Digest Fields

• New request and response header and trailer field definitions.

  • Section 2 (Content-Digest),
  • Section 3 (Repr-Digest), and
  • Section 4 (Want-Content-Digest and Want-Repr-Digest).

• Considerations specific to representation data integrity.

  • Section 3.1 (State-changing requests),
  • Section 3.2 (Content-Location),
  • Appendix A contains worked examples of representation data in message exchanges, and
  • Appendixes B and C contain worked examples of Repr-Digest and Want-Repr-Digest fields in message exchanges.
• Section 5 presents hash algorithm considerations and defines registration procedures for future entries.

The HTTP fields defined in this document can be used for HTTP integrity. Senders choose a hashing algorithm and calculate a digest from an input related to the HTTP message. The algorithm identifier and digest are transmitted in an HTTP field. Receivers can validate the digest for integrity purposes. Hashing algorithms are registered in the "Hash Algorithms for HTTP Digest Fields" registry (see Section 7.2).

Selecting the data on which digests are calculated depends on the use case of the HTTP messages. This document provides different fields for HTTP representation data and HTTP content.

There are use cases where a simple digest of the HTTP content bytes is required. The Content-Digest request and response header and trailer field is defined to support digests of content (Section 6.4 of [HTTP]); see Section 2.

For more advanced use cases, the Repr-Digest request and response header and trailer field (Section 3) is defined. It contains a digest value computed by applying a hashing algorithm to selected representation data (Section 8.1 of [HTTP]). Basing Repr-Digest on the selected representation makes it straightforward to apply it to use cases where the message content requires some sort of manipulation to be considered as representation of the resource or the content conveys a partial representation of a resource, such as range requests (see Section 14 of [HTTP]).

Content-Digest and Repr-Digest support hashing algorithm agility. The Want-Content-Digest and Want-Repr-Digest fields allow endpoints to express interest in Content-Digest and Repr-Digest, respectively, and to express algorithm preferences in either.

Content-Digest and Repr-Digest are collectively termed "Integrity fields". Want-Content-Digest and Want-Repr-Digest are collectively termed "Integrity preference fields".

Integrity fields are tied to the Content-Encoding and Content-Type header fields. Therefore, a given resource may have multiple different digest values when transferred with HTTP.

Integrity fields apply to HTTP message content or HTTP representations. They do not apply to HTTP messages or fields. However, they can be combined with other mechanisms that protect metadata, such as digital signatures, in order to protect the phases of an HTTP exchange in whole or in part. For example, HTTP Message Signatures [SIGNATURES] could be used to sign Integrity fields, thus providing coverage for HTTP content or representation data.

This specification does not define means for authentication, authorization, or privacy.

[RFC3230] defined the Digest and Want-Digest HTTP fields for HTTP integrity. It also coined the terms "instance" and "instance manipulation" in order to explain concepts, such as selected representation data (Section 8.1 of [HTTP]), that are now more universally defined and implemented as HTTP semantics.

Experience has shown that implementations of [RFC3230] have interpreted the meaning of "instance" inconsistently, leading to interoperability issues. The most common issue relates to the mistake of calculating the digest using (what we now call) message content, rather than using (what we now call) representation data as was originally intended. Interestingly, time has also shown that a digest of message content can be beneficial for some use cases, so it is difficult to detect if non-conformance to [RFC3230] is intentional or unintentional.

In order to address potential inconsistencies and ambiguity across implementations of Digest and Want-Digest, this document obsoletes [RFC3230]. The Integrity fields (Sections 2 and 3) and Integrity preference fields (Section 4) defined in this document are better aligned with current HTTP semantics and have names that more clearly articulate the intended usages.

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.

This document uses the Augmented BNF defined in [RFC5234] and updated by [RFC7405]. This includes the rules CR (carriage return), LF (line feed), and CRLF (CR LF).

This document uses the following terminology from Section 3 of [STRUCTURED-FIELDS] to specify syntax and parsing: Boolean, Byte Sequence, Dictionary, Integer, and List.

The definitions "representation", "selected representation", "representation data", "representation metadata", "user agent", and "content" in this document are to be interpreted as described in [HTTP].

This document uses the line folding strategies described in [FOLDING].

Hashing algorithm names respect the casing used in their definition document (e.g., SHA-1, CRC32c).

HTTP messages indicate hashing algorithms using an Algorithm Key (algorithms). Where the document refers to an Algorithm Key in prose, it is quoted (e.g., "sha", "crc32c").

The term "checksum" describes the output of applying an algorithm to a sequence of bytes, whereas "digest" is only used in relation to the value contained in the fields.

"Integrity fields" is the collective term for Content-Digest and Repr-Digest.

"Integrity preference fields" is the collective term for Want-Repr-Digest and Want-Content-Digest.

When the representation enclosed in a state-changing request does not describe the target resource, the representation digest MUST be computed on the representation data. This is the only possible choice because representation digest requires complete representation metadata (see Section 3).

In responses,

• if the representation describes the status of the request, Repr-Digest MUST be computed on the enclosed representation (see Appendix B.8);
• if there is a referenced resource, Repr-Digest MUST be computed on the selected representation of the referenced resource even if that is different from the target resource. This might or might not result in computing Repr-Digest on the enclosed representation.

The latter case is done according to the HTTP semantics of the given method, for example, using the Content-Location header field (see Section 8.7 of [HTTP]). In contrast, the Location header field does not affect Repr-Digest because it is not representation metadata.

For example, in PATCH requests, the representation digest will be computed on the patch document because the representation metadata refers to the patch document and not the target resource (see Section 2 of [PATCH]). In responses, instead, the representation digest will be computed on the selected representation of the patched resource.

When a state-changing method returns the Content-Location header field, the enclosed representation refers to the resource identified by its value and Repr-Digest is computed accordingly. An example is given in Appendix B.7.

This document specifies a data integrity mechanism that protects HTTP representation data or content, but not HTTP header and trailer fields, from certain kinds of corruption.

Integrity fields are not intended to be a general protection against malicious tampering with HTTP messages. In the absence of additional security mechanisms, an on-path malicious actor can either remove a digest value entirely or substitute it with a new digest value computed over manipulated representation data or content. This attack can be mitigated by combining mechanisms described in this document with other approaches such as Transport Layer Security (TLS) or digital signatures (for example, HTTP Message Signatures [SIGNATURES]).

Integrity fields can help detect representation data or content modification due to implementation errors, undesired "transforming proxies" (see Section 7.7 of [HTTP]), or other actions as the data passes across multiple hops or system boundaries. Even a simple mechanism for end-to-end representation data integrity is valuable because a user agent can validate that resource retrieval succeeded before handing off to an HTML parser, video player, etc., for parsing.

Note that using these mechanisms alone does not provide end-to-end integrity of HTTP messages over multiple hops since metadata could be manipulated at any stage. Methods to protect metadata are discussed in Section 6.3.

Digital signatures are widely used together with checksums to provide the certain identification of the origin of a message [FIPS186-5]. Such signatures can protect one or more HTTP fields and there are additional considerations when Integrity fields are included in this set.

There are no restrictions placed on the type or format of digital signature that Integrity fields can be used with. One possible approach is to combine them with HTTP Message Signatures [SIGNATURES].

Digests explicitly depend on the "representation metadata" (e.g., the values of Content-Type, Content-Encoding, etc.). A signature that protects Integrity fields but not other "representation metadata" can expose the communication to tampering. For example, an actor could manipulate the Content-Type field-value and cause a digest validation failure at the recipient, preventing the application from accessing the representation. Such an attack consumes the resources of both endpoints. See also Section 3.2.

Signatures are likely to be deemed an adversarial setting when applying Integrity fields; see Section 5. Repr-Digest offers an interesting possibility when combined with signatures. In the scenario where there is no content to send, the digest of an empty string can be included in the message and, if signed, can help the recipient detect if content was added either as a result of accident or purposeful manipulation. The opposite scenario is also supported; including an Integrity field for content and signing it can help a recipient detect where the content was removed.

Any mangling of Integrity fields might affect signature validation. Examples of such mangling include de-duplicating digests or combining different field values (see Section 5.2 of [HTTP]).

Before sending Integrity fields in a trailer section, the sender should consider that intermediaries are explicitly allowed to drop any trailer (see Section 6.5.2 of [HTTP]).

When Integrity fields are used in a trailer section, the field-values are received after the content. Eager processing of content before the trailer section prevents digest validation, possibly leading to processing of invalid data.

One of the benefits of using Integrity fields in a trailer section is that it allows hashing of bytes as they are sent. However, it is possible to design a hashing algorithm that requires processing of content in such a way that would negate these benefits. For example, Merkle Integrity Content Encoding [MICE] requires content to be processed in reverse order. This means the complete data needs to be available, which means there is negligible processing difference in sending an Integrity field in a header versus a trailer section.

Content coding mechanisms can support different encoding parameters, meaning that the same input content can produce different outputs. For example, GZIP supports multiple compression levels. Such encoding parameters are generally not communicated as representation metadata. For instance, different compression levels would all use the same "Content-Encoding: gzip" field. Other examples include where encoding relies on nonces or timestamps, such as the aes128gcm content coding defined in [RFC8188].

Since it is possible for there to be variation within content coding, the checksum conveyed by the Integrity fields cannot be used to provide a proof of integrity "at rest" unless the whole content is persisted.

The security properties of hashing algorithms are not fixed. Algorithm agility (see [RFC7696]) is achieved by providing implementations with flexibility to choose hashing algorithms from the IANA Hash Algorithms for HTTP Digest Fields registry; see Section 7.2.

Transition from weak algorithms is supported by negotiation of hashing algorithm using Want-Content-Digest or Want-Repr-Digest (see Section 4) or by sending multiple digests from which the receiver chooses. A receiver that depends on a digest for security will be vulnerable to attacks on the weakest algorithm it is willing to accept. Endpoints are advised that sending multiple values consumes resources that may be wasted if the receiver ignores them (see Section 3).

While algorithm agility allows the migration to stronger algorithms, it does not prevent the use of weaker algorithms. Integrity fields do not provide any mitigations for downgrade or substitution attacks (see Section 1 of [RFC6211]) of the hashing algorithm. To protect against such attacks, endpoints could restrict their set of supported algorithms to stronger ones and protect the fields' values by using TLS and/or digital signatures.

Integrity field validation consumes computational resources. In order to avoid resource exhaustion, implementations can restrict validation of the algorithm types, the number of validations, or the size of content. In these cases, skipping validation entirely or ignoring validation failure of a more-preferred algorithm leaves the possibility of a downgrade attack (see Section 6.6).

IANA has created the new "Hash Algorithms for HTTP Digest Fields" registry at <https://www.iana.org/assignments/http-digest-hash-alg/> and populated it with the entries in Table 2. The procedure for new registrations is provided in Section 5.

This registry is deprecated since it lists the algorithms that can be used with the Digest and Want-Digest fields defined in [RFC3230], which has been obsoleted by RFC 9530. While registration is not closed, new registrations are encouraged to use the Hash Algorithms for HTTP Digest Fields registry instead.

In this example, the message content conveys complete representation data. This means that in the response, Content-Digest and Repr-Digest are both computed over the JSON object {"hello": "world"} followed by an LF; thus, they have the same value.

In this example, a HEAD request is used to retrieve the checksum of a resource.

The response Content-Digest field-value is computed on empty content. Repr-Digest is calculated over the JSON object {"hello": "world"} followed by an LF, which is not shown because there is no content.

In this example, the client makes a range request and the server responds with partial content.

NOTE: '\' line wrapping per RFC 8792

Repr-Digest: \
  sha-256=:RK/0qy18MlBSVnWgjwz6lZEWjP/lF5HF9bvEF8FabDg=:

NOTE: '\' line wrapping per RFC 8792

Content-Digest: \
  sha-256=:jjcgBDWNAtbYUXI37CVG3gRuGOAjaaDRGpIUFsdyepQ=:

The request contains a Repr-Digest field-value calculated on the enclosed representation. It also includes an Accept-Encoding: br header field that advertises that the client supports Brotli encoding.

The response includes a Content-Encoding: br that indicates the selected representation is Brotli-encoded. The Repr-Digest field-value is therefore different compared to the request.

For presentation purposes, the response body is displayed as a sequence of hex-encoded bytes because it contains non-printable characters.

The request Repr-Digest field-value is calculated on the enclosed content, which is the JSON object {"hello": "world"} followed by an LF.

The response Repr-Digest field-value depends on the representation metadata header fields, including Content-Encoding: br, even when the response does not contain content.

NOTE: '\' line wrapping per RFC 8792

PUT /items/123 HTTP/1.1
Host: foo.example
Content-Type: application/json
Content-Length: 19
Accept-Encoding: br
Repr-Digest: \
  sha-256=:RK/0qy18MlBSVnWgjwz6lZEWjP/lF5HF9bvEF8FabDg==:

{"hello": "world"}

The response contains two digest values using different algorithms.

For presentation purposes, the response body is displayed as a sequence of hex-encoded bytes because it contains non-printable characters.

The request Repr-Digest field-value is computed on the enclosed representation (see Section 3.1), which is the JSON object {"title": "New Title"} followed by an LF.

The representation enclosed in the response is a multiline JSON object followed by an LF. It refers to the resource identified by Content-Location (see Section 6.4.2 of [HTTP]); thus, an application can use Repr-Digest in association with the resource referenced by Content-Location.

The request Repr-Digest field-value is computed on the enclosed representation (see Section 3.1), which is the JSON object {"title": "New Title"} followed by an LF.

The representation enclosed in the response describes the status of the request, so Repr-Digest is computed on that enclosed representation. It is a multiline JSON object followed by an LF.

Response Repr-Digest has no explicit relation with the resource referenced by Location.

This case is analogous to a POST request where the target resource reflects the target URI.

The PATCH request uses the application/merge-patch+json media type defined in [RFC7396]. Repr-Digest is calculated on the content that corresponds to the patch document and is the JSON object {"title": "New Title"} followed by an LF.

The response Repr-Digest field-value is computed on the complete representation of the patched resource. It is a multiline JSON object followed by an LF.

Note that a 204 No Content response without content, but with the same Repr-Digest field-value, would have been legitimate too. In that case, Content-Digest would have been computed on an empty content.

In error responses, the representation data does not necessarily refer to the target resource. Instead, it refers to the representation of the error.

In the following example, a client sends the same request from Figure 26 to patch the resource located at /books/123. However, the resource does not exist and the server generates a 404 response with a body that describes the error in accordance with [RFC9457].

The response Repr-Digest field-value is computed on this enclosed representation. It is a multiline JSON object followed by an LF.

An origin server sends Repr-Digest as trailer field, so it can calculate digest-value while streaming content and thus mitigate resource consumption. The Repr-Digest field-value is the same as in Appendix B.1 because Repr-Digest is designed to be independent of the use of one or more transfer codings (see Section 3).

In the response content below, the string "\r\n" represents the CRLF bytes.

The client requests a digest and prefers "sha". The server is free to reply with "sha-256" anyway.

The client requests a "sha" digest because that is the only algorithm it supports. The server is not obliged to produce a response containing a "sha" digest; it instead uses a different algorithm.

Appendix C.2 is an example where a server ignores the client's preferred digest algorithm. Alternatively, a server can also reject the request and return a response with an error status code such as 4xx or 5xx. This specification does not prescribe any requirement on status code selection; the following example illustrates one possible option.

In this example, the client requests a "sha" Repr-Digest, and the server returns an error with problem details [RFC9457] contained in the content. The problem details contain a list of the hashing algorithms that the server supports. This is purely an example; this specification does not define any format or requirements for such content.