Convert an XML document to a Struct.

xmlToStruct {/full} {/whitespace}  StructTarget, XmlDocument

xmlToStruct {/full} {/validate } { /schema}   StructTarget, XmlDocument, SchemaList


  • /full—include comments, doctype declarations, and namespace declarations, if they are present, and leave leading and trailing whitespace in character sequences within elements. By default, the constructs are ignored and whitespace stripped. See Conversion with /full and Namespace Handling.
  • /whitespace—convert whitespace between elements to Struct members and leave leading and trailing whitespace in element values or comments. By default, such whitespace is ignored or stripped.

    This switch is only applicable if /validate or /schema are not specified. DTDs and schemas handle whitespace correctly, causing only whitespace that the DTD or schema does not deem significant to be discarded.

  • /validate—validate the XML source against its DTD or schema. A DTD can be provided as an embedded DTD or as a reference. A schema can be provided as a reference or in the SchemaList parameter. A referenced DTD or schema is expected to be located at the provided URL. If only a DTD or schema name is provided, a file with that name is expected in the working directory.

    By default, the XmlSource is only checked to ensure it is well-formed. If /schema is specified, /validate is implicitly specified. If the XML document does not conform to the DTD or schema, validation fails, error -1504 is returned, and the struct output parameter is left unchanged..

  • /schema—include schema information, if present, during conversion and validate the XML against it. By default this information is ignored. See Conversion with /schema and Data Types


  • StructTarget—variable or parameter of type struct or any to hold the returned output
  • XmlDocument—XML document to parse; value can be a string, file name, or a URL. The XML document can only contain one XML root element.

    The format of the parameter determines the type. A string must begin with the open tag angle bracket (<), and a URL must have the format Protocol://Address. Otherwise, the parameter is interpreted as a file name.

  • SchemaList—namespace and location of one or more schemas. For a single schema, SchemaList can be a file name, URL, or a Uniface index list. For multiple schemas, it must be a Uniface list of name-value pairs in the format:


    NAMESPACE cannot exceed 4096 characters; LOCATION=cannot exceed 260 characters.

Return Values

0 is returned in $status if the conversion was successful. However, non-fatal errors may have occurred during conversion, because everything that is not recognized or usable is ignored. Warnings about such conditions are made available in DETAIL sublist of $procReturnContext. See $procReturnContext for xmlToStruct.

Common Errors Returned in $procerror After xmlToStruct
Error Error Constant Meaning
-4 IOSERR_OPEN_FAILURE Specified source does not exist or cannot be opened; or specified URL is invalid or cannot be read; or the NAME or LOCATION are too long.
-13 UIOSERR_OS_COMMAND The specified file name specified by XmlDocument is too long.
-1403 UPROCERR_OPERAND The Struct Target is not a variable or parameter of type struct or any
-1503 UXMLERR_PARSE The XML source is not well-formed XML
-1504 UXMLERR_VALIDATE The XML source does not conform to the DTD.


Allowed in all Uniface component types.

On iSeries, there is no schema support, so /schema and SchemaList cannot be used.


Use the xmlToStruct statement to transform an XML document to a Struct. This Struct:

  • Always contains a nameless root Struct that represents the XML document. It can optionally hold XML declaration tags.
  • The XML root element is always the first named child struct of this nameless struct .

Thus, <movie>The Matrix</movie> is converted to the following Struct:

  [movie] = "The Matrix"

Note:  xmlToStruct cannot be used to convert XML snippets.

An XML document always has a root node and it may have constructs that appear at the same level as the root node, such as the XML declaration, the DOCTYPE declaration, and comments. Some of these constructs are handled as annotations but others may be treated as Structs if /full is used.

During conversion xmlToStruct, transforms the XML document to one Struct at the top level, which holds the root node and its sibling constructs. In addition to converting XML constructs to Struct nodes, xmlToStruct sets annotations in $tags Struct of each Struct node.

Viewing the Struct

This can be clearly seen in the string representing the Struct that is returned by $dbgString, which can be used during development to view and analyze the Struct. For example:

; INPUT_FLD is xmlstream
; OUTPUT_FLD is string
; $InputStruct$ is struct component variable

xmlToStruct /full $InputStruct$, INPUT_FLD
OUTPUT_FLD = "%%($InputStruct$->$dbgString)%%%"

If INPUT_FLD contains:

<?xml version="1.0" encoding="UTF-8"?>
<div class="note">Text can be <b>bold</b> or <em>italic</em></div>

the following is put into OUTPUT_FLD:

      [xmlClass] = element
    Text can be 
    [b] = bold
        [xmlClass] = element
    [em] = italic
        [xmlClass] = element

For each XML construct, there is a set of annotations that may be set, depending on the XML contents.

Getting Information About Structs

Other Struct functions enable you to access and manipulate the Struct members in Proc.

For example:

$1 = $InputStruct$->$tags->xmlEncoding

Result: $1 = "UTF-8"

$2 = $InputStruct$->$collSize 

Result: $2 = 1, meaning this Struct contains only one member. The $tags Struct is not counted.

For more information, see Struct Annotations and Struct Annotations for XML. For more information on the conversion between XML and Structs, see Uniface XML Constructs.

Conversion with /full

By default XML comments, DOCTYPE declaration, and namespace declarations are ignored when converting to Structs.

To include them, use the /full switch. This ensures that all XML constructs, including comments, doctype declaration, and namespace declarations (if present), are also converted to Structs. This is critical if you need to reconstruct an equivalent XML document from the Struct.

Note:  The use of /full does not imply /schema or /whitespace.

If /full is specified, and <![CDATA[...]]> occurs within the character data, it is converted to a separate nameless scalar Struct with the xmlClass tags set to CDATA. Otherwise, the <![CDATA[...]]> piece is incorporated into one scalar Struct for the whole character value; it becomes indistinguishable from the rest of the data.

Namespace Handling

Qualified names are those consisting of namespace URI that qualify a local name. For example, the local name schema can be qualified by the namespace "http://www.w3.org/2001/XMLSchema" which together provide a unique identifier.

Namespaces can be defined with an alias. For example, xmlns:s="http://www.w3.org/2001/XMLSchema" means that within the scope of this definition s is the namespace http://www.w3.org/2001/XMLSchema".

An alias can be used in a qualified name. For example, s:schema and its semantics are completely defined by the XML Schema Definition and should not be confused with any other meaning schema.

By default, xmlToStruct recognizes qualified names only in element and attribute names, and records the namespace declaration in the $tags members of the element or attribute Struct node.

When xmlToStruct/full is used, it also recognizes:

  • Namespace declarations that are not actually used in a qualified name
  • Qualified names that are used as values of elements and attributes. This type of construction is used in specialized XML documents that describe or manipulate other XML documents.

For example, in the following XML namespace definition:

<ns1:element Callout 1 xmlns:ns1="www/namespace/example1" Callout 2
                xmlns:ns2="www/namespace/example2" Callout 3 ns2:attribute=a_value> Callout 4
  1.  element is a local name within the ns1 namespace.
  2.  ns1 is an alias or abbreviation of the namespace www/namespace/example1.
  3.  ns2 is an alias or abbreviation of the namespace www/namespace/example2.
  4.  attribute is a local name within the ns2 namespace.

By default, xmlToStruct registers this in the Struct node as:

[element] Callout 1
        [xmlNamespaceURI]=www/namespace/example1 Callout 2
            [xmlClass]=attribute Callout 4
            [xmlNamespaceURI]=www/namespace/example2 Callout 3

The default behaviour records only the namespace declarations that are actually used, so if there were a third namespace declaration such as xmlns:ns3="www/namespace/example3", the Struct nodes would not contain it because it is not used in those nodes.

In the following XML example, the attribute contains a namespace definition as a value:

<ns1:element xmlns:ns1="www/namespace/example1"
             xmlns:ns3="www/namespace/example3" ns2:attribute=”ns3:a_value">

By default, xmlToStruct produces:


Notice that the value of the attribute is ns3:a_value, but there is no definition of ns3.

In these cases, you must use the /full qualifier to ensure that all of the original namespace declarations are recorded in the Struct nodes. Thus xmlToStruct/full produces:

        [xmlClass] = namespace-declaration

Conversion with /schema

If /schema is used, the specified schema (or schemas) determine which whitespace is significant and which is not.

If schemas are specified, each Namespace+Location pair in the SchemaList is added to the list of schema locations as if they were specified in the XML source itself using the schemaLocation= keyword.

There can be only one location without accompanying namespace in the list, because the XML parser supports only one location specified by the noNamespaceSchemaLocation keyword.

For example:

  • Index list has one item, a schema location:
    xmlToStruct vStruct, "file.xml", "schema.xsd"
  • Index list has one item, an associated list with only a schema location; this is equivalent to the previous command:
    xmlToStruct vStruct, "file.xml", "LOCATION=schema.xsd"

When multiple schemas are specified, Namespace+Location pairs must be passed as an associated list. For example:

vSchemaList = "" ; initialize outer list
vItems = ""      ; initialize sublist 

; create Namespace+Location sublist
putitem/id vItems, "NAMESPACE", "http://www.shiporder.com"
putitem/id vItems, "LOCATION", "shiporder.xsd"
putitem vSchemaList, -1, vItems ; add items list to outer list

; add a second pair:
vItems = ""
putitem/id vItems, "NAMESPACE", "http://www.customs.com"
putitem/id vItems, "LOCATION", "custom.xsd"
putitem vSchemaList, -1, vItems

xmlToStruct vStruct, "file.xml", vSchemaList

Data Types

If /schema is not specified, the Uniface data type of every element and attribute is string.

If the /schema switch is specified, elements and attributes with predefined data types are included in the Struct as scalar Structs of matching Uniface data types. The values themselves are converted to Uniface format.

Note:  Supplying a schema list in the third parameter does not imply that schema information is to be included; the parameter may be supplied just to satisfy /validate, not because you want the schema information.

XML - Uniface Data Type Mapping
Uniface XML Data Types Remarks
string string, normalized string, token, language, name, NCName, token, NMTOKENS, ID, IDREF, IDREFS, ENTITY, ENTITIES  
duration, gYearMonth, gYear, gMonthDay, gDay, gMonth There is no equivalent in Uniface for these XML data types
date, time, datetime date, time, dateTime If the value has the Z or [+|-]hh:mm timezone suffix, that timezone and the local machine timezone are used to calculate the corresponding time in the local timezone
numeric decimal, integer,

long, int, short, byte,

nonPositiveInteger, negativeInteger

nonNegativeInteger, positiveInteger, unsignedLong, unsignedInt, unsignedShort, unsignedByte


boolean boolean The value is T or F
float float, double If the special values INF, -INF or NaN are encountered, it is a string with that value
raw base64Binary, hexBinary Value is the result of the base64-to-raw and hex-to-raw conversion performed by xmlToStruct
  NULL An element can be declared with the attribute nillable="true" in a schema. In an XML document that has the value of an element set to NULL, this is done by adding the attribute xs:nil="true" to the element. The resulting Struct contains a member named nil with xmlClass="attribute" for each element whose value is NULL. The value of the struct member itself will be an empty string.

$procReturnContext for xmlToStruct

$procReturnContext contains context and error information about the conversion in the form of a nested Uniface list.

Context=xmlToStruct ;}
{Infos=Number ;
{Warnings=Number ;} 
{Errors=Number ;}
{DETAILS=ID=MsgNum !!;SEVERITY=Type !!;MNEM=Mnemonic !!;DESCRIPTION=ErrorDescription !!;CURRENTSTRUCT=Struct !!;ADDITIONAL=TAGNAME=Name !!!;TAGVALUE=Value !!!;EXPECTED=ExpectedValue}  { !;ID= ...}
Items Returned by $ProcReturnContext for xmlToStruct
Item Description
Context Value indicating the previously executed command that set $procReturnContext, in this case, xmlToStruct



Number of messages, warnings, and non-fatal errors generated during processing
DETAILS Details about any messages, warnings, and non-fatal errors encountered during processing, structured as a Uniface sublist
ID Message number
MESSAGE Message text
SEVERITY Importance of the issue; one of INFO, WARNING, or ERROR

Mnemonic for the specified (numeric) ID:

USTRUCTERR_TAGVALUE_NOT_APPLICABLE—Annotation xmlClass has an unknown or illegal value (based on the current context)

DESCRIPTION Short description of the issue.
CURRENTSTRUCT List of all preceding parents, starting from the top. Each parent is described by its name (which can be empty) and index number. The top-level parent has no index number.
ADDITIONAL Uniface sublist of additional information about the Struct (member) causing the message. This information is provided if there is more detailed information to report, such as unexpected tags or tag values.
TAGNAME Name of the annotation tag.
TAGVALUE Value of the tag specified by TAGNAME.
EXPECTED Expected object for the context
Struct valid on XML document level
Struct valid on XML element level
DTD declaration
DTD attribute declaration
Version Change
9.5.01 Introduced
9.6.01 Added /schema qualifier