Text Data Description Languages

Document Type Definitions (DTDs)

Developed in SGML

• Collection of entities
• Element and attribute declarations
• Element "content model" declared by regular expression
• Simplistic id-idref constraints
• Atomic content restricted to strings => no other datatypes
• Predates namespaces

See movies.dtd

Namespaces

mechanism to identify names uniquely:

• URI as unique identifier
• prefix as local name

See Bray's example

XML Schema

3-part specification: primer, structures, datatypes

• 44 Simple types (atomic content)
• "nil" (e.g., <phone loc="work" xsi:nil="true"/>)
• Complex types (element structures)
  • Derivation by extension
  • Derivation by restriction
• Element and attribute declarations map names to types
• More flexible constraint definitions
  • minOccurs .. maxOccurs
  • ordered vs. unordered substructure
  • nillable
  • uniqueness with respect to a scope
    <xs:unique name="uniqueMovie">
      <xs:selector xpath="./movie"/>
      <xs:field xpath="title"/>
      <xs:field xpath="yearReleased"/>
    </xs:unique>
  • key-keyref allows scoping and compound keys
• Named attribute groups

See movies.xsd

Relax NG

ISO standard (Dec. 2003), developed by James Clark and Murata Makoto through OASIS

http://www.relaxng.org/

• Simpler than XML Schema, but with support for datatypes, namespaces
• Element "content model" and attribute values modelled as tree-regular grammars
• More expressiveness for unordered and mixed content

Two syntaxes:

XML Syntax

<grammar xmlns="http://relaxng.org/ns/structure/1.0">

<start>

<ref name="AddressBook"/>

</start>

<define name="AddressBook">

<zeroOrMore>

<ref name="Card"/>

</zeroOrMore>

</define>

<define name="Card">

<element name="name">

<text/>

</element>

<element name="email">

<attribute name "workOrHome">

<choice>

<value>work</value>

<value>home</value>

</choice>

</attribute>

<text/>

</element>

<optional>

<element name="note">

<text/>

</element>

</optional>

</define>

</grammar>

Compact Syntax

grammar {

start = AddressBook

AddressBook = element addressBook { Card* }

Card = element card { Name, Email, element note { text} ? }

Name = element name { text }

Email = element email { attribute workOrHome { ("work" | "home") }, text }

}

Relax NG specification in Relax NG

References and related reading

Querying XML, Chapter 4, 5, 7.3, 7.4

Bray98
W3C Namespaces	Bray99
W3C XML Schema0	W3C XML Schema1	W3C XML Schema2
Relax NG Spec	Relax NG Tutorial