ABAP Keyword Documentation → ABAP − Reference → Data Interfaces and Communication Interfaces → ABAP and XML → Transformations for XML
CALL TRANSFORMATION
Other versions: 7.31 | 7.40 | 7.54
Syntax
CALL TRANSFORMATION ID
| trans|(name)
[PARAMETERS {p1 = e1 p2 = e2 ...}|(ptab)]
[OPTIONS options]
SOURCE {XML src_xml}
| {{bn1 = e1 bn2 = e2 ...}|(stab)}
RESULT {XML rslt_xml}
| { {bn1 = f1 bn2 = f2 ...}|(rtab) }.
Alternatives
1. ... ID ...
2. ... trans|(name) ...
Extras
1a. ... SOURCE {XML src_xml}|{{bn1 = e1 bn2 = e2 ...}|(stab)}
1b. ... RESULT {XML rslt_xml}|{{bn1 = f1 bn2 = f2 ...}|(rtab)}
2. ... PARAMETERS {{p1 = e1 p2 = e2 ...}|(ptab)}
Effect
This statement calls the specified XSL transformation (XSLT) or a
simple transformation
(ST). The source of the transformation is specified behind SOURCE
and the
result is stored behind RESULT
. PARAMETERS
can
be used to pass parameters to the transformation. The addition
OPTIONS
can be used to specify transformation options options
. The possible transformation types are:
- from XML to XML (for XSLT only)
- from XML to ABAP (for XSLT and simple transformations only)
- from ABAP to XML (for XSLT and simple transformations only)
-
from ABAP to ABAP (for XSLT only)
JSON-XML, which maps JSON data to XML, can be used to execute all categories of transformations for JSON that are also possible for XML. More information is available in Transformations for JSON.
Notes
-
The characters <, >, &,
", and ', which have a separate meaning
in XML syntax, are handled automatically for XML data. When XML data is written, the required XML notation
<, <, >, &,
", or ' is created in
values of elements or attributes. When XML data is read, the XML notation is transformed to the appropriate
characters. It is not usually necessary to apply the escape character function
escape
or a similar method to character-like XML results of the statementCALL TRANSFORMATION
and would actually produce incorrect results. An exception to this rule is creating HTML data with XSLT in cases where it contains JavaScript parts. - When JSON data is written, special characters in character-like values are prefixed automatically with the escape character \. Conversely, the escape character is removed from character-like values when JSON data is read.
-
When a transformation is performed, methods from the class libraries for XML are used internally.
- In the case of XSLT, a DOM is always needed as a source. This DOM is constructed and processed using the iXML Library. In the case of the identity transformation ID, the iXML Library is used only if appropriate objects are specified as source or target.
- ST is processed serially, using the sXML Library.
Alternative 1
... ID ...
Effect
Specifies the identity transformation ID directly. This transformation exists as an XSL transformation in the repository, but is executed by the runtime environment in an optimized ID engine, depending on the transformation source or target. For more information, see:
Executable Example
Executable Example of an Identity Transformation
Alternative 2
... trans|(name) ...
Effect
Specifies any transformation. The transformation addition can be specified either statically or dynamically:
-
trans
trans
.
-
(name)
name
.
The specified transformation must exist as an XSLT program or as a simple transformation in the repository. If a dynamically specified transformation is not found, an exception of the class CX_INVALID_TRANSFORMATION is raised.
Executable Example
Executable Example of a Simple Transformation
Addition 1a
... SOURCE {XML src_xml}|{{bn1 = e1 bn2 = e2 ... }|(stab)}
Effect
Specifying the source.
Transformation of XML Data
When XML src_xml
is specified, the
XML data in src_xml
is transformed. src_xml
can exist in the following forms:
-
As a data object with the type
string
or as a standard table with a flat character-like row type, which can contain the following data:
- Character-like XML data in XML 1.0 format.
- Character-like JSON data. When any transformation
trans|(name)
is used, the JSON data is converted to the associated JSON XML format and passed to this transformation. The predefined identity transformation ID handles the JSON data directly.
-
As a data object with the type
xstring
or as a standard table with a flat byte-like row type, which can contain the following data:
- Byte-like representation of the characters in XML data in XML 1.0 format, in accordance with a code page such as UTF-8.
- XML data in Binary XML format.
- Byte-like representation of the characters in
JSON data, in accordance with a code page such as UTF-8. When any transformation
trans|(name)
is used, the JSON data is converted to the associated JSON XML format and passed to this transformation. The predefined identity transformation ID handles the JSON data directly.
-
As an interface reference variable for objects in the iXML Library (XSLT only):
- Interface reference variable with the type IF_IXML_ISTREAM, pointing to an input stream that reads XML data.
- Interface reference variable with the type IF_IXML_NODE, pointing to a node or a full XML document in DOM representation.
-
As a reference variable of the type IF_SXML_READER or CL_SXML_READER for
XML Reader in the
sXML Library. The transformation moves the reader by
one node and processes the entire matching subtree. If the transformation is successful, the reader is located at the end of the processed data. Any valid reader can be specified:
- XML readers for XML data in XML 1.0 format.
- XML readers for XML data in Binary XML format.
- JSON readers for JSON data. When any transformation
trans|(name)
is used, the JSON reader returns the JSON data in JSON-XML format. When a predefined identity transformation ID is used, the JSON data is handled directly.
Notes
-
The interface IF_SXML_READER and the class CL_SXML_READER are part of the API of the
sXML Library, whose objects can be accessed using
CALL TRANSFORMATION
. - Transformations of XML data in XOP format are only possible with an XOP reader in the sXML Library. All other formats can be also specified in appropriate strings or internal tables.
-
The interfaces IF_IXML_ISTREAM and IF_IXML_NODE are part of the API of the
iXML Library, whose objects can be accessed using
CALL TRANSFORMATION
. -
For compatibility reasons,
src_xml
can also be a class reference variable of the type CL_FX_READER. The class CL_FX_READER, however, is scheduled for replacement by CL_SXML_READER. - When XML data is specified in strings or internal tables, it is best to use byte strings or byte-like row types and the UTF-8 code page for the representation.
-
The methods CONVERT of the interfaces IF_ABAP_CONV_IN and IF_ABAP_CONV_OUT
of objects created using the class CL_ABAP_CONV_CODEPAGE can be used to transform character-like XML data or JSON data to a byte-like representation in accordance with a code page.
Transformation of ABAP Data
bn1 = e1 ...
or (stab)
is used to specify the ABAP data e1 e2 ...
to transform.
- When the identity transformation ID is called, the ABAP data are either serialized to the canonical XML format asXML or to the canonical JSON format asJSON, depending on the target.
-
When an XSLT program is called, the ABAP data is serialized to the
canonical XML representation, which is then used
as the source of the XSL transformation.
bn1 bn2 ...
is used to specify the names of the XML elements that represent the ABAP data objects in the canonical XML representation. -
When a simple transformation is called, the transformation has read-only access to the ABAP data using the names
bn1 bn2 ...
.
Instead of using a static parameter list, the data objects can be passed dynamically as value pairs in the columns of an internal table stab
that has the type ABAP_TRANS_SRCBIND_TAB from the
type group ABAP.
Serializations convert the elementary components of the ABAP data objects to asXML or asJSON, in accordance with the mappings for built-in ABAP types or for further XML schema data types. This can raise the exceptions described there, some of which can be bypassed by using transformation options.
The case of the names in the XML data or JSON data depends on how they are specified in the ABAP runtime environment. If specified statically
(b1
, b2
, ...), uppercase is used; if specified dynamically in stab
, the case used there is used.
Note
When a simple transformation is called, the serialization always creates an internal XML writer object, which can be addressed using the interface IF_SXML_WRITER from sXML Library. The object is accessed using attributes of the ST statement tt:call-method.
Executable Examples
- For an overview of all possible XML sources, see XML Sources of Transformations.
-
For deserializations of ABAP data from an XML reader in the sXML Library, see sXML-Library, Transformation from XML Reader.
Addition 1b
... RESULT {XML rslt_xml}|{{bn1 = f1 bn2 = f2 ...}|(rtab)}
Effect
Specifying the target.
Transformation to XML Data
When XML rslt_xml
is specified, a transformation to
XML data takes place and this data
is passed to rslt_xml
. rslt_xml
can exist in the following forms:
-
As a variable with the type
string
or as a standard table with a flat character-like row type. The XML data returned by the transformation is stored in XML 1.0 format as a character string in the string or in the table. Here, a byte order mark (BOM) is used as a prefix. -
As a variable with the type
xstring
or as a standard table with a flat byte-like row type. The XML data returned by the transformation is saved in XML 1.0 format and as a byte string in the string or table. The byte string represents the characters of the XML data in accordance with the code page UTF-8. A byte order mark (BOM) is not used as a prefix. -
As an inline declaration
DATA(var)
, where a variable with the typexstring
is created. -
As an interface reference variable for objects in the iXML Library (XSLT only):
- Interface reference variable with the type IF_IXML_OSTREAM, pointing to an output stream that writes XML data.
- Interface reference variable with the type IF_IXML_DOCUMENT that points to an XML document in DOM representation.
-
As a reference variable with the type IF_SXML_WRITER or CL_SXML_WRITER for
XML writers in the
sXML Library. The result of the transformation is added to the current element of the writer as a subtree. Any valid writer can be specified:
- XML writers for XML data in XML 1.0 format.
- XML writers for XML data in Binary XML format.
- JSON writers for JSON data. A transformation
trans|(name)
must create XML data in JSON-XML format. The predefined identity transformation ID creates the corresponding data directly.
Notes
-
The interface IF_SXML_WRITER and the class CL_SXML_WRITER are part of the API of the
sXML Library, whose objects can be accessed using
CALL TRANSFORMATION
. - Transformations to formats other than XML 1.0 are only possible with the appropriate writers in the sXML Library.
-
The interfaces IF_IXML_OSTREAM and IF_IXML_DOCUMENT are part of the API of the
iXML Library, whose objects can be accessed using
CALL TRANSFORMATION
. -
For compatibility reasons,
rslt_xml
can also be a class reference variable of the type CL_FX_WRITER. However, the class CL_FX_WRITER is replaced by CL_SXML_WRITER. - When XML data is saved in strings or internal tables, it is best to use byte strings or byte-like row types and the UTF-8 code page for the representation.
- XML data that is written to Unicode systems as character-like data cannot be processed as character-like data in a non-Unicode system, due to the byte order mark (BOM) that is predefined by the XML standard.
-
The methods CONVERT of the interfaces IF_ABAP_CONV_OUT and IF_ABAP_CONV_IN
of objects created using the class CL_ABAP_CONV_CODEPAGE can be used to transform character-like XML data or JSON data to a byte-like representation in accordance with a code page.
Security Note
If HTML data containing JavaScript is created using the output method "html" of an XSL transformation, the syntax characters <, >, &, ", and ' in the JavaScript parts are not transformed to their XML notation. To prevent cross site scripting (XSS) in HTML data that contains dynamic JavaScript parts, appropriate measures must be taken. These measures are best applied in the XSL transformation that creates the data.
Transformation to ABAP Data
bn1 = f1 ...
or (rtab)
are used to specify the ABAP target fields f1 f2 ...
into which the XML data is transformed.
-
When the identity transformation ID is called, the source must either be a
canonical XML representation or a
canonical JSON representation for ABAP data. The canonical
display format is deserialized in the ABAP data objects.
bn1 bn2 ...
is used to specify the names of the XML elements that represent the ABAP data objects in the canonical XML representation, andf1 f2 ...
is used to specify ABAP data objects (appropriate to the data type) into which they are deserialized. - When an XSLT program is called, the result must represent the XSL transformation into a canonical XML representation, which is deserialized into the ABAP data objects (like with the identity transformation).
-
When a simple transformation is called, the transformation performs writes on the ABAP data using the names
bn1 bn2 ...
.
Instead of using a static parameter list, the data objects can be passed dynamically as value pairs to the columns of the internal table rtab
that has the type ABAP_TRANS_RESBIND_TAB from the
type group ABAP.
Deserialization converts the elementary XML or JSON values to elementary ABAP data types in accordance with the mappings for built-in ABAP types or for further XML schema data types. This can raise the exceptions described there, some of which can be bypassed by using transformation options.
The case used in the XML or JSON data must match exactly the case specified in the ABAP runtime environment. If specified statically
(b1
, b2
, ...), uppercase is used; if specified dynamically in rtab
, the case used there is used.
When deserializing from a canonical representation, if no corresponding asXML element or asJON object
component exists for an ABAP object, the data object retains its previous value. If an empty XML element
or an empty JSON array is assigned to an elementary or tabular ABAP data object, the ABAP data object
is set to its type-dependent initial value. If an empty XML element or empty JSON object is assigned
to a structure, this has the same effect as non-existent elements on the structure components. The structure
components retain their previous values. The transformation option
clear
with the value "all" can be used to initialize all data objects that have nonexistent or empty elements assigned to them.
In the case of the deserialization of internal tables with non-unique table keys, the order of the duplicate rows in relation to these keys is not retained.
Note
When a simple transformation is called, deserialization always creates an internal XML reader object, which can be addressed using the interface IF_SXML_READER from the sXML Library. The object is accessed using attributes of the ST statement tt:call-method.
Executable Examples
- For an overview of all possible XML sources, see XML Sources of Transformations.
- For transformations of XML data to DOM representation in the iXML Library, see iXML Library, DOM Representation.
- For transformations of ABAP data to an XML writer in the sXML Library, see sXML Library, Transformation to XML Writer..
-
For deserializations of missing or empty asXML elements or asJSON object components, see:
-
For transformations of syntax characters to XML targets, see Transformation of Syntax Characters.
Addition 2
... PARAMETERS {{p1 = e1 p2 = e2 ...}|(ptab)}
Effect
This addition can be used to bind ABAP data objects e1 e2 ...
to the parameters p1 p2 ...
of an
XSL transformation or a
simple transformation. All elementary data objects and object references are permitted.
Instead of using a static parameter list, the parameters can be specified dynamically as value pairs in the columns of the internal table ptab
that has one of the following types from the
type group ABAP:
- ABAP_TRANS_PARMBIND_TAB for specifying strings,
- ABAP_TRANS_OBJBIND_TAB for specifying object references (for XSLT only),
-
ABAP_TRANS_PARM_OBJ_BIND_TAB for specifying data references (for XSLT only).
Parameter Passing to XSL Transformations
The specified parameters must be defined as input parameters in an XSL transformation as follows:
<xsl:param name="..." sap:type="..." /
>
Specify the parameter name for the attribute name
in uppercase letters. Before
a binding can take place, the case used in the XML document must match exactly the case specified in the ABAP runtime environment. If specified statically
(b1
, b2
, ...), uppercase is used; if specified dynamically in rtab
, the case used there is used.
For the optional attribute type
, one of the type names string
,
number
, boolean
, nodeset
,
xstring
, or object(...)
can be defined, with a
global ABAP class specified in parentheses behind object
. The namespace before
type
must be defined as xmlns:sap="http://www.sap.com/sapxsl"
, with the prefix sap
proposed as a convention.
If no type is specified in the transformation, the data types of elementary parameters are mapped to XSL types in accordance with the following table.
ABAP Data Type | XSL Parameter Type |
---|---|
c , d , n , string |
string |
i (b , s ), f , p |
number |
x , xstring |
string , where the content is represented in Base 64. |
If the XSL types shown in the table above are specified explicitly in the XSL transformation, appropriate elementary ABAP parameters must be specified that can be converted to the XSL type:
-
The XSL type
boolean
expects ABAP parameters of the typec
with the length one. A blank is interpreted as "false" and any other character as "true". -
The XSL type
xstring
expects ABAP parameters of the typex
orxstring
and the content is represented as hexadecimal. -
The XSL types
nodeset
andobject
expect an object reference variable that points to an instance of a class. The typenodeset
expects appropriate object properties.
If a parameter does not match the XSL type, a non-handleable exception is raised. If a parameter defined in the XSL transformation is not passed, it is set to a default value in the transformation. Any specified parameters that are not defined in the XSL transformation are ignored.
Notes
-
The XSL types
string
,number
,boolean
, andnodeset
are standard XSL types.xstring
andobject
represent special SAP enhancements. The typexstring
enables byte strings to be represented in hexadecimal instead of base 64. The typeobject
enables ABAP methods to be called from the XSLT program. -
The special addition
OBJECTS
can be used to pass object references, but this addition is replaced byPARAMETERS
.
Parameter Passing to ST Programs
In an ST program, the parameters must be defined as follows as parameters of the main template outside a template:
<tt:parameter name="..." [kind="..."]
[[s-val="..."][d-val="..."]]|[val="..."] />
When an ST program is called, output parameters and input/output parameters are supported as well as
input parameters. The type of the parameter is defined using kind
in the
ST program, where "in"
, "out"
, and "inout"
can be specified. "inout" is the default. The val
attributes can be used to define replacement values.
Exceptions
Handleable Exceptions
CX_TRANSFORMATION_ERROR
-
Cause: Common superclass of all exception classes for CALL
TRANSFORMATION. The direct subclasses are CX_ST_ERROR for simple
transformations and CX_XSLT_EXCEPTION for XSL transformations. When the
statement
CALL TRANSFORMATION
is executed, exceptions of the class CX_TRANSFORMATION_ERROR and almost all of its subclasses can be handled. The resumability of an exception of the subclass CX_ST_DESERIALIZATION_ERROR can be defined using the transformation optionexceptions = 'resumable'
.
- Exceptions like those raised in mappings from ABAP data types to asXML cannot be handled directly. These exceptions raise exceptions of the class CX_TRANSFORMATION_ERROR (or its subclasses) and can be handled using these.
- Exceptions raised in ABAP methods called in transformations (such as transformations of objects using the methods
SERIALIZE_HELPER and
DESERIALIZE_HELPER or when
accessing ABAP Objects from ST) raise exceptions
of the class CX_TRANSFORMATION_ERROR (or its subclasses), if they are declared in the interface of the method
RAISING
. The resumability of the original exception is applied here. Non-declared exceptions, on the other hand, raise the exception CX_SY_NO_HANDLER, which cannot be handled using CX_TRANSFORMATION_ERROR (or its subclasses).
Subclasses of CX_ST_ERROR
CX_ST_CALL_ERROR
-
Cause: Error when calling the transformation
CX_ST_CALL_METHOD_ERROR
-
Cause: Exception in an ABAP method called from the transformation
CX_ST_CONDITION
-
Cause: Error in a transformation condition
CX_ST_CONSTRAINT_ERROR
-
Cause: Data object to be serialized or deserialized is too long.
CX_ST_DESERIALIZATION_ERROR
-
Cause: Error in the deserialization
CX_ST_FORMAT_ERROR
-
Cause: Format error in a node
CX_ST_INVALID_XML
-
Cause: A serialization would produce invalid XML.
CX_ST_MATCH
-
Cause: Deserialization match failed
CX_ST_REF_ACCESS
-
Cause: Error in data node access
CX_ST_RUNTIME_ERROR
-
Cause: Error in execution of the transformation
CX_ST_SERIALIZATION_ERROR
-
Cause: Error in the serialization
CX_ST_VALIDATION_ERROR
-
Cause: Failed validation with the attribute
xsd-type
in the ST statements tt:value, tt:write, or tt:read. This exception cannot be caught directly.
Subclasses of CX_XSLT_EXCEPTION
CX_XSLT_ABAP_CALL_ERROR
-
Cause: Error when calling an ABAP method from XSLT
Runtime errorXSLT_METH_ERROR
CX_XSLT_DESERIALIZATION_ERROR
-
Cause: Error in the deserialization
Runtime error:XML_NOT_SUPPORTED
CX_XSLT_FORMAT_ERROR
-
Cause: Document is not in asXML format
Runtime error:XML_FORMAT_ERROR
CX_XSLT_RUNTIME_ERROR
-
Cause: Error in XSLT processor
Runtime error:UNCAUGHT_EXCEPTION
CX_XSLT_SERIALIZATION_ERROR
-
Cause: Error in the serialization
Runtime error:XML_NOT_SUPPORTED