XML XPath

XPath is a small language for finding elements and attributes inside an XML document, like an address for data.

The Example Document

All examples on this page use this file, shop.xml:

Example

xml

<?xml version="1.0" encoding="UTF-8"?>
<shop name="Saddar Mobiles">
  <product id="p1" category="phone">
    <name>Galaxy A15</name>
    <price>42000</price>
    <stock>8</stock>
  </product>
  <product id="p2" category="accessory">
    <name>Fast Charger</name>
    <price>2500</price>
    <stock>30</stock>
  </product>
  <product id="p3" category="phone">
    <name>Redmi 13</name>
    <price>38500</price>
    <stock>0</stock>
  </product>
</shop>

Paths

XPath paths look like folder paths:

XPath	Selects
`/shop`	The root element
`/shop/product`	All product elements under the root
`/shop/product/name`	All name elements of products
`//name`	Every name element anywhere in the document
`.`	The current element
`..`	The parent of the current element
`@id`	The id attribute
`/shop/@name`	The name attribute of shop

One slash / means "a direct child"
Two slashes // mean "anywhere below"

Example

text

//product/name/text()

Output

Galaxy A15
Fast Charger
Redmi 13

text() selects the text inside an element.

Predicates

A predicate is a condition in square brackets. It filters the results:

XPath	Selects
`//product[1]`	The first product (XPath counts from 1)
`//product[last()]`	The last product
`//product[@category='phone']`	Products whose category is phone
`//product[price > 10000]`	Products with price above 10000
`//product[stock = 0]/name`	Names of out-of-stock products
`//product[@id='p2']/price`	The price of product p2

Example

text

//product[@category='phone' and stock > 0]/name/text()

Output

Galaxy A15

XPath counts from 1, not from 0 like most programming languages. //product[1] is the first product.

Wildcards

XPath	Selects
`/shop/*`	All child elements of shop
`//product/@*`	All attributes of every product
`//*`	Every element in the document

Useful Functions

Function	Example	Result here
`count()`	`count(//product)`	3
`sum()`	`sum(//stock)`	38
`contains()`	`//product[contains(name, 'Charger')]`	The charger product
`starts-with()`	`//product[starts-with(name, 'R')]`	Redmi 13
`string-length()`	`string-length(//product[1]/name)`	10

Example

text

sum(//product[@category='phone']/stock)

Output

Trying XPath

In Chrome or Edge DevTools, open the Console on any page and type $x("//h1") to test XPath on HTML
xmllint on the command line: xmllint --xpath "//product/name/text()" shop.xml
Python, JavaScript, PHP, Java and C# all support XPath; the next lesson shows how

Where XPath is Used

Reading data from XML files and API responses
Web scraping and automated testing tools (Selenium, Playwright) use XPath to find page elements
XSLT, a language that transforms XML into HTML or other formats, is built on XPath

Exercise

Using shop.xml, which XPath selects the names of products that are out of stock?

/product[stock]/name//product[stock = 0]/name//name[0]//stock = 0