Extracting the body HTML and clean comments using PHP and Regex

Question

I want to clean the comments and some other garbage or tags from the <body> section in HTML using PHP and regex but my code not work:

$str=preg_replace_callback('/<body>(.*?)<\/body>/s', 
    function($matches){
        return '<body>'.preg_replace(array(
            '/<!--(.|\s)*?-->/',
        ),
        array(
            '',
        ), $matches[1]).'</body>';
    }, $str);

The problem is that nothing happens. Comments will remain where they are or any cleaning to do, nothing happens. Can you help? Thanks!

EDIT:

Thanks to @mhall I figureout that my regex not work becouse of attributes in <body> tag. I use his code and update this:

$str = preg_replace_callback('/(?=<body(.*?)>)(.*?)(?<=<\/body>)/s',
    function($matches) {
        return preg_replace('/<!--.*?-->/s', '', $matches[2]);
    }, $str);

This work PERFECT!

Thanks people!

Why do you want to clean out the comments? You could use `DOMDocument` or another document parser to do this more easily. — Explosion Pills, May 01 '15 at 15:25
Without talking about how not pertinent it is to use regex, I think your problem comes from the `(.*?)<\/body>` part. By default, the "." doesn't include line breaks. You might want to replace it by `[\s\S]` ## EDIT: Nevermind, didn't see the "s" flag. — Tiller, May 01 '15 at 15:47
@ExplosionPills I want a simple way to clean up some things from HTML. — Ivijan Stefan Stipić, May 01 '15 at 15:59
Works for me (PHP 5.5.14), but it drops the ``/`` tags as well. What string are you trying with? — mhall, May 01 '15 at 16:01
@nomistic I know that explanation. There are things that can be used in regex. — Ivijan Stefan Stipić, May 01 '15 at 16:01
I agree with @ExplosionPills here... RegEx is not good for parsing dom. Use a dom parser. — Edward J Beckett, May 01 '15 at 16:12
It is what it is... if you need the library use it. You wouldn't hammer a nail with a screwdriver... ;) — Edward J Beckett, May 01 '15 at 16:17
You could also fix the attribute problem by just writing ``... — alexis, May 01 '15 at 16:25

score 2 · Accepted Answer · answered May 01 '15 at 16:07

Try this. Made a modification on the preg_replace_callback not to include the body tags and replaced (.|\s) with a .* in preg_replace. Also dropped the array syntax from that and added a /s modifier:

$str = <<<EOS
<html>
    <body>
        <p>
             Here is some <!-- One comment --> text
             with a few <!--
                Another comment
             -->
             Comments in it
        </p>
    </body>
</html>
EOS;

$str = preg_replace_callback('/(?=<body>)(.*?)(?<=<\/body>)/s',
    function($matches) {
        return preg_replace('/<!--.*?-->/s', '', $matches[1]);
    }, $str);

echo $str, PHP_EOL;

Output:

<html>
    <body>
        <p>
             Here is some  text
             with a few 
             Comments in it
        </p>
    </body>
</html>

Does your body tag have any class declarations or such, or is it just a plain ``? — mhall, May 01 '15 at 16:14

alexis · Answer 2 · 2015-05-01T16:36:17.643

0

Aren't you making it too complicated? You don't need to jump in and out via a callback, since preg_replace will make replacements at every match:

$parts = explode("<body", $str, 2);
$clean = preg_replace('/<!--.*?-->/s', '', $parts[1]);
$str = $parts[0]."<body".$clean;

Splitting the string into head and body excludes the head from substitution without a lot of messy regexps. Note the s after the pattern: '/.../s'. This makes the dot in the regexp match embedded newlines along with other characters.

edited May 01 '15 at 16:36

answered May 01 '15 at 16:20

alexis

48,685
16
101
161

No becouse in tag I need to keep some comments for browser switcher. – Ivijan Stefan Stipić May 01 '15 at 16:23
Oh, I see. But still it would be much cleaner to split the string in two with `$parts = explode(" – alexis May 01 '15 at 16:31

Extracting the body HTML and clean comments using PHP and Regex

2 Answers2