Unicode Help

[Former Member]

Hi All,

We are presently working on mgration from 4.6 to ECC 6 in which the programs are to be converted into Unicode compilent programs.

I would like to know if anyone of you have a generic list of things we need to change in the program.

Eg : WS_UPLOAD/DOWNLOAD is to be replaced by GUI_UPLOAD/DOWNLOAD.

There could be N Number of such changes.

If you can share this, it would be great help.

Secondly,

We have around 2000 programs that are to be converted to Unicode programs.

Is there any way using which we can convert the programs in one stroke.

Eg: I have 450 programs using WS_DOWNLOAD. Do I have to edit all the 450 programs individually or is there anything which I can use to make the changes in all the 450 programs in one stroke for a perticular change.

Thanks for reading my post.

reagrds,

sachin

Message was edited by:

Sachin123

[Former Member]

Hi Sachin,

The Unicode errors are displayed when you give the program name as input in TCODE UCCHECK

It will display you the Unicode errors involved

ws_upload, ws_download, open dataset, type mismatch are some of the very common errors

After changing the program, you need to ensure the functionality. If the program doesnt work after you make corrections, the whole purpose of having the program done is lost. So kindly prioritize and schedule your time

hope this helps you

[rainer_hbenthal]

Hi, TA UCCHECK is not available in 4.6! So he cant use this. Transforming thousands of programs into unicode without having a unicode environment is extremly dangerous. I suggest to do the update in two steps or if this is not possible, make a client copy, migrate to 4.7 or higher.

And do not forget: TA UCCHECK only checks syntax, not the semantic! Program stll can fail, interfaces might not wokred anymore together. Migrating to unicode without having a unicode-compliant scenario is hazardous.

[Former Member]

Hi Rainer,

Thanks for the reply.

What I think is copying the programs from 4.6 system to ECC system as it is and as they are.

And then making them unicode compatible.

In ECC, I have access to UCCHECK.

But I would like to know if we can make a certain change in all the programs in one go.

Like I said, I have to replace WS_DOWNLOAD with GUI_DOWNLOAD.

regards,

[rainer_hbenthal]

We changed all Download/Upload functionality, even gui_upload and dui_download. We created a class which handles all the download/upload stuff. This class finally calls SAP class cl_gui_frontend_services=>gui_download which is most recommended- So at least we hat all programms unicode compliant regarind up/download, and every report has the same behaviour und the same look&feel to our customers. If now something will change we only need to adopt our wrapping class.

But up/dlownload ist the most boaring thing in making abaps unicode compliant.

More risky are tricky type conversions and copies in or out of structures.

Do not forget: changing all the programs is 25% of the work, testing is 75%!

[uwe_schieferstein]

Hello Sachin

As Rainer already said unicode-enabled programs can still result in failures on a Unicode system.

The ALV list in transaction UCCHECK is quite useful for doing the unicode-enabling of the program objects.

Take special care when your programs have to deal with files (reading and/or writing). On Unicode system we are obliged to add the following options to the DATASET statements:

OPEN DATASET '<my file on ApplServer>' FOR OUTPUT [INPUT]
                                   ENCODING [DEFAULT ... UTF8]
                                   IN TEXT MODE.

If you are using the statement

OPEN DATASET '<my file on ApplServer>' FOR OUTPUT
                                   ENCODING DEFAULT
                                   IN TEXT MODE.

the system will write the file in <b>UTF-8</b> encoding on a Unicode system and

in the default encoding (e.g. <b>ISO-8859-1</b>, <> UTF-8) on a non-Unicode system.

Thus, any file-handling program must be tested particularly careful on the Unicode system. In addition, you have to check whether receivers of files are able to cope with UTF-8 or not.

Regards

Uwe