| RFC 9285 | Base45 | July 2022 |
| Fältström, et al. | Informational | [Page] |
This document describes the Base45 encoding scheme, which is built upon the Base64, Base32, and Base16 encoding schemes.¶
This document is not an Internet Standards Track specification; it is published for informational purposes.¶
This document is a product of the Internet Engineering Task Force (IETF). It represents the consensus of the IETF community. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG). Not all documents approved by the IESG are candidates for any level of Internet Standard; see Section 2 of RFC 7841.¶
Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at https://www.rfc-editor.org/info/rfc9285.¶
Copyright (c) 2022 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
A QR code is used to encode text as a graphical image. Depending on the characters used in the text, various encoding options for a QR code exist, e.g., Numeric, Alphanumeric, and Byte mode. Even in Byte mode, a typical QR code reader tries to interpret a byte sequence as text encoded in UTF-8 or ISO/IEC 8859-1. Thus, QR codes cannot be used to encode arbitrary binary data directly. Such data has to be converted into an appropriate text before that text could be encoded as a QR code. Compared to already established Base64, Base32, and Base16 encoding schemes that are described in [RFC4648], the Base45 scheme described in this document offers a more compact QR code encoding.¶
One important difference from those others and Base45 is the key table and that the padding with '=' is not required.¶
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here.¶
Encoded data is to be interpreted as described in [RFC4648] with the exception that a different alphabet is selected.¶
QR codes have a limited ability to store binary data. In practice, binary data have to be encoded in characters according to one of the modes already defined in the standard for QR codes. The easiest mode to use in called Alphanumeric mode (see Section 7.3.4 and Table 2 of [ISO18004]. Unfortunately Alphanumeric mode uses 45 different characters which implies neither Base32 nor Base64 are very effective encodings.¶
A 45-character subset of US-ASCII is used; the 45 characters usable in a QR code in Alphanumeric mode (see Section 7.3.4 and Table 2 of [ISO18004]). Base45 encodes 2 bytes in 3 characters, compared to Base64 that encodes 3 bytes in 4 characters.¶
For encoding, two bytes [a, b] MUST be interpreted as a number n in Base256, i.e. as an unsigned integer over 16 bits so that the number n = (a*256) + b.¶
This number n is converted to Base45 [c, d, e] so that n = c + (d*45) + (e*45*45). Note the order of c, d and e which are chosen so that the left-most [c] is the least significant.¶
The values c, d and e are then looked up in Table 1 to produce a three character string. The process is reversed when decoding.¶
For encoding a single byte [a], it MUST be interpreted as a Base256 number, i.e. as an unsigned integer over 8 bits. That integer MUST be converted to Base45 [c d] so that a = c + (45*d). The values c and d are then looked up in Table 1 to produce a two-character string.¶
A byte string [a b c d ... x y z] with arbitrary content and arbitrary length MUST be encoded as follows: From left to right pairs of bytes MUST be encoded as described above. If the number of bytes is even, then the encoded form is a string with a length that is evenly divisible by 3. If the number of bytes is odd, then the last (rightmost) byte MUST be encoded on two characters as described above.¶
For decoding a Base45 encoded string the inverse operations are performed.¶
If binary data is to be stored in a QR code, the suggested mechanism is to use the Alphanumeric mode that uses 11 bits for 2 characters as defined in Section 7.3.4 of [ISO18004]. The ECI mode indicator for this encoding is 0010.¶
On the other hand if the data is to be sent via some other transport, a transport encoding suitable for that transport should be used instead of Base45. For example, it is not recommended to first encode data in Base45 and then encode the resulting string in Base64 if the data is to be sent via email. Instead, the Base45 encoding should be removed, and the data itself should be encoded in Base64.¶
The Alphanumeric mode is defined to use 45 characters as specified in this alphabet.¶
| Value Encoding | Value Encoding | Value Encoding | Value Encoding |
|---|---|---|---|
| 00 0 | 12 C | 24 O | 36 Space |
| 01 1 | 13 D | 25 P | 37 $ |
| 02 2 | 14 E | 26 Q | 38 % |
| 03 3 | 15 F | 27 R | 39 * |
| 04 4 | 16 G | 28 S | 40 + |
| 05 5 | 17 H | 29 T | 41 - |
| 06 6 | 18 I | 30 U | 42 . |
| 07 7 | 19 J | 31 V | 43 / |
| 08 8 | 20 K | 32 W | 44 : |
| 09 9 | 21 L | 33 X | |
| 10 A | 22 M | 34 Y | |
| 11 B | 23 N | 35 Z |
It should be noted that although the examples are all text, Base45 is an encoding for binary data where each octet can have any value 0-255.¶
Option (a): added line breaks¶
Encoding example 1:¶
Encoding example 2:¶
Encoding example 3:¶
Decoding example 1:¶
This document has no IANA actions.¶
When implementing encoding and decoding it is important to be very careful so that buffer overflow or similar does not occur. This of course includes the calculations in Base45 and lookup in the table of characters (Table 1). A decoder must also be robust regarding input, including proper handling of any octet value 0-255, including the NUL character (ASCII 0).¶
It should be noted that Base64 and some other encodings pad the string so that the encoding starts with an aligned number of characters while Base45 specifically avoids padding. Because of this, special care has to be taken when odd numbers of octets are to be encoded. Similarly, care must be taken if the number of characters to decode are not evenly divisible by 3.¶
Base encodings use a specific, reduced alphabet to encode binary data. Non-alphabet characters could exist within base-encoded data, caused by data corruption or by design. Non-alphabet characters may be exploited as a "covert channel", where non-protocol data can be sent for nefarious purposes. Non-alphabet characters might also be sent in order to exploit implementation errors leading to, for example, buffer overflow attacks.¶
Implementations MUST reject any input that is not a valid encoding. For example, it MUST reject the input (encoded data) if it contains characters outside the base alphabet (in Table 1) when interpreting base-encoded data.¶
Even though a Base45-encoded string contains only characters from the alphabet in Table 1, cases like the following have to be considered: The string "FGW" represents 65535 (FFFF in Base16), which is a valid encoding of 16 bits. A slightly different encoded string of the same length, "GGW", would represent 65536 (10000 in Base16), which is represented by more than 16 bits. Implementations MUST also reject the encoded data if it contains a triplet of characters that, when decoded, results in an unsigned integer that is greater than 65535 (ffff in Base16).¶
It should be noted that the resulting string after encoding to Base45 might include non-URL-safe characters so if the URL including the Base45 encoded data has to be URL-safe, one has to use percent-encoding.¶
The authors thank Mark Adler, Anders Ahl, Alan Barrett, Sam Spens Clason, Alfred Fiedler, Tomas Harreveld, Erik Hellman, Joakim Jardenberg, Michael Joost, Erik Kline, Christian Landgren, Anders Lowinger, Mans Nilsson, Jakob Schlyter, Peter Teufl, and Gaby Whitehead for the feedback. Also, everyone who has been working with Base64 over a long period of years and has proven the implementations are stable.¶