I'll suppose the key is the first word on the line,

$ perl -pi -e's/^TOM\b.*\n//' userlist.txt
$
[download]

Isn't this deleting the pattern from the line, rather than deleting the line itself?

If it's Unix/Linux, you might consider using grep or fgrep with the "-v" switch to skip lines containing a certain pattern.

For example, to remove lines containing "TOM" anywhere in the line, use fgrep ("fixed grep") like this:

% fgrep -v TOM userlist.txt > edited.txt
[download]

% grep -v "\<TOM\>" userlist.txt > edited.txt
[download]

Sorry guys, i meant in the script, not from command i am running a script that can add users and remove users. add users is done, but the remove users, i'm having trouble figuring out..

What do you have so far?  Can you show some code, even it isn't quite working, so we can see which part needs to be tweaked?

And what are the specific requirements?  Removing a line (or lines?) that exactly matches/match a given name?  Does it have to work for multiple names, or just one name?  Can the name be anywhere in the line(s)?  Can there be other names in the line, or not?  Does whitespace matter?

If you can show us the code that you have so far, and pinpoint the exact requirements, I'm sure we can get you farther towards your goal.

---

s''(q.S:$/9=(T1';s;(..)(..);$..=substr+crypt($1,$2),2,3;eg;print$..$/

my $replacestring = "TOM";
open(my $infile,"<","inputfile.txt") or die $!;
open (my $outfile,">","outputfile.txt") or die $!;
while (<$infile>)
{
    $_=~s/$replacestring//g;
    print $outfile $_;
}
close $infile;
close $outfile;
[download]

That will read the content of 'inputfile.txt', and write out the content to 'outputfile.txt', removing $replacestring along the way. The key bit is $_=~s/$replacestring//g - that replaces any instance of $replacestring in the $_ variable with blank.

The remaining step is to rename 'inputfile.txt' as something else, and rename 'outputfile.txt' as 'inputfile.txt'.

Update: Something else you might want to look at: your post says that you want to remove a string, but later you say a line. This code removes the entire line - I'll let you work out why ;-)

my $string = "TOM";
open(my $infile,"<","inputfile.txt") or die $!;
open (my $outfile,">","outputfile.txt") or die $!;
while (<$infile>)
{
    if ($_ !~/$string/)
    {
       print $outfile $_;
    }
}
close $infile;
close $outfile;
[download]

Hi Your solution seems to be great. But i have a different problem. I have a file 'A' which contains say n lines each having unique name. Now in another file 'B' I need to search for these n strings from 'A' and delete those lines in this 'B' and create a new file 'C'. Can any one help me on this

Your post isn't going to attract much attention here at the bottom of an old thread. Could I suggest that you repost here with a more detailed version of your question (and what code you've got so far).

--------------------------------------------------------------

"If there is such a phenomenon as absolute evil, it consists in treating another human being as a thing."
John Brunner, "The Shockwave Rider".

What you want to do is open the original file and open a new output file say "userlist-output.txt" for writing. Then you just loop over the original file's contents and when you find the line you want to omit, simply don't output it to the new output file.

Thanks for finding a solution guys and gals. Just saved me the headache of having to remove the word "remove" appended on the end of each line in a document with 300 lines.

#!perl
print "What is the name of the file you want to remove a word ar phras
+e from? \nPlease include the extension. \n";
chomp(my $rem_name = <STDIN>);
print "Please enter a word ar phrase you want to remove: \n";
chomp(my $replacestring = <STDIN>);
open(my $infile,"<","$rem_name") or die $!;
open (my $outfile,">","outputfile.txt") or die $!;
while (<$infile>)
{
    $_=~s/$replacestring//g;
    print $outfile $_;
}
close $infile;
close $outfile;
[download]

You could have saved yourself another headache and a buncha work if you had enclosed the entire code block in  <code> ... </code> or  <c> ... </c> tags rather than paragraph tags and lots of <br>s. Please see Markup in the Monastery, Writeup Formatting Tips, Perl Monks Approved HTML tags and How do I post a question effectively?.

---

Give a man a fish:  <%-(-(-(-<