RE: CSV file problems

[BMay@xxxxxxxxx]

Sheri,

In your CSV file, are the item numbers inclosed in double quotes?

Brian May
Project Lead
Management Information Systems
Garan, Incorporated
Starkville, Mississippi



"Rowe, Sheri" <srowe@xxxxxxxxx> 
Sent by: midrange-l-bounces@xxxxxxxxxxxx
05/20/2010 02:11 PM
Please respond to
Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>


To
"Midrange Systems Technical Discussion" <midrange-l@xxxxxxxxxxxx>
cc

Subject
RE: CSV file problems






Scott,

I have a question about CCSID.  I created a CSV file in the IFS for
emailing.  The raw data looks good, but when I pull it up in Excel 2 of
the item numbers look incorrect.  When I check the format of the
incorrect cells in Excel, it is 'Scientific'.  Other item numbers in the
same column that are correct, have a format of 'General'.

I tried creating the CSV with CCSID of 37 and 1208 and 1252 but it still
gives me incorrect data for 2 items.

The 2 item numbers are 5E021 and 5E011.

Is this purely an excel issue with CSV files, or is it something I can
fix from the iSeries?

When I import the file directly from the iSeries to Excel the 2 item
numbers in question are correct.

Sheri 

-----Original Message-----
From: Scott Klement [mailto:midrange-l@xxxxxxxxxxxxxxxx] 
Sent: Thursday, May 20, 2010 2:01 PM
To: Midrange Systems Technical Discussion
Subject: Re: CSV file problems

Pat,

A CCSID identifies a particular variation of ASCII or EBCDIC.  It's just

a ID number for a particular (precise) character set.

Just like when you create a customer record in your database, you refer 
to it by it's customer number (or account number).  IBM did the same 
sort of thing for character sets, they assigned a number to each one, 
and they refer to it by that number.

Here are a bunch of them

   37 = EBCDIC for USA (and various others)
  273 = EBCDIC for Germany, Austria
  285 = EBCDIC for UK
  297 = EBCDIC for France
        (you get the idea... there are lots of EBCDICs)
  437 = An old flavor of ASCII for MS-DOS in the USA
  819 = ISO-8859-1, an ISO standard flavor of ASCII covering most/all
        Latin-1 characters
1252 = A (more or less) superset of ISO-8859-1 used for Latin-1
        ASCII Microsoft Windows.
1208 = UTF-8 Unicode
1200 = UTF-16 Unicode

You say that you are outputting 437 right now...  and that may be close 
enough (for basic letters and numbers it'll be no problem) but certain 
special characters may be mistranslated.  Nobody really uses 437
anymore.

I'd suggest that 1208 is probably the right CCSID for Excel.  Most 
everything in Windows expects Unicode these days.  But if you need a 
single-byte character set, I'd say 1252 is more likely to be correct 
than 437.

But, anyway...   I hope you understand that CCSIDs aren't really 
languages, they are just specific flavors of ASCII, EBCDIC or Unicode. 
Just numbers to identify them.