I am assuming you are using ver 8's Monarch utility.
I would try replacing ASCII 13 (or is it ASCII 13 and ASCII 11 together?) with another character with the "replace all instances of a string with another" function. Use some character that is not used in the EDI as a new delimiter. Then take the processed file back through the EDI prep with the new delimiter.
Let us know if this works or not.
I am using version 8. I replaced all CR characters with single-left quote (`) characters. My manual EDI formatting options were ...
segment delimiter: `
element separator: |
sub element separator: ^
release character: (none)
The utility did two things.
1. Added a space between all of the empty element separators whose element was null.
2. Added a CRLF after each of my single-left quote segment delimiters.
Unless someone has any ideas, I think I am better off trying to do something with my original HL7 file.
I know nothing about HL7, but did find [url="http://itc.musc.edu/wiki/HL7"]A web page[/url] that states the segment delimiter is a pipe character & not CR.
Sorry I can't help with this.
/b[/quote]That's an interesting find Bruce.
Assuming it is accurate rather than a humorous criticism of the standards, it seems that, in reality, almost anything goes for a type 2 HL7.
I'm not sure it would help much but I would be tempted to use the Examine.exe utility to see what is there in terms of character count in a few records. At least it would be very clear whether the records are consistent and indicate precisely what characters are in use.