URLs

sridhar_dvjs,
i'm not fully sure what yo want here as your question looks short of information. Perhaps you could explain it a bit more.

However, it seems you would like to get rid of ../ within your URL's.  These characters tell the operating system to go to the preceeding directory.  If you take these out you will may navigate to the desired directory path.

if you wish to take them out you could use a regular expressions.

peewee

sridhar_dvjs,

an example regex is below:

#!/usr/bin/perl
print "start\n";
 
my $input = 'http://domain/dir1/dir4/../dir2/../../dir3/file.html';
 
#http://domain/dir1/dir2/../../dir3/file.html
#http://domain/dir1/../dir3/file.html
 
print "$input\n";
$input =~ s/\/\.\.//ig;
print "$input\n";
 
print "end\n";

Something like this might work:

#!/usr/bin/perl
##########################################################################
@url = ('http://domain/dir1/dir4/../dir2/../../dir3/file.html',
'http://domain/dir1/dir2/../../dir3/file.html',
'http://domain/dir1/../dir3/file.html');

foreach $l(@url){
     
     $l =~ s/(\.\.\/)//gis;
print "$l\n";
}

Hope this helps..
windfall

@url = ('http://domain/dir1/dir4/../dir2/../../dir3/file.html',
'http://domain/dir1/dir2/../../dir3/file.html',
'http://domain/dir1/../dir3/file.html');
foreach( @url ){
    1 while s![^/]+/\.\./!!;
    print "$_\n";
}
#assuming no symbolic links

sridhar_dvjs,

"..I have several URLs like..."

Where are these URL's in? Are these URL's inside HTML or other types of files?

If yes, then can you post a sample file here?

"..I guess a one liner can do the job!!..."

Do you have command prompt access to run any command?

Please provide as much more detail as you can.

This will help you get a more accurate answer, faster.

What the poster is asking is "how can I change a directory path with ../'s in it to the path without ../'s?"  In other words, how can I get from

  /this/that/../foo/bar/blat/../it.txt

to

  /this/foo/bar/it.txt

Here is such code:

  $URL = "http://www.pobox.com/../~japhy/regexes/../

  # remove "leading" ../'s
  $URL =~ s{(http://[^/]+/)(?:\.\.(?:/|$))+}{$1};

  # remove other /..'s
  1 while $URL =~ s{[^/]+/\.\.(?:/|$)}{}g;

This code works with URLs such as "http://www.foo.bar/blat/..", even though the ".." doesn't have a trailing slash.