses cleanly to the original line.

Use cases & limitations

You reach for encoding when you are building HTML by hand — a templating layer without automatic escaping, an email body, an RSS or Atom feed, or a value going into a CMS field that renders raw. The everything non-ASCII mode earns its keep for content that a legacy system might re-encode as Latin-1, where a bare é could break but é never will. Decoding is the everyday companion: pasting a log line, an API response, or scraped markup and wanting to read what it actually says.

The honest limit is that this escapes for two contexts only — text between tags and values inside quoted attributes. It is not a general sanitiser: a javascript: URL in an href, an unquoted attribute, or anything inside a <script> block needs context-specific handling that entity encoding does not provide. For encoding data destined for a query string or path segment, the URL encoder / decoder is the right tool instead, and if you are round-tripping text through binary-safe transport, the Base64 encoder / decoder covers that. To see how encoded snippets render, drop them into the Markdown preview.