如何获取img标签而不用DOMDocument :: saveHTML从字符串中更改它？

Question

I have string like:

...。すっきりと整理され...のです。<br /><br /><img src="/photo/images/2016/03/25/0_3316.jpg" alt="漢字漢字漢字" /><br /><br />「あのと...

And I'm trying to get img tag by DOMDocument::saveHTML:

$topic_conv = mb_convert_encoding(htmlspecialchars_decode($topic['Topic']['body']), 'HTML-ENTITIES', 'UTF-8');
$doc->loadHTML($topic_conv);
$img_tags = $doc->getElementsByTagName('img');
foreach ($img_tags as $tag) {
    $img_alt = $tag->getAttribute('alt');
    if (empty($img_alt)) continue;
    $img = $doc->saveHTML($tag);
    var_dump($img);
    str_replace($img, $img . "<span style='{$img_style_caption}'>" . $img_alt . "</span>", htmlspecialchars_decode($topic['Topic']['body'])));
}

The output is: <img src="/photo/images/2016/03/25/0_3316.jpg" alt="漢字漢字漢字">

It's different from the original string - there is no SPACE and / before the closing bracket, it should have /> instead of just >. And my next line of code str_replace doesn't work, because it doesn't match.

Maybe, someone know how can I get img tag as it is in the original string?

Answer 1

If you still need it, then try like this:

str_replace($img, $img . "<span style='{$img_style_caption}'>" . $img_alt . " /" . "</span>", htmlspecialchars_decode($topic['Topic']['body'])));

PS: code is from my head (so is untested). Let me know if any issues...