regex to select between and while ignoring all text inside any <>

Question

I have the following two types of text:

Type one:

<div class="meta-name">Corporate Officers</div>
<div class="meta-data"><table border="0" cellspacing="0" cellpadding="0" width="171">
<col width="171"></col>
<tbody>
<tr height="19">
<td width="171" height="19">Officer One</td>
</tr>
</tbody>
</table> 
</div>
</div>

Type two:

<div class="meta-name">Corporate Officers</div>
<div class="meta-data">Officer Two</div>
</div>
<pre>

I'm using php with preg_match_all. I need a single expression that will return Officer One and Officer Two from the above. I'm using Corporate Officers< /div> as the first anchor and< /div> as the second, but I can't find Keith Dennis inside all that table gibberish.

How do I return text between anchor1 and anchor2 while ignoring all text inside any brackets <> between?

I saw these threads but wasn't able to make their solutions work for me: RegEx: extract everything until X where X is not between two braces

everything, but everything between [ and ]

Answer 1

With SimpleXML:

$xml = new SimpleXMLElement('<div>
    <div class="meta-name">
        Corporate Officers
    </div>
    <div class="meta-data">
        <table border="0" cellspacing="0" cellpadding="0" width="171">
            <col width="171" />
            <tbody>
                <tr height="19">
                    <td width="171" height="19">
                        Officer One
                    </td>
                </tr>
            </tbody>
        </table>
    </div>
</div>
');

$results = array();
foreach($xml->children() as $node) {
    if($node->getName() == 'div') {
        $attributes = $node->attributes();
        $classes = explode(' ', $attributes['class']);
        if(in_array('meta-name', $classes) || in_array('meta-data', $classes)) {
            $results[] = getText($node);
        }
    }
}

function getText($node) {
    $text = trim(sprintf('%s', $node));
    if(strlen($text) !== 0) {
        return $text;
    }

    foreach($node->children() as $child) {
        if($text = getText($child)) {
            return $text;
        }
    }

    return null;
}

var_dump($results);

As a general rule of thumb, never use Regex to parse HTML.

Answer 2

About 80% of regex questions is about xml/html/xhtml. And about 75% of the answer is to not use a regex. Why? Because while it may seem to work for your example it will be fragile and may break with a slight change of the input.

Please take a look at this beautiful tool. If you can't use it then come back and we will provide with help.

Answer 3

Try this regex:

'~<div\b[^>]*>Corporate\s+Officers</div>\s*<div\b[^>]*>(?:<(?!/?div\b)[^>]*>|\s+)*\K[^<]+~'

This is based on the assumption that there's no other text content in the HTML between the opening <div> tags and the names you're looking for. The first part is self-explanatory:

<div\b[^>]*>Corporate\s+Officers</div>\s*<div\b[^>]*>

I'm assuming the "Corporate Officers" text is sufficient to locate the starting point, but you can reinsert the class attributes if necessary. After that,

(?:<(?!/?div\b)[^>]*>|\s+)*

...consumes any number of tags other than <div> or </div> tags, along with any intervening whitespace. Then \K comes along and says forget all that, the real match starts here. [^<]+ consumes everything up to the beginning of the next tag, and that's all you see in the match results. It's as if everything before the \K was really a positive lookbehind, but without all the restrictions.

Here's a demo.