Internet-Draft Retrofit Structured Fields for HTTP December 2021
Nottingham Expires 5 June 2022 [Page]
Workgroup:
Network Working Group
Internet-Draft:
draft-nottingham-http-structure-retrofit
Published:
Intended Status:
Informational
Expires:
Author:
M. Nottingham

Retrofit Structured Fields for HTTP

Abstract

This specification defines how a selection of existing HTTP fields can be handled as Structured Fields.

Discussion Venues

This note is to be removed before publishing as an RFC.

information can be found at https://mnot.github.io/I-D/.

Source for this draft and an issue tracker can be found at https://github.com/mnot/I-D/labels/http-structure-retrofit.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on 5 June 2022.

Table of Contents

1. Introduction

Structured Field Values for HTTP [STRUCTURED-FIELDS] introduced a data model with associated parsing and serialisation algorithms for HTTP field values. Header fields that are defined as Structured Fields can realise a number of benefits, including:

However, a field needs to be defined as a Structured Field for these benefits to be realised. Many existing fields are not, making up the bulk of header and trailer fields seen in HTTP traffic on the Internet.

This specification defines how a selection of existing HTTP fields can be handled as Structured Fields, so that these benefits can be realised -- thereby making them Retrofit Structured Fields.

It does so using two techniques. Section 2 lists compatible fields -- those that can be handled as if they were Structured Fields due to the similarity of their defined syntax to that in Structured Fields. Section 3 lists mapped fields -- those whose syntax needs to be transformed into an underlying data model which is then mapped into that defined by Structured Fields.

While implementations can parse and serialise Compatible Fields as Structured Fields subject to the caveats in Section 2, a sender cannot generate mapped fields from Section 3 and expect them to be understood and acted upon by the recipient without prior negotiation. This specification does not define such a mechanism.

1.1. Notational Conventions

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.

2. Compatible Fields

HTTP fields with the following names can usually have their values handled as Structured Fields according to the listed parsing and serialisation algorithms in [STRUCTURED-FIELDS], subject to the listed caveats.

The listed types are chosen for compatibility with the defined syntax of the field as well as with actual Internet traffic (see Appendix A). However, not all instances of these fields will successfully parse. This might be because the field value is clearly invalid, or it might be because it is valid but not parseable as a Structured Field.

As such, an application using this specification will need to consider how to handle these field values. Depending on its requirements, it might be advisable to reject such values, treat them as opaque strings, or attempt to recover a structured value from them in an ad hoc fashion.

Note the following caveats:

Parameters:

HTTP parameter names are case-insensitive (as per Section 5.6.6 of [HTTP]), but Structured Fields require them to be all-lowercase. Although the vast majority of parameters seen in typical traffic are all-lowercase, compatibility can be improved by force-lowercasing parameters when encountered.

Empty Field Values:

Empty and whitespace-only field values are considered errors in Structured Fields. For compatible fields, an empty field indicates that the field should be silently ignored.

Alt-Svc:

Some ALPN tokens (e.g., h3-Q43) do not conform to key's syntax. Since the final version of HTTP/3 uses the h3 token, this shouldn't be a long-term issue, although future tokens may again violate this assumption.

Cache-Control, Expect-CT, Pragma, Prefer, Preference-Applied, Surrogate-Control:

These Dictionary-based fields consider the key to be case-insensitive, but Structured Fields requires keys to be all-lowercase. Although the vast majority of values seen in typical traffic are all-lowercase, compatibility can be improved by force-lowercasing these Dictionary keys when encountered.

Content-Length:

Content-Length is defined as a List because it is not uncommon for implementations to mistakenly send multiple values. See Section 8.6 of [HTTP] for handling requirements.

Retry-After:

Only the delta-seconds form of Retry-After is supported; a Retry-After value containing a http-date will need to be either converted into delta-seconds or represented as a raw value.

3. Mapped Fields

HTTP fields with the following names can have their values represented in Structured Fields by mapping them into its data types and then serialising the result using an alternative field name.

For example, the Date HTTP header field carries a string representing a date:

Date: Sun, 06 Nov 1994 08:49:37 GMT

Its value is more efficiently represented as an integer number of delta seconds from the Unix epoch (00:00:00 UTC on 1 January 1970, minus leap seconds). Thus, the example above would be mapped as:

SF-Date: 784072177

As in Section 2, these fields are unable to represent values that are not Structured Fields, and so an application using this specification will need to how to support such values. Typically, serialising them using the original field name is sufficient.

Each field name listed below indicates a replacement field name and a means of mapping its original value into a Structured Field.

3.1. URLs

The following field names (paired with their replacement field names) have values that can be represented as Structured Fields by considering the original field's value as a string.

  • Content-Location - SF-Content-Location
  • Location - SF-Location
  • Referer - SF-Referer

For example, a Location field could be represented as:

SF-Location: "https://example.com/foo"

3.2. Dates

The following field names (paired with their replacement field names) have values that can be represented as Structured Fields by parsing their payload according to [RFC7231], Section 7.1.1.1, and representing the result as an integer number of seconds delta from the Unix Epoch (00:00:00 UTC on 1 January 1970, minus leap seconds).

  • Date - SF-Date
  • Expires - SF-Expires
  • If-Modified-Since - SF-IMS
  • If-Unmodified-Since - SF-IUS
  • Last-Modified - SF-LM

For example, an Expires field could be represented as:

SF-Expires: 1571965240

3.3. ETags

The following field names (paired with their replacement field names) have values that can be represented as Structured Fields by representing the entity-tag as a string, and the weakness flag as a boolean "w" parameter on it, where true indicates that the entity-tag is weak; if 0 or unset, the entity-tag is strong.

  • ETag - SF-ETag

For example:

SF-ETag: "abcdef"; w=?1

If-None-Match is a list of the structure described above.

  • If-None-Match - SF-INM

For example:

SF-INM: "abcdef"; w=?1, "ghijkl"

3.5. Cookies

The field-values of the Cookie and Set-Cookie fields [RFC6265] can be represented in Structured Fields as a List with parameters and a Dictionary, respectively.

The serialisation is almost identical, except that the Expires parameter is always a string (as it can contain a comma), multiple cookie-strings can appear in Set-Cookie, and cookie-pairs are delimited in Cookie by a comma, rather than a semicolon.

  • Set-Cookie: SF-Set-Cookie
  • Cookie: SF-Cookie
SF-Set-Cookie: lang=en-US; expires="Wed, 09 Jun 2021 10:18:14 GMT";
               samesite=Strict
SF-Cookie: SID=31d4d96e407aad42, lang=en-US

4. IANA Considerations

IANA is asked to register the following entries in the HTTP Field Name Registry with a status of "permanent" and referring to this document:

5. Security Considerations

Section 2 identifies existing HTTP fields that can be parsed and serialised with the algorithms defined in [STRUCTURED-FIELDS]. Variances from other implementations might be exploitable, particularly if they allow an attacker to target one implementation in a chain (e.g., an intermediary). However, given the considerable variance in parsers already deployed, convergence towards a single parsing algorithm is likely to have a net security benefit in the longer term.

Section 3 defines alternative representations of existing fields. Because downstream consumers might interpret the message differently based upon whether they recognise the alternative representation, implementations are prohibited from generating such fields unless they have negotiated support for them with their peer. This specification does not define such a mechanism, but any such definition needs to consider the implications of doing so carefully.

6. Normative References

[HTTP]
Fielding, R. T., Nottingham, M., and J. Reschke, "HTTP Semantics", Work in Progress, Internet-Draft, draft-ietf-httpbis-semantics-19, , <https://datatracker.ietf.org/doc/html/draft-ietf-httpbis-semantics-19>.
[RFC2119]
Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, , <https://www.rfc-editor.org/rfc/rfc2119>.
[RFC6265]
Barth, A., "HTTP State Management Mechanism", RFC 6265, DOI 10.17487/RFC6265, , <https://www.rfc-editor.org/rfc/rfc6265>.
[RFC7231]
Fielding, R., Ed. and J. Reschke, Ed., "Hypertext Transfer Protocol (HTTP/1.1): Semantics and Content", RFC 7231, DOI 10.17487/RFC7231, , <https://www.rfc-editor.org/rfc/rfc7231>.
[RFC8174]
Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, , <https://www.rfc-editor.org/rfc/rfc8174>.
[RFC8288]
Nottingham, M., "Web Linking", RFC 8288, DOI 10.17487/RFC8288, , <https://www.rfc-editor.org/rfc/rfc8288>.
[STRUCTURED-FIELDS]
Nottingham, M. and P-H. Kamp, "Structured Field Values for HTTP", RFC 8941, DOI 10.17487/RFC8941, , <https://www.rfc-editor.org/rfc/rfc8941>.

Appendix A. Data Supporting Field Compatibility

To help guide decisions about compatible fields, the HTTP response headers captured by the HTTP Archive https://httparchive.org in September 2021 (representing more than 528,000,000 HTTP exchanges) were parsed as Structured Fields using the types listed in Section 2, with the indicated number of successful header instances, failures, and the resulting failure rate:

accept                                 9,099 /        34 =   0.372%*
accept-encoding                      116,708 /        58 =   0.050%*
accept-language                      127,710 /        95 =   0.074%*
accept-patch                             281 /         0 =   0.000%
accept-ranges                    289,341,375 /     7,776 =   0.003%
access-control-allow-credentials  36,159,371 /     2,671 =   0.007%
access-control-allow-headers      25,980,519 /    23,181 =   0.089%
access-control-allow-methods      32,071,437 /    17,424 =   0.054%
access-control-allow-origin      165,719,859 /   130,247 =   0.079%
access-control-expose-headers     20,787,683 /     1,973 =   0.009%
access-control-max-age             9,549,494 /     9,846 =   0.103%
access-control-request-headers       165,882 /       503 =   0.302%*
access-control-request-method        346,135 /    30,680 =   8.142%*
age                              107,395,872 /    36,649 =   0.034%
allow                                579,822 /       281 =   0.048%
alt-svc                           56,773,977 / 4,914,119 =   7.966%
cache-control                    395,402,834 / 1,146,080 =   0.289%
connection                       112,017,641 /     3,491 =   0.003%
content-encoding                 225,568,224 /       237 =   0.000%
content-language                   3,339,291 /     1,744 =   0.052%
content-length                   422,415,406 /       126 =   0.000%
content-type                     503,950,894 /   507,133 =   0.101%
cross-origin-resource-policy     102,483,430 /       799 =   0.001%
expect                                     0 /        53 = 100.000%*
expect-ct                         54,129,244 /    80,333 =   0.148%
host                                  57,134 /     1,486 =   2.535%*
keep-alive                        50,606,877 /     1,509 =   0.003%
origin                                32,438 /     1,396 =   4.126%*
pragma                            66,321,848 /    97,328 =   0.147%
preference-applied                       189 /         0 =   0.000%
referrer-policy                   14,274,787 /     8,091 =   0.057%
retry-after                          523,533 /     7,585 =   1.428%
surrogate-control                    282,846 /       976 =   0.344%
te                                         1 /         0 =   0.000%
timing-allow-origin               91,979,983 /         8 =   0.000%
trailer                                1,171 /         0 =   0.000%
transfer-encoding                 15,098,518 /         0 =   0.000%
vary                             246,483,644 /    69,607 =   0.028%
x-content-type-options           166,063,072 /   237,255 =   0.143%
x-frame-options                   56,863,322 / 1,014,464 =   1.753%
x-xss-protection                 132,739,109 /   347,133 =   0.261%

Note that this data set only includes response headers, although some request headers are present, indicated with an asterisk (because, the Web). Also, Dictionary and Parameter keys have not been force-lowercased, with the result that any values containing uppercase keys are considered to fail.

The top thirty header fields in that data set that were not considered compatible are (* indicates that the field is mapped in Section 3):

Author's Address

Mark Nottingham
Prahran
VIC
Australia