http://qs321.pair.com?node_id=102967


in reply to appending a unique marker to each url in a file

I think using a regex on the entire file may get a little complicated.

I would use HTML::TokeParser to pull the links out of your file and then modify them from there, like so

#!/usr/bin/perl -w
use HTML::TokeParser;
use strict;

my $i=0;
## set marker definintions
my @markers = qw/ anfnf11 iopi1p83288 9032-jjjf /;
my $htmlfile = "index.html";
my $content;

## get contents of your html file
open (FILE," $htmlfile") || die "Cannot open HTML file for parsing!: $
+!\n";
while(<FILE>) {
        $content .= $_;
}
close(FILE);

my $parse = HTML::TokeParser->new(\$content);
while (my $token = $parse->get_tag("a")) {
        my $url = $token->[1]{href} || "-";     ## put link into $url
        my $text = $parse->get_trimmed_text("/a");      ## put link de
+sc into $text
        if ($markers[$i]) {
                print "<a href=$url/$markers[$i]>$text</a>\n";
        } else {
                ## no more markers...
        }
        $i++;
}
exit;

[download]

Update: fixed the way data was put into $content courtesy of Hofmator.

-p

Replies are listed 'Best First'.
Re: Re: appending a unique marker to each url in a file
by Hofmator (Curate) on Aug 08, 2001 at 14:14 UTC 

    open (FILE," $htmlfile") || die "Cannot open HTML file for parsing!: $
+!\n";
while(<FILE>) {
  $content="$content$_\n";
}
close(FILE);

    [download]
    This construct is not ideal. You are interpolating the variable $content into a new string for every line. You should use concatenation and just append to the string: 
    while(<FILE>) {
  $content .= "$_\n";
}

    [download]
    But what you are doing now is slurping in the whole file and adding an extra newline at the end of each line (for which I see absolutely no reason). The same thing can be achieved by undefing $/ like this: 
    {
  local $/; # undefs $/ for this block of code only

  open (FILE," $htmlfile") || die "Cannot open HTML file for parsing!:
+ $!\n";

  $content = <FILE>; # reads in whole file

  $content =~ s/\n/\n\n/g; # if really necessary to duplicate newlines

  close(FILE);
}

    [download]

    -- Hofmator

[reply]
[d/l]
[select]

      How about just $content = join'', <FILE>; No need to mess with $/ and have to remember to localise it. Wo betide he who forgets to localise $, $" $/ $\

      Update

      As chipmunk points out this is slower than undef $/ for the gory details see Re: Re: Re: Re: Re: appending a unique marker to each url in a file. For big files the difference is significant, for small ones it is negligible but who wants to paint themselves into a scaling corner? It is better to undef $/, just remember to localise it.

      Ugh posted bad code again.

      cheers

      tachyon

      s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print

[reply]
        Using $content = join '', <FILE>; rather than local $/; $content = <FILE> is fine, as long as you don't care that the latter is roughly an order of magnitude faster.
[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom