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

I'm trying to s/// on the plaintext in a variable containing HTML. If naively I do $html =~ s/PERL/Perl/g, lots of links and HTML elements become broken. I just want the text to be replaced.

I looked on CPAN and found HTML::FormatText, which converts the HTML to plaintext. But it won't convert back to HTML after I do the substitution. Can I use HTML::Parser, HTML::TokeParser, HTML::TokeParser::Simple, or HTML::TreeBuilder to identify the plaintext, manipulate it how I see fit, and reassemble the structure into the original HTML, with plaintext modifications? If so, how?

I looked at Sean M. Burke's article in TPJ, and it mentions the as_text() method, but I'm at wits end how to put the modified text back in the HTML document. Can someone provide an example? I just want to apply a regex to the plain text portion of $html, how can I do this? Thanks in advance, -jc

Replies are listed 'Best First'.
Re: Manipulating plaintext within HTML
by Ovid (Cardinal) on Jun 26, 2003 at 19:37 UTC

    If you use HTML::TokeParser::Simple, just pass a reference to the variable to the constructor.

    use HTML::TokeParser::Simple;
my $parser = HTML::TokeParser::Simple->new(\$html);

my $new_html = '';
while (my $token = $parser->get_token) {
    my $text = $token->as_is;
    $new_html .= $token->is_text ? munge_text($text) : $text;
}
print $new_html;

sub munge_text {
    # put your text munging stuff here
}

    [download]

    Cheers,
    Ovid

    New address of my CGI Course.
    Looking for work. Here's my resume. Will work for food (plus salary).

[reply]
[d/l]
      Thanks, thats perfect. :) Your module is pretty nifty.

      (If anyone is curious, I'm matching arbitrary English words with the regex (?<!&)(a-zA-Z'+). The look-behind assertion skips HTML entities, leaving only true English words.)

      Thanks again,

      -jc

[reply]

Back to Seekers of Perl Wisdom