infong/php-readability
1
0
Fork 0
mirror of https://github.com/j0k3r/php-readability.git synced 2026-09-27 06:26:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #104 from jtojnar/html-shadowing

Browse Source 
Fix discarding `html[lang]`
This commit is contained in:
Jérémy Benoist
2025-03-04 10:20:28 +01:00
committed by GitHub
parent a18cd0f2a9 efbbc86df9
commit 7413a38ff0
2 changed files with 89 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/Readability.php
+42 -1
View File
@@ -1422,7 +1422,7 @@ class Readability implements LoggerAwareInterface
unset($tidy);
}
$this->html = '<meta charset="utf-8">' . (string) $this->html;
$this->html = self::ensureMetaCharset((string) $this->html);
if ('html5lib' === $this->parser || 'html5' === $this->parser) {
$this->dom = (new HTML5())->loadHTML($this->html);
@@ -1510,4 +1510,45 @@ class Readability implements LoggerAwareInterface
)
&& !$node->hasAttribute('hidden');
}
/**
* Tries to insert `meta[charset]` tag into the proper place in the passed HTML document.
*
* `DOMDocument::loadHTML` will parse HTML documents as ISO-8859-1 if there is no `meta[charset]` tag.
* This means that UTF-8-encoded HTML fragments such as those coming from JSON-LD `articleBody` field would be parsed with incorrect encoding.
* Unfortunately, we cannot just put the tag at the start of the HTML fragment, since that would cause parser to auto-insert a `html` element, losing the attributes of the original `html` tag.
*
* @param string $html UTF-8 encoded document
*/
private static function ensureMetaCharset(string $html): string
{
$charsetTag = '<meta charset="utf-8">';
// Only look at first 1024 bytes since, according to HTML5 specification,
// that’s where <meta> elements declaring a character encoding must be located.
// https://developer.mozilla.org/en-US/docs/Web/HTML/Element/meta#charset
$start = substr($html, 0, 1000);
if (1 === preg_match('/<meta[^>]+charset/i', $start)) {
// <meta> tag is already present, no need for modification.
return $html;
}
if (1 === preg_match('/<head[^>]*>/i', $start)) {
// <head> tag was located, <meta> tags go there.
$html = preg_replace('/<head[^>]*>/i', '$0' . $charsetTag, $html, 1);
return $html;
}
if (1 === preg_match('/<html[^>]*>/i', $start)) {
// <html> tag was located, let’s put it inside and have parser create <head>.
$html = preg_replace('/<html[^>]*>/i', '$0' . $charsetTag, $html, 1);
return $html;
}
// Fallback – just plop the <meta> at the start of the fragment.
return $charsetTag . $html;
}
}