jsonschema-core.xml

<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!ENTITY RFC2119 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
<!ENTITY RFC2616 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2616.xml">
<!ENTITY RFC3986 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3986.xml">
<!ENTITY RFC4627 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4627.xml">
<!ENTITY RFC5988 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.5988.xml">
]>
<?rfc toc="yes"?>
<?rfc symrefs="yes"?>
<?rfc compact="yes"?>
<?rfc subcompact="no"?>
<?rfc strict="no"?>
<?rfc rfcedstyle="yes"?>
<rfc category="info" docName="draft-zyp-json-schema-05" ipr="trust200902">
    <front>
        <title abbrev="JSON Schema">JSON Schema: core definitions and terminology</title>

        <author fullname="Francis Galiegue" initials="F" surname="Galiegue" role="editor">
            <address>
                <email>fgaliegue@gmail.com</email>
            </address>
        </author>

        <author fullname="Kris Zyp" initials="K" surname="Zyp" role="editor">
            <organization>SitePen (USA)</organization>
            <address>
                <postal>
                    <street>530 Lytton Avenue</street>
                    <city>Palo Alto, CA 94301</city>
                    <country>USA</country>
                </postal>
                <phone>+1 650 968 8787</phone>
                <email>kris@sitepen.com</email>
            </address>
        </author>

        <author fullname="Gary Court" initials="G" surname="Court">
            <address>
                <postal>
                    <street></street>
                    <city>Calgary, AB</city>
                    <country>Canada</country>
                </postal>
                <email>gary.court@gmail.com</email>
            </address>
        </author>

        <date year="2015"/>
        <workgroup>Internet Engineering Task Force</workgroup>
        <keyword>JSON</keyword>
        <keyword>Schema</keyword>
        <keyword>Hyper Schema</keyword>
        <keyword>Hypermedia</keyword>

        <abstract>
            <t>
                JSON Schema defines the media type "application/schema+json", a JSON based format
                for defining the structure of JSON data. JSON Schema provides a contract for what
                JSON data is required for a given application and how to interact with it. JSON
                Schema is intended to define validation, documentation, hyperlink navigation, and
                interaction control of JSON data.
            </t>
        </abstract>
    </front>

    <middle>
        <section title="Introduction">
            <t>
                JSON Schema is a JSON media type for defining the structure of JSON data. JSON
                Schema provides a contract for what JSON data is required for a given application
                and how to interact with it. JSON Schema is intended to define validation,
                documentation, hyperlink navigation, and interaction control of JSON data.
            </t>

            <t>
                This specification defines JSON Schema core terminology and mechanisms; related
                specifications build upon this specification and define different applications of
                JSON Schema.
            </t>
        </section>

        <section title="Conventions and Terminology">
            <t>
                <!-- The text in this section has been copied from the official boilerplate,
                and should not be modified.-->

                The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD",
                "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be
                interpreted as described in <xref target="RFC2119">RFC 2119</xref>.
            </t>

            <t>
                The terms "JSON", "JSON text", "JSON value", "member", "element", "object", "array",
                "number", "string", "boolean", "true", "false", and "null" in this document are to
                be interpreted as defined in <xref target="RFC4627">RFC 4627</xref>.
            </t>
        </section>

        <section title="Core terminology">

            <section title="Property, item">
                <t>
                    When refering to a JSON Object, as defined by <xref target="RFC4627"/>, the
                    terms "member" and "property" may be used interchangeably.
                </t>

                <t>
                    When refering to a JSON Array, as defined by <xref target="RFC4627"/>, the terms
                    "element" and "item" may be used interchangeably.
                </t>
            </section>

            <section title="JSON Schema, keywords">
                <t>
                    A JSON Schema is a JSON document, and that document MUST be an object. Object
                    members (or properties) defined by JSON Schema (this specification, or related
                    specifications) are called keywords, or schema keywords.
                </t>

                <t>
                    If an instance is described by a schema, and a keyword is defined in the schema, then within this document that instance is informally said to be "described by" the keyword as well.
                </t>

                <t>
                    A JSON Schema MAY contain properties which are not schema keywords.
                </t>
            </section>

            <section title="Empty schema">
                <t>
                    An empty schema is a JSON Schema with no properties, or with properties which
                    are not schema keywords.
                </t>
            </section>

            <section title="Root schema, subschema">
                <t>
                    This example of a JSON Schema has no subschemas:
                </t>
                <figure>
                    <artwork>
<![CDATA[
{
    "title": "root"
}
]]>
                    </artwork>
                </figure>

                <t>
                    JSON Schemas can also be nested, as in this example:
                </t>
                <figure>
                    <artwork>
<![CDATA[
{
    "title": "root",
    "otherSchema": {
        "title": "nested",
        "anotherSchema": {
            "title": "alsoNested"
        }
    }
}
]]>
                    </artwork>
                </figure>

                <t>
                    In this example, "nested" and "alsoNested" are subschemas, and "root" is a root
                    schema.
                </t>
            </section>

            <section title="JSON Schema primitive types">
                <t>
                    JSON Schema defines seven primitive types for JSON values:
                    <list style="hanging">
                        <t hangText="array">A JSON array.</t>
                        <t hangText="boolean">A JSON boolean.</t>
                        <t hangText="integer">A JSON number without a fraction or exponent part.</t>
                        <t hangText="number">Any JSON number. Number includes integer.</t>
                        <t hangText="null">The JSON null value.</t>
                        <t hangText="object">A JSON object.</t>
                        <t hangText="string">A JSON string.</t>
                    </list>
                </t>
            </section>

            <section title="JSON value equality">
                <t>
                    Two JSON values are said to be equal if and only if:

                    <list>
                        <t>both are nulls; or</t>
                        <t>both are booleans, and have the same value; or</t>
                        <t>both are strings, and have the same value; or</t>
                        <t>both are numbers, and have the same mathematical value; or</t>
                        <t>both are arrays, and:
                            <list>
                                <t>have the same number of items; and</t>
                                <t>items at the same index are equal according to this definition;
                                or</t>
                            </list>
                        </t>
                        <t>both are objects, and:
                            <list>
                                <t>have the same set of property names; and</t>
                                <t>values for a same property name are equal according to this
                                definition.</t>
                            </list>
                        </t>
                    </list>
                </t>
            </section>

            <section title="Instance">
                <t>
                    An instance is any JSON value. An instance may be described by one or more schemas.
                </t>

                <t>
                    An instance may also be referred to as "JSON instance", or "JSON data".
                </t>
            </section>

        </section>

        <section title="Overview">
            <t>
                This document proposes a new media type "application/schema+json" to identify JSON
                Schema for describing JSON data. JSON Schemas are themselves written in JSON.  This,
                and related specifications, define keywords allowing to describe this data in terms
                of allowable values, textual descriptions and interpreting relations with other
                resources. The following sections are a summary of features defined by related
                specifications.
            </t>

            <section title="Validation">
                <t>
                    JSON Schema describes the structure of a JSON document (for instance, required properties and array length constraints).
                    Applications can use this information to validate instances (check that constraints are met), or inform interfaces to collect user input such that the constraints are satisfied.
                </t>
                <t>Validation behaviour and keywords are specified in <xref target="json-schema-validation">a separate document</xref>.</t>
            </section>

            <section title="Hypermedia and linking">
                <t>
                    JSON Hyper-Schema describes the hypertext structure of a JSON document.
                    This includes link relations from the instance to other resources, interpretation of instances as multimedia data, and specification of access method (e.g. HTTP GET/POST) and submission data required to use an API.
                </t>
                <t>Hyper-schema behaviour and keywords are specified in <xref target="json-hyper-schema">a separate document</xref>.</t>
            </section>
        </section>

        <section title="General considerations">

            <section title="Applicability to all JSON values">
                <t>
                    It is acknowledged that an instance may be any valid JSON value as defined by <xref target="RFC4627"/>.
                    As such, JSON Schema does not mandate that an instance be of a particular type: JSON Schema can describe any JSON value, including null.
                </t>
            </section>

            <section title="Programming language independence">
                <t>
                    JSON Schema is programming language agnostic. The only limitations are the ones
                    expressed by <xref target="RFC4627"/> and those of the host programming
                    language.
                </t>
            </section>

            <section title="JSON Schema and HTTP">
                <t>
                    This specification acknowledges the role of <xref target="RFC2616">HTTP</xref>
                    as the dominant protocol in use on the Internet, and the wealth of
                    official specifications related to it.
                </t>

                <t>
                    This specification uses a subset of these specifications to recommend a set of
                    mechanisms, usable by this protocol, to associate JSON instances to one or more
                    schemas.
                </t>
            </section>

            <section title="JSON Schema and other protocols">
                <t>
                    JSON Schema does not define any semantics for the client-server interface for
                    any other protocols than HTTP. These semantics are application dependent, or
                    subject to agreement between the parties involved in the use of JSON Schema for
                    their own needs.
                </t>
            </section>

            <section title="Mathematical integers">
                <t>
                    It is acknowledged by this specification that some programming languages, and
                    their associated parsers, use different internal representations for floating
                    point numbers and integers, while others do not.
                </t>

                <t>
                    As a consequence, for interoperability reasons, JSON values used in the context
                    of JSON Schema, whether that JSON be a JSON Schema or an instance, SHOULD ensure
                    that mathematical integers be represented as integers as defined by this
                    specification.
                </t>
            </section>

            <section title="Extending JSON Schema">
                <t>
                    Implementations MAY choose to define additional keywords to JSON Schema. Save
                    for explicit agreement, schema authors SHALL NOT expect these additional
                    keywords to be supported by peer implementations. Implementations SHOULD ignore
                    keywords they do not support.
                </t>
                <t>
                    Authors of extensions to JSON Schema are encouraged to write their own meta-schemas, which extend the existing meta-schemas using "allOf".
                    This extended meta-schema SHOULD be referenced using the "$schema" keyword, to allow tools to follow the correct behaviour.
                </t>
            </section>

            <section title="Security considerations">
                <t>
                    Both schemas and instances are JSON values. As such, all security considerations
                    defined in <xref target="RFC4627">RFC 4627</xref> apply.
                </t>
            </section>

        </section>

        <section title='The "$schema" keyword'>
            <section title="Purpose">
                <t>
                    The "$schema" keyword is both used as a JSON Schema version identifier and the
                    location of a resource which is itself a JSON Schema, which describes any schema
                    written for this particular version.
                </t>

                <t>
                    This keyword SHOULD be located at the root of a JSON Schema. The value of this
                    keyword MUST be a <xref target="RFC3986">URI</xref> and a valid <xref
                    target="json-reference">JSON Reference</xref>; this URI MUST be both absolute
                    and normalized. The resource located at this URI MUST successfully describe
                    itself. It is RECOMMENDED that schema authors include this keyword in their
                    schemas.
                </t>

                <t>
                    The following values are predefined:

                    <list style="hanging">
                        <t hangText="http://json-schema.org/schema#">JSON Schema written against
                        the current version of the specification.</t>
                        <t hangText="http://json-schema.org/hyper-schema#">JSON Schema written
                        against the current version of the specification.</t>
                        <t hangText="http://json-schema.org/draft-05/schema#">JSON Schema written
                        against this version.</t>
                        <t hangText="http://json-schema.org/draft-05/hyper-schema#">JSON Schema hyperschema
                        written against this version.</t>
                        <t hangText="http://json-schema.org/draft-04/schema#"> JSON Schema written
                        against <xref target="json-schema-04">JSON Schema, draft v4</xref>.</t>
                        <t hangText="http://json-schema.org/draft-04/hyper-schema#"> JSON Schema hyperschema
                        written against <xref target="json-schema-04">JSON Schema, draft v3</xref>.</t>
                        <t hangText="http://json-schema.org/draft-03/schema#"> JSON Schema written
                        against <xref target="json-schema-03">JSON Schema, draft v3</xref>.</t>
                        <t hangText="http://json-schema.org/draft-03/hyper-schema#"> JSON Schema hyperschema
                        written against <xref target="json-schema-03">JSON Schema, draft
                        v3</xref>.</t>
                    </list>
                </t>
            </section>

        </section>

        <section title="URI resolution scopes and dereferencing">
            <section title="Definition">
                <t>
                    JSON Schema uses <xref target="json-reference">JSON Reference</xref> as a
                    mechanism for schema addressing. It extends this specification in two ways:

                    <list>
                        <t>JSON Schema offers facilities to alter the base URI against which a
                        reference must resolve by the means of the "id" keyword;</t>
                        <t>it offers schemas a mechanism to declare their own URIs, placing no limits on the structure of the URI</t>
                    </list>

                </t>

                <t>
                    Altering the URI within a schema is called defining a new resolution scope.
                    The initial resolution scope of a schema is the URI of the schema itself, or a suitable substitute URI if none is known.
                </t>
            </section>

            <section title='URI resolution scope alteration with the "id" keyword'>

                <section title="Valid values">
                    <t>
                        The value for this keyword MUST be a string, and MUST be a valid URI (relative or absolute).
                        This URI SHOULD be normalized, and SHOULD NOT be an empty fragment (#) or the empty string.
                    </t>
                 </section>

                <section title="Usage">
                    <t>
                        The "id" keyword is used to alter the resolution scope.
                        When an id is encountered, an implementation MUST resolve this id against
                        the most immediate parent scope. The resolved URI will be the new resolution
                        scope for this subschema and all its children, until another "id" is
                        encountered.
                    </t>

                    <t>
                        When using "id" to alter resolution scopes, schema authors SHOULD ensure
                        that resolution scopes are unique within the schema.
                    </t>

                    <t>
                        This schema will be taken as an example:
                    </t>

                    <!--
                        FIXME: http again as a scheme, maybe another one? It can be any scheme after
                        all
                    -->
                    <figure>
                        <artwork>
<![CDATA[
{
    "id": "http://x.y.z/rootschema.json#",
    "schema1": {
        "id": "#foo"
    },
    "schema2": {
        "id": "otherschema.json",
        "nested": {
            "id": "#bar"
        },
        "alsonested": {
            "id": "t/inner.json#a"
        }
    },
    "schema3": {
        "id": "some://where.else/completely#"
    }
}
]]>
                        </artwork>
                    </figure>

                    <t>
                        Subschemas at the following URI-encoded <xref target="json-pointer">JSON
                        Pointer</xref>s (starting from the root schema) define the following
                        resolution scopes:

                        <list style="hanging">
                            <t hangText="# (document root)">http://x.y.z/rootschema.json#</t>
                            <t hangText="#/schema1">http://x.y.z/rootschema.json#foo</t>
                            <t hangText="#/schema2">http://x.y.z/otherschema.json#</t>
                            <t hangText="#/schema2/nested">http://x.y.z/otherschema.json#bar</t>
                            <t hangText="#/schema2/alsonested">http://x.y.z/t/inner.json#a</t>
                            <t hangText="#/schema3">some://where.else/completely#</t>
                        </list>
                    </t>

                </section>

                <section title="Methods for dereferencing">
                    <t>
                        When resolving a URI to obtain a schema, tools MUST support the lookup of documents based on this URI, and the use of JSON Pointer fragments to locate particular locations within a schema document.
                        This is known as "canonical referencing".
                     </t>
                     <t>
                         Tools MAY also take note of the URIs schemas provide for themselves using "id", and use these values for schema dereferencing as well.
                         This is known as "inline referencing".
                    </t>

                    <t>
                        For example, consider this schema:
                    </t>

                    <figure>
                        <artwork>
<![CDATA[
{
    "id": "http://my.site/myschema#",
    "definitions": {
        "schema1": {
            "id": "schema1",
            "type": "integer"
        },
        "schema2", {
            "type": "array",
            "items": { "$ref": "schema1" }
        }
    }
}
]]>
                        </artwork>
                    </figure>

                    <t>
                        When an implementation encounters the "schema1" reference, it resolves it
                        against the most immediate parent scope, leading to URI
                        "http://my.site/schema1#". The way to process this URI will differ
                        according to the chosen dereferencing mode:

                        <list>
                            <t>if canonical dereferencing is used, the implementation will
                            dereference this URI, and fetch the content at this URI;</t>
                            <t>if inline dereferencing is used, the implementation MAY notice that
                            URI scope "http://my.site/schema1#" is already defined within the
                            schema, and choose to use the appropriate subschema.</t>
                        </list>
                    </t>

                </section>

                <section title="Inline dereferencing and fragments">
                    <t>
                        When using inline dereferencing, a resolution scope may lead to a URI which
                        has a non empty fragment part which is not a JSON Pointer, as in this
                        example:
                    </t>

                    <figure>
                        <artwork>
<![CDATA[
{
    "id": "http://some.site/schema#",
    "not": { "$ref": "#inner" },
    "definitions": {
        "schema1": {
            "id": "#inner",
            "type": "boolean"
        }
    }
}
]]>
                        </artwork>
                    </figure>

                    <t>
                        An implementation choosing to support inline dereferencing SHOULD be able to use this kind of reference.
                    </t>
                </section>

            </section>

            <section title="Interoperability considerations">
                <t>
                    Inline dereferencing can produce canonical URIs which differ from the canonical URI
                    of the root schema. Schema authors SHOULD ensure that implementations using
                    canonical dereferencing obtain the same content as implementations using inline
                    dereferencing.
                </t>

                <t>
                    Extended JSON References using fragments which are not JSON Pointers are not dereferenceable by implementations that do not support inline dereferencing.
                    Schema authors should bear in mind that this support is optional before using it.
                </t>
            </section>
        </section>

        <section title="Recommended correlation mechanisms for use with the HTTP protocol">

            <t>
                It is acknowledged by this specification that the majority of interactive JSON
                Schema processing will be over HTTP. This section therefore gives recommendations
                for materializing an instance/schema correlation using mechanisms currently
                available for this protocol. An instance is said to be described by one (or more)
                schema(s).
            </t>

            <section title='Correlation by means of the "Link" header'>
                <t>
                    It is RECOMMENDED that instances be associated with JSON Schemas using the link relation "describedby", as defined by <xref target="RFC5988">RFC 5988, section 5.3</xref>.
                </t>

                <t>
                    In HTTP, such links can be declared in a format-independent way using the Link header.  An example of such a header would be:
                </t>

                <figure>
                    <artwork>
<![CDATA[
Link: <http://example.com/my-hyper-schema#>; rel="describedBy"
]]>
                    </artwork>
                </figure>

            </section>

            <section title='Correlation by means of the "Content-Type" header'>
                <t>
                    Instances may also specify a schema using the "profile" MIME type parameter in the Content-Type header, when the MIME type MUST be "application/json", or any other subtype.
                    An instance associated with a JSON Schema using this "profile" parameter is considered to be described by that schema.
                </t>

                <t>
                    An example of such a header would be:
                </t>

                <figure>
                    <artwork>
<![CDATA[
Content-Type: application/my-media-type+json;
          profile="http://example.com/my-hyper-schema#"
]]>
                    </artwork>
                </figure>

            </section>

        </section>

        <section title="IANA Considerations">
            <t>
                The proposed MIME media type for JSON Schema is defined as follows:

                <list>
                    <t>type name: application;</t>
                    <t>subtype name: schema+json.</t>
                </list>
            </t>

        </section>
    </middle>

    <back>
        <!-- References Section -->
        <references title="Normative References">
            &RFC2119;
        </references>

        <references title="Informative References">
            &RFC2616;
            &RFC3986;
            &RFC4627;
            &RFC5988;
           <reference anchor="json-schema-validation"
                target="http://tools.ietf.org/html/draft-fge-json-schema-validation-01">
                <front>
                    <title>JSON Schema: interactive and non interactive validation</title>
                    <author initials="F." surname="Galiegue">
                        <organization/>
                    </author>
                    <author initials="G." surname="Court">
                        <organization/>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2013" month="January"/>
                </front>
            </reference>
           <reference anchor="json-hyper-schema"
                target="http://tools.ietf.org/html/draft-luff-json-hyper-schema-01">
                <front>
                    <title>JSON Hyper-Schema: Hypertext definitions for JSON Schema</title>
                    <author initials="G." surname="Luff">
                        <organization/>
                    </author>
                    <author initials="G." surname="Court">
                        <organization/>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2013" month="January"/>
                </front>
            </reference>
            <reference anchor="json-reference"
                target="http://tools.ietf.org/html/draft-pbryan-zyp-json-ref-03">
                <front>
                    <title>JSON Reference (work in progress)</title>
                    <author initials="P." surname="Bryan">
                        <organization>ForgeRock US, Inc.</organization>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2012" month="September"/>
                </front>
            </reference>
            <reference anchor="json-pointer"
                target="http://tools.ietf.org/html/draft-ietf-appsawg-json-pointer-07">
                <front>
                    <title>JSON Pointer (work in progress)</title>
                    <author initials="P." surname="Bryan">
                        <organization>ForgeRock US, Inc.</organization>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2012" month="September"/>
                </front>
            </reference>
            <reference anchor="json-schema-03"
                target="http://tools.ietf.org/html/draft-zyp-json-schema-03">
                <front>
                    <title>JSON Schema, draft 3</title>
                    <author initials="G." surname="Court">
                        <organization/>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2012" month="September"/>
                </front>
            </reference>
            <reference anchor="json-schema-04"
                target="http://tools.ietf.org/html/draft-zyp-json-schema-04">
                <front>
                    <title>JSON Schema, draft 3</title>
                    <author initials="G." surname="Court">
                        <organization/>
                    </author>
                    <author initials="K." surname="Zyp">
                        <organization>SitePen (USA)</organization>
                    </author>
                    <date year="2013" month="January"/>
                </front>
            </reference>
        </references>

        <section title="ChangeLog">
            <t>
                <list style="hanging">
                    <t hangText="draft-00">
                        <list style="symbols">
                            <t>Initial draft.</t>
                            <t>Salvaged from draft v3.</t>
                            <t>Mandate the use of JSON Reference, JSON Pointer.</t>
                            <t>Define the role of "id". Define URI resolution scope.</t>
                            <t>Add interoperability considerations.</t>
                        </list>
                    </t>
                </list>
            </t>
        </section>
    </back>
</rfc>