French Character transfer issue in the file to Partner non-SAP-systems

[Sekhar-San]

Hi 

We recently did an unicode conversion of our SAP ECC6.0, we are on EHP8. Post that, when we generate an extract of master data that has French characters like accents, these characters look good on screen and in the file display in AL11 directory. When we download they look good to. However when the file is transferred systematically to another partner system via FTP or SFPT or MFT, the French characters are distracted on the target server. 

Example: PLàINFIéLD COùRTï in SAP is shown as PLàINFIéLD COùRTï  on the partner systems.  

There was no upgrade nor any change on the partner systems. Only after post Unicode conversion of SAP this started happening. 

In our extraction program, we have this statement included as part of the mandatory step towards unicode conversion. 

"OPEN DATASET OUTFILE FOR OUTPUT IN TEXT MODE message ws_msg ENCODING DEFAULT"

Any leads in how to fix this issue to ensure Partner systems receive the correct French Characters from SAP file. Thank you in advance.

[JimSpath]

While it has been at least a decade since I had hands-on involvement in Unicode system conversions, and my area was in the data transfer optimization, I recall we had to convert systems in phases as the business world could not stop entirely for any appreciable time. I remember that after R/3 systems were converted much testing was needed as legacy non-Unicode SAP code bases have a variety of character encodings primarily based on their age and crustiness. The statement:

OUTPUT IN TEXT MODE message ws_msg ENCODING DEFAULT

I think you need to know the encoding on both sides, and assume one side is definitely Unicode.

See, perhaps:

https://help.sap.com/doc/abapdocu_752_index_htm/7.52/en-US/abapopen_dataset_encoding.htm

"... Writes German umlaut characters to a non-Unicode code page. This code page is then extracted from the database table TCP0C and used to open the file as a a legacy text file."

DATA(dset) = 'test.dat'.
OPEN DATASET dset FOR OUTPUT IN TEXT MODE ENCODING NON-UNICODE.

"Code page maintenance is more guidelines than actual rules."

[nils_buerckel · Accepted solution]

Hi Sekhar-San,

in a Unicode system when using default encoding the files are created based on UTF-8.

Hence I assume your partner is expecting Non-Unicode (Latin-1)?

You can adapt the ENCODING option to NON-UNICODE, but then the file encoding will depend on  the current ABAP linguistic environment (e.g. the logon language).

For details see e.g. SAP note 1322715 > Attachment  Requirements_of_ABAP_Programs_in_Unicode_Systems.pdf

Chapter 5.8

Best regards,

Nils Buerckel