xml:id Version 1.0 -- Review Version

1 Introduction

[XML 1.0] and [XML 1.1] provide a mechanism for annotating elements with unique identifiers. This mechanism consists of declaring the type of an attribute as "ID", after which the parser will validate that

the ID value matches the allowed lexical form,
the value is unique within the XML document, and that
each element has at most one single unique identifier

Declarations in either the internal or external DTD subset of an XML document can declare attributes to be of type ID. However, some specifications, However, , forbid an internal subset, and processing the external DTD subset is optional for conformant XML processors; and some specifications, notably [SOAP], forbid DTDsprocessors, completely, leaving no guarantee that all consumers of the XML XML document will be able to successfully recognize the identifiers.

Identifiers can be declared through external mechanisms as well. Of particular interest is [XML Schemas] which provides a type "xs:ID" with the same uniqueness and validity constraints as XML. However, there are no guarantees that consumers will have the "correct" schema available, nor that they will process it if they do.

A mechanism allowing unique element identifiers to be recognized by all conformant XML processors, whether they validate or not, is desirable in making XML sub-resource linking robust. This specification allows authors to identify elements with IDs that can be recognized by any processor without regard to how, or if, any internal or external declarations are available.

An additional problem is that DTD-based and XML Schema-based identifiers are exposed through different conceptual mechanisms - the attribute type infoset property, and the type definition family of properties respectively. A uniform mechanism for recognizing identifiers is desirable.

This specification provides such a mechanism: it describes the semantics of xml:id attributes. This specification has been designed to be a separate layer in processing and to be compatible with existing validation technologies. Implementors are encouraged to support xml:id processing and to make ID type assignment the default behavior of their processors.

It has been a guiding principle in the design of this specification that the result of xml:id processing should be the same as if an appropriate declaration has been seen and used by the processor.

2 Terminology

[Definition: The key words must, must not, required, shall, shall not, should, should not, recommended, may, and optional in this specification are to be interpreted as described in [IETF RFC 2119].]

[Definition: An xml:id processor is a software module that works in conjunction with an XML processor to provide access to the IDs in an XML document.]

[Definition: An xml:id error is a non-fatal error that occurs when an xml:id processor finds that a document has violated the constraints of this specification.]

Validation is the process of comparing an XML document (or part of an XML document) against a grammar or set of rules to determine if the actual structure of the document satisfies the constraints of the grammar or the rules. Some validation technologies also perform type assignment, determining not only if the document satisfies the specified constraints but also determining, for example, which (elements and/or) attributes are of type “ID”.

Although often performed together, validation and type assignment are not the same process. A non-validating XML 1.0 processor, for example, can perform type assignment using only declarations from the internal subset, without ever having any information about the structural validity of the document.

[Definition: The process of ID type assignment causes an xml:id attribute value to be an ID.] This is often achieved by making the type of the attribute to be "ID" in the infoset or post-schema-validation infoset (PSVI), but that is not the only possible mechanism.

Note:

Application-level processing of IDs, including which elements can actually be addressed by which ID values, is beyond the scope of this specification.

3 Syntax

Per [Namespaces in XML] (and [Namespaces in XML 1.1]), prefixes beginning “xml” are reserved for use by XML and XML-related specifications. This specification licenses the use of the attribute “xml:id” for use as a common syntax for identifiers in XML with the semantics specified herein.

Authors of XML documents are encouraged to name their ID attributes "xml:id" to increase the interoperability of these identifiers on the Web.

In namespace-aware XML processors, the "xml" prefix is bound to the namespace name http://www.w3.org/XML/1998/namespace as described in Namespaces in XML [Namespaces in XML] (and [Namespaces in XML 1.1]). Note that xml:id can be still used by non-namespace-aware XML processors.

4 Processing `xml:id` Attributes

Each xml:id attribute is processed in the following way:

The attribute's value is normalized according to the rules for attribute-value normalization on attributes of type ID. For more details, see E Attribute Value Normalization on IDs.
The infoset normalized value property is updated with the normalized value.
ID type assignment is performed with the normalized value.

An xml:id processor must assure that the following constraints hold for all xml:id attributes:

The normalized value of the attribute is an NCName according to the Namespaces in XML Recommendation which has the same version as the document in which this attribute occurs (NCName for XML 1.0, or NCName for XML 1.1).
The declared type of the attribute, if it has one, is “ID”. All declarations for xml:id attributes must specify “ID” as the type of the attribute.

An xml:id processor should assure that the following constraints hold:

The values of all xml:id attributes and all attributes of type “ID” within a document are unique.

An xml:id error occurs for any xml:id attribute that does not satisfy the constraints.

The xml:id processor performs ID type assignment on all xml:id attributes, even those that do not satisfy the constraints.

An xml:id processor should update the infoset references infoset property, as described in Section 2.3 of [XML Information Set], and update any implementation-specific structures used for cross-referencing to reflect the results of ID assignment.

Many validation technologies impose the constraint that an XML element can have at most one attribute of type ID. That constraint is not imposed by xml:id processing.

This specification defines xml:id processing, but it is up to the application to determine when such processing occurs. Users of applications that provide facilities for modifying XML documents may reasonably expect xml:id processing to occur whenever a change is made to an ID value.

5 Informing the Application

WhenID type assignment may be performed when xml:id attributes are processed. If ID type assignment occurs, then the xml:id processor must report the assigned xml:id attributes to the application. How this is reported is implementation dependent.

For applications that operate conceptually on the Infoset, an xml:id processor can use the attribute type Infoset property:
The xml:id processor may report the results of ID type assignment in a DTD compatible manner by setting the attribute type infoset property of the attribute to ID.
For applications that operate conceptually on the PSVI, an xml:id processor can use the type definition family of PSVI properties:
The xml:id processor may report the results of ID type assignment in an XML Schema compatible manner by setting the PSVI type definition property of the attribute to xs:ID.
For applications that operate on data models defined in other ways, the mechanisms are implementation dependent:
The xml:id processor may report the results of ID type assignment in some other way.

The key requirement is that the application be made aware of the results of ID type assignment.

6 Errors

A violation of the constraints in this specification results in an xml:id error. Such errors are not fatal, but should be reported by the xml:id processor. In the interest of interoperability, it is strongly recommended that xml:id errors not be silently ignored.

7 Conformance

7.1 Conformance to xml:id

Conformance to xml:id for applications that rely on XML processors using validation technologies consists in the use of the xml:id construct as explained in 4 Processing xml:id Attributes and by conformance to both the constraints of this specification and the rules of the validation technology.

Conformance to xml:id for applications that rely on non-validating XML processors is defined by the recognition of xml:id attributes as explained in 4 Processing xml:id Attributes and by conformance to the constraints of this specification.

Conformance to constraints that “must” be assured is mandatory. It is recommended that applications assure the other constraints as well. This specification defines no simply optional constraints.

A document is conformant to this specification if it generates no xml:id errors.

7.2 XML Information Set Conformance

This specification conforms to the [XML Information Set]. The following information items must be present in the input infosets to enable correct processing:

Element Information Items with attributes property.
Attribute Information Items with namespace name, local name and normalized value properties.

In addition, the following properties might be present in the output infoset:

attribute type properties on Attribute Information Items.

8 Extensibility

This specification is not extensible. There are no provisions for application designers to alter the name of the xml:id attribute, the set of attribute values that are considered IDs, the location(s) where they can occur, or make any other extensions.

A References

IETF RFC 2119: RFC 2119: Key words for use in RFCs to Indicate Requirement Levels. Internet Engineering Task Force, 1997. (See http://www.ietf.org/rfc/rfc2119.txt.)
XML 1.0: Tim Bray, Jean Paoli, C.M. Sperberg-McQueen, et. al., editors. Extensible Markup Language (XML) 1.0 (Third Edition). World Wide Web Consortium, 2004. (See http://www.w3.org/TR/REC-xml/.)
XML 1.1: Tim Bray, Jean Paoli, C.M. Sperberg-McQueen, et. al., editors. Extensible Markup Language (XML) 1.1. World Wide Web Consortium, 2004. (See http://www.w3.org/TR/xml11/.)
XML Information Set: John Cowan and Richard Tobin, editors. XML Information Set (Second Edition). World Wide Web Consortium, 2004. (See http://www.w3.org/TR/xml-infoset/.)
Namespaces in XML: Tim Bray, Dave Hollander, and Andrew Layman, editors. Namespaces in XML. World Wide Web Consortium, 1999. (See http://www.w3.org/TR/REC-xml-names/.)
Namespaces in XML 1.1: Tim Bray, Dave Hollander, Andrew Layman, and Richard Tobin, editors. Namespaces in XML 1.1. World Wide Web Consortium, 2004. (See http://www.w3.org/TR/xml-names11/.)

B References (Non-Normative)

XML Schemas: Henry S. Thompson, David Beech, Murray Maloney, Noah Mendelsohn, editors. XML Schema Part 1: Structures. World Wide Web Consortium, 2001. (See http://www.w3.org/TR/xmlschema-1/.)
SOAP: Martin Gudgin, Marc Hadley, Noah Mendelsohn, et. al., editors. SOAP Version 1.2 Part 1: Messaging Framework. World Wide Web Consortium, 2003. (See http://www.w3.org/TR/soap12-part1/.)

C Impact on Canonicalization (Non-Normative)

The Canonical XML Version 1.0 specification describes a process whereby attributes in the xml: namespace are inherited in a canonicalized document. While this produces a reasonable result with xml:lang or xml:space attributes, processing xml:id attributes in this way is likely to produce documents that contain xml:id errors, specifically xml:id attribute values that are not unique.

This is an apparent flaw in the design of Canonical XML. The Exclusive XML Canonicalization Version 1.0 specification does not have this feature and may be more appropriate for documents containing IDs.

D Validation Technologies (Non-Normative)

This appendix describes how xml:id processing interacts with selected validation technologies.

D.1 With DTD Validation

DTD authors are encouraged to use xml:id attributes when providing identifiers for elements declared in their DTDS.

The following DTD fragment illustrates a sample declaration for the xml:id attribute:

<!ATTLIST someElement
    xml:id     ID          #IMPLIED
>

DTD authors are encouraged to declare attributes named xml:id with the type ID. A document that uses xml:id attributes that have a declared type other than ID will always generate xml:id errors.

Consumers of documents validated using properly declared xml:id attributes can recognize IDs through the attribute type property.

D.2 With XML Schema Validation

XML Schema authors are encouraged to use xml:id attributes when providing identifiers for elements declared in their schemas. Note that this can most easily be accomplished by importing the schema for the XML namespace and using the attribute declaration it contains.

The following XML Schema fragment for the XML namespace illustrates a sample declaration for the xml:id attribute:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"
           targetNamespace="http://www.w3.org/XML/1998/namespace">

    <xs:attribute name="id" type="xs:ID"/>

</xs:schema>

XML Schema authors are encouraged to declare attributes named xml:id with the type xs:ID. A document that uses xml:id attributes that have a declared type other than xs:ID will always generate xml:id errors.

Consumers of documents validating the xml:id attributes against an appropriate schema for the XML namespace can recognize IDs through the type definition family of PSVI properties.

Applications can recognize xml:id attributes as IDs by conceptually using a Minimally Conforming Schema Processor and the schema above.

Note that the effects of a Minimally Conforming Schema Processor, processing the above schema, are approximated by simply looking for attributes named xml:id, ensuring the value of such attributes has the correct lexical form (NCName), and the value is unique within the document.

D.3 With RELAX NG Validation

RELAX NG Grammar authors are encouraged to use xml:id attributes when providing identifiers for elements declared in their schemas.

The following RELAX NG fragment illustrates a sample declaration for the xml:id attribute:

<optional xmlns="http://relaxng.org/ns/structure/1.0"
              datatypeLibrary="http://www.w3.org/2001/XMLSchema-datatypes">
  <attribute name="xml:id">
    <data type="ID"/>
  </attribute>
</optional>

RELAX NG Grammar authors are encouraged to declare attributes named xml:id with the type xs:ID. A document that uses xml:id attributes that have a declared type other than xs:ID will always generate xml:id errors.

E Attribute Value Normalization on IDs (Non-Normative)

[XML 1.0]Parsers requires parsers to normalize all attribute values. Normalization expands character references, expands entity references, and cleans up line end characters. Attributes of type ID are subject to additional normalization rules: removing leading and trailing space characters and replacing sequences of spaces with a single space.

The xml:id processor has to assure that both steps of normalization are performed all attributes named xml:id. In particular, the parser may not have performed the additional normalization steps required for attributes of type ID because the attribute may not be declared or may be declared as an ID.

Consider the following document:

<!DOCTYPE doc [
<!ATTLIST doc xml:id ID #IMPLIED>
]>
<doc xml:id="  one
">
<para xml:id="  two
"></para>
</doc>

The initial value of xml:id on doc will be “one” because the parser knew that it was an ID. The initial value on para will be “ two ”. Because the parser didn't know it was an ID, it will not have performed the additional normalizations required.

After xml:id processing, the value of the xml:id attributes on doc and para will be “one” and “two”, respectively. These properly normalized values will be stored in the normalized value property in the infoset. Performing xml:id processing changes the infoset if there are incompletely normalized xml:id attributes.