Re: Odd character as sub-element separator

Tue, 07 Nov 2000 11:03:00 -0800

Mark,
I would go and ask them to change it. It can't hurt.
Companies are of two distinct camps on this issue. Some use printable characters and
others non-printable characters. It appears that your trading partner is of the earlier description.
My immediate question is what format is the data really in. You will need to hex dump the
data file to determine this. I find it highly unlikely that the single byte is trashed, unless the conversion has a problem in it somewhere. There has to be a specific place
where the EBCDIC to ASCII conversion happens.
I have seen ASCII systems with a reversed order collating sequences. In the event that this
were the case, the whole file would be trashed, not just the single bit.
As a move toward testing, edit your trading partner's file to change the byte in question to be another character you can process. The sub-element separator should not be anywhere in the data portion of the file as that would produce errors. If your not getting sub-elements, which is the norm in X-12, you will only need to change the byte in the ISA segment of the
file. This strategy should get you beyond the sub-element separator while you work the issue
of getting the sub-element worked out with the trading partner. One word of warning, editing an X-12 file can be tricky. Hex dump the final product and re-check the file before you process it. HEX dump is the only way that you can see what's really there printable or non-printable.
Good luck....
Mark
-----Original Message-----
From: Mark Dixon [mailto:[EMAIL PROTECTED]]
Sent: Monday, November 06, 2000 7:55 PM
To: [EMAIL PROTECTED]
Subject: Odd character as sub-element seperator
Newbie Question:
I have just started testing X12 (4010) with a large company that has
apparently been doing EDI for a long time. I have just downloaded the first
batch of test messages from an IBM Information Exchange network over the
AT&T secure IP network. I had a problem when I tried translating the test
messages to a flat mapping to view the contents:
There is an oddity in the ISA: The sub-element separator is specified as
0xAD, that is a minus sign with the high-order bit on.
1. Why would this be? Perhaps I can answer my own question there: my guess
is that the output is from an IBM mainframe and this the result of a wierd
EBCDIC to ASCII conversion.
2. Is it valid? Ie. should I try to insist that they change it?
3. Any suggestions on what I should do about it? (In this case it is
probably me that will have to adapt, not the trading partner because they
are very big and my organisation is very small).
Cheers, Mark Dixon, Anysize Software Pty Ltd
e-mail: [EMAIL PROTECTED]
Phone: +61 8 9249 4703 Fax: +61 8 9249 4052
=======================================================================
To signoff the EDI-L list, mailto:[EMAIL PROTECTED]
To subscribe, mailto:[EMAIL PROTECTED]