Guidelines for the conversion of print 
originals to DTBook XML






Version 1.3
2nd October 2007


Table of contents
About the guidelines	4
Guideline contents	4
Example collection	4
A short description of TPB’s production process	4
1	About the standard	5
1.1	The DAISY Consortium and the DAISY standard	5
1.1.1	DTBook	5
1.1.2	Which version of the standard is to be used?	5
2	Guidelines for scanning text and images	6
2.1	Scanning	6
2.2	OCR and proofreading	6
2.3	Images from scanned print originals	6
2.3.1	General	6
2.3.2	Handling of specific image types	6
2.3.3	Colour images	7
2.3.4	Greyscale images	7
2.3.5	Images containing text	7
2.3.6	Editing images	7
2.3.7	Delivery	8
3	General guidelines for mark up	8
3.1	TPB applies a subset of the DTBook standard	8
3.2	A complement to the DTD	8
3.3	Elements allowing both text and elements as children (%flow)	9
3.3.1	Nested lists	9
3.4	Placement of ”floating” elements	10
3.5	The XML file in general	10
3.5.1	File name	10
3.5.2	Processing instructions	10
3.5.3	XML declaration	10
3.5.4	Document type declaration	10
3.5.5	Character encoding	10
3.5.6	xml:lang	11
3.5.7	Representation of punctuation, typographic characters and special characters.
	11
3.5.8	Hyphenation	13
3.5.9	Empty elements	13
3.6	Explanation of wording	14
4	Detailed guidelines for markup	15
4.1	Guidelines for mark up of structural elements	15
4.1.1	dtbook	15
4.1.2	head	15
4.1.3	book	15
4.1.4	frontmatter	15
4.1.5	bodymatter	17
4.1.6	rearmatter	17
4.1.7	level1-6	17
4.2	Guidelines for the mark up of block elements - listed alphabetically	19
4.2.1	annoref & annotation	19
4.2.2	blockquote	19
4.2.3	doctitle	19
4.2.4	docauthor	19
4.2.5	div and the class=”pgroup” attribute	19
4.2.6	h1 - h6	21
4.2.7	imggroup	21
4.2.8	meta	23
4.2.9	note & noteref	24
4.2.10	p	25
4.2.11	pagenum	26
4.2.12	poem	27
4.2.13	sidebar	27
4.3	Guidelines for mark up of inline elements – listed alphabetically	27
4.3.1	br	27
4.3.2	code	27
4.3.3	em	27
4.3.4	strong	28
4.3.5	sub	29
4.3.6	sup	29
4.4	Guidelines for the mark up of lists	29
4.4.1	list	29
4.4.2	dl	31
4.5	Guidelines for the mark up of tables	32
4.5.1	Table	32
4.5.2	Caption	33
4.5.3	tr	33
4.5.4	th	33
4.5.5	td	33
5	Guidelines for the mark up of unified content	33
6	Guidelines for the notation and mark up of mathematics, physics and simpler chemistry
	33
6.1	Mark up of formula based content	34
6.1.1	Block	34
6.1.2	Inline	34
6.2	Notation	35
7	Changes	36
Version 1.3 - 2007-10-02	36
Authors:	37


About the guidelines
These guidelines describe those quality requirements established by TPB to be applied by 
suppliers producing text in DTBook XML format from print originals.
Guideline contents
The guidelines are divided into five sections:

1.	About the standard
2.	Guidelines for scanning text and images
3.	General guidelines for mark up
4.	Detailed guidelines for mark up
5.	Guidelines for the mark up of unified content
6.	Guidelines for the notation and mark up of mathematics, physics and simpler 
chemistry
Example collection
A collection of examples has been established illustrating principles described within these 
guidelines for a range of different material. The collection will be expanded continually 
during the contract period as solutions to problems are established. 

The collection is available in both .PDF and .xml formats and is distributed by TPB.
A short description of TPB’s production process
Since 2005, TPB has had a unit producing files in DTBook XML format. Said production has 
required the development of a process similar in detail to those services to be purchased by 
TPB. 
The following is a short description of TPB’s production process:

1.	Removal of the print original’s binding.
2.	a) Printed pages converted, via scanning, to PDF format or
 b) Printed pages converted, via scanning, directly to RTF format. This method 
involves an OCR process that is suitable for only the simplest material.
3.	PDF file zoned, zones then OCR processed using Abbyy FineReader 8.0. 
Resulting files saved in Microsoft Word 2003 .doc format.
4.	The saved file is examined for a subset of DTBook elements that are marked 
up using formatting templates for Microsoft Office Word 2003 developed at 
TPB.
5.	The edited file is converted to a basic DTBook XML.
6.	The basic DTBook XML is refined in Altova XMLSpy, utilising a number of 
XSL transformations and operator oversight, until the required level of 
DTBook conformity and quality is achieved.
7.	Late-stage correction is performed using the Authentic view in XMLSpy. 
Authentic can also be used as a free-standing program


1	About the standard
1.1	The DAISY Consortium and the DAISY standard
The DAISY Consortium assumed the role of maintenance agency for the DAISY 3 standard 
in 2005, the formal name for the standard is DAISY/NISO z.39.68-2005. The standard has 
been published in its entirety and can be found at http://www.daisy.org/z3986/2005/z3986-
2005.html
1.1.1	DTBook
DTBook is that part of the DAISY 3 standard describing the text content of a DAISY book. 

The use of the DTBook standard means a radical change for TPB. For the first time 
documents that are richly marked up using a homogenous format can act as a source 
document for the production of all text based media at TPB. While the DTBook file was 
developed for the production of DAISY talking books, it can also be used, though not 
exclusively, for the following media:
*	Braille volumes
*	TPB’s proprietary e-book format Textview
*	Automatic conversion to HTML, PDF and other e-book formats.

The DTBook standard has two main influences:
*	The Chicago Manual of Style, the de facto-standard for the design of printed materials 
and the naming of elements of content
*	XHTML 1.0, the accepted standard for mark up of web pages, developed by the World 
Wide Web Consortium
The DTBook standard is not meant to be a blanket solution for all possible content. It is, 
instead, a collection of the most common and applicable elements present in texts. In 
particular, teaching materials tend to include content that does not lend itself to accurate mark 
up using the elements available in the DTBook standard. Examples of such content are 
included in these guidelines.
1.1.2	Which version of the standard is to be used?
Production on behalf of TPB shall use the current DAISY standard, i.e. DAISY/NISO 
z.39.68-2005. See also www.daisy.org



2	Guidelines for scanning text and images
2.1	Scanning
The properties of print originals that affect scanning will vary from title to title and settings 
require adapting to suit each case. Two passes, a text specific and an image specific, might be 
necessary to achieve necessary quality, depending on the scanner being used (See Section 2.3 
Scanning and editing images). Text content, regardless of background colour, must be 
scanned in as a black and white image. Images must be reproduced in accordance with the 
print original. The majority of print originals can have their binding removed to facilitate 
scanning. The exception to this regards titles that have been loaned to TPB as opposed to 
being purchased. These need to be scanned using a flatbed scanner. A closer description of the 
handling of titles loaned to TPB will be included with relevant orders.
2.2	OCR and proofreading
OCR (Optical Character Recognition) involves the conversion of images of text to machine-
editable text.

Text produced using OCR must be proofread. The aim of proofreading is to reduce the 
amount of errors introduced into the text by the scanning and OCR processes. Proofreading 
does not include the correction of spelling errors present in the print original. TPB places no 
demands for the correction of obvious errors in the print original, leaving this decision to each 
producer. The resulting digital text is expected to mirror the print original (with reservation 
for the preceding sentence) to 100%. This requirement results in the right for TPB to return 
material found to have proofreading errors.
2.3	Images from scanned print originals
To facilitate a simplified handling of images in print originals the following guidelines are to 
be applied:
2.3.1	General
The practice of including images as a supplement to text content is prevalent.  The relevance 
of included imagery to the text in question cannot be judged generally though three elements 
can be used as a guide:
*	What is the nature of the text?
-	Social sciences, Humanities, Biography and so on
*	Is the image directly referenced in the text?
*	Is the image necessary for the understanding or appreciation of the text?
2.3.2	Handling of specific image types
Four types of image can occur in print originals:
*	Informational – i.e. images pertaining to the subject of the text that represent facts in a 
visual manner, e.g. diagrams and charts
*	Biographical-Ornamental – i.e. images that while they are not informational in the 
manner of diagrams or charts are connected to and may even be referenced by the text, 
e.g. portrait and still life photography
*	Logotypes –  a design or symbol used by a company to advertise its products
*	Vignettes – graphics used to separate sections or chapters, decorate borders or 
flyleaves, jackets or panels
*	Iconic – i.e. an image that stands for its object by virtue of a resemblance or analogy to 
it 
*	Formatting – i.e. images that have no connection to the subject matter and are purely 
an artefact of layout and design

Handling of the preceding types of images should take place as follows:
*	Informational – should be scanned and linked to via DTBook markup
*	Biographical-Ornamental - should be scanned and linked to via DTBook markup
*	Logotypes – images of this type are not to be included in the DTBook markup and 
consequently not delivered as images
*	Vignettes – images of this type are not to be included in the DTBook markup and 
consequently not delivered as images
*	Iconic
o	Inline – where iconic images occur in an inline context they are considered to 
be crucial to the understanding of the text.  In these cases producers of 
DTBook files are required to contact the ordering entity for advice
o	Other – iconic images not occurring in an inline context should be judged for 
their relevance to the text.  In these cases producers of DTBook files are 
recommended to contact the ordering entity for advice
*	Formatting – images of this type are not to be included in the DTBook markup and 
consequently not delivered as images
2.3.3	Colour images
Colour images present in the print original are to be reproduced with no observable 
degradation in colour rendering.
2.3.4	Greyscale images
Greyscale images present in the print original are to be reproduced without introducing visible 
compression artefacts, e.g. banding.
2.3.5	Images containing text
Images in the print original containing a preponderance of text should be reproduced using the 
same properties applied for scanning of text in general, optimally 400dpi though no less than 
300dpi.  Results must be subsequently proofread to ensure reproduction quality. 
2.3.6	Editing images
Scanned images often require some degree of editing.
*	Images are to be cropped, leaving them free of all text external to the image itself, e.g. 
captions, headers and footers etc.
*	Images skewed as a result of the scanning process must be rectified. 
2.3.7	Delivery
1.	Images size
o	All reasonable attempts should be made by producers to maintain consistency 
of size between resulting images and their source in the print original
o	Resulting images must not exceed 600 pixels in width regardless of orientation 
or the preceding point
2.	The ratio of height to width in resulting images must agree with the original.
3.	Images must be delivered in JPEG format. 
3	General guidelines for mark up
Marking up text using an XML grammar involves the placement of so-called tags around 
defined elements within a document. Text elements might have particular attributes that need 
to represented, this can be achieved by including them in the relevant element tag. Rules for 
element tags and their allowed attributes are collected in a ’Document Type Definition’ or 
’DTD’. Elements present in a text include, though not exclusively:
*	headings
*	page numbers
*	paragraphs
*	notes
*	tables
*	lists
*	sidebars
3.1	TPB applies a subset of the DTBook standard
TPB intends to apply a subset of those elements and attributes found in the DTBook standard 
that are described in the DAISY Structure Guidelines. All elements selected in this subset are 
described in Section 4 Detailed guidelines for markup. 

An exception to this concerns production requiring adaptation of a pedagogic nature, 
primarily production on behalf of SIT – The Swedish Institute for Special Needs Education. 
Such production may include instructions for the supplier directing the use of particular 
attribute values for certain types of content.

In those cases where the DAISY Structure Guidelines recommend markup that has not been 
described by TPB a choice of element/attribute providing the most applicable markup should 
be made. 
3.2	A complement to the DTD
The dtbook-2005-2.dtd, used to validate markup produced for TPB, allows a great deal of 
leeway in its application; this is a condition of its format. In the majority of cases, a document 
requires further editing after passing validation. The following example is typical, TPB 
requires that roman numerals be present in the text node (i.e. the text contained within 
element tags) of <pagenum page=”front”> markup. The DTD is unable to do so and can 
therefore not be used to validate such a requirement. See Section 4 Detailed guidelines for 
markup.

The DTD also exists as a so-called LiveDTD, a hypertext document that provides a more 
user-friendly version of the standard. A LiveDTD för dtbook-2005-1.dtd can be found at 
http://se.daisy.org/riktlinjer/. Conversion software to create a LiveDTD can downloaded from 
http://www.sagehill.net/livedtd/download.html
3.3	Elements allowing both text and elements as children (%flow)
Another example of leeway within the structure of the DTD concerns %flow in elements, e.g.:
*	<sidebar>
*	<list>
*	<table>
Block and inline elements are required not to be combined in an inappropriate manner, even 
when allowed by the DTD.

Two examples of acceptable allowed markup using <sidebar>:

Example 1:

<sidebar>
Oh, see. Oh, see Jane. Funny, funny Jane.
</sidebar>
 

Example 2: 

<sidebar>
<p> Oh, see. Oh, see Jane. Funny, funny Jane.</p>
</sidebar>

An example of  valid to the DTD though unacceptable markup using sidebar:

<sidebar>
Oh, see. Oh, see Jane. <p>Funny, funny Jane.</p>
</sidebar>
The example above is also valid according to the DTD, though mixing text and block 
elements. Yet it is semantically unclear and such mark up is to be avoided.
3.3.1	Nested lists
A list item may contain text and a new list. However, as above, this is undesirable and instead 
the <p> element should be employed as follows: 

<list>
  <li>Text</li>
  <li><p>Text</p>
    <list>
      <li>Text</li>
      <li>Text</li>
    </list>
  </li>
</list>
3.4	Placement of ”floating” elements
The correct placement of <table>, <sidebar>, <imggroup> and <annotation> in 
DTBook documents is not a given. It might be the case that an image or table 'floats', i.e. it is 
external to the flow of the text. In such cases, relevant elements are to be placed at their most 
logical contextual anchor point within the text flow. If such an anchor point cannot be 
identified the relevant elements are to be positioned as close as possible to the closure tag of 
the current containing mark up. Such relocation must not break DTD rules for placement and 
must attempt to maintain optimal conditions for reading comprehension.
3.5	The XML file in general
3.5.1	File name
DTBook files produced on behalf of TPB must be given the production number provided by 
TPB when ordered and have the .xml extension. File names and paths must exclusively 
contain the following characters: 0-9, a-z, underscore, and hyphen.
3.5.2	Processing instructions
Processing instructions, e.g. style sheet paths, must not be included in the delivered file.
3.5.3	XML declaration
The following XML declaration must be used:
<?xml version="1.0" encoding="utf-8"?>
3.5.4	Document type declaration
The following document type declaration must be included:
<!DOCTYPE dtbook PUBLIC "-//NISO//DTD dtbook 2005-2//EN" 
"http://www.daisy.org/z3986/2005/dtbook-2005-2.dtd">
NOTE: The abovementioned declaration may change as the standard is updated.
3.5.5	Character encoding
DTBook documents produced on behalf of TPB must be saved using the UTF-8 character 
encoding and not include a byte order mark (BOM).
3.5.6	xml:lang
Longer text extracts (at least one whole sentence) whose language is not that stated in the root 
element are required to be marked up with the xml:lang attribute. Single words or shorter 
inline quotes are not required to be marked up using the xml:lang attribute.

Examples of elements that may require the xml:lang attribute:

<blockquote>
<p>
<list>
<table>
<sidebar>
<note>
<annotation>

Segments of the print original such as references, indices and appendices, may include terms, 
tiles, names and the like that also deviate from the documents main language. These do not 
require marking up with the xml:lang attribute.
3.5.7	Representation of punctuation, typographic characters and special characters.
This section describes those requirements placed upon producers with regard to the 
representation of particular types of character sensitive to the scanning process and OCR.

TPB requires producers to contact ordering entities for a clarification of requirements when 
punctuation, typography and special characters beyond the scope of these guidelines occur.

This section of the guidelines may be expanded and/or updated to facilitate the effective 
production of DTBook files.
3.5.7.1	Introduction
Due to the nature of production based on print originals errors may be introduced into 
DTBook files.  TPB considers the accurate conversion of all characters present in print 
originals of the highest importance.  A particular group of characters has shown itself to be 
vulnerable to error:  punctuation, typographic  and special characters.
3.5.7.2	Punctuation
TPB requires that particular care be paid to the accurate representation of the following 
punctuation characters and any whitespace surrounding them:

1.	Minus character (U+2212)
2.	Hyphen character (U+2010)
3.	Figure dash character (U+2012)
4.	En dash character (U+2013)
5.	Em dash character (U+2014)
6.	Quotation dash character (U+2015)
3.5.7.2.1	Punctuation character representation for TPB
TPB allows producers the following choice:

*	that certain of the abovementioned group of characters be substituted or
*	that the abovementioned group of characters be correctly represented in the resulting 
DTBook file in their entirety.

Note that TPB does not require Producers to interpret the usage of the listed characters in 
those circumstances where typographer choice differs from usage descriptions provided 
herein.
3.5.7.2.2	Substitution
TPB requires the following representation:

*	The Minus character (U+2212), Hyphen character (U+2010) and Figure dash 
character (U+2012) can acceptably be represented with Hyphen-Minus (U+002D).   
However, if producers choose to make such a substitution TPB requires that all 
instances of each of the abovementioned characters be substituted in the file; 
producers are required to ensure that mixing of methods does not occur.
*	En dash characters are required to be represented using En dash (U+2013).
*	Em dash characters are to be represented using Em dash (U+2014).
*	Quotation dash characters (U+2015) can acceptably be represented using En dash 
(U+2013). 
However, if producers choose to make such a substitution TPB requires that all 
instances of each of the abovementioned characters be substituted in the file; 
producers are required to ensure that mixing of methods does not occur.
3.5.7.2.3	Non-substitution
TPB requires that the following characters be accurately represented:

7.	Minus character (U+2212)
8.	Hyphen character (U+2010)
9.	Figure dash character (U+2012)
10.	En dash character (U+2013)
11.	Em dash character (U+2014)
12.	Quotation dash character (U+2015)
3.5.7.3	Typographic characters
TPB requires producers to contact ordering entities for a clarification of requirements when 
punctuation, typography and special characters beyond the scope of these guidelines occur.
3.5.7.4	Special characters
3.5.7.4.1	Arrows
Print originals containing arrows in the text are required to have such characters represented 
accurately.  Producers are allowed to represent such characters using the appropriate Unicode 
code point from the span U+2190 to U+2199.
3.5.7.5	Other
3.5.7.5.1	Phonetics
TPB requires producers to contact ordering entities for a clarification of requirements when 
phonetic symbols are used in print originals.
3.5.7.5.2	Pictograms, Ideograms and Logograms
TPB requires producers to contact ordering entities for a clarification of requirements when 
characters belonging to the headed groups occur; commonly Chinese, Japanese or Korean, 
though hieroglyphics are also an example of such.
3.5.8	Hyphenation
Hyphenation, occurring due to line breaks or page breaks at margins in the print original, is 
expected to be corrected in the resulting XML document.
3.5.9	Empty elements
Empty elements may not occur in the XML document, exceptions to this are: 
<img>
<br>
<meta>
<link>
<col>
<th>
<td>
<dd>


3.6	Explanation of wording

Wording
Explanation
<p>
The element name is coloured red and bracketed by “less 
than” and “greater than” characters.
Attribute
Attribute names are coloured red.
”attribute value”
Attribute values are written with red text and quotation 
marks.
h[x] 
Elements and possible variables – brackets contain possible 
values
”should”
Implies compliance by suppliers, though invites judgement 
calls based on best practice. Non-compliance implies no 
recourse to return of material.
”is to be”, ”shall”, ”require”
Implies compliance without the need for judgement by 
suppliers. Non-compliance invites return of delivered 
materials as incomplete.

<p>Text</p>

Code examples are enclosed in a box.



4	Detailed guidelines for markup
4.1	Guidelines for mark up of structural elements
4.1.1	dtbook
The <dtbook> element is the DTBook documents root element and consequently contains all 
other element tags.

Attributes:
version, xml:lang, xmlns

The following values are to be used for the version, xml:lang and xmlns attributes
*	version="2005-2". This value may change if TPB adopts an updated version of the 
standard for production purposes. TPB will inform suppliers of any such change 
during the contract period.
*	xml:lang ("sv","en-GB" etc.). TPB requires the use of the IETF RFC 3066 standard 
for language identification.  
If the language is included in the ISO 639-1 standard TPB requires the use of the two-
letter code. If the language is not included TPB require the use of the three-letter code 
found in the ISO 639-2 standard.  
Print originals containing a mix of languages should be given an xml:lang attribute 
reflecting the majority language; this may often be reflected in that language 
represented by the title of the print original. 
English requires a more detailed two part description, i.e. the requisite language code 
must be coupled with the appropriate country code to differentiate between American 
and British English, i.e.: xml:lang="en-US"/ xml:lang="en-GB" 
Print originals containing both British and American forms of English and not having 
a discrete balance of usage are allowed to be marked up with the language code only, 
i..e. xml:lang="en"
*	xmlns. The following namespace is required: 
http://www.daisy.org/z3986/2005/dtbook/
4.1.2	head
The <head> element comes immediately after the <dtbook> and contains the documents 
metadata, <meta>.
4.1.3	book
The <book> element comes after (though at the same level of hierarchy as) <head> inside 
<dtbook>. <book> can be compared to <body> in HTML and contains the contents of the 
print original.
4.1.4	frontmatter
<frontmatter> contains that information presented before core content in the print original, 
e.g. copyright information, foreword, preface, table of contents and the like. 

To simplify production of adapted materials from the DTBook document TPB requires that 
content contained in preliminary pages is structured in a uniform manner, see Sections 4.1.4.1 
to 4.1.4.4 below. 
4.1.4.1	Title and author
The first elements contained by <frontmatter> must always be <doctitle> followed by 
<docauthor>. This applies regardless of how the print originals title page or cover is 
formatted. See Section 4.2 Guidelines for mark up of block elements - listed alphabetically.
4.1.4.2	Title page
Details of the title, author, and publisher can often be found repeated at the beginning of the 
print original. Such repeated content is superfluous is to be excluded from the XML 
document. <doctitle> and <docauthor> contain the necessary information.
4.1.4.3	Colophon and subsequent content
The print original’s colophon, dedication and the like are to be marked up using the <level1> 
element and one of the class attribute values described in Section 4.1.7.1 class attributes for  
<level1>.  

The colophon is required to have details specific to the printing of the work removed.  Typical 
texts to be removed are:
*	Publisher adverts
*	Publisher contact details
*	Details of paper type and quality
*	Printer’s details including location
*	Number combinations detailing printing year 

Details that are required to be retained in the colophon are:
*	Publisher details
*	Copyright texts
*	Copyright year
*	Print year
*	Edition
*	ISBN
*	Previously published works
*	Translator
o	In those cases where details concerning the translator of a work are found 
elsewhere, e.g. on the title page, the text is to be moved to the colophon
*	Editor
*	Illustrator/photographer
4.1.4.4	Introduction
If an ”Introduction” or material of a similar nature is numbered in accordance with the print 
originals other headings then such material should be included in the <bodymatter> 
element. Otherwise, this type of material is required to be included in the <frontmatter> 
element.
4.1.5	bodymatter
The <bodymatter> element contains the printed materials core content. This can be defined 
as the parts, chapters sub-headings found in the print original. Note that content such as 
epilogue, conclusions and the like are to be contained within the <bodymatter> element.
4.1.6	rearmatter
<rearmatter> shall contain all content appended to the core text. Examples of material that 
can be found in <rearmatter> include, though not exclusively: appendices, glossaries, 
references, notes and indices.
4.1.6.1	Back cover and dust jacket copy
Both the back cover copy and copy placed on the dust jacket flaps of the print original are to 
be included in <rearmatter> and be marked up using <level1> and a class attribute 
according to the following:

<level1 class=”backCoverText”>
<p>[rear cover copy]</p>
</level1>
<level1 class=”other”>
<p>[dust jacket flap copy]</p>
</level1>

No heading is required to be introduced.

Dust jacket copy is required to be included in the same level and included after any back 
cover text.  An exception to this is copy consisting of advertising and/or review.  Text about 
or by the author is required to be included, as is an image of the author.
4.1.7	level1-6
The <level[x]> element represents the heading levels of the print original. <level[x]> 
elements are to be associated with their respective headings in the print original, if such exist. 
[x] mirrors the structure of the headings and must have a value from the span 1 to 6.

The <level> element , i.e. the level element not assigned a structure number, is not to be 
used.

A relatively common layout practice is the inconsequent use of fonts when setting headings. 
This practice leads to discrepancies between a print original’s actual structure as opposed to 
the visual representation of it’s structure. For example, an introduction may have a font that is 
applied to sub-headings in the remaining text.  Application of <level[x]> must not represent 
this habit – attempts will produce DTD invalid markup. Note also that use of otherwise empty 
<level[x]> elements as a container for inappropriate <level[x]> mark-up in an attempt to 
achieve validity is also disallowed.
4.1.7.1	class attributes for  <level1>
The class- attribute is required to be applied to the <level1> element and is used to describe 
the type of content contained by the element. The exception to the above concerns print 
original’s utilising a primary structure of Part [x]. In such cases the class attribute is also to 
be applied to <level2> elements in <bodymatter>.
4.1.7.1.1	frontmatter
The following values are to be applied to the class attribute in <level1> contained in 
<frontmatter>:
*	”briefToc” – used to identify any table of contents providing an outline in those 
cases where two table of contents occur
*	"colophon" –used to identify content in the beginning of the print original pertaining 
to the publisher, printer and the like
*	”dedication” – used to identify author/editor dedications in the print original
*	"glossary" – used to identify glossaries, lists of abbreviations and the like
*	"introduction" – used to identify an introduction not numbered in accordance with 
the print original’s other chapter numbering (see below)
*	"other" – used to identify content within <level1> contained in <frontmatter> not 
corresponding to the above, e.g. ”List of figures” and so on
*	"preface" – used to identify an author/editor preface
*	"toc" – used to identify the print original’s table of contents
4.1.7.1.2	bodymatter
The following values are to be applied to the class attribute in <level1> and <level2> 
contained in <bodymatter>:
*	"part" – used to identify content when a primary structure of Part [x] is used
*	"chapter" – used to identify chapter structure and any introduction included in the 
chapter numbering scheme
*	"introduction" – used to identify an introduction not numbered according to the 
chapter numbering scheme. See Section 4.1.4.4 Introduction
*	"other" – used to identify content within <level1> contained in <bodymatter>, 
e.g. prologue, epilogue and the like
o	Note that in shorter works that do not contain a great deal of hierarchical 
structure, in particular children’s literature, it may be the case that only one (1) 
<level1> can be defined in the <bodymatter>.  Such should be marked 
using the "other" attribute in the following manner: <level1 
class="other">
4.1.7.1.3	rearmatter
*	The following values are to be applied to the class attribute in <level1> contained in 
<rearmatter>:
*	"appendix" – used to identify appendices 
*	"backCoverText" – is to be used to identify back cover and dust jacket copy. Dust 
jacket copy is required to be included in the same level and included after any back 
cover text.  An exception to this is copy consisting of advertising and/or review.  Text 
about or by the author is required to be included, as is an image of the author.
*	"bibliography" – used to identify a bibliography
*	"colophon" – used to identify content the print original, pertaining to the publisher, 
printer and the like, that is on occasion included in back matter
*	"footnotes" and "rearnotes" – used to identify a section of notes
*	"glossary" – used to identify glossaries
*	"index" – used to identify indices
*	"other" – to be used to identify <level1> elements in <rearmatter>
*	"toc" – used to identify the print original’s table of contents
4.2	Guidelines for the mark up of block elements - listed 
alphabetically
4.2.1	annoref & annotation
The <annoref> and <annotation> elements resemble the <noteref> and <note> 
elements. However, annotation (considered a floating element) is to be used when marking up 
shorter marginalia in the print original. A reference number or character (<annoref>) is 
associated, principally found in text parallel to the marginalia. Se also <noteref> and 
<note>.
4.2.1.1	annoref
The <annoref> element is required to have the following attribute:
- idref, this refers to the id attribute applied to the <annotation> element. Note that the 
value must be prefixed with the # character to fulfil URI requirements.
4.2.1.2	annotation
Each <annotation> is required to have at least one associated <annoref>.
4.2.1.2.1	Placement of annotation
See Section 3.4. Placement of ”floating” elements
4.2.2	blockquote
The <blockquote> element is used to mark up quotes broken out of the text flow, i.e. 
constituting their own paragraph. Note that the xml:lang attribute is required to be applied to 
the <blockquote> element if the quote is in a different language than that identified in the 
root element. Inline quotes are not required to be marked up with <q> though are required to 
be reproduced in accordance with the print original, e.g. with <em> or quotation marks.
4.2.3	doctitle
The first element within <frontmatter> must be <doctitle> and use the following content 
format: [title] – [sub-title].
Any text pertaining to the edition of the work is not to be included.
4.2.4	docauthor
The second element within <frontmatter>, after <doctitle>, must be <docauthor> and is 
required to use the following content format: [author/editor first name + other names + 
surname]. If the print original has more than one author/editor, each name is to be marked up 
using individual <docauthor> elements. Comma separation or other text before or between 
author names (e.g. ”by”, ”and”) is not to be included, regardless of usage in the print original. 
The following markup is required in those instances where an editor fulfils the role of author:

<docauthor>Gregory Stones (ed.)</docauthor>
4.2.5	div and the class=”pgroup” attribute
It is common practice to utilise indentation or empty lines to indicate a shift between 
paragraphs. It is the case that authors may apply both practices in an attempt to associate 
groups of paragraphs. TPB require this formatting to be marked up.

The <div> element and the applied class=”pgroup” attribute are to be used to identify all 
such formatting contained within any <level[x]> element. The following example uses <p> 
elements associated using indentation in the print original. Note that content association can 
be made using other methods.

<level[x]>
<h[x]>[Heading]</h[x]>
<div class=”pgroup”>
<p>[Paragraph]</p>
</div>
(empty line)
<div class=”pgroup”>
<p>[Possible indented paragraph]</p>
<p>[Indented paragraph]</p>
<p>[Indented paragraph]</p>
</div>
(empty line)
<div class=”pgroup”>
<p>[Possible indented paragraph]</p>
<p>[Indented paragraph]</p>
<p>[Indented paragraph]</p>
</div>
</level[x]>
4.2.5.1	Exceptions to class=”pgroup” mark up
The <div class=”pgroup”> element is not to be used if a <level[x]> element does not 
contain associated <p> elements.

It is common for empty lines to separate paragraphs from other types of content, e.g. sidebars, 
quotes and the like. <div class=”pgroup”> is not to be used in these cases. 
For example:

<level[x]>
<h[x]>Heading</h[x]>
<p>[Possible indented paragraph]</p>
<p>[Indented paragraph]</p>
<p>[Indented paragraph]</p>
(empty line)
<blockquote/>
(empty line)
<p>[Possible indented paragraph]</p>
<p>[Indented paragraph]</p>
<p>[Indented paragraph]</p>
</level[x]>
4.2.6	h1 - h6
The <h1>-<h6> elements are used to identify headings in the print original. Note that 
<h[x]> must be contained within their respective <level[x]> element.
Typographic formatting of a heading is not required to be mirrored in markup. An exception 
to this rule concerns single words in headings that are italicised or bold, this are to be marked 
up using <em> and <strong> respectively.
This also applies to sidebar, list and table headings (<hd>, <h[x]> and <caption>).

For example:

<h2>A word is <em>italicised</em></h2>

Standard rules apply for note or annotation references that appear in headings.

For example:

<h2>A word has a <noteref>note reference</noteref> in the heading</h2>

In cases where the author of a text is included directly under the heading TPB requires that 
these are marked up using <p>.  Such text is not to be included in the preceding <h[x]> 
element.

Chapters not possessing a heading in the print original are required to be identified with the 
appropriate <level[x]> element and do not require markup with <h[x]>.
4.2.7	imggroup
An image, its caption, and any associated image description must always be contained within 
the <imggroup> element. Exceptions to this rule are, for example, icons or images without 
captions or image descriptions that are contained within table cells. 
An example of <imggroup> mark-up:
<imggroup>
<img alt="image" src="[x]).jpg" />
<caption>Caption text</caption>
<prodnote render="optional">Image description</prodnote>
</imggroup>
4.2.7.1	Placement of imggroup
TPB requires that the <imggroup> element not be placed in an inline context, e.g. contained 
within the <p> element. It follows then that an image placed inline, i.e. in the text itself, of 
the print original should be relocated when the requisite <imggroup> mark-up is made. The 
most appropriate placement in the majority of cases is directly after end of paragraph. See 
also Section 3.4 Placement of ”floating” elements
Images stretching over two (2) pages are to be scanned as an individual image per page, i.e. 
images are required to be divided in two at the point of  change from recto to verso.  The 
following mark-up is to be employed:

<imggroup>
<img alt="image" src="[x].jpg" />
<caption>Caption text</caption>
<prodnote render="optional">Image description</prodnote>
</imggroup>
<pagenum id="page-[x]" page="normal">[x]</pagenum>
<imggroup>
<img alt="image" src="[x].jpg" />
<caption>Caption text</caption>
<prodnote render="optional">Image description</prodnote>
</imggroup>

NOTE: Remain aware of those requirements contained in Section 3.4 Placement of ”floating” 
elements
4.2.7.2	img
The <img> element requires the following attributes:
alt which must include the value=”image”
src which must include a reference to an image file.  
Images must be located in the same catalogue as the DTBook file.  
Filenames for images must only contain the following characters: 0-9, a-z, underscore and 
hyphen.
4.2.7.3	caption
The <caption> element contains that text associated with an image. <caption> must always 
be placed directly after the appropriate <img> element. If the <caption> text describes a 
series of images it must be placed after the last image in the series.
4.2.7.4	prodnote
A <prodnote> element contained within an image group is used to mark up an image 
description. However, standard practice is to provide image descriptions after production of 
the DTBook file. TPB therefore requires that a "dummy" text be placed within <prodnote> 
elements for image groups. The <prodnote> element is to have the render=”optional” 
attribute applied and is considered mandatory where image group mark-up occurs. The 
<prodnote> is to be marked up as follows:
<prodnote render=”optional”>Image  description</prodnote>
It is important to remember not to mix both block and inline mark up in elements that can 
contain %flow. See Section 3.3 Elements allowing both text and elements as children 
(%flow).
Image description prodnotes are always placed last in image groups, after any <caption> 
that may occur.
4.2.8	meta
The <meta> element contains information about the DTBook document. Meta information is 
used for identification purposes and to aid index compilation.
4.2.8.1	Table of attributes, contents and schemes

Attribute
name=””
content=””
scheme=””
Content
"dtb:uid"
[identification]

"dc:Title"
[title] : [subtitle]

"dc:Creator"
[surname], [first name] 
[other name]

"dc:Date"
[Date of file completion]
[YYYY]-[MM]-[DD]
"dc:Publisher"
[ordering entity]

"dc:Language"
[xx(x)](-[XX])

“tpb:Supplier”
[Company Name]

“tpb:SuppliedDate”
[Date of file completion]
[YYYY]-[MM]-[DD]
“dc:Identifier”
[identification]
[ordering entity]

4.2.8.2	Meta element format:
<meta name="dtb:uid" content="[identification]"/> Contains the document’s 
identification and is supplied by TPB when production is ordered.

<meta name="dc:Title" content="[title] : [subtitle]"/> Contains the title of the print 
original and is to be taken from order details provided by TPB.  In cases where more than one 
author is present TPB require that a <meta> element be inserted for each author.

<meta name="dc:Creator" content="[surname], [first name] [other name]"/> 
Contains the authors name and is to be formatted "surname, first name". Individual <meta> 
mark up is to be provided for each author if more than one is indicated.

<meta name="dc:Date" content="[YYYY]-[MM]-[DD]"/> Indicates the date production 
of the DTBook file is completed, to be formatted as follows: Year (four figures) – Month (two 
figures) – Day (two figures)

<meta name="dc:Publisher" content="[TPB]"/> Used to indicate the ordering entity.

<meta name="dc:Language" content="[xx(x)](-[XX])"/> Used to indicate the main 
language of the document. TPB requires the use of the IETF RFC 3066 standard for language 
identification. If the language is included in the ISO 639-1 standard TPB requires the use of 
the two-letter code. If the language is not included TPB require the use of the three-letter code 
found in the ISO 639-2 standard. English requires a more detailed two part description, i.e. the 
requisite language code must be coupled with the appropriate country code to differentiate 
between American and British English, i.e.: xml:lang="en-US"/ xml:lang="en-GB".

<meta name=”tpb:Supplier” content=”[Company Name]” /> Included only once and 
must contain the name of the contracted supplier of the DTBook file.

<meta name=”tpb:SuppliedDate” content=”[Date of file completion]” 
scheme=”YYYY-MM-DD” /> Included once on initial delivery of the DTBook file.  In the 
event of returns of faulty DTBook files to the Supplier for correction TPB require that an 
additional instance be included with the new delivery date as it’s content.  Deletion of the 
initial instance is not allowed.  An additional instance of this element is required for each 
return of a faulty DTBook file for correction.

Certain DTBook file productions, details of which are supplied when ordered, require the 
following <meta> elements:
<meta name="dc:Identifier" content="[identification]" scheme="[ordering 
entity]"/>
The value of [identification] is supplied when ordered and can differ from the [identification] 
value for dtb:uid.
4.2.9	note & noteref
Notes are comprised of two connected elements, <noteref> and <note>. The actual text of 
the note is marked up using the <note> element and the note reference is marked up using 
the <noteref> element.

Three types of note can be identified in the print original:
*	notes placed in the footer of the page
*	collected at the end of the respective chapter
*	in a note section appended to the main text
Note references to the above occur within the text.
4.2.9.1	noteref
The following attributes are to be applied to the <noteref> element:
- idref refers to the notes id attribute. Note that the value must be prefixed with the # 
character to fulfil URI requirements.
- class this must contain one of the following values, "endnote" for notes occurring at the 
end of a chapter or "rearnote" for notes appended to the main text.
Example:
<noteref idref="#fn_2_1" class="rearnote">1</noteref>

4.2.9.1.1	Note references stating an interval
Note references stating an interval are to be handled as follows; each number in the interval is 
to be marked up with an individual <noteref>, though separators (hyphens, commas and the 
like) are not to be included. 
For example:

Print original:
”In the midway1-3 of this our mortal life…”

DTBook:
In the midway<noteref>1</noteref><noteref>2</noteref>
<noteref>3</noteref> of this our mortal life…
4.2.9.2	Note
Each <note> element is required to have at least one associated <noteref> element.

The following attributes are required to be applied to <note> elements:
- id refers to an associated idref attribute applied to a <noteref> element.
- class is required to have the same value as that applied to the <noteref> element.
4.2.9.3	Placement of notes
All notes present in the DTBook document, excepting chapter notes (endnotes), are to be 
placed last in <rearmatter> in an appended section of notes. If rear cover copy is present, 
the note section should be placed before this. 
4.2.9.3.1	Footnotes and table footer notes
Notes placed in, for example, page footers or table footers are to be converted to rearmatter 
notes in the DTBook document, <note class=”rearnote”>. These notes are to be placed in 
a section appended to the main text in the rearmatter section. If rear cover copy is present, the 
note section should be placed before this. This note section is to be marked up separately 
(<level1 class=”rearnotes”>).  Notes are required to be sorted in the same order as their 
respective initial <noteref>.

Example of a note section:

<level1 class=”rearnotes”>
<note class=”rearnote” id=”fn_1”>
<p>Note text</p>
…
</level1>
4.2.9.3.2	Endnotes
As stated, notes already placed at the end of a chapter are not required to be moved.
4.2.9.3.3	Notes originally  placed in back matter (rearnotes)
Notes already present in the back matter of the print original are not required to be moved 
though they are to be marked up using <note class=”rearnote”>. The note section itself is 
to be marked up using <level1 class=”rearnotes”>, existing headings are to be preserved 
and no new headings inserted.
4.2.10	p

The <p> element identifies a paragraph.
4.2.11	pagenum
The <pagenum> element identifies the change from recto to verso and consequently the 
change in pagination, the text node contains, barring exceptions, the page identification; 
almost exclusively a number. 
All pages present in the print original included in the pagination, even those not provided with 
a printed number, are required to be marked up with the <pagenum> element. 
TPB’s requirements are illustrated by the following; if page 148 is followed by an un-
numbered page which is in turn followed by page 150 TPB requires that the un-numbered 
page be numbered 149.
4.2.11.1	An exception with regard to flyleaves and redundant content
An exception to Section 4.2.11 concerns print originals containing flyleaves and/or 
superfluous duplicate information about author and title (see Section 4.1.4.2 Title page).  Such 
pages occurring are not to be identified with <pagenum>.  Generally speaking mark-up with 
<pagenum> should begin at <level1 class=”colophon”>.
4.2.11.2	Attributes
page with the values "front", "normal", "special"
id with the value format "page-[page number]"

Three types of page are described in the DTBook standard: front, normal and special.

The attribute value "front" is to be used where the print original utilises roman numerals in 
its front matter. Roman numerals are to retain that letter case used in the print original. 
<pagenum page=”front”/> is not allowed outside of the <frontmatter/> element; though 
<frontmatter/> may contain <pagenum page=”normal”/>.

The attribute value "normal" is to be used for pages occurring in the body and back matter of 
the print original.

The "special" attribute value is rarely applicable. An example is an appendix not numbered 
in a standard manner, e.g. a1, a2, a3 and so on. The "special" value is also to be used in 
those cases where un-numbered inserts, for example suites of photography, occur; these are to 
be marked up as follows:

<pagenum page=”special” id=”unnum-[page no.]”>Un-numbered page</page>

In those cases where pagination of a text cannot be effectively represented using the above 
rules the Supplier is required to contact TPB.
4.2.11.3	Placement of pagenum
The <pagenum> element can be used as either block or inline and can therefore be placed 
precisely where the print original changes page regardless of content mark-up, e.g. within a 
paragraph, list or table.
The <pagenum> element is also to be placed within words breaking over pages. It is 
allowed though, to place the <pagenum> element directly after words hyphenated due to 
page change.  If such a relocation of the <pagenum> element is made TPB requires that 
eventual  hyphens, with reservation for those found in compound words, be removed.  
Relocated <pagenum> elements should always be preceded by a space character.

When a page begins with a <level[x]> element the <pagenum> element is to be placed 
between the <level[x]> element and the <h[x]> element.

I de fall en boks sidnumrering är väldigt speciell och inte inkluderas i ovanstående 
resonemang, t.ex. om sidnumren repeteras (t.ex. s. 1-18 för Del 1, s. 1-33 för Del 2 och s. 1-28 
för Del 3), ska producenten kontakta TPB för vägledning om uppmärkningen.
4.2.11.4	Print originals free of pagination
Un-paginated print originals, particularly shorter children’s literature and picture books, are 
required to be provided with pagination.  Note that standard verso-recto order is to be 
maintained.
4.2.12	poem
The <poem> element is to be used to mark up poems, song texts and the like.

The following elements are to be used where appropriate within the <poem> element:
<hd>, <linegroup> and <line>.
4.2.13	sidebar
The <sidebar> element is used to mark up sidebar content, i.e. content that while 
contextually coherent may graphically or positionally differentiated from the main flow of the 
text. The <sidebar> element can contain the majority of other block elements, e.g. p, list, 
table and so on. Headings occurring within the <sidebar> element are to be marked up using 
the <hd> element.
4.2.13.1	Attribute
TPB requires that sidebar has the render attribute applied. The render attribute is to have 
the value ”optional”.
4.2.13.2	Placement of sidebar
See Section 3.4 Placement of ”floating” elements.
4.3	Guidelines for mark up of inline elements – listed alphabetically
4.3.1	br
The <br/> element identifies a line break. It is to be used as an exception and only where 
appropriate structural mark-up is unavailable.
4.3.2	code
The <code> element is used to identify program code.
4.3.3	em
The <em> element identifies italicised text. 
The <em> element is not to be used to emphasize headings in those instances where the use 
of italics is entirely typographic, though individual words in headings that are italicised are 
required to be marked up.
4.3.3.1	Handling spaces and punctuation in conjunction with <em>
Two usages of  italicisation occur that can generate mark-up errors when using <em>, partial 
italicisation of text within a sentence and italicisation of an entire sentence.

In those cases where only parts of a sentence are italicised TPB requires the following rules to 
be applied when using <em>:
*	The start <em> tag must not be followed by any whitespace character, new line, 
carriage return or punctuation
*	The closing <em> tag must not be preceded by any whitespace character, new line, 
carriage return or punctuation

The following example illustrates correct usage:

<p>This is a sentence <em>where part of the text</em> is italicised.  This sentence has no 
italicisation.</p>

In those cases where an entire sentence is italicised TPB requires the following rules to be 
applied when using <em>:
*	The start <em> tag must not be followed by any whitespace character, new line, 
carriage return or punctuation
*	The closing <em> tag must not be preceded by any whitespace character, new line or 
carriage return
*	The closing <em> tag must be preceded by associated punctuation

<p>This sentence is not italicised.  <em>This is sentence is completely italicised.</em>   
This sentence is not italicised.</p>
4.3.4	strong
The <strong> element identifies bold text. 
The <strong> element is not to be used to emphasize headings in those instances where the 
use of italics is entirely typographic, though individual words in headings that are italicised 
are required to be marked up.
4.3.4.1	Handling spaces and punctuation in conjunction with <strong>
Two usages of  italicisation occur that can generate mark-up errors when using <strong>, 
partial italicisation of text within a sentence and italicisation of an entire sentence.

In those cases where only parts of a sentence are italicised TPB requires the following rules to 
be applied when using <strong>:
*	The start <strong> tag must not be followed by any whitespace character, new line, 
carriage return or punctuation
*	The closing <strong> tag must not be preceded by any whitespace character, new 
line, carriage return or punctuation

The following example illustrates correct usage:

<p>This is a sentence <strong>where part of the text</strong> is in boldface.  This 
sentence does not contain bold text.</p>

In those cases where an entire sentence is italicised TPB requires the following rules to be 
applied when using <strong>:
*	The start <strong> tag must not be followed by any whitespace character, new line, 
carriage return or punctuation
*	The closing <strong> tag must not be preceded by any whitespace character, new 
line or carriage return
*	The closing <strong> tag must be preceded by associated punctuation

<p>This sentence does not contain bold text.  <strong>This is sentence is completely in 
boldface.</strong>  This sentence does not contain bold text.</p>
4.3.5	sub
The <sub> element identifies subscripted text, for example H2O is required to be marked up 
as H<sub>2</sub>O.
Mark-up with <sub> must not be contain any whitespace character, new line, carriage return 
or punctuation.
4.3.6	sup
The <sup> element identifies superscripted text, for example X2 is required to be marked up 
as X<sup>2</sup>.
Mark-up with <sup> must not be contain any whitespace character, new line, carriage return 
or punctuation.
Note that use of <sup> to identify note references is incorrect, see Section 4.2.9 note & 
noteref
4.4	Guidelines for the mark up of lists
4.4.1	list
The <list> element is used to mark up lists and organised registers. A list item may or may 
not be identified with a number or bullet of some kind.

A list item can consist of more than one paragraph (<p>). 

The <list> element must not occur within the <p> element. The <p> element is to be closed 
before beginning <list> mark-up and a new <p> element opened, if necessary, on completion 
of the <list> mark-up.

Borderline cases occur, for example, where an apparent list stretches over a number of pages, 
where <list> mark-up is inappropriate. Such cases must be judged as and when they occur, 
with the readers understanding of the text guiding any decision made.

Lists of terms and definitions are not to be marked up using <list>. In these cases, i.e. 
definition lists, the <dl> element is required to be used. See Section 4.4.2 dl.

Standard lists consist mainly of two or more <li> elements. Where headings occur they are to 
be marked up using the <hd> element.

Lists can contain the <pagenum> element.
4.4.1.1	Attributes
The <list> requires the following attributes to be applied:

The type attribute identifies three varieties of list: 

*	”ol” identifies ordered lists; each item is preceded by an alphanumeric character in 
series. If ”ol” is applied, TPB requires that the alphanumeric character be removed 
from the text node of the list item.
o	The enum attribute is required by TPB for lists applying type=”ol” and has 
the values:
?	”1”  for numbers
?	”a” for lower case characters
?	”A” for upper case characters
?	”i” for lower case roman numerals and
?	 ”I” for upper case roman numerals.
o	Note that ”ol” lists without a specified enum value default to enum=”1”
*	”ul” identifies un-ordered lists; each item is preceded by a symbol. If ”ul” is applied, 
TPB requires that the symbol and any associated whitespace be removed from the text 
node of the list item.
*	”pl” identifies lists not possessing any type of  leading character, e.g. bullets or 
numerals.  Typical examples of this kind of list are a table of contents and indexes 
In certain circumstances ”pl” is to be used to identify lists that do possess leading 
characters where it is important to preserve the specific character in use.  For example, 
if  the text refers to particular symbol types used ”pl” is required to be used in such 
cases instead of ”ul” or ”ol” with the character and associated whitespace preserved.
4.4.1.2	Content to be marked up with list
The following content is required to be marked up with <list>:
*	Table of contents, see Section 4.4.14 lic
*	Indices – TPB requires that these be marked up as one list with nesting as appropriate.  No 
nesting is expected for those lists divided alphabetically
*	References – TPB requires that these be marked up as one list, no nesting is required to 
identify reference type
4.4.1.3	li
The <li> element is to be used to mark up individual items in a list.
4.4.1.4	lic
For tables of contents, brief tables of contents, lists of illustrations, lists of tables and the like 
TPB requires the use of the <lic> element for mark up of constituent parts of a list item. 
Indices and other registers do not require the use of the <lic> element. The purpose of such 
mark-up is to differentiate between a heading and its page number. This is achieved with <lic 
class=”entry”> and <lic class=”pagenum”> respectively.
See the following example:

<level1 class="toc">
<h1>Content</h1>
<list type="pl">
<li><lic class="entry">Chapter 1</lic><lic class="pagenum">1</lic></li>
<li><lic class="entry">Chapter 2</lic><lic class="pagenum">15</lic></li>
<li><lic class="entry">Chapter 3</lic><lic class="pagenum">30</lic></li>
</list>
</level>
4.4.1.5	hd
The <hd> element is used to mark up headings not associated with a level element. 
Occurrence of this type of heading is generally limited to the <list> and <sidebar> 
elements. 
4.4.2	dl
The <dl> element is used to mark up lists of terms and their definitions, examples range from 
glossaries to lists of acronyms and the like. 

Definition lists are generally made up of coupled <dt> and <dd> elements.

<dl>
<dt>Term</dt>
<dd>Definition</dd>
</dl>

The <dl> element is not allowed in the <p> element. Preceding <p> elements are required to 
be closed before marking up definition lists. A new <p> should be opened, if necessary, on 
completion of the definition list mark-up.
4.4.2.1	dt
The <dt> element is used to mark up those terms occurring in definition lists.
4.4.2.2	dd
The <dd> element is used to mark up definitions corresponding to relevant terms in a 
definition list. In those cases where empty definitions occur, essentially in lists that illustrate 
that a term has no definition, the following mark-up should be used, <dd/> or <dd></dd>. 
Such mark up may also occur where definition lists apply empty tags for formatting purposes.
4.5	Guidelines for the mark up of tables
4.5.1	Table
The <table> element is required when content in the print original is presented in tabular 
format or table-like format. Suppliers are required to judge whether table content can be 
defined as informative and illustrative or simply pictorial, i.e. whether the table in question 
provides the reader with information necessary for understanding an argument or can be seen 
as incidental. In those cases where table content is considered incidental TPB require that the 
content be scanned as an image and provided with a <prodnote>. An illustration of the 
above is a train timetable where, unless specific cells are referenced by the text, the content 
can acceptably be included as an image with relevant image description mark-up.

Table captions require mark up with the <caption> element, this element is placed directly 
after the <table> element regardless of placement in the print original.
4.5.1.1	Empty table cells
The following practice is to be applied when marking up empty table cells occurring in the 
print original, either <td/> or <td></td>.
4.5.1.2	Tables covering more than one page
In those instances where tables stretch over more than one page the following mark up is to be 
applied: the <pagenum> tag is placed within a <p> tag that is itself contained within the 
first <td> or <th> element on the new page.  Note that the text node of the <td> or <th> 
must also be included within the <p> tag.
For example:

<td>
<p><pagenum id="page-[x]" page="[x]">[x]</pagenum>
Text node</p>
</td>

4.5.1.3	Placement of tables
See Section 3.4. Placement of ”floating” elements.
4.5.1.4	Table notes
Notes found in or associated with tables are required to be moved to a note section according 
to that practice outlined in section 4.2.9 note & noteref.
4.5.1.5	Tables within tables
Tables occurring within table cells are not required to be marked up using the <table> 
element. TPB requires that such content be marked up using elements appropriate to the 
nested table’s content, e.g. <p>, <list> and so on.
4.5.2	Caption
The following mark up is to be applied where two captions occur in a table: a <caption> 
element is opened, each caption is then contained with a <p> element, and the <caption> 
element is then closed. 
4.5.3	tr
The <tr> element is used to mark up table rows.
4.5.4	th
The <th> element is used to mark up column or row headings. In those cases where a <th> 
element stretches over more than one column or row, TPB requires that the colspan and 
rowspan attributes be applied. 
4.5.5	td
The <td> element is used to mark up table cell content. 
In those cases where a <td> element stretches over more than one column or row TPB 
requires that the colspan and rowspan attributes be applied. 
5	Guidelines for the mark up of unified content
Certain print originals contain text and image content that are integrated to such a degree that 
separation of the two is considered overly problematic. In such cases of unified content TPB 
requires that the entire image, including text content, be captured as a JPEG file. TPB also 
recommends that the entire page be captured if this is considered optimal.

The <caption> element is considered unlikely to occur in such material, though TPB 
requires inclusion where necessary.

The <prodnote> element used for image descriptions continues to be required as in standard 
image mark-up. Text content is required to be included using applicable mark-up.

Note that TPB requires that marked up text content be placed before the <imggroup> 
element. Examples of such mark-up are included in the Example Collection.
6	Guidelines for the notation and mark up of 
mathematics, physics and simpler chemistry
Structurally complex formulae found in mathematics, physics, chemistry and other disciplines 
using such, are to be handled using the notation currently applied on behalf of TPB and SIT 
for production of Textview formatted titles. 

This facilitates production and delivery of different e-text formats, e.g. Textview and Full text 
DAISY DTB.

The DTBook file can also be utilised Braille production, though a great deal of editing is 
required to realise this.

The mathematical notation currently in use has been developed for both the sighted and 
visually-impaired who use tactile presentation, and while requirements placed by speech 
synthesis have been taken into account current notation is a compromise.

The notation described in these guidelines cannot be used to express formulae in all 
circumstances. Structurally complex formulae unsuitable to notation are required to be 
scanned in and marked up as images, see Section 4.2.7 imggroup. In certain circumstances, in 
particular for the blind, tactile images may also be produced to complement the finished 
adapted title.

Future production requiring adaptation of mathematics, physics, chemistry, and the like will 
be based entirely on available XML grammars, specifically MathML. A project aimed at 
delivering production methodologies for the above has been started by TPB, successful 
development also been made by the DAISY Consortium. A proposed goal for the TPB project 
is delivery of MathML solutions for production in time for the next purchasing round.

Production of DTBook documents containing formula based content will require manual 
editing in the post-scanning process.
6.1	Mark up of formula based content
All occurrences of formula based content to be adapted using notation is to be marked up 
according to the following:
6.1.1	Block
Formulae formatted as paragraphs, i.e. separated with blank lines, are to be marked up as 
follows:
<p class=”tpb-specialnotation”>

Example:

 

<p class=”tpb-specialnotation”>
sum{k=0;3;(-2)^k}= (-2)^0 + (-2)^1 + (-2)^2 + (-2)^3 = -5
</p>


6.1.2	Inline
Formulae found in an inline context are to be marked up as follows:
<span class=”tpb-specialnotation”>
Example:

 

<p>And that being able to simplify oneself – simplify. Let us say that we are developing an 
algorithm or implementing someone else’s. Let us also say that we have also arrived at a point 
when a certain task is to be performed regarding whether <span class=”tpb-
specialnotation”>y ej= z</span> or whether <span class=”tpb-specialnotation”>x 
= 0</span> or whether both <span class=”tpb-specialnotation”>x ej= 0</span> and 
<span class=”tpb-specialnotation”>y = z</span>. At this point it is a benefit if one is 
conversant with the rules of logic so that one can assure oneself that this task should 
<em>always</em> be performed.</p>
6.2	Notation
The notation required by TPB/SIT is defined in Appendix 5 Mathematical Notation.

TPB require that a dialogue be established between the supplier and TPB in those instances 
where the notation defined in Appendix 2 does not cover formulae found in the print original 
but suppliers judge that conversion is necessary and possible.
Note that the notation described in Appendix 5 Mathematical Notation and Section 7 
Guidelines for the notation and mark up of mathematics, physics and simpler chemistry, are 
not required unless specifically ordered.  In those cases where no order of notation is made 
but simpler mathematics and the like appear in the print original it is required to be marked up 
using appropriate elements, e.g. <sup> for powers.


7	Changes
Version 1.3 - 2007-10-02
2.3.2 Handling of specific image types 
Clarification regarding image types.
2.3.7 Delivery 
Clarification regarding image size.
3.5.7 Representation of punctuation, 
typographic characters and special 
characters. 
Re-definition of punctuation handling.

4.1.4.3 Colophon and subsequent 
content
Clarification of content that should be retained or 
removed.
4.2.11 pagenum
New text requiring Suppliers to contact TPB with 
regard to special pagination.

4.2.13 sidebar
Clarification.



Authors:
Linus Ericson 
Markus Gylling 
Joel Håkansson 
Tomas Johansson 
Jesper Klein 
Cecilia Mattsson 
Gita Nabavi 
Björn Nyqvist 
Anne Stigell 
Richard Stones 
Björn Westling
Guidelines for the conversion of print originals to DTBook XML
.	
1

4
TPB, The Swedish Library of Talking Books and Braille
Guidelines for conversion of print originals to DTBook xml	
TPB – The Swedish Library of Talking Books and Braille	37