/**
* Filter a single element
*
* @param DOMElement $element
* @return void
*/
public function filterElement(DOMElement $element)
{
// Only decorate non-empty paragraphs
if (trim($element->textContent) && $element->getProperty('type') === false) {
$element->setProperty('type', 'para');
}
}
/**
* Filter a single element
*
* @param DOMElement $element
* @return void
*/
public function filterElement(DOMElement $element)
{
if ($element->hasAttribute('rowspan') && $element->getAttribute('rowspan') > 1) {
$attributes = $element->getProperty('attributes');
$attributes['morerows'] = $element->getAttribute('rowspan') - 1;
$element->setProperty('attributes', $attributes);
}
// @todo: Handle colspan, too - even it is quite complex to express in
// docbook.
}
/**
* Filter a single element.
*
* @param DOMElement $element
* @return void
*/
public function filterElement(DOMElement $element)
{
$element->setProperty('type', 'ulink');
$attrs = $element->getProperty('attributes');
if (!is_array($attrs)) {
$attrs = array();
}
// @todo: Can we convert more attributes here? Maybe <ulink type="…"/>?
$attrs['url'] = $element->getAttributeNS(ezcDocumentOdt::NS_XLINK, 'href');
$element->setProperty('attributes', $attrs);
}
/**
* Filter a single element.
*
* @param DOMElement $element
* @return void
*/
public function filterElement(DOMElement $element)
{
$element->setProperty('type', 'footnote');
$citations = $element->getElementsByTagNameNS(ezcDocumentOdt::NS_ODT_TEXT, 'note-citation');
// Should be only 1, foreach to remove all
foreach ($citations as $cite) {
$attrs = $element->getProperty('attributes');
if ($attrs === false) {
$attrs = array();
}
$attrs['label'] = $cite->nodeValue;
$element->setProperty('attributes', $attrs);
$element->removeChild($cite);
}
}
/**
* Shows a string representation of the current node.
*
* Is only there for debugging purposes
*
* @param DOMElement $element
* @param bool $newLine
* @access private
*/
protected function showCurrentNode(DOMElement $element, $newLine = true)
{
if ($element->parentNode && $element->parentNode instanceof DOMElement) {
$this->showCurrentNode($element->parentNode, false);
}
echo '> ', $element->tagName;
if ($element->getProperty('type') !== false) {
echo ' (', $element->getProperty('type'), ')';
}
echo $newLine ? "\n" : ' ';
}
/**
* Calculate content factors
*
* Try to calculate some kind of probability for each node in the document,
* that the respective node is the root of the actual document content.
*
* @param DOMElement $element
* @return float
*/
protected function calculateContentFactors(DOMElement $element)
{
$textLength = 0;
$childElements = 0;
$childFactors = 0;
$childTypeBonus = 0;
foreach ($element->childNodes as $child) {
switch ($child->nodeType) {
case XML_ELEMENT_NODE:
++$childElements;
$childFactors += $this->calculateContentFactors($child);
if (isset($this->bonus[$child->tagName])) {
$childTypeBonus += $this->bonus[$child->tagName];
}
break;
case XML_TEXT_NODE:
$textLength += strlen(trim($child->wholeText));
break;
}
}
// Use an exponential metric on text amount.
$textFactor = max(1, pow($textLength / 50, 4));
$factor = $textFactor * (($childFactors + $childTypeBonus) / max(1, abs(10 - $childElements)));
if ($factor > $this->maximumImportance && $element->getProperty('type') === 'section') {
$this->maximumImportance = $factor;
$this->mostImportantNode = $element;
}
// $attributes = $element->getProperty( 'attributes' );
// $attributes['factor'] = $factor;
// $element->setProperty( 'attributes', $attributes );
return $factor;
}
/**
* Recursively transform annotated ODT elements to docbook
*
* @param DOMElement $odt
* @param DOMElement $docbook
* @param bool $significantWhitespace
* @return void
*/
protected function transformToDocbook(DOMElement $odt, DOMElement $docbook, $significantWhitespace = false)
{
if (($spaces = $odt->getProperty('spaces')) !== false) {
$docbook->appendChild(new DOMText($spaces));
}
if (($tagName = $odt->getProperty('type')) !== false) {
$node = new DOMElement($tagName);
$docbook->appendChild($node);
$docbook = $node;
if (($attributes = $odt->getProperty('attributes')) !== false) {
foreach ($attributes as $name => $value) {
$node->setAttribute($name, $value);
}
}
}
$numChildren = $odt->childNodes->length;
for ($i = 0; $i < $numChildren; ++$i) {
$child = $odt->childNodes->item($i);
switch ($child->nodeType) {
case XML_ELEMENT_NODE:
$this->transformToDocbook($child, $docbook, $significantWhitespace || $odt->getProperty('whitespace') === 'significant');
break;
case XML_TEXT_NODE:
$docbook->appendChild(new DOMText($child->data));
break;
case XML_CDATA_SECTION_NODE:
$docbook->appendChild($docbook->ownerDocument->createCDATASection($child->data));
break;
// case XML_ENTITY_NODE:
// Seems not required, as entities in the source document
// are automatically transformed back to their text
// targets.
// break;
// case XML_ENTITY_NODE:
// Seems not required, as entities in the source document
// are automatically transformed back to their text
// targets.
// break;
case XML_COMMENT_NODE:
$comment = new DOMElement('comment', $child->data);
$docbook->appendChild($comment);
break;
}
}
}
/**
* Recursively transform annotated XHtml elements to docbook
*
* @param DOMElement $xhtml
* @param DOMElement $docbook
* @param bool $significantWhitespace
* @return void
*/
protected function transformToDocbook(DOMElement $xhtml, DOMElement $docbook, $significantWhitespace = false)
{
if (($tagName = $xhtml->getProperty('type')) !== false) {
$node = new DOMElement($tagName);
$docbook->appendChild($node);
$docbook = $node;
if (($attributes = $xhtml->getProperty('attributes')) !== false) {
foreach ($attributes as $name => $value) {
$node->setAttribute($name, htmlspecialchars($value));
}
}
}
foreach ($xhtml->childNodes as $child) {
switch ($child->nodeType) {
case XML_ELEMENT_NODE:
$this->transformToDocbook($child, $docbook, $significantWhitespace || $xhtml->getProperty('whitespace') === 'significant');
break;
case XML_TEXT_NODE:
// Skip pure whitespace text nodes, except for
// intentionally converted <br> elements.
if (trim($text = $child->data) === '' && !$significantWhitespace && $xhtml->getProperty('whitespace') !== 'significant') {
continue;
}
if ($xhtml->getProperty('whitespace') === 'significant' || $significantWhitespace) {
// Don't normalize inside nodes with significant whitespaces.
$text = new DOMText($text);
$docbook->appendChild($text);
} else {
if ($this->isInlineElement($docbook)) {
$text = new DOMText(preg_replace('(\\s+)', ' ', $text));
$docbook->appendChild($text);
} else {
// Wrap contents into a paragraph, if we are yet
// outside of an inline element.
$text = new DOMText(trim(preg_replace('(\\s+)', ' ', $text)));
$para = $docbook->ownerDocument->createElement('para');
$para->appendChild($text);
$docbook->appendChild($para);
}
}
break;
case XML_CDATA_SECTION_NODE:
// $data = new DOMCharacterData();
// $data->appendData( $child->data );
// $docbook->appendChild( $data );
break;
case XML_ENTITY_NODE:
// Seems not required, as entities in the source document
// are automatically transformed back to their text
// targets.
break;
case XML_COMMENT_NODE:
// Ignore comments
break;
$comment = new DOMElement('comment', $child->data);
$docbook->appendChild($comment);
break;
}
}
}
