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

Hi, can anyone help with this little problem I have?

I'm trying to automatically hyperlink url's from some plain text and I have a simple (but not perfect) regular expression that does this.

However when I try and use it with a url that wraps onto a new line, the url itself contains a new line and stops the browser going to the right page.

Allow me to demonstrate with this bit of code: 

#!/usr/bin/perl
use strict;

my $output = "";

open(FILE, "file.txt") || die "$!";
while(<FILE>)
{
  $output .= $_;
}
close FILE;

$output =~ s#(http://[^\!\"\£\$\^\*\(\)\{\}\[\]\;\:\'\@\,\<\> ]+)#\<a 
+href="$1" target="_blank"\>$1\</a\>#gis;

print $output;

[download]
If file.txt has the following: 
The quick brown fox at http://fish.org would like to go home and http:
+//see
through.com/pants/foo/bar what there is to eat

[download]
then my output is: 
The quick brown fox at <a href="http://fish.org" target="_blank">http:
+//fish.org</a> would like to go home and <a href="http://see
through.com" target="_blank">http://see
through.com</a> what there is to eat

[download]
So my question is, is there any way i can remove that \n that is in the middle of the a href?

Many thanks!

Replies are listed 'Best First'.
Re: Automatically hyperlinking text fails with newlines
by suaveant (Parson) on Sep 20, 2001 at 21:05 UTC
    try... 
    while(<FILE>)
{
  chomp;
  $output .= $_;
}

    [download]
    which will remove newlines from $_

                    - Ant
                    - Some of my best work - Fish Dinner

[reply]
[d/l]
      Hi, thanks for this.

      However I only want to remove the \n from inside the hyperlink whilst still preserving the newlines elsewhere. In other words the output should be: 

      The quick brown fox at <a href="http://fish.org" target="_blank">http:
+//fish.org</a> would like to go home and <a href="http://seethrough.c
+om" target="_blank">http://see
through.com</a> what there is to eat

      [download]
      (http://www.seethrough.com is preserved within the href tag, but still has the newline for the display)
[reply]
[d/l]
        well then... you could pass $1 as a function, since you are matching the hyperlink with the \n, similar to... 
        s/regex/make_link($1)/eg;
sub make_link {
  my $url = $_[0];
  $url =~ s/\s+//g;
  qq`<A HREF="$url">$url</A>`;
}

        [download]

                        - Ant
                        - Some of my best work - Fish Dinner

[reply]
[d/l]
        Greetings.

        This being the case, I would (still) slurp the entire file, not replace \n, and put \n among the acceptable characters in the regexp (and I *think* you may have to use the /s modifier). I would then replace the \n still embedded in the href.

        Not perfect, because this:

        Ehi, check this out on http://cnn.com\n
dude!
        would yield: http://cnn.comdude. In other words, if you want ignorable newlines within URIs, then URIS must always be separated by real whitespace from the surrounding text.

        Cheers,
        alf

[reply]
Re: Automatically hyperlinking text fails with newlines
by alien_life_form (Pilgrim) on Sep 20, 2001 at 21:20 UTC
    What about:

    ...
#error checking and stuff left as an exercise.
$/=undef; #slurp files.
$lines=<FILE>;
close(FILE);

$lines =~ s/\n//g;

#now do your thing.

    Note that this joins lines, so if you count on newlines as whitespace, you're out of luck with this.

    Cheers,
    alf

[reply]

Back to Seekers of Perl Wisdom