Unicode ABAP programs

[Former Member]

Hi ,

I want to know the steps to be followed AFTER syntax errors are removed from UCCHECK programs with error status.

Shall I test the programs for runtime errors before enabling Unicode flag?

or

first I should activate Unicode flag in UCHHECK for all programs and then do testing to check runtime errors and correctness of data.

Please clarify.

Thanks,

Yogita

[Former Member]

HI first do the run time errors.

if u r system is unicode system there is no need to

correct the unicode issurs.

[former_member194669]

Hi,

After the Unicode conversion is complete, tests of the applications follow. For instance, a part of the custom ABAP objects can be tested. If possible, interface tests should also be performed at this point. Normally, however this is only fully possible in the QAS, because in the sandbox system the interfaces must first be set up.

a®

[Former Member]

So you mean to say is Unicode flag should be activated after tesing the programs?

Also can you please let me know regarding interface testing regarding Unicode?

Thanks,

Yogita

[uwe_schieferstein]

Hello Yogita

Irrespective of whether you are already on a Unicode system or not you should always set the Unicode flag.

The major problem with possible runtime errors (UCCHECK warnings) is: How do you set up situations under Unicode where runtime errors are provoked? A very tricky, if not almost impossible task.

Therefore, if you have to convert many programs forget about the UCCHECK warnings because you neither have the time nor the means to test them.

However, it is very important to test the <b>file interface </b>(READ DATASET) because the READ DATASET statement has the option ENCODING (DEFAULT / UTF-8 / NON-UNICODE) as mandatory addition under Unicode.

Regards

Uwe

[Former Member]

There is a system parameter BASIS can set that makes the unicode flag mandatory in programs - once this is set, programs without the flag will short dump. This can be done in non-unicode system as a preliminary step before eventual move to unicode.

Setting this means the unicode flag will not be missed on any used programs - testing will soon show the missing flag.

It also means that leaving the flag not set on obsolete programs prevents their accidental execution and highlights if they are really obsolete or not.

following this strategy means that you do not want to set the flag globally - only in programs you know are required.

Once the paramenter is set, the first syntax error in any program will be unicode flag not set, and this needs to be fixed before further errors can be found/fixed.

Extended check (SLIN) and source code inspector both have options to highlight unicode issues in programs. Not all need to be fixed - eg use of TYPE instead of LIKE only needs fixing in ABAP Objects, not report programs. Good practice would be to fix all, but resource / time limits mean you fix the importants ones instead.

Andrew