Introduction to XSLT - TEItei.oucs.ox.ac.uk/Talks/2011-07-dhox/presentations/xsl-01.pdf · XSLT The...

transcript

Introduction to XSLT

TEI@Oxford

July 2011

Summer School 2011 1/59

Publishing XML files using XSLT

Our work will be divided into four partsBasic XSLT Target: make HTML from TEI documents

More complex XSLT Target: making more complex HTML,sorting and summarizing

Using TEI XSL family Target: customize existing library ofstylesheets

TEI, ODT, DOCX, and ePub Target: transforming TEI to andfrom word-processor and epub formats

… depending on how fast or slow we go, and what the class wantsto talk about …It is assumed that we are working on TEI XML documents.

What is the XSL family?

XPath: a language for expressing paths through XML treesXSLT: a programming language for transforming XMLXSL FO: an XML vocabulary for describing formatted pages

The XSLT language isexpressed in XMLuses namespaces to distinguish output from instructionspurely functionalreads and writes XML trees

It was designed to generate XSL FO, but now widely used togenerate HTML.

What is a transformation?Take this:.

<persName><forename>Milo</forename><surname>Casagrande</surname>

</persName><persName><forename>Corey</forename><surname>Burger</surname>

</persName><persName><forename>Naaman</forename><surname>Campbell</surname>

</persName>

and make this:.

<item n="1"><name>Burger</name>

</item><item n="2"><name>Campbell</name>

</item><item n="3"><name>Casagrande</name>

</item>

A text example

Take this.

<div type="recipe" n="34"><head>Pasta for beginners</head><list><item>Pasta</item><item>Grated cheese</item>

</list><p>Cook the pasta and mix with the cheese</p>

</div>

and make this.

<html><h1>34: Pasta for beginners</h1><p>Ingredients: Pasta Grated cheese</p><p>Cook the pasta and mix with the cheese</p>

</html>

How do you express that in XSL?.

<xsl:stylesheetxpath-default-namespace="http://www.tei-

c.org/ns/1.0" version="2.0"><xsl:template match="div"><html><h1><xsl:value-of select="@n"/>:

<xsl:value-of select="head"/></h1><p>Ingredients:<xsl:apply-templates select="list/item"/></p><p><xsl:value-of select="p"/>

</p></html>

</xsl:template></xsl:stylesheet>

Note: the namespace declaration linking xsl: tohttp://www.w3.org/1999/XSL/Transform is not shownin these examples.

Structure of an XSL file

<xsl:stylesheetxpath-default-namespace="http://www.tei-

c.org/ns/1.0" version="2.0"><xsl:template match="div">

</xsl:template><xsl:template match="p">

</xsl:template>

</xsl:stylesheet>

The div and p are XPath expressions, which specify which bit ofthe document is matched by the template.Any element not starting with xsl: in a template body is put intothe output.

The Golden Rules of XSLT

...1 If there is no template matching an element, we go on andprocess the elements inside it

...2 If there are no elements to process by Rule 1, any text insidethe element is output

...3 Children elements are not processed by a template unless youexplicitly say so

...4 xsl:apply-templates select="XX" looks fortemplates which match element ”XX”; xsl:value-ofselect="XX" simply gets any text from that element

...5 The order of templates in your program file is immaterial

...6 You can process any part of the document from any template

...7 Everything is well-formed XML. Everything!

Important magic

Our examples and exercises all start with two important attributeson <stylesheet>:

<xsl:stylesheetxmlns:xsl="http://www.w3.org/1999/XSL/Transform"xpath-default-namespace="http://www.tei-c.org/ns/1.0"version="2.0">....

which indicates...1 In our XPath expressions, any element name without a

namespace is assumed to be in the TEI namespace...2 We want to use version 2.0 of the XSLT specification. This

means that we must use the Saxon processor for our work.

A simple test file

<text><front><div><p>Material up front</p>

</div></front><body><div><head>Introduction</head><p rend="it">Some sane words</p><p>Rather more surprising words</p>

</div></body><back><div><p>Material in the back</p>

</div></back>

</text>

Feature: apply-templates.

<xsl:stylesheet version="2.0"xpath-default-namespace="http://www.tei-c.org/ns/1.0">

<xsl:template match="/"><html><xsl:apply-templates/>

</html></xsl:template>

</xsl:stylesheet>

<xsl:template match="TEI"><xsl:apply-templates select="text"/>

</xsl:template>

<xsl:template match="text"><h1>FRONT MATTER</h1><xsl:apply-templates select="front"/><h1>BODY MATTER</h1><xsl:apply-templates select="body"/>

</xsl:template>

Feature: value-of

Templates for paragraphs and headings:.

<xsl:template match="p"><p><xsl:apply-templates/>

</p></xsl:template><xsl:template match="div"><h2><xsl:value-of select="head"/>

</h2><xsl:apply-templates/>

</xsl:template><xsl:template match="div/head"/>

Notice how we avoid getting the heading text twice.Why did we need to qualify it to deal with just <head> inside<div>?

More complex patterns

The select attribute can point to any part of the document. UsingXPath expressions, we can find:

/ the root of document (outside the root element)* any elementtext()name an element called ‘name’@name an attribute called ‘name’

Example of complete path in <value-of>:.

<xsl:value-ofselect="/TEI/teiHeader/fileDesc/titleStmt/title"/>

XPath is the basis of most other XML querying and transformationlanguages.

It is a syntax for accessing parts of an XML documentIt uses a path structure to define XML elementsIt has a library of standard functionsIt is a W3C Standard and one of the main components ofXQuery and XSLT

Example text

<body n="anthology"><div type="poem"><head>The SICK ROSE </head><lg type="stanza"><l n="1">O Rose thou art sick.</l><l n="2">The invisible worm,</l><l n="3">That flies in the night </l><l n="4">In the howling storm:</l>

</lg><lg type="stanza"><l n="5">Has found out thy bed </l><l n="6">Of crimson joy:</l><l n="7">And his dark secret love </l><l n="8">Does thy life destroy.</l>

</lg></div>

</body>

XML Structure

body type=“anthology”

div type=“poem”

div type=“shortpoem”

lg type=“stanza”

lg type=“couplet”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Really attributes (and text) are separate nodes!

/body/div/head

div type= “poem”

div type= “shortpoem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

XPath locates any matching nodes

/body/div/lg ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div/lg

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div/@type ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

@ = attributes

/body/div/@type

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

type=“poem”

type=“shortpoem”

/body/div/lg/l ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div/lg/l

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div/lg/l[@n=“2”] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Square Brackets Filter Selection

/body/div/lg/l[@n=“2”]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div[@type=“poem”]/head ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

/body/div[@type=“poem”]/head

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//lg[@type=“stanza”] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

// = any descendant

//lg[@type=“stanza”]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//div[@type=“poem”]//l ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//div[@type=“poem”]//l

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//l[5] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Square brackets can also filter by counting

//l[5]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//lg/../@type ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Paths are relative: .. = parent

//lg/../@type

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

type=“poem”

type=“shortpoem”

//l[@n > 5] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Numerical operations can be useful.

//l[@n > 5]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//div[head]/lg/l[@n=“2”] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

Notice the deleted <head> !

//div[head]/lg/l[@n=“2”]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

//l[ancestor::div/@type=“shortpoem”] ?

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

ancestor:: is an unabbreviated axis name

//l[ancestor::div/@type=“shortpoem”]

div type=“poem”

l n=“4”

l n=“6”

l n=“2”

l n=“3”

l n=“7”

l n=“1”

l n=“8”

l n=“5”

l n=“1”

l n=“2”l n=“2”

l n=“2”

XPath: More About Paths

A location path results in a node-setPaths can be absolute (/div/lg[1]/l)Paths can be relative (l/../../head)Formal Syntax: (axisname::nodetest[predicate])For example:child::div[contains(head, 'ROSE')]

XPath: Axes

ancestor:: Contains all ancestors (parent, grandparent, etc.)of the current node

ancestor-or-self:: Contains the current node plus all itsancestors (parent, grandparent, etc.)

attribute:: Contains all attributes of the current nodechild:: Contains all children of the current node

descendant:: Contains all descendants (children,grandchildren, etc.) of the current node

descendant-or-self:: Contains the current node plus all itsdescendants (children, grandchildren, etc.)

XPath: Axes (2)

following:: Contains everything in the document after theclosing tag of the current node

following-sibling:: Contains all siblings after the currentnode

parent:: Contains the parent of the current nodepreceding:: Contains everything in the document that is

before the starting tag of the current nodepreceding-sibling:: Contains all siblings before the current

nodeself:: Contains the current node

Axis examplesancestor::lg = all <lg> ancestorsancestor-or-self::div = all <div> ancestors orcurrentattribute::n = n attribute of current nodechild::l = <l> elements directly under current nodedescendant::l = <l> elements anywhere under currentnodedescendant-or-self::div = all <div> children orcurrentfollowing-sibling::l[1] = next <l> element at thislevelpreceding-sibling::l[1] = previous <l> element atthis levelself::head = current <head> element

XPath: Predicates

child::lg[attribute::type='stanza']child::l[@n='4']child::div[position()=3]child::div[4]child::l[last()]child::lg[last()-1]

XPath: Abbreviated Syntax

nothing is the same as child::, so lg is short forchild::lg@ is the same as attribute::, so @type is short forattribute::type. is the same as self::, so ./head is short forself::node()/child::head.. is the same as parent::, so ../lg is short forparent::node()/child::lg// is the same as descendant-or-self::, so div//l isshort for child::div/descendant-or-self::node()/child::l

Example of context-dependent matches

Compare.

<xsl:template match="head"> ....</xsl:template>

<xsl:template match="div/head"> ...</xsl:template><xsl:template match="figure/head"> ....</xsl:template>

Priorities when templates conflict

It is possible for it to be ambiguous which template is to be used:.

<xsl:template match="person/name">…</xsl:template><xsl:template match="name">…</xsl:template>

when the processor meets a <name>, which template is used?

Solving priorities

There is a priority attribute on <template>; the higher the value,the more inclined the XSLT engine is to use it:.

<xsl:template match="name" priority="1"><xsl:apply-templates/>

</xsl:template><xsl:template match="person/name" priority="2"> A name</xsl:template>

Template priority generally

The more normal rule is that the most specific template wins..

<xsl:template match="*"></xsl:template><xsl:template match="tei:*"></xsl:template><xsl:template match="p"></xsl:template><xsl:template match="div/p"></xsl:template><xsl:template match="div/p/@n"></xsl:template>

Pushing and pullingXSLT stylesheets can be characterized as being of two types:

push In this type of stylesheet, there is a different templatefor every element, communication via<xsl:apply-templates> and the overall result isassembled from bits in each template. It is sometimeshard to visualize the final design. Common fordata-oriented processing where the structure is fixed.

pull In this type, there is a master template (usuallymatching /) with the main structure of the output,and specific <xsl:for-each> or <xsl:value-of>commands to grab what is needed for each part. Thetemplates tend to get large and unwieldy. Commonfor document-oriented processing where the inputdocument structure varies.

Attribute value template

What if we want to turn... ..

. .<ref target="http://www.oucs.ox.ac.uk/">OUCS</ref>

into... ..

. .<a href="http://www.oucs.ox.ac.uk/"/>

? What we cannot do is.

<xsl:template match="ref"><a href="@target"><xsl:apply-templates/>

</a></xsl:template>

This would give the @href attribute the value ‘@target’.

For example

Instead we use {} to indicate that the expression must beevaluated:.

<xsl:template match="ref"><a href="{@target}"><xsl:apply-templates/>

</a></xsl:template>

This would give the @href attribute whatever value the attribute@target has!

Feature: for-eachIf we want to avoid lots of templates, we can do in-line loopingover a set of elements. For example:.

<xsl:template match="listPerson"><ul><xsl:for-each select="person"><li><xsl:value-of select="persName"/>

</li></xsl:for-each>

</ul></xsl:template>

contrast with.

<xsl:template match="listPerson"><ul><xsl:apply-templates select="person"/>

</ul></xsl:template><xsl:template match="person"><li><xsl:value-of select="persName"/>

</li></xsl:template>

Feature: ifWe can make code conditional on a test being passed:.

<xsl:template match="person"><xsl:if test="@sex='1'"><li><xsl:value-of select="persName"/>

</li></xsl:if>

</xsl:template>

contrast with.

<xsl:template match="person[@sex='1']"><li><xsl:value-of select="persName"/>

</li></xsl:template><xsl:template match="person"/>

The @test can use any XPath facilities.

Feature: choose

We can make a multi-value choice conditional on what we find inthe text:.

<xsl:template match="person"><xsl:apply-templates/><xsl:choose><xsl:when test="@sex='1'">(male)</xsl:when><xsl:when test="@sex='2'">(female)</xsl:when><xsl:when test="not(@sex)">(no sex specified)</xsl:when><xsl:otherwise>(unknown sex)</xsl:otherwise>

</xsl:choose></xsl:template>

Summary

Now you can...1 Write templates which match any element or attribute...2 Pick out text from anywhere...3 Write code conditional on something in the text

Introduction to XSLT - TEItei.oucs.ox.ac.uk/Talks/2011-07-dhox/presentations/xsl-01.pdf · XSLT The...

Documents