fdd6da537662858e51a06651bee682138d4f7072
6 * This contains the functions necessary to detect and decode MIME
9 * @copyright © 1999-2007 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 $msg->is_flagged
= true;
97 if (strtolower($flag) == '$mdnsent') {
98 $msg->is_mdnsent
= true;
106 // listEntities($msg);
112 /* This starts the parsing of a particular structure. It is called recursively,
113 * so it can be passed different structures. It returns an object of type
115 * First, it checks to see if it is a multipart message. If it is, then it
116 * handles that as it sees is necessary. If it is just a regular entity,
117 * then it parses it and adds the necessary header information (by calling out
118 * to mime_get_elements()
121 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
122 /* Do a bit of error correction. If we couldn't find the entity id, just guess
123 * that it is the first one. That is usually the case anyway.
127 $cmd = "FETCH $id BODY[]";
129 $cmd = "FETCH $id BODY[$ent_id]";
132 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
134 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
136 $topline = trim(array_shift($data));
137 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH.*/i', $topline)) ;
139 $wholemessage = implode('', $data);
140 if (ereg('\\{([^\\}]*)\\}', $topline, $regs)) {
141 $ret = substr($wholemessage, 0, $regs[1]);
142 /* There is some information in the content info header that could be important
143 * in order to parse html messages. Let's get them here.
145 // if ($ret{0} == '<') {
146 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
148 } else if (ereg('"([^"]*)"', $topline, $regs)) {
150 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
151 $ret = $wholemessage;
153 global $where, $what, $mailbox, $passed_id, $startMessage;
154 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
155 if (isset($where) && isset($what)) {
156 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
158 $par .= '&startMessage=' . $startMessage . '&show_more=0';
160 $par .= '&response=' . urlencode($response) .
161 '&message=' . urlencode($message) .
162 '&topline=' . urlencode($topline);
165 '<table width="80%"><tr>' .
166 '<tr><td colspan="2">' .
167 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
169 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
170 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
171 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
172 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
173 "</table><br /></tt></font><hr />";
175 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
177 $wholemessage = implode('', $data);
179 $ret = $wholemessage;
184 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout') {
186 /* Don't kill the connection if the browser is over a dialup
187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
191 if (!ini_get('safe_mode')) {
194 /* in case of base64 encoded attachments, do not buffer them.
195 Instead, echo the decoded attachment directly to screen */
196 if (strtolower($encoding) == 'base64') {
198 $query = "FETCH $id BODY[]";
200 $query = "FETCH $id BODY[$ent_id]";
202 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
204 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
205 if (is_resource($rStream)) {
206 fputs($rStream,decodeBody($body,$encoding));
208 echo decodeBody($body, $encoding);
213 TODO, use the same method for quoted printable.
214 However, I assume that quoted printable attachments aren't that large
215 so the performancegain / memory usage drop will be minimal.
216 If we decide to add that then we need to adapt sqimap_fread because
217 we need to split te result on \n and fread doesn't stop at \n. That
218 means we also should provide $results from sqimap_fread (by ref) to
219 te function and set $no_return to false. The $filter function for
220 quoted printable should handle unsetting of $results.
223 TODO 2: find out how we write to the output stream php://stdout. fwrite
224 doesn't work because 'php://stdout isn't a stream.
230 /* -[ END MIME DECODING ]----------------------------------------------------------- */
232 /* This is here for debugging purposes. It will print out a list
233 * of all the entity IDs that are in the $message object.
235 function listEntities ($message) {
237 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br />';
238 for ($i = 0; isset($message->entities
[$i]); $i++
) {
240 $msg = listEntities($message->entities
[$i]);
250 function getPriorityStr($priority) {
251 $priority_level = substr($priority,0,1);
253 switch($priority_level) {
254 /* Check for a higher then normal priority. */
257 $priority_string = _("High");
260 /* Check for a lower then normal priority. */
263 $priority_string = _("Low");
266 /* Check for a normal priority. */
269 $priority_level = '3';
270 $priority_string = _("Normal");
274 return $priority_string;
277 /* returns a $message object for a particular entity id */
278 function getEntity ($message, $ent_id) {
279 return $message->getEntity($ent_id);
283 * Extracted from strings.php 23/03/2002
286 function translateText(&$body, $wrap_at, $charset) {
287 global $where, $what; /* from searching */
288 global $color; /* color theme */
290 // require_once(SM_PATH . 'functions/url_parser.php');
292 $body_ary = explode("\n", $body);
293 for ($i=0; $i < count($body_ary); $i++
) {
294 $line = $body_ary[$i];
295 if (strlen($line) - 2 >= $wrap_at) {
296 sqWordWrap($line, $wrap_at, $charset);
298 $line = charset_decode($charset, $line);
299 $line = str_replace("\t", ' ', $line);
308 if ($line[$pos] == ' ') {
310 } else if (strpos($line, '>', $pos) === $pos) {
319 $line = '<span class="quote1">' . $line . '</span>';
321 $line = '<span class="quote2">' . $line . '</span>';
324 $body_ary[$i] = $line;
326 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
330 * This returns a parsed string called $body. That string can then
331 * be displayed as the actual message in the HTML. It contains
332 * everything needed, including HTML Tags, Attachments at the
335 * Since 1.2.0 function uses message_body hook.
336 * Till 1.3.0 function included output of formatAttachments().
338 * @param resource $imap_stream imap connection resource
339 * @param object $message squirrelmail message object
340 * @param array $color squirrelmail color theme array
341 * @param integer $wrap_at number of characters per line
342 * @param string $ent_num (since 1.3.0) message part id
343 * @param integer $id (since 1.3.0) message id
344 * @param string $mailbox (since 1.3.0) imap folder name
345 * @param boolean $clean (since 1.5.1) Do not output stuff that's irrelevant for the printable version.
346 * @return string html formated message text
348 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX', $clean=FALSE) {
349 /* This if statement checks for the entity to show as the
350 * primary message. To add more of them, just put them in the
351 * order that is their priority.
353 global $startMessage, $languages, $squirrelmail_language,
354 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
355 $use_iframe, $iframe_height, $download_and_unsafe_link,
356 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
359 // workaround for not updated config.php
360 if (! isset($use_iframe)) $use_iframe = false;
362 // If there's no "view_unsafe_images" setting in the user's preferences,
363 // turn unsafe images off by default.
364 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
365 $view_unsafe_images = false;
369 $urlmailbox = urlencode($mailbox);
370 $body_message = getEntity($message, $ent_num);
371 if (($body_message->header
->type0
== 'text') ||
372 ($body_message->header
->type0
== 'rfc822')) {
373 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
374 $body = decodeBody($body, $body_message->header
->encoding
);
376 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
377 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
378 if (mb_detect_encoding($body) != 'ASCII') {
379 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
383 /* As of 1.5.2, $body is passed (and modified) by reference */
384 do_hook('message_body', $body);
386 /* If there are other types that shouldn't be formatted, add
390 if ($body_message->header
->type1
== 'html') {
391 if ($show_html_default <> 1) {
392 $entity_conv = array(' ' => ' ',
401 $body = strtr($body, $entity_conv);
402 $body = strip_tags($body);
404 translateText($body, $wrap_at,
405 $body_message->header
->getParameter('charset'));
406 } elseif ($use_iframe && ! $clean) {
407 // $clean is used to remove iframe in printable view.
410 * If we don't add html message between iframe tags,
411 * we must detect unsafe images and modify $has_unsafe_images.
413 $html_body = magicHTML($body, $id, $message, $mailbox);
414 // Convert character set in order to display html mails in different character set
415 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
417 // creating iframe url
418 $iframeurl=sqm_baseuri().'src/view_html.php?'
419 . 'mailbox=' . $urlmailbox
420 . '&passed_id=' . $id
421 . '&ent_id=' . $ent_num
422 . '&view_unsafe_images=' . (int) $view_unsafe_images;
425 $oTemplate->assign('iframe_url', $iframeurl);
426 $oTemplate->assign('iframe_height', $iframe_height);
427 $oTemplate->assign('html_body', $html_body);
429 $body = $oTemplate->fetch('read_html_iframe.tpl');
431 // old way of html rendering
433 * convert character set. charset_decode does not remove html special chars
434 * applied by magicHTML functions and does not sanitize them second time if
435 * fourth argument is true.
437 $charset = $body_message->header
->getParameter('charset');
438 if (!empty($charset)) {
439 $body = charset_decode($charset,$body,false,true);
441 $body = magicHTML($body, $id, $message, $mailbox);
444 translateText($body, $wrap_at,
445 $body_message->header
->getParameter('charset'));
448 // if this is the clean display (i.e. printer friendly), stop here.
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. Separated from
509 * formatAttachments() below so that the same array can be given to the
510 * print-friendly version.
513 * @param object $message SquirrelMail message object
514 * @param array $exclude_id message parts that are not attachments.
515 * @param string $mailbox mailbox name
516 * @param integer $id message id
518 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
519 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
521 $att_ar = $message->getAttachments($exclude_id);
522 $urlMailbox = urlencode($mailbox);
524 $attachments = array();
525 foreach ($att_ar as $att) {
526 $ent = $att->entity_id
;
527 $header = $att->header
;
528 $type0 = strtolower($header->type0
);
529 $type1 = strtolower($header->type1
);
532 $links['download link']['text'] = _("Download");
533 $links['download link']['href'] = $base_uri .
534 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
536 if ($type0 =='message' && $type1 == 'rfc822') {
537 $default_page = $base_uri . 'src/read_body.php';
538 $rfc822_header = $att->rfc822_header
;
539 $filename = $rfc822_header->subject
;
540 if (trim( $filename ) == '') {
541 $filename = 'untitled-[' . $ent . ']' ;
543 $from_o = $rfc822_header->from
;
544 if (is_object($from_o)) {
545 $from_name = decodeHeader($from_o->getAddress(false));
546 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
547 // something weird happens when a digest message is opened and you return to the digest
548 // now the from object is part of an array. Probably the parseHeader call overwrites the info
549 // retrieved from the bodystructure in a different way. We need to fix this later.
550 // possible starting point, do not fetch header we already have and inspect how
551 // the rfc822_header object behaves.
552 $from_name = decodeHeader($from_o[0]->getAddress(false));
554 $from_name = _("Unknown sender");
556 $description = _("From").': '.$from_name;
558 $default_page = $base_uri . 'src/download.php';
559 $filename = $att->getFilename();
560 if ($header->description
) {
561 $description = decodeHeader($header->description
);
567 $display_filename = $filename;
568 if (isset($passed_ent_id)) {
569 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
571 $passed_ent_id_link = '';
573 $defaultlink = $default_page . "?startMessage=$startMessage"
574 . "&passed_id=$id&mailbox=$urlMailbox"
575 . '&ent_id='.$ent.$passed_ent_id_link;
576 if ($where && $what) {
577 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
579 // IE does make use of mime content sniffing. Forcing a download
580 // prohibit execution of XSS inside an application/octet-stream attachment
581 if ($type0 == 'application' && $type1 == 'octet-stream') {
582 $defaultlink .= '&absolute_dl=true';
585 /* This executes the attachment hook with a specific MIME-type.
586 * If that doesn't have results, it tries if there's a rule
587 * for a more generic type. Finally, a hook for ALL attachment
588 * types is run as well.
590 // First remember the default link.
591 $defaultlink_orig = $defaultlink;
593 /* The API for this hook has changed as of 1.5.2 so that all plugin
594 arguments are passed in an array instead of each their own plugin
595 argument, and arguments are passed by reference, so instead of
596 returning any changes, changes should simply be made to the original
597 arguments themselves. */
598 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
599 &$defaultlink, &$display_filename, &$where, &$what);
600 do_hook("attachment $type0/$type1", $temp);
601 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
602 /* The API for this hook has changed as of 1.5.2 so that all plugin
603 arguments are passed in an array instead of each their own plugin
604 argument, and arguments are passed by reference, so instead of
605 returning any changes, changes should simply be made to the original
606 arguments themselves. */
607 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
608 &$defaultlink, &$display_filename, &$where, &$what);
609 do_hook("attachment $type0/*", $temp);
611 /* The API for this hook has changed as of 1.5.2 so that all plugin
612 arguments are passed in an array instead of each their own plugin
613 argument, and arguments are passed by reference, so instead of
614 returning any changes, changes should simply be made to the original
615 arguments themselves. */
616 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
617 &$defaultlink, &$display_filename, &$where, &$what);
618 // Do not let a generic plugin change the default link if a more
619 // specialized one already did it...
620 if ($defaultlink != $defaultlink_orig) {
624 do_hook("attachment */*", $temp);
626 $this_attachment = array();
627 $this_attachment['Name'] = decodeHeader($display_filename);
628 $this_attachment['Description'] = $description;
629 $this_attachment['DefaultHREF'] = $defaultlink;
630 $this_attachment['DownloadHREF'] = $links['download link']['href'];
631 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
632 $this_attachment['Size'] = $header->size
;
633 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
634 $this_attachment['OtherLinks'] = array();
635 foreach ($links as $val) {
636 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
638 if (empty($val['text']) && empty($val['extra']))
642 $temp['HREF'] = $val['href'];
643 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
644 $this_attachment['OtherLinks'][] = $temp;
646 $attachments[] = $this_attachment;
655 * Displays attachment links and information
657 * Since 1.3.0 function is not included in formatBody() call.
659 * Since 1.0.2 uses attachment $type0/$type1 hook.
660 * Since 1.2.5 uses attachment $type0/* hook.
661 * Since 1.5.0 uses attachments_bottom hook.
662 * Since 1.5.2 uses templates and does *not* return a value.
664 * @param object $message SquirrelMail message object
665 * @param array $exclude_id message parts that are not attachments.
666 * @param string $mailbox mailbox name
667 * @param integer $id message id
669 function formatAttachments($message, $exclude_id, $mailbox, $id) {
672 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
674 $oTemplate->assign('attachments', $attach);
675 $oTemplate->display('read_attachments.tpl');
678 function sqimap_base64_decode(&$string) {
680 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
681 // fly decoding (to reduce memory usage) you have to check if the
682 // data has incomplete pairs
684 // Remove the noise in order to check if the 4 bytes pairs are complete
685 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
688 $iMod = strlen($string) %
4;
690 $sStringRem = substr($string,-$iMod);
691 // Check if $sStringRem contains padding characters
692 if (substr($sStringRem,-1) != '=') {
693 $string = substr($string,0,-$iMod);
698 $string = base64_decode($string);
703 * Decodes encoded message body
705 * This function decodes the body depending on the encoding type.
706 * Currently quoted-printable and base64 encodings are supported.
707 * decode_body hook was added to this function in 1.4.2/1.5.0
708 * @param string $body encoded message body
709 * @param string $encoding used encoding
710 * @return string decoded string
713 function decodeBody($body, $encoding) {
715 $body = str_replace("\r\n", "\n", $body);
716 $encoding = strtolower($encoding);
718 $encoding_handler = do_hook('decode_body', $encoding);
721 // plugins get first shot at decoding the body
723 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
724 $body = $encoding_handler('decode', $body);
726 } elseif ($encoding == 'quoted-printable' ||
727 $encoding == 'quoted_printable') {
729 * quoted_printable_decode() function is broken in older
730 * php versions. Text with \r\n decoding was fixed only
731 * in php 4.3.0. Minimal code requirement 4.0.4 +
732 * str_replace("\r\n", "\n", $body); call.
734 $body = quoted_printable_decode($body);
735 } elseif ($encoding == 'base64') {
736 $body = base64_decode($body);
739 // All other encodings are returned raw.
746 * This function decodes strings that are encoded according to
747 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
748 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
750 * @param string $string header string that has to be made readable
751 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
752 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
753 * @param boolean $decide decide if string can be utfencoded. defaults to false
754 * @return string decoded header string
756 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
757 global $languages, $squirrelmail_language,$default_charset;
758 if (is_array($string)) {
759 $string = implode("\n", $string);
762 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
763 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
764 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
765 // Do we need to return at this point?
772 $aString = explode(' ',$string);
774 foreach ($aString as $chunk) {
775 if ($encoded && $chunk === '') {
777 } elseif ($chunk === '') {
782 /* if encoded words are not separated by a linear-space-white we still catch them */
785 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
786 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
787 if ($iLastMatch !== $j) {
797 $ret .= htmlspecialchars($res[1]);
801 $encoding = ucfirst($res[3]);
803 /* decide about valid decoding */
804 if ($decide && is_conversion_safe($res[2])) {
806 $can_be_encoded=true;
808 $can_be_encoded=false;
813 $replace = base64_decode($res[4]);
815 if ($can_be_encoded) {
816 /* convert string to different charset,
817 * if functions asks for it (usually in compose)
819 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
821 // convert string to html codes in order to display it
822 $ret .= charset_decode($res[2],$replace);
826 $replace = htmlspecialchars($replace);
832 $replace = str_replace('_', ' ', $res[4]);
833 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
836 if ($can_be_encoded) {
837 /* convert string to different charset,
838 * if functions asks for it (usually in compose)
840 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
842 // convert string to html codes in order to display it
843 $replace = charset_decode($res[2], $replace);
847 $replace = htmlspecialchars($replace);
866 if (!$encoded && $htmlsafe) {
867 $ret .= htmlspecialchars($chunk);
873 /* remove the first added space */
876 $ret = substr($ret,5);
878 $ret = substr($ret,1);
888 * Function uses XTRA_CODE _encodeheader function, if such function exists.
890 * Function uses Q encoding by default and encodes a string according to RFC
891 * 1522 for use in headers if it contains 8-bit characters or anything that
892 * looks like it should be encoded.
894 * Function switches to B encoding and encodeHeaderBase64() function, if
895 * string is 8bit and multibyte character set supported by mbstring extension
896 * is used. It can cause E_USER_NOTICE errors, if interface is used with
897 * multibyte character set unsupported by mbstring extension.
899 * @param string $string header string, that has to be encoded
900 * @return string quoted-printable encoded string
901 * @todo make $mb_charsets system wide constant
903 function encodeHeader ($string) {
904 global $default_charset, $languages, $squirrelmail_language;
906 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
907 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
908 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
911 // Use B encoding for multibyte charsets
912 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
913 if (in_array($default_charset,$mb_charsets) &&
914 in_array($default_charset,sq_mb_list_encodings()) &&
915 sq_is8bit($string)) {
916 return encodeHeaderBase64($string,$default_charset);
917 } elseif (in_array($default_charset,$mb_charsets) &&
918 sq_is8bit($string) &&
919 ! in_array($default_charset,sq_mb_list_encodings())) {
920 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
921 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
924 // Encode only if the string contains 8-bit characters or =?
925 $j = strlen($string);
926 $max_l = 75 - strlen($default_charset) - 7;
929 $iEncStart = $enc_init = false;
930 $cur_l = $iOffset = 0;
931 for($i = 0; $i < $j; ++
$i) {
940 if ($iEncStart === false) {
944 if ($cur_l > ($max_l-2)) {
945 /* if there is an stringpart that doesn't need encoding, add it */
946 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
947 $aRet[] = "=?$default_charset?Q?$ret?=";
953 $ret .= sprintf("=%02X",ord($string{$i}));
958 if ($iEncStart !== false) {
959 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
960 $aRet[] = "=?$default_charset?Q?$ret?=";
968 if ($iEncStart !== false) {
970 if ($cur_l > $max_l) {
971 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
972 $aRet[] = "=?$default_charset?Q?$ret?=";
983 $k = ord($string{$i});
985 if ($iEncStart === false) {
986 // do not start encoding in the middle of a string, also take the rest of the word.
987 $sLeadString = substr($string,0,$i);
988 $aLeadString = explode(' ',$sLeadString);
989 $sToBeEncoded = array_pop($aLeadString);
990 $iEncStart = $i - strlen($sToBeEncoded);
991 $ret .= $sToBeEncoded;
992 $cur_l +
= strlen($sToBeEncoded);
995 /* first we add the encoded string that reached it's max size */
996 if ($cur_l > ($max_l-2)) {
997 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
998 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1005 $ret .= sprintf("=%02X", $k);
1007 if ($iEncStart !== false) {
1009 if ($cur_l > $max_l) {
1010 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1011 $aRet[] = "=?$default_charset?Q?$ret?=";
1017 $ret .= $string{$i};
1026 if ($iEncStart !== false) {
1027 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1028 $aRet[] = "=?$default_charset?Q?$ret?=";
1030 $aRet[] = substr($string,$iOffset);
1032 $string = implode('',$aRet);
1038 * Encodes string according to rfc2047 B encoding header formating rules
1040 * It is recommended way to encode headers with character sets that store
1041 * symbols in more than one byte.
1043 * Function requires mbstring support. If required mbstring functions are missing,
1044 * function returns false and sets E_USER_WARNING level error message.
1046 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1047 * that mbstring functions will generate E_WARNING errors, if unsupported
1048 * character set is used. mb_encode_mimeheader function provided by php
1049 * mbstring extension is not used in order to get better control of header
1052 * Used php code functions - function_exists(), trigger_error(), strlen()
1053 * (is used with charset names and base64 strings). Used php mbstring
1054 * functions - mb_strlen and mb_substr.
1056 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1057 * encoding), rfc 2822 (header folding)
1059 * @param string $string header string that must be encoded
1060 * @param string $charset character set. Must be supported by mbstring extension.
1061 * Use sq_mb_list_encodings() to detect supported charsets.
1062 * @return string string encoded according to rfc2047 B encoding formating rules
1064 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1065 * @todo Do we want to control max length of header?
1066 * @todo Do we want to control EOL (end-of-line) marker?
1067 * @todo Do we want to translate error message?
1069 function encodeHeaderBase64($string,$charset) {
1071 * Check mbstring function requirements.
1073 if (! function_exists('mb_strlen') ||
1074 ! function_exists('mb_substr')) {
1075 // set E_USER_WARNING
1076 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1081 // initial return array
1085 * header length = 75 symbols max (same as in encodeHeader)
1086 * remove $charset length
1087 * remove =? ? ?= (5 chars)
1088 * remove 2 more chars (\r\n ?)
1090 $iMaxLength = 75 - strlen($charset) - 7;
1092 // set first character position
1095 // loop through all characters. count characters and not bytes.
1096 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1097 // encode string from starting character to current character.
1098 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1100 // Check encoded string length
1101 if(strlen($encoded_string)>$iMaxLength) {
1102 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1103 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1105 // set new starting character
1106 $iStartCharNum = $iCharNum-1;
1108 // encode last char (in case it is last character in string)
1109 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1110 } // if string is shorter than max length - add next character
1113 // add last encoded string to array
1114 $aRet[] = $encoded_string;
1116 // set initial return string
1119 // loop through encoded strings
1120 foreach($aRet as $string) {
1121 // TODO: Do we want to control EOL (end-of-line) marker
1122 if ($sRet!='') $sRet.= " ";
1124 // add header tags and encoded string to return string
1125 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1131 /* This function trys to locate the entity_id of a specific mime element */
1132 function find_ent_id($id, $message) {
1133 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1134 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1135 $ret = find_ent_id($id, $message->entities
[$i]);
1137 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1138 // if (sq_check_save_extension($message->entities[$i])) {
1139 return $message->entities
[$i]->entity_id
;
1141 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1143 * This is part of a fix for Outlook Express 6.x generating
1144 * cid URLs without creating content-id headers
1147 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1148 return $message->entities
[$i]->entity_id
;
1156 function sq_check_save_extension($message) {
1157 $filename = $message->getFilename();
1158 $ext = substr($filename, strrpos($filename,'.')+
1);
1159 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1160 return in_array($ext, $save_extensions);
1165 ** HTMLFILTER ROUTINES
1169 * This function checks attribute values for entity-encoded values
1170 * and returns them translated into 8-bit strings so we can run
1173 * @param $attvalue A string to run entity check against.
1174 * @return Nothing, modifies a reference value.
1176 function sq_defang(&$attvalue){
1179 * Skip this if there aren't ampersands or backslashes.
1181 if (strpos($attvalue, '&') === false
1182 && strpos($attvalue, '\\') === false){
1186 // before deent, translate the dangerous unicode characters and ... to safe values
1187 // otherwise the regular expressions do not match.
1193 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1194 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1195 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1196 } while ($m == true);
1197 $attvalue = stripslashes($attvalue);
1201 * Kill any tabs, newlines, or carriage returns. Our friends the
1202 * makers of the browser with 95% market value decided that it'd
1203 * be funny to make "java[tab]script" be just as good as "javascript".
1205 * @param attvalue The attribute value before extraneous spaces removed.
1206 * @return attvalue Nothing, modifies a reference value.
1208 function sq_unspace(&$attvalue){
1210 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1211 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1212 Array('', '', '', '', ''), $attvalue);
1217 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1218 * IE as regular characters.
1220 * @param attvalue The attribute value before dangerous characters are translated.
1221 * @return attvalue Nothing, modifies a reference value.
1222 * @author Marc Groot Koerkamp.
1224 function sq_fixIE_idiocy(&$attvalue) {
1226 $attvalue = str_replace("\0", "", $attvalue);
1228 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1230 // IE has the evil habit of accepting every possible value for the attribute expression.
1231 // The table below contains characters which are parsed by IE if they are used in the "expression"
1233 $aDangerousCharsReplacementTable = array(
1234 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1235 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1236 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1237 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1238 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1239 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1240 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1241 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1242 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1243 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1244 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1245 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1246 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1247 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1248 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1249 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1250 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1251 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1252 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1253 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1254 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1255 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1256 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1257 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1258 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1259 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1260 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1261 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1262 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1263 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1264 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1265 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1266 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1267 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1268 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1269 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1270 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1271 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1272 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1273 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1274 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1275 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1276 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1277 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1278 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1279 "\xCA\x9F", /* L UNICODE IPA Extension */
1280 "\xCA\x80", /* R UNICODE IPA Extension */
1281 "\xC9\xB4"), /* N UNICODE IPA Extension */
1282 array('l', 'l', 'r','r','n','n',
1283 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1284 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1285 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1286 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1288 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1290 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1294 * This function returns the final tag out of the tag name, an array
1295 * of attributes, and the type of the tag. This function is called by
1296 * sq_sanitize internally.
1298 * @param $tagname the name of the tag.
1299 * @param $attary the array of attributes and their values
1300 * @param $tagtype The type of the tag (see in comments).
1301 * @return a string with the final tag representation.
1303 function sq_tagprint($tagname, $attary, $tagtype){
1304 $me = 'sq_tagprint';
1307 $fulltag = '</' . $tagname . '>';
1309 $fulltag = '<' . $tagname;
1310 if (is_array($attary) && sizeof($attary)){
1312 while (list($attname, $attvalue) = each($attary)){
1313 array_push($atts, "$attname=$attvalue");
1315 $fulltag .= ' ' . join(" ", $atts);
1326 * A small helper function to use with array_walk. Modifies a by-ref
1327 * value and makes it lowercase.
1329 * @param $val a value passed by-ref.
1330 * @return void since it modifies a by-ref value.
1332 function sq_casenormalize(&$val){
1333 $val = strtolower($val);
1337 * This function skips any whitespace from the current position within
1338 * a string and to the next non-whitespace value.
1340 * @param $body the string
1341 * @param $offset the offset within the string where we should start
1342 * looking for the next non-whitespace character.
1343 * @return the location within the $body where the next
1344 * non-whitespace char is located.
1346 function sq_skipspace($body, $offset){
1347 $me = 'sq_skipspace';
1348 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1349 if (sizeof($matches{1})){
1350 $count = strlen($matches{1});
1357 * This function looks for the next character within a string. It's
1358 * really just a glorified "strpos", except it catches if failures
1361 * @param $body The string to look for needle in.
1362 * @param $offset Start looking from this position.
1363 * @param $needle The character/string to look for.
1364 * @return location of the next occurance of the needle, or
1365 * strlen($body) if needle wasn't found.
1367 function sq_findnxstr($body, $offset, $needle){
1368 $me = 'sq_findnxstr';
1369 $pos = strpos($body, $needle, $offset);
1370 if ($pos === FALSE){
1371 $pos = strlen($body);
1377 * This function takes a PCRE-style regexp and tries to match it
1378 * within the string.
1380 * @param $body The string to look for needle in.
1381 * @param $offset Start looking from here.
1382 * @param $reg A PCRE-style regex to match.
1383 * @return Returns a false if no matches found, or an array
1384 * with the following members:
1385 * - integer with the location of the match within $body
1386 * - string with whatever content between offset and the match
1387 * - string with whatever it is we matched
1389 function sq_findnxreg($body, $offset, $reg){
1390 $me = 'sq_findnxreg';
1393 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1394 if (!isset($matches{0}) ||
!$matches{0}){
1397 $retarr{0} = $offset +
strlen($matches{1});
1398 $retarr{1} = $matches{1};
1399 $retarr{2} = $matches{2};
1405 * This function looks for the next tag.
1407 * @param $body String where to look for the next tag.
1408 * @param $offset Start looking from here.
1409 * @return false if no more tags exist in the body, or
1410 * an array with the following members:
1411 * - string with the name of the tag
1412 * - array with attributes and their values
1413 * - integer with tag type (1, 2, or 3)
1414 * - integer where the tag starts (starting "<")
1415 * - integer where the tag ends (ending ">")
1416 * first three members will be false, if the tag is invalid.
1418 function sq_getnxtag($body, $offset){
1419 $me = 'sq_getnxtag';
1420 if ($offset > strlen($body)){
1423 $lt = sq_findnxstr($body, $offset, "<");
1424 if ($lt == strlen($body)){
1429 * blah blah <tag attribute="value">
1432 $pos = sq_skipspace($body, $lt+
1);
1433 if ($pos >= strlen($body)){
1434 return Array(false, false, false, $lt, strlen($body));
1437 * There are 3 kinds of tags:
1438 * 1. Opening tag, e.g.:
1440 * 2. Closing tag, e.g.:
1442 * 3. XHTML-style content-less tag, e.g.:
1443 * <img src="blah" />
1446 switch (substr($body, $pos, 1)){
1453 * A comment or an SGML declaration.
1455 if (substr($body, $pos+
1, 2) == "--"){
1456 $gt = strpos($body, "-->", $pos);
1458 $gt = strlen($body);
1462 return Array(false, false, false, $lt, $gt);
1464 $gt = sq_findnxstr($body, $pos, ">");
1465 return Array(false, false, false, $lt, $gt);
1470 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1480 * Look for next [\W-_], which will indicate the end of the tag name.
1482 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1483 if ($regary == false){
1484 return Array(false, false, false, $lt, strlen($body));
1486 list($pos, $tagname, $match) = $regary;
1487 $tagname = strtolower($tagname);
1490 * $match can be either of these:
1491 * '>' indicating the end of the tag entirely.
1492 * '\s' indicating the end of the tag name.
1493 * '/' indicating that this is type-3 xhtml tag.
1495 * Whatever else we find there indicates an invalid tag.
1500 * This is an xhtml-style tag with a closing / at the
1501 * end, like so: <img src="blah" />. Check if it's followed
1502 * by the closing bracket. If not, then this tag is invalid
1504 if (substr($body, $pos, 2) == "/>"){
1508 $gt = sq_findnxstr($body, $pos, ">");
1509 $retary = Array(false, false, false, $lt, $gt);
1513 return Array($tagname, false, $tagtype, $lt, $pos);
1517 * Check if it's whitespace
1519 if (!preg_match('/\s/', $match)){
1521 * This is an invalid tag! Look for the next closing ">".
1523 $gt = sq_findnxstr($body, $lt, ">");
1524 return Array(false, false, false, $lt, $gt);
1530 * At this point we're here:
1531 * <tagname attribute='blah'>
1534 * At this point we loop in order to find all attributes.
1540 while ($pos <= strlen($body)){
1541 $pos = sq_skipspace($body, $pos);
1542 if ($pos == strlen($body)){
1546 return Array(false, false, false, $lt, $pos);
1549 * See if we arrived at a ">" or "/>", which means that we reached
1550 * the end of the tag.
1553 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1557 $pos +
= strlen($matches{1});
1558 if ($matches{2} == "/>"){
1562 return Array($tagname, $attary, $tagtype, $lt, $pos);
1566 * There are several types of attributes, with optional
1567 * [:space:] between members.
1569 * attrname[:space:]=[:space:]'CDATA'
1571 * attrname[:space:]=[:space:]"CDATA"
1573 * attr[:space:]=[:space:]CDATA
1577 * We leave types 1 and 2 the same, type 3 we check for
1578 * '"' and convert to """ if needed, then wrap in
1579 * double quotes. Type 4 we convert into:
1582 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1583 if ($regary == false){
1585 * Looks like body ended before the end of tag.
1587 return Array(false, false, false, $lt, strlen($body));
1589 list($pos, $attname, $match) = $regary;
1590 $attname = strtolower($attname);
1592 * We arrived at the end of attribute name. Several things possible
1594 * '>' means the end of the tag and this is attribute type 4
1595 * '/' if followed by '>' means the same thing as above
1596 * '\s' means a lot of things -- look what it's followed by.
1597 * anything else means the attribute is invalid.
1602 * This is an xhtml-style tag with a closing / at the
1603 * end, like so: <img src="blah" />. Check if it's followed
1604 * by the closing bracket. If not, then this tag is invalid
1606 if (substr($body, $pos, 2) == "/>"){
1610 $gt = sq_findnxstr($body, $pos, ">");
1611 $retary = Array(false, false, false, $lt, $gt);
1615 $attary{$attname} = '"yes"';
1616 return Array($tagname, $attary, $tagtype, $lt, $pos);
1620 * Skip whitespace and see what we arrive at.
1622 $pos = sq_skipspace($body, $pos);
1623 $char = substr($body, $pos, 1);
1625 * Two things are valid here:
1626 * '=' means this is attribute type 1 2 or 3.
1627 * \w means this was attribute type 4.
1628 * anything else we ignore and re-loop. End of tag and
1629 * invalid stuff will be caught by our checks at the beginning
1634 $pos = sq_skipspace($body, $pos);
1636 * Here are 3 possibilities:
1637 * "'" attribute type 1
1638 * '"' attribute type 2
1639 * everything else is the content of tag type 3
1641 $quot = substr($body, $pos, 1);
1643 $regary = sq_findnxreg($body, $pos+
1, "\'");
1644 if ($regary == false){
1645 return Array(false, false, false, $lt, strlen($body));
1647 list($pos, $attval, $match) = $regary;
1649 $attary{$attname} = "'" . $attval . "'";
1650 } else if ($quot == '"'){
1651 $regary = sq_findnxreg($body, $pos+
1, '\"');
1652 if ($regary == false){
1653 return Array(false, false, false, $lt, strlen($body));
1655 list($pos, $attval, $match) = $regary;
1657 $attary{$attname} = '"' . $attval . '"';
1660 * These are hateful. Look for \s, or >.
1662 $regary = sq_findnxreg($body, $pos, "[\s>]");
1663 if ($regary == false){
1664 return Array(false, false, false, $lt, strlen($body));
1666 list($pos, $attval, $match) = $regary;
1668 * If it's ">" it will be caught at the top.
1670 $attval = preg_replace("/\"/s", """, $attval);
1671 $attary{$attname} = '"' . $attval . '"';
1673 } else if (preg_match("|[\w/>]|", $char)) {
1675 * That was attribute type 4.
1677 $attary{$attname} = '"yes"';
1680 * An illegal character. Find next '>' and return.
1682 $gt = sq_findnxstr($body, $pos, ">");
1683 return Array(false, false, false, $lt, $gt);
1689 * The fact that we got here indicates that the tag end was never
1690 * found. Return invalid tag indication so it gets stripped.
1692 return Array(false, false, false, $lt, strlen($body));
1696 * Translates entities into literal values so they can be checked.
1698 * @param $attvalue the by-ref value to check.
1699 * @param $regex the regular expression to check against.
1700 * @param $hex whether the entites are hexadecimal.
1701 * @return True or False depending on whether there were matches.
1703 function sq_deent(&$attvalue, $regex, $hex=false){
1707 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1708 preg_match_all($regex, $attvalue, $matches);
1709 if (is_array($matches) && sizeof($matches[0]) > 0){
1711 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1712 $numval = $matches[1][$i];
1714 $numval = hexdec($numval);
1716 $repl{$matches[0][$i]} = chr($numval);
1718 $attvalue = strtr($attvalue, $repl);
1726 * This function runs various checks against the attributes.
1728 * @param $tagname String with the name of the tag.
1729 * @param $attary Array with all tag attributes.
1730 * @param $rm_attnames See description for sq_sanitize
1731 * @param $bad_attvals See description for sq_sanitize
1732 * @param $add_attr_to_tag See description for sq_sanitize
1733 * @param $message message object
1734 * @param $id message id
1735 * @return Array with modified attributes.
1737 function sq_fixatts($tagname,
1747 while (list($attname, $attvalue) = each($attary)){
1749 * See if this attribute should be removed.
1751 foreach ($rm_attnames as $matchtag=>$matchattrs){
1752 if (preg_match($matchtag, $tagname)){
1753 foreach ($matchattrs as $matchattr){
1754 if (preg_match($matchattr, $attname)){
1755 unset($attary{$attname});
1762 * Workaround for IE quirks
1764 sq_fixIE_idiocy($attvalue);
1767 * Remove any backslashes, entities, and extraneous whitespace.
1770 $oldattvalue = $attvalue;
1771 sq_defang($attvalue);
1772 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1773 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1774 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1775 $attvalue = "idiocy";
1776 $attary{$attname} = $attvalue;
1778 sq_unspace($attvalue);
1781 * Now let's run checks on the attvalues.
1782 * I don't expect anyone to comprehend this. If you do,
1783 * get in touch with me so I can drive to where you live and
1784 * shake your hand personally. :)
1786 foreach ($bad_attvals as $matchtag=>$matchattrs){
1787 if (preg_match($matchtag, $tagname)){
1788 foreach ($matchattrs as $matchattr=>$valary){
1789 if (preg_match($matchattr, $attname)){
1791 * There are two arrays in valary.
1793 * Second one is replacements
1795 list($valmatch, $valrepl) = $valary;
1797 preg_replace($valmatch, $valrepl, $attvalue);
1798 if ($newvalue != $attvalue){
1799 $attary{$attname} = $newvalue;
1800 $attvalue = $newvalue;
1806 if ($attname == 'style') {
1807 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1808 // 8bit and control characters in style attribute values can be used for XSS, remove them
1809 $attary{$attname} = '"disallowed character"';
1811 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1812 if (count($aMatch)) {
1813 foreach($aMatch[1] as $sMatch) {
1815 $urlvalue = $sMatch;
1816 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1817 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1822 * Use white list based filtering on attributes which can contain url's
1824 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1825 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1826 $attary{$attname} = $attvalue;
1830 * See if we need to append any attributes to this tag.
1832 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1833 if (preg_match($matchtag, $tagname)){
1834 $attary = array_merge($attary, $addattary);
1841 * This function filters url's
1843 * @param $attvalue String with attribute value to filter
1844 * @param $message message object
1845 * @param $id message id
1846 * @param $mailbox mailbox
1847 * @param $sQuote quoting characters around url's
1849 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1850 $attvalue = trim($attvalue);
1851 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1852 // remove the double quotes
1853 $sQuote = $attvalue[0];
1854 $attvalue = trim(substr($attvalue,1,-1));
1857 // If there's no "view_unsafe_images" setting in the user's preferences,
1858 // turn unsafe images off by default.
1859 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1860 $view_unsafe_images = false;
1862 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1865 * Replace empty src tags with the blank image. src is only used
1866 * for frames, images, and image inputs. Doing a replace should
1867 * not affect them working as should be, however it will stop
1868 * IE from being kicked off when src for img tags are not set
1870 if ($attvalue == '') {
1871 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1873 // first, disallow 8 bit characters and control characters
1874 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1877 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1880 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1884 $aUrl = parse_url($attvalue);
1885 if (isset($aUrl['scheme'])) {
1886 switch(strtolower($aUrl['scheme'])) {
1891 if ($attname != 'href') {
1892 if ($view_unsafe_images == false) {
1893 $attvalue = $sQuote . $secremoveimg . $sQuote;
1895 if (isset($aUrl['path'])) {
1896 // validate image extension.
1897 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1898 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1899 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1902 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1909 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1910 * One day MS might actually make it match something useful, for now, falling
1911 * back to using cid2http, so we can grab the blank.png.
1913 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1917 * Turn cid: urls into http-friendly ones.
1919 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1922 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1926 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1927 // parse_url did not lead to satisfying result
1928 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1936 * This function edits the style definition to make them friendly and
1937 * usable in SquirrelMail.
1939 * @param $message the message object
1940 * @param $id the message id
1941 * @param $content a string with whatever is between <style> and </style>
1942 * @param $mailbox the message mailbox
1943 * @return a string with edited content.
1945 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1946 $me = 'sq_fixstyle';
1947 // workaround for </style> in between comments
1948 $iCurrentPos = $pos;
1953 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
1960 if ($sToken == '<') {
1970 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1975 $content .= $sToken;
1983 if ($sToken == '<') {
1985 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
1986 $i = strpos($body,'-->',$i+
3);
1987 if ($i === false) { // no end comment
2005 if ($bSucces == FALSE){
2006 return array(FALSE, strlen($body));
2012 * First look for general BODY style declaration, which would be
2014 * body {background: blah-blah}
2015 * and change it to .bodyclass so we can just assign it to a <div>
2017 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2018 $secremoveimg = '../images/' . _("sec_remove_eng.png");
2020 * Fix url('blah') declarations.
2022 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2023 // "url(\\1$secremoveimg\\2)", $content);
2025 // first check for 8bit sequences and disallowed control characters
2026 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2027 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2028 return array($content, $newpos);
2031 // IE Sucks hard. We have a special function for it.
2032 sq_fixIE_idiocy($content);
2034 // remove @import line
2035 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2037 // translate ur\l and variations (IE parses that)
2038 // TODO check if the sq_fixIE_idiocy function already handles this.
2039 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2040 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2041 if (count($aMatch)) {
2042 $aValue = $aReplace = array();
2043 foreach($aMatch[1] as $sMatch) {
2045 $urlvalue = $sMatch;
2046 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2047 $aValue[] = $sMatch;
2048 $aReplace[] = $urlvalue;
2050 $content = str_replace($aValue,$aReplace,$content);
2054 * Remove any backslashes, entities, and extraneous whitespace.
2056 $contentTemp = $content;
2057 sq_defang($contentTemp);
2058 sq_unspace($contentTemp);
2061 * Fix stupid css declarations which lead to vulnerabilities
2064 $match = Array('/\/\*.*\*\//',
2068 '/include-source/i',
2071 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2072 $contentNew = preg_replace($match, $replace, $contentTemp);
2073 if ($contentNew !== $contentTemp) {
2074 // insecure css declarations are used. From now on we don't care
2075 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2076 $content = $contentNew;
2078 return array($content, $newpos);
2083 * This function converts cid: url's into the ones that can be viewed in
2086 * @param $message the message object
2087 * @param $id the message id
2088 * @param $cidurl the cid: url.
2089 * @param $mailbox the message mailbox
2090 * @return a string with a http-friendly url
2092 function sq_cid2http($message, $id, $cidurl, $mailbox){
2094 * Get rid of quotes.
2096 $quotchar = substr($cidurl, 0, 1);
2097 if ($quotchar == '"' ||
$quotchar == "'"){
2098 $cidurl = str_replace($quotchar, "", $cidurl);
2102 $cidurl = substr(trim($cidurl), 4);
2104 $match_str = '/\{.*?\}\//';
2106 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2108 $linkurl = find_ent_id($cidurl, $message);
2109 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2110 unsafe link image */
2114 * This is part of a fix for Outlook Express 6.x generating
2115 * cid URLs without creating content-id headers. These images are
2116 * not part of the multipart/related html mail. The html contains
2117 * <img src="cid:{some_id}/image_filename.ext"> references to
2118 * attached images with as goal to render them inline although
2119 * the attachment disposition property is not inline.
2122 if (empty($linkurl)) {
2123 if (preg_match('/{.*}\//', $cidurl)) {
2124 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2125 if (!empty($cidurl)) {
2126 $linkurl = find_ent_id($cidurl, $message);
2131 if (!empty($linkurl)) {
2132 $httpurl = $quotchar . SM_PATH
. 'src/download.php?absolute_dl=true&' .
2133 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2134 '&ent_id=' . $linkurl . $quotchar;
2137 * If we couldn't generate a proper img url, drop in a blank image
2138 * instead of sending back empty, otherwise it causes unusual behaviour
2140 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2147 * This function changes the <body> tag into a <div> tag since we
2148 * can't really have a body-within-body.
2150 * @param $attary an array of attributes and values of <body>
2151 * @param $mailbox mailbox we're currently reading (for cid2http)
2152 * @param $message current message (for cid2http)
2153 * @param $id current message id (for cid2http)
2154 * @return a modified array of attributes to be set for <div>
2156 function sq_body2div($attary, $mailbox, $message, $id){
2157 $me = 'sq_body2div';
2158 $divattary = Array('class' => "'bodyclass'");
2160 $has_bgc_stl = $has_txt_stl = false;
2162 if (is_array($attary) && sizeof($attary) > 0){
2163 foreach ($attary as $attname=>$attvalue){
2164 $quotchar = substr($attvalue, 0, 1);
2165 $attvalue = str_replace($quotchar, "", $attvalue);
2168 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2169 $styledef .= "background-image: url('$attvalue'); ";
2172 $has_bgc_stl = true;
2173 $styledef .= "background-color: $attvalue; ";
2176 $has_txt_stl = true;
2177 $styledef .= "color: $attvalue; ";
2181 // Outlook defines a white bgcolor and no text color. This can lead to
2182 // white text on a white bg with certain themes.
2183 if ($has_bgc_stl && !$has_txt_stl) {
2184 $styledef .= "color: $text; ";
2186 if (strlen($styledef) > 0){
2187 $divattary{"style"} = "\"$styledef\"";
2194 * This is the main function and the one you should actually be calling.
2195 * There are several variables you should be aware of an which need
2196 * special description.
2198 * Since the description is quite lengthy, see it here:
2199 * http://linux.duke.edu/projects/mini/htmlfilter/
2201 * @param $body the string with HTML you wish to filter
2202 * @param $tag_list see description above
2203 * @param $rm_tags_with_content see description above
2204 * @param $self_closing_tags see description above
2205 * @param $force_tag_closing see description above
2206 * @param $rm_attnames see description above
2207 * @param $bad_attvals see description above
2208 * @param $add_attr_to_tag see description above
2209 * @param $message message object
2210 * @param $id message id
2211 * @return sanitized html safe to show on your pages.
2213 function sq_sanitize($body,
2215 $rm_tags_with_content,
2225 $me = 'sq_sanitize';
2226 $rm_tags = array_shift($tag_list);
2228 * Normalize rm_tags and rm_tags_with_content.
2230 @array_walk
($tag_list, 'sq_casenormalize');
2231 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2232 @array_walk
($self_closing_tags, 'sq_casenormalize');
2234 * See if tag_list is of tags to remove or tags to allow.
2235 * false means remove these tags
2236 * true means allow these tags
2239 $open_tags = Array();
2240 $trusted = "\n<!-- begin sanitized html -->\n";
2241 $skip_content = false;
2243 * Take care of netscape's stupid javascript entities like
2246 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2248 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2249 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2250 $free_content = substr($body, $curpos, $lt-$curpos);
2252 * Take care of <style>
2254 if ($tagname == "style" && $tagtype == 1){
2255 list($free_content, $curpos) =
2256 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2257 if ($free_content != FALSE){
2258 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2259 $trusted .= $free_content;
2260 $trusted .= sq_tagprint($tagname, false, 2);
2264 if ($skip_content == false){
2265 $trusted .= $free_content;
2267 if ($tagname != FALSE){
2269 if ($skip_content == $tagname){
2271 * Got to the end of tag we needed to remove.
2274 $skip_content = false;
2276 if ($skip_content == false){
2277 if ($tagname == "body"){
2280 if (isset($open_tags{$tagname}) &&
2281 $open_tags{$tagname} > 0){
2282 $open_tags{$tagname}--;
2290 * $rm_tags_with_content
2292 if ($skip_content == false){
2294 * See if this is a self-closing type and change
2295 * tagtype appropriately.
2298 && in_array($tagname, $self_closing_tags)){
2302 * See if we should skip this tag and any content
2305 if ($tagtype == 1 &&
2306 in_array($tagname, $rm_tags_with_content)){
2307 $skip_content = $tagname;
2309 if (($rm_tags == false
2310 && in_array($tagname, $tag_list)) ||
2311 ($rm_tags == true &&
2312 !in_array($tagname, $tag_list))){
2316 * Convert body into div.
2318 if ($tagname == "body"){
2320 $attary = sq_body2div($attary, $mailbox,
2324 if (isset($open_tags{$tagname})){
2325 $open_tags{$tagname}++
;
2327 $open_tags{$tagname}=1;
2331 * This is where we run other checks.
2333 if (is_array($attary) && sizeof($attary) > 0){
2334 $attary = sq_fixatts($tagname,
2348 if ($tagname != false && $skip_content == false){
2349 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2354 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2355 if ($force_tag_closing == true){
2356 foreach ($open_tags as $tagname=>$opentimes){
2357 while ($opentimes > 0){
2358 $trusted .= '</' . $tagname . '>';
2364 $trusted .= "<!-- end sanitized html -->\n";
2369 * This is a wrapper function to call html sanitizing routines.
2371 * @param $body the body of the message
2372 * @param $id the id of the message
2376 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2377 * into internal SM compose links
2378 * (optional; default = TRUE)
2379 * @return a string with html safe to display in the browser.
2381 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2383 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2385 global $attachment_common_show_images, $view_unsafe_images,
2388 * Don't display attached images in HTML mode.
2392 $attachment_common_show_images = false;
2407 $rm_tags_with_content = Array(
2417 $self_closing_tags = Array(
2425 $force_tag_closing = true;
2427 $rm_attnames = Array(
2438 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2439 $bad_attvals = Array(
2442 "/^src|background/i" =>
2445 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2446 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2447 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2450 "\\1$secremoveimg\\2",
2451 "\\1$secremoveimg\\2",
2452 "\\1$secremoveimg\\2",
2455 "/^href|action/i" =>
2458 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2459 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2460 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2475 "/include-source/i",
2476 "/position\s*:\s*absolute/i",
2477 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2478 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2479 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2480 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2481 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2500 // If there's no "view_unsafe_images" setting in the user's preferences,
2501 // turn unsafe images off by default.
2502 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
2503 $view_unsafe_images = false;
2506 if (!$view_unsafe_images){
2508 * Remove any references to http/https if view_unsafe_images set
2511 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2512 '/^([\'\"])\s*https*:.*([\'\"])/si');
2513 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2514 "\\1$secremoveimg\\1");
2515 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2516 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2517 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2518 "url(\\1$secremoveimg\\1)");
2521 $add_attr_to_tag = Array(
2523 Array('target'=>'"_blank"',
2524 'title'=>'"'._("This external link will open in a new window").'"'
2527 $trusted = sq_sanitize($body,
2529 $rm_tags_with_content,
2539 if (strpos($trusted,$secremoveimg)){
2540 $has_unsafe_images = true;
2543 // we want to parse mailto's in HTML output, change to SM compose links
2544 // this is a modified version of code from url_parser.php... but Marc is
2545 // right: we need a better filtering implementation; adding this randomly
2546 // here is not a great solution
2548 if ($take_mailto_links) {
2549 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2550 global $MailTo_PReg_Match;
2551 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2552 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2553 foreach ($regs[0] as $i => $mailto_before) {
2554 $mailto_params = $regs[10][$i];
2555 // get rid of any tailing quote since we have to add send_to to the end
2557 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2558 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2559 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2560 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2562 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2563 $to = 'to=' . $regs[1][$i];
2564 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2565 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2567 if ($mailto_params) //already some params, append to them
2568 $mailto_params .= '&' . $to;
2570 $mailto_params .= '?' . $to;
2574 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2576 // we'll already have target=_blank, no need to allow comp_in_new
2577 // here (which would be a lot more work anyway)
2579 global $compose_new_win;
2580 $temp_comp_in_new = $compose_new_win;
2581 $compose_new_win = 0;
2582 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2583 $compose_new_win = $temp_comp_in_new;
2585 // remove <a href=" and anything after the next quote (we only
2586 // need the uri, not the link HTML) in compose uri
2588 $comp_uri = substr($comp_uri, 9);
2589 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2590 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2599 * function SendDownloadHeaders - send file to the browser
2601 * Original Source: SM core src/download.php
2602 * moved here to make it available to other code, and separate
2603 * front end from back end functionality.
2605 * @param string $type0 first half of mime type
2606 * @param string $type1 second half of mime type
2607 * @param string $filename filename to tell the browser for downloaded file
2608 * @param boolean $force whether to force the download dialog to pop
2609 * @param optional integer $filesize send the Content-Header and length to the browser
2612 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2613 global $languages, $squirrelmail_language;
2614 $isIE = $isIE6plus = false;
2616 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2618 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2619 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2623 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2624 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2628 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2629 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2631 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2633 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace(' ', ' ', $filename));
2636 // A Pox on Microsoft and it's Internet Explorer!
2638 // IE has lots of bugs with file downloads.
2639 // It also has problems with SSL. Both of these cause problems
2640 // for us in this function.
2642 // See this article on Cache Control headers and SSL
2643 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2645 // The best thing you can do for IE is to upgrade to the latest
2647 //set all the Cache Control Headers for IE
2649 $filename=rawurlencode($filename);
2650 header ("Pragma: public");
2651 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2652 header ("Cache-Control: post-check=0, pre-check=0", false);
2653 header ("Cache-Control: private");
2655 //set the inline header for IE, we'll add the attachment header later if we need it
2656 header ("Content-Disposition: inline; filename=$filename");
2660 // Try to show in browser window
2661 header ("Content-Disposition: inline; filename=\"$filename\"");
2662 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2664 // Try to pop up the "save as" box
2666 // IE makes this hard. It pops up 2 save boxes, or none.
2667 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2668 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2669 // But, according to Microsoft, it is "RFC compliant but doesn't
2670 // take into account some deviations that allowed within the
2671 // specification." Doesn't that mean RFC non-compliant?
2672 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2674 // all browsers need the application/octet-stream header for this
2675 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2677 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2678 // Do not have quotes around filename, but that applied to
2679 // "attachment"... does it apply to inline too?
2680 header ("Content-Disposition: attachment; filename=\"$filename\"");
2682 if ($isIE && !$isIE6plus) {
2683 // This combination seems to work mostly. IE 5.5 SP 1 has
2684 // known issues (see the Microsoft Knowledge Base)
2686 // This works for most types, but doesn't work with Word files
2687 header ("Content-Type: application/download; name=\"$filename\"");
2688 header ("Content-Type: application/force-download; name=\"$filename\"");
2689 // These are spares, just in case. :-)
2690 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2691 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2692 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2694 // This is to prevent IE for MIME sniffing and auto open a file in IE
2695 header ("Content-Type: application/force-download; name=\"$filename\"");
2697 // another application/octet-stream forces download for Netscape
2698 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2702 //send the content-length header if the calling function provides it
2703 if ($filesize > 0) {
2704 header("Content-Length: $filesize");
2707 } // end fn SendDownloadHeaders