cajeta.codec.Base64 — Base64 (RFC 4648) byte ⇄ text codec. Supports both alphabets in §4 (standard, + /) and §5 (URL- and filename-safe, - _), with explicit control over padding. Decoding accepts either alphabet transparently, so a caller need not know which variant produced the text (an HTTP Authorization: Basic … header vs. a JWT segment). A character outside the alphabet or a stripped length of 1 mod 4 raises Base64Exception with the offending byte offset; trailing = padding is consumed and ignored, so unpadded URL-safe input round-trips without the caller re-adding =.

int8[] data = heap int8[3];
data[0] = (int8) 77;
data[1] = (int8) 97;
data[2] = (int8) 110;
String b #= Base64.encode(data, (int64) data.count());        // "TWFu"
String u #= Base64.encodeUrlSafe(data, (int64) data.count()); // "TWFu" (no '=')
int8[] raw #= Base64.decode(b);       // either alphabet, padded or not

Methods

Signature
`static #String encode(int8[] data, int64 len)` ⚑	Encode `data[0..len)` to standard Base64 (§4) with `=` padding
`static #String encodeUrlSafe(int8[] data, int64 len)` ⚑	Encode to URL-/filename-safe Base64 (§5, `-`/`_`) without padding — the JWT / query-parameter form
`static #String encode(int8[] data, int64 len, boolean urlSafe, boolean pad)`	General encode: pick the alphabet and padding explicitly
`static #int8[] decode(String s)` ⚑	Decode a Base64 `String` (either alphabet, padded or not) to raw bytes; throws `Base64Exception` on garbage
`static #int8[] decodeBytes(int8[] data, int64 len)`	Decode `data[0..len)` (encoded text held as bytes) to raw bytes

⚑ = @EntryPoint

Base64

Methods

See also