public function translateTemplate($untranslated, $moduleName, $templateName, $language = null)
{
// $translations is expected to be an array
$translations = Locator::getInstance()->getTemplateTranslation($moduleName, $templateName, $language);
if (is_array($translations)) {
$translations = array();
}
return TextProcessor::doTextVariation($untranslated, '[[', ']]', $translations);
}
/**
* Output information about a variable
*
* Context-sensitive replacement for the native var_dump() function.
* If the xdebug extension is installed or if text/plain HTTP content
* headers have been sent then var_dump() is used as is. Otherwise the
* var_dump() call is wrapped with <pre> containter tags.
*
* @param mixed $var
*/
public static function dump($var)
{
// Check if it's actually okay to output anything
if (true === DEBUG) {
$text = new TextProcessor();
// Don't want to shoot our bolt too soon
ob_start();
if (extension_loaded('xdebug')) {
// Bliss, nothing extra for us to do
var_dump($var);
} elseif (headers_sent()) {
$http = headers_list();
// Try to discover what kind of context we're in
foreach ($http as $index => $header) {
list($key, $val) = explode(': ', $header);
if ('Content-type' == $key) {
if ('text/plain' == substr($val, 0, 10)) {
var_dump($var);
} elseif ('text/html' == substr($val, 0, 9)) {
echo $text->wrapWithTag($var, 'pre');
}
break;
}
}
}
// Gather what output we may already have
$info = ob_get_clean();
if (false === empty($info)) {
// We have output already
echo $info;
} else {
if ('text/html' == ini_get('default_mimetype')) {
// If the default mimetype is HTML
echo $text->wrapWithTag($var, 'pre');
} else {
// Tried everything else, resort to simple var_dump
var_dump($var);
}
}
}
}
/**
* Used to extract the title, description and links from
* a string consisting of Word Doc data (2004 or earlier).
*
* @param string $page the web-page contents
* @param string $url the url where the page contents came from,
* used to canonicalize relative links
*
* @return array a summary of the contents of the page
*
*/
function process($page, $url)
{
$text = "";
if (is_string($page)) {
$text = self::extractASCIIText($page);
}
if ($text == "") {
$text = $url;
}
$summary = parent::process($text, $url);
return $summary;
}
/**
* Used to extract the title, description and links from
* a string consisting of rss news feed data.
*
* @param string $page web-page contents
* @param string $url the url where the page contents came from,
* used to canonicalize relative links
*
* @return array a summary of the contents of the page
*
*/
function process($page, $url)
{
$summary = NULL;
if (is_string($page)) {
$dom = self::dom($page);
if ($dom !== false) {
$summary[self::TITLE] = $url;
$summary[self::DESCRIPTION] = "Sitemap of " . $url;
$summary[self::LANG] = "en-US";
$summary[self::LINKS] = self::links($dom, $url);
if (strlen($summary[self::DESCRIPTION] . $summary[self::TITLE]) == 0 && count($summary[self::LINKS]) == 0) {
//maybe not a sitemap? treat as text still try to get urls
$summary = parent::process($page, $url);
}
$summary[self::JUST_METAS] = true;
} else {
$summary = parent::process($page, $url);
$summary[self::JUST_METAS] = true;
}
}
return $summary;
}
/**
* Returns summary of body of a web page based on crude regex matching
* used as a fall back if dom parsing did not work.
*
* @param string $page to extract description from
* @return string a title of the page
*/
static function crudeDescription($page)
{
$body = parent::getBetweenTags($page, 0, "<body", "</body");
$body = preg_replace("/\\</", " <", $body);
$body = strip_tags("<body" . $body[1] . "</body>");
if ($body == "") {
return $body;
}
$body = preg_replace("/\\s+/", " ", $body);
return mb_substr($body, 0, self::$max_description_len);
}
/**
* Parses a mediawiki document to produce an HTML equivalent
*
* @param string $document a document which might have mediawiki markup
* @param bool $parse_head_vars header variables are an extension of
* mediawiki syntax used to add meta variable and titles to
* the head tag of an html document. This flag controls whether to
* supprot this extension or not
* @param bool $handle_big_files for indexing purposes Yioop by default
* truncates long documents before indexing them. If true, this
* method does not do this default truncation. The true value
* is more useful when using Yioop's built-in wiki.
* @return string HTML document obtained by parsing mediawiki
* markup in $document
*/
function parse($document, $parse_head_vars = true, $handle_big_files = false)
{
$head = "";
$page_type = "standard";
$head_vars = array();
$draw_toc = true;
if ($parse_head_vars && !$this->minimal) {
$document_parts = explode("END_HEAD_VARS", $document);
if (count($document_parts) > 1) {
$head = $document_parts[0];
$document = $document_parts[1];
$head_lines = preg_split("/\n\\s*\n/", $head);
foreach ($head_lines as $line) {
$semi_pos = strpos($line, ";") ? strpos($line, ";") : strlen($line);
$line = substr($line, 0, $semi_pos);
$line_parts = explode("=", $line);
if (count($line_parts) == 2) {
$head_vars[trim(addslashes($line_parts[0]))] = addslashes(trim($line_parts[1]));
}
}
if (isset($head_vars['page_type'])) {
$page_type = $head_vars['page_type'];
}
if (isset($head_vars['toc'])) {
$draw_toc = $head_vars['toc'];
}
}
}
$document = preg_replace_callback("/<nowiki>(.+?)<\\/nowiki>/s", "base64EncodeCallback", $document);
$document = preg_replace_callback("/<pre>(.+?)<\\/pre>/s", "spaceEncodeCallback", $document);
if (!$this->minimal) {
if ($draw_toc && $page_type != "presentation") {
$toc = $this->makeTableOfContents($document);
}
list($document, $references) = $this->makeReferences($document);
}
$document = preg_replace_callback('/(\\A|\\n){\\|(.*?)\\n\\|}/s', "makeTableCallback", $document);
if ($page_type == "presentation") {
$lines = explode("\n", $document);
$out_document = "";
$slide = "";
$div = "<div class='slide'>";
foreach ($lines as $line) {
if (trim($line) == "....") {
$slide = $this->processRegexes($slide);
$out_document .= $div . $this->cleanLinksAndParagraphs($slide) . "</div>";
$slide = "";
} else {
$slide .= $line . "\n";
}
}
$document = $out_document . $div . $this->processRegexes($slide) . "</div>";
} else {
if ($handle_big_files) {
$document = $this->processRegexes($document);
$document = $this->cleanLinksAndParagraphs($document);
} else {
if (strlen($document) > 0.9 * MAX_GROUP_PAGE_LEN) {
$document = substr($document, 0, 0.9 * MAX_GROUP_PAGE_LEN);
}
$document = $this->processRegexes($document);
$document = $this->cleanLinksAndParagraphs($document);
}
}
if (!$this->minimal) {
$document = $this->insertReferences($document, $references);
if ($draw_toc && $page_type != "presentation") {
$document = $this->insertTableOfContents($document, $toc);
}
}
$document = preg_replace_callback("/<nowiki>(.+?)<\\/nowiki>/s", "base64DecodeCallback", $document);
if ($head != "" && $parse_head_vars) {
$document = $head . "END_HEAD_VARS" . $document;
}
if (!$handle_big_files && strlen($document) > 0.9 * MAX_GROUP_PAGE_LEN) {
$document = substr($document, 0, 0.9 * MAX_GROUP_PAGE_LEN);
TextProcessor::closeDanglingTags($document);
$document .= "...";
}
return $document;
}
/**
* Used to extract the title, description and links from
* a string consisting of svg image. If the image is small
* enough, an attempt is made to generate a thumbnail
*
* @param string $page web-page contents
* @param string $url the url where the page contents came from,
* used to canonicalize relative links
*
* @return array a summary of the contents of the page
*
*/
function process($page, $url)
{
if (is_string($page)) {
self::closeDanglingTags($page);
$dom = self::dom($page);
if ($dom !== false && isset($dom->documentElement)) {
$summary[self::TITLE] = "";
$summary[self::DESCRIPTION] = self::description($dom);
$summary[self::LINKS] = array();
$summary[self::PAGE] = "<html><body><div><img src='data:image/svg+xml;base64," . base64_encode($page) . "' alt='" . $summary[self::DESCRIPTION] . "' /></div></body></html>";
if (strlen($page) < self::MAX_THUMB_LEN) {
$thumb_string = self::createThumb($dom);
$summary[self::THUMB] = 'data:image/svg+xml;base64,' . base64_encode($thumb_string);
}
} else {
$summary = parent::process($page, $url);
}
}
return $summary;
}
/**
* Used to extract the title, description and links from
* a string consisting of rss news feed data.
*
* @param string $page web-page contents
* @param string $url the url where the page contents came from,
* used to canonicalize relative links
*
* @return array a summary of the contents of the page
*
*/
function process($page, $url)
{
$summary = NULL;
if (is_string($page)) {
self::closeDanglingTags($page);
$dom = self::dom($page);
$root_name = isset($dom->documentElement->nodeName) ? $dom->documentElement->nodeName : "";
unset($dom);
$XML_PROCESSORS = array("rss" => "RssProcessor", "html" => "HtmlProcessor", "sitemapindex" => "SitemapProcessor", "urlset" => "SitemapProcessor", "svg" => "SvgProcessor");
if (isset($XML_PROCESSORS[$root_name])) {
$processor_name = $XML_PROCESSORS[$root_name];
$processor = new $processor_name($this->plugin_instances);
$summary = $processor->process($page, $url);
} else {
$summary = parent::process($page, $url);
}
}
return $summary;
}
/**
* Computes a summary based on a string of a binary Powerpoint document
* (as opposed to the modern xml powerpoint format).
*
* Text is extracted from the Powerpoint document using a crude finite
* state machine that was developed by looking at a few Powerpoint
* documents in a Hex editor. Then the TextProcessor:: process() method
* is used to make a summary
*
* @param string $page string of a Powerpoint document
* @param string $url location the document came from, not used by
* TextProcessor at this point. Some of its subclasses override
* this method and use url to produce complete links for
* relative links within a document
*
* @return array a summary of (title, description,links, and content) of
* the information in $page
*/
function process($page, $url)
{
$text = "";
if (is_string($page)) {
$text_objects = array();
$cur_id = 0;
$state = self::PPT_IGNORING;
$cur_char_pos = 0;
$len = strlen($page);
while ($cur_char_pos < $len) {
$ascii = ord($page[$cur_char_pos]);
switch ($state) {
case self::PPT_IGNORING:
if ($ascii == 0) {
$state = self::ZEROONE_IGNORING;
}
break;
case self::ZEROONE_IGNORING:
if ($ascii == 0) {
$state = self::ZEROTWO_IGNORING;
} else {
$state = self::PPT_IGNORING;
}
break;
case self::ZEROTWO_IGNORING:
if ($ascii == 168) {
$state = self::FIRST_CHAR_TEXT_SEG;
} else {
if ($ascii != 0) {
$state = self::PPT_IGNORING;
}
}
break;
case self::FIRST_CHAR_TEXT_SEG:
if ($ascii == 15) {
$state = self::READ_LEN_TEXT_SEG;
$text_len = 0;
$text_len_pos = 0;
} else {
$state = self::PPT_IGNORING;
}
break;
case self::READ_LEN_TEXT_SEG:
if ($text_len_pos < 4) {
$text_len += $ascii << $text_len_pos * 8;
$text_len_pos++;
} else {
$state = self::SCAN_TEXT_SEG;
$scan_text_pos = 0;
$out_text = chr($ascii);
}
break;
case self::SCAN_TEXT_SEG:
if (strpos($out_text, "lick to edit Master title style") > 0) {
$state = self::ALWAYS_IGNORE;
} else {
if ($scan_text_pos < $text_len) {
if ($ascii >= 32 && $ascii <= 126 || $ascii == 10) {
$out_text .= chr($ascii);
$scan_text_pos++;
}
} else {
$text_objects[$cur_id] = $out_text;
$cur_id++;
$state = self::PPT_IGNORING;
}
}
break;
case self::ALWAYS_IGNORE:
break;
}
$cur_char_pos++;
}
$text = implode("\n", $text_objects);
}
if ($text == "") {
$text = $url;
}
$summary = parent::process($text, $url);
return $summary;
}
/**
* Used to extract the title, description and links from
* a string consisting of rss or atom news feed data.
*
* @param string $page web-page contents
* @param string $url the url where the page contents came from,
* used to canonicalize relative links
*
* @return array a summary of the contents of the page
*
*/
function process($page, $url)
{
$summary = NULL;
if (is_string($page)) {
$dom = self::dom($page);
$atom = false;
$feed_nodes = $dom->getElementsByTagName('feed');
if ($feed_nodes->length > 0) {
$atom = true;
}
if ($dom !== false) {
$summary[self::TITLE] = self::title($dom, $atom);
$summary[self::DESCRIPTION] = self::description($dom, $atom);
$summary[self::LANG] = self::lang($dom, $summary[self::DESCRIPTION]);
$summary[self::LINKS] = self::links($dom, $url, $atom);
if (strlen($summary[self::DESCRIPTION] . $summary[self::TITLE]) == 0 && count($summary[self::LINKS]) == 0) {
//maybe not rss or atom? treat as text still try to get urls
$summary = parent::process($page, $url);
}
}
}
return $summary;
}
