Semantic HTML

<?php
$html = $_GET['html'];
$lines = preg_split('/[\n]/', $html, -1, PREG_SPLIT_NO_EMPTY);

foreach ($lines as $key => $value) {
    if (strpos($value, 'div') !== false) {
        //extract div
        preg_match_all('/<div[^>]*>/', $value, $matches);
        if (count($matches[0]) > 0) {
            $div = $matches[0][0];
            //extract id and class attributes
            preg_match_all('/(id\s*=\s*"[^"]+")|(class\s*=\s*"[^"]+")/', $div, $matches2);
            $atr = $matches2[0][0];
            //exact tag name
            preg_match_all('/"[^"]+"/', $atr, $matches3);
            $tag = $matches3[0][0];
            //remove quotes from tag name
            $tag = preg_replace('/"/', '', $tag);
            //replace div with tag name
            $lines[$key] = preg_replace('/div/', $tag, $value);
            //remove id and class
            $lines[$key] = preg_replace('/(id\s*=\s*"[^"]+"\s*)|(class\s*=\s*"[^"]+"\s*)/', '', $lines[$key]);
            //clear excessive spaces at the end
            $lines[$key] = preg_replace('/\s*>/', '>', $lines[$key]);
            //make multiple spaces single
            $lines[$key] = preg_replace('/\s{2,}/', ' ', $lines[$key]);
        } else {
            //extract /div
            preg_match_all('/<\/div>\s*<!--\s*[^--]*-->/', $value, $matches);
            $div = $matches[0][0];
            //extract tag name from comment
            preg_match_all('/(\s*(nav)\s*)|(\s*(main)\s*)|(\s*(header)\s*)|(\s*(section)\s*)|(\s*(article)\s*)|(\s*(aside)\s*)|(\s*(footer)\s*)/', $div, $matches2);

            $tag = trim($matches2[0][0]);
            //replace /div with tag name
            $lines[$key] = preg_replace('/div/', $tag, $value);
            //remove comment tag
            $lines[$key] = preg_replace('/<!--\s*[^-]*-->/', '', $lines[$key]);
        }
    } else {
        continue;
    }
}
$result = implode("\n", $lines);
echo $result;
//var_dump($lines);