XPath Tester (XPath 1.0 for XML and HTML)

XPath selects parts of an XML or HTML document with path expressions such as //book[price > 10]/title. Scrapers, XSLT, Selenium and Playwright locators, SOAP clients and many config tools rely on it, and a typo silently selects nothing. Paste a document, type an expression and see exactly which nodes match — or the number, string or boolean an expression like count() returns — using your browser's own XPath 1.0 engine.

How to use this tool

1. Paste XML or HTML and choose how to parse it.
2. Type an XPath expression and press Calculate.
3. For namespaced XML, prefixes declared in the document are bound for you; add others as prefix=URI lines.
4. Read the matches (text content, type, name and serialized markup) or the computed value.

Formula

/ child · // descendant · @ attribute · [predicate] · text() · count() sum() string() contains() starts-with() position() last()

This is XPath 1.0, the version browsers implement (XPath 2.0/3.1 functions such as matches() or lower-case() are not available). Node-sets are listed in document order. An unprefixed name never matches an element in a default namespace (xmlns="…") — bind a prefix to that URI and use it, as XPath 1.0 requires.

Example

In the sample catalog, //book[price > 10]/title matches 2 nodes: "XML Developer's Guide" (44.95) and "Maeve Ascendant" (12.50). count(//book) returns the number 3, count(//book[@lang]) returns 2 (two books have a lang attribute), and //book/@id lists bk101, bk102 and bk103.