projects / squirrelmail.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
Add proper RFC 3461 DSN functionality (previously we relied only on the Return-Receip...
[squirrelmail.git] / functions / mime.php
CommitLineData
59177427 1<?php
2ba13803 2
35586184 3/**
8bd0068d 4 * mime.php
5 *
8bd0068d 6 * This contains the functions necessary to detect and decode MIME
7 * messages.
8 *
77a1e3d1 9 * @copyright 1999-2022 The SquirrelMail Project Team
4b4abf93 10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
8bd0068d 11 * @version $Id$
12 * @package squirrelmail
13 */
b74ba498 14
202bcbcc 15/**
16 * dependency information
17 functions dependency
18 mime_structure
19 class/mime/Message.class.php
20 Message::parseStructure
21 functions/page_header.php
22 displayPageHeader
23 functions/display_messages.php
24 plain_error_message
25 mime_fetch_body
26 functions/imap_general.php
27 sqimap_run_command
28 mime_print_body_lines
29
30
31
32functions/imap.php
33functions/attachment_common.php
34functions/display_messages.php
35
36magicHtml => url_parser
37translateText => url_parser
38
39*/
40
8beafbbc 41
7c7b74b3 42/* -------------------------------------------------------------------------- */
43/* MIME DECODING */
44/* -------------------------------------------------------------------------- */
b74ba498 45
d6c32258 46/**
8bd0068d 47 * Get the MIME structure
48 *
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.
52 */
a4a70693 53function mime_structure ($bodystructure, $flags=array()) {
c9d78ab4 54
3d8371be 55 /* Isolate the body structure and remove beginning and end parenthesis. */
a4a70693 56 $read = trim(substr ($bodystructure, strpos(strtolower($bodystructure), 'bodystructure') + 13));
451f74a2 57 $read = trim(substr ($read, 0, -1));
22efa9fb 58 $i = 0;
59 $msg = Message::parseStructure($read,$i);
2b665f28 60
9de42168 61 if (!is_object($msg)) {
3d8371be 62 global $color, $mailbox;
a48eba8f 63 displayPageHeader( $color, $mailbox );
5e8de8b6 64 $errormessage = _("SquirrelMail could not decode the bodystructure of the message");
8bd0068d 65 $errormessage .= '<br />'._("The bodystructure provided by your IMAP server:").'<br /><br />';
3047e291 66 $errormessage .= '<pre>' . sm_encode_html_special_chars($read) . '</pre>';
ce8c6f42 67 plain_error_message( $errormessage );
3d8371be 68 echo '</body></html>';
9de42168 69 exit;
70 }
a4a70693 71 if (count($flags)) {
7a9e9c89 72 foreach ($flags as $flag) {
e1115979 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.
1710ad65 74 $char = strtoupper($flag[1]);
7a9e9c89 75 switch ($char) {
3d8371be 76 case 'S':
77 if (strtolower($flag) == '\\seen') {
78 $msg->is_seen = true;
79 }
80 break;
81 case 'A':
82 if (strtolower($flag) == '\\answered') {
83 $msg->is_answered = true;
84 }
85 break;
86 case 'D':
87 if (strtolower($flag) == '\\deleted') {
88 $msg->is_deleted = true;
89 }
90 break;
91 case 'F':
92 if (strtolower($flag) == '\\flagged') {
93 $msg->is_flagged = true;
94 }
1a753239 95 else if (strtolower($flag) == '$forwarded') {
96 $msg->is_forwarded = true;
97 }
3d8371be 98 break;
99 case 'M':
100 if (strtolower($flag) == '$mdnsent') {
101 $msg->is_mdnsent = true;
102 }
103 break;
104 default:
105 break;
7a9e9c89 106 }
107 }
451f74a2 108 }
7a9e9c89 109 // listEntities($msg);
3d8371be 110 return $msg;
451f74a2 111}
b74ba498 112
22efa9fb 113
114
3d8371be 115/* This starts the parsing of a particular structure. It is called recursively,
8bd0068d 116 * so it can be passed different structures. It returns an object of type
117 * $message.
118 * First, it checks to see if it is a multipart message. If it is, then it
119 * handles that as it sees is necessary. If it is just a regular entity,
120 * then it parses it and adds the necessary header information (by calling out
121 * to mime_get_elements()
122 */
451f74a2 123
4d592352 124function mime_fetch_body($imap_stream, $id, $ent_id=1, $fetch_size=0) {
3d8371be 125 /* Do a bit of error correction. If we couldn't find the entity id, just guess
8bd0068d 126 * that it is the first one. That is usually the case anyway.
127 */
7c7b74b3 128
09a4bde3 129 if (!$ent_id) {
08b7f7cc 130 $cmd = "FETCH $id BODY[]";
1035e159 131 } else {
08b7f7cc 132 $cmd = "FETCH $id BODY[$ent_id]";
09a4bde3 133 }
3d8371be 134
4d592352 135 if ($fetch_size!=0) $cmd .= "<0.$fetch_size>";
da2415c1 136
6201339c 137 $data = sqimap_run_command ($imap_stream, $cmd, true, $response, $message, TRUE);
77b88425 138 do {
3d8371be 139 $topline = trim(array_shift($data));
8439f61d 140 } while($topline && ($topline[0] == '*') && !preg_match('/\* [0-9]+ FETCH .*BODY.*/i', $topline)) ;
141 // Matching with "BODY" above is difficult: in most cases "FETCH \(BODY" would work
142 // but some servers may put other things in the same result, perhaps something such
143 // as "* 23 FETCH (FLAGS (\Seen) BODY[1] {174}". There is some small chance that
144 // if the character sequence "BODY" appears in a response where it isn't actually
145 // a FETCH response data item name, the current regex will break things. The better
146 // way to do this would be to parse the response correctly and not use a regex.
a4a70693 147
451f74a2 148 $wholemessage = implode('', $data);
b7910e12 149 if (preg_match('/\{([^\}]*)\}/', $topline, $regs)) {
3d8371be 150 $ret = substr($wholemessage, 0, $regs[1]);
151 /* There is some information in the content info header that could be important
8bd0068d 152 * in order to parse html messages. Let's get them here.
153 */
1710ad65 154// if ($ret[0] == '<') {
6201339c 155// $data = sqimap_run_command ($imap_stream, "FETCH $id BODY[$ent_id.MIME]", true, $response, $message, TRUE);
0600bdf1 156// }
b7910e12 157 } else if (preg_match('/"([^"]*)"/', $topline, $regs)) {
451f74a2 158 $ret = $regs[1];
06bcb9c3 159 } else if ((stristr($topline, 'nil') !== false) && (empty($wholemessage))) {
160 $ret = $wholemessage;
451f74a2 161 } else {
162 global $where, $what, $mailbox, $passed_id, $startMessage;
3d8371be 163 $par = 'mailbox=' . urlencode($mailbox) . '&amp;passed_id=' . $passed_id;
451f74a2 164 if (isset($where) && isset($what)) {
3d8371be 165 $par .= '&amp;where=' . urlencode($where) . '&amp;what=' . urlencode($what);
a3daaaf3 166 } else {
3d8371be 167 $par .= '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
451f74a2 168 }
e5ea9327 169 $par .= '&amp;response=' . urlencode($response) .
8bd0068d 170 '&amp;message=' . urlencode($message) .
171 '&amp;topline=' . urlencode($topline);
a019eeb8 172
8bd0068d 173 echo '<tt><br />' .
02474e43 174 '<table width="80%"><tr>' .
175 '<tr><td colspan="2">' .
176 _("Body retrieval error. The reason for this is most probably that the message is malformed.") .
177 '</td></tr>' .
178 '<tr><td><b>' . _("Command:") . "</td><td>$cmd</td></tr>" .
179 '<tr><td><b>' . _("Response:") . "</td><td>$response</td></tr>" .
180 '<tr><td><b>' . _("Message:") . "</td><td>$message</td></tr>" .
181 '<tr><td><b>' . _("FETCH line:") . "</td><td>$topline</td></tr>" .
182 "</table><br /></tt></font><hr />";
346817d4 183
6201339c 184 $data = sqimap_run_command ($imap_stream, "FETCH $passed_id BODY[]", true, $response, $message, TRUE);
451f74a2 185 array_shift($data);
186 $wholemessage = implode('', $data);
a019eeb8 187
346817d4 188 $ret = $wholemessage;
a3daaaf3 189 }
3d8371be 190 return $ret;
451f74a2 191}
d4467150 192
c53195bb 193// TODO: Needs documentation. $ent_id default is usually 1
194function mime_print_body_lines ($imap_stream, $id, $ent_id, $encoding, $rStream='php://stdout', $force_crlf='') {
1035e159 195
3d8371be 196 /* Don't kill the connection if the browser is over a dialup
8bd0068d 197 * and it would take over 30 seconds to download it.
198 * Don't call set_time_limit in safe mode.
199 */
b7206e1d 200
3d8371be 201 if (!ini_get('safe_mode')) {
b7206e1d 202 set_time_limit(0);
203 }
7c7b74b3 204 /* in case of base64 encoded attachments, do not buffer them.
8bd0068d 205 Instead, echo the decoded attachment directly to screen */
7c7b74b3 206 if (strtolower($encoding) == 'base64') {
207 if (!$ent_id) {
7591f143 208 $query = "FETCH $id BODY[]";
7c7b74b3 209 } else {
7591f143 210 $query = "FETCH $id BODY[$ent_id]";
7c7b74b3 211 }
7591f143 212 sqimap_run_command($imap_stream,$query,true,$response,$message,TRUE,'sqimap_base64_decode',$rStream,true);
1d142b8d 213 } else {
7591f143 214 $body = mime_fetch_body ($imap_stream, $id, $ent_id);
215 if (is_resource($rStream)) {
fa26ab45 216 fputs($rStream,decodeBody($body, $encoding, $force_crlf));
7591f143 217 } else {
fa26ab45 218 echo decodeBody($body, $encoding, $force_crlf);
7591f143 219 }
1d142b8d 220 }
346817d4 221
da2415c1 222 /*
8bd0068d 223 TODO, use the same method for quoted printable.
224 However, I assume that quoted printable attachments aren't that large
225 so the performancegain / memory usage drop will be minimal.
226 If we decide to add that then we need to adapt sqimap_fread because
227 we need to split te result on \n and fread doesn't stop at \n. That
228 means we also should provide $results from sqimap_fread (by ref) to
229 te function and set $no_return to false. The $filter function for
230 quoted printable should handle unsetting of $results.
231 */
da2415c1 232 /*
8bd0068d 233 TODO 2: find out how we write to the output stream php://stdout. fwrite
234 doesn't work because 'php://stdout isn't a stream.
235 */
7c7b74b3 236
5d9c6f73 237 return;
451f74a2 238}
beb9e459 239
451f74a2 240/* -[ END MIME DECODING ]----------------------------------------------------------- */
d4467150 241
3d8371be 242/* This is here for debugging purposes. It will print out a list
8bd0068d 243 * of all the entity IDs that are in the $message object.
244 */
451f74a2 245function listEntities ($message) {
3d8371be 246 if ($message) {
3c621ba1 247 echo "<tt>" . $message->entity_id . ' : ' . $message->type0 . '/' . $message->type1 . ' parent = '. $message->parent->entity_id. '<br />';
3d8371be 248 for ($i = 0; isset($message->entities[$i]); $i++) {
249 echo "$i : ";
250 $msg = listEntities($message->entities[$i]);
251
252 if ($msg) {
253 echo "return: ";
254 return $msg;
255 }
256 }
a4a70693 257 }
451f74a2 258}
f0c4dc12 259
f792c641 260function getPriorityStr($priority) {
3d8371be 261 $priority_level = substr($priority,0,1);
262
263 switch($priority_level) {
264 /* Check for a higher then normal priority. */
265 case '1':
266 case '2':
267 $priority_string = _("High");
268 break;
269
270 /* Check for a lower then normal priority. */
271 case '4':
272 case '5':
273 $priority_string = _("Low");
274 break;
275
276 /* Check for a normal priority. */
277 case '3':
278 default:
279 $priority_level = '3';
280 $priority_string = _("Normal");
281 break;
282
283 }
284 return $priority_string;
f792c641 285}
286
451f74a2 287/* returns a $message object for a particular entity id */
288function getEntity ($message, $ent_id) {
a4a70693 289 return $message->getEntity($ent_id);
451f74a2 290}
8beafbbc 291
3d8371be 292/* translateText
8bd0068d 293 * Extracted from strings.php 23/03/2002
294 */
da4c66e8 295
296function translateText(&$body, $wrap_at, $charset) {
3d8371be 297 global $where, $what; /* from searching */
298 global $color; /* color theme */
da4c66e8 299
202bcbcc 300 // require_once(SM_PATH . 'functions/url_parser.php');
da4c66e8 301
302 $body_ary = explode("\n", $body);
da4c66e8 303 for ($i=0; $i < count($body_ary); $i++) {
4c1f4be3 304 $line = rtrim($body_ary[$i],"\r");
305
da4c66e8 306 if (strlen($line) - 2 >= $wrap_at) {
c7aff938 307 sqWordWrap($line, $wrap_at, $charset);
da4c66e8 308 }
309 $line = charset_decode($charset, $line);
310 $line = str_replace("\t", ' ', $line);
311
312 parseUrl ($line);
313
3d8371be 314 $quotes = 0;
da4c66e8 315 $pos = 0;
3d8371be 316 $j = strlen($line);
da4c66e8 317
3d8371be 318 while ($pos < $j) {
da4c66e8 319 if ($line[$pos] == ' ') {
3d8371be 320 $pos++;
da4c66e8 321 } else if (strpos($line, '&gt;', $pos) === $pos) {
322 $pos += 4;
3d8371be 323 $quotes++;
da4c66e8 324 } else {
325 break;
326 }
327 }
3d8371be 328
83c94382 329 if ($quotes % 2) {
d0814c02 330 $line = '<span class="quote1">' . $line . '</span>';
4c25967c 331 } elseif ($quotes) {
d0814c02 332 $line = '<span class="quote2">' . $line . '</span>';
da4c66e8 333 }
3d8371be 334
da4c66e8 335 $body_ary[$i] = $line;
336 }
337 $body = '<pre>' . implode("\n", $body_ary) . '</pre>';
338}
339
a2bfcbce 340/**
da1b55ad 341 * This returns a parsed string called $body. That string can then
342 * be displayed as the actual message in the HTML. It contains
343 * everything needed, including HTML Tags, Attachments at the
344 * bottom, etc.
f8a1ed5a 345 *
da1b55ad 346 * Since 1.2.0 function uses message_body hook.
347 * Till 1.3.0 function included output of formatAttachments().
348 *
349 * @param resource $imap_stream imap connection resource
350 * @param object $message squirrelmail message object
351 * @param array $color squirrelmail color theme array
352 * @param integer $wrap_at number of characters per line
353 * @param string $ent_num (since 1.3.0) message part id
354 * @param integer $id (since 1.3.0) message id
355 * @param string $mailbox (since 1.3.0) imap folder name
da1b55ad 356 * @return string html formated message text
357 */
a540f994 358function formatBody($imap_stream, $message, $color, $wrap_at, $ent_num, $id, $mailbox='INBOX') {
3d8371be 359 /* This if statement checks for the entity to show as the
8bd0068d 360 * primary message. To add more of them, just put them in the
361 * order that is their priority.
362 */
ce68b76b 363 global $startMessage, $languages, $squirrelmail_language,
40a34e57 364 $show_html_default, $sort, $has_unsafe_images, $passed_ent_id,
14c85e39 365 $use_iframe, $iframe_height, $download_and_unsafe_link,
955bfc8f 366 $download_href, $unsafe_image_toggle_href, $unsafe_image_toggle_text,
551c7b53 367 $oTemplate, $nbsp;
2c25d36a 368
369 // workaround for not updated config.php
370 if (! isset($use_iframe)) $use_iframe = false;
77bfbd2e 371
d31f73f1 372 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
373 // images off by default.
afbf184c 374 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d03c24f4 375
cc34b00d 376 $body = '';
23bcec6f 377 $urlmailbox = urlencode($mailbox);
451f74a2 378 $body_message = getEntity($message, $ent_num);
379 if (($body_message->header->type0 == 'text') ||
8bd0068d 380 ($body_message->header->type0 == 'rfc822')) {
3d8371be 381 $body = mime_fetch_body ($imap_stream, $id, $ent_num);
451f74a2 382 $body = decodeBody($body, $body_message->header->encoding);
e842b215 383
384 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 385 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode')) {
e842b215 386 if (mb_detect_encoding($body) != 'ASCII') {
33a55f5a 387 $body = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decode',$body);
e842b215 388 }
389 }
d849b570 390
391 /* As of 1.5.2, $body is passed (and modified) by reference */
392 do_hook('message_body', $body);
23bcec6f 393
3d8371be 394 /* If there are other types that shouldn't be formatted, add
8bd0068d 395 * them here.
396 */
3d8371be 397
451f74a2 398 if ($body_message->header->type1 == 'html') {
3d8371be 399 if ($show_html_default <> 1) {
85015544 400 $entity_conv = array('&nbsp;' => ' ',
8bd0068d 401 '<p>' => "\n",
402 '<P>' => "\n",
403 '<br>' => "\n",
404 '<BR>' => "\n",
405 '<br />' => "\n",
406 '<BR />' => "\n",
407 '&gt;' => '>',
408 '&lt;' => '<');
85015544 409 $body = strtr($body, $entity_conv);
3d8371be 410 $body = strip_tags($body);
85015544 411 $body = trim($body);
412 translateText($body, $wrap_at,
8bd0068d 413 $body_message->header->getParameter('charset'));
a540f994 414 } elseif ($use_iframe) {
84410f31 415 /**
416 * If we don't add html message between iframe tags,
417 * we must detect unsafe images and modify $has_unsafe_images.
f8a1ed5a 418 */
758a7889 419 $html_body = magicHTML($body, $id, $message, $mailbox);
b6c52e61 420 // Convert character set in order to display html mails in different character set
421 $html_body = charset_decode($body_message->header->getParameter('charset'),$html_body,false,true);
84410f31 422
2c25d36a 423 // creating iframe url
424 $iframeurl=sqm_baseuri().'src/view_html.php?'
f8a1ed5a 425 . 'mailbox=' . $urlmailbox
2c25d36a 426 . '&amp;passed_id=' . $id
427 . '&amp;ent_id=' . $ent_num
428 . '&amp;view_unsafe_images=' . (int) $view_unsafe_images;
429
79d58d4c 430 global $oTemplate;
431 $oTemplate->assign('iframe_url', $iframeurl);
51554708 432 $oTemplate->assign('iframe_height', $iframe_height);
79d58d4c 433 $oTemplate->assign('html_body', $html_body);
2b665f28 434
79d58d4c 435 $body = $oTemplate->fetch('read_html_iframe.tpl');
a3daaaf3 436 } else {
2c25d36a 437 // old way of html rendering
b6c52e61 438 /**
758a7889 439 * convert character set. charset_decode does not remove html special chars
b6c52e61 440 * applied by magicHTML functions and does not sanitize them second time if
758a7889 441 * fourth argument is true.
442 */
567dc524 443 $charset = $body_message->header->getParameter('charset');
444 if (!empty($charset)) {
445 $body = charset_decode($charset,$body,false,true);
446 }
447 $body = magicHTML($body, $id, $message, $mailbox);
a3daaaf3 448 }
451f74a2 449 } else {
3d8371be 450 translateText($body, $wrap_at,
8bd0068d 451 $body_message->header->getParameter('charset'));
451f74a2 452 }
a2bfcbce 453
9ebace19 454 /*
455 * Previously the links for downloading and unsafe images were printed
456 * under the mail. By putting the links in a global variable we can
457 * print it in the toolbar where it belongs. Since the original code was
458 * in this place it's left here. It might be possible to move it to some
459 * other place if that makes sense. The possibility to do so has not
460 * been evaluated yet.
461 */
462
463 // Initialize the global variable to an empty string.
464 // 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.
40a34e57 465 $download_and_unsafe_link = '';
466
9ebace19 467 // Prepare and build a link for downloading the mail.
83cf04bd 468 $link = 'passed_id=' . $id . '&amp;ent_id='.$ent_num.
8bd0068d 469 '&amp;mailbox=' . $urlmailbox .'&amp;sort=' . $sort .
470 '&amp;startMessage=' . $startMessage . '&amp;show_more=0';
08b7f7cc 471 if (isset($passed_ent_id)) {
472 $link .= '&amp;passed_ent_id='.$passed_ent_id;
473 }
14c85e39 474 $download_href = SM_PATH . 'src/download.php?absolute_dl=true&amp;' . $link;
9ebace19 475
476 // Always add the link for downloading the mail as a file to the global
477 // variable.
955bfc8f 478 $download_and_unsafe_link .= "$nbsp|$nbsp"
479 . create_hyperlink($download_href, _("Download this as a file"));
9ebace19 480
481 // Find out the right text to use in the link depending on the
482 // circumstances. If the unsafe images are displayed the link should
483 // hide them, if they aren't displayed the link should only appear if
484 // the mail really contains unsafe images.
7aad7b77 485 if ($view_unsafe_images) {
23f617b8 486 $text = _("Hide Unsafe Images");
7aad7b77 487 } else {
08b7f7cc 488 if (isset($has_unsafe_images) && $has_unsafe_images) {
489 $link .= '&amp;view_unsafe_images=1';
490 $text = _("View Unsafe Images");
491 } else {
492 $text = '';
493 }
3d8371be 494 }
9ebace19 495
496 // Only create a link for unsafe images if there's need for one. If so:
497 // add it to the global variable.
83cf04bd 498 if($text != '') {
14c85e39 499 $unsafe_image_toggle_href = SM_PATH . 'src/read_body.php?'.$link;
500 $unsafe_image_toggle_text = $text;
955bfc8f 501 $download_and_unsafe_link .= "$nbsp|$nbsp"
502 . create_hyperlink($unsafe_image_toggle_href, $text);
83cf04bd 503 }
3d8371be 504 }
505 return $body;
451f74a2 506}
b74ba498 507
da1b55ad 508/**
a540f994 509 * Generate attachments array for passing to templates.
2b665f28 510 *
d67f519a 511 * @since 1.5.2
da1b55ad 512 * @param object $message SquirrelMail message object
513 * @param array $exclude_id message parts that are not attachments.
514 * @param string $mailbox mailbox name
515 * @param integer $id message id
da1b55ad 516 */
d67f519a 517function buildAttachmentArray($message, $exclude_id, $mailbox, $id) {
caf0ab1d 518 global $where, $what, $startMessage, $color, $passed_ent_id,
519 $base_uri, $block_svg_download;
451f74a2 520
23bcec6f 521 $att_ar = $message->getAttachments($exclude_id);
23bcec6f 522 $urlMailbox = urlencode($mailbox);
523
d67f519a 524 $attachments = array();
23bcec6f 525 foreach ($att_ar as $att) {
fdc9d9b5 526 $ent = $att->entity_id;
2e25760a 527 $header = $att->header;
f0c4dc12 528 $type0 = strtolower($header->type0);
529 $type1 = strtolower($header->type1);
caf0ab1d 530 if ($block_svg_download && strpos($type1, 'svg') === 0)
531 continue;
532
2e25760a 533 $name = '';
d0187bd6 534 $links = array();
21dab2dc 535 $links['download link']['text'] = _("Download");
202bcbcc 536 $links['download link']['href'] = $base_uri .
8bd0068d 537 "src/download.php?absolute_dl=true&amp;passed_id=$id&amp;mailbox=$urlMailbox&amp;ent_id=$ent";
2b665f28 538
2e25760a 539 if ($type0 =='message' && $type1 == 'rfc822') {
202bcbcc 540 $default_page = $base_uri . 'src/read_body.php';
2e25760a 541 $rfc822_header = $att->rfc822_header;
098ea084 542 $filename = $rfc822_header->subject;
6cc08d8b 543 if (trim( $filename ) == '') {
544 $filename = 'untitled-[' . $ent . ']' ;
08b7f7cc 545 }
2e25760a 546 $from_o = $rfc822_header->from;
547 if (is_object($from_o)) {
04ea844e 548 $from_name = decodeHeader($from_o->getAddress(false));
7e697748 549 } elseif (is_array($from_o) && count($from_o) && is_object($from_o[0])) {
550 // something weird happens when a digest message is opened and you return to the digest
551 // now the from object is part of an array. Probably the parseHeader call overwrites the info
552 // retrieved from the bodystructure in a different way. We need to fix this later.
553 // possible starting point, do not fetch header we already have and inspect how
554 // the rfc822_header object behaves.
555 $from_name = decodeHeader($from_o[0]->getAddress(false));
2e25760a 556 } else {
557 $from_name = _("Unknown sender");
f0c4dc12 558 }
d0187bd6 559 $description = _("From").': '.$from_name;
23bcec6f 560 } else {
202bcbcc 561 $default_page = $base_uri . 'src/download.php';
02474e43 562 $filename = $att->getFilename();
f810c0b2 563 if ($header->description) {
098ea084 564 $description = decodeHeader($header->description);
f810c0b2 565 } else {
3d8371be 566 $description = '';
567 }
2e25760a 568 }
569
570 $display_filename = $filename;
3d8371be 571 if (isset($passed_ent_id)) {
572 $passed_ent_id_link = '&amp;passed_ent_id='.$passed_ent_id;
573 } else {
574 $passed_ent_id_link = '';
575 }
576 $defaultlink = $default_page . "?startMessage=$startMessage"
8bd0068d 577 . "&amp;passed_id=$id&amp;mailbox=$urlMailbox"
578 . '&amp;ent_id='.$ent.$passed_ent_id_link;
2e25760a 579 if ($where && $what) {
8bd0068d 580 $defaultlink .= '&amp;where='. urlencode($where).'&amp;what='.urlencode($what);
2e25760a 581 }
7e2ff844 582 // IE does make use of mime content sniffing. Forcing a download
583 // prohibit execution of XSS inside an application/octet-stream attachment
584 if ($type0 == 'application' && $type1 == 'octet-stream') {
585 $defaultlink .= '&amp;absolute_dl=true';
586 }
2b665f28 587
3d8371be 588 /* This executes the attachment hook with a specific MIME-type.
53901c7b 589 * It also allows plugins to run if there's a rule for a more
590 * generic type. Finally, a hook for ALL attachment types is
591 * run as well.
8bd0068d 592 */
8dca4d22 593 // First remember the default link.
594 $defaultlink_orig = $defaultlink;
595
d849b570 596 /* The API for this hook has changed as of 1.5.2 so that all plugin
597 arguments are passed in an array instead of each their own plugin
598 argument, and arguments are passed by reference, so instead of
599 returning any changes, changes should simply be made to the original
600 arguments themselves. */
53901c7b 601 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
9f6cc07c 602 &$defaultlink, &$display_filename, &$where, &$what,
c2eacf5e 603 &$type0, &$type1);
9c3b2d22 604 do_hook("attachment $type0/$type1", $temp);
53901c7b 605 /* The API for this hook has changed as of 1.5.2 so that all plugin
606 arguments are passed in an array instead of each their own plugin
607 argument, and arguments are passed by reference, so instead of
608 returning any changes, changes should simply be made to the original
609 arguments themselves. */
610 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
c2eacf5e 611 &$defaultlink, &$display_filename, &$where, &$what,
612 &$type0, &$type1);
53901c7b 613 // Do not let a generic plugin change the default link if a more
614 // specialized one already did it...
615 if ($defaultlink != $defaultlink_orig) {
616 $dummy = '';
617 $temp[5] = &$dummy;
2e25760a 618 }
53901c7b 619 do_hook("attachment $type0/*", $temp);
d849b570 620 /* The API for this hook has changed as of 1.5.2 so that all plugin
621 arguments are passed in an array instead of each their own plugin
622 argument, and arguments are passed by reference, so instead of
623 returning any changes, changes should simply be made to the original
624 arguments themselves. */
9c3b2d22 625 $temp = array(&$links, &$startMessage, &$id, &$urlMailbox, &$ent,
c2eacf5e 626 &$defaultlink, &$display_filename, &$where, &$what,
627 &$type0, &$type1);
8dca4d22 628 // Do not let a generic plugin change the default link if a more
629 // specialized one already did it...
630 if ($defaultlink != $defaultlink_orig) {
631 $dummy = '';
632 $temp[5] = &$dummy;
633 }
9c3b2d22 634 do_hook("attachment */*", $temp);
77b88425 635
d67f519a 636 $this_attachment = array();
637 $this_attachment['Name'] = decodeHeader($display_filename);
638 $this_attachment['Description'] = $description;
639 $this_attachment['DefaultHREF'] = $defaultlink;
640 $this_attachment['DownloadHREF'] = $links['download link']['href'];
641 $this_attachment['ViewHREF'] = isset($links['attachment_common']) ? $links['attachment_common']['href'] : '';
24e8917e 642
643 // base64 encoded file sizes are misleading, so approximate real size
644 if (!empty($header->encoding) && strtolower($header->encoding) == 'base64')
645 $this_attachment['Size'] = $header->size / 4 * 3;
646 else
647 $this_attachment['Size'] = $header->size;
648
3047e291 649 $this_attachment['ContentType'] = sm_encode_html_special_chars($type0 .'/'. $type1);
d67f519a 650 $this_attachment['OtherLinks'] = array();
3d8371be 651 foreach ($links as $val) {
336995ea 652 if ($val['text']==_("Download")) {
653 $this_attachment['DownloadHREF'] = $val['href'];
d0187bd6 654 continue;
336995ea 655 }
656 if ($val['text']==_("View")) {
657 $this_attachment['ViewHREF'] = $val['href'];
658 continue;
659 }
4ed45b62 660
661 // This makes no sense - If 'text' and 'extra' are just concatenated,
662 // there is no point in having 'extra'.... I am going to assume this
663 // was a mistake and am changing 'extra' to be what I think it was
664 // meant to be: additional tag attributes. However, I'm not checking
665 // extensively for plugins that were using this the wrong way (but why would they?)
666 if (empty($val['text']))
d0187bd6 667 continue;
2b665f28 668
d67f519a 669 $temp = array();
670 $temp['HREF'] = $val['href'];
4ed45b62 671 $temp['Text'] = $val['text'];
672 $temp['Extra'] = (empty($val['extra']) ? '' : $val['extra']);
d67f519a 673 $this_attachment['OtherLinks'][] = $temp;
2e25760a 674 }
d67f519a 675 $attachments[] = $this_attachment;
2b665f28 676
3d8371be 677 unset($links);
2e25760a 678 }
2b665f28 679
d67f519a 680 return $attachments;
681}
682
683/**
684 * Displays attachment links and information
685 *
686 * Since 1.3.0 function is not included in formatBody() call.
687 *
688 * Since 1.0.2 uses attachment $type0/$type1 hook.
689 * Since 1.2.5 uses attachment $type0/* hook.
690 * Since 1.5.0 uses attachments_bottom hook.
691 * Since 1.5.2 uses templates and does *not* return a value.
692 *
693 * @param object $message SquirrelMail message object
694 * @param array $exclude_id message parts that are not attachments.
695 * @param string $mailbox mailbox name
696 * @param integer $id message id
697 */
698function formatAttachments($message, $exclude_id, $mailbox, $id) {
699 global $oTemplate;
2b665f28 700
d67f519a 701 $attach = buildAttachmentArray($message, $exclude_id, $mailbox, $id);
d0187bd6 702
703 $oTemplate->assign('attachments', $attach);
704 $oTemplate->display('read_attachments.tpl');
451f74a2 705}
b74ba498 706
7c7b74b3 707function sqimap_base64_decode(&$string) {
7c0ec1d8 708
b17a8968 709 // Base64 encoded data goes in pairs of 4 bytes. To achieve on the
7c0ec1d8 710 // fly decoding (to reduce memory usage) you have to check if the
711 // data has incomplete pairs
712
b17a8968 713 // Remove the noise in order to check if the 4 bytes pairs are complete
7c0ec1d8 714 $string = str_replace(array("\r\n","\n", "\r", " "),array('','','',''),$string);
715
42ce44f8 716 $sStringRem = '';
7c0ec1d8 717 $iMod = strlen($string) % 4;
718 if ($iMod) {
719 $sStringRem = substr($string,-$iMod);
b17a8968 720 // Check if $sStringRem contains padding characters
7c0ec1d8 721 if (substr($sStringRem,-1) != '=') {
722 $string = substr($string,0,-$iMod);
723 } else {
724 $sStringRem = '';
725 }
726 }
7c7b74b3 727 $string = base64_decode($string);
7c0ec1d8 728 return $sStringRem;
7c7b74b3 729}
730
fdf7cef1 731/**
fa26ab45 732 * Decodes encoded string (usually message body)
733 *
734 * This function decodes a string (usually the message body)
735 * depending on the encoding type. Currently quoted-printable
736 * and base64 encodings are supported.
737 *
738 * The decode_body hook was added to this function in 1.4.2/1.5.0.
739 * The $force_crlf parameter was added in 1.5.2.
740 *
741 * @param string $string The encoded string
742 * @param string $encoding used encoding
743 * @param string $force_crlf Whether or not to force CRLF or LF
744 * line endings (or to leave as is).
745 * If given as "LF", line endings will
746 * all be converted to LF; if "CRLF",
747 * line endings will all be converted
748 * to CRLF. If given as an empty value,
9785376e 749 * the global $force_crlf_default will
fa26ab45 750 * be consulted (it can be specified in
751 * config/config_local.php). Otherwise,
752 * any other value will cause the string
753 * to be left alone. Note that this will
754 * be overridden to "LF" if not using at
755 * least PHP version 4.3.0. (OPTIONAL;
756 * default is empty - consult global
757 * default value)
758 *
759 * @return string The decoded string
fdf7cef1 760 *
fdf7cef1 761 * @since 1.0
fa26ab45 762 *
fdf7cef1 763 */
fa26ab45 764function decodeBody($string, $encoding, $force_crlf='') {
765
766 global $force_crlf_default;
767 if (empty($force_crlf)) $force_crlf = $force_crlf_default;
768 $force_crlf = strtoupper($force_crlf);
769
770 // must force line endings to LF due to broken
771 // quoted_printable_decode() in PHP versions
772 // before 4.3.0 (see below)
773 //
774 if (!check_php_version(4, 3, 0) || $force_crlf == 'LF')
775 $string = str_replace("\r\n", "\n", $string);
776 else if ($force_crlf == 'CRLF')
777 $string = str_replace("\n", "\r\n", $string);
83be314a 778
b583c3e8 779 $encoding = strtolower($encoding);
3d8371be 780
d849b570 781 $encoding_handler = do_hook('decode_body', $encoding);
5166f86a 782
783
fa26ab45 784 // plugins get first shot at decoding the string
5166f86a 785 //
786 if (!empty($encoding_handler) && function_exists($encoding_handler)) {
fa26ab45 787 $string = $encoding_handler('decode', $string);
5166f86a 788
fdf7cef1 789 } elseif ($encoding == 'quoted-printable' ||
8bd0068d 790 $encoding == 'quoted_printable') {
fa26ab45 791
792 // quoted_printable_decode() function is broken in older
793 // php versions. Text with \r\n decoding was fixed only
794 // in php 4.3.0. Minimal code requirement is PHP 4.0.4+
795 // and the above call to: str_replace("\r\n", "\n", $string);
796 //
797 $string = quoted_printable_decode($string);
798
fdf7cef1 799 } elseif ($encoding == 'base64') {
fa26ab45 800 $string = base64_decode($string);
b583c3e8 801 }
3d8371be 802
b583c3e8 803 // All other encodings are returned raw.
fa26ab45 804 return $string;
451f74a2 805}
806
9f7f68c3 807/**
8bd0068d 808 * Decodes headers
809 *
e1115979 810 * This function decodes strings that are encoded according to
8bd0068d 811 * RFC1522 (MIME Part Two: Message Header Extensions for Non-ASCII Text).
812 * Patched by Christian Schmidt <christian@ostenfeld.dk> 23/03/2002
813 *
814 * @param string $string header string that has to be made readable
815 * @param boolean $utfencode change message in order to be readable on user's charset. defaults to true
bcae324b 816 * @param boolean $htmlsafe preserve spaces and sanitize html special characters. defaults to true
8bd0068d 817 * @param boolean $decide decide if string can be utfencoded. defaults to false
818 * @return string decoded header string
819 */
bcae324b 820function decodeHeader ($string, $utfencode=true,$htmlsafe=true,$decide=false) {
caf0ab1d 821 global $languages, $squirrelmail_language,$default_charset, $fix_broken_base64_encoded_messages;
79e07c7e 822 if (is_array($string)) {
823 $string = implode("\n", $string);
824 }
da2415c1 825
10dec454 826 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 827 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader')) {
33a55f5a 828 $string = call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_decodeheader', $string);
08b7f7cc 829 // Do we need to return at this point?
830 // return $string;
83be314a 831 }
79e07c7e 832 $i = 0;
08b7f7cc 833 $iLastMatch = -2;
db65b6b0 834 $encoded = true;
0a06275a 835
821d1f14 836// FIXME: spaces are allowed inside quoted-printable encoding, but the following line will bust up any such encoded strings
098ea084 837 $aString = explode(' ',$string);
08b7f7cc 838 $ret = '';
098ea084 839 foreach ($aString as $chunk) {
358a78a1 840 if ($encoded && $chunk === '') {
08b7f7cc 841 continue;
358a78a1 842 } elseif ($chunk === '') {
08b7f7cc 843 $ret .= ' ';
844 continue;
845 }
098ea084 846 $encoded = false;
08b7f7cc 847 /* if encoded words are not separated by a linear-space-white we still catch them */
848 $j = $i-1;
7e6ca3e8 849
08b7f7cc 850 while ($match = preg_match('/^(.*)=\?([^?]*)\?(Q|B)\?([^?]*)\?=(.*)$/Ui',$chunk,$res)) {
851 /* if the last chunk isn't an encoded string then put back the space, otherwise don't */
852 if ($iLastMatch !== $j) {
bcae324b 853 if ($htmlsafe) {
9f7f68c3 854 $ret .= '&#32;';
08b7f7cc 855 } else {
856 $ret .= ' ';
857 }
858 }
859 $iLastMatch = $i;
860 $j = $i;
bcae324b 861 if ($htmlsafe) {
3047e291 862 $ret .= sm_encode_html_special_chars($res[1]);
cb718de0 863 } else {
864 $ret .= $res[1];
865 }
098ea084 866 $encoding = ucfirst($res[3]);
d6f584fc 867
868 /* decide about valid decoding */
869 if ($decide && is_conversion_safe($res[2])) {
8bd0068d 870 $utfencode=true;
871 $can_be_encoded=true;
d6f584fc 872 } else {
8bd0068d 873 $can_be_encoded=false;
d6f584fc 874 }
098ea084 875 switch ($encoding)
876 {
8bd0068d 877 case 'B':
caf0ab1d 878 // fix broken base64-encoded strings (remove end = padding,
879 // change any = to + in middle of string, add padding back
880 // to the end)
881 if ($fix_broken_base64_encoded_messages) {
882 $encoded_string_minus_padding = strtr(rtrim($res[4], '='), '=', '+');
883 $res[4] = str_pad($encoded_string_minus_padding, strlen($res[4]), '=');
884 }
8bd0068d 885 $replace = base64_decode($res[4]);
886 if ($utfencode) {
887 if ($can_be_encoded) {
888 /* convert string to different charset,
889 * if functions asks for it (usually in compose)
890 */
bcae324b 891 $ret .= charset_convert($res[2],$replace,$default_charset,$htmlsafe);
8bd0068d 892 } else {
893 // convert string to html codes in order to display it
894 $ret .= charset_decode($res[2],$replace);
895 }
fab65ca9 896 } else {
bcae324b 897 if ($htmlsafe) {
3047e291 898 $replace = sm_encode_html_special_chars($replace);
8bd0068d 899 }
900 $ret.= $replace;
fab65ca9 901 }
8bd0068d 902 break;
903 case 'Q':
904 $replace = str_replace('_', ' ', $res[4]);
6c4e2e9b 905 $replace = preg_replace_callback('/=([0-9a-f]{2})/i',
9cfb0b2e 906 (check_php_version(5, 3, 0)
907 ? function($matches) { return chr(hexdec($matches[1])); }
908 : create_function ('$matches', 'return chr(hexdec($matches[1]));')
909 ),
910 $replace);
8bd0068d 911 if ($utfencode) {
912 if ($can_be_encoded) {
913 /* convert string to different charset,
914 * if functions asks for it (usually in compose)
915 */
bcae324b 916 $replace = charset_convert($res[2], $replace,$default_charset,$htmlsafe);
8bd0068d 917 } else {
918 // convert string to html codes in order to display it
919 $replace = charset_decode($res[2], $replace);
920 }
921 } else {
bcae324b 922 if ($htmlsafe) {
3047e291 923 $replace = sm_encode_html_special_chars($replace);
8bd0068d 924 }
098ea084 925 }
8bd0068d 926 $ret .= $replace;
927 break;
928 default:
929 break;
79e07c7e 930 }
098ea084 931 $chunk = $res[5];
932 $encoded = true;
08b7f7cc 933 }
934 if (!$encoded) {
bcae324b 935 if ($htmlsafe) {
9f7f68c3 936 $ret .= '&#32;';
08b7f7cc 937 } else {
938 $ret .= ' ';
da2415c1 939 }
08b7f7cc 940 }
dc3d13a7 941
bcae324b 942 if (!$encoded && $htmlsafe) {
3047e291 943 $ret .= sm_encode_html_special_chars($chunk);
dc3d13a7 944 } else {
945 $ret .= $chunk;
946 }
098ea084 947 ++$i;
948 }
fd81e884 949 /* remove the first added space */
950 if ($ret) {
bcae324b 951 if ($htmlsafe) {
9f7f68c3 952 $ret = substr($ret,5);
fd81e884 953 } else {
954 $ret = substr($ret,1);
955 }
956 }
da2415c1 957
08b7f7cc 958 return $ret;
451f74a2 959}
960
9f7f68c3 961/**
a24cf710 962 * Encodes header
8bd0068d 963 *
a24cf710 964 * Function uses XTRA_CODE _encodeheader function, if such function exists.
a24cf710 965 *
758a7889 966 * Function uses Q encoding by default and encodes a string according to RFC
967 * 1522 for use in headers if it contains 8-bit characters or anything that
a24cf710 968 * looks like it should be encoded.
8bd0068d 969 *
758a7889 970 * Function switches to B encoding and encodeHeaderBase64() function, if
971 * string is 8bit and multibyte character set supported by mbstring extension
972 * is used. It can cause E_USER_NOTICE errors, if interface is used with
f270a6eb 973 * multibyte character set unsupported by mbstring extension.
974 *
8bd0068d 975 * @param string $string header string, that has to be encoded
976 * @return string quoted-printable encoded string
f270a6eb 977 * @todo make $mb_charsets system wide constant
8bd0068d 978 */
451f74a2 979function encodeHeader ($string) {
6fbd125b 980 global $default_charset, $languages, $squirrelmail_language;
83be314a 981
10dec454 982 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 983 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader')) {
33a55f5a 984 return call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_encodeheader', $string);
83be314a 985 }
793cc001 986
a24cf710 987 // Use B encoding for multibyte charsets
f270a6eb 988 $mb_charsets = array('utf-8','big5','gb2313','euc-kr');
989 if (in_array($default_charset,$mb_charsets) &&
990 in_array($default_charset,sq_mb_list_encodings()) &&
991 sq_is8bit($string)) {
992 return encodeHeaderBase64($string,$default_charset);
993 } elseif (in_array($default_charset,$mb_charsets) &&
994 sq_is8bit($string) &&
995 ! in_array($default_charset,sq_mb_list_encodings())) {
996 // Add E_USER_NOTICE error here (can cause 'Cannot add header information' warning in compose.php)
997 // trigger_error('encodeHeader: Multibyte character set unsupported by mbstring extension.',E_USER_NOTICE);
998 }
a24cf710 999
451f74a2 1000 // Encode only if the string contains 8-bit characters or =?
3d8371be 1001 $j = strlen($string);
098ea084 1002 $max_l = 75 - strlen($default_charset) - 7;
1003 $aRet = array();
451f74a2 1004 $ret = '';
c96c32f4 1005 $iEncStart = $enc_init = false;
0d53f0f9 1006 $cur_l = $iOffset = 0;
3d8371be 1007 for($i = 0; $i < $j; ++$i) {
1710ad65 1008 switch($string[$i])
c96c32f4 1009 {
6b76cffa 1010 case '"':
8bd0068d 1011 case '=':
1012 case '<':
1013 case '>':
1014 case ',':
1015 case '?':
1016 case '_':
1017 if ($iEncStart === false) {
1018 $iEncStart = $i;
1019 }
1020 $cur_l+=3;
1021 if ($cur_l > ($max_l-2)) {
1022 /* if there is an stringpart that doesn't need encoding, add it */
08b7f7cc 1023 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 1024 $aRet[] = "=?$default_charset?Q?$ret?=";
1025 $iOffset = $i;
1026 $cur_l = 0;
1027 $ret = '';
1028 $iEncStart = false;
08b7f7cc 1029 } else {
1710ad65 1030 $ret .= sprintf("=%02X",ord($string[$i]));
c96c32f4 1031 }
8bd0068d 1032 break;
1033 case '(':
1034 case ')':
1035 if ($iEncStart !== false) {
08b7f7cc 1036 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
8bd0068d 1037 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1038 $iOffset = $i;
8bd0068d 1039 $cur_l = 0;
1040 $ret = '';
1041 $iEncStart = false;
c96c32f4 1042 }
8bd0068d 1043 break;
1044 case ' ':
c96c32f4 1045 if ($iEncStart !== false) {
098ea084 1046 $cur_l++;
1047 if ($cur_l > $max_l) {
08b7f7cc 1048 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
c96c32f4 1049 $aRet[] = "=?$default_charset?Q?$ret?=";
c96c32f4 1050 $iOffset = $i;
1051 $cur_l = 0;
098ea084 1052 $ret = '';
8bd0068d 1053 $iEncStart = false;
08b7f7cc 1054 } else {
8bd0068d 1055 $ret .= '_';
c96c32f4 1056 }
3d8371be 1057 }
8bd0068d 1058 break;
1059 default:
1710ad65 1060 $k = ord($string[$i]);
8bd0068d 1061 if ($k > 126) {
1062 if ($iEncStart === false) {
1063 // do not start encoding in the middle of a string, also take the rest of the word.
1064 $sLeadString = substr($string,0,$i);
1065 $aLeadString = explode(' ',$sLeadString);
1066 $sToBeEncoded = array_pop($aLeadString);
1067 $iEncStart = $i - strlen($sToBeEncoded);
1068 $ret .= $sToBeEncoded;
1069 $cur_l += strlen($sToBeEncoded);
1070 }
1071 $cur_l += 3;
1072 /* first we add the encoded string that reached it's max size */
1073 if ($cur_l > ($max_l-2)) {
1074 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1075 $aRet[] = "=?$default_charset?Q?$ret?= "; /* the next part is also encoded => separate by space */
1076 $cur_l = 3;
1077 $ret = '';
1078 $iOffset = $i;
1079 $iEncStart = $i;
1080 }
1081 $enc_init = true;
1082 $ret .= sprintf("=%02X", $k);
1083 } else {
1084 if ($iEncStart !== false) {
1085 $cur_l++;
1086 if ($cur_l > $max_l) {
1087 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1088 $aRet[] = "=?$default_charset?Q?$ret?=";
1089 $iEncStart = false;
1090 $iOffset = $i;
1091 $cur_l = 0;
1092 $ret = '';
1093 } else {
1710ad65 1094 $ret .= $string[$i];
8bd0068d 1095 }
1096 }
1097 }
1098 break;
f7b3ba37 1099 }
451f74a2 1100 }
793cc001 1101
c96c32f4 1102 if ($enc_init) {
1103 if ($iEncStart !== false) {
1104 $aRet[] = substr($string,$iOffset,$iEncStart-$iOffset);
1105 $aRet[] = "=?$default_charset?Q?$ret?=";
1106 } else {
1107 $aRet[] = substr($string,$iOffset);
1108 }
1109 $string = implode('',$aRet);
451f74a2 1110 }
3d8371be 1111 return $string;
451f74a2 1112}
b74ba498 1113
f270a6eb 1114/**
1115 * Encodes string according to rfc2047 B encoding header formating rules
1116 *
758a7889 1117 * It is recommended way to encode headers with character sets that store
f270a6eb 1118 * symbols in more than one byte.
1119 *
1120 * Function requires mbstring support. If required mbstring functions are missing,
1121 * function returns false and sets E_USER_WARNING level error message.
1122 *
758a7889 1123 * Minimal requirements - php 4.0.6 with mbstring extension. Please note,
1124 * that mbstring functions will generate E_WARNING errors, if unsupported
f270a6eb 1125 * character set is used. mb_encode_mimeheader function provided by php
1126 * mbstring extension is not used in order to get better control of header
1127 * encoding.
1128 *
758a7889 1129 * Used php code functions - function_exists(), trigger_error(), strlen()
1130 * (is used with charset names and base64 strings). Used php mbstring
f270a6eb 1131 * functions - mb_strlen and mb_substr.
1132 *
758a7889 1133 * Related documents: rfc 2045 (BASE64 encoding), rfc 2047 (mime header
f270a6eb 1134 * encoding), rfc 2822 (header folding)
1135 *
1136 * @param string $string header string that must be encoded
758a7889 1137 * @param string $charset character set. Must be supported by mbstring extension.
f270a6eb 1138 * Use sq_mb_list_encodings() to detect supported charsets.
1139 * @return string string encoded according to rfc2047 B encoding formating rules
1140 * @since 1.5.1
1141 * @todo First header line can be wrapped to $iMaxLength - $HeaderFieldLength - 1
1142 * @todo Do we want to control max length of header?
1143 * @todo Do we want to control EOL (end-of-line) marker?
1144 * @todo Do we want to translate error message?
1145 */
1146function encodeHeaderBase64($string,$charset) {
1147 /**
1148 * Check mbstring function requirements.
1149 */
1150 if (! function_exists('mb_strlen') ||
1151 ! function_exists('mb_substr')) {
1152 // set E_USER_WARNING
1153 trigger_error('encodeHeaderBase64: Required mbstring functions are missing.',E_USER_WARNING);
1154 // return false
1155 return false;
1156 }
1157
1158 // initial return array
1159 $aRet = array();
1160
1161 /**
1162 * header length = 75 symbols max (same as in encodeHeader)
1163 * remove $charset length
1164 * remove =? ? ?= (5 chars)
1165 * remove 2 more chars (\r\n ?)
1166 */
1167 $iMaxLength = 75 - strlen($charset) - 7;
1168
1169 // set first character position
1170 $iStartCharNum = 0;
1171
1172 // loop through all characters. count characters and not bytes.
1173 for ($iCharNum=1; $iCharNum<=mb_strlen($string,$charset); $iCharNum++) {
1174 // encode string from starting character to current character.
1175 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1176
1177 // Check encoded string length
1178 if(strlen($encoded_string)>$iMaxLength) {
1179 // if string exceeds max length, reduce number of encoded characters and add encoded string part to array
1180 $aRet[] = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum-1,$charset));
1181
1182 // set new starting character
1183 $iStartCharNum = $iCharNum-1;
1184
1185 // encode last char (in case it is last character in string)
1186 $encoded_string = base64_encode(mb_substr($string,$iStartCharNum,$iCharNum-$iStartCharNum,$charset));
1187 } // if string is shorter than max length - add next character
1188 }
1189
1190 // add last encoded string to array
1191 $aRet[] = $encoded_string;
1192
1193 // set initial return string
1194 $sRet = '';
1195
1196 // loop through encoded strings
1197 foreach($aRet as $string) {
1198 // TODO: Do we want to control EOL (end-of-line) marker
1199 if ($sRet!='') $sRet.= " ";
1200
1201 // add header tags and encoded string to return string
1202 $sRet.= '=?'.$charset.'?B?'.$string.'?=';
1203 }
1204
1205 return $sRet;
1206}
1207
691a2d25 1208/* This function trys to locate the entity_id of a specific mime element */
3d8371be 1209function find_ent_id($id, $message) {
a171b359 1210 for ($i = 0, $ret = ''; $ret == '' && $i < count($message->entities); $i++) {
1211 if ($message->entities[$i]->header->type0 == 'multipart') {
3d8371be 1212 $ret = find_ent_id($id, $message->entities[$i]);
451f74a2 1213 } else {
3d8371be 1214 if (strcasecmp($message->entities[$i]->header->id, $id) == 0) {
d8cffbab 1215// if (sq_check_save_extension($message->entities[$i])) {
8bd0068d 1216 return $message->entities[$i]->entity_id;
da2415c1 1217// }
c8f5f606 1218 } elseif (!empty($message->entities[$i]->header->parameters['name'])) {
1219 /**
1220 * This is part of a fix for Outlook Express 6.x generating
1221 * cid URLs without creating content-id headers
1222 * @@JA - 20050207
1223 */
1224 if (strcasecmp($message->entities[$i]->header->parameters['name'], $id) == 0) {
1225 return $message->entities[$i]->entity_id;
1226 }
3d8371be 1227 }
a3daaaf3 1228 }
451f74a2 1229 }
3d8371be 1230 return $ret;
451f74a2 1231}
a3daaaf3 1232
e5e9381a 1233function sq_check_save_extension($message) {
1234 $filename = $message->getFilename();
1235 $ext = substr($filename, strrpos($filename,'.')+1);
1236 $save_extensions = array('jpg','jpeg','gif','png','bmp');
3d8371be 1237 return in_array($ext, $save_extensions);
e5e9381a 1238}
1239
1240
691a2d25 1241/**
8bd0068d 1242 ** HTMLFILTER ROUTINES
1243 */
451f74a2 1244
2dd879b8 1245/**
0493ed11 1246 * This function checks attribute values for entity-encoded values
1247 * and returns them translated into 8-bit strings so we can run
1248 * checks on them.
8bd0068d 1249 *
0493ed11 1250 * @param $attvalue A string to run entity check against.
1251 * @return Nothing, modifies a reference value.
8bd0068d 1252 */
0493ed11 1253function sq_defang(&$attvalue){
1254 $me = 'sq_defang';
2dd879b8 1255 /**
0493ed11 1256 * Skip this if there aren't ampersands or backslashes.
8bd0068d 1257 */
0493ed11 1258 if (strpos($attvalue, '&') === false
1259 && strpos($attvalue, '\\') === false){
1260 return;
2dd879b8 1261 }
0493ed11 1262 $m = false;
2b665f28 1263 // before deent, translate the dangerous unicode characters and ... to safe values
1264 // otherwise the regular expressions do not match.
1265
1266
1267
0493ed11 1268 do {
1269 $m = false;
1270 $m = $m || sq_deent($attvalue, '/\&#0*(\d+);*/s');
1271 $m = $m || sq_deent($attvalue, '/\&#x0*((\d|[a-f])+);*/si', true);
1272 $m = $m || sq_deent($attvalue, '/\\\\(\d+)/s', true);
1273 } while ($m == true);
1274 $attvalue = stripslashes($attvalue);
2dd879b8 1275}
1276
1277/**
8bd0068d 1278 * Kill any tabs, newlines, or carriage returns. Our friends the
1279 * makers of the browser with 95% market value decided that it'd
1280 * be funny to make "java[tab]script" be just as good as "javascript".
1281 *
1282 * @param attvalue The attribute value before extraneous spaces removed.
0493ed11 1283 * @return attvalue Nothing, modifies a reference value.
8bd0068d 1284 */
0493ed11 1285function sq_unspace(&$attvalue){
1286 $me = 'sq_unspace';
1287 if (strcspn($attvalue, "\t\r\n\0 ") != strlen($attvalue)){
1288 $attvalue = str_replace(Array("\t", "\r", "\n", "\0", " "),
1289 Array('', '', '', '', ''), $attvalue);
2dd879b8 1290 }
2dd879b8 1291}
1292
2b665f28 1293/**
88de4926 1294 * Translate all dangerous Unicode or Shift_JIS characters which are accepted by
2b665f28 1295 * IE as regular characters.
1296 *
1297 * @param attvalue The attribute value before dangerous characters are translated.
1298 * @return attvalue Nothing, modifies a reference value.
1299 * @author Marc Groot Koerkamp.
1300 */
1301function sq_fixIE_idiocy(&$attvalue) {
1302 // remove NUL
1303 $attvalue = str_replace("\0", "", $attvalue);
1304 // remove comments
1305 $attvalue = preg_replace("/(\/\*.*?\*\/)/","",$attvalue);
1306
88de4926 1307 // IE has the evil habit of accepting every possible value for the attribute expression.
1308 // The table below contains characters which are parsed by IE if they are used in the "expression"
2b665f28 1309 // attribute value.
1310 $aDangerousCharsReplacementTable = array(
1311 array('&#x029F;', '&#0671;' ,/* L UNICODE IPA Extension */
1312 '&#x0280;', '&#0640;' ,/* R UNICODE IPA Extension */
1313 '&#x0274;', '&#0628;' ,/* N UNICODE IPA Extension */
567dc524 1314 '&#xFF25;', '&#65317;' ,/* Unicode FULLWIDTH LATIN CAPITAL LETTER E */
1315 '&#xFF45;', '&#65349;' ,/* Unicode FULLWIDTH LATIN SMALL LETTER E */
2b665f28 1316 '&#xFF38;', '&#65336;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER X */
1317 '&#xFF58;', '&#65368;',/* Unicode FULLWIDTH LATIN SMALL LETTER X */
1318 '&#xFF30;', '&#65328;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER P */
1319 '&#xFF50;', '&#65360;',/* Unicode FULLWIDTH LATIN SMALL LETTER P */
1320 '&#xFF32;', '&#65330;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER R */
1321 '&#xFF52;', '&#65362;',/* Unicode FULLWIDTH LATIN SMALL LETTER R */
1322 '&#xFF33;', '&#65331;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER S */
1323 '&#xFF53;', '&#65363;',/* Unicode FULLWIDTH LATIN SMALL LETTER S */
1324 '&#xFF29;', '&#65321;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER I */
1325 '&#xFF49;', '&#65353;',/* Unicode FULLWIDTH LATIN SMALL LETTER I */
1326 '&#xFF2F;', '&#65327;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER O */
1327 '&#xFF4F;', '&#65359;',/* Unicode FULLWIDTH LATIN SMALL LETTER O */
1328 '&#xFF2E;', '&#65326;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER N */
1329 '&#xFF4E;', '&#65358;',/* Unicode FULLWIDTH LATIN SMALL LETTER N */
1330 '&#xFF2C;', '&#65324;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER L */
1331 '&#xFF4C;', '&#65356;',/* Unicode FULLWIDTH LATIN SMALL LETTER L */
1332 '&#xFF35;', '&#65333;',/* Unicode FULLWIDTH LATIN CAPITAL LETTER U */
1333 '&#xFF55;', '&#65365;',/* Unicode FULLWIDTH LATIN SMALL LETTER U */
1334 '&#x207F;', '&#8319;' ,/* Unicode SUPERSCRIPT LATIN SMALL LETTER N */
567dc524 1335 "\xEF\xBC\xA5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER E */ // in unicode this is some Chinese char range
1336 "\xEF\xBD\x85", /* Shift JIS FULLWIDTH LATIN SMALL LETTER E */
1337 "\xEF\xBC\xB8", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER X */
1338 "\xEF\xBD\x98", /* Shift JIS FULLWIDTH LATIN SMALL LETTER X */
1339 "\xEF\xBC\xB0", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER P */
1340 "\xEF\xBD\x90", /* Shift JIS FULLWIDTH LATIN SMALL LETTER P */
1341 "\xEF\xBC\xB2", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER R */
1342 "\xEF\xBD\x92", /* Shift JIS FULLWIDTH LATIN SMALL LETTER R */
1343 "\xEF\xBC\xB3", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER S */
1344 "\xEF\xBD\x93", /* Shift JIS FULLWIDTH LATIN SMALL LETTER S */
1345 "\xEF\xBC\xA9", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER I */
1346 "\xEF\xBD\x89", /* Shift JIS FULLWIDTH LATIN SMALL LETTER I */
1347 "\xEF\xBC\xAF", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER O */
1348 "\xEF\xBD\x8F", /* Shift JIS FULLWIDTH LATIN SMALL LETTER O */
1349 "\xEF\xBC\xAE", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER N */
1350 "\xEF\xBD\x8E", /* Shift JIS FULLWIDTH LATIN SMALL LETTER N */
1351 "\xEF\xBC\xAC", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER L */
1352 "\xEF\xBD\x8C", /* Shift JIS FULLWIDTH LATIN SMALL LETTER L */
1353 "\xEF\xBC\xB5", /* Shift JIS FULLWIDTH LATIN CAPITAL LETTER U */
1354 "\xEF\xBD\x95", /* Shift JIS FULLWIDTH LATIN SMALL LETTER U */
1355 "\xE2\x81\xBF", /* Shift JIS FULLWIDTH SUPERSCRIPT N */
1356 "\xCA\x9F", /* L UNICODE IPA Extension */
1357 "\xCA\x80", /* R UNICODE IPA Extension */
1358 "\xC9\xB4"), /* N UNICODE IPA Extension */
2b665f28 1359 array('l', 'l', 'r','r','n','n',
567dc524 1360 'E','E','e','e','X','X','x','x','P','P','p','p','R','R','r','r','S','S','s','s','I','I',
1361 'i','i','O','O','o','o','N','N','n','n','L','L','l','l','U','U','u','u','n','n',
1362 'E','e','X','x','P','p','R','r','S','s','I','i','O','o','N','n','L','l','U','u','n','l','r','n'));
2b665f28 1363 $attvalue = str_replace($aDangerousCharsReplacementTable[0],$aDangerousCharsReplacementTable[1],$attvalue);
1364
88de4926 1365 // Escapes are useful for special characters like "{}[]()'&. In other cases they are
1366 // used for XSS.
2b665f28 1367 $attvalue = preg_replace("/(\\\\)([a-zA-Z]{1})/",'$2',$attvalue);
1368}
1369
691a2d25 1370/**
8bd0068d 1371 * This function returns the final tag out of the tag name, an array
1372 * of attributes, and the type of the tag. This function is called by
1373 * sq_sanitize internally.
1374 *
1375 * @param $tagname the name of the tag.
1376 * @param $attary the array of attributes and their values
1377 * @param $tagtype The type of the tag (see in comments).
1378 * @return a string with the final tag representation.
1379 */
691a2d25 1380function sq_tagprint($tagname, $attary, $tagtype){
b583c3e8 1381 $me = 'sq_tagprint';
3d8371be 1382
691a2d25 1383 if ($tagtype == 2){
1384 $fulltag = '</' . $tagname . '>';
1385 } else {
1386 $fulltag = '<' . $tagname;
1387 if (is_array($attary) && sizeof($attary)){
1388 $atts = Array();
c71c845c 1389 foreach ($attary as $attname => $attvalue){
691a2d25 1390 array_push($atts, "$attname=$attvalue");
1391 }
1392 $fulltag .= ' ' . join(" ", $atts);
1393 }
1394 if ($tagtype == 3){
b583c3e8 1395 $fulltag .= ' /';
691a2d25 1396 }
b583c3e8 1397 $fulltag .= '>';
451f74a2 1398 }
691a2d25 1399 return $fulltag;
451f74a2 1400}
a3daaaf3 1401
691a2d25 1402/**
8bd0068d 1403 * A small helper function to use with array_walk. Modifies a by-ref
1404 * value and makes it lowercase.
1405 *
1406 * @param $val a value passed by-ref.
1407 * @return void since it modifies a by-ref value.
1408 */
691a2d25 1409function sq_casenormalize(&$val){
1410 $val = strtolower($val);
1411}
451f74a2 1412
691a2d25 1413/**
8bd0068d 1414 * This function skips any whitespace from the current position within
1415 * a string and to the next non-whitespace value.
1416 *
1417 * @param $body the string
1418 * @param $offset the offset within the string where we should start
1419 * looking for the next non-whitespace character.
1420 * @return the location within the $body where the next
1421 * non-whitespace char is located.
1422 */
691a2d25 1423function sq_skipspace($body, $offset){
b583c3e8 1424 $me = 'sq_skipspace';
3d8371be 1425 preg_match('/^(\s*)/s', substr($body, $offset), $matches);
aab65dbd 1426 if (!empty($matches[1])){
1427 $offset += strlen($matches[1]);
451f74a2 1428 }
691a2d25 1429 return $offset;
451f74a2 1430}
a3daaaf3 1431
691a2d25 1432/**
8bd0068d 1433 * This function looks for the next character within a string. It's
1434 * really just a glorified "strpos", except it catches if failures
1435 * nicely.
1436 *
1437 * @param $body The string to look for needle in.
1438 * @param $offset Start looking from this position.
1439 * @param $needle The character/string to look for.
1440 * @return location of the next occurance of the needle, or
1441 * strlen($body) if needle wasn't found.
1442 */
691a2d25 1443function sq_findnxstr($body, $offset, $needle){
3d8371be 1444 $me = 'sq_findnxstr';
691a2d25 1445 $pos = strpos($body, $needle, $offset);
1446 if ($pos === FALSE){
1447 $pos = strlen($body);
451f74a2 1448 }
691a2d25 1449 return $pos;
451f74a2 1450}
a3daaaf3 1451
691a2d25 1452/**
8bd0068d 1453 * This function takes a PCRE-style regexp and tries to match it
1454 * within the string.
1455 *
1456 * @param $body The string to look for needle in.
1457 * @param $offset Start looking from here.
1458 * @param $reg A PCRE-style regex to match.
1459 * @return Returns a false if no matches found, or an array
1460 * with the following members:
1461 * - integer with the location of the match within $body
1462 * - string with whatever content between offset and the match
1463 * - string with whatever it is we matched
1464 */
691a2d25 1465function sq_findnxreg($body, $offset, $reg){
b583c3e8 1466 $me = 'sq_findnxreg';
691a2d25 1467 $matches = Array();
1468 $retarr = Array();
7d06541f 1469 preg_match("%^(.*?)($reg)%si", substr($body, $offset), $matches);
1710ad65 1470 if (!isset($matches[0]) || !$matches[0]){
691a2d25 1471 $retarr = false;
1472 } else {
1710ad65 1473 $retarr[0] = $offset + strlen($matches[1]);
1474 $retarr[1] = $matches[1];
1475 $retarr[2] = $matches[2];
691a2d25 1476 }
1477 return $retarr;
1478}
a3daaaf3 1479
691a2d25 1480/**
8bd0068d 1481 * This function looks for the next tag.
1482 *
1483 * @param $body String where to look for the next tag.
1484 * @param $offset Start looking from here.
1485 * @return false if no more tags exist in the body, or
1486 * an array with the following members:
1487 * - string with the name of the tag
1488 * - array with attributes and their values
1489 * - integer with tag type (1, 2, or 3)
1490 * - integer where the tag starts (starting "<")
1491 * - integer where the tag ends (ending ">")
1492 * first three members will be false, if the tag is invalid.
1493 */
691a2d25 1494function sq_getnxtag($body, $offset){
b583c3e8 1495 $me = 'sq_getnxtag';
691a2d25 1496 if ($offset > strlen($body)){
1497 return false;
1498 }
1499 $lt = sq_findnxstr($body, $offset, "<");
1500 if ($lt == strlen($body)){
1501 return false;
1502 }
1503 /**
8bd0068d 1504 * We are here:
1505 * blah blah <tag attribute="value">
1506 * \---------^
1507 */
691a2d25 1508 $pos = sq_skipspace($body, $lt+1);
1509 if ($pos >= strlen($body)){
1510 return Array(false, false, false, $lt, strlen($body));
1511 }
1512 /**
8bd0068d 1513 * There are 3 kinds of tags:
1514 * 1. Opening tag, e.g.:
1515 * <a href="blah">
1516 * 2. Closing tag, e.g.:
1517 * </a>
1518 * 3. XHTML-style content-less tag, e.g.:
1519 * <img src="blah" />
1520 */
691a2d25 1521 $tagtype = false;
1522 switch (substr($body, $pos, 1)){
3d8371be 1523 case '/':
1524 $tagtype = 2;
1525 $pos++;
1526 break;
1527 case '!':
1528 /**
8bd0068d 1529 * A comment or an SGML declaration.
1530 */
3d8371be 1531 if (substr($body, $pos+1, 2) == "--"){
1532 $gt = strpos($body, "-->", $pos);
1533 if ($gt === false){
1534 $gt = strlen($body);
1535 } else {
1536 $gt += 2;
1537 }
1538 return Array(false, false, false, $lt, $gt);
bb8d0799 1539 } else {
3d8371be 1540 $gt = sq_findnxstr($body, $pos, ">");
1541 return Array(false, false, false, $lt, $gt);
1542 }
1543 break;
1544 default:
1545 /**
8bd0068d 1546 * Assume tagtype 1 for now. If it's type 3, we'll switch values
1547 * later.
1548 */
3d8371be 1549 $tagtype = 1;
1550 break;
691a2d25 1551 }
a3daaaf3 1552
0493ed11 1553 $tag_start = $pos;
691a2d25 1554 $tagname = '';
1555 /**
8bd0068d 1556 * Look for next [\W-_], which will indicate the end of the tag name.
1557 */
691a2d25 1558 $regary = sq_findnxreg($body, $pos, "[^\w\-_]");
1559 if ($regary == false){
1560 return Array(false, false, false, $lt, strlen($body));
1561 }
1562 list($pos, $tagname, $match) = $regary;
1563 $tagname = strtolower($tagname);
1564
1565 /**
8bd0068d 1566 * $match can be either of these:
1567 * '>' indicating the end of the tag entirely.
1568 * '\s' indicating the end of the tag name.
1569 * '/' indicating that this is type-3 xhtml tag.
1570 *
1571 * Whatever else we find there indicates an invalid tag.
1572 */
691a2d25 1573 switch ($match){
3d8371be 1574 case '/':
691a2d25 1575 /**
8bd0068d 1576 * This is an xhtml-style tag with a closing / at the
1577 * end, like so: <img src="blah" />. Check if it's followed
1578 * by the closing bracket. If not, then this tag is invalid
1579 */
3d8371be 1580 if (substr($body, $pos, 2) == "/>"){
1581 $pos++;
1582 $tagtype = 3;
1583 } else {
1584 $gt = sq_findnxstr($body, $pos, ">");
1585 $retary = Array(false, false, false, $lt, $gt);
1586 return $retary;
1587 }
1588 case '>':
1589 return Array($tagname, false, $tagtype, $lt, $pos);
1590 break;
1591 default:
1592 /**
8bd0068d 1593 * Check if it's whitespace
1594 */
3d8371be 1595 if (!preg_match('/\s/', $match)){
1596 /**
8bd0068d 1597 * This is an invalid tag! Look for the next closing ">".
1598 */
7d06541f 1599 $gt = sq_findnxstr($body, $lt, ">");
3d8371be 1600 return Array(false, false, false, $lt, $gt);
1601 }
1602 break;
691a2d25 1603 }
3d8371be 1604
691a2d25 1605 /**
8bd0068d 1606 * At this point we're here:
1607 * <tagname attribute='blah'>
1608 * \-------^
1609 *
1610 * At this point we loop in order to find all attributes.
1611 */
691a2d25 1612 $attname = '';
0493ed11 1613 $atttype = false;
691a2d25 1614 $attary = Array();
1615
1616 while ($pos <= strlen($body)){
1617 $pos = sq_skipspace($body, $pos);
1618 if ($pos == strlen($body)){
1619 /**
8bd0068d 1620 * Non-closed tag.
1621 */
691a2d25 1622 return Array(false, false, false, $lt, $pos);
1623 }
1624 /**
8bd0068d 1625 * See if we arrived at a ">" or "/>", which means that we reached
1626 * the end of the tag.
1627 */
691a2d25 1628 $matches = Array();
164800ad 1629 if (preg_match("%^(\s*)(>|/>)%s", substr($body, $pos), $matches)) {
c828931c 1630 /**
8bd0068d 1631 * Yep. So we did.
1632 */
1710ad65 1633 $pos += strlen($matches[1]);
1634 if ($matches[2] == "/>"){
c828931c 1635 $tagtype = 3;
1636 $pos++;
1637 }
1638 return Array($tagname, $attary, $tagtype, $lt, $pos);
1639 }
a3daaaf3 1640
cca46357 1641 /**
8bd0068d 1642 * There are several types of attributes, with optional
1643 * [:space:] between members.
1644 * Type 1:
1645 * attrname[:space:]=[:space:]'CDATA'
1646 * Type 2:
1647 * attrname[:space:]=[:space:]"CDATA"
1648 * Type 3:
1649 * attr[:space:]=[:space:]CDATA
1650 * Type 4:
1651 * attrname
1652 *
1653 * We leave types 1 and 2 the same, type 3 we check for
1654 * '"' and convert to "&quot" if needed, then wrap in
1655 * double quotes. Type 4 we convert into:
1656 * attrname="yes".
1657 */
3f7c623f 1658 $regary = sq_findnxreg($body, $pos, "[^:\w\-_]");
691a2d25 1659 if ($regary == false){
1660 /**
8bd0068d 1661 * Looks like body ended before the end of tag.
1662 */
691a2d25 1663 return Array(false, false, false, $lt, strlen($body));
cca46357 1664 }
691a2d25 1665 list($pos, $attname, $match) = $regary;
1666 $attname = strtolower($attname);
1667 /**
8bd0068d 1668 * We arrived at the end of attribute name. Several things possible
1669 * here:
1670 * '>' means the end of the tag and this is attribute type 4
1671 * '/' if followed by '>' means the same thing as above
1672 * '\s' means a lot of things -- look what it's followed by.
1673 * anything else means the attribute is invalid.
1674 */
691a2d25 1675 switch($match){
3d8371be 1676 case '/':
691a2d25 1677 /**
8bd0068d 1678 * This is an xhtml-style tag with a closing / at the
1679 * end, like so: <img src="blah" />. Check if it's followed
1680 * by the closing bracket. If not, then this tag is invalid
1681 */
3d8371be 1682 if (substr($body, $pos, 2) == "/>"){
691a2d25 1683 $pos++;
3d8371be 1684 $tagtype = 3;
691a2d25 1685 } else {
3d8371be 1686 $gt = sq_findnxstr($body, $pos, ">");
1687 $retary = Array(false, false, false, $lt, $gt);
1688 return $retary;
1689 }
1690 case '>':
1710ad65 1691 $attary[$attname] = '"yes"';
3d8371be 1692 return Array($tagname, $attary, $tagtype, $lt, $pos);
1693 break;
1694 default:
1695 /**
8bd0068d 1696 * Skip whitespace and see what we arrive at.
1697 */
3d8371be 1698 $pos = sq_skipspace($body, $pos);
1699 $char = substr($body, $pos, 1);
1700 /**
8bd0068d 1701 * Two things are valid here:
1702 * '=' means this is attribute type 1 2 or 3.
1703 * \w means this was attribute type 4.
1704 * anything else we ignore and re-loop. End of tag and
1705 * invalid stuff will be caught by our checks at the beginning
1706 * of the loop.
1707 */
3d8371be 1708 if ($char == "="){
1709 $pos++;
1710 $pos = sq_skipspace($body, $pos);
691a2d25 1711 /**
8bd0068d 1712 * Here are 3 possibilities:
1713 * "'" attribute type 1
1714 * '"' attribute type 2
1715 * everything else is the content of tag type 3
1716 */
3d8371be 1717 $quot = substr($body, $pos, 1);
1718 if ($quot == "'"){
1719 $regary = sq_findnxreg($body, $pos+1, "\'");
1720 if ($regary == false){
1721 return Array(false, false, false, $lt, strlen($body));
1722 }
1723 list($pos, $attval, $match) = $regary;
1724 $pos++;
1710ad65 1725 $attary[$attname] = "'" . $attval . "'";
3d8371be 1726 } else if ($quot == '"'){
1727 $regary = sq_findnxreg($body, $pos+1, '\"');
1728 if ($regary == false){
1729 return Array(false, false, false, $lt, strlen($body));
1730 }
1731 list($pos, $attval, $match) = $regary;
1732 $pos++;
1710ad65 1733 $attary[$attname] = '"' . $attval . '"';
3d8371be 1734 } else {
1735 /**
8bd0068d 1736 * These are hateful. Look for \s, or >.
1737 */
3d8371be 1738 $regary = sq_findnxreg($body, $pos, "[\s>]");
1739 if ($regary == false){
1740 return Array(false, false, false, $lt, strlen($body));
1741 }
1742 list($pos, $attval, $match) = $regary;
1743 /**
8bd0068d 1744 * If it's ">" it will be caught at the top.
1745 */
3d8371be 1746 $attval = preg_replace("/\"/s", "&quot;", $attval);
1710ad65 1747 $attary[$attname] = '"' . $attval . '"';
7e235a1a 1748 }
3d8371be 1749 } else if (preg_match("|[\w/>]|", $char)) {
691a2d25 1750 /**
8bd0068d 1751 * That was attribute type 4.
1752 */
1710ad65 1753 $attary[$attname] = '"yes"';
3d8371be 1754 } else {
1755 /**
8bd0068d 1756 * An illegal character. Find next '>' and return.
1757 */
3d8371be 1758 $gt = sq_findnxstr($body, $pos, ">");
1759 return Array(false, false, false, $lt, $gt);
451f74a2 1760 }
3d8371be 1761 break;
691a2d25 1762 }
1763 }
1764 /**
8bd0068d 1765 * The fact that we got here indicates that the tag end was never
1766 * found. Return invalid tag indication so it gets stripped.
1767 */
691a2d25 1768 return Array(false, false, false, $lt, strlen($body));
1769}
1770
1771/**
0493ed11 1772 * Translates entities into literal values so they can be checked.
8bd0068d 1773 *
0493ed11 1774 * @param $attvalue the by-ref value to check.
1775 * @param $regex the regular expression to check against.
1776 * @param $hex whether the entites are hexadecimal.
1777 * @return True or False depending on whether there were matches.
8bd0068d 1778 */
0493ed11 1779function sq_deent(&$attvalue, $regex, $hex=false){
b583c3e8 1780 $me = 'sq_deent';
0493ed11 1781 $ret_match = false;
2b665f28 1782 // remove comments
1783 //$attvalue = preg_replace("/(\/\*.*\*\/)/","",$attvalue);
0493ed11 1784 preg_match_all($regex, $attvalue, $matches);
1785 if (is_array($matches) && sizeof($matches[0]) > 0){
1786 $repl = Array();
1787 for ($i = 0; $i < sizeof($matches[0]); $i++){
1788 $numval = $matches[1][$i];
1789 if ($hex){
1790 $numval = hexdec($numval);
a3daaaf3 1791 }
1710ad65 1792 $repl[$matches[0][$i]] = chr($numval);
691a2d25 1793 }
0493ed11 1794 $attvalue = strtr($attvalue, $repl);
1795 return true;
1796 } else {
1797 return false;
691a2d25 1798 }
691a2d25 1799}
1800
1801/**
8bd0068d 1802 * This function runs various checks against the attributes.
1803 *
1804 * @param $tagname String with the name of the tag.
1805 * @param $attary Array with all tag attributes.
1806 * @param $rm_attnames See description for sq_sanitize
1807 * @param $bad_attvals See description for sq_sanitize
1808 * @param $add_attr_to_tag See description for sq_sanitize
1809 * @param $message message object
1810 * @param $id message id
1811 * @return Array with modified attributes.
1812 */
da2415c1 1813function sq_fixatts($tagname,
1814 $attary,
691a2d25 1815 $rm_attnames,
1816 $bad_attvals,
1817 $add_attr_to_tag,
1818 $message,
b3af12ef 1819 $id,
3d8371be 1820 $mailbox
691a2d25 1821 ){
b583c3e8 1822 $me = 'sq_fixatts';
c71c845c 1823 foreach ($attary as $attname => $attvalue){
691a2d25 1824 /**
8bd0068d 1825 * See if this attribute should be removed.
1826 */
691a2d25 1827 foreach ($rm_attnames as $matchtag=>$matchattrs){
1828 if (preg_match($matchtag, $tagname)){
1829 foreach ($matchattrs as $matchattr){
1830 if (preg_match($matchattr, $attname)){
1710ad65 1831 unset($attary[$attname]);
691a2d25 1832 continue;
1833 }
451f74a2 1834 }
451f74a2 1835 }
691a2d25 1836 }
2b665f28 1837 /**
1838 * Workaround for IE quirks
1839 */
1840 sq_fixIE_idiocy($attvalue);
1841
691a2d25 1842 /**
8bd0068d 1843 * Remove any backslashes, entities, and extraneous whitespace.
1844 */
2b665f28 1845
1846 $oldattvalue = $attvalue;
0493ed11 1847 sq_defang($attvalue);
2b665f28 1848 if ($attname == 'style' && $attvalue !== $oldattvalue) {
1849 // entities are used in the attribute value. In 99% of the cases it's there as XSS
1850 // i.e.<div style="{ left:exp&#x0280;essio&#x0274;( alert('XSS') ) }">
1851 $attvalue = "idiocy";
1710ad65 1852 $attary[$attname] = $attvalue;
2b665f28 1853 }
0493ed11 1854 sq_unspace($attvalue);
af861a34 1855
691a2d25 1856 /**
8bd0068d 1857 * Now let's run checks on the attvalues.
1858 * I don't expect anyone to comprehend this. If you do,
1859 * get in touch with me so I can drive to where you live and
1860 * shake your hand personally. :)
1861 */
691a2d25 1862 foreach ($bad_attvals as $matchtag=>$matchattrs){
1863 if (preg_match($matchtag, $tagname)){
1864 foreach ($matchattrs as $matchattr=>$valary){
1865 if (preg_match($matchattr, $attname)){
1866 /**
8bd0068d 1867 * There are two arrays in valary.
1868 * First is matches.
1869 * Second one is replacements
1870 */
691a2d25 1871 list($valmatch, $valrepl) = $valary;
da2415c1 1872 $newvalue =
691a2d25 1873 preg_replace($valmatch, $valrepl, $attvalue);
1874 if ($newvalue != $attvalue){
1710ad65 1875 $attary[$attname] = $newvalue;
567dc524 1876 $attvalue = $newvalue;
691a2d25 1877 }
1878 }
1879 }
451f74a2 1880 }
a3daaaf3 1881 }
567dc524 1882 if ($attname == 'style') {
1883 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1884 // 8bit and control characters in style attribute values can be used for XSS, remove them
1710ad65 1885 $attary[$attname] = '"disallowed character"';
567dc524 1886 }
1887 preg_match_all("/url\s*\((.+)\)/si",$attvalue,$aMatch);
1888 if (count($aMatch)) {
1889 foreach($aMatch[1] as $sMatch) {
1890 // url value
1891 $urlvalue = $sMatch;
1892 sq_fix_url($attname, $urlvalue, $message, $id, $mailbox,"'");
1710ad65 1893 $attary[$attname] = str_replace($sMatch,$urlvalue,$attvalue);
567dc524 1894 }
1895 }
691a2d25 1896 }
ff940ebc 1897 /**
567dc524 1898 * Use white list based filtering on attributes which can contain url's
ff940ebc 1899 */
caf0ab1d 1900 else if ($attname == 'href' || $attname == 'xlink:href' || $attname == 'src'
1901 || $attname == 'poster' || $attname == 'formaction'
1902 || $attname == 'background' || $attname == 'action') {
567dc524 1903 sq_fix_url($attname, $attvalue, $message, $id, $mailbox);
1710ad65 1904 $attary[$attname] = $attvalue;
ff940ebc 1905 }
a3daaaf3 1906 }
691a2d25 1907 /**
8bd0068d 1908 * See if we need to append any attributes to this tag.
1909 */
691a2d25 1910 foreach ($add_attr_to_tag as $matchtag=>$addattary){
1911 if (preg_match($matchtag, $tagname)){
1912 $attary = array_merge($attary, $addattary);
1913 }
1914 }
1915 return $attary;
451f74a2 1916}
a3daaaf3 1917
567dc524 1918/**
1919 * This function filters url's
1920 *
1921 * @param $attvalue String with attribute value to filter
1922 * @param $message message object
1923 * @param $id message id
1924 * @param $mailbox mailbox
1925 * @param $sQuote quoting characters around url's
1926 */
1927function sq_fix_url($attname, &$attvalue, $message, $id, $mailbox,$sQuote = '"') {
1928 $attvalue = trim($attvalue);
1929 if ($attvalue && ($attvalue[0] =='"'|| $attvalue[0] == "'")) {
1930 // remove the double quotes
1931 $sQuote = $attvalue[0];
1932 $attvalue = trim(substr($attvalue,1,-1));
1933 }
1934
d31f73f1 1935 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
1936 // images off by default.
afbf184c 1937 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
d31f73f1 1938
6ab8321f 1939 global $use_transparent_security_image;
1940 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
1941 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
567dc524 1942
1943 /**
1944 * Replace empty src tags with the blank image. src is only used
1945 * for frames, images, and image inputs. Doing a replace should
1946 * not affect them working as should be, however it will stop
1947 * IE from being kicked off when src for img tags are not set
1948 */
1949 if ($attvalue == '') {
1950 $attvalue = '"' . SM_PATH . 'images/blank.png"';
1951 } else {
1952 // first, disallow 8 bit characters and control characters
1953 if (preg_match('/[\0-\37\200-\377]+/',$attvalue)) {
1954 switch ($attname) {
1955 case 'href':
1956 $attvalue = $sQuote . 'http://invalid-stuff-detected.example.com' . $sQuote;
1957 break;
1958 default:
1959 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
1960 break;
1961 }
1962 } else {
1963 $aUrl = parse_url($attvalue);
1964 if (isset($aUrl['scheme'])) {
1965 switch(strtolower($aUrl['scheme'])) {
d75e755b 1966 case 'mailto':
567dc524 1967 case 'http':
1968 case 'https':
1969 case 'ftp':
1970 if ($attname != 'href') {
1971 if ($view_unsafe_images == false) {
1972 $attvalue = $sQuote . $secremoveimg . $sQuote;
1973 } else {
1974 if (isset($aUrl['path'])) {
8bbbc757 1975
1976 // No one has been able to show that image URIs
1977 // can be exploited, so for now, no restrictions
1978 // are made at all. If this proves to be a problem,
1979 // the commented-out code below can be of help.
1980 // (One consideration is that I see nothing in this
1981 // function that specifically says that we will
1982 // only ever arrive here when inspecting an image
1983 // tag, although that does seem to be the end
1984 // result - e.g., <script src="..."> where malicious
1985 // image URIs are in fact a problem are already
1986 // filtered out elsewhere.
1987 /* ---------------------------------
567dc524 1988 // validate image extension.
1989 $ext = strtolower(substr($aUrl['path'],strrpos($aUrl['path'],'.')));
1990 if (!in_array($ext,array('.jpeg','.jpg','xjpeg','.gif','.bmp','.jpe','.png','.xbm'))) {
8bbbc757 1991 // If URI is to something other than
1992 // a regular image file, get the contents
1993 // and try to see if it is an image.
1994 // Don't use Fileinfo (finfo_file()) because
1995 // we'd need to make the admin configure the
1996 // location of the magic.mime file (FIXME: add finfo_file() support later?)
1997 //
1998 $mime_type = '';
1999 if (function_exists('mime_content_type')
2000 && ($FILE = @fopen($attvalue, 'rb', FALSE))) {
2001
2002 // fetch file
2003 //
2004 $file_contents = '';
2005 while (!feof($FILE)) {
2006 $file_contents .= fread($FILE, 8192);
2007 }
2008 fclose($FILE);
2009
2010 // store file locally
2011 //
2012 global $attachment_dir, $username;
2013 $hashed_attachment_dir = getHashedDir($username, $attachment_dir);
2014 $localfilename = GenerateRandomString(32, '', 7);
2015 $full_localfilename = "$hashed_attachment_dir/$localfilename";
2016 while (file_exists($full_localfilename)) {
2017 $localfilename = GenerateRandomString(32, '', 7);
2018 $full_localfilename = "$hashed_attachment_dir/$localfilename";
2019 }
2020 $FILE = fopen("$hashed_attachment_dir/$localfilename", 'wb');
2021 fwrite($FILE, $file_contents);
2022 fclose($FILE);
2023
2024 // get mime type and remove file
2025 //
2026 $mime_type = mime_content_type("$hashed_attachment_dir/$localfilename");
2027 unlink("$hashed_attachment_dir/$localfilename");
2028 }
2029 // debug: echo "$attvalue FILE TYPE IS $mime_type<HR>";
2030 if (substr(strtolower($mime_type), 0, 5) != 'image') {
2031 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2032 }
567dc524 2033 }
8bbbc757 2034 --------------------------------- */
567dc524 2035 } else {
2036 $attvalue = $sQuote . SM_PATH . 'images/blank.png'. $sQuote;
2037 }
2038 }
cf088f55 2039 } else {
2040 $attvalue = $sQuote . $attvalue . $sQuote;
567dc524 2041 }
2042 break;
2043 case 'outbind':
2044 /**
2045 * "Hack" fix for Outlook using propriatary outbind:// protocol in img tags.
2046 * One day MS might actually make it match something useful, for now, falling
2047 * back to using cid2http, so we can grab the blank.png.
2048 */
cf088f55 2049 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2050 break;
2051 case 'cid':
2052 /**
2053 * Turn cid: urls into http-friendly ones.
2054 */
cf088f55 2055 $attvalue = $sQuote . sq_cid2http($message, $id, $attvalue, $mailbox) . $sQuote;
567dc524 2056 break;
2057 default:
2058 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2059 break;
2060 }
2061 } else {
b1fbb25f 2062 if (!isset($aUrl['path']) || $aUrl['path'] != $secremoveimg) {
567dc524 2063 // parse_url did not lead to satisfying result
2064 $attvalue = $sQuote . SM_PATH . 'images/blank.png' . $sQuote;
2065 }
2066 }
2067 }
2068 }
2069}
2070
691a2d25 2071/**
8bd0068d 2072 * This function edits the style definition to make them friendly and
2073 * usable in SquirrelMail.
2074 *
2075 * @param $message the message object
2076 * @param $id the message id
2077 * @param $content a string with whatever is between <style> and </style>
2078 * @param $mailbox the message mailbox
2079 * @return a string with edited content.
2080 */
e60a299a 2081function sq_fixstyle($body, $pos, $message, $id, $mailbox){
b583c3e8 2082 $me = 'sq_fixstyle';
7e2ff844 2083 // workaround for </style> in between comments
2084 $iCurrentPos = $pos;
2085 $content = '';
2086 $sToken = '';
2087 $bSucces = false;
2088 $bEndTag = false;
2089 for ($i=$pos,$iCount=strlen($body);$i<$iCount;++$i) {
1710ad65 2090 $char = $body[$i];
7e2ff844 2091 switch ($char) {
2092 case '<':
10c0caeb 2093 $sToken = $char;
7e2ff844 2094 break;
2095 case '/':
2096 if ($sToken == '<') {
2097 $sToken .= $char;
2098 $bEndTag = true;
2099 } else {
2100 $content .= $char;
2101 }
2102 break;
2103 case '>':
2104 if ($bEndTag) {
2105 $sToken .= $char;
2106 if (preg_match('/\<\/\s*style\s*\>/i',$sToken,$aMatch)) {
2107 $newpos = $i + 1;
2108 $bSucces = true;
2109 break 2;
2110 } else {
2111 $content .= $sToken;
2112 }
2113 $bEndTag = false;
2114 } else {
2115 $content .= $char;
2116 }
2117 break;
2118 case '!':
2119 if ($sToken == '<') {
2120 // possible comment
1710ad65 2121 if (isset($body[$i+2]) && substr($body,$i,3) == '!--') {
7e2ff844 2122 $i = strpos($body,'-->',$i+3);
2b665f28 2123 if ($i === false) { // no end comment
2124 $i = strlen($body);
2125 }
7e2ff844 2126 $sToken = '';
2127 }
2128 } else {
2129 $content .= $char;
2130 }
2131 break;
2132 default:
2133 if ($bEndTag) {
2134 $sToken .= $char;
2135 } else {
2136 $content .= $char;
2137 }
2138 break;
2139 }
2140 }
2141 if ($bSucces == FALSE){
7d06541f 2142 return array(FALSE, strlen($body));
2143 }
7e2ff844 2144
2b665f28 2145
2146
691a2d25 2147 /**
9ebace19 2148 * First look for general BODY style declaration, which would be
2149 * like so:
2150 * body {background: blah-blah}
2151 * and change it to .bodyclass so we can just assign it to a <div>
2152 */
7732acc6 2153 // $content = preg_replace("|body(\s*\{.*?\})|si", ".bodyclass\\1", $content);
d2f3a6fb 2154 // Nah, this is even better - try to preface all CSS selectors with
2155 // our <div> class ID "bodyclass" then correct generic "body" selectors
2156 // TODO: this works pretty good but breaks stuff like this:
2157 // @media print { body { font-size: 10pt; } }
2158 // but there isn't an easy way to make this regex skip @media
2159 // definitions... though lots of the ones in the wild will be
2160 // correctly handled because they tend to end with a parenthesis, like:
2161 // @media screen and (max-width:480px) { ...
2162 $content = preg_replace('/([a-z0-9._-][a-z0-9 >+~|:._-]*\s*(?:,|{.*?}))/si', '.bodyclass $1', $content);
2163 $content = str_replace('.bodyclass body', '.bodyclass', $content);
6ab8321f 2164
2165 global $use_transparent_security_image;
2166 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2167 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2168
691a2d25 2169 /**
0493ed11 2170 * Fix url('blah') declarations.
2171 */
2172 // $content = preg_replace("|url\s*\(\s*([\'\"])\s*\S+script\s*:.*?([\'\"])\s*\)|si",
2173 // "url(\\1$secremoveimg\\2)", $content);
2b665f28 2174
567dc524 2175 // first check for 8bit sequences and disallowed control characters
2176 if (preg_match('/[\16-\37\200-\377]+/',$content)) {
2177 $content = '<!-- style block removed by html filter due to presence of 8bit characters -->';
2178 return array($content, $newpos);
2179 }
2180
2b665f28 2181 // IE Sucks hard. We have a special function for it.
2182 sq_fixIE_idiocy($content);
2183
2184 // remove @import line
2185 $content = preg_replace("/^\s*(@import.*)$/mi","\n<!-- @import rules forbidden -->\n",$content);
2186
5b4884be 2187 // translate ur\l and variations (IE parses that)
2b665f28 2188 // TODO check if the sq_fixIE_idiocy function already handles this.
5b4884be 2189 $content = preg_replace("/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i", 'url', $content);
567dc524 2190 preg_match_all("/url\s*\((.+)\)/si",$content,$aMatch);
2191 if (count($aMatch)) {
2192 $aValue = $aReplace = array();
2193 foreach($aMatch[1] as $sMatch) {
2194 // url value
2195 $urlvalue = $sMatch;
2196 sq_fix_url('style',$urlvalue, $message, $id, $mailbox,"'");
2197 $aValue[] = $sMatch;
2198 $aReplace[] = $urlvalue;
0493ed11 2199 }
567dc524 2200 $content = str_replace($aValue,$aReplace,$content);
691a2d25 2201 }
567dc524 2202
9ebace19 2203 /**
2204 * Remove any backslashes, entities, and extraneous whitespace.
2205 */
0493ed11 2206 $contentTemp = $content;
2207 sq_defang($contentTemp);
2208 sq_unspace($contentTemp);
a3daaaf3 2209
691a2d25 2210 /**
8bd0068d 2211 * Fix stupid css declarations which lead to vulnerabilities
2212 * in IE.
39352565 2213 *
2214 * Also remove "position" attribute, as it can easily be set
2215 * to "fixed" or "absolute" with "left" and "top" attributes
2216 * of zero, taking over the whole content frame. It can also
2217 * be set to relative and move itself anywhere it wants to,
2218 * displaying content in areas it shouldn't be allowed to touch.
8bd0068d 2219 */
caf0ab1d 2220 $match = Array('/\/\*.*\*\//', // removes /* blah blah */
5db90261 2221 '/expression/i',
0493ed11 2222 '/behaviou*r/i',
2223 '/binding/i',
2b665f28 2224 '/include-source/i',
2225 '/javascript/i',
39352565 2226 '/script/i',
2227 '/position/i');
2228 $replace = Array('','idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', 'idiocy', '');
0493ed11 2229 $contentNew = preg_replace($match, $replace, $contentTemp);
2230 if ($contentNew !== $contentTemp) {
2231 // insecure css declarations are used. From now on we don't care
2232 // anymore if the css is destroyed by sq_deent, sq_unspace or sq_unbackslash
2233 $content = $contentNew;
2234 }
7d06541f 2235 return array($content, $newpos);
691a2d25 2236}
a3daaaf3 2237
0493ed11 2238
691a2d25 2239/**
8bd0068d 2240 * This function converts cid: url's into the ones that can be viewed in
2241 * the browser.
2242 *
2243 * @param $message the message object
2244 * @param $id the message id
2245 * @param $cidurl the cid: url.
2246 * @param $mailbox the message mailbox
2247 * @return a string with a http-friendly url
2248 */
b3af12ef 2249function sq_cid2http($message, $id, $cidurl, $mailbox){
691a2d25 2250 /**
8bd0068d 2251 * Get rid of quotes.
2252 */
691a2d25 2253 $quotchar = substr($cidurl, 0, 1);
2dd879b8 2254 if ($quotchar == '"' || $quotchar == "'"){
2255 $cidurl = str_replace($quotchar, "", $cidurl);
2256 } else {
2257 $quotchar = '';
2258 }
691a2d25 2259 $cidurl = substr(trim($cidurl), 4);
0493ed11 2260
2261 $match_str = '/\{.*?\}\//';
2262 $str_rep = '';
2263 $cidurl = preg_replace($match_str, $str_rep, $cidurl);
2264
e5e9381a 2265 $linkurl = find_ent_id($cidurl, $message);
9ebace19 2266 /* in case of non-safe cid links $httpurl should be replaced by a sort of
2267 unsafe link image */
e5e9381a 2268 $httpurl = '';
c8f5f606 2269
8bd0068d 2270 /**
2271 * This is part of a fix for Outlook Express 6.x generating
2272 * cid URLs without creating content-id headers. These images are
2273 * not part of the multipart/related html mail. The html contains
2274 * <img src="cid:{some_id}/image_filename.ext"> references to
2275 * attached images with as goal to render them inline although
2276 * the attachment disposition property is not inline.
2277 */
c8f5f606 2278
2279 if (empty($linkurl)) {
2280 if (preg_match('/{.*}\//', $cidurl)) {
2281 $cidurl = preg_replace('/{.*}\//','', $cidurl);
2282 if (!empty($cidurl)) {
2283 $linkurl = find_ent_id($cidurl, $message);
2284 }
2285 }
2286 }
f8a1ed5a 2287
c8f5f606 2288 if (!empty($linkurl)) {
fa26ab45 2289 $httpurl = $quotchar . sqm_baseuri() . 'src/download.php?absolute_dl=true&amp;' .
8bd0068d 2290 "passed_id=$id&amp;mailbox=" . urlencode($mailbox) .
2291 '&amp;ent_id=' . $linkurl . $quotchar;
c8f5f606 2292 } else {
2293 /**
2294 * If we couldn't generate a proper img url, drop in a blank image
2295 * instead of sending back empty, otherwise it causes unusual behaviour
2296 */
bc017c1d 2297 $httpurl = $quotchar . SM_PATH . 'images/blank.png' . $quotchar;
e5e9381a 2298 }
f8a1ed5a 2299
691a2d25 2300 return $httpurl;
2301}
2302
2303/**
8bd0068d 2304 * This function changes the <body> tag into a <div> tag since we
2305 * can't really have a body-within-body.
2306 *
2307 * @param $attary an array of attributes and values of <body>
2308 * @param $mailbox mailbox we're currently reading (for cid2http)
2309 * @param $message current message (for cid2http)
2310 * @param $id current message id (for cid2http)
2311 * @return a modified array of attributes to be set for <div>
2312 */
2dd879b8 2313function sq_body2div($attary, $mailbox, $message, $id){
b583c3e8 2314 $me = 'sq_body2div';
3d8371be 2315 $divattary = Array('class' => "'bodyclass'");
b583c3e8 2316 $text = '#000000';
c189a963 2317 $has_bgc_stl = $has_txt_stl = false;
b583c3e8 2318 $styledef = '';
691a2d25 2319 if (is_array($attary) && sizeof($attary) > 0){
2320 foreach ($attary as $attname=>$attvalue){
2321 $quotchar = substr($attvalue, 0, 1);
2322 $attvalue = str_replace($quotchar, "", $attvalue);
2323 switch ($attname){
3d8371be 2324 case 'background':
8bd0068d 2325 $attvalue = sq_cid2http($message, $id, $attvalue, $mailbox);
3d8371be 2326 $styledef .= "background-image: url('$attvalue'); ";
2327 break;
2328 case 'bgcolor':
c189a963 2329 $has_bgc_stl = true;
3d8371be 2330 $styledef .= "background-color: $attvalue; ";
2331 break;
2332 case 'text':
c189a963 2333 $has_txt_stl = true;
3d8371be 2334 $styledef .= "color: $attvalue; ";
2335 break;
691a2d25 2336 }
a3daaaf3 2337 }
c189a963 2338 // Outlook defines a white bgcolor and no text color. This can lead to
2339 // white text on a white bg with certain themes.
2340 if ($has_bgc_stl && !$has_txt_stl) {
2341 $styledef .= "color: $text; ";
2342 }
691a2d25 2343 if (strlen($styledef) > 0){
1710ad65 2344 $divattary["style"] = "\"$styledef\"";
691a2d25 2345 }
2346 }
2347 return $divattary;
2348}
a3daaaf3 2349
691a2d25 2350/**
8bd0068d 2351 * This is the main function and the one you should actually be calling.
2352 * There are several variables you should be aware of an which need
2353 * special description.
2354 *
2355 * Since the description is quite lengthy, see it here:
2356 * http://linux.duke.edu/projects/mini/htmlfilter/
2357 *
2358 * @param $body the string with HTML you wish to filter
2359 * @param $tag_list see description above
2360 * @param $rm_tags_with_content see description above
2361 * @param $self_closing_tags see description above
2362 * @param $force_tag_closing see description above
2363 * @param $rm_attnames see description above
2364 * @param $bad_attvals see description above
2365 * @param $add_attr_to_tag see description above
2366 * @param $message message object
2367 * @param $id message id
dfc64d14 2368 * @param $recursively_called boolean flag for recursive calls into this function (optional; default FALSE)
8bd0068d 2369 * @return sanitized html safe to show on your pages.
2370 */
da2415c1 2371function sq_sanitize($body,
8bd0068d 2372 $tag_list,
2373 $rm_tags_with_content,
2374 $self_closing_tags,
2375 $force_tag_closing,
2376 $rm_attnames,
2377 $bad_attvals,
2378 $add_attr_to_tag,
2379 $message,
2380 $id,
dfc64d14 2381 $mailbox,
2382 $recursively_called=FALSE
8bd0068d 2383 ){
b583c3e8 2384 $me = 'sq_sanitize';
dfc64d14 2385
2386 /**
2387 * See if tag_list is of tags to remove or tags to allow.
2388 * false means remove these tags
2389 * true means allow these tags
2390 */
2391 $orig_tag_list = $tag_list;
7d06541f 2392 $rm_tags = array_shift($tag_list);
dfc64d14 2393
691a2d25 2394 /**
8bd0068d 2395 * Normalize rm_tags and rm_tags_with_content.
2396 */
7d06541f 2397 @array_walk($tag_list, 'sq_casenormalize');
691a2d25 2398 @array_walk($rm_tags_with_content, 'sq_casenormalize');
2399 @array_walk($self_closing_tags, 'sq_casenormalize');
dfc64d14 2400
691a2d25 2401 $curpos = 0;
2402 $open_tags = Array();
2dd879b8 2403 $trusted = "\n<!-- begin sanitized html -->\n";
691a2d25 2404 $skip_content = false;
bb8d0799 2405 /**
8bd0068d 2406 * Take care of netscape's stupid javascript entities like
2407 * &{alert('boo')};
2408 */
bb8d0799 2409 $body = preg_replace("/&(\{.*?\};)/si", "&amp;\\1", $body);
691a2d25 2410
7d06541f 2411 while (($curtag = sq_getnxtag($body, $curpos)) != FALSE){
691a2d25 2412 list($tagname, $attary, $tagtype, $lt, $gt) = $curtag;
dfc64d14 2413
2414 /**
2415 * RCDATA and RAWTEXT tags are handled differently:
2416 * next instance of closing tag is used, whether or not
2417 * the HTML is well formed before that
2418 */
2419 global $rcdata_rawtext_tags;
2420 if (!$recursively_called
2421 && in_array($tagname, $rcdata_rawtext_tags)
2422 && $tagtype === 1){
2423 $closing_tag = false;
2424 $closing_tag_offset = $curpos;
2425 // seek out the closing tag for the current RCDATA/RAWTEXT tag
2426 while (1) {
2427 // first we need to move forward to next available closing tag
2428 // (intentionally leave off the closing > and let sq_getnxtag() validate a proper tag syntax)
2429 $next_tag = sq_findnxreg($body, $closing_tag_offset, "</\s*$tagname");
2430 if ($next_tag === false) {
2431 $closing_tag = false;
2432 break;
2433 }
2434 // but then we have to make sure it's a well-formed tag
2435 $closing_tag = sq_getnxtag($body, $next_tag[0]);
2436 if ($closing_tag === false)
2437 break;
2438 else if ($closing_tag[0] !== false
2439 // these should be redundant
2440 && $closing_tag[0] === $tagname && $closing_tag[2] === 2) {
2441 $trusted .= sq_sanitize(substr($body, $curpos, $closing_tag[4] - $curpos + 1),
2442 $orig_tag_list, $rm_tags_with_content, $self_closing_tags,
2443 $force_tag_closing, $rm_attnames, $bad_attvals, $add_attr_to_tag,
2444 $message, $id, $mailbox, true);
2445 $curpos = $closing_tag[4] + 1;
2446 continue 2;
2447 }
2448 $closing_tag_offset = $next_tag[0] + 1;
2449 }
2450 if ($closing_tag === false)
2451 { /* no-op... there was no closing tag for this RCDATA/RAWTEXT tag - we could probably set $curpos to the end of $body, but this HTML is malformed anyway and should just fall apart on its own */ }
2452 }
2453
691a2d25 2454 $free_content = substr($body, $curpos, $lt-$curpos);
2455 /**
8bd0068d 2456 * Take care of <style>
2457 */
7d06541f 2458 if ($tagname == "style" && $tagtype == 1){
da2415c1 2459 list($free_content, $curpos) =
e60a299a 2460 sq_fixstyle($body, $gt+1, $message, $id, $mailbox);
7d06541f 2461 if ($free_content != FALSE){
1d398b1d 2462 if ( !empty($attary) ) {
2463 $attary = sq_fixatts($tagname,
2464 $attary,
2465 $rm_attnames,
2466 $bad_attvals,
2467 $add_attr_to_tag,
2468 $message,
2469 $id,
2470 $mailbox
2471 );
2472 }
7d06541f 2473 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2474 $trusted .= $free_content;
2475 $trusted .= sq_tagprint($tagname, false, 2);
2476 }
2477 continue;
691a2d25 2478 }
2479 if ($skip_content == false){
2480 $trusted .= $free_content;
691a2d25 2481 }
2482 if ($tagname != FALSE){
2483 if ($tagtype == 2){
2484 if ($skip_content == $tagname){
2485 /**
8bd0068d 2486 * Got to the end of tag we needed to remove.
2487 */
691a2d25 2488 $tagname = false;
2489 $skip_content = false;
2490 } else {
2491 if ($skip_content == false){
c828931c 2492 if ($tagname == "body"){
2493 $tagname = "div";
2dd879b8 2494 }
1710ad65 2495 if (isset($open_tags[$tagname]) &&
2496 $open_tags[$tagname] > 0){
2497 $open_tags[$tagname]--;
691a2d25 2498 } else {
2dd879b8 2499 $tagname = false;
691a2d25 2500 }
691a2d25 2501 }
2502 }
2503 } else {
2504 /**
8bd0068d 2505 * $rm_tags_with_content
2506 */
691a2d25 2507 if ($skip_content == false){
2508 /**
8bd0068d 2509 * See if this is a self-closing type and change
2510 * tagtype appropriately.
2511 */
691a2d25 2512 if ($tagtype == 1
8bd0068d 2513 && in_array($tagname, $self_closing_tags)){
2dd879b8 2514 $tagtype = 3;
691a2d25 2515 }
2516 /**
8bd0068d 2517 * See if we should skip this tag and any content
2518 * inside it.
2519 */
691a2d25 2520 if ($tagtype == 1 &&
8bd0068d 2521 in_array($tagname, $rm_tags_with_content)){
691a2d25 2522 $skip_content = $tagname;
2523 } else {
da2415c1 2524 if (($rm_tags == false
8bd0068d 2525 && in_array($tagname, $tag_list)) ||
2526 ($rm_tags == true &&
2527 !in_array($tagname, $tag_list))){
691a2d25 2528 $tagname = false;
2529 } else {
2dd879b8 2530 /**
8bd0068d 2531 * Convert body into div.
2532 */
2dd879b8 2533 if ($tagname == "body"){
2534 $tagname = "div";
da2415c1 2535 $attary = sq_body2div($attary, $mailbox,
8bd0068d 2536 $message, $id);
2dd879b8 2537 }
691a2d25 2538 if ($tagtype == 1){
1710ad65 2539 if (isset($open_tags[$tagname])){
2540 $open_tags[$tagname]++;
691a2d25 2541 } else {
1710ad65 2542 $open_tags[$tagname]=1;
691a2d25 2543 }
2544 }
2545 /**
8bd0068d 2546 * This is where we run other checks.
2547 */
691a2d25 2548 if (is_array($attary) && sizeof($attary) > 0){
2549 $attary = sq_fixatts($tagname,
8bd0068d 2550 $attary,
2551 $rm_attnames,
2552 $bad_attvals,
2553 $add_attr_to_tag,
2554 $message,
2555 $id,
2556 $mailbox
2557 );
691a2d25 2558 }
2559 }
2560 }
691a2d25 2561 }
2562 }
2563 if ($tagname != false && $skip_content == false){
2564 $trusted .= sq_tagprint($tagname, $attary, $tagtype);
2565 }
691a2d25 2566 }
2567 $curpos = $gt+1;
a3daaaf3 2568 }
691a2d25 2569 $trusted .= substr($body, $curpos, strlen($body)-$curpos);
2570 if ($force_tag_closing == true){
2571 foreach ($open_tags as $tagname=>$opentimes){
2572 while ($opentimes > 0){
2573 $trusted .= '</' . $tagname . '>';
2574 $opentimes--;
2575 }
2576 }
2577 $trusted .= "\n";
2578 }
2579 $trusted .= "<!-- end sanitized html -->\n";
2580 return $trusted;
2581}
451f74a2 2582
691a2d25 2583/**
8bd0068d 2584 * This is a wrapper function to call html sanitizing routines.
2585 *
2586 * @param $body the body of the message
2587 * @param $id the id of the message
c189a963 2588
2589 * @param $message
2590 * @param $mailbox
2591 * @param boolean $take_mailto_links When TRUE, converts mailto: links
2592 * into internal SM compose links
2593 * (optional; default = TRUE)
8bd0068d 2594 * @return a string with html safe to display in the browser.
2595 */
c189a963 2596function magicHTML($body, $id, $message, $mailbox = 'INBOX', $take_mailto_links =true) {
2597
202bcbcc 2598 // require_once(SM_PATH . 'functions/url_parser.php'); // for $MailTo_PReg_Match
c189a963 2599
691a2d25 2600 global $attachment_common_show_images, $view_unsafe_images,
dfc64d14 2601 $has_unsafe_images, $allow_svg_display, $rcdata_rawtext_tags,
2602 $remove_rcdata_rawtext_tags_and_content;
2603
2604 $rcdata_rawtext_tags = array(
2605 "noscript",
2606 "noframes",
2607 "noembed",
2608 "textarea",
2609 // also "title", "xmp", "script", "iframe", "plaintext" which we already remove below
2610 );
2611
691a2d25 2612 /**
8bd0068d 2613 * Don't display attached images in HTML mode.
2b665f28 2614 *
d0187bd6 2615 * SB: why?
8bd0068d 2616 */
691a2d25 2617 $attachment_common_show_images = false;
2618 $tag_list = Array(
dfc64d14 2619 false, // remove these tags
8bd0068d 2620 "meta",
2621 "html",
2622 "head",
2623 "base",
2624 "link",
2625 "frame",
2626 "iframe",
2627 "plaintext",
caf0ab1d 2628 "marquee",
8bd0068d 2629 );
691a2d25 2630
2631 $rm_tags_with_content = Array(
8bd0068d 2632 "script",
caf0ab1d 2633 "object",
8bd0068d 2634 "applet",
2635 "embed",
2636 "title",
2637 "frameset",
0493ed11 2638 "xmp",
caf0ab1d 2639 "xml",
8bd0068d 2640 );
c88cef21 2641 if (!$allow_svg_display)
caf0ab1d 2642 $rm_tags_with_content[] = 'svg';
dfc64d14 2643 /**
2644 * SquirrelMail will parse RCDATA and RAWTEXT tags and handle them as the special
2645 * case that they are, but if you prefer to remove them and their contents entirely
2646 * (in most cases, should be a safe thing with minimal impact), you can add the
2647 * following to config/config_local.php
2648 * $remove_rcdata_rawtext_tags_and_content = TRUE;
2649 */
2650 if ($remove_rcdata_rawtext_tags_and_content)
2651 $rm_tags_with_content = array_merge($rm_tags_with_content, $rcdata_rawtext_tags);
691a2d25 2652
2653 $self_closing_tags = Array(
8bd0068d 2654 "img",
2655 "br",
2656 "hr",
2657 "input",
caf0ab1d 2658 "outbind",
8bd0068d 2659 );
691a2d25 2660
2dd879b8 2661 $force_tag_closing = true;
691a2d25 2662
2663 $rm_attnames = Array(
8bd0068d 2664 "/.*/" =>
2665 Array(
2666 "/target/i",
2667 "/^on.*/i",
2668 "/^dynsrc/i",
2669 "/^data.*/i",
caf0ab1d 2670 "/^lowsrc.*/i",
8bd0068d 2671 )
2672 );
691a2d25 2673
6ab8321f 2674 global $use_transparent_security_image;
2675 if ($use_transparent_security_image) $secremoveimg = '../images/spacer.png';
2676 else $secremoveimg = '../images/' . _("sec_remove_eng.png");
2677
691a2d25 2678 $bad_attvals = Array(
8bd0068d 2679 "/.*/" =>
691a2d25 2680 Array(
0a6ec9b5 2681 "/^src|background/i" =>
8bd0068d 2682 Array(
691a2d25 2683 Array(
8bd0068d 2684 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2685 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2686 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
691a2d25 2687 ),
8bd0068d 2688 Array(
2689 "\\1$secremoveimg\\2",
2690 "\\1$secremoveimg\\2",
2691 "\\1$secremoveimg\\2",
8bd0068d 2692 )
2693 ),
0a6ec9b5 2694 "/^href|action/i" =>
8bd0068d 2695 Array(
0a6ec9b5 2696 Array(
8bd0068d 2697 "/^([\'\"])\s*\S+script\s*:.*([\'\"])/si",
2698 "/^([\'\"])\s*mocha\s*:*.*([\'\"])/si",
2699 "/^([\'\"])\s*about\s*:.*([\'\"])/si"
0a6ec9b5 2700 ),
691a2d25 2701 Array(
8bd0068d 2702 "\\1#\\1",
2703 "\\1#\\1",
2704 "\\1#\\1"
02474e43 2705 )
8bd0068d 2706 ),
2707 "/^style/i" =>
2708 Array(
2709 Array(
5db90261 2710 "/\/\*.*\*\//",
8bd0068d 2711 "/expression/i",
2712 "/binding/i",
2713 "/behaviou*r/i",
2714 "/include-source/i",
39352565 2715
2716 // position:relative can also be exploited
2717 // to put content outside of email body area
2718 // and position:fixed is similarly exploitable
2719 // as position:absolute, so we'll remove it
2720 // altogether....
2721 //
2722 // Does this screw up legitimate HTML messages?
2723 // If so, the only fix I see is to allow position
2724 // attributes (any values? I think we still have
2725 // to block static and fixed) only if $use_iframe
2726 // is enabled (1.5.0+)
2727 //
2728 // was: "/position\s*:\s*absolute/i",
2729 //
2730 "/position\s*:/i",
2731
1d935bc2 2732 "/(\\\\)?u(\\\\)?r(\\\\)?l(\\\\)?/i",
8bd0068d 2733 "/url\s*\(\s*([\'\"])\s*\S+script\s*:.*([\'\"])\s*\)/si",
2734 "/url\s*\(\s*([\'\"])\s*mocha\s*:.*([\'\"])\s*\)/si",
2735 "/url\s*\(\s*([\'\"])\s*about\s*:.*([\'\"])\s*\)/si",
39352565 2736 "/(.*)\s*:\s*url\s*\(\s*([\'\"]*)\s*\S+script\s*:.*([\'\"]*)\s*\)/si",
8bd0068d 2737 ),
2738 Array(
5db90261 2739 "",
8bd0068d 2740 "idiocy",
2741 "idiocy",
2742 "idiocy",
2743 "idiocy",
567dc524 2744 "idiocy",
5b4884be 2745 "url",
8bd0068d 2746 "url(\\1#\\1)",
2747 "url(\\1#\\1)",
2748 "url(\\1#\\1)",
8bd0068d 2749 "\\1:url(\\2#\\3)"
2750 )
691a2d25 2751 )
8bd0068d 2752 )
691a2d25 2753 );
9ebace19 2754
d31f73f1 2755 // If there's no "view_unsafe_images" variable in the URL, turn unsafe
2756 // images off by default.
afbf184c 2757 sqgetGlobalVar('view_unsafe_images', $view_unsafe_images, SQ_GET, FALSE);
9ebace19 2758
691a2d25 2759 if (!$view_unsafe_images){
2760 /**
8bd0068d 2761 * Remove any references to http/https if view_unsafe_images set
2762 * to false.
2763 */
1710ad65 2764 array_push($bad_attvals['/.*/']['/^src|background/i'][0],
8bd0068d 2765 '/^([\'\"])\s*https*:.*([\'\"])/si');
1710ad65 2766 array_push($bad_attvals['/.*/']['/^src|background/i'][1],
8bd0068d 2767 "\\1$secremoveimg\\1");
1710ad65 2768 array_push($bad_attvals['/.*/']['/^style/i'][0],
7ef0b415 2769 '/url\([\'\"]?https?:[^\)]*[\'\"]?\)/si');
1710ad65 2770 array_push($bad_attvals['/.*/']['/^style/i'][1],
8bd0068d 2771 "url(\\1$secremoveimg\\1)");
691a2d25 2772 }
451f74a2 2773
691a2d25 2774 $add_attr_to_tag = Array(
8bd0068d 2775 "/^a$/i" =>
c25f2fbb 2776 Array('target'=>'"_blank"',
02474e43 2777 'title'=>'"'._("This external link will open in a new window").'"'
8bd0068d 2778 )
2779 );
da2415c1 2780 $trusted = sq_sanitize($body,
8bd0068d 2781 $tag_list,
2782 $rm_tags_with_content,
2783 $self_closing_tags,
2784 $force_tag_closing,
2785 $rm_attnames,
2786 $bad_attvals,
2787 $add_attr_to_tag,
2788 $message,
2789 $id,
2790 $mailbox
2791 );
567dc524 2792 if (strpos($trusted,$secremoveimg)){
691a2d25 2793 $has_unsafe_images = true;
da2415c1 2794 }
c189a963 2795
2796 // we want to parse mailto's in HTML output, change to SM compose links
2797 // this is a modified version of code from url_parser.php... but Marc is
2798 // right: we need a better filtering implementation; adding this randomly
2799 // here is not a great solution
2800 //
2801 if ($take_mailto_links) {
2802 // parseUrl($trusted); // this even parses URLs inside of tags... too aggressive
2803 global $MailTo_PReg_Match;
7196da62 2804 // some mailers (Microsoft, surprise surprise) produce mailto strings without being
2805 // inside an anchor (link) tag, so we have to make sure the regex looks for the
2806 // quote before mailto, and we'll also try to convert the non-links back into links
2807 $MailTo_PReg_Match = '/([\'"])?mailto:' . substr($MailTo_PReg_Match, 1) ;
c189a963 2808 if ((preg_match_all($MailTo_PReg_Match, $trusted, $regs)) && ($regs[0][0] != '')) {
2809 foreach ($regs[0] as $i => $mailto_before) {
7196da62 2810 $mailto_params = $regs[11][$i];
2811
2812 // get rid of any leading quote we may have captured but don't care about
2813 //
2814 $mailto_before = ltrim($mailto_before, '"\'');
2815
c189a963 2816 // get rid of any tailing quote since we have to add send_to to the end
2817 //
7196da62 2818 $mailto_before = rtrim($mailto_before, '"\'');
2819 $mailto_params = rtrim($mailto_params, '"\'');
c189a963 2820
7196da62 2821 if ($regs[2][$i]) { //if there is an email addr before '?', we need to merge it with the params
2822 $to = 'to=' . $regs[2][$i];
c189a963 2823 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
2824 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
2825 else {
2826 if ($mailto_params) //already some params, append to them
2827 $mailto_params .= '&amp;' . $to;
2828 else
2829 $mailto_params .= '?' . $to;
2830 }
2831 }
2832
2833 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
2834
2835 // we'll already have target=_blank, no need to allow comp_in_new
2836 // here (which would be a lot more work anyway)
2837 //
2838 global $compose_new_win;
2839 $temp_comp_in_new = $compose_new_win;
2840 $compose_new_win = 0;
2841 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
2842 $compose_new_win = $temp_comp_in_new;
2843
2844 // remove <a href=" and anything after the next quote (we only
2845 // need the uri, not the link HTML) in compose uri
2846 //
7196da62 2847 // but only do this if the original mailto was in a real anchor tag
2848 //
2849 if (!empty($regs[1][$i])) {
2850 $comp_uri = substr($comp_uri, 9);
2851 $comp_uri = substr($comp_uri, 0, strpos($comp_uri, '"', 1));
2852 }
c189a963 2853 $trusted = str_replace($mailto_before, $comp_uri, $trusted);
2854 }
2855 }
2856 }
2857
691a2d25 2858 return $trusted;
451f74a2 2859}
a4a70693 2860
da2415c1 2861/**
8bd0068d 2862 * function SendDownloadHeaders - send file to the browser
2863 *
2864 * Original Source: SM core src/download.php
2865 * moved here to make it available to other code, and separate
2866 * front end from back end functionality.
2867 *
2868 * @param string $type0 first half of mime type
2869 * @param string $type1 second half of mime type
2870 * @param string $filename filename to tell the browser for downloaded file
2871 * @param boolean $force whether to force the download dialog to pop
2872 * @param optional integer $filesize send the Content-Header and length to the browser
2873 * @return void
2874 */
02474e43 2875function SendDownloadHeaders($type0, $type1, $filename, $force, $filesize=0) {
2876 global $languages, $squirrelmail_language;
cfffd60b 2877 $isIE = $isIE6plus = false;
02474e43 2878
2879 sqgetGlobalVar('HTTP_USER_AGENT', $HTTP_USER_AGENT, SQ_SERVER);
2880
2881 if (strstr($HTTP_USER_AGENT, 'compatible; MSIE ') !== false &&
8bd0068d 2882 strstr($HTTP_USER_AGENT, 'Opera') === false) {
cfffd60b 2883 $isIE = true;
02474e43 2884 }
2885
cfffd60b 2886 if (preg_match('/compatible; MSIE ([0-9]+)/', $HTTP_USER_AGENT, $match) &&
2887 ((int)$match[1]) >= 6 && strstr($HTTP_USER_AGENT, 'Opera') === false) {
2888 $isIE6plus = true;
02474e43 2889 }
2890
2891 if (isset($languages[$squirrelmail_language]['XTRA_CODE']) &&
8bd0068d 2892 function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename')) {
02474e43 2893 $filename =
8bd0068d 2894 call_user_func($languages[$squirrelmail_language]['XTRA_CODE'] . '_downloadfilename', $filename, $HTTP_USER_AGENT);
02474e43 2895 } else {
1a4bc4a6 2896 $filename = preg_replace('/[\\\\\/:*?"<>|;]/', '_', str_replace('&nbsp;', ' ', $filename));
02474e43 2897 }
2898
2899 // A Pox on Microsoft and it's Internet Explorer!
2900 //
2901 // IE has lots of bugs with file downloads.
2902 // It also has problems with SSL. Both of these cause problems
2903 // for us in this function.
2904 //
2905 // See this article on Cache Control headers and SSL
2906 // http://support.microsoft.com/default.aspx?scid=kb;en-us;323308
2907 //
2908 // The best thing you can do for IE is to upgrade to the latest
2909 // version
2910 //set all the Cache Control Headers for IE
2911 if ($isIE) {
2912 $filename=rawurlencode($filename);
2913 header ("Pragma: public");
8bd0068d 2914 header ("Cache-Control: no-store, max-age=0, no-cache, must-revalidate"); // HTTP/1.1
c2503a4e 2915 // does nothing - see: https://blogs.msdn.microsoft.com/ieinternals/2009/07/20/internet-explorers-cache-control-extensions/
2916 // header ("Cache-Control: post-check=0, pre-check=0", false);
8bd0068d 2917 header ("Cache-Control: private");
02474e43 2918
2919 //set the inline header for IE, we'll add the attachment header later if we need it
2920 header ("Content-Disposition: inline; filename=$filename");
2921 }
2922
2923 if (!$force) {
2924 // Try to show in browser window
2925 header ("Content-Disposition: inline; filename=\"$filename\"");
2926 header ("Content-Type: $type0/$type1; name=\"$filename\"");
2927 } else {
2928 // Try to pop up the "save as" box
2929
2930 // IE makes this hard. It pops up 2 save boxes, or none.
2931 // http://support.microsoft.com/support/kb/articles/Q238/5/88.ASP
2932 // http://support.microsoft.com/default.aspx?scid=kb;EN-US;260519
2933 // But, according to Microsoft, it is "RFC compliant but doesn't
2934 // take into account some deviations that allowed within the
2935 // specification." Doesn't that mean RFC non-compliant?
2936 // http://support.microsoft.com/support/kb/articles/Q258/4/52.ASP
2937
2938 // all browsers need the application/octet-stream header for this
2939 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2940
2941 // http://support.microsoft.com/support/kb/articles/Q182/3/15.asp
2942 // Do not have quotes around filename, but that applied to
2943 // "attachment"... does it apply to inline too?
2944 header ("Content-Disposition: attachment; filename=\"$filename\"");
2945
cfffd60b 2946 if ($isIE && !$isIE6plus) {
02474e43 2947 // This combination seems to work mostly. IE 5.5 SP 1 has
2948 // known issues (see the Microsoft Knowledge Base)
2949
2950 // This works for most types, but doesn't work with Word files
2951 header ("Content-Type: application/download; name=\"$filename\"");
7e2ff844 2952 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2953 // These are spares, just in case. :-)
2954 //header("Content-Type: $type0/$type1; name=\"$filename\"");
2955 //header("Content-Type: application/x-msdownload; name=\"$filename\"");
2956 //header("Content-Type: application/octet-stream; name=\"$filename\"");
7e2ff844 2957 } else if ($isIE) {
2958 // This is to prevent IE for MIME sniffing and auto open a file in IE
2959 header ("Content-Type: application/force-download; name=\"$filename\"");
02474e43 2960 } else {
2961 // another application/octet-stream forces download for Netscape
2962 header ("Content-Type: application/octet-stream; name=\"$filename\"");
2963 }
2964 }
2965
2966 //send the content-length header if the calling function provides it
2967 if ($filesize > 0) {
2968 header("Content-Length: $filesize");
2969 }
07c49f57 2970
8d863f64 2971} // end fn SendDownloadHeaders
Unnamed repository; edit this file 'description' to name the repository.