Overview

• The main points from yesterday
• Suggested solution to yesterday's exercise
• Schema languages for XML
• DTD syntax
• Linking DTDs to XML documents
• Exercise

XML—repetition

• A markup language marks up the structure of documents
  • Enabling computers to process the documents
  • Separating structure and content from presentation
• The XML specification:
  1. Rules and syntax of well-formed XML documents
  2. Describes how to create a markup language with a Document Type Definition (DTD)
     • An XML-based markup language is called an application of XML
• XML documents are made up of elements and attributes
• An XML processor must reject XML documents which are not well-formed

Suggested solution to yesterday's exercise

reviews.xml

Schema

• Formally describes an XML application (i.e. a markup language)
  • Which elements and attributes can be used
  • Constraints on the structure and content of document instances
• Some XML processors can check if a document conforms to a schema
  • An XML document that does, is called a valid document
• A schema can be used by XML editors, "forcing" the document writer to create a valid document

Different schema languages

• DTD (Document Type Definition)
  Approved by W3C (part of the XML specification)
• XML Schema Language
  Approved by W3C (a recommendation in May 2001)
• RELAX NG
• Schematron

XHTML—an XML application

• Extensible HyperText Markup Language
• A W3C Recommendation
• "HTML reformulated in XML"
  • Lowercase characters in element names (<h1>, </h1>)
  • Empty elements must be closed (<br />)
  • Attribute values must be quoted (<table border="0">)
• Web browsers can read and display XHTML documents
• Defined as a DTD: http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd (XHTML 1.0 Strict)

DTD

• A set of declarations
  • Declares which elements can be used in the language
  • Declares the content model of each element
  • Declares which attributes can be used for the different elements, and their legal values
  • Declares entities
• The order of the declarations is of no importance
• All XML documents must be well-formed, but validity is not mandatory
  • An XML processor should not reject a well-formed XML document, even if it's not valid

Our example document

<?xml version="1.0" encoding="UTF-8"?>
<books>
   <book number="b1">
      <title>How to Build a Digital Library</title>
      <authors>
         <author>
            <first_name>Ian H.</first_name>
            <family_name>Witten</family_name>
         </author>
         <author>
            <first_name>David</first_name>
            <family_name>Bainbridge</family_name>
         </author>
      </authors>
      <year>2003</year>
      <publisher>Morgan Kaufmann Publishers</publisher>
      <isbn>1-55860-790-0</isbn>
   </book>
   <book number="b2">
      <title>Understanding Digital Libraries</title>
      <edition>2</edition>
      <authors>
         <author>
            <first_name>Michael</first_name>
            <family_name>Lesk</family_name>
         </author>
      </authors>
      <year>2005</year>
      <publisher>Morgan Kaufmann Publishers</publisher>
      <isbn>1-55860-924-5</isbn>
   </book>
</books>

A possible DTD for this document

<!ELEMENT books (book+)>
<!ELEMENT book (title, edition?, authors, year, publisher, isbn)>
<!ELEMENT title (#PCDATA)>
<!ELEMENT edition (#PCDATA)>
<!ELEMENT authors (author+)>
<!ELEMENT author (first_name, family_name)>
<!ELEMENT first_name (#PCDATA)>
<!ELEMENT family_name (#PCDATA)>
<!ELEMENT year (#PCDATA)>
<!ELEMENT publisher (#PCDATA)>
<!ELEMENT isbn (#PCDATA)>

<!ATTLIST book number ID #REQUIRED>

Element declarations

• The general syntax:
  <!ELEMENT element_name content_model>
• The most common content_models are:
  • Parsed character data
  • Child elements
  • Mixed content
  • Empty elements

Element declarations: Parsed character data

• Parsed character data can include:
  • Plain characters
  • General entities, like &lt;
    The XML processor will replace the entities with their full meaning in the document tree
• Is called #PCDATA in a DTD
  • Example: <!ELEMENT first_name (#PCDATA)>
  • The #PCDATA entries in DTDs must be surrounded by parentheses

Element declarations: Child elements

• Elements which contain no character data, but only child elements
• Example:
  <!ELEMENT author (first_name, family_name)>
  • The elements in the XML document must be in exactly the order specified in the DTD, meaning that this is not valid:
    <author>
       <family_name>Witten</family_name>
       <first_name>Ian H.</first_name>
    </author>
  • The frequency of child element instances can be specified by a suffix
    • ?: Zero or one element instances are allowed
    • *: Zero or more element instances are allowed
    • +: One or more element instances are allowed
    • With no suffix, exactly one element instance is allowed
  • Example: <!ELEMENT book (title, edition?, authors, year, publisher, isbn)>

Element declarations: Mixed content

• Elements with content that can contain parsed character data mixed with child elements
• Markup example: <paragraph>But Finnis<note>2</note> states that the works of Rushdie are not postmodern. Marx uses the term '<term>the deconstructive paradigm of reality</term>' to denote not, in fact, structuralism, but poststructuralism.</paragraph>
• The DTD element declaration: <!ELEMENT paragraph (#PCDATA | note | term)*>

Element declarations: Empty elements

• Empty elements have no content, but can have attributes
• Example (forced line break in XHTML): <!ELEMENT br EMPTY>
  • In markup: <p>This is a<br/>line break.</p>

Attribute declarations

• The general syntax:
  <!ATTLIST element_name attribute_name type description
  ...>
• The most common attribute types:
  • CDATA: Character data, including whitespace
  • NMTOKEN: name token—Character data, excluding whitespace
  • ID: Unique id numbers (primary keys)
  • IDREF: Reference to unique id numbers (foreign keys)
  • Enumerations: a list of possible values
• The most common descriptions:
  • #IMPLIED: The attribute is optional
  • #REQUIRED: The attribute is required
  • Literal: a default value is given as a quoted string

Attribute declarations—some examples

• ID: <!ATTLIST book number ID #REQUIRED>
  • Valid markup (start tag): <book number="b1">
    • The number attribute must be unique in the document
  • Not valid markup: <book>
• CDATA: <!ATTLIST homepage type CDATA #IMPLIED>
  • Valid markup: <homepage type="personal">
  • Also valid: <homepage>
• Enumeration: <!ATTLIST homepage lang (no | en) "no">
  • Valid markup: <homepage lang="en">
  • Valid markup: <homepage>
    • The XML processor will add the attribute value no
  • Not valid markup: <homepage lang="ge">

General entities

• You can define your own entities for content that is used many times in documents
  • Works like the predefined character entities
• General syntax: <!ENTITY entity_name replacement_text>
• Example: <!ENTITY hioa "Oslo and Akershus University College of Applied Sciences">
• In markup: <message>Welcome to &hioa;</message>
  • The XML processor will replace every occurrence of the entity with the replacement text

Advanced topics

• Parameter entities
  • For repeated text in DTD declarations
  • Example: <!ENTITY % block "p | ul | blockquote">
  • Usage: <!ELEMENT chapter (h1, (%block;)+)>
  • Result: <!ELEMENT chapter (h1, (p | ul | blockquote)+)>
  • Used a lot in the XHTML DTD!
• Modularisation
  • Importing modules from external sources
  • Conditional includes (INCLUDE and IGNORE)
  • XHTML has many modules, enabling inclusion of parts of the language (like the character entities or table definitions) in other XML applications

A generic XML document

1. <?xml version="1.0"?>
2. <!DOCTYPE book_reviews SYSTEM
      "http://www.jbi.hio.no/book_reviews.dtd"
3. [
      <!ENTITY part1 SYSTEM "part1.xml">
      <!ENTITY ouc "Oslo University College">
   ]>
4. <book_reviews>
      <!-- The actual document -->
   </book_reviews>

• 1, 2 and 3 together is the document prolog
• 2 and 3 together is the document type declaration

1. XML declaration
2. External subset
3. Internal subset
4. The root (or document) element

The document type declaration

• Two parts: External and internal subset
  • Both optional
  • Declarations in the internal subset take precedence over declarations in the external subset
• We will only use an external DTD in our exercise:

  <?xml version="1.0" encoding="UTF-8"?>
  <!DOCTYPE books SYSTEM "books.dtd">
  <books>
     <book number="b1">
        <title>How to Build a Digital Library</title>
        <authors>
           <author>
        ...
     </book>
     <book number="b2">
        ...
     </book>
  </books>

Two topics from yesterday

• Namespaces
  • Namespaces are used to prevent naming conflicts when elements from two different XML languages are combined in one document
    • The XML languages are defined in a schema language, like DTD
• The standalone attribute in the XML declaration
  • Example: <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
    • no: The XML processor has to read external files to make the XML document complete, i.e. there are general entities and/or default values for attributes defined in an external DTD
    • yes: The XML document does not depend on external declarations

Exercise

• Create an external DTD for the book reviews XML document
  • Use XML Copy Editor to create and edit the XML document
  • Save the file in the same directory as the XML document
  • Reference the DTD in the document prolog of the XML document
  • Check that the XML document is valid according to the DTD
• Tip: Start with the declaration of the root element