public function more($content, $post)
{
$more_text = 'Read the rest →';
$max_paragraphs = 1;
$showmore = false;
$matches = preg_split('/<!--\\s*more\\s*-->/is', $content, 2, PREG_SPLIT_NO_EMPTY);
if (count($matches) > 1) {
$summary = $matches[0];
$remainder = $matches[1];
if (trim($remainder) != '') {
$showmore = true;
}
} else {
$ht = new HtmlTokenizer($content, false);
$set = $ht->parse();
$stack = array();
$para = 0;
$token = $set->current();
$summary = new HTMLTokenSet(false);
$remainder = new HTMLTokenSet(false);
$set->rewind();
for ($token = $set->current(); $set->valid(); $token = $set->next()) {
if ($token['type'] == HTMLTokenizer::NODE_TYPE_ELEMENT_OPEN) {
$stack[$token['name']] = $token['name'];
}
if ($para < $max_paragraphs) {
$summary[] = $token;
}
if ($para >= $max_paragraphs) {
$remainder[] = $token;
$showmore = true;
}
if ($token['type'] == HTMLTokenizer::NODE_TYPE_ELEMENT_CLOSE) {
if (isset($stack[$token['name']])) {
while (end($stack) != $token['name']) {
array_pop($stack);
}
array_pop($stack);
}
if (count($stack) == 0) {
$para++;
}
}
}
}
if ($post->slug == Controller::get_var('slug')) {
$content = $summary . '<div id="more" class="moreanchor">' . 'Continues here →' . '</div>' . $remainder;
} elseif ($showmore == true) {
$content = $summary . '<p class="more"><a href="' . $post->permalink . '#more">' . $more_text . '</a></p>';
} else {
$content = $summary . $remainder;
}
return $content;
}
/**
* Extract an entity like "<!DOCTYPE html>".
*
* @return array|string
*/
private function extractEntity()
{
$entity = array(0 => '', 1 => [], 2 => '', 'html' => '');
$token = $this->tokenizer->current();
if ($token !== '<!') {
// Sanity check
throw new \Exception('Sanity check failed. Expected a "<!" token');
}
while ($this->tokenizer->valid()) {
$token = $this->tokenizer->current();
$this->tokenizer->next();
$entity['html'] .= is_array($token) ? $token[1] : $token;
if ($token[0] == 'T_DTD_ENTITY' || $token == '<!') {
$entity[0] .= is_array($token) ? $token[1] : $token;
} elseif ($token[0] == 'T_DTD_ATTRIBUTES') {
$entity[1][] = $token[1];
} elseif ($token == '>') {
$entity[2] = $token;
return $entity;
} else {
$this->warnings[] = 'TagIterator: Unexpected token in entity: "' . (is_array($token) ? '[' . $token[0] . '] ' . $token[1] : $token) . '"';
return $entity['html'];
}
}
}
public function attValueEnd($ch)
{
$this->c_quote = $this->att_quote;
$c_val = $this->stack;
$res = parent::attValueEnd($ch);
if ($this->debug) {
fx::debug($res === false, $c_val);
}
if ($res !== false) {
$this->c_att['value'] = $c_val;
$this->addAtt();
$this->stack = '';
}
$this->c_quote = null;
return $res;
}
/**
* Splits a text in smaller parts having a length lower than $maxLength
* Texts would be cut after a period.
*
* @TODO Improve me to work with HTML
* @param string $text Text to split in smaller parts
* @param int $maxLength Maximum length of a text part
* @param boolean $html If true HTML tags will be handled properly
* @return array
*/
protected function _splitText($text, $maxLength, $html = false)
{
if (strlen($text) <= $maxLength) {
$texts = array($text);
} else {
if ($html) {
App::import('Lib', 'I18n.HtmlTokenizer');
$Tokenizer = new HtmlTokenizer($text);
$texts = $Tokenizer->tokens($maxLength);
} else {
$sentences = preg_split("/[\\.][\\s]+/", $text);
$texts = array('');
$i = 0;
foreach ($sentences as $sentence) {
if (empty($sentence)) {
continue;
}
$sentence .= '. ';
if (empty($texts[$i]) && strlen($sentence) >= $maxLength) {
// Cut the string before the latest word
while (strlen($sentence) >= $maxLength) {
$sentencePart = substr($sentence, 0, $maxLength);
$sentencePart = substr($sentencePart, 0, strrpos($sentencePart, ' ') + 1);
$texts[$i++] = $sentencePart;
$sentence = substr($sentence, strlen($sentencePart));
}
$texts[$i++] = $sentence;
$texts[$i] = '';
} elseif (strlen($texts[$i]) + strlen($sentence) < $maxLength) {
$texts[$i] .= $sentence;
} else {
$i++;
$texts[$i] = $sentence;
}
}
// Removes the ". " of the latest text if it was not finished with a period
if (substr(trim($text), -1) !== '.') {
$texts[count($texts) - 1] = substr($texts[count($texts) - 1], 0, -2);
}
}
}
return $texts;
}
/**
* Returns a shortened version of whatever is passed in.
* @param string $value A string to shorten
* @param integer $count Maximum words to display [100]
* @param integer $max_paragraphs Maximum paragraphs to display [1]
* @return string The string, shortened
*/
public static function summarize( $text, $count = 100, $max_paragraphs = 1 )
{
$ellipsis = '…';
$showmore = false;
$ht = new HtmlTokenizer($text, false);
$set = $ht->parse();
$stack = array();
$para = 0;
$token = $set->current();
$summary = new HTMLTokenSet();
$set->rewind();
$remaining_words = $count;
// $bail lets the loop end naturally and close all open elements without adding new ones.
$bail = false;
for ( $token = $set->current(); $set->valid(); $token = $set->next() ) {
if ( !$bail && $token['type'] == HTMLTokenizer::NODE_TYPE_ELEMENT_OPEN ) {
$stack[] = $token;
}
if ( !$bail ) {
switch ( $token['type'] ) {
case HTMLTokenizer::NODE_TYPE_TEXT:
$words = preg_split( '/(\\s+)/u', $token['value'], -1, PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY );
// word count is doubled because spaces between words are captured as their own array elements via PREG_SPLIT_DELIM_CAPTURE
$words = array_slice( $words, 0, $remaining_words * 2 );
$remaining_words -= count( $words ) / 2;
$token['value'] = implode( '', $words );
if ( $remaining_words <= 0 ) {
$token['value'] .= $ellipsis;
$summary[] = $token;
$bail = true;
}
else {
$summary[] = $token;
}
break;
case HTMLTokenizer::NODE_TYPE_ELEMENT_CLOSE;
// don't handle this case here
break;
default:
$summary[] = $token;
break;
}
}
if ( $token['type'] == HTMLTokenizer::NODE_TYPE_ELEMENT_CLOSE ) {
do {
$end = array_pop( $stack );
$end['type'] = HTMLTokenizer::NODE_TYPE_ELEMENT_CLOSE;
$end['attrs'] = null;
$end['value'] = null;
$summary[] = $end;
} while ( ( $bail || $end['name'] != $token['name'] ) && count( $stack ) > 0 );
if ( count( $stack ) == 0 ) {
$para++;
}
if ( $bail || $para >= $max_paragraphs ) {
break;
}
}
}
return (string) $summary;
}
protected function parseHTML($content)
{
$tokenizer = HtmlTokenizer::create(StringInputStream::create($content))->lowercaseTags(true)->lowercaseAttributes(true);
$insideHead = false;
while ($token = $tokenizer->nextToken()) {
if (!$insideHead) {
if ($token instanceof SgmlOpenTag && $token->getId() == 'head') {
$insideHead = true;
continue;
}
}
if ($insideHead) {
if ($token instanceof SgmlEndTag && $token->getId() == 'head') {
break;
}
if ($token instanceof SgmlOpenTag && $token->getId() == 'link' && $token->hasAttribute('rel') && $token->hasAttribute('href')) {
if ($token->getAttribute('rel') == 'openid.server') {
$this->server = HttpUrl::create()->parse($token->getAttribute('href'));
}
if ($token->getAttribute('rel') == 'openid.delegate') {
$this->realId = HttpUrl::create()->parse($token->getAttribute('href'));
}
}
if ($token instanceof SgmlOpenTag && $token->getId() == 'meta' && $token->hasAttribute('content') && $token->hasAttribute('http-equiv') && mb_strtolower($token->getAttribute('http-equiv')) == self::HEADER_XRDS_LOCATION) {
$this->loadXRDS($token->getAttribute('content'));
return $this;
}
}
}
return $this;
}
/**
* Auxiliary function to extract tokens from an child node
*
* @param DomNode $child the child node to extract tokens from
* @param array tokes array of tokes previously extracted
* @param int $length maximum size for each token extracted from the HTML document
* @return array list of tokens
*/
protected function _tokenizeChild($child, $tokens, $length)
{
$attributes = $this->_getNodeAttributes($child);
if ($attributes) {
$attributes = ' ' . $attributes;
}
$tokens[] = "<{$child->tagName}{$attributes}>";
foreach ($child->childNodes as $c) {
$cContent = $this->_getTokenContents($c);
if (!$c->hasChildNodes()) {
if (!trim($cContent)) {
continue;
}
if (isset($child->tagName) && $child->tagName == 'code') {
$cContent = str_replace('??>', '?>', $cContent);
}
$tokens[] = $cContent;
continue;
}
$tokenizer = new HtmlTokenizer($cContent);
$newTokens = $tokenizer->tokens($length, $c);
if (array_sum(array_map('strlen', $newTokens)) <= $length) {
$tokens[] = join('', $newTokens);
} else {
$tokens = array_merge($tokens, $newTokens);
}
}
$tokens[] = "</{$child->tagName}>";
return $tokens;
}
