search for string and replace backwards

Shell Programming and Scripting
1

I'm new to Unix scripting and I'm not sure if this can be done. Example:

search (grep) in a file for 'Control ID' and then replace with 4 blanks 7 bytes before 'Control ID.

input
"xxxxxx1234xxxxxxxControl IDxxxxxx"

output:
"xxxxxx xxxxxxxControl IDxxxxxx"

thanks!

2

Hi.

Using sed:

sed "s/\(.*\)....\(.\{7\}Control ID.*\)/\1 \2/" input_file
 
xxxxxx xxxxxxxControl IDxxxxxx
3

thanks a lot..

4

Scott,

the command works ok, but 1 space replaces 4 bytes which compresses the file:

sed "s/\(.\)....\(.\{7\}Control ID.\)/\1 \2/"

In:
abcde****1234567Control ID890

to:
abcde 1234567Control ID890

need:
abcdebbbb1234567Control ID890

with b= space

Thanks!

5

Hi.

Add some spaces between \1 and \2

sed "s/\(.*\)....\(.\{7\}Control ID.*\)/\1    \2/" file2
xxxxxx    xxxxxxxControl IDxxxxxx
6

Thanks Scott! that worked..

---------- Post updated at 11:12 AM ---------- Previous update was at 09:26 AM ----------

when I try to run this on my production file it does not work. I think it's the file size or record length issue. Does sed have a limit for record lenght? should I try awk instead?

7

Hi.

No, it doesn't.

I just ran it on a file with a record (line) length of 750,000, and it worked fine.

Please state what the problem is (error message, etc.), and show what you did.

Thanks.

8

no error messages. nothing happens, looks like it can't find 'Control ID".

When I run this on a small file it works. Could be end of line characters.. not sure.

Background:
This is a AFP file which I converted from ebcdic to ASCII.

when I do a "wc AFP.test1.ascii"

I receive 41 of these messages:
wc: AFP.test1.ascii:41: Invalid or incomplete multibyte or wide character
then...
40 606 10286 AFP.test1.ascii

Not sure if this helps or not.