search pattern and replace x-y characters in nth line after every match

Shell Programming and Scripting
1

Hi,

I am looking for any script which can do the following.
have to read a pattern from fileA and copy it to fileB.

fileA:
...
...
Header
...
...
..p1
...
...

fileB:
....
....
Header
...
...
..p2
....
...

The position of "p1" is fixed relative to "Header"
say x th charecter in nth line after Header.
"p1" size is also fixed 5 chars if it helps.

p2 is location and size are same as p1
i have to read "p1" from fileA and replace "p2" in fileB with "p1"

How can we do this ??

Thanks in advance.

2

This will work for a line, maybe more details will help me...

replace.sh:

#!/usr/bin/awk -f
FNR==1 { fnum++ }
/^HEADER$/ {h=1;l=0}
h {l++}
l==3 && fnum==1 { p1=$0; next }
l==3 && fnum==2 { print p1; next }
fnum==2 { print $0 }

[mute@geek ~/test]$ cat fileA
HEADER
some kinda of data aAAaaa
this is the new data aaAAAA
more data etc aaaa

[mute@geek ~/test]$ cat fileB
HEADER
some kinda of data in BBBB
this is the old data bbBBBb
more data etc bBBbBB

[mute@geek ~/test]$ ./replace.sh fileA fileB
HEADER
some kinda of data in BBBB
this is the new data aaAAAA
more data etc bBBbBB
3

Thanks scott for your quick reply.
In my case p1 starts at char 5 in line 4 after HEADER. p1 is 7 charecter long.

And each of the fileA and fileB will have same number of such occurances.
I have to copy from fileA to fileB in the same order.

4

This is what I came up with.

#!/usr/bin/awk -f
FNR==1 { ++fnum; c = 0; }
# keep track of line# after header
h {l++}
/^HEADER$/ { h = 1; l = 0; }
# store p1
l==4 && fnum==1 { p1[c++]=substr($0,5,7); next }
# replace p2 with p1
l==4 && fnum==2 { print substr($0,1,4) p1[c++] substr($0,12); next }
# print non-matching fileB files as-is
fnum==2 { print $0 }