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', $force_crlf='') {
186 /* Don't kill the connection if the browser is over a dialup
187 * and it would take over 30 seconds to download it.
188 * Don't call set_time_limit in safe mode.
191 if (!ini_get('safe_mode')) {
194 /* in case of base64 encoded attachments, do not buffer them.
195 Instead, echo the decoded attachment directly to screen */
196 if (strtolower($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, $force_crlf));
208 echo decodeBody($body, $encoding, $force_crlf);
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 $mailbox (since 1.3.0) imap folder name
345 * @return string html formated message text
347 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
348 /* This if statement checks for the entity to show as the
349 * primary message. To add more of them, just put them in the
350 * order that is their priority.
352 global $startMessage, $languages, $squirrelmail_language,
353 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
354 $use_iframe, $iframe_height, $download_and_unsafe_link,
355 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
358 // workaround for not updated config.php
359 if (! isset($use_iframe)) $use_iframe = false;
361 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
362 // images off by default.
363 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
366 $urlmailbox = urlencode($mailbox);
367 $body_message = getEntity($message, $ent_num);
368 if (($body_message->header
->type0
== 'text') ||
369 ($body_message->header
->type0
== 'rfc822')) {
370 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
371 $body = decodeBody($body, $body_message->header
->encoding
);
373 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
374 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
375 if (mb_detect_encoding($body) != 'ASCII') {
376 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
380 /* As of 1.5.2, $body is passed (and modified) by reference */
381 do_hook('message_body', $body);
383 /* If there are other types that shouldn't be formatted, add
387 if ($body_message->header
->type1
== 'html') {
388 if ($show_html_default <> 1) {
389 $entity_conv = array(' ' => ' ',
398 $body = strtr($body, $entity_conv);
399 $body = strip_tags($body);
401 translateText($body, $wrap_at,
402 $body_message->header
->getParameter('charset'));
403 } elseif ($use_iframe) {
405 * If we don't add html message between iframe tags,
406 * we must detect unsafe images and modify $has_unsafe_images.
408 $html_body = magicHTML($body, $id, $message, $mailbox);
409 // Convert character set in order to display html mails in different character set
410 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
412 // creating iframe url
413 $iframeurl=sqm_baseuri().'src/view_html.php?'
414 . 'mailbox=' . $urlmailbox
415 . '&passed_id=' . $id
416 . '&ent_id=' . $ent_num
417 . '&view_unsafe_images=' . (int) $view_unsafe_images;
420 $oTemplate->assign('iframe_url', $iframeurl);
421 $oTemplate->assign('iframe_height', $iframe_height);
422 $oTemplate->assign('html_body', $html_body);
424 $body = $oTemplate->fetch('read_html_iframe.tpl');
426 // old way of html rendering
428 * convert character set. charset_decode does not remove html special chars
429 * applied by magicHTML functions and does not sanitize them second time if
430 * fourth argument is true.
432 $charset = $body_message->header
->getParameter('charset');
433 if (!empty($charset)) {
434 $body = charset_decode($charset,$body,false,true);
436 $body = magicHTML($body, $id, $message, $mailbox);
439 translateText($body, $wrap_at,
440 $body_message->header
->getParameter('charset'));
444 * Previously the links for downloading and unsafe images were printed
445 * under the mail. By putting the links in a global variable we can
446 * print it in the toolbar where it belongs. Since the original code was
447 * in this place it's left here. It might be possible to move it to some
448 * other place if that makes sense. The possibility to do so has not
449 * been evaluated yet.
452 // Initialize the global variable to an empty string.
453 // 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.
454 $download_and_unsafe_link = '';
456 // Prepare and build a link for downloading the mail.
457 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
458 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
459 '&startMessage=' . $startMessage . '&show_more=0';
460 if (isset($passed_ent_id)) {
461 $link .= '&passed_ent_id='.$passed_ent_id;
463 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
465 // Always add the link for downloading the mail as a file to the global
467 $download_and_unsafe_link .= "$nbsp|$nbsp"
468 . create_hyperlink($download_href, _("Download this as a file"));
470 // Find out the right text to use in the link depending on the
471 // circumstances. If the unsafe images are displayed the link should
472 // hide them, if they aren't displayed the link should only appear if
473 // the mail really contains unsafe images.
474 if ($view_unsafe_images) {
475 $text = _("Hide Unsafe Images");
477 if (isset($has_unsafe_images) && $has_unsafe_images) {
478 $link .= '&view_unsafe_images=1';
479 $text = _("View Unsafe Images");
485 // Only create a link for unsafe images if there's need for one. If so:
486 // add it to the global variable.
488 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
489 $unsafe_image_toggle_text = $text;
490 $download_and_unsafe_link .= "$nbsp|$nbsp"
491 . create_hyperlink($unsafe_image_toggle_href, $text);
498 * Generate attachments array for passing to templates.
501 * @param object $message SquirrelMail message object
502 * @param array $exclude_id message parts that are not attachments.
503 * @param string $mailbox mailbox name
504 * @param integer $id message id
506 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
507 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
509 $att_ar = $message->getAttachments($exclude_id);
510 $urlMailbox = urlencode($mailbox);
512 $attachments = array();
513 foreach ($att_ar as $att) {
514 $ent = $att->entity_id
;
515 $header = $att->header
;
516 $type0 = strtolower($header->type0
);
517 $type1 = strtolower($header->type1
);
520 $links['download link']['text'] = _("Download");
521 $links['download link']['href'] = $base_uri .
522 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
524 if ($type0 =='message' && $type1 == 'rfc822') {
525 $default_page = $base_uri . 'src/read_body.php';
526 $rfc822_header = $att->rfc822_header
;
527 $filename = $rfc822_header->subject
;
528 if (trim( $filename ) == '') {
529 $filename = 'untitled-[' . $ent . ']' ;
531 $from_o = $rfc822_header->from
;
532 if (is_object($from_o)) {
533 $from_name = decodeHeader($from_o->getAddress(false));
534 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
535 // something weird happens when a digest message is opened and you return to the digest
536 // now the from object is part of an array. Probably the parseHeader call overwrites the info
537 // retrieved from the bodystructure in a different way. We need to fix this later.
538 // possible starting point, do not fetch header we already have and inspect how
539 // the rfc822_header object behaves.
540 $from_name = decodeHeader($from_o[0]->getAddress(false));
542 $from_name = _("Unknown sender");
544 $description = _("From").': '.$from_name;
546 $default_page = $base_uri . 'src/download.php';
547 $filename = $att->getFilename();
548 if ($header->description
) {
549 $description = decodeHeader($header->description
);
555 $display_filename = $filename;
556 if (isset($passed_ent_id)) {
557 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
559 $passed_ent_id_link = '';
561 $defaultlink = $default_page . "?startMessage=$startMessage"
562 . "&passed_id=$id&mailbox=$urlMailbox"
563 . '&ent_id='.$ent.$passed_ent_id_link;
564 if ($where && $what) {
565 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
567 // IE does make use of mime content sniffing. Forcing a download
568 // prohibit execution of XSS inside an application/octet-stream attachment
569 if ($type0 == 'application' && $type1 == 'octet-stream') {
570 $defaultlink .= '&absolute_dl=true';
573 /* This executes the attachment hook with a specific MIME-type.
574 * If that doesn't have results, it tries if there's a rule
575 * for a more generic type. Finally, a hook for ALL attachment
576 * types is run as well.
578 // First remember the default link.
579 $defaultlink_orig = $defaultlink;
581 /* The API for this hook has changed as of 1.5.2 so that all plugin
582 arguments are passed in an array instead of each their own plugin
583 argument, and arguments are passed by reference, so instead of
584 returning any changes, changes should simply be made to the original
585 arguments themselves. */
586 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
587 &$defaultlink, &$display_filename, &$where, &$what);
588 do_hook("attachment $type0/$type1", $temp);
589 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
590 /* The API for this hook has changed as of 1.5.2 so that all plugin
591 arguments are passed in an array instead of each their own plugin
592 argument, and arguments are passed by reference, so instead of
593 returning any changes, changes should simply be made to the original
594 arguments themselves. */
595 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
596 &$defaultlink, &$display_filename, &$where, &$what);
597 do_hook("attachment $type0/*", $temp);
599 /* The API for this hook has changed as of 1.5.2 so that all plugin
600 arguments are passed in an array instead of each their own plugin
601 argument, and arguments are passed by reference, so instead of
602 returning any changes, changes should simply be made to the original
603 arguments themselves. */
604 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
605 &$defaultlink, &$display_filename, &$where, &$what);
606 // Do not let a generic plugin change the default link if a more
607 // specialized one already did it...
608 if ($defaultlink != $defaultlink_orig) {
612 do_hook("attachment */*", $temp);
614 $this_attachment = array();
615 $this_attachment['Name'] = decodeHeader($display_filename);
616 $this_attachment['Description'] = $description;
617 $this_attachment['DefaultHREF'] = $defaultlink;
618 $this_attachment['DownloadHREF'] = $links['download link']['href'];
619 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
620 $this_attachment['Size'] = $header->size
;
621 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
622 $this_attachment['OtherLinks'] = array();
623 foreach ($links as $val) {
624 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
626 if (empty($val['text']) && empty($val['extra']))
630 $temp['HREF'] = $val['href'];
631 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
632 $this_attachment['OtherLinks'][] = $temp;
634 $attachments[] = $this_attachment;
643 * Displays attachment links and information
645 * Since 1.3.0 function is not included in formatBody() call.
647 * Since 1.0.2 uses attachment $type0/$type1 hook.
648 * Since 1.2.5 uses attachment $type0/* hook.
649 * Since 1.5.0 uses attachments_bottom hook.
650 * Since 1.5.2 uses templates and does *not* return a value.
652 * @param object $message SquirrelMail message object
653 * @param array $exclude_id message parts that are not attachments.
654 * @param string $mailbox mailbox name
655 * @param integer $id message id
657 function formatAttachments($message, $exclude_id, $mailbox, $id) {
660 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
662 $oTemplate->assign('attachments', $attach);
663 $oTemplate->display('read_attachments.tpl');
666 function sqimap_base64_decode(&$string) {
668 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
669 // fly decoding (to reduce memory usage) you have to check if the
670 // data has incomplete pairs
672 // Remove the noise in order to check if the 4 bytes pairs are complete
673 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
676 $iMod = strlen($string) %
4;
678 $sStringRem = substr($string,-$iMod);
679 // Check if $sStringRem contains padding characters
680 if (substr($sStringRem,-1) != '=') {
681 $string = substr($string,0,-$iMod);
686 $string = base64_decode($string);
691 * Decodes encoded string (usually message body)
693 * This function decodes a string (usually the message body)
694 * depending on the encoding type. Currently quoted-printable
695 * and base64 encodings are supported.
697 * The decode_body hook was added to this function in 1.4.2/1.5.0.
698 * The $force_crlf parameter was added in 1.5.2.
700 * @param string $string The encoded string
701 * @param string $encoding used encoding
702 * @param string $force_crlf Whether or not to force CRLF or LF
703 * line endings (or to leave as is).
704 * If given as "LF", line endings will
705 * all be converted to LF; if "CRLF",
706 * line endings will all be converted
707 * to CRLF. If given as an empty value,
708 * the global $default_force_crlf will
709 * be consulted (it can be specified in
710 * config/config_local.php). Otherwise,
711 * any other value will cause the string
712 * to be left alone. Note that this will
713 * be overridden to "LF" if not using at
714 * least PHP version 4.3.0. (OPTIONAL;
715 * default is empty - consult global
718 * @return string The decoded string
723 function decodeBody($string, $encoding, $force_crlf='') {
725 global $force_crlf_default;
726 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
727 $force_crlf = strtoupper($force_crlf);
729 // must force line endings to LF due to broken
730 // quoted_printable_decode() in PHP versions
731 // before 4.3.0 (see below)
733 if (!check_php_version(4, 3, 0) ||
$force_crlf == 'LF')
734 $string = str_replace("\r\n", "\n", $string);
735 else if ($force_crlf == 'CRLF')
736 $string = str_replace("\n", "\r\n", $string);
738 $encoding = strtolower($encoding);
740 $encoding_handler = do_hook('decode_body', $encoding);
743 // plugins get first shot at decoding the string
745 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
746 $string = $encoding_handler('decode', $string);
748 } elseif ($encoding == 'quoted-printable' ||
749 $encoding == 'quoted_printable') {
751 // quoted_printable_decode() function is broken in older
752 // php versions. Text with \r\n decoding was fixed only
753 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
754 // and the above call to: str_replace("\r\n", "\n", $string);
756 $string = quoted_printable_decode($string);
758 } elseif ($encoding == 'base64') {
759 $string = base64_decode($string);
762 // All other encodings are returned raw.
769 * This function decodes strings that are encoded according to
770 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
771 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
773 * @param string $string header string that has to be made readable
774 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
775 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
776 * @param boolean $decide decide if string can be utfencoded. defaults to false
777 * @return string decoded header string
779 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
780 global $languages, $squirrelmail_language,$default_charset;
781 if (is_array($string)) {
782 $string = implode("\n", $string);
785 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
786 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
787 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
788 // Do we need to return at this point?
795 $aString = explode(' ',$string);
797 foreach ($aString as $chunk) {
798 if ($encoded && $chunk === '') {
800 } elseif ($chunk === '') {
805 /* if encoded words are not separated by a linear-space-white we still catch them */
808 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
809 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
810 if ($iLastMatch !== $j) {
820 $ret .= htmlspecialchars($res[1]);
824 $encoding = ucfirst($res[3]);
826 /* decide about valid decoding */
827 if ($decide && is_conversion_safe($res[2])) {
829 $can_be_encoded=true;
831 $can_be_encoded=false;
836 $replace = base64_decode($res[4]);
838 if ($can_be_encoded) {
839 /* convert string to different charset,
840 * if functions asks for it (usually in compose)
842 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
844 // convert string to html codes in order to display it
845 $ret .= charset_decode($res[2],$replace);
849 $replace = htmlspecialchars($replace);
855 $replace = str_replace('_', ' ', $res[4]);
856 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
859 if ($can_be_encoded) {
860 /* convert string to different charset,
861 * if functions asks for it (usually in compose)
863 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
865 // convert string to html codes in order to display it
866 $replace = charset_decode($res[2], $replace);
870 $replace = htmlspecialchars($replace);
889 if (!$encoded && $htmlsafe) {
890 $ret .= htmlspecialchars($chunk);
896 /* remove the first added space */
899 $ret = substr($ret,5);
901 $ret = substr($ret,1);
911 * Function uses XTRA_CODE _encodeheader function, if such function exists.
913 * Function uses Q encoding by default and encodes a string according to RFC
914 * 1522 for use in headers if it contains 8-bit characters or anything that
915 * looks like it should be encoded.
917 * Function switches to B encoding and encodeHeaderBase64() function, if
918 * string is 8bit and multibyte character set supported by mbstring extension
919 * is used. It can cause E_USER_NOTICE errors, if interface is used with
920 * multibyte character set unsupported by mbstring extension.
922 * @param string $string header string, that has to be encoded
923 * @return string quoted-printable encoded string
924 * @todo make $mb_charsets system wide constant
926 function encodeHeader ($string) {
927 global $default_charset, $languages, $squirrelmail_language;
929 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
930 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
931 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
934 // Use B encoding for multibyte charsets
935 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
936 if (in_array($default_charset,$mb_charsets) &&
937 in_array($default_charset,sq_mb_list_encodings()) &&
938 sq_is8bit($string)) {
939 return encodeHeaderBase64($string,$default_charset);
940 } elseif (in_array($default_charset,$mb_charsets) &&
941 sq_is8bit($string) &&
942 ! in_array($default_charset,sq_mb_list_encodings())) {
943 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
944 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
947 // Encode only if the string contains 8-bit characters or =?
948 $j = strlen($string);
949 $max_l = 75 - strlen($default_charset) - 7;
952 $iEncStart = $enc_init = false;
953 $cur_l = $iOffset = 0;
954 for($i = 0; $i < $j; ++
$i) {
963 if ($iEncStart === false) {
967 if ($cur_l > ($max_l-2)) {
968 /* if there is an stringpart that doesn't need encoding, add it */
969 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
970 $aRet[] = "=?$default_charset?Q?$ret?=";
976 $ret .= sprintf("=%02X",ord($string{$i}));
981 if ($iEncStart !== false) {
982 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
983 $aRet[] = "=?$default_charset?Q?$ret?=";
991 if ($iEncStart !== false) {
993 if ($cur_l > $max_l) {
994 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
995 $aRet[] = "=?$default_charset?Q?$ret?=";
1006 $k = ord($string{$i});
1008 if ($iEncStart === false) {
1009 // do not start encoding in the middle of a string, also take the rest of the word.
1010 $sLeadString = substr($string,0,$i);
1011 $aLeadString = explode(' ',$sLeadString);
1012 $sToBeEncoded = array_pop($aLeadString);
1013 $iEncStart = $i - strlen($sToBeEncoded);
1014 $ret .= $sToBeEncoded;
1015 $cur_l +
= strlen($sToBeEncoded);
1018 /* first we add the encoded string that reached it's max size */
1019 if ($cur_l > ($max_l-2)) {
1020 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1021 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1028 $ret .= sprintf("=%02X", $k);
1030 if ($iEncStart !== false) {
1032 if ($cur_l > $max_l) {
1033 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1034 $aRet[] = "=?$default_charset?Q?$ret?=";
1040 $ret .= $string{$i};
1049 if ($iEncStart !== false) {
1050 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1051 $aRet[] = "=?$default_charset?Q?$ret?=";
1053 $aRet[] = substr($string,$iOffset);
1055 $string = implode('',$aRet);
1061 * Encodes string according to rfc2047 B encoding header formating rules
1063 * It is recommended way to encode headers with character sets that store
1064 * symbols in more than one byte.
1066 * Function requires mbstring support. If required mbstring functions are missing,
1067 * function returns false and sets E_USER_WARNING level error message.
1069 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1070 * that mbstring functions will generate E_WARNING errors, if unsupported
1071 * character set is used. mb_encode_mimeheader function provided by php
1072 * mbstring extension is not used in order to get better control of header
1075 * Used php code functions - function_exists(), trigger_error(), strlen()
1076 * (is used with charset names and base64 strings). Used php mbstring
1077 * functions - mb_strlen and mb_substr.
1079 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1080 * encoding), rfc 2822 (header folding)
1082 * @param string $string header string that must be encoded
1083 * @param string $charset character set. Must be supported by mbstring extension.
1084 * Use sq_mb_list_encodings() to detect supported charsets.
1085 * @return string string encoded according to rfc2047 B encoding formating rules
1087 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1088 * @todo Do we want to control max length of header?
1089 * @todo Do we want to control EOL (end-of-line) marker?
1090 * @todo Do we want to translate error message?
1092 function encodeHeaderBase64($string,$charset) {
1094 * Check mbstring function requirements.
1096 if (! function_exists('mb_strlen') ||
1097 ! function_exists('mb_substr')) {
1098 // set E_USER_WARNING
1099 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1104 // initial return array
1108 * header length = 75 symbols max (same as in encodeHeader)
1109 * remove $charset length
1110 * remove =? ? ?= (5 chars)
1111 * remove 2 more chars (\r\n ?)
1113 $iMaxLength = 75 - strlen($charset) - 7;
1115 // set first character position
1118 // loop through all characters. count characters and not bytes.
1119 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1120 // encode string from starting character to current character.
1121 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1123 // Check encoded string length
1124 if(strlen($encoded_string)>$iMaxLength) {
1125 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1126 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1128 // set new starting character
1129 $iStartCharNum = $iCharNum-1;
1131 // encode last char (in case it is last character in string)
1132 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1133 } // if string is shorter than max length - add next character
1136 // add last encoded string to array
1137 $aRet[] = $encoded_string;
1139 // set initial return string
1142 // loop through encoded strings
1143 foreach($aRet as $string) {
1144 // TODO: Do we want to control EOL (end-of-line) marker
1145 if ($sRet!='') $sRet.= " ";
1147 // add header tags and encoded string to return string
1148 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1154 /* This function trys to locate the entity_id of a specific mime element */
1155 function find_ent_id($id, $message) {
1156 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1157 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1158 $ret = find_ent_id($id, $message->entities
[$i]);
1160 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1161 // if (sq_check_save_extension($message->entities[$i])) {
1162 return $message->entities
[$i]->entity_id
;
1164 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1166 * This is part of a fix for Outlook Express 6.x generating
1167 * cid URLs without creating content-id headers
1170 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1171 return $message->entities
[$i]->entity_id
;
1179 function sq_check_save_extension($message) {
1180 $filename = $message->getFilename();
1181 $ext = substr($filename, strrpos($filename,'.')+
1);
1182 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1183 return in_array($ext, $save_extensions);
1188 ** HTMLFILTER ROUTINES
1192 * This function checks attribute values for entity-encoded values
1193 * and returns them translated into 8-bit strings so we can run
1196 * @param $attvalue A string to run entity check against.
1197 * @return Nothing, modifies a reference value.
1199 function sq_defang(&$attvalue){
1202 * Skip this if there aren't ampersands or backslashes.
1204 if (strpos($attvalue, '&') === false
1205 && strpos($attvalue, '\\') === false){
1209 // before deent, translate the dangerous unicode characters and ... to safe values
1210 // otherwise the regular expressions do not match.
1216 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1217 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1218 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1219 } while ($m == true);
1220 $attvalue = stripslashes($attvalue);
1224 * Kill any tabs, newlines, or carriage returns. Our friends the
1225 * makers of the browser with 95% market value decided that it'd
1226 * be funny to make "java[tab]script" be just as good as "javascript".
1228 * @param attvalue The attribute value before extraneous spaces removed.
1229 * @return attvalue Nothing, modifies a reference value.
1231 function sq_unspace(&$attvalue){
1233 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1234 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1235 Array('', '', '', '', ''), $attvalue);
1240 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1241 * IE as regular characters.
1243 * @param attvalue The attribute value before dangerous characters are translated.
1244 * @return attvalue Nothing, modifies a reference value.
1245 * @author Marc Groot Koerkamp.
1247 function sq_fixIE_idiocy(&$attvalue) {
1249 $attvalue = str_replace("\0", "", $attvalue);
1251 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1253 // IE has the evil habit of accepting every possible value for the attribute expression.
1254 // The table below contains characters which are parsed by IE if they are used in the "expression"
1256 $aDangerousCharsReplacementTable = array(
1257 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1258 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1259 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1260 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1261 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1262 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1263 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1264 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1265 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1266 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1267 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1268 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1269 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1270 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1271 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1272 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1273 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1274 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1275 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1276 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1277 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1278 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1279 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1280 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1281 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1282 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1283 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1284 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1285 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1286 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1287 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1288 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1289 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1290 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1291 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1292 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1293 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1294 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1295 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1296 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1297 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1298 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1299 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1300 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1301 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1302 "\xCA\x9F", /* L UNICODE IPA Extension */
1303 "\xCA\x80", /* R UNICODE IPA Extension */
1304 "\xC9\xB4"), /* N UNICODE IPA Extension */
1305 array('l', 'l', 'r','r','n','n',
1306 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1307 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1308 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1309 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1311 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1313 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1317 * This function returns the final tag out of the tag name, an array
1318 * of attributes, and the type of the tag. This function is called by
1319 * sq_sanitize internally.
1321 * @param $tagname the name of the tag.
1322 * @param $attary the array of attributes and their values
1323 * @param $tagtype The type of the tag (see in comments).
1324 * @return a string with the final tag representation.
1326 function sq_tagprint($tagname, $attary, $tagtype){
1327 $me = 'sq_tagprint';
1330 $fulltag = '</' . $tagname . '>';
1332 $fulltag = '<' . $tagname;
1333 if (is_array($attary) && sizeof($attary)){
1335 while (list($attname, $attvalue) = each($attary)){
1336 array_push($atts, "$attname=$attvalue");
1338 $fulltag .= ' ' . join(" ", $atts);
1349 * A small helper function to use with array_walk. Modifies a by-ref
1350 * value and makes it lowercase.
1352 * @param $val a value passed by-ref.
1353 * @return void since it modifies a by-ref value.
1355 function sq_casenormalize(&$val){
1356 $val = strtolower($val);
1360 * This function skips any whitespace from the current position within
1361 * a string and to the next non-whitespace value.
1363 * @param $body the string
1364 * @param $offset the offset within the string where we should start
1365 * looking for the next non-whitespace character.
1366 * @return the location within the $body where the next
1367 * non-whitespace char is located.
1369 function sq_skipspace($body, $offset){
1370 $me = 'sq_skipspace';
1371 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1372 if (sizeof($matches{1})){
1373 $count = strlen($matches{1});
1380 * This function looks for the next character within a string. It's
1381 * really just a glorified "strpos", except it catches if failures
1384 * @param $body The string to look for needle in.
1385 * @param $offset Start looking from this position.
1386 * @param $needle The character/string to look for.
1387 * @return location of the next occurance of the needle, or
1388 * strlen($body) if needle wasn't found.
1390 function sq_findnxstr($body, $offset, $needle){
1391 $me = 'sq_findnxstr';
1392 $pos = strpos($body, $needle, $offset);
1393 if ($pos === FALSE){
1394 $pos = strlen($body);
1400 * This function takes a PCRE-style regexp and tries to match it
1401 * within the string.
1403 * @param $body The string to look for needle in.
1404 * @param $offset Start looking from here.
1405 * @param $reg A PCRE-style regex to match.
1406 * @return Returns a false if no matches found, or an array
1407 * with the following members:
1408 * - integer with the location of the match within $body
1409 * - string with whatever content between offset and the match
1410 * - string with whatever it is we matched
1412 function sq_findnxreg($body, $offset, $reg){
1413 $me = 'sq_findnxreg';
1416 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1417 if (!isset($matches{0}) ||
!$matches{0}){
1420 $retarr{0} = $offset +
strlen($matches{1});
1421 $retarr{1} = $matches{1};
1422 $retarr{2} = $matches{2};
1428 * This function looks for the next tag.
1430 * @param $body String where to look for the next tag.
1431 * @param $offset Start looking from here.
1432 * @return false if no more tags exist in the body, or
1433 * an array with the following members:
1434 * - string with the name of the tag
1435 * - array with attributes and their values
1436 * - integer with tag type (1, 2, or 3)
1437 * - integer where the tag starts (starting "<")
1438 * - integer where the tag ends (ending ">")
1439 * first three members will be false, if the tag is invalid.
1441 function sq_getnxtag($body, $offset){
1442 $me = 'sq_getnxtag';
1443 if ($offset > strlen($body)){
1446 $lt = sq_findnxstr($body, $offset, "<");
1447 if ($lt == strlen($body)){
1452 * blah blah <tag attribute="value">
1455 $pos = sq_skipspace($body, $lt+
1);
1456 if ($pos >= strlen($body)){
1457 return Array(false, false, false, $lt, strlen($body));
1460 * There are 3 kinds of tags:
1461 * 1. Opening tag, e.g.:
1463 * 2. Closing tag, e.g.:
1465 * 3. XHTML-style content-less tag, e.g.:
1466 * <img src="blah" />
1469 switch (substr($body, $pos, 1)){
1476 * A comment or an SGML declaration.
1478 if (substr($body, $pos+
1, 2) == "--"){
1479 $gt = strpos($body, "-->", $pos);
1481 $gt = strlen($body);
1485 return Array(false, false, false, $lt, $gt);
1487 $gt = sq_findnxstr($body, $pos, ">");
1488 return Array(false, false, false, $lt, $gt);
1493 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1503 * Look for next [\W-_], which will indicate the end of the tag name.
1505 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1506 if ($regary == false){
1507 return Array(false, false, false, $lt, strlen($body));
1509 list($pos, $tagname, $match) = $regary;
1510 $tagname = strtolower($tagname);
1513 * $match can be either of these:
1514 * '>' indicating the end of the tag entirely.
1515 * '\s' indicating the end of the tag name.
1516 * '/' indicating that this is type-3 xhtml tag.
1518 * Whatever else we find there indicates an invalid tag.
1523 * This is an xhtml-style tag with a closing / at the
1524 * end, like so: <img src="blah" />. Check if it's followed
1525 * by the closing bracket. If not, then this tag is invalid
1527 if (substr($body, $pos, 2) == "/>"){
1531 $gt = sq_findnxstr($body, $pos, ">");
1532 $retary = Array(false, false, false, $lt, $gt);
1536 return Array($tagname, false, $tagtype, $lt, $pos);
1540 * Check if it's whitespace
1542 if (!preg_match('/\s/', $match)){
1544 * This is an invalid tag! Look for the next closing ">".
1546 $gt = sq_findnxstr($body, $lt, ">");
1547 return Array(false, false, false, $lt, $gt);
1553 * At this point we're here:
1554 * <tagname attribute='blah'>
1557 * At this point we loop in order to find all attributes.
1563 while ($pos <= strlen($body)){
1564 $pos = sq_skipspace($body, $pos);
1565 if ($pos == strlen($body)){
1569 return Array(false, false, false, $lt, $pos);
1572 * See if we arrived at a ">" or "/>", which means that we reached
1573 * the end of the tag.
1576 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1580 $pos +
= strlen($matches{1});
1581 if ($matches{2} == "/>"){
1585 return Array($tagname, $attary, $tagtype, $lt, $pos);
1589 * There are several types of attributes, with optional
1590 * [:space:] between members.
1592 * attrname[:space:]=[:space:]'CDATA'
1594 * attrname[:space:]=[:space:]"CDATA"
1596 * attr[:space:]=[:space:]CDATA
1600 * We leave types 1 and 2 the same, type 3 we check for
1601 * '"' and convert to """ if needed, then wrap in
1602 * double quotes. Type 4 we convert into:
1605 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1606 if ($regary == false){
1608 * Looks like body ended before the end of tag.
1610 return Array(false, false, false, $lt, strlen($body));
1612 list($pos, $attname, $match) = $regary;
1613 $attname = strtolower($attname);
1615 * We arrived at the end of attribute name. Several things possible
1617 * '>' means the end of the tag and this is attribute type 4
1618 * '/' if followed by '>' means the same thing as above
1619 * '\s' means a lot of things -- look what it's followed by.
1620 * anything else means the attribute is invalid.
1625 * This is an xhtml-style tag with a closing / at the
1626 * end, like so: <img src="blah" />. Check if it's followed
1627 * by the closing bracket. If not, then this tag is invalid
1629 if (substr($body, $pos, 2) == "/>"){
1633 $gt = sq_findnxstr($body, $pos, ">");
1634 $retary = Array(false, false, false, $lt, $gt);
1638 $attary{$attname} = '"yes"';
1639 return Array($tagname, $attary, $tagtype, $lt, $pos);
1643 * Skip whitespace and see what we arrive at.
1645 $pos = sq_skipspace($body, $pos);
1646 $char = substr($body, $pos, 1);
1648 * Two things are valid here:
1649 * '=' means this is attribute type 1 2 or 3.
1650 * \w means this was attribute type 4.
1651 * anything else we ignore and re-loop. End of tag and
1652 * invalid stuff will be caught by our checks at the beginning
1657 $pos = sq_skipspace($body, $pos);
1659 * Here are 3 possibilities:
1660 * "'" attribute type 1
1661 * '"' attribute type 2
1662 * everything else is the content of tag type 3
1664 $quot = substr($body, $pos, 1);
1666 $regary = sq_findnxreg($body, $pos+
1, "\'");
1667 if ($regary == false){
1668 return Array(false, false, false, $lt, strlen($body));
1670 list($pos, $attval, $match) = $regary;
1672 $attary{$attname} = "'" . $attval . "'";
1673 } else if ($quot == '"'){
1674 $regary = sq_findnxreg($body, $pos+
1, '\"');
1675 if ($regary == false){
1676 return Array(false, false, false, $lt, strlen($body));
1678 list($pos, $attval, $match) = $regary;
1680 $attary{$attname} = '"' . $attval . '"';
1683 * These are hateful. Look for \s, or >.
1685 $regary = sq_findnxreg($body, $pos, "[\s>]");
1686 if ($regary == false){
1687 return Array(false, false, false, $lt, strlen($body));
1689 list($pos, $attval, $match) = $regary;
1691 * If it's ">" it will be caught at the top.
1693 $attval = preg_replace("/\"/s", """, $attval);
1694 $attary{$attname} = '"' . $attval . '"';
1696 } else if (preg_match("|[\w/>]|", $char)) {
1698 * That was attribute type 4.
1700 $attary{$attname} = '"yes"';
1703 * An illegal character. Find next '>' and return.
1705 $gt = sq_findnxstr($body, $pos, ">");
1706 return Array(false, false, false, $lt, $gt);
1712 * The fact that we got here indicates that the tag end was never
1713 * found. Return invalid tag indication so it gets stripped.
1715 return Array(false, false, false, $lt, strlen($body));
1719 * Translates entities into literal values so they can be checked.
1721 * @param $attvalue the by-ref value to check.
1722 * @param $regex the regular expression to check against.
1723 * @param $hex whether the entites are hexadecimal.
1724 * @return True or False depending on whether there were matches.
1726 function sq_deent(&$attvalue, $regex, $hex=false){
1730 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1731 preg_match_all($regex, $attvalue, $matches);
1732 if (is_array($matches) && sizeof($matches[0]) > 0){
1734 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1735 $numval = $matches[1][$i];
1737 $numval = hexdec($numval);
1739 $repl{$matches[0][$i]} = chr($numval);
1741 $attvalue = strtr($attvalue, $repl);
1749 * This function runs various checks against the attributes.
1751 * @param $tagname String with the name of the tag.
1752 * @param $attary Array with all tag attributes.
1753 * @param $rm_attnames See description for sq_sanitize
1754 * @param $bad_attvals See description for sq_sanitize
1755 * @param $add_attr_to_tag See description for sq_sanitize
1756 * @param $message message object
1757 * @param $id message id
1758 * @return Array with modified attributes.
1760 function sq_fixatts($tagname,
1770 while (list($attname, $attvalue) = each($attary)){
1772 * See if this attribute should be removed.
1774 foreach ($rm_attnames as $matchtag=>$matchattrs){
1775 if (preg_match($matchtag, $tagname)){
1776 foreach ($matchattrs as $matchattr){
1777 if (preg_match($matchattr, $attname)){
1778 unset($attary{$attname});
1785 * Workaround for IE quirks
1787 sq_fixIE_idiocy($attvalue);
1790 * Remove any backslashes, entities, and extraneous whitespace.
1793 $oldattvalue = $attvalue;
1794 sq_defang($attvalue);
1795 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1796 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1797 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1798 $attvalue = "idiocy";
1799 $attary{$attname} = $attvalue;
1801 sq_unspace($attvalue);
1804 * Now let's run checks on the attvalues.
1805 * I don't expect anyone to comprehend this. If you do,
1806 * get in touch with me so I can drive to where you live and
1807 * shake your hand personally. :)
1809 foreach ($bad_attvals as $matchtag=>$matchattrs){
1810 if (preg_match($matchtag, $tagname)){
1811 foreach ($matchattrs as $matchattr=>$valary){
1812 if (preg_match($matchattr, $attname)){
1814 * There are two arrays in valary.
1816 * Second one is replacements
1818 list($valmatch, $valrepl) = $valary;
1820 preg_replace($valmatch, $valrepl, $attvalue);
1821 if ($newvalue != $attvalue){
1822 $attary{$attname} = $newvalue;
1823 $attvalue = $newvalue;
1829 if ($attname == 'style') {
1830 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1831 // 8bit and control characters in style attribute values can be used for XSS, remove them
1832 $attary{$attname} = '"disallowed character"';
1834 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1835 if (count($aMatch)) {
1836 foreach($aMatch[1] as $sMatch) {
1838 $urlvalue = $sMatch;
1839 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1840 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1845 * Use white list based filtering on attributes which can contain url's
1847 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1848 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1849 $attary{$attname} = $attvalue;
1853 * See if we need to append any attributes to this tag.
1855 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1856 if (preg_match($matchtag, $tagname)){
1857 $attary = array_merge($attary, $addattary);
1864 * This function filters url's
1866 * @param $attvalue String with attribute value to filter
1867 * @param $message message object
1868 * @param $id message id
1869 * @param $mailbox mailbox
1870 * @param $sQuote quoting characters around url's
1872 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1873 $attvalue = trim($attvalue);
1874 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1875 // remove the double quotes
1876 $sQuote = $attvalue[0];
1877 $attvalue = trim(substr($attvalue,1,-1));
1880 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1881 // images off by default.
1882 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
1884 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1887 * Replace empty src tags with the blank image. src is only used
1888 * for frames, images, and image inputs. Doing a replace should
1889 * not affect them working as should be, however it will stop
1890 * IE from being kicked off when src for img tags are not set
1892 if ($attvalue == '') {
1893 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1895 // first, disallow 8 bit characters and control characters
1896 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1899 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1902 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1906 $aUrl = parse_url($attvalue);
1907 if (isset($aUrl['scheme'])) {
1908 switch(strtolower($aUrl['scheme'])) {
1913 if ($attname != 'href') {
1914 if ($view_unsafe_images == false) {
1915 $attvalue = $sQuote . $secremoveimg . $sQuote;
1917 if (isset($aUrl['path'])) {
1919 // No one has been able to show that image URIs
1920 // can be exploited, so for now, no restrictions
1921 // are made at all. If this proves to be a problem,
1922 // the commented-out code below can be of help.
1923 // (One consideration is that I see nothing in this
1924 // function that specifically says that we will
1925 // only ever arrive here when inspecting an image
1926 // tag, although that does seem to be the end
1927 // result - e.g., <script src="..."> where malicious
1928 // image URIs are in fact a problem are already
1929 // filtered out elsewhere.
1930 /* ---------------------------------
1931 // validate image extension.
1932 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1933 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1934 // If URI is to something other than
1935 // a regular image file, get the contents
1936 // and try to see if it is an image.
1937 // Don't use Fileinfo (finfo_file()) because
1938 // we'd need to make the admin configure the
1939 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1942 if (function_exists('mime_content_type')
1943 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1947 $file_contents = '';
1948 while (!feof($FILE)) {
1949 $file_contents .= fread($FILE, 8192);
1953 // store file locally
1955 global $attachment_dir, $username;
1956 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1957 $localfilename = GenerateRandomString(32, '', 7);
1958 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1959 while (file_exists($full_localfilename)) {
1960 $localfilename = GenerateRandomString(32, '', 7);
1961 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1963 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1964 fwrite($FILE, $file_contents);
1967 // get mime type and remove file
1969 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1970 unlink("$hashed_attachment_dir/$localfilename");
1972 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1973 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1974 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1977 --------------------------------- */
1979 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1983 $attvalue = $sQuote . $attvalue . $sQuote;
1988 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1989 * One day MS might actually make it match something useful, for now, falling
1990 * back to using cid2http, so we can grab the blank.png.
1992 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
1996 * Turn cid: urls into http-friendly ones.
1998 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2001 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2005 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
2006 // parse_url did not lead to satisfying result
2007 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2015 * This function edits the style definition to make them friendly and
2016 * usable in SquirrelMail.
2018 * @param $message the message object
2019 * @param $id the message id
2020 * @param $content a string with whatever is between <style> and </style>
2021 * @param $mailbox the message mailbox
2022 * @return a string with edited content.
2024 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
2025 $me = 'sq_fixstyle';
2026 // workaround for </style> in between comments
2027 $iCurrentPos = $pos;
2032 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
2039 if ($sToken == '<') {
2049 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2054 $content .= $sToken;
2062 if ($sToken == '<') {
2064 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
2065 $i = strpos($body,'-->',$i+
3);
2066 if ($i === false) { // no end comment
2084 if ($bSucces == FALSE){
2085 return array(FALSE, strlen($body));
2091 * First look for general BODY style declaration, which would be
2093 * body {background: blah-blah}
2094 * and change it to .bodyclass so we can just assign it to a <div>
2096 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2097 $secremoveimg = '../images/' . _("sec_remove_eng.png");
2099 * Fix url('blah') declarations.
2101 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2102 // "url(\\1$secremoveimg\\2)", $content);
2104 // first check for 8bit sequences and disallowed control characters
2105 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2106 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2107 return array($content, $newpos);
2110 // IE Sucks hard. We have a special function for it.
2111 sq_fixIE_idiocy($content);
2113 // remove @import line
2114 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2116 // translate ur\l and variations (IE parses that)
2117 // TODO check if the sq_fixIE_idiocy function already handles this.
2118 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2119 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2120 if (count($aMatch)) {
2121 $aValue = $aReplace = array();
2122 foreach($aMatch[1] as $sMatch) {
2124 $urlvalue = $sMatch;
2125 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2126 $aValue[] = $sMatch;
2127 $aReplace[] = $urlvalue;
2129 $content = str_replace($aValue,$aReplace,$content);
2133 * Remove any backslashes, entities, and extraneous whitespace.
2135 $contentTemp = $content;
2136 sq_defang($contentTemp);
2137 sq_unspace($contentTemp);
2140 * Fix stupid css declarations which lead to vulnerabilities
2143 $match = Array('/\/\*.*\*\//',
2147 '/include-source/i',
2150 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2151 $contentNew = preg_replace($match, $replace, $contentTemp);
2152 if ($contentNew !== $contentTemp) {
2153 // insecure css declarations are used. From now on we don't care
2154 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2155 $content = $contentNew;
2157 return array($content, $newpos);
2162 * This function converts cid: url's into the ones that can be viewed in
2165 * @param $message the message object
2166 * @param $id the message id
2167 * @param $cidurl the cid: url.
2168 * @param $mailbox the message mailbox
2169 * @return a string with a http-friendly url
2171 function sq_cid2http($message, $id, $cidurl, $mailbox){
2173 * Get rid of quotes.
2175 $quotchar = substr($cidurl, 0, 1);
2176 if ($quotchar == '"' ||
$quotchar == "'"){
2177 $cidurl = str_replace($quotchar, "", $cidurl);
2181 $cidurl = substr(trim($cidurl), 4);
2183 $match_str = '/\{.*?\}\//';
2185 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2187 $linkurl = find_ent_id($cidurl, $message);
2188 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2189 unsafe link image */
2193 * This is part of a fix for Outlook Express 6.x generating
2194 * cid URLs without creating content-id headers. These images are
2195 * not part of the multipart/related html mail. The html contains
2196 * <img src="cid:{some_id}/image_filename.ext"> references to
2197 * attached images with as goal to render them inline although
2198 * the attachment disposition property is not inline.
2201 if (empty($linkurl)) {
2202 if (preg_match('/{.*}\//', $cidurl)) {
2203 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2204 if (!empty($cidurl)) {
2205 $linkurl = find_ent_id($cidurl, $message);
2210 if (!empty($linkurl)) {
2211 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&' .
2212 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2213 '&ent_id=' . $linkurl . $quotchar;
2216 * If we couldn't generate a proper img url, drop in a blank image
2217 * instead of sending back empty, otherwise it causes unusual behaviour
2219 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2226 * This function changes the <body> tag into a <div> tag since we
2227 * can't really have a body-within-body.
2229 * @param $attary an array of attributes and values of <body>
2230 * @param $mailbox mailbox we're currently reading (for cid2http)
2231 * @param $message current message (for cid2http)
2232 * @param $id current message id (for cid2http)
2233 * @return a modified array of attributes to be set for <div>
2235 function sq_body2div($attary, $mailbox, $message, $id){
2236 $me = 'sq_body2div';
2237 $divattary = Array('class' => "'bodyclass'");
2239 $has_bgc_stl = $has_txt_stl = false;
2241 if (is_array($attary) && sizeof($attary) > 0){
2242 foreach ($attary as $attname=>$attvalue){
2243 $quotchar = substr($attvalue, 0, 1);
2244 $attvalue = str_replace($quotchar, "", $attvalue);
2247 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2248 $styledef .= "background-image: url('$attvalue'); ";
2251 $has_bgc_stl = true;
2252 $styledef .= "background-color: $attvalue; ";
2255 $has_txt_stl = true;
2256 $styledef .= "color: $attvalue; ";
2260 // Outlook defines a white bgcolor and no text color. This can lead to
2261 // white text on a white bg with certain themes.
2262 if ($has_bgc_stl && !$has_txt_stl) {
2263 $styledef .= "color: $text; ";
2265 if (strlen($styledef) > 0){
2266 $divattary{"style"} = "\"$styledef\"";
2273 * This is the main function and the one you should actually be calling.
2274 * There are several variables you should be aware of an which need
2275 * special description.
2277 * Since the description is quite lengthy, see it here:
2278 * http://linux.duke.edu/projects/mini/htmlfilter/
2280 * @param $body the string with HTML you wish to filter
2281 * @param $tag_list see description above
2282 * @param $rm_tags_with_content see description above
2283 * @param $self_closing_tags see description above
2284 * @param $force_tag_closing see description above
2285 * @param $rm_attnames see description above
2286 * @param $bad_attvals see description above
2287 * @param $add_attr_to_tag see description above
2288 * @param $message message object
2289 * @param $id message id
2290 * @return sanitized html safe to show on your pages.
2292 function sq_sanitize($body,
2294 $rm_tags_with_content,
2304 $me = 'sq_sanitize';
2305 $rm_tags = array_shift($tag_list);
2307 * Normalize rm_tags and rm_tags_with_content.
2309 @array_walk
($tag_list, 'sq_casenormalize');
2310 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2311 @array_walk
($self_closing_tags, 'sq_casenormalize');
2313 * See if tag_list is of tags to remove or tags to allow.
2314 * false means remove these tags
2315 * true means allow these tags
2318 $open_tags = Array();
2319 $trusted = "\n<!-- begin sanitized html -->\n";
2320 $skip_content = false;
2322 * Take care of netscape's stupid javascript entities like
2325 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2327 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2328 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2329 $free_content = substr($body, $curpos, $lt-$curpos);
2331 * Take care of <style>
2333 if ($tagname == "style" && $tagtype == 1){
2334 list($free_content, $curpos) =
2335 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2336 if ($free_content != FALSE){
2337 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2338 $trusted .= $free_content;
2339 $trusted .= sq_tagprint($tagname, false, 2);
2343 if ($skip_content == false){
2344 $trusted .= $free_content;
2346 if ($tagname != FALSE){
2348 if ($skip_content == $tagname){
2350 * Got to the end of tag we needed to remove.
2353 $skip_content = false;
2355 if ($skip_content == false){
2356 if ($tagname == "body"){
2359 if (isset($open_tags{$tagname}) &&
2360 $open_tags{$tagname} > 0){
2361 $open_tags{$tagname}--;
2369 * $rm_tags_with_content
2371 if ($skip_content == false){
2373 * See if this is a self-closing type and change
2374 * tagtype appropriately.
2377 && in_array($tagname, $self_closing_tags)){
2381 * See if we should skip this tag and any content
2384 if ($tagtype == 1 &&
2385 in_array($tagname, $rm_tags_with_content)){
2386 $skip_content = $tagname;
2388 if (($rm_tags == false
2389 && in_array($tagname, $tag_list)) ||
2390 ($rm_tags == true &&
2391 !in_array($tagname, $tag_list))){
2395 * Convert body into div.
2397 if ($tagname == "body"){
2399 $attary = sq_body2div($attary, $mailbox,
2403 if (isset($open_tags{$tagname})){
2404 $open_tags{$tagname}++
;
2406 $open_tags{$tagname}=1;
2410 * This is where we run other checks.
2412 if (is_array($attary) && sizeof($attary) > 0){
2413 $attary = sq_fixatts($tagname,
2427 if ($tagname != false && $skip_content == false){
2428 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2433 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2434 if ($force_tag_closing == true){
2435 foreach ($open_tags as $tagname=>$opentimes){
2436 while ($opentimes > 0){
2437 $trusted .= '</' . $tagname . '>';
2443 $trusted .= "<!-- end sanitized html -->\n";
2448 * This is a wrapper function to call html sanitizing routines.
2450 * @param $body the body of the message
2451 * @param $id the id of the message
2455 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2456 * into internal SM compose links
2457 * (optional; default = TRUE)
2458 * @return a string with html safe to display in the browser.
2460 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2462 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2464 global $attachment_common_show_images, $view_unsafe_images,
2467 * Don't display attached images in HTML mode.
2471 $attachment_common_show_images = false;
2486 $rm_tags_with_content = Array(
2496 $self_closing_tags = Array(
2504 $force_tag_closing = true;
2506 $rm_attnames = Array(
2517 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2518 $bad_attvals = Array(
2521 "/^src|background/i" =>
2524 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2525 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2526 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2529 "\\1$secremoveimg\\2",
2530 "\\1$secremoveimg\\2",
2531 "\\1$secremoveimg\\2",
2534 "/^href|action/i" =>
2537 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2538 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2539 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2554 "/include-source/i",
2555 "/position\s*:\s*absolute/i",
2556 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2557 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2558 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2559 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2560 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2579 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2580 // images off by default.
2581 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
2583 if (!$view_unsafe_images){
2585 * Remove any references to http/https if view_unsafe_images set
2588 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2589 '/^([\'\"])\s*https*:.*([\'\"])/si');
2590 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2591 "\\1$secremoveimg\\1");
2592 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2593 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2594 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2595 "url(\\1$secremoveimg\\1)");
2598 $add_attr_to_tag = Array(
2600 Array('target'=>'"_blank"',
2601 'title'=>'"'._("This external link will open in a new window").'"'
2604 $trusted = sq_sanitize($body,
2606 $rm_tags_with_content,
2616 if (strpos($trusted,$secremoveimg)){
2617 $has_unsafe_images = true;
2620 // we want to parse mailto's in HTML output, change to SM compose links
2621 // this is a modified version of code from url_parser.php... but Marc is
2622 // right: we need a better filtering implementation; adding this randomly
2623 // here is not a great solution
2625 if ($take_mailto_links) {
2626 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2627 global $MailTo_PReg_Match;
2628 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2629 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2630 foreach ($regs[0] as $i => $mailto_before) {
2631 $mailto_params = $regs[10][$i];
2632 // get rid of any tailing quote since we have to add send_to to the end
2634 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2635 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2636 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2637 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2639 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2640 $to = 'to=' . $regs[1][$i];
2641 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2642 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2644 if ($mailto_params) //already some params, append to them
2645 $mailto_params .= '&' . $to;
2647 $mailto_params .= '?' . $to;
2651 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2653 // we'll already have target=_blank, no need to allow comp_in_new
2654 // here (which would be a lot more work anyway)
2656 global $compose_new_win;
2657 $temp_comp_in_new = $compose_new_win;
2658 $compose_new_win = 0;
2659 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2660 $compose_new_win = $temp_comp_in_new;
2662 // remove <a href=" and anything after the next quote (we only
2663 // need the uri, not the link HTML) in compose uri
2665 $comp_uri = substr($comp_uri, 9);
2666 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2667 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2676 * function SendDownloadHeaders - send file to the browser
2678 * Original Source: SM core src/download.php
2679 * moved here to make it available to other code, and separate
2680 * front end from back end functionality.
2682 * @param string $type0 first half of mime type
2683 * @param string $type1 second half of mime type
2684 * @param string $filename filename to tell the browser for downloaded file
2685 * @param boolean $force whether to force the download dialog to pop
2686 * @param optional integer $filesize send the Content-Header and length to the browser
2689 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2690 global $languages, $squirrelmail_language;
2691 $isIE = $isIE6plus = false;
2693 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2695 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2696 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2700 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2701 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2705 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2706 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2708 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2710 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace(' ', ' ', $filename));
2713 // A Pox on Microsoft and it's Internet Explorer!
2715 // IE has lots of bugs with file downloads.
2716 // It also has problems with SSL. Both of these cause problems
2717 // for us in this function.
2719 // See this article on Cache Control headers and SSL
2720 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2722 // The best thing you can do for IE is to upgrade to the latest
2724 //set all the Cache Control Headers for IE
2726 $filename=rawurlencode($filename);
2727 header ("Pragma: public");
2728 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2729 header ("Cache-Control: post-check=0, pre-check=0", false);
2730 header ("Cache-Control: private");
2732 //set the inline header for IE, we'll add the attachment header later if we need it
2733 header ("Content-Disposition: inline; filename=$filename");
2737 // Try to show in browser window
2738 header ("Content-Disposition: inline; filename=\"$filename\"");
2739 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2741 // Try to pop up the "save as" box
2743 // IE makes this hard. It pops up 2 save boxes, or none.
2744 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2745 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2746 // But, according to Microsoft, it is "RFC compliant but doesn't
2747 // take into account some deviations that allowed within the
2748 // specification." Doesn't that mean RFC non-compliant?
2749 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2751 // all browsers need the application/octet-stream header for this
2752 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2754 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2755 // Do not have quotes around filename, but that applied to
2756 // "attachment"... does it apply to inline too?
2757 header ("Content-Disposition: attachment; filename=\"$filename\"");
2759 if ($isIE && !$isIE6plus) {
2760 // This combination seems to work mostly. IE 5.5 SP 1 has
2761 // known issues (see the Microsoft Knowledge Base)
2763 // This works for most types, but doesn't work with Word files
2764 header ("Content-Type: application/download; name=\"$filename\"");
2765 header ("Content-Type: application/force-download; name=\"$filename\"");
2766 // These are spares, just in case. :-)
2767 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2768 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2769 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2771 // This is to prevent IE for MIME sniffing and auto open a file in IE
2772 header ("Content-Type: application/force-download; name=\"$filename\"");
2774 // another application/octet-stream forces download for Netscape
2775 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2779 //send the content-length header if the calling function provides it
2780 if ($filesize > 0) {
2781 header("Content-Length: $filesize");
2784 } // end fn SendDownloadHeaders