UNIX for Dummies Questions & Answers

Hi there,

I have a folder with many files where the first line has almost the same pattern in all the files and I want to replace this one with another which is going to be the same in all the files.

So the first line of all the files begins like
UNB+IATA:1+CHUAGT87HUUPSS/BUD01:PIMA

and I want to replace this one
UNB+IATA:1+REUAGT87GLSEHDQ/FRA16:PIMA

I tried the following but it doesn't see to do anything.

Code:

grep 'UNB+IATA:1+[A-Z,0-9]\{1,\}\/\{0,\}[A-Z,0-9]\{0,\}:PIMA' *.msg|sed 's/UNB+IATA:1+[A-Z,0-9]\{1,\}\/\{0,\}[A-Z,0-9]\{0,\}:PIMA/UNB+IATA:1+REUAGT87GLSEHDQ/FRA16:PIMA'

I tried various modification but always I was getting errors like

sed: -e expression #1, char 85: unknown option to `s' or
sed: -e expression #1, char 2: extra characters after command

can somebody please advise?

Something like this?

Code:

awk 'NR==1{print "UNB+IATA:1+REUAGT87GLSEHDQ/FRA16:PIMA";next}1' file > outfile

Code:

sed '1 {;s@UNB+IATA:1+CHUAGT87HUUPSS/BUD01:PIMA@UNB+IATA:1+REUAGT87GLSEHDQ/FRA16:PIMA@;}'

This will change the mentioned text only in the first line. Wrap this into a file-by-file-loop and be done.

Code:

ls | while read file ; do
     sed '.....' ${file} > ${file}.new
     mv ${file}.new ${file}
done

I hope this helps.

bakunin

PS: the reason why your regexp failed is the "/" inside of your replacement string. Probably it would work if you escape all instances of it (replcace "/" with "\/").

But the problem was that my sed was not working. Because UNB+IATA:1+CHUAGT87HUUPSS/BUD01:PIMA is not in every file the same. It just has the same pattern. Additionally it is only the beginning of the line and NOT the whole line. That's why I tried the code I gave in my first post.
My sed needs a correction first before I try the ways that you told me.

Thanks

Edit: Sorry I just saw your PS. I will check and tell you.

Yeap, works like a charm. Thanks a lot for your help