XML functions

xf:attribute() / xxf:attribute()

This function is available in both the xf and xxf namespaces, and we recommend you use it with the xf namespace.

xf:attribute(
    $qname as xs:anyAtomicType
) as attribute()

xf:attribute(
    $qname as xs:anyAtomicType,
    $value as xs:anyAtomicType?
) as attribute()

The xf:attribute() function returns a new XML attribute with the qualified name provided as first argument. If the qualified name is not of type xs:QName, its string value is used. If it has a prefix, it is resolved using in-scope namespaces. The second argument is an optional value for the attribute. It defaults to the empty string.

<!-- Add an attribute called "id" with a value of "first-name" to element "section" -->
<xf:insert
    context="section"
    origin="xf:attribute('id', 'first-name')"/>

The first argument can be of type xs:QName:

<!-- Add an attribute called "id" with a value of "foo:bar" to element "section"
     and resolve the namespaces on element $element -->
<xf:insert
    context="section"
    origin="xf:attribute(resolve-QName('foo:bar', $element), 'first-name')"/>

xxf:call-xpl()

xxf:call-xpl(
    $xplURL        as xs:string,
    $inputNames    as xs:string*,
    $inputElements as element()*,
    $outputNames   as xs:string+
) as document-node()*

This function lets you call an XPL pipeline.

• $xplURL is the URL of the pipeline. It must be an absolute URL.

• $inputNames is a sequence of strings, each one representing the name of an input of the pipeline that you want to connect.

• $inputElements is a sequence of elements to be used as input for the pipeline. The $inputNames and $inputElements sequences must have the same length. For each element in $inputElements, a document is created and connected to an input of the pipeline. Elements are matched to input name by position, for instance the element at position 3 of $inputElements is connected to the input with the name specified at position 3 in $inputNames.

• $outputNames is a sequence of output names to read.

The function returns a sequence of document nodes corresponding the output of the pipeline. The returned sequence will have the same length as $outputNames and will correspond to the pipeline output with the name specified on $outputNames based on position.

The example below shows a call to the xxf:call-xpl function, calling a pipeline with two inputs and one output:

xxf:call-xpl(
    'oxf:/examples/sandbox/xpath/run-xpath.xpl',
    (
        'input',
        'xpath'
    ),
    (
        instance('instance')/input,
        instance('instance')/xpath
    ),
    'html'
)

xxf:classes()

xxf:classes() as xs:string*
xxf:classes($el as node()) as xs:string*

Returns for the context element or the given element if provided, all the classes on the class attribute.

xxf:create-document()

xxf:create-document() as document-node()

The xxf:create-document() creates a new empty XML document. You can then insert new data into that document with the xf:insert action.

<xf:var name="new" value="xxf:create-document()"/>
<xf:insert context="$new" origin="instance('my-data')"/>

xf:element() / xxf:element()

This function is available in both the xf and xxf namespaces, and we recommend you use it with the xf namespace.

xf:element(
    $element-name as xs:anyAtomicType
) as element()

xf:element(
    $element-name as xs:anyAtomicType,
    $content      as item()*
) as element()

The xf:element() function returns a new XML element with the qualified name provided. If the qualified name is not of type xs:QName, its string value is used. If it has a prefix, it is resolved using in-scope namespaces.

<!-- Insert an element called "value" as a child of element "section" -->
<xf:insert
    context="section"
    origin="xf:element('value')">

The second, optional argument can take a sequence of items specifying attributes and content for the new element:

<!-- Insert an element called "value" as a child of element "section",
     with an attribute and text content -->
<xf:insert
    context="section"
    origin="xf:element('value', (xf:attribute('id', 'my-value'), 'John'))"/>

The first argument can be of type xs:QName:

<!-- Insert an element called "foo:bar" as a child of element "section"
     and resolve the namespaces on element $element -->
<xf:insert
    context="section"
    origin="xf:element(resolve-QName('foo:bar', $element))"/>

xxf:evaluate()

xxf:evaluate(
    $xpath as xs:string
) as item()*

The xxf:evaluate() function allows you to evaluate XPath expressions dynamically. For example:

<xf:input ref="xxf:evaluate(concat('instance(''my-instance'')/document', my-xpath))">
    <xf:label>...</xf:label>
</xf:input>

xxf:extract-document()

xxf:extract-document(
  $element as element()
) as document-node()

xxf:extract-document(
  $element               as element(),
  $excludeResultPrefixes as xs:boolean
) as document-node()

xxf:extract-document(
  $element               as element(),
  $excludeResultPrefixes as xs:boolean,
  $readonly as xs:boolean
) as document-node()

The xxf:extract-document() function extracts a new XML document from a document fragment under an enclosing element. For example with the following instance:

<xf:instance id="my-instance">
    <library>
        <book>
            <title>Jacques le fataliste et son maître</title>
            <author>Denis Diderot</author>
        </book>
    </library>
</xf:instance>

The expression:

xxf:extract-document(instance('my-instance')/book, '', false())

returns a new XML document rooted at the <book> element:

<book>
    <title>Jacques le fataliste et son maître</title>
    <author>Denis Diderot</author>
</book>

• $excludeResultPrefixes: optional parameter; contains a list of space-separated namespace prefixes to exclude. Defaults to the empty string.

• $readonly: optional parameter; when set to true(), return a readonly instance. Defaults to false().

xxf:form-urlencode()

xxf:form-urlencode($document as node()) as xs:string

Performs application/x-www-form-urlencoded encoding on an XML document as per the XForms specification.

xxf:has-class()

xxf:has-class($class-name as xs:string) as xs:boolean
xxf:has-class($class-name as xs:string, $el as node()) as xs:boolean

Returns whether the context element, or the given element, has a class attribute containing the specified class name.

xxf:mutable-document()

xxf:mutable-document(
    $node as node()
) as document-node()

The xxf:mutable-document() function takes a document as input and returns a mutable document, i.e. a document on which you can for example use xf:setvalue.

<xf:action ev:event="xforms-submit-serialize">
  <!-- Get initial document to submit -->
  <xf:var
    name="request-document"
    value="xxf:mutable-document(saxon:parse(/my/request))"/>
  <!-- Set value -->
  <xf:setvalue
    ref="$request-document/my/first-name">Joe</xf:setvalue>
  <!-- Serialize request document -->
  <xf:setvalue
    ref="event('submission-body')"
    value="saxon:serialize($request-document, instance('my-output-instance'))"/>
</xf:action>

Note that by compatibility with the XSLT document() and XPath 2.0 doc() functions, and unlike the instance() function, xxf:mutable-document() returns a document node, not a document element.

xxf:serialize()

xxf:serialize(
    $item as node(),
    $format as xs:string?
) as xs:string

The xxf:serialize() function allows you to serialize an XML node to XML, HTML, XHTML or text. For example:

<xf:bind ref="my-html" calculate="xxf:serialize(instance('my-instance'), 'html')"="">

xxf:sort()

xxf:sort(
    $sequence   as item()*,
    $sort-key   as item(),
    $datatype   as xs:string?,
    $order      as xs:string?,
    $case-order as xs:string?,
    $lang       as xs:string?,
    $collation  as xs:string?,
    $stable     as xs:string?
) as item()*

The second argument differs from the exf:sort() function in that it doesn't take a plain string but a literal expression, for example:

<xf:itemset
  ref="
    xxf:sort(
        instance('samples-instance')/file,
        @name,
        'text',
        'ascending'
    )">
  ...
</xf:itemset>

The last 3 arguments are available since Orbeon Forms 2021.1.9, 2022.1.4, and 2023.1. They follow the XSLT 2.0 semantics for the corresponding attributes of <xsl:sort> (see recommendation). Those additional arguments are of particular significance when you need the sorting to follow particular, often language-dependent, rules. For instance, xxf:sort(//first-name, ., 'text', 'ascending', 'upper-first', 'pl') sorts the <first-name> in the current document according to the rules of the Polish language.