not parsing properly??

wisemonkey has asked for the wisdom of the Perl Monks concerning the following question:

Hi everyone I need to delete particular files from directory structure, So I'm using modified version of File::Found function which from perl monks as below

#!/usr/bin/perl
use warnings;
use strict;
use diagnostics;
use File::Find;

sub eachFile {
  my $filename = $_;
  my $fullpath = $File::Find::name;

  if (-e $filename) { print "$filename exists!\n"; }
  if ($filename =~ /index.html.*/){
        print "$filename need to be deleted\n";
        system "rm $filename";
        }
}

find (\&eachFile, "./");
[download]

examples of filenames are:

 index.html
index.html?C=D;O=D
[download]

etc. Output of given code comes out to be:

index.html?C=S;O=D exists!
index.html?C=S;O=D need to be deleted
rm: cannot remove `index.html?C=S': No such file or directory
[download]

I don't understand why its being so, I appreciate your suggestions

Comment on not parsing properly?? Select or Download Code

Replies are listed 'Best First'.
Re: not parsing properly?? by toolic (Bishop) on Apr 17, 2011 at 23:37 UTC
Your system statement passes the string `rm index.html?C=S;O=D` to the unix shell. Since the semicolon has special meaning to the shell (it is a command separator), the shell breaks your string up into 2 commands: `rm index.html?C=S O=D` [download] Since you do not have a file named "index.html?C=S", you get the "No such file or directory" message. Try to replace your system with unlink: `unlink $filename or die $!;` [download] If that doesn't work, you may have to resort to some type of escaping heroics using backslashes or quotes.	[reply] [d/l] [select]
Re^2: not parsing properly?? by wisemonkey (Novice) on Apr 17, 2011 at 23:49 UTC
Awesome thanks a ton. That fixed it, Currently script is running. I'll let you know as its done deleting all junk :)	[reply]
Re^3: not parsing properly?? by wisemonkey (Novice) on Apr 18, 2011 at 01:07 UTC
yup cleaned and confirmed, thanks again :)	[reply]
Re: not parsing properly?? by ikegami (Patriarch) on Apr 17, 2011 at 23:32 UTC
You are creating a shell command without properly converting the file name into a shell literal. You could avoid the shell by using the multi-argument form of `system`, or you could delete the file using `unlink` instead of running another program.	[reply]
Re: not parsing properly?? by anonymized user 468275 (Curate) on Apr 18, 2011 at 13:30 UTC
In addition to the semicolon issue already mentioned, I think it is worth pointing out that the regexp covers a slightly larger range of possibilities than specifically intended. "." in a regexp means any character, so the "." between index and html needs to be escaped with a "\" to avoid, e.g. indexXhtml.* being included . Probably not important this time, but it might be another time. One world, one people	[reply]