XPath Operators

XPath Queries

/ (child operator)

Refers to the root of the XML document when used at the beginning of the XPath expression. The child operator is used to specify the next child to select. The expression "/Employees/Employee", for, example says, start at the root of the XML document, select the Employees node and then select all the Employee child nodes within the Employees node. This will return the two Employee nodes in the sample XML document.

// (recursive descendant operator)

The recursive descendant operator indicates to include all descendant nodes in the search. Using the operator at the beginning of the XPath expression means you start from the root of the XML document. The expression "//LastName" starts at the root and finds any LastName node. The expression "/Employees//LastName" selects the Employees node and then, within that node, finds any LastName node. It yields the same result, but searches in a different way.

* (wildcard operator)

The wildcard operator finds any node. The expression "/*" finds any node under the root, which in our case is Employees. The expression "/Employees/*" means find any node under the Employees node, which in our case results with the two Employee nodes. Now what is the difference between the "/Employees" and "/Employees/*" expression? The first expression returns the Employees node but the second node finds any node under the Employees node, meaning it returns the two Employee nodes. The expression "//*" means to select any node including descendant nodes, so it will effectively list every single node in the complete XML document.

. (current context operator)

The current context operator refers to the current context. For example, you have written some code that selected the Employees node and then from there you run the expression "./Employee", which means it starts out from the currently selected Employees node and then selects the two Employee nodes. The expression "Employee" would yield the same result because it also starts out from the current context. Similar the expression ".//LastName" means start from the current context, the Employees node, and find any LastName node including any descendant nodes.

.. (parent operator)

The parent operator refers to the parent. For example, the expression "/Employees/Employee/.." returns the Employees node because you navigate down to the Employee nodes and then tell it to return its parent, which is the Employees node.

@ (attribute operator)

The attribute operator refers to an attribute instead of an element. The expression "/Employees//@ID" selects any ID attribute it finds under the Employees node. Now, keep in mind that the XPath query always returns the selected node. In the case of an attribute, the node below it is its value. So, the expression really two returns nodes, each with the value of each selected attribute. Furthermore, you can use the wildcard operator with attributes, so "/Employees//@*" means any attribute underneath the Employees node.

[ ] (filter operator)

You can apply a filter operator to filter the selected nodes. This works with attributes and with elements. The expression "/Employees/Employee[@ID=1]" returns any Employee node under the Employees node that has an ID attribute with the value one. You also can apply filters that just say that an attribute or element with that name needs to be present. For example, the expression "/Employees/Employee[WebAddress]" returns Employee nodes that have a WebAddress node as child. The expression "/Employees/Employee[FirstName='Klaus']" returns the Employee node that has a FirstName node with the value Klaus.

text() function

The "text()" function refers to the text of the selected node or attribute. The expression "//Employee//text()" does not list all the descendant nodes of all Employee nodes but rather the value for each descendant node. The expression "//Employee/FirstName[text()='Klaus']" lists all FirstName nodes which have a value of Klaus.

[ ] (collection operator)

When your expression returns more then one node with the same name, you have a collection returned. The expression "//Employee" returns two Employee nodes, which is nothing more than a collection of Employee nodes. You can apply a collection operator and specify which item from the collection you want to select. Keep in mind that the index starts at one. The expression "//Employee[2]" returns the second Employee node. The order of the selected nodes is the same order as in the XML document. You can use the collection operator in any blend, such as "//Employee[1]/LastName", which selects the first Employee node and then from there the LastName node.

( ) (group operator)

The collection operator can sometimes have some odd side effects. Assume you have two Employee nodes and each has two Job nodes. What does the expression "//Employee/Job[1]" return? It returns the first Job node for each selected Employee node. But, using the group operator allows you to apply explicit precedence to selections. The expression "(//Employee/Job)[4]" first selects all Job nodes for all Employee nodes and from that collection it returns the fourth node. The group operator can only be applied to the top level expression; for example, "//Employees/(Employee/FirstName)" is invalid.

comment() function

Returns a comment node. The expression "//comment()" returns any comment node in the XML. The expression "/Employees/comment()" returns the comment nodes under the Employees node.

node() function

XML documents consist of elements, attributes, and their values, each being a node. So, in XPath expressions you can use a node() function instead of a node name or the text() function. It is a generic way to address a node. The expressions "//Employee/JobTitle/node()" and "//Employee/JobTitle/text()" return the same result, the value of both JobTitle nodes. But, "//Employee//node()" will not just return the elements but also the values of each element, because both are nodes.
| (union or set operator)

Returns the union of one or more location paths. The expression "//LastName
//FirstName" returns all the LastName and FirstName nodes. It preserves the order of the elements as in the XML and does not return any duplicates. The two location paths "//Employee[@ID=1]
//Employee[FirstName='Klaus']" return the same nodes but the union of these two returns just the one unique node.