6 * This contains the functions necessary to detect and decode MIME
9 * @copyright 1999-2019 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 $errormessage .= '<pre>' . sm_encode_html_special_chars($read) . '</pre>';
67 plain_error_message( $errormessage );
68 echo '</body></html>';
72 foreach ($flags as $flag) {
73 //FIXME: please document why it is we have to check the first char of the flag but we then go ahead and do a full string comparison anyway. Is this a speed enhancement? If not, let's keep it simple and just compare the full string and forget the switch block.
74 $char = strtoupper($flag{1});
77 if (strtolower($flag) == '\\seen') {
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered
= true;
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted
= true;
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged
= true;
95 else if (strtolower($flag) == '$forwarded') {
96 $msg->is_forwarded
= true;
100 if (strtolower($flag) == '$mdnsent') {
101 $msg->is_mdnsent
= true;
109 // listEntities($msg);
115 /* This starts the parsing of a particular structure. It is called recursively,
116 * so it can be passed different structures. It returns an object of type
118 * First, it checks to see if it is a multipart message. If it is, then it
119 * handles that as it sees is necessary. If it is just a regular entity,
120 * then it parses it and adds the necessary header information (by calling out
121 * to mime_get_elements()
124 function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
125 /* Do a bit of error correction. If we couldn't find the entity id, just guess
126 * that it is the first one. That is usually the case anyway.
130 $cmd = "FETCH $id BODY[]";
132 $cmd = "FETCH $id BODY[$ent_id]";
135 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
137 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
139 $topline = trim(array_shift($data));
140 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH .*BODY.*/i', $topline)) ;
141 // Matching with "BODY" above is difficult: in most cases "FETCH \(BODY" would work
142 // but some servers may put other things in the same result, perhaps something such
143 // as "* 23 FETCH (FLAGS (\Seen) BODY[1] {174}". There is some small chance that
144 // if the character sequence "BODY" appears in a response where it isn't actually
145 // a FETCH response data item name, the current regex will break things. The better
146 // way to do this would be to parse the response correctly and not use a regex.
148 $wholemessage = implode('', $data);
149 if (preg_match('/\{([^\}]*)\}/', $topline, $regs)) {
150 $ret = substr($wholemessage, 0, $regs[1]);
151 /* There is some information in the content info header that could be important
152 * in order to parse html messages. Let's get them here.
154 // if ($ret{0} == '<') {
155 // $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
157 } else if (preg_match('/"([^"]*)"/', $topline, $regs)) {
159 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
160 $ret = $wholemessage;
162 global $where, $what, $mailbox, $passed_id, $startMessage;
163 $par = 'mailbox=' . urlencode($mailbox) . '&passed_id=' . $passed_id;
164 if (isset($where) && isset($what)) {
165 $par .= '&where=' . urlencode($where) . '&what=' . urlencode($what);
167 $par .= '&startMessage=' . $startMessage . '&show_more=0';
169 $par .= '&response=' . urlencode($response) .
170 '&message=' . urlencode($message) .
171 '&topline=' . urlencode($topline);
174 '<table width="80%"><tr>' .
175 '<tr><td colspan="2">' .
176 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
178 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
179 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
180 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
181 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
182 "</table><br /></tt></font><hr />";
184 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
186 $wholemessage = implode('', $data);
188 $ret = $wholemessage;
193 function mime_print_body_lines ($imap_stream, $id, $ent_id=1, $encoding, $rStream='php://stdout', $force_crlf='') {
195 /* Don't kill the connection if the browser is over a dialup
196 * and it would take over 30 seconds to download it.
197 * Don't call set_time_limit in safe mode.
200 if (!ini_get('safe_mode')) {
203 /* in case of base64 encoded attachments, do not buffer them.
204 Instead, echo the decoded attachment directly to screen */
205 if (strtolower($encoding) == 'base64') {
207 $query = "FETCH $id BODY[]";
209 $query = "FETCH $id BODY[$ent_id]";
211 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
213 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
214 if (is_resource($rStream)) {
215 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
217 echo decodeBody($body, $encoding, $force_crlf);
222 TODO, use the same method for quoted printable.
223 However, I assume that quoted printable attachments aren't that large
224 so the performancegain / memory usage drop will be minimal.
225 If we decide to add that then we need to adapt sqimap_fread because
226 we need to split te result on \n and fread doesn't stop at \n. That
227 means we also should provide $results from sqimap_fread (by ref) to
228 te function and set $no_return to false. The $filter function for
229 quoted printable should handle unsetting of $results.
232 TODO 2: find out how we write to the output stream php://stdout. fwrite
233 doesn't work because 'php://stdout isn't a stream.
239 /* -[ END MIME DECODING ]----------------------------------------------------------- */
241 /* This is here for debugging purposes. It will print out a list
242 * of all the entity IDs that are in the $message object.
244 function listEntities ($message) {
246 echo "<tt>" . $message->entity_id
. ' : ' . $message->type0
. '/' . $message->type1
. ' parent = '. $message->parent
->entity_id
. '<br />';
247 for ($i = 0; isset($message->entities
[$i]); $i++
) {
249 $msg = listEntities($message->entities
[$i]);
259 function getPriorityStr($priority) {
260 $priority_level = substr($priority,0,1);
262 switch($priority_level) {
263 /* Check for a higher then normal priority. */
266 $priority_string = _("High");
269 /* Check for a lower then normal priority. */
272 $priority_string = _("Low");
275 /* Check for a normal priority. */
278 $priority_level = '3';
279 $priority_string = _("Normal");
283 return $priority_string;
286 /* returns a $message object for a particular entity id */
287 function getEntity ($message, $ent_id) {
288 return $message->getEntity($ent_id);
292 * Extracted from strings.php 23/03/2002
295 function translateText(&$body, $wrap_at, $charset) {
296 global $where, $what; /* from searching */
297 global $color; /* color theme */
299 // require_once(SM_PATH . 'functions/url_parser.php');
301 $body_ary = explode("\n", $body);
302 for ($i=0; $i < count($body_ary); $i++
) {
303 $line = rtrim($body_ary[$i],"\r");
305 if (strlen($line) - 2 >= $wrap_at) {
306 sqWordWrap($line, $wrap_at, $charset);
308 $line = charset_decode($charset, $line);
309 $line = str_replace("\t", ' ', $line);
318 if ($line[$pos] == ' ') {
320 } else if (strpos($line, '>', $pos) === $pos) {
329 $line = '<span class="quote1">' . $line . '</span>';
331 $line = '<span class="quote2">' . $line . '</span>';
334 $body_ary[$i] = $line;
336 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
340 * This returns a parsed string called $body. That string can then
341 * be displayed as the actual message in the HTML. It contains
342 * everything needed, including HTML Tags, Attachments at the
345 * Since 1.2.0 function uses message_body hook.
346 * Till 1.3.0 function included output of formatAttachments().
348 * @param resource $imap_stream imap connection resource
349 * @param object $message squirrelmail message object
350 * @param array $color squirrelmail color theme array
351 * @param integer $wrap_at number of characters per line
352 * @param string $ent_num (since 1.3.0) message part id
353 * @param integer $id (since 1.3.0) message id
354 * @param string $mailbox (since 1.3.0) imap folder name
355 * @return string html formated message text
357 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
358 /* This if statement checks for the entity to show as the
359 * primary message. To add more of them, just put them in the
360 * order that is their priority.
362 global $startMessage, $languages, $squirrelmail_language,
363 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
364 $use_iframe, $iframe_height, $download_and_unsafe_link,
365 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
368 // workaround for not updated config.php
369 if (! isset($use_iframe)) $use_iframe = false;
371 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
372 // images off by default.
373 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
376 $urlmailbox = urlencode($mailbox);
377 $body_message = getEntity($message, $ent_num);
378 if (($body_message->header
->type0
== 'text') ||
379 ($body_message->header
->type0
== 'rfc822')) {
380 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
381 $body = decodeBody($body, $body_message->header
->encoding
);
383 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
384 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
385 if (mb_detect_encoding($body) != 'ASCII') {
386 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
390 /* As of 1.5.2, $body is passed (and modified) by reference */
391 do_hook('message_body', $body);
393 /* If there are other types that shouldn't be formatted, add
397 if ($body_message->header
->type1
== 'html') {
398 if ($show_html_default <> 1) {
399 $entity_conv = array(' ' => ' ',
408 $body = strtr($body, $entity_conv);
409 $body = strip_tags($body);
411 translateText($body, $wrap_at,
412 $body_message->header
->getParameter('charset'));
413 } elseif ($use_iframe) {
415 * If we don't add html message between iframe tags,
416 * we must detect unsafe images and modify $has_unsafe_images.
418 $html_body = magicHTML($body, $id, $message, $mailbox);
419 // Convert character set in order to display html mails in different character set
420 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
422 // creating iframe url
423 $iframeurl=sqm_baseuri().'src/view_html.php?'
424 . 'mailbox=' . $urlmailbox
425 . '&passed_id=' . $id
426 . '&ent_id=' . $ent_num
427 . '&view_unsafe_images=' . (int) $view_unsafe_images;
430 $oTemplate->assign('iframe_url', $iframeurl);
431 $oTemplate->assign('iframe_height', $iframe_height);
432 $oTemplate->assign('html_body', $html_body);
434 $body = $oTemplate->fetch('read_html_iframe.tpl');
436 // old way of html rendering
438 * convert character set. charset_decode does not remove html special chars
439 * applied by magicHTML functions and does not sanitize them second time if
440 * fourth argument is true.
442 $charset = $body_message->header
->getParameter('charset');
443 if (!empty($charset)) {
444 $body = charset_decode($charset,$body,false,true);
446 $body = magicHTML($body, $id, $message, $mailbox);
449 translateText($body, $wrap_at,
450 $body_message->header
->getParameter('charset'));
454 * Previously the links for downloading and unsafe images were printed
455 * under the mail. By putting the links in a global variable we can
456 * print it in the toolbar where it belongs. Since the original code was
457 * in this place it's left here. It might be possible to move it to some
458 * other place if that makes sense. The possibility to do so has not
459 * been evaluated yet.
462 // Initialize the global variable to an empty string.
463 // FIXME: To have $download_and_unsafe_link as a global variable might not be needed since the use of separate variables ($download_href, $unsafe_image_toggle_href, and $unsafe_image_toggle_text) for the templates was introduced.
464 $download_and_unsafe_link = '';
466 // Prepare and build a link for downloading the mail.
467 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
468 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
469 '&startMessage=' . $startMessage . '&show_more=0';
470 if (isset($passed_ent_id)) {
471 $link .= '&passed_ent_id='.$passed_ent_id;
473 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
475 // Always add the link for downloading the mail as a file to the global
477 $download_and_unsafe_link .= "$nbsp|$nbsp"
478 . create_hyperlink($download_href, _("Download this as a file"));
480 // Find out the right text to use in the link depending on the
481 // circumstances. If the unsafe images are displayed the link should
482 // hide them, if they aren't displayed the link should only appear if
483 // the mail really contains unsafe images.
484 if ($view_unsafe_images) {
485 $text = _("Hide Unsafe Images");
487 if (isset($has_unsafe_images) && $has_unsafe_images) {
488 $link .= '&view_unsafe_images=1';
489 $text = _("View Unsafe Images");
495 // Only create a link for unsafe images if there's need for one. If so:
496 // add it to the global variable.
498 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
499 $unsafe_image_toggle_text = $text;
500 $download_and_unsafe_link .= "$nbsp|$nbsp"
501 . create_hyperlink($unsafe_image_toggle_href, $text);
508 * Generate attachments array for passing to templates.
511 * @param object $message SquirrelMail message object
512 * @param array $exclude_id message parts that are not attachments.
513 * @param string $mailbox mailbox name
514 * @param integer $id message id
516 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
517 global $where, $what, $startMessage, $color, $passed_ent_id,
518 $base_uri, $block_svg_download;
520 $att_ar = $message->getAttachments($exclude_id);
521 $urlMailbox = urlencode($mailbox);
523 $attachments = array();
524 foreach ($att_ar as $att) {
525 $ent = $att->entity_id
;
526 $header = $att->header
;
527 $type0 = strtolower($header->type0
);
528 $type1 = strtolower($header->type1
);
529 if ($block_svg_download && strpos($type1, 'svg') === 0)
534 $links['download link']['text'] = _("Download");
535 $links['download link']['href'] = $base_uri .
536 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
538 if ($type0 =='message' && $type1 == 'rfc822') {
539 $default_page = $base_uri . 'src/read_body.php';
540 $rfc822_header = $att->rfc822_header
;
541 $filename = $rfc822_header->subject
;
542 if (trim( $filename ) == '') {
543 $filename = 'untitled-[' . $ent . ']' ;
545 $from_o = $rfc822_header->from
;
546 if (is_object($from_o)) {
547 $from_name = decodeHeader($from_o->getAddress(false));
548 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
549 // something weird happens when a digest message is opened and you return to the digest
550 // now the from object is part of an array. Probably the parseHeader call overwrites the info
551 // retrieved from the bodystructure in a different way. We need to fix this later.
552 // possible starting point, do not fetch header we already have and inspect how
553 // the rfc822_header object behaves.
554 $from_name = decodeHeader($from_o[0]->getAddress(false));
556 $from_name = _("Unknown sender");
558 $description = _("From").': '.$from_name;
560 $default_page = $base_uri . 'src/download.php';
561 $filename = $att->getFilename();
562 if ($header->description
) {
563 $description = decodeHeader($header->description
);
569 $display_filename = $filename;
570 if (isset($passed_ent_id)) {
571 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
573 $passed_ent_id_link = '';
575 $defaultlink = $default_page . "?startMessage=$startMessage"
576 . "&passed_id=$id&mailbox=$urlMailbox"
577 . '&ent_id='.$ent.$passed_ent_id_link;
578 if ($where && $what) {
579 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
581 // IE does make use of mime content sniffing. Forcing a download
582 // prohibit execution of XSS inside an application/octet-stream attachment
583 if ($type0 == 'application' && $type1 == 'octet-stream') {
584 $defaultlink .= '&absolute_dl=true';
587 /* This executes the attachment hook with a specific MIME-type.
588 * It also allows plugins to run if there's a rule for a more
589 * generic type. Finally, a hook for ALL attachment types is
592 // First remember the default link.
593 $defaultlink_orig = $defaultlink;
595 /* The API for this hook has changed as of 1.5.2 so that all plugin
596 arguments are passed in an array instead of each their own plugin
597 argument, and arguments are passed by reference, so instead of
598 returning any changes, changes should simply be made to the original
599 arguments themselves. */
600 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
601 &$defaultlink, &$display_filename, &$where, &$what);
602 do_hook("attachment $type0/$type1", $temp);
603 /* The API for this hook has changed as of 1.5.2 so that all plugin
604 arguments are passed in an array instead of each their own plugin
605 argument, and arguments are passed by reference, so instead of
606 returning any changes, changes should simply be made to the original
607 arguments themselves. */
608 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
609 &$defaultlink, &$display_filename, &$where, &$what);
610 // Do not let a generic plugin change the default link if a more
611 // specialized one already did it...
612 if ($defaultlink != $defaultlink_orig) {
616 do_hook("attachment $type0/*", $temp);
617 /* The API for this hook has changed as of 1.5.2 so that all plugin
618 arguments are passed in an array instead of each their own plugin
619 argument, and arguments are passed by reference, so instead of
620 returning any changes, changes should simply be made to the original
621 arguments themselves. */
622 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
623 &$defaultlink, &$display_filename, &$where, &$what);
624 // Do not let a generic plugin change the default link if a more
625 // specialized one already did it...
626 if ($defaultlink != $defaultlink_orig) {
630 do_hook("attachment */*", $temp);
632 $this_attachment = array();
633 $this_attachment['Name'] = decodeHeader($display_filename);
634 $this_attachment['Description'] = $description;
635 $this_attachment['DefaultHREF'] = $defaultlink;
636 $this_attachment['DownloadHREF'] = $links['download link']['href'];
637 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
638 $this_attachment['Size'] = $header->size
;
639 $this_attachment['ContentType'] = sm_encode_html_special_chars($type0 .'/'. $type1);
640 $this_attachment['OtherLinks'] = array();
641 foreach ($links as $val) {
642 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
644 if (empty($val['text']) && empty($val['extra']))
648 $temp['HREF'] = $val['href'];
649 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
650 $this_attachment['OtherLinks'][] = $temp;
652 $attachments[] = $this_attachment;
661 * Displays attachment links and information
663 * Since 1.3.0 function is not included in formatBody() call.
665 * Since 1.0.2 uses attachment $type0/$type1 hook.
666 * Since 1.2.5 uses attachment $type0/* hook.
667 * Since 1.5.0 uses attachments_bottom hook.
668 * Since 1.5.2 uses templates and does *not* return a value.
670 * @param object $message SquirrelMail message object
671 * @param array $exclude_id message parts that are not attachments.
672 * @param string $mailbox mailbox name
673 * @param integer $id message id
675 function formatAttachments($message, $exclude_id, $mailbox, $id) {
678 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
680 $oTemplate->assign('attachments', $attach);
681 $oTemplate->display('read_attachments.tpl');
684 function sqimap_base64_decode(&$string) {
686 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
687 // fly decoding (to reduce memory usage) you have to check if the
688 // data has incomplete pairs
690 // Remove the noise in order to check if the 4 bytes pairs are complete
691 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
694 $iMod = strlen($string) %
4;
696 $sStringRem = substr($string,-$iMod);
697 // Check if $sStringRem contains padding characters
698 if (substr($sStringRem,-1) != '=') {
699 $string = substr($string,0,-$iMod);
704 $string = base64_decode($string);
709 * Decodes encoded string (usually message body)
711 * This function decodes a string (usually the message body)
712 * depending on the encoding type. Currently quoted-printable
713 * and base64 encodings are supported.
715 * The decode_body hook was added to this function in 1.4.2/1.5.0.
716 * The $force_crlf parameter was added in 1.5.2.
718 * @param string $string The encoded string
719 * @param string $encoding used encoding
720 * @param string $force_crlf Whether or not to force CRLF or LF
721 * line endings (or to leave as is).
722 * If given as "LF", line endings will
723 * all be converted to LF; if "CRLF",
724 * line endings will all be converted
725 * to CRLF. If given as an empty value,
726 * the global $force_crlf_default will
727 * be consulted (it can be specified in
728 * config/config_local.php). Otherwise,
729 * any other value will cause the string
730 * to be left alone. Note that this will
731 * be overridden to "LF" if not using at
732 * least PHP version 4.3.0. (OPTIONAL;
733 * default is empty - consult global
736 * @return string The decoded string
741 function decodeBody($string, $encoding, $force_crlf='') {
743 global $force_crlf_default;
744 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
745 $force_crlf = strtoupper($force_crlf);
747 // must force line endings to LF due to broken
748 // quoted_printable_decode() in PHP versions
749 // before 4.3.0 (see below)
751 if (!check_php_version(4, 3, 0) ||
$force_crlf == 'LF')
752 $string = str_replace("\r\n", "\n", $string);
753 else if ($force_crlf == 'CRLF')
754 $string = str_replace("\n", "\r\n", $string);
756 $encoding = strtolower($encoding);
758 $encoding_handler = do_hook('decode_body', $encoding);
761 // plugins get first shot at decoding the string
763 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
764 $string = $encoding_handler('decode', $string);
766 } elseif ($encoding == 'quoted-printable' ||
767 $encoding == 'quoted_printable') {
769 // quoted_printable_decode() function is broken in older
770 // php versions. Text with \r\n decoding was fixed only
771 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
772 // and the above call to: str_replace("\r\n", "\n", $string);
774 $string = quoted_printable_decode($string);
776 } elseif ($encoding == 'base64') {
777 $string = base64_decode($string);
780 // All other encodings are returned raw.
787 * This function decodes strings that are encoded according to
788 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
789 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
791 * @param string $string header string that has to be made readable
792 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
793 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
794 * @param boolean $decide decide if string can be utfencoded. defaults to false
795 * @return string decoded header string
797 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
798 global $languages, $squirrelmail_language,$default_charset, $fix_broken_base64_encoded_messages;
799 if (is_array($string)) {
800 $string = implode("\n", $string);
803 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
804 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
805 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
806 // Do we need to return at this point?
813 // FIXME: spaces are allowed inside quoted-printable encoding, but the following line will bust up any such encoded strings
814 $aString = explode(' ',$string);
816 foreach ($aString as $chunk) {
817 if ($encoded && $chunk === '') {
819 } elseif ($chunk === '') {
824 /* if encoded words are not separated by a linear-space-white we still catch them */
827 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
828 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
829 if ($iLastMatch !== $j) {
839 $ret .= sm_encode_html_special_chars($res[1]);
843 $encoding = ucfirst($res[3]);
845 /* decide about valid decoding */
846 if ($decide && is_conversion_safe($res[2])) {
848 $can_be_encoded=true;
850 $can_be_encoded=false;
855 // fix broken base64-encoded strings (remove end = padding,
856 // change any = to + in middle of string, add padding back
858 if ($fix_broken_base64_encoded_messages) {
859 $encoded_string_minus_padding = strtr(rtrim($res[4], '='), '=', '+');
860 $res[4] = str_pad($encoded_string_minus_padding, strlen($res[4]), '=');
862 $replace = base64_decode($res[4]);
864 if ($can_be_encoded) {
865 /* convert string to different charset,
866 * if functions asks for it (usually in compose)
868 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
870 // convert string to html codes in order to display it
871 $ret .= charset_decode($res[2],$replace);
875 $replace = sm_encode_html_special_chars($replace);
881 $replace = str_replace('_', ' ', $res[4]);
882 $replace = preg_replace_callback('/=([0-9a-f]{2})/i',
883 create_function ('$matches', 'return chr(hexdec($matches[1]));'),
886 if ($can_be_encoded) {
887 /* convert string to different charset,
888 * if functions asks for it (usually in compose)
890 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
892 // convert string to html codes in order to display it
893 $replace = charset_decode($res[2], $replace);
897 $replace = sm_encode_html_special_chars($replace);
916 if (!$encoded && $htmlsafe) {
917 $ret .= sm_encode_html_special_chars($chunk);
923 /* remove the first added space */
926 $ret = substr($ret,5);
928 $ret = substr($ret,1);
938 * Function uses XTRA_CODE _encodeheader function, if such function exists.
940 * Function uses Q encoding by default and encodes a string according to RFC
941 * 1522 for use in headers if it contains 8-bit characters or anything that
942 * looks like it should be encoded.
944 * Function switches to B encoding and encodeHeaderBase64() function, if
945 * string is 8bit and multibyte character set supported by mbstring extension
946 * is used. It can cause E_USER_NOTICE errors, if interface is used with
947 * multibyte character set unsupported by mbstring extension.
949 * @param string $string header string, that has to be encoded
950 * @return string quoted-printable encoded string
951 * @todo make $mb_charsets system wide constant
953 function encodeHeader ($string) {
954 global $default_charset, $languages, $squirrelmail_language;
956 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
957 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
958 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
961 // Use B encoding for multibyte charsets
962 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
963 if (in_array($default_charset,$mb_charsets) &&
964 in_array($default_charset,sq_mb_list_encodings()) &&
965 sq_is8bit($string)) {
966 return encodeHeaderBase64($string,$default_charset);
967 } elseif (in_array($default_charset,$mb_charsets) &&
968 sq_is8bit($string) &&
969 ! in_array($default_charset,sq_mb_list_encodings())) {
970 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
971 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
974 // Encode only if the string contains 8-bit characters or =?
975 $j = strlen($string);
976 $max_l = 75 - strlen($default_charset) - 7;
979 $iEncStart = $enc_init = false;
980 $cur_l = $iOffset = 0;
981 for($i = 0; $i < $j; ++
$i) {
991 if ($iEncStart === false) {
995 if ($cur_l > ($max_l-2)) {
996 /* if there is an stringpart that doesn't need encoding, add it */
997 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
998 $aRet[] = "=?$default_charset?Q?$ret?=";
1004 $ret .= sprintf("=%02X",ord($string{$i}));
1009 if ($iEncStart !== false) {
1010 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1011 $aRet[] = "=?$default_charset?Q?$ret?=";
1019 if ($iEncStart !== false) {
1021 if ($cur_l > $max_l) {
1022 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1023 $aRet[] = "=?$default_charset?Q?$ret?=";
1034 $k = ord($string{$i});
1036 if ($iEncStart === false) {
1037 // do not start encoding in the middle of a string, also take the rest of the word.
1038 $sLeadString = substr($string,0,$i);
1039 $aLeadString = explode(' ',$sLeadString);
1040 $sToBeEncoded = array_pop($aLeadString);
1041 $iEncStart = $i - strlen($sToBeEncoded);
1042 $ret .= $sToBeEncoded;
1043 $cur_l +
= strlen($sToBeEncoded);
1046 /* first we add the encoded string that reached it's max size */
1047 if ($cur_l > ($max_l-2)) {
1048 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1049 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1056 $ret .= sprintf("=%02X", $k);
1058 if ($iEncStart !== false) {
1060 if ($cur_l > $max_l) {
1061 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1062 $aRet[] = "=?$default_charset?Q?$ret?=";
1068 $ret .= $string{$i};
1077 if ($iEncStart !== false) {
1078 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1079 $aRet[] = "=?$default_charset?Q?$ret?=";
1081 $aRet[] = substr($string,$iOffset);
1083 $string = implode('',$aRet);
1089 * Encodes string according to rfc2047 B encoding header formating rules
1091 * It is recommended way to encode headers with character sets that store
1092 * symbols in more than one byte.
1094 * Function requires mbstring support. If required mbstring functions are missing,
1095 * function returns false and sets E_USER_WARNING level error message.
1097 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1098 * that mbstring functions will generate E_WARNING errors, if unsupported
1099 * character set is used. mb_encode_mimeheader function provided by php
1100 * mbstring extension is not used in order to get better control of header
1103 * Used php code functions - function_exists(), trigger_error(), strlen()
1104 * (is used with charset names and base64 strings). Used php mbstring
1105 * functions - mb_strlen and mb_substr.
1107 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1108 * encoding), rfc 2822 (header folding)
1110 * @param string $string header string that must be encoded
1111 * @param string $charset character set. Must be supported by mbstring extension.
1112 * Use sq_mb_list_encodings() to detect supported charsets.
1113 * @return string string encoded according to rfc2047 B encoding formating rules
1115 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1116 * @todo Do we want to control max length of header?
1117 * @todo Do we want to control EOL (end-of-line) marker?
1118 * @todo Do we want to translate error message?
1120 function encodeHeaderBase64($string,$charset) {
1122 * Check mbstring function requirements.
1124 if (! function_exists('mb_strlen') ||
1125 ! function_exists('mb_substr')) {
1126 // set E_USER_WARNING
1127 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1132 // initial return array
1136 * header length = 75 symbols max (same as in encodeHeader)
1137 * remove $charset length
1138 * remove =? ? ?= (5 chars)
1139 * remove 2 more chars (\r\n ?)
1141 $iMaxLength = 75 - strlen($charset) - 7;
1143 // set first character position
1146 // loop through all characters. count characters and not bytes.
1147 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1148 // encode string from starting character to current character.
1149 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1151 // Check encoded string length
1152 if(strlen($encoded_string)>$iMaxLength) {
1153 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1154 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1156 // set new starting character
1157 $iStartCharNum = $iCharNum-1;
1159 // encode last char (in case it is last character in string)
1160 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1161 } // if string is shorter than max length - add next character
1164 // add last encoded string to array
1165 $aRet[] = $encoded_string;
1167 // set initial return string
1170 // loop through encoded strings
1171 foreach($aRet as $string) {
1172 // TODO: Do we want to control EOL (end-of-line) marker
1173 if ($sRet!='') $sRet.= " ";
1175 // add header tags and encoded string to return string
1176 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1182 /* This function trys to locate the entity_id of a specific mime element */
1183 function find_ent_id($id, $message) {
1184 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1185 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1186 $ret = find_ent_id($id, $message->entities
[$i]);
1188 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1189 // if (sq_check_save_extension($message->entities[$i])) {
1190 return $message->entities
[$i]->entity_id
;
1192 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1194 * This is part of a fix for Outlook Express 6.x generating
1195 * cid URLs without creating content-id headers
1198 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1199 return $message->entities
[$i]->entity_id
;
1207 function sq_check_save_extension($message) {
1208 $filename = $message->getFilename();
1209 $ext = substr($filename, strrpos($filename,'.')+
1);
1210 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1211 return in_array($ext, $save_extensions);
1216 ** HTMLFILTER ROUTINES
1220 * This function checks attribute values for entity-encoded values
1221 * and returns them translated into 8-bit strings so we can run
1224 * @param $attvalue A string to run entity check against.
1225 * @return Nothing, modifies a reference value.
1227 function sq_defang(&$attvalue){
1230 * Skip this if there aren't ampersands or backslashes.
1232 if (strpos($attvalue, '&') === false
1233 && strpos($attvalue, '\\') === false){
1237 // before deent, translate the dangerous unicode characters and ... to safe values
1238 // otherwise the regular expressions do not match.
1244 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1245 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1246 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1247 } while ($m == true);
1248 $attvalue = stripslashes($attvalue);
1252 * Kill any tabs, newlines, or carriage returns. Our friends the
1253 * makers of the browser with 95% market value decided that it'd
1254 * be funny to make "java[tab]script" be just as good as "javascript".
1256 * @param attvalue The attribute value before extraneous spaces removed.
1257 * @return attvalue Nothing, modifies a reference value.
1259 function sq_unspace(&$attvalue){
1261 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1262 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1263 Array('', '', '', '', ''), $attvalue);
1268 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1269 * IE as regular characters.
1271 * @param attvalue The attribute value before dangerous characters are translated.
1272 * @return attvalue Nothing, modifies a reference value.
1273 * @author Marc Groot Koerkamp.
1275 function sq_fixIE_idiocy(&$attvalue) {
1277 $attvalue = str_replace("\0", "", $attvalue);
1279 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1281 // IE has the evil habit of accepting every possible value for the attribute expression.
1282 // The table below contains characters which are parsed by IE if they are used in the "expression"
1284 $aDangerousCharsReplacementTable = array(
1285 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1286 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1287 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1288 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1289 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1290 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1291 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1292 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1293 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1294 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1295 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1296 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1297 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1298 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1299 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1300 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1301 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1302 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1303 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1304 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1305 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1306 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1307 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1308 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1309 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1310 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1311 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1312 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1313 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1314 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1315 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1316 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1317 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1318 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1319 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1320 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1321 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1322 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1323 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1324 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1325 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1326 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1327 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1328 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1329 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1330 "\xCA\x9F", /* L UNICODE IPA Extension */
1331 "\xCA\x80", /* R UNICODE IPA Extension */
1332 "\xC9\xB4"), /* N UNICODE IPA Extension */
1333 array('l', 'l', 'r','r','n','n',
1334 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1335 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1336 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1337 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1339 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1341 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1345 * This function returns the final tag out of the tag name, an array
1346 * of attributes, and the type of the tag. This function is called by
1347 * sq_sanitize internally.
1349 * @param $tagname the name of the tag.
1350 * @param $attary the array of attributes and their values
1351 * @param $tagtype The type of the tag (see in comments).
1352 * @return a string with the final tag representation.
1354 function sq_tagprint($tagname, $attary, $tagtype){
1355 $me = 'sq_tagprint';
1358 $fulltag = '</' . $tagname . '>';
1360 $fulltag = '<' . $tagname;
1361 if (is_array($attary) && sizeof($attary)){
1363 while (list($attname, $attvalue) = each($attary)){
1364 array_push($atts, "$attname=$attvalue");
1366 $fulltag .= ' ' . join(" ", $atts);
1377 * A small helper function to use with array_walk. Modifies a by-ref
1378 * value and makes it lowercase.
1380 * @param $val a value passed by-ref.
1381 * @return void since it modifies a by-ref value.
1383 function sq_casenormalize(&$val){
1384 $val = strtolower($val);
1388 * This function skips any whitespace from the current position within
1389 * a string and to the next non-whitespace value.
1391 * @param $body the string
1392 * @param $offset the offset within the string where we should start
1393 * looking for the next non-whitespace character.
1394 * @return the location within the $body where the next
1395 * non-whitespace char is located.
1397 function sq_skipspace($body, $offset){
1398 $me = 'sq_skipspace';
1399 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1400 if (sizeof($matches{1})){
1401 $count = strlen($matches{1});
1408 * This function looks for the next character within a string. It's
1409 * really just a glorified "strpos", except it catches if failures
1412 * @param $body The string to look for needle in.
1413 * @param $offset Start looking from this position.
1414 * @param $needle The character/string to look for.
1415 * @return location of the next occurance of the needle, or
1416 * strlen($body) if needle wasn't found.
1418 function sq_findnxstr($body, $offset, $needle){
1419 $me = 'sq_findnxstr';
1420 $pos = strpos($body, $needle, $offset);
1421 if ($pos === FALSE){
1422 $pos = strlen($body);
1428 * This function takes a PCRE-style regexp and tries to match it
1429 * within the string.
1431 * @param $body The string to look for needle in.
1432 * @param $offset Start looking from here.
1433 * @param $reg A PCRE-style regex to match.
1434 * @return Returns a false if no matches found, or an array
1435 * with the following members:
1436 * - integer with the location of the match within $body
1437 * - string with whatever content between offset and the match
1438 * - string with whatever it is we matched
1440 function sq_findnxreg($body, $offset, $reg){
1441 $me = 'sq_findnxreg';
1444 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1445 if (!isset($matches{0}) ||
!$matches{0}){
1448 $retarr{0} = $offset +
strlen($matches{1});
1449 $retarr{1} = $matches{1};
1450 $retarr{2} = $matches{2};
1456 * This function looks for the next tag.
1458 * @param $body String where to look for the next tag.
1459 * @param $offset Start looking from here.
1460 * @return false if no more tags exist in the body, or
1461 * an array with the following members:
1462 * - string with the name of the tag
1463 * - array with attributes and their values
1464 * - integer with tag type (1, 2, or 3)
1465 * - integer where the tag starts (starting "<")
1466 * - integer where the tag ends (ending ">")
1467 * first three members will be false, if the tag is invalid.
1469 function sq_getnxtag($body, $offset){
1470 $me = 'sq_getnxtag';
1471 if ($offset > strlen($body)){
1474 $lt = sq_findnxstr($body, $offset, "<");
1475 if ($lt == strlen($body)){
1480 * blah blah <tag attribute="value">
1483 $pos = sq_skipspace($body, $lt+
1);
1484 if ($pos >= strlen($body)){
1485 return Array(false, false, false, $lt, strlen($body));
1488 * There are 3 kinds of tags:
1489 * 1. Opening tag, e.g.:
1491 * 2. Closing tag, e.g.:
1493 * 3. XHTML-style content-less tag, e.g.:
1494 * <img src="blah" />
1497 switch (substr($body, $pos, 1)){
1504 * A comment or an SGML declaration.
1506 if (substr($body, $pos+
1, 2) == "--"){
1507 $gt = strpos($body, "-->", $pos);
1509 $gt = strlen($body);
1513 return Array(false, false, false, $lt, $gt);
1515 $gt = sq_findnxstr($body, $pos, ">");
1516 return Array(false, false, false, $lt, $gt);
1521 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1531 * Look for next [\W-_], which will indicate the end of the tag name.
1533 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1534 if ($regary == false){
1535 return Array(false, false, false, $lt, strlen($body));
1537 list($pos, $tagname, $match) = $regary;
1538 $tagname = strtolower($tagname);
1541 * $match can be either of these:
1542 * '>' indicating the end of the tag entirely.
1543 * '\s' indicating the end of the tag name.
1544 * '/' indicating that this is type-3 xhtml tag.
1546 * Whatever else we find there indicates an invalid tag.
1551 * This is an xhtml-style tag with a closing / at the
1552 * end, like so: <img src="blah" />. Check if it's followed
1553 * by the closing bracket. If not, then this tag is invalid
1555 if (substr($body, $pos, 2) == "/>"){
1559 $gt = sq_findnxstr($body, $pos, ">");
1560 $retary = Array(false, false, false, $lt, $gt);
1564 return Array($tagname, false, $tagtype, $lt, $pos);
1568 * Check if it's whitespace
1570 if (!preg_match('/\s/', $match)){
1572 * This is an invalid tag! Look for the next closing ">".
1574 $gt = sq_findnxstr($body, $lt, ">");
1575 return Array(false, false, false, $lt, $gt);
1581 * At this point we're here:
1582 * <tagname attribute='blah'>
1585 * At this point we loop in order to find all attributes.
1591 while ($pos <= strlen($body)){
1592 $pos = sq_skipspace($body, $pos);
1593 if ($pos == strlen($body)){
1597 return Array(false, false, false, $lt, $pos);
1600 * See if we arrived at a ">" or "/>", which means that we reached
1601 * the end of the tag.
1604 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1608 $pos +
= strlen($matches{1});
1609 if ($matches{2} == "/>"){
1613 return Array($tagname, $attary, $tagtype, $lt, $pos);
1617 * There are several types of attributes, with optional
1618 * [:space:] between members.
1620 * attrname[:space:]=[:space:]'CDATA'
1622 * attrname[:space:]=[:space:]"CDATA"
1624 * attr[:space:]=[:space:]CDATA
1628 * We leave types 1 and 2 the same, type 3 we check for
1629 * '"' and convert to """ if needed, then wrap in
1630 * double quotes. Type 4 we convert into:
1633 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1634 if ($regary == false){
1636 * Looks like body ended before the end of tag.
1638 return Array(false, false, false, $lt, strlen($body));
1640 list($pos, $attname, $match) = $regary;
1641 $attname = strtolower($attname);
1643 * We arrived at the end of attribute name. Several things possible
1645 * '>' means the end of the tag and this is attribute type 4
1646 * '/' if followed by '>' means the same thing as above
1647 * '\s' means a lot of things -- look what it's followed by.
1648 * anything else means the attribute is invalid.
1653 * This is an xhtml-style tag with a closing / at the
1654 * end, like so: <img src="blah" />. Check if it's followed
1655 * by the closing bracket. If not, then this tag is invalid
1657 if (substr($body, $pos, 2) == "/>"){
1661 $gt = sq_findnxstr($body, $pos, ">");
1662 $retary = Array(false, false, false, $lt, $gt);
1666 $attary{$attname} = '"yes"';
1667 return Array($tagname, $attary, $tagtype, $lt, $pos);
1671 * Skip whitespace and see what we arrive at.
1673 $pos = sq_skipspace($body, $pos);
1674 $char = substr($body, $pos, 1);
1676 * Two things are valid here:
1677 * '=' means this is attribute type 1 2 or 3.
1678 * \w means this was attribute type 4.
1679 * anything else we ignore and re-loop. End of tag and
1680 * invalid stuff will be caught by our checks at the beginning
1685 $pos = sq_skipspace($body, $pos);
1687 * Here are 3 possibilities:
1688 * "'" attribute type 1
1689 * '"' attribute type 2
1690 * everything else is the content of tag type 3
1692 $quot = substr($body, $pos, 1);
1694 $regary = sq_findnxreg($body, $pos+
1, "\'");
1695 if ($regary == false){
1696 return Array(false, false, false, $lt, strlen($body));
1698 list($pos, $attval, $match) = $regary;
1700 $attary{$attname} = "'" . $attval . "'";
1701 } else if ($quot == '"'){
1702 $regary = sq_findnxreg($body, $pos+
1, '\"');
1703 if ($regary == false){
1704 return Array(false, false, false, $lt, strlen($body));
1706 list($pos, $attval, $match) = $regary;
1708 $attary{$attname} = '"' . $attval . '"';
1711 * These are hateful. Look for \s, or >.
1713 $regary = sq_findnxreg($body, $pos, "[\s>]");
1714 if ($regary == false){
1715 return Array(false, false, false, $lt, strlen($body));
1717 list($pos, $attval, $match) = $regary;
1719 * If it's ">" it will be caught at the top.
1721 $attval = preg_replace("/\"/s", """, $attval);
1722 $attary{$attname} = '"' . $attval . '"';
1724 } else if (preg_match("|[\w/>]|", $char)) {
1726 * That was attribute type 4.
1728 $attary{$attname} = '"yes"';
1731 * An illegal character. Find next '>' and return.
1733 $gt = sq_findnxstr($body, $pos, ">");
1734 return Array(false, false, false, $lt, $gt);
1740 * The fact that we got here indicates that the tag end was never
1741 * found. Return invalid tag indication so it gets stripped.
1743 return Array(false, false, false, $lt, strlen($body));
1747 * Translates entities into literal values so they can be checked.
1749 * @param $attvalue the by-ref value to check.
1750 * @param $regex the regular expression to check against.
1751 * @param $hex whether the entites are hexadecimal.
1752 * @return True or False depending on whether there were matches.
1754 function sq_deent(&$attvalue, $regex, $hex=false){
1758 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1759 preg_match_all($regex, $attvalue, $matches);
1760 if (is_array($matches) && sizeof($matches[0]) > 0){
1762 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1763 $numval = $matches[1][$i];
1765 $numval = hexdec($numval);
1767 $repl{$matches[0][$i]} = chr($numval);
1769 $attvalue = strtr($attvalue, $repl);
1777 * This function runs various checks against the attributes.
1779 * @param $tagname String with the name of the tag.
1780 * @param $attary Array with all tag attributes.
1781 * @param $rm_attnames See description for sq_sanitize
1782 * @param $bad_attvals See description for sq_sanitize
1783 * @param $add_attr_to_tag See description for sq_sanitize
1784 * @param $message message object
1785 * @param $id message id
1786 * @return Array with modified attributes.
1788 function sq_fixatts($tagname,
1798 while (list($attname, $attvalue) = each($attary)){
1800 * See if this attribute should be removed.
1802 foreach ($rm_attnames as $matchtag=>$matchattrs){
1803 if (preg_match($matchtag, $tagname)){
1804 foreach ($matchattrs as $matchattr){
1805 if (preg_match($matchattr, $attname)){
1806 unset($attary{$attname});
1813 * Workaround for IE quirks
1815 sq_fixIE_idiocy($attvalue);
1818 * Remove any backslashes, entities, and extraneous whitespace.
1821 $oldattvalue = $attvalue;
1822 sq_defang($attvalue);
1823 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1824 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1825 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1826 $attvalue = "idiocy";
1827 $attary{$attname} = $attvalue;
1829 sq_unspace($attvalue);
1832 * Now let's run checks on the attvalues.
1833 * I don't expect anyone to comprehend this. If you do,
1834 * get in touch with me so I can drive to where you live and
1835 * shake your hand personally. :)
1837 foreach ($bad_attvals as $matchtag=>$matchattrs){
1838 if (preg_match($matchtag, $tagname)){
1839 foreach ($matchattrs as $matchattr=>$valary){
1840 if (preg_match($matchattr, $attname)){
1842 * There are two arrays in valary.
1844 * Second one is replacements
1846 list($valmatch, $valrepl) = $valary;
1848 preg_replace($valmatch, $valrepl, $attvalue);
1849 if ($newvalue != $attvalue){
1850 $attary{$attname} = $newvalue;
1851 $attvalue = $newvalue;
1857 if ($attname == 'style') {
1858 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1859 // 8bit and control characters in style attribute values can be used for XSS, remove them
1860 $attary{$attname} = '"disallowed character"';
1862 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1863 if (count($aMatch)) {
1864 foreach($aMatch[1] as $sMatch) {
1866 $urlvalue = $sMatch;
1867 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1868 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1873 * Use white list based filtering on attributes which can contain url's
1875 else if ($attname == 'href' ||
$attname == 'xlink:href' ||
$attname == 'src'
1876 ||
$attname == 'poster' ||
$attname == 'formaction'
1877 ||
$attname == 'background' ||
$attname == 'action') {
1878 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1879 $attary{$attname} = $attvalue;
1883 * See if we need to append any attributes to this tag.
1885 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1886 if (preg_match($matchtag, $tagname)){
1887 $attary = array_merge($attary, $addattary);
1894 * This function filters url's
1896 * @param $attvalue String with attribute value to filter
1897 * @param $message message object
1898 * @param $id message id
1899 * @param $mailbox mailbox
1900 * @param $sQuote quoting characters around url's
1902 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1903 $attvalue = trim($attvalue);
1904 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1905 // remove the double quotes
1906 $sQuote = $attvalue[0];
1907 $attvalue = trim(substr($attvalue,1,-1));
1910 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1911 // images off by default.
1912 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
1914 global $use_transparent_security_image;
1915 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
1916 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
1919 * Replace empty src tags with the blank image. src is only used
1920 * for frames, images, and image inputs. Doing a replace should
1921 * not affect them working as should be, however it will stop
1922 * IE from being kicked off when src for img tags are not set
1924 if ($attvalue == '') {
1925 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1927 // first, disallow 8 bit characters and control characters
1928 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1931 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1934 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1938 $aUrl = parse_url($attvalue);
1939 if (isset($aUrl['scheme'])) {
1940 switch(strtolower($aUrl['scheme'])) {
1945 if ($attname != 'href') {
1946 if ($view_unsafe_images == false) {
1947 $attvalue = $sQuote . $secremoveimg . $sQuote;
1949 if (isset($aUrl['path'])) {
1951 // No one has been able to show that image URIs
1952 // can be exploited, so for now, no restrictions
1953 // are made at all. If this proves to be a problem,
1954 // the commented-out code below can be of help.
1955 // (One consideration is that I see nothing in this
1956 // function that specifically says that we will
1957 // only ever arrive here when inspecting an image
1958 // tag, although that does seem to be the end
1959 // result - e.g., <script src="..."> where malicious
1960 // image URIs are in fact a problem are already
1961 // filtered out elsewhere.
1962 /* ---------------------------------
1963 // validate image extension.
1964 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1965 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1966 // If URI is to something other than
1967 // a regular image file, get the contents
1968 // and try to see if it is an image.
1969 // Don't use Fileinfo (finfo_file()) because
1970 // we'd need to make the admin configure the
1971 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1974 if (function_exists('mime_content_type')
1975 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1979 $file_contents = '';
1980 while (!feof($FILE)) {
1981 $file_contents .= fread($FILE, 8192);
1985 // store file locally
1987 global $attachment_dir, $username;
1988 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1989 $localfilename = GenerateRandomString(32, '', 7);
1990 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1991 while (file_exists($full_localfilename)) {
1992 $localfilename = GenerateRandomString(32, '', 7);
1993 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1995 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1996 fwrite($FILE, $file_contents);
1999 // get mime type and remove file
2001 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
2002 unlink("$hashed_attachment_dir/$localfilename");
2004 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
2005 if (substr(strtolower($mime_type), 0, 5) != 'image') {
2006 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2009 --------------------------------- */
2011 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
2015 $attvalue = $sQuote . $attvalue . $sQuote;
2020 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
2021 * One day MS might actually make it match something useful, for now, falling
2022 * back to using cid2http, so we can grab the blank.png.
2024 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2028 * Turn cid: urls into http-friendly ones.
2030 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2033 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2037 if (!isset($aUrl['path']) ||
$aUrl['path'] != $secremoveimg) {
2038 // parse_url did not lead to satisfying result
2039 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2047 * This function edits the style definition to make them friendly and
2048 * usable in SquirrelMail.
2050 * @param $message the message object
2051 * @param $id the message id
2052 * @param $content a string with whatever is between <style> and </style>
2053 * @param $mailbox the message mailbox
2054 * @return a string with edited content.
2056 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
2057 $me = 'sq_fixstyle';
2058 // workaround for </style> in between comments
2059 $iCurrentPos = $pos;
2064 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
2071 if ($sToken == '<') {
2081 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2086 $content .= $sToken;
2094 if ($sToken == '<') {
2096 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
2097 $i = strpos($body,'-->',$i+
3);
2098 if ($i === false) { // no end comment
2116 if ($bSucces == FALSE){
2117 return array(FALSE, strlen($body));
2123 * First look for general BODY style declaration, which would be
2125 * body {background: blah-blah}
2126 * and change it to .bodyclass so we can just assign it to a <div>
2128 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2130 global $use_transparent_security_image;
2131 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2132 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2135 * Fix url('blah') declarations.
2137 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2138 // "url(\\1$secremoveimg\\2)", $content);
2140 // first check for 8bit sequences and disallowed control characters
2141 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2142 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2143 return array($content, $newpos);
2146 // IE Sucks hard. We have a special function for it.
2147 sq_fixIE_idiocy($content);
2149 // remove @import line
2150 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2152 // translate ur\l and variations (IE parses that)
2153 // TODO check if the sq_fixIE_idiocy function already handles this.
2154 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2155 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2156 if (count($aMatch)) {
2157 $aValue = $aReplace = array();
2158 foreach($aMatch[1] as $sMatch) {
2160 $urlvalue = $sMatch;
2161 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2162 $aValue[] = $sMatch;
2163 $aReplace[] = $urlvalue;
2165 $content = str_replace($aValue,$aReplace,$content);
2169 * Remove any backslashes, entities, and extraneous whitespace.
2171 $contentTemp = $content;
2172 sq_defang($contentTemp);
2173 sq_unspace($contentTemp);
2176 * Fix stupid css declarations which lead to vulnerabilities
2179 * Also remove "position" attribute, as it can easily be set
2180 * to "fixed" or "absolute" with "left" and "top" attributes
2181 * of zero, taking over the whole content frame. It can also
2182 * be set to relative and move itself anywhere it wants to,
2183 * displaying content in areas it shouldn't be allowed to touch.
2185 $match = Array('/\/\*.*\*\//', // removes /* blah blah */
2189 '/include-source/i',
2193 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
2194 $contentNew = preg_replace($match, $replace, $contentTemp);
2195 if ($contentNew !== $contentTemp) {
2196 // insecure css declarations are used. From now on we don't care
2197 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2198 $content = $contentNew;
2200 return array($content, $newpos);
2205 * This function converts cid: url's into the ones that can be viewed in
2208 * @param $message the message object
2209 * @param $id the message id
2210 * @param $cidurl the cid: url.
2211 * @param $mailbox the message mailbox
2212 * @return a string with a http-friendly url
2214 function sq_cid2http($message, $id, $cidurl, $mailbox){
2216 * Get rid of quotes.
2218 $quotchar = substr($cidurl, 0, 1);
2219 if ($quotchar == '"' ||
$quotchar == "'"){
2220 $cidurl = str_replace($quotchar, "", $cidurl);
2224 $cidurl = substr(trim($cidurl), 4);
2226 $match_str = '/\{.*?\}\//';
2228 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2230 $linkurl = find_ent_id($cidurl, $message);
2231 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2232 unsafe link image */
2236 * This is part of a fix for Outlook Express 6.x generating
2237 * cid URLs without creating content-id headers. These images are
2238 * not part of the multipart/related html mail. The html contains
2239 * <img src="cid:{some_id}/image_filename.ext"> references to
2240 * attached images with as goal to render them inline although
2241 * the attachment disposition property is not inline.
2244 if (empty($linkurl)) {
2245 if (preg_match('/{.*}\//', $cidurl)) {
2246 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2247 if (!empty($cidurl)) {
2248 $linkurl = find_ent_id($cidurl, $message);
2253 if (!empty($linkurl)) {
2254 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&' .
2255 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2256 '&ent_id=' . $linkurl . $quotchar;
2259 * If we couldn't generate a proper img url, drop in a blank image
2260 * instead of sending back empty, otherwise it causes unusual behaviour
2262 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2269 * This function changes the <body> tag into a <div> tag since we
2270 * can't really have a body-within-body.
2272 * @param $attary an array of attributes and values of <body>
2273 * @param $mailbox mailbox we're currently reading (for cid2http)
2274 * @param $message current message (for cid2http)
2275 * @param $id current message id (for cid2http)
2276 * @return a modified array of attributes to be set for <div>
2278 function sq_body2div($attary, $mailbox, $message, $id){
2279 $me = 'sq_body2div';
2280 $divattary = Array('class' => "'bodyclass'");
2282 $has_bgc_stl = $has_txt_stl = false;
2284 if (is_array($attary) && sizeof($attary) > 0){
2285 foreach ($attary as $attname=>$attvalue){
2286 $quotchar = substr($attvalue, 0, 1);
2287 $attvalue = str_replace($quotchar, "", $attvalue);
2290 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2291 $styledef .= "background-image: url('$attvalue'); ";
2294 $has_bgc_stl = true;
2295 $styledef .= "background-color: $attvalue; ";
2298 $has_txt_stl = true;
2299 $styledef .= "color: $attvalue; ";
2303 // Outlook defines a white bgcolor and no text color. This can lead to
2304 // white text on a white bg with certain themes.
2305 if ($has_bgc_stl && !$has_txt_stl) {
2306 $styledef .= "color: $text; ";
2308 if (strlen($styledef) > 0){
2309 $divattary{"style"} = "\"$styledef\"";
2316 * This is the main function and the one you should actually be calling.
2317 * There are several variables you should be aware of an which need
2318 * special description.
2320 * Since the description is quite lengthy, see it here:
2321 * http://linux.duke.edu/projects/mini/htmlfilter/
2323 * @param $body the string with HTML you wish to filter
2324 * @param $tag_list see description above
2325 * @param $rm_tags_with_content see description above
2326 * @param $self_closing_tags see description above
2327 * @param $force_tag_closing see description above
2328 * @param $rm_attnames see description above
2329 * @param $bad_attvals see description above
2330 * @param $add_attr_to_tag see description above
2331 * @param $message message object
2332 * @param $id message id
2333 * @return sanitized html safe to show on your pages.
2335 function sq_sanitize($body,
2337 $rm_tags_with_content,
2347 $me = 'sq_sanitize';
2348 $rm_tags = array_shift($tag_list);
2350 * Normalize rm_tags and rm_tags_with_content.
2352 @array_walk
($tag_list, 'sq_casenormalize');
2353 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2354 @array_walk
($self_closing_tags, 'sq_casenormalize');
2356 * See if tag_list is of tags to remove or tags to allow.
2357 * false means remove these tags
2358 * true means allow these tags
2361 $open_tags = Array();
2362 $trusted = "\n<!-- begin sanitized html -->\n";
2363 $skip_content = false;
2365 * Take care of netscape's stupid javascript entities like
2368 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2370 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2371 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2372 $free_content = substr($body, $curpos, $lt-$curpos);
2374 * Take care of <style>
2376 if ($tagname == "style" && $tagtype == 1){
2377 list($free_content, $curpos) =
2378 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2379 if ($free_content != FALSE){
2380 if ( !empty($attary) ) {
2381 $attary = sq_fixatts($tagname,
2391 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2392 $trusted .= $free_content;
2393 $trusted .= sq_tagprint($tagname, false, 2);
2397 if ($skip_content == false){
2398 $trusted .= $free_content;
2400 if ($tagname != FALSE){
2402 if ($skip_content == $tagname){
2404 * Got to the end of tag we needed to remove.
2407 $skip_content = false;
2409 if ($skip_content == false){
2410 if ($tagname == "body"){
2413 if (isset($open_tags{$tagname}) &&
2414 $open_tags{$tagname} > 0){
2415 $open_tags{$tagname}--;
2423 * $rm_tags_with_content
2425 if ($skip_content == false){
2427 * See if this is a self-closing type and change
2428 * tagtype appropriately.
2431 && in_array($tagname, $self_closing_tags)){
2435 * See if we should skip this tag and any content
2438 if ($tagtype == 1 &&
2439 in_array($tagname, $rm_tags_with_content)){
2440 $skip_content = $tagname;
2442 if (($rm_tags == false
2443 && in_array($tagname, $tag_list)) ||
2444 ($rm_tags == true &&
2445 !in_array($tagname, $tag_list))){
2449 * Convert body into div.
2451 if ($tagname == "body"){
2453 $attary = sq_body2div($attary, $mailbox,
2457 if (isset($open_tags{$tagname})){
2458 $open_tags{$tagname}++
;
2460 $open_tags{$tagname}=1;
2464 * This is where we run other checks.
2466 if (is_array($attary) && sizeof($attary) > 0){
2467 $attary = sq_fixatts($tagname,
2481 if ($tagname != false && $skip_content == false){
2482 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2487 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2488 if ($force_tag_closing == true){
2489 foreach ($open_tags as $tagname=>$opentimes){
2490 while ($opentimes > 0){
2491 $trusted .= '</' . $tagname . '>';
2497 $trusted .= "<!-- end sanitized html -->\n";
2502 * This is a wrapper function to call html sanitizing routines.
2504 * @param $body the body of the message
2505 * @param $id the id of the message
2509 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2510 * into internal SM compose links
2511 * (optional; default = TRUE)
2512 * @return a string with html safe to display in the browser.
2514 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2516 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2518 global $attachment_common_show_images, $view_unsafe_images,
2519 $has_unsafe_images, $allow_svg_display;
2521 * Don't display attached images in HTML mode.
2525 $attachment_common_show_images = false;
2539 $rm_tags_with_content = Array(
2549 if (!$allow_svg_display)
2550 $rm_tags_with_content[] = 'svg';
2552 $self_closing_tags = Array(
2560 $force_tag_closing = true;
2562 $rm_attnames = Array(
2573 global $use_transparent_security_image;
2574 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2575 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2577 $bad_attvals = Array(
2580 "/^src|background/i" =>
2583 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2584 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2585 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2588 "\\1$secremoveimg\\2",
2589 "\\1$secremoveimg\\2",
2590 "\\1$secremoveimg\\2",
2593 "/^href|action/i" =>
2596 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2597 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2598 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2613 "/include-source/i",
2615 // position:relative can also be exploited
2616 // to put content outside of email body area
2617 // and position:fixed is similarly exploitable
2618 // as position:absolute, so we'll remove it
2621 // Does this screw up legitimate HTML messages?
2622 // If so, the only fix I see is to allow position
2623 // attributes (any values? I think we still have
2624 // to block static and fixed) only if $use_iframe
2625 // is enabled (1.5.0+)
2627 // was: "/position\s*:\s*absolute/i",
2631 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2632 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2633 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2634 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2635 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si",
2654 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2655 // images off by default.
2656 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
2658 if (!$view_unsafe_images){
2660 * Remove any references to http/https if view_unsafe_images set
2663 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2664 '/^([\'\"])\s*https*:.*([\'\"])/si');
2665 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2666 "\\1$secremoveimg\\1");
2667 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2668 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2669 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2670 "url(\\1$secremoveimg\\1)");
2673 $add_attr_to_tag = Array(
2675 Array('target'=>'"_blank"',
2676 'title'=>'"'._("This external link will open in a new window").'"'
2679 $trusted = sq_sanitize($body,
2681 $rm_tags_with_content,
2691 if (strpos($trusted,$secremoveimg)){
2692 $has_unsafe_images = true;
2695 // we want to parse mailto's in HTML output, change to SM compose links
2696 // this is a modified version of code from url_parser.php... but Marc is
2697 // right: we need a better filtering implementation; adding this randomly
2698 // here is not a great solution
2700 if ($take_mailto_links) {
2701 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2702 global $MailTo_PReg_Match;
2703 // some mailers (Microsoft, surprise surprise) produce mailto strings without being
2704 // inside an anchor (link) tag, so we have to make sure the regex looks for the
2705 // quote before mailto, and we'll also try to convert the non-links back into links
2706 $MailTo_PReg_Match = '/([\'"])?mailto:' . substr($MailTo_PReg_Match, 1) ;
2707 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2708 foreach ($regs[0] as $i => $mailto_before) {
2709 $mailto_params = $regs[11][$i];
2711 // get rid of any leading quote we may have captured but don't care about
2713 $mailto_before = ltrim($mailto_before, '"\'');
2715 // get rid of any tailing quote since we have to add send_to to the end
2717 $mailto_before = rtrim($mailto_before, '"\'');
2718 $mailto_params = rtrim($mailto_params, '"\'');
2720 if ($regs[2][$i]) { //if there is an email addr before '?', we need to merge it with the params
2721 $to = 'to=' . $regs[2][$i];
2722 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2723 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2725 if ($mailto_params) //already some params, append to them
2726 $mailto_params .= '&' . $to;
2728 $mailto_params .= '?' . $to;
2732 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2734 // we'll already have target=_blank, no need to allow comp_in_new
2735 // here (which would be a lot more work anyway)
2737 global $compose_new_win;
2738 $temp_comp_in_new = $compose_new_win;
2739 $compose_new_win = 0;
2740 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2741 $compose_new_win = $temp_comp_in_new;
2743 // remove <a href=" and anything after the next quote (we only
2744 // need the uri, not the link HTML) in compose uri
2746 // but only do this if the original mailto was in a real anchor tag
2748 if (!empty($regs[1][$i])) {
2749 $comp_uri = substr($comp_uri, 9);
2750 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2752 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2761 * function SendDownloadHeaders - send file to the browser
2763 * Original Source: SM core src/download.php
2764 * moved here to make it available to other code, and separate
2765 * front end from back end functionality.
2767 * @param string $type0 first half of mime type
2768 * @param string $type1 second half of mime type
2769 * @param string $filename filename to tell the browser for downloaded file
2770 * @param boolean $force whether to force the download dialog to pop
2771 * @param optional integer $filesize send the Content-Header and length to the browser
2774 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2775 global $languages, $squirrelmail_language;
2776 $isIE = $isIE6plus = false;
2778 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2780 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2781 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2785 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2786 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2790 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2791 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2793 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2795 $filename = preg_replace('/[\\\\\/:*?"<>|;]/', '_', str_replace(' ', ' ', $filename));
2798 // A Pox on Microsoft and it's Internet Explorer!
2800 // IE has lots of bugs with file downloads.
2801 // It also has problems with SSL. Both of these cause problems
2802 // for us in this function.
2804 // See this article on Cache Control headers and SSL
2805 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2807 // The best thing you can do for IE is to upgrade to the latest
2809 //set all the Cache Control Headers for IE
2811 $filename=rawurlencode($filename);
2812 header ("Pragma: public");
2813 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2814 // does nothing - see: https://blogs.msdn.microsoft.com/ieinternals/2009/07/20/internet-explorers-cache-control-extensions/
2815 // header ("Cache-Control: post-check=0, pre-check=0", false);
2816 header ("Cache-Control: private");
2818 //set the inline header for IE, we'll add the attachment header later if we need it
2819 header ("Content-Disposition: inline; filename=$filename");
2823 // Try to show in browser window
2824 header ("Content-Disposition: inline; filename=\"$filename\"");
2825 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2827 // Try to pop up the "save as" box
2829 // IE makes this hard. It pops up 2 save boxes, or none.
2830 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2831 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2832 // But, according to Microsoft, it is "RFC compliant but doesn't
2833 // take into account some deviations that allowed within the
2834 // specification." Doesn't that mean RFC non-compliant?
2835 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2837 // all browsers need the application/octet-stream header for this
2838 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2840 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2841 // Do not have quotes around filename, but that applied to
2842 // "attachment"... does it apply to inline too?
2843 header ("Content-Disposition: attachment; filename=\"$filename\"");
2845 if ($isIE && !$isIE6plus) {
2846 // This combination seems to work mostly. IE 5.5 SP 1 has
2847 // known issues (see the Microsoft Knowledge Base)
2849 // This works for most types, but doesn't work with Word files
2850 header ("Content-Type: application/download; name=\"$filename\"");
2851 header ("Content-Type: application/force-download; name=\"$filename\"");
2852 // These are spares, just in case. :-)
2853 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2854 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2855 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2857 // This is to prevent IE for MIME sniffing and auto open a file in IE
2858 header ("Content-Type: application/force-download; name=\"$filename\"");
2860 // another application/octet-stream forces download for Netscape
2861 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2865 //send the content-length header if the calling function provides it
2866 if ($filesize > 0) {
2867 header("Content-Length: $filesize");
2870 } // end fn SendDownloadHeaders