Use of new ABAP 7.4 syntax

ABAP 7.4 has many nice syntax improvements. These enhancements increase the efficiency of the code written, making it easier for developers to use and to read the code.

Key features:

  • Inline Declarations
  • Table Expressions
  • Constructor Operator NEW
  • Value Operator VALUE
  • FOR operator
  • Reduction operator REDUCE
  • Conditional operators COND and SWITCH
  • CORRESPONDING operator
  • Strings
  • Filter

Inline Declaration(DATA)

Inline declarations are a new concept of declaring variables and field symbols. This enables programmer to declare the variable at the time of they used. With this new format the compiler knows that the data store the value in the form of string variable. No need to explicitly declare the string variable. So, the line of code got reduced.

Before 7.4

DATA zlv_text TYPE string. 
zlv_text = 'ABC’.

With 7.4

DATA(zlv_text) = 'ABC'.

Inline Declaration(Loop at into work area)

Every time we need to loop over an internal table we need work area . Many times this work area is not in use when loop processing is completed. Hence inline declaration make it easy to declare it and use it when it is required.

Before 7.40

Types: Begin of zlts_itab, 
Matnr type matnr,
            Sernr type gernr,
            end of zlts_itab,

DATA : zlt_itab type table of zlts_itab,  
             zls_wa type zlts_itab.
LOOP AT zlt_itab INTO zls_wa.
......
ENDLOOP. 

With 7.40

LOOP At zlt_ itab INTO DATA(zls_wa). 
…………
ENDLOOP.

Inline Declaration(Field Symbol)

Like work area we can use inline declaration for Field symbol. Here we can use angular bracket and FIELD_SYMBOL keyword as before. Read a line of internal table and assign it to a field symbol.

Before 7.40

FIELD_SYMBOLS <zls_test> type zlts_itab.
LOOP AT zlt_itab assigning <zls_test>
.....
ENDLOOP.

Read table zlt_ itab assigning <zls_test>.

With 7.40

LOOP At zlt_ itab assigning FIELD_SYMBOL (<zls_test>).
…………

ENDLOOP.
Read table zlt_itab assigning FIELD-SYMBOL(<zls_test>).

Inline Declaration(Select into table)

We need to put @symbol in front of the variables using in the select statement while using new feature. That helps compiler to understand that we are not referring the field of the database but the variable in the program.

No need to declare the internal table separately.

Please note when we use new feature we need to put comma between field while retrieving data from DB table, and also put into statement at the end.

Before 7.40

DATA : zlv_fld1 type matnr  value ‘number’,
zlt_itab TYPE TABLE OF mara.

SELECT * FROM mara
INTO TABLE zlt_itab
WHERE matnr = zlv_fld1.

With 7.40

SELECT * FROM mara 
INTO TABLE @DATA(zlt_itab)
WHERE matnr = @zlv_fld1.

Inline Declaration(Call Method)

Inline ABAP DATA declarations are a new concept introduced in release 7.4 which allows you to declare your internal table variables or work areas within the code that uses them.

No need to declare the internal table separately.

Before 7.40

DATA zlv_a1 TYPE ...
DATA zlv_a2 TYPE ...

zlo_oref->meth( IMPORTING p1 = zlv_a1
                             IMPORTING p2 = zlv_a2 ).

With 7.40

zlo_oref->meth( IMPORTING p1 =DATA(zlv_a1)
IMPORTING p2 = DATA(zlv_a2) ).

Table Expression(Read Table Index)

Previously we saw we can read the internal table to work area without declaring work area before the read statement.

In new feature we can see we might never have to use Read statement at all.

If we want to read 3rd line from Internal table we need to pass 3 into the variable idx = 3.

Before 7.40

READ TABLE zlt_itab INDEX zlv_indx  INTO zls_wa.

With 7.40

zls_wa =zlt_itab[ zlv_idx ].

Table Expression(Read Table using key and with key)

In this example we are reading internal table with key and if the passing parameter values and it satisfied then work area will be populated. If not then sy-subrc will not equal to 0.

But in new statement we can simply write work area equal to…it simple replace the read statement with direct assignment of table details to work area.

Before 7.40

1. READ TABLE zlt_itab INDEX zlv_idx
USING KEY key INTO zls_wa.

2. READ TABLE zlt_itab
               WITH KEY col1 =
                        col2 = …  INTO zls_wa.

3. READ TABLE zlt_itab WITH TABLE KEY key
COMPONENTS col1 = …
                              col2 = …INTO zls_wa.

With 7.40

zls_wa = zlt_itab[ KEY key INDEX zlv_idx ].

zls_wa = zlt_itab[ col1 = … col2 = … ].

zls_wa = zlt_itab[ KEY key col1 = …col2 = … ].

Table Expression (Does record exist?)

Whenever we have data in internal table it is very common practice to check the data which we require from table that is already exist or not?

We use here TRANSPORTING NO FIELDS statement  as we are not interested about the data but it is exist or not in the table.

In new statement we use line_exist statement to do this. We use the statement then internal table and the parameters if it is successful then compiler go inside.

Before 7.40

READ TABLE zlt_itab ... TRANSPORTING NO FIELDS.
IF sy-subrc = 0.   
...
ENDIF.

With 7.40

IF line_exists( zlt_itab[ ... ] ).      
...
ENDIF.

Table Expression(Get table index)

This example is used to read the index table index. We use line_index in the new statement.

Before 7.40

DATA zv_idx type sy-tabix.
READ TABLE ...
TRANSPORTING NO FIELDS.

zv_idx = sy-tabix.

With 7.40

DATA(zv_idx) = line_index(zlt _itab[ ... ] ).

String processing

This is frequently used in ABAP code. The data from the DB needs to be formatted before display and vice versa.

Concatenation of string: the changes are as followed:

  • String template: Create character string out of literals texts, expressions and control character, The goal is to display data in more human readable format.
  • We use pipe symbol and && operator to concatenate string.
  • To replace the write to statement.
  • A string template is defined by using the |(pipe) symbol at the beginning and end of a template.
  • If we add some value with string template like {expression} and control character.
  • Embedded expression are defined within the string template with curly brackets. Note a space between bracket and expression is obligatory.
  • An expression can be variable, functional method, predefined function or calculation expression.

Before 7.4

Data zlv_qmnum type string.
Data zlv_string type string.

zlv_qmnum = zls_qmel-qmnum.

Concatenate ‘Your notification no:’
                           zlv_qmnum
                           into zlv_string
                           separated by space.

Write :/ zlv_string.

With 7.4

Data zlv_qmnum type string.
Data zlv_string type string.

zlv_qmnum = zls_qmel-qmnum.
zlv_string = |’Your Notification No:’|&& zlv_qmnum.
zlv_string = ’Your Notification No:’.

Same as (as soon as we add pipe symbol it became string template).

zlv_string = |’Your Notification No:’.|.
zlv_string = |{ a numeric variable }|.
zlv_string = |’The return code is: ({ sy-subrc }|.
zlv_string = |’The length of text : {( text-001 )} is { strlen(text-001) }.

Chaining Operator: The chaining operator && can be used to create one character string out of multiple other strings  and string template.

In this example a number text , a space, an existing character string and a new string template are concatenated into new character string. The code convert the amount field in a display format as per user settings.

Character_string = ‘Text literal(002)’&& ‘ ’&& character_string && |{ amount_field NUMBER = USER }|.

Conversion operator

Conversion operator is used to convert value into specified type. It is suitable to avoid the declaration of helper variable.

For example :let us assume that a method expects string but we have data in character. We would need to move the  value to the string variable and then pass the helper variable to the method call. With CONV variable helper variable not required. The value can be converted directly during method call.

Before 7.40

Data zlv_cust_name type c length 20.
Data:zlv_helper type string.

zlv_helper = zlv_cust_name.

Cl_func->process_func(ziv_input = zlv_helper).

With 7.40

Data zlv_cust_name type c length 20.
Cl_func->process_func(ziv_input =CONV string( zlv_cust_name ).

Cl_func->process_func(ziv_input =CONV #(zlv_cust_name ).

Casting operator

The casting operator CAST is a constructor operator that performs a down cast or up cast for the object and create a reference variable as a result.

Syntax of casting operator is :

Cast #/type( [let_exp] dobj)

Type can be class or interface. The # character is a symbol for the operand type.

With 7.40

CLASS zcl1 DEFINITION.
ENDCLASS.
CLASS zcl2 DEFINITION INHERITING FROM    zcl1.
ENDCLASS.

DATA: zlo_oref1 TYPE REF TO zcl1,
      zlo_oref2 TYPE REF TO zcl2.

IF zlo_oref1 IS INSTANCE OF zcl2.
  zlo_oref2 ?= zlo_oref1.
  zlo_oref2 =  CAST #( zlo_oref1 ).
ENDIF.

Value operator

The VALUE operator in SAP ABAP 7.4 is used to create and initialize data objects of a specified type. It is particularly useful for constructing values for structured types, table types, and controlling the type of results in table expressions.

Initial Value for All Types:

The VALUE operator can be used to create initial values for any non-generic data types

DATA zlv_initial TYPE i
zlv_initial = VALUE i( )

Structures

For structures, the VALUE operator allows you to specify values for individual fields

TYPES: BEGIN OF zlts_struct,
field1 TYPE string,
field2 TYPE string,
END OF zlts_struct

DATA zls_struct TYPE zlts_struct

zls_struct = VALUE zlts_struct( field1 = 'Value1' field2 = 'Value2' )

Internal Tables

The VALUE operator can also be used to construct internal tables

DATA zlt_table TYPE TABLE OF string
zlt_table = VALUE #( ( 'Row1' ) ( 'Row2' ) ( 'Row3' ) )

For operator

For operator is used to loop at an internal table. For each loop the row is read and assigned to a work area or field symbol. This is similar to the same FOR loop in C language.

Transfer data from one internal table to another internal table:

New Syntax:

Data: zlt_r_equipment TYPE RANGE OF equnr,
zlt_r_equipment = VALUE #(  FOR zls_equipment IN zlt_r_equipment
                    ( sign   = zls_equipment-sign
                      option = zls_equipment-option
                      low    = |{ zls_equipment-low   ALPHA = IN }|
                      high   = |{ zls_equipment-high ALPHA = IN }| ) ).
).

Reduction operator

Reduce operator creates a results of specified data type after going through iteration. In Classical ABAP if we had to evaluate the data in an internal table then we had to loop through the internal table, evaluate the condition, and then take appropriate action. This could be done much simpler way using REDUCE.

Data(zlv_lines) = REDUCE i ( init x = 0 for wa_sales in lt_sales where ( salesoffice = ‘IN’) Next x= x+1 ) .

Conditional operator

It is an accepted practice in ABAP to use CASE statement instead of IF statement. CASE statement made the code readable but had an issue that it was not able to evaluate multiple condition.so we use if else statement. In 7.40 we use COND operator for the same.

Data(zlv_text) = COND # ( when zlv_qmtxt = ‘X1’and zlv_artpr = ‘XX’ then ‘Customer’
when zlv_qmtxt = ‘01’and zlv_artpr = ‘XX’ then ‘test1’
when zlv_qmtxt = ‘02’and zlv_artpr = ‘XX’ then ‘test2’).

Write:/zlv_text

Switch operator

Switch operator is a conditional operator like CASE but more powerful and less code. It is used to switch from one value to another based on condition.

Old syntax:

data: zlv_indicator like scal-indicator, zlv_day(10) type c.
    case zlv_indicator.
    when 1.
      zlv_day = 'Monday'.
    when 2.
      zlv_day = 'Tuesday'.
    when 3.
      zlv_day = 'Wednesday'.
    when 4.
      zlv_day = 'Thursday'.
    when 5.
      zlv_day = 'Friday'.
    when 6.
      zlv_day = 'Saturday'.
    when 7.
      zlv_day = 'Sunday'.
else.
     Raise exception type zcx_day_problem.
endcase.

New Syntax:

DATA(zlv_day) = SWITCH char10( zlv_indicator
when 1 THEN 'Monday'
   when 2 THEN 'Tuesday'
   when 3 THEN 'Wednesday'
   when 4 THEN 'Thursday'
   when 5 THEN 'Friday'
   when 6 THEN 'Saturday'
   when 7 THEN 'Sunday'
   ELSE THROW zcx_day_problem( ) ).

Corresponding operator

Corresponding operator allows the copy of data from one internal table to another internal table just like move corresponding but provides more options on which columns are copied.

TYPES : BEGIN OF zlts_demo1,
col1 TYPE c,
          col2 TYPE c,
        END OF zlts_demo1,

      BEGIN OF zlts_demo2,
          col1 TYPE c,
          col3 TYPE c,
          col4 TYPE c,
        END OF zlts_demo2.

Data: zlt_itab1 TYPE STANDARD TABLE OF zlts_demo1,
      zlt_itab2 TYPE STANDARD TABLE OF zlts_demo2.

zlt_itab1 = VALUE #( ( col1 = 'A' col2 = 'B' )
                 ( col1 = 'P' col2 = 'Q' )
                  ( col1 = 'N' col2 = 'P' ) ).

zlt_itab2 = CORRESPONDING #(zlt_itab1 ).

cl_demo_output=>write_data( zlt_itab1 ).
cl_demo_output=>write_data( zlt_itab2 ).

cl_demo_output=>display( ).

Filter

A new FILTER operator is available which can used on ABAP internal tables to filter the data (or) to retrieve subset of data into a new internal table. As of ABAP 7.4 this keyword is available to use in the system.

Old Syntax:

Data: zlt_qmel_all type standard table of qmel,
zlt_qmel_zz type standard table of qmel.

Select * from qmel into table @zlt_qmel_all.

If sy-subrc = 0.
Loop at zlt_qmel_all into data data(zls_qmel) where qmart = ‘00’.
Append zls_qmel to zlt_qmel_zz.
Clear : zls_qmel.
Endloop.
Endif.

New Syntax:

zlt_qmel_zz = FILTER #(zlt_qmel_all using key qmart where qmart = ‘00’)

SNC encryption tips and tricks

This blog will give tips and tricks around the topic SNC encryption.

SNC encryption exists for both SAP GUI and RFC connections.

Formal documentation about SNC can be found on help.sap.com.

SAP GUI client encryption

Central OSS notes for SAP GUI client encryption:

How to check if all GUI’s are using SNC encryption? The audit log can register unencrypted us of the GUI: 2122578 – New: Security Audit Log event for unencrypted GUI / RFC connections. Activate this in the main client(s) as well as in client 000 (3577840 – Information about Security Audit Log event BUJ are required).

Use of insecure SAP GUI

Use of insecure SAP GUI can be detected by using the SAP audit log events. Event BUJ is recording the insecure use: 2122578 – New: Security Audit Log event for unencrypted GUI / RFC connections and 3577840 – Information about Security Audit Log event BUJ are required.

See OSS note 3552348 – Record failed SAP GUI SNC logon attempts in Security Audit Log for attempts.

SAP GUI SNC log on enforcing

As explained in OSS note 3249205 – Difference between snc/only_encrypted_gui and snc/accept_insecure_gui – SAP for Me parameter snc/only_encrypted_gui can be set to 1 to reject any non-SNC GUI connection. Parameter snc/accept_insecure_gui determines if user password logon is still allowed (using SNC), or only password less SSO.

SAP RFC encryption

Generic SAP to SAP RFC encryption is explained in OSS note 2653733 – Enabling SNC on RFCs between AS ABAP and 3373138 – SNC for SM59 destinations that use load balancing.

Specific use case: SNC for STMS

Note 3025554 – SNC for STMS explains the SNC setup for RFC needed in STMS. If not setup properly, you might get the error as described in this OSS note 3477342 – RFC communication error with system/destination : 00024 error during logon.

Specific use case: SNC for JAVA and MII

Note 3394750 – SNC configuration issue between SAP MII Java and ERP explains the SNC setup for RFC needed in JAVA MII. Which refers also to the generic JAVA to ABAP SNC setup note 2573413 – How to configure SNC from 7.1x onwards AS Java to AS ABAP.

Specific use case: CPI-DS

Note 3280758 – Enabling SNC between CPI-DS and ABAP backend fails with “Test failed for the default configuration ‘default'” gives hints on SNC for CPI-DS.

Specific use case: SNC for SAP Router

For SNC for SAP router read this OSS note: 525751 – Installation of the SNC SAPRouter as NT Service.

Good blog on SNC setup for SAP router: link, and standard SAP help content on SCN for SAP router.

Specific note: 3464887 – SAPRouter SNC error -> SNCERR_BAD_NT_PREFIX.

Specific use case: SNC for SAP support

SAP support will need to access your systems via the remote support connection to give support to solve bugs, execute remote analysis, etc. If you want to force this connection to SNC, follow the instructions from OSS note 2562127 – R/3 Support with SNC/SSO.

SNC issue solving notes

List of notes to help solve issues:

Idoc change pointer setup (ALE)

Idoc change pointers can be used to set up master data distribution. Most used objects are materials, customers, vendors, GL accounts. This setup is also known as the ALE (Application Link Enabling) setup.

General activation of change pointers

Start transaction BD61 to activate change pointers in general (this is once off general activation):

Per message type the change pointer activation is done in transaction BD51:

In transaction BD52 you can see which field trigger a change pointer for that specific message type:

If you want to know (or steer) the processing function module behind, start transaction BD60 and check the details:

Distribution model and Idoc partner profile setup

In transaction BD54 you define logical systems. In our example we will use the definition SOURCE and TARGET:

The SOURCE system definition is normally connected to the main client in the SCC4 transaction:

Now we can model the data flow in BD64 distribution model.

Create the Model View first:

Then add the message type with sender, receiver and message type:

So the end result looks like this:

In WE20 in the source system, now we set up the partner profile:

The receiver port must be defined in WE21 (ports in idoc processing):

The RFC destination is maintained in SM59 as usual and contains the technical data of the target system.

In the target system the setup of the ALE model needs to be done as well, and the partner profile needs to be on the inbound side:

Testing the setup

To test the setup create a material or change one. This should trigger a change pointer.

Run program RBDMIDOC or start transaction BD21 to evaluate the change pointers:

If you run first time, best to clear all the old items with program RBDCPCLR2.If the activation was done years ago, you otherwise end up with a lot of unwanted Idocs.

When running the program for each material master change (not only yours) an Idoc is created. You can check in WE02, WE05 or WLF_IDOC if the Idoc is created correctly.

OSS notes:

And look for application specific issues in ALE: (example note):

Generic clean up

With this setup there are two generic clean ups needed:

  • Clean up old change pointers (program RBDCPCLR2)
  • Clean up old Idocs (program RSETESTD)

See blog SAP database growth control: technical cleanup – Saptechnicalguru.com for reference.

ABAP Clean Core development

ABAP Clean Core is a development concept. It is not to be confused with ABAP clean code.

The ABAP Clean Core is fully explained in this very extensive SAP document: Extend SAP S/4HANA in the cloud and on premise with ABAP based extensions.

SAP has a positioning of development patterns and tools from the past. This is written in OSS note 3578329 – Frameworks, Technologies and Development Patterns in Context of Clean Core Extensibility.

This blog will focus on the initial phase to get insights into your existing code and to brief you on the main changes.

Prepare the ATC run for Clean Core

First step is to prepare the ATC runs by applying OSS note 3565942 – ATC Checks “Usage of APIs” and “Allowed Enhancement Technologies” (don’t forget this action After implementation, press ‘Import Parameters’ for the ATC check object SYCM_USAGE_OF_APIS (in ADT for Eclipse).).

Then use the Notes Analyzer to apply correction notes from OSS note 3627152 – SAP Note Analyzer Files for ATC Checks Related to Clean Core. And check OSS note 3605376 – ATC Findings for usage of not release APIs for key user items.

ABAP CVA security checks is part of this variant. If you run on SAP cloud this is part of license, on premise it is not (so you need separate license). If you don’t have the license use ABAP Eclipse ADT to remove the check from ATC variant ABAP_CLOUD_DEVELOPMENT_DEFAULT.

To make sure all new items are loaded, start transaction SCI and choose the option Utilities / Import Check Variants.

ATC runs for Clean Core

Run the ATC for variant ABAP_CLOUD_DEVELOPMENT_DEFAULT.

Remark: please read the document Extend SAP S/4HANA in the cloud and on premise with ABAP based extensions if there a newer version defined!

Now run this one (or older ABAP_CLOUD_READINESS) on this simple test program:

REPORT ztest.

DATA: zlt_mara TYPE TABLE OF mara.
DATA: zls_mara TYPE mara.

SELECT * FROM mara INTO zls_mara.
ENDSELECT.

Result:

Already in this small piece of code 2 showstoppers:

  • You cannot use SE38 programs any more with REPORT statement
  • Direct table reads (in this case MARA) are forbidden in Clean Core

When you run on your existing code base, you will find many issues. On average a single old fashioned written ABAP code easily generates 100 clean core findings or more.

Forbidden to use in Clean Core

What else is not allowed?

Short list:

  • ALV grid output
  • Enjoy screens
  • SAP script
  • Smartforms
  • Webdynpro
  • Non released function modules
  • Batch input
  • Many more

New technology

So what do I need to use?

  • Data selection: CDS views
  • User interaction: FIORI (or FIORI elements) including FIORI key user extensibility
  • Data and processing logic: RAP (restful application programming) framework
  • Use released API’s (see the Cloudification Repository Viewer, link and explanation)
  • SAP extension points

BTP side by side developments

BTP side by side developments are an option. These developments are meant for loosely coupled scenarios. Also check if the BTP platform availability restrictions meet your SLA requirements.

Summary

Starting with a Green Field new SAP implementation, there is a good chance you will succeed in applying the Clean Core principles.

With an existing implementation there are a couple of bottlenecks:

  • Most likely skill set of you current ABAP developers is not up to par to meet the Clean Core development skill set
  • Small changes to existing code you will keep doing old style, since the effort to convert is out of balance with the effort of doing old style
  • The new style of coding is more structured and takes more initial effort (which will be regained at the end in the maintenance phase), but this can be a hard sell. Old fashioned ABAP with selection screen, direct SQL and ALV output is very effective and efficient to write in terms of man days work.
Tip: try the new development style for a completely new development. The chance of success will rise.