UNIX for Dummies Questions & Answers

I'm just getting back into UNIX - but I was impressed - I'd been working on it for 3 days - he came up with it in less than an hour

Who got it right on the first try? I could not get sed to do it. I gave up and let sed do the first phase and awk do the second phase. Even that took a few tries. The final result makes a few asumptions that might not have been true.

1 the first record can be ignored
2 there are never 2 record separators on one text line
3 the desired data will always be on the same line as a record separator

Had any of those been false, this would not have completely worked.

I think I'm getting a complex

Well it almost worked. The problem I'm having is that some of the data wraps around multiple lines and for them it's pulling blanks. The 3rd and 5th lines in the sample text show the problem. I need the 5th field and on alot of the records it's just not there. The only things I can figure out to tell it it's a multiple line record require the same character in the first position of a line. Any ideas? ^

856^18530001~BSN^14^486165^040927^181323^0004~NTE^856^18530002~BSN^14^486399^040927^181323^0004~NTE^
856^18
856^18530004~BSN^14^489012^040927^181323^0004~NTE^
856^18530005~BSN^1
856^18540001~BSN^14^488594^040927^181941^0004~NTE^
856^18540002~BSN^14^489016^040927^181941^0004~NTE^
856^18550001~BSN^14^488567^040927^182539^0004~NTE^
856^18550002~BSN^14^489102^040927^182539^0004~NTE^
856^18560001~BSN^14^484633^040927^183156^0004~NTE^

I don't follow you. I edited your post to add code tags to easily see the lines...

Line 3 starts out...
ISA^00^D011-56018

I have line 5 as:
PER^BD^JANET ARROYO~N1^ST^^9

Neither line contains a ~ST sequence. Only lines 4, 7, and 15 have the ~ST sequence. So my command produces 3 lines of ouput:
440597
440598
440601

True, on the sample file - but on some of the ~ST is near the end of the line and continues on the next line - when this is the case the ~ST is on one line and the 440601 would on the next line. I think what I need to figure out how to do is to tell it to continue to the next line for the record until it hits the ~ST again. Does that make sense? When I run the SED part of the command you sent me I get the results I just posted. I ran that seperately because I was seeing blanks in the file where I should have had records.

Try this...

sed -n '/\~ST\^/{N;s/\n//;s/.*\~ST\^//p;}' < inputfile | awk -v FS=^ '$1=855 { print $5 }'