]> MIT CSAIL

32 Vassar Street, Room 32-G692 Cambridge Massachusetts 02139 USA rivest@mit.edu https://www.csail.mit.edu/person/ronald-l-rivest

Independent

2386 Panoramic Circle Apopka Florida 32703 US +1-508-333-2270 d3e3e3@gmail.com

Applications and Real Time Network Working Group Sexp This memo specifies a data structure representation that is suitable for representing arbitrary, complex data structures. It was devised in 1996/1997 to support SPKI (RFC 2692) certificates with the intent that it be more widely applicable and has been used elsewhere. There are many implementations in a variety of programming languages. Uses of this representation herein are referred to as "S-expressions". This memo makes precise the encodings of these S-expressions: it gives a "canonical form" for them, describes two "transport" representations, and also describe an "advanced" format for display to people.

Introduction This memo specifies a data structure representation that is suitable for representing arbitrary, complex data structures. It was devised in 1996/1997 to support SPKI certificates with the intent that it be more widely applicable (see , History) and has been used elsewhere. Uses of this representation herein are referred to as "S-expressions". This memo makes precise the encodings of these S-expressions: it gives a "canonical form" for them, describes two "transport" representations, and also describe an "advanced" format for display to people. There are many implementations of S-expression in a variety of programming languages including Python, Ruby, and C (see ). These S-expressions are either byte-strings ("octet-strings") or lists of simpler S-expressions. Here is a sample S-expression: (snicker "abc" (#03# |YWJj|)) It is a list of length three containing the following:

• the octet-string "snicker"
• the octet-string "abc"
• a sub-list containing two elements: the hexadecimal constant #03# (i.e., 0x03) and the base-64 constant |YWJj| (which is the same as "abc")

This document specifies how to construct and use these S-expressions. They are independent of any particular application. The design goals for S-expressions were as follows:

generality:

S-expressions should be good at representing arbitrary data.

readability:

It should be easy for someone to examine and understand the structure of an S-expression.

economy:

S-expressions should represent data compactly.

tranportability:

S-expressions should be easy to transport over communication media (such as email) that are known to be less than perfect.

flexibility:

S-expressions should make it relatively simple to modify and extend data structures.

canonicalization:

It should be easy to produce a unique "canonical" form of an S-expression, for digital signature purposes.

efficiency:

S-expressions should admit in-memory representations that allow efficient processing.

Implementors of new applications / protocols may wish to consider potential alternative representations such as , CBOR , or JSON .

Uses of S-Expressions The S-expressions specified herein are in active use today between GnuPG and Ribose's RNP . Ribose has implemented C++ software to compose and parse these S-expressions . The GNU software is here and there are other implementations (see ). They are used/referenced in the following RFCs:

• for
• XML-Signature Syntax and Processing

In addition, S-Expressions are the inspiration for the encodings in other protocols. For example, or Section 6 of .

Historical Note The S-expression technology described here was originally developed for "SDSI" (the Simple Distributed Security Infrastructure by Lampson and Rivest ) in 1996, although the origins clearly date back to McCarthy's programming language. It was further refined and improved during the merger of SDSI and SPKI during the first half of 1997. S-expressions are more readable and flexible than, Bernstein's "net-strings" , which were developed contemporaneously.

Conventions Used in This Document 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 when, and only when, they appear in all capitals, as shown here.

S-expressions -- informal introduction Informally, an S-expression is either:

• an octet-string, or
• a finite list of simpler S-expressions.

An octet-string is a finite sequence of eight-bit octets. There may be many different but equivalent ways of representing an octet-string abc -- as a token "abc" -- as a quoted string #616263# -- as a hexadecimal string 3:abc -- as a length-prefixed "verbatim" encoding |YWJj| -- as a base-64 encoding of the octet-string "abc" {MzphYmM=} -- as a base-64 encoding of the verbatim encoding (that is, an encoding of "3:abc") The above encodings are all equivalent in that they all denote the same octet-string. Details of these encodings are given below, and how to give a "display type" to a simple-string is also described. A list is a finite sequence of zero or more simpler S-expressions. A list is represented by using parentheses to surround the sequence of encodings of its elements, as in: (abc (de #6667#) "ghi jkl") As can be seen, there is variability possible in the encoding of an S-expression. In some applications, it is desirable to standardize or restrict the encodings; in other cases, it is desirable to have no restrictions. The following are the target cases these s-expressions aim to handle:

• a "transport" or "basic" encoding for transporting the S-expression between computers.
• a "canonical" encoding, used when signing the S-expression.
• an "advanced" encoding used for input/output to people.
• an "in-memory" encoding used for processing the S-expression in the computer.

In this document, related encoding techniques for each of these uses are provided.

Character set This document describes encodings of S-expressions. Except when giving "verbatim" encodings, the character set used is limited to the following characters in ASCII :

Alphabetic:

A B ... Z a b ... z

Numeric:

0 1 ... 9

Whitespace:

space, horizontal tab, vertical tab, form-feed carriage-return, line-feed

The following graphics characters, which are called "pseudo-alphabetic" in this document:

- hyphen or minus . period / slash _ underscore : colon * asterisk + plus = equal

The following graphics characters, which are "reserved punctuation":

The following characters are unused and unavailable, except in "verbatim" and "quoted string" encodings:

greater than ? question mark ]]>

Octet-string representation types This section describes in detail the ways in which an octet-string may be represented. Recall that an octet-string is any finite sequence of octets, and that the octet-string may have length zero.

Verbatim representation A verbatim encoding of an octet-string consists of three parts:

• the length (number of octets) of the octet-string, given in decimal, most significant digit first, with no leading zeros.
• a colon ":"
• the octet-string itself, verbatim.

There are no blanks or whitespace separating the parts. No "escape sequences" are interpreted in the octet-string. This encoding is also called a "binary" or "raw" encoding. Here are some sample verbatim encodings: 3:abc 7:subject 4::::: 12:hello world! 10:abcdefghij 0:

Quoted-string representation The quoted-string representation of an octet-string consists of:

• an optional decimal length field
• an initial double-quote (")
• the octet-string with "C" escape conventions (\n, etc.)
• a final double-quote (")

The specified length is the length of the resulting string after any backslash escape sequences have been converted to the octet value they denote. The string does not have any "terminating NULL" that includes, and the length does not count such a character. The length is optional. The escape conventions within the quoted string are as follows (these follow the "C" programming language conventions, with an extension for ignoring line terminators of just LF, CRLF, or LFCR and more restrictive octal and hexadecimal value formats): \a -- audible alert (bell) \b -- backspace \t -- horizontal tab \v -- vertical tab \n -- new-line \f -- form-feed \r -- carriage-return \" -- double-quote \' -- single-quote \? -- question mark \\ -- back-slash \ooo -- character with octal value ooo (all three digits MUST be present) \xhh -- character with hexadecimal value hh (both digits MUST be present) \<carriage-return> -- causes carriage-return to be ignored. \<line-feed> -- causes linefeed to be ignored. \<carriage-return><line-feed> -- causes CRLF to be ignored. \<line-feed><carriage-return> -- causes LFCR to be ignored. Here are some examples of quoted-string encodings: "subject" "hi there" 7"subject" "\xFE is the same octet as \376" 3"\n\n\n" "This has\n two lines." "This has \ one." ""

Token representation An octet-string that meets the following conditions may be given directly as a "token":

• it does not begin with a digit;
• it contains only characters that are: alphabetic (upper or lower case), numeric, or one of the following eight "pseudo-alphabetic" punctuation marks:  -  .  /  _  :  *  +  =
• it is length 1 or greater.

Note: Upper and lower case are not equivalent. A token may begin with punctuation, including ":". Here are some examples of token representations: subject not-before class-of-1997 //microsoft.com/names/smith *

Hexadecimal representation An octet-string may be represented with a hexadecimal encoding consisting of:

• an (optional) decimal length of the octet-string
• a sharp-sign "#"
• a hexadecimal encoding of the octet-string, with each octet represented with two hexadecimal digits, most significant digit first. There MUST be an even number of such digits.
• a sharp-sign "#"

There may be whitespace inserted in the midst of the hexadecimal encoding arbitrarily; it is ignored. It is an error to have characters other than whitespace and hexadecimal digits. Here are some examples of hexadecimal encodings: #616263# -- represents "abc" 3#616263# -- also represents "abc" # 616 263 # -- also represents "abc" ## -- represents the zero length string

Base-64 representation of octet-strings An octet-string may be represented in a base-64 encoding consisting of:

• an (optional) decimal length of the octet-string
• a vertical bar "|"
• the base-64 encoding of the octet string.
• a final vertical bar "|"

Base-64 encoding produces four characters of output for each three octets of input. If the length of the input divided by three leaves a remainder of one or two, it produces an output block of length four ending in two or one equals signs, respectively. These equals signs MUST be included on output but input routines MAY accept inputs where one or two equals signs are dropped. Whitespace inserted in the midst of the base-64 encoding is ignored. It is an error to have characters other than whitespace and base-64 characters. Here are some examples of base-64 encodings: |YWJj| -- represents "abc" | Y W J j | -- also represents "abc" 3|YWJj| -- also represents "abc" |YWJjZA==| -- represents "abcd" |YWJjZA| -- also represents "abcd" || -- represents the zero length string Note the difference between this base-64 encoding of an octet-string using verticle bars ("| |") and the base-64 encoding of an S-expression using curly braces ("{ }") in .

Display-hint The purposes of a display-hint is to provide information on how to display an octet-string to a user. It has no other function. Many of the MIME types work here. A display-hint is an octet-string surrounded by square brackets. There may be whitespace separating the octet-string from the surrounding brackets. Any of the legal formats may be used for the octet-string but a dsaplay-hint may not be applied to a display-hint string, that is, display-hints may not be nested recursively. Every octet-string is either preceded by a single "display-hint" or not so preceded. Here are some examples of display-hints: [image/gif] [charset=unicode-1-1] [text/richtext] ["text/plain; charset=iso-8859-1"] [application/postscript] [application/octet-stream] [audio/basic] ["http://abc.com/display-types/funky.html"] Unless some other value is specified for the application of use, an octet-string that has no display-hint may be considered to have a pre-specified "default" MIME type as follows: "application/octet-stream" When an S-Expression is being encoded in one of the representations described in , any display-hint present is included. If a display-hint is the default, it is not surpressed nor is the default display-hint included in the representation for an octet-string without a display-hint.

Equality of octet-strings It is RECOMMENDED that two octet-strings be considered "equal" for most computational and algorithmic purposed if and only if they have the same display-hint and the same data octet-strings. Note that octet-strings are "case-sensitive"; the octet-string "abc" is not equal to the octet-string "ABC". An untyped octet-string can be compared to another octet-string (typed or not) by considering it as a typed octet-string with the default type specified for the applications or, in the absence of such specificaion, the default type specified in .

Lists Just as with octet-strings, there are variations in representing a list. Whitespace may be used to separate list elements, but they are only required to separate two octet-strings when otherwise the two octet-strings might be interpreted as one, as when one token follows another. Also, whitespace may follow the initial left parenthesis, or precede the final right parenthesis. Here are some examples of encodings of lists: (a bob c) ( a ( bob c ) ( ( d e ) ( e f ) ) ) (11:certificate(6:issuer3:bob)(7:subject5:alice)) ({ODpFeGFtcGxlIQ==} "1997" murphy 3:XC+) ()

S-expression representation types There are three "types" of representations:

• canonical
• basic transport
• advanced transport

The first two MUST be supported by any implementation; the last is OPTIONAL. As part of both basic and advanced transport representations, the base-64 representation of an S-expression may be used as described in .

Base-64 representation of S-expressions An S-expression may be represented in a base-64 encoding consisting of:

• an opening curly brace "{"
• the base-64 encoding of the S-expression.
• a final closing curly brace "}"

Base-64 encoding produces four characters of output for each three octets of input. If the length of the input divided by three leaves a remainder of one or two, it produces an output block of length four ending in two or one equals signs, respectively. These equals signs MUST be included on output but input routines MAY accept inputs where one or two equals signs are dropped. Whitespace inserted in the midst of the base-64 encoding is ignored. It is an error to have characters other than whitespace and base-64 characters. Note the difference between this base-64 encoding of an S-expression using curly braces ("{ }") and the base-64 encoding of an octet-string using verticle bars ("| |") in .

Canonical representation This canonical representation is used for digital signature purposes and transport over channels not sensitive to specific octet values. It is uniquely defined for each S-expression. It is not particularly readable, but that is not the point. It is intended to be very easy to parse, to be reasonably economical, and to be unique for any S-expression. (See .) The "canonical" form of an S-expression represents each octet-string in verbatim mode, and represents each list with no blanks separating elements from each other or from the surrounding parentheses (see also ). Here are some examples of canonical representations of S-expressions: (6:issuer3:bob) (4:icon[12:image/bitmap]9:xxxxxxxxx) (7:subject(3:ref5:alice6:mother)) 10:foo)]}>bar 0:

Basic transport representation There are two forms of the "basic transport" representation:

• the canonical representation
• an base-64 representation of the canonical representation, surrounded by braces (see ).

The basic transport representations (see ) are intended to provide a universal means of representing S-expressions for transport from one machine to another. The base-64 encoding would be appropriate if the channel over which the S-expression is being sent might be sensitive to octets of some special values, such as an octet of all zero bits (NULL) or an octet of all one bits (DEL), or the channel is sensitive to "line length" such that occasional line terminating white space is needed. Here are two examples of an S-expression represented in basic transport mode: (1:a1:b1:c) {KDE6YTE6YjE 6YykK} The second example above is the same S-expression as the first encoded in base-64.

Advanced transport representation The "advanced transport" representation is intended to provide more flexible and readable notations for documentation, design, debugging, and (in some cases) user interface. The advanced transport representation allows all of the octet-string representation forms described above in Section 4: quoted strings, base-64, hexadecimal, tokens, representations of strings with omitted lengths, and so on. It also allow use of the base-64 representation of S-expressions. (See ).

ABNF for syntax ABNF is the Augmented Backus-Naur Form for syntax specifications as defined in . The ABNF for advanced representataion of S-expressions is given first and the basic and canonical forms derived therefrom. The rule names below in all caps are defined in Appendix B.1 of .

ABNF for advanced transport

ABNF for canonical

ABNF for basic transport

Restricted S-expressions This document has described S-expressions in general form. Application writers may wish to restrict their use of S-expressions in various ways as well as to specify a different default display-hint. Here are some possible restrictions that might be considered:

• no advanced representations (only canonical and basic)
• no display-hints
• no lengths on hexadecimal, quoted-strings, or base-64 encodings
• no empty lists
• no empty octet-strings
• no lists having another list as its first element
• no base-64 or hexadecimal encodings
• fixed limits on the size of octet-strings

As provided in , conformant implementations will support canonical and basic representation but support for advanced representation is not generally required. Thus advanced representation can only be used in applications which mandate its support or where a capability discovery mechanism indicates support.

In-memory representations For processing, the S-expression would typically be parsed and represented in memory in a way that is more amenable to efficient processing. This document suggests two alternatives:

• "list-structure"
• "array-layout"

These are only sketched here, as they are only suggestive. The code illustrates these styles in more detail.

List-structure memory representation Here there are separate records for simple-strings, strings, and lists or list nodes. An S-expression of the form ("abc" "de") could be encoded as two records for the simple-strings, two for the strings, and two for the list elements, where a record is a relatively small block of memory and, except for simple-string, might have pointers in it to other records. This is a fairly conventional representation as discussed in Section 4 of .

Array-layout memory representation Here each S-expression is represented as a contiguous array of octets. The first octet codes the "type" of the S-expression: 01 octet-string 02 octet-string with display-hint 03 beginning of list (and 00 is used for "end of list") Each of the three types is immediately followed by a k-octet integer indicating the size (in octets) of the following representation. Here k is an integer that depends on the implementation, it might be anywhere from 2 to 8, but would be fixed for a given implementation; it determines the size of the objects that can be handled. The transport and canonical representations are independent of the choice of k made by the implementation. Although the lengths of lists are not given in the usual S-expression notations, it is easy to fill them in when parsing; when you reach a right-parenthesis you know how long the list representation was, and where to go back to fill in the missing length.

Octet-string This is represented as follows: ]]> For example (here k = 2) 01 0003 a b c

Octet-string with display-hint This is represented as follows: 01 /* for display-type */ 01 /* for octet-string */ ]]> For example, the S-expression [gif] #61626364# would be represented as (with k = 2) 02 000d 01 0003 g i f 01 0004 61 62 63 64

List This is represented as ... 00 ]]> For example, the list (abc [d]ef (g)) is represented in memory as (with k = 2) 03 001b 01 0003 a b c 02 0009 01 0001 d 01 0002 e f 03 0005 01 0001 g 00 00

Security Considerations As a pure data representation format, there are few security considerations to S-expressions. A canonical form is required for the reliable verification of digital signatures. This is provided in . For convenience, the default display-hint (see ) can be specified for an application. Note that if S-expressions containing uptyped octet-strings represented for that application are processed by a different application, those untyped octet-string may be treated as if they had a different display-hint.

IANA Considerations This document requires no IANA actions.

Normative References Informative References Universität Bremen TZI Free Software Foundation, Inc. Random Contrarian Insurgent Organization GnuPG The Computer Center and Research Laboratory of Electronics, Massachusetts Institute of Technology Massachusetts Institute of Technology Ribose Group Inc. MIT Lab for Computer Science Textuality and Netscape Microsoft W3C Sun Microsystems

Implementations At this time there multiple implementations, many open source, available that are intended to read and parse some or all of the various S-expression formats specified here. In particular, see the following likely incomplete list:

• Project GNU's .
• Ribose's RNP in C++.
• Github project of J. P. Malkiewicz in C.
• The Inferno implementation .
• Small Fast X-Expression Library .
• S-expression Processor in Ruby.
• Canonical S-expressions (OCAML).

Change History RFC Editor Note: Please delete this section before publication.

-00 Changes This sub-section summarizes significant changes between the original 1997 -00 version of this document and the 2023 -00 version submitted to the IETF.

1. Convert to XML v3.
2. Update Ron Rivest author information and, with his permission, add Donald Eastlake as an author.
3. Add minimal "IANA Considerations" and "Security Considerations" sections.
4. Since implementation requirements terminology is used, add the usual paragraph about it as a sub-section of Section 1 and add references to and .
5. Divide references into Normative and Informational and update base-64 reference to be to .
6. Add a couple of sentences to the "Historical note" section about the history of -00 versions of the draft.

Changes from -00 to -01

1. Fix glitches and errors in the BNF.
2. Add Acknowledgements section to list Marc Petit-Huguenin (who provided BNF improvements) and John Klensin.
3. Update code references in and add to Informative References section. Note: The code in the Malkiewicz github repository may be the code that was originally at http://theory.lcs.mit.edu/~rivest/sexp.html
4. Add this Change History Appendix.
5. Move "Historical Notes" which were formerly a separate section at the end of the document up to be a sub-section of Section 1.
6. Add references to , , and .
7. Add simple security considerations.
8. Minor editorial fixes/improvements.

Changes from -01 to -02

1. Change default MIME Type in to have charset=utf-8 .
2. Change BNF to ABNF and add reference to .
3. Move Marc Petit-Huguenin to a Contributors section for his work on the ABNF.

Changes from -02 to -03

1. Add current S-expression usage Section 1.2.
2. Add the white book as a reference.
3. Add reference to the Ribose RNP code .
4. Minor editorial improvements.

Changes from -03 to -04 Trivial keep-alive update.

Changes from -04 to -05

1. Add reference to implementation.
2. Eliminate remaining references to being a "proposal".
3. Emphasize that a particular application can specify a different default display-hint.
4. Add reference to for ASCII.
5. Minor editorial improvements.

Changes from -05 to -06

1. Move implementations list to Appendix A. Add numerous implementations.
2. Change default display-hint to "application/octet-stream".
3. Expand Abstract and include most of Abstract in the Introduction.
4. Use different tokens for the top level rule in the three ABNF encodings so that the rules would not collide if all were used. Fix ABNF for "printable".
5. Add an illustration of list-structure memory representation.
6. Editorial improvements.

Changes from -06 to -07

1. Re-order some top level sections.
2. Replace "list-structure" memory figure with explanation and reference.
3. Re-organize ABNF to give full ABNF for advanced transport first and then mostly derive canonical and basic from advanced.
4. Correct reference to to be to Appendix B.1, not Appendix A.
5. Attempt to clarify the difference between canonicalization and equality.
6. Add the explicit on base-64 representation of S-expressions.
7. Globally hyphenate "octet-string" and "display-hing", generally replace "byte" with "octet".
8. Add some more examples here and there.
9. Fix typos. Other editorial improvements.

Acknowledgements Special thanks to Daniel K. Gillmore for his extensive comments. The comments and suggestions of the following are gratefully acknowledged: John Klensin and Caleb Malchik.

Contributors Special thanks to the following contributor, particularly for their extensive work and advice on the ABNF: Impedance Mismatch LLC

marc@petit-huguenin.org