| Internet Engineering Task Force (IETF) | M. Nottingham |
| Request for Comments: 8288 | October 2017 |
| Obsoletes: 5988 | |
| Category: Standards Track | |
| ISSN: 2070-1721 |
This specification defines a model for the relationships between resources on the Web ("links") and the type of those relationships ("link relation types").¶
It also defines the serialisation of such links in HTTP headers with the Link header field.¶
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/rfc8288.¶
Copyright (c) 2017 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 Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.¶
This document may contain material from IETF Documents or IETF Contributions published or made publicly available before November 10, 2008. The person(s) controlling the copyright in some of this material may not have granted the IETF Trust the right to allow modifications of such material outside the IETF Standards Process. Without obtaining an adequate license from the person(s) controlling the copyright in such materials, this document may not be modified outside the IETF Standards Process, and derivative works of it may not be created outside the IETF Standards Process, except to format it for publication as an RFC or to translate it into languages other than English.¶
This specification defines a model for the relationships between resources on the Web ("links") and the type of those relationships ("link relation types").¶
HTML [W3C.REC-html5-20141028] and Atom [RFC4287] both have well-defined concepts of linking; Section 2 generalises this into a framework that encompasses linking in these formats and (potentially) elsewhere.¶
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 Backus-Naur Form (ABNF) [RFC5234] notation of [RFC7230], including the #rule, and explicitly includes the following rules from it: quoted-string, token, SP (space), BWS (bad whitespace), OWS (optional whitespace), RWS (required whitespace), LOALPHA, DIGIT.¶
Additionally, the following rules are included: ¶
The requirements regarding conformance and error handling highlighted in [RFC7230], Section 2.5 apply to this document.¶
In this specification, a link is a typed connection between two resources and is comprised of:¶
A link can be viewed as a statement of the form "link context has a link relation type resource at link target, which has target attributes".¶
For example, "https://www.example.com/" has a "canonical" resource at "https://example.com", which has a "type" of "text/html".¶
Link contexts and link targets are both Internationalized Resource Identifiers (IRIs) [RFC3987]. However, in the common case, the link context will also be a URI [RFC3986], because many protocols (such as HTTP) do not support dereferencing IRIs. Likewise, the link target will sometimes be converted to a URI (see [RFC3987], Section 3.1) in serialisations that do not support IRIs (such as the Link header field defined in Section 3).¶
This specification does not place restrictions on the cardinality of links; there can be multiple links to and from a particular target and multiple links of the same or different types between a given context and target. Likewise, the relative ordering of links in any particular serialisation, or between serialisations (e.g., the Link header field and in-content links), is not specified or significant in this specification; applications that wish to consider ordering significant can do so.¶
Links are conveyed in link serialisations; they are the "bytes on the wire", and can occur in various forms. For example, Atom [RFC4287] and HTML [W3C.REC-html5-20141028] both defined serialisations of links into their respective formats, and Section 3 defines how to serialise links in HTTP header fields.¶
This specification does not define a general syntax for links across different serialisations, nor does it mandate a specific context for any given link; it is expected that serialisations of links will specify both aspects.¶
Finally, links are used by link applications. Generally, an application will define the link relation type(s) it uses, along with the serialisation(s) that they might occur within. For example, the application "Web browsing" looks for the "stylesheet" link relation type in the HTML link serialisation (and optionally in the Link header field), whereas the application "AtomPub" uses the "edit" and "edit-media" link relations in the Atom serialisation.¶
In the simplest case, a link relation type identifies the semantics of a link. For example, a link with the relation type "copyright" indicates that the current link context has a copyright resource at the link target.¶
Link relation types can also be used to indicate that the target resource has particular attributes, or exhibits particular behaviours; for example, a "service" link implies that the link target can be used as part of a defined protocol (in this case, a service description).¶
Relation types are not to be confused with media types [RFC2046]; they do not identify the format of the representation that results when the link is dereferenced. Rather, they only describe how the current context is related to another resource.¶
Relation types SHOULD NOT infer any additional semantics based upon the presence or absence of another link relation type, or its own cardinality of occurrence. An exception to this is the combination of the "alternate" and "stylesheet" registered relation types, which has special meaning in HTML for historical reasons.¶
There are two kinds of relation types: registered and extension.¶
Well-defined relation types can be registered as tokens for convenience and/or to promote reuse by other applications, using the procedure in Section 2.1.1.1.¶
Registered relation type names MUST conform to the reg-rel-type rule (see Section 3.3) and MUST be compared character by character in a case-insensitive fashion. They SHOULD be appropriate to the specificity of the relation type; that is, if the semantics are highly specific to a particular application, the name should reflect that, so that more general names are available for less-specific use.¶
Registered relation types MUST NOT constrain the media type of the link context and MUST NOT constrain the available representation media types of the link target. However, they can specify the behaviours and properties of the target resource (e.g., allowable HTTP methods, and request and response media types that are required be supported).¶
Historically, registered relation types have been identified with a URI [RFC3986] by prefixing their names with an application-defined base URI (e.g., see Appendix A.2). This practice is NOT RECOMMENDED, because the resulting strings will not be considered equivalent to the registered relation types by other applications. Applications that do use such URIs internally MUST NOT use them in link serialisations that do not explicitly accommodate them.¶
The "Link Relations" registry is located at <https://www.iana.org/assignments/link-relations/>. Registration requests can be made by following the instructions located there or by sending an email to the <link-relations@ietf.org> mailing list.¶
Registration requests consist of at least the following information:¶
The expert(s) can define additional fields to be collected in the registry.¶
General requirements for registered relation types are described in Section 2.1.1.¶
Registrations MUST reference a freely available, stable specification.¶
Note that relation types can be registered by third parties (including the expert(s)), if the expert(s) determines that an unregistered relation type is widely deployed and not likely to be registered in a timely manner otherwise. Such registrations still are subject to the requirements defined, including the need to reference a specification.¶
Relation types are registered using the Specification Required policy (see Section 4.6 of [RFC8126]), which implies review and approval by a designated expert.¶
The goal of the registry is to reflect common use of links on the Internet. Therefore, the expert(s) should be strongly biased towards approving registrations, unless they are abusive, frivolous, not likely to be used on the Internet, or actively harmful to the Internet and/or the Web (not merely aesthetically displeasing or architecturally dubious). As stated in Section 2.1.1, the expert(s) can withhold registration of names that are too general for the proposed application.¶
The expert(s) will clearly identify any issues that cause a registration to be refused. Advice about the semantics of a proposed link relation type can be given, but if it does not block registration, this should be explicitly stated.¶
When a request is approved, the expert(s) will inform IANA, and the registration will be processed. The IESG is the final arbiter of any objection.¶
Applications that don't wish to register a relation type can use an extension relation type, which is a URI [RFC3986] that uniquely identifies the relation type. Although the URI can point to a resource that contains a definition of the semantics of the relation type, clients SHOULD NOT automatically access that resource to avoid overburdening its server.¶
The URI used for an extension relation type SHOULD be under the control of the person or party defining it or be delegated to them.¶
When extension relation types are compared, they MUST be compared as strings (after converting to URIs if serialised in a different format) in a case-insensitive fashion, character by character. Because of this, all-lowercase URIs SHOULD be used for extension relations.¶
Note that while extension relation types are required to be URIs, a serialisation of links can specify that they are expressed in another form, as long as they can be converted to URIs.¶
Target attributes are a list of key/value pairs that describe the link or its target; for example, a media type hint.¶
They can be defined both by individual link relation types and by link serialisations.¶
This specification does not attempt to coordinate the name of target attributes, their cardinality, or use. Those creating and maintaining serialisations SHOULD coordinate their target attributes to avoid conflicts in semantics or syntax and MAY define their own registries of target attributes.¶
The names of target attributes SHOULD conform to the token rule, but SHOULD NOT include any of the characters "%", "'", or "*", for portability across serialisations and MUST be compared in a case-insensitive fashion.¶
Target attribute definitions SHOULD specify:¶
This specification does define target attributes for use in the Link HTTP header field in Section 3.4.¶
The Link header field provides a means for serialising one or more links into HTTP headers.¶
The ABNF for the field value is:¶
Link = #link-value link-value = "<" URI-Reference ">" *( OWS ";" OWS link-param ) link-param = token BWS [ "=" BWS ( token / quoted-string ) ]
Note that any link-param can be generated with values using either the token or the quoted-string syntax; therefore, recipients MUST be able to parse both forms. In other words, the following parameters are equivalent:¶
x=y x="y"
Previous definitions of the Link header did not equate the token and quoted-string forms explicitly; the title parameter was always quoted, and the hreflang parameter was always a token. Senders wishing to maximize interoperability will send them in those forms.¶
Individual link-params specify their syntax in terms of the value after any necessary unquoting (as per [RFC7230], Section 3.2.6).¶
This specification establishes the link-params "rel", "anchor", and "rev" (which are part of the general link model), as well as "hreflang", "media", "title", "title*", and "type" (which are target attributes defined by the serialisation).¶
Each link-value conveys one target IRI as a URI-Reference (after conversion to one, if necessary; see [RFC3987], Section 3.1) inside angle brackets ("<>"). If the URI-Reference is relative, parsers MUST resolve it as per [RFC3986], Section 5. Note that any base IRI appearing in the message's content is not applied.¶
By default, the context of a link conveyed in the Link header field is the URL of the representation it is associated with, as defined in [RFC7231], Section 3.1.4.1, and is serialised as a URI.¶
When present, the anchor parameter overrides this with another URI, such as a fragment of this resource, or a third resource (i.e., when the anchor value is an absolute URI). If the anchor parameter's value is a relative URI, parsers MUST resolve it as per [RFC3986], Section 5. Note that any base URI from the body's content is not applied.¶
The ABNF for the anchor parameter's value is:¶
URI-Reference ; Section 4.1 of [RFC3986]
Link application can choose to ignore links with an anchor parameter. For example, the application in use might not allow the link context to be assigned to a different resource. In such cases, the entire link is to be ignored; link applications MUST NOT process the link without applying the anchor.¶
Note that depending on HTTP status code and response headers, the link context might be "anonymous" (i.e., no link context is available). For example, this is the case on a 404 response to a GET request.¶
The relation type of a link conveyed in the Link header field is conveyed in the "rel" parameter's value. The rel parameter MUST be present but MUST NOT appear more than once in a given link-value; occurrences after the first MUST be ignored by parsers.¶
The rel parameter can, however, contain multiple link relation types. When this occurs, it establishes multiple links that share the same context, target, and target attributes.¶
The "rev" parameter has been used in the past to indicate that the semantics of the relationship are in the reverse direction. That is, a link from A to B with REL="X" expresses the same relationship as a link from B to A with REV="X". rev is deprecated by this specification because it often confuses authors and readers; in most cases, using a separate relation type is preferable.¶
The ABNF for the rel and rev parameters' values is:¶
relation-type *( 1*SP relation-type )
where:¶
Note that extension relation types are REQUIRED to be absolute URIs in Link header fields and MUST be quoted when they contain characters not allowed in tokens, such as a semicolon (";") or comma (",") (as these characters are used as delimiters in the header field itself).¶
The Link header field defines several target attributes specific to this serialisation and also allows extension target attributes. Target attributes are serialised in the Link header field as parameters (see [RFC7231], Section 3.1.1.1 for the definition of their syntax).¶
The "hreflang", "media", "title", "title*", and "type" link-params can be translated to serialisation-defined target attributes for the link.¶
The "hreflang" attribute, when present, is a hint indicating what the language of the result of dereferencing the link should be. Note that this is only a hint; for example, it does not override the Content-Language header field of a HTTP response obtained by actually following the link. Multiple hreflang attributes on a single link-value indicate that multiple languages are available from the indicated resource.¶
The ABNF for the hreflang parameter's value is:¶
Language-Tag
The "media" attribute, when present, is used to indicate intended destination medium or media for style information (see [W3C.REC-html5-20141028], Section 4.2.4). Its value MUST be quoted if it contains a semicolon (";") or comma (","). There MUST NOT be more than one media attribute in a link-value; occurrences after the first MUST be ignored by parsers.¶
The ABNF for the media parameter's value is:¶
media-query-list
The "title" attribute, when present, is used to label the destination of a link such that it can be used as a human-readable identifier (e.g., a menu entry) in the language indicated by the Content-Language header field (if present). The title attribute MUST NOT appear more than once in a given link; occurrences after the first MUST be ignored by parsers.¶
The "title*" link-param can be used to encode this attribute in a different character set and/or contain language information as per [RFC8187]. The title* link-param MUST NOT appear more than once in a given link-value; occurrences after the first MUST be ignored by parsers. If the attribute does not contain language information, its language is indicated by the Content-Language header field (when present).¶
If both the title and title* link-params appear in a link, applications SHOULD use the title* link-param's value for the title attribute.¶
The "type" attribute, when present, is a hint indicating what the media type of the result of dereferencing the link should be. Note that this is only a hint; for example, it does not override the Content-Type header field of a HTTP response obtained by actually following the link. The type attribute MUST NOT appear more than once in a given link-value; occurrences after the first MUST be ignored by parsers.¶
The ABNF for the type parameter's value is:¶
type-name "/" subtype-name ; see Section 4.2 of [RFC6838]
Other link-params are link-extensions and are to be considered as target attributes.¶
Such target attributes MAY be defined to use the encoding in [RFC8187] (e.g., "example" and "example*"). When both forms are present, they SHOULD be considered to be the same target attribute; applications SHOULD use the value of the name ending in "*" (after [RFC8187] decoding) but MAY fall back to the other value if there is an error in decoding it, or if they do not support decoding.¶
For example:¶
Link: <http://example.com/TheBook/chapter2>; rel="previous";
title="previous chapter"
indicates that "chapter2" is previous to this resource in a logical navigation path.¶
Similarly,¶
Link: </>; rel="http://example.net/foo"
indicates that the root resource ("/") is related to this resource with the extension relation type "http://example.net/foo".¶
This link:¶
Link: </terms>; rel="copyright"; anchor="#foo"
indicates that the linked copyright terms only apply to the portion of the document indicated by the (media type-specific) fragment identifier "foo".¶
The example below shows an instance of the Link header field encoding multiple links and also the use of the encoding from RFC 8187 to encode both non-ASCII characters and language information.¶
Link: </TheBook/chapter2>;
rel="previous"; title*=UTF-8'de'letztes%20Kapitel,
</TheBook/chapter4>;
rel="next"; title*=UTF-8'de'n%c3%a4chstes%20Kapitel
Here, both links have titles encoded in UTF-8, both use the German language ("de"), and the second link contains the Unicode code point U+00E4 ("LATIN SMALL LETTER A WITH DIAERESIS").¶
Note that link-values can convey multiple links between the same link target and link context; for example:¶
Link: <http://example.org/>;
rel="start http://example.net/relation/other"
Here, the link to "http://example.org/" has the registered relation type "start" and the extension relation type "http://example.net/relation/other".¶
Finally, this header field:¶
Link: <https://example.org/>; rel="start",
<https://example.org/index>; rel="index"
is equivalent to these:¶
Link: <https://example.org/>; rel="start" Link: <https://example.org/index>; rel="index"
This specification updates the "Message Headers" registry entry for "Link" in HTTP [RFC3864] to refer to this document.¶
Header Field Name: Link Protocol: http Status: standard Reference: RFC 8288
This specification updates the registration procedures for the "Link Relation Types" registry; see Section 2.1.1.1. Also, all references to RFC 5988 in that registry have been replaced with references to this document.¶
IANA will direct any incoming requests regarding the registry to this document and, if defined, the processes established by the expert(s); typically, this will mean referring them to the registry Web page.¶
Note that the expert(s) is allowed (as per Section 2.1.1.1) to define additional fields to be collected in the registry.¶
Per this specification, IANA has removed the "Link Relation Application Data" registry, as it has not been used, and future use is not anticipated.¶
The content of the Link header field is not secure, private, or integrity-guaranteed. Use of Transport Layer Security (TLS) with HTTP [RFC2818] is currently the only end-to-end way to provide these properties.¶
Link applications ought to consider the attack vectors opened by automatically following, trusting, or otherwise using links gathered from HTTP header fields.¶
For example, Link header fields that use the "anchor" parameter to associate a link's context with another resource cannot be trusted since they are effectively assertions by a third party that could be incorrect or malicious. Applications can mitigate this risk by specifying that such links should be discarded unless some relationship between the resources is established (e.g., they share the same authority).¶
Dereferencing links has a number of risks, depending on the application in use. For example, the Referer header [RFC7231] can expose information about the application's state (including private information) in its value. Likewise, cookies [RFC6265] are another mechanism that, if used, can become an attack vector. Applications can mitigate these risks by carefully specifying how such mechanisms should operate.¶
Link targets may need to be converted to URIs in order to express them in serialisations that do not support IRIs. This includes the Link HTTP header field.¶
Similarly, the anchor parameter of the Link header field does not support IRIs; therefore, IRIs must be converted to URIs before inclusion there.¶
Relation types are defined as URIs, not IRIs, to aid in their comparison. It is not expected that they will be displayed to end users.¶
Note that registered Relation Names are required to be lowercase ASCII letters.¶
Header fields (Section 3) are only one serialisation of links; other specifications have defined alternative serialisations.¶
HTML motivated the original syntax of the Link header field, and many of the design decisions in this document are driven by a desire to stay compatible with it.¶
In HTML, the link element can be mapped to links as specified here by using the "href" attribute for the target URI, and "rel" to convey the relation type, as in the Link header field. The context of the link is the URI associated with the entire HTML document. HTML also defines several attributes on links that can be seen as target attributes, including "media", "hreflang", "type", and "sizes".¶
Section 4.8 of HTML5 [W3C.REC-html5-20141028] defines modern HTML links. That document links to the Microformats Wiki as a registry; over time, the IANA registry ought to mirror its contents and, ideally, eventually replace it (although that depends on the HTML community).¶
Surveys of existing HTML content have shown that unregistered link relation types that are not URIs are (perhaps inevitably) common. Consuming HTML implementations ought not consider such unregistered short links to be errors, but rather relation types with a local scope (i.e., their meaning is specific and perhaps private to that document).¶
Finally, the HTML specification gives a special meaning when the "alternate" relation types coincide with other relation types in the same link. Such links ought to be serialised in the Link header field using a single list of relation-types (e.g., rel="alternate stylesheet") to preserve this relationship.¶
Atom [RFC4287] is a link serialisation that conveys links in the atom:link element, with the "href" attribute indicating the link target and the "rel" attribute containing the relation type. The context of the link is either a feed locator or an entry ID, depending on where it appears; generally, feed-level links are obvious candidates for transmission as a Link header field.¶
When serialising an atom:link into a Link header field, it is necessary to convert link targets (if used) to URIs.¶
Atom defines extension relation types in terms of IRIs. This specification redefines them as URIs, to simplify and reduce errors in their comparison.¶
Atom allows registered link relation types to be serialised as absolute URIs using a prefix, "http://www.iana.org/assignments/relation/". This prefix is specific to the Atom serialisation.¶
Furthermore, link relation types are always compared in a case-sensitive fashion; therefore, registered link relation types SHOULD be converted to their registered form (usually, lowercase) when serialised in an Atom document.¶
Note also that while the Link header field allows multiple relations to be serialised in a single link, atom:link does not. In this case, a single link-value may map to several atom:link elements.¶
As with HTML, atom:link defines some attributes that are not explicitly mirrored in the Link header field syntax, but they can also be used as link-extensions to maintain fidelity.¶
This appendix outlines a set of non-normative algorithms: for parsing the Link header(s) out of a header set, for parsing a Link header field value, and algorithms for parsing generic parts of the field value.¶
These algorithms are more permissive than the ABNF defining the syntax might suggest; the error handling embodied in them is a reasonable approach, but not one that is required. As such they are advisory only, and in cases where there is disagreement, the correct behaviour is defined by the body of this specification.¶
This algorithm can be used to parse the Link header fields that a HTTP header set contains. Given a header_set of (string field_name, string field_value) pairs, assuming ASCII encoding, it returns a list of link objects.¶
This algorithm parses zero or more comma-separated link-values from a Link header field. Given a string field_value, assuming ASCII encoding, it returns a list of link objects.¶
This algorithm parses the parameters from a header field value. Given input, an ASCII string, it returns a list of (string parameter_name, string parameter_value) tuples that it contains. input is modified to remove the parsed parameters.¶
This algorithm parses a quoted string, as per [RFC7230], Section 3.2.6. Given input, an ASCII string, it returns an unquoted string. input is modified to remove the parsed string.¶
This specification has the following differences from its predecessor, RFC 5988:¶