pub fn base64_decode_standard(text: &str) -> Option<Vec<u8>>Expand description
What:
Decode standard padded base64 (RFC 4648 section 4) into bytes,
or nothing when the text is not exactly what base64_standard would print for some bytes.
Why:
A journal that holds the exact bytes of a name that is not UTF-8 is read back by the wrapper;
a text that
decodes in more than one way (stray padding, non-zero spare bits) is a damaged record,
not a name.
In TS you’d write (pseudocode):
function base64Decode(text: string): Uint8Array | undefined { const bytes = Buffer.from(text, 'base64'); return bytes.toString('base64') === text ? bytes : undefined; }