Guide towards a simple conversion of an XML file to ABAP Internal table, using XML parsing.

Applies to:

SAP ECC 6.0. And further.

This article elaborates the conversion process of an XML file residing in the SAP Application server, into an ABAP internal table.

An effortless approach, unlike the other intricate methods available for conversion, has been presented here. In this particular concept, initially, the XML file is converted into a string. Hence, we use the XML parsing technique to convert that string into an ‘x’ string and finally into an internal table.

A simple conversion idea, along with the supporting code blocks, has been demonstrated in this document.

Author(s):    Aastha Mehrotra

Company:    Larsen & Toubro Infotech

Created on:  05January, 2012

Author Bio

Aastha Mehrotra has been engaged with Larsen and Toubro Infotech for two years and three months, her acquaintance with the IT industry being the same.

The author has worked on various projects in SAP with ABAP. She is also trained in a couple of other ABAP technologies like WebDynpro for ABAP and Workflows. However, the milestone in her career by far, has been her distinguished work in the field of Data Extraction from MDM to SAP R/3.  Aastha did a major research on retrieving flat as well as lookup data from MDM into SAP R/3 system, using the MDM ABAP APIs. She also applied her know how on MDM ABAP APIs, in two implementation projects. She has also imparted trainings on Setting up connection between MDM and SAP R/3 as well as Extracting data from MDM to SAP R/3.


Various practices and procedures are available for XML-ABAP conversion. However, I discovered that, none of those takes us entirely across the process of converting an XML residing in the Application Server into an internal table in ABAP. Certain other methods, which might actually take us across, are difficult to comprehend and involve a number of complicated steps.

I used this method while working on an object which required me to

  • Fetch a request XML file from the Application server.
  • Convert it into an internal table in ABAP.
  • Extract data from ERP corresponding to the data in the internal table (from the XML file).
  • Generate a Response file consisting of both XML file data as well as data from ERP.

This document shall let you know the entire process involving the evolution of the XML file in the Application server to an Internal Table.

The XML file

Following is an archetype of the XML file which needs to be converted into an internal table in ABAP.

This XML file carries data for three fields- HomePernr, UNAME and USERID. We need to convert the XML file into an internal table comprising of these three fields, and holding the three records from XML.

We would be considering the example of the following file through the course of this document. This would assist us convert any XML containing any number of fields and records into an internal table, in an uncomplicated manner.

Note: FileFeed happens to be the root node for the XML.

EmployeeRequests is the element node containing the separate Employee records within it.

Employeerequest is the element node for each record. 

HomePERNR, UNAME and USERID are the value nodes containing the values for the records.

Step wise conversion of the XML to Internal Table

The entire mechanism would be elaborated by this document in several steps, supported with the appropriate code snippets.

The series of execution would be as follows:

  • Open the XML in order to read the data.
  • Transfer the contents into an internal table.
  • Concatenate the lines of the internal table into a string.
  • Convert the string into a hexadecimal string – The ‘X’ string.
  • Parse the ‘x’ string to convert the data into an XML table.
  • Read the XML table to transfer the data into the table with the required structure.

Note: The code snippets are actually parts of a single code, in a sequence. Hence, these can be conjoined together to achieve the entire code for XML_ABAP conversion.

We shall advance with the specification of the step wise conversion process. However, we must be aware of the variables to be declared, in advance. This would enhance our understanding of the steps and would provide clarity to the context.

Type declarations for the variables

Following are the declarations for all the variables to be used in the code snippets throughout the document. This would help us avoid all the confusion pertaining to the type of variables used.

*      Declaring the file type

DATA:  g_unixfilename TYPE zpathfile.                 “UNIX file path.

*      Declaring the structure for the XML internal table


         raw(2000) TYPE c,

       END OF ty_xml.

*      Declaring the XML internal table

DATA:  g_t_xml_tab TYPE TABLE OF ty_xml INITIAL SIZE 0.

*      Declaring the work area for the XML internal table

DATA:  wa_xml_tab TYPE ty_xml.

*      Declaring the string to contain the data for the XML internal table

DATA:  g_str TYPE string.

*      Declaring the string to contain x string

DATA:  g_xmldata TYPE xstring.

*      Declaring the table to contain the parsed data

DATA:  g_t_xml_info TYPE TABLE OF smum_xmltb INITIAL SIZE 0.

*      Declaring the work area for the internal table containing the parsed data.

DATA:  g_s_xml_info LIKE LINE OF g_t_xml_info.

*      Declaring the table to contain the returned messages from the parsing FM

DATA:  g_t_return TYPE STANDARD TABLE OF bapiret2.

*      Declaring the work area for the return table

DATA:  wa_return LIKE LINE OF g_t_return.

*      Declaring the structure for the table containing fields in the XML file

TYPES: BEGIN OF struc_people,


         Uname(4) TYPE c,


       END OF struc_people.

*      Declaring the internal table containing the fields in the XML file

DATA:  g_t_employeerequest TYPE TABLE OF struc_people.

*      Declaring the work area for the internal table containing the fields in the      *      XML file

DATA:  g_s_employeerequest LIKE LINE OF g_t_employeerequest.

  •      Open the XML to read data

The first breakthrough would be to open the XML file and read the data using the OPEN DATASET statement.

The addition IN TEXT MODE, to the OPEN DATASET opens the file as a text file. The addition ENCODING defines how the characters are represented in the text file. While writing the data into a text file, the content of a data object is converted to the representation entered after ENCODING, and transferred to the file.

* Open the XML file for reading data


  IF sy-subrc NE 0.

    MESSAGE ‘Error opening the XML file’ TYPE ‘E’.


  •   Transfer the contents into an internal table of a specific type.

The next furtherance would be to move the contents of the file to an internal table. Following is the piece of code which would help us Read the data from the file into an internal table.

We use the READ DATASET statement to read the data from the file.



* Transfer the contents from the file to the work area of the internal table

      READ DATASET g_unixfilename INTO wa_xml_tab.

      IF sy-subrc EQ 0.

        CONDENSE wa_xml_tab.

*       Append the contents of the work area to the internal table

        APPEND wa_xml_tab TO g_t_xml_tab.






* Close the file after reading the data

  CLOSE DATASET g_unixfilename.

  •   Concatenate the lines of the internal table into a string.

Next, we move the lines of the internal table into a string. Hence, we get a string containing the data of the entire XML file.


*Transfer the contents from the internal table to a string

IF NOT g_t_xml_tab IS INITIAL.

    CONCATENATE LINES OF g_t_xml_tab INTO g_str SEPARATED BY space.


  •   Converting the normal string into an ‘X’ string.

We need to convert the string thus formed into a ‘x’ string or a hexadecimal string using the function Module ‘SCMS_STRING_TO_XSTRING’.The type xstring allows a hexadecimal display of byte chains instead of the presentation to the base of 64.


* The function module is used to convert string to xstring



    text   = g_str


    buffer = g_xmldata


    failed = 1

    OTHERS = 2.

  IF sy-subrc 0.

     MESSAGE ‘Error in the XML file’ TYPE ‘E’.


  •     Parsing the x string in order to convert the data into an XML table.

In case we have an XML string that needs to be converted into an object, then the XML string needs to be parsed. The parsing would convert the string into a table. The internal table that would be returned is g_t_xml_info. G_t_xml_info is of type SMUM_XML_TB. The structure for SMUM_XML_TB is in the screen-shot below:

The table SMUM_XML_TB has four fields: HIER, TYPE, CNAME and CVALUE. The XML data would be transferred into the four fields.

Moving forward, we would come to know that how exactly is the data allocated to the fields of this table?



* This function module is used to parse the XML and get the

* data in the form of a table



     xml_input = g_xmldata


     xml_table = g_t_xml_info

     return    = g_t_return


      OTHERS    = 0.

*”If XML parsing is not successful, return table will contain error messages

  READ TABLE g_t_return INTO wa_return WITH KEY type = ‘E’.

  IF sy-subrc EQ 0.

    MESSAGE ‘Error converting the input XML file’ TYPE ‘E’.


    REFRESH g_t_return.


  •   Read the XML table to transfer the data into the required table.

We are acquainted with the fields of table g_t_xml_info HIER, TYPE, CNAME and CVALUE.

We now, need to move the data from this table into the required, final table. For instance in the scenario we have been considering, the final table would consist of the three fields HomePernr, UNAME and USERID. The transfer of data from the table g_t_xml_info to our final table would be done in accordance with the values held by the four fields of the g_t_xml_info table.

While parsing, the values are assigned to the table g_t_xml_info. The values are moved depending upon the node being converted to the table.

Let me elaborate about the values held by the table g_t_xml_info.

1. Values held by the field HIER in XML table

The field ‘Hier’ would hold the value ‘1’ for the element in the root node. Since, Root node is the first level of the XML hierarchy.

‘Hier’ would hold the value ‘2’ for the fields in the Element node, which is at the second level of hierarchy. In our example EmployeeRequests is at the second level of hierarchy.

It would hold ‘3’ for Element nodes at the third level of hierarchy in the XML file that is EmployeeRequest here.

The value would be ‘4’ for all the Value nodes.  Here the value nodes would be HomePernr, UNAME and USERID.

Hence, we conclude that the value of ‘Hier’ depends upon the level of hierarchy that particular node holds in the XML file.

2.  Values held by the field TYPE in XML table

The value that the field ‘Type’ holds for all the elements in the Header of the XML file would be ‘A’. It would be ‘V’ for the value nodes. For all the other nodes, it would hold a blank.

3. Values held by the field CNAME in XML table

Cname contains the names of the nodes. Hence, considering our example here, if the Hier is ‘2’ the Type would be blank the Cname would be ‘EmployeeRequests’ and the Cvalue would be blank since EmployeeRequests holds no value.

4. Values held by the field CVALUE in XML table

Cvalue contains the values held by the elements of the various nodes of an XML file.  Taking an example of our case here-

For HIER ‘4’ the Type would be ‘V’, Cname can be ‘HomePernr’, ‘UNAME’ or ‘USERID’ and the Cvalue would be the values held by these records.

Here is a screen-shot of the values contained in the table g_t_xml_info, considering our example.

Now that we are acquainted with the pattern of values contained in g_t_xml_info, we can move these values to the required structure.

We require only the values contained in the value nodes and we have to transfer these to their respective fields i.e. HomePernr, UNAME and USERID. Since, we need the values only for the value nodes we can directly move the values in the final table for a value of HIER equal to ‘3’.

Note: The value of HIER for which we need the data, can be manipulated in accordance with the position of the value nodes in the hierarchy of XML.

We use the following code snippet to move the data:


* Moving the data from the g_t_xml_info table to the required table

IF NOT g_t_xml_info IS INITIAL.

  LOOP AT g_t_xml_info INTO    g_s_xml_info WHERE hier EQ 3.

      tabix = sy-tabix + 1.

      READ TABLE g_t_xml_info INTO g_s_xml_info INDEX tabix.

      g_s_employeerequest-homepernr = g_s_xml_info-cvalue.

      tabix = tabix + 1.

      READ TABLE g_t_xml_info INTO g_s_xml_info INDEX tabix.

      g_s_employeerequest-Uname = g_s_xml_info-cvalue.

      tabix = tabix + 1.

      READ TABLE g_t_xml_info INTO g_s_xml_info INDEX tabix.

      g_s_employeerequest-Userid = g_s_xml_info-cvalue.

      APPEND g_s_employeerequest TO g_t_employeerequest.



Here, the value ‘3’ for the field ‘Hier’ marks the beginning of a new record. Hence, we move the values to the g_t_employeerequest table whenever the value for ‘Hier’ is ‘3’.

This way we get the required values in the internal table g_t_employeerequest.

Thus, we get the values in our desired structure. We populated the final table in accordance with the location

This marks the end of our journey from the XML residing in the application server to the internal table in SAP ABAP.

Advantages/Disadvantages of using the method SMUM_XML_PARSE over XSLT (call transformation id)

  • As apparent from the description of the entire process, the Function Module SMUM_XML_PARSE is beyond question an undemanding approach towards the XML-ABAP conversion.  SMUM_XML_PARSE is an uncomplicated, unreleased, effortless and undocumented version of the powerful, released and documented iXML.

  • This function module can be used for complex XML structures by deciding upon suitable ways to segregate the data from the XML table in to the required internal table. This is entirely in our hands since; we are the ones to decide upon the structure of the final internal table required and to decide upon the HIER values of the XML table for which we need the data in the required table..

  • To turn the XML into an internal table or any other series of ABAP objects we can make use of ABAP’s CALL TRANSFORMATION keyword. Though, this unlike SMUM_XML_PARSE is very well documented in the help documentations and makes use of a XSLT template to transform XML into ABAP objects and vice versa. However, while making use of CALL TRANSFORMATION we need to create the XML schema or the XSLT program in the transaction SE80 which has to be in a definite format, failing which there can be a number of anomalies. Besides, this is a time consuming process. This adds to the complexity quotient of the CALL TRANSFORMATION method.

  •   Quite often, a peculiar problem that occurs while reading the xml file using the CALL TRANSFORMATION method is- ‘format not compatible with the internal table’.  Hence, in order to get rid of this particular issue we further need to apply another transformation to convert the data from the internal table into the xml file. Then only, do we get the format of XML which might be utilized for conversion and is compatible with the given internal table.

Related Content

SAP Community Network Wiki – Code Gallery – XML TO ABAP INTERNAL TABL CONVERSION

Parsing an XML and Inserting to ABAP



New NetWeaver Information at

Very Helpfull

User Rating: Be the first one !