diff options
Diffstat (limited to 'src/HTMLParser.php')
| -rw-r--r-- | src/HTMLParser.php | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/src/HTMLParser.php b/src/HTMLParser.php index 5589e83..f35a950 100644 --- a/src/HTMLParser.php +++ b/src/HTMLParser.php @@ -37,6 +37,8 @@ class HTMLParser 'prevLink' => '/(prev|earl|old|new|<|«)/i', 'whitespace' => '/^\s*$/', 'hasContent' => '/\S$/', + // \x{00A0} is the unicode version of + 'onlyWhitespace' => '/\x{00A0}|\s+/u' ]; private $defaultTagsToScore = [ @@ -1261,7 +1263,7 @@ class HTMLParser $iframeCount = $paragraph->getElementsByTagName('iframe')->length; $totalCount = $imgCount + $embedCount + $objectCount + $iframeCount; - if ($totalCount === 0 && !trim($paragraph->textContent)) { + if ($totalCount === 0 && !preg_replace($this->regexps['onlyWhitespace'], '', $paragraph->textContent)) { // TODO must be done via readability $paragraph->parentNode->removeChild($paragraph); } |