IDNA

Internationalized Domain Names in Applications, or IDNA, is a family of standards that allow domain names to contain non-ASCII characters from languages around the world. It enables use of scripts such as Arabic, Cyrillic, Chinese, Hindi, and many others in user interfaces and applications while preserving compatibility with the existing DNS, which is ASCII-based.

IDNA separates the user-facing form (U-label) from the DNS form (A-label). When a user enters a domain in Unicode, the application converts each label to ASCII using Punycode, producing an A-label with an xn-- prefix that can be queried in DNS. This encoding is defined by RFC 3492; the mapping to and from Unicode labels is regulated by the IDNA standards.

IDNA began with IDNA2003, defined in RFC 3490, 3491, and 3492, and was revised by IDNA2008, published as RFC 5890–5893. IDNA2008 introduced stricter rules for allowed characters, handling of combining marks, and restrictions to reduce security and compatibility issues. The standards specify preparation steps, label processing, and how to handle RTL scripts, contextual rules, and prohibition of certain characters.

Applications such as web browsers, email clients, and other Internet software implement IDNA to display Unicode strings to users while performing DNS queries using ASCII representations. Registry and registrar policies, as well as software implementations, determine which scripts are allowed for registrable domain labels and how homographs are treated.

Security and usability considerations include risks of spoofing via visually similar characters, the need for script restrictions, and the possibility of confusion between homoglyphs. Ongoing work in the IETF maintains and updates the IDNA standards in response to evolving Unicode versions and security research.