XPath tutorial - basics

From EduTech Wiki
Jump to: navigation, search

Draft


1 Introduction

XPath is a language for addressing parts of an XML document. Basic understanding of XPath is needed for XSLT and XQuery programming. In this piece, we shall show how to create XSLT stylesheets that use some moderately complex XPath expressions.

Learning Objectives

  • Understand how to use XPath expressions, in order use XSLT and XQuery more effectively
  • Learn some XSLT programming constructions (conditions and loops)
  • Being able to cope with most XML to HTML transformations

Prerequisites

Next steps

Materials

  • We use the same XML document in most examples. You can download the file or look at a wiki page:
xpath-jungle.xml
XPath tutorial - basics/XML example code

Disclaimer

  • This is an unfinished, not very nice, introductory XPath tutorial. Cut/paste from slides with a few fixes. It needs more work, since right now it's more like a list of XPath features ... - Daniel K. Schneider
  • There may be typos (sorry) and mistakes (sorry again)

2 Introduction to XML Path Language

2.1 Definition and history

XPath is a language for addressing parts of an XML document, In support of this primary purpose, it also provides basic facilities for manipulation of strings, numbers and booleans.

Its 2.0 editition was described as “XPath 2.0 is an expression language that allows the processing of values conforming to the data model defined in XQuery/XPath Data Model (XDM). The data model provides a tree representation of XML documents as well as atomic values such as integers, strings, and booleans, and sequences that may contain both references to nodes in an XML document and atomic values. The result of an XPath expression may be a selection of nodes from the input documents, or an atomic value, or more generally, any sequence allowed by the data model. The name of the language derives from its most distinctive feature, the path expression, which provides a means of hierarchic addressing of the nodes in an XML tree.” (XML Path Language (XPath) 2.0 ,W3C Recommendation 23 January 2007, retrieved 16:38, 9 February 2010 (UTC).

  • XPath uses a compact non-XML syntax (to facilitate use of XPath within URIs and XML attribute values).
  • XPath gets its name from its use of a path notation as in URLs for navigating through the hierarchical structure of an XML document.
  • XPath was defined at the same time as XSLT (nov 1999)
  • Initally, it was developped to support XSLT and XPointer (XML Pointer Language used for XLink, XInclude, etc.). Today it is also used by XQuery and other applications. Many programming languages support an XPath library, e.g. PHP5.

In plain English, XPath allows retrieving parts of an XML document. XPath expressions can be as simple as the name of an XML element or so complicated that only experts can understand it ....

Specifications

XPath 1.0 is used by XSLT 1.0, i.e. the XSLT processor included in virtually every web browser as of Jan 2010 and since the early 2000's in IE/Mozilla.

XPath 2.0 is a superset of XPath 1.0 and is used by XSLT 2.0, XQuery and other specifications.

Reference manuals

2.2 XSLT, XQuery and XPath

Each time a given XSLT or XQuery instruction needs to address (refers to) parts of an XML document, we use XPath expressions. XPath expressions also can contain functions, simple math and boolean expressions.

Within XSLT, XPath expressions are typicially used in match, select and test attributes:

Xpath expressions in an XSLT template

Below is an XQuery example taken from the XQuery tutorial - basics

for $t in fn:doc("catalog09.xml")//c3msbrick
let $n := count($t//c3mssoft)
where ($n > 1)
order by $n
return <result> {$t/title/text()} owns {$n} bricks </result>

2.3 Playtime

Most XML editors will display the XPath of a selected element. In addition, you can search with XPath expressions, i.e. test an expression that you then would use in your XSLT or XQuery code.

Below is a screenshot of the XML Exchanger editor. As you can see we entered the //participant expression in the search box. The result is a so-called node-set that is displayed in the XPath pane at the bottom.

Xpath seach in the Exchanger editor

3 The XPath Syntax and the document model

3.1 Xpath Syntax

XPath expressions can be quite simple or very complex. An Xpath expression may include a location path (i.e. "where to look for"), node tests (i.e. identifying a node) and predicates(i.e. additional tests).

There are two notations for location paths.

(1) abbreviated

This simple notation allows to locate itself, children, parents, attributes and combinations of these. Going up or down is called an axis. The abbreviated form has limited axis.

para means "all "para" child elements of the current context

(2) unabbreviated

Unabbreviated location path allows to search in more axis then just parents, children, and siblings.

child::para" is identical to para above.

The abbreviated location path look a bit like file paths. E.g. the following expression:

/section/title

means: find all title nodes below section nodes

Syntax overview of the primary (relatively simple) XPath expressions

The picture below is not entirely correct, i.e. the "green" elements are part of the so called node-test Xml-xpath-3.png

The result of an Xpath expression can be various data types, e.g. sets of nodes, a single node, a number, etc. Most often the result, is a set of nodes.


The formal specification of an XML Path
  • is very complex, i.e. has about 39 clauses and is very difficult to understand
  • Some expressions shown here are beyond the scope of this tutorial, don't panic !

3.2 The document model of XPath

  • XPath sees an XML document as a tree structure
  • Each information (XML elements, attributes, text, etc.) is called a node. This is fairly similar to the W3C DOM model an XML or XSLT processor would use.
Nodes that XPath can see
  • root node
    • ATTENTION: The root is not necessarily the XML root element. E.g. processing instructions like a stylesheet declaration are also nodes.
  • Elements and attributes
  • Special nodes like comments, processing instructions, namespace declarations.
Nodes XPath can't see
  • XPath looks at the final document, therefore it can't see entities and document type declarations....
The XML context
  • What a given XPath expression means, is always defined by a given XML context, i.e. the current node in the XML tree a processor is looking at.

4 Using simple location path

Below, we present a few expressions for locating nodes using the simple abbreviated syntax. As we said before, location paths can be horribly complex, but simple location path look a bit like file path that you would use in HTML links or in an operating system like Unix, Windows or MacOS.

4.1 List of simple location path

Document root node - returns the document root (which is not necessarily the XML root!)
/
Direct child element
XML_element_name
Direct child of the root node
/XML_element_name
Child of a child
XML_element_name/XML_element_name
Descendant of the root
//XML_element_name
Descendant of a node
XML_element_name//XML_element_name
Parent of a node
../
A far cousin of a node
../../XML_element_name/XML_element_name/XML_element_name

4.2 Example - Extracting titles from an XML file with XSLT

Let us recall that we use the same XML document in most examples. You can download the file or look at a wiki page:

xpath-jungle.xml
XPath tutorial - basics/XML example code
<?xml version="1.0"?>
<project>
 <title>The Xpath project</title>
 <participants>
  <participant>
    <FirstName>Daniel</FirstName>
    <qualification>8</qualification>
    <description>Daniel will be the tutor</description>
    <FoodPref picture="dolores_001.jpg">Sea Food</FoodPref>
  </participant>
  <participant>
    <FirstName>Jonathan</FirstName>
    <qualification>5</qualification>
    <FoodPref picture="dolores_002.jpg">Asian</FoodPref>
  </participant>
  <participant>
   <FirstName>Bernadette</FirstName>
   <qualification>8</qualification>
    <description>Bernadette is an arts major</description>
  </participant>
  <participant>
   <FirstName>Nathalie</FirstName>
   <qualification>2</qualification>
  </participant>
 </participants>
 <problems>
  <problem>
    <title>Initial problem</title>
    <description>We have to learn something about Location Path</description>
    <difficulty level="5">This problem should not be too hard</difficulty>
  </problem>
  <solutions>
   <item val="low">Buy a XSLT book</item>
   <item val="low">Find an XSLT website</item>
   <item val="high">Register for a XSLT course and do exercices</item>
  </solutions>
   <problem>
    <title>Next problem</title>
    <description>We have to learn something about predicates</description>
    <difficulty level="6">This problem is a bit more difficult</difficulty>
  </problem>
  <solutions>
   <item val="low">Buy a XSLT book</item>
   <item val="medium">Read the specification and do some exercises</item>
   <item val="high">Register for a XPath course and do exercices</item>
  </solutions>
 </problems>
</project>
Task

We would like to get a simple list of problem titles

Solution

XSLT template (file: xpath-jungle-1.xsl

<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                version="1.0">
 
  <xsl:output method="html"/>
 
  <xsl:template match="/project">
    <html>
      <body bgcolor="#FFFFFF">
        <h1><xsl:value-of select="title" /></h1>
        Here are the titles of our problems: <ul>
        <xsl:apply-templates select="problems/problem" /> 
      </ul>
      </body>
    </html>
  </xsl:template>
 
<xsl:template match="problems/problem">
  <li><xsl:value-of select="title" /></li>
</xsl:template>
 
</xsl:stylesheet>
(1) XSLT template for the root element
  • The XPath of the "match" means: apply the template to the project element node, which is a direct child of the root node
  • The execution context of this template is therefore the element "project"
  • The xsl:apply-templates expression will now tell the processor to find a rule for the problems/problem descendant.
(2) XSLT template for the problems/problem element

This second rule will be triggered by the first rule, because problems/problem is indeed a location that can be found in project element. The processor then can extract the value of the title element.

Alternatively we could have written this rule as:

<xsl:template match="problems/problem/title">
  <li><xsl:apply-templates/></li>
</xsl:template>

or

<xsl:template match="problems/problem/title">
  <li><xsl:value-of select="."/></li>
</xsl:template>
(3) Result HTML
 <html>
    <body bgcolor="#FFFFFF">
       <h1>The Xpath project</h1>
       Here are the titles of our problems:
       <ul>
          ''<li>Initial problem</li>''
          ''<li>Next problem</li>''
       </ul>
    </body>
 </html>

Live example:

4.3 Attribute Location Paths

Of course, XPath also also to locate attributes. We shall show the principle, using a few examples.

(1) To find an attribute of a child element in the current context use:

@attribute_name

Example:

@val

(2) Find attributes of an element in a longer location path starting from root

/element_name/element_name/@attribute_name

Example:

/project/problems/solutions/item/@val

(3) Find attributes in the whole document: //@attribute_name

As you can see you can combine element location with attribute identification.

4.4 Example - Create an html img link from an attribute

XML fragment

Same as above

Task

Display a list of First Names plus their food preferences

XSLT (File xpath-jungle-2.xsl
<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                version="1.0">
 
  <xsl:output method="html"/>
 
  <xsl:template match="/">
    <html>
      <body bgcolor="#FFFFFF">
        <h1>What do we know about our participants ?</h1>
        Here are some food preferences: <ul>
        <xsl:apply-templates select=".//participant" /> 
      </ul>
      </body>
    </html>
  </xsl:template>
 
<xsl:template match="participant">
  <li><xsl:value-of select="FirstName"/>
  <xsl:apply-templates select="FoodPref"/>
  </li>
</xsl:template>
 
<xsl:template match="FoodPref">
  prefers <xsl:value-of select="."/>. 
  <img src="{@picture}"/> <br clear="all"/>
</xsl:template>
 
</xsl:stylesheet>
  • The second rule will display names of participants and launch a template for FoodPref
  • Note: Not all participants have a FoodPref element. If it is absent it will just be ignored.
  • The third rule (FoodPref) displays the text (contents) of FoodPref and then makes an HTML img tag


Parts of the result
   <h1>What do we know about our participants ?</h1>
       Here are some food preferences:
       <ul>
          <li>Daniel prefers Sea Food.
             <img src="dolores_001.jpg"><br clear="all"></li>
          <li>Jonathan
             prefers Asian.
             <img src="dolores_002.jpg"><br clear="all"></li>
          <li>Bernadette</li>
          <li>Nathalie</li>
       </ul>

Live example:

4.5 Location wildcards

  • Sometimes (but not often!), it is useful to work with wildcards
  • You have to understand that only one rule will be applied per element. Rules with wildcards have less priority and btw. this is why "your rules" are applied before the system default rules.
Find all child nodes of type XML element
*
Find all child nodes (including comments, etc.)
node()
Find all element attributes
@*
Find all text nodes
text()
Combine locations
use the "|" operator, an example is just below.

Example: XSLT includes two built-in default rules. They rely on using these wildcards.

This rule applies to the document root and all other elements

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

Text and attribute values are just copied

 <xsl:template match="text()|@*">
   <xsl:value-of select="."/>
 </xsl:template>

5 XPaths with predicates

Let us now scale up a bit.

  • A predicate is an expression that can be true or false
  • It is appended within [...] to a given location path and will refine results
  • More than one predicate can be appended to and within (!) a location path
  • Expressions can contain mathematical or boolean operators
Find element number N in a list
Syntax: XML_element_name [ N ]
/project/participants/participant[2]
/project/participants/participant[2]/FirstName
Find elements that have a given attribute
Synatx: XML_element_name [ @attribute_name ]
//difficulty[@level]
Find elements that have a given element as child
Syntax XML_element_name [ XML_element_name ]//participant[FoodPref]
Note: this is not the same as //participant/FoodPref. The latter would return a list of FoodPref elements, whereas the former returns a list of participant
Mathematical expressions
  • Use the standard operators, except div instead of / (for obvious reasons)
- * div mod
  • mod is interesting if you want to display a long list in table format
5 mod 2 returns 1, as will "7 mod 2" and "3 mod 2"
Boolean operators (comparison, and, or)
  • List of operators (according to precedence)
<=, <, >=, >
=, !=
and, or


Examples
  • Return all exercise titles with a grade bigger than 5.
//exercise[note>5]/title
  • Find elements that have a given attribute with a given value
Recall of the Syntax: XML_element_name [ @attribute_name="value"]
//solutions/item[@val="low"]
  • Example XSLT template that will match all item elements with val="low".
<xsl:template match="'//item[@val='low']">
    <xsl:value-of select="." />
</xsl:template>

Usually expressions also contain functions as we shall see below, examples:

  • Return the last five elements of a list
author [(last() - 4) <= position()) and (position() <= last())]
  • Return all Participant nodes with a contents of FirstName bigger than 7 characters:
"//Participant[string-length(FirstName)>=8]"

5.1 Example: Retrieve selected elements

The following example will retrieve the following:

  • All persons who do have a food preference
  • All items that have an a "priority" attribute set to "high"
The XSLT stylesheet (file xpath-jungle-3.xsl)
<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                version="1.0">
 
  <xsl:output method="html"/>
 
  <xsl:template match="/">
    <html>
      <body bgcolor="#FFFFFF">
        <h1>Retrieve selected elements</h1>
        Here is the name of participant two: 
        <ul><li><xsl:value-of select=".//participant[2]/FirstName"/></li></ul>
        Here are all participant's firstnames that have a food preference: 
        <ul><xsl:apply-templates select=".//participant[FoodPref]"/></ul>
        Here are all items that have a value of "high"
        <ul><xsl:apply-templates select=".//item[@val='high']"/></ul>
      </body>
    </html>
  </xsl:template>
 
<xsl:template match="participant">
  <li><xsl:value-of select="FirstName"/></li>
</xsl:template>
 
<xsl:template match="item">
  <li><xsl:value-of select="."/></li>
</xsl:template>
 
</xsl:stylesheet>
HTML result
<html>
   <body bgcolor="#FFFFFF">
      <h1>Retrieve selected elements</h1>
      Here is the name of participant two:
      <ul>
         <li>Jonathan</li>
      </ul>
      Here are all participant's firstnames that have a food preference:
      <ul>
         <li>Daniel</li>
         <li>Jonathan</li>
      </ul>
      Here are all items that have a value of "high"
      <ul>
         <li>Register for a XSLT course and do exercices</li>
         <li>Register for a XPath course and do exercices</li>
      </ul>
   </body>
</html>

Live example:

6 XPath functions

XPath defines a certain number of functions. You can recognize a function because it has appended "()".

Functions are programming constructs that will return various kinds of informations, e.g.

  • true / false
  • a number
  • a string
  • a list of nodes

It is not obvious to understand what all of these functions do. For example, there are restrictions on how you can use functions (stick to examples or the reference)

last()
last() gives the number or nodes within a context
position()
position() returns the position of an element with respect to other children in the same parent
Warning: The result will include empty nodes create from whitespaces between elements. To avoid this you should strip the parent nodes using
 <xsl:strip-space elements="name_of_the_parent_node"/>
count(node-set)
count gives the number of nodes in a node set
We got <xsl:value-of select="count(//problem)"/> problems, 
i.e. it will count N elements retrieved and not the child elements of problem
Live example: xpath-jungle-6.xml - xpath-jungle-6.xsl
starts-with(string, string)
returns TRUE if the second string is part of the first and starts off the first
//Participant[starts-with(Firstname,'Berna')]"
contains(string, string)
returns TRUE if the second string is part of the first
//Participant[contains(FirstName,'nat')]
string-length(string)
returns the length of a string
number(string)
transforms a string into a number
sum(node-set)
computes the sum of a given set of nodes.
If necessary, does string conversion with number()
round(number)
round a number, e.g. 1.4 becomes 1 and 1.7 becomes 2
translate(string1, string2, string3)
translates string1 by substituting string2 elements with string3 elements

6.1 Example: Computation of an average

  • We would like to compute the average of participant's qualifications
 <participant><FirstName>Daniel</FirstName>
              ''<qualification>8</qualification>''
    </participant>
The XSLT stylesheet (file xpath-jungle-4.xsl
  • We compute the sum of a node-set and then divide by the number of nodes
<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                version="1.0">
 
  <xsl:output method="html"/>
 
  <xsl:template match="/">
    <html>
      <body bgcolor="#FFFFFF">
        <h1>Qualification level of participants</h1>
        Average is
        <xsl:value-of select="sum(.//participant/qualification) div count(.//participant/qualification)"/>
      </body>
    </html>
  </xsl:template>
</xsl:stylesheet>
HTML result
 <html>
    <body bgcolor="#FFFFFF">
       <h1>Qualification level of participants</h1>
       Average is
       5.75
    </body>
 </html>

Live example:

6.2 Example: Find first names containing 'nat'

The XSLT stylesheet (file xpath-jungle-5.xsl
<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                version="1.0">
 
  <xsl:output method="html"/>
 
  <xsl:template match="/">
    <html>
      <body bgcolor="#FFFFFF">
        <h1>Do we have a "nat" ?</h1>
        First Names that contain "nat":
        <ul><xsl:apply-templates select=".//participant[contains(FirstName,'nat')]"/></ul>
        First Names that contain "nat" and "Nat":
        <ul><xsl:apply-templates select=".//participant[contains(translate(FirstName,'N','n'),'nat')]"/></ul>
      </body>
    </html>
  </xsl:template>
 
<xsl:template match="participant">
  <li><xsl:value-of select="FirstName"/></li>
</xsl:template>
 
</xsl:stylesheet>

Live example

7 Union of XPaths

Union Xpaths combine more than one XPath (and all the resulting nodes are returned). A typical example that we already introduced above is the default rule which means that the template matches either the root element (i.e. "/" or just any element),

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

Often, this construct is used to simplify apply-templates or even templates themselves. E.g. the following rules applies to both "description" and "para" elements.

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

8 List of commonly used XPath expressions

Syntax
element

(Type of path)

Example path

Example matches

name

child element name

project

<project> ...... </project>

/

child / child

project/title

<project> <title> ... </title>

/

(root element)

//

descendant

project//title

<project><problem> <title>....</title>

//title

<root>... <title>..</title> (any place)

*

"wildcard"

*/title

<bla> <title>..</title> and <bli> <title>...</title>

|

"or operator

title|head

<title>...</title> or <head> ...</head>

*|/|@*

All elements: root, children and attributes

.

current element

.

../

parent element

../problem

<project>

@attr

attribute name

@id

<xyz id="test">...</xyz>

element/@attr

attribute of child

project/@id

<project id="test" ...> ... </project>

@attr='value'

value of attribute

list[@type='ol']

<list type="ol"> ...... </list>

position()

position of element
in parent

position()

last()

number of elements within a context

last()

position()!=last()


Important:

  • The XML standard requires that an XML parser returns emtpy space between elements as empty nodes. In other words, functions such as location() or last() will also count empty nodes ! Tell XSLT to remove these within the elements within which you need to count.

Good code:

<xsl:strip-space elements="name_of_the_parent"/>

Bad code:

you forgot to use this ..

Example fragment where position in a list is used to compute position on the screen with SVG in HTML5:

<xsl:strip-space elements="list"/>
<xsl:template match="list">
    <rect x="10" y="105" width="{10 * count(item)}" 
	  height="5" fill="black" stroke="red"/>
    <xsl:apply-templates/>
</xsl:template>

Live example:

9 Links

9.1 Introductory tutorials

9.2 Other