bf9701eeb6d353204d51a7fc2405c13eb767ccd8
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 $nbsp = $oTemplate->fetch('non_breaking_space.tpl');
361 // workaround for not updated config.php
362 if (! isset($use_iframe)) $use_iframe = false;
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('html_body', $html_body);
428 $body = $oTemplate->fetch('read_html_iframe.tpl');
430 // old way of html rendering
432 * convert character set. charset_decode does not remove html special chars
433 * applied by magicHTML functions and does not sanitize them second time if
434 * fourth argument is true.
436 $charset = $body_message->header
->getParameter('charset');
437 if (!empty($charset)) {
438 $body = charset_decode($charset,$body,false,true);
440 $body = magicHTML($body, $id, $message, $mailbox);
443 translateText($body, $wrap_at,
444 $body_message->header
->getParameter('charset'));
447 // if this is the clean display (i.e. printer friendly), stop here.
452 $download_and_unsafe_link = '';
454 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
455 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
456 '&startMessage=' . $startMessage . '&show_more=0';
457 if (isset($passed_ent_id)) {
458 $link .= '&passed_ent_id='.$passed_ent_id;
460 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
461 $download_and_unsafe_link .= "$nbsp|$nbsp"
462 . create_hyperlink($download_href, _("Download this as a file"));
463 if ($view_unsafe_images) {
464 $text = _("Hide Unsafe Images");
466 if (isset($has_unsafe_images) && $has_unsafe_images) {
467 $link .= '&view_unsafe_images=1';
468 $text = _("View Unsafe Images");
474 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
475 $unsafe_image_toggle_text = $text;
476 $download_and_unsafe_link .= "$nbsp|$nbsp"
477 . create_hyperlink($unsafe_image_toggle_href, $text);
484 * Generate attachments array for passing to templates. Separated from
485 * formatAttachments() below so that the same array can be given to the
486 * print-friendly version.
489 * @param object $message SquirrelMail message object
490 * @param array $exclude_id message parts that are not attachments.
491 * @param string $mailbox mailbox name
492 * @param integer $id message id
494 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
495 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
497 $att_ar = $message->getAttachments($exclude_id);
498 $urlMailbox = urlencode($mailbox);
500 $attachments = array();
501 foreach ($att_ar as $att) {
502 $ent = $att->entity_id
;
503 $header = $att->header
;
504 $type0 = strtolower($header->type0
);
505 $type1 = strtolower($header->type1
);
508 $links['download link']['text'] = _("Download");
509 $links['download link']['href'] = $base_uri .
510 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
512 if ($type0 =='message' && $type1 == 'rfc822') {
513 $default_page = $base_uri . 'src/read_body.php';
514 $rfc822_header = $att->rfc822_header
;
515 $filename = $rfc822_header->subject
;
516 if (trim( $filename ) == '') {
517 $filename = 'untitled-[' . $ent . ']' ;
519 $from_o = $rfc822_header->from
;
520 if (is_object($from_o)) {
521 $from_name = decodeHeader($from_o->getAddress(false));
522 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
523 // something weird happens when a digest message is opened and you return to the digest
524 // now the from object is part of an array. Probably the parseHeader call overwrites the info
525 // retrieved from the bodystructure in a different way. We need to fix this later.
526 // possible starting point, do not fetch header we already have and inspect how
527 // the rfc822_header object behaves.
528 $from_name = decodeHeader($from_o[0]->getAddress(false));
530 $from_name = _("Unknown sender");
532 $description = _("From").': '.$from_name;
534 $default_page = $base_uri . 'src/download.php';
535 $filename = $att->getFilename();
536 if ($header->description
) {
537 $description = decodeHeader($header->description
);
543 $display_filename = $filename;
544 if (isset($passed_ent_id)) {
545 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
547 $passed_ent_id_link = '';
549 $defaultlink = $default_page . "?startMessage=$startMessage"
550 . "&passed_id=$id&mailbox=$urlMailbox"
551 . '&ent_id='.$ent.$passed_ent_id_link;
552 if ($where && $what) {
553 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
555 // IE does make use of mime content sniffing. Forcing a download
556 // prohibit execution of XSS inside an application/octet-stream attachment
557 if ($type0 == 'application' && $type1 == 'octet-stream') {
558 $defaultlink .= '&absolute_dl=true';
561 /* This executes the attachment hook with a specific MIME-type.
562 * If that doesn't have results, it tries if there's a rule
563 * for a more generic type. Finally, a hook for ALL attachment
564 * types is run as well.
566 /* The API for this hook has changed as of 1.5.2 so that all plugin
567 arguments are passed in an array instead of each their own plugin
568 argument, and arguments are passed by reference, so instead of
569 returning any changes, changes should simply be made to the original
570 arguments themselves. */
571 do_hook("attachment $type0/$type1", $temp=array(&$links,
572 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
573 &$display_filename, &$where, &$what));
574 if(count($links) <= 1) {
575 /* The API for this hook has changed as of 1.5.2 so that all plugin
576 arguments are passed in an array instead of each their own plugin
577 argument, and arguments are passed by reference, so instead of
578 returning any changes, changes should simply be made to the original
579 arguments themselves. */
580 do_hook("attachment $type0/*", $temp=array(&$links,
581 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
582 &$display_filename, &$where, &$what));
584 /* The API for this hook has changed as of 1.5.2 so that all plugin
585 arguments are passed in an array instead of each their own plugin
586 argument, and arguments are passed by reference, so instead of
587 returning any changes, changes should simply be made to the original
588 arguments themselves. */
589 do_hook("attachment */*", $temp=array(&$links,
590 &$startMessage, &$id, &$urlMailbox, &$ent, &$defaultlink,
591 &$display_filename, &$where, &$what));
593 $this_attachment = array();
594 $this_attachment['Name'] = decodeHeader($display_filename);
595 $this_attachment['Description'] = $description;
596 $this_attachment['DefaultHREF'] = $defaultlink;
597 $this_attachment['DownloadHREF'] = $links['download link']['href'];
598 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
599 $this_attachment['Size'] = $header->size
;
600 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
601 $this_attachment['OtherLinks'] = array();
602 foreach ($links as $val) {
603 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
605 if (empty($val['text']) && empty($val['extra']))
609 $temp['HREF'] = $val['href'];
610 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
611 $this_attachment['OtherLinks'][] = $temp;
613 $attachments[] = $this_attachment;
622 * Displays attachment links and information
624 * Since 1.3.0 function is not included in formatBody() call.
626 * Since 1.0.2 uses attachment $type0/$type1 hook.
627 * Since 1.2.5 uses attachment $type0/* hook.
628 * Since 1.5.0 uses attachments_bottom hook.
629 * Since 1.5.2 uses templates and does *not* return a value.
631 * @param object $message SquirrelMail message object
632 * @param array $exclude_id message parts that are not attachments.
633 * @param string $mailbox mailbox name
634 * @param integer $id message id
636 function formatAttachments($message, $exclude_id, $mailbox, $id) {
639 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
641 $oTemplate->assign('attachments', $attach);
642 $oTemplate->display('read_attachments.tpl');
645 function sqimap_base64_decode(&$string) {
647 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
648 // fly decoding (to reduce memory usage) you have to check if the
649 // data has incomplete pairs
651 // Remove the noise in order to check if the 4 bytes pairs are complete
652 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
655 $iMod = strlen($string) %
4;
657 $sStringRem = substr($string,-$iMod);
658 // Check if $sStringRem contains padding characters
659 if (substr($sStringRem,-1) != '=') {
660 $string = substr($string,0,-$iMod);
665 $string = base64_decode($string);
670 * Decodes encoded message body
672 * This function decodes the body depending on the encoding type.
673 * Currently quoted-printable and base64 encodings are supported.
674 * decode_body hook was added to this function in 1.4.2/1.5.0
675 * @param string $body encoded message body
676 * @param string $encoding used encoding
677 * @return string decoded string
680 function decodeBody($body, $encoding) {
682 $body = str_replace("\r\n", "\n", $body);
683 $encoding = strtolower($encoding);
685 $encoding_handler = do_hook('decode_body', $encoding);
688 // plugins get first shot at decoding the body
690 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
691 $body = $encoding_handler('decode', $body);
693 } elseif ($encoding == 'quoted-printable' ||
694 $encoding == 'quoted_printable') {
696 * quoted_printable_decode() function is broken in older
697 * php versions. Text with \r\n decoding was fixed only
698 * in php 4.3.0. Minimal code requirement 4.0.4 +
699 * str_replace("\r\n", "\n", $body); call.
701 $body = quoted_printable_decode($body);
702 } elseif ($encoding == 'base64') {
703 $body = base64_decode($body);
706 // All other encodings are returned raw.
713 * This function decodes strings that are encoded according to
714 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
715 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
717 * @param string $string header string that has to be made readable
718 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
719 * @param boolean $htmlsave preserve spaces and sanitize html special characters. defaults to true
720 * @param boolean $decide decide if string can be utfencoded. defaults to false
721 * @return string decoded header string
723 function decodeHeader ($string, $utfencode=true,$htmlsave=true,$decide=false) {
724 global $languages, $squirrelmail_language,$default_charset;
725 if (is_array($string)) {
726 $string = implode("\n", $string);
729 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
730 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
731 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
732 // Do we need to return at this point?
739 $aString = explode(' ',$string);
741 foreach ($aString as $chunk) {
742 if ($encoded && $chunk === '') {
744 } elseif ($chunk === '') {
749 /* if encoded words are not separated by a linear-space-white we still catch them */
752 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
753 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
754 if ($iLastMatch !== $j) {
764 $ret .= htmlspecialchars($res[1]);
768 $encoding = ucfirst($res[3]);
770 /* decide about valid decoding */
771 if ($decide && is_conversion_safe($res[2])) {
773 $can_be_encoded=true;
775 $can_be_encoded=false;
780 $replace = base64_decode($res[4]);
782 if ($can_be_encoded) {
783 /* convert string to different charset,
784 * if functions asks for it (usually in compose)
786 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsave);
788 // convert string to html codes in order to display it
789 $ret .= charset_decode($res[2],$replace);
793 $replace = htmlspecialchars($replace);
799 $replace = str_replace('_', ' ', $res[4]);
800 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
803 if ($can_be_encoded) {
804 /* convert string to different charset,
805 * if functions asks for it (usually in compose)
807 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsave);
809 // convert string to html codes in order to display it
810 $replace = charset_decode($res[2], $replace);
814 $replace = htmlspecialchars($replace);
833 if (!$encoded && $htmlsave) {
834 $ret .= htmlspecialchars($chunk);
840 /* remove the first added space */
843 $ret = substr($ret,5);
845 $ret = substr($ret,1);
855 * Function uses XTRA_CODE _encodeheader function, if such function exists.
857 * Function uses Q encoding by default and encodes a string according to RFC
858 * 1522 for use in headers if it contains 8-bit characters or anything that
859 * looks like it should be encoded.
861 * Function switches to B encoding and encodeHeaderBase64() function, if
862 * string is 8bit and multibyte character set supported by mbstring extension
863 * is used. It can cause E_USER_NOTICE errors, if interface is used with
864 * multibyte character set unsupported by mbstring extension.
866 * @param string $string header string, that has to be encoded
867 * @return string quoted-printable encoded string
868 * @todo make $mb_charsets system wide constant
870 function encodeHeader ($string) {
871 global $default_charset, $languages, $squirrelmail_language;
873 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
874 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
875 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
878 // Use B encoding for multibyte charsets
879 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
880 if (in_array($default_charset,$mb_charsets) &&
881 in_array($default_charset,sq_mb_list_encodings()) &&
882 sq_is8bit($string)) {
883 return encodeHeaderBase64($string,$default_charset);
884 } elseif (in_array($default_charset,$mb_charsets) &&
885 sq_is8bit($string) &&
886 ! in_array($default_charset,sq_mb_list_encodings())) {
887 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
888 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
891 // Encode only if the string contains 8-bit characters or =?
892 $j = strlen($string);
893 $max_l = 75 - strlen($default_charset) - 7;
896 $iEncStart = $enc_init = false;
897 $cur_l = $iOffset = 0;
898 for($i = 0; $i < $j; ++
$i) {
907 if ($iEncStart === false) {
911 if ($cur_l > ($max_l-2)) {
912 /* if there is an stringpart that doesn't need encoding, add it */
913 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
914 $aRet[] = "=?$default_charset?Q?$ret?=";
920 $ret .= sprintf("=%02X",ord($string{$i}));
925 if ($iEncStart !== false) {
926 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
927 $aRet[] = "=?$default_charset?Q?$ret?=";
935 if ($iEncStart !== false) {
937 if ($cur_l > $max_l) {
938 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
939 $aRet[] = "=?$default_charset?Q?$ret?=";
950 $k = ord($string{$i});
952 if ($iEncStart === false) {
953 // do not start encoding in the middle of a string, also take the rest of the word.
954 $sLeadString = substr($string,0,$i);
955 $aLeadString = explode(' ',$sLeadString);
956 $sToBeEncoded = array_pop($aLeadString);
957 $iEncStart = $i - strlen($sToBeEncoded);
958 $ret .= $sToBeEncoded;
959 $cur_l +
= strlen($sToBeEncoded);
962 /* first we add the encoded string that reached it's max size */
963 if ($cur_l > ($max_l-2)) {
964 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
965 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
972 $ret .= sprintf("=%02X", $k);
974 if ($iEncStart !== false) {
976 if ($cur_l > $max_l) {
977 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
978 $aRet[] = "=?$default_charset?Q?$ret?=";
993 if ($iEncStart !== false) {
994 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
995 $aRet[] = "=?$default_charset?Q?$ret?=";
997 $aRet[] = substr($string,$iOffset);
999 $string = implode('',$aRet);
1005 * Encodes string according to rfc2047 B encoding header formating rules
1007 * It is recommended way to encode headers with character sets that store
1008 * symbols in more than one byte.
1010 * Function requires mbstring support. If required mbstring functions are missing,
1011 * function returns false and sets E_USER_WARNING level error message.
1013 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1014 * that mbstring functions will generate E_WARNING errors, if unsupported
1015 * character set is used. mb_encode_mimeheader function provided by php
1016 * mbstring extension is not used in order to get better control of header
1019 * Used php code functions - function_exists(), trigger_error(), strlen()
1020 * (is used with charset names and base64 strings). Used php mbstring
1021 * functions - mb_strlen and mb_substr.
1023 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1024 * encoding), rfc 2822 (header folding)
1026 * @param string $string header string that must be encoded
1027 * @param string $charset character set. Must be supported by mbstring extension.
1028 * Use sq_mb_list_encodings() to detect supported charsets.
1029 * @return string string encoded according to rfc2047 B encoding formating rules
1031 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1032 * @todo Do we want to control max length of header?
1033 * @todo Do we want to control EOL (end-of-line) marker?
1034 * @todo Do we want to translate error message?
1036 function encodeHeaderBase64($string,$charset) {
1038 * Check mbstring function requirements.
1040 if (! function_exists('mb_strlen') ||
1041 ! function_exists('mb_substr')) {
1042 // set E_USER_WARNING
1043 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1048 // initial return array
1052 * header length = 75 symbols max (same as in encodeHeader)
1053 * remove $charset length
1054 * remove =? ? ?= (5 chars)
1055 * remove 2 more chars (\r\n ?)
1057 $iMaxLength = 75 - strlen($charset) - 7;
1059 // set first character position
1062 // loop through all characters. count characters and not bytes.
1063 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1064 // encode string from starting character to current character.
1065 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1067 // Check encoded string length
1068 if(strlen($encoded_string)>$iMaxLength) {
1069 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1070 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1072 // set new starting character
1073 $iStartCharNum = $iCharNum-1;
1075 // encode last char (in case it is last character in string)
1076 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1077 } // if string is shorter than max length - add next character
1080 // add last encoded string to array
1081 $aRet[] = $encoded_string;
1083 // set initial return string
1086 // loop through encoded strings
1087 foreach($aRet as $string) {
1088 // TODO: Do we want to control EOL (end-of-line) marker
1089 if ($sRet!='') $sRet.= " ";
1091 // add header tags and encoded string to return string
1092 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1098 /* This function trys to locate the entity_id of a specific mime element */
1099 function find_ent_id($id, $message) {
1100 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1101 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1102 $ret = find_ent_id($id, $message->entities
[$i]);
1104 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1105 // if (sq_check_save_extension($message->entities[$i])) {
1106 return $message->entities
[$i]->entity_id
;
1108 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1110 * This is part of a fix for Outlook Express 6.x generating
1111 * cid URLs without creating content-id headers
1114 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1115 return $message->entities
[$i]->entity_id
;
1123 function sq_check_save_extension($message) {
1124 $filename = $message->getFilename();
1125 $ext = substr($filename, strrpos($filename,'.')+
1);
1126 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1127 return in_array($ext, $save_extensions);
1132 ** HTMLFILTER ROUTINES
1136 * This function checks attribute values for entity-encoded values
1137 * and returns them translated into 8-bit strings so we can run
1140 * @param $attvalue A string to run entity check against.
1141 * @return Nothing, modifies a reference value.
1143 function sq_defang(&$attvalue){
1146 * Skip this if there aren't ampersands or backslashes.
1148 if (strpos($attvalue, '&') === false
1149 && strpos($attvalue, '\\') === false){
1153 // before deent, translate the dangerous unicode characters and ... to safe values
1154 // otherwise the regular expressions do not match.
1160 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1161 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1162 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1163 } while ($m == true);
1164 $attvalue = stripslashes($attvalue);
1168 * Kill any tabs, newlines, or carriage returns. Our friends the
1169 * makers of the browser with 95% market value decided that it'd
1170 * be funny to make "java[tab]script" be just as good as "javascript".
1172 * @param attvalue The attribute value before extraneous spaces removed.
1173 * @return attvalue Nothing, modifies a reference value.
1175 function sq_unspace(&$attvalue){
1177 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1178 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1179 Array('', '', '', '', ''), $attvalue);
1184 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1185 * IE as regular characters.
1187 * @param attvalue The attribute value before dangerous characters are translated.
1188 * @return attvalue Nothing, modifies a reference value.
1189 * @author Marc Groot Koerkamp.
1191 function sq_fixIE_idiocy(&$attvalue) {
1193 $attvalue = str_replace("\0", "", $attvalue);
1195 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1197 // IE has the evil habit of accepting every possible value for the attribute expression.
1198 // The table below contains characters which are parsed by IE if they are used in the "expression"
1200 $aDangerousCharsReplacementTable = array(
1201 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1202 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1203 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1204 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1205 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1206 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1207 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1208 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1209 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1210 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1211 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1212 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1213 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1214 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1215 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1216 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1217 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1218 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1219 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1220 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1221 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1222 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1223 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1224 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1225 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1226 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1227 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1228 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1229 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1230 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1231 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1232 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1233 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1234 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1235 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1236 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1237 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1238 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1239 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1240 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1241 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1242 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1243 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1244 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1245 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1246 "\xCA\x9F", /* L UNICODE IPA Extension */
1247 "\xCA\x80", /* R UNICODE IPA Extension */
1248 "\xC9\xB4"), /* N UNICODE IPA Extension */
1249 array('l', 'l', 'r','r','n','n',
1250 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1251 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1252 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1253 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1255 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1257 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1261 * This function returns the final tag out of the tag name, an array
1262 * of attributes, and the type of the tag. This function is called by
1263 * sq_sanitize internally.
1265 * @param $tagname the name of the tag.
1266 * @param $attary the array of attributes and their values
1267 * @param $tagtype The type of the tag (see in comments).
1268 * @return a string with the final tag representation.
1270 function sq_tagprint($tagname, $attary, $tagtype){
1271 $me = 'sq_tagprint';
1274 $fulltag = '</' . $tagname . '>';
1276 $fulltag = '<' . $tagname;
1277 if (is_array($attary) && sizeof($attary)){
1279 while (list($attname, $attvalue) = each($attary)){
1280 array_push($atts, "$attname=$attvalue");
1282 $fulltag .= ' ' . join(" ", $atts);
1293 * A small helper function to use with array_walk. Modifies a by-ref
1294 * value and makes it lowercase.
1296 * @param $val a value passed by-ref.
1297 * @return void since it modifies a by-ref value.
1299 function sq_casenormalize(&$val){
1300 $val = strtolower($val);
1304 * This function skips any whitespace from the current position within
1305 * a string and to the next non-whitespace value.
1307 * @param $body the string
1308 * @param $offset the offset within the string where we should start
1309 * looking for the next non-whitespace character.
1310 * @return the location within the $body where the next
1311 * non-whitespace char is located.
1313 function sq_skipspace($body, $offset){
1314 $me = 'sq_skipspace';
1315 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1316 if (sizeof($matches{1})){
1317 $count = strlen($matches{1});
1324 * This function looks for the next character within a string. It's
1325 * really just a glorified "strpos", except it catches if failures
1328 * @param $body The string to look for needle in.
1329 * @param $offset Start looking from this position.
1330 * @param $needle The character/string to look for.
1331 * @return location of the next occurance of the needle, or
1332 * strlen($body) if needle wasn't found.
1334 function sq_findnxstr($body, $offset, $needle){
1335 $me = 'sq_findnxstr';
1336 $pos = strpos($body, $needle, $offset);
1337 if ($pos === FALSE){
1338 $pos = strlen($body);
1344 * This function takes a PCRE-style regexp and tries to match it
1345 * within the string.
1347 * @param $body The string to look for needle in.
1348 * @param $offset Start looking from here.
1349 * @param $reg A PCRE-style regex to match.
1350 * @return Returns a false if no matches found, or an array
1351 * with the following members:
1352 * - integer with the location of the match within $body
1353 * - string with whatever content between offset and the match
1354 * - string with whatever it is we matched
1356 function sq_findnxreg($body, $offset, $reg){
1357 $me = 'sq_findnxreg';
1360 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1361 if (!isset($matches{0}) ||
!$matches{0}){
1364 $retarr{0} = $offset +
strlen($matches{1});
1365 $retarr{1} = $matches{1};
1366 $retarr{2} = $matches{2};
1372 * This function looks for the next tag.
1374 * @param $body String where to look for the next tag.
1375 * @param $offset Start looking from here.
1376 * @return false if no more tags exist in the body, or
1377 * an array with the following members:
1378 * - string with the name of the tag
1379 * - array with attributes and their values
1380 * - integer with tag type (1, 2, or 3)
1381 * - integer where the tag starts (starting "<")
1382 * - integer where the tag ends (ending ">")
1383 * first three members will be false, if the tag is invalid.
1385 function sq_getnxtag($body, $offset){
1386 $me = 'sq_getnxtag';
1387 if ($offset > strlen($body)){
1390 $lt = sq_findnxstr($body, $offset, "<");
1391 if ($lt == strlen($body)){
1396 * blah blah <tag attribute="value">
1399 $pos = sq_skipspace($body, $lt+
1);
1400 if ($pos >= strlen($body)){
1401 return Array(false, false, false, $lt, strlen($body));
1404 * There are 3 kinds of tags:
1405 * 1. Opening tag, e.g.:
1407 * 2. Closing tag, e.g.:
1409 * 3. XHTML-style content-less tag, e.g.:
1410 * <img src="blah" />
1413 switch (substr($body, $pos, 1)){
1420 * A comment or an SGML declaration.
1422 if (substr($body, $pos+
1, 2) == "--"){
1423 $gt = strpos($body, "-->", $pos);
1425 $gt = strlen($body);
1429 return Array(false, false, false, $lt, $gt);
1431 $gt = sq_findnxstr($body, $pos, ">");
1432 return Array(false, false, false, $lt, $gt);
1437 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1447 * Look for next [\W-_], which will indicate the end of the tag name.
1449 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1450 if ($regary == false){
1451 return Array(false, false, false, $lt, strlen($body));
1453 list($pos, $tagname, $match) = $regary;
1454 $tagname = strtolower($tagname);
1457 * $match can be either of these:
1458 * '>' indicating the end of the tag entirely.
1459 * '\s' indicating the end of the tag name.
1460 * '/' indicating that this is type-3 xhtml tag.
1462 * Whatever else we find there indicates an invalid tag.
1467 * This is an xhtml-style tag with a closing / at the
1468 * end, like so: <img src="blah" />. Check if it's followed
1469 * by the closing bracket. If not, then this tag is invalid
1471 if (substr($body, $pos, 2) == "/>"){
1475 $gt = sq_findnxstr($body, $pos, ">");
1476 $retary = Array(false, false, false, $lt, $gt);
1480 return Array($tagname, false, $tagtype, $lt, $pos);
1484 * Check if it's whitespace
1486 if (!preg_match('/\s/', $match)){
1488 * This is an invalid tag! Look for the next closing ">".
1490 $gt = sq_findnxstr($body, $lt, ">");
1491 return Array(false, false, false, $lt, $gt);
1497 * At this point we're here:
1498 * <tagname attribute='blah'>
1501 * At this point we loop in order to find all attributes.
1507 while ($pos <= strlen($body)){
1508 $pos = sq_skipspace($body, $pos);
1509 if ($pos == strlen($body)){
1513 return Array(false, false, false, $lt, $pos);
1516 * See if we arrived at a ">" or "/>", which means that we reached
1517 * the end of the tag.
1520 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1524 $pos +
= strlen($matches{1});
1525 if ($matches{2} == "/>"){
1529 return Array($tagname, $attary, $tagtype, $lt, $pos);
1533 * There are several types of attributes, with optional
1534 * [:space:] between members.
1536 * attrname[:space:]=[:space:]'CDATA'
1538 * attrname[:space:]=[:space:]"CDATA"
1540 * attr[:space:]=[:space:]CDATA
1544 * We leave types 1 and 2 the same, type 3 we check for
1545 * '"' and convert to """ if needed, then wrap in
1546 * double quotes. Type 4 we convert into:
1549 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1550 if ($regary == false){
1552 * Looks like body ended before the end of tag.
1554 return Array(false, false, false, $lt, strlen($body));
1556 list($pos, $attname, $match) = $regary;
1557 $attname = strtolower($attname);
1559 * We arrived at the end of attribute name. Several things possible
1561 * '>' means the end of the tag and this is attribute type 4
1562 * '/' if followed by '>' means the same thing as above
1563 * '\s' means a lot of things -- look what it's followed by.
1564 * anything else means the attribute is invalid.
1569 * This is an xhtml-style tag with a closing / at the
1570 * end, like so: <img src="blah" />. Check if it's followed
1571 * by the closing bracket. If not, then this tag is invalid
1573 if (substr($body, $pos, 2) == "/>"){
1577 $gt = sq_findnxstr($body, $pos, ">");
1578 $retary = Array(false, false, false, $lt, $gt);
1582 $attary{$attname} = '"yes"';
1583 return Array($tagname, $attary, $tagtype, $lt, $pos);
1587 * Skip whitespace and see what we arrive at.
1589 $pos = sq_skipspace($body, $pos);
1590 $char = substr($body, $pos, 1);
1592 * Two things are valid here:
1593 * '=' means this is attribute type 1 2 or 3.
1594 * \w means this was attribute type 4.
1595 * anything else we ignore and re-loop. End of tag and
1596 * invalid stuff will be caught by our checks at the beginning
1601 $pos = sq_skipspace($body, $pos);
1603 * Here are 3 possibilities:
1604 * "'" attribute type 1
1605 * '"' attribute type 2
1606 * everything else is the content of tag type 3
1608 $quot = substr($body, $pos, 1);
1610 $regary = sq_findnxreg($body, $pos+
1, "\'");
1611 if ($regary == false){
1612 return Array(false, false, false, $lt, strlen($body));
1614 list($pos, $attval, $match) = $regary;
1616 $attary{$attname} = "'" . $attval . "'";
1617 } else if ($quot == '"'){
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 . '"';
1627 * These are hateful. Look for \s, or >.
1629 $regary = sq_findnxreg($body, $pos, "[\s>]");
1630 if ($regary == false){
1631 return Array(false, false, false, $lt, strlen($body));
1633 list($pos, $attval, $match) = $regary;
1635 * If it's ">" it will be caught at the top.
1637 $attval = preg_replace("/\"/s", """, $attval);
1638 $attary{$attname} = '"' . $attval . '"';
1640 } else if (preg_match("|[\w/>]|", $char)) {
1642 * That was attribute type 4.
1644 $attary{$attname} = '"yes"';
1647 * An illegal character. Find next '>' and return.
1649 $gt = sq_findnxstr($body, $pos, ">");
1650 return Array(false, false, false, $lt, $gt);
1656 * The fact that we got here indicates that the tag end was never
1657 * found. Return invalid tag indication so it gets stripped.
1659 return Array(false, false, false, $lt, strlen($body));
1663 * Translates entities into literal values so they can be checked.
1665 * @param $attvalue the by-ref value to check.
1666 * @param $regex the regular expression to check against.
1667 * @param $hex whether the entites are hexadecimal.
1668 * @return True or False depending on whether there were matches.
1670 function sq_deent(&$attvalue, $regex, $hex=false){
1674 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1675 preg_match_all($regex, $attvalue, $matches);
1676 if (is_array($matches) && sizeof($matches[0]) > 0){
1678 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1679 $numval = $matches[1][$i];
1681 $numval = hexdec($numval);
1683 $repl{$matches[0][$i]} = chr($numval);
1685 $attvalue = strtr($attvalue, $repl);
1693 * This function runs various checks against the attributes.
1695 * @param $tagname String with the name of the tag.
1696 * @param $attary Array with all tag attributes.
1697 * @param $rm_attnames See description for sq_sanitize
1698 * @param $bad_attvals See description for sq_sanitize
1699 * @param $add_attr_to_tag See description for sq_sanitize
1700 * @param $message message object
1701 * @param $id message id
1702 * @return Array with modified attributes.
1704 function sq_fixatts($tagname,
1714 while (list($attname, $attvalue) = each($attary)){
1716 * See if this attribute should be removed.
1718 foreach ($rm_attnames as $matchtag=>$matchattrs){
1719 if (preg_match($matchtag, $tagname)){
1720 foreach ($matchattrs as $matchattr){
1721 if (preg_match($matchattr, $attname)){
1722 unset($attary{$attname});
1729 * Workaround for IE quirks
1731 sq_fixIE_idiocy($attvalue);
1734 * Remove any backslashes, entities, and extraneous whitespace.
1737 $oldattvalue = $attvalue;
1738 sq_defang($attvalue);
1739 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1740 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1741 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1742 $attvalue = "idiocy";
1743 $attary{$attname} = $attvalue;
1745 sq_unspace($attvalue);
1748 * Now let's run checks on the attvalues.
1749 * I don't expect anyone to comprehend this. If you do,
1750 * get in touch with me so I can drive to where you live and
1751 * shake your hand personally. :)
1753 foreach ($bad_attvals as $matchtag=>$matchattrs){
1754 if (preg_match($matchtag, $tagname)){
1755 foreach ($matchattrs as $matchattr=>$valary){
1756 if (preg_match($matchattr, $attname)){
1758 * There are two arrays in valary.
1760 * Second one is replacements
1762 list($valmatch, $valrepl) = $valary;
1764 preg_replace($valmatch, $valrepl, $attvalue);
1765 if ($newvalue != $attvalue){
1766 $attary{$attname} = $newvalue;
1767 $attvalue = $newvalue;
1773 if ($attname == 'style') {
1774 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1775 // 8bit and control characters in style attribute values can be used for XSS, remove them
1776 $attary{$attname} = '"disallowed character"';
1778 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1779 if (count($aMatch)) {
1780 foreach($aMatch[1] as $sMatch) {
1782 $urlvalue = $sMatch;
1783 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1784 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1789 * Use white list based filtering on attributes which can contain url's
1791 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1792 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1793 $attary{$attname} = $attvalue;
1797 * See if we need to append any attributes to this tag.
1799 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1800 if (preg_match($matchtag, $tagname)){
1801 $attary = array_merge($attary, $addattary);
1808 * This function filters url's
1810 * @param $attvalue String with attribute value to filter
1811 * @param $message message object
1812 * @param $id message id
1813 * @param $mailbox mailbox
1814 * @param $sQuote quoting characters around url's
1816 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1817 $attvalue = trim($attvalue);
1818 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1819 // remove the double quotes
1820 $sQuote = $attvalue[0];
1821 $attvalue = trim(substr($attvalue,1,-1));
1824 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
1825 $view_unsafe_images = false;
1827 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1830 * Replace empty src tags with the blank image. src is only used
1831 * for frames, images, and image inputs. Doing a replace should
1832 * not affect them working as should be, however it will stop
1833 * IE from being kicked off when src for img tags are not set
1835 if ($attvalue == '') {
1836 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1838 // first, disallow 8 bit characters and control characters
1839 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1842 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1845 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1849 $aUrl = parse_url($attvalue);
1850 if (isset($aUrl['scheme'])) {
1851 switch(strtolower($aUrl['scheme'])) {
1856 if ($attname != 'href') {
1857 if ($view_unsafe_images == false) {
1858 $attvalue = $sQuote . $secremoveimg . $sQuote;
1860 if (isset($aUrl['path'])) {
1861 // validate image extension.
1862 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1863 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1864 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1867 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1874 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1875 * One day MS might actually make it match something useful, for now, falling
1876 * back to using cid2http, so we can grab the blank.png.
1878 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1882 * Turn cid: urls into http-friendly ones.
1884 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
1887 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1891 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
1892 // parse_url did not lead to satisfying result
1893 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
1901 * This function edits the style definition to make them friendly and
1902 * usable in SquirrelMail.
1904 * @param $message the message object
1905 * @param $id the message id
1906 * @param $content a string with whatever is between <style> and </style>
1907 * @param $mailbox the message mailbox
1908 * @return a string with edited content.
1910 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
1911 $me = 'sq_fixstyle';
1912 // workaround for </style> in between comments
1913 $iCurrentPos = $pos;
1918 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
1925 if ($sToken == '<') {
1935 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
1940 $content .= $sToken;
1948 if ($sToken == '<') {
1950 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
1951 $i = strpos($body,'-->',$i+
3);
1952 if ($i === false) { // no end comment
1970 if ($bSucces == FALSE){
1971 return array(FALSE, strlen($body));
1977 * First look for general BODY style declaration, which would be
1979 * body {background: blah-blah}
1980 * and change it to .bodyclass so we can just assign it to a <div>
1982 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
1983 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1985 * Fix url('blah') declarations.
1987 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
1988 // "url(\\1$secremoveimg\\2)", $content);
1990 // first check for 8bit sequences and disallowed control characters
1991 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
1992 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
1993 return array($content, $newpos);
1996 // IE Sucks hard. We have a special function for it.
1997 sq_fixIE_idiocy($content);
1999 // remove @import line
2000 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2002 // translate ur\l and variations (IE parses that)
2003 // TODO check if the sq_fixIE_idiocy function already handles this.
2004 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2005 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2006 if (count($aMatch)) {
2007 $aValue = $aReplace = array();
2008 foreach($aMatch[1] as $sMatch) {
2010 $urlvalue = $sMatch;
2011 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2012 $aValue[] = $sMatch;
2013 $aReplace[] = $urlvalue;
2015 $content = str_replace($aValue,$aReplace,$content);
2019 * Remove any backslashes, entities, and extraneous whitespace.
2021 $contentTemp = $content;
2022 sq_defang($contentTemp);
2023 sq_unspace($contentTemp);
2026 * Fix stupid css declarations which lead to vulnerabilities
2029 $match = Array('/\/\*.*\*\//',
2033 '/include-source/i',
2036 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2037 $contentNew = preg_replace($match, $replace, $contentTemp);
2038 if ($contentNew !== $contentTemp) {
2039 // insecure css declarations are used. From now on we don't care
2040 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2041 $content = $contentNew;
2043 return array($content, $newpos);
2048 * This function converts cid: url's into the ones that can be viewed in
2051 * @param $message the message object
2052 * @param $id the message id
2053 * @param $cidurl the cid: url.
2054 * @param $mailbox the message mailbox
2055 * @return a string with a http-friendly url
2057 function sq_cid2http($message, $id, $cidurl, $mailbox){
2059 * Get rid of quotes.
2061 $quotchar = substr($cidurl, 0, 1);
2062 if ($quotchar == '"' ||
$quotchar == "'"){
2063 $cidurl = str_replace($quotchar, "", $cidurl);
2067 $cidurl = substr(trim($cidurl), 4);
2069 $match_str = '/\{.*?\}\//';
2071 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2073 $linkurl = find_ent_id($cidurl, $message);
2074 /* in case of non-save cid links $httpurl should be replaced by a sort of
2075 unsave link image */
2079 * This is part of a fix for Outlook Express 6.x generating
2080 * cid URLs without creating content-id headers. These images are
2081 * not part of the multipart/related html mail. The html contains
2082 * <img src="cid:{some_id}/image_filename.ext"> references to
2083 * attached images with as goal to render them inline although
2084 * the attachment disposition property is not inline.
2087 if (empty($linkurl)) {
2088 if (preg_match('/{.*}\//', $cidurl)) {
2089 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2090 if (!empty($cidurl)) {
2091 $linkurl = find_ent_id($cidurl, $message);
2096 if (!empty($linkurl)) {
2097 $httpurl = $quotchar . SM_PATH
. 'src/download.php?absolute_dl=true&' .
2098 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2099 '&ent_id=' . $linkurl . $quotchar;
2102 * If we couldn't generate a proper img url, drop in a blank image
2103 * instead of sending back empty, otherwise it causes unusual behaviour
2105 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2112 * This function changes the <body> tag into a <div> tag since we
2113 * can't really have a body-within-body.
2115 * @param $attary an array of attributes and values of <body>
2116 * @param $mailbox mailbox we're currently reading (for cid2http)
2117 * @param $message current message (for cid2http)
2118 * @param $id current message id (for cid2http)
2119 * @return a modified array of attributes to be set for <div>
2121 function sq_body2div($attary, $mailbox, $message, $id){
2122 $me = 'sq_body2div';
2123 $divattary = Array('class' => "'bodyclass'");
2125 $has_bgc_stl = $has_txt_stl = false;
2127 if (is_array($attary) && sizeof($attary) > 0){
2128 foreach ($attary as $attname=>$attvalue){
2129 $quotchar = substr($attvalue, 0, 1);
2130 $attvalue = str_replace($quotchar, "", $attvalue);
2133 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2134 $styledef .= "background-image: url('$attvalue'); ";
2137 $has_bgc_stl = true;
2138 $styledef .= "background-color: $attvalue; ";
2141 $has_txt_stl = true;
2142 $styledef .= "color: $attvalue; ";
2146 // Outlook defines a white bgcolor and no text color. This can lead to
2147 // white text on a white bg with certain themes.
2148 if ($has_bgc_stl && !$has_txt_stl) {
2149 $styledef .= "color: $text; ";
2151 if (strlen($styledef) > 0){
2152 $divattary{"style"} = "\"$styledef\"";
2159 * This is the main function and the one you should actually be calling.
2160 * There are several variables you should be aware of an which need
2161 * special description.
2163 * Since the description is quite lengthy, see it here:
2164 * http://linux.duke.edu/projects/mini/htmlfilter/
2166 * @param $body the string with HTML you wish to filter
2167 * @param $tag_list see description above
2168 * @param $rm_tags_with_content see description above
2169 * @param $self_closing_tags see description above
2170 * @param $force_tag_closing see description above
2171 * @param $rm_attnames see description above
2172 * @param $bad_attvals see description above
2173 * @param $add_attr_to_tag see description above
2174 * @param $message message object
2175 * @param $id message id
2176 * @return sanitized html safe to show on your pages.
2178 function sq_sanitize($body,
2180 $rm_tags_with_content,
2190 $me = 'sq_sanitize';
2191 $rm_tags = array_shift($tag_list);
2193 * Normalize rm_tags and rm_tags_with_content.
2195 @array_walk
($tag_list, 'sq_casenormalize');
2196 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2197 @array_walk
($self_closing_tags, 'sq_casenormalize');
2199 * See if tag_list is of tags to remove or tags to allow.
2200 * false means remove these tags
2201 * true means allow these tags
2204 $open_tags = Array();
2205 $trusted = "\n<!-- begin sanitized html -->\n";
2206 $skip_content = false;
2208 * Take care of netscape's stupid javascript entities like
2211 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2213 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2214 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2215 $free_content = substr($body, $curpos, $lt-$curpos);
2217 * Take care of <style>
2219 if ($tagname == "style" && $tagtype == 1){
2220 list($free_content, $curpos) =
2221 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2222 if ($free_content != FALSE){
2223 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2224 $trusted .= $free_content;
2225 $trusted .= sq_tagprint($tagname, false, 2);
2229 if ($skip_content == false){
2230 $trusted .= $free_content;
2232 if ($tagname != FALSE){
2234 if ($skip_content == $tagname){
2236 * Got to the end of tag we needed to remove.
2239 $skip_content = false;
2241 if ($skip_content == false){
2242 if ($tagname == "body"){
2245 if (isset($open_tags{$tagname}) &&
2246 $open_tags{$tagname} > 0){
2247 $open_tags{$tagname}--;
2255 * $rm_tags_with_content
2257 if ($skip_content == false){
2259 * See if this is a self-closing type and change
2260 * tagtype appropriately.
2263 && in_array($tagname, $self_closing_tags)){
2267 * See if we should skip this tag and any content
2270 if ($tagtype == 1 &&
2271 in_array($tagname, $rm_tags_with_content)){
2272 $skip_content = $tagname;
2274 if (($rm_tags == false
2275 && in_array($tagname, $tag_list)) ||
2276 ($rm_tags == true &&
2277 !in_array($tagname, $tag_list))){
2281 * Convert body into div.
2283 if ($tagname == "body"){
2285 $attary = sq_body2div($attary, $mailbox,
2289 if (isset($open_tags{$tagname})){
2290 $open_tags{$tagname}++
;
2292 $open_tags{$tagname}=1;
2296 * This is where we run other checks.
2298 if (is_array($attary) && sizeof($attary) > 0){
2299 $attary = sq_fixatts($tagname,
2313 if ($tagname != false && $skip_content == false){
2314 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2319 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2320 if ($force_tag_closing == true){
2321 foreach ($open_tags as $tagname=>$opentimes){
2322 while ($opentimes > 0){
2323 $trusted .= '</' . $tagname . '>';
2329 $trusted .= "<!-- end sanitized html -->\n";
2334 * This is a wrapper function to call html sanitizing routines.
2336 * @param $body the body of the message
2337 * @param $id the id of the message
2341 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2342 * into internal SM compose links
2343 * (optional; default = TRUE)
2344 * @return a string with html safe to display in the browser.
2346 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2348 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2350 global $attachment_common_show_images, $view_unsafe_images,
2353 * Don't display attached images in HTML mode.
2357 $attachment_common_show_images = false;
2372 $rm_tags_with_content = Array(
2382 $self_closing_tags = Array(
2390 $force_tag_closing = true;
2392 $rm_attnames = Array(
2403 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2404 $bad_attvals = Array(
2407 "/^src|background/i" =>
2410 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2411 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2412 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2415 "\\1$secremoveimg\\2",
2416 "\\1$secremoveimg\\2",
2417 "\\1$secremoveimg\\2",
2420 "/^href|action/i" =>
2423 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2424 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2425 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2440 "/include-source/i",
2441 "/position\s*:\s*absolute/i",
2442 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2443 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2444 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2445 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2446 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2464 if( !sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
) ) {
2465 $view_unsafe_images = false;
2467 if (!$view_unsafe_images){
2469 * Remove any references to http/https if view_unsafe_images set
2472 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2473 '/^([\'\"])\s*https*:.*([\'\"])/si');
2474 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2475 "\\1$secremoveimg\\1");
2476 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2477 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2478 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2479 "url(\\1$secremoveimg\\1)");
2482 $add_attr_to_tag = Array(
2484 Array('target'=>'"_blank"',
2485 'title'=>'"'._("This external link will open in a new window").'"'
2488 $trusted = sq_sanitize($body,
2490 $rm_tags_with_content,
2500 if (strpos($trusted,$secremoveimg)){
2501 $has_unsafe_images = true;
2504 // we want to parse mailto's in HTML output, change to SM compose links
2505 // this is a modified version of code from url_parser.php... but Marc is
2506 // right: we need a better filtering implementation; adding this randomly
2507 // here is not a great solution
2509 if ($take_mailto_links) {
2510 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2511 global $MailTo_PReg_Match;
2512 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2513 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2514 foreach ($regs[0] as $i => $mailto_before) {
2515 $mailto_params = $regs[10][$i];
2516 // get rid of any tailing quote since we have to add send_to to the end
2518 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2519 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2520 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2521 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2523 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2524 $to = 'to=' . $regs[1][$i];
2525 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2526 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2528 if ($mailto_params) //already some params, append to them
2529 $mailto_params .= '&' . $to;
2531 $mailto_params .= '?' . $to;
2535 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2537 // we'll already have target=_blank, no need to allow comp_in_new
2538 // here (which would be a lot more work anyway)
2540 global $compose_new_win;
2541 $temp_comp_in_new = $compose_new_win;
2542 $compose_new_win = 0;
2543 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2544 $compose_new_win = $temp_comp_in_new;
2546 // remove <a href=" and anything after the next quote (we only
2547 // need the uri, not the link HTML) in compose uri
2549 $comp_uri = substr($comp_uri, 9);
2550 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2551 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2560 * function SendDownloadHeaders - send file to the browser
2562 * Original Source: SM core src/download.php
2563 * moved here to make it available to other code, and separate
2564 * front end from back end functionality.
2566 * @param string $type0 first half of mime type
2567 * @param string $type1 second half of mime type
2568 * @param string $filename filename to tell the browser for downloaded file
2569 * @param boolean $force whether to force the download dialog to pop
2570 * @param optional integer $filesize send the Content-Header and length to the browser
2573 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2574 global $languages, $squirrelmail_language;
2575 $isIE = $isIE6plus = false;
2577 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2579 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2580 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2584 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2585 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2589 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2590 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2592 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2594 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace(' ', ' ', $filename));
2597 // A Pox on Microsoft and it's Internet Explorer!
2599 // IE has lots of bugs with file downloads.
2600 // It also has problems with SSL. Both of these cause problems
2601 // for us in this function.
2603 // See this article on Cache Control headers and SSL
2604 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2606 // The best thing you can do for IE is to upgrade to the latest
2608 //set all the Cache Control Headers for IE
2610 $filename=rawurlencode($filename);
2611 header ("Pragma: public");
2612 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2613 header ("Cache-Control: post-check=0, pre-check=0", false);
2614 header ("Cache-Control: private");
2616 //set the inline header for IE, we'll add the attachment header later if we need it
2617 header ("Content-Disposition: inline; filename=$filename");
2621 // Try to show in browser window
2622 header ("Content-Disposition: inline; filename=\"$filename\"");
2623 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2625 // Try to pop up the "save as" box
2627 // IE makes this hard. It pops up 2 save boxes, or none.
2628 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2629 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2630 // But, according to Microsoft, it is "RFC compliant but doesn't
2631 // take into account some deviations that allowed within the
2632 // specification." Doesn't that mean RFC non-compliant?
2633 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2635 // all browsers need the application/octet-stream header for this
2636 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2638 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2639 // Do not have quotes around filename, but that applied to
2640 // "attachment"... does it apply to inline too?
2641 header ("Content-Disposition: attachment; filename=\"$filename\"");
2643 if ($isIE && !$isIE6plus) {
2644 // This combination seems to work mostly. IE 5.5 SP 1 has
2645 // known issues (see the Microsoft Knowledge Base)
2647 // This works for most types, but doesn't work with Word files
2648 header ("Content-Type: application/download; name=\"$filename\"");
2649 header ("Content-Type: application/force-download; name=\"$filename\"");
2650 // These are spares, just in case. :-)
2651 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2652 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2653 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2655 // This is to prevent IE for MIME sniffing and auto open a file in IE
2656 header ("Content-Type: application/force-download; name=\"$filename\"");
2658 // another application/octet-stream forces download for Netscape
2659 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2663 //send the content-length header if the calling function provides it
2664 if ($filesize > 0) {
2665 header("Content-Length: $filesize");
2668 } // end fn SendDownloadHeaders