Encode any Elm string in Base64.

encode "Hello world!"
--> "SGVsbG8gd29ybGQh"

The encoding process appends = or == so that the resulting string's length is always a multiple of 4. This is sometimes referred to as padded encoding.

encode "Hi"
--> "SGk="

During encoding, strings are converted to UTF-8. The encoder correctly handles any valid Elm string, including characters outside the BMP, like Emoji.

encode "👍"
--> "8J+RjQ=="

Decode Base64-encoded strings into Elm strings.

decode "SGVsbG8gd29ybGQh"
--> Ok "Hello world!"

When decoding fails, an Err is returned to explain what went wrong. Currently, this may happen under two separate circumstances: either the input is not a valid Base64 string, or it contains invalid character sequences that cannot be converted to an UTF-16 string.

decode "What is this"
--> Err "Invalid base64"


decode "/Ng9"
--> Err "Invalid UTF-16"

Decoding assumes that the Base64 was encoded from an UTF-8 string. Note that this library converts the input to UTF-8 when encoding, so decoding should always succeed with strings generated by encode. Trailing = characters may be omitted.

decode "8J+RjQ"
--> Ok "👍"


decode "8J+RjQ=="
--> Ok "👍"