replace certain chars in every line of a text file

[Steve Dubinskiy_1]

Hello, I need some help
I need to replace chars 7-10 in each line of a file with something else.. I looked and SED and TR but can't find the proper syntax.

Please help
thanx,
Steve

[A. Clay Stephenson]

Here's a simple method in awk but Perl would do this easily as well:

First create an awk script, my.awk.

{
if (length($0) >= 7)
{
s1 = substr($0,1,6)
s2 = "NEW!"
s3 = substr($0,11)
printf("%s%s%s\n",s1,s2,s3)
}
else printf("%s\n",$0)
}

Now:

awk -f my.awk oldfile > newfile.

[Martin Johnson]

In vi:

:= #Gives total lines in file (n)

:1,ns/str1/str2/g #replaces str1 with str2 in the entire file

HTH
Marty

[Paul Sperry]

in vi

:%s/old/new/g

[Carlos Fernandez Riera]

I cant help you, but try this one:

http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0x001d63f96280d711abdc0090277a778c,00.html
I've saved it.

Maybe someday i will learn martian.

[V.Tamilvanan]

Hi,

#sed 's/Old/New/g' inputfile >outputfile

HTH

[Martin Robinson · Accepted solution]

Since you specifically want to replace definite column positions, s/old/new/ won't do what you want. This will:

cat file | sed 's/^\(.......\).../\1XXX/'

This means substitute, matching a pattern that starts at the beginning of the line ^, match any 7 characters and keep them, because they are enclosed in \( and \). Then match the next three characters. Replace this with whatever the first 7 characters were using \1 then add the new stuff you want, XXX. There is also no special processing required for lines shorter than 10 characters, as the pattern will simply fail to match.

[H.Merijn Brand (procura]

# perl -pi -e'substr($_,6,3)="Boo!"' your_file

Enjoy, have FUN! H.Merijn

[Karl Mumford]

cut -b 1-6 infile | sed 's/$/4NEW/' >f1
cut -b 11- infile >f2
paste f1 f2 >outfile

[V. V. Ravi Kumar_1]

Hi,
You can achieve this as follows

for i in `cat `
do
echo $i > (tmpfile1)
str=`cut -c 7-10 tmpfile1`
sed 's/'$str'/xxxx/' tmpfile1 >> /tmp/
done

where "xxxx" is your substitute string

Regards

[Steve Dubinskiy_1]

Wow,

didn't expect so many replies..
thanx a bunch


Steve

[Steve Dubinskiy_1]

also, wanted to thank Martin Robinson.. his advice worked for what I needed.

using sed 's/Old/New/g' wouldn't work in my case because I didn't have anything to match to.. the only constant in my case was that I needed to replace from char 7 to char 10.

I didn't try using suggested AWK scripts, because I'm not very familiar with awk and didn't know what it was doing ( and I don't want to run scripts that I don't know what they do )

anyhow.. I want to thank everybody for their suggestions.. its great to know that I can turn to these forums if I ever need any more help.

thanx,
Steve

[Brian Kinney]

Steve -
now you should go back and assign points to these answers. It's another way to say thanks to those who've answered your questions.

[John Hall]

The following uses "ex" to accomplish your task. This line should all be typed on the same line as a single command. It looks a little cryptic but, hey it's Unix, and it works:

#-----------------------
echo "1,\$s/\(......\)..../\1abcd/\n:wq!" | /usr/bin/ex xyz
#-----------------------

where "abcd" is the new text that you want to replace the old text, and "xyz" is the file to be modified.