URL Encoder / Decoder — Percent-Encoding for Safe Transmission of Any Character

URLs are restricted to a specific set of characters defined in RFC 3986: letters, digits, and the characters - _ . ~ in the unreserved set, plus a small set of reserved characters that have structural meaning in URL syntax (: / ? # [ ] @ ! $ & ' ( ) * + , ; =). Any character outside this set — spaces, non-ASCII characters like accented letters and CJK characters, and characters that conflict with URL syntax — must be percent-encoded: replaced with a percent sign followed by the two-digit hexadecimal value of the character's UTF-8 encoding. A space becomes %20. The equals sign in a query parameter value becomes %3D. The ampersand that separates query parameters becomes %26. A Chinese character like 你 becomes %E4%BD%A0. Without this encoding, URLs break at every character that conflicts with URL syntax rules.

Query string encoding and path encoding follow different rules that create a common source of bugs. In a query string, the space character can be encoded as either %20 (standard percent encoding) or + (form encoding, the convention HTML forms use). Path components must use %20 — a + in a URL path is a literal plus sign, not a space. API clients and server-side URL parsers that mix these conventions produce mismatched encoding: a server that decodes + as space in paths will misinterpret literal plus signs in filenames. The encoder distinguishes between query string encoding (using + for spaces by default, following form encoding convention) and path encoding (using %20 for spaces, following strict RFC 3986), with a toggle to switch between conventions for contexts where the distinction matters.

Internationalized Domain Names (IDN) and non-ASCII URL paths require an additional encoding step called Punycode for the hostname component and UTF-8 percent encoding for the path. A domain like café.com cannot be transmitted in DNS as-is — the é character is not in the DNS protocol's character set. Punycode converts it to xn--caf-dma.com, the ASCII-compatible encoding that DNS can transmit. The path /über/straße becomes /%C3%BCber/stra%C3%9Fe in percent-encoded form. Browsers display the decoded human-readable form in the address bar while transmitting the encoded form over the network — which is why URLs look readable in the browser but produce encoding surprises when copied into API calls or server logs. The encoder handles both IDN and path encoding to produce the wire-format URL that servers actually receive.