CODEWITHSUNDEEP
phphtmldom
satin
satin

Reputation: 707

Fixing unclosed HTML tags

I am working on some blog layout and I need to create an abstract of each post (say 15 of the lastest) to show on the homepage. Now the content I use is already formatted in html tags by the textile library. Now if I use substr to get 1st 500 chars of the post, the main problem that I face is how to close the unclosed tags.

e.g

<div>.......................</div>
<div>...........
     <p>............</p>
     <p>...........| 500 chars
     </p>
<div>

What I get is two unclosed tags <p> and <div> , p wont create much trouble , but div just messes with the whole page layout. So any suggestion how to track the opening tags and close them manually or something?

Upvotes: 15

Views: 27237

Answers (4)

Richard
Richard

Reputation: 2100

I found a solution which uses DOMDocument but does not add extra tags to your strings; just fixes malformed HTML. See answer here: https://stackoverflow.com/a/79081559/492132

Original github (not mine) here: https://gist.github.com/hubgit/1322324

Upvotes: 0

Kevin Newman
Kevin Newman

Reputation: 928

You can use DOMDocument to do it, but be careful of string encoding issues. Also, you'll have to use a complete HTML document, then extract the components you want. Here's an example:

function make_excerpt ($rawHtml, $length = 500) {
  // append an ellipsis and "More" link
  $content = substr($rawHtml, 0, $length)
    . '&hellip; <a href="/link-to-somewhere">More &gt;</a>';

  // Detect the string encoding
  $encoding = mb_detect_encoding($content);

  // pass it to the DOMDocument constructor
  $doc = new DOMDocument('', $encoding);

  // Must include the content-type/charset meta tag with $encoding
  // Bad HTML will trigger warnings, suppress those
  @$doc->loadHTML('<html><head>'
    . '<meta http-equiv="content-type" content="text/html; charset='
    . $encoding . '"></head><body>' . trim($content) . '</body></html>');

  // extract the components we want
  $nodes = $doc->getElementsByTagName('body')->item(0)->childNodes;
  $html = '';
  $len = $nodes->length;
  for ($i = 0; $i < $len; $i++) {
    $html .= $doc->saveHTML($nodes->item($i));
  }
  return $html;
}

$html = "<p>.......................</p>
  <p>...........
    <p>............</p>
    <p>...........| 500 chars";

// output fixed html
echo make_excerpt($html, 500);

Outputs:

<p>.......................</p>
  <p>...........
    </p>
<p>............</p>
    <p>...........| 500 chars… <a href="/link-to-somewhere">More &gt;</a></p>

If you are using WordPress you should wrap the substr() invocation in a call to wpautop - wpautop(substr(...)). You may also wish to test the length of the $rawHtml passed to the function, and skip appending the "More" link if it isn't long enough.

Upvotes: 3

Jerry
Jerry

Reputation: 1238

As ajreal said, DOMDocument is a solution.

Example :

$str = "
<html>
 <head>
  <title>test</title>
 </head>
 <body>
  <p>error</i>
 </body>
</html>
";

$doc = new DOMDocument();
@$doc->loadHTML($str);
echo $doc->saveHTML();

Advantage : natively included in PHP, contrary to PHP Tidy.

Upvotes: 19

ajreal
ajreal

Reputation: 47321

There are lots of methods that can be used:

  1. Use a proper HTML parser, like DOMDocument
  2. Use PHP Tidy to repair the un-closed tag
  3. Some would suggest HTML Purifier

Upvotes: 18

Related Questions