http://qs321.pair.com?node_id=11145864


in reply to URL resolve (operations on path etc.)

use warnings;
use strict;
use URI;

my $x = 'http://a.com/a/b/c/d/e.json?x=1&y=2';
my $y = '../../../';
my $u = URI->new($y);
print $u->abs($x), "\n";

__END__

http://a.com/a/

[download]

Replies are listed 'Best First'.
Re^2: URL resolve (operations on path etc.)
by bliako (Monsignor) on Jul 31, 2022 at 08:03 UTC

    Thanks haukex++ (I swear last night abs was not working!!!!!!)

    EDIT!: ouchhhh I was doing the reverse!!!!!:

    # this is wrong, note new($x)!! abs($y) it should be the reverse
my $u = URI->new($x); # this is wrong
print $u->abs($y), "\n"; # this is wrong

    [download]

    Edit2: that's definetely an X-Y problem hehe

[reply]
[d/l]
[select]
Re^2: URL resolve (operations on path etc.)
by ikegami (Patriarch) on Aug 01, 2022 at 22:50 UTC
[reply]
[d/l]
Re^2: URL resolve (operations on path etc.)
by harangzsolt33 (Chaplain) on Jul 31, 2022 at 11:31 UTC
    Could this be resolved with a regex maybe? Like for example: 

    #!/usr/bin/perl

use strict;
use warnings;

my $PATH = 'http://a.com/a/b/c/d/e.json?something' . '/' . '/../../../
+';
$PATH =~ tr|/||s;
my $PREV = 0;
my $LEN;
for (;;)
{
  $PATH =~ s/\/[^\/]+\/\.\.//;
  $LEN = length($PATH);
  last if ($PREV == $LEN);
  $PREV = $LEN;
}
print $PATH;

    [download]

    I put the regex in a for(;;) loop, because adding global flag didn't do what I want:

    $PATH =~ s/\/[^\/]+\/\.\.//g;

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom