westhoff_schlitt.css Validate and Transform XML using XML-Schema and XML-Stylesheets Jakob Westhoff (jakob@php.net) Tobias Schlitt (toby@php.net) Tue, 2007-04-17 University of Dortmund
  • About us
  • What's all about this XSD?
  • A simple example
  • Simple types
  • Complex types
  • Predefined data types
  • What's all about this XSL?
  • What can I do with XSLT?
  • What is XPath?
  • XPath terminologie
  • XPath syntax
  • Starting over with XSLT
  • The basic structure
  • ]]>
  • and ]]>
  • and ]]>
  • The result
  • Conclusion

Jakob Westhoff

Contact: jakob@php.net

  • PHP Developer for more than 5 years
  • Currently studying information science at the university of Dortmund
  • Founder of the PHP-Usergroup-Dortmund
  • Freelancing Software Architect and Author

Tobias Schlitt

Contact: toby@php.net

  • PHP Developer for more than 6 years
  • Company trained IT Specialist (Deutsche Bank)
  • Working for eZ Systems on the eZ Components project
  • Member of Zend PHP Certification Board
  • Freelancing Software Architect, Author, Trainer
  • XSD = XML Schema Definition
  • Define the legal building blocks of an XML document
    • Elements
    • Attributes
    • Order/number of child elements
    • Allowed data types
    • ...
  • Easily validate a given XML file to a schema
  • W3C Recommendation since 02. May 2001
  • Successors of DTD

A simple XML document

Jakob Westhoff Tobias Schlitt XSD is the successor of DTD Hi Toby! XSD is definetly more advanced than the somehow outdated document type definitions ]]>

An appropriate XSD file

]]>

Asscoiate the xsd with the document

... ]]>
  • Simple type elements have certain limitations
    • Only character data allowed
    • No child elements are allowed
  • Some useful predefined types
    • xs:string
    • xs:integer
    • xs:decimal
    • xs:boolean
    • xs:date
    • ...
]]> Any string content true|false ]]>
  • Restrictions
    • Are applied to already known predefined types
    • e.g. Limit the range of an integer or restrict integers to an enumeration

Limit integer range

]]>

Restrict strings with enumerations

]]>

Regular expression pattern matching

]]>
  • Complex types may contain
    • Nothing
    • Character data
    • Other elements
    • Both
    • Attributes

Complex type example

Tobias Schlitt 26 fellow student good friend of mine ]]> ]]>

Split type and element definition

]]>

Adding attributes to predefined simple types

]]>

Mix character data with elements

You are a great audience ]]> ]]>
  • Indicators
    • Order indicators
    • Occurance indicators
    • Group indicators
  • Order Indicators
    • Sequence
      • Child elements need specific order
    • Choice
      • Only one of the specified elements is needed
    • All
      • All of the elements are needed
      • Order not specified
      • min and maxOccurs may only be 1 or 0
... ]]> ... ]]>
  • Occurance Indicators
    • Already known
    • minOccurs
    • maxOccurs

Occurance indicators may be applied to order indicators

coffee beer limonade ]]> ]]>
  • Used to define related elements
  • Can be referenced inside the xsd multiple times
  • Work for element and attribute definitions

Something you already know

]]>

Can also be defined as a group

]]>

Now it can be referenced any where in the schema

]]>

Attribute group example

]]> ]]> ]]>

XSL = eXtensible Stylesheet Language.

  • W3C recommendation since October 2001
  • XSL consists of:
    • XSLT = XSL Transformations
    • XSL-FO = XSL Formatting Objects
    • XPath = Navigation and selection language for XML elements
  • When saying XSL most people mean XSLT.

XSLT = eXtensible Stylesheet Language Transformations.

  • W3C recommendation since November 16th 1999
  • Current version is 2.0 (since 2007-01-23)
  • XML based
  • Declarative, funktional-applicative programming language
  • Used for transformations between XML formats
  • Originally defined to transform any XML into XSL-FO
  • Makes use of XPath
  • Commonly used to transform between any XML formats
  • All modern browsers support XSLT (more or less)
  • Little known fact: XSLT is turing complete

Attention: Recursion is everything in a functional language!
  • XML formats are descriptive
  • When defining a format, ignore everything but the pure meaning of data!
  • To allow others to work with your data: Transform it into their language!
  • Example: Transforming a data storage format into XHTML to view it in a browser

Some simple XML document:

Simple test document This is the first paragraph. This is the second paragraph. ]]>

A stylesheet to transform this document into XHTML:

<xsl:value-of select="title" />

]]>

The result:

Simple test document

This is the first paragraph.

This is the second paragraph.

]]>

Before you can learn XSLT, you should know XPath.

  • Language to address and select XML elements
  • W3C recommendation since November 1999
  • Current version is 2.0 (since 2007-01-23)
  • Used for navigation in XML structures
  • Enables you to select information from an XML document

You will see many XPath expressions during this session and I will explain them in place!

In XPath (almost) everything is a node.

  • The root node of an XML document is referred to as the document node. ]]>
  • Any tag inside the document node is referred to as an element node This is a paragraph

    ]]>
  • Attributes of an element node are referred to as attribute nodes ]]>
  • Everything that is not a node is called an atomic value

Every XPath expression matches a set of nodes (0..n).

XML is a tree structure, the usual terms apply:

  • Parent
  • Child / children
  • Sibling
  • Ancestor
  • Descendant

The XPath syntax is quite simple, but very powerfull.

  • Works similar to path references in the file system (Unix syntax)
  • Allows traversal on the whole XML tree
  • Additionally supports more than 100 special functions for:
    • string manipulation
    • math calculations
    • date and time comparison
    • node and sequence manipulation
    • ...

Some basic examples (here for XHTML):

  • Tag matches:
    • Relative match: ul/li
    • Absolute match: /html/body/p
    • Global match: //img
    • Anywhere match: /html/body/p//img
    • Alternative match: ul|ol
  • Attribute matches:
    • Existence match: //img[@border]
    • Value match: //img[@border='1']
  • Relation matches:
    • Current node match: //p/img/.
    • Parent node match: //p/img/..
    • Position match: //ul/li[1]

To learn XSLT, we will examine this presentation system xmlpress.

The basic XML structure looks like this:

default.css Validate and Transform XML using XML-Schema and XML-Stylesheets Jakob Westhoff (jakob@php.net) Tobias Schlitt (toby@php.net) Tue, 2007-04-17 University of Dortmund

Before you can learn XSLT, you should know XPath.

  • Language to address and select XML elements
  • W3C recommendation since November 1999
  • Current version is 2.0 (since 2007-01-23)
  • Used for navigation in XML structures
  • Enables you to select information from an XML document

You will see many XPath expressions during this session and I will explain them in place!

 ]]>

In following we will build the XSLT to transform this into XHTML, step by step.

What every XSLT stylesheet needs:

  • XML declaration
  • The XSL namespace declaration
  • The document root node ]]>
  • At least 1 ]]> node

Optionally:

  • A defined output method (meta info for the processor)
  • More template definitions
]]>

]]>

  • A template applies to a set of nodes
  • The node set is defined by the match attribute
  • The match attribute is an XPath expression
  • The selected node set is submitted to the template

xmlpress uses the following templates:

  • Process the root node ]]>: ]]>
  • Process the global options: ]]>
  • Process misc formatting tags: ]]>
  • Special processing form image tags: ]]>

... there are some more, but those look very similar...

]]>

  • Prints the value of a selected node
  • Attribute "select" defines, what to print using XPath
  • XPath expression is local to the current template match

]]>

  • Indicates to perform apply the given template name
  • Applies template to nodes selected by "match" attribute
  • Result is inserted at this place (remember: recursion!)

The main template performs the following actions:

  • Print the basic XHTML structure
  • Extract global information and place it correctly into XHTML
  • Call sub templates for more complex operations
<xsl:value-of select="general/title[1]" />