|
|
 next
|
  Subject: converter:EDI:val=no AND chr=...?????
Author: Tony Lavinio
Date: 03 Aug 2009 11:20 PM
|
1) It happens because those characters cause the input for the COM01 to be
completed, so it is at the start of where the COM02 element would be
that triggers the error message.
2) It has to check them, because it is impossible to write out valid EDI
if the input is not convertible. Although we let you turn off many
types of validation, we don't let you disable anything that will produce
syntactically invalid EDI downstream.
3) UNOC corresponds to ISO-8859-1. It's not clear which character set
you should be using; if we could see a sample of the file that is
causing trouble we could help.
The various definitions of the character sets in EDIFACT can be found
here: http://www.gefeg.com/jswg/cl/v41/40107/cl1.htm
For more clarification on the contents of each set, Wikipedia is good
source because it shows exactly which characters go in each position.
Now, if the delimiters are set in the usual fashion, there should be
no problems with ', ? or + in input files if the input is XML, or even
if the input is EDI and the characters are escaped with the defined
release character. If they are not escaped in incoming EDI, then an
EDI parser ends up with an ambiguous condition, and cannot continue.
If you want to send a copy of the file for us to look at, zip it
(so that intervening mail transport agents don't mess with any of the
non-ASCII characters that might be present) and send it to
stylus-field-report@progress.com.
Does this help?
|
|
|
|
top