6 * This contains the functions necessary to detect and decode MIME
9 * @copyright 1999-2012 The SquirrelMail Project Team
10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
12 * @package squirrelmail
16 * dependency information
19 class/mime/Message.class.php
20 Message::parseStructure
21 functions/page_header.php
23 functions/display_messages.php
26 functions/imap_general.php
33 functions/attachment_common.php
34 functions/display_messages.php
36 magicHtml => url_parser
37 translateText => url_parser
42 /* -------------------------------------------------------------------------- */
44 /* -------------------------------------------------------------------------- */
47 * Get the MIME structure
49 * This function gets the structure of a message and stores it in the "message" class.
50 * It will return this object for use with all relevant header information and
51 * fully parsed into the standard "message" object format.
53 function mime_structure ($bodystructure, $flags=array()) {
55 /* Isolate the body structure and remove beginning and end parenthesis. */
56 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') +
13));
57 $read = trim(substr ($read, 0, -1));
59 $msg = Message
::parseStructure($read,$i);
61 if (!is_object($msg)) {
62 global $color, $mailbox;
63 displayPageHeader( $color, $mailbox );
64 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
65 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
66 $errormessage .= '<pre>' . htmlspecialchars($read) . '</pre>';
67 plain_error_message( $errormessage );
68 echo '</body></html>';
72 foreach ($flags as $flag) {
73 //FIXME: please document why it is we have to check the first char of the flag but we then go ahead and do a full string comparison anyway. Is this a speed enhancement? If not, let's keep it simple and just compare the full string and forget the switch block.
74 $char = strtoupper($flag{1});
77 if (strtolower($flag) == '\\seen') {
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered
= true;
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted
= true;
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged
= true;
95 else if (strtolower($flag) == '$forwarded') {
96 $msg->is_forwarded
= true;
100 if (strtolower($flag) == '$mdnsent') {
101 $msg->is_mdnsent
= true;
109 // listEntities($msg);
115 /* This starts the parsing of a particular structure. It is called recursively,
116 * so it can be passed different structures. It returns an object of type
118 * First, it checks to see if it is a multipart message. If it is, then it
119 * handles that as it sees is necessary. If it is just a regular entity,
120 * then it parses it and adds the necessary header information (by calling out
121 * to mime_get_elements()
124 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
125 /* Do a bit of error correction. If we couldn't find the entity id, just guess
126 * that it is the first one. That is usually the case anyway.
130 $cmd = "FETCH $id BODY[]";
132 $cmd = "FETCH $id BODY[$ent_id]";
135 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
137 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
139 $topline = trim(array_shift($data));
140 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH .*BODY.*/i', $topline)) ;
141 // Matching with "BODY" above is difficult: in most cases "FETCH \(BODY" would work
142 // but some servers may put other things in the same result, perhaps something such
143 // as "* 23 FETCH (FLAGS (\Seen) BODY[1] {174}". There is some small chance that
144 // if the character sequence "BODY" appears in a response where it isn't actually
145 // a FETCH response data item name, the current regex will break things. The better
146 // way to do this would be to parse the response correctly and not use a regex.
148 $wholemessage = implode('', $data);
149 if (preg_match('/\{([^\}]*)\}/', $topline, $regs)) {
150 $ret = substr($wholemessage, 0, $regs[1]);
151 /* There is some information in the content info header that could be important
152 * in order to parse html messages. Let's get them here.
154 // if ($ret{0} == '<') {
155 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
157 } else if (preg_match('/"([^"]*)"/', $topline, $regs)) {
159 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
160 $ret = $wholemessage;
162 global $where, $what, $mailbox, $passed_id, $startMessage;
163 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
164 if (isset($where) && isset($what)) {
165 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
167 $par .= '&startMessage=' . $startMessage . '&show_more=0';
169 $par .= '&response=' . urlencode($response) .
170 '&message=' . urlencode($message) .
171 '&topline=' . urlencode($topline);
174 '<table width="80%"><tr>' .
175 '<tr><td colspan="2">' .
176 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
178 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
179 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
180 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
181 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
182 "</table><br /></tt></font><hr />";
184 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
186 $wholemessage = implode('', $data);
188 $ret = $wholemessage;
193 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout', $force_crlf='') {
195 /* Don't kill the connection if the browser is over a dialup
196 * and it would take over 30 seconds to download it.
197 * Don't call set_time_limit in safe mode.
200 if (!ini_get('safe_mode')) {
203 /* in case of base64 encoded attachments, do not buffer them.
204 Instead, echo the decoded attachment directly to screen */
205 if (strtolower($encoding) == 'base64') {
207 $query = "FETCH $id BODY[]";
209 $query = "FETCH $id BODY[$ent_id]";
211 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
213 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
214 if (is_resource($rStream)) {
215 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
217 echo decodeBody($body, $encoding, $force_crlf);
222 TODO, use the same method for quoted printable.
223 However, I assume that quoted printable attachments aren't that large
224 so the performancegain / memory usage drop will be minimal.
225 If we decide to add that then we need to adapt sqimap_fread because
226 we need to split te result on \n and fread doesn't stop at \n. That
227 means we also should provide $results from sqimap_fread (by ref) to
228 te function and set $no_return to false. The $filter function for
229 quoted printable should handle unsetting of $results.
232 TODO 2: find out how we write to the output stream php://stdout. fwrite
233 doesn't work because 'php://stdout isn't a stream.
239 /* -[ END MIME DECODING ]----------------------------------------------------------- */
241 /* This is here for debugging purposes. It will print out a list
242 * of all the entity IDs that are in the $message object.
244 function listEntities ($message) {
246 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br />';
247 for ($i = 0; isset($message->entities
[$i]); $i++
) {
249 $msg = listEntities($message->entities
[$i]);
259 function getPriorityStr($priority) {
260 $priority_level = substr($priority,0,1);
262 switch($priority_level) {
263 /* Check for a higher then normal priority. */
266 $priority_string = _("High");
269 /* Check for a lower then normal priority. */
272 $priority_string = _("Low");
275 /* Check for a normal priority. */
278 $priority_level = '3';
279 $priority_string = _("Normal");
283 return $priority_string;
286 /* returns a $message object for a particular entity id */
287 function getEntity ($message, $ent_id) {
288 return $message->getEntity($ent_id);
292 * Extracted from strings.php 23/03/2002
295 function translateText(&$body, $wrap_at, $charset) {
296 global $where, $what; /* from searching */
297 global $color; /* color theme */
299 // require_once(SM_PATH . 'functions/url_parser.php');
301 $body_ary = explode("\n", $body);
302 for ($i=0; $i < count($body_ary); $i++
) {
303 $line = rtrim($body_ary[$i],"\r");
305 if (strlen($line) - 2 >= $wrap_at) {
306 sqWordWrap($line, $wrap_at, $charset);
308 $line = charset_decode($charset, $line);
309 $line = str_replace("\t", ' ', $line);
318 if ($line[$pos] == ' ') {
320 } else if (strpos($line, '>', $pos) === $pos) {
329 $line = '<span class="quote1">' . $line . '</span>';
331 $line = '<span class="quote2">' . $line . '</span>';
334 $body_ary[$i] = $line;
336 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
340 * This returns a parsed string called $body. That string can then
341 * be displayed as the actual message in the HTML. It contains
342 * everything needed, including HTML Tags, Attachments at the
345 * Since 1.2.0 function uses message_body hook.
346 * Till 1.3.0 function included output of formatAttachments().
348 * @param resource $imap_stream imap connection resource
349 * @param object $message squirrelmail message object
350 * @param array $color squirrelmail color theme array
351 * @param integer $wrap_at number of characters per line
352 * @param string $ent_num (since 1.3.0) message part id
353 * @param integer $id (since 1.3.0) message id
354 * @param string $mailbox (since 1.3.0) imap folder name
355 * @return string html formated message text
357 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
358 /* This if statement checks for the entity to show as the
359 * primary message. To add more of them, just put them in the
360 * order that is their priority.
362 global $startMessage, $languages, $squirrelmail_language,
363 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
364 $use_iframe, $iframe_height, $download_and_unsafe_link,
365 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
368 // workaround for not updated config.php
369 if (! isset($use_iframe)) $use_iframe = false;
371 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
372 // images off by default.
373 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
376 $urlmailbox = urlencode($mailbox);
377 $body_message = getEntity($message, $ent_num);
378 if (($body_message->header
->type0
== 'text') ||
379 ($body_message->header
->type0
== 'rfc822')) {
380 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
381 $body = decodeBody($body, $body_message->header
->encoding
);
383 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
384 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
385 if (mb_detect_encoding($body) != 'ASCII') {
386 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
390 /* As of 1.5.2, $body is passed (and modified) by reference */
391 do_hook('message_body', $body);
393 /* If there are other types that shouldn't be formatted, add
397 if ($body_message->header
->type1
== 'html') {
398 if ($show_html_default <> 1) {
399 $entity_conv = array(' ' => ' ',
408 $body = strtr($body, $entity_conv);
409 $body = strip_tags($body);
411 translateText($body, $wrap_at,
412 $body_message->header
->getParameter('charset'));
413 } elseif ($use_iframe) {
415 * If we don't add html message between iframe tags,
416 * we must detect unsafe images and modify $has_unsafe_images.
418 $html_body = magicHTML($body, $id, $message, $mailbox);
419 // Convert character set in order to display html mails in different character set
420 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
422 // creating iframe url
423 $iframeurl=sqm_baseuri().'src/view_html.php?'
424 . 'mailbox=' . $urlmailbox
425 . '&passed_id=' . $id
426 . '&ent_id=' . $ent_num
427 . '&view_unsafe_images=' . (int) $view_unsafe_images;
430 $oTemplate->assign('iframe_url', $iframeurl);
431 $oTemplate->assign('iframe_height', $iframe_height);
432 $oTemplate->assign('html_body', $html_body);
434 $body = $oTemplate->fetch('read_html_iframe.tpl');
436 // old way of html rendering
438 * convert character set. charset_decode does not remove html special chars
439 * applied by magicHTML functions and does not sanitize them second time if
440 * fourth argument is true.
442 $charset = $body_message->header
->getParameter('charset');
443 if (!empty($charset)) {
444 $body = charset_decode($charset,$body,false,true);
446 $body = magicHTML($body, $id, $message, $mailbox);
449 translateText($body, $wrap_at,
450 $body_message->header
->getParameter('charset'));
454 * Previously the links for downloading and unsafe images were printed
455 * under the mail. By putting the links in a global variable we can
456 * print it in the toolbar where it belongs. Since the original code was
457 * in this place it's left here. It might be possible to move it to some
458 * other place if that makes sense. The possibility to do so has not
459 * been evaluated yet.
462 // Initialize the global variable to an empty string.
463 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
464 $download_and_unsafe_link = '';
466 // Prepare and build a link for downloading the mail.
467 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
468 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
469 '&startMessage=' . $startMessage . '&show_more=0';
470 if (isset($passed_ent_id)) {
471 $link .= '&passed_ent_id='.$passed_ent_id;
473 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
475 // Always add the link for downloading the mail as a file to the global
477 $download_and_unsafe_link .= "$nbsp|$nbsp"
478 . create_hyperlink($download_href, _("Download this as a file"));
480 // Find out the right text to use in the link depending on the
481 // circumstances. If the unsafe images are displayed the link should
482 // hide them, if they aren't displayed the link should only appear if
483 // the mail really contains unsafe images.
484 if ($view_unsafe_images) {
485 $text = _("Hide Unsafe Images");
487 if (isset($has_unsafe_images) && $has_unsafe_images) {
488 $link .= '&view_unsafe_images=1';
489 $text = _("View Unsafe Images");
495 // Only create a link for unsafe images if there's need for one. If so:
496 // add it to the global variable.
498 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
499 $unsafe_image_toggle_text = $text;
500 $download_and_unsafe_link .= "$nbsp|$nbsp"
501 . create_hyperlink($unsafe_image_toggle_href, $text);
508 * Generate attachments array for passing to templates.
511 * @param object $message SquirrelMail message object
512 * @param array $exclude_id message parts that are not attachments.
513 * @param string $mailbox mailbox name
514 * @param integer $id message id
516 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
517 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
519 $att_ar = $message->getAttachments($exclude_id);
520 $urlMailbox = urlencode($mailbox);
522 $attachments = array();
523 foreach ($att_ar as $att) {
524 $ent = $att->entity_id
;
525 $header = $att->header
;
526 $type0 = strtolower($header->type0
);
527 $type1 = strtolower($header->type1
);
530 $links['download link']['text'] = _("Download");
531 $links['download link']['href'] = $base_uri .
532 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
534 if ($type0 =='message' && $type1 == 'rfc822') {
535 $default_page = $base_uri . 'src/read_body.php';
536 $rfc822_header = $att->rfc822_header
;
537 $filename = $rfc822_header->subject
;
538 if (trim( $filename ) == '') {
539 $filename = 'untitled-[' . $ent . ']' ;
541 $from_o = $rfc822_header->from
;
542 if (is_object($from_o)) {
543 $from_name = decodeHeader($from_o->getAddress(false));
544 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
545 // something weird happens when a digest message is opened and you return to the digest
546 // now the from object is part of an array. Probably the parseHeader call overwrites the info
547 // retrieved from the bodystructure in a different way. We need to fix this later.
548 // possible starting point, do not fetch header we already have and inspect how
549 // the rfc822_header object behaves.
550 $from_name = decodeHeader($from_o[0]->getAddress(false));
552 $from_name = _("Unknown sender");
554 $description = _("From").': '.$from_name;
556 $default_page = $base_uri . 'src/download.php';
557 $filename = $att->getFilename();
558 if ($header->description
) {
559 $description = decodeHeader($header->description
);
565 $display_filename = $filename;
566 if (isset($passed_ent_id)) {
567 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
569 $passed_ent_id_link = '';
571 $defaultlink = $default_page . "?startMessage=$startMessage"
572 . "&passed_id=$id&mailbox=$urlMailbox"
573 . '&ent_id='.$ent.$passed_ent_id_link;
574 if ($where && $what) {
575 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
577 // IE does make use of mime content sniffing. Forcing a download
578 // prohibit execution of XSS inside an application/octet-stream attachment
579 if ($type0 == 'application' && $type1 == 'octet-stream') {
580 $defaultlink .= '&absolute_dl=true';
583 /* This executes the attachment hook with a specific MIME-type.
584 * It also allows plugins to run if there's a rule for a more
585 * generic type. Finally, a hook for ALL attachment types is
588 // First remember the default link.
589 $defaultlink_orig = $defaultlink;
591 /* The API for this hook has changed as of 1.5.2 so that all plugin
592 arguments are passed in an array instead of each their own plugin
593 argument, and arguments are passed by reference, so instead of
594 returning any changes, changes should simply be made to the original
595 arguments themselves. */
596 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
597 &$defaultlink, &$display_filename, &$where, &$what);
598 do_hook("attachment $type0/$type1", $temp);
599 /* The API for this hook has changed as of 1.5.2 so that all plugin
600 arguments are passed in an array instead of each their own plugin
601 argument, and arguments are passed by reference, so instead of
602 returning any changes, changes should simply be made to the original
603 arguments themselves. */
604 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
605 &$defaultlink, &$display_filename, &$where, &$what);
606 // Do not let a generic plugin change the default link if a more
607 // specialized one already did it...
608 if ($defaultlink != $defaultlink_orig) {
612 do_hook("attachment $type0/*", $temp);
613 /* The API for this hook has changed as of 1.5.2 so that all plugin
614 arguments are passed in an array instead of each their own plugin
615 argument, and arguments are passed by reference, so instead of
616 returning any changes, changes should simply be made to the original
617 arguments themselves. */
618 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
619 &$defaultlink, &$display_filename, &$where, &$what);
620 // Do not let a generic plugin change the default link if a more
621 // specialized one already did it...
622 if ($defaultlink != $defaultlink_orig) {
626 do_hook("attachment */*", $temp);
628 $this_attachment = array();
629 $this_attachment['Name'] = decodeHeader($display_filename);
630 $this_attachment['Description'] = $description;
631 $this_attachment['DefaultHREF'] = $defaultlink;
632 $this_attachment['DownloadHREF'] = $links['download link']['href'];
633 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
634 $this_attachment['Size'] = $header->size
;
635 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
636 $this_attachment['OtherLinks'] = array();
637 foreach ($links as $val) {
638 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
640 if (empty($val['text']) && empty($val['extra']))
644 $temp['HREF'] = $val['href'];
645 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
646 $this_attachment['OtherLinks'][] = $temp;
648 $attachments[] = $this_attachment;
657 * Displays attachment links and information
659 * Since 1.3.0 function is not included in formatBody() call.
661 * Since 1.0.2 uses attachment $type0/$type1 hook.
662 * Since 1.2.5 uses attachment $type0/* hook.
663 * Since 1.5.0 uses attachments_bottom hook.
664 * Since 1.5.2 uses templates and does *not* return a value.
666 * @param object $message SquirrelMail message object
667 * @param array $exclude_id message parts that are not attachments.
668 * @param string $mailbox mailbox name
669 * @param integer $id message id
671 function formatAttachments($message, $exclude_id, $mailbox, $id) {
674 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
676 $oTemplate->assign('attachments', $attach);
677 $oTemplate->display('read_attachments.tpl');
680 function sqimap_base64_decode(&$string) {
682 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
683 // fly decoding (to reduce memory usage) you have to check if the
684 // data has incomplete pairs
686 // Remove the noise in order to check if the 4 bytes pairs are complete
687 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
690 $iMod = strlen($string) %
4;
692 $sStringRem = substr($string,-$iMod);
693 // Check if $sStringRem contains padding characters
694 if (substr($sStringRem,-1) != '=') {
695 $string = substr($string,0,-$iMod);
700 $string = base64_decode($string);
705 * Decodes encoded string (usually message body)
707 * This function decodes a string (usually the message body)
708 * depending on the encoding type. Currently quoted-printable
709 * and base64 encodings are supported.
711 * The decode_body hook was added to this function in 1.4.2/1.5.0.
712 * The $force_crlf parameter was added in 1.5.2.
714 * @param string $string The encoded string
715 * @param string $encoding used encoding
716 * @param string $force_crlf Whether or not to force CRLF or LF
717 * line endings (or to leave as is).
718 * If given as "LF", line endings will
719 * all be converted to LF; if "CRLF",
720 * line endings will all be converted
721 * to CRLF. If given as an empty value,
722 * the global $force_crlf_default will
723 * be consulted (it can be specified in
724 * config/config_local.php). Otherwise,
725 * any other value will cause the string
726 * to be left alone. Note that this will
727 * be overridden to "LF" if not using at
728 * least PHP version 4.3.0. (OPTIONAL;
729 * default is empty - consult global
732 * @return string The decoded string
737 function decodeBody($string, $encoding, $force_crlf='') {
739 global $force_crlf_default;
740 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
741 $force_crlf = strtoupper($force_crlf);
743 // must force line endings to LF due to broken
744 // quoted_printable_decode() in PHP versions
745 // before 4.3.0 (see below)
747 if (!check_php_version(4, 3, 0) ||
$force_crlf == 'LF')
748 $string = str_replace("\r\n", "\n", $string);
749 else if ($force_crlf == 'CRLF')
750 $string = str_replace("\n", "\r\n", $string);
752 $encoding = strtolower($encoding);
754 $encoding_handler = do_hook('decode_body', $encoding);
757 // plugins get first shot at decoding the string
759 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
760 $string = $encoding_handler('decode', $string);
762 } elseif ($encoding == 'quoted-printable' ||
763 $encoding == 'quoted_printable') {
765 // quoted_printable_decode() function is broken in older
766 // php versions. Text with \r\n decoding was fixed only
767 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
768 // and the above call to: str_replace("\r\n", "\n", $string);
770 $string = quoted_printable_decode($string);
772 } elseif ($encoding == 'base64') {
773 $string = base64_decode($string);
776 // All other encodings are returned raw.
783 * This function decodes strings that are encoded according to
784 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
785 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
787 * @param string $string header string that has to be made readable
788 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
789 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
790 * @param boolean $decide decide if string can be utfencoded. defaults to false
791 * @return string decoded header string
793 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
794 global $languages, $squirrelmail_language,$default_charset;
795 if (is_array($string)) {
796 $string = implode("\n", $string);
799 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
800 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
801 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
802 // Do we need to return at this point?
809 $aString = explode(' ',$string);
811 foreach ($aString as $chunk) {
812 if ($encoded && $chunk === '') {
814 } elseif ($chunk === '') {
819 /* if encoded words are not separated by a linear-space-white we still catch them */
822 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
823 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
824 if ($iLastMatch !== $j) {
834 $ret .= htmlspecialchars($res[1]);
838 $encoding = ucfirst($res[3]);
840 /* decide about valid decoding */
841 if ($decide && is_conversion_safe($res[2])) {
843 $can_be_encoded=true;
845 $can_be_encoded=false;
850 $replace = base64_decode($res[4]);
852 if ($can_be_encoded) {
853 /* convert string to different charset,
854 * if functions asks for it (usually in compose)
856 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
858 // convert string to html codes in order to display it
859 $ret .= charset_decode($res[2],$replace);
863 $replace = htmlspecialchars($replace);
869 $replace = str_replace('_', ' ', $res[4]);
870 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
873 if ($can_be_encoded) {
874 /* convert string to different charset,
875 * if functions asks for it (usually in compose)
877 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
879 // convert string to html codes in order to display it
880 $replace = charset_decode($res[2], $replace);
884 $replace = htmlspecialchars($replace);
903 if (!$encoded && $htmlsafe) {
904 $ret .= htmlspecialchars($chunk);
910 /* remove the first added space */
913 $ret = substr($ret,5);
915 $ret = substr($ret,1);
925 * Function uses XTRA_CODE _encodeheader function, if such function exists.
927 * Function uses Q encoding by default and encodes a string according to RFC
928 * 1522 for use in headers if it contains 8-bit characters or anything that
929 * looks like it should be encoded.
931 * Function switches to B encoding and encodeHeaderBase64() function, if
932 * string is 8bit and multibyte character set supported by mbstring extension
933 * is used. It can cause E_USER_NOTICE errors, if interface is used with
934 * multibyte character set unsupported by mbstring extension.
936 * @param string $string header string, that has to be encoded
937 * @return string quoted-printable encoded string
938 * @todo make $mb_charsets system wide constant
940 function encodeHeader ($string) {
941 global $default_charset, $languages, $squirrelmail_language;
943 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
944 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
945 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
948 // Use B encoding for multibyte charsets
949 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
950 if (in_array($default_charset,$mb_charsets) &&
951 in_array($default_charset,sq_mb_list_encodings()) &&
952 sq_is8bit($string)) {
953 return encodeHeaderBase64($string,$default_charset);
954 } elseif (in_array($default_charset,$mb_charsets) &&
955 sq_is8bit($string) &&
956 ! in_array($default_charset,sq_mb_list_encodings())) {
957 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
958 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
961 // Encode only if the string contains 8-bit characters or =?
962 $j = strlen($string);
963 $max_l = 75 - strlen($default_charset) - 7;
966 $iEncStart = $enc_init = false;
967 $cur_l = $iOffset = 0;
968 for($i = 0; $i < $j; ++
$i) {
978 if ($iEncStart === false) {
982 if ($cur_l > ($max_l-2)) {
983 /* if there is an stringpart that doesn't need encoding, add it */
984 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
985 $aRet[] = "=?$default_charset?Q?$ret?=";
991 $ret .= sprintf("=%02X",ord($string{$i}));
996 if ($iEncStart !== false) {
997 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
998 $aRet[] = "=?$default_charset?Q?$ret?=";
1006 if ($iEncStart !== false) {
1008 if ($cur_l > $max_l) {
1009 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1010 $aRet[] = "=?$default_charset?Q?$ret?=";
1021 $k = ord($string{$i});
1023 if ($iEncStart === false) {
1024 // do not start encoding in the middle of a string, also take the rest of the word.
1025 $sLeadString = substr($string,0,$i);
1026 $aLeadString = explode(' ',$sLeadString);
1027 $sToBeEncoded = array_pop($aLeadString);
1028 $iEncStart = $i - strlen($sToBeEncoded);
1029 $ret .= $sToBeEncoded;
1030 $cur_l +
= strlen($sToBeEncoded);
1033 /* first we add the encoded string that reached it's max size */
1034 if ($cur_l > ($max_l-2)) {
1035 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1036 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1043 $ret .= sprintf("=%02X", $k);
1045 if ($iEncStart !== false) {
1047 if ($cur_l > $max_l) {
1048 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1049 $aRet[] = "=?$default_charset?Q?$ret?=";
1055 $ret .= $string{$i};
1064 if ($iEncStart !== false) {
1065 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1066 $aRet[] = "=?$default_charset?Q?$ret?=";
1068 $aRet[] = substr($string,$iOffset);
1070 $string = implode('',$aRet);
1076 * Encodes string according to rfc2047 B encoding header formating rules
1078 * It is recommended way to encode headers with character sets that store
1079 * symbols in more than one byte.
1081 * Function requires mbstring support. If required mbstring functions are missing,
1082 * function returns false and sets E_USER_WARNING level error message.
1084 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1085 * that mbstring functions will generate E_WARNING errors, if unsupported
1086 * character set is used. mb_encode_mimeheader function provided by php
1087 * mbstring extension is not used in order to get better control of header
1090 * Used php code functions - function_exists(), trigger_error(), strlen()
1091 * (is used with charset names and base64 strings). Used php mbstring
1092 * functions - mb_strlen and mb_substr.
1094 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1095 * encoding), rfc 2822 (header folding)
1097 * @param string $string header string that must be encoded
1098 * @param string $charset character set. Must be supported by mbstring extension.
1099 * Use sq_mb_list_encodings() to detect supported charsets.
1100 * @return string string encoded according to rfc2047 B encoding formating rules
1102 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1103 * @todo Do we want to control max length of header?
1104 * @todo Do we want to control EOL (end-of-line) marker?
1105 * @todo Do we want to translate error message?
1107 function encodeHeaderBase64($string,$charset) {
1109 * Check mbstring function requirements.
1111 if (! function_exists('mb_strlen') ||
1112 ! function_exists('mb_substr')) {
1113 // set E_USER_WARNING
1114 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1119 // initial return array
1123 * header length = 75 symbols max (same as in encodeHeader)
1124 * remove $charset length
1125 * remove =? ? ?= (5 chars)
1126 * remove 2 more chars (\r\n ?)
1128 $iMaxLength = 75 - strlen($charset) - 7;
1130 // set first character position
1133 // loop through all characters. count characters and not bytes.
1134 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1135 // encode string from starting character to current character.
1136 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1138 // Check encoded string length
1139 if(strlen($encoded_string)>$iMaxLength) {
1140 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1141 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1143 // set new starting character
1144 $iStartCharNum = $iCharNum-1;
1146 // encode last char (in case it is last character in string)
1147 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1148 } // if string is shorter than max length - add next character
1151 // add last encoded string to array
1152 $aRet[] = $encoded_string;
1154 // set initial return string
1157 // loop through encoded strings
1158 foreach($aRet as $string) {
1159 // TODO: Do we want to control EOL (end-of-line) marker
1160 if ($sRet!='') $sRet.= " ";
1162 // add header tags and encoded string to return string
1163 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1169 /* This function trys to locate the entity_id of a specific mime element */
1170 function find_ent_id($id, $message) {
1171 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1172 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1173 $ret = find_ent_id($id, $message->entities
[$i]);
1175 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1176 // if (sq_check_save_extension($message->entities[$i])) {
1177 return $message->entities
[$i]->entity_id
;
1179 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1181 * This is part of a fix for Outlook Express 6.x generating
1182 * cid URLs without creating content-id headers
1185 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1186 return $message->entities
[$i]->entity_id
;
1194 function sq_check_save_extension($message) {
1195 $filename = $message->getFilename();
1196 $ext = substr($filename, strrpos($filename,'.')+
1);
1197 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1198 return in_array($ext, $save_extensions);
1203 ** HTMLFILTER ROUTINES
1207 * This function checks attribute values for entity-encoded values
1208 * and returns them translated into 8-bit strings so we can run
1211 * @param $attvalue A string to run entity check against.
1212 * @return Nothing, modifies a reference value.
1214 function sq_defang(&$attvalue){
1217 * Skip this if there aren't ampersands or backslashes.
1219 if (strpos($attvalue, '&') === false
1220 && strpos($attvalue, '\\') === false){
1224 // before deent, translate the dangerous unicode characters and ... to safe values
1225 // otherwise the regular expressions do not match.
1231 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1232 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1233 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1234 } while ($m == true);
1235 $attvalue = stripslashes($attvalue);
1239 * Kill any tabs, newlines, or carriage returns. Our friends the
1240 * makers of the browser with 95% market value decided that it'd
1241 * be funny to make "java[tab]script" be just as good as "javascript".
1243 * @param attvalue The attribute value before extraneous spaces removed.
1244 * @return attvalue Nothing, modifies a reference value.
1246 function sq_unspace(&$attvalue){
1248 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1249 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1250 Array('', '', '', '', ''), $attvalue);
1255 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1256 * IE as regular characters.
1258 * @param attvalue The attribute value before dangerous characters are translated.
1259 * @return attvalue Nothing, modifies a reference value.
1260 * @author Marc Groot Koerkamp.
1262 function sq_fixIE_idiocy(&$attvalue) {
1264 $attvalue = str_replace("\0", "", $attvalue);
1266 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1268 // IE has the evil habit of accepting every possible value for the attribute expression.
1269 // The table below contains characters which are parsed by IE if they are used in the "expression"
1271 $aDangerousCharsReplacementTable = array(
1272 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1273 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1274 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1275 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1276 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1277 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1278 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1279 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1280 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1281 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1282 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1283 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1284 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1285 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1286 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1287 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1288 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1289 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1290 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1291 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1292 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1293 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1294 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1295 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1296 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1297 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1298 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1299 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1300 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1301 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1302 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1303 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1304 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1305 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1306 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1307 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1308 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1309 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1310 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1311 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1312 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1313 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1314 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1315 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1316 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1317 "\xCA\x9F", /* L UNICODE IPA Extension */
1318 "\xCA\x80", /* R UNICODE IPA Extension */
1319 "\xC9\xB4"), /* N UNICODE IPA Extension */
1320 array('l', 'l', 'r','r','n','n',
1321 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1322 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1323 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1324 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1326 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1328 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1332 * This function returns the final tag out of the tag name, an array
1333 * of attributes, and the type of the tag. This function is called by
1334 * sq_sanitize internally.
1336 * @param $tagname the name of the tag.
1337 * @param $attary the array of attributes and their values
1338 * @param $tagtype The type of the tag (see in comments).
1339 * @return a string with the final tag representation.
1341 function sq_tagprint($tagname, $attary, $tagtype){
1342 $me = 'sq_tagprint';
1345 $fulltag = '</' . $tagname . '>';
1347 $fulltag = '<' . $tagname;
1348 if (is_array($attary) && sizeof($attary)){
1350 while (list($attname, $attvalue) = each($attary)){
1351 array_push($atts, "$attname=$attvalue");
1353 $fulltag .= ' ' . join(" ", $atts);
1364 * A small helper function to use with array_walk. Modifies a by-ref
1365 * value and makes it lowercase.
1367 * @param $val a value passed by-ref.
1368 * @return void since it modifies a by-ref value.
1370 function sq_casenormalize(&$val){
1371 $val = strtolower($val);
1375 * This function skips any whitespace from the current position within
1376 * a string and to the next non-whitespace value.
1378 * @param $body the string
1379 * @param $offset the offset within the string where we should start
1380 * looking for the next non-whitespace character.
1381 * @return the location within the $body where the next
1382 * non-whitespace char is located.
1384 function sq_skipspace($body, $offset){
1385 $me = 'sq_skipspace';
1386 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1387 if (sizeof($matches{1})){
1388 $count = strlen($matches{1});
1395 * This function looks for the next character within a string. It's
1396 * really just a glorified "strpos", except it catches if failures
1399 * @param $body The string to look for needle in.
1400 * @param $offset Start looking from this position.
1401 * @param $needle The character/string to look for.
1402 * @return location of the next occurance of the needle, or
1403 * strlen($body) if needle wasn't found.
1405 function sq_findnxstr($body, $offset, $needle){
1406 $me = 'sq_findnxstr';
1407 $pos = strpos($body, $needle, $offset);
1408 if ($pos === FALSE){
1409 $pos = strlen($body);
1415 * This function takes a PCRE-style regexp and tries to match it
1416 * within the string.
1418 * @param $body The string to look for needle in.
1419 * @param $offset Start looking from here.
1420 * @param $reg A PCRE-style regex to match.
1421 * @return Returns a false if no matches found, or an array
1422 * with the following members:
1423 * - integer with the location of the match within $body
1424 * - string with whatever content between offset and the match
1425 * - string with whatever it is we matched
1427 function sq_findnxreg($body, $offset, $reg){
1428 $me = 'sq_findnxreg';
1431 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1432 if (!isset($matches{0}) ||
!$matches{0}){
1435 $retarr{0} = $offset +
strlen($matches{1});
1436 $retarr{1} = $matches{1};
1437 $retarr{2} = $matches{2};
1443 * This function looks for the next tag.
1445 * @param $body String where to look for the next tag.
1446 * @param $offset Start looking from here.
1447 * @return false if no more tags exist in the body, or
1448 * an array with the following members:
1449 * - string with the name of the tag
1450 * - array with attributes and their values
1451 * - integer with tag type (1, 2, or 3)
1452 * - integer where the tag starts (starting "<")
1453 * - integer where the tag ends (ending ">")
1454 * first three members will be false, if the tag is invalid.
1456 function sq_getnxtag($body, $offset){
1457 $me = 'sq_getnxtag';
1458 if ($offset > strlen($body)){
1461 $lt = sq_findnxstr($body, $offset, "<");
1462 if ($lt == strlen($body)){
1467 * blah blah <tag attribute="value">
1470 $pos = sq_skipspace($body, $lt+
1);
1471 if ($pos >= strlen($body)){
1472 return Array(false, false, false, $lt, strlen($body));
1475 * There are 3 kinds of tags:
1476 * 1. Opening tag, e.g.:
1478 * 2. Closing tag, e.g.:
1480 * 3. XHTML-style content-less tag, e.g.:
1481 * <img src="blah" />
1484 switch (substr($body, $pos, 1)){
1491 * A comment or an SGML declaration.
1493 if (substr($body, $pos+
1, 2) == "--"){
1494 $gt = strpos($body, "-->", $pos);
1496 $gt = strlen($body);
1500 return Array(false, false, false, $lt, $gt);
1502 $gt = sq_findnxstr($body, $pos, ">");
1503 return Array(false, false, false, $lt, $gt);
1508 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1518 * Look for next [\W-_], which will indicate the end of the tag name.
1520 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1521 if ($regary == false){
1522 return Array(false, false, false, $lt, strlen($body));
1524 list($pos, $tagname, $match) = $regary;
1525 $tagname = strtolower($tagname);
1528 * $match can be either of these:
1529 * '>' indicating the end of the tag entirely.
1530 * '\s' indicating the end of the tag name.
1531 * '/' indicating that this is type-3 xhtml tag.
1533 * Whatever else we find there indicates an invalid tag.
1538 * This is an xhtml-style tag with a closing / at the
1539 * end, like so: <img src="blah" />. Check if it's followed
1540 * by the closing bracket. If not, then this tag is invalid
1542 if (substr($body, $pos, 2) == "/>"){
1546 $gt = sq_findnxstr($body, $pos, ">");
1547 $retary = Array(false, false, false, $lt, $gt);
1551 return Array($tagname, false, $tagtype, $lt, $pos);
1555 * Check if it's whitespace
1557 if (!preg_match('/\s/', $match)){
1559 * This is an invalid tag! Look for the next closing ">".
1561 $gt = sq_findnxstr($body, $lt, ">");
1562 return Array(false, false, false, $lt, $gt);
1568 * At this point we're here:
1569 * <tagname attribute='blah'>
1572 * At this point we loop in order to find all attributes.
1578 while ($pos <= strlen($body)){
1579 $pos = sq_skipspace($body, $pos);
1580 if ($pos == strlen($body)){
1584 return Array(false, false, false, $lt, $pos);
1587 * See if we arrived at a ">" or "/>", which means that we reached
1588 * the end of the tag.
1591 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1595 $pos +
= strlen($matches{1});
1596 if ($matches{2} == "/>"){
1600 return Array($tagname, $attary, $tagtype, $lt, $pos);
1604 * There are several types of attributes, with optional
1605 * [:space:] between members.
1607 * attrname[:space:]=[:space:]'CDATA'
1609 * attrname[:space:]=[:space:]"CDATA"
1611 * attr[:space:]=[:space:]CDATA
1615 * We leave types 1 and 2 the same, type 3 we check for
1616 * '"' and convert to """ if needed, then wrap in
1617 * double quotes. Type 4 we convert into:
1620 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1621 if ($regary == false){
1623 * Looks like body ended before the end of tag.
1625 return Array(false, false, false, $lt, strlen($body));
1627 list($pos, $attname, $match) = $regary;
1628 $attname = strtolower($attname);
1630 * We arrived at the end of attribute name. Several things possible
1632 * '>' means the end of the tag and this is attribute type 4
1633 * '/' if followed by '>' means the same thing as above
1634 * '\s' means a lot of things -- look what it's followed by.
1635 * anything else means the attribute is invalid.
1640 * This is an xhtml-style tag with a closing / at the
1641 * end, like so: <img src="blah" />. Check if it's followed
1642 * by the closing bracket. If not, then this tag is invalid
1644 if (substr($body, $pos, 2) == "/>"){
1648 $gt = sq_findnxstr($body, $pos, ">");
1649 $retary = Array(false, false, false, $lt, $gt);
1653 $attary{$attname} = '"yes"';
1654 return Array($tagname, $attary, $tagtype, $lt, $pos);
1658 * Skip whitespace and see what we arrive at.
1660 $pos = sq_skipspace($body, $pos);
1661 $char = substr($body, $pos, 1);
1663 * Two things are valid here:
1664 * '=' means this is attribute type 1 2 or 3.
1665 * \w means this was attribute type 4.
1666 * anything else we ignore and re-loop. End of tag and
1667 * invalid stuff will be caught by our checks at the beginning
1672 $pos = sq_skipspace($body, $pos);
1674 * Here are 3 possibilities:
1675 * "'" attribute type 1
1676 * '"' attribute type 2
1677 * everything else is the content of tag type 3
1679 $quot = substr($body, $pos, 1);
1681 $regary = sq_findnxreg($body, $pos+
1, "\'");
1682 if ($regary == false){
1683 return Array(false, false, false, $lt, strlen($body));
1685 list($pos, $attval, $match) = $regary;
1687 $attary{$attname} = "'" . $attval . "'";
1688 } else if ($quot == '"'){
1689 $regary = sq_findnxreg($body, $pos+
1, '\"');
1690 if ($regary == false){
1691 return Array(false, false, false, $lt, strlen($body));
1693 list($pos, $attval, $match) = $regary;
1695 $attary{$attname} = '"' . $attval . '"';
1698 * These are hateful. Look for \s, or >.
1700 $regary = sq_findnxreg($body, $pos, "[\s>]");
1701 if ($regary == false){
1702 return Array(false, false, false, $lt, strlen($body));
1704 list($pos, $attval, $match) = $regary;
1706 * If it's ">" it will be caught at the top.
1708 $attval = preg_replace("/\"/s", """, $attval);
1709 $attary{$attname} = '"' . $attval . '"';
1711 } else if (preg_match("|[\w/>]|", $char)) {
1713 * That was attribute type 4.
1715 $attary{$attname} = '"yes"';
1718 * An illegal character. Find next '>' and return.
1720 $gt = sq_findnxstr($body, $pos, ">");
1721 return Array(false, false, false, $lt, $gt);
1727 * The fact that we got here indicates that the tag end was never
1728 * found. Return invalid tag indication so it gets stripped.
1730 return Array(false, false, false, $lt, strlen($body));
1734 * Translates entities into literal values so they can be checked.
1736 * @param $attvalue the by-ref value to check.
1737 * @param $regex the regular expression to check against.
1738 * @param $hex whether the entites are hexadecimal.
1739 * @return True or False depending on whether there were matches.
1741 function sq_deent(&$attvalue, $regex, $hex=false){
1745 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1746 preg_match_all($regex, $attvalue, $matches);
1747 if (is_array($matches) && sizeof($matches[0]) > 0){
1749 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1750 $numval = $matches[1][$i];
1752 $numval = hexdec($numval);
1754 $repl{$matches[0][$i]} = chr($numval);
1756 $attvalue = strtr($attvalue, $repl);
1764 * This function runs various checks against the attributes.
1766 * @param $tagname String with the name of the tag.
1767 * @param $attary Array with all tag attributes.
1768 * @param $rm_attnames See description for sq_sanitize
1769 * @param $bad_attvals See description for sq_sanitize
1770 * @param $add_attr_to_tag See description for sq_sanitize
1771 * @param $message message object
1772 * @param $id message id
1773 * @return Array with modified attributes.
1775 function sq_fixatts($tagname,
1785 while (list($attname, $attvalue) = each($attary)){
1787 * See if this attribute should be removed.
1789 foreach ($rm_attnames as $matchtag=>$matchattrs){
1790 if (preg_match($matchtag, $tagname)){
1791 foreach ($matchattrs as $matchattr){
1792 if (preg_match($matchattr, $attname)){
1793 unset($attary{$attname});
1800 * Workaround for IE quirks
1802 sq_fixIE_idiocy($attvalue);
1805 * Remove any backslashes, entities, and extraneous whitespace.
1808 $oldattvalue = $attvalue;
1809 sq_defang($attvalue);
1810 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1811 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1812 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1813 $attvalue = "idiocy";
1814 $attary{$attname} = $attvalue;
1816 sq_unspace($attvalue);
1819 * Now let's run checks on the attvalues.
1820 * I don't expect anyone to comprehend this. If you do,
1821 * get in touch with me so I can drive to where you live and
1822 * shake your hand personally. :)
1824 foreach ($bad_attvals as $matchtag=>$matchattrs){
1825 if (preg_match($matchtag, $tagname)){
1826 foreach ($matchattrs as $matchattr=>$valary){
1827 if (preg_match($matchattr, $attname)){
1829 * There are two arrays in valary.
1831 * Second one is replacements
1833 list($valmatch, $valrepl) = $valary;
1835 preg_replace($valmatch, $valrepl, $attvalue);
1836 if ($newvalue != $attvalue){
1837 $attary{$attname} = $newvalue;
1838 $attvalue = $newvalue;
1844 if ($attname == 'style') {
1845 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1846 // 8bit and control characters in style attribute values can be used for XSS, remove them
1847 $attary{$attname} = '"disallowed character"';
1849 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1850 if (count($aMatch)) {
1851 foreach($aMatch[1] as $sMatch) {
1853 $urlvalue = $sMatch;
1854 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1855 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1860 * Use white list based filtering on attributes which can contain url's
1862 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1863 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1864 $attary{$attname} = $attvalue;
1868 * See if we need to append any attributes to this tag.
1870 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1871 if (preg_match($matchtag, $tagname)){
1872 $attary = array_merge($attary, $addattary);
1879 * This function filters url's
1881 * @param $attvalue String with attribute value to filter
1882 * @param $message message object
1883 * @param $id message id
1884 * @param $mailbox mailbox
1885 * @param $sQuote quoting characters around url's
1887 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1888 $attvalue = trim($attvalue);
1889 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1890 // remove the double quotes
1891 $sQuote = $attvalue[0];
1892 $attvalue = trim(substr($attvalue,1,-1));
1895 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1896 // images off by default.
1897 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
1899 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1902 * Replace empty src tags with the blank image. src is only used
1903 * for frames, images, and image inputs. Doing a replace should
1904 * not affect them working as should be, however it will stop
1905 * IE from being kicked off when src for img tags are not set
1907 if ($attvalue == '') {
1908 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1910 // first, disallow 8 bit characters and control characters
1911 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1914 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1917 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1921 $aUrl = parse_url($attvalue);
1922 if (isset($aUrl['scheme'])) {
1923 switch(strtolower($aUrl['scheme'])) {
1928 if ($attname != 'href') {
1929 if ($view_unsafe_images == false) {
1930 $attvalue = $sQuote . $secremoveimg . $sQuote;
1932 if (isset($aUrl['path'])) {
1934 // No one has been able to show that image URIs
1935 // can be exploited, so for now, no restrictions
1936 // are made at all. If this proves to be a problem,
1937 // the commented-out code below can be of help.
1938 // (One consideration is that I see nothing in this
1939 // function that specifically says that we will
1940 // only ever arrive here when inspecting an image
1941 // tag, although that does seem to be the end
1942 // result - e.g., <script src="..."> where malicious
1943 // image URIs are in fact a problem are already
1944 // filtered out elsewhere.
1945 /* ---------------------------------
1946 // validate image extension.
1947 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1948 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1949 // If URI is to something other than
1950 // a regular image file, get the contents
1951 // and try to see if it is an image.
1952 // Don't use Fileinfo (finfo_file()) because
1953 // we'd need to make the admin configure the
1954 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1957 if (function_exists('mime_content_type')
1958 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1962 $file_contents = '';
1963 while (!feof($FILE)) {
1964 $file_contents .= fread($FILE, 8192);
1968 // store file locally
1970 global $attachment_dir, $username;
1971 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1972 $localfilename = GenerateRandomString(32, '', 7);
1973 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1974 while (file_exists($full_localfilename)) {
1975 $localfilename = GenerateRandomString(32, '', 7);
1976 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1978 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1979 fwrite($FILE, $file_contents);
1982 // get mime type and remove file
1984 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1985 unlink("$hashed_attachment_dir/$localfilename");
1987 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1988 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1989 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1992 --------------------------------- */
1994 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1998 $attvalue = $sQuote . $attvalue . $sQuote;
2003 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
2004 * One day MS might actually make it match something useful, for now, falling
2005 * back to using cid2http, so we can grab the blank.png.
2007 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2011 * Turn cid: urls into http-friendly ones.
2013 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2016 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2020 if (!isset($aUrl['path']) ||
$aUrl['path'] != $secremoveimg) {
2021 // parse_url did not lead to satisfying result
2022 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2030 * This function edits the style definition to make them friendly and
2031 * usable in SquirrelMail.
2033 * @param $message the message object
2034 * @param $id the message id
2035 * @param $content a string with whatever is between <style> and </style>
2036 * @param $mailbox the message mailbox
2037 * @return a string with edited content.
2039 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
2040 $me = 'sq_fixstyle';
2041 // workaround for </style> in between comments
2042 $iCurrentPos = $pos;
2047 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
2054 if ($sToken == '<') {
2064 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2069 $content .= $sToken;
2077 if ($sToken == '<') {
2079 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
2080 $i = strpos($body,'-->',$i+
3);
2081 if ($i === false) { // no end comment
2099 if ($bSucces == FALSE){
2100 return array(FALSE, strlen($body));
2106 * First look for general BODY style declaration, which would be
2108 * body {background: blah-blah}
2109 * and change it to .bodyclass so we can just assign it to a <div>
2111 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2112 $secremoveimg = '../images/' . _("sec_remove_eng.png");
2114 * Fix url('blah') declarations.
2116 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2117 // "url(\\1$secremoveimg\\2)", $content);
2119 // first check for 8bit sequences and disallowed control characters
2120 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2121 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2122 return array($content, $newpos);
2125 // IE Sucks hard. We have a special function for it.
2126 sq_fixIE_idiocy($content);
2128 // remove @import line
2129 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2131 // translate ur\l and variations (IE parses that)
2132 // TODO check if the sq_fixIE_idiocy function already handles this.
2133 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2134 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2135 if (count($aMatch)) {
2136 $aValue = $aReplace = array();
2137 foreach($aMatch[1] as $sMatch) {
2139 $urlvalue = $sMatch;
2140 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2141 $aValue[] = $sMatch;
2142 $aReplace[] = $urlvalue;
2144 $content = str_replace($aValue,$aReplace,$content);
2148 * Remove any backslashes, entities, and extraneous whitespace.
2150 $contentTemp = $content;
2151 sq_defang($contentTemp);
2152 sq_unspace($contentTemp);
2155 * Fix stupid css declarations which lead to vulnerabilities
2158 * Also remove "position" attribute, as it can easily be set
2159 * to "fixed" or "absolute" with "left" and "top" attributes
2160 * of zero, taking over the whole content frame. It can also
2161 * be set to relative and move itself anywhere it wants to,
2162 * displaying content in areas it shouldn't be allowed to touch.
2164 $match = Array('/\/\*.*\*\//',
2168 '/include-source/i',
2172 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
2173 $contentNew = preg_replace($match, $replace, $contentTemp);
2174 if ($contentNew !== $contentTemp) {
2175 // insecure css declarations are used. From now on we don't care
2176 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2177 $content = $contentNew;
2179 return array($content, $newpos);
2184 * This function converts cid: url's into the ones that can be viewed in
2187 * @param $message the message object
2188 * @param $id the message id
2189 * @param $cidurl the cid: url.
2190 * @param $mailbox the message mailbox
2191 * @return a string with a http-friendly url
2193 function sq_cid2http($message, $id, $cidurl, $mailbox){
2195 * Get rid of quotes.
2197 $quotchar = substr($cidurl, 0, 1);
2198 if ($quotchar == '"' ||
$quotchar == "'"){
2199 $cidurl = str_replace($quotchar, "", $cidurl);
2203 $cidurl = substr(trim($cidurl), 4);
2205 $match_str = '/\{.*?\}\//';
2207 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2209 $linkurl = find_ent_id($cidurl, $message);
2210 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2211 unsafe link image */
2215 * This is part of a fix for Outlook Express 6.x generating
2216 * cid URLs without creating content-id headers. These images are
2217 * not part of the multipart/related html mail. The html contains
2218 * <img src="cid:{some_id}/image_filename.ext"> references to
2219 * attached images with as goal to render them inline although
2220 * the attachment disposition property is not inline.
2223 if (empty($linkurl)) {
2224 if (preg_match('/{.*}\//', $cidurl)) {
2225 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2226 if (!empty($cidurl)) {
2227 $linkurl = find_ent_id($cidurl, $message);
2232 if (!empty($linkurl)) {
2233 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&' .
2234 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2235 '&ent_id=' . $linkurl . $quotchar;
2238 * If we couldn't generate a proper img url, drop in a blank image
2239 * instead of sending back empty, otherwise it causes unusual behaviour
2241 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2248 * This function changes the <body> tag into a <div> tag since we
2249 * can't really have a body-within-body.
2251 * @param $attary an array of attributes and values of <body>
2252 * @param $mailbox mailbox we're currently reading (for cid2http)
2253 * @param $message current message (for cid2http)
2254 * @param $id current message id (for cid2http)
2255 * @return a modified array of attributes to be set for <div>
2257 function sq_body2div($attary, $mailbox, $message, $id){
2258 $me = 'sq_body2div';
2259 $divattary = Array('class' => "'bodyclass'");
2261 $has_bgc_stl = $has_txt_stl = false;
2263 if (is_array($attary) && sizeof($attary) > 0){
2264 foreach ($attary as $attname=>$attvalue){
2265 $quotchar = substr($attvalue, 0, 1);
2266 $attvalue = str_replace($quotchar, "", $attvalue);
2269 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2270 $styledef .= "background-image: url('$attvalue'); ";
2273 $has_bgc_stl = true;
2274 $styledef .= "background-color: $attvalue; ";
2277 $has_txt_stl = true;
2278 $styledef .= "color: $attvalue; ";
2282 // Outlook defines a white bgcolor and no text color. This can lead to
2283 // white text on a white bg with certain themes.
2284 if ($has_bgc_stl && !$has_txt_stl) {
2285 $styledef .= "color: $text; ";
2287 if (strlen($styledef) > 0){
2288 $divattary{"style"} = "\"$styledef\"";
2295 * This is the main function and the one you should actually be calling.
2296 * There are several variables you should be aware of an which need
2297 * special description.
2299 * Since the description is quite lengthy, see it here:
2300 * http://linux.duke.edu/projects/mini/htmlfilter/
2302 * @param $body the string with HTML you wish to filter
2303 * @param $tag_list see description above
2304 * @param $rm_tags_with_content see description above
2305 * @param $self_closing_tags see description above
2306 * @param $force_tag_closing see description above
2307 * @param $rm_attnames see description above
2308 * @param $bad_attvals see description above
2309 * @param $add_attr_to_tag see description above
2310 * @param $message message object
2311 * @param $id message id
2312 * @return sanitized html safe to show on your pages.
2314 function sq_sanitize($body,
2316 $rm_tags_with_content,
2326 $me = 'sq_sanitize';
2327 $rm_tags = array_shift($tag_list);
2329 * Normalize rm_tags and rm_tags_with_content.
2331 @array_walk
($tag_list, 'sq_casenormalize');
2332 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2333 @array_walk
($self_closing_tags, 'sq_casenormalize');
2335 * See if tag_list is of tags to remove or tags to allow.
2336 * false means remove these tags
2337 * true means allow these tags
2340 $open_tags = Array();
2341 $trusted = "\n<!-- begin sanitized html -->\n";
2342 $skip_content = false;
2344 * Take care of netscape's stupid javascript entities like
2347 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2349 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2350 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2351 $free_content = substr($body, $curpos, $lt-$curpos);
2353 * Take care of <style>
2355 if ($tagname == "style" && $tagtype == 1){
2356 list($free_content, $curpos) =
2357 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2358 if ($free_content != FALSE){
2359 if ( !empty($attary) ) {
2360 $attary = sq_fixatts($tagname,
2370 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2371 $trusted .= $free_content;
2372 $trusted .= sq_tagprint($tagname, false, 2);
2376 if ($skip_content == false){
2377 $trusted .= $free_content;
2379 if ($tagname != FALSE){
2381 if ($skip_content == $tagname){
2383 * Got to the end of tag we needed to remove.
2386 $skip_content = false;
2388 if ($skip_content == false){
2389 if ($tagname == "body"){
2392 if (isset($open_tags{$tagname}) &&
2393 $open_tags{$tagname} > 0){
2394 $open_tags{$tagname}--;
2402 * $rm_tags_with_content
2404 if ($skip_content == false){
2406 * See if this is a self-closing type and change
2407 * tagtype appropriately.
2410 && in_array($tagname, $self_closing_tags)){
2414 * See if we should skip this tag and any content
2417 if ($tagtype == 1 &&
2418 in_array($tagname, $rm_tags_with_content)){
2419 $skip_content = $tagname;
2421 if (($rm_tags == false
2422 && in_array($tagname, $tag_list)) ||
2423 ($rm_tags == true &&
2424 !in_array($tagname, $tag_list))){
2428 * Convert body into div.
2430 if ($tagname == "body"){
2432 $attary = sq_body2div($attary, $mailbox,
2436 if (isset($open_tags{$tagname})){
2437 $open_tags{$tagname}++
;
2439 $open_tags{$tagname}=1;
2443 * This is where we run other checks.
2445 if (is_array($attary) && sizeof($attary) > 0){
2446 $attary = sq_fixatts($tagname,
2460 if ($tagname != false && $skip_content == false){
2461 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2466 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2467 if ($force_tag_closing == true){
2468 foreach ($open_tags as $tagname=>$opentimes){
2469 while ($opentimes > 0){
2470 $trusted .= '</' . $tagname . '>';
2476 $trusted .= "<!-- end sanitized html -->\n";
2481 * This is a wrapper function to call html sanitizing routines.
2483 * @param $body the body of the message
2484 * @param $id the id of the message
2488 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2489 * into internal SM compose links
2490 * (optional; default = TRUE)
2491 * @return a string with html safe to display in the browser.
2493 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2495 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2497 global $attachment_common_show_images, $view_unsafe_images,
2500 * Don't display attached images in HTML mode.
2504 $attachment_common_show_images = false;
2519 $rm_tags_with_content = Array(
2529 $self_closing_tags = Array(
2537 $force_tag_closing = true;
2539 $rm_attnames = Array(
2550 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2551 $bad_attvals = Array(
2554 "/^src|background/i" =>
2557 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2558 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2559 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2562 "\\1$secremoveimg\\2",
2563 "\\1$secremoveimg\\2",
2564 "\\1$secremoveimg\\2",
2567 "/^href|action/i" =>
2570 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2571 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2572 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2587 "/include-source/i",
2589 // position:relative can also be exploited
2590 // to put content outside of email body area
2591 // and position:fixed is similarly exploitable
2592 // as position:absolute, so we'll remove it
2595 // Does this screw up legitimate HTML messages?
2596 // If so, the only fix I see is to allow position
2597 // attributes (any values? I think we still have
2598 // to block static and fixed) only if $use_iframe
2599 // is enabled (1.5.0+)
2601 // was: "/position\s*:\s*absolute/i",
2605 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2606 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2607 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2608 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2609 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si",
2628 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2629 // images off by default.
2630 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
2632 if (!$view_unsafe_images){
2634 * Remove any references to http/https if view_unsafe_images set
2637 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2638 '/^([\'\"])\s*https*:.*([\'\"])/si');
2639 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2640 "\\1$secremoveimg\\1");
2641 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2642 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2643 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2644 "url(\\1$secremoveimg\\1)");
2647 $add_attr_to_tag = Array(
2649 Array('target'=>'"_blank"',
2650 'title'=>'"'._("This external link will open in a new window").'"'
2653 $trusted = sq_sanitize($body,
2655 $rm_tags_with_content,
2665 if (strpos($trusted,$secremoveimg)){
2666 $has_unsafe_images = true;
2669 // we want to parse mailto's in HTML output, change to SM compose links
2670 // this is a modified version of code from url_parser.php... but Marc is
2671 // right: we need a better filtering implementation; adding this randomly
2672 // here is not a great solution
2674 if ($take_mailto_links) {
2675 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2676 global $MailTo_PReg_Match;
2677 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2678 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2679 foreach ($regs[0] as $i => $mailto_before) {
2680 $mailto_params = $regs[10][$i];
2681 // get rid of any tailing quote since we have to add send_to to the end
2683 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2684 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2685 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2686 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2688 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2689 $to = 'to=' . $regs[1][$i];
2690 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2691 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2693 if ($mailto_params) //already some params, append to them
2694 $mailto_params .= '&' . $to;
2696 $mailto_params .= '?' . $to;
2700 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2702 // we'll already have target=_blank, no need to allow comp_in_new
2703 // here (which would be a lot more work anyway)
2705 global $compose_new_win;
2706 $temp_comp_in_new = $compose_new_win;
2707 $compose_new_win = 0;
2708 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2709 $compose_new_win = $temp_comp_in_new;
2711 // remove <a href=" and anything after the next quote (we only
2712 // need the uri, not the link HTML) in compose uri
2714 $comp_uri = substr($comp_uri, 9);
2715 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2716 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2725 * function SendDownloadHeaders - send file to the browser
2727 * Original Source: SM core src/download.php
2728 * moved here to make it available to other code, and separate
2729 * front end from back end functionality.
2731 * @param string $type0 first half of mime type
2732 * @param string $type1 second half of mime type
2733 * @param string $filename filename to tell the browser for downloaded file
2734 * @param boolean $force whether to force the download dialog to pop
2735 * @param optional integer $filesize send the Content-Header and length to the browser
2738 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2739 global $languages, $squirrelmail_language;
2740 $isIE = $isIE6plus = false;
2742 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2744 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2745 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2749 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2750 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2754 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2755 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2757 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2759 $filename = preg_replace('/[\\\\\/:*?"<>|;]/', '_', str_replace(' ', ' ', $filename));
2762 // A Pox on Microsoft and it's Internet Explorer!
2764 // IE has lots of bugs with file downloads.
2765 // It also has problems with SSL. Both of these cause problems
2766 // for us in this function.
2768 // See this article on Cache Control headers and SSL
2769 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2771 // The best thing you can do for IE is to upgrade to the latest
2773 //set all the Cache Control Headers for IE
2775 $filename=rawurlencode($filename);
2776 header ("Pragma: public");
2777 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2778 header ("Cache-Control: post-check=0, pre-check=0", false);
2779 header ("Cache-Control: private");
2781 //set the inline header for IE, we'll add the attachment header later if we need it
2782 header ("Content-Disposition: inline; filename=$filename");
2786 // Try to show in browser window
2787 header ("Content-Disposition: inline; filename=\"$filename\"");
2788 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2790 // Try to pop up the "save as" box
2792 // IE makes this hard. It pops up 2 save boxes, or none.
2793 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2794 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2795 // But, according to Microsoft, it is "RFC compliant but doesn't
2796 // take into account some deviations that allowed within the
2797 // specification." Doesn't that mean RFC non-compliant?
2798 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2800 // all browsers need the application/octet-stream header for this
2801 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2803 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2804 // Do not have quotes around filename, but that applied to
2805 // "attachment"... does it apply to inline too?
2806 header ("Content-Disposition: attachment; filename=\"$filename\"");
2808 if ($isIE && !$isIE6plus) {
2809 // This combination seems to work mostly. IE 5.5 SP 1 has
2810 // known issues (see the Microsoft Knowledge Base)
2812 // This works for most types, but doesn't work with Word files
2813 header ("Content-Type: application/download; name=\"$filename\"");
2814 header ("Content-Type: application/force-download; name=\"$filename\"");
2815 // These are spares, just in case. :-)
2816 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2817 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2818 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2820 // This is to prevent IE for MIME sniffing and auto open a file in IE
2821 header ("Content-Type: application/force-download; name=\"$filename\"");
2823 // another application/octet-stream forces download for Netscape
2824 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2828 //send the content-length header if the calling function provides it
2829 if ($filesize > 0) {
2830 header("Content-Length: $filesize");
2833 } // end fn SendDownloadHeaders