HTML::Scrubber(3) Perl extension for scrubbing/sanitizing html

VERSION

version 0.11

SYNOPSIS


use HTML::Scrubber;
my $scrubber = HTML::Scrubber->new( allow => [ qw[ p b i u hr br ] ] );
print $scrubber->scrub('<p><b>bold</b> <em>missing</em></p>');
# output is: <p><b>bold</b> </p>
# more complex input
my $html = q[
<style type="text/css"> BAD { background: #666; color: #666;} </style>
<script language="javascript"> alert("Hello, I am EVIL!"); </script>
<HR>
a => <a href=1>link </a>
br => <br>
b => <B> bold </B>
u => <U> UNDERLINE </U>
];
print $scrubber->scrub($html);
$scrubber->deny( qw[ p b i u hr br ] );
print $scrubber->scrub($html);

DESCRIPTION

If you want to ``scrub'' or ``sanitize'' html input in a reliable and flexible fashion, then this module is for you.

I wasn't satisfied with HTML::Sanitizer because it is based on HTML::TreeBuilder, so I thought I'd write something similar that works directly with HTML::Parser.

METHODS

First a note on documentation: just study the EXAMPLE below. It's all the documentation you could need

Also, be sure to read all the comments as well as How does it work?.

If you're new to perl, good luck to you.

comment

    warn "comments are  ", $p->comment ? 'allowed' : 'not allowed';
    $p->comment(0);  # off by default

process

    warn "process instructions are  ", $p->process ? 'allowed' : 'not allowed';
    $p->process(0);  # off by default

script

    warn "script tags (and everything in between) are supressed"
        if $p->script;      # off by default
    $p->script( 0 || 1 );

** Please note that this is implemented using HTML::Parser's ignore_elements function, so if "script" is set to true, all script tags encountered will be validated like all other tags.

style

    warn "style tags (and everything in between) are supressed"
        if $p->style;       # off by default
    $p->style( 0 || 1 );

** Please note that this is implemented using HTML::Parser's ignore_elements function, so if "style" is set to true, all style tags encountered will be validated like all other tags.

allow

    $p->allow(qw[ t a g s ]);

deny

    $p->deny(qw[ t a g s ]);

rules

    $p->rules(
        img => {
            src => qr{^(?!http://)}i, # only relative image links allowed
            alt => 1,                 # alt attribute allowed
            '*' => 0,                 # deny all other attributes
        },
        a => {
            href => sub { ... },      # check or adjust with a callback
        },
        b => 1,
        ...
    );

Updates set of attribute rules. Each rule can be 1/0, regular expression or a callback. Values longer than 1 char are treated as regexps. Callback is called with the following arguments: this object, tag name, attribute name and attribute value, should return empty list to drop attribute, "undef" to keep it without value or a new scalar value.

default

    print "default is ", $p->default();
    $p->default(1);      # allow tags by default
    $p->default(
        undef,           # don't change
        {                # default attribute rules
            '*' => 1,    # allow attributes by default
        }
    );

scrub_file

    $html = $scrubber->scrub_file('foo.html');   ## returns giant string
    die "Eeek $!" unless defined $html;  ## opening foo.html may have failed
    $scrubber->scrub_file('foo.html', 'new.html') or die "Eeek $!";
    $scrubber->scrub_file('foo.html', *STDOUT)
        or die "Eeek $!"
            if fileno STDOUT;

scrub

    print $scrubber->scrub($html);  ## returns giant string
    $scrubber->scrub($html, 'new.html') or die "Eeek $!";
    $scrubber->scrub($html', *STDOUT)
        or die "Eeek $!"
            if fileno STDOUT;

default handler, used by both _scrub and _scrub_fh Moved all the common code (basically all of it) into a single routine for ease of maintenance

default handler, does the scrubbing if we're scrubbing out to a file. Now calls _scrub_str and pushes that out to a file.

default handler, does the scrubbing if we're returning a giant string. Now calls _scrub_str and appends that to the output string.

How does it work?

When a tag is encountered, HTML::Scrubber allows/denies the tag using the explicit rule if one exists.

If no explicit rule exists, Scrubber applies the default rule.

If an explicit rule exists, but it's a simple rule(1), the default attribute rule is applied.

EXAMPLE

    #!/usr/bin/perl -w
    use HTML::Scrubber;
    use strict;
    my @allow = qw[ br hr b a ];
    my @rules = (
        script => 0,
        img => {
            src => qr{^(?!http://)}i, # only relative image links allowed
            alt => 1,                 # alt attribute allowed
            '*' => 0,                 # deny all other attributes
        },
    );
    my @default = (
        0   =>    # default rule, deny all tags
        {
            '*'           => 1, # default rule, allow all attributes
            'href'        => qr{^(?:http|https|ftp)://}i,
            'src'         => qr{^(?:http|https|ftp)://}i,
    #   If your perl doesn't have qr
    #   just use a string with length greater than 1
            'cite'        => '(?i-xsm:^(?:http|https|ftp):)',
            'language'    => 0,
            'name'        => 1, # could be sneaky, but hey ;)
            'onblur'      => 0,
            'onchange'    => 0,
            'onclick'     => 0,
            'ondblclick'  => 0,
            'onerror'     => 0,
            'onfocus'     => 0,
            'onkeydown'   => 0,
            'onkeypress'  => 0,
            'onkeyup'     => 0,
            'onload'      => 0,
            'onmousedown' => 0,
            'onmousemove' => 0,
            'onmouseout'  => 0,
            'onmouseover' => 0,
            'onmouseup'   => 0,
            'onreset'     => 0,
            'onselect'    => 0,
            'onsubmit'    => 0,
            'onunload'    => 0,
            'src'         => 0,
            'type'        => 0,
        }
    );
    my $scrubber = HTML::Scrubber->new();
    $scrubber->allow( @allow );
    $scrubber->rules( @rules ); # key/value pairs
    $scrubber->default( @default );
    $scrubber->comment(1); # 1 allow, 0 deny
    ## preferred way to create the same object
    $scrubber = HTML::Scrubber->new(
        allow   => \@allow,
        rules   => \@rules,
        default => \@default,
        comment => 1,
        process => 0,
    );
    require Data::Dumper,die Data::Dumper::Dumper($scrubber) if @ARGV;
    my $it = q[
        <?php   echo(" EVIL EVIL EVIL "); ?>    <!-- asdf -->
        <hr>
        <I FAKE="attribute" > IN ITALICS WITH FAKE="attribute" </I><br>
        <B> IN BOLD </B><br>
        <A NAME="evil">
            <A HREF="javascript:alert('die die die');">HREF=JAVA &lt;!&gt;</A>
            <br>
            <A HREF="image/bigone.jpg" ONMOUSEOVER="alert('die die die');">
                <IMG SRC="image/smallone.jpg" ALT="ONMOUSEOVER JAVASCRIPT">
            </A>
        </A> <br>
    ];
    print "#original text",$/, $it, $/;
    print
        "#scrubbed text (default ",
        $scrubber->default(), # no arguments returns the current value
        " comment ",
        $scrubber->comment(),
        " process ",
        $scrubber->process(),
        " )",
        $/,
        $scrubber->scrub($it),
        $/;
    $scrubber->default(1); # allow all tags by default
    $scrubber->comment(0); # deny comments
    print
        "#scrubbed text (default ",
        $scrubber->default(),
        " comment ",
        $scrubber->comment(),
        " process ",
        $scrubber->process(),
        " )",
        $/,
        $scrubber->scrub($it),
        $/;
    $scrubber->process(1);        # allow process instructions (dangerous)
    $default[0] = 1;              # allow all tags by default
    $default[1]->{'*'} = 0;       # deny all attributes by default
    $scrubber->default(@default); # set the default again
    print
        "#scrubbed text (default ",
        $scrubber->default(),
        " comment ",
        $scrubber->comment(),
        " process ",
        $scrubber->process(),
        " )",
        $/,
        $scrubber->scrub($it),
        $/;

FUN

If you have Test::Inline (and you've installed HTML::Scrubber), try

    pod2test Scrubber.pm >scrubber.t
    perl scrubber.t

INSTALLATION

See perlmodinstall for information and options on installing Perl modules.

BUGS AND LIMITATIONS

You can make new bug reports, and view existing ones, through the web interface at <http://rt.cpan.org/Public/Dist/Display.html?Name=HTML-Scrubber>.

AVAILABILITY

The project homepage is <https://metacpan.org/release/HTML-Scrubber>.

The latest version of this module is available from the Comprehensive Perl Archive Network (CPAN). Visit <http://www.perl.com/CPAN/> to find a CPAN site near you, or see <https://metacpan.org/module/HTML::Scrubber/>.

AUTHORS

COPYRIGHT AND LICENSE

This software is copyright (c) 2013 by Ruslan Zakirov, Nigel Metheringham, 2003-2004 D. H..

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.