RFC 9277




Internet Engineering Task Force (IETF)                     M. Richardson
Request for Comments: 9277                      Sandelman Software Works
Category: Standards Track                                     C. Bormann
ISSN: 2070-1721                                   Universität Bremen TZI
                                                             August 2022


  On Stable Storage for Items in Concise Binary Object Representation
                                 (CBOR)

Abstract



   This document defines a stored ("file") format for Concise Binary
   Object Representation (CBOR) data items that is friendly to common
   systems that recognize file types, such as the Unix file(1) command.

Status of This Memo



   This is an Internet Standards Track document.

   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).  Further information on
   Internet Standards is available in 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/rfc9277.

Copyright Notice



   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.

Table of Contents



   1.  Introduction
     1.1.  Terminology
     1.2.  Requirements for a Magic Number
   2.  Protocol
     2.1.  The CBOR-Protocol-Specific Tag
     2.2.  Enveloping Method: CBOR Tag Wrapped
       2.2.1.  Example
     2.3.  Enveloping Method: Labeled CBOR Sequence
       2.3.1.  Example
   3.  Security Considerations
   4.  IANA Considerations
     4.1.  Labeled CBOR Sequence Tag
     4.2.  CBOR-Labeled Non-CBOR Data Tag
     4.3.  CBOR Tags for CoAP Content-Format Numbers
   5.  References
     5.1.  Normative References
     5.2.  Informative References
   Appendix A.  Advice to Protocol Designer
     A.1.  Is the on-wire format new?
     A.2.  Can many items be trivially concatenated?
     A.3.  Are there tags at the start?
   Appendix B.  CBOR Tags for CoAP Content Formats
     B.1.  Content-Format Tag Examples
   Appendix C.  Example from Openswan
   Appendix D.  Using CBOR Labels for Non-CBOR Data
     D.1.  Content-Format Tag Examples
   Acknowledgements
   Contributors

   Authors' Addresses



1.  Introduction



   Since very early in computing, operating systems have sought ways to
   mark which files could be processed by which programs.  In Unix,
   everything is a stream of bytes; identifying the contents of a stream
   of bytes became a heuristic activity.

   For instance, the Unix file(1) command, which has existed since 1973
   [FILE], has been able to identify many file formats based upon the
   contents of the file for decades.

   Many systems (Linux, macOS, Windows) will select the correct
   application based upon the file contents if the system cannot
   determine it by other means.  For instance, in classical Mac OS, a
   resource fork was maintained separately from the file data that
   included file type information; this way, the OS ideally never needed
   to know anything about the file data contents to determine the media
   type.

   Many other systems do this by using file extensions.  Many common Web
   servers derive the media-type information from file extensions.

   Having a media type associated with the file contents can avoid some
   of the brittleness of this approach.  When files become disconnected
   from their type information, such as when attempting to do forensics
   on a damaged system, being able to identify the type of information
   stored in a file can become very important.

   A common way to identify the type of a file from its contents is to
   place a "magic number" at the start of the file contents [MAGIC].  In
   the media type registration template [RFC6838], a magic number is
   asked for, if available, as is a file extension.

   A challenge for the file(1) command is often that it can be confused
   by recognizing the overall encoding but not the content being
   encoded.  For instance, an Android Package Kit APK (as used to
   transfer and store an application) may be identified as a ZIP file.
   Additionally, both OpenOffice and MSOffice files are ZIP files of XML
   files; the identification may stop at identifying them as ZIP files.

   As CBOR becomes a more and more common encoding for a wide variety of
   artifacts, identifying them as just "CBOR" is probably not
   sufficient.  This document provides a way to encode a magic number
   into the beginning of a CBOR format file.  As a CBOR format may use a
   single CBOR data item or a CBOR sequence of data items [RFC8742], two
   possible methods of enveloping data are presented; a CBOR Protocol
   designer will specify one.  (A CBOR Protocol is a specification that
   uses CBOR as its encoding.)

   This document also gives advice to designers of CBOR Protocols on
   choosing one of these mechanisms for identifying their contents.
   This advice is informative.

   A third method is also proposed by which a CBOR format tag is
   prepended to identify non-CBOR files.  Further information on this
   method appears in Appendix D because it is not about identifying
   media types containing CBOR-encoded data items.  This includes a
   simple way to derive a magic number for content-formats as defined in
   [RFC7252], even if the file is not in CBOR form.

   Examples of CBOR Protocols currently under development include
   Concise Software Identification Tags [CoSWID] and Entity Attestation
   Tokens [EAT].  CBOR Object Signing and Encryption (COSE) itself
   [RFC8152] is considered infrastructure.  The encoding of public keys
   in CBOR as _C509_ as described in [C509-CERT] would benefit from
   being an identified CBOR Protocol.

   A major inspiration for this document is observing the disarray in
   certain ASN.1-based systems where most files are Privacy-Enhanced
   Mail (PEM) encoded; these files are all identified by the extension
   "pem", which confounds public keys, private keys, certificate
   requests, and S/MIME content.

   While the envelopes defined in this specification add information to
   how data conforming to CBOR Protocols are stored in files, there is
   no requirement that either type of envelope be transferred on the
   wire.  However, there are some protocols that may benefit from having
   such a magic number on the wire if they are presently using a
   different (legacy) encoding scheme.  The presence of the identifiable
   magic sequence can be used to signal that a CBOR Protocol is being
   used as opposed to a legacy scheme.

1.1.  Terminology



   Byte is a synonym for octet.  The term "byte string" refers to the
   data item defined in [STD94].

   The term "file" is understood to stand in a general way for a stored
   representation that is somewhat detached from the original context of
   usage of that representation; its usage in this document encompasses
   similar units of storage that may have different identification
   schemes such as partitions or media blocks.

   The term "diagnostic notation" refers to the human-readable notation
   for CBOR data items defined in Section 8 of [STD94] and Appendix G of
   [RFC8610].

   The term "CDDL" (Concise Data Definition Language) refers to the
   language defined in [RFC8610].

   The function TN(ct) is defined in Appendix B.

1.2.  Requirements for a Magic Number



   Ideally, a magic number is a fingerprint that is unique to a specific
   CBOR Protocol, is present in the first few (small multiple of 4)
   bytes of the file and does not change when the contents change, and
   does not depend upon the length of the file.

   Less ideal solutions have a pattern that needs to be matched, but in
   which some bytes need to be ignored.  While the Unix file(1) command
   can be told to ignore certain bytes, this can lead to ambiguities.

2.  Protocol



   This section presents two enveloping methods.  Both use CBOR tags in
   a way that results in a deterministic first 8 to 12 bytes.  The
   Protocol designer determines which one to use; see Appendix A for
   some guidance.

2.1.  The CBOR-Protocol-Specific Tag



   In both enveloping methods, CBOR Protocol designers need to obtain a
   CBOR tag for each kind of object that they might store in files.  As
   there are more than 4 billion available 4-byte tags, there should be
   little issue in allocating a few to each available CBOR Protocol.

   The IANA policy for 4-byte CBOR tags is First Come First Served
   [RFC8126] so only a simple interaction (e.g., via Web or email) with
   IANA is required.  The interaction includes filling in the small
   template provided in Section 9.2 of [STD94].  In the template, a
   reference to this specification (RFC 9277) alongside the Description
   of semantics is suggested.

   Allocation of the CBOR tag needs to be initiated by the designer of
   the CBOR Protocol, who can provide a proposed tag number.  In order
   to be in the 4-byte range, and so that there are no leading zero
   bytes in the 4-byte encoding of the tag number, the value needs to be
   in the range 0x01000000 (decimal 16777216) to 0xFFFFFFFF (decimal
   4294967295) inclusive.  It is further suggested to avoid values that
   have an embedded zero byte in the 4 bytes of their binary
   representation (such as 0x12003456), as these may confuse
   implementations that treat the magic number as a C string.

   The use of a sequence of four ASCII [RFC20] codes which are mnemonic
   to the protocol is encouraged, but not required (there may be reasons
   to encode other information into the tag; see Appendix B for an
   example).  For instance, Appendix C uses "OPSN", which translates to
   the tag number 1330664270 registered for it.

   In [IANA.CORE-PARAMETERS], the Constrained Application Protocol
   (CoAP) defines the "CoAP Content-Formats" registry to assign Content-
   Format Numbers (Section 12.3 of [RFC7252]) to Content Types in a
   specific Content Coding.  For CBOR data items that form a
   representation that is already described by such a Content-Format
   Number, a tag number has proactively been allocated in Section 4.3
   (see Appendix B for details and examples).

2.2.  Enveloping Method: CBOR Tag Wrapped



   The CBOR Tag Wrapped method is appropriate for use with CBOR
   Protocols that encode a single CBOR data item.  This data item is
   enveloped into two nested tags:

   The outer tag is a self-described CBOR tag, 55799, as described in
   Section 3.4.6 of [STD94].

   The tag content of the outer tag is a second CBOR tag whose tag
   number has been allocated to describe the specific Protocol involved,
   as discussed in Section 2.1.  The tag content of this inner tag is
   the single CBOR data item.

   This method wraps the CBOR data item as CBOR tags usually do.
   Applications that need to send the stored CBOR data item across a
   constrained network may wish to remove the two tags if the type is
   understood from the protocol context, e.g., from a CoAP Content-
   Format Option (Section 5.10.3 of [RFC7252]).  Therefore, a CBOR
   Protocol specification may pick the specific cases where the CBOR Tag
   Wrapped enveloping method is to be used.  For instance, it might
   specify its use for storing the representation in a local file or for
   Web access, but not within protocol messages that already provide the
   necessary context.

2.2.1.  Example



   To construct an example without registering a new tag, we use the
   Content-Format ID assigned for application/senml+cbor (112) [RFC8428]
   of the "CoAP Content-Formats" registry [IANA.CORE-PARAMETERS]).

   Using the technique described in Appendix B, this translates into the
   tag TN(112) = 1668546929.

   With this tag, the SenML-CBOR pack [{0: "current", 6: 3, 2: 1.5}]
   would be enveloped as follows (in diagnostic notation):

   55799(1668546929([{0: "current", 6: 3, 2: 1.5}]))

   Or in hex:

   d9 d9f7                       # tag(55799)
      da 63740171                # tag(1668546929)
         81                      # array(1)
            a3                   # map(3)
               00                # unsigned(0)
               67                # text(7)
                  63757272656e74 # "current"
               06                # unsigned(6)
               03                # unsigned(3)
               02                # unsigned(2)
               f9 3e00           # primitive(15872)

   At the representation level, the unique fingerprint for application/
   senml+cbor is composed of the 8 bytes d9d9f7da63740171 hex, after
   which the unadorned CBOR data (81... for the SenML data) is appended.

2.3.  Enveloping Method: Labeled CBOR Sequence



   The Labeled CBOR Sequence method is appropriate for use with CBOR
   Sequences as described in [RFC8742].

   This method prepends a newly constructed, separate data item to the
   CBOR Sequence, the _label_.

   The label is a nesting of two tags, similar to but distinct from the
   CBOR Tag Wrapped methods, with an inner tag content of a constant
   byte string.  The total length of the label is 12 bytes.

   1.  The outer tag is the self-described CBOR Sequence tag, 55800.

   2.  The inner tag is a CBOR tag from the First Come First Served
       space that uniquely identifies the CBOR Protocol.  As with the
       CBOR Tag Wrapped method, the use of a 4-byte tag that encodes
       without zero bytes is encouraged.

   3.  The tag content is a 3-byte CBOR byte string containing
       0x42_4f_52 ('BOR' in diagnostic notation).



   The outer tag in the label identifies the file as being a CBOR
   Sequence and does so with all the desirable properties explained in
   Section 3.4.6 of [STD94].  Specifically, it does not appear to
   conflict with any known file types, and it is not valid Unicode in
   any Unicode encoding.

   The inner tag in the label identifies which CBOR Protocol is used, as
   described above.

   The inner tag content is a constant byte string that is represented
   as 0x43_42_4f_52, the ASCII characters "CBOR", which is the CBOR-
   encoded data item for the 3-byte string 0x42_4f_52 ('BOR' in
   diagnostic notation).

   The actual CBOR Protocol data then follows as the next data item(s)
   in the CBOR Sequence, without a need for any further specific tag.
   The use of a CBOR Sequence allows the application to trivially remove
   the first item with the two tags.

   Should this file be reviewed by a human (directly in an editor or in
   a hexdump display), it will include the ASCII characters "CBOR"
   prominently.  This value is also included simply because the inner
   nested tag needs to tag something.

2.3.1.  Example



   To construct an example without registering a new tag, we use ID 272
   as assigned for application/missing-blocks+cbor-seq of the "CoAP
   Content-Formats" registry [RFC9177].

   Using the technique described in Appendix B, this translates into the
   tag TN(272) = 1668547090.

   This is a somewhat contrived example, as this is not a media type
   that is likely to be committed to storage.  Nonetheless, with this
   tag, missing blocks list 0, 8, 15 would be enveloped as (in
   diagnostic notation):

   55800(1668547090('BOR')),
   0,
   8,
   15

   Or in hex:

   # CBOR sequence with 4 elements
   d9 d9f8         # tag(55800)
      da 63740212  # tag(1668547090)
         43        # bytes(3)
            424f52 # "BOR"
   00 # unsigned(0)
   08 # unsigned(8)
   0f # unsigned(15)

   At the representation level, the unique fingerprint for application/
   missing-blocks+cbor-seq is composed of the 8 bytes d9d9f8da63740212
   hex, after which the unadorned CBOR sequence (00... for the missing
   block list given) is appended.

3.  Security Considerations

   This document provides a way to identify CBOR Protocol objects.
   Clearly identifying CBOR contents in files may have a variety of
   impacts.

   The most obvious is that it may allow malware to identify interesting
   stored objects, and then exfiltrate or corrupt them.

   Protective applications (that check data) cannot rely on the
   applications they try to protect (that use the data) to make exactly
   the same decisions in recognizing file formats.  (This is an instance
   of a check versus use issue.)  For example, end-point assessment
   technologies should not solely rely on the labeling approaches
   described in this document to decide whether to inspect a given file.
   Similarly, depending on operating system configurations and related
   properties of the execution environment, the labeling might influence
   the default application used to process a file in a way that may not
   be predicted by a protective application.

4.  IANA Considerations



   These IANA considerations are entirely about CBOR tags in the
   "Concise Binary Object Representation (CBOR) Tags" registry
   [IANA.CBOR-TAGS].

   Section 4.1 documents the allocation for a CBOR tag to be used in a
   CBOR sequence to identify the sequence (an example for using this tag
   is found in Appendix C).  Section 4.2 documents the allocation for a
   CBOR tag to be used in the CBOR-Labeled Non-CBOR Data Enveloping
   Method (Appendix D, which also shows examples).  Section 4.3
   allocates a CBOR tag for each actual or potential CoAP Content-Format
   number (examples are in Appendix B).

4.1.  Labeled CBOR Sequence Tag



   IANA has allocated tag 55800 for the Labeled CBOR Sequence Enveloping
   Method from the "CBOR Tags" registry.  IANA has updated this tag
   registration to point to this document.

   This tag is from the First Come First Served area.

   The value has been picked to have properties similar to the 55799 tag
   (Section 3.4.6 of [STD94]).

   The hexadecimal representation of the encoded tag head is 0xd9_d9_f8.

   This is not valid UTF-8: the first 0xd9 introduces a 3-byte sequence
   in UTF-8, but the 0xd9 as the second value is not a valid second byte
   for UTF-8.

   This is not valid UTF-16: the byte sequence 0xd9d9 (in either endian
   order) puts this value into the UTF-16 high-half zone, which would
   signal that this is a 32-bit Unicode value.  However, the following
   16-bit big-endian value 0xf8_xx is not a valid second sequence
   according to [RFC2781].  On a little-endian system, it would be
   necessary to examine the fourth byte to determine if it is valid.
   That next byte is determined by the subsequent encoding, and
   Section 3.4.6 of [STD94] has already determined that no valid CBOR
   encodings result in valid UTF-16.

   Data Item:
      tagged byte string

   Semantics:
      indicates that the file contains CBOR Sequences

4.2.  CBOR-Labeled Non-CBOR Data Tag



   IANA has allocated tag 55801 for the CBOR-Labeled Non-CBOR Data
   Enveloping Method (Appendix D) from the "CBOR Tags" registry.  IANA
   updated this tag registration to point to this document.

   This tag is from the First Come First Served area.

   The value has been picked to have properties similar to the 55799 tag
   (Section 3.4.6 of [STD94]).

   The hexadecimal representation of the encoded tag head is 0xd9_d9_f9.

   This is not valid UTF-8: the first 0xd9 introduces a 3-byte sequence
   in UTF-8, but the 0xd9 as the second value is not a valid second byte
   for UTF-8.

   This is not valid UTF-16: the byte sequence 0xd9d9 (in either endian
   order) puts this value into the UTF-16 high-half zone, which would
   signal that this is a 32-bit Unicode value.  However, the following
   16-bit big-endian value 0xf9_xx is not a valid second sequence
   according to [RFC2781].  On a little-endian system, it would be
   necessary to examine the fourth byte to determine if it is valid.
   That next byte is determined by the subsequent encoding, and
   Section 3.4.6 of [STD94] has already determined that no valid CBOR
   encodings result in valid UTF-16.

   Data Item:
      tagged byte string

   Semantics:
      indicates that the file starts with a CBOR-Labeled Non-CBOR Data
      label.

4.3.  CBOR Tags for CoAP Content-Format Numbers



   IANA allocated the tag numbers 1668546817 (0x63740101) to 1668612095
   (0x6374ffff) as follows:

   Data Item:
      byte string or any CBOR data item (see Appendix B)

   Semantics:
      the representation of content-format ct < 65025 is indicated by
      tag number
      TN(ct) = 0x63740101 + (ct / 255) * 256 + ct % 255

   Reference:
      RFC 9277

   The "CoAP Content-Formats" registry [IANA.CORE-PARAMETERS] is defined
   in Section 12.3 of [RFC7252].

5.  References



5.1.  Normative References



   [C]        International Organization for Standardization,
              "Information technology -- Programming languages -- C",
              ISO/IEC 9899:2018, Fourth Edition, June 2018,
              <https://www.iso.org/standard/74528.html>.

   [RFC8742]  Bormann, C., "Concise Binary Object Representation (CBOR)
              Sequences", RFC 8742, DOI 10.17487/RFC8742, February 2020,
              <https://www.rfc-editor.org/info/rfc8742>.

   [STD94]    Bormann, C. and P. Hoffman, "Concise Binary Object
              Representation (CBOR)", STD 94, RFC 8949,
              DOI 10.17487/RFC8949, December 2020,
              <https://www.rfc-editor.org/info/std94>.

5.2.  Informative References



   [C509-CERT]
              Mattsson, J. P., Selander, G., Raza, S., Höglund, J., and
              M. Furuhed, "CBOR Encoded X.509 Certificates (C509
              Certificates)", Work in Progress, Internet-Draft, draft-
              ietf-cose-cbor-encoded-cert-04, 10 July 2022,
              <https://datatracker.ietf.org/doc/html/draft-ietf-cose-
              cbor-encoded-cert-04>.

   [CoSWID]   Birkholz, H., Fitzgerald-McKay, J., Schmidt, C., and D.
              Waltermire, "Concise Software Identification Tags", Work
              in Progress, Internet-Draft, draft-ietf-sacm-coswid-22, 20
              July 2022, <https://datatracker.ietf.org/doc/html/draft-
              ietf-sacm-coswid-22>.

   [EAT]      Lundblade, L., Mandyam, G., and J. O'Donoghue, "The Entity
              Attestation Token (EAT)", Work in Progress, Internet-
              Draft, draft-ietf-rats-eat-14, 10 July 2022,
              <https://datatracker.ietf.org/doc/html/draft-ietf-rats-
              eat-14>.

   [FILE]     Wikipedia, "file (command)", 2 July 2022,
              <https://en.wikipedia.org/w/
              index.php?title=File_(command)&oldid=1096086462>.

   [IANA.CBOR-TAGS]
              IANA, "Concise Binary Object Representation (CBOR) Tags",
              <https://www.iana.org/assignments/cbor-tags>.

   [IANA.CORE-PARAMETERS]
              IANA, "Constrained RESTful Environments (CoRE)
              Parameters",
              <https://www.iana.org/assignments/core-parameters>.

   [MAGIC]    Bell Labs, "archive (library) file format", Unix
              Programmer's Manual, First Edition: File Formats, 3
              November 1971,
              <https://www.bell-labs.com/usr/dmr/www/man51.pdf#page=4>.

   [RFC20]    Cerf, V., "ASCII format for network interchange", STD 80,
              RFC 20, DOI 10.17487/RFC0020, October 1969,
              <https://www.rfc-editor.org/info/rfc20>.

   [RFC2781]  Hoffman, P. and F. Yergeau, "UTF-16, an encoding of ISO
              10646", RFC 2781, DOI 10.17487/RFC2781, February 2000,
              <https://www.rfc-editor.org/info/rfc2781>.

   [RFC6838]  Freed, N., Klensin, J., and T. Hansen, "Media Type
              Specifications and Registration Procedures", BCP 13,
              RFC 6838, DOI 10.17487/RFC6838, January 2013,
              <https://www.rfc-editor.org/info/rfc6838>.

   [RFC7252]  Shelby, Z., Hartke, K., and C. Bormann, "The Constrained
              Application Protocol (CoAP)", RFC 7252,
              DOI 10.17487/RFC7252, June 2014,
              <https://www.rfc-editor.org/info/rfc7252>.

   [RFC8017]  Moriarty, K., Ed., Kaliski, B., Jonsson, J., and A. Rusch,
              "PKCS #1: RSA Cryptography Specifications Version 2.2",
              RFC 8017, DOI 10.17487/RFC8017, November 2016,
              <https://www.rfc-editor.org/info/rfc8017>.

   [RFC8126]  Cotton, M., Leiba, B., and T. Narten, "Guidelines for
              Writing an IANA Considerations Section in RFCs", BCP 26,
              RFC 8126, DOI 10.17487/RFC8126, June 2017,
              <https://www.rfc-editor.org/info/rfc8126>.

   [RFC8152]  Schaad, J., "CBOR Object Signing and Encryption (COSE)",
              RFC 8152, DOI 10.17487/RFC8152, July 2017,
              <https://www.rfc-editor.org/info/rfc8152>.

   [RFC8428]  Jennings, C., Shelby, Z., Arkko, J., Keranen, A., and C.
              Bormann, "Sensor Measurement Lists (SenML)", RFC 8428,
              DOI 10.17487/RFC8428, August 2018,
              <https://www.rfc-editor.org/info/rfc8428>.

   [RFC8610]  Birkholz, H., Vigano, C., and C. Bormann, "Concise Data
              Definition Language (CDDL): A Notational Convention to
              Express Concise Binary Object Representation (CBOR) and
              JSON Data Structures", RFC 8610, DOI 10.17487/RFC8610,
              June 2019, <https://www.rfc-editor.org/info/rfc8610>.

   [RFC9110]  Fielding, R., Ed., Nottingham, M., Ed., and J. Reschke,
              Ed., "HTTP Semantics", STD 97, RFC 9110,
              DOI 10.17487/RFC9110, June 2022,
              <https://www.rfc-editor.org/info/rfc9110>.

   [RFC9177]  Boucadair, M. and J. Shallow, "Constrained Application
              Protocol (CoAP) Block-Wise Transfer Options Supporting
              Robust Transmission", RFC 9177, DOI 10.17487/RFC9177,
              March 2022, <https://www.rfc-editor.org/info/rfc9177>.

   [X.690]    ITU-T, "Information technology - ASN.1 encoding rules:
              Specification of Basic Encoding Rules (BER), Canonical
              Encoding Rules (CER) and Distinguished Encoding Rules
              (DER)", ITU-T Recommendation X.690, ISO/IEC 8825-1,
              February 2021, <https://www.itu.int/rec/T-REC-X.690>.

Appendix A.  Advice to Protocol Designer



   This document introduces a choice between wrapping a single CBOR data
   item into a pair of identifying CBOR tags or prepending an
   identifying encoded CBOR data item (which, in turn, contains a pair
   of identifying CBOR tags) to a CBOR Sequence (which might be a single
   data item).

   Which should a protocol designer use?

   In this discussion, one assumes that there is an object stored in a
   file, perhaps specified by a system operator in a configuration file.

   For example: a private key used in COSE operations, a public key/
   certificate in C509 [C509-CERT] or CBOR format, a recorded sensor
   reading stored for later transmission, or a COVID-19 vaccination
   certificate that needs to be displayed in QR code form.

   Both the Labeled CBOR Sequence and the wrapped tag can be trivially
   removed by an application before sending the CBOR content out on the
   wire.

   The Labeled CBOR Sequence can be slightly easier to remove as, in
   most cases, CBOR parsers will return it as a unit and then return the
   actual CBOR item, which could be anything at all and could include
   CBOR tags that _do_ need to be sent on the wire.

   On the other hand, having the Labeled CBOR Sequence in the file
   requires that all programs that expect to examine that file be able
   to skip what appears to be a CBOR item with two tags nested around a
   3-byte byte string.  The 3-byte entry is not of the format the
   program would normally have processed, so it may be a surprise.  On
   the other hand, CBOR parsers are generally tolerant of tags that
   appear: many of them will process extra tags, making unknown tags
   available as meta information.  A program that is not expecting those
   tags may just ignore them.

   As an example of where there was a problem with previous security
   systems, "PEM" format certificate files grew to be able to contain
   multiple certificates by simple concatenation.  The PKCS1 format
   [RFC8017] could also contain a private key object followed by one or
   more certificate objects, but only when in PEM format.  Annoyingly,
   when in binary DER format ([X.690], which like CBOR is self-
   delimiting), concatenation of certificates was not compatible with
   most programs as they did not expect to read more than one item in
   the file.

   The use of CBOR Tag Wrapped format is easier to retrofit to an
   existing format with existing and unchangeable stored format for a
   single CBOR data item.  This new sequence of tags is expected to be
   trivially ignored by many existing programs when reading CBOR from
   files or similar units of storage, even if the program only supports
   decoding a single data item (and not a CBOR sequence).  But, a naive
   program might also then transmit the additional tags across the
   network.  Removing the CBOR Tag Wrapped format requires knowledge of
   the two tags involved.  Other tags present might be needed.

   For a representation matching a specific media-type that is carried
   in a CBOR byte string, the byte string head will already have to be
   removed for use as such a representation, so it should be easy to
   remove the enclosing tag heads as well.  This is of particular
   interest with the predefined tags provided in Appendix B for media
   types with CoAP Content-Format numbers.

   Some considerations in the form of survey questions follow.

A.1.  Is the on-wire format new?



   If the on-wire format is new, then it could be specified with the
   CBOR Tag Wrapped format if the extra 8 bytes are not a problem.  The
   stored format is then identical to the on-wire format.

   If the 8 bytes are a problem on the wire (and they often are if CBOR
   is being considered), then the Labeled CBOR Sequence format should be
   adopted for the stored format.

A.2.  Can many items be trivially concatenated?



   If the programs that read the contents of the file already expect to
   process all of the CBOR data items in the file (not just the first),
   then the Labeled CBOR Sequence format may be easily retrofitted.

   The programs involved may throw errors or warnings on the Labeled
   CBOR Sequence if they have not yet been updated, but this may not be
   a problem.

   There are situations where multiple objects may be concatenated into
   a single file.  If each object is preceded by a Labeled CBOR Sequence
   label, then there may be multiple such labels in the file.

   A protocol based on CBOR Sequences may specify that Labeled CBOR
   Sequence labels can occur within a CBOR Sequence, possibly even to
   switch to data items following in the sequence that are of a
   different type.

   If the CBOR-Sequence-based protocol does not define the semantics for
   or at least tolerate embedded labels, care must be taken when
   concatenating Labeled CBOR Sequences to remove the label from all but
   the first part.

      |  As an example from legacy PEM-encoded PKIX certificates, many
      |  programs accept a series of PKIX certificates in a single file
      |  in order to set up a certificate chain.  The file would contain
      |  not just the End-Entity (EE) certificate, but also any
      |  subordinate certification authorities (CAs) needed to validate
      |  the EE.  This mechanism actually only works for PEM-encoded
      |  certificates, and not DER-encoded certificates.  One of the
      |  reasons for this specification is to make sure that CBOR-
      |  encoded certificates do not suffer from this problem.
      |
      |  As an example of mixing of types, some TLS server programs also
      |  can accept both their PEM-encoded private key and their PEM-
      |  encoded certificate in the same file.

   If only one item is ever expected in the file, the use of the Labeled
   CBOR Sequence may present an implementation hurdle to programs that
   previously just read a single data item and used it.

A.3.  Are there tags at the start?



   If the Protocol expects to use other tags at its top level, then the
   use of the CBOR Tag Wrapped format may be easy to explain at the same
   place in the protocol description.

Appendix B.  CBOR Tags for CoAP Content Formats



   Section 5.10.3 of [RFC7252] defines the concept of a Content-Format,
   which is a short, 16-bit unsigned integer that identifies a specific
   content type (media type plus (optionally) parameters), optionally
   together with a content coding (see Section 8.4.1 of [RFC9110]).

   Outside of a transfer protocol that indicates the Content-Format for
   a representation, it may be necessary to identify the Content-Format
   of the representation when it is stored in a file, in firmware, or
   when debugging.

   This specification allocates CBOR tag numbers 1668546817 (0x63740101)
   to 1668612095 (0x6374FFFF) for the tagging of representations of
   specific content formats.

   Using tags from this range, a byte string that is to be interpreted
   as a representation of Content-Format number ct, with ct < 65025
   (255*255), can be identified by enclosing it in a tag with tag number
   TN(ct) where:

      TN(ct) = 0x63740101 + (ct / 255) * 256 + ct % 255.

   (where +, *, / and % stand for integer addition, multiplication,
   division, and remainder as in the programming language C [C].)

      |  This formula avoids the use of zero bytes in the representation
      |  of the tag number.
      |  
      |  Note that no tag numbers are assigned for Content-Format
      |  numbers in the following range:
      |  
      |     65025 ≤ ct ≤ 65535
      |  
      |  (This range is in the range reserved for Experimental Use
      |  [RFC8126] by Section 12.3 of [RFC7252].  The overlap of 25 code
      |  points between this experimental range with the range this
      |  appendix defines tag numbers for can be used for experiments
      |  that want to employ a tag number.)

   Exceptionally, when used immediately as tag content of one of the
   tags 55799, 55800, or 55801, the tag content is as follows:

   Tag 55799 (Section 2.2):  One of:

      1.  The CBOR data item within the representation (without byte-
          string wrapping).  This only works for Content-Formats that
          are represented by a single CBOR data item in identity
          content-coding.

      2.  The data items in the CBOR sequence within the representation,
          without byte string wrapping, but wrapped in a CBOR array.
          This works for Content-Formats that are represented by a CBOR
          sequence in identity content-coding.

   Tags 55800 (Section 2.3) or 55801 (Appendix D):  the byte string
      'BOR', signifying that the representation of the given content-
      format follows in the file, in the way defined for these tags.

B.1.  Content-Format Tag Examples



   The "CoAP Content-Formats" registry [IANA.CORE-PARAMETERS] defines
   content formats that can be used as examples:

   *  As discussed in Section 2.2.1, Content-Format ID 112 represents
      the application/senml+cbor media type (no parameters).  The
      corresponding tag number is TN(112) = 1668546929.

      The following CDDL snippet can be used to identify application/
      senml+cbor representations:

      senml-cbor = #6.1668546929(bstr)

      Note that a byte string is used as the type of the tag content
      because a media type representation in general can be any byte
      string.

   *  Content-Format ID 272 represents the application/missing-
      blocks+cbor-seq media type, which is a CBOR sequence [RFC9177].

      The corresponding tag number is TN(272) = 1668547090.

      The following CDDL snippet can be used to identify application/
      missing-blocks+cbor-seq representations as embedded in a CBOR byte
      string:

      missing-blocks = #6.1668547090(bstr)

Appendix C.  Example from Openswan



   The Openswan IPsec project has a daemon ("pluto") and two control
   programs ("addconn" and "whack").  They communicate via a Unix-domain
   socket, over which a C-structure containing pointers to strings is
   serialized using a bespoke mechanism.  This is normally not a problem
   as the structure is compiled by the same compiler; but when there are
   upgrades, it is possible for the daemon and the control programs to
   get out of sync by the bespoke serialization.  As a result, there are
   extra compensations to deal with shutting the daemon down.  During
   testing, it is sometimes the case that upgrades are backed out.

   In addition, when doing unit testing, the easiest way to load policy
   is to use the normal policy-reading process, but that is not normally
   loaded in the daemon.  Instead, the IPC that is normally sent across
   the wire is compiled, serialized, and placed in a file.  The above
   magic number is included in the file and on the IPC in order to
   distinguish the "shutdown" command CBOR operation.

   In order to reduce the problems due to serialization, the
   serialization is being changed to CBOR.  Additionally, this change
   allows the IPC to be described by CDDL and any implementation
   language to be used that can encode CBOR.

   IANA has allocated the tag 1330664270 or 0x4f_50_53_4e for this
   purpose.  As a result, each file and each IPC is prefixed with a CBOR
   Sequence tag.

   In diagnostic notation:

   55800(1330664270(h'424F52'))

   Or in hex:

   d9 d9f8         # tag(55800)
      da 4f50534e  # tag(1330664270)
         43        # bytes(3)
            424f52 # "BOR"

Appendix D.  Using CBOR Labels for Non-CBOR Data



   The CBOR-Labeled Non-CBOR data method is appropriate for adding a
   magic number to a Non-CBOR data format, particularly one that can be
   described by a Content-Format tag (Appendix B).

   This method prepends a CBOR data item to the Non-CBOR data; this data
   item is called the "header" and, similar to the Labeled CBOR-Sequence
   label, consists of two nested tags around a constant byte string for
   a total of 12 bytes.

   1.  The outer tag is the CBOR-Labeled Non-CBOR Data tag, 55801.

   2.  The inner tag is a CBOR tag from the First Come First Served
       space that uniquely identifies the CBOR Protocol.  As with CBOR
       Tag Wrapped, the use of a 4-byte tag is encouraged that encodes
       without zero bytes.

   3.  The tag content is a 3-byte CBOR byte string containing
       0x42_4F_52 ('BOR' in diagnostic notation).

   The outer tag in the label identifies the file as being prefixed by a
   Non-CBOR data label and does so with all the desirable properties
   explained in Section 3.4.6 of [STD94].  Specifically, it does not
   appear to conflict with any known file types, and it is not valid
   Unicode in any Unicode encoding.

   The inner tag in the label identifies which Non-CBOR Protocol is
   used.

   The inner tag content is a constant byte string that is represented
   as 0x43_42_4f_52, the ASCII characters "CBOR", which is the CBOR-
   encoded data item for the 3-byte string 0x42_4f_52 ('BOR' in
   diagnostic notation).

   The actual Non-CBOR Protocol data then follow directly appended to
   the CBOR representation of the header.  This allows the application
   to trivially remove the header item with the two nested tags and the
   byte string.

   As with the Labeled CBOR Sequence {#sequences}, this choice of the
   tag content places the ASCII characters "CBOR" prominently into the
   header.

D.1.  Content-Format Tag Examples



   The "CoAP Content-Formats" registry [IANA.CORE-PARAMETERS] defines
   content formats that can be used as examples:

   *  Content-Format ID 432 represents the application/td+json media
      type (no parameters).

      The corresponding tag number is TN(432) = 1668547250.

      The following CDDL snippet can be used to identify a CBOR-Labeled
      Non-CBOR data for application/td+json representations:

      td-json-header = #6.55801(#6.1668547250('BOR'))

   *  Content-Format 11050 represents the application/json media type in
      deflate content-coding.

      The corresponding tag number is TN(11050) = 1668557910.

      The following CDDL snippet can be used to identify a CBOR-Labeled
      Non-CBOR data for application/json representations compressed in
      deflate content-coding:

      json-deflate-header = #6.55801(#6.1668557910('BOR'))

Acknowledgements



   The CBOR WG brainstormed this protocol on January 20, 2021 via a
   number of productive email exchanges on the mailing list.

Contributors

   Josef 'Jeff' Sipek
   Email: jeffpc@josefsipek.net


Authors' Addresses



   Michael Richardson
   Sandelman Software Works
   Email: mcr+ietf@sandelman.ca


   Carsten Bormann
   Universität Bremen TZI
   Postfach 330440
   D-28359 Bremen
   Germany
   Phone: +49-421-218-63921
   Email: cabo@tzi.org