CO-PA external data transfer (KEFC): unicode error?

[Former Member]

Hello experts,

I am trying to read data from a comma-separated file in transaction KEFC (CO-PA external data transfer). The system then issues a message saying that the file cannot be imported in this format (message DU027). The long text of the message suggests to create the file in another format, but it is not clear what that format should be.

Upon debugging, I discovered that this message is issued before the input file has even been opened. The data is to be read in a custom structure, containing QUAN- and CURR-fields. In form routing check_unicode (include: RKCDFILEINCFOR) a check is done if the system is a unicode system. If so and if non-character data is to be read, the error message is issued.

Does anyone know why this is and even more importantly, how it can be solved?

Thanks!

Alfio Esposito

[Former Member]

Hello experts,

I am trying to read data from a comma-separated file in transaction KEFC (CO-PA external data transfer). The system then issues a message saying that the file cannot be imported in this format (message DU027). The long text of the message suggests to create the file in another format, but it is not clear what that format should be.

Upon debugging, I discovered that this message is issued before the input file has even been opened. The data is to be read in a custom structure, containing QUAN- and CURR-fields. In form routing check_unicode (include: RKCDFILEINCFOR) a check is done if the system is a unicode system. If so and if non-character data is to be read, the error message is issued.

Does anyone know why this is and even more importantly, how it can be solved?

Thanks!

Alfio Esposito

[Former Member]

I also faced the same issue. I tried changing the QUAN and CURR fields to CHAR fields in the sender structure. but when i tried to load the input file using KEFC transaction, it said error in content and in the log i could find it saying untable to consider 1234,56 as a packed decimal number.

Anyone knows how to get over this, please help.

Thanks

Balaji

[Former Member]

Hello Balaji,

Thanks for your suggestion. I have changed the QUAN and CURR fields into CHAR16-fields, and for me it now seems to work. That is, I can import the file using KEFC. The functional team still has to check if all data is transferred correctly.

I am not getting the packed number message that you are getting. I am using a period as decimal separator in the file, not a comma as seems to be the case in your example. Perhaps that solves your problem.

Regards,

Alfio

[Former Member]

The functional guy has looked at the imported data and all is correct. The period as decimal separator works fine. So the issue is now solved.

[former_member249672]

Hello,

We have the same problem as you. We follow what you said, ie changing the sender structure by SE11 (all quantities and units are now char fields) but we always have the same problem.

Did you change anything in the customizing, where you affect each field from the sender structure to the receiver structure ?

Did you do anything else than changing the sender structure ?

Thank you for your answer.

Vanessa Roulier

[Former Member]

Hello Vanessa,

I only changed the QUAN- and CURR-fields of the sender structure. No other changes in customizing were done.

What i did do additionally is implement SAP note 1139909, although that note is not directly related to the unicode error that we originally got.

Maybe this helps you.

Regards,

Alfio Esposito

[former_member249672]

Thank you for your help.

The note didn't change anything.

It doesn't work for us.

Could you please looak in your unicode system and tell me:

- if in the sender structure, I have the field BUKRS on CHAR 4, in the receiver structure (KEFD), I see BUKRS on CHAR 8 (idem for all fields, other example MATERIAL on CHAR 18 in the sender structure is CHAR 36 in the receiver structure).

Could you please look in your receiver structure (by KEFD) and tell me if you have all the fields with length double than before unicode ?

And idem in the sender structure ?

Thank you for your help.

I hope we'll find a solution soon....

Vanessa Roulier

[Former Member]

Hello Experts,

We are facing the same issue explained in this post and after going through it I got clear the only solution is to change the data types (for example from CURR to CHAR).

However this does not solve our problem because modify our custom structure will affect adapting other programs, etc and is not feasible.

Did you find other possible solution? If so, could you please let me know?

I even looked for OSS notes but without success...

Thanks very much!

Edu.

[Former Member]

Please advice if any of the config settings may resolve this issue.

I was looking at the config and found RULES while defining the structures. But i was not able to

found anything which resolves this problem.

If anybody have looks upon the config and found the suitable solution(except changing the structure ONLY) then please share it.

Gaurav

[Former Member]

Hello,

I've got the same problem!

Is there any solution now, without changing data types in the sender structure?

Because every time, if someone will apply a new "value field" (TCODE KEA6 / KEA0) there will be the same problem!

@SAINT-GOBAIN : ...

I think the data type DEC is also a problem for unicode

Thanks in advance!