Questions on sed and grep

[Lakshmipathi]

Hi all,
I facing 2 problems:
problem 1:
How to grep lines ?
I want to grep a line with 2 underscores in a file like this.

L1field1_L1field2_L1field3_L1field4
L2field1_L2field2_L2field3
L3field1_L3field2

The line with two underscores,which is :
L2field1_L2field2_L2field3

problem 2:
How to use sed with \ character ?
I want to replace word called "word_123" into "word\123"
replace _ with \ using sed.

Cheers

[Franklin52]

To replace the underscore with a backslash in lines with 2 underscores:

awk 'BEGIN{FS="_"}NF==3{print $1 "\\" $2 "\\"$3;next}{print}'

Regards

[ghostdog74]

shell script

#!/bin/sh
while read -r line
do
 before=`printf $line|wc -c`
 after=`printf $line | tr -d '_' | wc -c`
 n=`echo "$before - $after" | bc`
 if [ $n -eq 2 ];then
    echo $line | sed 's/_/\\/g' 
 fi 
done < "file"

[wowbag1]

Hello, out of interest unless you really need or want to use sed, Perl on the command line is very quick.
at prompt write:
perl -p -i -e 's/word_123/word\\123/g' then the file name

[Franklin52]

Hello, out of interest unless you really need or want to use sed, Perl on the command line is very quick.
at prompt write:
perl -p -i -e 's/word_123/word\\123/g' then the file name

The OP only wants to edit lines with 2 underscores.

Regards

[Lakshmipathi]

Thanks guyz for your help