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

I am attempting to write a perl script which outputs the contents of the html tags to a seperate file. Unfortunatley sometimes/most of the time, the close </html> tag is on a seperate line. I tried using the chomp command to remove all line returns, but my code is still return no results. Can anybody help.

while (<INPUT>)
{
   my $TheLine = $_;
   chomp $TheLine;

   if ($TheLine =~ /<html>[^\n]*<\/html>/i)
        {
        $TheLine =~ s/<html>([^\n]*)<\/html>/$1/i;
        print OUTPUT "$1\n";
    }

}
close INPUT;
close OUTPUT;

[download]

Replies are listed 'Best First'.
Re: Search and replace in html
by tall_man (Parson) on May 08, 2003 at 23:48 UTC
[reply]
Re: Search and replace in html
by Limbic~Region (Chancellor) on May 09, 2003 at 00:06 UTC
    Anonymous Monk,
    There are a plethora of modules on CPAN that could help you, I would suggest looking at the following search. Roll your own solutions with unknown data sources are likely to fail. With that said - let's assume your HTML is perfectly formatted and you want everything between the start and end HTML tags to include other HTML tags. 
    #!/usr/bin/perl -w
use strict;
open (INPUT,"file") or die "Unable to open input : $!";
open (OUTPUT,">output") or die "Unable to open outpu : $!";
select OUTPUT;
$\ = "\n";

my $foundstart;
while (<INPUT>) {
    chomp;
    next unless ($foundstart || /<html *>/i);    
    if (/<html *>/i && ! $foundstart) {
        $_ =~ s/^.*?<html *>(.*)$/$1/i;
        $foundstart++;
        next unless($_);        
    }
    if ($_ =~ m|</html *>|i) {
        $_ =~ s|^(.*?)</html *>.*$|$1|i;
        print if($_);
        last;
    }
    print;    
}
close INPUT;
close OUTPUT;

    [download]

    Cheers - L~R

[reply]
[d/l]
      Thanks guys, it has started to get me on my way. I have now found that the files contains multiple html references. Would it simply be a matter of just changing the if to a while.
[reply]
        Anonymous Monk,
        No - you should not try to roll your own unless you are 100% sure of your data. That is what I was trying to point out. Follow tall_man's advice or find a module you like using the search I provided.

        Cheers - L~R

[reply]
Re: Search and replace in html
by kilinrax (Deacon) on May 08, 2003 at 23:32 UTC

    Sounds like you could want to set the input record separator ('$/') to undef, then you'll pull the file in one huge chunk rather than line-by-line.

    Try either of the following lines: 

    undef $/;
local $/;

    [download]

    The first line will set '$/' to undef for the rest of the file, the second only for the enclosing scope (arguably better).

    You might want to try reading perlvar to get a better idea what '$/' is, and maybe pick a better value to set it to.

[reply]
[d/l]
Re: Search and replace in html
by LameNerd (Hermit) on May 08, 2003 at 23:31 UTC
    Do you want something like this? 
    #!/usr/bin/perl -w
use strict;

while(<DATA>) {
    next if /<HTML.*?>/gi;
    next if /<\/HTML.*?>/gi;
    print;
}
__DATA__
<HTML>
<HEAD><TITLE>Homepage</TITLE></HEAD>
<BODY>
<a href='blah.html'> man blah.pl</a><BR>
<a href='blah.html'> man blablablah.sh </a><BR>
<a href='blah.html'> man blablablablah.sh </a><BR>
</BODY>
</HTML>

    [download]
    update ... or maybe ... 
    #!/usr/bin/perl -w
use strict;

while(<DATA>) {
    s/<HTML.*?>//gi;
    s/<\/HTML.*?>//gi;
    print;
}
__DATA__
<HTML><HEAD><TITLE>Homepage</TITLE></HEAD>
<BODY>
<a href='blah.html'> man blah.pl</a><BR>
<a href='blah.html'> man blablablah.sh </a><BR>
<a href='blah.html'> man blablablablah.sh </a><BR>
</BODY>
</HTML>

    [download]
[reply]
[d/l]
[select]
      LameNerd,
      Try your code with: 
      __DATA__
asdfasdf asdfasdf
asdfasdf<htMl    >asdfasdf
blah

</htmlasdf>
foo
bar
</html   >asdfasdf asdfasdf

      [download]
      I am not saying that Anonymous Monk should even be attempting to do this as a roll your own solution (go CPAN) - just thought I would point out a weakness or two.

      Cheers - L~R

[reply]
[d/l]
        The output is ... 
        asdfasdf asdfasdf
asdfasdfasdfasdf
blah


foo
bar
asdfasdf asdfasdf

        [download]
        What's wrong with that? It got rid of the html tags?
        I think that is all Anonymous Monk wanted to accomplish.
        That is also why I stated in my original post ...

        Do you want something like this?
[reply]
[d/l]

Back to Seekers of Perl Wisdom