A Base62 encoder / decoder with support for leading zero bytes.
Normally, during the conversion of base10 (decimal) to base62, the input data is interpreted as one large number:
[0x00, 0x13, 0x37] => 0x001337 => 4919 (decimal)
As leading zeroes do not count to the value of a number (001337 = 1337),
they are ignored while converting the number to base62.
encode_num] function does.encode_data] keeps these leading zeroes.This is achieved by prepending a 0x01 byte to the data before encoding,
thus creating a number that starts with a 1: 001337 => 1001337 (No zeroes
are removed)
The leading 0x01 is removed after the data has been decoded from bse62 back
to base10: 1001337 => 001337
txt
0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz
This is the same alphabet that
CyberChef
uses by default: [0-9][A-Z][a-z].
Wikipedia/Base62 suggests another
alphabet ([A-Z][a-z][0-9]) but I found that many online converters use
either [0-9][A-Z][a-z] or [0-9][a-z][A-Z]. And as I love
CyberChef, I decided to use their
default alphabet aswell. I also think that starting with numbers is more
natural as base62 is actually a number system like decimal (which is actually
base10) or hexa-decimal (base16).
This method will prepend 0x01 to the data before encoding it.
```rust let data = vec![0x13, 0x37]; let encoded = bs62::encode_data(&data);
assert_eq!(encoded, "IKN") ```
This method expects a leading 0x01 in the byte array after decoding. It
removes the first byte before returning the byte array.
```rust let encoded = "IKN"; let data = bs62::decode_data(&encoded)?;
asserteq!(data, vec![0x13u8, 0x37]); ```
```rust let num = 1337; let encoded = bs62::encode_num(&num);
assert_eq!(encoded, "LZ") ```
```rust let num = 1337; let encoded = bs62::encode_num(&num);
assert_eq!(encoded, "LZ") ```
bs62 is licensed under MIT.
See the LICENSE.txt file in this repository for more information.