Wrap in <a> tags specific characters in HTML tables with regex & PHP

I have a text... and in it I have paragraphs and tables... I need to replace every X (a single Japanese kanji character to be precise... but it could be any character) which is in one of the tables with <a href="http://example.com/#X">X</a>, but only those X that are in the tables, n开发者_StackOverflow中文版ot outside of them.

There can be several X in a single table so preg_replace('#<td>(X)#','replacewith',$source) wouldn't work as it replaces only one of the X.

Any ideas? Thanks.

$startIndex = strpos($source, '<table');

while ($startIndex !== false) {
    $endIndex = strpos($source, '</table>', $startIndex);

    $excerpt = substr($source, $startIndex, $endIndex - $startIndex);
    $excerpt = preg_replace('/(X|Y|Z)/', '<a href="http://example.com/#$1">$1</a>', $excerpt);
    $source = substr_replace($source, $excerpt, $startIndex, $endIndex - $startIndex);

    if (strlen($source) < $endIndex)
        $startIndex = false;
    else
        $startIndex = strpos($source, '<table', $endIndex);
}

Edit: fixed, tested, works.

parsing html with regex is really not pretty. have a look here how its done and adapt: How to replace text URLs and exclude URLs in HTML tags?