Convert an internationalized domain to Punycode

The job

An internationalized domain name shows up as xn--… in configs, certificates and DNS records. Converting each label to Punycode (RFC 3492) is how the unicode name travels, and decoding it back reveals what a suspicious xn-- string actually says.

How to do it

Paste the domain — with unicode labels or xn-- labels — pick the direction, and convert. Each dot-separated label is processed independently, so mixing ascii and unicode parts is fine.

• Only labels with non-ASCII characters get the xn-- prefix; plain-ASCII labels pass through.
• Full-width dots (。 and ．) count as separators — a paste from Japanese text still parses.

Worked example

bücher.example becomes xn--bcher-kva.example for DNS, and xn--wgv71a119e.jp decodes back to 日本語.jp — useful when a certificate or phishing email hides behind the encoded form.

Limits

This is label-level Punycode, not a full URL parser — paths and query strings are rejected. Labels are lowercased as DNS is case-insensitive, and there is no homograph detection: a decoded lookalike domain still needs a human check.