1.0: Overview

1.1: Purpose and Scope

In order for electronic-book technology to achieve widespread success in the marketplace, Reading Systems need to have convenient access to a large number and variety of titles. ~~The~~ Another related specification, the Open Publication Structure (OPS) Specification, describes a standard for representing the content of electronic publications and is meant to reduce barriers to the proliferation of content. Specifically, the specification is intended to:

Give publication tool providers and content providers (e.g. publishers, authors, and others who have content to be displayed) minimal and common guidelines that ensure fidelity, accuracy, accessibility, and adequate presentation of electronic content over various Reading Systems; and
Build on established content format standards; and
Define a standard means of content description in order for electronic books to move smoothly through the distribution chain.

This document, the Open Packaging Format (OPF) Specification, defines the mechanism by which the various components of an OPS publication are tied together and provides additional structure and semantics to the electronic publication. Specifically, OPF:

Describes and references all components of the electronic publication (e.g. markup files, images, navigation structures).
Provides publication-level metadata.
Specifies the linear reading-order of the publication.
Provides fallback information to use when unsupported extensions to OPS are employed.
Provides a mechanism to specify a declarative table of contents (the NCX).

This OPF specification is separate from the OPS markup specification to modularize the described packaging methodology and the described content. This will help facilitate the use of this packaging technology by other standards bodies (e.g. ~~Daisy~~DAISY) in non-OPS contexts.

A third specification, the OEBPS Container Format (OCF) Specification, defines the standard mechanism by which all components of an electronic publication can be packaged together into a single file for transmission, delivery and archival.

1.2: Definitions

Content Provider

A publisher, author, or other information provider who provides a publication to one or more Reading Systems in the form described in this specification and the OPS specification.

Deprecated

A feature that is permitted, but not recommended, by this specification. Such features might be removed in future revisions. Conformant Reading Systems must support deprecated features.

Extended Module

A module of a modularized XML vocabulary (i.e. a set of named modules is defined in its specification) that is not mandated to be supported by its specification (e.g. the XHTML ruby or forms modules in the OPS context).

Inline XML Island

An inline XML Island is an XML document fragment using a non-Preferred Vocabulary or using an Extended Module that exists within an XHTML Preferred Vocabulary document within an OPS Publication.

NCX

A declarative table of contents (the Navigation Center eXtended or NCX).

OCF

The OEBPS Container Format defines a mechanism by which all components of an OPS Publication can be combined into a single file-system entity.

OEBPS

The Open ~~Publication~~ Publication Structure. Previous versions of this specification (OPF) and its related specification, OPS, were unified into the single OEBPS specification. For this version, OEBPS was broken into separate OPF and OPS specifications to aid modular adoption of the specifications. OEBPS 1.2 was the highest version of the previous unified specification.

OPF

The Open Packaging Format — this standard — defines the mechanism by which all components of a published work conforming to the OPS standard including metadata, reading order and navigational information are packaged into an OPS Publication.

OPF Package Document

An XML Document that describes an OPS Publication and references all files used by the OPS Publication that are not part of the OPF Package Document itself. It identifies all other files in the Publication and provides descriptive information about them. The OPF Package Document is defined by this specification and is valid to the OPF Package Schema defined herein.

The "root file" of the OPF Package Document should use the .opf extension. This XML file may refer to other XML files via XML’s general entity mechanism, but those files must not use the .opf file extension. This construction could be used to simplify the creation of OPF Package Documents for very large Publications. However, the most common case is for the OPF Package Document to be a single XML file using the .opf extension.

OPS

The Open Publication Structure — the sister-standard to this standard — defines the markup necessary to construct OPS Content Documents.

OPS Content Document

An XHTML, DTBook, or out-of-line XML document that conforms to the OPS specification that can legally appear in the OPF Package Document spine.

OPS Core Media Type

A MIME media type, defined in the OPS Specification, that all Reading Systems must support.

OPS Publication

A collection of OPS Content Documents, an OPF Package Document, and other files, typically in a variety of media types, including structured text and graphics, that constitute a cohesive unit for publication.

Out-of-Line XML Island

An Out-Of-Line XML Island is an XML document that exists within an OPS Publication and is either not authored using a Preferred Vocabulary or is authored using a Preferred Vocabulary but uses Extended Modules. It is an entirely separate, complete, and valid XML document.

Preferred Vocabulary

XML consisting only of OPS-supported XHTML markup and/or DTBook markup.

Reader

A person who reads a publication.

Reading System

A combination of hardware and/or software that accepts OPS Publications (likely packaged in an OCF Container) and makes them available to consumers of the content. Great variety is possible in the architecture of Reading Systems. A Reading System may be implemented entirely on one device, or it may be split among several computers. In particular, a Reading Device that is a component of a Reading System need not directly accept OPS Publications, but all Reading Systems must do so. Reading Systems may include additional processing functions, such as compression, indexing, encryption, rights management, and distribution.

XML Document

An XML Document is a complete and valid XML document as defined in XML 1.1 (http://www.w3.org/TR/xml11/).

XML Document Fragment

Referred to as either a document fragment or as an XML Document Fragment, as defined in Document Object Model Level 1 (http://www.w3.org/TR/REC-DOM-Level-1/) but with the additional requirement that they be well-formed.

XML Island

An Inline XML Island or an Out-of-Line XML island.

XML Namespaces

Referred to as XML namespaces, or just namespaces, these must conform to XML Namespaces (http://www.w3.org/TR/xml-names11/).

1.3: Relationship to Other Specifications

This specification combines subsets and applications of other specifications. Together, these facilitate the construction, organization, presentation, and unambiguous interchange of electronic documents:

Extensible Markup Language (XML) 1.1 (Second Edition) specification (http://www.w3.org/TR/xml11/); and
Namespaces in XML 1.0 (Second Edition) (http://www.w3.org/TR/xml-names11/); and
The OPS Specification (http://www.idpf.org/ops/ops2.0/download/); and
XHTML™ 1.1 - Module-based XHTML - Second Edition specification (http://www.w3.org/TR/xhtml11/); and
Specifications for the Digital Talking Book (DTB) (http://www.niso.org/standards/resources/Z39-86-2005.html); and
Dublin Core Metadata Element Set, Version 1.1 specification (http://dublincore.org/documents/2004/12/20/dces/) and the MARC relator code list (http://www.loc.gov/marc/relators/); and
Unicode Standard, Version 4.0. Reading, Mass.: Addison-Wesley, 2003, as updated from time to time by the publication of new versions. (See http://www.unicode.org/unicode/standard/versions for the latest version and additional information on versions of the standard and of the Unicode Character Database); and
Particular MIME media types (http://www.ietf.org/rfc/rfc4288.txt and http://www.iana.org/assignments/media-types/index.html); and
Web Content Accessibility Guidelines 1.0 (http://www.w3.org/TR/WCAG10/); and
RFC 2119: Key words for use in RFCs to Indicate Requirement Levels. (http://www.ietf.org/rfc/rfc2119.txt).

1.3.1: Relationship to XML

~~OPS~~OPF is based on XML because of XML’s generality and simplicity, and because XML documents are likely to adapt well to future technologies and uses. XML also provides well-defined rules for the syntax of documents, which decreases the cost to implementers and reduces incompatibility across systems. Further, XML is extensible: it is not tied to any particular type of document or set of element types, it supports internationalization, and it encourages document markup that can represent a document’s internal parts more directly, making them amenable to automated formatting and other types of computer processing.

Reading Systems must be XML processors as defined in XML 1.1. All OPF Package Documents must be valid XML documents according to the OPF Package Schema.

1.3.2: Relationship to XML Namespaces

Reading Systems must process XML namespaces according to the XML Namespaces Recommendation at http://www.w3.org/TR/xml-names11/.

Namespace prefixes distinguish identical names that are drawn from different XML vocabularies. An XML namespace declaration in an XML document associates a namespace prefix with a unique URI. The prefix can then be employed on element or attribute names in the document. Alternatively, a namespace declaration in an XML document may identify a URI as the default namespace, applicable to elements lacking a namespace prefix. The XML namespace prefix is separated from the suffix element or attribute name by a colon.

The namespace for the OPF Package Document is http://www.idpf.org/2007/opf, and must be declared at the root of all OPF Package Documents. In addition, to be processed as an OPF 2.0 package, a version attribute with a value of 2.0 must be specified on ~~all~~the package elements. A package element that omits the version attribute must be processed as an OEBPS 1.2 package.

Example:

 <package version="2.0" xmlns="http://www.idpf.org/2007/opf">
                ...
 </package>

1.3.3: Relationship to Dublin Core

~~The~~ Dublin Core metadata is designed to minimize the cataloging burden on authors and publishers, while providing enough metadata to be useful. This specification supports the set of Dublin Core 1.1 metadata elements (http://dublincore.org/documents/2004/12/20/dces/), supplemented with a small set of additional attributes addressing areas where more specific information ~~could be~~is useful. For example, the OPF role attribute added to the Dublin Core creator and contributor elements allows for much more detailed specification of contributors to a publication, including their roles expressed via relator codes.

Content providers must include a minimum set of metadata elements, defined in Section 2.2, and should incorporate additional metadata to enable readers to discover publications of interest.

1.3.4: Relationship to Unicode

OPF Package Documents may use the entire Unicode character set, in UTF-8 or UTF-16 encodings, as defined by Unicode (see http://www.unicode.org/unicode/standard/versions). The use of Unicode facilitates internationalization and multilingual documents. However, Reading Systems are not required to provide glyphs for all Unicode characters.

Reading Systems must parse all UTF-8 and UTF-16 characters properly (as required by XML). Reading Systems may decline to display some characters, but must be capable of signaling in some fashion that undisplayable characters are present. Reading Systems must not display Unicode characters merely as if they were 8-bit characters. For example, the biohazard symbol (0x2623) need not be supported by including the correct glyph, but must not be parsed or displayed as if its component bytes were the two characters "&#" (0x0026 0x0023).

To aid Reading Systems in implementing consistent searching and sorting behavior it is required that Unicode Normalization Form C (NFC) be used (See http://www.w3.org/TR/charmod-norm/).

1.4: Conformance

The keywords "must", "must not", "required", "shall", "shall not", "should", "recommended", "may", and "optional" in this document must be interpreted as described in RFC 2119.

This section defines conformance for OPF Package Documents, and Reading Systems that process those Documents.

1.4.1: Package Conformance

This specification defines conformance for both individual OPF Package Documents and for a collection of files including exactly one OPF Package Document that collectively are referred to as an OPS Publication.

1.4.1.1: Package Conformance

Each conformant OPF Package Document must meet these necessary conditions:

it is a well-formed XML document (as defined in XML 1.1); and
it is encoded in UTF-8 or UTF-16; and
it is a valid XML document according to the OPF Package Schema defined in Appendix A; and
it may consist of one or more XML files, but only one may use the file extension .opf.

1.4.1.2: Publication Conformance

A collection of files is a conforming OPS Publication if ~~and only if~~:

it includes a single OPF Package Document that obeys the Package Conformance Requirements listed above; and
only one file in the Publication may use the file extension .opf. If such a file is present, it must be the "root file" of the OPF Package Document; and
the OPF Package Document includes one and only one manifest entry corresponding to each other file in the OPS Publication excluding the file or files that make up the OPF Package Document itself; and
the manifest entry for each file in the publication specifies a MIME media type for the file (see http://www.ietf.org/rfc/rfc2046.txt); and
each file whose manifest entry identifies it as being in one of the OPS Core Media Types conforms as defined for those MIME media types; and
each file listed in the spine of the OPF Package Document must conform to the OPS Content Document requirements defined in the OPS specification; and
~~if the publication contains one or more documents which are either DTBook or an Out-Of-Line XML Island,~~ an NCX must be included; and
the metadata element or deprecated dc-metadata element contains at least one identifier element, at least one title element, and at least one language element drawn from the Dublin Core tag set; and
the unique-identifier attribute of the package element is a correct XML IDREF to a Dublin Core identifier element; and
any extended values specified for the Dublin Core creator and contributor elements' OPF role attribute must be taken from the registered MARC Relator Code list or must begin with oth.; and
any extended values specified for the guide element's type attribute begin with other.; and
the version attribute of the package element is specified with a value of 2.0; and
the package element's namespace must be http://www.idpf.org/2007/opf.

This specification and the OPS specification impose additional conformance constraints on the Package Document and on OPS Content Documents.

1.4.2: Reading System Conformance

This specification defines conformance for a Reading System when presented with an OPS Publication. OPS Content documents have further conformance requirements that can be found in the OPS specification. A Reading System is conformant if and only if it processes documents as follows:

When presented with an OPF Package Document the Reading System must
1. process all elements and attributes as described in Section 2 of this specification; and
2. ignore all elements and attributes not described in Section 2 of this specification; and
3. verify the existence of the appropriate namespace specifications, as defined in the Relationship to XML Namespaces section above.
When providing navigation via the OPF spine, the Reading System must not render content that is not an OPS Content Document.
When presented with an OEBPS 1.2 Package, the Reading System must process it as a conformant OEBPS 1.2 Reading System would. Note that only the OEBPS 1.2 Package must be processed as an OEBPS 1.2 Reading System would, not the content documents referred to in the Package.
When presented with an OEBPS 1.2 Publication, a Reading System should process it as a conformant OEBPS 1.2 Reading System would. Such a Reading System can claim an optional level of Reading System conformance, "Backward Compatibility Conformance."

1.4.3: Compatibility of OPF Version 2.0

Version 2.0 of OPF is not meant to be a substantially "new" specification. However, version 2.0 does add one significant functional enhancement in addition to a number of other changes from OEBPS version 1.2. Specifically, the following are the most substantive additions:

XML 1.1 is incorporated.
XML namespace processing inis now required.
Support for ~~Daisy~~DAISY’s Navigation Center eXtended (NCX) has been added to enhance ease of navigation and accessibility of publications.
~~The leading character of~~ Dublin Core element names must now be lower case to conform to Dublin Core’s XML implementation recommendations.
The tours element has been deprecated.

While most changes from version OEBPS 1.2 to OPF 2.0 have been done via deprecation rather than removal of previous functionality, the OEBPS 1.2 Package is not a fully compatible subset of OPF 2.0 (e.g. new namespace processing requirements).

1.5: Accessibility

This specification incorporates features that ensure content can be made accessible to, and usable by, persons with reading disabilities. Existing accessibility features developed by the World Wide Web Consortium (W3C) are incorporated into this specification.

Publications should be authored in accordance with the W3C Web Content Accessibility Guidelines 1.0 (http://www.w3.org/TR/1999/WAI-WEBCONTENT-19990505/) or, if it is released while the Working Group is active, the Web Content Accessibility Guidelines 2.0 (the current draft is available at http://www.w3.org/TR/WCAG20/) to ensure that the broadest possible set of users will have access to books delivered in this format.

In addition, recommendations from the W3C HTML 4.0 Guidelines for Mobile Access (http://www.w3.org/TR/NOTE-html40-mobile/) and the W3C Web Accessibility Initiative's proposed User Agent Guidelines (http://www.w3.org/TR/WD-WAI-USERAGENT/) ought to be reviewed and applied by OPF implementers to ensure that Reading Systems will be in conformance with accessibility requirements.

2.0: The OPF Package Document

A publication conforming to this specification must include exactly one XML OPF Package Document, which specifies the OPS Content Documents, images, and other objects that make up the OPS Publication and how they relate to each other.

The OPF Package Document should be named ~~using~~ending in the extension ".opf", in order to make it readily identifiable within the group of files making up the publication. The OPS Package Document is of MIME media type application/oebps-package+xml. This specification does not define means for physically bundling files together to make one data transfer object (such as using zip or tar); the OEBPS Container Format (OCF) specifies this functionality.

This specification neither precludes nor requires the inclusion of the OPF Package Schema in a Publication.

The major parts of the OPF Package Document are:

Package Name

A unique identifier for the OPS Publication as a whole.

Metadata

Publication metadata (title, author, publisher, etc.).

Manifest

A list of files (documents, images, style sheets, etc.) that make up the publication. The manifest also includes fallback declarations for files of types not supported by this specification.

Spine

An arrangement of documents providing a linear reading order.

Tours (Deprecated)

A set of alternate reading sequences through the publication, such as selective views for various reading purposes, reader expertise levels, etc.

Guide

A set of references to fundamental structural features of the publication, such as table of contents, foreword, bibliography, etc.

An OPF Package Document must be a valid XML document conforming to the OPF Package schema (Appendix A). An informal outline of the package is as follows:

<?xml version="1.0"?>
<package version="2.0" xmlns="http://www.idpf.org/2007/opf" unique-identifier="BookId">
        metadata
        manifest
        spine
        guide
</package>

The following sections describe the parts of the OPF Package Document.

2.1: Package Identity

The package element is the root element in an OPF Package Document; all other elements are nested within it.

The package element must specify a value for its unique-identifier attribute. The unique-identifier attribute's value specifies which Dublin Core identifier element, described in Section 2.2.10, provides the package's preferred, or primary, identifier. The OPF Package Document's author is responsible for choosing a primary identifier that is unique to one and only one particular package (i.e., the set of files referenced from the package document's manifest).

Notwithstanding the requirement for uniqueness, Reading Systems must not fail catastrophically if they encounter two distinct packages with the same purportedly unique primary identifier.

2.2: Publication Metadata

The required metadata element is used to provide information about the publication as a whole. It may contain Dublin Core metadata elements directly or within a (now deprecated) dc-metadata sub-element. Supplemental metadata can also be specified directly or within a (now deprecated) x-metadata sub-element.

Reading Systems must allow the specification of the deprecated dc-metadata and x-metadata elements. Newly created OPS 2.0 packages should not create dc-metadata or x-metadata elements. If the dc-metadata element is used, all dc elements must go in dc-metadata and all other metadata elements, if any, must go into x-metadata. If the dc-metadata element is not used, all metadata elements must go directly in the metadata element.

The required metadata or dc-metadata (deprecated) elements contain specific publication-level metadata as defined by the Dublin Core Metadata Initiative (http://www.dublincore.org/). The descriptions below are included for convenience, and the Dublin Core's own definitions take precedence (see http://dublincore.org/documents/2004/12/20/dces/).

One or more optional instances of a meta element, analogous to the XHTML 1.1 meta element but applicable to the publication as a whole, may be placed within the metadata element or within the deprecated x-metadata element. This allows content providers to express arbitrary metadata beyond the data described by the Dublin Core specification. Individual OPS Content Documents may include the meta element directly (as in XHTML 1.1) for document-specific metadata. This specification uses the OPF Package Document alone as the basis for expressing publication-level Dublin Core metadata.

For example:

<metadata xmlns:dc="http://purl.org/dc/elements/1.1/"
          xmlns:opf="http://www.idpf.org/2007/opf">
   <dc:title>Tale of Two Cities</dc:title>
   <dc:creator opf:role="aut">Charles Dickens</dc:creator>
   ...
   <meta name="price" content="USD 19.99" />
</metadata>

The XML namespace mechanism (see http://www.w3.org/TR/REC-xml-names11/) is used to identify the elements used for Dublin Core metadata without conflict. The metadata or dc-metadata (deprecated) elements may contain any number of instances of any Dublin Core elements. Dublin Core metadata elements may occur in any order; in fact, multiple instances of the same element type (e.g. multiple Dublin Core creator elements) can be interspersed with other metadata elements without change of meaning.

Each Dublin Core field is represented by an element whose content is the field's value. At least one of each of Dublin Core title, identifier and language must be included in the metadata element. Dublin Core elements, like any other elements in the OPF Package Document, may have an id attribute specified. At least one Dublin Core identifier, which is referenced from the package unique-identifier attribute, must have an id specified.

Because the Dublin Core metadata fields for creator and contributor do not distinguish roles of specific contributors (such as author, editor, and illustrator), this specification adds an optional role attribute for this purpose. See Section 2.2.6 for a discussion of role.

To facilitate machine processing of Dublin Core creator and contributor fields, this specification adds the optional file-as attribute for those elements. This attribute is used to specify a normalized form of the contents. See Section 2.2.2 for a discussion of file-as.

This specification also adds a scheme attribute to the Dublin Core identifier element to provide a structural mechanism to separate an identifier value from the system or authority that generated or defined that identifier value. See Section 2.2.10 for a discussion of scheme.

This specification also adds an event attribute to the Dublin Core date element to enable content providers to distinguish various publication specific dates (for example, creation, publication, modification). See Section 2.2.7 for a discussion of event. For example:

<package version="2.0" xmlns="http://www.idpf.org/2007/opf"
         unique-identifier="BookId">
        <metadata xmlns:dc="http://purl.org/dc/elements/1.1/"
                xmlns:opf="http://www.idpf.org/2007/opf">
           <dc:title>Alice in Wonderland</dc:title>
           <dc:language>en</dc:language>
           <dc:identifier id="BookId" opf:scheme="ISBN">
            123456789X
         </dc:identifier>
           <dc:creator opf:role="aut">Lewis Carroll</dc:creator>
</metadata>
         ...
</package>

There are no attributes for the elements within metadata defined by Dublin Core — only the elements' contents are so defined. In the above example, the specification of the OPF namespace on the metadata element is present to resolve the scheme and role attributes used in the identifier and creator elements, respectively.

For compatibility with Guidelines for implementing Dublin Core in XML (http://dublincore.org/documents/dc-xml-guidelines/) this specification allows xsi:type attribute for metadata items that can be given using some sort of encoding scheme and xml:lang attribute when an item can be given using human-readable text. Elements that allow xsi:type attribute are identifier, language, date, format and type. Elements that allow xml:lang attribute are: title, contributor, coverage, creator, description, publisher, relation, rights, source, and subject. This specification does not impose any specific rules for these attributes (with possible exception of heuristics that use xml:lang described below).

The following subsections describe the individual Dublin Core metadata elements.

2.2.1: <title> </title>

The title of the publication. An OPF Package Document must include at least one instance of this element type, however multiple instances are permitted. Any Reading System that displays title metadata should display the content of the most appropriate title element(s). Determination of the most appropriate titles is not defined by this specification, but may include the available fonts, an examination of xml:lang attributes, or other heuristics. In the absence of such an algorithm, conforming Reading Systems should consider either the first title element or all the title elements as the most appropriate.

2.2.2: <creator> </creator>

A primary creator or author of the publication. Additional contributors whose contributions are secondary to those listed in creator elements should be named in contributor elements.

Publications with multiple co-authors should provide multiple creator elements, each containing one author. The order of creator elements is presumed to define the order in which the creators' names should be presented by the Reading System.

This specification recommends that the content of the creator elements hold the text for a single name as it would be presented to the Reader.

This specification adds to the creator element two optional attributes: role and file-as. The set of values for role are identical to those defined in Section 2.2.6 for the contributor element. The file-as attribute should be used to specify a normalized form of the contents, suitable for machine processing. For example, one might find

<dc:creator opf:file-as="King, Martin Luther Jr." opf:role="aut">
        Rev. Dr. Martin Luther King Jr.
</dc:creator>

Any Reading System that displays creator metadata should display the content of the most appropriate creator element(s). Determination of the most appropriate creators is not defined by this specification, but may include the available fonts, an examination of xml:lang attributes, or other heuristics. In the absence of such an algorithm, conforming Reading Systems should display the contents of all creator elements, in the order provided, with appropriate separating spacing and/or punctuation.

2.2.3: <subject> </subject>

Multiple instances of the subject element are supported, each including an arbitrary phrase or keyword. This specification makes no attempt to standardize subject naming schemes, such as the Library of Congress Subject Heading System.

2.2.4: <description> </description>

Description of the publication's content.

2.2.5: <publisher> </publisher>

The publisher as defined by the Dublin Core Metadata Element Set (http://dublincore.org/documents/2004/12/20/dces/).

2.2.6: <contributor> </contributor>

A party whose contribution to the publication is secondary to those named in creator elements.

Other than significance of contribution, the semantics of this element are identical to those of creator. Reading Systems are free to choose to display creator information without accompanying contributor information.

This specification adds to the contributor element two optional attributes: role and file-as. The file-as attribute is defined as for creator, and is documented in Section 2.2.2.

The normative list of values used for the role attribute is defined by the MARC relator code list (http://www.loc.gov/marc/relators/). When roles are specified, the 3-character registered MARC values must be used when applicable. Although that list is extensive, other values may be added if a desired role is not covered by those predefined values. Such values must begin with oth., and shall be considered subdivisions of the other relator code. Like other constructs in this specification, these values are case-sensitive and must be coded entirely in lower-case.

For convenience, some relator code values are listed here as examples. Consult the MARC code list cited above for the complete list.

Adapter [adp]	Use for a person who 1) reworks a musical composition, usually for a different medium, or 2) rewrites novels or stories for motion pictures or other audiovisual medium.
Annotator [ann]	Use for a person who writes manuscript annotations on a printed item.
Arranger [arr]	Use for a person who transcribes a musical composition, usually for a different medium from that of the original; in an arrangement the musical substance remains essentially unchanged.
Artist [art]	Use for a person (e.g., a painter) who conceives, and perhaps also implements, an original graphic design or work of art, if specific codes (e.g., [egr], [etr]) are not desired. For book illustrators, prefer Illustrator [ill].
Associated name [asn]	Use as a general relator for a name associated with or found in an item or collection, or which cannot be determined to be that of a Former owner [fmo] or other designated relator indicative of provenance.
Author [aut]	Use for a person or corporate body chiefly responsible for the intellectual or artistic content of a work. This term may also be used when more than one person or body bears such responsibility.
Author in quotations or text extracts [aqt]	Use for a person whose work is largely quoted or extracted in a works to which he or she did not contribute directly. Such quotations are found particularly in exhibition catalogs, collections of photographs, etc.
Author of afterword, colophon, etc. [aft]	Use for a person or corporate body responsible for an afterword, postface, colophon, etc. but who is not the chief author of a work.
Author of introduction, etc. [aui]	Use for a person or corporate body responsible for an introduction, preface, foreword, or other critical matter, but who is not the chief author.
Bibliographic antecedent [ant]	Use for the author responsible for a work upon which the work represented by the catalog record is based. This can be appropriate for adaptations, sequels, continuations, indexes, etc.
Book producer [bkp]	Use for the person or firm responsible for the production of books and other print media, if specific codes (e.g., [bkd], [egr], [tyd], [prt]) are not desired.
Collaborator [clb]	Use for a person or corporate body that takes a limited part in the elaboration of a work of another author or that brings complements (e.g., appendices, notes) to the work of another author.
Commentator [cmm]	Use for a person who provides interpretation, analysis, or a discussion of the subject matter on a recording, motion picture, or other audiovisual medium. Compiler [com] Use for a person who produces a work or publication by selecting and putting together material from the works of various persons or bodies.
Designer [dsr]	Use for a person or organization responsible for design if specific codes (e.g., [bkd], [tyd]) are not desired.
Editor [edt]	Use for a person who prepares for publication a work not primarily his/her own, such as by elucidating text, adding introductory or other critical matter, or technically directing an editorial staff.
Illustrator [ill]	Use for the person who conceives, and perhaps also implements, a design or illustration, usually to accompany a written text.
Lyricist [lyr]	Use for the writer of the text of a song.
Metadata contact [mdc]	Use for the person or organization primarily responsible for compiling and maintaining the original description of a metadata set (e.g., geospatial metadata set).
Musician [mus]	Use for the person who performs music or contributes to the musical content of a work when it is not possible or desirable to identify the function more precisely.
Narrator [nrt]	Use for the speaker who relates the particulars of an act, occurrence, or course of events.
Other [oth]	Use for relator codes from other lists which have no equivalent in the MARC list or for terms which have not been assigned a code.
Photographer [pht]	Use for the person or organization responsible for taking photographs, whether they are used in their original form or as reproductions.
Printer [prt]	Use for the person or organization who prints texts, whether from type or plates.
Redactor [red]	Use for a person who writes or develops the framework for an item without being intellectually responsible for its content.
Reviewer [rev]	Use for a person or corporate body responsible for the review of book, motion picture, performance, etc.
Sponsor [spn]	Use for the person or agency that issued a contract, or under whose auspices a work has been written, printed, published, etc.
Thesis advisor [ths]	Use for the person under whose supervision a degree candidate develops and presents a thesis, memoir, or text of a dissertation.
Transcriber [trc]	Use for a person who prepares a handwritten or typewritten copy from original material, including from dictated or orally recorded material.
Translator [trl]	Use for a person who renders a text from one language into another, or from an older form of a language into the modern form.

2.2.7: <date> </date>

Date of publication, in the format defined by "Date and Time Formats" at http://www.w3.org/TR/NOTE-datetime and by ISO 8601 on which it is based. In particular, dates without times are represented in the form YYYY[-MM[-DD]]: a required 4-digit year, an optional 2-digit month, and if the month is given, an optional 2-digit day of month.

The date element has one optional OPF event attribute. The set of values for event are not defined by this specification; possible values may include: creation, publication, and modification.

2.2.8: <type> </type>

type includes terms describing general categories, functions, genres, or aggregation levels for content. The advised best practice is to select a value from a controlled vocabulary.

2.2.9: <format> </format>

The media type or dimensions of the resource. Best practice is to use a value from a controlled vocabulary (e.g. MIME media types).

2.2.10: <identifier> </identifier>

A string or number used to uniquely identify the resource. An OPF Package Document must include at least one instance of this element type, however multiple instances are permitted.

At least one identifier must have an id specified (the value being of the XML "ID" data type), so it can be referenced from the package unique-identifier attribute described in Section 2.1.

The identifier element has an optional OPF scheme attribute defined by this specification. The scheme attribute names the system or authority that generated or assigned the text contained within the identifier element, for example "ISBN" or "DOI." The values of the scheme attribute are case sensitive only when the particular scheme requires it.

This specification does not standardize or endorse any particular publication identifier scheme. Specific uses of URLs or ISBNs are not yet addressed by this specification. Identifier schemes are not currently defined by Dublin Core.

2.2.11: <source> </source>

Information regarding a prior resource from which the publication was derived; see the Dublin Core Metadata Element Set (http://dublincore.org/documents/2004/12/20/dces/).

2.2.12: <language> </language>

Identifies a language of the intellectual content of the Publication. An OPF Package Document must include at least one instance of this element type, however multiple instances are permitted. The content of this element must comply with RFC 3066 (see http://www.ietf.org/rfc/rfc3066.txt), or its successor on the IETF Standards Track. The Dublin Core permits other descriptions as well; this specification does not.

2.2.13: <relation> </relation>

An identifier of an auxiliary resource and its relationship to the publication.

2.2.14: <coverage> </coverage>

The extent or scope of the publication’s content. The advised best practice is to select a value from a controlled vocabulary; see the Dublin Core Metadata Element Set (http://dublincore.org/documents/2004/12/20/dces/).

2.2.15: <rights> </rights>

A statement about rights, or a reference to one. In this specification, the copyright notice and any further rights description should appear directly.

This specification does not address the manner in which a Content Provider specifies to a secure distributor any licensing terms under which readership rights or copies of the content could be sold.

2.3: Manifest

The required manifest must provides a list of all the files that are parts of the publication (e.g. Content Documents, style sheets, image files, any embedded font files, any included schemas). The manifest element must contain one or more item elements. Each item describes a document, an image file, a style sheet, or other component that is considered part of the publication. The manifest must not include item elements referring to the file or files that make up the OPF Package Document.

Each item element contained within a manifest element must have the attributes id, href (a URI; if relative, the URI is interpreted as relative to the OPF Package Document file containing the reference), and media-type (specifying the item’s MIME media type).

The order of item elements in the manifest is not significant.

For example:

<manifest>
        <item id="intro" href="introduction.html"
                media-type="application/xhtml+xml" />
        <item id="c1" href="chapter-1.html"
                media-type="application/xhtml+xml" />
        <item id="c2" href="chapter-2.html"
                media-type=application/xhtml+xml" />
        <item id="toc" href="contents.xml"
                media-type="application/xhtml+xml" />
        <item id="oview" href="arch.png"
                media-type="image/png" />
</manifest>

The URIs in href attributes of item elements in the manifest must not use fragment identifiers.

A single resource (href) must not be listed in the manifest more than once.

A Content Document’s root element must be consistent with the media-type of the associated item element within the manifest.

2.3.1: Fallback Items

The OPS specification defines a set of OPS Core Media Types that all conforming Reading Systems must support (e.g. XHTML, PNG, SVG). For a publication that uses only those media types, the manifest merely lists the publication's component files directly. However, content providers may construct publications that reference items of additional media types. In order for such publications to be read by all conforming Reading Systems, content providers must provide alternative "fallback" items for each such item. For every item that is not an OPS Core Media Type, at least one of its associated fallback items must either be of a type drawn from the set of OPS Core Media Types or, in some cases, CSS styling may be provided for documents containing non-preferred XML vocabularies.

This specification and the OPS specification jointly define four different mechanisms for specifying OPS Core Media Type fallbacks. These are as follows:

For inline "replaced" resources referenced via the object element, this specification relies on that element’s inherent replacement capabilities, described in Section 2.3.6 of the OPS specification.
For inline "replaced" resources referenced via the img element, the text value of the alt or title attributes provides a valid fallback, described in Section 2.3.4 of the OPS specification.
For Inline XML Islands a switch-based fallback mechanism is provided, described in Section 2.6.3.1 of the OPS specification.
For non-inline destinations, whether referenced from a document or a package, and for inline "replaced" resources referenced via the img element, the various attributes of the package item element are used to provide fallback information. This is defined in this section of this specification.

For the purpose of fallback specification, the file with the media type application/x-dtbncx+xml that specifies the publication’s NCX (see below) should be considered a Core Media Type, thus fallback information must not be provided for this file.

2.3.1.1: Items That Are Not OPS Core Media Types

An item that specifies a resource that is not an OPS Core Media Type ~~and is not an Out-Of-Line XML Island~~ (e.g. a non-core image type, a text file, a PDF file) must have a fallback specified. In this case, its fallback must be identified with a fallback attribute pointing to another item. See Section 2.3.1.2 for fallback requirements for Out-of-Line XML Islands.

An item identifies a fallback item using its fallback attribute, which must specify the ID of the item element that identifies the fallback. Items referenced from fallback attributes may each specify a fallback attribute in turn, forming a multi-level fallback chain. For example:

<manifest>
        <item id="item1"
                href="FunDoc.txt"
                media-type="text/plain"
                fallback="fall1" />
        <item id="fall1" fallback="fall2"
                href="FunDoc.pdf"
                media-type="application/pdf" />
        <item id="fall2"
                href="FunDoc.html"
                media-type="application/xhtml+xml" />
        <item ...>
</manifest>

If a fallback attribute points to an item that also has a fallback attribute, a Reading System must continue down the fallback chain until it reaches a reference to an item with a media type it can display (or as specified below, it reaches an item with a fallback-style attribute). A Reading System may continue further, and may display any item from the chain. In the absence of element-specific (i.e. img and object) fallback information, every item in a publication that does not have one of the OPS Core Media Types must, directly or indirectly, specify a fallback to an item that does have one of the OPS Core Media Types (or as specified below, an item with a fallback-style attribute).

Fallback chains must terminate; circular references are not permitted. Nevertheless, Reading Systems should not fail catastrophically if they encounter such a loop.

2.3.1.2: Items That Are Out-Of-Line XML Islands

An item that specifies a resource that is an Out-Of-Line XML Island (an XML document that is not authored in a Preferred Vocabulary). An item is an Out-Of-Line XML Island if:

It specifies a resource that is an XML document not authored in a Preferred Vocabulary (i.e. an XML document with a media-type that is neither application/xhtml+xml, application/x-dtbook+xml nor the deprecated text/x-eob1-document); or
It specifies a resource that is an XML document authored in a Preferred Vocabulary and incorporates the use of Extended Modules.

More information is needed, and more freedom is provided, for fallback determination and processing of Out-Of-Line XML Islands. The namespace of an Out-Of-Line XML Island item must be specified with the required-namespace attribute and its fallback must be identified with either a fallback attribute pointing to another item or by providing CSS styling that can be used to render the island via the fallback-style attribute.

If the fallback attribute is specified, Reading System processing is identical to the Non-OPS Core Media Types handling described above.

If the fallback-style attribute is specified, a Reading System may choose to process the Out-Of-Line XML Island (even though it can not natively process the vocabulary or Extended Modules used in the island) using the stylesheet specified by the fallback-style attribute's value which must contain a reference to the id of the item containing an href to the stylesheet desired for the island.

Both fallback and fallback-style attributes may be specified, in which case the Reading System may choose to either follow the fallback chain or to process the Out-Of-Line XML Island with the supplied CSS stylesheet.

An Out-Of-Line XML Island authored in a Preferred Vocabulary by definition incorporates the use of Extended Modules. In this case, and with non-Preferred Vocabulary islands utilizing Extended Modules, the required-modules attribute must be present along with the required-namespace attribute.

The attribute value for required-modules must be a comma-separated list containing the name(s) of the Extended Modules used in the Out-of-Line XML Island. The names of the modules are not case-sensitive, unless specifically defined otherwise in the XML vocabulary specification. Spaces in module names must be replaced with "-" for listing in the required-modules attribute value. For XHTML, in the context of OPS, the Extended Modules include ruby, forms, server-side-image-map, and intrinsic-events.

Note that listing the names of non-Extended Modules in a required-modules attribute value is also allowed; such modules are always considered to be supported if the XML vocabulary is supported. This can be useful both for clarity and in the case where there is a possibility that some modules could become optional in the later revisions of the specification (e.g. the currently deprecated Style Attribute XHTML module in OPS).

It is allowed, and sometimes useful, to provide a required-modules attribute on an item specifying a non-Preferred Vocabulary Out-Of-Line XML Island — either for clarity or to specify Extended Modules needed from the non-Preferred Vocabulary.

Most importantly, a Reading System that can natively process the non-Preferred Vocabulary (or Extended Modules) used for an Out-Of-Line XML Island may choose to use its integral understanding to natively process the document. However, fallback information must be provided for Reading Systems that do not have such native processing ability.

For example:

<manifest>
        <item id="item1"
              href="Doc1.hpy"
              media-type="text/happy+xml"
              required-namespace="http://happy.com/ns/happy1/"
              fallback="item2" />
        <item id="item2"
              href="Doc1.less-hpy"
              media-type="text/less-happy+xml"
              required-namespace="http://happy.com/ns/happy2/"
              fallback="item2.5"
              fallback-style="css1" />
        <item id="item2.5"
              href="Doc1.htm"
              media-type="application/xhtml+xml"
              required-namespace="http://www.w3.org/1999/xhtml"
                required-modules="ruby, server-side-image-map"
              fallback="item3" />
        <item id="item3"
              href="Doc1.dtb"
              media-type="application/x-dtbook+xml" />
        <item id="item4"
              href="Doc2.hpy"
              media-type="text/happy+xml"
              required-namespace="http://happy.com/ns/happy1/"
              fallback-style="css1" />
        <item id="css1"
              href="happy.css"
              media-type="text/css" />
</manifest>

In the above example when processing item1, a Reading System could choose to render item1 natively, item2 natively, item2 with only styling from css1, item2.5 natively (assuming the Ruby and Server Side Image Map XHTML Extended Modules are supported by the Reading System), or item3 natively. When processing item4, a Reading System could choose to render item4 natively or item4 with only styling from css1.

Inclusion of the required-namespace attribute is not required in item elements referring to XML documents authored in Preferred Vocabularies unless Extended Modules are used, in which case both required-namespace and required-modules attributes must be provided.

2.4: Spine

Following the manifest, there must be one spine element, which defines the primary linear reading order of the publication. It specifies an ordered list of one or more OPS Content Documents drawn from the manifest, using itemref elements contained within the spine element.

A publication must specify exactly one spine. Reading Systems must treat the file named in the first itemref element within the spine as the first file to be rendered in the reading of the publication. The successive files named in successive itemref elements are those that are to be rendered using "next-page" style functionality that could be available in the Reading System.

The spine must refer only to item elements with media types of application/xhtml+xml, application/x-dtbook+xml, the deprecated text/x-eob1-document, or an Out-Of-Line XML Island (with required fallback). Content with other media types may be referenced via OPS Content Documents, which should provide text alternates and other information to enhance accessibility as appropriate.

The spine need not include references to every one of the manifest’s item elements that reference OPS Content Documents, because there are means other than the spine for accessing documents in the publication. For example, hypertext links may provide access to documents not in the spine, as may tours and guides (see below).

~~For example:~~


<manifest>
        <item id="toc"
                href="contents.html"
                media-type="application/html+xml" />
        <item id="c1"
                href="chap1.html"
                media-type="application/html+xml" />
        <item id="c2"
                href="chap2.dtb"
                media-type="application/x-dtbook+xml" />
        <item id="c3"
                href="chap3.html"
                media-type="application/html+xml" />
        <item id="f1" href="fig1.jpg" media-type="image/jpeg" />
        <item id="f2" href="fig2.jpg" media-type="image/jpeg" />
        <item id="f3" href="fig3.jpg" media-type="image/jpeg" />
        <item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml" />
</manifest>
<spine toc="ncx">
        <itemref idref="toc" />
        <itemref idref="c1" />
        <itemref idref="c2" />
        <itemref idref="c3" />
</spine>

In the above example, suppose the document referenced by id c1 is being viewed by a reader. When the end of the document is reached, the next document in linear order would be the document referenced by id c2.

~~The NCX section below specifies the requirements for the navigation control file and the spine toc attribute.~~

Content is considered reachable if it is referenced directly or indirectly through hyperlinks or navigation structure (NCX). Referencing content through img or object elements does not make an item reachable. All content reachable in the publication must be referenced with itemref elements.

Each item must only be referenced by a single element (itemref). Content referenced by an itemref element that omits the optional linear attribute (or provides the linear attribute with a value of yes) is called primary content. Content referenced by an itemref element that includes the optional linear attribute with a value of no is referred to as auxiliary content.

Reading Systems must display all primary (and optionally auxiliary) content items in a way that allows linear (forward and backward) navigation through all the items in the order in which they are listed in the spine. The decision on whether to include auxiliary content in the linear navigation order should be based on the Reading System design and device capabilities. This specification neither encourages nor discourages such inclusion. Reading Systems must display in some fashion all auxiliary content items if such items can be reached through hyperlinks, even if the auxiliary content items are not included in the linear navigation order.

~~For example:~~


<spine toc="ncx">
   <itemref idref="intro" />
   <itemref idref="chapter1" />
   <itemref idref="chapter1_answerkey" linear="no" />
   <itemref idref="chapter2" />
   <itemref idref="chapter2_answerkey" linear="no" />
   <itemref idref="conclusion" />
   <itemref idref="about_this_book" linear="no" />
</spine>

~~Note that this design (the addition of the linear attribute) is backward compatible with OEBPS 1.2, assuming that the OEBPS 1.2 Reading System ignores unknown attributes.~~

Following manifest, there must be one and only one spine element, which contains one or more itemref elements. Each itemref references an OPS Content Document designated in the manifest. The order of the itemref elements organizes the associated OPS Content Documents into the linear reading order of the publication.

Each itemref in spine must not reference media types other than OPS Content Documents (or documents whose fallback chain includes an OPS Content Document). An OPS Content Document must be of one of the following media types: application/xhtml+xml, application/x-dtbook+xml, the deprecated text/x-oeb1-document, and Out-Of-Line XML Island (with required fallback.) When a document with a media type not from this list (or a document whose fallback chain doesn't include a document with a media type from this list) is referenced in spine, Reading Systems must not include it as part of the spine.

As items appearing in the spine must either be OPS Content Documents or items with a fallback chain that includes an OPS Content, it is possible to have a fallback chain for a spine item that "falls over" another OPS Core Media type. For example, a spine itemref could reference a PDF document, that falls back to a PNG image, that in turn falls back to a OPS XHTML Content Document. It is valid for this item to appear in the spine because the fallback chain includes (in this case terminates with) an OPS Content Document.

In addition, a specific spine item (from the perspective of its id attribute value in manifest) must not appear more than once in spine.

All OPS Content Documents that are part of the publication (i.e. are listed in the manifest) which are potentially reachable by any reference mechanism allowed in this specification must be included in the spine. Such reference mechanisms include, as a partial list, hypertext links within OPS Content Documents, and references by the NCX, Tours and Guide.

Should a Reading System encounter, by such reference, an OPS Content Document not listed in spine as required in this specification, the Reading System should add it to spine (the placement at the discretion of the Reading System) and assign the value of the linear attribute to no (see next.)

For each itemref, the publication author may specify the optional linear attribute to designate whether the associated OPS Content Document is primary (linear="yes", which is the default when linear is not present) or auxiliary (linear="no".) It is important that the publication author include some kind of internal reference, such as a hypertext link, to any OPS Content Document that is declared to be auxiliary; it is recommended that references be added to NCX for all auxiliary content. At least one itemref in spine must be declared primary.

Specifying whether an OPS Content Document is primary or auxiliary is useful for Reading Systems which opt to present auxiliary content differently than primary content. For example, a Reading System might opt to render auxiliary content in a popup window apart from the main window which presents the primary content. (For an example of the types of content that may be considered auxiliary, refer to the example below and the subsequent discussion.)

Reading Systems are not required to differentiate between primary and auxiliary content, and for the requirements and recommendations given in this section may consider all OPS Content Documents in spine to be primary, regardless of the value of the linear attribute.

The linear attribute also maintains compatibility with OEBPS 1.x, where not all reachable OEBPS content documents were required to be listed in the spine. For upgrading an OEBPS Publication to OPS, every unlisted, reachable content document in the OEBPS 1.x Publication should be assigned linear="no".

Reading Systems are to use the ordered itemref information in spine to present the publication during reading. Reading Systems must recognize the first primary OPS Content Document in spine to be the beginning of the main reading order of the publication. Successive primary OPS Content Documents form the remainder of the main reading order in the same order given in spine. Reading Systems may use "next-page" style functionality when moving from one primary OPS Content Document to the next primary one in spine.

The spine element must include the toc attribute, whose value is the the id attribute value of the required NCX document declared in manifest (see Section 2.4.1.)

Example illustrating spine and the optional linear attribute:

<manifest>
     <item id="intro"
           href="intro.html"
           media-type="application/xhtml+xml" />
     <item id="c1"
           href="chap1.html"
           media-type="application/xhtml+xml" />
     <item id="c1-answerkey"
           href="chap1-answerkey.html"
           media-type="application/xhtml+xml" />
     <item id="c2"
           href="chap2.dtb"
           media-type="application/x-dtbook+xml" />
     <item id="c2-answerkey"
           href="chap2-answerkey.html"
           media-type="application/xhtml+xml" />
     <item id="c3"
           href="chap3.html"
           media-type="application/xhtml+xml" />
     <item id="c3-answerkey"
           href="chap3-answerkey.html"
           media-type="application/xhtml+xml" />
     <item id="note"
           href="note.html"
           media-type="application/xhtml+xml" />
     <item id="f1"
           href="fig1.jpg"
           media-type="image/jpeg" />
     <item id="f2"
           href="fig2.jpg"
           media-type="image/jpeg" />
     <item id="f3"
           href="fig3.jpg"
           media-type="image/jpeg" />
     <item id="ncx"
           href="toc.ncx"
           media-type="application/x-dtbncx+xml" />
</manifest>
<spine toc="ncx">
     <itemref idref="intro" />
     <itemref idref="c1" />
     <itemref idref="c1-answerkey" linear="no" />
     <itemref idref="c2" />
     <itemref idref="c2-answerkey" linear="no" />
     <itemref idref="c3" />
     <itemref idref="c3-answerkey" linear="no" />
     <itemref idref="note" linear="no" />
</spine>

In the above example, the publication author set linear="no" on four of the eight OPS Content Documents listed in spine, designating these content documents to be auxiliary. Three of the four are "answer keys," and the fourth is a note of some sort; all four are auxiliary to the main flow of the book and may be viewed separately from the main flow.

Reading Systems which recognize and render auxiliary content separate from primary content will set the main reading order to be the four primary OPS Content Documents: intro, c1, c2 and c3. The auxiliary content documents will be rendered by such Reading Systems, upon activation (such as through a hypertext link or entry in NCX), in some manner distinct from the main reading order. It is important that the publication author provide the necessary references to the auxiliary content documents, otherwise this content might not be reachable in some auxiliary-aware Reading Systems.

Reading Systems which opt to ignore linear="no" and set all itemref to be primary, as allowed in this specification, will assign all eight OPS Content Documents to the main reading order in the order given. This is especially useful for Reading Systems which provide print output, where it is important that all the information in the OPS Content Documents be printed in an author-determined linear order.

A Reading System may, at its discretion, provide both rendering options to the user.

2.4.1: Declarative Table of Contents — the NCX

In order to enable ease of navigation and provide greater accessibility, the OPF Package Document supports a "Navigation Center eXtended," the NCX. This is a concept and implementation that has been standardized by the DAISY Consortium.

This specification uses the NCX defined in the DAISY/NISO Standard, formally the ANSI/NISO Z39.86-2005, Specifications for the Digital Talking Book. The NCX is a portion (Section 8) of this comprehensive multimedia standard. The DAISY Consortium maintains the NCX DTD. No modifications to the DTD are necessary for use with OPF. In the future the DAISY/NISO Advisory Committee will consider modularizing the NCX and changing terminology to be more in line with ebooks, multimedia publications and other electronic document usage.

Some optional elements and metadata items are not needed to implement the NCX for this specification. The sections below have been changed to normatively reference the DAISY/NISO standard for the NCX rather than duplicating it here. All "exceptions" are described in Section 2.4.2, below.

2.4.1.1: Introduction

The Navigation Control file for XML applications (NCX) exposes the hierarchical structure of a Publication to allow the user to navigate through it. The NCX is similar to a table of contents in that it enables the reader to jump directly to any of the major structural elements of the document, i.e. part, chapter, or section, but it will often contain more elements of the document than the publisher chooses to include in the original print table of contents. It can be visualized as a collapsible tree familiar to PC users. Its development was motivated by the need to provide quick access to the main structural elements of a document without the need to parse the entire documents. Other elements such as pages, footnotes, figures, tables, etc. can be included in separate, nonhierarchical lists and can be accessed by the user as well.

It is important to emphasize that these navigation features are intended as a convenience for users who want them, and not a burden to those who do not. The alternative guide to the book may be provided for those users not requiring the navigation features of the NCX.

2.4.1.2: Key NCX Requirements

Reading Systems must support NCX.

~~Publications containing only XHTML OPS Content Documents should include an NCX (this "should" could transition to a "must" in future versions of this specification).~~

OPS Publications ~~containing DTBook or Out-of-Line XML Island OPS Content Documents~~ must include an NCX.

Like all other Publication components, the NCX must be included as an item in the manifest (with media-type of application/x-dtbncx+xml). The NCX-referencing item must not contain any fallback information (required-namespace, fallback or fallback-style attributes).

If a Publication includes an NCX, the item that describes the NCX must be referenced by the spine toc attribute.

The NCX file must be a valid XML document conforming to ncx-2005-1.dtd and comply with the additional normative requirements defined in http://www.niso.org/standards/resources/Z39-86-2005.html. The version and xmlns attributes on the ncx element must be explicitly specified in the document instance, using values drawn from the above-named DTD.

2.4.2: NCX Exceptions in Usage for Publications

The NCX as defined in the ANSI/NISO Z39.86-2005 Standard Section 8 is ideal for OPS applications, however, some exceptions are noted herein. In the standard, the links from the NCX to the Publication point to SMIL (http://www.w3.org/TR/2005/REC-SMIL2-20050107/) documents. For OPS Publications, the links will point to an XML elements in the source OPS Content Document. This difference causes the following exceptions to be noted from Section 8 in that standard:

smilCustomTest is not used in the OPF application of the NCX;
audio is not used in the OPS application of the NCX;
clipBegin (%SMILtimeVal, REQUIRED): is used for identifying the start of an audio segment, and hence not used in the OPF application of the NCX;
clipEnd (%SMILtimeVal, REQUIRED): is used for identifying the end of an audio segment, and hence not used in the OPF application of the NCX;
content Description: Pointer into SMIL file to beginning of the item referenced by the navPoint or navTarget. Note: In the OPF application of the NCX the pointer is to an XML element and not a SMIL location.
DTBs Spanning Multiple Media Units are not relevant in the Publication context because OPS Publications are significantly smaller than multimedia files.
Examples show links to SMIL files, but in the OPF application of NCX the links will be to XML elements. Also, the examples show audio references with clipBegin and clipEnd, which are not used in the OPF application of the NCX.
The required NCX metadata items "dtb:totalPageCount" and "dtb:maxPageNumber" may not be relevant when content is not related to a rendition with "paper pages". In this case, their values may be specified as zero and the Reading System may ignore them.
playOrder attributes remain required. Although playOrder attributes are not used to sequence SMIL content, they should contain valid values that reflect the linear document reading order. This could, for example, be used when one is navigating a pageList to find a corresponding location in the navMap.

2.4.3: XML Islands in the Spine

XML Islands may be referenced from the spine. In the event that a Reading System cannot display the XML Island correctly, then the standard fallback methodology defined in the Open Publication Structure must be used.

In short, the Reading System must display the chosen fallback for an XML Island in the event that the island itself cannot be displayed.

2.5: Tours [Deprecated]

Much as a tour-guide might assemble points of interest into a set of sightseers' tours, a content provider could assemble selected parts of a publication into a set of tours to enable convenient navigation.

An OPS Package Document may, but need not, contain one tours element, which in turn contains one or more tour elements. Each tour must have a title attribute, intended for presentation to the user. Reading Systems may use tours to provide various access sequences to parts of the publication, such as selective views for various reading purposes, reader expertise levels, etc. Because Reading Systems are not required to implement tour support, content providers should also provide other means of accessing content referenced from tours.

Each tour element contains one or more site elements, each of which must have an href attribute and a title attribute. The href attribute must refer to an OPS Content Document included in the manifest, and may include a fragment identifier as defined in section 4.1 of RFC 2396 (see http://www.ietf.org/rfc/rfc2396.txt). Each site element specifies a starting point from which the reader can explore freely. Reading Systems may use the bounds of the referenced element to determine the scope of the site. If a fragment identifier is not used, the scope is considered to be the entire document. This specification does not require Reading Systems to mark or otherwise identify the entire scope of a referenced element. The order of site elements is presumed to be significant, and should be used by Reading Systems to aid navigation.

For example:

<tours>
        <tour id="tour1" title="Chicken Recipes">
                <site title="Chicken Fingers"
                  href="appetizers.html#r3" />
                <site title="Chicken a la King"
                  href="entrees.html#r5" />
        </tour>
        <tour id="tour2" title="Vegan Recipes">
                <site title="Hummus" href ="appetizer.html#r6" />
                <site title="Lentil Casserole" href="lentils.html" />
        </tour>
</tours>

2.6: Guide

Within the package there may be one guide element, containing one or more reference elements. The guide element identifies fundamental structural components of the publication, to enable Reading Systems to provide convenient access to them.

For example:

<guide>
        <reference type="toc" title="Table of Contents"
                 href="toc.html" />
        <reference type="loi" title="List Of Illustrations"
                 href="toc.html#figures" />
        <reference type="other.intro" title="Introduction"
                 href="intro.html" />
</guide>

The structural components of the books are listed in reference elements contained within the guide element. These components could refer to the table of contents, list of illustrations, foreword, bibliography, and many other standard parts of the book. Reading Systems are not required to use the guide element in any way.

Each reference must have an href attribute referring to an OPS Content Document included in the manifest, and which may include a fragment identifier as defined in section 4.1 of RFC 2396 (see http://www.ietf.org/rfc/rfc2396.txt). Reading Systems may use the bounds of the referenced element to determine the scope of the reference. If a fragment identifier is not used, the scope is considered to be the entire document. This specification does not require Reading Systems to mark or otherwise identify the entire scope of a referenced element.

The required type attribute describes the publication component referenced by the href attribute. The values for the type attributes must be selected from the list defined below when applicable. Other types may be used when none of the predefined types are applicable; their names must begin with the string other.. The value for the type attribute is case-sensitive.

The following list of type values is derived from the 13th edition of the Chicago Manual of Style:

cover	the book cover(s), jacket information, etc.
title-page	page with possibly title, author, publisher, and other metadata
toc	table of contents
index	back-of-book style index
glossary
acknowledgements
bibliography
colophon
copyright-page
dedication
epigraph
foreword
loi	list of illustrations
lot	list of tables
notes
preface
text	First "real" page of content (e.g. "Chapter 1")

Appendices