/**
* process one entity
*
* This function may process either the global message, or any embedded entity.
*
* According to RFC 2046, the processing depends on content type.
*
* The five discrete top-level media types are:
*
* [*] text -- textual information. The subtype "plain" in
* particular indicates plain text containing no
* formatting commands or directives of any sort. Plain
* text is intended to be displayed "as-is".
* YACS attemps to create a page out of provided text.
*
* [*] image -- image data. "Image" requires a display device
* (such as a graphical display, a graphics printer, or a
* FAX machine) to view the information.
* YACS stores the image entity in the file system, and
* creates an article to reference it.
*
* [*] audio -- audio data. "Audio" requires an audio output
* device (such as a speaker or a telephone) to "display"
* the contents.
* YACS stores the audio entity in the file system, and
* creates an article to reference it.
*
* [*] video -- video data. "Video" requires the capability
* to display moving images, typically including
* specialized hardware and software.
* YACS stores the video entity in the file system, and
* creates an article to reference it.
*
* [*] application -- some other kind of data, typically
* either uninterpreted binary data or information to be
* processed by an application.
* YACS stores the entity in the file system, and
* creates an article to reference it.
*
* [*] multipart -- data consisting of multiple entities of
* independent data types.
* YACS separate entities, attempts to locate an article,
* and attach all other entities to it.
*
* [*] message -- an encapsulated message. A body of media
* type "message" is itself all or a portion of some kind
* of message object.
* YACS processes only sub types 'message/rfc822' and 'message/delivery-status', as plain text.
*
* @link http://www.faqs.org/rfcs/rfc2046.html MIME Media Types
*
* @param string the entity to be processed
* @param array poster attributes
* @param string the main anchor to use
* @param string reference of the object to be extended, if any
* @return string reference to the created object, or NULL
*
* @see files/files.php
* @see images/images.php
*/
public static function process_entity($entity, $user, $anchor, $target = NULL)
{
global $context;
// sanity check
if (!trim($entity)) {
return NULL;
}
// split headers and body
if (!($position = strpos($entity, "\n\n"))) {
Logger::remember('agents/messages.php: Can not split header and body', $entity);
return NULL;
}
// parse and decode all headers
$entity_headers = Messages::parse_headers(substr($entity, 0, $position));
if (!$entity_headers) {
$entity_headers[] = array('name' => 'Content-Type', 'value' => 'text/plain; charset=us-ascii');
}
// determine content encoding
$content_transfer_encoding = '7bit';
foreach ($entity_headers as $header) {
if (preg_match('/Content-Transfer-Encoding/i', $header['name'])) {
$content_transfer_encoding = $header['value'];
break;
}
}
// decode the entity
$content = Messages::decode(substr($entity, $position + 2), $content_transfer_encoding);
// determine content type
$content_type = 'text/plain';
foreach ($entity_headers as $header) {
if (preg_match('/Content-Type/i', $header['name'])) {
$content_type = $header['value'];
break;
}
}
// nothing has been created yet
$reference = NULL;
// text type -- create a page or a comment out of it
if (preg_match('/^text\\//i', $content_type)) {
// transcode content, if necessary
if (preg_match('/-(8859|1252)/', $content_type)) {
$content = utf8::from_iso8859($content);
}
$reference = Messages::submit_page($entity_headers, $content, $user, $anchor, $target);
// image type -- save as an image file and create a page to reference it
} elseif (preg_match('/^image\\//i', $content_type)) {
$reference = Messages::submit_image($entity_headers, $content, $user, $anchor, $target);
// audio, video or application types -- create a page and attache the file to it
} elseif (preg_match('/^(audio|video|application)\\//i', $content_type)) {
$reference = Messages::submit_file($entity_headers, $content, $user, $anchor, $target);
// multipart -- split entities and process each of them
} elseif (preg_match('/^multipart\\//i', $content_type)) {
// look for the boundary
if (!preg_match('/boundary="*([a-zA-Z0-9\'\\(\\)\\+_,-\\.\\/:=\\?]+)"*\\s*/i', $content_type, $matches)) {
Logger::remember('agents/messages.php: No multipart boundary in ' . $content_type, $content_type);
return NULL;
}
// split entities
$entities = explode('--' . $matches[1], $content);
if (@count($entities) < 3) {
Logger::remember('agents/messages.php: Empty multipart message', $content);
return NULL;
}
// if multipart/alternative, select only one entity
if (preg_match('/^multipart\\/alternative/i', $content_type)) {
// ignore preamble and epilogue parts; start from the end
for ($index = count($entities) - 1; $index > 1; $index--) {
// stop when we actually create some content
if ($reference = Messages::process_entity($entities[count($entities) - 2], $user, $anchor, $target)) {
break;
}
}
// we have a mix of entities to save
} else {
// skip preamble and epilogue
for ($index = 1; $index < count($entities) - 1; $index++) {
// process one entity
$reference = Messages::process_entity($entities[$index], $user, $anchor, $target);
// post everything at the same place
if (!$anchor) {
$anchor = $reference;
}
// combine all elements at one single place
if (!$target) {
$target = $reference;
}
}
}
// message/delivery-status -- create a page or a comment out of it
} elseif (preg_match('/^message\\/delivery-status/i', $content_type)) {
$reference = Messages::submit_page($entity_headers, nl2br($content), $user, $anchor, $target);
// message/rfc822 -- process it as the main entity
} elseif (preg_match('/^message\\/rfc822/i', $content_type)) {
$reference = Messages::process_entity($entity_headers, $content, $user, $anchor, $target);
// unknown type
} else {
Logger::remember('agents/messages.php: Do not know how to process type ' . $content_type);
return NULL;
}
// job done
return $reference;
}
function parse_cdata($parser, $text)
{
// transcode non-UTF-8 data
if ($this->encoding != 'UTF-8') {
$text = utf8::from_iso8859($text);
}
// skip entry, feed, entries first time we see them
if ($this->elements_stack[0] == $this->current_field or !$this->current_field) {
return;
}
// we are describing a feed
if ($this->elements_stack[0] == 'feed') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->feed[$this->current_name_space][$this->current_field])) {
$this->feed[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->feed[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->feed[$this->current_field])) {
$this->feed[$this->current_field] .= $text;
} else {
$this->feed[$this->current_field] = $text;
}
}
// we are describing one information entry
} elseif ($this->elements_stack[0] == 'entry') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->current_entry[$this->current_name_space][$this->current_field])) {
$this->current_entry[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->current_entry[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->current_entry[$this->current_field])) {
$this->current_entry[$this->current_field] .= $text;
} else {
$this->current_entry[$this->current_field] = $text;
}
}
// echo '.'.$this->current_entry[$this->current_field].BR;
// we are describing a textinput
} elseif ($this->elements_stack[0] == 'textinput') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->textinput[$this->current_name_space][$this->current_field])) {
$this->textinput[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->textinput[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->textinput[$this->current_field])) {
$this->textinput[$this->current_field] .= $text;
} else {
$this->textinput[$this->current_field] = $text;
}
}
// we are describing an image
} elseif ($this->elements_stack[0] == 'image') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->image[$this->current_name_space][$this->current_field])) {
$this->image[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->image[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->image[$this->current_field])) {
$this->image[$this->current_field] .= $text;
} else {
$this->image[$this->current_field] = $text;
}
}
}
}
function parse_cdata($parser, $text)
{
// transcode non-UTF-8 data
if ($this->encoding != 'UTF-8') {
$text = utf8::from_iso8859($text);
}
// skip item, channel, items first time we see them
if ($this->elements_stack[0] == $this->current_field or !$this->current_field) {
return;
}
// we are describing a channel
if ($this->elements_stack[0] == 'channel') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->channel[$this->current_name_space][$this->current_field])) {
$this->channel[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->channel[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->channel[$this->current_field])) {
$this->channel[$this->current_field] .= $text;
} else {
$this->channel[$this->current_field] = $text;
}
}
// we are describing one information item
} elseif ($this->elements_stack[0] == 'item') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->current_item[$this->current_name_space][$this->current_field])) {
$this->current_item[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->current_item[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->current_item[$this->current_field])) {
$this->current_item[$this->current_field] .= $text;
} else {
$this->current_item[$this->current_field] = $text;
}
}
// we are describing a textinput
} elseif ($this->elements_stack[0] == 'textinput') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->textinput[$this->current_name_space][$this->current_field])) {
$this->textinput[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->textinput[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->textinput[$this->current_field])) {
$this->textinput[$this->current_field] .= $text;
} else {
$this->textinput[$this->current_field] = $text;
}
}
// we are describing an image
} elseif ($this->elements_stack[0] == 'image') {
if (isset($this->current_name_space) && $this->current_name_space) {
if (isset($this->image[$this->current_name_space][$this->current_field])) {
$this->image[$this->current_name_space][$this->current_field] .= $text;
} else {
$this->image[$this->current_name_space][$this->current_field] = $text;
}
} else {
if (isset($this->image[$this->current_field])) {
$this->image[$this->current_field] .= $text;
} else {
$this->image[$this->current_field] = $text;
}
}
}
}
