While Power Automate documentation does not explicitly state the XPath version supported by the xPath() function, there are sources that confirm Power Automate supports XPath 1.0. Sources like... community discussions, or failed tests when using XPath 2.0 😉
For the XPath reference, including supported syntax and functions, see XPath Reference
the bookstore
Most of the examples I found for using xpath in Power Automate cover the very basic example of the bookstore.
Let's extend the example with additional books and chapter information.
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book>
<title lang="en">Harry Potter</title>
<author>J K. Rowling</author>
<year>2005</year>
<price currency="USD">29.99</price>
<info>
<chapters>
<chapter>
<name>The Boy Who Lived</name>
<page>1</page>
<description>Introduction to Harry Potter's early life.</description>
</chapter>
<chapter>
<name>The Vanishing Glass</name>
<page>15</page>
<description>Harry's encounter with a snake at the zoo.</description>
</chapter>
</chapters>
</info>
</book>
<book>
<title lang="de">Der Hobbit</title>
<author>J.R.R. Tolkien</author>
<year>1937</year>
<price currency="USD">19.99</price>
<info>
<chapters>
<chapter lcid="1031">
<name>Eine unerwartete Party</name>
<page>1</page>
<description>Bilbo Beutlin trifft die Zwerge.</description>
</chapter>
<chapter lcid="1031">
<name>Gebratenes Hammelfleisch</name>
<page>35</page>
<description>Die Zwerge begegnen Trollen.</description>
</chapter>
</chapters>
</info>
</book>
<book>
<title lang="en">1984</title>
<author>George Orwell</author>
<year>1949</year>
<price currency="USD">14.99</price>
<info>
<chapters>
<chapter>
<name>The Principles of Newspeak</name>
<page>1</page>
<description>Introduction to the language of Oceania.</description>
</chapter>
<chapter>
<name>War is Peace</name>
<page>50</page>
<description>Explanation of the Party's slogans.</description>
</chapter>
</chapters>
</info>
</book>
</bookstore>
Important! Remember to parse the string to xml object using
xml()function.
Selecting Nodes
| xPath expression | Result |
|---|---|
xpath(variables('oBookstoreXml'),'//book/*') |
all direct descendants of 'book' nodes |
xpath(variables('oBookstoreXml'),'//book/*[@*]') |
all direct descendants of 'book' nodes, which have at least one attribute of any kind |
xpath(variables('oBookstoreXml'),'//book[title[@lang="en"]]') |
all 'book' nodes where title has parameter lang='en |
xpath(variables('oBookstoreXml'),'//book/*[self::title or self::author]') |
all title and author elements of all book nodes * |
* The
|operator, used for for selecting several paths, is not supported in xPath 1.0.
XPath functions
I'm using Select action in all the examples.
For a list of supported XPath functions, see XPath Functions.
Example 1
Select all direct descendants of 'book' nodes and read nodes information. This example used the following XPath functions: name(), string(), count(), concat(), position().
From: xpath(variables('oBookstoreXml'),'//book/*')
Map:
| value | expression |
|---|---|
| nodeName | xpath(item(), 'name(/*)') |
| nodeValueAsString | xpath(item(), 'string(/*)') |
| chapterCount | xpath(item(), 'count(//chapter)') |
| chapterInformation | xpath(item(), 'concat("This node has ",count(//chapter), " chapter child nodes")') |
| nodePosition | xpath(item(), 'position()') |
Results
[
{
"nodeName": "title",
"nodeValueAsString": "Harry Potter",
"chapterCount": 0,
"chapterInformation": "This node has 0 chapter child nodes",
"nodePosition": 1
},
{
"nodeName": "author",
"nodeValueAsString": "J K. Rowling",
"chapterCount": 0,
"chapterInformation": "This node has 0 chapter child nodes",
"nodePosition": 1
},
//....
{
"nodeName": "info",
"nodeValueAsString": "\n \n \n The Boy Who Lived\n 1\n Introduction to Harry Potter's early life.\n \n \n The Vanishing Glass\n 15\n Harry's encounter with a snake at the zoo.\n \n \n",
"chapterCount": 2,
"chapterInformation": "This node has 2 chapter child nodes",
"nodePosition": 1
},
//...
]
The position() function returns 1 for each node, because the xpath expressions in the Map section only have access to the chunk returned by the From query.
Some string functions, .e.g. upper-case(), would throw an error because they are not supported in XPath 1.0.
Example 2
Select all direct descendants of 'book' nodes, which have at least one attribute of any kind. This example focuses on working with attributes: filtering based on their presence and accessing their values
From: xpath(variables('oBookstoreXml'),'//book/*[@*]')
Map:
| value | expression |
|---|---|
| nodeName | xpath(item(), 'name(/*)') |
| attributeLang | xpath(item(), '//@lang') |
| attributeLangValue | xpath(item(), 'string(//@lang)') |
| anyAttribute | xpath(item(), '//@*') |
Results
[
{
"nodeName": "title",
"attributeLang": [
"lang=\"en\""
],
"attributeLangValue": "en",
"anyAttribute": [
"lang=\"en\""
]
},
//...
]
Example 3
Select all 'book' nodes where title has parameter lang='en', perform more complex transformations to retrieve results as string, array or a JSON object.
From: xpath(variables('oBookstoreXml'),'//book/*[@*]')
Map:
| value | expression |
|---|---|
| titleAsArray | xpath(item(), '//title/text()') |
| chapterAsArray | xpath(item(), '//chapter/name/text()') |
| titlesAsString | xpath(item(), 'string(//title)') |
| chapterAsString | xpath(item(), 'string(//chapter/name)') |
| chaptersAfterPage10AsArray | xpath(item(), '//chapter[page>10]/name/text()') |
| chaptersAndPages | xpath(item(), '//chapter/*[self::name or self::page]/text()') |
| titleAndChaptersAsJSON | json(concat('{"title":"',xpath(item(), string(//title)'),'","chapters":',xpath(item(), '//chapter/name/text()'),'}')) |
Results
[
{
"titleAsArray": [
"Harry Potter"
],
"chapterAsArray": [
"The Boy Who Lived",
"The Vanishing Glass"
],
"titlesAsString": "Harry Potter",
"chapterAsString": "The Boy Who Lived",
"chaptersAfterPage10AsArray": [
"The Vanishing Glass"
],
"chaptersAndPages": [
"The Boy Who Lived",
"1",
"The Vanishing Glass",
"15"
],
"titleAndChaptersAsJSON": {
"title": "Harry Potter",
"chapters": [
"The Boy Who Lived",
"The Vanishing Glass"
]
}
},
//...
]
string() function converts nodes to strings, while /text() is a node test that selects text nodes directly.
The xpath(item(), '//chapter/*[self::name or self::page]/text()') expression selects multiple paths in a way supported by XPath 1.0.
The xPath expression used by titleAndChaptersAsJSON is a bit more involved and is using both, Power Automate and XPath functions.
json(
concat(
'{"title":"',
xpath(item(), 'string(//title)'),
'","chapters":',
xpath(item(), '//chapter/name/text()'),
'}')
)
The json() function allows us to return an object. the concat() function generates string representing the object. Both of these functions are Power Automate functions.
The value returned as title is retrieved using string(//title) XPath expression ensuring that the returned value is a string.
The chapters list is retrieved using //chapter/name/text() XPath expression, returning a list of string values.
What didn't work.
The xpath(item(),'//book/concat(title/text(), " - ", author/text())') expression would be useful in transforming XML to an object with a simplified structure. Unfortunately, XPath 1.0 does not support the concat function directly within the path expression.
Beyond XML
xpath() is an incredibly useful and powerful function, that can speed up content transformations. And its application goes beyond to parsing XML files. Whenever the Select action is not enough, and I want to avoid using Apply to each, I transform JSON to XML and use xPath() for more power 🚀🚀🚀
Value: xml(json(concat('{ "root": { "arr": ',variables('jsonResponse'), '}}')))
If you are unsure when to use
SelectvsApply to eachaction in your workflows, see Streamlining your Power Automate Flows with the Select Action.
YOu may download the solution with Power Automate workflow including the above examples from Power-Automate xPath 101 GitHub repo.
Stay tuned for my next example, showing the sheer power of Select and xPath working together.
Top comments (0)