Shell Programming and Scripting

Greetings,

I need to take a fixed length file, similar to the following:

<input file>

1233 e 612 i
43378 f 03 x
22 17 e 9899
a 323e a6 z7

read in the character in position 6, and if that character = e, delete that line from the file.

<output file>

43378 f 03 x
22 17 e 9899

and if possible, output the deleted lines to a new "log" file.

<log file>

1233 e 612 i
a 323e a6 z7

The old file must maintain its structure, the log file wouldn't.

I'm on an AIX platform, and have sh/ksh/bash available to script in.
I'm sure it can be done with some combination of cut/sed/awk, I just haven't been able to get it right as of yet. Any help would be greatly appreciated!

Thanks in advance.
DB

sed solution(s) :

Code:

sed '/^.....e/d' inputfile
sed /^.\{5\}e/d' inputfile

Jean-Pierre.

Also note the w command in sed for writing stuff to a named file.

You don't really need sed, though.

Code:

grep ^.....e >ditch
grep -v ^.....e >keep

Thank you Jean-Pierre, that works great, much appreciated!

-DB

Thank you Era!