c151ecf6f5cb2d6d30c52a7426fad33f3827ccc0
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( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
363 $view_unsafe_images = false;
367 $urlmailbox = urlencode($mailbox);
368 $body_message = getEntity($message, $ent_num);
369 if (($body_message->header
->type0
== 'text') ||
370 ($body_message->header
->type0
== 'rfc822')) {
371 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
372 $body = decodeBody($body, $body_message->header
->encoding
);
374 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
375 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
376 if (mb_detect_encoding($body) != 'ASCII') {
377 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
381 /* As of 1.5.2, $body is passed (and modified) by reference */
382 do_hook('message_body', $body);
384 /* If there are other types that shouldn't be formatted, add
388 if ($body_message->header
->type1
== 'html') {
389 if ($show_html_default <> 1) {
390 $entity_conv = array(' ' => ' ',
399 $body = strtr($body, $entity_conv);
400 $body = strip_tags($body);
402 translateText($body, $wrap_at,
403 $body_message->header
->getParameter('charset'));
404 } elseif ($use_iframe && ! $clean) {
405 // $clean is used to remove iframe in printable view.
408 * If we don't add html message between iframe tags,
409 * we must detect unsafe images and modify $has_unsafe_images.
411 $html_body = magicHTML($body, $id, $message, $mailbox);
412 // Convert character set in order to display html mails in different character set
413 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
415 // creating iframe url
416 $iframeurl=sqm_baseuri().'src/view_html.php?'
417 . 'mailbox=' . $urlmailbox
418 . '&passed_id=' . $id
419 . '&ent_id=' . $ent_num
420 . '&view_unsafe_images=' . (int) $view_unsafe_images;
423 $oTemplate->assign('iframe_url', $iframeurl);
424 $oTemplate->assign('iframe_height', $iframe_height);
425 $oTemplate->assign('html_body', $html_body);
427 $body = $oTemplate->fetch('read_html_iframe.tpl');
429 // old way of html rendering
431 * convert character set. charset_decode does not remove html special chars
432 * applied by magicHTML functions and does not sanitize them second time if
433 * fourth argument is true.
435 $charset = $body_message->header
->getParameter('charset');
436 if (!empty($charset)) {
437 $body = charset_decode($charset,$body,false,true);
439 $body = magicHTML($body, $id, $message, $mailbox);
442 translateText($body, $wrap_at,
443 $body_message->header
->getParameter('charset'));
446 // if this is the clean display (i.e. printer friendly), stop here.
451 $download_and_unsafe_link = '';
453 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
454 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
455 '&startMessage=' . $startMessage . '&show_more=0';
456 if (isset($passed_ent_id)) {
457 $link .= '&passed_ent_id='.$passed_ent_id;
459 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
460 $download_and_unsafe_link .= "$nbsp|$nbsp"
461 . create_hyperlink($download_href, _("Download this as a file"));
462 if ($view_unsafe_images) {
463 $text = _("Hide Unsafe Images");
465 if (isset($has_unsafe_images) && $has_unsafe_images) {
466 $link .= '&view_unsafe_images=1';
467 $text = _("View Unsafe Images");
473 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
474 $unsafe_image_toggle_text = $text;
475 $download_and_unsafe_link .= "$nbsp|$nbsp"
476 . create_hyperlink($unsafe_image_toggle_href, $text);
483 * Generate attachments array for passing to templates. Separated from
484 * formatAttachments() below so that the same array can be given to the
485 * print-friendly version.
488 * @param object $message SquirrelMail message object
489 * @param array $exclude_id message parts that are not attachments.
490 * @param string $mailbox mailbox name
491 * @param integer $id message id
493 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
494 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
496 $att_ar = $message->getAttachments($exclude_id);
497 $urlMailbox = urlencode($mailbox);
499 $attachments = array();
500 foreach ($att_ar as $att) {
501 $ent = $att->entity_id
;
502 $header = $att->header
;
503 $type0 = strtolower($header->type0
);
504 $type1 = strtolower($header->type1
);
507 $links['download link']['text'] = _("Download");
508 $links['download link']['href'] = $base_uri .
509 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
511 if ($type0 =='message' && $type1 == 'rfc822') {
512 $default_page = $base_uri . 'src/read_body.php';
513 $rfc822_header = $att->rfc822_header
;
514 $filename = $rfc822_header->subject
;
515 if (trim( $filename ) == '') {
516 $filename = 'untitled-[' . $ent . ']' ;
518 $from_o = $rfc822_header->from
;
519 if (is_object($from_o)) {
520 $from_name = decodeHeader($from_o->getAddress(false));
521 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
522 // something weird happens when a digest message is opened and you return to the digest
523 // now the from object is part of an array. Probably the parseHeader call overwrites the info
524 // retrieved from the bodystructure in a different way. We need to fix this later.
525 // possible starting point, do not fetch header we already have and inspect how
526 // the rfc822_header object behaves.
527 $from_name = decodeHeader($from_o[0]->getAddress(false));
529 $from_name = _("Unknown sender");
531 $description = _("From").': '.$from_name;
533 $default_page = $base_uri . 'src/download.php';
534 $filename = $att->getFilename();
535 if ($header->description
) {
536 $description = decodeHeader($header->description
);
542 $display_filename = $filename;
543 if (isset($passed_ent_id)) {
544 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
546 $passed_ent_id_link = '';
548 $defaultlink = $default_page . "?startMessage=$startMessage"
549 . "&passed_id=$id&mailbox=$urlMailbox"
550 . '&ent_id='.$ent.$passed_ent_id_link;
551 if ($where && $what) {
552 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
554 // IE does make use of mime content sniffing. Forcing a download
555 // prohibit execution of XSS inside an application/octet-stream attachment
556 if ($type0 == 'application' && $type1 == 'octet-stream') {
557 $defaultlink .= '&absolute_dl=true';
560 /* This executes the attachment hook with a specific MIME-type.
561 * If that doesn't have results, it tries if there's a rule
562 * for a more generic type. Finally, a hook for ALL attachment
563 * types is run as well.
565 // First remember the default link.
566 $defaultlink_orig = $defaultlink;
568 /* The API for this hook has changed as of 1.5.2 so that all plugin
569 arguments are passed in an array instead of each their own plugin
570 argument, and arguments are passed by reference, so instead of
571 returning any changes, changes should simply be made to the original
572 arguments themselves. */
573 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
574 &$defaultlink, &$display_filename, &$where, &$what);
575 do_hook("attachment $type0/$type1", $temp);
576 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
577 /* The API for this hook has changed as of 1.5.2 so that all plugin
578 arguments are passed in an array instead of each their own plugin
579 argument, and arguments are passed by reference, so instead of
580 returning any changes, changes should simply be made to the original
581 arguments themselves. */
582 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
583 &$defaultlink, &$display_filename, &$where, &$what);
584 do_hook("attachment $type0/*", $temp);
586 /* The API for this hook has changed as of 1.5.2 so that all plugin
587 arguments are passed in an array instead of each their own plugin
588 argument, and arguments are passed by reference, so instead of
589 returning any changes, changes should simply be made to the original
590 arguments themselves. */
591 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
592 &$defaultlink, &$display_filename, &$where, &$what);
593 // Do not let a generic plugin change the default link if a more
594 // specialized one already did it...
595 if ($defaultlink != $defaultlink_orig) {
599 do_hook("attachment */*", $temp);
601 $this_attachment = array();
602 $this_attachment['Name'] = decodeHeader($display_filename);
603 $this_attachment['Description'] = $description;
604 $this_attachment['DefaultHREF'] = $defaultlink;
605 $this_attachment['DownloadHREF'] = $links['download link']['href'];
606 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
607 $this_attachment['Size'] = $header->size
;
608 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
609 $this_attachment['OtherLinks'] = array();
610 foreach ($links as $val) {
611 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
613 if (empty($val['text']) && empty($val['extra']))
617 $temp['HREF'] = $val['href'];
618 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
619 $this_attachment['OtherLinks'][] = $temp;
621 $attachments[] = $this_attachment;
630 * Displays attachment links and information
632 * Since 1.3.0 function is not included in formatBody() call.
634 * Since 1.0.2 uses attachment $type0/$type1 hook.
635 * Since 1.2.5 uses attachment $type0/* hook.
636 * Since 1.5.0 uses attachments_bottom hook.
637 * Since 1.5.2 uses templates and does *not* return a value.
639 * @param object $message SquirrelMail message object
640 * @param array $exclude_id message parts that are not attachments.
641 * @param string $mailbox mailbox name
642 * @param integer $id message id
644 function formatAttachments($message, $exclude_id, $mailbox, $id) {
647 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
649 $oTemplate->assign('attachments', $attach);
650 $oTemplate->display('read_attachments.tpl');
653 function sqimap_base64_decode(&$string) {
655 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
656 // fly decoding (to reduce memory usage) you have to check if the
657 // data has incomplete pairs
659 // Remove the noise in order to check if the 4 bytes pairs are complete
660 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
663 $iMod = strlen($string) %
4;
665 $sStringRem = substr($string,-$iMod);
666 // Check if $sStringRem contains padding characters
667 if (substr($sStringRem,-1) != '=') {
668 $string = substr($string,0,-$iMod);
673 $string = base64_decode($string);
678 * Decodes encoded message body
680 * This function decodes the body depending on the encoding type.
681 * Currently quoted-printable and base64 encodings are supported.
682 * decode_body hook was added to this function in 1.4.2/1.5.0
683 * @param string $body encoded message body
684 * @param string $encoding used encoding
685 * @return string decoded string
688 function decodeBody($body, $encoding) {
690 $body = str_replace("\r\n", "\n", $body);
691 $encoding = strtolower($encoding);
693 $encoding_handler = do_hook('decode_body', $encoding);
696 // plugins get first shot at decoding the body
698 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
699 $body = $encoding_handler('decode', $body);
701 } elseif ($encoding == 'quoted-printable' ||
702 $encoding == 'quoted_printable') {
704 * quoted_printable_decode() function is broken in older
705 * php versions. Text with \r\n decoding was fixed only
706 * in php 4.3.0. Minimal code requirement 4.0.4 +
707 * str_replace("\r\n", "\n", $body); call.
709 $body = quoted_printable_decode($body);
710 } elseif ($encoding == 'base64') {
711 $body = base64_decode($body);
714 // All other encodings are returned raw.
721 * This function decodes strings that are encoded according to
722 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
723 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
725 * @param string $string header string that has to be made readable
726 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
727 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
728 * @param boolean $decide decide if string can be utfencoded. defaults to false
729 * @return string decoded header string
731 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
732 global $languages, $squirrelmail_language,$default_charset;
733 if (is_array($string)) {
734 $string = implode("\n", $string);
737 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
738 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
739 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
740 // Do we need to return at this point?
747 $aString = explode(' ',$string);
749 foreach ($aString as $chunk) {
750 if ($encoded && $chunk === '') {
752 } elseif ($chunk === '') {
757 /* if encoded words are not separated by a linear-space-white we still catch them */
760 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
761 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
762 if ($iLastMatch !== $j) {
772 $ret .= htmlspecialchars($res[1]);
776 $encoding = ucfirst($res[3]);
778 /* decide about valid decoding */
779 if ($decide && is_conversion_safe($res[2])) {
781 $can_be_encoded=true;
783 $can_be_encoded=false;
788 $replace = base64_decode($res[4]);
790 if ($can_be_encoded) {
791 /* convert string to different charset,
792 * if functions asks for it (usually in compose)
794 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
796 // convert string to html codes in order to display it
797 $ret .= charset_decode($res[2],$replace);
801 $replace = htmlspecialchars($replace);
807 $replace = str_replace('_', ' ', $res[4]);
808 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
811 if ($can_be_encoded) {
812 /* convert string to different charset,
813 * if functions asks for it (usually in compose)
815 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
817 // convert string to html codes in order to display it
818 $replace = charset_decode($res[2], $replace);
822 $replace = htmlspecialchars($replace);
841 if (!$encoded && $htmlsave) {
842 $ret .= htmlspecialchars($chunk);
848 /* remove the first added space */
851 $ret = substr($ret,5);
853 $ret = substr($ret,1);
863 * Function uses XTRA_CODE _encodeheader function, if such function exists.
865 * Function uses Q encoding by default and encodes a string according to RFC
866 * 1522 for use in headers if it contains 8-bit characters or anything that
867 * looks like it should be encoded.
869 * Function switches to B encoding and encodeHeaderBase64() function, if
870 * string is 8bit and multibyte character set supported by mbstring extension
871 * is used. It can cause E_USER_NOTICE errors, if interface is used with
872 * multibyte character set unsupported by mbstring extension.
874 * @param string $string header string, that has to be encoded
875 * @return string quoted-printable encoded string
876 * @todo make $mb_charsets system wide constant
878 function encodeHeader ($string) {
879 global $default_charset, $languages, $squirrelmail_language;
881 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
882 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
883 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
886 // Use B encoding for multibyte charsets
887 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
888 if (in_array($default_charset,$mb_charsets) &&
889 in_array($default_charset,sq_mb_list_encodings()) &&
890 sq_is8bit($string)) {
891 return encodeHeaderBase64($string,$default_charset);
892 } elseif (in_array($default_charset,$mb_charsets) &&
893 sq_is8bit($string) &&
894 ! in_array($default_charset,sq_mb_list_encodings())) {
895 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
896 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
899 // Encode only if the string contains 8-bit characters or =?
900 $j = strlen($string);
901 $max_l = 75 - strlen($default_charset) - 7;
904 $iEncStart = $enc_init = false;
905 $cur_l = $iOffset = 0;
906 for($i = 0; $i < $j; ++
$i) {
915 if ($iEncStart === false) {
919 if ($cur_l > ($max_l-2)) {
920 /* if there is an stringpart that doesn't need encoding, add it */
921 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
922 $aRet[] = "=?$default_charset?Q?$ret?=";
928 $ret .= sprintf("=%02X",ord($string{$i}));
933 if ($iEncStart !== false) {
934 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
935 $aRet[] = "=?$default_charset?Q?$ret?=";
943 if ($iEncStart !== false) {
945 if ($cur_l > $max_l) {
946 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
947 $aRet[] = "=?$default_charset?Q?$ret?=";
958 $k = ord($string{$i});
960 if ($iEncStart === false) {
961 // do not start encoding in the middle of a string, also take the rest of the word.
962 $sLeadString = substr($string,0,$i);
963 $aLeadString = explode(' ',$sLeadString);
964 $sToBeEncoded = array_pop($aLeadString);
965 $iEncStart = $i - strlen($sToBeEncoded);
966 $ret .= $sToBeEncoded;
967 $cur_l +
= strlen($sToBeEncoded);
970 /* first we add the encoded string that reached it's max size */
971 if ($cur_l > ($max_l-2)) {
972 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
973 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
980 $ret .= sprintf("=%02X", $k);
982 if ($iEncStart !== false) {
984 if ($cur_l > $max_l) {
985 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
986 $aRet[] = "=?$default_charset?Q?$ret?=";
1001 if ($iEncStart !== false) {
1002 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1003 $aRet[] = "=?$default_charset?Q?$ret?=";
1005 $aRet[] = substr($string,$iOffset);
1007 $string = implode('',$aRet);
1013 * Encodes string according to rfc2047 B encoding header formating rules
1015 * It is recommended way to encode headers with character sets that store
1016 * symbols in more than one byte.
1018 * Function requires mbstring support. If required mbstring functions are missing,
1019 * function returns false and sets E_USER_WARNING level error message.
1021 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1022 * that mbstring functions will generate E_WARNING errors, if unsupported
1023 * character set is used. mb_encode_mimeheader function provided by php
1024 * mbstring extension is not used in order to get better control of header
1027 * Used php code functions - function_exists(), trigger_error(), strlen()
1028 * (is used with charset names and base64 strings). Used php mbstring
1029 * functions - mb_strlen and mb_substr.
1031 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1032 * encoding), rfc 2822 (header folding)
1034 * @param string $string header string that must be encoded
1035 * @param string $charset character set. Must be supported by mbstring extension.
1036 * Use sq_mb_list_encodings() to detect supported charsets.
1037 * @return string string encoded according to rfc2047 B encoding formating rules
1039 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1040 * @todo Do we want to control max length of header?
1041 * @todo Do we want to control EOL (end-of-line) marker?
1042 * @todo Do we want to translate error message?
1044 function encodeHeaderBase64($string,$charset) {
1046 * Check mbstring function requirements.
1048 if (! function_exists('mb_strlen') ||
1049 ! function_exists('mb_substr')) {
1050 // set E_USER_WARNING
1051 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1056 // initial return array
1060 * header length = 75 symbols max (same as in encodeHeader)
1061 * remove $charset length
1062 * remove =? ? ?= (5 chars)
1063 * remove 2 more chars (\r\n ?)
1065 $iMaxLength = 75 - strlen($charset) - 7;
1067 // set first character position
1070 // loop through all characters. count characters and not bytes.
1071 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1072 // encode string from starting character to current character.
1073 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1075 // Check encoded string length
1076 if(strlen($encoded_string)>$iMaxLength) {
1077 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1078 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1080 // set new starting character
1081 $iStartCharNum = $iCharNum-1;
1083 // encode last char (in case it is last character in string)
1084 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1085 } // if string is shorter than max length - add next character
1088 // add last encoded string to array
1089 $aRet[] = $encoded_string;
1091 // set initial return string
1094 // loop through encoded strings
1095 foreach($aRet as $string) {
1096 // TODO: Do we want to control EOL (end-of-line) marker
1097 if ($sRet!='') $sRet.= " ";
1099 // add header tags and encoded string to return string
1100 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1106 /* This function trys to locate the entity_id of a specific mime element */
1107 function find_ent_id($id, $message) {
1108 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1109 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1110 $ret = find_ent_id($id, $message->entities
[$i]);
1112 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1113 // if (sq_check_save_extension($message->entities[$i])) {
1114 return $message->entities
[$i]->entity_id
;
1116 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1118 * This is part of a fix for Outlook Express 6.x generating
1119 * cid URLs without creating content-id headers
1122 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1123 return $message->entities
[$i]->entity_id
;
1131 function sq_check_save_extension($message) {
1132 $filename = $message->getFilename();
1133 $ext = substr($filename, strrpos($filename,'.')+
1);
1134 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1135 return in_array($ext, $save_extensions);
1140 ** HTMLFILTER ROUTINES
1144 * This function checks attribute values for entity-encoded values
1145 * and returns them translated into 8-bit strings so we can run
1148 * @param $attvalue A string to run entity check against.
1149 * @return Nothing, modifies a reference value.
1151 function sq_defang(&$attvalue){
1154 * Skip this if there aren't ampersands or backslashes.
1156 if (strpos($attvalue, '&') === false
1157 && strpos($attvalue, '\\') === false){
1161 // before deent, translate the dangerous unicode characters and ... to safe values
1162 // otherwise the regular expressions do not match.
1168 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1169 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1170 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1171 } while ($m == true);
1172 $attvalue = stripslashes($attvalue);
1176 * Kill any tabs, newlines, or carriage returns. Our friends the
1177 * makers of the browser with 95% market value decided that it'd
1178 * be funny to make "java[tab]script" be just as good as "javascript".
1180 * @param attvalue The attribute value before extraneous spaces removed.
1181 * @return attvalue Nothing, modifies a reference value.
1183 function sq_unspace(&$attvalue){
1185 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1186 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1187 Array('', '', '', '', ''), $attvalue);
1192 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1193 * IE as regular characters.
1195 * @param attvalue The attribute value before dangerous characters are translated.
1196 * @return attvalue Nothing, modifies a reference value.
1197 * @author Marc Groot Koerkamp.
1199 function sq_fixIE_idiocy(&$attvalue) {
1201 $attvalue = str_replace("\0", "", $attvalue);
1203 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1205 // IE has the evil habit of accepting every possible value for the attribute expression.
1206 // The table below contains characters which are parsed by IE if they are used in the "expression"
1208 $aDangerousCharsReplacementTable = array(
1209 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1210 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1211 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1212 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1213 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1214 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1215 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1216 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1217 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1218 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1219 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1220 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1221 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1222 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1223 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1224 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1225 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1226 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1227 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1228 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1229 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1230 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1231 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1232 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1233 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1234 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1235 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1236 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1237 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1238 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1239 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1240 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1241 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1242 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1243 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1244 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1245 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1246 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1247 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1248 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1249 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1250 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1251 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1252 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1253 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1254 "\xCA\x9F", /* L UNICODE IPA Extension */
1255 "\xCA\x80", /* R UNICODE IPA Extension */
1256 "\xC9\xB4"), /* N UNICODE IPA Extension */
1257 array('l', 'l', 'r','r','n','n',
1258 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1259 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1260 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1261 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1263 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1265 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1269 * This function returns the final tag out of the tag name, an array
1270 * of attributes, and the type of the tag. This function is called by
1271 * sq_sanitize internally.
1273 * @param $tagname the name of the tag.
1274 * @param $attary the array of attributes and their values
1275 * @param $tagtype The type of the tag (see in comments).
1276 * @return a string with the final tag representation.
1278 function sq_tagprint($tagname, $attary, $tagtype){
1279 $me = 'sq_tagprint';
1282 $fulltag = '</' . $tagname . '>';
1284 $fulltag = '<' . $tagname;
1285 if (is_array($attary) && sizeof($attary)){
1287 while (list($attname, $attvalue) = each($attary)){
1288 array_push($atts, "$attname=$attvalue");
1290 $fulltag .= ' ' . join(" ", $atts);
1301 * A small helper function to use with array_walk. Modifies a by-ref
1302 * value and makes it lowercase.
1304 * @param $val a value passed by-ref.
1305 * @return void since it modifies a by-ref value.
1307 function sq_casenormalize(&$val){
1308 $val = strtolower($val);
1312 * This function skips any whitespace from the current position within
1313 * a string and to the next non-whitespace value.
1315 * @param $body the string
1316 * @param $offset the offset within the string where we should start
1317 * looking for the next non-whitespace character.
1318 * @return the location within the $body where the next
1319 * non-whitespace char is located.
1321 function sq_skipspace($body, $offset){
1322 $me = 'sq_skipspace';
1323 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1324 if (sizeof($matches{1})){
1325 $count = strlen($matches{1});
1332 * This function looks for the next character within a string. It's
1333 * really just a glorified "strpos", except it catches if failures
1336 * @param $body The string to look for needle in.
1337 * @param $offset Start looking from this position.
1338 * @param $needle The character/string to look for.
1339 * @return location of the next occurance of the needle, or
1340 * strlen($body) if needle wasn't found.
1342 function sq_findnxstr($body, $offset, $needle){
1343 $me = 'sq_findnxstr';
1344 $pos = strpos($body, $needle, $offset);
1345 if ($pos === FALSE){
1346 $pos = strlen($body);
1352 * This function takes a PCRE-style regexp and tries to match it
1353 * within the string.
1355 * @param $body The string to look for needle in.
1356 * @param $offset Start looking from here.
1357 * @param $reg A PCRE-style regex to match.
1358 * @return Returns a false if no matches found, or an array
1359 * with the following members:
1360 * - integer with the location of the match within $body
1361 * - string with whatever content between offset and the match
1362 * - string with whatever it is we matched
1364 function sq_findnxreg($body, $offset, $reg){
1365 $me = 'sq_findnxreg';
1368 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1369 if (!isset($matches{0}) ||
!$matches{0}){
1372 $retarr{0} = $offset +
strlen($matches{1});
1373 $retarr{1} = $matches{1};
1374 $retarr{2} = $matches{2};
1380 * This function looks for the next tag.
1382 * @param $body String where to look for the next tag.
1383 * @param $offset Start looking from here.
1384 * @return false if no more tags exist in the body, or
1385 * an array with the following members:
1386 * - string with the name of the tag
1387 * - array with attributes and their values
1388 * - integer with tag type (1, 2, or 3)
1389 * - integer where the tag starts (starting "<")
1390 * - integer where the tag ends (ending ">")
1391 * first three members will be false, if the tag is invalid.
1393 function sq_getnxtag($body, $offset){
1394 $me = 'sq_getnxtag';
1395 if ($offset > strlen($body)){
1398 $lt = sq_findnxstr($body, $offset, "<");
1399 if ($lt == strlen($body)){
1404 * blah blah <tag attribute="value">
1407 $pos = sq_skipspace($body, $lt+
1);
1408 if ($pos >= strlen($body)){
1409 return Array(false, false, false, $lt, strlen($body));
1412 * There are 3 kinds of tags:
1413 * 1. Opening tag, e.g.:
1415 * 2. Closing tag, e.g.:
1417 * 3. XHTML-style content-less tag, e.g.:
1418 * <img src="blah" />
1421 switch (substr($body, $pos, 1)){
1428 * A comment or an SGML declaration.
1430 if (substr($body, $pos+
1, 2) == "--"){
1431 $gt = strpos($body, "-->", $pos);
1433 $gt = strlen($body);
1437 return Array(false, false, false, $lt, $gt);
1439 $gt = sq_findnxstr($body, $pos, ">");
1440 return Array(false, false, false, $lt, $gt);
1445 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1455 * Look for next [\W-_], which will indicate the end of the tag name.
1457 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1458 if ($regary == false){
1459 return Array(false, false, false, $lt, strlen($body));
1461 list($pos, $tagname, $match) = $regary;
1462 $tagname = strtolower($tagname);
1465 * $match can be either of these:
1466 * '>' indicating the end of the tag entirely.
1467 * '\s' indicating the end of the tag name.
1468 * '/' indicating that this is type-3 xhtml tag.
1470 * Whatever else we find there indicates an invalid tag.
1475 * This is an xhtml-style tag with a closing / at the
1476 * end, like so: <img src="blah" />. Check if it's followed
1477 * by the closing bracket. If not, then this tag is invalid
1479 if (substr($body, $pos, 2) == "/>"){
1483 $gt = sq_findnxstr($body, $pos, ">");
1484 $retary = Array(false, false, false, $lt, $gt);
1488 return Array($tagname, false, $tagtype, $lt, $pos);
1492 * Check if it's whitespace
1494 if (!preg_match('/\s/', $match)){
1496 * This is an invalid tag! Look for the next closing ">".
1498 $gt = sq_findnxstr($body, $lt, ">");
1499 return Array(false, false, false, $lt, $gt);
1505 * At this point we're here:
1506 * <tagname attribute='blah'>
1509 * At this point we loop in order to find all attributes.
1515 while ($pos <= strlen($body)){
1516 $pos = sq_skipspace($body, $pos);
1517 if ($pos == strlen($body)){
1521 return Array(false, false, false, $lt, $pos);
1524 * See if we arrived at a ">" or "/>", which means that we reached
1525 * the end of the tag.
1528 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1532 $pos +
= strlen($matches{1});
1533 if ($matches{2} == "/>"){
1537 return Array($tagname, $attary, $tagtype, $lt, $pos);
1541 * There are several types of attributes, with optional
1542 * [:space:] between members.
1544 * attrname[:space:]=[:space:]'CDATA'
1546 * attrname[:space:]=[:space:]"CDATA"
1548 * attr[:space:]=[:space:]CDATA
1552 * We leave types 1 and 2 the same, type 3 we check for
1553 * '"' and convert to """ if needed, then wrap in
1554 * double quotes. Type 4 we convert into:
1557 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1558 if ($regary == false){
1560 * Looks like body ended before the end of tag.
1562 return Array(false, false, false, $lt, strlen($body));
1564 list($pos, $attname, $match) = $regary;
1565 $attname = strtolower($attname);
1567 * We arrived at the end of attribute name. Several things possible
1569 * '>' means the end of the tag and this is attribute type 4
1570 * '/' if followed by '>' means the same thing as above
1571 * '\s' means a lot of things -- look what it's followed by.
1572 * anything else means the attribute is invalid.
1577 * This is an xhtml-style tag with a closing / at the
1578 * end, like so: <img src="blah" />. Check if it's followed
1579 * by the closing bracket. If not, then this tag is invalid
1581 if (substr($body, $pos, 2) == "/>"){
1585 $gt = sq_findnxstr($body, $pos, ">");
1586 $retary = Array(false, false, false, $lt, $gt);
1590 $attary{$attname} = '"yes"';
1591 return Array($tagname, $attary, $tagtype, $lt, $pos);
1595 * Skip whitespace and see what we arrive at.
1597 $pos = sq_skipspace($body, $pos);
1598 $char = substr($body, $pos, 1);
1600 * Two things are valid here:
1601 * '=' means this is attribute type 1 2 or 3.
1602 * \w means this was attribute type 4.
1603 * anything else we ignore and re-loop. End of tag and
1604 * invalid stuff will be caught by our checks at the beginning
1609 $pos = sq_skipspace($body, $pos);
1611 * Here are 3 possibilities:
1612 * "'" attribute type 1
1613 * '"' attribute type 2
1614 * everything else is the content of tag type 3
1616 $quot = substr($body, $pos, 1);
1618 $regary = sq_findnxreg($body, $pos+
1, "\'");
1619 if ($regary == false){
1620 return Array(false, false, false, $lt, strlen($body));
1622 list($pos, $attval, $match) = $regary;
1624 $attary{$attname} = "'" . $attval . "'";
1625 } else if ($quot == '"'){
1626 $regary = sq_findnxreg($body, $pos+
1, '\"');
1627 if ($regary == false){
1628 return Array(false, false, false, $lt, strlen($body));
1630 list($pos, $attval, $match) = $regary;
1632 $attary{$attname} = '"' . $attval . '"';
1635 * These are hateful. Look for \s, or >.
1637 $regary = sq_findnxreg($body, $pos, "[\s>]");
1638 if ($regary == false){
1639 return Array(false, false, false, $lt, strlen($body));
1641 list($pos, $attval, $match) = $regary;
1643 * If it's ">" it will be caught at the top.
1645 $attval = preg_replace("/\"/s", """, $attval);
1646 $attary{$attname} = '"' . $attval . '"';
1648 } else if (preg_match("|[\w/>]|", $char)) {
1650 * That was attribute type 4.
1652 $attary{$attname} = '"yes"';
1655 * An illegal character. Find next '>' and return.
1657 $gt = sq_findnxstr($body, $pos, ">");
1658 return Array(false, false, false, $lt, $gt);
1664 * The fact that we got here indicates that the tag end was never
1665 * found. Return invalid tag indication so it gets stripped.
1667 return Array(false, false, false, $lt, strlen($body));
1671 * Translates entities into literal values so they can be checked.
1673 * @param $attvalue the by-ref value to check.
1674 * @param $regex the regular expression to check against.
1675 * @param $hex whether the entites are hexadecimal.
1676 * @return True or False depending on whether there were matches.
1678 function sq_deent(&$attvalue, $regex, $hex=false){
1682 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1683 preg_match_all($regex, $attvalue, $matches);
1684 if (is_array($matches) && sizeof($matches[0]) > 0){
1686 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1687 $numval = $matches[1][$i];
1689 $numval = hexdec($numval);
1691 $repl{$matches[0][$i]} = chr($numval);
1693 $attvalue = strtr($attvalue, $repl);
1701 * This function runs various checks against the attributes.
1703 * @param $tagname String with the name of the tag.
1704 * @param $attary Array with all tag attributes.
1705 * @param $rm_attnames See description for sq_sanitize
1706 * @param $bad_attvals See description for sq_sanitize
1707 * @param $add_attr_to_tag See description for sq_sanitize
1708 * @param $message message object
1709 * @param $id message id
1710 * @return Array with modified attributes.
1712 function sq_fixatts($tagname,
1722 while (list($attname, $attvalue) = each($attary)){
1724 * See if this attribute should be removed.
1726 foreach ($rm_attnames as $matchtag=>$matchattrs){
1727 if (preg_match($matchtag, $tagname)){
1728 foreach ($matchattrs as $matchattr){
1729 if (preg_match($matchattr, $attname)){
1730 unset($attary{$attname});
1737 * Workaround for IE quirks
1739 sq_fixIE_idiocy($attvalue);
1742 * Remove any backslashes, entities, and extraneous whitespace.
1745 $oldattvalue = $attvalue;
1746 sq_defang($attvalue);
1747 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1748 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1749 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1750 $attvalue = "idiocy";
1751 $attary{$attname} = $attvalue;
1753 sq_unspace($attvalue);
1756 * Now let's run checks on the attvalues.
1757 * I don't expect anyone to comprehend this. If you do,
1758 * get in touch with me so I can drive to where you live and
1759 * shake your hand personally. :)
1761 foreach ($bad_attvals as $matchtag=>$matchattrs){
1762 if (preg_match($matchtag, $tagname)){
1763 foreach ($matchattrs as $matchattr=>$valary){
1764 if (preg_match($matchattr, $attname)){
1766 * There are two arrays in valary.
1768 * Second one is replacements
1770 list($valmatch, $valrepl) = $valary;
1772 preg_replace($valmatch, $valrepl, $attvalue);
1773 if ($newvalue != $attvalue){
1774 $attary{$attname} = $newvalue;
1775 $attvalue = $newvalue;
1781 if ($attname == 'style') {
1782 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1783 // 8bit and control characters in style attribute values can be used for XSS, remove them
1784 $attary{$attname} = '"disallowed character"';
1786 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1787 if (count($aMatch)) {
1788 foreach($aMatch[1] as $sMatch) {
1790 $urlvalue = $sMatch;
1791 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1792 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1797 * Use white list based filtering on attributes which can contain url's
1799 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1800 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1801 $attary{$attname} = $attvalue;
1805 * See if we need to append any attributes to this tag.
1807 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1808 if (preg_match($matchtag, $tagname)){
1809 $attary = array_merge($attary, $addattary);
1816 * This function filters url's
1818 * @param $attvalue String with attribute value to filter
1819 * @param $message message object
1820 * @param $id message id
1821 * @param $mailbox mailbox
1822 * @param $sQuote quoting characters around url's
1824 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1825 $attvalue = trim($attvalue);
1826 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1827 // remove the double quotes
1828 $sQuote = $attvalue[0];
1829 $attvalue = trim(substr($attvalue,1,-1));
1832 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1833 $view_unsafe_images = false;
1835 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1838 * Replace empty src tags with the blank image. src is only used
1839 * for frames, images, and image inputs. Doing a replace should
1840 * not affect them working as should be, however it will stop
1841 * IE from being kicked off when src for img tags are not set
1843 if ($attvalue == '') {
1844 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1846 // first, disallow 8 bit characters and control characters
1847 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1850 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1853 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1857 $aUrl = parse_url($attvalue);
1858 if (isset($aUrl['scheme'])) {
1859 switch(strtolower($aUrl['scheme'])) {
1864 if ($attname != 'href') {
1865 if ($view_unsafe_images == false) {
1866 $attvalue = $sQuote . $secremoveimg . $sQuote;
1868 if (isset($aUrl['path'])) {
1869 // validate image extension.
1870 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1871 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1872 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1875 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1882 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1883 * One day MS might actually make it match something useful, for now, falling
1884 * back to using cid2http, so we can grab the blank.png.
1886 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1890 * Turn cid: urls into http-friendly ones.
1892 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1895 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1899 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1900 // parse_url did not lead to satisfying result
1901 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1909 * This function edits the style definition to make them friendly and
1910 * usable in SquirrelMail.
1912 * @param $message the message object
1913 * @param $id the message id
1914 * @param $content a string with whatever is between <style> and </style>
1915 * @param $mailbox the message mailbox
1916 * @return a string with edited content.
1918 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1919 $me = 'sq_fixstyle';
1920 // workaround for </style> in between comments
1921 $iCurrentPos = $pos;
1926 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
1933 if ($sToken == '<') {
1943 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1948 $content .= $sToken;
1956 if ($sToken == '<') {
1958 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
1959 $i = strpos($body,'-->',$i+
3);
1960 if ($i === false) { // no end comment
1978 if ($bSucces == FALSE){
1979 return array(FALSE, strlen($body));
1985 * First look for general BODY style declaration, which would be
1987 * body {background: blah-blah}
1988 * and change it to .bodyclass so we can just assign it to a <div>
1990 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1991 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1993 * Fix url('blah') declarations.
1995 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1996 // "url(\\1$secremoveimg\\2)", $content);
1998 // first check for 8bit sequences and disallowed control characters
1999 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2000 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2001 return array($content, $newpos);
2004 // IE Sucks hard. We have a special function for it.
2005 sq_fixIE_idiocy($content);
2007 // remove @import line
2008 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2010 // translate ur\l and variations (IE parses that)
2011 // TODO check if the sq_fixIE_idiocy function already handles this.
2012 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2013 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2014 if (count($aMatch)) {
2015 $aValue = $aReplace = array();
2016 foreach($aMatch[1] as $sMatch) {
2018 $urlvalue = $sMatch;
2019 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2020 $aValue[] = $sMatch;
2021 $aReplace[] = $urlvalue;
2023 $content = str_replace($aValue,$aReplace,$content);
2027 * Remove any backslashes, entities, and extraneous whitespace.
2029 $contentTemp = $content;
2030 sq_defang($contentTemp);
2031 sq_unspace($contentTemp);
2034 * Fix stupid css declarations which lead to vulnerabilities
2037 $match = Array('/\/\*.*\*\//',
2041 '/include-source/i',
2044 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2045 $contentNew = preg_replace($match, $replace, $contentTemp);
2046 if ($contentNew !== $contentTemp) {
2047 // insecure css declarations are used. From now on we don't care
2048 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2049 $content = $contentNew;
2051 return array($content, $newpos);
2056 * This function converts cid: url's into the ones that can be viewed in
2059 * @param $message the message object
2060 * @param $id the message id
2061 * @param $cidurl the cid: url.
2062 * @param $mailbox the message mailbox
2063 * @return a string with a http-friendly url
2065 function sq_cid2http($message, $id, $cidurl, $mailbox){
2067 * Get rid of quotes.
2069 $quotchar = substr($cidurl, 0, 1);
2070 if ($quotchar == '"' ||
$quotchar == "'"){
2071 $cidurl = str_replace($quotchar, "", $cidurl);
2075 $cidurl = substr(trim($cidurl), 4);
2077 $match_str = '/\{.*?\}\//';
2079 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2081 $linkurl = find_ent_id($cidurl, $message);
2082 /* in case of non-save cid links $httpurl should be replaced by a sort of
2083 unsave link image */
2087 * This is part of a fix for Outlook Express 6.x generating
2088 * cid URLs without creating content-id headers. These images are
2089 * not part of the multipart/related html mail. The html contains
2090 * <img src="cid:{some_id}/image_filename.ext"> references to
2091 * attached images with as goal to render them inline although
2092 * the attachment disposition property is not inline.
2095 if (empty($linkurl)) {
2096 if (preg_match('/{.*}\//', $cidurl)) {
2097 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2098 if (!empty($cidurl)) {
2099 $linkurl = find_ent_id($cidurl, $message);
2104 if (!empty($linkurl)) {
2105 $httpurl = $quotchar . SM_PATH
. 'src/download.php?absolute_dl=true&' .
2106 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2107 '&ent_id=' . $linkurl . $quotchar;
2110 * If we couldn't generate a proper img url, drop in a blank image
2111 * instead of sending back empty, otherwise it causes unusual behaviour
2113 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2120 * This function changes the <body> tag into a <div> tag since we
2121 * can't really have a body-within-body.
2123 * @param $attary an array of attributes and values of <body>
2124 * @param $mailbox mailbox we're currently reading (for cid2http)
2125 * @param $message current message (for cid2http)
2126 * @param $id current message id (for cid2http)
2127 * @return a modified array of attributes to be set for <div>
2129 function sq_body2div($attary, $mailbox, $message, $id){
2130 $me = 'sq_body2div';
2131 $divattary = Array('class' => "'bodyclass'");
2133 $has_bgc_stl = $has_txt_stl = false;
2135 if (is_array($attary) && sizeof($attary) > 0){
2136 foreach ($attary as $attname=>$attvalue){
2137 $quotchar = substr($attvalue, 0, 1);
2138 $attvalue = str_replace($quotchar, "", $attvalue);
2141 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2142 $styledef .= "background-image: url('$attvalue'); ";
2145 $has_bgc_stl = true;
2146 $styledef .= "background-color: $attvalue; ";
2149 $has_txt_stl = true;
2150 $styledef .= "color: $attvalue; ";
2154 // Outlook defines a white bgcolor and no text color. This can lead to
2155 // white text on a white bg with certain themes.
2156 if ($has_bgc_stl && !$has_txt_stl) {
2157 $styledef .= "color: $text; ";
2159 if (strlen($styledef) > 0){
2160 $divattary{"style"} = "\"$styledef\"";
2167 * This is the main function and the one you should actually be calling.
2168 * There are several variables you should be aware of an which need
2169 * special description.
2171 * Since the description is quite lengthy, see it here:
2172 * http://linux.duke.edu/projects/mini/htmlfilter/
2174 * @param $body the string with HTML you wish to filter
2175 * @param $tag_list see description above
2176 * @param $rm_tags_with_content see description above
2177 * @param $self_closing_tags see description above
2178 * @param $force_tag_closing see description above
2179 * @param $rm_attnames see description above
2180 * @param $bad_attvals see description above
2181 * @param $add_attr_to_tag see description above
2182 * @param $message message object
2183 * @param $id message id
2184 * @return sanitized html safe to show on your pages.
2186 function sq_sanitize($body,
2188 $rm_tags_with_content,
2198 $me = 'sq_sanitize';
2199 $rm_tags = array_shift($tag_list);
2201 * Normalize rm_tags and rm_tags_with_content.
2203 @array_walk
($tag_list, 'sq_casenormalize');
2204 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2205 @array_walk
($self_closing_tags, 'sq_casenormalize');
2207 * See if tag_list is of tags to remove or tags to allow.
2208 * false means remove these tags
2209 * true means allow these tags
2212 $open_tags = Array();
2213 $trusted = "\n<!-- begin sanitized html -->\n";
2214 $skip_content = false;
2216 * Take care of netscape's stupid javascript entities like
2219 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2221 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2222 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2223 $free_content = substr($body, $curpos, $lt-$curpos);
2225 * Take care of <style>
2227 if ($tagname == "style" && $tagtype == 1){
2228 list($free_content, $curpos) =
2229 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2230 if ($free_content != FALSE){
2231 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2232 $trusted .= $free_content;
2233 $trusted .= sq_tagprint($tagname, false, 2);
2237 if ($skip_content == false){
2238 $trusted .= $free_content;
2240 if ($tagname != FALSE){
2242 if ($skip_content == $tagname){
2244 * Got to the end of tag we needed to remove.
2247 $skip_content = false;
2249 if ($skip_content == false){
2250 if ($tagname == "body"){
2253 if (isset($open_tags{$tagname}) &&
2254 $open_tags{$tagname} > 0){
2255 $open_tags{$tagname}--;
2263 * $rm_tags_with_content
2265 if ($skip_content == false){
2267 * See if this is a self-closing type and change
2268 * tagtype appropriately.
2271 && in_array($tagname, $self_closing_tags)){
2275 * See if we should skip this tag and any content
2278 if ($tagtype == 1 &&
2279 in_array($tagname, $rm_tags_with_content)){
2280 $skip_content = $tagname;
2282 if (($rm_tags == false
2283 && in_array($tagname, $tag_list)) ||
2284 ($rm_tags == true &&
2285 !in_array($tagname, $tag_list))){
2289 * Convert body into div.
2291 if ($tagname == "body"){
2293 $attary = sq_body2div($attary, $mailbox,
2297 if (isset($open_tags{$tagname})){
2298 $open_tags{$tagname}++
;
2300 $open_tags{$tagname}=1;
2304 * This is where we run other checks.
2306 if (is_array($attary) && sizeof($attary) > 0){
2307 $attary = sq_fixatts($tagname,
2321 if ($tagname != false && $skip_content == false){
2322 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2327 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2328 if ($force_tag_closing == true){
2329 foreach ($open_tags as $tagname=>$opentimes){
2330 while ($opentimes > 0){
2331 $trusted .= '</' . $tagname . '>';
2337 $trusted .= "<!-- end sanitized html -->\n";
2342 * This is a wrapper function to call html sanitizing routines.
2344 * @param $body the body of the message
2345 * @param $id the id of the message
2349 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2350 * into internal SM compose links
2351 * (optional; default = TRUE)
2352 * @return a string with html safe to display in the browser.
2354 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2356 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2358 global $attachment_common_show_images, $view_unsafe_images,
2361 * Don't display attached images in HTML mode.
2365 $attachment_common_show_images = false;
2380 $rm_tags_with_content = Array(
2390 $self_closing_tags = Array(
2398 $force_tag_closing = true;
2400 $rm_attnames = Array(
2411 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2412 $bad_attvals = Array(
2415 "/^src|background/i" =>
2418 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2419 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2420 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2423 "\\1$secremoveimg\\2",
2424 "\\1$secremoveimg\\2",
2425 "\\1$secremoveimg\\2",
2428 "/^href|action/i" =>
2431 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2432 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2433 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2448 "/include-source/i",
2449 "/position\s*:\s*absolute/i",
2450 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2451 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2452 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2453 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2454 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2472 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
2473 $view_unsafe_images = false;
2475 if (!$view_unsafe_images){
2477 * Remove any references to http/https if view_unsafe_images set
2480 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2481 '/^([\'\"])\s*https*:.*([\'\"])/si');
2482 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2483 "\\1$secremoveimg\\1");
2484 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2485 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2486 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2487 "url(\\1$secremoveimg\\1)");
2490 $add_attr_to_tag = Array(
2492 Array('target'=>'"_blank"',
2493 'title'=>'"'._("This external link will open in a new window").'"'
2496 $trusted = sq_sanitize($body,
2498 $rm_tags_with_content,
2508 if (strpos($trusted,$secremoveimg)){
2509 $has_unsafe_images = true;
2512 // we want to parse mailto's in HTML output, change to SM compose links
2513 // this is a modified version of code from url_parser.php... but Marc is
2514 // right: we need a better filtering implementation; adding this randomly
2515 // here is not a great solution
2517 if ($take_mailto_links) {
2518 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2519 global $MailTo_PReg_Match;
2520 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2521 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2522 foreach ($regs[0] as $i => $mailto_before) {
2523 $mailto_params = $regs[10][$i];
2524 // get rid of any tailing quote since we have to add send_to to the end
2526 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2527 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2528 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2529 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2531 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2532 $to = 'to=' . $regs[1][$i];
2533 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2534 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2536 if ($mailto_params) //already some params, append to them
2537 $mailto_params .= '&' . $to;
2539 $mailto_params .= '?' . $to;
2543 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2545 // we'll already have target=_blank, no need to allow comp_in_new
2546 // here (which would be a lot more work anyway)
2548 global $compose_new_win;
2549 $temp_comp_in_new = $compose_new_win;
2550 $compose_new_win = 0;
2551 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2552 $compose_new_win = $temp_comp_in_new;
2554 // remove <a href=" and anything after the next quote (we only
2555 // need the uri, not the link HTML) in compose uri
2557 $comp_uri = substr($comp_uri, 9);
2558 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2559 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2568 * function SendDownloadHeaders - send file to the browser
2570 * Original Source: SM core src/download.php
2571 * moved here to make it available to other code, and separate
2572 * front end from back end functionality.
2574 * @param string $type0 first half of mime type
2575 * @param string $type1 second half of mime type
2576 * @param string $filename filename to tell the browser for downloaded file
2577 * @param boolean $force whether to force the download dialog to pop
2578 * @param optional integer $filesize send the Content-Header and length to the browser
2581 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2582 global $languages, $squirrelmail_language;
2583 $isIE = $isIE6plus = false;
2585 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2587 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2588 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2592 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2593 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2597 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2598 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2600 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2602 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace(' ', ' ', $filename));
2605 // A Pox on Microsoft and it's Internet Explorer!
2607 // IE has lots of bugs with file downloads.
2608 // It also has problems with SSL. Both of these cause problems
2609 // for us in this function.
2611 // See this article on Cache Control headers and SSL
2612 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2614 // The best thing you can do for IE is to upgrade to the latest
2616 //set all the Cache Control Headers for IE
2618 $filename=rawurlencode($filename);
2619 header ("Pragma: public");
2620 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2621 header ("Cache-Control: post-check=0, pre-check=0", false);
2622 header ("Cache-Control: private");
2624 //set the inline header for IE, we'll add the attachment header later if we need it
2625 header ("Content-Disposition: inline; filename=$filename");
2629 // Try to show in browser window
2630 header ("Content-Disposition: inline; filename=\"$filename\"");
2631 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2633 // Try to pop up the "save as" box
2635 // IE makes this hard. It pops up 2 save boxes, or none.
2636 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2637 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2638 // But, according to Microsoft, it is "RFC compliant but doesn't
2639 // take into account some deviations that allowed within the
2640 // specification." Doesn't that mean RFC non-compliant?
2641 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2643 // all browsers need the application/octet-stream header for this
2644 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2646 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2647 // Do not have quotes around filename, but that applied to
2648 // "attachment"... does it apply to inline too?
2649 header ("Content-Disposition: attachment; filename=\"$filename\"");
2651 if ($isIE && !$isIE6plus) {
2652 // This combination seems to work mostly. IE 5.5 SP 1 has
2653 // known issues (see the Microsoft Knowledge Base)
2655 // This works for most types, but doesn't work with Word files
2656 header ("Content-Type: application/download; name=\"$filename\"");
2657 header ("Content-Type: application/force-download; name=\"$filename\"");
2658 // These are spares, just in case. :-)
2659 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2660 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2661 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2663 // This is to prevent IE for MIME sniffing and auto open a file in IE
2664 header ("Content-Type: application/force-download; name=\"$filename\"");
2666 // another application/octet-stream forces download for Netscape
2667 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2671 //send the content-length header if the calling function provides it
2672 if ($filesize > 0) {
2673 header("Content-Length: $filesize");
2676 } // end fn SendDownloadHeaders