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 $line = rtrim($body_ary[$i],"\r");
296 if (strlen($line) - 2 >= $wrap_at) {
297 sqWordWrap($line, $wrap_at, $charset);
299 $line = charset_decode($charset, $line);
300 $line = str_replace("\t", ' ', $line);
309 if ($line[$pos] == ' ') {
311 } else if (strpos($line, '>', $pos) === $pos) {
320 $line = '<span class="quote1">' . $line . '</span>';
322 $line = '<span class="quote2">' . $line . '</span>';
325 $body_ary[$i] = $line;
327 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
331 * This returns a parsed string called $body. That string can then
332 * be displayed as the actual message in the HTML. It contains
333 * everything needed, including HTML Tags, Attachments at the
336 * Since 1.2.0 function uses message_body hook.
337 * Till 1.3.0 function included output of formatAttachments().
339 * @param resource $imap_stream imap connection resource
340 * @param object $message squirrelmail message object
341 * @param array $color squirrelmail color theme array
342 * @param integer $wrap_at number of characters per line
343 * @param string $ent_num (since 1.3.0) message part id
344 * @param integer $id (since 1.3.0) message id
345 * @param string $mailbox (since 1.3.0) imap folder name
346 * @return string html formated message text
348 function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
349 /* This if statement checks for the entity to show as the
350 * primary message. To add more of them, just put them in the
351 * order that is their priority.
353 global $startMessage, $languages, $squirrelmail_language,
354 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
355 $use_iframe, $iframe_height, $download_and_unsafe_link,
356 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
359 // workaround for not updated config.php
360 if (! isset($use_iframe)) $use_iframe = false;
362 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
363 // images off by default.
364 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
367 $urlmailbox = urlencode($mailbox);
368 $body_message = getEntity($message, $ent_num);
369 if (($body_message->header
->type0
== 'text') ||
370 ($body_message->header
->type0
== 'rfc822')) {
371 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
372 $body = decodeBody($body, $body_message->header
->encoding
);
374 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
375 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
376 if (mb_detect_encoding($body) != 'ASCII') {
377 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
381 /* As of 1.5.2, $body is passed (and modified) by reference */
382 do_hook('message_body', $body);
384 /* If there are other types that shouldn't be formatted, add
388 if ($body_message->header
->type1
== 'html') {
389 if ($show_html_default <> 1) {
390 $entity_conv = array(' ' => ' ',
399 $body = strtr($body, $entity_conv);
400 $body = strip_tags($body);
402 translateText($body, $wrap_at,
403 $body_message->header
->getParameter('charset'));
404 } elseif ($use_iframe) {
406 * If we don't add html message between iframe tags,
407 * we must detect unsafe images and modify $has_unsafe_images.
409 $html_body = magicHTML($body, $id, $message, $mailbox);
410 // Convert character set in order to display html mails in different character set
411 $html_body = charset_decode($body_message->header
->getParameter('charset'),$html_body,false,true);
413 // creating iframe url
414 $iframeurl=sqm_baseuri().'src/view_html.php?'
415 . 'mailbox=' . $urlmailbox
416 . '&passed_id=' . $id
417 . '&ent_id=' . $ent_num
418 . '&view_unsafe_images=' . (int) $view_unsafe_images;
421 $oTemplate->assign('iframe_url', $iframeurl);
422 $oTemplate->assign('iframe_height', $iframe_height);
423 $oTemplate->assign('html_body', $html_body);
425 $body = $oTemplate->fetch('read_html_iframe.tpl');
427 // old way of html rendering
429 * convert character set. charset_decode does not remove html special chars
430 * applied by magicHTML functions and does not sanitize them second time if
431 * fourth argument is true.
433 $charset = $body_message->header
->getParameter('charset');
434 if (!empty($charset)) {
435 $body = charset_decode($charset,$body,false,true);
437 $body = magicHTML($body, $id, $message, $mailbox);
440 translateText($body, $wrap_at,
441 $body_message->header
->getParameter('charset'));
445 * Previously the links for downloading and unsafe images were printed
446 * under the mail. By putting the links in a global variable we can
447 * print it in the toolbar where it belongs. Since the original code was
448 * in this place it's left here. It might be possible to move it to some
449 * other place if that makes sense. The possibility to do so has not
450 * been evaluated yet.
453 // Initialize the global variable to an empty string.
454 // 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.
455 $download_and_unsafe_link = '';
457 // Prepare and build a link for downloading the mail.
458 $link = 'passed_id=' . $id . '&ent_id='.$ent_num.
459 '&mailbox=' . $urlmailbox .'&sort=' . $sort .
460 '&startMessage=' . $startMessage . '&show_more=0';
461 if (isset($passed_ent_id)) {
462 $link .= '&passed_ent_id='.$passed_ent_id;
464 $download_href = SM_PATH
. 'src/download.php?absolute_dl=true&' . $link;
466 // Always add the link for downloading the mail as a file to the global
468 $download_and_unsafe_link .= "$nbsp|$nbsp"
469 . create_hyperlink($download_href, _("Download this as a file"));
471 // Find out the right text to use in the link depending on the
472 // circumstances. If the unsafe images are displayed the link should
473 // hide them, if they aren't displayed the link should only appear if
474 // the mail really contains unsafe images.
475 if ($view_unsafe_images) {
476 $text = _("Hide Unsafe Images");
478 if (isset($has_unsafe_images) && $has_unsafe_images) {
479 $link .= '&view_unsafe_images=1';
480 $text = _("View Unsafe Images");
486 // Only create a link for unsafe images if there's need for one. If so:
487 // add it to the global variable.
489 $unsafe_image_toggle_href = SM_PATH
. 'src/read_body.php?'.$link;
490 $unsafe_image_toggle_text = $text;
491 $download_and_unsafe_link .= "$nbsp|$nbsp"
492 . create_hyperlink($unsafe_image_toggle_href, $text);
499 * Generate attachments array for passing to templates.
502 * @param object $message SquirrelMail message object
503 * @param array $exclude_id message parts that are not attachments.
504 * @param string $mailbox mailbox name
505 * @param integer $id message id
507 function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
508 global $where, $what, $startMessage, $color, $passed_ent_id, $base_uri;
510 $att_ar = $message->getAttachments($exclude_id);
511 $urlMailbox = urlencode($mailbox);
513 $attachments = array();
514 foreach ($att_ar as $att) {
515 $ent = $att->entity_id
;
516 $header = $att->header
;
517 $type0 = strtolower($header->type0
);
518 $type1 = strtolower($header->type1
);
521 $links['download link']['text'] = _("Download");
522 $links['download link']['href'] = $base_uri .
523 "src/download.php?absolute_dl=true&passed_id=$id&mailbox=$urlMailbox&ent_id=$ent";
525 if ($type0 =='message' && $type1 == 'rfc822') {
526 $default_page = $base_uri . 'src/read_body.php';
527 $rfc822_header = $att->rfc822_header
;
528 $filename = $rfc822_header->subject
;
529 if (trim( $filename ) == '') {
530 $filename = 'untitled-[' . $ent . ']' ;
532 $from_o = $rfc822_header->from
;
533 if (is_object($from_o)) {
534 $from_name = decodeHeader($from_o->getAddress(false));
535 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
536 // something weird happens when a digest message is opened and you return to the digest
537 // now the from object is part of an array. Probably the parseHeader call overwrites the info
538 // retrieved from the bodystructure in a different way. We need to fix this later.
539 // possible starting point, do not fetch header we already have and inspect how
540 // the rfc822_header object behaves.
541 $from_name = decodeHeader($from_o[0]->getAddress(false));
543 $from_name = _("Unknown sender");
545 $description = _("From").': '.$from_name;
547 $default_page = $base_uri . 'src/download.php';
548 $filename = $att->getFilename();
549 if ($header->description
) {
550 $description = decodeHeader($header->description
);
556 $display_filename = $filename;
557 if (isset($passed_ent_id)) {
558 $passed_ent_id_link = '&passed_ent_id='.$passed_ent_id;
560 $passed_ent_id_link = '';
562 $defaultlink = $default_page . "?startMessage=$startMessage"
563 . "&passed_id=$id&mailbox=$urlMailbox"
564 . '&ent_id='.$ent.$passed_ent_id_link;
565 if ($where && $what) {
566 $defaultlink .= '&where='. urlencode($where).'&what='.urlencode($what);
568 // IE does make use of mime content sniffing. Forcing a download
569 // prohibit execution of XSS inside an application/octet-stream attachment
570 if ($type0 == 'application' && $type1 == 'octet-stream') {
571 $defaultlink .= '&absolute_dl=true';
574 /* This executes the attachment hook with a specific MIME-type.
575 * If that doesn't have results, it tries if there's a rule
576 * for a more generic type. Finally, a hook for ALL attachment
577 * types is run as well.
579 // First remember the default link.
580 $defaultlink_orig = $defaultlink;
582 /* The API for this hook has changed as of 1.5.2 so that all plugin
583 arguments are passed in an array instead of each their own plugin
584 argument, and arguments are passed by reference, so instead of
585 returning any changes, changes should simply be made to the original
586 arguments themselves. */
587 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
588 &$defaultlink, &$display_filename, &$where, &$what);
589 do_hook("attachment $type0/$type1", $temp);
590 if(count($links) <= 1 && $defaultlink == $defaultlink_orig) {
591 /* The API for this hook has changed as of 1.5.2 so that all plugin
592 arguments are passed in an array instead of each their own plugin
593 argument, and arguments are passed by reference, so instead of
594 returning any changes, changes should simply be made to the original
595 arguments themselves. */
596 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
597 &$defaultlink, &$display_filename, &$where, &$what);
598 do_hook("attachment $type0/*", $temp);
600 /* The API for this hook has changed as of 1.5.2 so that all plugin
601 arguments are passed in an array instead of each their own plugin
602 argument, and arguments are passed by reference, so instead of
603 returning any changes, changes should simply be made to the original
604 arguments themselves. */
605 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
606 &$defaultlink, &$display_filename, &$where, &$what);
607 // Do not let a generic plugin change the default link if a more
608 // specialized one already did it...
609 if ($defaultlink != $defaultlink_orig) {
613 do_hook("attachment */*", $temp);
615 $this_attachment = array();
616 $this_attachment['Name'] = decodeHeader($display_filename);
617 $this_attachment['Description'] = $description;
618 $this_attachment['DefaultHREF'] = $defaultlink;
619 $this_attachment['DownloadHREF'] = $links['download link']['href'];
620 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ?
$links['attachment_common']['href'] : '';
621 $this_attachment['Size'] = $header->size
;
622 $this_attachment['ContentType'] = htmlspecialchars($type0 .'/'. $type1);
623 $this_attachment['OtherLinks'] = array();
624 foreach ($links as $val) {
625 if ($val['text']==_("Download") ||
$val['text'] == _("View"))
627 if (empty($val['text']) && empty($val['extra']))
631 $temp['HREF'] = $val['href'];
632 $temp['Text'] = (empty($val['text']) ?
'' : $val['text']) . (empty($val['extra']) ?
'' : $val['extra']);
633 $this_attachment['OtherLinks'][] = $temp;
635 $attachments[] = $this_attachment;
644 * Displays attachment links and information
646 * Since 1.3.0 function is not included in formatBody() call.
648 * Since 1.0.2 uses attachment $type0/$type1 hook.
649 * Since 1.2.5 uses attachment $type0/* hook.
650 * Since 1.5.0 uses attachments_bottom hook.
651 * Since 1.5.2 uses templates and does *not* return a value.
653 * @param object $message SquirrelMail message object
654 * @param array $exclude_id message parts that are not attachments.
655 * @param string $mailbox mailbox name
656 * @param integer $id message id
658 function formatAttachments($message, $exclude_id, $mailbox, $id) {
661 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
663 $oTemplate->assign('attachments', $attach);
664 $oTemplate->display('read_attachments.tpl');
667 function sqimap_base64_decode(&$string) {
669 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
670 // fly decoding (to reduce memory usage) you have to check if the
671 // data has incomplete pairs
673 // Remove the noise in order to check if the 4 bytes pairs are complete
674 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
677 $iMod = strlen($string) %
4;
679 $sStringRem = substr($string,-$iMod);
680 // Check if $sStringRem contains padding characters
681 if (substr($sStringRem,-1) != '=') {
682 $string = substr($string,0,-$iMod);
687 $string = base64_decode($string);
692 * Decodes encoded string (usually message body)
694 * This function decodes a string (usually the message body)
695 * depending on the encoding type. Currently quoted-printable
696 * and base64 encodings are supported.
698 * The decode_body hook was added to this function in 1.4.2/1.5.0.
699 * The $force_crlf parameter was added in 1.5.2.
701 * @param string $string The encoded string
702 * @param string $encoding used encoding
703 * @param string $force_crlf Whether or not to force CRLF or LF
704 * line endings (or to leave as is).
705 * If given as "LF", line endings will
706 * all be converted to LF; if "CRLF",
707 * line endings will all be converted
708 * to CRLF. If given as an empty value,
709 * the global $force_crlf_default will
710 * be consulted (it can be specified in
711 * config/config_local.php). Otherwise,
712 * any other value will cause the string
713 * to be left alone. Note that this will
714 * be overridden to "LF" if not using at
715 * least PHP version 4.3.0. (OPTIONAL;
716 * default is empty - consult global
719 * @return string The decoded string
724 function decodeBody($string, $encoding, $force_crlf='') {
726 global $force_crlf_default;
727 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
728 $force_crlf = strtoupper($force_crlf);
730 // must force line endings to LF due to broken
731 // quoted_printable_decode() in PHP versions
732 // before 4.3.0 (see below)
734 if (!check_php_version(4, 3, 0) ||
$force_crlf == 'LF')
735 $string = str_replace("\r\n", "\n", $string);
736 else if ($force_crlf == 'CRLF')
737 $string = str_replace("\n", "\r\n", $string);
739 $encoding = strtolower($encoding);
741 $encoding_handler = do_hook('decode_body', $encoding);
744 // plugins get first shot at decoding the string
746 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
747 $string = $encoding_handler('decode', $string);
749 } elseif ($encoding == 'quoted-printable' ||
750 $encoding == 'quoted_printable') {
752 // quoted_printable_decode() function is broken in older
753 // php versions. Text with \r\n decoding was fixed only
754 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
755 // and the above call to: str_replace("\r\n", "\n", $string);
757 $string = quoted_printable_decode($string);
759 } elseif ($encoding == 'base64') {
760 $string = base64_decode($string);
763 // All other encodings are returned raw.
770 * This function decodes strings that are encoded according to
771 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
772 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
774 * @param string $string header string that has to be made readable
775 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
776 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
777 * @param boolean $decide decide if string can be utfencoded. defaults to false
778 * @return string decoded header string
780 function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
781 global $languages, $squirrelmail_language,$default_charset;
782 if (is_array($string)) {
783 $string = implode("\n", $string);
786 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
787 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
788 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
789 // Do we need to return at this point?
796 $aString = explode(' ',$string);
798 foreach ($aString as $chunk) {
799 if ($encoded && $chunk === '') {
801 } elseif ($chunk === '') {
806 /* if encoded words are not separated by a linear-space-white we still catch them */
809 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
810 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
811 if ($iLastMatch !== $j) {
821 $ret .= htmlspecialchars($res[1]);
825 $encoding = ucfirst($res[3]);
827 /* decide about valid decoding */
828 if ($decide && is_conversion_safe($res[2])) {
830 $can_be_encoded=true;
832 $can_be_encoded=false;
837 $replace = base64_decode($res[4]);
839 if ($can_be_encoded) {
840 /* convert string to different charset,
841 * if functions asks for it (usually in compose)
843 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
845 // convert string to html codes in order to display it
846 $ret .= charset_decode($res[2],$replace);
850 $replace = htmlspecialchars($replace);
856 $replace = str_replace('_', ' ', $res[4]);
857 $replace = preg_replace('/=([0-9a-f]{2})/ie', 'chr(hexdec("\1"))',
860 if ($can_be_encoded) {
861 /* convert string to different charset,
862 * if functions asks for it (usually in compose)
864 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
866 // convert string to html codes in order to display it
867 $replace = charset_decode($res[2], $replace);
871 $replace = htmlspecialchars($replace);
890 if (!$encoded && $htmlsafe) {
891 $ret .= htmlspecialchars($chunk);
897 /* remove the first added space */
900 $ret = substr($ret,5);
902 $ret = substr($ret,1);
912 * Function uses XTRA_CODE _encodeheader function, if such function exists.
914 * Function uses Q encoding by default and encodes a string according to RFC
915 * 1522 for use in headers if it contains 8-bit characters or anything that
916 * looks like it should be encoded.
918 * Function switches to B encoding and encodeHeaderBase64() function, if
919 * string is 8bit and multibyte character set supported by mbstring extension
920 * is used. It can cause E_USER_NOTICE errors, if interface is used with
921 * multibyte character set unsupported by mbstring extension.
923 * @param string $string header string, that has to be encoded
924 * @return string quoted-printable encoded string
925 * @todo make $mb_charsets system wide constant
927 function encodeHeader ($string) {
928 global $default_charset, $languages, $squirrelmail_language;
930 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
931 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
932 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
935 // Use B encoding for multibyte charsets
936 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
937 if (in_array($default_charset,$mb_charsets) &&
938 in_array($default_charset,sq_mb_list_encodings()) &&
939 sq_is8bit($string)) {
940 return encodeHeaderBase64($string,$default_charset);
941 } elseif (in_array($default_charset,$mb_charsets) &&
942 sq_is8bit($string) &&
943 ! in_array($default_charset,sq_mb_list_encodings())) {
944 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
945 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
948 // Encode only if the string contains 8-bit characters or =?
949 $j = strlen($string);
950 $max_l = 75 - strlen($default_charset) - 7;
953 $iEncStart = $enc_init = false;
954 $cur_l = $iOffset = 0;
955 for($i = 0; $i < $j; ++
$i) {
964 if ($iEncStart === false) {
968 if ($cur_l > ($max_l-2)) {
969 /* if there is an stringpart that doesn't need encoding, add it */
970 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
971 $aRet[] = "=?$default_charset?Q?$ret?=";
977 $ret .= sprintf("=%02X",ord($string{$i}));
982 if ($iEncStart !== false) {
983 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
984 $aRet[] = "=?$default_charset?Q?$ret?=";
992 if ($iEncStart !== false) {
994 if ($cur_l > $max_l) {
995 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
996 $aRet[] = "=?$default_charset?Q?$ret?=";
1007 $k = ord($string{$i});
1009 if ($iEncStart === false) {
1010 // do not start encoding in the middle of a string, also take the rest of the word.
1011 $sLeadString = substr($string,0,$i);
1012 $aLeadString = explode(' ',$sLeadString);
1013 $sToBeEncoded = array_pop($aLeadString);
1014 $iEncStart = $i - strlen($sToBeEncoded);
1015 $ret .= $sToBeEncoded;
1016 $cur_l +
= strlen($sToBeEncoded);
1019 /* first we add the encoded string that reached it's max size */
1020 if ($cur_l > ($max_l-2)) {
1021 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1022 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1029 $ret .= sprintf("=%02X", $k);
1031 if ($iEncStart !== false) {
1033 if ($cur_l > $max_l) {
1034 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1035 $aRet[] = "=?$default_charset?Q?$ret?=";
1041 $ret .= $string{$i};
1050 if ($iEncStart !== false) {
1051 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1052 $aRet[] = "=?$default_charset?Q?$ret?=";
1054 $aRet[] = substr($string,$iOffset);
1056 $string = implode('',$aRet);
1062 * Encodes string according to rfc2047 B encoding header formating rules
1064 * It is recommended way to encode headers with character sets that store
1065 * symbols in more than one byte.
1067 * Function requires mbstring support. If required mbstring functions are missing,
1068 * function returns false and sets E_USER_WARNING level error message.
1070 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1071 * that mbstring functions will generate E_WARNING errors, if unsupported
1072 * character set is used. mb_encode_mimeheader function provided by php
1073 * mbstring extension is not used in order to get better control of header
1076 * Used php code functions - function_exists(), trigger_error(), strlen()
1077 * (is used with charset names and base64 strings). Used php mbstring
1078 * functions - mb_strlen and mb_substr.
1080 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
1081 * encoding), rfc 2822 (header folding)
1083 * @param string $string header string that must be encoded
1084 * @param string $charset character set. Must be supported by mbstring extension.
1085 * Use sq_mb_list_encodings() to detect supported charsets.
1086 * @return string string encoded according to rfc2047 B encoding formating rules
1088 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1089 * @todo Do we want to control max length of header?
1090 * @todo Do we want to control EOL (end-of-line) marker?
1091 * @todo Do we want to translate error message?
1093 function encodeHeaderBase64($string,$charset) {
1095 * Check mbstring function requirements.
1097 if (! function_exists('mb_strlen') ||
1098 ! function_exists('mb_substr')) {
1099 // set E_USER_WARNING
1100 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING
);
1105 // initial return array
1109 * header length = 75 symbols max (same as in encodeHeader)
1110 * remove $charset length
1111 * remove =? ? ?= (5 chars)
1112 * remove 2 more chars (\r\n ?)
1114 $iMaxLength = 75 - strlen($charset) - 7;
1116 // set first character position
1119 // loop through all characters. count characters and not bytes.
1120 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++
) {
1121 // encode string from starting character to current character.
1122 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1124 // Check encoded string length
1125 if(strlen($encoded_string)>$iMaxLength) {
1126 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1127 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1129 // set new starting character
1130 $iStartCharNum = $iCharNum-1;
1132 // encode last char (in case it is last character in string)
1133 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1134 } // if string is shorter than max length - add next character
1137 // add last encoded string to array
1138 $aRet[] = $encoded_string;
1140 // set initial return string
1143 // loop through encoded strings
1144 foreach($aRet as $string) {
1145 // TODO: Do we want to control EOL (end-of-line) marker
1146 if ($sRet!='') $sRet.= " ";
1148 // add header tags and encoded string to return string
1149 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1155 /* This function trys to locate the entity_id of a specific mime element */
1156 function find_ent_id($id, $message) {
1157 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities
); $i++
) {
1158 if ($message->entities
[$i]->header
->type0
== 'multipart') {
1159 $ret = find_ent_id($id, $message->entities
[$i]);
1161 if (strcasecmp($message->entities
[$i]->header
->id
, $id) == 0) {
1162 // if (sq_check_save_extension($message->entities[$i])) {
1163 return $message->entities
[$i]->entity_id
;
1165 } elseif (!empty($message->entities
[$i]->header
->parameters
['name'])) {
1167 * This is part of a fix for Outlook Express 6.x generating
1168 * cid URLs without creating content-id headers
1171 if (strcasecmp($message->entities
[$i]->header
->parameters
['name'], $id) == 0) {
1172 return $message->entities
[$i]->entity_id
;
1180 function sq_check_save_extension($message) {
1181 $filename = $message->getFilename();
1182 $ext = substr($filename, strrpos($filename,'.')+
1);
1183 $save_extensions = array('jpg','jpeg','gif','png','bmp');
1184 return in_array($ext, $save_extensions);
1189 ** HTMLFILTER ROUTINES
1193 * This function checks attribute values for entity-encoded values
1194 * and returns them translated into 8-bit strings so we can run
1197 * @param $attvalue A string to run entity check against.
1198 * @return Nothing, modifies a reference value.
1200 function sq_defang(&$attvalue){
1203 * Skip this if there aren't ampersands or backslashes.
1205 if (strpos($attvalue, '&') === false
1206 && strpos($attvalue, '\\') === false){
1210 // before deent, translate the dangerous unicode characters and ... to safe values
1211 // otherwise the regular expressions do not match.
1217 $m = $m ||
sq_deent($attvalue, '/\�*(\d+);*/s');
1218 $m = $m ||
sq_deent($attvalue, '/\�*((\d|[a-f])+);*/si', true);
1219 $m = $m ||
sq_deent($attvalue, '/\\\\(\d+)/s', true);
1220 } while ($m == true);
1221 $attvalue = stripslashes($attvalue);
1225 * Kill any tabs, newlines, or carriage returns. Our friends the
1226 * makers of the browser with 95% market value decided that it'd
1227 * be funny to make "java[tab]script" be just as good as "javascript".
1229 * @param attvalue The attribute value before extraneous spaces removed.
1230 * @return attvalue Nothing, modifies a reference value.
1232 function sq_unspace(&$attvalue){
1234 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1235 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1236 Array('', '', '', '', ''), $attvalue);
1241 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
1242 * IE as regular characters.
1244 * @param attvalue The attribute value before dangerous characters are translated.
1245 * @return attvalue Nothing, modifies a reference value.
1246 * @author Marc Groot Koerkamp.
1248 function sq_fixIE_idiocy(&$attvalue) {
1250 $attvalue = str_replace("\0", "", $attvalue);
1252 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1254 // IE has the evil habit of accepting every possible value for the attribute expression.
1255 // The table below contains characters which are parsed by IE if they are used in the "expression"
1257 $aDangerousCharsReplacementTable = array(
1258 array('ʟ', 'ʟ' ,/* L UNICODE IPA Extension */
1259 'ʀ', 'ʀ' ,/* R UNICODE IPA Extension */
1260 'ɴ', 'ɴ' ,/* N UNICODE IPA Extension */
1261 'E', 'E' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1262 'e', 'e' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
1263 'X', 'X',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1264 'x', 'x',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1265 'P', 'P',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1266 'p', 'p',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1267 'R', 'R',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1268 'r', 'r',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1269 'S', 'S',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1270 's', 's',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1271 'I', 'I',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1272 'i', 'i',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1273 'O', 'O',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1274 'o', 'o',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1275 'N', 'N',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1276 'n', 'n',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1277 'L', 'L',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1278 'l', 'l',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1279 'U', 'U',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1280 'u', 'u',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1281 'ⁿ', 'ⁿ' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
1282 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1283 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1284 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1285 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1286 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1287 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1288 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1289 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1290 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1291 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1292 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1293 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1294 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1295 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1296 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1297 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1298 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1299 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1300 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1301 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1302 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1303 "\xCA\x9F", /* L UNICODE IPA Extension */
1304 "\xCA\x80", /* R UNICODE IPA Extension */
1305 "\xC9\xB4"), /* N UNICODE IPA Extension */
1306 array('l', 'l', 'r','r','n','n',
1307 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1308 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1309 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
1310 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1312 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1314 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1318 * This function returns the final tag out of the tag name, an array
1319 * of attributes, and the type of the tag. This function is called by
1320 * sq_sanitize internally.
1322 * @param $tagname the name of the tag.
1323 * @param $attary the array of attributes and their values
1324 * @param $tagtype The type of the tag (see in comments).
1325 * @return a string with the final tag representation.
1327 function sq_tagprint($tagname, $attary, $tagtype){
1328 $me = 'sq_tagprint';
1331 $fulltag = '</' . $tagname . '>';
1333 $fulltag = '<' . $tagname;
1334 if (is_array($attary) && sizeof($attary)){
1336 while (list($attname, $attvalue) = each($attary)){
1337 array_push($atts, "$attname=$attvalue");
1339 $fulltag .= ' ' . join(" ", $atts);
1350 * A small helper function to use with array_walk. Modifies a by-ref
1351 * value and makes it lowercase.
1353 * @param $val a value passed by-ref.
1354 * @return void since it modifies a by-ref value.
1356 function sq_casenormalize(&$val){
1357 $val = strtolower($val);
1361 * This function skips any whitespace from the current position within
1362 * a string and to the next non-whitespace value.
1364 * @param $body the string
1365 * @param $offset the offset within the string where we should start
1366 * looking for the next non-whitespace character.
1367 * @return the location within the $body where the next
1368 * non-whitespace char is located.
1370 function sq_skipspace($body, $offset){
1371 $me = 'sq_skipspace';
1372 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
1373 if (sizeof($matches{1})){
1374 $count = strlen($matches{1});
1381 * This function looks for the next character within a string. It's
1382 * really just a glorified "strpos", except it catches if failures
1385 * @param $body The string to look for needle in.
1386 * @param $offset Start looking from this position.
1387 * @param $needle The character/string to look for.
1388 * @return location of the next occurance of the needle, or
1389 * strlen($body) if needle wasn't found.
1391 function sq_findnxstr($body, $offset, $needle){
1392 $me = 'sq_findnxstr';
1393 $pos = strpos($body, $needle, $offset);
1394 if ($pos === FALSE){
1395 $pos = strlen($body);
1401 * This function takes a PCRE-style regexp and tries to match it
1402 * within the string.
1404 * @param $body The string to look for needle in.
1405 * @param $offset Start looking from here.
1406 * @param $reg A PCRE-style regex to match.
1407 * @return Returns a false if no matches found, or an array
1408 * with the following members:
1409 * - integer with the location of the match within $body
1410 * - string with whatever content between offset and the match
1411 * - string with whatever it is we matched
1413 function sq_findnxreg($body, $offset, $reg){
1414 $me = 'sq_findnxreg';
1417 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1418 if (!isset($matches{0}) ||
!$matches{0}){
1421 $retarr{0} = $offset +
strlen($matches{1});
1422 $retarr{1} = $matches{1};
1423 $retarr{2} = $matches{2};
1429 * This function looks for the next tag.
1431 * @param $body String where to look for the next tag.
1432 * @param $offset Start looking from here.
1433 * @return false if no more tags exist in the body, or
1434 * an array with the following members:
1435 * - string with the name of the tag
1436 * - array with attributes and their values
1437 * - integer with tag type (1, 2, or 3)
1438 * - integer where the tag starts (starting "<")
1439 * - integer where the tag ends (ending ">")
1440 * first three members will be false, if the tag is invalid.
1442 function sq_getnxtag($body, $offset){
1443 $me = 'sq_getnxtag';
1444 if ($offset > strlen($body)){
1447 $lt = sq_findnxstr($body, $offset, "<");
1448 if ($lt == strlen($body)){
1453 * blah blah <tag attribute="value">
1456 $pos = sq_skipspace($body, $lt+
1);
1457 if ($pos >= strlen($body)){
1458 return Array(false, false, false, $lt, strlen($body));
1461 * There are 3 kinds of tags:
1462 * 1. Opening tag, e.g.:
1464 * 2. Closing tag, e.g.:
1466 * 3. XHTML-style content-less tag, e.g.:
1467 * <img src="blah" />
1470 switch (substr($body, $pos, 1)){
1477 * A comment or an SGML declaration.
1479 if (substr($body, $pos+
1, 2) == "--"){
1480 $gt = strpos($body, "-->", $pos);
1482 $gt = strlen($body);
1486 return Array(false, false, false, $lt, $gt);
1488 $gt = sq_findnxstr($body, $pos, ">");
1489 return Array(false, false, false, $lt, $gt);
1494 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1504 * Look for next [\W-_], which will indicate the end of the tag name.
1506 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1507 if ($regary == false){
1508 return Array(false, false, false, $lt, strlen($body));
1510 list($pos, $tagname, $match) = $regary;
1511 $tagname = strtolower($tagname);
1514 * $match can be either of these:
1515 * '>' indicating the end of the tag entirely.
1516 * '\s' indicating the end of the tag name.
1517 * '/' indicating that this is type-3 xhtml tag.
1519 * Whatever else we find there indicates an invalid tag.
1524 * This is an xhtml-style tag with a closing / at the
1525 * end, like so: <img src="blah" />. Check if it's followed
1526 * by the closing bracket. If not, then this tag is invalid
1528 if (substr($body, $pos, 2) == "/>"){
1532 $gt = sq_findnxstr($body, $pos, ">");
1533 $retary = Array(false, false, false, $lt, $gt);
1537 return Array($tagname, false, $tagtype, $lt, $pos);
1541 * Check if it's whitespace
1543 if (!preg_match('/\s/', $match)){
1545 * This is an invalid tag! Look for the next closing ">".
1547 $gt = sq_findnxstr($body, $lt, ">");
1548 return Array(false, false, false, $lt, $gt);
1554 * At this point we're here:
1555 * <tagname attribute='blah'>
1558 * At this point we loop in order to find all attributes.
1564 while ($pos <= strlen($body)){
1565 $pos = sq_skipspace($body, $pos);
1566 if ($pos == strlen($body)){
1570 return Array(false, false, false, $lt, $pos);
1573 * See if we arrived at a ">" or "/>", which means that we reached
1574 * the end of the tag.
1577 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
1581 $pos +
= strlen($matches{1});
1582 if ($matches{2} == "/>"){
1586 return Array($tagname, $attary, $tagtype, $lt, $pos);
1590 * There are several types of attributes, with optional
1591 * [:space:] between members.
1593 * attrname[:space:]=[:space:]'CDATA'
1595 * attrname[:space:]=[:space:]"CDATA"
1597 * attr[:space:]=[:space:]CDATA
1601 * We leave types 1 and 2 the same, type 3 we check for
1602 * '"' and convert to """ if needed, then wrap in
1603 * double quotes. Type 4 we convert into:
1606 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
1607 if ($regary == false){
1609 * Looks like body ended before the end of tag.
1611 return Array(false, false, false, $lt, strlen($body));
1613 list($pos, $attname, $match) = $regary;
1614 $attname = strtolower($attname);
1616 * We arrived at the end of attribute name. Several things possible
1618 * '>' means the end of the tag and this is attribute type 4
1619 * '/' if followed by '>' means the same thing as above
1620 * '\s' means a lot of things -- look what it's followed by.
1621 * anything else means the attribute is invalid.
1626 * This is an xhtml-style tag with a closing / at the
1627 * end, like so: <img src="blah" />. Check if it's followed
1628 * by the closing bracket. If not, then this tag is invalid
1630 if (substr($body, $pos, 2) == "/>"){
1634 $gt = sq_findnxstr($body, $pos, ">");
1635 $retary = Array(false, false, false, $lt, $gt);
1639 $attary{$attname} = '"yes"';
1640 return Array($tagname, $attary, $tagtype, $lt, $pos);
1644 * Skip whitespace and see what we arrive at.
1646 $pos = sq_skipspace($body, $pos);
1647 $char = substr($body, $pos, 1);
1649 * Two things are valid here:
1650 * '=' means this is attribute type 1 2 or 3.
1651 * \w means this was attribute type 4.
1652 * anything else we ignore and re-loop. End of tag and
1653 * invalid stuff will be caught by our checks at the beginning
1658 $pos = sq_skipspace($body, $pos);
1660 * Here are 3 possibilities:
1661 * "'" attribute type 1
1662 * '"' attribute type 2
1663 * everything else is the content of tag type 3
1665 $quot = substr($body, $pos, 1);
1667 $regary = sq_findnxreg($body, $pos+
1, "\'");
1668 if ($regary == false){
1669 return Array(false, false, false, $lt, strlen($body));
1671 list($pos, $attval, $match) = $regary;
1673 $attary{$attname} = "'" . $attval . "'";
1674 } else if ($quot == '"'){
1675 $regary = sq_findnxreg($body, $pos+
1, '\"');
1676 if ($regary == false){
1677 return Array(false, false, false, $lt, strlen($body));
1679 list($pos, $attval, $match) = $regary;
1681 $attary{$attname} = '"' . $attval . '"';
1684 * These are hateful. Look for \s, or >.
1686 $regary = sq_findnxreg($body, $pos, "[\s>]");
1687 if ($regary == false){
1688 return Array(false, false, false, $lt, strlen($body));
1690 list($pos, $attval, $match) = $regary;
1692 * If it's ">" it will be caught at the top.
1694 $attval = preg_replace("/\"/s", """, $attval);
1695 $attary{$attname} = '"' . $attval . '"';
1697 } else if (preg_match("|[\w/>]|", $char)) {
1699 * That was attribute type 4.
1701 $attary{$attname} = '"yes"';
1704 * An illegal character. Find next '>' and return.
1706 $gt = sq_findnxstr($body, $pos, ">");
1707 return Array(false, false, false, $lt, $gt);
1713 * The fact that we got here indicates that the tag end was never
1714 * found. Return invalid tag indication so it gets stripped.
1716 return Array(false, false, false, $lt, strlen($body));
1720 * Translates entities into literal values so they can be checked.
1722 * @param $attvalue the by-ref value to check.
1723 * @param $regex the regular expression to check against.
1724 * @param $hex whether the entites are hexadecimal.
1725 * @return True or False depending on whether there were matches.
1727 function sq_deent(&$attvalue, $regex, $hex=false){
1731 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
1732 preg_match_all($regex, $attvalue, $matches);
1733 if (is_array($matches) && sizeof($matches[0]) > 0){
1735 for ($i = 0; $i < sizeof($matches[0]); $i++
){
1736 $numval = $matches[1][$i];
1738 $numval = hexdec($numval);
1740 $repl{$matches[0][$i]} = chr($numval);
1742 $attvalue = strtr($attvalue, $repl);
1750 * This function runs various checks against the attributes.
1752 * @param $tagname String with the name of the tag.
1753 * @param $attary Array with all tag attributes.
1754 * @param $rm_attnames See description for sq_sanitize
1755 * @param $bad_attvals See description for sq_sanitize
1756 * @param $add_attr_to_tag See description for sq_sanitize
1757 * @param $message message object
1758 * @param $id message id
1759 * @return Array with modified attributes.
1761 function sq_fixatts($tagname,
1771 while (list($attname, $attvalue) = each($attary)){
1773 * See if this attribute should be removed.
1775 foreach ($rm_attnames as $matchtag=>$matchattrs){
1776 if (preg_match($matchtag, $tagname)){
1777 foreach ($matchattrs as $matchattr){
1778 if (preg_match($matchattr, $attname)){
1779 unset($attary{$attname});
1786 * Workaround for IE quirks
1788 sq_fixIE_idiocy($attvalue);
1791 * Remove any backslashes, entities, and extraneous whitespace.
1794 $oldattvalue = $attvalue;
1795 sq_defang($attvalue);
1796 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1797 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1798 // i.e.<div style="{ left:expʀessioɴ( alert('XSS') ) }">
1799 $attvalue = "idiocy";
1800 $attary{$attname} = $attvalue;
1802 sq_unspace($attvalue);
1805 * Now let's run checks on the attvalues.
1806 * I don't expect anyone to comprehend this. If you do,
1807 * get in touch with me so I can drive to where you live and
1808 * shake your hand personally. :)
1810 foreach ($bad_attvals as $matchtag=>$matchattrs){
1811 if (preg_match($matchtag, $tagname)){
1812 foreach ($matchattrs as $matchattr=>$valary){
1813 if (preg_match($matchattr, $attname)){
1815 * There are two arrays in valary.
1817 * Second one is replacements
1819 list($valmatch, $valrepl) = $valary;
1821 preg_replace($valmatch, $valrepl, $attvalue);
1822 if ($newvalue != $attvalue){
1823 $attary{$attname} = $newvalue;
1824 $attvalue = $newvalue;
1830 if ($attname == 'style') {
1831 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1832 // 8bit and control characters in style attribute values can be used for XSS, remove them
1833 $attary{$attname} = '"disallowed character"';
1835 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1836 if (count($aMatch)) {
1837 foreach($aMatch[1] as $sMatch) {
1839 $urlvalue = $sMatch;
1840 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1841 $attary{$attname} = str_replace($sMatch,$urlvalue,$attvalue);
1846 * Use white list based filtering on attributes which can contain url's
1848 else if ($attname == 'href' ||
$attname == 'src' ||
$attname == 'background') {
1849 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1850 $attary{$attname} = $attvalue;
1854 * See if we need to append any attributes to this tag.
1856 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1857 if (preg_match($matchtag, $tagname)){
1858 $attary = array_merge($attary, $addattary);
1865 * This function filters url's
1867 * @param $attvalue String with attribute value to filter
1868 * @param $message message object
1869 * @param $id message id
1870 * @param $mailbox mailbox
1871 * @param $sQuote quoting characters around url's
1873 function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1874 $attvalue = trim($attvalue);
1875 if ($attvalue && ($attvalue[0] =='"'||
$attvalue[0] == "'")) {
1876 // remove the double quotes
1877 $sQuote = $attvalue[0];
1878 $attvalue = trim(substr($attvalue,1,-1));
1881 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1882 // images off by default.
1883 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
1885 $secremoveimg = '../images/' . _("sec_remove_eng.png");
1888 * Replace empty src tags with the blank image. src is only used
1889 * for frames, images, and image inputs. Doing a replace should
1890 * not affect them working as should be, however it will stop
1891 * IE from being kicked off when src for img tags are not set
1893 if ($attvalue == '') {
1894 $attvalue = '"' . SM_PATH
. 'images/blank.png"';
1896 // first, disallow 8 bit characters and control characters
1897 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1900 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1903 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1907 $aUrl = parse_url($attvalue);
1908 if (isset($aUrl['scheme'])) {
1909 switch(strtolower($aUrl['scheme'])) {
1914 if ($attname != 'href') {
1915 if ($view_unsafe_images == false) {
1916 $attvalue = $sQuote . $secremoveimg . $sQuote;
1918 if (isset($aUrl['path'])) {
1920 // No one has been able to show that image URIs
1921 // can be exploited, so for now, no restrictions
1922 // are made at all. If this proves to be a problem,
1923 // the commented-out code below can be of help.
1924 // (One consideration is that I see nothing in this
1925 // function that specifically says that we will
1926 // only ever arrive here when inspecting an image
1927 // tag, although that does seem to be the end
1928 // result - e.g., <script src="..."> where malicious
1929 // image URIs are in fact a problem are already
1930 // filtered out elsewhere.
1931 /* ---------------------------------
1932 // validate image extension.
1933 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1934 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
1935 // If URI is to something other than
1936 // a regular image file, get the contents
1937 // and try to see if it is an image.
1938 // Don't use Fileinfo (finfo_file()) because
1939 // we'd need to make the admin configure the
1940 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1943 if (function_exists('mime_content_type')
1944 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
1948 $file_contents = '';
1949 while (!feof($FILE)) {
1950 $file_contents .= fread($FILE, 8192);
1954 // store file locally
1956 global $attachment_dir, $username;
1957 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
1958 $localfilename = GenerateRandomString(32, '', 7);
1959 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1960 while (file_exists($full_localfilename)) {
1961 $localfilename = GenerateRandomString(32, '', 7);
1962 $full_localfilename = "$hashed_attachment_dir/$localfilename";
1964 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
1965 fwrite($FILE, $file_contents);
1968 // get mime type and remove file
1970 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
1971 unlink("$hashed_attachment_dir/$localfilename");
1973 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
1974 if (substr(strtolower($mime_type), 0, 5) != 'image') {
1975 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1978 --------------------------------- */
1980 $attvalue = $sQuote . SM_PATH
. 'images/blank.png'. $sQuote;
1984 $attvalue = $sQuote . $attvalue . $sQuote;
1989 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
1990 * One day MS might actually make it match something useful, for now, falling
1991 * back to using cid2http, so we can grab the blank.png.
1993 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
1997 * Turn cid: urls into http-friendly ones.
1999 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
2002 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2006 if (!(isset($aUrl['path']) && $aUrl['path'] == $secremoveimg)) {
2007 // parse_url did not lead to satisfying result
2008 $attvalue = $sQuote . SM_PATH
. 'images/blank.png' . $sQuote;
2016 * This function edits the style definition to make them friendly and
2017 * usable in SquirrelMail.
2019 * @param $message the message object
2020 * @param $id the message id
2021 * @param $content a string with whatever is between <style> and </style>
2022 * @param $mailbox the message mailbox
2023 * @return a string with edited content.
2025 function sq_fixstyle($body, $pos, $message, $id, $mailbox){
2026 $me = 'sq_fixstyle';
2027 // workaround for </style> in between comments
2028 $iCurrentPos = $pos;
2033 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++
$i) {
2040 if ($sToken == '<') {
2050 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2055 $content .= $sToken;
2063 if ($sToken == '<') {
2065 if (isset($body{$i+
2}) && substr($body,$i,3) == '!--') {
2066 $i = strpos($body,'-->',$i+
3);
2067 if ($i === false) { // no end comment
2085 if ($bSucces == FALSE){
2086 return array(FALSE, strlen($body));
2092 * First look for general BODY style declaration, which would be
2094 * body {background: blah-blah}
2095 * and change it to .bodyclass so we can just assign it to a <div>
2097 $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
2098 $secremoveimg = '../images/' . _("sec_remove_eng.png");
2100 * Fix url('blah') declarations.
2102 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2103 // "url(\\1$secremoveimg\\2)", $content);
2105 // first check for 8bit sequences and disallowed control characters
2106 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2107 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2108 return array($content, $newpos);
2111 // IE Sucks hard. We have a special function for it.
2112 sq_fixIE_idiocy($content);
2114 // remove @import line
2115 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2117 // translate ur\l and variations (IE parses that)
2118 // TODO check if the sq_fixIE_idiocy function already handles this.
2119 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
2120 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2121 if (count($aMatch)) {
2122 $aValue = $aReplace = array();
2123 foreach($aMatch[1] as $sMatch) {
2125 $urlvalue = $sMatch;
2126 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2127 $aValue[] = $sMatch;
2128 $aReplace[] = $urlvalue;
2130 $content = str_replace($aValue,$aReplace,$content);
2134 * Remove any backslashes, entities, and extraneous whitespace.
2136 $contentTemp = $content;
2137 sq_defang($contentTemp);
2138 sq_unspace($contentTemp);
2141 * Fix stupid css declarations which lead to vulnerabilities
2144 $match = Array('/\/\*.*\*\//',
2148 '/include-source/i',
2151 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy');
2152 $contentNew = preg_replace($match, $replace, $contentTemp);
2153 if ($contentNew !== $contentTemp) {
2154 // insecure css declarations are used. From now on we don't care
2155 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2156 $content = $contentNew;
2158 return array($content, $newpos);
2163 * This function converts cid: url's into the ones that can be viewed in
2166 * @param $message the message object
2167 * @param $id the message id
2168 * @param $cidurl the cid: url.
2169 * @param $mailbox the message mailbox
2170 * @return a string with a http-friendly url
2172 function sq_cid2http($message, $id, $cidurl, $mailbox){
2174 * Get rid of quotes.
2176 $quotchar = substr($cidurl, 0, 1);
2177 if ($quotchar == '"' ||
$quotchar == "'"){
2178 $cidurl = str_replace($quotchar, "", $cidurl);
2182 $cidurl = substr(trim($cidurl), 4);
2184 $match_str = '/\{.*?\}\//';
2186 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2188 $linkurl = find_ent_id($cidurl, $message);
2189 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2190 unsafe link image */
2194 * This is part of a fix for Outlook Express 6.x generating
2195 * cid URLs without creating content-id headers. These images are
2196 * not part of the multipart/related html mail. The html contains
2197 * <img src="cid:{some_id}/image_filename.ext"> references to
2198 * attached images with as goal to render them inline although
2199 * the attachment disposition property is not inline.
2202 if (empty($linkurl)) {
2203 if (preg_match('/{.*}\//', $cidurl)) {
2204 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2205 if (!empty($cidurl)) {
2206 $linkurl = find_ent_id($cidurl, $message);
2211 if (!empty($linkurl)) {
2212 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&' .
2213 "passed_id=$id&mailbox=" . urlencode($mailbox) .
2214 '&ent_id=' . $linkurl . $quotchar;
2217 * If we couldn't generate a proper img url, drop in a blank image
2218 * instead of sending back empty, otherwise it causes unusual behaviour
2220 $httpurl = $quotchar . SM_PATH
. 'images/blank.png' . $quotchar;
2227 * This function changes the <body> tag into a <div> tag since we
2228 * can't really have a body-within-body.
2230 * @param $attary an array of attributes and values of <body>
2231 * @param $mailbox mailbox we're currently reading (for cid2http)
2232 * @param $message current message (for cid2http)
2233 * @param $id current message id (for cid2http)
2234 * @return a modified array of attributes to be set for <div>
2236 function sq_body2div($attary, $mailbox, $message, $id){
2237 $me = 'sq_body2div';
2238 $divattary = Array('class' => "'bodyclass'");
2240 $has_bgc_stl = $has_txt_stl = false;
2242 if (is_array($attary) && sizeof($attary) > 0){
2243 foreach ($attary as $attname=>$attvalue){
2244 $quotchar = substr($attvalue, 0, 1);
2245 $attvalue = str_replace($quotchar, "", $attvalue);
2248 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
2249 $styledef .= "background-image: url('$attvalue'); ";
2252 $has_bgc_stl = true;
2253 $styledef .= "background-color: $attvalue; ";
2256 $has_txt_stl = true;
2257 $styledef .= "color: $attvalue; ";
2261 // Outlook defines a white bgcolor and no text color. This can lead to
2262 // white text on a white bg with certain themes.
2263 if ($has_bgc_stl && !$has_txt_stl) {
2264 $styledef .= "color: $text; ";
2266 if (strlen($styledef) > 0){
2267 $divattary{"style"} = "\"$styledef\"";
2274 * This is the main function and the one you should actually be calling.
2275 * There are several variables you should be aware of an which need
2276 * special description.
2278 * Since the description is quite lengthy, see it here:
2279 * http://linux.duke.edu/projects/mini/htmlfilter/
2281 * @param $body the string with HTML you wish to filter
2282 * @param $tag_list see description above
2283 * @param $rm_tags_with_content see description above
2284 * @param $self_closing_tags see description above
2285 * @param $force_tag_closing see description above
2286 * @param $rm_attnames see description above
2287 * @param $bad_attvals see description above
2288 * @param $add_attr_to_tag see description above
2289 * @param $message message object
2290 * @param $id message id
2291 * @return sanitized html safe to show on your pages.
2293 function sq_sanitize($body,
2295 $rm_tags_with_content,
2305 $me = 'sq_sanitize';
2306 $rm_tags = array_shift($tag_list);
2308 * Normalize rm_tags and rm_tags_with_content.
2310 @array_walk
($tag_list, 'sq_casenormalize');
2311 @array_walk
($rm_tags_with_content, 'sq_casenormalize');
2312 @array_walk
($self_closing_tags, 'sq_casenormalize');
2314 * See if tag_list is of tags to remove or tags to allow.
2315 * false means remove these tags
2316 * true means allow these tags
2319 $open_tags = Array();
2320 $trusted = "\n<!-- begin sanitized html -->\n";
2321 $skip_content = false;
2323 * Take care of netscape's stupid javascript entities like
2326 $body = preg_replace("/&(\{.*?\};)/si", "&\\1", $body);
2328 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
2329 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
2330 $free_content = substr($body, $curpos, $lt-$curpos);
2332 * Take care of <style>
2334 if ($tagname == "style" && $tagtype == 1){
2335 list($free_content, $curpos) =
2336 sq_fixstyle($body, $gt+
1, $message, $id, $mailbox);
2337 if ($free_content != FALSE){
2338 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2339 $trusted .= $free_content;
2340 $trusted .= sq_tagprint($tagname, false, 2);
2344 if ($skip_content == false){
2345 $trusted .= $free_content;
2347 if ($tagname != FALSE){
2349 if ($skip_content == $tagname){
2351 * Got to the end of tag we needed to remove.
2354 $skip_content = false;
2356 if ($skip_content == false){
2357 if ($tagname == "body"){
2360 if (isset($open_tags{$tagname}) &&
2361 $open_tags{$tagname} > 0){
2362 $open_tags{$tagname}--;
2370 * $rm_tags_with_content
2372 if ($skip_content == false){
2374 * See if this is a self-closing type and change
2375 * tagtype appropriately.
2378 && in_array($tagname, $self_closing_tags)){
2382 * See if we should skip this tag and any content
2385 if ($tagtype == 1 &&
2386 in_array($tagname, $rm_tags_with_content)){
2387 $skip_content = $tagname;
2389 if (($rm_tags == false
2390 && in_array($tagname, $tag_list)) ||
2391 ($rm_tags == true &&
2392 !in_array($tagname, $tag_list))){
2396 * Convert body into div.
2398 if ($tagname == "body"){
2400 $attary = sq_body2div($attary, $mailbox,
2404 if (isset($open_tags{$tagname})){
2405 $open_tags{$tagname}++
;
2407 $open_tags{$tagname}=1;
2411 * This is where we run other checks.
2413 if (is_array($attary) && sizeof($attary) > 0){
2414 $attary = sq_fixatts($tagname,
2428 if ($tagname != false && $skip_content == false){
2429 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2434 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2435 if ($force_tag_closing == true){
2436 foreach ($open_tags as $tagname=>$opentimes){
2437 while ($opentimes > 0){
2438 $trusted .= '</' . $tagname . '>';
2444 $trusted .= "<!-- end sanitized html -->\n";
2449 * This is a wrapper function to call html sanitizing routines.
2451 * @param $body the body of the message
2452 * @param $id the id of the message
2456 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2457 * into internal SM compose links
2458 * (optional; default = TRUE)
2459 * @return a string with html safe to display in the browser.
2461 function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2463 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
2465 global $attachment_common_show_images, $view_unsafe_images,
2468 * Don't display attached images in HTML mode.
2472 $attachment_common_show_images = false;
2487 $rm_tags_with_content = Array(
2497 $self_closing_tags = Array(
2505 $force_tag_closing = true;
2507 $rm_attnames = Array(
2518 $secremoveimg = "../images/" . _("sec_remove_eng.png");
2519 $bad_attvals = Array(
2522 "/^src|background/i" =>
2525 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2526 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2527 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2530 "\\1$secremoveimg\\2",
2531 "\\1$secremoveimg\\2",
2532 "\\1$secremoveimg\\2",
2535 "/^href|action/i" =>
2538 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2539 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2540 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
2555 "/include-source/i",
2556 "/position\s*:\s*absolute/i",
2557 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
2558 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2559 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2560 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
2561 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si"
2580 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2581 // images off by default.
2582 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET
, FALSE);
2584 if (!$view_unsafe_images){
2586 * Remove any references to http/https if view_unsafe_images set
2589 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[0],
2590 '/^([\'\"])\s*https*:.*([\'\"])/si');
2591 array_push($bad_attvals{'/.*/'}{'/^src|background/i'}[1],
2592 "\\1$secremoveimg\\1");
2593 array_push($bad_attvals{'/.*/'}{'/^style/i'}[0],
2594 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
2595 array_push($bad_attvals{'/.*/'}{'/^style/i'}[1],
2596 "url(\\1$secremoveimg\\1)");
2599 $add_attr_to_tag = Array(
2601 Array('target'=>'"_blank"',
2602 'title'=>'"'._("This external link will open in a new window").'"'
2605 $trusted = sq_sanitize($body,
2607 $rm_tags_with_content,
2617 if (strpos($trusted,$secremoveimg)){
2618 $has_unsafe_images = true;
2621 // we want to parse mailto's in HTML output, change to SM compose links
2622 // this is a modified version of code from url_parser.php... but Marc is
2623 // right: we need a better filtering implementation; adding this randomly
2624 // here is not a great solution
2626 if ($take_mailto_links) {
2627 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2628 global $MailTo_PReg_Match;
2629 $MailTo_PReg_Match = '/mailto:' . substr($MailTo_PReg_Match, 1) ;
2630 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2631 foreach ($regs[0] as $i => $mailto_before) {
2632 $mailto_params = $regs[10][$i];
2633 // get rid of any tailing quote since we have to add send_to to the end
2635 if (substr($mailto_before, strlen($mailto_before) - 1) == '"')
2636 $mailto_before = substr($mailto_before, 0, strlen($mailto_before) - 1);
2637 if (substr($mailto_params, strlen($mailto_params) - 1) == '"')
2638 $mailto_params = substr($mailto_params, 0, strlen($mailto_params) - 1);
2640 if ($regs[1][$i]) { //if there is an email addr before '?', we need to merge it with the params
2641 $to = 'to=' . $regs[1][$i];
2642 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2643 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2645 if ($mailto_params) //already some params, append to them
2646 $mailto_params .= '&' . $to;
2648 $mailto_params .= '?' . $to;
2652 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2654 // we'll already have target=_blank, no need to allow comp_in_new
2655 // here (which would be a lot more work anyway)
2657 global $compose_new_win;
2658 $temp_comp_in_new = $compose_new_win;
2659 $compose_new_win = 0;
2660 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2661 $compose_new_win = $temp_comp_in_new;
2663 // remove <a href=" and anything after the next quote (we only
2664 // need the uri, not the link HTML) in compose uri
2666 $comp_uri = substr($comp_uri, 9);
2667 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2668 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2677 * function SendDownloadHeaders - send file to the browser
2679 * Original Source: SM core src/download.php
2680 * moved here to make it available to other code, and separate
2681 * front end from back end functionality.
2683 * @param string $type0 first half of mime type
2684 * @param string $type1 second half of mime type
2685 * @param string $filename filename to tell the browser for downloaded file
2686 * @param boolean $force whether to force the download dialog to pop
2687 * @param optional integer $filesize send the Content-Header and length to the browser
2690 function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2691 global $languages, $squirrelmail_language;
2692 $isIE = $isIE6plus = false;
2694 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER
);
2696 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
2697 strstr($HTTP_USER_AGENT, 'Opera') === false) {
2701 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2702 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2706 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
2707 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
2709 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
2711 $filename = ereg_replace('[\\/:\*\?"<>\|;]', '_', str_replace(' ', ' ', $filename));
2714 // A Pox on Microsoft and it's Internet Explorer!
2716 // IE has lots of bugs with file downloads.
2717 // It also has problems with SSL. Both of these cause problems
2718 // for us in this function.
2720 // See this article on Cache Control headers and SSL
2721 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2723 // The best thing you can do for IE is to upgrade to the latest
2725 //set all the Cache Control Headers for IE
2727 $filename=rawurlencode($filename);
2728 header ("Pragma: public");
2729 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
2730 header ("Cache-Control: post-check=0, pre-check=0", false);
2731 header ("Cache-Control: private");
2733 //set the inline header for IE, we'll add the attachment header later if we need it
2734 header ("Content-Disposition: inline; filename=$filename");
2738 // Try to show in browser window
2739 header ("Content-Disposition: inline; filename=\"$filename\"");
2740 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2742 // Try to pop up the "save as" box
2744 // IE makes this hard. It pops up 2 save boxes, or none.
2745 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2746 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2747 // But, according to Microsoft, it is "RFC compliant but doesn't
2748 // take into account some deviations that allowed within the
2749 // specification." Doesn't that mean RFC non-compliant?
2750 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2752 // all browsers need the application/octet-stream header for this
2753 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2755 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2756 // Do not have quotes around filename, but that applied to
2757 // "attachment"... does it apply to inline too?
2758 header ("Content-Disposition: attachment; filename=\"$filename\"");
2760 if ($isIE && !$isIE6plus) {
2761 // This combination seems to work mostly. IE 5.5 SP 1 has
2762 // known issues (see the Microsoft Knowledge Base)
2764 // This works for most types, but doesn't work with Word files
2765 header ("Content-Type: application/download; name=\"$filename\"");
2766 header ("Content-Type: application/force-download; name=\"$filename\"");
2767 // These are spares, just in case. :-)
2768 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2769 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2770 //header("Content-Type: application/octet-stream; name=\"$filename\"");
2772 // This is to prevent IE for MIME sniffing and auto open a file in IE
2773 header ("Content-Type: application/force-download; name=\"$filename\"");
2775 // another application/octet-stream forces download for Netscape
2776 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2780 //send the content-length header if the calling function provides it
2781 if ($filesize > 0) {
2782 header("Content-Length: $filesize");
2785 } // end fn SendDownloadHeaders