Mainframe, MVS and zOS Discussion

Hi,

We are sending some files to an end user. They are facing trouble with some of the shipments when they sort the files using excel. For some of the data they are facing problems where they are nor easily able to transform through “text-to-columns” feature in excel.

They have got a couple of Nordic characters in the files, are you familiar with such complications from any of your experience where you've faced such a situations and solved?

You may need to investigate the code page used for the mainframe data set. If it is not a standard code page, and if you are using FTP for the file transfer, you can set the conversion using an FTP option.

Thanks. I am not very aware about the code page concept, could you please elaborate it?

For FTP, based on what you have said, I have searched what keyword can be used and I find it should be SITE, can you please confirm. I have looked at it here: http://www-01.ibm.com/support/docview.w ... s8N1016959

But I am not sure how to use it, not able to find a good example.

not able to find a good example

i.e. one that gives the EXACT solution to MY problem.

I am not very aware about the code page concept, could you please elaborate it

I would hate to think how many hits you would get when googling 'code page'.

The forum is not a tutorial/teaching service.

There are various collating sequences defined; each is called a code page. They vary by national characters and currency symbols (and sometimes sequence of characters -- EBCDIC and ASCII both have code pages but they put characters in different sequences). Start by Googling z/os code page -- there are "only" 10,300,000 results returned for me with that search -- and start reading.

Whether or not SITE is the right FTP option depends upon how the FTP is being initiated. If the mainframe is initiating the FTP session, you would use LOCSITE. If the other end is initiating the FTP session, you would use SITE (or more likely QUOTE SITE) to pass the option from the server to the mainframe.

Thanks Robert. I have used the following FTP card:
I need one more help.

1. There is a screen which can be accessed from a lot of geographical locations. The data is entered at CODE037 location and then saved to a DB2 table. Now a $ sign is stored in DB2. But when we access this data in Nordic region, using a CICS screen, $ is displayed "like A". Nordic region is using CODE page 1142 while fetching the data on CICS.

2. Now, using a COBOL program we are creating a file which is FTPed in Nordic region. But the FTP file has the data as $ only, they want to show it "like A".

What should be the approach for it? I am thinking as the data is stored as $, it'll remain $ as we should take care about the character conversion while saving the data in DB2 table. While fetching it later, changing code page might not affect it. Is that correct?

changing code page might not affect it. Is that correct?

Maybe, maybe not. The $ in EBCDIC 037 is X'5B'. X'5B' in EBCDIC 1142 (also known as 277) is Å so the hexadecimal representation is not changing even though the display of the hexadecimal value is. You may need to experiment to see how you need the data coded in DB2 and transferred with FTP to get what you want. Code page 1142 has the euro symbol at X'5A' while 277 has ¤ (currency character) at X'5A'.

Thanks.

I am testing it. Will keep posted about the results.