TypeScript icon, indicating that this package has built-in type declarations

1.1.0 • Public • Published


Base62 encoder/decoder


npm install --save @fry/base62


encode() expects data that can be turned into a Buffer (e.g. a string, an Array, ArrayBuffer, or a Buffer) and it returns the Base62 encoded string.

decode() returns the decoded data in a Buffer which has good readers for binary data as well as a fast toString() that can return the data encoded in utf8 (default), base64, hex, etc.

Binary input

    const base62 = require('@fry/base62');

    const encoded = base62.encode([0x2a, 0x2a]);
    // encoded => AYA

    const decoded = base62.decode(encoded);
    // decoded => <Buffer 2a 2a>

    decoded[0]; // => 0x2a
    decoded.readUInt16BE(); // => 0x2a2a

String input

    const encoded = base62.encode('Memento vivere');
    // encoded => JMLjPMvqRo1sQNPbSc5

    const decoded = base62.decode(encoded);
    // decoded => <Buffer 4d 65 6d 65 6e 74 6f 20 76 69 76 65 72 65>

    decoded.toString(); // => Memento vivere

Custom symbol table

There is no standard for Base62 encoding so various encoders may use different symbol tables. This module will by default use alphanumerical characters in their natural order: 0-9A-Za-z.

To use another set of characters first initiate them with prepareSymbolTable() and then supply the resulting object as a second parameter to the encode/decode functions, e.g.:

    const symbols = base62.prepareSymbolTable('ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789');
    const encoded = base62.encode([0x53, 0xfe, 0x92], symbols);
    // encoded => U98kC

    const decoded = base62.decode(encoded, symbols);
    // decoded => <Buffer 53 fe 92>

Base62 encoding process

The Base62 symbol table is limited to 62 characters, which is just shy of 6 bits of information. This means that special care needs to be taken in order to encode any given stream of bytes. This module takes the approach of reading a sextet (6-bit chunk) from the binary stream and encode it into the symbol table as long as the chunk is not in risk of overflowing. If the sextet is in the ”overflow zone“ then only 5 bits are used while encoding that particular chunk, indicated in the output by the presence of one of the symbol table’s last two characters. The overflow bit is left in the stream until the next sextet will be read. The last chunk may also be left-padded with zeroes during the encoding process which the decoder will detect and reverse to ensure byte-alignment of the underlying binary stream.

Happy encoding!

/ Fry


npm i @fry/base62

DownloadsWeekly Downloads






Unpacked Size

16.6 kB

Total Files


Last publish


  • fry