11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Products
Products Industries
Industries Support
Support Training
Training Community
Community Developer
Developer Partner
Partner About
About 

Home / Community / Blogs + Actions

ABAP Test Cockpit Checks

October 19, 2017 | 831 Views |

Ana Maria Butiseaca

more by this author

ABAP Testing and Analysis

ABAP Development | abap test cockpit | abap_checks | abap;sci | Code Inspector; ATC

share
0 share
3 tweet share
4 like
2

Follow

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 1/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

The purpose of this blog is to describe ABAP Test Cockpit (ATC) checks. ATC is based on Code
Inspector (SCI), it reuses Code Inspector checks and significantly improves the code quality by offering
the uniform quality criteria for the whole development landscape and establishing new quality assurance
processes (quality gates, exemptions, mass regression tests) to minimize errors in productive systems.

A general overview of ATC can be found in the blog ABAP Test Cockpit – an Introduction to SAP’s new
ABAP Quality Assurance Tool. More details about remote code analysis using ATC you can read in the
blog series about Remote Code Analysis in ATC .

I focus in this blog on the main ATC checks categories:

Performance Checks
Security Checks (CVA)
Robust Programming Checks
Programming Convention Checks
User Interface Checks

Please note: you can find the SAP S/4HANA related checks in the blog SAP S/4HANA System
Conversion – Custom code adaptation process

Robust Programming Checks

This category contains:

Append entries to SORTED TABLE in specific position can be done using APPEND or INSERT with
INDEX to insert entries into internal tables of type SORTED. If, however, the sort order is not kept, an
exception which can´t be handled is raised at runtime.

This check finds the statements specified.

Search DB Operations in Pool/Cluster Tables Check finds all SQL accesses to physical pool or
cluster tables.

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 2/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Direct access to the physical (not logical) pool or cluster is suspicious. After depooling/declustering
these accesses are hard errors, because after depooling/declustering the logical table is converted to a
transparent table, but the physical pool or cluster is still there but empty.

All accesses with SQL to these tables are senseless after depooling/declustering and shall be corrected.

Empty Unused Procedures Check identifies the procedures in ABAP programs that are empty and
have never been called in the current system.

Here, procedures also include ABAP events such as START-OF-SELECTION.

The test returns the different types of procedures in the results list as follows:

METH: Methods
FORM: Forms
FUNC: Function modules
MODI: Modules (INPUT)
MODO: Modules (OUTPUT)
SSEL: START-OF-SELECTION (can also be a REPORT statement)
ESEL: END-OF-SELECTION
TOPA: TOP-OF-PAGE
TOPS: TOP-OF-PAGE DURING LINE-SELECTION.
ENPA: END-OF-PAGE
LDPR: LOAD-OF-PROGRAM
INIT: INITIALIZATION
ATUC: AT USER-COMMAND.
LISE: AT LINE-SELECTION.
ATPF: AT PFnn
GET: GET dbtab.
GETL: GET dbtab LATE.

Messages from the test can be hidden using the pseudo comment.

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 3/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Analyze ABAP source code with the RND parser identifies such source code, either so that you can
improve it, or in order to analyze unexpected behavior of such tools.

Many tools that process ABAP source code (such as pretty printer and code completion) do not use the
ABAP compiler but a component called RND parser. There are situations where the ABAP compiler
accepts source code but the RND parser does not, which means that above mentioned tools might
behave unexpectedly.

The test delivers the following result messages:

TOKEN – RND parser does not recognize the token: check whether you can replace obsolete
with modern syntax or whether you can rename variables so that special characters (punctuation
characters except underscore etc.) are avoided

TIME_OUT – RND parser could not analyze the code in reasonable time: this is probably a bug,
create a ticket for BC-ABA-LA.

Complex WHERE Condition in the SELECT Statement check searches for complex WHERE
conditions when SELECT is used. It can search for:

OR branches of differing lengths, meaning they have different numbers of comparison fields
OR branches containing different fields.
WHERE conditions that contain a high number of AND joins
ABAP JOINs that join a large number of tables.

Test for support of Field Extension searches for statements where problems occurs during a field
extension. The check parameters are:

Full-Names of the types which should be extended

New Length of the types
Comment for the field extension. The comment is included in the check messages.

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 4/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Test to check handling of type INT8 searches for statements that have to be changed for the
introduction of the new type INT8 (8 Byte Integer). The check parameters are:

Only local calls: with this parameter, you can specify if external procedure calls should be
analyzed too.
Non-local analysis depth: with this parameter, you can specify the maximal evaluation depth into
external programs.

Search problematic statements for result of SELECT/OPEN CURSOR without ORDER BY

searches for SELECT and OPEN CURSOR statements where no ORDER BY clause is specified.

Afterwards problematic statements are searched which use the results of these SELECT or OPEN
CURSOR statements. Problematic statements are statements which depend on the order of the entries
of an internal table. The problematic statements are listed further below.

If the result of a SELECT or OPEN CURSOR statement is sorted by performing a SORT-Statement or

by assigning to a sorted table the searching is aborted.

SELECT SINGLE statements are analyzed independently of problematic statements afterwards. They
are problematic if they are ambiguous, i.e. if the WHERE condition does not specify a full key of the
database table or if the FROM clause contains a JOIN. If the result of an ambiguous SELECT SINGLE
statement is not used no message is thrown.

The check parameters are:

Only local calls: with this parameter you can specify if external procedure call should be analyzed
too
Report access to Pool/Cluster table: Controls whether the messages SEL_POOL/CLUST or
OPEN_POOL/CLUST are raised
Non-local analysis depth: With this parameter you can specify the maximal evaluation depth into
external programs.

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 5/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Tenuous use of FOR ALL ENTRIES check tries by some heuristics to judge from static code analysis if
the program flow ensures content or not. These heuristics are not capable to analyze the full complexity
of data flows in ABAP however. The heuristics try to avoid false positives. This also implies there are
false negative findings; there constellations where code flaws are not reported.

Programming Convention Checks

This category contains:

ABAP naming convention checks which examine whether the naming conventions specified by the
test parameters are met for the various objects (such as types, classes, or fields) defined in an ABAP
program.

Extended check on naming conventions in ABAP programs examines whether the naming
conventions specified by the test parameters are met for the various objects (such as types, classes, or
fields) defined in an ABAP program.

Pretty Print State check in ABAP Programs establishes whether the ABAP code is formatted with
respect to the configured Pretty Print variants.

ABAP unit test conventions check helps you to detect flaws in the code of Unit Tests. Unit Tests are
special code fragments that execute the domain code in order to detect functional errors within the
domain code. The test fragments – typically local classes – underlie some constraints. For example,
they must not require interactive user input in order to ensure the capability of fully automated test
execution.

User Interface Checks

This category contains:

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 6/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Web Dynpro Component – Active Components check uses technology, such as Java, Adobe Flash,
Microsoft Silverlight or others, to provide enriched functionality to the user.

Due to security constraints and the ongoing improvement of HTML, CSS and JavaScript, some of these
technologies are no longer necessary or can be replaced by native functions. Additionally, browser
vendors have started to block such plugin technologies.

This check identifies UI elements which are active components to provide an opportunity to analyze the
usage of these elements. The rules of the check are:

Usage of Adobe Flash based component which requires a Adobe Flash plugin to be available in
the Web browser of the user.
Usage of Adobe Forms based component which requires the Adobe Reader plugin to be available
in the Web browser of the user.
Usage of Java applet based component which requires the installation of the JAVA runtime on the
user machine. Additionally, a plugin is required in the Web browser of the user
Usage of Microsoft ActiveX based component which requires the usage of Microsoft Internet
Explorer as Web browser.
Usage of Microsoft Office integration component which requires the installation of Microsoft Office
and a Microsoft-based Web browser, e.g. Internet Explorer.
Usage of Microsoft Silverlight based component which requires the Microsoft Silverlight plugin to
be available in the Web browser of the user.

Web Dynpro Component – Standard Check delegates the analysis of Web Dynpro Components to
the technical check also used by the IDE. The IDs of findings are mapped generically by the priority
used in the IDE. If you compare the information about the findings in the IDE and in the Code Inspector
adapter, the IDE contains more information.

Web Dynpro Conventions for Source Text allows you to enhance the declaration of the user interface
using code in certain places.

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 7/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

This Code Inspector test enables you to identify these problematic statements and avoid problems
before they arise. These checks cover the following areas:

Database access
File access
Lists & Dynpro technology
Changes to the program flow
Low-level commands
Including source code
Direct method call
Querying the parameter interface

Dynpro Checks are checks on dynpros and resource consumption.

This check contains all usability and accessibility checks for dynpros and the scope of the check is
identical to that in the menu function Check -> Layout in Screen Painter (used for dynpro editing in the
integrated SAP GUI).

‘SAP Script’ Scanner analyzes SAP Script content (SE61-documentation).

The following issues can be identified by the scanner:

Inconsistencies between transport key repository (TADIR) and document repository

Inactive document content
Inconsistent document content, like malformed tags
Invalid meta data, like missing master language.

Alert Moderator

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 8/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

4 Comments

Vener Martinez

October 24, 2017 at 7:24 am

Other useful feature of ATC is Metrics and Statistics -> Procedural Metrics, in here you can control the nesting depth and
number of executable statement: subroutines, function modules and methods, to prevent bloated codes inside routines, benefit
will be easy to read/maintain codes. Just my opinion.

Regards,

Vener S. Martinez

Julian Phillips

October 26, 2017 at 8:37 am

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 9/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Always good to get more information on ATC. The problem that we have is that we need to run the inspector for all objects in a
release, this way we can check all objects in the release once they are imported into our test enviornment – a release can
include hundreds of transports. Currently there is no straightforward way to achieve this, unless anyone knows differently.

Matias Villagarcia

October 26, 2017 at 9:03 pm

You can create object sets, you can get all the main programs (ATC uses main programs) in a TXT file and use
that.

It still requires some work, because you have to get the main program for all the modified objects, but this can be
done with a custom report, and you would have to do this only once.

Unless there is a standard FM or something that can do this.

Julian Phillips

November 8, 2017 at 12:17 pm

Well we mostly use classes and methods, and frequently its only a few methods that are changed in any given class. Then we
have a team of 80+ developers and that means a lot of custom objects to manually handle. It would be so much better if we
could just run ATC across a list of transports!

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 10/11
11/20/2017 ABAP Test Cockpit Checks | SAP Blogs

Add Comment

Share & Follow Privacy Terms of Use Legal Disclosure Copyright Trademark Sitemap Newsletter

https://blogs.sap.com/2017/10/19/abap-test-cockpit-checks/ 11/11