Base16 writes each byte as two hexadecimal digits, so valid byte-oriented hex always has an even number of nibbles. This decoder enforces that invariant, normalizes letter case, and shows both the byte array and UTF-8 interpretation when the bytes form valid text.
Hex carries no type information. The same characters can be a digest, a color, an integer, a public-key component, or arbitrary bytes. tryb therefore lowers generic-hex confidence at common hash lengths and CSS color lengths instead of pretending that successful decoding proves intent.
A 0x prefix is strong evidence that the author meant hexadecimal, while an unprefixed string of only decimal digits is much weaker. Control bytes are reported even when the overall sequence is valid UTF-8 because invisible bytes can change terminal, log, and parser behavior.
Edge cases worth knowing
- Odd nibble counts
- abc contains one complete byte and half of another, so it is rejected rather than silently padded.
- Digest-sized values
- A 32-, 40-, 64-, 96-, or 128-character value remains available as bytes, but the dedicated hash interpretation ranks higher.
- Invalid UTF-8
- The bytes still decode as hex even when they are not UTF-8. tryb shows a null text interpretation rather than inserting replacement characters.
Common mistakes
- Assuming every hexadecimal string is text rather than arbitrary binary data.
- Padding an odd nibble count without knowing whether the missing half-byte belongs at the start or end.
- Calling a value a hash solely because it is hexadecimal; length is only a clue, not proof of algorithm or provenance.
- Treating hex as encryption or secret storage.