sed trouble

hfawzy · 08-06-2006, 02:08 PM

Hi,
I'm having some troubles replacing a string in a file using sed.
Here is what I want : I want to replace the first match occured with 1 the second match with 2, etc... till the end of the file.
For example if the file looks like this :

Quote:

Sample text..
Here is some text...
Some text

I want it to look like this after applying the above replacing system to string "text":

Quote:

Sample 1..
Here is some 2...
Some 3

How can I do that ?
Thank you.

sdexp · 08-06-2006, 02:32 PM

Code:

i=1
for word in `echo "Sample text..
Here is some text...
Some text"`
do 
   echo $word | sed 's/text/$i/g'
   if [ $word = "text" ]
   then
      let "i+=1"
   fi
done

I haven't tested it, but think it will work, maybe with a few modifications.

hfawzy · 08-06-2006, 03:20 PM

Thanks for the reply. This almost solved the problem.
Problem is that $i incrments, even if the word is not "text". I want to increment $i only if "text" is found and has been replaced.

homey · 08-06-2006, 05:21 PM

Will this work for you?

Code:

perl -ne 'print if /text/' < file.txt | perl -pe 's/text/$./'

spirit receiver · 08-06-2006, 05:36 PM

I think this is what hfawzy wants. It makes a backup of the original file.

Code:

perl -i.bak -e '$i=1; while(<>){$i++ if(s/text/$i/);print}' filename

sdexp · 08-07-2006, 10:37 PM

Yes, you caught me dosing off there hfawzy.

I did correct it though in that post.

Here it is again:

Code:

i=1
for word in `echo "Sample text..
Here is some text...
Some text"`
do 
   echo $word | sed 's/text/$i/g'
   if [ $word = "text" ]
   then
      let "i+=1"
   fi
done