<!--
Original authors:

Harvey Bingham
George Kerscher
Michael Moodie
David Pawson

Assisted by DAISY Consortium and NISO DTB Committee work teams.

1. Purpose

The Digital Talking Book Document Type Definition (DTD) provides
the means to mark up the text of a document to permit support for
the combination of professional narration and navigation into that
narration. It also facilitates the output of a document's content in
a variety of accessible formats. The markup tags in the book convey
its content in structure, and contain some metadata about the book
content and its structure.

The Document Type Definition names and defines the allowable element
types, their allowable content, and their attributes. Correct markup
of the text of the book permits the textual material to be synchronized
using SMIL [SMIL2.0] files with the professionally narrated version of
that book. The synchronization can permit concurrent display of the
text being narrated. The textual content can be searched in context to
locate material desired for narration.

More detailed documentation of this dtbook dtd [DTBOOKV20053DTD] is
available as an html document. See [DTBOOKV20053DOC].

1.1. Prior Related Work

The DAISY (Digital Accessible Information SYstem) Consortium
contributed substantially to the development of this DTD.
This application of XML is the next generation after several DAISY
versions of 2.X specifications, see [DAISY202].

The DAISY Statement of Principles for the Creation and Production
of Accessible Books and Materials [DAISY-2-GUIDELINES] represents
the minimum standard to be met by Libraries of the Blind and producers
of alternative format materials.

Its Navigation Control Center (NCC) provided for synchronizing
document structure with narration.

The NCC evolved into an XML application called the "Navigation Control
File for XML applications" (NCX). Its content is derived from
the markup of documents tagged using the dtbook DTD. Richer
structuring capability is one of the objectives of that DTD. The
Synchronized Multimedia Integration Language [SMIL2.0] is used
to provide synchronized narrations and text. The NCX provides
navigation using the identified elements of documents tagged to this DTD.

The dtbook DTD includes many, but not all, of the element types found
in both the [HTML401STRICT] and [XHTML11STRICT] strict DTDs. HTML
authoring tools permit those additional element tags, and may ignore
the additional tags that are dtbook-specific. The lowercase names
from XHTML are used, rather than the uppercase names from HTML.

1.2. Evolution from HTML and XHTML

Dtbook-2005-3 has 83 element types. It shares 46 element types with the
HTML4.0 Strict DTD [HTML401STRICT] (as adjusted to use the lower-case
names consonant with the XHTML Strict DTD [XHTML11STRICT]). It omits
31 element types from them, and has 32 unique element types.

Endtag markup is sometimes optional in HTML. It is required for use with
xhtml and dtbook. Any XML application [XML12] requires endtags, or their
abbreviated form for empty elements, such as "<link />". The benefit of
including endtags is that the tagged document has dependable structure
that can be validated against the dtbook dtd.

Some tools available for browsing HTML may be used with dtbook
material, at the expense of their discarding or ignoring some specific
tagging and attributes that are not part of HTML 4.0. A CSS-based
stylesheet [CSS1] or [CSS2] that identifies the presentation expectations
for the HTML and non-HTML tags, or a filter to map those tags onto
suitable HTML tags can provide appropriate visual presentation.

2. Document Prolog

A Digital Talking Book document is an XML application. Therefore, it
must begin with the XML declaration, followed by the DOCTYPE
declaration.

2.1. XML Declaration

The XML declaration identifies the version of XML, and the
optional character set encoding for the document:

<?xml version="1.0" encoding="UTF-8" ?>

2.2. Character Set Encodings (removed as redundant; unicode suffices)

2.3. DOCTYPE Declaration

The document type declaration, the DOCTYPE, follows. It has several forms.
The simpler form assumes that the proper version of the dtbook DTD
is in the same directory as the dtbook file itself.

<!DOCTYPE dtbook SYSTEM
"dtbook-2005-3.dtd">

A more general form provides the PUBLIC URI from which the SYSTEM
filename can be substituted, should that system copy be missing:

<!DOCTYPE dtbook PUBLIC
"http://www.daisy.org/z3986/2005/dtbook-2005-3.dtd"
"dtbook-2005-3.dtd">

That assumes the URI can be reached, which may not be true for
portable dtbook players.

The still more general form recommended for xml applications [XML12] is:

<!DOCTYPE dtbook PUBLIC
"-//NISO//DTD dtbook 2005-3//EN"
"http://www.daisy.org/z3986/2005/dtbook-2005-3.dtd">

where the Formal Public Identifier (FPI) on the second line is converted
to the URI where it may be publicly found:

http://www.daisy.org/z3986/2005/dtbook-2005-3.dtd

The [OASIS-TR9401] Entity Management Catalog provides an indirect
means to provide that mapping from FPI to the dtd.

That catalog is more generally useful to provide the mapping from
any external entity names (such as modules) to URIs where they may
be found.

Note that the reference above is to a particular version of the DTD,
distinguished by the string "2005-3".

2.4. Digital Talking Book File MIME Type

A Digital Talking Book document is tagged to the dtbook XML
application. Its MIME media-type is "text/xml". The tagged book
filename should have suffix ".xml". See [RFC2045].

3. Modular Extension to the DTD

The dtbook DTD has four parameter entities defined that provide means
to allow an individual book to modularly extend the content models:

<!ENTITY % externalblock "">
<!ENTITY % externalinline "">
<!ENTITY % externalFlow "">
<!ENTITY % externalNamespaces "">

These parameter entities appear in corresponding block and inline
content models. With this "" content they have no effect on books
tagged to the dtbook DTD. In a book that needs a modular extension,
values are given by redefinition in the internal subset of that book.
This extends the dtbook DTD without having to change it.

A book can augment the dtbook DTD by including other declarations
or parameter entity references in the internal subset of declarations.
The internal subset may occur in square brackets following the
ExternalID and before the concluding ">" of the initial DOCTYPE
declaration that identifies the dtbook DTD.

Those additional markup declarations in the internal subset
take preference over any in the dtbook DTD itself. The effective
DTD is thereby augmented by the parameter entity values and any other
declarations of the book's internal subset. When a given parameter
entity declaration appears more than once in the external modules and
the dtbook DTD, the first occurrence of that declaration is the one
that takes effect, with modules in the internal subset being processed
in order, before the DTD itself.

For example:

<!DOCTYPE dtbook SYSTEM
"dtbook.dtd"
[
<!ENTITY % dramaModule SYSTEM "drama.dtd">
%dramaModule;
<!ENTITY % externalblock "| d:drama">
<!ENTITY % externalinline "| d:character">
<!ENTITY % externalFlow "| d:stagedir">
<!ENTITY % externalNamespaces "xmlns:d CDATA #FIXED 'http://www.sample.org/drama'">
]>

The "%dramaModule;" invocation causes all declarations made within
dramaModule to become the initial part of the dtbook DTD. Within the
book, the empty entity declarations for % externalblock,
% externalinline, % externalFlow, and %externalNamespaces
are replaced by these new definitions. Thus the
block element d:drama can appear wherever block elements may occur in
dtbook, d:character can appear wherever inline elements may occur, and
d:stagedir can appear in either block or inline contexts. Note that
each extension element must appear in no more than one of these three
entities. The namespace attribute xmlns:d can occur on any element with
the fixed value of "http://www.sample.org/drama".

More than one module may be needed and included in a book, for example
both poem and drama can appear in the internal subset of the book.
For example, the internal subset of the book could contain:

Such external modules need to include the definitions of any parameter
entities that are used in the modules since their definitions are needed
before they can be expanded in their references. They cannot depend
on parameter entities in the SystemLiteral or PubidLiteral.

Note that arbitrary external modules from other sources may not have
all the needed attributes. XML allows augmentation of ATTLISTs in the
internal subset. Additional attribute names can be added to an
associated element type. Any redefinitions of a particular named
attribute resulting from presence in the internal subset have
precedence.

Note that tools and players processing any extended markup that affects
navigation structure will need to know of those modular extensions.

The form above for augmenting the dtbook dtd through the document's
internal subset does not require the XML namespace mechanism, with
its namespace-specific prefix on element and attribute names to
disambiguate any potential name collisions. However, use of XML
namespaces [XML-NAMES] is recommended.

4. References

These references are informative. The bracketed names here are targets for
indirect reference from the corresponding bracketed names in other parts
of this document or in descriptions within this section.

[CSS1] Cascading Style Sheets, Level 1. Rec-CSS1-1999011 Revised 11 Jan 1999

http://www.w3.org/TR/REC-CSS1

[CSS2] Cascading Style Sheets, Level 2 CSS2 Specification REC-CSS2-19980512

http://www.w3.org/TR/REC-CSS2

[DAISY202] The DAISY 2.02 Specification for the DAISY Digital Talking
Book (DTB) format, which enables navigation within a sequential
and hierarchical structure consisting of (marked-up) text synchronized
with audio.

http://www.daisy.org/dtbook/spec/2/final/d202/daisy_202.html

[DAISY-2-GUIDELINES] The DAISY 2.02 Specification for
the Creation and Production of Accessible Books and Materials,
Version 0.99 1999-09-23 represents minimum standard to be met by
Libraries for the Blind and producers of alternative format materials:

http://www.daisy.org/dtbook/guidelines/draft/principles.htm

[DTBOOKV20053DTD] The dtbook DTD version 2005-3 (this DTD) is available at:

http://www.daisy.org/z3986/2005/dtbook-2005-3.dtd

Note that some browsers do not permit downloading a file with suffix dtd.

[DTBOOKV20053DOC] Digital Talking Book Expanded Document Type Definition
Documentation for Version 2005-3 of this DTD is available as an
HTML 4.0 document:

http://www.daisy.org/z3986/2005/dtbook/dtbookdoc.html

Should revisions occur, a new directory with node named for the year
of change will contain the revisions.

Any prior specific version of the dtbook dtd and its documentation will
persist.

[DTBOOK3] The last public beta version was dtbook3-07.dtd (2001-01-31).

http://www.loc.gov/nls/z3986/background/dtbk3_old_dtds/dtbk3-07.dtd

and its expanded documentation:

http://www.loc.gov/nls/z3986/background/dtbk3_old_dtds/dtbk3-07doc.htm

Those and prior versions are available at:

http://www.loc.gov/nls/z3986/background/dtbk3_old_dtds/index.html

The history of changes prior to this version, including those
in internal drafts through dtbk110.dtd and before is in:

http://www.loc.gov/nls/z3986/background/dtbook-dtd-changes.txt

In that directory also are the old dtdbk3 dtds, some of which have
been used for test markup, and their documentation. See its
index.html for the list. (Caution: some browsers may not
permit downloading DTDs.)

http://www.loc.gov/nls/z3986/background/index.html

[HTML401STRICT] "HTML 4.0 Strict DTD," 1999-12-24, Dave Raggett,
Arnaud Le hors, and Ian Jacobs. Dtbook110 was originally based on
the HTML 4.0 Strict DTD with design adaptation for dtbook110.
A principal adaptation is to use lower-case names for element types
and attribute names. For expanded discussion, see [HTML401].

http://www.w3.org/TR/1999/REC-html401-19991224/strict.dtd

[HTML401] "HTML 4.01 Specification" W3C Recommendation 24 December 1999
Documentation of the element types that come from the HTML 4.0 Strict
DTD [HTML401STRICT] is available at:

http://www.w3.org/TR/1999/REC-html401-19991224/

Dtbook110 (and now Dtbook122) is partially harmonized with the
[XHTML11STRICT] DTD.

The XHTML camelCase parameter entity names are retained, and comments
and references following those parameter entities explain them. The
lower-case element and attribute names are used. The simplified table
content model of just table rows is included.

[ISO10646] "Information Technology - Universal Multiple-Octet Coded
Character Set (UCS) - Part 1: Architecture and Basic Multilingual
Plane", ISO/IEC 10646-1:1993. The current specification also takes
into consideration the first five amendments to ISO/IEC 10646-1:1993.

[ISO8859] "Information Processing - 8-bit single-byte coded graphic
character sets - Part 1: Latin alphabet No. 1," ISO 8859-1:1987.
Other suffixes "-2 through -9" correspond to other character sets
in the family.

[JIS] "JIS Character Sets" describes the history of JIS, and the
several character sets for KANJI, KANA and other characters.

http://www.io.com/~kazushi/encoding/jis.html

[ANSINISOZ39-86-2002] Specifications for the Digital Talking Book.

http://www.niso.org

[ANSINISOZ39-86-2005] Specifications for the Digital Talking Book.

http://www.niso.org

[NLS-Z3986] Development of ANSI/NISO Z39.86
Contains links to the DTDs developed for ANSI/NISO Z39.86,
Specifications for the Digital Talking Book

http://www.loc.gov/nls/z3986/index.html

[OASIS-TR9401] Entity Management, OASIS Technical Resolution 9401:1997
(Amendment 2 to TR 9401). Paul Grosso, 1997 September 10.

http://www.oasis-open.org/specs/tr9401.html

[RFC1556] "Handling of Bi-directional Texts in MIME," H. Nussbacher,
December 1993.

http://www.cis.ohio-state.edu/cgi-bin/rfc/rfc1556.html

[RFC1942] "HTML Tables", D. Raggett, May 1996

http://www.ietf.org/rfc/rfc1942.txt

Contains detailed descriptions of table elements and their
inheritance of attribute values. Adjustment for XML application is
required: end-tags are necessary, not optional, attribute values
must be quoted.

[RFC2045] "Multipurpose Internet Mail Extensions (MIME) Part One:
Format of Internet Message Bodies", N. Freed and N. Borenstein,
November 1996. Note that this RFC obsoletes RFC1521, RFC1522, and RFC1590.
The %ContentType; and %ContentTypes; media types and the
%Charset; and %Charsets; character encoding values are from [RFC2045].

http://www.cis.ohio-state.edu/cgi-bin/rfc/rfc2045.html

[RFC2046] "Multipurpose Internet Mail Extensions (MIME) Part Two:
Media Types," N. Freed, November 1996. Source for %ContentType; and
%ContentTypes; permitted values:

http://www.cis.ohio-state.edu/cgi-bin/rfc/rfc2046.html

[RFC2396] "Uniform Resource Identifiers (URI): Generic Syntax,"
T. Berners-Lee, R. Fielding, L. Masinter, August 1998. Note that this RFC
revises and replaces the generic definitions in RFC 1738 and RFC 1808.

http://www.cis.ohio-state.edu/cgi-bin/rfc/rfc2396.html

[RFC4646] "Tags for Identifying Languages,"
A. Phillips, Ed., September 2006

http://www.faqs.org/rfcs/rfc4646.html

[SMIL2.0] The Synchronized Multimedia Integration Language SMIL 2.0
W3C Recommendation 07 August 2001 is available at:

http://www.w3.org/TR/2001/REC-smil20-20010807/smil20.html

[XHTML11] "XHTML (tm) 1.0: The Extensible HyperText Markup Language,"
W3C Recommendation 26 January 2000, A reformulation of HTML4 in XML 1.0
includes case-sensitive names, lower-case for elements and their
attributes (but not parameter entity names) and in some cases
equivalent content models that do not require SGML inclusions
and exclusion exceptions (as occurred in the HTML4.0 strict
DTD [HTML401STRICT]) is available at:

http://www.w3.org/TR/xhtml/

[XML-NAMES] "Namespaces in XML 1.1" World Wide Web Consortium
W3C Recommendation 4-Febrary-2004

http://www.w3.org/TR/REC-xml-names11-20040204

[XHTML11STRICT] Expanded documentation of the element types that come
from the XHTML11 strict.dtd and its other DTDs is available within
the zip file:

http://www.w3.org/TR/xhtml1/DTD/xhtml1/xhtml1.zip

Note: some browsers cannot download a dtd directly.

[XML12] This dtbook-2005-3.dtd is an application of the Extensible Markup
Language XML 1.0 (Second Edition) W3C Recommendation 6 October 2000.
It is available at:

http://www.w3.org/TR/REC-xml

-->

<!-- Some comments start with a pattern followed by a colon:

Use: element type and its use.

Attuse: attribute use for associated element type.

HB: date object comment on change by Michael Moodie.

Other comments without such a pattern are dividing lines,
details about the DTD structure, or about dtbook objects.
-->

<!ENTITY % Character "CDATA" >

<!ENTITY % Charset "CDATA" >

<!ENTITY % ContentType "CDATA" >

<!ENTITY % LanguageCode "NMTOKEN" >

<!ENTITY % Number "CDATA" >

<!ENTITY % LinkTypes "CDATA" >

<!ENTITY % MediaDesc "CDATA" >

<!ENTITY % Text "CDATA" >

<!ENTITY % URI "CDATA" >

<!ENTITY % externalblock "" >

<!ENTITY % externalinline "" >

<!ENTITY % externalFlow "" >

<!ENTITY % externalNamespaces "" >

<!ENTITY % dtbookblocknoimggroup
"author | prodnote | sidebar | note |
annotation %externalblock;" >

<!ENTITY % inlineinblock
"a | cite | samp | kbd | pagenum" >

<!ENTITY % coreattrs
"id ID #IMPLIED
class CDATA #IMPLIED
title %Text; #IMPLIED
xml:space (default|preserve) #IMPLIED
%externalNamespaces;"
>

<!-- coreattrs are attributes permissible for most elements

id document-wide unique id
class space separated list of classes used for rendering
title advisory title/amplification
xml:space whitespace handling (see http://www.w3.org/TR/xml11/#sec-white-space)
-->

<!ENTITY % i18n
"xml:lang %LanguageCode; #IMPLIED
dir (ltr|rtl) #IMPLIED" >

<!-- i18n internationalization attributes
xml:lang language code (as per XML 1.0 spec)
dir direction for weak/neutral text
ltr=left to right
rtl=right to left

xhtml recommendation: use xml:lang such as "en-US", on the major
containing block, to provide source language for
the #IMPLIED values of its descendent elements.
See [RFC1556] for handling bi-directional text in MIME.
-->

<!ENTITY % showin
"showin (xxx|xxp|xlx|xlp|bxx|bxp|blx|blp) #IMPLIED" >

<!--showin attribute applies for text elements to permit identification
of the kinds of display appropriate for the element, so presentation
choice by the reader among alternative readings can be provided, when
appropriate. Values of showin are coded with three letters in order:
"b"=Braille, "l"=Largeprint, and "p"=Print; or "x"=inappropriate:

Value Braille Largeprint Print Interpretation

"xxx" hide
"xxp" p print only
"xlx" l largeprint only
"xlp" l p largeprint and print
"bxx" b braille only
"bxp" b p braille and print
"blx" b l braille and largeprint
"blp" b l p braille, largeprint, and print

There is no default value; this attribute value is implied
from the most immediate ancestor that specifies a value.
The usual default for showin is 'blp'. If only one showin
value is needed it should be included with <book>.

Different content for the same element (usually <prodnote>) meeting
different needs is possible, with showin serving as a switch to
differentiate among them. Both largeprint and print are appropriate
for screen rendering as well as printing. Different corresponding
styles may be appropriate.

It is possible to include equivalent content from any major structure
below <book> to provide the different content suitable for different
media. These would be independent, sharing no direct content, possibly
having common references to images, with different accompanying text
descriptions.
-->

<!ENTITY % attrs
"%coreattrs;
%i18n;
smilref CDATA #IMPLIED
%showin;" >

<!-- %attrs; is part of most attribute lists. It includes

%coreattrs; from which come the four #IMPLIED attributes:
id, class, title, and xml:space

%i18n; from which come the implied attributes: xml:lang, and dir

smilref is a pointer to a [SMIL2.0] file, normally to the time container
(SMIL <par> or <seq>) containing the media object that references this
element. However, in a text-only DTB consisting of a sequence of
text media objects, smilref points to the media object that
references this element. smilref allows resumption of SMIL
presentation at the proper location after navigation via dtbook file.
All smilref values are expected to be added to an augmented
version of the <dtbook> during production.

%showin; (See entity declaration.)
-->

<!ENTITY % attrsrqd
"id ID #REQUIRED
class CDATA #IMPLIED
title %Text; #IMPLIED
xml:space (default|preserve) #IMPLIED
smilref CDATA #IMPLIED
%i18n;
%showin;
%externalNamespaces; " >
<!-- %attrsrqd; includes required id and implied class,
title, and xml:space (see %coreattrs above)

%i18n; from which come the implied attributes: xml:lang, and dir

%showin; (See entity declaration.)
-->

<!ENTITY % dtbookcontent
"head, book" >

<!ELEMENT dtbook (%dtbookcontent;) >

<!--Attuse: dtbook

"version" is required, and contains the specific
version of the dtd, so that the dtd version for any dtbook can
be recognized.

"xmlns" is the URL where this dtd and its related materials can
be found.

"%i18n;" internationalization attributes characterize the <book>.
Those values may be adjusted for language changes within it.
-->

<!ATTLIST dtbook
version CDATA #FIXED '2005-3'
xmlns %URI; #FIXED 'http://www.daisy.org/z3986/2005/dtbook/'
%i18n;
%externalNamespaces;
>

<!ENTITY % headmisc
"meta | link " >

<!ELEMENT head (%headmisc;)* >

<!ATTLIST head
%i18n;
profile %URI; #IMPLIED
>

<!ELEMENT link EMPTY >

<!--Attuse: link

Each attribute use indicated by a parameter entity is
defined in the comment following its definition.
-->

<!ATTLIST link
%attrs;
charset %Charset; #IMPLIED
href %URI; #IMPLIED
hreflang %LanguageCode; #IMPLIED
type %ContentType; #IMPLIED
rel %LinkTypes; #IMPLIED
rev %LinkTypes; #IMPLIED
media %MediaDesc; #IMPLIED
>

<!ELEMENT meta EMPTY >

<!--Attuse: meta

"http-equiv" connects the content attribute
value to an http header field.

"name" value identifies the specific kind of
content value.

"content" indicates the value for that "name",
possibly constrained by the semantics for the individual names.

"scheme" indicates a predetermined format for interpreting
the content value, such as the Dublin Core.
-->

<!ATTLIST meta
%i18n;
http-equiv NMTOKEN #IMPLIED
name NMTOKEN #IMPLIED
content CDATA #REQUIRED
scheme CDATA #IMPLIED
>

<!ELEMENT book (frontmatter?, bodymatter?, rearmatter?) >

<!ATTLIST book
%attrs;
>

<!ELEMENT frontmatter (doctitle, covertitle?, docauthor*, (level | level1)*) >

<!ATTLIST frontmatter
%attrs;
>

<!ELEMENT bodymatter (level | level1)+ >

<!ATTLIST bodymatter
%attrs;
>

<!ELEMENT rearmatter (level | level1)+ >

<!ATTLIST rearmatter
%attrs;
>

<!ELEMENT level ((hd, (%docblockorinline; | level)+) |
((%docblockorinline; | level)+, (hd, (%docblockorinline; | level)+)?))
>

<!--Attuse: level

"class" identifies the actual name (e.g., part,
chapter, section, subsection) of the structure it marks.
"depth" indicates the nesting depth, starting at 1.
-->

<!ATTLIST level
%attrs;
depth CDATA #IMPLIED
>

<!ELEMENT level1 ((h1, (%docblockorinline; | level2)+) |
((%docblockorinline; | level2)+, (h1, (%docblockorinline; | level2)+)?))
>

<!ATTLIST level1
%attrs;
>

<!ELEMENT level2 ((h2, (%docblockorinline; | level3)+) |
((%docblockorinline; | level3)+, (h2, (%docblockorinline; | level3)+)?))
>

<!ATTLIST level2
%attrs;
>

<!ELEMENT level3 ((h3, (%docblockorinline; | level4)+) |
((%docblockorinline; | level4)+, (h3, (%docblockorinline; | level4)+)?))
>

<!ATTLIST level3
%attrs;
>

<!ELEMENT level4 ((h4, (%docblockorinline; | level5)+) |
((%docblockorinline; | level5)+, (h4, (%docblockorinline; | level5)+)?))
>

<!ATTLIST level4
%attrs;
>

<!ELEMENT level5 ((h5, (%docblockorinline; | level6)+) |
((%docblockorinline; | level6)+, (h5, (%docblockorinline; | level6)+)?))
>

<!ATTLIST level5
%attrs;
>

<!ELEMENT level6 ((h6, (%docblockorinline;)+) |
((%docblockorinline;)+, (h6, (%docblockorinline;)+)?))
>

<!ATTLIST level6
%attrs;
>

<!ENTITY % special
"a | img | imggroup | br | q | sub | sup | span | bdo " >

<!ENTITY % specialnoa
"img | imggroup | br | q | sub | sup | span | bdo " >

<!ENTITY % inline
"#PCDATA | %phrase; | %special; | %dtbookinline;" >

<!ENTITY % inlinenoa
"#PCDATA | %phrase; | %specialnoa; | %dtbookinline;" >

<!ENTITY % inlinew
"#PCDATA | %phrase; | %special; %externalinline; %externalFlow;" >

<!ENTITY % flow
"%inlinenoprodnote; | %blocknoimggroup; %externalFlow;" >

<!ENTITY % flownopagenum
"%inlinenopagenum; | %blocknoimggroup; %externalFlow; " >

<!ELEMENT br EMPTY >

<!--Attuse: br

The %coreattrs; only appear, as there is no content
to which the more general %attrs; apply.
-->

<!ATTLIST br
%coreattrs;
>

<!ELEMENT line (%inline; | linenum)* >

<!ATTLIST line
%attrs;
>

<!ELEMENT linenum (#PCDATA) >

<!ATTLIST linenum
%attrs;
>

<!ELEMENT address (%inline; | line)* >

<!ATTLIST address
%attrs;
>

<!ELEMENT div (%docblockorinline;)+ >

<!--Attuse: div

"class" value can identify the
actual name (e.g., part, chapter, letter) of the structure it marks.
-->

<!ATTLIST div
%attrs;
>

<!ELEMENT title (%inline;)* >

<!ATTLIST title
%attrs;
>

<!ELEMENT author (%inline;)* >

<!ATTLIST author
%attrs;
>

<!ELEMENT prodnote (%flow;)* >

<!--Attuse: prodnote

"imgref" identifies the space-separated id value(s)
on pertinent images <img>.

"render" indicates that the content is
'required' or 'optional' for the user. If optional, some user
preference may allow skipping over the content. But <prodnote
render='required'> is essential content for the user. An
audible cue could announce the presence of the <prodnote>.
-->

<!ATTLIST prodnote
%attrs;
imgref IDREFS #IMPLIED
render (required | optional) #REQUIRED
>

<!ELEMENT sidebar (%flow; | hd)* >

<!ATTLIST sidebar
%attrs;
render (required | optional) #REQUIRED
>

<!ELEMENT note (%block; | %inlineinblock;)+ >

<!ATTLIST note
%attrsrqd;
>

<!ELEMENT annotation (%block; | %inlineinblock;)+ >

<!ATTLIST annotation
%attrsrqd;
>

<!ELEMENT epigraph (%flow;)* >

<!ATTLIST epigraph
%attrs;
>

<!ELEMENT byline (%inline;)* >

<!ATTLIST byline
%attrs;
>

<!ELEMENT dateline (%inline;)* >

<!ATTLIST dateline
%attrs;
>

<!ATTLIST linegroup
%attrs;
>

<!ATTLIST poem
%attrs;
>

<!ELEMENT a (%inlinenoa;)* >

<!--Attuse: a

"href" value may have three forms:

1) '#idref', in the <dtbook>, to the element type having the
referenced id value;

2) 'uri', a uniform resource identifier to a resource, typically a
document, see [RFC2396], possibly restricted to work with only
the <dtbook> document content, as referenced content is
expected to be available on the same media, and off-media
references may not be available;

3) 'uri#xxx', in the resource uri, the element with id='xxx'.

Uses of the remaining attributes other than %attrs; are:

"type" is advisory content MIME type of the target, see [RFC2046];

"hreflang" is the language code of the href target, see [RFC4646];

"rel" is a list of forward link type(s), the relationship(s)
expressed by the href value to the target, space-separated
if multiple;

"rev" is a list of reverse link types, the relationship(s)
to this location from the href target, space-separated
if multiple;

"accesskey"=accessibility key character shortcut;

"tabindex"=tabbing order.

"external" An external link points to media content that is not part
of the DTB. The external media content must be rendered by an
external application, whether or not that content is renderable by
the DTB player.
-->

<!ATTLIST a
%attrs;
type %ContentType; #IMPLIED
href %URI; #IMPLIED
hreflang %LanguageCode; #IMPLIED
rel %LinkTypes; #IMPLIED
rev %LinkTypes; #IMPLIED
accesskey %Character; #IMPLIED
tabindex %Number; #IMPLIED
external (true | false) 'false'
>

<!ELEMENT em (%inline;)* >

<!ATTLIST em
%attrs;
>

<!ELEMENT strong (%inline;)* >

<!ATTLIST strong
%attrs;
>

<!ELEMENT dfn (%inline;)* >

<!ATTLIST dfn
%attrs;
>

<!ELEMENT kbd (%inline;)* >

<!ATTLIST kbd
%attrs;
>

<!ELEMENT code (%inline;)* >

<!--Attuse: code

See http://www.w3.org/TR/xml11/#sec-white-space for information on xml:space.
-->

<!ATTLIST code
id ID #IMPLIED
class CDATA #IMPLIED
title %Text; #IMPLIED
xml:space (default|preserve) 'preserve'
%externalNamespaces;
%i18n;
smilref CDATA #IMPLIED
%showin;
>

<!ELEMENT samp (%inline;)* >

<!--Attuse: samp

See http://www.w3.org/TR/xml11/#sec-white-space for information on xml:space.
-->

<!ATTLIST samp
id ID #IMPLIED
class CDATA #IMPLIED
title %Text; #IMPLIED
xml:space (default|preserve) 'preserve'
%externalNamespaces;
%i18n;
smilref CDATA #IMPLIED
%showin;
>

<!ELEMENT cite (%inline; | title | author)* >

<!ATTLIST cite
%attrs;
>

<!ELEMENT abbr (%inline;)* >

<!--Attuse: abbr

"title" value may expand that abbreviation.
-->

<!ATTLIST abbr
%attrs;
>

<!ELEMENT acronym (%inline;)* >

<!--Attuse: acronym

"title" value may expand that acronym.
"pronounce" value 'yes' indicates that the
acronym is pronounceable as a word (for example, NATO);
'no' that the acronym is best presented as a sequence
of letters (for examples, "XML" or "US").
-->

<!ATTLIST acronym
%attrs;
pronounce (yes | no) #IMPLIED
>

<!ELEMENT sub (%inline;)* >

<!ATTLIST sub
%attrs;
>

<!ELEMENT sup (%inline;)* >

<!ATTLIST sup
%attrs;
>

<!ELEMENT span (%inline;)* >

<!ATTLIST span
%attrs;
>

<!ELEMENT bdo (%inline;)* >

<!--Attuse: bdo

"lang" indicates the language of the content.

"dir" indicates the writing direction: 'ltr' is
left-to-right, 'rtl' is right-to-left.
-->

<!ATTLIST bdo
%coreattrs;
xml:lang %LanguageCode; #IMPLIED
dir (ltr | rtl) #REQUIRED
smilref CDATA #IMPLIED
%showin;
>

<!ELEMENT sent (%inlines;)* >

<!ATTLIST sent
%attrs;
>

<!ELEMENT w (%inlinew;)* >

<!ATTLIST w
%attrs;
>

<!ELEMENT pagenum (#PCDATA) >

<!--Attuse: pagenum

"page" allows three kinds of page numbering
schemes to be identified:
"front" (for roman-numeral pages at the start of a book),
"normal" (for pages identified by arabic numerals),
or "special" (for all other kinds of pages)
For multi-page continuous content,
such as large <img> or <table>, put the sequence of <pagenum> on
the page where that content starts.
-->

<!ATTLIST pagenum
%attrsrqd;
page (front | normal | special) 'normal'
>

<!ELEMENT noteref (#PCDATA) >

<!--Attuse: noteref

"idref" is a URI pointing to the target note, either within
same document or in another document. For example:
<noteref idref='#yyy'> refers to <note id='yyy'> within same document.
<noteref idref='foobar.xml#xxx'> refers to <note id='xxx'> in document foobar.xml

"type" provides advisory content MIME type of
the target, see [RFC2046].
-->

<!ATTLIST noteref
%attrs;
idref %URI; #REQUIRED
type %ContentType; #IMPLIED
>

<!ELEMENT annoref (#PCDATA) >

<!--Attuse: annoref

"idref" is a URI pointing to the target annotation, either within
same document or in another document. For example:
<annoref idref='#yyy'> refers to <annotation id='yyy'> within same document.
<annoref idref='foobar.xml#xxx'> refers to <annotation id='xxx'> in document foobar.xml

"type" provides advisory content MIME
type of the targeted id, see [RFC2046].
-->

<!ATTLIST annoref
%attrs;
idref %URI; #REQUIRED
type %ContentType; #IMPLIED
>

<!ELEMENT q (%inline;)* >

<!--Attuse: q

"cite" may provide a URI reference.
-->

<!ATTLIST q
%attrs;
cite %URI; #IMPLIED
>

<!ENTITY % Length "CDATA" >

<!ENTITY % MultiLength "CDATA" >

<!ENTITY % Pixels "CDATA" >

<!ELEMENT img EMPTY >

<!--Attuse: img

"src" specifies by URI the location of the image file.

"alt" is used to supply a short description of the <img>.

"longdesc" generally contains a pointer to a related
<prodnote> that contains a detailed description of the <img>.

The attributes "height" and "width" provide visual sizing
information, measured in pixels.
-->

<!ATTLIST img
%attrs;
src %URI; #REQUIRED
alt %Text; #REQUIRED
longdesc %URI; #IMPLIED
height %Length; #IMPLIED
width %Length; #IMPLIED
>

<!--Use: imggroup provides a container for one or more <img> and associated
<caption>(s) and <prodnote>(s). A <prodnote> may contain a description
of the image. The content model allows:

1) multiple <img> if they share a caption, with the ids of
each <img> in the <caption imgref="id1 id2 ...">,

2) multiple <caption> if several captions refer to a single
<img id="xxx"> where each caption has the same
<caption imgref="xxx">,

3) multiple <prodnote> if different versions are needed for different
media (e.g., large print, braille, or print). If several <prodnote>
refer to a single <img id="xxx">, each prodnote has the same <prodnote
imgref="xxx">.

4) one or more <pagenum> if the image group spans several pages.
-->

<!ELEMENT imggroup (prodnote | img | caption | pagenum)+ >

<!ATTLIST imggroup
%attrs;
>

<!ELEMENT p (%inline; | list | dl)* >

<!ATTLIST p
%attrs;
>

<!ELEMENT doctitle (%inline;)* >

<!ATTLIST doctitle
%attrs;
>

<!ELEMENT docauthor (%inline;)* >

<!ATTLIST docauthor
%attrs;
>

<!ELEMENT covertitle (%inline;)* >

<!ATTLIST covertitle
%attrs;
>

<!ELEMENT h1 (%inline;)* >

<!ATTLIST h1
%attrs;
>

<!ELEMENT h2 (%inline;)* >

<!ATTLIST h2
%attrs;
>

<!ELEMENT h3 (%inline;)* >

<!ATTLIST h3
%attrs;
>

<!ELEMENT h4 (%inline;)* >
<!ATTLIST h4
%attrs;
>

<!ELEMENT h5 (%inline;)* >

<!ATTLIST h5
%attrs;
>

<!ELEMENT h6 (%inline;)* >

<!ATTLIST h6
%attrs;
>

<!ELEMENT bridgehead (%inline;)* >

<!ATTLIST bridgehead
%attrs;
>

<!ELEMENT hd (%inline;)* >

<!ATTLIST hd
%attrs;
>

<!ELEMENT blockquote (pagenum | %block;)* >

<!--Attuse: blockquote

"cite" permits inclusion of the
URI from which the <blockquote> came.
-->

<!ATTLIST blockquote
%attrs;
cite %URI; #IMPLIED
>

<!ELEMENT dl (dt | dd | pagenum)+ >

<!ATTLIST dl
%attrs;
>

<!ELEMENT dt (%inline;)* >

<!ATTLIST dt
%attrs;
>

<!ELEMENT dd (%flow;)* >

<!ATTLIST dd
%attrs;
>

<!ELEMENT list (hd | prodnote | li | pagenum)+ >

<!--Attuse: list

"type" indicates whether the list items
<li> are ordered 'ol' or
unordered 'ul' or
preformatted 'pl'.

"depth" indicates nesting depth of lists within list items <li>
of ancestor lists, starting at 1.

"enum" indicates the kind of enumeration:
'1'=integer,
'a'=lowercase,
'A'=uppercase,
'i'=lowercase Roman, or
'I'=uppercase Roman.

"start" value indicates the ordered list integer ordinal, impllcitly 1,
to denote the first ordered list item. Its value is determined for the
indicated enum type. A start value is useful when there is an injection
of a different element type, such as <pageno>.
-->

<!ATTLIST list
%attrs;
type (ol | ul | pl) #REQUIRED
depth CDATA #IMPLIED
enum (1 | a | A | i | I) #IMPLIED
start CDATA #IMPLIED
>

<!ELEMENT li (%flow; | lic)* >

<!ATTLIST li
%attrs;
>

<!ELEMENT lic (%inline;)* >

<!--Attuse: lic

class attribute may be used to identify the particular
component of a list item <li>. For example, in a table of contents
class values might include "section", and "pagenumber".
-->

<!ATTLIST lic
%attrs;
>

<!-- The XHTML <table> model is used, including the presentational
attributes that have little meaning in Digital Talking Books,
but may be useful for concurrent display in different media.
That model is derived from IETF HTML table standard, see [RFC1942].

Note: The XHTML <table> model has been enhanced from HTML to allow
a simple <table> of one or more rows <tr>.
-->

<!ENTITY % Scope
"(row | col | rowgroup | colgroup)" >

<!ENTITY % TRules
"(none | groups | rows | cols | all)" >

<!ENTITY % cellhalign
"align (left|center|right|justify|char) #IMPLIED
char %Character; #IMPLIED
charoff %Length; #IMPLIED" >
<!-- % cellhalign
cellhalign sets horizontal alignment of content in
a table cell.

char indicates a character expected in each table
cell of a column that text should align on. The default is
the decimal point for the current language.

charoff sets the alignment offset of the first character
to align on, as specified with char.

Inheritance order for horizontal alignment is <th>|<td>,
<tr>, <thead>|<tfoot>|<tbody>, <col>, <colgroup>, default. The
default value is from the user agent, and may be affected
by user preference. The recommended default alignment for
<th> is center, and for <td> is left.
-->

<!ENTITY % cellvalign
"valign (top|middle|bottom|baseline) #IMPLIED" >
<!-- % cellvalign
valign sets vertical alignment of content in a table cell.

Inheritance order for vertical alignment is <th>|<td>, <col>,
<colgroup>, <tr>, <thead>|<tfoot>|<tbody>, default. The
default value is from the user agent, and may be affected
by user preference. The recommended default is middle.
-->

<!ELEMENT table (caption?, (col* | colgroup*), thead?, tfoot?, (tbody+| (tr|pagenum)+)) >

<!--Attuse: table

"summary" value provides a textual summary.

The attributes: "width", "border", "frame", "rules", "cellspacing",
and "cellpadding" provide visual presentation guidance. See their
explanation in the comment following those parameter entity
declarations.
-->

<!ATTLIST table
%attrs;
summary %Text; #IMPLIED
width %Length; #IMPLIED
border %Pixels; #IMPLIED
frame %TFrame; #IMPLIED
rules %TRules; #IMPLIED
cellspacing %Length; #IMPLIED
cellpadding %Length; #IMPLIED
>

<!ELEMENT caption (%flow;)* >

<!--Attuse: caption

"imgref" value (or space-separated
id values) identifies the id values of the <img>(s) to which the
caption applies. Only references
to images within the same document are allowed.
Note: imgref is an extension for dtbook.
-->

<!ATTLIST caption
%attrs;
imgref IDREFS #IMPLIED
>

<!ELEMENT thead (tr)+ >

<!ATTLIST thead
%attrs;
%cellhalign;
%cellvalign;
>

<!ELEMENT tfoot (tr)+ >

<!ATTLIST tfoot
%attrs;
%cellhalign;
%cellvalign;
>

<!ELEMENT tbody ((tr|pagenum)+) >

<!ATTLIST tbody
%attrs;
%cellhalign;
%cellvalign;
>

<!ELEMENT colgroup (col)* >

<!--Attuse: colgroup

"span" causes the attributes of one
col element to apply to more than one column, extending in the
writing direction for the language. Span is ignored if
any <col> are present.

"width" specifies the width of the columns, e.g.

width='64' width in screen pixels
width='0.5*' relative width of 0.5

"width" may contain a space- or comma-separated list of widths
for each <col>, or percentages if values end in '%', or relative
'nn*' to indicate a fractional portion of the remaining
widths after more explicit forms have been specified for other
columns in the colgroup (presumably the sum encompasses all
the columns). '0*' indicates minimal acceptable width based
on column content.

The <col> in a <colgroup> may inherit attribute values
from it, or the closest enclosing ancestor element:
<thead>|<tfoot>|<tbody>. If none, then the values are up to the
user agent, possibly by user preference.
-->

<!ATTLIST colgroup
%attrs;
span NMTOKEN '1'
width %MultiLength; #IMPLIED
%cellhalign;
%cellvalign;
>

<!ELEMENT col EMPTY >

<!--Attuse: col
"span" indicates how many columns the <col>
extends, in the writing direction of the <table>. The
attribute values apply to <th> and <td> that start in the
column, even if they extend into the next column(s), by
span value more than 1, and that next <col> may have different
attribute values.

"width" may contain a space- or comma-separated list of widths
for each <col> of the span, or percentages if values end in '%',
or relative 'nn*' to indicate a fractional portion of the remaining
widths after more explicit forms have been specified for other
columns in the span (presumably the sum encompasses all
the columns of this span, and of other <colgroup>s). '0*' indicates
minimal acceptable width based on individual <col> content.

<col> may inherit attribute values from the most immediate
enclosing ancestor element: <colgroup>, <thead>|<tbody>|<tfoot>.
If none, it is up to the user agent, possibly by user preference.
-->

<!ATTLIST col
%attrs;
span NMTOKEN '1'
width %MultiLength; #IMPLIED
%cellhalign;
%cellvalign;
>

<!ELEMENT tr (th | td)+ >

<!--Attuse: tr

%cellvalign; values provide default vertical alignment
values for <th> and <td> in the row, overriding any from
<thead>|<tbody>|<tfoot>, <col>, <colgroup>
If none, it is up to the user agent, possibly by user preference.

%cellhalign; values provide default horizontal alignment from
<thead>|<tbody>|<tfoot>
If none, it is up to the user agent, possibly by user preference.

-->

<!ATTLIST tr
%attrs;
%cellhalign;
%cellvalign;
>

<!ELEMENT th (%flownopagenum;)* >

<!--Attuse: th

The uses of attributes other than %attrs; %cellvalign; and %cellhalign;
are shown below. See [HTML401STRICT] for details and examples.

"abbr" provides an abbreviated name for a <th> cell that can be used
when referring to that <th> cell. Its default value is the cell content.

"axis" is used to place cells into conceptual categories in order to
provide improved access to information.

"headers" provides the id value(s), used with <td> cells, to reference
one or more cells with <th id="xxx"> that contain headings that
collectively describe or qualify the content of the cell, for example
<td headers="id1 id2">.

"scope" identifies one of
(row | rowgroup | column | colgroup)
to which the header information applies.

"rowspan" indicates the total number of rows that the cell extends, by
default 1. All spanned cells share these attributes.

"colspan" indicates the total number of columns the cell extends,
by default 1, in the writing direction of the table. All spanned
cells share these attributes.
-->

<!ATTLIST th
%attrs;
abbr %Text; #IMPLIED
axis CDATA #IMPLIED
headers IDREFS #IMPLIED
scope %Scope; #IMPLIED
rowspan NMTOKEN '1'
colspan NMTOKEN '1'
%cellhalign;
%cellvalign;
>

<!ELEMENT td (%flownopagenum;)* >

<!--Attuse: td

The uses of attributes other than %attrs; %cellhalign; and %cellvalign;
are shown below. See [HTML401STRICT] for details and examples.

"abbr" provides an abbreviated name for a <th> cell that can be used
when referring to that <th> cell. Its default value is the cell content.

"axis" is used to place cells into conceptual categories in order to
provide improved access to information.

"scope" identifies one of
(row | rowgroup | column | colgroup)
to which the header information applies.

"rowspan" indicates the total number of rows that the cell extends, by
default 1. All spanned cells share these attributes.

"colspan" indicates the total number of columns the cell extends,
by default 1, in the writing direction of the table. All spanned
cells share these attributes.
-->

<!ATTLIST td
%attrs;
abbr %Text; #IMPLIED
axis CDATA #IMPLIED
headers IDREFS #IMPLIED
scope %Scope; #IMPLIED
rowspan NMTOKEN '1'
colspan NMTOKEN '1'
%cellhalign;
%cellvalign;
>