Hi All, i have a problem with using sed in a perl script. Requirement: Read a csv file (file1) containing certain pattern, match the pattern in another text file (file2) and delete it where ever its occurring.

file1:

abc/def
sde/dfg/htf
[download]

file2:

abc/def = 100
sde/dfg/htf = 230
dad/fry = 000
[download]

in the end only dad/fry = 000 should remain in file2.

I cannot install any package, so have to use the default perl options.

Since my file1 contains special character "/" , i have escaped it in my code.

my sed command is working fine if I try to run it without loop, i.e for one pattern only. As soon as I add it in loop, it fails.

my code:

my $var = "sde\\/dfg\\/htf";
`sed -i '/$var/d' ADT_DVEBMGS33_ldciadt`;
[download]

The above works fine.

i tried to encorporate in a loop

my $len = scalar @delete;
for ( my $i = 0; $i <=$len; $i++){
        my $first = $delete[$i];
        chomp $first;
        $first =~ s:/:\\\\/:g;
        `sed -i "/$first/d" file2`;
}
[download]

error

<c> sed: -e expression #1, char 10: extra characters after command sed: -e expression #1, char 0: no previous regular expression