/**
* Detect valid formatting elements within a paragraph
*
* @param \Candybanana\HtmlToCarbonJson\Element
* @return string
*/
protected function detectFormatting(Element $element)
{
if ($element->hasChildren()) {
$formats = [];
foreach ($element->getChildren() as $node) {
if (!$node->isText() && in_array($node->getTagName(), $this->config['allowedFormattingTags'])) {
$type = $this->config['formattingTags'][$node->getTagName()];
if (empty($node->getValue())) {
continue;
}
list($from, $to) = $this->getTextPosition($element->getValue(), $node->getValue());
$format = ['type' => $type, 'from' => $from, 'to' => $to];
// if <a> tag, add attributes
if ($type == 'a') {
$format['attrs'] = ['href' => $node->getAttribute('href')];
}
$formats[] = $format;
}
}
return !empty($formats) ? $formats : null;
}
}
/**
* Extract the equivalent of 'Sections' in Carbon from the document
*
* @param \Candybanana\HtmlToCarbonJson\Element
* @return boolean
*/
protected function extractSections(Element $element)
{
// recursively iterate until we get to the innermost child
if ($element->hasChildren()) {
foreach ($element->getChildren() as $child) {
// we've found our section, return
if ($this->extractSections($child)) {
return;
}
}
}
// is this a block element with children?
// 30/10/16: hacky fix here for LIs containing children. @todo: Proper fix = custom isBlock() function.
if ($element->getTagName() !== 'li' && $element->isBlock() && $element->hasChildren()) {
// does it have a parent? If so that's our section
if ($parent = $element->getParent()) {
$this->sectionsHtml[] = $parent;
// remove it from the DOM
// @todo: check if this is the root element - if so error out
$parent->remove();
return true;
}
}
// we're done with this node
return false;
}
