agriget/vendor/andreskrey/Readability/Nodes/NodeUtility.php
2019-03-22 10:17:29 -04:00

161 lines
5.5 KiB
PHP

<?php
namespace andreskrey\Readability\Nodes;
use andreskrey\Readability\Nodes\DOM\DOMDocument;
use andreskrey\Readability\Nodes\DOM\DOMElement;
use andreskrey\Readability\Nodes\DOM\DOMNode;
/**
* Class NodeUtility.
*/
class NodeUtility
{
/**
* Collection of regexps to check the node usability.
*
* @var array
*/
public static $regexps = [
'unlikelyCandidates' => '/-ad-|banner|breadcrumbs|combx|comment|community|cover-wrap|disqus|extra|foot|header|legends|menu|related|remark|replies|rss|shoutbox|sidebar|skyscraper|social|sponsor|supplemental|ad-break|agegate|pagination|pager|popup|yom-remote/i',
'okMaybeItsACandidate' => '/and|article|body|column|main|shadow/i',
'extraneous' => '/print|archive|comment|discuss|e[\-]?mail|share|reply|all|login|sign|single|utility/i',
'byline' => '/byline|author|dateline|writtenby|p-author/i',
'replaceFonts' => '/<(\/?)font[^>]*>/gi',
'normalize' => '/\s{2,}/',
'videos' => '/\/\/(www\.)?((dailymotion|youtube|youtube-nocookie|player\.vimeo|v\.qq)\.com|(archive|upload\.wikimedia)\.org|player\.twitch\.tv)/i',
'nextLink' => '/(next|weiter|continue|>([^\|]|$)|»([^\|]|$))/i',
'prevLink' => '/(prev|earl|old|new|<|«)/i',
'whitespace' => '/^\s*$/',
'hasContent' => '/\S$/',
'positive' => '/article|body|content|entry|hentry|h-entry|main|page|pagination|post|text|blog|story/i',
'negative' => '/hidden|^hid$| hid$| hid |^hid |banner|combx|comment|com-|contact|foot|footer|footnote|masthead|media|meta|outbrain|promo|related|scroll|share|shoutbox|sidebar|skyscraper|sponsor|shopping|tags|tool|widget/i',
// \x{00A0} is the unicode version of &nbsp;
'onlyWhitespace' => '/\x{00A0}|\s+/u'
];
/**
* Imported from the Element class on league\html-to-markdown.
*
* @param $node
*
* @return DOMElement
*/
public static function nextElement($node)
{
$next = $node;
while ($next
&& $next->nodeType !== XML_ELEMENT_NODE
&& $next->isWhitespace()) {
$next = $next->nextSibling;
}
return $next;
}
/**
* Changes the node tag name. Since tagName on DOMElement is a read only value, this must be done creating a new
* element with the new tag name and importing it to the main DOMDocument.
*
* @param DOMNode $node
* @param string $value
* @param bool $importAttributes
*
* @return DOMNode
*/
public static function setNodeTag($node, $value, $importAttributes = true)
{
$new = new DOMDocument('1.0', 'utf-8');
$new->appendChild($new->createElement($value));
$children = $node->childNodes;
/** @var $children \DOMNodeList $i */
for ($i = 0; $i < $children->length; $i++) {
$import = $new->importNode($children->item($i), true);
$new->firstChild->appendChild($import);
}
if ($importAttributes) {
// Import attributes from the original node.
foreach ($node->attributes as $attribute) {
$new->firstChild->setAttribute($attribute->nodeName, $attribute->nodeValue);
}
}
// The import must be done on the firstChild of $new, since $new is a DOMDocument and not a DOMElement.
$import = $node->ownerDocument->importNode($new->firstChild, true);
$node->parentNode->replaceChild($import, $node);
return $import;
}
/**
* Removes the current node and returns the next node to be parsed (child, sibling or parent).
*
* @param DOMNode $node
*
* @return DOMNode
*/
public static function removeAndGetNext($node)
{
$nextNode = self::getNextNode($node, true);
$node->parentNode->removeChild($node);
return $nextNode;
}
/**
* Remove the selected node.
*
* @param $node DOMElement
*
* @return void
**/
public static function removeNode($node)
{
$parent = $node->parentNode;
if ($parent) {
$parent->removeChild($node);
}
}
/**
* Returns the next node. First checks for children (if the flag allows it), then for siblings, and finally
* for parents.
*
* @param DOMNode $originalNode
* @param bool $ignoreSelfAndKids
*
* @return DOMNode
*/
public static function getNextNode($originalNode, $ignoreSelfAndKids = false)
{
/*
* Traverse the DOM from node to node, starting at the node passed in.
* Pass true for the second parameter to indicate this node itself
* (and its kids) are going away, and we want the next node over.
*
* Calling this in a loop will traverse the DOM depth-first.
*/
// First check for kids if those aren't being ignored
if (!$ignoreSelfAndKids && $originalNode->firstChild) {
return $originalNode->firstChild;
}
// Then for siblings...
if ($originalNode->nextSibling) {
return $originalNode->nextSibling;
}
// And finally, move up the parent chain *and* find a sibling
// (because this is depth-first traversal, we will have already
// seen the parent nodes themselves).
do {
$originalNode = $originalNode->parentNode;
} while ($originalNode && !$originalNode->nextSibling);
return ($originalNode) ? $originalNode->nextSibling : $originalNode;
}
}