IBM Mainframe Forum

N tilde should be X'A0' while the # symbol is X'23'; I'm not sure why they are getting crossed.

Hi Robert,

Thanks alot. Can you suggest if there is any way to transfer these files without losing spanish characters.

Thanks,
Nagesh.

First, mainframes have "files" in Unix System Services and on tape only. Disk allocations are ALWAYS "data sets" and calling them "files" is just plain WRONG.

Second, FTP is the way to transfer data sets to other platforms. Some sites purchase other solutions (for example, Connect:Direct) but unless your site has purchased such a solution, you will need to use FTP.

Third, you may have to experiment -- possibly dozens of times -- with FTP jobs to get the parameters right for the transfer you want to do. SBD=(IBM-284,?) is most likely what you'll need but what code page to replace the ? is what you need to figure out.

Fourth, one possibility would be to transfer the data set as binary, then write your own translation program on the Linux system to convert EBCDIC to ASCII (preserving the Spanish characters).

You keep saying spanish charaters, But I highly doubt it is spanish.

To test your scenario.

Go to the Mainframe source file. Have only two lines of data and the data should produce the "spanish charaters" when sent.
Do a HEX ON in the file and post it here. That would give the input file.

For the output file, if you have NOTEPAD++, then open the transferred file using "Hex-Editor Plugin for Notepad++"
This would give you hexadecimal values that were transferred. Post the hex format of the output file.

We would be able to identify the root cause once you have the input & output data.

Hi Vasanthz,

Thanks alot for quick reply.
Mainframe Hex on giving me below output.

FRANCISCO VILLASEÑOR
CDCDCCECD4ECDDCEC6DD
69153923605933125969

Please find the attachment that I tried to see the special character with conversion table.

If anyone had any syntax to convert EBCIDIC to UTF-8 that would help alot. I am planning to load this on to oralce and these are inserted as it is, I mean the special character is not decoded while loading with sqlloader.



Thanks,
Nagesh.

ok.
How does the file look on the receiving side? with hex data

Hi,

I was able to recreate your issue.

Sent


using standard FTP.


Ñ was converted to x'ac' which is ¬

Sent using the below control cards,


Ñ was converted to x'd1' which is Ñ . http://www.simotime.com/asc2ebc1.htm confirms that the translation is correct.

Can you try this option? Or have you tried this already and facing issues with this?

Thanks alot Vasanthz. I will try the above options and will post the results.

Thanks,
Nagesh.